Argumentation Library

Fernando Leal
Hubert Marraud

How Philosophers
Argue
An Adversarial Collaboration
on the Russell–Copleston Debate
Argumentation Library

Volume 41

Series Editor
Frans H. van Eemeren, University of Amsterdam, Amsterdam, The Netherlands

Editorial Board
Fernando Leal Carretero, University of Guadalajara, Guadalajara, Jalisco, Mexico
Maurice A Finocchiaro, Department of Philosophy, University of Nevada at Las
Vegas, Las Vegas, NV, USA
Bart Garssen, Faculty of Humanities, TAR, University of Amsterdam, Amsterdam,
Noord-Holland, The Netherlands
Sally Jackson, Communication, University of Illinois Urbana-Champaign, Urbana,
IL, USA
Wu Peng, School of Foreign Languages, Jiangsu University, Zhenjiang, China
Sara Rubinelli, University of Luzern, Nottwil, Luzern, Switzerland
Takeshi Suzuki, School of Information and Communication, Meiji University,
Chiyoda-ku, Tokyo, Japan
Cristián Santibañez Yañez, Faculdad de Psicologia, University of Concepción,
Concepción, Chile
David Zarefsky, School of Communication, Northwestern University, Evanston,
IL, USA
Sara Greco, IALS, Università della Svizzera Italiana, Lugano, Ticino, Switzerland
Since 1986 Springer, formerly Kluwer Academic Publishers, publishes the inter-
national interdisciplinary journal Argumentation. This journal is a medium for
distributing contributions to the study of argumentation from all schools of thought.
From a journal that published guest-edited issues devoted to specific themes, Argu-
mentation has developed into a regular journal providing a platform for discussing all
theoretical aspects of argumentative discourse. Since 1999 the journal has an accom-
panying book series consisting of volumes containing substantial contributions to
the study of argumentation.
The Argumentation Library aims to be a high quality book series consisting of
monographs and edited volumes. It publishes texts offering important theoretical
insights in certain major characteristics of argumentative discourse in order to inform
the international community of argumentation theorists of recent developments in the
field. The insights concerned may pertain to the process of argumentation but also to
aspects of argumentative texts resulting from this process. This means that books will
be published not only on various types of argumentative procedures, but also on the
features of enthymematic argumentation, argumentation structures, argumentation
schemes and fallacies.
Contributions to the series can be made by scholars from a broad variety of
disciplines, ranging from law to history, from linguistics to theology, and from science
to sociology. In particular, contributions are invited from argumentation theorists
with a background in informal or formal logic, modern or classical rhetoric, and
discourse analysis or speech communication. A prerequisite in all cases is that the
contribution involved is original and provides the forum of argumentation theorists
with an exemplary specimen of advanced scholarship. The Argumentation Library
should enrich the study of argumentation with insights that enhance its quality and
constitute a fruitful starting point for further research and application.
All proposals will be carefully taken into consideration by the editors. They are to
be submitted in fourfold. If the prospects for including a certain project in the series
are realistic, the author(s) will be invited to send at least three representative chapters
of their manuscript for review to the editors. In case the manuscript is then judged
eligible for publication, the complete manuscript will be reviewed by outside expert
referees. Only then a final decision can be taken concerning publication.
This book series is indexed in SCOPUS.
Authors interested in submitting a proposal or completed manuscript can contact
either christi.lue@springer.com or the Series Editor.

More information about this series at https://link.springer.com/bookseries/5642

Fernando Leal · Hubert Marraud

How Philosophers Argue

An Adversarial Collaboration
on the Russell–Copleston Debate
Fernando Leal Hubert Marraud
Universidad de Guadalajara Universidad Autónoma de Madrid
Guadalajara, Mexico Madrid, Spain

ISSN 1566-7650 ISSN 2215-1907 (electronic)

Argumentation Library
ISBN 978-3-030-85367-9 ISBN 978-3-030-85368-6 (eBook)
https://doi.org/10.1007/978-3-030-85368-6

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature
Switzerland AG 2022
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse
of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface

We shall avail ourselves of the privilege prefaces offer to tell the story of how this
book was born. Readers uninterested in personal narrative are advised to stop reading
this preface and go immediately to the introduction. Actually, there are two stories
to tell, so it is better to keep them separate.
Fernando Leal: About 15 years ago, I was charged with teaching a course on
argumentation for students in an M.A. in philosophy. After trying several mixtures
of methods taken out of the two main toolboxes at our disposal, namely the history
of philosophy and the logical approach of the analytic philosophers, I discovered
the English textbook on pragma-dialectics, published in 2002 and updated in 2017.
The idea of exercises was very palatable to me, because my course had, from the
very beginning, been organized as a workshop, where little packets of ‘theory’ were
immediately put into practice on actual philosophical texts.
Being at first somewhat cautious about the new approach, I had my class work
through the actual exercises of that book and only added a few philosophical texts,
which I selected with an eye to the greatest variety possible. But as I gained mastery
of the pragma-dialectical methods, I came to replace philosophical texts for all the
original exercises, whilst retaining the character of the exercises, in particular the
instructions for their solution. In due time, certain peculiarities of the philosophical
texts demanded more or less substantial modifications both to the way the material
was presented and to the exercise instructions.
As time passed, I became increasingly aware that I had unwittingly entered the
field of ‘argumentation in context’, the last phase of the pragma-dialectical research
programme, in which attention is paid to the special features of the communica-
tive activity which we call ‘philosophical argumentation’. This new and burgeoning
research area requires taking the general framework of standard pragma-dialectics
and adapting it to the way argumentation has been institutionalized in philosophy.
Given that any and all adaptations must be based on the empirical evidence of actual
cases of philosophical argumentation, I began by building up a large database of

v
vi Preface

philosophical argumentative texts of between 20 and 800 words, yet the time came
to experiment with longer and more complex examples.
The first seriously long text set for the workshop was precisely the debate between
Bertrand Russell and Frederick Copleston that is at the centre of this volume. I divided
the debate in several parts and gave it to my students to read with the following
instruction: Identify the stages of a critical discussion and specify which particular
operations within any particular stage are being performed in which utterance or set
of utterance within the text. The weekly discussions we had in class about the results
obtained and sent in by the students provided one of the richest teaching experiences
that I can remember. One thing was apparent, though: the students invariably selected
certain bits of the text as evidence for a particular operation within a particular
discussion stage and ignored or cut out all the other bits as irrelevant to the model.
This cutting out and paring away of smaller or larger portions of text is, of course, a
common experience in the analysis of arguments, independently of the theory.
Intrigued by the fact that I myself had incurred in this pruning of texts as I prepared
myself for each workshop session, I took a whole summer holiday to go through each
word, phrase, sentence, paragraph, and turn in the Russell-Copleston debate with the
sole purpose of finding out whether the ideal model of critical discussion could
actually account for everything in the text. I was guided by the intuition that this
should be the case, that no pruning was necessary, at least if the concept of critical
discussion was really a workable model of the whole argumentative process that
takes place when people argue.
The result of the summer exercise was encouraging, yet there was one little glitch.
When arguments are analysed in argumentation theory, all authors coincide in looking
at them from what would seem a ‘solipsistic’ perspective, almost as though they
were pieces of reasoning carried out in Olympic solitude. If this is questionable for
so-called ‘monologic’ texts, it is simply outrageous in the case of an oral dispute
such as the debate between Russell and Copleston. As most readers of this book will
probably know, some authors have tried to remedy the situation by means of dialogue
profiles (Walton) or dialectical profiles (van Eemeren and co-workers), a device to
keep track of what a party can do. This proposal goes in the right direction, but it does
not engage all the way with the traces left by the dialogical process in the text. The
theory of Hubert Marraud, argument dialectics, was a step forward, although it is not
well known in the international field of argumentation studies, given that most of the
materials (single papers and a textbook) have only been published in Spanish, due
to the close analysis of linguistic and speech phenomena that the method requires.
This book was born from the idea of doing a contrastive analytic study. In order to
highlight the power of Marraud’s methods, I have kept strictly within ‘solipsistic’
argument reconstruction in my analysis of the debate.
Hubert Marraud: I arrived at the theory of argumentation at the beginning of the
twenty-first century, coming from formal logic. Teaching formal logic in a philos-
ophy school is often frustrating, and repeated attempts had led me to conclude that
Preface vii

it did not make much sense. Johnson and Blair point out that the stimulus for the
emergence of informal logic came from three types of critiques: pedagogical crit-
icism, internal criticism, and empirical criticism. Johnson and Blair thus describe
pedagogical criticism: «The pedagogical critique arose from the emerging belief
that traditional formal deductive logic does not provide the student with the adequate
tools for the analysis and the evaluation of argumentation as that occurs in ordinary
discourse.» (Johnson & Blair 2002, p. 340). So it was the pedagogical critique that
led me to the theory of argumentation.
However, I did not come into the theory of argumentation through informal logic,
but through linguistics, and more precisely through the theory of l’argumentation
dans la langue by Ducrot and Anscombre. At that time I was working with sequent
calculi, and the discovery of argumentative operators and connectors fascinated me.
I think that initially I understood l’argumentation dans la langue from the distinc-
tion between logical rules and structural rules, and Kosta Došen’s idea that logical
constants function as punctuation marks. The crux of the matter is that logical connec-
tors do not join statements to form complex statements (as logical operators do), but
arguments to form complex arguments. Argument connectors and similar devices
allow us to set up a logical conception of argument structures and to redraw the
boundaries between logic and dialectics. I call “argument dialectics” the resulting
approach.
In general, linguists lack a systematic theory of inter-argumentative relations, and
their analyses of connectors are poor from the point of view of argument theory. At the
same time, logicians, formal and informal, do not pay due attention to language, and
therefore they put reasoning above argument. A properly argumentative interpreta-
tion of connectors allows us to get rid of many logical prejudices. In the first place, the
theorist of argumentation does not analyze timeless relationships between abstract
entities, but relationships between linguistic segments introduced by connectors and
other similar resources. This openly contradicts the inferential conception of argu-
ments, according to which what characterizes a good argument (from a logical point
of view) is that the conclusion follows from the premises. Moreover, when one exam-
ines the way in which we give reasons using these devices, one notices that there are
not only good and bad arguments, but better and worse ones. The result is the replace-
ment of a qualitative concept of good argument (such as validity) with a comparative
concept (such as force). This allows us to recognize that we build complex arguments
reinforcing, opposing and comparing some arguments with others.
The foregoing ideas are not completely new in argumentation theory, although
their effects have been quite limited for the moment, due to a penchant for syncretism.
First, to preserve the inferentialist conception of arguments and the qualitative
concept of good argument, many authors restrict the comparative, non-qualitative
character to a special type of inferences and arguments; namely “conductive reason-
ing”. Second, the inferentialist conception of arguments is rooted in the understanding
viii Preface

of logic as a theory of the different types of inferences (abductive, deductive, induc-

tive, etc.). This implies limiting the subject matter of logic to intra-argumentative rela-
tions, making inter-argumentative relations a subject for dialectical analysis. Thus,
the emphasis is on the process within which these structures arise and the func-
tions the various argument structures fulfil in this process. In this way, Johnson’s
celebrated concept of dialectical tier, which covers alternative positions and stan-
dard objections, is defined in terms of the dialectical obligations of the arguers. My
“argument dialectics”, instead, focusing on reasons vs. inferences, recovers a logical
approach to argumentative structures and inter-argumentative relations, embracing at
the same time a naturalized logic related to the interactive construction of arguments
by discursive agents.
This book would not have been possible without Fernando’s decisive impulse,
although seen more broadly it can be said that it is the product of the discussions
on argumentation theory that we are holding since 2013. The process of writing
the book has taken place under especially difficult conditions, both because of the
global pandemic and because of my personal circumstances. Here again, Fernando’s
encouragement has been decisive in not throwing in the towel. Adversarial collabo-
ration with Fernando has been an intellectually stimulating and demanding exercise,
which has forced me to refine my intuitions and to ask myself about the relation-
ship of my dialectic of arguments with pragma-dialectics. The notions of “interac-
tively constructed macro-argument” and of the “logical structure underlying critical
discussions” are the direct result of my attempts to explain to myself this relationship.

Acknowledgments

Fernando Leal: My heartfelt thanks go to Frans van Eemeren, who has always encour-
aged me to put my thoughts on paper and who believed in this project from the start,
and to Kenneth Blackwell, at the McMaster University Russell Archives in Canada,
for sending me important archival materials, correcting some mistakes, and being
infallibly attentive, courteous, and helpful to a perfect stranger. I am, of course,
immensely grateful to Hubert Marraud, who has been such a good sport over the
years, accepting all the textual challenges I have thrown on his path. I also want
to express my gratitude to Joaquín Galindo for taking the time, in the midst of the
pandemic, in order to read, with a fine logical comb, the second draft of my analysis
in full (poor soul) and for his valuable comments. This is also the place to thank the
graduate students of my two 2018 workshops on philosophical argumentation, which
were the locus of inception of this book. They worked hard with me and somehow
managed to cope with my perhaps excessive love of details. Last but actually first,
there are no words to express what I owe to my beloved wife Judith, without whom
I would have, not once by many times, given up on this project.
Hubert Marraud: I have been lucky enough to contribute to the birth of a group of
Spanish-speaking argumentation theorists, from which the Ibero-American Society
of Argumentation has emerged, and from which I have learned a lot. Among its
Preface ix

members I would like to express special thanks, in addition to Fernando, to Joaquín

Galindo Castañeda, José Ángel Gascón, Paula Olmos Gómez and Luis Vega Reñón.
Finally, I would like to thank FEDER/Ministerio de Ciencia, Innovación y Univer-
sidades, Agencia Estatal de Investigación, for their support. This research has been
funded by the Project “Argumentative practices and the pragmatics of reasons”
(Parg-Praz), reference number PGC2018-095941-B-I00.

Guadalajara, Mexico Fernando Leal

Madrid, Spain Hubert Marraud
Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 The Russell-Copleston Debate . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Why This Debate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.3 Why in This Way: A Word on Adversarial Collaboration . . . . . . 15
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2 Argumentation Theories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.1 Argumentation as Argument: A-Theories . . . . . . . . . . . . . . . . . . . 21
2.2 Argumentation as Argument Delivery and as Argument
Exchange: D-Theories and E-Theories . . . . . . . . . . . . . . . . . . . . . 31
2.3 Argumentation as Argumentative Process: P-Theories . . . . . . . . 38
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

Part I A P-Theoretical Analysis of the Debate, by Fernando Leal

3 Description of the Method Followed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.1 Questions in Philosophical Argumentation . . . . . . . . . . . . . . . . . . 53
3.2 Disagreements in Philosophy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
3.3 Philosophical Argumentation in the Ideal Model . . . . . . . . . . . . . 73
3.4 The Russell-Copleston Debate in a Nutshell . . . . . . . . . . . . . . . . . 81
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
4 Analysis of Segment I: Start of the Debate . . . . . . . . . . . . . . . . . . . . . . . 93
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
5 Analysis of Segment II: Discussion of Copleston’s
Metaphysical Argument . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
6 Analysis of Segment III: Discussion of Copleston’s Religious
Argument . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217

xi
xii Contents

7 Analysis of Segment IV: Discussion of Copleston’s Moral

Argument . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
8 Analysis of Segment V: Summing-Up of the Arguments . . . . . . . . . . . 267
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277

Part II An E-Theoretical Analysis of the Debate, by Hubert

Marraud
9 Argument Dialectic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
9.1 Logic or the Theory of Argument . . . . . . . . . . . . . . . . . . . . . . . . . . 281
9.2 Arguer Dialectics Versus Argument Dialectics . . . . . . . . . . . . . . . 283
9.3 Dialectical Rules Versus Logical Norms . . . . . . . . . . . . . . . . . . . . 284
9.4 The Scope of Logical Rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
9.5 Logic, Dialectic and the Macrostructure of Argument . . . . . . . . . 289
9.6 Criticism and Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294
10 Argumentation Structures and Operations . . . . . . . . . . . . . . . . . . . . . . 297
10.1 Argumentative Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
10.2 Simple Arguments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
10.3 Concatenation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
10.4 Warrants and Backings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
10.5 Analogous Arguments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
11 Counterarguments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
11.1 Counterarguments and Reasons . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
11.2 Main Kinds of Counterarguments . . . . . . . . . . . . . . . . . . . . . . . . . . 313
11.3 Dismissal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314
11.4 Objection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316
11.5 Rebuttal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
11.6 Refutation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
11.7 Weighing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
11.8 Rebuttals, Refutations and Standards of Proof . . . . . . . . . . . . . . . 326
11.9 Counterargument Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
11.10 The Order of Counterargumentation . . . . . . . . . . . . . . . . . . . . . . . 327
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328
12 Co-Oriented Reasons and Modifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
12.1 Co-Oriented Reasons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
12.2 Conjunction and Disjunction of Arguments . . . . . . . . . . . . . . . . . 334
12.3 Modifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 336
12.4 Modified Arguments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345
Contents xiii

13 Intertwined Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347

13.1 Argumentative Structures and Dialectical Profiles . . . . . . . . . . . . 347
13.2 A User’s Guide of Argument Structure Diagrams . . . . . . . . . . . . 350
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359
14 An Argument Dialectical Analysis of the Russell-Copleston
Debate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
14.1 The Argument from Contingency . . . . . . . . . . . . . . . . . . . . . . . . . . 362
14.2 Religious Experience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401
14.3 The Moral Argument . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
14.4 Concluding Stage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 434
15 Final Cross-Examination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
15.1 Comments on Part I (Hubert Marraud) . . . . . . . . . . . . . . . . . . . . . 437
15.2 Comments on Part II (Fernando Leal) . . . . . . . . . . . . . . . . . . . . . . 446
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 452

Appendix: Text of the Russell-Copleston Debate . . . . . . . . . . . . . . . . . . . . . . 455

Chapter 1
Introduction

Abstract The purpose of this introduction is, first, to describe the text of the famous
Russell-Copleston 1948 debate, which we propose to analyze from the vantage point
of two different argumentation theories, pragma-dialectics (van Eemeren, 2018) and
argument dialectics (Marraud, 2020c); next to justify why we chose precisely that
text; and finally, to explain briefly the way in which we performed the analysis,
namely by what is called "adversarial collaboration" (Kahneman, 2013).

The field of argumentation studies is awash with detailed analysis of texts in which
argumentation takes place from various theoretical perspectives. Yet most of those
texts are quite short, say, between a single sentence and perhaps two or three pages,
with an average of as little as 150 words. The only exception known to us is Mauricio
Finocchiaro’s Galileo and the art of reasoning: Rhetorical foundations of logic and
scientific method (1980).
This extraordinary book actually tackles no less a monument to scientific and
philosophical scholarship than Galileo’s Dialogue on the two chief world systems
(1632), in many ways the founding document of the scientific revolution. Although
an astounding achievement in many ways, which this is not the place to discuss,
Finocchiaro does not tackle the impossible task of following the text in all its many
turns and twists, and of going through all the utterances and all the steps contained in
Galileo’s Dialogue. Of necessity, such a detailed analysis would be larger than the text
analysed, and that text is quite long, almost 500 pages in the 1897 Edizione Nazionale.
If our experience is a guide, Finocchiaro would have had to write ten volumes of the
size of his Galileo had he wanted to perform a detailed textual analysis.
Most texts analysed in papers, treatises, and textbooks in critical thinking, informal
logic, and argumentation theory are, besides, either made up by the analysts or
extracted from larger written pieces. This is good pedagogical practice, and we have
nothing to say against it. Nonetheless, we have often heard from students the question
whether the various methods of analysis can be equally applied to a longer text, and
in fact we have for a long time been equally curious about how those methods would
fare when faced with a longer and more complex specimen of argumentation. Now,
if a complete text should be subjected to a close analysis in the light of modern
argumentation theory, then, for the reasons outlined above, the selected text would

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 1

F. Leal and H. Marraud, How Philosophers Argue, Argumentation Library 41,
https://doi.org/10.1007/978-3-030-85368-6_1
2 1 Introduction

have to be considerably shorter than a book, at most perhaps the length of a self-
contained book chapter or an ordinary academic paper.
The authors of the present volume happen to teach argumentation to philosophy
students, so it was natural to choose a philosophical text for the purpose. In the
following, we shall describe the text we chose (Sect. 1.1), and then explain why we
chose that text (Sect. 1.2) and how we went about analysing it (Sect. 1.3).

1.1 The Russell-Copleston Debate

From 1946 to 1967, the legendary British Broadcasting Corporation, or BBC, had
three radio programmes or channels. The first (called ‘Home Service’) was mainly
dedicated to the news, to talks and interviews on current affairs and ‘topical’ subjects,
to jazz and that part of the classical music that had achieved popularity. The second
programme (‘the Light Programme’) was addressed to popular music and entertain-
ment. The Third Programme (this was its name) was devoted to less popular classical
music, and most specially to the arts, to science, and to culture generally. Naturally
enough, the first two radio channels were developed with the general public in mind,
but
on Sunday 29 September, 1946, at 6 pm, the BBC launched an experimental new service
called the Third Programme. The Third Programme was radio’s most uncompromising
attempt ever to engage the brain. Over the next forty years, it would absorb most of the
serious content that would previously have been assigned to the Home Service. The hope
was that never again would speakers be urged to treat their subjects ‘in a rather light way’.
(Games 2015, p. 50)

The rest is history. The Third Programme was so well conceived and realized that it
was quickly imitated worldwide, and it indeed became, without exaggeration, ‘the
envy of the world’ (Carpenter 1996).
Bertrand Russell (1872–1970), then in his mid-seventies and the most famous
English-speaking philosopher, was an assiduous collaborator of the Third Programme
from the beginning (Moorehead 1992, pp. 456–458). At the end of a humorous talk
for ‘the corporate customers’ of Hazell Watson & Viney, printers and bookbinders, in
1951, Russell was unfavourably comparing the reach of books to that of broadcasts,
and, probably seeing the dismay in his listeners, quickly said: “Perhaps the Third
Programme will so stimulate the intellectual life of the nation that it will read more
books than ever before.” One can almost see the twinkle in the old man’s eye as he
added: “Let us hope so, since as yet there is no tax on hope” (Blackwell and Ruja
1993, p. xxxix).
Sometime in 1947, a BBC producer called Langford suggested that a debate could
be held in the Third Programme on the question: ‘Is the world created and ordered by
a conscious intelligence and purpose or is it part of wholly self-determined system?’
(Slater 1997, p. 522). We have been unable to ascertain who Langford was, how he
1.1 The Russell-Copleston Debate 3

came up with this question, or what his purpose was in suggesting the debate.1 It is
probably safe to surmise that Langford wanted Bertrand Russell in it, probably on
the strength of his fame as an atheist and a public figure. And so it came that Basil
Taylor, an art historian working for the Home Talks Department of the BBC, wrote
to Russell on 27 November 1947 that the Third Programme had given 40 min for
such a debate (Slater 1997, p. 521).
It must have been a huge problem to find somebody who would dare to face the
formidable Russell in such a debate. Enter Frederick C. Copleston, then a 40-years
old teacher at the University of London (Heythrop College) and a Jesuit priest, who
was not afraid and in fact quite enjoyed debating:
In 1947, as a relatively young man, Copleston bravely accepted an invitation from BBC radio
to debate Bertrand Russell on the existence of God. He felt at the time that he was debating
a national monument, but the exchange was memorable and later published. The following
year he similarly debated A. J. Ayer on logical positivism, and later still, jousted with Ayer
on BBC television on the subject of the self. (Felt 1994, pp. 237–238)
In 1948 Bertrand Russell and I discussed the existence of God on the Third Programme of
the BBC. How many people had been invited by the BBC to debate with Russell before I was
approached, I do not know. Anyway, having been invited, I rashly accepted. If I remember
correctly, the topic, as originally stated by the BBC, was the question whether the world
was self-sufficient. It was natural enough that Russell and I should have interpreted this as
a question about the existence of God. (Copleston 1993, p. 133)

Of course, Frederick C. Copleston (1907–1994) is nowadays for us simply the

historian of philosophy in the English language. Yet in 1947 he had only published a
few papers in obscure journals, a book on Nietzsche, another one on Schopenhauer,
and, most importantly, the first of the nine volumes of his History of philosophy,
just 1 year before (Copleston 1946). As an interesting detail, he had also reviewed
Russell’s History of Western Philosophy, published the same year as Copleston’s first
volume (Russell 1946; Copleston 1947; Madigan 1987, pp. 418–419, 436).
In one word, at the time that he received and accepted the invitation to debate
Russell on Langford’s question, Copleston was at most a promising young man who,
after studies at Heythrop and ordained as a Jesuit priest, had continued his education
in Germany, and planned to pursue his doctorate in Rome, a dream he could not
fulfill on account of the war (Cameron n.d.). Anyway, in 1947 he was practically an
unknown quantity, nowhere Russell’s match in age or prestige. But it is a tribute to
his performance and debating skills that none less than the Director General of the
BBC has this to say in a memo dated the next day:
Copleston is a find. He was the first man I have heard who could stand in the same ring as
Russell on these matters and not seem out of place.
(…)
The outstanding success of the broadcast was its temper. A feeling of intellectual integrity
allied to a determination to examine any proposition the other man put forward has not come
through so forcibly in any previous broadcast I have heard. Whether all the credit for this

1The producer might be A. W. Langford, who seemed to have been working for the BBC since
1946 (Games 2015, p. 363).
4 1 Introduction

should go to the two speakers I do not know, but whoever was responsible deserves very
high marks. (Haley 1948)

This is high praise indeed. As for the last part of the memo, it emerges from
the extant documentation, now at the Russell Archives, that this debate was very
carefully prepared. In the Talks Booking Requisition, signed by Basil Taylor on
7 January 1948, Langford’s somewhat convoluted question is registered simply as
‘The Existence of God’, for 28 January from 9:45 to 10:25 p.m., which means 40 min
(Taylor 1948a). In an ‘extract’ dated the same day, it is specified that the debaters
will
attend a script conference lasting about three hours, and the script from which they will
broadcast will result from this. It will, however, almost certainly be necessary for them to do
a considerable amount of work on the transcript arising from the script conference. (Taylor
1948b)

The script conference is described by Taylor in his letter of 27 January to Russell

as a matter of taking down by dictaphone what Copleston and Russell would say in
such a way that the final script would be “an edited version of the transcript” (Slater
1997, p. 521).
The booking for radio time was thus done 3 weeks in advance of the broadcast.
On 8 January there is a letter from Taylor to Russell, inviting him for lunch with
Copleston at 12:45, an hour and a half before the script conference on Wednesday 14
January, which was scheduled for 2 h starting at 2:15 (Taylor 1948c). Either because
he was the younger man or because he was going to defend the affirmative in the
debate, Copleston seems to have been asked to produce some notes for Russell to
read and respond to. The set of Copleston’s notes seems to have been lost, but we
can reconstruct both their general content and Russell’s reaction to them by a letter
to Basil Taylor on 15 December 1947:
With regard to Father Copleston’s points, I should reply:

(1) The importance of the question of God’s existence is not in debate, but has no bearing
on the question of the true answer.
(2) I do not think the word ‘transcendent’ has any clear meaning. In any case, I do not see
how human conduct can imply anything as to the rest of the universe.
(3) The proofs of the existence of God, so far as known to me, rest on an antiquated logic
or an illegitimate objectification of human ethics.
(4) I cannot admit ‘necessary being’, or indeed anything ‘necessary’ in the realm of
existence.
(5) Does not arise unless God’s existence is admitted.
(6) I should wish for a definition of ‘intelligible’. I think, as understood by Father Cople-
ston, there is no reason to think anything ‘intelligible’, and, if there were, much human
experience would point rather to Satan. (As quoted in Slater 1997, p. 522)

Russell’s observations are pretty clear and allow one to imagine what Copleston must
have written:
1.1 The Russell-Copleston Debate 5

(ad 1) Given that some listeners to the Third Programme might be inclined to
say that the question of God’s existence was purely academic and in fact
otiose, it was important for Copleston to wring from Russell an admission
that this was not so, but on the contrary, one’s position on the issue always
has weighty practical consequences. In spite of Russell’s attempt to elimi-
nate this question from the discussion, Copleston introduced it, right at the
beginning of the broadcast (see Appendix, Turns 5 and 7), and so forced
Russell to make a partial concession.
(ad 2) It is far more difficult to make out what Copleston said which provoked
Russell’s reply. The word ‘transcendent’, is in any case, used sparingly
and late in the actual debate (see Appendix, Turns 53, 73, and 121). None
of the appearances of the word elicits Russell’s protest at all. As for the
connection between human conduct “and the rest of the universe”, it is the
moral argument (Turns 92–136) that is probably meant.
(ad 3) This is perhaps the single most important note, for it declares the ultimate
ground of Russell’s opposition, not so much to Copleston’s two main argu-
ments in themselves as rather to the philosophical worldview that under-
pinned those arguments. For Russell, all metaphysics must be based on
logic. Given that modern mathematical logic is notably different in many
important ways to the traditional logic that Copleston was trained in, it was
only natural to reject the metaphysics based on it. The ‘antiquated logic’
that Russell refers to will become one major point of contention during the
debate: it is supposed to vitiate the way Copleston argues, and the latter will
protest against what he thinks is something like an usurpation of functions,
in which logic is meant to replace metaphysics. As for the moral argument,
Copleston defended a non-naturalistic version of ethics, whereas Russell,
after a short spell in which he followed Moore’s anti-naturalism, tried all
his life to produce some form of ethics in which moral rules and norms are
a matter of convention and convenience, with some survival value at most.
(ad 4) This is perhaps the point about which Russell makes the greatest fuss during
the debate. That God is a being that not only exists, as everything else,
but exists necessarily, is, of course, central to traditional Roman Catholic
theology. Given this centrality, there is no way in which the clash of world-
views on this point could have been attenuated or avoided. Russell could
perhaps have been more straightforward by saying that anything that exists
might not have existed, which is something that Copleston would readily
have agreed on, adding that this is what he meant by using the term ‘contin-
gent’. Russell will stubbornly reject the term ‘contingent’, even though he
could have accepted it as shorthand for the italicized proposition above. At
that point, the exact point of disagreement would have been clearly revealed:
the further step from the italicized proposition to the conclusion that there-
fore, there must be something which made it exist is not allowed by Russell.
The further question, however, concerning the grounds for prohibiting that
step, and in particular whether those grounds are logical in nature, could
have been discussed more fully in the debate.
6 1 Introduction

(ad 5) Russell’s note obviously refers to a question raised by Copleston which

would only become debatable if God’s existence is first admitted. Given that
Copleston’s question, whatever it was, cannot be the same practical issue
mentioned in (1), we can only surmise that Copleston, in his original note,
wanted to discuss those attributes of God that go beyond the metaphysical
argument, namely ‘that He is personal, good, and so on’ (see Appendix,
Turn 137). For in fact, when Copleston tries to base the love of the good on
the love of God, he admits that “the validity of such an interpretation of a
man’s conduct depends on the recognition of God’s existence” (ibid., Turn
93). It is difficult to imagine, on the basis of the debate as it was actually
broadcast, what other question might have been in Copleston’s mind.
(ad 6) This is a curious note. On the one hand, Russell wants a definition of ‘intel-
ligible’ as applied to what is going on in the world, which means that he
fails to understand Copleston’s meaning. On the other hand, the phrase “as
understood by Father Copleston” implies that Russell knows what the word
means for Copleston. When we consider that the context in which these
matters are discussed in the debate can only be the moral argument (and
not, for instance, the argument from design, which was not at issue in the
debate), it is quite likely that the two philosophers are talking about different
things: whilst Copleston is referring to the lofty aspirations of Christian-
inspired moral thinking, Russell prefers to focus, with a joking reference to
the works of the devil, on the conduct actually exhibited by the majority of
people.

As readers of the debate can check for themselves, the only extant set of notes made
prior to the script conference, namely Russell’s, give us a very good glimpse, almost
an abstract, of what the debate actually contains.
Now, the script conference is supposed to have been a kind of general rehearsal
for the actual debate, 2 weeks in advance of the broadcast. The debaters would,
of course, have made notes as the discussion went on, and the conversation would
have been carefully recorded in order to produce a faithful transcript for the debaters
to read, correct, and modify, so as to have, at the end of the day, a proper script
to be read from during the debate in real time on the day of the broadcast. This
was the “considerable amount of work” which Taylor predicted was going to be
necessary. It is pretty clear that the idea behind the script conference was to prevent
wild improvisation, in part because of the time constraints of the debate, but possibly
also to adapt its unavoidably somewhat technical contents to the educated public of
the Third Programme.
Fortunately, we have Copleston’s recollection of what transpired during the script
conference:
Russell and I were set down on either side of the microphone and asked to get on with it.
Finding myself confronted with an elderly philosopher with a worldwide reputation, I was
naturally somewhat nervous. However, we managed to keep going, even if at one point it
seemed that the discussion might dry up. (Copleston 1993, p. 134)
1.1 The Russell-Copleston Debate 7

It seems that the script conference was pretty impressive, because on 15 January
Taylor wrote to Russell to thank him for the manuscript of his notes, and by the way
commented that “the [planned] broadcast is, if I may be impertinent enough to say
so, most extraordinarily interesting and it should not only (obviously) be valuable in
itself but make a highly provocative opening for the series” (Taylor 1948d).2 Anyway,
the invitation for lunch had the obvious purpose to introduce Copleston to Russell,
so as to break the ice; and the script conference was arranged in such a way that they
could talk freely and at length, in order to shorten it later and at leisure, with the help
of the transcript, so as not to overstep the 40 min allowed by the Third Programme
officials.3
Alas, the plan failed in a bad way. Writing on 19 January, Basil Taylor broke to
Russell the unpleasant news:
Dear Lord Russell,
I enclose a copy of the transcript of last week’s discussion. I am afraid that a most unfortunate
accident occurred last Wednesday—our Control Room were feeding the discussion into the
wrong dictaphone for the first twenty minutes and consequently no transcript is available of
what was said during that time. It was not until afterwards that this information was given
to me.
I fear that it will be exceedingly difficult to make a script of the first part of the discussion,
as the argument did spring from much of what was said in the first quarter of an hour. There
seem to me two alternatives—either that we should record the first part of the discussion
impromptu from notes made at the rehearsal beforehand, and that the second half, on the
moral justification for the existence of God should be read from a script, or that I should make
some attempt to concoct a script for the first half and that Father Copleston and yourself
should attempt to supply the material of the first bit impromptu.
I have had an opportunity of speaking to Father Copleston about this and he is prepared to
adopt either procedure. I am extremely sorry that such a tiresome technical fault should have
occurred.
In the meantime I will proceed with the editing of the second half of the script and perhaps
you would let me know if you would prefer me also to make a version of what remains of
the first half. In either case I hope to have this finished by Wednesday or Thursday.
Yours truly, etc. (Taylor 1948e; see also Slater 1997, pp. 522–523)

2 It is not quite clear what ‘series’ Taylor refers to. In any case, with the manuscript of his notes,
Russell had asked ‘if any more explanation were desirable’. Taylor’s letter contains two long para-
graphs suggesting many clarifications about topics that never made into the broadcast: the doctrines
of the ‘Philosophic Radicals’, those of Bentham and Malthus, political liberty in ‘the intellectual
and artistic spheres’, and some references to past European history that could be confused with the
more recent events of World War Two (Taylor 1948e). The evidence thus indicates that everybody
was very conscious about the need to prune the debate from obscure and complicated allusions.
3 Copleston (1993, p. 134): “On the appointed day I turned up in the foyer of Broadcasting House.

Though I had not met Russell before, I immediately recognized him from his photographs. He was
sitting in a corner, and I went up and introduced myself. In the course of conversation he expressed
the hope that I would not introduce emotion into the debate.” There was obvious prejudice on
Russell’s part. “From my memory of the 1948 debate…, Russell was a bit ruffled at the end, as I had
suggested (or he thought I had suggested—I forget which) that he was a logical positivist. He was
not of course.” (From an interview with Kenneth Blackwell on 10 May 1976, McMaster Russell
Archives.).
8 1 Introduction

Copleston’s recollection of what happened fills in some details, which help

understand certain aspects of the debate:
What we said was supposed to be recorded by invisible machines, after which it would be
typed, texts being sent to us for boiling down to the requisite length. As things turned out,
the performance of the machines was defective, and we received texts interspersed with
considerable blank spaces, with the request that we would reconstruct omitted passages in
time for our next meeting, to be followed by the actual recording. As one might expect in
the case of a relatively young and unknown person, I spend a good deal of time and trouble
on reconstruction and, incidentally, in using what opportunities presented themselves for
improving my position. Russell, however, did nothing at all. The result was that when we
met on the appointed day, we had to work at the text through lunch. Russell maintained that
we were supposed to have parity of space, and that as, owing to my work of reconstruction,
I had more material than he had, I should have to do most of the cutting. He admitted that
he was to blame for not having attempted to reconstruct what was missing in his sections of
the text, but he none the less expected me to do the cutting. Short of creating a scene, there
was nothing to be done but for me to take a pencil and excise passages while trying to eat
my lunch. Russell just modified a few of his statements. After lunch the agreed result was
recorded. (Copleston 1993, pp. 134–135)

It might indeed be speculated whether this technical mishap, together with

Russell’s neglect of his duty, may account for the undeniable fact that the first half
of the debate, about the metaphysical argument, contains far more technical details
and exhibits far more improvisation (even embarrassment at some point) than the
second half, featuring the religious and the moral arguments. This discrepancy was
already noted by Haley in the memo referred to above:
[Last night’s debate between Copleston and Russell] was not quite well proportioned. The
argument about contingencies [sic], first causes, and the Necessary Being was too long for
hearing.
It was also perhaps a little too difficult for the broad run of even a Third Programme audience.
Kant, Plotinus, Sartre, and Leibnitz [sic], quoted and poured at that rate took some absorbing.
…
… Was not the broadcast really on two subjects to two different audiences? Certainly the
second half was far more intelligible than the first.
It was also the better half. The greatest tribute one could pay to it was that as it developed
one got terribly anxious lest the debate should end at 10:25 as the Radio Times said it would.
There could be no better proof of its effectiveness than that. (Haley 1948)

As we explained before, the debate was scheduled to last only 40 min, from 9:45
to 10:25 in the evening. Yet, there is a fragment of the original recording available on
YouTube that lasts about 15 min, and after comparing the contents of that fragment
with the transcript, we estimate the duration of the broadcast to be close to 60 min.
As the Director General of the BBC himself correctly notes, it is a great tribute to
the interest awakened especially by the second half of the debate, that whoever was
in charge that night did not dare to cut it short in favour of whatever else was coming
up next in the radio schedule.
This interest in the Russell-Copleston debate has been kept alive for years and
decades since it was broadcast in 1948. A transcription of the debate was published a
few months later in Humanitas (Slater 1997, p. 521), an obscure journal published by
1.1 The Russell-Copleston Debate 9

left-wing Dominicans from Manchester (Corrin 2013, Chap. 8). It has been reprinted
many times later, including as a chapter in Russell (1957) and in Hicks (1964), apart
from quite a few anthologies. The transcription was soon translated into Spanish in
1949, German in 1950, and Norwegian in 1957 (Madigan 1987, pp. 420–421). Both
the English text and some translations are available on several internet pages. As
for the recording of the debate, it seems to have been much in demand, for it was
rebroadcast in April, 1959.4 An excerpt of about a third of it was used for an Open
University course, and a copy of that excerpt has found its way onto YouTube. A
copy of the whole recording must be in the BBC archives, but due to the coronavirus
pandemic we have not obtained access to it.5
Russell himself seems to have had no further interest in the matter, since his auto-
biography (1967, 1968, 1969) contains no mention whatsoever either of the debate
itself or of Father Copleston or indeed of his relationship to the Third Programme.
Given that Russell’s autobiography comprises three big volumes and given that there
are many far more trivial episodes in Russell’s life than this debate, the exclusion
of any mention of what was, after all, a famous event in Britain’s intellectual life,
does indicate that Russell did not want to keep that particular memory alive. Cople-
ston, however, did leave a detailed account of the debate that we have used here and
elsewhere in the book (Copleston 1993, pp. 133–138).
As for the secondary literature, we are sad to say that it is practically non-existent.
Apart from a book review of a translation of the debate, in which some arguments
are paraphrased and cursorily discussed (Henderson 1980), we have found nothing
but passing mentions of this or that aspect in papers and books, without any close
textual analysis. In our experience, however, this old debate still lives, in the present,
in textbook anthologies as well as in the heads and mouths of undergraduate students
who find it today as exciting as did the original listeners in 1948.

1.2 Why This Debate

Let us start with the fact that professional philosophers have not paid any close
attention to the Russell-Copleston debate. Some of them may have little interest in
the question itself of the existence of God, or at least in the importance of either the

4 Russell seems to have been “glad it is considered worthy of this repetition”. This piece of infor-
mation on the debate can easily be retrieved from the Russell Archives at McMaster University, by
just typing ‘Copleston’ on their excellent search engine (https://www.bracers.mcmaster.ca/).
5 We used the fragment on YouTube to do a fresh transcription of the parts conserved, but a task for

the future would be to obtain access to the full recording in order to make sure of the exact wording
of the debate. All transcriptions, published and posted on the internet, have approximately the same
content. However, a Russell biographer reports many things that are certainly in none of the printed
versions (Moorehead 1992, p. 459).
10 1 Introduction

cosmological or the moral argument that feature in that question.6 But even those
who take seriously either the question or the arguments or both, must consider the
debate per se as rather mediocre and unsophisticated from a technical or professional
point of view.
Philosophers are justified in feeling proud of the two great branches of scholarship
they have created. On the one hand, nobody can feel but awe before the historiography
of philosophy with its impressive philological developments, its critical editions, its
long tradition of ever more refined translations, and the continuous growth of careful
and erudite interpretation as evinced by the sheer number of commentaries and
monographs on all sorts of authors, periods, topics. On the other hand, we must also
admire the degree of technical sophistication attained by what is still usual to call
analytic philosophy, in part thanks to developments in mathematical disciplines, such
as logic, set theory, decision theory, game theory, and the calculus of probability, but in
part simply by the enormous amount of rigour in argumentation in which philosophy
students are disciplined from their first term, and which makes publication in the
best peer-reviewed journals of the field such a hard endeavour.
It is interesting that the two parties in the debate are paradigmatic representatives
of the two branches, great scholars and even pioneers in their respective branch. Yet, a
philosopher trained in the historiography of philosophy may admire Copleston’s nine-
volume History of philosophy whilst pointing out that there is little erudition in what
he brings to bear during the BBC debate. As for Russell, we have all been told how
original and daring his paper ‘On denoting’ (1905) is and how he taught us, like no one
before, how to use the marvellous tools of the predicate calculus to solve complicated
issues in ontology, epistemology, and the philosophy of language. And yet there are
no logical fireworks in the meagre offerings timidly handled in front of the BBC
microphones. Therefore, all things considered, we agree with the tacit judgment of
posterity bestowed by professional philosophers upon the Russell-Copleston debate
in the past 70 years.
Whatever the case may be in those exalted circles, we nevertheless find the debate
enormously attractive from the perspective, and for the purposes, of argumentation
theory. In fact, from that perspective and for those purposes the Russell-Copleston
debate is a true, unrecognized jewel. We want to go through all the reasons for our
partiality and we hope that, by the end of our course, we shall have convinced most
of our readers. In particular, we beg them to remember, whilst reading the following
list, that the most important mission of argumentation studies is to educate the young.
It is in the light of that task that we find our choice most compelling.
i. First of all, it is an actual public debate, where what they say on the occasion
is recorded and will be transcribed for other people to read. Philosophers, of
course, discuss all the time, but it is not very common that they participate in
a debate with those features. And when they do, it is quite common to see the
senior party dominate the proceedings, whereas the other parties behave rather

6 In the debate there is also talk about the religious argument, but that argument receives little more
than a perfunctory hearing there. Russell could even be argued to have laughed it off. Besides, it
sorts of blends into the moral argument around Turn 92 (see Appendix for the text).
1.2 Why This Debate 11

submissively. In fact, this happens even when the discussion is not recorded!
In the Russell-Copleston debate there was clearly a senior-junior situation.
Russell almost doubled his debating partner in years; and his philosophical
achievements were (and still are) justly considered far more significant than
those of Copleston. However, Copleston, whilst treating his partner with great
respect, does not let himself be imposed or overawed. In an exhibition of true
Socratic humility, he asks Russell many more questions than the other way
around; and most of those questions are actual requests for information or
agreement, not disguised assertions. At the same time, Copleston is very active
all the time, putting forward standpoints, arguments, and counterarguments
for consideration and discussion. As Haley quite rightly said at the time, he
“could stand in the same ring as Russell on these matters and not seem out
of place”. Copleston may make mistakes, as any discussant is bound to from
time to time, but he is never just going through the motions and avoiding
trouble out of timidity or amour-propre.
ii. Our second reason follows directly from the first. In papers and books within
argumentation studies, most of us think most of the time of dialogue as the
primordial field where argumentation takes place. Yet very often we analyse
texts which are not, on the face of it, dialogical. There is much to be said in
favour of the position that any argumentative text is in some large sense part of
a dialogue. There is, strictly speaking, no monologue in argumentation, for we
all argue with or for an interlocutor or a group of them. Well and good, but we
have all listened to students’ complaints about the enormous gap separating
our talk of dialogue, discussion, and conversation from the solitary pieces
of writing we use to apply our concepts and methods. The Russell-Copleston
debate is not only a real dialogue that took place in time and space, but we can
actually listen to at least the excerpt of it available on YouTube. In fact, one of
the questions that come up in argumentation studies is that many a discussion
between two people is less about one party convincing the other than about
convincing the audience. The debate examined in this book is probably a great
example of that, once we agree that both philosophers at best hope to win the
BBC listeners to their side of the issue.7
iii. The two disputants are very intelligent, very articulate, and very courteous.
The combination of these three qualities must be exceedingly rare. We all
known that some people are intelligent but not quite articulate or articulate but
not particularly intelligent. That Russell and Copleston were both is probably

7 “[I]t is probably true to say that such debates tend to leave people’s beliefs what they were before.
The fact that someone judges that one or other participant got the better of the contest does not
necessarily affect his or her personal beliefs in regard to the substantial issue, here the existence of
God. In other words, the Copleston-Russell debate tended to be a performance, in regard to which
listeners could award good or bad marks to the participants as they thought fit. (…) In confrontation
with Bertrand Russell I had the impression that each of us got down into his own trench, so to speak,
and sniped at the other over the parapet. Neither was going to change his position. It was a question
of who could snipe the more effectively, in the estimation of the audience at any rate.” (Copleston
1993, pp. 137–138.).
12 1 Introduction

out of the question, even though some readers may differ on the question of
who exhibited more of one or the other quality. Apart from that, it is most
noticeable, both in reading the transcription and in listening to what is available
from the original recording, that they are both perfect specimens of the genus
British gentleman. To give just one piece of evidence: although we do not
have the entire recording, of course, so that it is hard to be sure, it seems that
there are only two clear interruptions, incidentally both by Russell (Turns 74
and 132).8 Moreover, one of those rare interruptions is actually announced as
such, and with the utmost politeness (see Appendix, Turn 74). The excellent
combination of high intelligence, articulacy, and courteousness evinced by
Copleston and Russell makes this debate an example worth studying and
perhaps even imitating. In particular, the ideal of reasonableness developed
in the pragma-dialectical model of critical discussion is hard to keep for us
human beings. The Russell-Copleston debate, however, comes pretty close,
perhaps as close as it is possible for mortals.
iv. The content of the debate is a topic that has never ceased to be of general
interest, no matter what position we have about it. For even those who look
at the questions of God’s existence with total indifference, or are lukewarm
about its importance, must perforce recognize that it is not just one issue
among many others, both historically speaking and with regard to the bulk
of humankind. In fact, it has, so to speak, recently become a hot topic again
thanks to the good offices of the new militant atheists (see Hitchens et al. 2019).
Every teacher has confronted the glassy eyes and half-suppressed yawning
of students trying hard to keep away through one more letter to the editor
or political speech. This is also the case with some philosophical texts that
should perhaps interest the kids but actually do not. For some reason, this is
never the case when discussing the existence or non-existence of God.
v. The level of discussion in the Russell-Copleston debate sits comfort-
ably between technical sophistication—either in mathematical logic or in
scholastic metaphysics—and a dumbed-down version. The public that the
creators of the Third Programme had in mind, that elusive beast known as the
‘educated layperson’, is, we think, exactly right for a course on argumenta-
tion theory. If the students have no proper training in philosophy, a few things
can be explained to them in a short time and without entering into too many
details. If they do, the debate will spur them on to show their prowess and
clarify things for the benefit of their fellow students. Many other texts we have
used in class are far more demanding than the Russell-Copleston debate, but
this does not imply that that debate is by any means philosophical light fare.
On the contrary, it involves some of the toughest and most intriguing issues
both in analytic philosophy and in the history of philosophical doctrines.

8 In fact, a timid attempt at interrupting Russell, on Copleston’s part, is detectable in the recording
(5 45 of https://www.youtube.com/watch?v=Kz2GjKPbQds&t=87s). The attempt does not go
anywhere, for Russell keeps talking.
1.2 Why This Debate 13

vi. It has a manageable size. At about nine thousand words, it has the length of
many academic papers in philosophy, yet its very nature as a dialogue makes it
much less severe in style. From Plato’s Socratic dialogues, extraordinary from
a literary point of view, to much less polished contemporary pieces by the likes
of John Perry (1978), Paul Feyerabend (1991), or Timothy Williamson (2015),
we can certainly find lots of similar-sized philosophical discussions; but the
truth of the matter is that they are all made up. We feel a quasi-monologic
author pulling the strings from behind. Man merkt die Absicht und man ist
verstimmt, to quote Goethe: “one is aware of the intention and the spell is
broken”. In contrast, Russell and Copleston are real characters actually having
a discussion. Of course, there was some preparation beforehand, yet there is
no detectable stiffness either in the extant recording or in the transcription.
This has a tremendous attraction for students. On the other hand, written
and printed philosophical discussions of the sort published in the excellent
new series of “Great Debates in Philosophy” and “Contemporary Debates
in Philosophy” (published by Blackwell and Wiley, respectively) are great
materials in many ways but incomparable in spontaneity and liveliness to our
debate. Compare, for instance the Russell-Copleston debate with Smart and
Haldane (2003).
vii. All philosophical debates feature, in some form and to a greater or smaller
extent, a clash between global worldviews (Weltanschauungen). When the
discussants share many points of agreement, the clash can shrink to insignif-
icance, e.g. two historians of philosophy who believe in Plato’s ‘unwritten
doctrines’ or two analytic philosophers of mind who are hard-headed inter-
nalists. Philosophers will even in such cases have many bones of contention
to argue about. However, between Russell and Copleston opens a real abyss.
This makes it a very interesting case study, for the struggle is never just about
this or that technical issue. Consider the striking description given by Jay
Rosenberg in his textbook on philosophical argumentation:
A [philosophical] disagreement is transferred from conclusions to premises and from
premises to presuppositions, ultimately pulling in whole complex families of beliefs
and commitments. Much of what is initially disagreeable about the enterprise may
become less so when seen in this light, that is, when the details of argument directed
at seemingly inconsequential puzzles come to be appreciated as tactical moments
in a larger philosophical development. Ultimately, any challenge is addressed not to
this or that individual thesis but to the consistency and coherence of a whole family
of beliefs in which the thesis is embedded.
What is genuinely at issue in a philosophical dispute, then, is not a particular state-
ment or claim but rather a rich, more or less systematic world view. A particular
philosophical encounter is like the collision of two icebergs. What lies beneath the
surface is larger than, and gives shape and force to, what is visible above the waters.
(Rosenberg 1996, pp. 50–51)
14 1 Introduction

A careful perusal of the Russell-Copleston debate will confirm the accu-

racy of this metaphor. The clash of worldviews makes it particularly inter-
esting reading and provides challenges for the analysis of argumentation more
exciting than those we find in many other philosophical writings.
viii. Philosophy is all about questions. When philosophers argue, they invariably
start from a question, such as whether God exists, and often finish with another
or several other questions. In the transition from the initial question to the
final ones, what we always witness is a proliferation of questions that rapidly
emerge, one from another, as soon as the parties begin to argue. Now, some
philosophers, such as Plato, Aristotle, Augustine, Aquinas, Descartes, Kant,
or Heidegger are upfront with their questions, whilst others tend to be perhaps
less passionate about them or more reserved, so that one is hard put to identify
what they are (think of Parmenides, Epicurus, Anselm, Hegel, or Quine). But
this is just a matter of style, for they all spend all or most of their time going
from one question to the next. The debate between Russell and Copleston
is in part straightforward in the way it starts with a question; but then the
constantly arising new questions are not always as direct. It is a wonderful
exercise to track them down and make them explicit whenever they are just
touched upon. The task of identification makes that exercise not too hard and
not too easy, and thus almost ideal for pedagogical purposes.
ix. The questions raised in the Russell-Copleston debate are unfailingly impor-
tant and interesting questions and never, as is not unheard of in the philo-
sophical literature, trivial, boring, or easily leading to merely verbal disputes.
Many students who chose to major in philosophy because of the intensity with
which they had been smitten by a philosophical question are surprised to find
themselves having to deal with much less exciting or even comprehensible
stuff. This is seldom the case here. For Russell and Copleston are not dealing
with the married life of philosophers or the choice between corn flakes and
eggs-and-bacon for breakfast. (One finds such examples in certain authors.)
Again, they are not discussing how many angels can dance on a pin or whether
the relation of wetness to H2 O is a matter of supervenience or superduperve-
nience.9 As for verbal disputes, some people believe that philosophy is full of
those but, even if this should really be the case—a moot point—no such thing
appears in this debate. On the contrary, all questions arising from the dispute
are real, and, whilst none of them can be considered unimportant, some have
great depth, e.g. the right relation between logic and metaphysics or the true
status of moral norms.
x. Both the teachers of argumentation and their students are interested in eval-
uation, be it of the arguments themselves or of the argumentative process as
such. Although this book is firmly focused on analysis, our readers will see

9 This is, of course, a joke. The questions used for illustration have a serious side to them (the
relation between spirit and space in the first example, the explainability of emergence in the second
one). The point of the joke is just to evoke the faces of sophomores when first encountering those
questions.
1.2 Why This Debate 15

that we touch on evaluation from time to time, even if perhaps not in the
usual manner. But the point here is that it is very difficult to listen to or read a
debate without asking oneself who won it or whose performance was better.
The present authors read the transcription in their somewhat remote youth,
whilst they were still more or less unconditional admirers of Bertrand Russell.
And so, it was only natural that we were at the time firmly convinced of the
immense intellectual superiority of Russell over Copleston. For us, there was
no question about who had won the argument and whose arguments were
better. It has been a refreshing experience, after all those years, to see that the
situation is far more complicated than that. All in all, these two great repre-
sentatives of the two main branches of professional philosophy are extremely
skilled; but they are both enchantingly human in their missteps, quirks, and
foibles as well. That makes the debate admirable and earthly at the same time.

Some of these reasons may appeal to some readers more than others; but we believe
that together they constitute a solid case in favour of the special interest that the debate
between Russell and Copleston on the existence of God has for argumentation studies
in general, for argumentation theory in particular, and even more for the teaching
of argumentation, informal logic, and critical thinking, especially though not exclu-
sively when our charges are students of philosophy. It is time to rescue the Russell-
Copleston debate from its relative obscurity in the professional circles of philosophers
and to present it to all non-philosophers who have a stake in argumentation.

1.3 Why in This Way: A Word on Adversarial

Collaboration

The field of argumentation studies boasts, if that is the right word, of quite a few
different theories of argumentation, each one of which could in principle be applied
to the Russell-Copleston debate. Some of those theories are the result of decade-
long teamwork; others are much more recent and the affair of one or only a few
researchers. Some theories have enjoyed a lot of exposure in the international field;
others are confined to one or a few countries, depending on how extensively they
use English as the medium of communication of ideas and results, but also on how
successful the authors of those theories have networked and participated in the main
conferences, journals, and publishing houses within the field. Some theories have
managed to produce textbooks for the college student market with lots of exercises
for the training of skills; others are restricted to more or less technical papers and
thick tomes. In some cases, we can identify a crisp and clear theoretical framework,
developed and expounded in detail; in others the theorizing remains a bit fuzzy. Some
authors are very careful to offer many examples, real or invented, in the light of which
one can see how the theory works in practice; others rest content with much more
abstract version of what argumentation is. There is input from traditional disciplines,
such as logic, syntax, semantics, the law, journalism, and intercollegiate debating;
16 1 Introduction

but there is also input from relatively new ones, such as communication, pragmatics,
cognitive science, and artificial intelligence.
In fine, it is a vibrant and sometimes confusing field—which is why, before actu-
ally presenting our analysis of the Russell-Copleston debate, we tried to create a
framework for orientation which may also help the reader to locate our theories.
This is the content of chapter 2. After that, the readers of this book will find two
parts, each of which comprises six chapters and presents a different analysis of the
same text. The truth of the matter is that, although we share many ideas and have read
more or less the same authors, we have had a different experience, both theoretical and
pedagogical, i.e. both in trying to make sense of what argumentation is and in trying
to teach argumentation to university students; and so, naturally enough, each of us
finds himself most comfortable working within a different theory of argumentation.
For Fernando Leal (FL), that theory is a slightly modified version of the pragma-
dialectical model of critical discussion (see Part I, Chaps. 3–8), for Hubert Marraud
(HM) it is a particular combination of argumentative semantics and informal logic
(see Part II, Chaps. 9–14).
The ideal model of critical discussion has been expounded many times, with
greater or lesser detail, and so can be safely considered to be well known for the
intended audience. And so, after a brief reminder of its principal points in chapter 2,
Sect. 2.3, FL proceeds, in Chap. 3, to present only the slight modifications introduced
in the model and to discuss some other particular features, not so much of the theory
as of the method of analysis applied to the debate. That analysis is the subject of
Chaps. 4–8.
The theoretical situation was much more complicated for HM, for once the insights
of argumentative semantics (a branch of linguistics rather than of argumentation
studies) are pressed into the service of a proper theory of argumentation, the ideas of
informal logic are transformed almost beyond recognition (see Chap. 2, Sect. 2.2).
Now, the emerging theory—argument dialectics—has so far been mainly expounded
in Spanish, due to the fact that the analytic focus on the special properties discourse
markers have in different languages. The few papers published in English (Marraud
2015, 2020a) have not given the theory enough exposure, which is why HM was
forced to make it the object of a long, detailed exposition in Chaps. 9–13. Once the
theory is understood, the actual work of analysis does not need so many pages. It is
the subject of Chap. 14.
Each part was written independently by its author, although we shared files and
commented on each other’s work, and hopefully learnt enough from those comments
to improve our solitary work. When we planned this book, we hoped to be able to
create a unified theoretical framework. Alas, we have not reached that point, and
the problem of theoretical unification has proved to be more difficult than we had
reckoned before embarking on our analyses. We still believe that our two frameworks
are complementary rather than incompatible; more concretely, that it is possible to
use HM’s argument dialectics to reconstruct all the argument exchanges that take
place, either in the argumentation stage of a critical discussion or in the other stages.
However, this is a task for the future. In the present, we can only conceive of our
1.3 Why in This Way: A Word on Adversarial Collaboration 17

respective toolboxes as alternative theoretical frameworks and analytic methods.

Hence, the idea of adversarial collaboration.
The term, which has the paradoxical air of an oxymoron, was invented by Daniel
Kahneman, an experimental psychologist who was rewarded with the Nobel Prize in
Economic Sciences in 2002 for his contributions to the understanding of systematic
errors in human reasoning when confronted with problems involving probabilities
(for a quick-and-dirty summary, see Kahneman 2013). His results, originally obtained
in collaboration with Amos Tversky, surprised many people and sparked off quite a
few controversies in the social and cognitive sciences (Lewis 2016).
Kahneman tells us that he has occasionally engaged in those controversies and,
as many other scholars before him, has found their outcome highly unsatisfactory:
Some years ago, I decided to do something about this. Unaware that someone else had
preceded me (Latham et al. 1988), I began to champion a procedure of adversarial collabo-
ration as a substitute for the format of critique-reply-rejoinder in which debates are currently
conducted in the social sciences. Adversarial collaboration involves a good-faith effort to
conduct debates by carrying out joint research… Because the contestants are not expected to
reach complete agreement at the end of the exercise, adversarial collaborations will usually
lead to an unusual type of joint publication, in which disagreements are laid out as part of a
jointly authored paper. (Kahneman 2003, p. 729)

The nature of the present endeavour is, at least at this point, not experimental, so
that the procedure could not be followed in all details, as set forth in the “suggestions
for adversarial collaboration” by Mellers et al. (2001, p. 270, Table 1). The following
guidelines, adapted from Kahneman’s “suggestions”, may help understand what we
did, in case others are interested in pursuing this remarkable method:
• Instead of criticizing a different theory of argumentation or different methods of
analysis, try doing joint research on the basis of a previously agreed procedure.
If theoretical or methodological differences are too extensive or deep, it might
be a good idea to appoint an arbiter. In practice, this would have to be a trusted
colleague who is not wedded to either theory/method, so that he or she will be
neutral in case mounting substantive or procedural disagreements threaten the
effort.
• For the joint research to make sense, it should not be just a matter of pursuing a
controversy in abstract terms, wielding pieces of general theory and throwing them
to each other, but agreement should be reached on a particular argumentative text
to which the method of analysis associated with each theory should be applied.
The result of the joint research should be two concrete analyses of the text, made
in such a way that they make clear the commitments of each theory/method.
• For this purpose, make sure to choose an argumentative text that is interesting
and important, both for the adversaries and for the intended audience. The exact
wording of the text must be carefully established, so as to make sure that both
researchers are working on the same text. Even if it could be argued that any
argumentative text, no matter how long, is in some sense a fragment extracted from
a more expansive whole, the piece chosen should be reasonably self-contained
and not excessively technical. If the joint research is successful, more complex
texts might be subjected to a renewed effort later on.
18 1 Introduction

• Both authors should recognize in advance that the results will not be a full reso-
lution of the initial theoretical or methodological disagreement. Moreover, they
should accept that the joint research effort is not a competition with definite losers
and winners, but rather a first step toward the ultimate goal of reaching a larger
and stronger theory/method that feeds from the original ones.
• The last goal is probably, in many cases, an unfulfillable dream. At the very least
then, the researchers should agree on a map of the theoretical and methodological
proposals available in the field, so as to locate the two contending approaches in as
precise and accurate manner as possible. Only in that way, could the researchers
as well as the readers have some measure of the distance still remaining in view
of a ‘complete’ theory of argumentation.

One intangible aspect of adversarial collaboration is that the researchers should be

friends or at least friendly. It would be too difficult to pursue adversarial collaboration
if there is hostility on one or both sides. In our case, this intangible was perfectly
fulfilled here, the present authors having been friends for almost a decade. This
fact, together with the conviction that our differences are not, at the end of the day,
unbridgeable, made unnecessary the appointment of an arbiter to mediate between
us.
Having said that, it is not superfluous to insist that our theoretical and method-
ological differences could not in fact be bridged, at least not to the point of allowing
us to attain a unified approach to argumentation. This explains both the title of this
book and its division into two parts. To complete the task, we have jointly written
a last chapter (Chap. 15), in which we try to lay out the similarities and differences
between the two approaches and a few ideas about how to bring them closer to each
other. One thing is clear: we have learned a lot from the exercise; and it is our shared
hope that the readers of this book will do so, too.

References

Blackwell, K., and Ruja, H. (1993). A bibliography of Bertrand Russell, vol. I. London: Routledge.
Cameron, J. (n.d.). ‘Frederick Copleston.’ Note at the Gifford Lectures website, retrieved on 20
June 2020 (https://www.giffordlectures.org/lecturers/frederick-charles-copleston).
Carpenter, H. (1996). The envy of the world: Fifty years of the Third Programme and Radio 3.
London: Weidenfeld and Nicolson.
Copleston, F. C. (1946). A history of philosophy, vol. I: Greece and Rome. London: Burns Oates
and Washbourne.
Copleston, F. C. (1947). Review of Russell (1946). The Wind and the Rain 4 (1), 57–62.
Copleston, F. C. (1993). Memoirs of a philosopher. Kansas City: Sheed & Ward.
Corrin, J. P. (2013). Catholic progressives in England after Vatican II. Notre Dame, IN: University
of Notre Dame Press.
Felt, J. W. (1994). In memoriam: Frederick C. Copleston (1907–1994). The Review of Metaphysics
48 (1), 237–238.
Feyerabend, P. (1991). Dialogues on knowledge. Oxford: Blackwell.
Games, D. (2015). Pevsner: The BBC years. Listening to the visual arts. Farnham, UK: Ashgate.
References 19

Haley, W. J. (1948). ‘Copleston/Russell discussion.’ Internal BBC memo dated 29 January 1948.
Bertrand Russell Archives, McMaster University, Hamilton, Ontario.
Hick, J. (Ed.) (1964). The existence of God. New York: Macmillan.
Hitchens, C., Dawkins, R., Harris, S., and Dennett, D. (2019). The four horsemen: The conversation
that sparked the atheist revolution. New York: Random House.
Kahneman, D. (2003). Experience of collaborative research. American Psychologist 58 (9), 723–
730.
Kahneman, D. (2013). Thinking fast and slow. New York: Farrar, Strauss & Giroux.
Latham, G. P., Erez, M., and Locke, E. A. (1988). Resolving scientific disputes by the joint design
of crucial experiments by the antagonists: Application to the Erez-Latham dispute regarding
participation in goal setting. Journal of Applied Psychology 73 (4), 753–772.
Lewis, M. (2016). The undoing project: A friendship that changed our minds. New York: Norton.
Madigan, P. (1987). Frederick C. Copleston: An 80th birthday bibliography. The Heythrop Journal
38, 418–438.
Marraud, H. (2015). Do arguers dream of logical standards? Arguers’ dialectic vs. arguments’
dialectic. Revista Iberoamericana de Argumentación 10, 1–18.
Marraud, H. (2020a). On the logical ways to counter an argument: A typology and some theoretical
consequences. In F. H. van Eemeren and B. Garssen (Eds.), From argument schemes to argumen-
tative relations in the wild: A variety of contributions to argumentation theory (pp. 149–166).
Cham: Springer.
Marraud, H. (2020b). En buena lógica: Una introducción a la teoría de la argumentación. [Logically
speaking: An introduction to argumentation theory.] Guadalajara (Mexico): Editorial UdeG.
Mellers, B., Hertwig, R., and Kahneman, D. (2001). Do frequency representations eliminate
conjunction effects? An exercise in adversarial collaboration. Psychological Science 12 (4),
269–275.
Moorehead, C. (1992). Bertrand Russell: A life. New York: Viking.
Perry, J. (1978). A dialogue on personal identity and immortality. Indianapolis: Hackett.
Rosenberg, J. (1996). The practice of philosophy: A handbook for beginners. 3rd edition. Upper
Saddle River, NJ: Prentice Hall.
Russell, B. (1905). On denoting. Mind (N.S.) 14 (56), 479–493.
Russell, B. (1946). History of Western philosophy. London: George Allen and Unwin.
Russell, B. (1957). Why I am not a Christian, and other essays on religion and related subjects.
Edited by Paul Edwards. London: George Allen & Unwin.
Russell, B. (1967). Autobiography 1872-1914. London: George Allen & Unwin.
Russell, B. (1968). Autobiography 1914-1944. London: George Allen & Unwin.
Russell, B. (1969). Autobiography 1944-1967. London: George Allen & Unwin.
Slater, J. G. (Ed.) (1997). The collected papers of Bertrand Russell, vol. 11: Last philosophical
testament 1943–1968. London: Routledge.
Smart, J. J. C., and Haldane, J. J. (2003). Atheism and theism. 2nd edition. Malden, MA: Blackwell.
Taylor, B. (1948a). Talks Booking Requisition dated 7 January 1948. Bertrand Russell Archives,
McMaster University, Hamilton, Ontario.
Taylor, B. (1948b). Extract filed at the Talks Department dated 7 January 1948. Bertrand Russell
Archives, McMaster University, Hamilton, Ontario.
Taylor, B. (1948c). Letter to the Earl Russell, F.R.S., dated 8 January 1948. Bertrand Russell
Archives, McMaster University, Hamilton, Ontario.
Taylor, B. (1948d). Letter to the Earl Russell, F.R.S., dated 15 January 1948. Bertrand Russell
Archives, McMaster University, Hamilton, Ontario.
Taylor, B. (1948e). Letter to the Earl Russell, F.R.S., dated 19 January 1948. Bertrand Russell
Archives, McMaster University, Hamilton, Ontario.
Williamson, T. (2015). Tetralogue: I’m right, you’re wrong. Oxford: Oxford University Press.
Chapter 2
Argumentation Theories

Abstract Argumentation theory is a house with many rooms that may seem bewil-
dering to the newcomer. Given the aims and procedure of this book, as outlined in
Chap. 1, we need to establish a map of the different theories. To draw that map is
the purpose of this chapter. Starting with the major fact of the lexical ambiguity
of ‘argument’ in English, we propose to distinguish three major types of theory,
as they focus on arguments as abstract objects (like propositions or numbers), on
the presentation and exchange of arguments in text and discourse, or on the whole
communicative situation (what we often call a ‘discussion’) in which such presenta-
tion and exchange takes place. The two theories applied in this book to the analysis
of the Russell-Copleston debate belong to the second and the third type.

Theories of argumentation have different targets—as far as we can say, mainly three,
corresponding roughly to the three main meanings of the word ‘argumentation’. Each
target is important in its own right, and so is each meaning; we shall therefore keep in
our analysis all three targets and all three meanings firmly in mind. However, before
going into the description of the concrete methods we use, it is first advisable to
describe with some detail what kind of object each theoretical target is; then to give
a rough idea of the kinds of problems associated with each target; finally, to sketch
the kinds of theories that have thus far tackled those problems.

2.1 Argumentation as Argument: A-Theories

The English word ‘argument’ is notoriously ambiguous. Colloquially, it appears

to occupy the middle ground between a discussion, which can be reasonable to a
greater or smaller extent, and a quarrel, which rarely is reasonable (even if quarrels
are otherwise often necessary to clear the air). In more formal and sedate settings, the
word ‘argument’ refers to an abstract object of a certain sort. This is quite a difficult
concept, which we shall explain in this section. The first sense is a quirk of English
which is not documented in any other Western language, a fact that has caused no
end of trouble, given the status of English as the international language of academia.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 21

F. Leal and H. Marraud, How Philosophers Argue, Argumentation Library 41,
https://doi.org/10.1007/978-3-030-85368-6_2
22 2 Argumentation Theories

We intend to leave completely aside the first sense as irrelevant and keep only the
second sense. Of course, what is of conceptual interest in the first sense will be paid
attention to as we discuss the other two meanings of ‘argumentation’ later on.
The most important fact about arguments is that they can be, and constantly
are, edited, extended, shortened, reformulated, elaborated upon, in short modified
in various ways. Those modifications are sometimes judged to be just different
‘versions’ of the same argument, and sometimes not. Judgments of sameness and
difference can be categorical or put forward as a matter of degree, context, or purpose
at hand; and they can be accepted or challenged; but the fact remains that, whenever
someone produces an argument by means of certain words, we can in general use
other words whilst claiming that they constitute the same argument, or at the very
least we can offer them to the original arguer and ask her whether she accepts our
words as an alternative version. This is just a fact of human communication—and a
very important fact. By taking it into account, we come to a more refined concept of
‘argument’, in which any argument is to some extent independent of any particular
formulation of it. It would indeed be strange to challenge that fact and to maintain
that an argument is identical to, and fixed by, a particular formulation.
This fact underlies and supports our occasional talk of such things as ‘the onto-
logical argument’ in philosophy or ‘the standard argument in favour of free trade’
in economics. These and numerous others can be formulated with endless variations
in such a way that we can reasonably claim to be talking about the same argument.
Of course, there comes a point in which a formulation, intended as an alternative to
given formulations, introduces such changes in the matter that we prefer to say that
we now have a new argument altogether. This possibility, however, does not destroy
the fact that the old argument can be expressed in many equivalent ways. In fact,
we often say that a new version is a clear improvement over older ones in that it
expresses better what the original argument was about.
All these things are well known; and arguments in this sense are far from being
alone in having such a feature. There are many other objects that behave like that,
for instance, propositions, numbers, and theories. The theory of universal gravitation
is not identical with the version which its creator, Isaac Newton, formulated for
the first time. In fact, there are uncountable versions of that theory, and many of
them are considered to be clearer or terser expressions of it. The number 25 can be
referred to by infinitely many arithmetical operations. And, in spite of some measure
of disagreement on the matter, we tend to think that the active and the passive version
of any sentence express the same proposition. Arguments are very much like that. Of
course, there may be some debate as to whether a given set of sentences is still the
same argument as another given set of sentences. But the same is true of theories and
propositions. Numbers are perhaps in a better position, although the fact that some
equations have more than two, and sometimes even infinitely many solutions, seem
to indicate that even in the realm of numbers there is some measure of referential
ambiguity.
Philosophers are, in particular, quite at home with the idea that arguments can be
expressed in many different ways. A recent textbook makes that very explicit:
2.1 Argumentation as Argument: A-Theories 23

[F]or almost any given philosophy class that we took as undergraduates, there were only a
handful of arguments, totaling no more than a few pages of carefully crafted notes, that we
needed to know. We imagined a rolodex of arguments in front of us, which we could spin
through with ease to find the argument and move on. Midterm or final examinations in one
of these classes would be reduced to presenting a philosopher’s argument, followed by a
critique—usually another philosopher’s argument. The ability to state an argument clearly
and concisely, in a term paper, for example, demonstrates that one succinctly understands
the material. The following arguments can be viewed as answers to such test questions…
(Bruce and Barbone 2011, p. 1)

The phrase ‘the following arguments’ refers to one hundred of those curious
objects that can be formulated in different ways and make up the bulk of the book
from which we have taken the above quotation. That book was indeed constructed
by asking about seventy professors of philosophy to produce a particular version of
one famous philosophical argument (in the sense of the word we are describing) that
would be clear, precise and didactic, so that students of philosophy could learn what
they need for their examinations without having to go to the trouble of reading the
originals:
The time has long passed when it was possible for one to read the entire Western philosophical
canon. Philosophy needs new didactic tools to address the fact that the quantity of influential
arguments will increase while the number of hours that a student at any level has will remain
relatively the same. Philosophy as a formal discipline will increasingly need to “get smart”
about how it selects which arguments deserve more attention than others in the classroom
and then how to teach them. Outside of the classroom, there are little-to-no resources that
function as study guides. Detailed study guides are made for everything—the Bible, calculus,
grammar, biology—except for philosophy. There are laminated sheets in bookstores that list
all the standard mathematical equations, sheets that have common Spanish verbs, … but not
one has arguments on the existence of God, free will, or moral responsibility. Many books
present important philosophical arguments, but it is often the case that these books outline
only a single argument or a string of related arguments. Encyclopedias of philosophy are great
for limited descriptions of philosophers and concepts, but there is a need for reference tools
that offer specific arguments. In the end, these secondary sources often bury the argument
in commentary and analysis and do not lend themselves to concise and efficient referencing.
It can take just as long to find an argument in the analysis as it would to go to the original
text. This volume acts as a compact and accessible companion to both sources. (Bruce and
Barbone 2011, p. 2)

Some readers may find the tone of this passage overblown and perhaps frivolous,
as though philosophy boiled down to a kind of grocery store where you can sample
the wares on offer and take what you need from the shelves; or rather like a set of
grocery stores, each one offering more or less the same wares, or overlapping subsets
of a universal set of wares, in different brands and at different prices. Nonetheless,
the conception presented in the above quotations underlies the practice of philosophy
everywhere. The emphasis on arguments is particularly pronounced in Anglo-Saxon
departments of philosophy and is currently associated with the label ‘analytic philos-
ophy’; but in so-called ‘Continental’ departments we find the same thing, just with
different undertones and slightly displaced priorities. Thus, everyone speaks of ‘the
ontological argument’ or of ‘the private language argument’ or of ‘the argument
supporting an ethics of discourse’. And those arguments are not particular passages
24 2 Argumentation Theories

in books in which certain authors (say, Anselm, Wittgenstein, Habermas) have written
certain words. Arguments are rather a kind of abstract object, of the same general
sort as theories, propositions, and numbers.
In the contemporary field of argumentation studies, there are two widespread
views of argument that approach the meaning we are discussing here, without being
identical to it. One comes from informal logic, and more widely from the critical
thinking tradition, the other from speech communication. It may be useful to ascertain
the similarities and differences between our conception and those views.
The informal logic concept is presented, with praiseworthy lucidity, by Ralph
Johnson in a 1981 paper addressed to future teachers of the subject (Johnson 2014,
pp. 53–72). In discussing the question of how to choose a method of analysis, he
proposes a series of criteria that such methods should fulfil. The first of them is ‘a
procedure for the elimination of extraneous material’. The underlying idea is that,
when someone produces an argument (in Johnson’s sense), she will surround it with
all sorts of stuff that don’t belong to the argument. Johnson is focused on what he calls
‘real-life arguments’, in which ‘people digress, throw in asides, get side-tracked, mix
in background information, and so on’. Later on, in the same paper (ibid., p. 65),
and in other papers (p. 116), Johnson will frankly speak of clutter, comprising all
those things that have to be pared away (as the inedible skin of a fruit) or weeded
out (as obnoxious plants or growths) by the careful handling of the text in which the
argument is embedded.
We shall come back to this idea of clutter later on, when we discuss argumentation
as something different from argument; but for now we just want to praise Johnson for
clearly pointing out this operation of elimination which we all in fact carry out, either
on our own or to the benefit of our students; moreover we want to praise Johnson for
his daring to put such an offhand and contemptuous label to the stuff we eliminate.
Calling a spade a spade always has the benefit of clearing the air. In another lucid
passage of the same 1981 paper, he complains about ‘the present state of poverty in
our discipline’, i.e. in informal logic, and says:
We do not yet possess an adequate theory of argument. Unless, that is, one conceives of an
argument as do authors of informal logic texts: a set of statements (sansclutter) among
which certain inference relationships supposedly obtain. It may be true that there are in
the real world of arguments such domesticated little creatures. Anyone who has worked at
all with arguments in their real setting knows that there are a great many ornery beasts whose
logical structure and behavior is much more complicated. We need to know their taxonomy
if we are to have any hope of taming them. As our concept of argument broadens to include
them, the need for a more adequate theory of argument becomes ever more pressing. (Johnson
2014, p. 70; bold lettering added.)

In order to explain the difference between Johnson’s conception and the one we
are proposing, consider Russell’s celebrated concept of number as the set of all
equivalent sets. If each member of a set is matched by one and only one member
of another set, then both classes are equivalent. In other words, they have the same
2.1 Argumentation as Argument: A-Theories 25

number of elements. Now, the concept expressed, in the above passage, by the noun
phrase in bold letters could, for ease of reference, be called a Johnson set.1
(1) Definition. Johnson set = df set of statements, without clutter, among which
certain inference relationships supposedly obtain

Thus, an argument is, within informal logic, just a Johnson set. If the reader bears in
mind what we said before about the fact that we often judge two different formulations
as expressing the same argument, then we can now introduce our concept of argument
as follows:
(2) Definition. Argument = df set of all equivalent Johnson sets.

For definition (2) to work, we would naturally need a definition of equivalence

between argument formulations. As far as we know, there is no work done on this
question. In philosophy of science, various kinds of equivalence between theories
have been discussed; and it is often a considerable mathematical problem to establish
that two theories are indeed equivalent; in fact, several kinds of equivalence have been
postulated, so that two theories can be equivalent in one sense and not equivalent in
another. In argumentation theory, we have not reached that level of reflection and
sophistication. Nonetheless, we do presuppose some concept of equivalence. Johnson
himself, when analysing the operation of ‘eliminating extraneous material’, includes
(ibid., p. 65) two relevant subroutines, namely, rephrasing over indexical statements
(e.g. replacing pronouns by names or descriptions) and reformulating statements ‘to
bring out their meaning’ or ‘to bring on the issue’. These two subroutines are intended
to produce equivalent formulations of the argument. We all know the kind of thing
Johnson is talking about here; there is no point in denying that. In fact, classical
rhetoric teaches that any change to a text belongs to one of four classes: addition,
subtraction, permutation, and alteration. What an analyst of arguments does is, then,
first to identify the Johnson set, which is, as we have explained, a linguistic unit
present in the text. In trying to find out what the argument is, however, the analyst
starts tinkering with the Johnson set by means of
(3) The fourfold method:
a. addition, when words, phrases, or sentences are felt to be missing if the
meaning or the issue is not clear enough;
b. subtraction, when information is deemed to be unnecessary or redundant,
some words, phrases, or sentences are deleted;
c. permutation, when the order in which words, phrases, or sentences as
arranged is thought to prevent the correct understanding of the argument
or parts of it; and

1 Observe that this definition is based in the idea of inference and not on the idea of reasons. Later on,
we shall have occasion to revise this concept of argument and incorporate into it the idea of reasons.
The point is that, by focusing on inferential relations, A-theories produce too narrow a concept of
argument for the purpose of theories of argumentation. See the last paragraphs of Sect. 2.2.
26 2 Argumentation Theories

d. alteration, when the phrasing chosen by the arguer is replaced by some

allegedly equivalent choice of words, phrases, or sentences.
This fourfold method (quadripartita ratio) is part and parcel of our attempts at
finding out what the argument is in a given text. We first detect that there is an argu-
ment in that text, then we identify the parts in the texts that somehow express the
argument, and then we reconstruct the argument by means of the fourfold method.
So, we have a sense of what it means to produce an equivalent formulation of some-
thing far more abstract, even if what we produce may be challenged as non-equivalent
and even if we would have great trouble saying how we do it. Besides, all equiva-
lence sets put forward are relative to context and purpose. Thus, Anselm’s various
Johnson sets can be considered equivalent to each other, or even to the Johnson sets
present in Descartes, within an elementary textbook of the history of philosophy, yet
emphatically not equivalent in a technical paper for a journal in formal logic.
This may suffice to distinguish two senses of ‘argument’ within informal logic,
one that is tied up with a particular formulation (chosen by an arguer for a variety of
reasons) and another that admits of equivalent formulations. In this way, argument
theories, or A-theories for short, differ in degree of abstraction. However, given that
in A-theories of the sort indicated by Johnson, i.e. in theories of Johnson sets, analysts
usually, perhaps unavoidably, introduce some modifications as part of their analyses,
they go some way towards a more abstract conception, although perhaps not as far
as philosophers want to go.
Now, if all A-theorists that come from the logical tradition are comfortable with
the conception just expounded, speech communication theorists, in trying to come to
terms with argumentation as argument, have come up with a very different concep-
tion. For them, the phrase ‘argument as product’ serves to separate it from ‘argument
as process’ and ‘argument as procedure’. This threefold distinction seems to have
emerged slowly in the 1970s from the work of Karen Rasmussen (1974), Wayne
Brockriede (1975, 1977), Daniel O’Keefe (1977), Joseph Wenzel (1979), and Jürgen
Habermas (1981, vol. I, pp. 47–50). The following early statement is particularly
clear:
Before looking for the clues that may lead to the discovery of where “argument” is, perhaps
I should state some of my biases so you may be less surprised if I don’t go instantly to where
you presume I could find the culprit without difficulty. My principal bias is a humanistic point
of view that denies an interest in logical systems, in messages, in reasoning, in evidence, or
in propositions—unless these things involve human activity rather directly. (…) Arguments
are not in statements but in people. Hence, a first clue on the whereabouts of argument:
people will find arguments in the vicinity of people.
Second, argument is not a “thing” to be looked for but a concept people use, a perspective they
take. Human activity does not usefully constitute an argument until some person perceives
what is happening as an argument. Although defining the term on this basis is not as neat
as speaking of necessary and sufficient conditions, seeing argument as a human activity
encourages persons to take into account the conceptual choices of the relevant people. Hence,
a second clue: only people can find and label and use an argument. (Brockriede 1975, p. 179)

Note that this passage could not be translated into any other European language
while preserving talk of ‘argument’. This is due to the quirk of English we mentioned
2.1 Argumentation as Argument: A-Theories 27

before. In French, German, Italian or Spanish, we would have to substitute ‘argumen-

tation’; but, in so doing, the self-confessed ‘humanistic bias’, which leads Brockriede
in this passage to reject the kind of argument informal logicians, like Johnson, are
interested in and looking for, becomes less intelligible.
We approve the motivation underlying the malaise of speech communication theo-
rists (and later of a philosopher like Habermas) in trying to ‘bring people back’ into
argumentation studies; but they run the risk of throwing the baby with the bathwater.
People make arguments to be sure, but once they make them, the arguments take a
life of their own, and they can be contemplated, admired, improved upon, precisely
like ‘things’, i.e. in the same way other abstract objects are treated. Einstein may have
invented the theory of special relativity, but once he published it and other physi-
cists could study it, work with it, changed bits here and there, it was not Einstein’s
theory anymore; it had become the possession of physicists and in the last instance
of humankind; and no particular formulation of it, not even Einstein’s own, had any
exclusive right on it. It is exactly like that with arguments. In fact, arguments, like
theories, change the way we live and think, they change us. A young student who
learns the theory of relativity is utterly transformed by that learning. And the exact
same thing is true of arguments. So, yes: people make arguments, that much we can
concede to Brockriede and all speech communication theorists and all philosophers
of communicative action; but we have to add that arguments make people as well.
When Daniel O’Keeffe (1977) criticized Brockriede’s paper by saying that the
latter had neglected to distinguish between ‘making an argument’ and ‘having an
argument’, and so introduced the well-known distinction between argument1 and
argument2 , Brockriede admitted the charge but added the following:
O’Keefe’s distinction perhaps can be expressed more sharply. His argument1 , making or
advancing arguments, may be represented by the noun “argument” since a thing has been
produced. His argument2 may be referenced better by “arguing” since persons are doing
something. Argument1 focuses on a product made, argument2 on a process engaged in.
(Brockriede 1977, p. 129)
This passage seems to give birth to the conception of argument as product, and
from the context it is clear that the ‘argument as product’ is made—produced—by the
arguer or arguers. But unless arguers are preternaturally aware of what they are doing,
then they may indeed be doing, making, producing something, a set of utterances,
that speech communication theorists find useful to call ‘argument’; but it definitely is
not the same as what we mean by that word. For starters, most people would probably
have trouble identifying the Johnson set in what they said, extracting it from their
whole utterance, let alone analysing it. A Johnson set is something that can only be
said to be ‘produced’ by an argumentation analyst. Of course, some arguers are, or,
after some prodding or training, can become sufficiently aware and reflective about
their communicative actions, so that they can identify their Johnson sets and perhaps
even rephrase them. But, if, as, and when they do so, they have ceased to be just
arguers, they have become analysts of their own arguments. If they can go on and
discuss questions of equivalence between their rephrased argument and the original
version they gave, or indeed other versions offered them, then they would be able to
zoom in on the argument in the sense we mean.
28 2 Argumentation Theories

In sum, we can distinguish three senses of ‘argument as product’, and only the
last one would correspond to the concept we are trying to submit.
(4) Three senses of ‘argument as product’:
a. a set of concrete utterances as the product of an arguer A which an analyst
B can identify as a Johnson set;
b. a Johnson set as the product of the analyst B who has come to analyse the
utterances of A; and
c. an equivalence set as the product of an analyst C, i.e. a set that includes the
Johnson set identified by the analyst B, the paraphrases of that Johnson
set which have been produced by B in the course of her analysis, and often
other paraphrases which C considers of theoretical interest.
Of course, as we already implied, it is possible in some cases that A and B, and
even A, B, and C are the same person. Nonetheless, if the persons may occasionally
be the same, the roles are clearly different; and if we use the word ‘argument’ for
the three products, that word means something different in each case. In (4a), the
product is just unanalysed talk, nothing else. In (4b), the Johnson set is already
the product of analysis, obtained by extraction from a larger set of utterances that
have been pared away as clutter. In (4c), we have a much more abstract product of
analysis, which implies both paraphrases (changes according to the fourfold method)
and an equivalence thesis across different versions of what is deemed to be the same
argument.2
Terminologically, we say that arguments as in (4b) and (4c) are sets of premises
and conclusion, something that is not the case in (4a).3 As a matter of fact, no
proper A-theory can be produced just on the basis of (4a). It is thus no wonder that
speech communication theorists mention this kind of argument-as-product only to
abandon it immediately as of no interest. It has indeed no theoretical interest. Informal
logicians, however, writers in the critical thinking tradition, and other argumentation
theorists who start from (4b) are in a better position to build an A-theory. Nonetheless,
there will be some severe limits to theorizing as long as research is restricted to the
conception of a Johnson set. In their practice, as they go about paraphrasing their
uncluttered Johnson sets in order to reveal their argument structure or to evaluate their
argumentative merits, they will go beyond (4b), but they will not have an A-theory to
cover that extended conception. Finally, as for (4c), this is the terrain in which analytic

2 The attentive reader will see that (4b) is a starting point towards (4c). The analyst starts with a
piece of text, that she has extracted from a larger text. This is her Johnson set. Then, she starts to add
bits here, subtract bits there, moving words around, and actually introducing new material. As these
modifications progress, new sets of statements emerge that the analyst considers to be equivalent
to the original Johnson set to some degree, in some context, or for some purpose. Thus is born an
equivalence set, which is the abstract object which is the proper target of an A-theory.
3 What we cannot say is that they are sets of reasons supporting a standpoint, for A-theories are not at

all dialectical. Their target is not the activity of arguing, even if the Johnson sets and equivalence sets
can only be extracted from acts of arguing. When we come to the broader sense of ‘argumentation’
in Sect. 2.2, we shall come back to the questions of reasons as a theoretical term on a completely
different level than premises.
2.1 Argumentation as Argument: A-Theories 29

philosophers and historians of philosophy feel at home and move easily around. The
problem, however, is that these scholars are not really interested in argumentation
theory. When they reflect upon what they do as philosophical arguers (sometimes
they call this ‘metaphilosophy’), philosophers of different traditions come up with
interesting ideas and brilliant aperçus about argumentation in philosophy.4 However,
no A-theory has been worked out in all details within philosophy.
So, in A-theory, we are stuck with the variegated collection of insights that we find
in the writings of several scholars. In order to bring some order in this collection, we
propose to consider the kinds of question that arise when considering both Johnson
sets and equivalence sets of Johnson sets. We can distinguish four groups of questions.
(5) Four groups of questions about arguments:
a. Questions of location. Is there an argument in this speech, text, conver-
sation, interview, exchange? If there is one, where is it in the speech,
etc.? Where does it begin and where does it end? These are, first, ques-
tions concerning the sheer detection of what seems to be an argument.
Secondly, we have questions of identification, pinning down the words
and utterances that are involved in, or part of, the argument, so that we can
perform the delicate operation of extracting the argument from Johnson’s
‘clutter’.
b. Questions of meaning. What exactly does the argument thus extracted say?
To whom or what does it refer with its different words and phrases? What is
its scope, its force, its connotations? Is it intended as an understatement,
a hyperbole, as irony, sarcasm, even a joke? What is the content and
amount of background knowledge that we need in order to understand the
argument? What are the issues, the problems, the questions that underlie
and trigger the argument and its several components? Has the argument
any missing premises, presuppositions, tacit or hidden propositions? Are
there any shared assumptions in the argument? Shared with whom? Does it
rest on a conceptual distinction which is not fully expressed? Are there any
terms that should be defined in order to get to the meaning of the argument?
What are they? Is there a false dichotomy lurking behind? How can we
find out? Some of these questions have been asked by argumentation
theorists of different stripes, others are more at home in philosophy, or
else in rhetoric, literary theory, or linguistics. As we try to answer those
questions, we begin to reformulate, rephrase, reconstruct the argument,
by changing and modifying it according to the fourfold method under (1).

4We find these ideas and aperçus, first, naturally enough, in textbooks designed to teach students
how to argue in philosophy (e.g. Cornman et al. 1992; Rosenberg 1996; Morton 2004; Cohen 2004;
Baggini and Fosl 2010; Martinich 2016; Harrell 2016). Then, we find them in books that discuss the
nature of philosophy (e.g. Stove 1991; Glymour 1992; Bouveresse 1996; Williamson 2007; Gutting
2009; Schnädelbach 2012; Stoljar 2017). Finally, there are a few authors who point towards a theory
of philosophical argumentation (e.g. Johnstone 1959, 1978; Passmore 1961; Rescher 2001, 2006,
2009; Powers 2012; Nelson 2016; see also Johnstone 1988).
30 2 Argumentation Theories

Some of these questions easily go over the thin line separating questions
of meaning and questions of value (see below under d).
c. Questions of structure. How is the argument thus reformulated arranged,
organized, exactly? What makes it tick? Has it a special form? How many
forms are there? How can we know? Are there kinds of argument and how
can we distinguish them? Questions like these were originally discussed
by Aristotle in his attempt at understanding the Socratic method as exhib-
ited in Plato’s dialogues. They gave rise to Aristotle’s dialectics, which
in turn produced what we call ‘logic’ (he called it ‘analytics’) as well
as his rhetoric. In the fullness of time, Aristotle’s logic, with some addi-
tions by other logicians, was mathematized, and we inherited a series of
concepts and distinctions (premises, conclusion, inference, consequence,
syllogism, enthymeme, deduction vs. induction, lógos vs. páthos vs. éthos,
and so on) and we added some new ones (linked vs. convergent, coordina-
tive vs. multiple, the Toulmin diagram, conductive, and so on). Aristotle
also taught us about tópoi, out of which were born the argument schemes.
About a decade and a half ago, the progress in computer power allowed
us to draw argument ‘maps’ or ‘diagrams’ of any size or complexity, and
according to differing views of argument structure.
d. Questions of value. Is the argument good, bad, strong, weak, cogent,
sound, even valid? Is it more so than other arguments in the same text
or in other texts? Is it consistent with other arguments? Are the premises
of the argument plausible, probable, acceptable, true? Are they relevant
to the conclusion that we want to draw from them? Does the conclusion
follow from the premises? In which way, with what force? Are there any
considerations that might weaken the inference? Is the argument mono-
tonic or nonmonotonic? Given that arguments are not proofs, what are the
proper criteria to judge their value? For a while, argumentation theorists
relied almost exclusively on logic, and analytic philosophers still feign
to do so, but at some point in the 1970s argumentation theorists at least
decided to go their own way and, without forgetting the important lessons
of modern formal logic, they set out to devise their own criteria of good-
ness. In the beginning as well, argumentation theorist unduly relied on the
traditional concept of ‘fallacy’, but they also came to criticize it firmly and
soundly. We have since then been trying to rescue what is still of value
in that concept without swallowing the incoherent story that goes with it.
Informal logicians and pragma-dialecticians have been especially active
in offering solutions (the acceptability-relevance-sufficiency criterion, the
rules of a critical discussion).
This list is the product of a long history. It is perhaps superfluous to add that,
although there is a kind of natural ordering from a to d in (5), we should not deceive
ourselves by the linearity imposed by listing. Each successive set of questions can
sometimes bring us back to an earlier or later set and modify our answers. It is
2.1 Argumentation as Argument: A-Theories 31

a dynamic set of questions with many loops backwards and forwards.5 What is
important is that the various answers given to those questions is what constitutes the
backbone and substance of our A-theories, targeted to Johnson sets and equivalence
sets, as set forth in (4).
Before going on to the other two senses of ‘argumentation’, it may be of interest to
note that, although no proper secondary literature has hitherto been produced on the
Russell-Copleston debate, the few scattered references to and comments on it have
all been A-theoretical in nature (Rowe 1970; Harrison 1974; Leahy 1976; Henderson
1980; Koons 1997; Landini 2009).

2.2 Argumentation as Argument Delivery and as Argument

Exchange: D-Theories and E-Theories

‘Let us have a look at Paul’s argumentation.’ This sentence may be an invitation to

do what we have just described: identifying and extracting a particular Johnson set
from a text, then finding the argument by means of equivalent formulations of that
particular Johnson set.6 When the above sentence is thus interpreted, ‘argumentation’
means as much as ‘argument’. In fact, often the use of ‘argumentation’ seems to imply
that there may be more than one argument; but still we are invited to look at those
several arguments as such. As opposed to this, the invitation can also refer to a very
different task: studying the precise way in which Paul’s argument or arguments are
presented to his intended audience. The word ‘argumentation’, in this case, does not
just refer to one or more Johnson sets or sets of equivalent Johnson sets, presumably
contained in Paul’s text, but to the argument delivery, to the particular way Paul has
delivered those arguments, to the choices he made in his arguing.
A theory of argumentation that engages in the study of argument delivery will
be called a D-theory, in order to distinguish it from an A-theory. Perhaps the most
obvious difference between both kinds of theories is the fact, discussed above, that A-
theories are bent on extracting the argument from the ‘clutter’ of the text—everything
that is not the argument. By so doing, A-theorists necessarily abstract from the

5 As we shall see in due time, some of these questions, in particular those having to do with particular
ways of delivering the argument, spill over to the next sense of argumentation. In fact, one of the
limitations of A-theories is that the three main rhetorical tasks—invention, disposition, elocution—
tend to be dismissed as clutter. This is not a reason to reject wholesale the A-theoretical approach but
rather an unavoidable consequence of the first sense of argumentation that we have been discussing.
It goes with the territory.
6 In this section, we shall use the somewhat stilted word ‘text’ to refer both to written pieces and to

oral speeches. In fact, the ‘text’ that this book is set to analyze represents a hybrid case. On the one
hand, the Russell-Copleston debate was delivered orally and broadcast so as to be listened to, not
read. On the other hand, the discussants had exchanged written notes before a general rehearsal of
the debate and probably had those notes during the rehearsal (see Chapter 1, Sect. 1.1). This was,
in its turn, recorded, transcribed, and mailed to the discussants before the actual broadcast, so that
we can imagine them using the transcription as they spoke, whilst improvising at certain points. In
Chap. 4, we shall come back to the finer details.
32 2 Argumentation Theories

dialectical situation in which actual people argue with each other. Paying attention
to the manner of delivery implies being interested in the give and take of actual
discussions. Does this mean that D-theorists have no concept of ‘clutter’ like the
A-theorists? No, because D-theorists also find it necessary to do an operation of
extraction. After all, when people discuss, they do more than just offer arguments.
Only a theory of the whole argumentative process, what we shall call a P-theory, is
fully interested in everything that is said during a discussion. But before we come to
that, let us have a look at the features in a text that have to do with the delivery of
arguments.
These features go beyond the list (5) of questions that characterizes A-theories.
They are common to all cases in which somebody argues, although more skilled
arguers naturally take greater advantage of them. However, any use of those features
can make a great difference between two texts supposedly making the same argument.
Although a full list of all the ‘tricks’ in argument delivery is probably not possible,
we can discern some order by considering the following list.
(6) Families of features of argument delivery:
a. The particular ideas the arguer actually uses in order to express her argu-
ment as opposed to all those other ideas which she leaves out of her text,
even actively and carefully avoids to use or mention.
b. The relative position in the text in which the argument appears, the way it is
embedded in the rest of the text, more generally the order and arrangement
of the different items comprising the text.
c. The style in which the argument is couched, including specific syntactic
discourse markers, punctuation (or its counterpart in oral presentation),
emphases, repetitions, insinuations, allusions, fancy words, colloqui-
alisms, and so on.
A careful look at this list will convince the reader that all these choices are unavoid-
able the moment an individual starts to argue with another, what she produces will
have to express certain ideas and suppress others, the expressed ideas will be arranged
in some order, and they will be expressed in a certain style and with a certain choice
of words, phrases, and sentences. This is just in the nature of language.7

7 Again, some arguments include, or even require, some additional linguistic and non-linguistic
signs. First of all, we have intonation, stress, volume, and tempo—phonological properties that
accompany language and are part of language. Secondly, we have, in living speech, gestures, gaze,
body posture, and other significant motions that also accompany language, although they are not
part of it. Finally, we have all sorts of additional signs, which may have or lack linguistic items
in them: pictures, drawings, diagrams, photographs, and other visual props (Groarke 1996, 2019;
Gross and Harmon 2014). It can even be argued that this shortlist may grow much larger and come
to include further sensory modalities (Groarke 2015). Moreover, occasionally some of these non-
linguistic signs may become the main part of the argumentative message, and in certain situations
perhaps the whole of it. This is all part and parcel of human communication, and it makes no sense
to deny it. Another question is when those signs that go beyond the purely verbal mode favored by
A-theorists actually belong to the argument, and when they are more profitably analyzed as part of
argument delivery.
2.2 Argumentation as Argument Delivery and as Argument … 33

The attentive reader will recognize in list (6) the three first tasks of classical
rhetoric: a corresponds to invention, b to disposition, and c to elocution. To each
task corresponds a series of questions which would be longer to enlist than (5). Any
good handbook of rhetoric, from the classical Greek and Roman ones, to modern
compendia such as Lausberg (1963) can be seen as raising and answering such
questions.8
It is indeed well known that each task in list (6), together with its set of questions,
did merit a great deal of attention in the writings of the old rhetoricians; to which we
only add that they all are equally important to D-theories. Notice, however, that, as far
as A-theories are concerned, the phenomena briefly referred to in (6) are often of no
importance whatsoever. In fact, the tendency of A-theorists is to consider rhetorical
devices (a term often perceived as loaded) to be either pure clutter or at best an
annoying obstacle that needs overcoming in order to zoom in on the true content of
the argument. This is the case of many of the questions of meaning described under
(5b). Thus, if an A-theorist believes that an utterance is ironic or metonymic, then, in
order to reconstruct the argument, she will often feel the need to replace the original
utterance by a literal rendering; if she encounters a rhetorical question, she will
substitute the corresponding assertion; if an aspect of the issue in dispute has been
suppressed by the arguer, then it just cannot be incorporated into the reconstructed
argument; and so on. All these things go with the territory and so are not defects of
A-theories.
For D-theorists, on the other hand, in principle any feature covered by list (6) is
a potentially invaluable clue about what is going on in an argumentative text. Yet,
whereas A-theorists have the advantage of being a large and, in spite of internal
disagreements, well-coordinated group, D-theorists are in the minority. Of course,
even the most recalcitrant A-theorist occasionally observes and comments on aspects
of argument delivery. Yet occasional observations and comments, however sharp they
may otherwise be, do not constitute a theoretical model, specifically targeting those
phenomena.
The area of research we have outlined was pioneered by a French linguist, Oswald
Ducrot, who in the late 1960s became intrigued by what is said and what is left unsaid
in speech (cf. 6a). The search for answers beyond the well-known concepts of presup-
position and implicature in Anglo-Saxon philosophy of language led to a whole
research programme, which Ducrot and his co-worker Jean-Claude Anscombre first
called ‘argumentation in language’, and more recently ‘argumentative semantics’.9

8 In fact, the phrase ‘argument delivery’ was chosen to label this particular meaning of the word
‘argumentation’ precisely because ‘delivery’ is the standard translation of the last task of the orator
in classical rhetoric. Only after finding what to say (inventio), arranging it in a certain order (dispo-
sition), and choosing the appropriate linguistic means (elocutio), can the orator deliver it to the
audience (actio). Between elocution and delivery, there is a task for memory, which in a script-based
culture such as ours has lost theoretical importance.
9 Anscombre and Ducrot (1976, 1983). For a brief English account of the whole research programme,

see van Eemeren et al. (2014, pp. 37, 480, 489–498). Readers who know French now have an
invaluable resource in the dedicated website https://semanticar.hypotheses.org/.
34 2 Argumentation Theories

Before we go on, it is important to say right away that this specific research area
belongs more to linguistics than to argumentation theory. Ducrot himself says:
The theory of argumentation in language… takes the word argumentation in a rather unusual
sense, which leads to a lot of misunderstandings. In this paper, I shall instead use the expres-
sion linguistic argumentation, which I shall sometimes abbreviate to argumentation. The
misunderstandings come from a tendency to read our research by giving the word argu-
mentation a completely different sense, for which I shall reserve the expression rhetorical
argumentation. (…) I shall not only distinguish the phenomena falling under these two
senses…, but I shall oppose them by showing that linguistic argumentation has nothing to
do with rhetorical argumentation. (…)
I shall understand by rhetorical argumentation the verbal activity aiming at making someone
believe something. (…)
The second term to be defined is the expression linguistic argumentation or just argumenta-
tion for short… I shall call thus the segments of discourse constituted by the chaining of two
propositions A and C, linked together, implicitly or explicitly, by a connector like therefore,
then, in consequence, and so on. I shall call A the argument and C the conclusion. This
definition can be extended to chains not of propositions but sequences of propositions, say
two paragraphs in an article. (Ducrot 2004, para. 1-3; italics in the original; our translation.)

At first sight, it would appear that the distinction offered here is similar to
the already mentioned distinction between argument-as-product and argument-as-
process, the former being equivalent to linguistic, the latter to rhetorical argumenta-
tion. But this would miss a crucial difference. A bit after the passage quoted, Ducrot
refers to the idea that linguistic argumentation is the means used in rhetorical argu-
mentation, i.e. that we use arguments to persuade people. Ducrot assures us that this
is not only false but impossible (ibid., para. 5). The argument he gives is that the
meaning of A ‘cannot be defined independently of the fact that A is seen to lead to
C’. So, there is nothing in A (e.g. truth) that could be transmitted to C (para. 8).10
This may sound a bit strange to A-theoretical ears, but the example offered by
Ducrot as illustration will bring us quickly to the core of the matter. Somebody says
to somebody else: ‘You drive too fast, you risk having an accident.’ There is no
explicit connector between the two sentences, but we can easily see the argumen-
tative connection. The idea that one can complete this argument by mentioning an
‘implicit principle’ (say, a Toulminian warrant) like ‘When people drive too fast,
they risk having an accident’ is, of course, perfectly ordinary from an A-theoretical
perspective, yet Ducrot finds this description absurd,
because the word too itself, present in the antecedent [= premise], cannot be understood
except in relation to the consequent [= conclusion]. What is driving ‘too fast’ if not driving
at a speed that risks of producing undesirable consequences? Speed itself is here characterized
by the fact that it can provoke an accident: ‘too fast’ means here ‘at a dangerous speed’.
(Ducrot 2004, para. 10; our translation.)

This is not the place to discuss Ducrot’s arguments against what we all call argu-
ments, but the example shows that, for him, practically any word, and not just the

10 Ducrot, as any scholar, is constantly arguing for his position and against the traditional ones,
so that readers may be excused if a slight suspicion of performative inconsistency starts to bother
them.
2.2 Argumentation as Argument Delivery and as Argument … 35

usual suspects (e.g. therefore, in consequence, it follows), may have argumentative

value, more precisely argumentative orientation, either in favor of or against a propo-
sition. Later on in the same text, he briefly discusses the French equivalents of little, a
little, near–far (ibid., para. 13–15), and he mentions work in progress on phrases like
to be interested and to be thirsty (para. 17). There seems to be no limit, which is why
he and Anscombre speak of ‘argumentation in language’, meaning that language, all
language—lexicon, syntax, and phonology—is suffused with an intrinsic orientation
to argumentation.
Where does this theoretical effort, coming from an unexpected quarter and focused
on linguistic problems, lead us to? Our thesis is that, if we understand Ducrot’s
remarks and arguments as belonging to a D-theory rather than an A-theory, then they
suddenly make a lot of sense. Given, however, that Ducrot declares his total lack of
interest in what we call argumentation (and he rhetorical argumentation), we cannot
expect from him and his co-workers and students to produce a D-theory which takes
its starting point from the analyses of ‘argumentation in language’. This would be a
different task altogether, and quite a challenging one.
This task was tackled by one of the present authors in the last decade (Marraud
2013, 2020a, b). However, the theory of argumentation that emerged from these
efforts is not a theory of argument delivery (a D-theory) anymore, but a theory of
argument exchange (an E-theory). This shift in theoretical target is so radical that it
deserves some comment.
Marraud’s full theoretical model will be expounded in Part II, so what follows is
just a brief sketch for the purposes of this chapter. Starting with an example, in her
excellent textbook, the argumentation theorist Trudy Govier presents the following
text:
There are lots of things that human beings do that our ancestors a thousand years ago already
did. Feelings shared include fear and mourning, and activities include joking and flirting. In
these sorts of areas human beings haven’t changed a lot. But when we begin to think instead
of things like science and technology, then at that point today’s human beings do seem very
different from our ancestors. We have more wealth, more power, and more understanding.
(Govier 2014, p. 13; bold print added, see below.)

The student is asked whether the text contains an argument, and if so, what the
conclusion is. The answer given by Govier at the end of her textbook is:
This passage offers an argument that people today are different in significant ways from
people thousands of years ago. The evidence is in differences in technology, powers, global
consciousness, and intellectual and material wealth. (Ibid., p. 386.)

The text is thus interpreted in purely A-theoretical terms as containing an argu-

ment, i.e. a set of premises and conclusion, which it certainly does. The reader can in
fact observe that Govier does some paraphrasing, which indicates that she is already
embarking in a process of abstraction away from the Johnson set contained in the
original example. By giving her solution to the exercises, she certainly implies that
her paraphrase is a rendering of the argument that is equivalent to the original passage.
The question, however, is whether this is all we can say from the point of view of
how that argument is delivered to its unknown audience.
36 2 Argumentation Theories

If we look more carefully at the text, we can see the little word ‘but’ (highlighted
in bold print above) separating the first three sentences from the last two. This little
word, an adversative conjunction, marks an opposition. When we ask ourselves what
the terms of the opposition are, the answer is: two arguments. The underlying issue
is whether people today are the same as people ‘thousands of years ago’. The first
three sentences contain an argument for the affirmative, the last two sentences an
argument for the negative. And the adversative conjunction in the middle serves to
indicate that the writer considers that the second argument is stronger than the first.
The writer thus clearly upholds the negative answer, as correctly analysed by Govier
in A-theoretical terms. The rest of the text is brushed aside as clutter; and clutter it is
indeed, if all we want to know is what the position of the writer is and what argument
she offers for it.
However, if we attend to the way the argument is delivered (or rather the argu-
ments), then we see something else that may be worth theorizing about. According
to the list (6) above, the choice of such a potent discourse marker as ‘but’ belongs to
style (6c). In this case, the overarching form of the text is ‘A but B’, where ‘A’ and
‘B’ are proxies for whole arguments, either in the A-theoretical sense (Johnson sets
or equivalence sets) or in Ducrot’s sense (see above).11 This is one of the lessons of
Ducrot’s argumentative semantics. The speaker has presented two reasons, one that is
oriented towards an affirmative answer, the other oriented against it. The conjunction
indicates that the second reason is stronger than the first. By the way, the use of ‘but’
in this way is only compatible with putting the weaker reason before the stronger one,
which is a choice of order (6b). Of course, the existence of some similarities between
people today and people thousands of years ago is not logically incompatible with the
existence of some differences between them. There is no contradiction in asserting
that people today are both the same as people thousands of years ago and different
from them. The incompatibility arises when these differences and similarities are
adduced to answer the question: are human beings of today essentially similar to
human beings of thousands of years ago? This requires deciding whether the differ-
ences are more important than the similarities or the other way around. Once the
question is formulated in this way, the idea of categorical opposition (either-or) has
been chosen in preference to the idea of a gradient or combination (both-and). This
choice of ideas expressed and suppressed is the first item in the list (6a). So, we can
see that all three phenomena of argument delivery are at work even in such a simple
example as this passage from Govier (2014).
Nonetheless, as soon as we notice that in Govier’s exercise a confrontation of
arguments is put forward, it becomes clear that the speaker in that text is arguing
with somebody. It does not matter whether that somebody is a real person or an

11 In texts about argumentation, the word ‘argument’ most often refers to the whole set
of propositions: premises and conclusion in the traditional A-theoretical sense, reasons and
thesis/position/standpoint in the different versions of dialectics (see also Sect. 2.3). However, some-
times the word ‘argument’ is used to refer just to the premises or reasons (as when someone asks:
‘What is your argument?’). The word ‘but’ usually introduces a sentence that expresses a premise
or reason, so that we can say that ‘but’ connects premises. In a larger sense, of course, we can also
say that it connects whole arguments.
2.2 Argumentation as Argument Delivery and as Argument … 37

imaginary one. The speaker is arguing with that person. Everything happens in the
text as if that real or imaginary person had argued that people a thousand years
ago were the same as people today because the feelings of human beings have not
changed. The speaker confronts that argument by counterarguing that, even if that is
true, the fact that technology, wealth, and understanding is so much larger and more
powerful overrides as a reason the other person’s argument. What we have here is
not just argument delivery (although it is that) but also an exchange of arguments.
In fact, the whole argument present in the text is a much larger and more complex
argument, consisting of a comparison of arguments in favour of one of them. Thus,
the theory of argument delivery (a D-theory) inherent in Ducrot’s argumentative
semantics develops into a theory of argument exchange (an E-theory), for which
arguments are constructed by the joint work of arguers.
This is, of course, more clearly the case when the argumentation occurs within
the framework of an actual discussion, such as the Russell-Copleston debate we shall
analyse later on; but it is also present in any apparently monologic text and can be
detected precisely by noting the features listed in (6), which we now see are also
features of argument exchange.
There is a great temptation which we should resist, however: the idea, namely,
that the theoretical development from a D-theory into an E-theory can be captured
by saying that the former belongs to rhetoric and the latter to dialectics. This form
of dividing the theoretical arena has become very popular ever since Joseph Wenzel
introduced it; but it may be misleading if we are not careful. One can certainly say
that A-theories are not dialectical at all, and also that D-theories are not dialectical
to any serious extent. Let us spell this out a little bit.
A-theories remain exclusively in the realm of logic, conceived as a theory of
argument in the sense outlined in Sect. 2.1. Their business is to extract Johnson
sets from texts and to produce, as the outcome of their analytic procedures, sets of
premises and conclusions. A-theories have only one use for rhetoric, namely, the
fourfold method. However, they do not incoporate that method into their theory of
argument. In fact, A-theorists rarely talk explicitly about that method. Johnson (2014)
is a laudable exception.
D-theories, in their turn, depend on A-theories for their concept of argument,
but they enrich that concept by a careful consideration of the way arguments are
presented, and so they are largely rhetorical.12 But since they are, as theories of
argument delivery, uninterested in the contrast and conflict of argument, they can
perhaps play around with dialectical concepts, but they are at bottom closed to the
properly dialectical perspective. This will become clearer in Part II, Chap. 9, when
the distinction between arguer dialectics and argument dialectics is introduced.

12 Notable examples of D-theoretical analyses of arguments are Finocchiaro (1980), McCloskey

(1985), Fahnestock (1999), and the work of Alan Gross (e.g. Gross 2006; Harmon and Gross 2010;
Gross and Harmon 2014; Gross 2018). It is a moot point, however, whether beyond the analysis we
can find an explicit theory of argument delivery in those works. Anyway, this is not the place for a
discussion on this question.
38 2 Argumentation Theories

Table 2.1 Argumentation theories as related to traditional disciplines

Argumentation theories Logic Rhetoric Dialectics
Theories of argument (A-theories) + – –
Theories of argument delivery (D-theories) + + –
Theories of argument exchange (E-theories) + + +

Finally, E-theories necessarily partake of logic, rhetoric, and dialectics. As for

logic, however, they reject the simple idea of argument contained in A-theories of
the sort described in Sect. 2.1. For E-theories, arguments are, as we illustrated with the
simple example in Govier (2014), constructed in a stepwise way, combining smaller
arguments into larger ones. The resulting super-arguments or macro-argumentations
can in some case have a modest size, as in the Govier example, but sometimes they
can reach a staggering complexity. The tools for making such macro-argumentations
will all be set forth in Part II and we shall see them at work in the analysis of the
Russell-Copleston debate. It is important to notice that macro-argumentations are
arguments, i.e. abstract objects; so, everything we said about arguments in Sect. 2.1
of this chapter, apply to macro-argumentations as well. The only thing is that they
are larger structures than ordinary arguments, and for the same reason A-theories
have not theorized about them. We need, therefore, a special logical study as part of
our E-theory. But an E- theory will also have to draw on both rhetoric, in the sense
of list (6), to analyze how the macro-argumention is constructed and on dialectics
to show how that macro-argumentation is both a joint product of the arguers and a
complex structure representing the dialectical relationships between the individual
arguments combined in it. Table 2.1 may be a way of representing the comparison
between argumentation theories.
This gives a new perspective on the question of how much clutter a given text has
from the point of view of different argumentation theories. It is clear that A-theories
will find more of it in a text than do D-theories, and these more than E-theories. Now
it is time to present a third type of theory, according to which an argumentative text
has no clutter at all. This requires a far more capacious concept of argumentation.

2.3 Argumentation as Argumentative Process: P-Theories

According to the third sense of argumentation we have to consider the argumenta-

tive process as it actually takes place, in its entirety. When people argue, they do
deliver arguments, of course, but they perform many other operations apart from
delivering arguments, which are part and parcel of the communicative process in a
non-accidental way. Those operations create the framework within which arguments
can be presented and sustain it all the way through. They are thus constitutive of
argumentation in a way that goes far beyond both the arguments themselves and the
way they are delivered.
2.3 Argumentation as Argumentative Process: P-Theories 39

A theory of all those operations is a P-theory (P for process). The prototype of such
a theory is pragma-dialectics, a research programme that started in the late 1970s
and continues to this day, thanks to the joint work of many people, under the leader-
ship of Frans van Eemeren.13 And the hard core of the pragma-dialectical research
programme is the ideal model of critical discussion. A critical discussion is defined
as a discussion in which the discussants are trying to resolve their disagreements on
the sole basis of the merits of the reasons offered in the discussion.
The model of critical discussion has been expounded many times before; but, as
far as we are aware, this is the first time that the single operations, in contradistinction
to the stages, are emphasized. Note that the ambition of the ideal model of critical
discussion is to comprehend all the different operations which parties in a discussion
engage in when trying to resolve their disagreement in a reasonable manner. The
operations are conceived as taking place within a sequence of stages, which I proceed
to describe as briefly as possible. The Roman numbers will help us keep track of the
operations.
(7) Operations performed during a critical discussion, organized by stages:

a. First comes the confrontation stage, in which

(I) one of the parties expresses a standpoint,
(II) another party indicates a lack of agreement with the standpoint, either
by flatly opposing it in some way or by expressing doubts or other
reservations, and
(III) both parties recognize the lack of agreement.

If each party expresses a standpoint, then the difference of opinion is called mixed,
and if only one party has a standpoint whereas the other party has only reservations
in relation to the standpoint, then the difference of opinion is called non-mixed. If
the standpoint contains several propositions on which there is disagreement, then
the difference of opinion is called multiple, if only one, then single. By combining
both criteria, four classes of differences of opinion emerge: single non-mixed, single
mixed, multiple non-mixed, multiple mixed. A good confrontation stage should allow
the two parties to identify to which class their disagreement belongs.
b. Then comes the opening stage, in which
(IV) the parties agree to try to resolve their disagreement by delivering
arguments,
(V) establish the substantive and

13 The word ‘process’ is liable to some misunderstanding. It suggests the dynamic aspect of arguing
as opposed to the static character of the argument as an abstract object (or a Johnson set). But we
should beware that, in a discussion, we have both the exchange of arguments and many other
communicative moves (such as questions, assignment of roles, agreement on rules, and so on).
Both have a dynamic and a static aspect. Both are processes, and analysis can extract objects from
such processes. However, the full argumentative process with which P-theories are concerned is
broader than the process of exchanging arguments. For more on this, see Chap. 3.
40 2 Argumentation Theories

(VI) procedural starting points, i.e. the propositions which they agree on and
the rules of discussion that they intend to abide by, and
(VII) they jointly assign the roles of ‘protagonist’ and ‘antagonist’ to each
other, the antagonist being the party who presents doubts and objec-
tions to the standpoint at issue and the protagonist being the party who
responds to those doubts and objections.
If the difference of opinion is mixed, then both parties will have both roles, but if
it is non-mixed, then one party must have the role of protagonist and the other the
role of antagonist.
Both these stages are supposed to precede the argumentation proper if the two
parties are reasonable and are trying reasonably to resolve their disagreement. For
no critical discussion can take place if, in the first place, there is no disagreement or
if it is not recognized as such, which is what the confrontation stage is about. Such
requirement is precisely fulfilled by operations (I)–(III). Again, no critical discussion
can take place if, in the second place, the parties do not previously express the mutual
intention of trying to resolve the disagreement by argumentation, and thus actually
open the discussion; and if they are reasonable about how to have the discussion,
then they need both to establish the starting points and to assign the roles to be
played by each party. The operations (IV)–(VII) thus constitute the opening stage
of a critical discussion. Of course, in all communicative situations there will be a
number of shared starting points and a presumption as to the burden of proof; and
in some special communicative situations (e.g. in the courtroom, in intercollegiate
debates, in the defense of a dissertation) all of this may be strictly defined by the
institution before the discussion starts; and yet there may be occasion to resort to the
opening stage to remind the parties of their duties and commitments.
c. Now we enter the argumentation stage, in which
(VIII1) the protagonist delivers one or several arguments that support her
standpoint, and, if those arguments should not convince the antago-
nist, then
(IX1) the latter presents her doubts or objections, to which
(VIII2) the protagonist responds to by producing a fresh set of arguments, to
which
(IX2) the antagonist counters with a fresh set of doubts or objections and
so on.
The operations (VIII)–(IX) can be iterated again and again until no more can be
said by the parties. In other words, operations (VIII)–(IX) are cyclical.
d. And so, we finally come to the concluding stage, in which
(X) either the protagonist admits that she cannot produce any more arguments
to counter the antagonist’s doubts and objections or the antagonist admits
that she cannot produce any more doubts or objections to the standpoint
in view of the arguments offered; the disagreement is declared by the
parties to be resolved in either case.
2.3 Argumentation as Argumentative Process: P-Theories 41

Except for the emphasis on the individual operations that take place in a critical
discussion, this model is the old model that everyone in the field or argumentation
studies is bound to know. However, we think that a few additional remarks may help
readers understand and appreciate how this picture works.
a. The model is called ideal for a reason. What is the ideal that the model tries to
flesh out? It is the ideal of what is reasonable. In other words, if only people
would be fully and completely reasonable in their dealings with each other,
then they would follow the model and so would resolve their disagreements.
Are people fully and completely reasonable? No, they are not. In their actual
arguments, people deviate from the ideal of what is reasonable in many ways—
they get angry with each other, they want to outshine each other, they wish to
win the argument, they become impatient, they have other things to do, and so
on. Still, we cannot know which behaviors deviate from the ideal of what is
reasonable, and how and to what extent, if we did not keep the ideal in mind. In
other words, we use the ideal as a benchmark, for the purpose of measurement,
in order to assess what people are trying to do at each point of the discussion,
and whether they are achieving or failing to achieve it.
b. The model is about discussion. Many discussions take place among people who
do not manage to resolve a difference of opinion; some do not even aim at doing
so; but no resolution of disagreement on the merits can happen without discus-
sion. Now, the concept of discussion is larger than that of argumentation in
the sense of argument delivery. Yet all argumentative discourse is embedded in
discussion. There is no such thing as presenting reasons, or asking for reasons,
unless there is a larger framework of discussion. Some readers may say, in
this case, that pragma-dialectics is not a theory of argumentation in the proper
sense but a theory of discussion. We have no objection to such readers, because
in the end this is a verbal dispute. The point is that there is something to be
called ‘argumentative discourse’ and there is something to be called ‘argumen-
tative process’, and these things are not exhausted by the questions of argument
delivery, let alone by arguments.
c. For a discussion to succeed in resolving a difference of opinion, the discussion
has to be critical. The meaning of the adjective ‘critical’, in its modern use, stands
in opposition to blind acceptance of what somebody else says, especially if in a
position of domination or authority (see already van Eemeren and Grootendorst,
1983, pp. 6, 16–18).14 People behave non-critically whenever they yield or give
up on their publicly stated commitments without further ado; whenever they do

14 There is a modern use and a classical use of critical and all akin terms (criticism, critique, critic,
criterion) and both are current. The classical use refers to the close study of texts. Criticism in
that sense comprises three operations: textual criticism (establishing who said what), interpretation
(what was meant by it), and evaluation (what are the good or bad, better or worse parts of the work,
according to a variety of criteria). More broadly, criticism can be applied to other cultural artifacts,
such as theater performance, musical compositions and performance, sculptures, paintings, and
other works, mainly in the ‘fine arts’ but increasingly in popular culture. The modern use, as far
as it is not totally adulterated to refer to negative reactions to something, is sketched above. For an
outline of the history of the term and its two meanings, see Leal (2003).
42 2 Argumentation Theories

not offer reasons when challenged; whenever they do not offer objections to an
unacceptable proposition; and so on. We can therefore say that the whole point
of the ideal model of critical discussion is simply to flesh out in detail what this
fundamental and commonly shared concept entails.
d. The description of the ideal model has only two parties engaged in a critical
discussion. But we should here remember that a party in a discussion is not
necessarily an individual: it could be a team (as in intercollegiate debates), a
group (as in Parliament) or even a school of thought (as in philosophy). This
already shows that a critical discussion may engage more than two people and
so be a ‘polylogue’ (Kerbrat-Orecchioni 2004; for ‘polylogues’ in pragma-
dialectics, see Lewiński and Aakhus 2014). Moreover, one could have a critical
discussion at the same time with three or even more parties at different levels.
For instance, it can happen that A puts forward a standpoint, B contradicts one
part of it, C a different part, D all of it, and E expresses doubts about what all the
others are saying. In such an example, there are mixed differences of opinion
between A and B, A and C, and A and D, whereas each of them has a non-mixed
difference of opinion with E. Again, the differences of opinion seem to be single
in the case of A-B and A-C, but multiple in the case of A-D, and so on (see also
van Eemeren, 2010, pp. 108–109). The polylogical critical discussion does not
even have to take place contemporaneously. In fact, philosophical discussions
can last for years, decades, centuries (Leal 2003).
e. In the ideal model of critical discussion the different operations are actions
performed by means of natural language. Each operation is realized by one
or several speech acts. Just to give an example, one of such speech acts is
the asking of questions, which in pragma-dialectics is conceptualized as the
directive speech act of ‘requesting’. Now, in the standard theory requesting
makes only two (see van Eemeren and Grootendorst, 2004, p. 67) or three
appearances during a critical discussion (van Eemeren, 2018, p. 42). If we use
the letters ‘A’ and ‘P’ for the roles of antagonist and protagonist, respectively,
then we have that (i) within the opening stage, A challenges P to defend P’s
standpoint, (ii) within the argumentation stage, A asks P to offer reasons for P’s
standpoint, and (iii) within any stage, one party can beg another party to clarify,
elaborate, amplify, illustrate or otherwise explain her use of words or phrases.
The point is that questions and requests are, as yet, no part of argumentation
theories of the sort described in Sects. 2.1 and 2.2.15 If they appear on a text,
they are either left aside as clutter, or they are swiftly converted into assertions.
Yet questions and requests are absolutely essential to argumentative discourse
and the argumentative process.

15 We are well aware that questions can be incorporated into a non-classical system of formal
logic as part of inferences (Wiśniewski 1995, 2013). This makes possible to introduce them into
A-theoretical analysis (for a recent attempt, see Hitchcock 2020). Again, D-theoretical models of
philosophical discussion can crucially include questions (Galindo 2020). If these recent approaches
are further developed, it will become possible to discuss in detail what differences in the treatment
of questions in the three types of theory remain. Our bet is that the main difference will be in the
use of speech act theory.
2.3 Argumentation as Argumentative Process: P-Theories 43

After these few remarks, we must come to the most important question. What
kind of theory is a P-theory such as pragma-dialectics? In the context of the present
chapter, this is the same as asking: What is the relationship of this P-theory to the
theories described above? On the one hand, the ideal model of critical discussion
looks at argumentation in a way that is utterly different from all preceding theoret-
ical frameworks in argumentation studies. As we have seen, A-theories concentrate
on arguments in the light of the questions set forth in list (5). For their part, D-
theories go beyond the arguments as such and raise new questions having to do with
the way arguments are presented, questions which can be comfortably classified
according to the classical tasks of rhetoric sketched in (6). As for pragma-dialectics,
it is certainly interested in both arguments and argument delivery. In fact, the whole
point of operation (VIII), and its iterations, is about argument delivery, but this is
not seen as an isolated operation, subsisting by itself and surrounded by all sorts of
irrelevant speech. On the contrary, for pragma-dialectics the cyclical operation (VIII)
only makes sense as part of a sequence of reasonable activities, including the equally
cyclical operation (IX) that is the match of (VIII), but also operations (I)–(VII), that
prepare the field for the interaction of (VIII) and (IX), as well as operation (X) that
closes that interaction. Pragma-dialectics thus is the first theoretical framework that
gives argument delivery its proper place as one among many other activities which
jointly constitute argumentative discourse. Interestingly enough, arguments are not
only delivered in the argumentation stage, but in principle also in the other stages.
This is a topic, however, that we shall discuss in more detail in Chap. 3 and shall see
at work in the analysis of the debate between Russell and Copleston given in Part I.
On the other hand, precisely because of its reliance on A-theories and D-theories,
which have been the main dish in the field, pragma-dialectics has little to say about
argument exchange as such, the target of E-theories. In the abstract form of the theory,
this is of course included, for it is explicitly a dialectical theory engaged in studying
the interaction between arguers. But, as happens to A- and D-theories, there is no
place for the study of the relationships (the dialectic) among the arguments, even
though those relationships leave traces in the form of discourse markers, disposition
of the text, and selection of which topics are expressed and which are suppressed,
thus all the different aspects studied by classical rhetoric. In spite of this, arguments
delivered are examined as though they were separate from each other—and not as
entangled with each other, which is how they actually appear as soon as we adopt an
E-theoretical perspective. The most challenging open question is how to incorporate
argument exchange into the pragmatic-dialectical framework.
As a final observation, we want to add that, if the ideal model of critical discussion
is the most developed framework to date of the argumentative process, it is not the
only one in the field of argumentation studies. At the risk of neglecting to quote
important work, we would like to mention at least one of the other frameworks,
namely that developed by Michael Gilbert over the past forty years.16 It is highly

16 Some other theorists in the field of argumentation studies have also insisted on the importance
of taking a broader view and have defended the idea that the whole communicative process is the
proper target of argumentation theory (e.g. Willard 1989; Tindale 1999, 2011). We do not go into
44 2 Argumentation Theories

original work which unfortunately has been widely misunderstood, partly because its
main concepts—multi-modal argumentation and coalescence (Gilbert 1994, 1995,
1997)—tend to be seen, and sometimes to be presented by Gilbert himself, from an
A-theoretical vantage point. We think that, if the typology here presented is on the
right track, many of Gilbert’s ideas would more naturally find its place within a larger
P-theoretical framework. In fact, long before the insights associated with multi-modal
argumentation and coalescence had been born as such, Gilbert had already offered
us certain concrete applications of those insight which deserve wider appreciation.
They nicely tie up with the kind of theory represented by pragma-dialectics.
The concrete applications sometimes appear in Gilbert’s papers or in his
systematic book (Gilbert 1997), but they are most at home in his two textbooks.
The first textbook marks his official transition from formal logic to argumentation
studies. It has gone through three editions (How to Win an Argument, 1978, 1996,
2008), but in all of them the division in three parts remains the same. The first
two parts present the nuts and bolts of Gilbert’s approach to argument analysis and
evaluation. Yet the third part contains specimens of argumentation together with a
running commentary. It is in that commentary that the secret of Gilbert’s approach
becomes evident. It is not in vain that Gilbert himself recommend his readers to go
through Part Three first.17 He wants ‘to make [his readers] aware of what is going
on in an argument’. Here, the word ‘argument’ means the same as ‘discussion’. And
‘what is going on in an argument’ refers to a much wider range of phenomena than
what A-theories and even D-theories have in mind. To give a very small sample
of what Gilbert means by that, let us have a look at the first four exchanges of the
first example (‘Pot Luck’, Gilbert 1996, Chap. 24; text on the left, comments on the
right; bold print added to highlight the parts in the exchange that are the object of
the comments):

Paul: Lord, I wish they’d legalize marijuana

Chloe: Are you sure that’s a good idea? Good. Chloe thus responds to Paul’s claim
with a question, not with another statement.
This is the right way.
Paul: Sure I’m sure. Don’t tell me you think it Paul is trying to intimidate Chloe: being the
should be illegal! You must be the last person last person to agree means being “out of it”.
in the world to believe that. This is a fallacy. Notice that Paul has not yet
given a reason for his position.
(continued)

them in this chapter, because their relatively abstract character makes its application to an extended
piece of argumentation such as the Russell-Copleston debate far more difficult.
17 In Gilbert’s second and most recent textbook (Arguing with People, 2014), the examples and

running commentary appear in Part 3, Sect. 7).

2.3 xm-replace_text {SimplePara} 45

(continued)
Paul: Lord, I wish they’d legalize marijuana
Chloe: If I’m the last person, how come it’s
still illegal? I’d like to know why it should be
legal.
After all, there’s no proof it isn’t harmful.
Good. This is an excellent response to the
fallacy.
Mistake. Chloe has offered a reason when she
did not have to—she did not start the
argument. If Paul attacks, Chloe may be on the
defensive.

And on it goes. This is not the place to expound Gilbert’s approach, rules, or
criteria for evaluation (for a recent update, the reader should consult his second text-
book, Gilbert 2014). All we want to show is that he is clearly interested in the argu-
ments offered and in their particular delivery, yet his way of analyzing the discussion
includes other aspects that are often ignored in other theoretical frameworks. Note
the parts in bold print: three questions, one exclamation. These kinds of utterance
would be considered clutter both from an A-theoretical and a D-theoretical perspec-
tive. For Gilbert’s theorizing, they are crucial, as is clear from the comments. They
would also be analyzed as crucial within the pragma-dialectical model of critical
discussion.

References

Anscombre, J.-C., and Ducrot, O. (1976). L’argumentation dans la langue. Langages 10 (42), 5–27.
Anscombre, J.-C., and Ducrot, O. (1983). L’argumentation dans la langue. Brussels: Pierre
Mardaga.
Baggini, J., and Peter S. F. (2010). The philosopher’s toolkit: A compendium of philosophical
concepts and methods. New York: Wiley.
Bouveresse, J. (1996). La demande philosophique: Que veut la philosophie et que peut-on vouloir
d’elle? [The philosophical query: What does philosophy want and what can one want from
philosophy?] Paris: Éditions de Minuit.
Brockriede, W. (1975). Where is argument? The Journal of the American Forensic Association [later
Argumentation and Advocacy] 11 (4), 179–182.
Brockriede, W. (1977). Characteristics of arguments and arguing. The Journal of the American
Forensic Association [later Argumentation and Advocacy] 13 (3), 129–132
Bruce, M., and Barbone, S. (2011). Just the arguments: 100 of the most important arguments in
Western philosophy. Malden, MA: Wiley Blackwell.
Cohen, D. (2004). Arguments and metaphors in philosophy. Lanham, MD: University Press of
America.
Cornman, J. W., Lehrer, K., and Pappas, G. S. (1992). Philosophical problems and arguments: An
introduction. 4th edition. Indianapolis: Hackett.
Ducrot, O. (2004). Argumentation rhétorique et argumentation linguistique [Rhetorical and
linguistic argumentation]. In M. Doury and S. Moirand (Eds.), L’argumentation aujourd’hui:
Positions théoriques en confrontation [Argumentation today: Confrontation of theoretical posi-
tions] (pp. 17–34). Paris: Presses Sorbonne Nouvelle. [Open access: https://books.openedition.
org/psn/756?lang=en.]
46 2 Argumentation Theories

Eemeren, F. H. van (2010). Strategic maneuvering in argumentative discourse. Amsterdam: John

Benjamins.
Eemeren, F. H. van (2018). Argumentation theory: A pragma-dialectical perspective. Cham:
Springer.
Eemeren, F. H. van, and Grootendorst, R. (1983). Speech acts in argumentative discussions.
Dordrecht: Foris.
Eemeren, F. H. van, and Grootendorst, R. (2004). A systematic theory of argumentation: The pragma-
dialectical approach. New York: Cambridge University Press.
Eeemeren, F. H. van, Garssen, B., Krabbe, E. C. W., Snoeck Henkemans, A. F., Verheij, B., and
Wagemans, J. H. M. (2014). Handbook of argumentation theory. Dordrecht: Springer.
Fahnestock, J. (1999). Rhetorical figures in science. New York: Oxford University Press.
Finocchiaro, M. (1980). Galileo and the art of reasoning: Rhetorical foundations of logic and
scientific method. Dordrecht: Reidel.
Galindo, J. (2020). Dialogical sequences, argumentative moves, and interrogative burden of proof
in philosophical argumentation. In F. H. van Eemeren and B. Garssen (Eds.), From argument
schemes to argumentative relations in the wild: A variety of contributions to argumentation
theory (pp. 79–96). Cham: Springer.
Gilbert, M. A. (1978). How to win an argument. New York: McGraw-Hill.
Gilbert, M. A. (1994). Multi-modal argumentation. Philosophy of the Social Sciences 24 (2), 159–
177.
Gilbert, M. A. (1995). Coalescent argumentation. Argumentation 9, 837–852.
Gilbert, M. A. (1996). How to win an argument. 2nd edition. New York: John Wiley.
Gilbert, M. A. (2008). How to win an argument. 3rd edition. Lanham, MD: University Press of
America.
Gilbert, M. A. (1997). Coalescent argumentation. Mahwah; NJ: Lawrence Erlbaum.
Gilbert, M. A. (2014). Arguing with people. Toronto: Broadview.
Glymour, Clark. (1992). Thinking things through: An introduction to philosophical issues and
achievements. Cambridge, MA: The MIT Press.
Govier, T. (2014). A practical study of argument. Enhanced 7th edition. Boston: Wadsworth.
Groarke, L. (1996). Logic, art, and argument. Informal Logic 18 (2–3), 105–129.
Groarke, L. (2015). Going multimodal: What is a mode of arguing and why does it matter?
Argumentation 29, 133–155.
Groarke, L. (2019). Depicting visual arguments: An ‘ART’ approach. In F. Puppo (Ed.), Informal
logic: A ‘Canadian’ approach to argument, 332–374. Windsor, ON: Windsor Studies in
Argumentation.
Gross, A. G. (2006). Starring the text: The place of rhetoric in science studies. Carbondale, IL:
Southern Illinois University Press.
Gross, A. G. (2018). The scientific sublime: Popular science unravels the mysteries of the universe.
New York: Oxford University Press.
Gross, A. G., and Harmon, J. E. (2014). Science from sight to insight: How scientists illustrate
meaning. Chicago: The University of Chicago Press.
Gutting, G. (2009). What philosophers know: Case studies in recent analytic philosophy. New York:
Cambridge University Press.
Habermas, J. (1981). Theorie des kommunikativen Handelns, 2 volumes. Frankfurt am Main:
Suhrkamp. [English translation: The theory of communicative action. Boston: Beacon Press,
1984.]
Harmon, J. E., and Gross, A. G. (2010). The craft of scientific communication. Chicago: The
University of Chicago Press.
Harrell, M. (2016). What is the argument? An introduction to philosophical argument analysis.
Cambridge, MA: The MIT Press.
Harrison, C. (1974). Totalities and the logic of first cause arguments. Philosophy and Phenomeno-
logical Research 35 (1), 1–19.
Henderson A., W. (1980). Spanish review of Russell and Copleston (1978). Diánoia 26, 297–301.
References 47

Hitchcock, D. (2020). Arguing for questions. In F. H. van Eemeren and B. Garssen (Eds.), From argu-
ment schemes to argumentative relations in the wild: A variety of contributions to argumentation
theory, 167–184. Cham: Springer.
Johnson, R. H. (2014). The rise of informal logic: Essays on argumentation, critical thinking,
reasoning, and politics. Windsor, ON: Windsor Studies in Argumentation.
Johnstone Jr., H. W. (1959). Philosophy and argument. State College, PA: Pennsylvania State
University Press.
Johnstone Jr., H. W. (1978). Validity and rhetoric in philosophical argument: An outlook in
transition. University Park (PA): The Dialogue Press of Man and World.
Johnstone Jr., H. W. (Ed.) (1988). Special issue on philosophical argumentation. Argumentation 2,
153–294.
Kerbrat-Orecchioni, C. (2004). Introducing polylogue. Journal of Pragmatics 36 (1), 1–24.
Koons, R. C. (1997). A new look at the cosmological argument. American Philosophical Quarterly
34 (2), 193–211.
Landini, G. (2009). Russell and the ontological argument. Russell: The Journal of Bertrand Russell
Studies (n.s.) 29, 101–128.
Lausberg, H. (1963). Handbuch der literarischen Rhetorik: Eine Grundlegung der Literaturwis-
senschaft. Munich: Max Hueber Verlag. English translation: Handbook of literary rhetoric: A
foundation for literary study. Leiden: Brill, 1998.
Leahy, L. (1976). Contingency, in the cosmological argument. Religious Studies 12 (1), 93–100.
Leal, F. (2003). ¿Qué es crítico? Apuntes para la historia de un término. [What is critical? Notes
towards the history of a term.] Revista Mexicana de Investigación Educativa 8 (17), 245–261.
Lewiński, M., and Aakhus, M. (2014). Argumentative polylogues in a dialectical framework: A
methodological inquiry. Argumentation 28 (2), 161–185.
Marraud, H. (2013). ¿Es lógic@? Análisis y evaluación de argumentos [Is it logic(al)? Analysis
and evaluation of arguments]. Madrid: Cátedra.
Marraud, H. (2020a). On the logical ways to counter an argument: A typology and some theoretical
consequences. In F. H. van Eemeren and B. Garssen (Eds.), From argument schemes to argumen-
tative relations in the wild: A variety of contributions to argumentation theory (pp. 149–166).
Cham: Springer.
Marraud, H. (2020b). En buena lógica: Una introducción a la teoría de la argumentación. [Logically
speaking: An introduction to the theory of argumentation]. Guadalajara: University of Guadalajara
Press.
Martinich, A. P. (2016). Philosophical writing: An introduction. 4th edition. New York: Wiley.
McCloskey, D. N. (1985). The rhetoric of economics. Madison: The University of Wisconsin Press.
Morton, A. (2004). Philosophy in practice: An introduction to the main questions. 2nd edition. New
York: Wiley.
Nelson, L. (2016). A theory of philosophical fallacies. Cham: Springer.
O’Keefe, D. (1977). Two concepts of argument. The Journal of the American Forensic Association
[Argumentation and Advocacy] 13, (3), 121–128.
Passmore, J. (1961). Philosophical reasoning. New York: Charles Scribner’s Sons.
Powers. L. (2012). Non-contradiction. London: College Publications.
Rasmussen, K. (1974). Implications of argumentation for aesthetic experience: A transactional
perspective. Ph. D. dissertation, University of Colorado.
Rescher, N. (2001). Philosophical reasoning: A study in the methodology of philosophizing. Oxford:
Blackwell.
Rescher, N. (2006). Philosophical dialectics: An essay on metaphilosophy. Albany, NY: State
University of New York Press.
Rescher, N. (2009). Aporetics: Rational deliberation in the face of inconsistency. Pittsburgh, PA:
The University of Pittsburgh Press.
Rosenberg, J. (1996). The practice of philosophy: A handbook for beginners. 3rd edition. Upper
Saddle River, NJ: Prentice Hall.
Rowe, W. L. (1970). Two criticisms of the cosmological argument. The Monist 54 (3), 441–459.
48 2 Argumentation Theories

Russell, B., and Copleston, F. C. (1978). Debate sobre la existencia de Dios [Debate on the existence
of God]. Valencia: Universidad de Valencia (Cuadernos Teorema).
Schnädelbach, H. (2012). Was Philosophen wissen—und was man von ihnen lernen kann [What
philosophers know—and what one can learn from them]. Munich: Beck.
Stoljar, D. (2017). Philosophical progress: In defence of a reasonable optimism. New York: Oxford
University Press.
Stove, D. (1991). The plato cult and other philosophical follies. Oxford: Blackwell.
Tindale, C. (1999). Acts of arguing: A rhetorical model of argumentation. Albany, NY: State
University of New York Press.
Tindale, C. (2011). Rhetorical argumentation: Principles of theory and practice. Thousand Oaks,
CA: Sage.
Wenzel, J. (1979). Jürgen Habermas and the dialectical perspective on argumentation. The Journal
of the American Forensic Association [Argumentation and Advocacy] 16 (2), 83–94.
Willard, C. A. (1989). A theory of argumentation. Tuscaloosa: The University of Alabama Press.
Williamson, T. (2007). The philosophy of philosophy. Oxford: Blackwell.
Wiśniewski, A. (1995). The posing of questions: Logical foundations of erotetic inferences.
Dordrecht: Kluwer.
Wiśniewski, A. (2013). Questions, inferences, and scenarios. London: College Publications.
 Part I
A P-Theoretical Analysis of the Debate,
by Fernando Leal
Chapter 3
Description of the Method Followed

Abstract Before presenting the actual pragma-dialectical analysis, a few words

are needed to explain the special role that questions and questioning plays in the
Russell–Copleston debate, and indeed in all philosophical argumentation. Starting
with the way standard pragma-dialectics deals with this issue, a few concepts are
added to tie up the matter of questions with the fundamental concepts of standpoint
and difference of opinion. Then the role of questioning within the ideal model of a
critical discussion is tackled. Finally, all these developments are used to present an
overview of the Russell–Copleston debate to be analysed in Chaps. 4–8.

The method followed in the ensuing analysis of the Russell-Copleston debate is

based on standard pragma-dialectics augmented by an erotetic perspective. Standard
pragma-dialectics is well-known. Its basic construct is the ideal model of critical
discussion. An erotetic perspective means making explicit the place of questioning
in that model.
The erotetic perspective has been part and parcel of the pragma-dialectical
approach from the very beginning, but it has remained largely implicit during the
development of the theory. Consider the following passage taken from the first
English exposition of the theory:
Suppose three people hear someone on television propound the view that women have a
logic of their own; suppose also that the three people, having heard this view propounded,
embark upon a (serious) discussion of the question ‘Have women a logic of their own?’
One of them then says ‘In my opinion it is true that women have a logic of their own’, the
second says ‘In my opinion it is not true that women have a logic of their own’, and the third
says ‘I do not know whether or not it is true that women have a logic of their own’. (van
Eemeren and Grootendorst 1983, p. 78; bold print added)

The reader will recognize in the three positions described here, respectively, the
affirmative answer to the question that triggers the discussion (call it Yes), the negative
answer (call it No), and an expression of noncommittal doubt (call it Maybe). The
question has thus brought three differences of opinion to the surface, which we may
represent by Table 3.1.1

1 There is, of course, a fourth difference of opinion involving, at the same time, the three parties in
the above example. This fact points to the concept of a trilogue and, more generally, of a polylogue
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 51
F. Leal and H. Marraud, How Philosophers Argue, Argumentation Library 41,
https://doi.org/10.1007/978-3-030-85368-6_3
52 3 Description of the Method Followed

Table 3.1 One question,

three parties, three differences
of opinion
Question: Have women a logic of their own?
Differences of opinion – Mixed: P1 yes, P2 no
– Non-mixed: P1 yes, P3 maybe
– Non-mixed: P2 no, P3 maybe

This is all standard pragma-dialectics: a mixed difference of opinion involves two

parties, each one of which upholds a standpoint; a non-mixed one involves one party
which upholds a standpoint and another party which does not and is content to voice a
noncommittal position. The only thing that is added to the standard model is precisely
the perspective of a triggering question which was raised by someone on television.
It is not necessary that the persons or characters, real or fictional, who raised the
question on television actually asked a question. It suffices that they said something
that was sufficiently controversial to have raised the question which triggered the
discussion of the viewers (Leal, 2021).
The distinction just mentioned is common in English and other languages. People
can ask questions in the strong sense of the word, i.e. request for information that
they need but don’t have at the moment, by using interrogative sentences as well as
a variety of other verbal means (‘please tell me whether…’, ‘I wonder why…’, and
so on). But questions are raised by things said and done by people in the course of
human interactions. If we don’t want to name the person or persons who by word
or deed raised a question, we can also use an impersonal expression and just say
that the question arises. These and other expressions, both in English and in other
languages, are perfectly ordinary, and we want to avail ourselves of this lexically
marked distinction in order to make clear that no critical discussion can take place
unless a question arises, even if no one in the situation actually asked the question.
The rest of this chapter will present the theory with an eye to its application to
the field of philosophical argumentation. If the reader thinks that the same theory
can also be applied to other fields, well and good. For us, it is enough if the special

(Kerbrat-Orecchioni, 2004). A trilogue is a discussion in which all three parties are engaged at the
same time. In a trilogical situation, there are seven combinations of possible disagreements: Yes-
No-Maybe (as in the example), Yes-Yes-Maybe, No-No-Maybe, Yes-Maybe-Maybe, No-Maybe-
Maybe, Yes-Yes-No, and No-No-Yes. The most famous discussion which has the Yes-No-Maybe
arrangement is Galileo’s Dialogue Concerning the Two Chief World Systems, in which there is
a mixed difference of opinion between Salviati, who defends the Copernican world system, and
Simplicio, who attacks it, whereas the younger Sagredo is undecided and thus has a non-mixed
difference of opinion with both Salviati and Simplicio. Standard pragma-dialectics is predicated
upon dilogical situations (two parties in disagreement), but it can be adapted to any polylogical
situation, as shown in Lewiński and Aakhus (2014). In this book, we shall remain focused on the
dilogical situation for ease of exposition. Besides, the Russell-Copleston debate to be analyzed is
ostensibly a dilogue not a polylogue. Nonetheless, it is arguable that all philosophical discussions
must be considered polylogical, as soon as we include the constant, if often implicit, reference to the
opinions of other philosophers concerning the issue or issues at hand (Leal 2019). In our analysis
we shall largely refrain to develop this aspect of the Russell-Copleston debate and content ourselves
with a few occasional pointers to the polylogical background.
3 Description of the Method Followed 53

application is grasped to the extent needed to appreciate its use in the special case of
the debate between Russell and Copleston that is at the centre of this volume.

3.1 Questions in Philosophical Argumentation

All philosophical texts are argumentative; but that does not mean that these texts
only contain arguments. In fact, philosophical arguments never exist on their own
but always as part of texts in which philosophers do many other things apart from
presenting and exchanging arguments: they ask questions, and sometimes they reject
questions as meaningless; they propose conjectures; they launch daring theses,
which they sometimes retract or dilute; they reject other philosophers’ theses as
false, absurd, banal, pointless, or nonsensical; they solve and dissolve problems;
they attempt to clarify or even to define terms; they reject terms as useless, merely
metaphorical, or just plain nonsense; they offer explanations and develop ideas; they
describe and classify concepts; they illustrate what they mean with examples; they
tell or make up stories and thought experiments, often weird and amusing ones; they
use metaphors and metonymies, either startling new and shining ones, other rather
trite and faded; they quote, allude, comment—and many other things besides.
The point is that, in order to grasp the arguments of philosophers, it is not enough
to pay close attention to them, even though that is very important. We should also
take those other things into account for argumentation-theoretical purposes. And,
when we deal with philosophers, the very first thing that we must listen to, before
we attend to their arguments, is the questions that intrigue and oppress them, those
they are laboriously trying to answer. Those questions are like a red thread guiding
us through the labyrinth of philosophical texts; and no philosophical argument can
be grasped except in the light of those questions.
Now, some philosophers put forward the questions they want answered, thus
apparently make understanding them easier. For instance, Jean-Paul Sartre uses a
question as a title for one of his celebrated essays (Sartre 1948, p. 55; transl. 1950):
(1) what is literature?

For purposes of clarity, we shall highlight questions with small capitals from this
point forward. Now, going back to Sartre, he was not satisfied with putting forward
a question to cover his long essay, but he also used questions to title its first three
chapters:
(2) Chapter I. what is writing?
(3) Chapter II. why writing?
(4) Chapter III. for whom does one write?

It might thus seem that (1) is the main question Sartre wants to answer in the essay
that bears that title; and it might perhaps seem that the questions (2), (3), and (4) are
54 3 Description of the Method Followed

secondary questions, to which it may be necessary to answer in order to arrive at a

satisfactory answer to (1), the main question.
Perhaps; but we should beware, and for various reasons. One is that philosophers
sometimes begin a piece of writing by putting forward a question in order to say
immediately that they will not answer it. Plato, for instance, starts his dialogue Meno
with the following question, thrown in with no warning by young Meno to old
Socrates (70A):
(5) is virtue teachable?

To this question a barrage of questions is added in a quick sequence (ibid.):

(6) or is virtue not teachable but only attainable through
practice?
(7) or is virtue neither attainable through practice nor learnable
but it comes to people as a gift of nature or in some other
way?

By putting forward all those questions, it would seem that Plato wants to answer all
of them; but Socrates replies to Meno (71A–C) that he cannot answer them because
he does not know the answer to a previous question, namely:
(8) what is virtue?

Moreover, Socrates proposes to Meno that they should both try to find an answer to
(8) together. Therefore, it might be thought that Plato’s dialogue is a philosophical
text whose purpose is to answer (8), and also that the reason the text starts with
questions (5), (6), and (7) is not to answer them in any way until some answer to (8)
is found. The whole thing would be a lesson in method. The trouble is that, as if Plato
wanted to bring us into deeper confusion, the dialogue proceeds in such a way that,
after many turns and bends, the dialogue partners come back to the initial questions
(5)–(7) and, under young Meno’s pressure, old Socrates consents to put forward an
answer to those.
It is thus a fact of life that one should not uncritically trust appearances when we
deal with philosophers. For the same reason, we cannot just take the fact that Sartre
used question (1) as the title of his essay and infer from it that his intention was to
answer it. The title question may, on the contrary, just have been an expedient to call
attention to other questions, such as (2)–(4). Or his intention may have been another
one.
There is a second reason why we should be very careful when philosophers starts
with a question and not assume right away that that is in fact the question that they
want to answer. For example, Timothy Williamson has an article with the following
title (Williamson 1995):
(9) is knowing a state of mind?
3.1 Questions in Philosophical Argumentation 55

Anybody starting to read this article might justifiedly think that (9) is the question
that Williamson is interested in answering; but look at the way he begins his paper
(p. 533; small capitals added):
(10) Let there be no vulgar suspense: the title will be answered in the affirmative.
What takes longer is to clarify its meaning.

Note that the small capitals are still used to highlight the question. By reading and
understanding Williamson, we know that he is not concerned with the question
whether knowing is a state of mind but rather with the question what the
meaning is of saying that knowing is a state of mind. There are indeed
many ways in which we verbalize a question. In (10), an abstract noun (‘meaning’)
is used; and a small moment’s reflection will show any reader that there are quite a
few other ways. We shall come back to that presently.
What was Williamson’s intention in putting forward his question in such a way?
Consider the stilted syllogism used in old logic textbooks to illustrate the traditional
syllogism:
All men are mortal.
Socrates is a man.
Therefore, Socrates is mortal.
The first premise is a truism, which we can also express by saying: ‘we all have to
die’, ‘death will reach us all’, ‘the last hour will come for all of us’, on appropriate
occasions. Nobody doubts that this is true. To ask whether human beings are
mortal is not really a question for any of us. We all know the sad answer. But
what it means that we are all going to die is a completely different
kind of question, in fact a philosophical one. It is a hard question, about which tons
of ink have been spilled in the course of the millennia. Something similar is what
Williamson wanted to convey; and it is quite possible that Sartre as well, when he
came up with the title of his essay, also wanted to convey that, although in one sense
we all know what literature is, the kind of question he is asking is something very
far from that low-level knowledge we all have. Philosophers do this sort of thing all
the time, and we should beware.
To ask a question at the same time that one has really a very different question
in sight, this other question being that whose answer one really cares about, is very
common in philosophy. What we should never forget is that, no matter what the
question is that the philosopher wants to answer, we’d better know what it is in order
to understand the philosopher’s arguments as well as the other things she says apart
from presenting and exchanging arguments. Philosophical questions are at the centre
of philosophical texts, always, everywhere.
Now, all questions (1)–(9) have been verbalized by means of interrogative
sentences, which is the canonical form that questions take in grammar. Some people
may understandably think that, if somebody produces an interrogative sentence, then
this is because she is asking a question; and conversely that, if somebody wants to
ask a question, the only way to do so is by using an interrogative sentence. Both
56 3 Description of the Method Followed

propositions are false, as we have seen with the question contained in (10). It is
therefore important, for the purposes of the kind of analysis that we shall perform on
the Russell-Copleston debate, to remember some well-known but easily forgettable
facts.
On the one hand, sometimes interrogative sentences are used to do something
else than asking a question. One famous case is the so-called ‘rhetorical question’,
which is usually defined as an assertion masquerading as a question. For instance,
Friedrich Nietzsche begins the preface to his book Beyond Good and Evil (1886) as
follows (bracketed small Roman numerals added):
(11) Supposing that truth is a woman—(i) what then? (ii) is the suspicion
not justified that all philosophers, in so far as they have
been dogmatic, have always had little knowledge of women?
(iii) that the awful seriousness and clumsy pushiness with
which they have so far tended to approach truth have been
heavy- handed and crude ways to get a woman’s interest?

The undertone of questions (ii) and (iii) in (11) shows that for Nietzsche the answers
are clear and well known. He formulates questions, but he does not really seem to
ask questions. He is rather declaring that dogmatic philosophers, i.e. philosophers
who are sure about the truth of what they say, have a series of defects, amusingly and
colourfully described, that prevent them from getting at the truth—such hopelessness
being the answer to (i). The fact that Nietzsche’s declaration is put into words under
the shape of questions does not change that.
Interrogative sentences do express questions, even though it is sometimes not
quite easy to get them. Consider the following passage, taken from Ortega y Gasset
(1932, Lesson V), in which he starts from a man in a room:
(12) What we ask ourselves is: (i) what is this room?, but not in the abstract
and without further ado, but rather we ask ourselves (ii) what this room
is as an ingredient in my life. As man finds himself living, he finds
himself in a circumstance or world. In this case, the circumstance is this room.
My life now is being in a room. We ask ourselves (iii) what this room is
insofar as it is where i am.

Question (i) is very easy to identify because it has the canonical form of a direct
interrogative sentence. It is a philosophical question put forward by Ortega for his
audience. But there are two other questions which are just a little bit hidden, although
we can recognize them as the grammatical objects of the verb ‘to ask oneself’. For
this reason, the sentences which play the role of object are called by grammarians
indirect interrogative sentences.
In (12) the give-away word is the verb ‘to ask’, but in other philosophical texts
the questions are more deeply embedded, because no such verb is present. We can
nonetheless recover them by paying a little bit of attention. This is, for instance the
case of the beginning of Aristotle’s treatise On the Soul (402a 22-402b 3), where the
pronoun ‘it’ refers to ‘the soul’ (psyche):
3.1 Questions in Philosophical Argumentation 57

(13) It is perhaps first of all necessary to distinguish (i) in which genus the
soul is and (ii) what it is. i mean (iii) whether it is a ‘this here’,
that is, a substance, or a quality, or a quantity, or some-
thing in some one of the othe delineated categories; then (iv)
whether it is one of the things which exist potentially or
rather whether it is a fully realized being. For this makes no
small difference. One must also inquire (v) whether it is divided or
undivided, and (vi) whether all souls are similar in kind or not;
but, if not similar in kind, then (vii) whether they differ in species or
in genus.

Aristotle asks here seven questions, all of them by means of indirect interrogative
sentences which are, however, not grammatical objects of the tell-tale verb ‘to ask’ but
of slightly less transparent verbs such as ‘to inquire’ or even ‘to distinguish’. These
verbs, and many others, should be closely examined when reading a philosophical
text, because they usually indicate that a question is being asked.
The construction verb+interrogative sentence (direct or indirect) stands out, even
if the verb is not always as transparent as ‘to ask’. In contrast, let us look back at
Williamson’s passage (10) above: the verb here is ‘to clarify’ (because the question
asked is one of clarification) and its grammatical object is an abstract noun in place
of an interrogative sentence.2
Now compare (10) with (13) for a further hint that we are dealing with a question:
Williamson says that ‘what takes longer is to clarify its meaning’. Aristotle says that
‘it is necessary to distinguish’ and ‘one must also inquire’. The three expressions
serve to point to a task to be done. And indeed, philosophers often refer to their
questions with words such as ‘task’, ‘difficulty’, ‘trouble’, ‘obstacle’, ‘problem’, and
the like. These are all a special family of abstract nouns (see note 2) whose function
is to highlight a philosophical question. Philosophers are thus prone to speak of
the problem of free will, the task of reducing qualia, the trouble with dualism, of
epistemological questions in general or of the question of being in particular. When
one reads phrases like this, we can be sure that somebody is asking something or has
asked something, even though it is not always easy to say what exactly the question
is.

2 An abstract noun denotes a noun that has been derived from a predicative word (a verb, an adjective,
or another noun) in such a way that it denotes an abstract object: ‘organization’ from ‘to organize’,
‘prettiness’ from ‘pretty’, ‘knighthood’ from ‘knight’. The crucial linguistic function of abstract
nouns is nominalization: the power to condense whole sentences so as to make them propositional
contents that can be objects of thought (Porzig 1942, Lees 1960). Consider the sentence: ‘Columbus
discovered America in 1492’. Thanks to the abstractive technique of derivation, we can speak of
‘The discovery of America by Columbus in 1492’ and even, by eliminating parts of the sentence,
just ‘The discovery of America’. Take now Williamson’s passage (10). The full question referred
to there would be something like: ‘what does it mean to say that knowing is a state
of mind?’ Thanks to the abstract noun ‘meaning’ we can say ‘its meaning’ whilst understanding
that the possessive pronoun ‘its’ refers to ‘saying that knowing is a state of mind’. In fact, we have
here not one nominalization but four (meaning, saying, knowing, state of mind, where even ‘mind’
can work as a non-derived abstract noun).
58 3 Description of the Method Followed

It is thus possible to become skilled and sophisticated about the use of words and
expressions by means of which a question comes to be expressed in philosophical
texts. However, that does not mean that we understand the question itself, let alone
that we know how to answer it or even that we can identify the answer, if any,
which the philosopher has put forward. Nonetheless, one has to begin somewhere;
and a good way to begin is to be on the alert for the questions that philosophers are
interested in, and this implies paying attention at the different ways those questions
are verbalized in their texts.
Now, philosophers, like everyone else, put forward questions of two broad kinds,
which grammarians refer to as open and closed. The latter are so called because
the way of asking gives us a restricted number of options from which we have to
choose. The so-called ‘multiple-choice questions’ in examinations are closed ques-
tions, because we have to pick one of the options given and are not allowed to add any
further option. The simplest closed questions are those in which the options are just
Yes or No, the affirmative and the negative. In contrast, open questions are so called
because the number of options is not pre-determined by a list of possible answers.
Whoever faces an open question is confronted with an unlimited landscape. Such
an indetermination is what makes answers to open questions in a test so much more
difficult to produce and to evaluate.
Let us have a look at the examples above from the perspective of the open-closed
distinction. From now on, we shall refer to question always by using the form of
the indirect interrogative sentence. First, all of Sartre’s questions (1)–(4) are clearly
open:
what literature is
what writing is
why write
for whom to write
None of them gives us a list of possible answers, so that we have to pick the right
one. The same is the case with question (8), raised by Socrates:
what virtue is
Compare the latter, however, with all of Meno’s questions (5)–(7):
whether virtue is teachable
whether virtue is only attainable through practice
whether virtue comes to people as a gift of nature
none of the above
The reader can see that Meno’s barrage of questions is actually like a multiple-
choice test. This means that Meno is putting forward just one closed question. Again,
question (9) in the title of Williamson’s paper is also closed:
whether knowing is a state of mind
3.1 Questions in Philosophical Argumentation 59

Only in this we are only given the choice between Yes and No, without any
alternative. However, as we have seen in passage (10), the question that really interests
Williamson is open:
what it means to say that knowing is a state of mind
If readers now look at the other questions used above for illustration, they will
ascertain that all of Ortega’s questions in (12) are open, whereas in the passage
(11), questions (ii) and (iii) are closed questions, yet his question (i) is open (‘what
then?’ can roughly be paraphrased as ‘what follows from supposing truth
is a woman?).
This is also the case with passage (13), where Aristotle presents us with a combi-
nation of open and closed questions. If we examine them carefully, we can observe
that there is a definite relation connecting them: when a philosopher poses an open
question, it is often the case that it is not as open as it may appear at first sight, for
the answers which the philosopher considers viable are restricted in such a way that
an open question is equivalent to certain closed questions.
Thus, take the two initial open questions in passage (13):
in what genus the soul is
what the soul is
These two seem prima facie not to be accompanied by a list, but this is a false
appearance, for Aristotle, immediately after posing them, says: ‘I mean’ (λšγω δ),
by which utterance he introduces four closed questions:
whether the soul is a substance
whether the soul is a quality
whether the soul is a quantity
whether the soul is another of the catergories that have been
distinguished
The last question is shorthand for a continuation of the list down the seven
remaining categories distinguished in Aristotle’s theory. So, we have again a multiple-
choice test, i.e. a single closed question with ten options. These options are Aristotle’s
ten categories, and they give content to the apparently open question what the soul
is (τί ™στι ¹ ψυχή). Of course, we are able to take passage (13) and reconstruct
question (iii) as an explanation or clarification of the content of question (ii), and not
as a different question, thanks to a certain amount of background information about
60 3 Description of the Method Followed

Aristotle’s concepts and methods.3 This is not particular to Aristotle but common to
all philosophical questions and indeed to questions altogether and as such.4
The analysis of this example teaches us that sometimes, possibly more often than
we may have suspected, an open question raised by a philosopher matches a closed
one. Take the three next questions in (13):
whether the soul exists potentially or is a fully realized being
whether the soul is divided or undivided
whether all souls are similar in kind or not
They are all closed question, which give us stark alternatives: potential vs. actual
existence, divisible versus indivisible, similar in kind vs. different in kind. All three
questions, (iv)–(vi) in passage (13), correspond to heavy-duty philosophical doctrines
of Aristotle, different from his theory of categories. Altogether, we can see that the
list of questions in (13) is not haphazard but follows a definite methodology. Now,
question (vi) has a notable feature: it is followed on its heels by question (vii). Both
questions are tied to the Aristotelian theory of genera and species. Very briefly,
soul—or rather psyche, which is perhaps more accurately rendered as ‘that which
makes something be alive’, which can be shortened to ‘principle of life’ or even just
‘life’—is a kind of being, so the question arises what kind.
If this question is asked today, people may try to answer in a rather nonchalant way,
taking the word ‘kind’ in a largely nontechnical sense. But for Aristotle, that question
involves a whole intellectual edifice, within which the concepts and distinctions we
have mentioned have a role to play in the construction of arguments. Without entering
into more details about Aristotle’s logic and metaphysics, suffice it to say that the
seven questions contained in passage (13) are a few nodes of a large, complex, and
tight network. None of them stands on its own feet. They are all interrelated. In fact,
a philosophical theory and methodology can be defined by such a network. This is as
true of Aristotle as of any other important philosopher and philosophical school or
movement. If the reader of a philosophical text is serious about wanting to understand

3 The term ‘background information’ is intended to cover both background knowledge (van Eemeren
and Grootendorst 1983, pp. 135, 166, 180; Kahane 1984, p. 171) and background beliefs (Kahane
1984, pp. 16–17, 35, 171). The old Platonic distinction between knowledge and belief belongs to
philosophical epistemology and, after the deluge of undecisive literature on the topic, we think it
advisable to set it aside for the purposes of argumentation theory. The word ‘information’, whose
home is in computer science, is austere and abstract enough for our needs (cf. ‘unstated information’
in Kahane 1984, p. 37).
4 Background information can actually be defined as what makes possible both to understand,

wholly or partly, the question at issue, and to ask further questions to improve one’s grasp of the
original question (Leal, 2021). So, it is not enough to say that one cannot understand a question or
ask a good question if one lacks background information, but it should also be said that one lacks
background information if one cannot understand a question or ask a good one. The two concepts
are interdefinable; but we suggest that the concept of question is basic to argumentation theory, so
we should use it to define background information and not the other way around. After all, one
major problem in argumentation theory is precisely that, although it is widely understood that the
concept of background information is essential to the interpretation of what people mean when they
argue, yet nobody has found a way to define it, let alone theorize about it.
3.1 Questions in Philosophical Argumentation 61

that text, including its arguments, she needs to get acquainted with the network of
questions of its author.
To give an example, philosophers are prone to ask questions of the form what
x is, where X can be truth, virtue, knowledge, the good, the just state, and whatnot.
However, the meaning of the question may vary enormously, for some philosophers
may be thinking of the essence of x, whereas others may rather be thinking of the
meaning of the word ‘x’, or even both, although perhaps not at the same time. Philoso-
phers may thus seem to be asking the same question, yet each one may in fact be
asking a different one. So, when Sartre asks what literature is, Socrates what
virtue is, Aristotle what the soul is, Ortega what this room is, there is no
guarantee that they are asking the same type of question, and so no guarantee, either,
that they would be satisfied with the same type of answer. If anything, we should
expect that this is rarely the case. And this is true of all the other questions used for
the sake of illustration, and in fact of all philosophical questions. The fact that we
can become proficient in detecting the question or questions in a philosophical text
does not imply that we ipso facto understand what its author was actually looking
for.
Nonetheless, it is crucial that we do detect the answers, even when the text appar-
ently does not contain any. We shall end this section with a couple of examples. They
are taken from the beginning of two of the most famous and enigmatic texts of the
whole history of philosophy:
(14) The Monad, of which we will speak here, is nothing else than a simple
substance, which goes into composed ones; simple, that is, without parts.
(15) The world is all that is the case.

The reader will probably recognize here the first, numbered propositions in
those curious pieces of writing which we know, respectively, as Leibniz’s
Monadology (written in 1714) and Wittgenstein’s Tractatus logico-philosophicus
(1921). Anybody who reads these two passages for the first time is nonplussed.
One’s reaction can only be to ask oneself what on earth this man is talking
about. This question is the right one, because it admits one’s colossal ignorance in
front of the text. And even if it takes time to find out the answer to that question,
one step in the right direction is to modify it a little bit and ask oneself what the
question is that this man is answering by this proposition. The inno-
cent reader does not know that question, and this is the main reason why she does
not understand the answer represented by a proposition so important to each author
that he put it right at the beginning of his treatise and numbered it for good measure.
The point is that philosophers not always follow the example of Aristotle in (13)
and formulate their questions in an explicit way (which does not guarantee, of course,
that we will understand the question only because it is explicit). One possible reason
for omitting the questions is that philosophers are writing for an audience which
they think will understand. We all know that human communication is, somewhat
paradoxically, both repetitive and incomplete. Some of the information we want
to impart is known by the recipient, so we leave it unexpressed; and some of the
62 3 Description of the Method Followed

information is so important for us and for the recipient that we express it several
times and in different ways. So, perhaps that was the reason why both Leibniz and
Wittgenstein thought it perfectly appropriate to start in this very abrupt manner. The
fact of the matter is that even a flat-out assertion makes only sense if it is viewed
as an answer to a question, so that trying to discover that question—or rather, the
network of questions—is the first step in understanding the text.

3.2 Disagreements in Philosophy

In philosophy, lack of agreement is a constant. We may even say it is the normal

situation. What is not normal, however, is when philosophers just accept the fact that
their opinions differ and leave things alone. On the contrary, they normally engage in
discussion. When the rare event happens that they do not discuss, we can practically be
sure that at least one of the philosophers involved thinks that the differences opening
up between them are so deep that it is not productive to have a discussion. The most
frequent attitude in those rare cases is that the other philosopher or philosophers
labour under such a profound mistake or confusion that it is nearly impossible for
them to see the error of their ways. A late scholastic adage has it that ‘it is impossible
to discuss against those who deny the principles’ (contra principia negantem non est
disputandum). This adage, like any other philosophical proposition, may be, and has
been, disputed (Nelson 1904; Popper 1976). However, most philosophers often act
as though they believed in it.
What happens in a discussion? We can distinguish two varieties, which we shall
calle the playful and the serious.
Discussion can be a kind of game with a winner. This is the case in competitive
debates, which are guided by Yes-or-No closed question, two individual or collective
parties (the ‘affirmative’ and the ‘negative’), a set of rules of conduct, and a jury that
decides who the winner is. For this kind of game, it does not matter in the least
whether the parties have a firm opinion on the issue. The only important thing is that
they prepare for the debate and show ability in defending the position assigned them
and in attacking their adversaries. Something similar, although vastly more chaotic,
takes place when a group of friends start a discussion on questions such as which is
the best football team or what is the best movie of all time. In most
cases, nobody really cares about the right answer, if any. What matter is winning the
discussion. In this playful use of discussion, we cannot say that people are trying to
resolve their difference of opinion.
The serious variety of discussion is when people actually want to make a decision
with which they can agree. This kind of discussion is definitely not a game and here
it does matter that agreement is reached. In those cases, four methods for coming to
a decision seem to be possible (van Eemeren and Snoeck Henkemans 2017, p. 24):
3.2 Disagreements in Philosophy 63

(16) a. Lottery. The decision is left to a random process. This first method makes
sense if the parties think there is no point in discussing or if they have tried
discussion for a while in vain.
b. Voting. This method is usually preceded by extended discussion, but it is
widely accepted by the parties that they cannot resolve their differences of
opinion, so a certain rule is chosen, e.g. relative majority.
c. Refereeing. An individual or a group is, from the start, appointed to play the
role of judge, arbiter, or referee. He or she or they will listen to the arguments
of all parties and make a decision which all parties are committed to accept.
This method can be combined with voting, as in parliaments or constitutional
courts. (Incidentally, this method is the only one that can be used for the
playful type of discussion, but there is a difference in agreeing with a position
decided by a judge and agreeing in who the winner in a debating game is.)
d. Persuasion. This method is used when the parties trust that it is possible
that the merits of the arguments offered by the parties will eventually lead to
resolve the difference of opinion. Arguments are also presented in the three
methods described above, but they play a different role in each. In all of
them, argument always shows that agreement between the parties cannot be
reached, but in voting and refereeing it is believed that voters and referees can
be, to some extent, be swayed by argument. In persuasion, however, it is the
arguments themselves which will directly prevail in the other party’s mind.
After surveying these four decision-making methods, two questions arise: (a)
which variety of discussion can be applied to philosophical argu-
mentation; (b) which method or combination of methods is appropriate
to it.
There is admittedly some justification to the widely shared impression among
non-philosophers that philosophers are just involved in a kind of very intricate game.
This impression stems from the fact that they do not manage to reach agreement,
or at least not for long. (Philosophical schools have a tendency to branch out in
many directions.) Moreover, the self-appointed task of philosophers is to question
many things, if not indeed everything. However, philosophy has such a long history
(two and a half millennia at least in the West), and philosophers are in general so
intelligent and refined, and their writings are occupied with such seemingly serious
and important matters, that it would be too frivolous to conclude that the whole
thing is nothing but an elaborate game. Besides, the intellectual constructions of
philosophers look like they have been built up in order to persuade other people, and
it must be conceded that they often concern questions like the following:
(17) how we ought to live
what our duties and responsibilities are
what the most iportant thing in life is
what is truly real

If these are not serious questions, then it would be difficult to say which questions
would then be serious.
64 3 Description of the Method Followed

So, for all these reasons, we have to conclude that philosophical discussions are to
be considered serious rather than just playful, concerned more with making decisions
and not just winning, those decisions being, incidentally, weighty. Nonetheless, we
observe again and again that a philosopher does not persuade many other philosophers
for a considerable amount of time, in contradistinction with other serious kinds of
discussions (alone or complemented by the methods of lottery, voting, or refereeing),
where we see people reaching agreement.
This is puzzling; but before we tackle the puzzle, we must first consider in a very
broad way what are the main features of philosophical discussions. Let us start with
the simplest case: a closed question that only admits of two answers, the affirmative
and the negative. Take the hoary old philosophical question of
(18) whether the world is eternal or not

There is then disagreement between any two philosophers when one of them, call
her Alpha, takes the affirmative position on this question, and another philosopher,
call her Beta, reacts by taking the negative position, in full knowledge that she is
contradicting Alpha. In logic, we say in such a case that both positions cannot at the
same time be either true or false. In other words, if one of the two position is true,
then the other is false; and the other way around, if one is false, the other is true. In
such a situation, we can say that there is a bilateral difference of opinion, meaning
that both sides, Alpha and Beta, are taking a definite position about question (18).
Is this the only possibility? No, because there may be a third philosopher, call
her, Gamma, who reacts by saying that she does not know whether the world
is eternal or not. The attentive reader will recognize the Alpha-Beta-Gamma
situation as formally the same described by the founders of pragma-dialectics with
which we opened this chapter. From now on, however, a difference will emerge which
we have to pin down in order to be able to apply pragma-dialectics to philosophy.
We can say that Gamma accepts question (18) as a perfectly intelligible and
sensible question, only expressing doubt with respect to which of the two
answers, the affirative or the negative, is the right one. In logical
terms, Gamma, in contradistinction to both Alpha and Beta, is not taking a position
with respect to question (18) because she can neither say that the affirmative answer
to (18) is true nor that it is false. There is a difference of opinion also between Alpha
and Gamma, only it is not bilateral but unilateral, for in this case it is only Alpha
who takes a stand on the question, whereas Gamma does not. The same is true of the
difference of opinion between Gamma and Beta.5

5 The distinction between ‘bilateral’ and ‘unilateral’ is the same as the standard pragma-dialectical
distinction between ‘mixed’ and ‘non-mixed’ differences of opinion mentioned in Chap. 2, Sect. 2.3.
We have two reasons for this slight terminological change. One is that the proposed expressions are
more intuitive than the traditional ones. This would not be a strong enough reason to counteract the
benefits of keeping a received terminology. The second reason, however, tips the balance: we can
refer comfortably to trilateral, and more generally multilateral differences of opinion when we go
beyond discussions with two parties (‘dilogues’) to those with three or many parties (‘trilogues’,
‘polylogues’).
3.2 Disagreements in Philosophy 65

At this point it is appropriate to ask ourselves what makes someone like Gamma
hesitate when faced with the option of siding with Alpha against Beta or with Beta
against Alpha. We can recognize two broad kinds of philosophical doubt.
The first kind concerns the meaning of the terms used to express question (5).
We must after all admit that both the term ‘world’ and the term ‘eternal’, when not
used in the cavalier way of everyday speech, but rather in a philosophical way, have
a tendency to cause trouble. Philosophically, the world is the whole set of all things;
and we say of something that it is eternal when it does have nor a beginning nor an
ending in time. In everyday speech, we deal with this or that thing, never with all
things; and again, the things we deal with, in everyday speech, have begun at some
point in time and they will also end sometime. In ordinary life, there is no place for
the world or for eternity in their philosophical sense. Therefore, Gamma might have
some scruples concerning the meaning of those terms, and so hesitate before giving
her assent to what Alpha asserts (that the world is eternal) or to what is asserted by
Beta (that it is not).6
This first kind of philosophical doubt is exactly the same we encountered in
Sect. 3.1, when we referred to the dialogue between Socrates and young Meno. Let
us look again at the latter’s initial question:
(19) whether virtue is teachable or not
Let us for the time being ignore the other options mentioned by Meno, which we
ascertained in Sect. 3.1. Socrates replies to Meno that there is no way to know the
answer to (19) if one does not first have an answer to another, more basic question,
namely:

(20) what virtue is

By the way, you can bet there will be other philosophers who might want to add
another question, namely:
(21) what teaching is

In other words, the first kind of philosophical doubt concern the fact that words do not
exist in isolation: they always are rather part of whole network of questions, among
which the philosopher moves and sometimes, like Socrates in Plato’s dialogue, dares
to say that one of them must be answered first, before we can pretend to answer the
other one.

6 It may be of interest to revert to the discussion, imagined by van Eemeren and Grootendorst, among
three people who have a difference of opinion around the question whether women have a
logic of their own. For we can imagine that one of them or a fourth party says: ‘Well, it rather
depends on what we mean by “logic”, does it not?’ By putting forward that new question,
the questioner has made a first step towards a philosophical discussion. Although philosophy is a
‘field’, with a hierarchy of prestige and proficiency, the fact of the matter is that it can appear in
principle in any conversation. A similar point, tracking philosophy in all sorts of non-professional
situations, is made by Romano (2012).
66 3 Description of the Method Followed

The second kind of philosophical doubt concerns not so much the meaning of the
terms contained in the question as the sort of argument that is required to persuade
somebody of one or the other answer. Very often, when we hear or read a philosophical
thesis, we have the impression that it would exceedingly difficult, if not altogether
impossible, to ‘prove’ such a thesis, that is, to construct an argument such that,
whoever would study it, would have to assent to its conclusion. In some cases, it
is difficult to escape the impression that any argument offered in support of the
thesis would have to be weak and deceitful. Think, for instance, of another classical
question:
(22) whether god exists or not

There might be Alphas who believe that God exists, and Betas who believe that God
does not exist. In fact, it is pretty easy to find members of each class. However, there
is scarcely anybody today who thinks that it is possible to prove (in some sense of
this word) either of those propositions. The more or less general consensus seems to
be that the God question is a question of faith, not of science or of knowledge, and
that there is no way prove that God exists or that God does not exist. We live, in the
present world, in a general climate of agnosticism, the term invented in the nineteenth
century for the stance that no answer to question (22) is forthcoming. In other words,
we are all Gammas now with respect to the question of god’s existence.
Anyway, it would seem that all is finished here. In standard pragma-dialectical
terms we would have two kinds of difference of opinion, Alpha vs. Beta and Alpha (or
Beta) vs. Gamma, bilateral and unilateral, mixed and non-mixed. Yet in philosophy
there is room at least for a fourth character, Delta, who gives voice to doubts, not
about Alpha’s assertion, but about the question itself and as such. Such doubts can
be more or less radical and extreme, but they all make the discussion move in a fresh
direction. For Delta, what is worth discussing is the pertinence of the question, not
the truth or falsehood of the answer. Take question (18), whether the world is
eternal or not. Gamma may hesitate between siding with Alpha (who says that
the world is eternal) or with Beta (who says that the world is not eternal), precisely
because she is not quite sure what the meaning of ‘world’ (or of ‘eternal’
or of both) is. The same is true of question (22), whether god exists or not.
In the case of (22), Gamma’s doubts may concern the term ‘God’, but it could come
to concern the term ‘exist’.
Now, the plot thickens, because, once Gamma comes to see that she cannot take
a stance on (18) or (22) because she does not quite understand one or several of the
terms in those questions, and indeed because she may reflect that there is something
wrongheaded in asking questions like (18) and (22), Gamma’s attitude may slide
down the slope and, bit by bit, come to resemble that of Delta. These two kinds of
philosophical doubt—about the terms and about the question itself—are the same
we described above. And this is not fiction we are talking about. It can be argued, and
indeed documented, that this is what has happened, again and again, in the history
of philosophy in regard to questions (18) and (22). Moreover, it could also be argued
3.2 Disagreements in Philosophy 67

Table 3.2 One question,

Question: whether god exists or not
three parties, six differences
of opinion Differences of First-order bilateral Alpha, Beta
opinion Alpha, Gamma
Beta, Gamma
Second-order bilateral Delta, Alpha
Delta, Beta
Delta, Gamma

that many, if not most, philosophical disputes are in the end about what questions
it is legitimate to ask.
If so, then Gamma’s dubitative reaction to Alpha’s (and Beta’s) assurance is
philosophically unstable and provisional. Either Gamma manages to overcome her
doubts and sides with one of the other two, or she ends up on Delta’s side. However,
in the texts we can easily find Gamma’s dubitative attitude, so it is justified to keep
the distinction between four attitudes and not try to reduce them to three. In this way,
if Gamma has not gone all the way down to where Delta is situated, for as long as
she still believes that the questions makes sense and can be answered, even if it is
difficult, we must accept that we have a disagreement of a different sort dividing
Alpha, Beta, and Gamma on one side and Delta on the other side. We can also say
that this difference of opinion is bilateral but, as it were, of a higher order. For it
does not concern the question of truth anymore but rather one of meaning or perhaps
of pertinence. The issue is not which answer is true and which false, but
rather whether the question should be asked at all. Table 3.2 puts it all
together.
Some readers may think that such an extended classification might only apply to
closed Yes-or-No questions, but not for the other two kinds described above, namely,
multiple-choice closed questions and open questions. However, a small moment’s
reflection is enough to see that all philosophical questions must end up as Yes-No
questions, as was perfectly understood by the Founding Fathers—Socrates, Plato,
and Aristotle. The medieval thinkers who wrote quaestiones also understood this.7
We must now direct our attention to a different point. Philosophers often express
themselves in a very compressed way, by discussing many questions in one fell
swoop. A good example of this is taken from Monarchia, a treatise on political philos-
ophy written by Dante Alighieri at the beginning of the fourteenth century, which
means in the middle of long disputes about the appropriate relationship between
the papacy and the ‘temporal’ monarchies—such as the Holy Roman Empire or the
newly emerging kingdoms in Spain and France. In that treatise, Dante tells us that
he will tackle three questions:

7The argument in brief is that any open question (e.g. ‘where are the car keys?’) can only
be answered by replacing it by a series of closed Yes–No questions (‘are they in my office
or not?’, ‘are they inside the car or not?’, and so on).
68 3 Description of the Method Followed

(23) a. whether temporal monarchies are necessary for the well-

being for the world
b. whether the ancient romans were right in abolishing their
temporal monarchy
c. whether a temporal monarchy’s authority comes directly
from god or rather from the pope as god’s vicar on earth.

Note that all these are closed questions. Questions (23a) and (23b) only admit Yes or
No for an answer, and (23c) offers just two choices. Each was a delicate proposition,
both philosophically and politically, at the time. In particular, question (23c) was the
most delicate, because the first option was the preferred one for the emperor and the
kings, whereas the second option was naturally favoured by the pope. In fact, Dante
would have called for serious trouble if his treatise had circulated widely during his
lifetime, something he took good care to avoid. His position with respect to (23c) was
close to what we now call the separation of state and church, i.e. that the authorities of
the monarchs comes directly from God, so that the pope has no authority to exert upon
the temporal monarchies. When Dante’s treatise became known to many people, it
was burned at the stake and the poor poet’s bones were well-nigh exhumed and burnt
together with it.8
Leaving aside all these rather awful facts, Dante’s intention is to give a definite
answer to each of the three questions in (23). Each such answer means that he is
taking a stand. In the course of his argumentation, Dante will also take a stand with
respect to a few other questions. For arguing in favour of a definite answer to any
question implies that the philosopher relies on previous definite answers to other
question in the relevant network. This is the way arguments are built up.
Now take the following argument, which is but a very small part of a long
argumentation:
Since every truth that is not a principle is made manifest from the truth of some principle, it
is necessary in any inquiry whatsoever to have knowledge of the principle, to which one can
recur analytically for the sake of the certainty of all propositions that are subsumed under
the principle. [Dante, Monarchia, II, 4; my translation]

We readily identified this piece of the text by the word ‘since’. If we say that
something is this since something is that, the latter is a reason that supports the
former. The reason to assert that it is necessary to have knowledge of the principle
in any inquiry is that every truth that is not a principle is made manifest from the
truth of some principle. Now, what is the question which Dante seems to be trying
to answer in this passage?
The following is one possibility:

8 Dante’s tract was written in Latin, the language of academia at the time, and not in Italian, as
Dante’s much better known poetical works. The writing took place shortly before Dante’s death
(1321) and long before Gutenberg’s invention of the printing press (1450). Publication thus was
a matter of dissemination through copies. Our information about the burning of the tract comes
from Bocaccio in 1329 (see Shaw 1996, Cassell 2004). In 1554 the tract was put in the index of
prohibited books of the Catholic Church, and in 1558 it was printed for the first time in Basel. It is
now part of the canon of Western political philosophy.
3.2 Disagreements in Philosophy 69

(24) how we should proceed when making an inquiry

Dante’s answer to (24) is broadly that the inquirer needs to know what the prin-
ciple of her inquiry is. The principle of an inquiry? What is Dante talking
about? To find out, let us attend to the reason he gives us for a rather mysterious
answer. That reason is, after some simplification and some guesswork, that any truth
to be obtained within an inquiry is proved by bringing it back or reducing it to the
principle of the inquiry. Bringing back? Reducing? To understand all that we must
consider quite a few other things.
Dante’s argument depends upon a whole way of thinking about another classical
question in philosophy:
(25) what the structure of scientific knowledge is

Specifically, Dante presupposes an Aristotelian concept of ‘science’ (scientia,

™πιστήμη) as a hierarchy of propositions which are ultimately derived from a prin-
ciple. This conception of science, i.e. this answer to question (25), was expounded
and defended by Aristotle for the first time in his Posterior Analytics and, as such, it
was, in Dante’s time, well known to all educated Europeans.9
The Aristotelian conception of science has become in the meantime a piece of
arcane knowledge, unfamiliar and largely unintelligible. (Just think of all that has
happened since the fourteenth century.) Very few people today have a grasp of the
implications of the Aristotelian conception, but even with a mediocre idea of it, most
people would find it difficult to accept as part of contemporary talk about political
philosophy or political science. The point is that our attitude towards Aristotle’s
standpoint is at most that of Gamma, for we are very much in doubt about Dante’s
answer to question (24) as well as about the argument he offers in its support. So,
we don’t only question the assertion that the inquirer needs to know the principle of
her inquiry, but also a whole set of propositions about the general question what it
means to have scientific knowledge as well as the particular question what
it means to have scientific knowledge of politics.
Of course, we could say that, in Dante’s case, his immediate readers accepted the
Aristotelian conception of science and so had no difference of opinion with Dante
on that point. Yet that is not the case. The history of the reception of Aristotle’s
conception of science is a very complicated historical question. Fortunately for the
purposes of this chapter, we don’t have to enter that battleground, for it is already
perfectly clear that nearly every contemporary reader does have such a difference of
opinion with Dante; and the point is that our difference of opinion does not concern
the answer to a single question, but it covers many other questions and their answers.
Situations such as this are pretty common in philosophy, for what a philosopher
asserts is always part of a set of questions and answers which entertain tight inter-
relations, so that, when a philosopher does not agree with what another philosopher

9 The interested reader can find several modern expositions of the Aristotelian conception of (what
he understood by) science. Among them, we highly recommend Barnes (1969), Hintikka (1972),
Irwin (1988), Ferejohn (1991).
70 3 Description of the Method Followed

says, the points around which that disagreement turns are quite numerous. It is there-
fore crucial to understanding the writings of philosophers that we try to detect the
various questions underlying them. Let us, for a start, abandon the simplification of
the above exposition of what Dante says and look again at the passage quoted above:
Since every truth that is not a principle is made manifest from the truth of some principle, it
is necessary in any inquiry whatsoever to have knowledge of the principle, to which one can
recur analytically for the sake of the certainty of all propositions that are subsumed under
the principle.

This long sentence contains two different conclusions: (a) that it is necessary in all
inquiries to have knowledge of the principle, and (b) that in any inquiry the principle
is that to which we can recur analytically for the sake of certainty of all propositions
assumed under the principle.
We must ask again: what are the questions which these two propositions are trying
to answer? Remember that Aristotle is the inventor of logic, including methodology.
One of his most insistent questions concerns the difference between theorit-
ical knowledge proper (e.g. that of a mathematician) and ordinary empir-
ical knowledge (e.g. that of a carpenter or a shoemaker). Aristotle thinks that
the former has a logical structure governed by principles, but the question is what
principles are these. One answer to the question is precisely (b), namely that
in any inquiry the principle is that to which we can recur analytically. That is what
explain the function which a principle has in an Aristotelian science. In thus distin-
guishing between (b) and (a), we can see that a reader of Dante can agree with him
on (a) without necessarily agreeing on (b). In other words, that reader can agree that
an inquiry rests on certain general propositions, which we may call ‘principles’,
but this does not force that reader’s hand in also having to agree that those general
propositions work in such a way that all other propositions contained in the inquiry
or emerging from it are proved or certified by, get absolutely certainty from, the
general propositions in consequence of some definite inferential process, which in
Aristotle’s case is deductive. It is in fact easy to imagine such a reader, at least in
the present. Modern physics, for instance, rests on general propositions but most
physicists would agree that their scientific knowledge is not certain or deductive in
any traditional sense of those words.
And we can go even further. In (b) a funny phrase occurs: ‘to recur analytically’.
Wihin that phrase, an interesting adverb occurs: ‘analytically’. What is the meaning
of such a phrase and such an adverb? It can be argued that, historically, the adverb
‘analytically’, as well as its cognates, the adjective ‘analytic’ and the verb ‘to analyze’,
have very specific meanings given to them by Aristotle. In fact, we can make explicit
what is implicit in them by saying (c) that the objects of any inquiry are subsumed
under categories in such a way that a scientific inference takes place by means of
definitions of the objects thus categorized.
Our readers can now appreciate how things have become horribly complicated.
Terms like ‘category’ and ‘definition’ belong to Aristotle’s logic and methodology,
disciplines which he gave no other name but—analytics. It is by distinguishing
between the supreme genera of things (the ‘categories’) that it becomes possible
3.2 Disagreements in Philosophy 71

to organize the sciences; and it is by ‘defining’, according to certain rules, what

principles these are that we can construct proofs and attain certain knowledge;
and all proofs are ‘analytic’ because, no matter what truth we discover about any
given object, we must be able to reduce it, by means of ‘analysis’, to ever more
general propositions, until we arrive at the most general of them, which is, quite
precisely, the ‘principle’ of the inquiry.
We do not need to enter into more of the gory details of that extraordinary concep-
tion of science that Aristotle created and had so much influence over so may excellent
authors to see that, the deeper we go into any argumentative text within the history
of philosophy, the richer will be the set of questions and answers that make that text
possible. Each and every philosophical argumentation is tightly packed in that sense.
The reader may ask why we insist on these things. Well, because, if we
put together the answers (a)–(c), we can appreciate that they are specific answers
to equally specific questions put forward within the Aristotelian ‘analytics’, with
which Dante and his immediate readers were perfectly familiar. In contrast, imagine
a different reader, say, a specialist in current political philosophy or political science,
highly skilled in mathematical models of politics, and let her read Dante. She could
easily come to accept both (a) and even (b) but resolutely express her disagreement
on (c). And, of course, if we would take apart (c), we would soon enough find
further underlying questions, which would lead us to more subtle and sophisticated
differences of opinion.
We hope our readers have now seen the point of all this long commentary on
Dante: as soon as we started analyzing his argument, it seems as though a single
question was at the bottom of it, namely (24), how we should proceed when
making an inquiry. Yet the historical sketch above (of a sort that could be done
on any argumentative text in the history of philosophy) reveals that Dante was taking
a position on three very different questions:
(26) a. whether the manner of conducting an inquiry in order
to obtain theoretical knowledge requires us to know the
principles us to know the principle of that inquiry
b. whether the function of a princple is to give us full
certainty about the truths to be discovered in the course
of the inquiry
c. whether the way in which in which that certainty is trans-
mitted from the principle to the turths discovered in the
course of the inquiry is ‘analytic’ in the sense of aristotle
Once we have made explicit these three specific questions, we become aware that
the possible bilateral differences of opinion are as depicted in Table 3.3.
In general, we can say that the differences of opinion between any two philoso-
phers may concern the answer to one, two, three or, more generally, many questions
from a given set. Note that Table 3.2 may expand ad libitum as we include further
questions down the road to Dante’s thought. And the same will happen with any
philosophical text, so that in the end all differences of opinion in philosophy will be
multiple, even if, at first, they may have appeared as single (Leal 2019).
72 3 Description of the Method Followed

Table 3.3 Three questions, four parties, three bilateral differences of opinion
Combination Questions (26) Bilateral difference of opinion with Alpha
a b c
Dante (Alpha) Yes Yes Yes
Beta 1 Yes Yes No Single
Beta 2 Yes No No Double
Beta 3 No No No Triple

So, we see that we need at least two criteria if we want to classify disagreements
in philosophy, as illustrated in Tables 3.2 and 3.3. If we concentrate on the important
points described in this section, then, whenever we want to understand a philosophical
text in which we find a difference of opinion, these are the questions we must ask
ourselves (they are all P-theoretical meta-questions):
(27) a. how many and which questions are at issue
b. how many and which parties are involved in the discussion
c. which parties are taking a definite stand on which question,
which are expressing doubts about which answer, and which
are expressing doubts about which questions
These questions may be more or less difficult to answer, depending on the nature
and complexity of the text and on the amount of background knowledge we need to
understand it.
The Russell-Copleston debate is relatively easy to analyse thanks to a set of
features the text has. The debate took place not very long ago, the parties in it are
quite well-known, they both speak clearly and reasonably, the text is relatively short
(less than ten thousand words), and it is clearly dialogical, so that we can identify
the parties involved. When one or several of these features are absent, things become
increasingly difficult. Yet, even in the case of such a debate, the reader needs to
be careful when tracking the questions at issue. The debate is nominally about one
question, whether god exists or not. Yet, as soon as the parties start discussing,
a whole lot of other, more or less closely associated questions, are raised (in fact,
close to one hundred). Some of them provoke new differences of opinion, some are
brushed aside, explicitly or implicitly, and each one requires background knowledge,
some of it of considerable sophistication. Again, at some points it is very explicit
which role one party is playing, but roles can become blurred or even shift. And
even though the debate is definitely a dilogue, i.e. a dialogue between two parties,
the discourse of each party gets entangled with other voices and positions, both from
the past and the present.
All these troubles that we shall have occasion to point out in our analysis get only
larger and deeper as soon as the text becomes more remote from our current concerns,
the language in which the parties speak is not English, the text is long, the style
obscure, the manners less forthcoming, strong emotions get the best of the parties,
their allusions and references are unclear. Anybody who knows something about the
3.2 Disagreements in Philosophy 73

myriad texts that adorn the history of philosophy will come up with examples that
offer us huge hurdles against our attempts at understanding. However, the questions
in (27) gives us a valuable guideline in these efforts.

3.3 Philosophical Argumentation in the Ideal Model

Being reasonable seems to imply that, when facing a difference of opinion, one should
try to resolve it by means of reasons. If my interlocutor and I have a difference of
opinion, it does not seem reasonable at all to leave things like that; it does not seem
reasonable to just say that anybody is entitled to her own opinion, and that’s that. It
may be expedient, of course, but not reasonable. It is true that my interlocutor has
her own reasons and I have mine; but reason is all about reasoning with the other,
comparing and contrasting her reasons with my reasons, in order to find whether
reason is on her side or mine. Hence the idea of a reasonable discussion, i.e. a
discussion in which reasons are compared and contrasted. Even though human beings
are not perfectly reasonable and oftentimes act against reason, it is also true that
sometimes we do pay attention to reasons, that we do use reasons in order to resolve
our differences of opinion. We are all imperfectly reasonable, but we are reasonable
none the less.
Now, the situation in philosophy is special if not indeed paradoxical: philosoph-
ical discussions never seem to end; agreement among philosophers seldom occurs;
and, when it does, it tends to be short-lived and extend itself to just a few issues.
To acknowledge this undeniable fact should, however, not mean to stop trying to
identify a reasonable procedure for discussion in philosophy. One such procedure
is the ideal model of critical discussion proposed by pragma-dialectics. Since this
model is predicated upon the idea of the resolution of differences of opinion on the
merits, some readers may find that its applicability to philosophy is, to say the least,
problematic (Leal 2019). Of course, the proof of the pudding is in the eating; and the
core of this book, Chaps. 4–8, where the model is applied to the Russell-Copleston
debate, will attempt such a proof. Nonetheless, it may be useful to say something
beforehand about the ideal model.
Perhaps the most important thing to say concerns the Opening Stage. The
operations of this stage (see Chap. 2, Sect. 2.3) look for agreement on four questions:
(28) a. whether the discussants shall try to resolve their differ-
ence of opinion by means of arguments
b. whether they can agree on certain substantive starting
points (propositions)
c. whether they can agree on certain procedural starting
points (rules of discussion)
d. whether they can agree on the assignment of the roles of
protagonist and antagonist
74 3 Description of the Method Followed

All discussions among human beings have some degree of institutionalization,

which includes some pre-established agreements on these four questions. From time
to time, though, once the difference of opinion is expressed and recognized by the
parties (this is what constitutes the Confrontation Stage), they may need to go back
and restate some of those agreements or arrange for fresh ones. That is what the
Opening Stage is about. Because of this character, the Opening Stage has the peculiar
feature that it can either take place immediately after the Confrontation Stage or later
on, as a kind of feedback loop. Thus, after the judge or a specially appointed clerk
announces the case to be tried in a criminal court, a few starting points may be either
argued for or, to speed the process, stipulated by the attorneys. However, it may
happen, during the interrogation of witnesses, that certain rules have to be brought to
bear even though they are part of previous agreements. Thus, an attorney may object
to a certain line of questioning by another attorney, so that the presiding judge has to
rule on that by recourse to a recognized code. Again, the judge may ask the spectators
to stop making noise, and the bailiff may enforce that. The bailiff may take some
people away from court—unruly witnesses, defendants, spectators—in obedience to
the judge’s injunction. Those brief or long interruptions of the proceedings—i.e. the
main argumentation on the question whether the defendant is guilty or not—are all
returns to the Opening Stage.
Agreements on (28) are analytically previous and necessary to any disagreement
on a given question, for nobody discusses about anything unless there is a measure
of agreement on how to conduct the discussion, which questions are not at issue,
and so on. This is a fact of human communication. Let us imagine that two philoso-
phers (or groups of philosophers) would agree on nothing at all, then they could not
communicate with each other; but, in such a case, no difference of opinion could
take place, either. To take a simple example, if a philosopher would express her
opinions in Mandarin to a German philosopher innocent of that language, barring an
interpreter, there is no way in which a difference of opinion would emerge, let alone
be recognized by the parties. The existence of a shared language, and that means the
mutual knowledge of a lexicon, a syntax, and a phonology, is an essential part of the
agreements that necessarily precede any critical discussion.
Now, the question arises as to what the points of agreement are in a
philosophical discussion. Before we attempt to answer that question, we should
remember something. In contradistinction to any other discussion that can take place
among human beings, philosophical discussions have the peculiar feature that, in
them, anything can be questioned. So, we should get used to the idea that, no matter
how firm we think our agreements to be, in principle any one of them may be put
in doubt when two philosophers engage in a critical discussion. In fact, part of the
shape of philosophical discussions, their twists and turns, have to do with this fact
of universal questionability.
The pre-agreements of human discussions are in general tacit. This is for the
simple reason that nobody knows how many of team are there or which
they are. In general, a tacit agreement gets verbalized because one of the parties
becomes aware of a certain obstacle to the reasonable continuation of the discussion.
To give a humdrum example, somebody starts to speak more loudly than usual, and
3.3 Philosophical Argumentation in the Ideal Model 75

then another participant suggests that all that yelling does is to produce heat and
make emotions run high. What was an implicit norm, that had been followed up to
that point, shows its importance in consequence of its unwitting violation by one of
the parties. This state of affairs can be observed in philosophy as well.
Another important feature of pre-agreements is that many of them are extremely
vague. Take again loudness. People can discuss within an acceptable range, not
too loud and not too low; but that is a range, not an exact magnitude. Besides, the
circumstances of a conversation can vary, so that speaking more or less loudly can
be more appropriate than keeping the voice within the initial range. Equally vague
and imprecise are in general the meanings of the words we use. For most practical
purposes, it does no matter much whether the meaning of a word is opaque or fuzzy, or
that words have a variety of meanings; but there are situations in which this becomes
so important that we actually need to talk about the meaning of words, and even to
make what we mean clearer or more precise. In philosophy this is notoriously the
case more often than in ordinary life.
The tacit and vague character of the agreements that precede any discussion
suggests a simile from a current craze. Let us think of a ‘cloud of agreement’ that
somehow floats above a discussion. Such a cloud of agreement makes possible to
have a discussion in a similar way that all our electronic devices—cell phones, tablets,
PCs and laptops—work by being constantly connected to the information cloud. This
happens all the time without our being aware of it; and, when we have a particular
explicit need (say, of knowing where we are geo-located), we access the relevant
information so as to review our situation in order to continue doing what we are
doing: we can thus surf in the cloud, download the agreements, examine them, even
modify them so as to get our critical discussion unstuck.
This can happen any time and from any stage we are in. Perhaps the main difference
between a philosophical discussion and a discussion of another kind is that stopping
the proceedings in order to access the cloud and review some agreement is vastly
more frequent in philosophy than in any other argumentative activity. This is so
because, as we said before, there is in philosophy no agreement that cannot be put
in doubt, even if we initially took it for granted, established, and unmovable. Of
course, some radical doubts have to be reasonably motivated by the experience of
the discussion.
We can divide the agreements ‘uploaded’ into the cloud and revisable at any point
into four classes:
(29) a. Agreements about the meaning of what is said.
b. Agreements on the questions that are not the subject of dispute, or those
that should not be the subject of dispute.
c. Agreements about the rules, norms, codes, and procedures that it is deemed
appropriate to follow in the discussion.
d. Agreements about the duties, right, obligations, and commitments of the
parties in dispute.
These four classes of downloadable and reviewable agreements are conceptually
independent of each other, although there is flow among them. Questions of meaning
76 3 Description of the Method Followed

sometimes, albeit not always, occur in the context of separating and ordering the
questions that are or are not to be discussed. In a similar manner, the separation and
ordering of questions can be part of a certain procedure to be followed in order to be
reasonable. Finally, the assignation and recognition of duties and responsibilities of
the parties can be a normative aspect of disputes. So, the distinction of classes has
an analytic purpose which does not preclude occasional overlaps.
It would be useful to try and illustrate each one of these classes of agreements.
In class (29a) we find, on the one hand, agreements concerning the terms, i.e. of
words and phrases which philosophers use to denote or refer to the objects—concrete
or abstract, real or ideal—they want to talk about. Discussions among philosophers
often start by questioning the meaning or reference of such terms. Three cases may
be distinguished: philosophers question either the ordinary use of ordinary terms, or
the non-ordinary use of ordinary terms, or the use of non-ordinary, invented terms.
In all those cases we shall find that a difference of opinion is associated to words or
phrases whose meaning another party finds not quite, or even not at all, intelligible.
This sometimes leads to say that certain disputes in philosophy are purely verbal or
terminological, so that they there is no real difference of opinion. This suggests that
all that is needed to increase agreement or make it more lasting is for philosophers
to consent to give up on their favourite words and phrases (Rorty 1984; Chalmers
2011). However, this suggestion should be taken with prudence for two reasons:
because most philosophers protest that their disagreement with another philosopher
is not verbal, and because it is quite hard to find undisputable cases of verbal dispute
in philosophy.
Be that as it may, questions of meaning are not exhausted by terms, i.e. by deno-
tative words and phrases. They rather extend to expressions and ways of talking that
are not used to denote. John Locke, for instance, was bent on refuting Descartes’s
thesis that some of our ideas are innate. He takes as an instance the principle of
non-contradiction, and he argues that, if an idea is innate, then already young chil-
dren would have it; but young children are not even capable of understanding such a
principle, let alone to assent to it. A defender of innates ideas would then reply, says
Locke, that such children would know and accept the principle of non-contradiction
‘when they come to the use of reason’. This is extremely suspect to Locke:
Doubtful Expressions, that have scarce any Signification, go for clear Reasons, to those, who
being pre-possessed, take not the Pains to examine even what they themselves say. For, to
apply this Answer with any tolerable Sense to our present Purpose, it must signify one of
these two Things; either, That as soon as Men come to the Use of Reason, these supposed
native Inscriptions come to be known, and observed by them: Or else, That the Use and
Exercise of Mens Reason, assists them in the Discovery of these Principles, and certainly
makes them known to them. (Locke 1689, I, ii, 7)

Locke objects to the clause ‘when the child comes to the use of reason’ as impre-
cise: if we try to make it more precise, we discover two possible senses. According to
the first sense, it is asserted that there is an age in which it occurs that the child grasps
the principle of non-contradiction. According to the second sense, it is asserted that
the use of reason is what helps the child to discover the principle of non-contradiction.
What then follows in Locke’s Essay is a long argument purported to prove that in
3.3 Philosophical Argumentation in the Ideal Model 77

none of these two senses can the Cartesian thesis of innate ideas be saved. This is
how making our ideas clearer and more precise is used by philosophers to attack the
positions of their adversaries. In this and other cases in which a philosopher ques-
tions the sense of certain expressions, what we have is an argumentative revision of
certain mostly tacit pre-agreements that were ‘in the cloud’.
Let us go on to class (29b). When Meno asks Socrates in which of various
ways virtue is acquired, Socrates replies that that question cannot be answered
unless another one is asked and answered first, namely what virtue is. What
happens in this exchange is that, although Meno’s original question seems intelligible
and in principle answerable, according to some tacit pre-agreements between the two
discussants, Socrates thinks it is necessary to enter the cloud in order to examine the
network of question to which Meno’s question belongs, for only in this way it can be
established what the right ordering of questions is. One can see from this
example that class (29b) is not a watertight compartment, utterly separated from class
(29a), for the ordering of questions can be related to the meaning of expressions, in
this case the meaning of ‘virtue’ (or rather ¢ρετή, in the language Socrates shared
with Meno).
A different example of (29b) comes from the British philosopher Gilbert Ryle in
the following passage:
It has for a long time been taken for an indisputable axiom that the Mind is in some important
sense tripartite, that is, that there are just three ultimate classes of mental processes. The
Mind or Soul, we are often told, has three parts, namely, Thought, Feeling and Will; or, more
solemnly, the Mind or Soul functions in three irreducibly different modes, the Cognitive
mode, the Emotional mode and the Conative mode. This traditional dogma is not only not
self-evident, it is such a welter of confusions and false inferences that it is best to give up
any attempt to re-fashion it. It should be treated as one of the curios of theory.
The main object of this chapter is not, however, to discuss the whole trinitarian theory of
mind but to discuss, and discuss destructively, one of its ingredients. I hope to refute the
doctrine that there exists a Faculty, immaterial Organ, or Ministry, corresponding to the
theory’s description of the ‘Will’ and, accordingly, that there occur processes, or operations,
corresponding to what it describes as ‘volitions’. I must however make it clear from the
start that this refutation will not invalidate the distinctions which we all quite properly
draw between voluntary and involuntary actions and between strong-willed and weak-willed
persons. It will, on the contrary, make clearer what is meant by ‘voluntary’ and ‘involuntary’,
by ‘strong-willed’ and ‘weak-willed’, by emancipating these ideas from bondage to an absurd
hypothesis. (Ryle 1949, Chap. III, Sect. 2)

In this patently polemic passage, Ryle announces that he has a multiple difference
of opinion covering the whole ‘trinitarian’ theory of mind, for this theory presum-
ably contains answers to a good many different questions. However, Ryle says he is
going to treat only one question within this multiple disagreemens, namely whether
there are processes or operations of the will or not. All other ques-
tions which, in the ‘trinitarian’ theory, are associated with this one must be brushed
aside, as if there was no disagreement on those. Ryle himself takes good care to say
that there is, but also that it is not appropriate to discuss them at this point.
Class (29c), in its turn, contains some of the most crucial disagreements that there
are or may in principle be among philosophers, those which have to do with the
78 3 Description of the Method Followed

rules of discussion, among them those we associate with the word ‘logic’. Among
analytic philosophers, for instance, a consensus existed for quite some time, and it
may even exist today, namely that classical logic (elementary formal logic, consisting
of two-valued propositional and quantificational logic with identity) is a necessary
and sufficient instrument for the solution of intricate philosophical problems. This
consensus has broken under the pressure of so-called ‘non-classical’ logics, which
either increase or decrease the scope and power of classical logic. One example of
non-classical logic is free logic (or existence-free logic) in which it is not assumed, as
in classical logic, that each singular term denotes, i.e. that there is a referent for that
term. One of the pioneers of free logic, the Italian philosopher Ermanno Bencivenga,
tells us:
There are contexts in which some people use expressions as singular terms without assuming
that they denote anything, or maybe even in the process of wondering whether they denote or
not. For example, an attempt by a person to prove that God exists—or that ‘God’ denotes—
might be conceived as a case in point. Whether these people are right or not, a logic allowing
for non-denoting singular terms would also allow for a more direct and faithful representation
(and evaluation) of their reasoning in those contexts. So this logic would be an instrument
of wider and simpler applicability than classical logic, and would not prejudge important
issues which it is inappropriate for logic to decide. (Bencivenga 2002, p. 151)

This proposed revision of a longstanding agreement is as important as the one

that replaced ‘traditional logic’ by classical logic. As we said before, in philosophy
everything is revisable, even logic, whose solidity would seem beyond doubt.
By the way, there are norms in (29c) which lie beyond logic in the strict sense of
the word. We can illustrate those with a discussion between two philosophers about
a third one, in this case Searle’s attack on Derrida (1971), a paper in which the latter
interprets and criticizes Austin’s theory of speech acts:
It would be a mistake, I think, to regard Derrida’s discussion of Austin as a confrontation
between two prominent philosophical traditions. This is not so much because Derrida has
failed to discuss the central theses in Austin’s theory of language, but rather because he has
misunderstood and misstated Austin’s position at several crucial points, as I shall attempt to
show, and thus the confrontation never quite takes place. (Searle 1977, p. 198)

This passage reminds readers of a norm that could be formulated as follows: if you
are going to discuss an author, you should better understand that author’s position
correctly. It is that norm that Searle retrieves from the cloud and shows Derrida the
way a referee shows a red card in soccer.
As was only to be expected, Derrida does not accept the red card. Instead he
returns the compliment:
I shall place in the margin… the following question. I address it to Searle. But where is he?
Do I know him? He may never even read this question. If he does, it will be after many
others, myself included, and perhaps without understanding it. Perhaps he will understand it
only in part and without judging it to be quite serious. (Derrida 1977, English transl. 1988,
p. 29)

The question Derrida says he will place in the margin is:

3.3 Philosophical Argumentation in the Ideal Model 79

What is the nature of the debate that seems to begin here? Where, here? Here? Is it a
debate? Does it take place? Has it begun already? When? Ever since Plato, whispers the
prompter promptly from the wings, and the actor repeats, ever since Plato. Is it still going
on? Is it finished? Does it pertain to philosophy; to serious philosophy? Does it pertain to
literature? The theatre? Morals? Politics? Psychoanalysis? Fiction? If it takes place, what
is its place? And these utterances—are they “serious” or not? “Literal” or not? “Fictional”
or not? “Citational” or not? “Used” or “mentioned”? “Standard” or not? “Void” or not? All
these words are, I assure you and you can verify it yourselves, “citations” of Searle. (Derrida
1988, pp. 29–30)

As anyone can see, this is not one but many questions, all of which will reveal
themselves to any attentive reader to be Derrida’s own methodological red cards. In
other words, other pre-agreements about norms and rules of discussion have been
retrieved from the cloud and opposed to Searle’s norm. We now have a disagreement
about norms which may be even more difficult to resolve than the original substantive
disagreement. We are bringing up this example only because it is a limiting case.
The parties in this dispute seem almost to be playing two different games.
This leads us to class (29d), which is about discussants’ duties and responsibili-
ties. Observe that in the previous sample, Searle wants to impose certain obligations
on his interlocutor and argue that he is not fulfilling them; and Derrida does exactly
the same. The example is extreme, because the obligations dictated by each party
correspond to different sets of agreements, which stem from different traditions. As
the commonplace has it: in this corner ‘analytic’ philosophy, in that corner ‘Conti-
nental’ philosophy. A series of labels and libels would then be used to describe each
opponent. Note also that the abyss separating the two traditions has, to a consider-
able extent, to do with the use of terms and expressions. This confirms what we said
before: the four classes do not lie in watertight compartments.
A naive way out would be to invoke the burden of proof. This is really a dead end,
because any philosopher worth her salt will ask to be allowed to present her case.
Searle and Derrida are as pugnacious as one could wish. They never tire of presenting
arguments. The question, however, is rather what counts as an argument. If
one party presents an argument, the other dismisses it as a web of fallacies if not indeed
incoherent trash. In terms of the classes (29), one could say that any representative
of the two camps thinks that those in the other camp (a) utter meaningless words,
(b) do not respect the ordering of questions, (c) do not follow the norms and rules of
discussion, and (d) do not assume their duties and responsibilities as discussants.
We shall close this section by referring to an example that is highly relevant
to the Searle-Derrida dispute. Plato’s dialogue Protagoras opposes Protagoras and
Socrates, two of the most famous characters in the cultural and intellectual life of
ancient Greece. The main hindrance of the debate has to do with the different methods
of argumentation that each has developed to the point of mastery: the sophistical art
of long, beautiful speeches versus the philosophical art of dialogue by question
and answer. The trouble starts as a case of (29c), namely norms of discussion; but
it quickly turns into a question of duties and responsibilities thanks to Socrates
pointing out two things. On the one hand, he, Socrates, has a bad memory, so he is
terribly handicapped in a conversation with Protagoras because, whenever he asks a
80 3 Description of the Method Followed

question, Protagoras gives a long speech as an answer. On the other hand, Protagoras
has explicitly claimed to be a master of all speech, long as well as short, so he can
handle swift and brief question-and-answer dialogue as much as long speeches. This
shifts the trouble to (29d), for it is now a matter of leveling the ground of the dispute
by choosing dialogue as the procedure that is fair to both parties. According to the
Socratic method, discussion takes place by assigning two roles to the parties, asking
questions and answering them. Readers will observe that we have now moved to
class (29b).
Both the role of questioner and the role of answerer require great skill. We
know how Aristotle tried to codify the Socratic game of question and answer. Since
Protagoras is well aware that he is not a match for Socrates in the dialogue game,
he tries to withdraw from it; but, after some back and forth, and the intervention of
the company of friends that is witnessing and hugely enjoying the encounter of the
two masters, Protagoras agrees to answer Socrates’s questions. When things are not
going so well for the sophist and he starts grumbling, the philosopher offers him to
change roles: ‘If Protagoras does not want to answer, let him ask questions and I
shall answer’ (Plato, Protagoras, 338C–D; compare Gorgias 462 A–B).
Summing it all up, when philosophers engage in a discussion, everything is up
for grabs. And all pre-agreements may be ‘downloaded’ and put on the table for
examination and revision. This constant revision ends up in philosophers going back
again and again to the Opening Stage, as we shall see in the analysis of the debate
between Copleston and Russell. We shall also have occasion to observe two other
features highlighted by our analysis.
One is that argumentation, more precisely the exchange of arguments, not only
takes place in the Argumentation Stage stricto sensu but also in the other stages
as well. Philosophers argue all the way and all the time. We have seen above some
examples of philosophical argumentation in the Opening Stage, but we can find them
even in the Confrontation and the Concluding Stage.
The other feature is that, even though a given discussion is triggered by a certain
question, many other questions arise and quite often they reveal many other disagree-
ments, as we have shown in Sects. 3.1 and 3.2. In the Russell-Copleston debate,
starting from the single question whether god exists or not, about a hundred
other questions come to light. Some of them are quickly abandoned by one or the
two philosophers without further ado, others are dismissed as beyond the pale, and
still other questions are taken up and discussed, even if in most cases an intractable
disagreement emerges. As the discussion unfolds, the careful reader can see how the
differences of opinion are not tied up to one question but have a certain coherence
that slowly adds up: something like a giant wall separating the worldviews of the
parties can be discerned. The important thing in philosophy is thus to keep a score
of the questions.
3.4 The Russell-Copleston Debate in a Nutshell 81

3.4 The Russell-Copleston Debate in a Nutshell

The Russell-Copleston debate is more or less neatly divided into five segments (where
Segment III more or less suddenly blends into Segment IV), as represented in Table
3.4.
The question that presides the whole debate is
(30) whether god exists or not

And yet, what (30) represents is not really one question, but at least three, given that
the background against which the discussants argue depends on the kind of argument
that the main Protagonist, Copleston, is trying to develop. Before going into that, it
is useful to insist that Copleston starts all his arguments from a factual premise. This
is important, because Russell’s strategy sometimes is (a) to try to turn Copleston’s
factual premise into a conceptual one, and sometimes (b) to throw the factual premise
into doubt, by means of an alternative description of the facts involved.
In Segment II of the debate, Copleston argues from the factual existence of contin-
gent beings, and the God he is talking about is the creator of the universe in the sense
of the first or ultimate cause of the existence of everything else, prima causa. So, the
more nuanced question for Segment II is rather
(31) whether there is a first cause of everything or not

Notice how (31) is subtly different from (30). But this is only the beginning, for in
Segment III Copleston argues that the experiences of the great mystics are a puzzle
that needs explanation, and that the best explanation for those experiences is God.
In Segment III, therefore, the question is, again, subtly different:
(32) whether the existence of god is the best explanation for the
mystical experiences or not

If God’s existence is indeed the best explanation for the fact of mystical experiences,
then God exists. Now, as this argument unfolds, it transpires that God is something
like—is in effect redefined as—the first or ultimate ground of goodness or of value,
so the actual question of Segment III would seem to be rather

Table 3.4 The five segments of the Russell-Copleston debate

Segment Turns Description
I 1–10 Start of the debate
II 11–70 Discussion of Copleston’s metaphysical argument or the argument from
contingent beings
III 71–91 Discussion of Copleston’s argument from religious, specifically mystical
experience
IV 92–136 Discussion of Copleston’s argument from moral experience
V 137–138 Summing-up of the arguments and counterarguments
82 3 Description of the Method Followed

(33) whether the existence of an ultimate ground of goodness is

the best explanation for the mystical experiences or not

Once this becomes clear, the path is cleared for Segment IV, which again starts with
a fact, namely moral experience, which also cries for explanation. This gives us the
question
(34) whether the existence of an ultimate ground of value is the
best explanation for the moral experience or not

Now, at some point Copleston introduces a Kantian-like distinction between the

content or matter of the moral law and its form, and he makes clear that it is the latter
which he wants to explain through God, so that the actual question of Segment IV
turns out to be
(35) whether the existence of an ultimate ground of goodness is
the best explanation for the form of the moral law or not

It is through the back and forth of argument and counterargument that the questions
at issue become thus specified. This is part of what philosophical argumentation is—
the clarification and redefinition of the question or questions that are the object of
dispute. If we should rest content with saying that Russell and Copleston are debating
the existence of God, as is commonly done, we would in effect be distorting and even
falsifying what they were actually up to. Therefore, I propose to reframe the fivefold
division of the debate as in Table 3.5
The questions that serve to describe Segments II, III, and IV we shall call the
presiding questions of that part of the debate. The three presiding questions are thus
the real coin behind the somewhat vague question (30) whether god exists or
not, but this is not always perfectly clear.
Now, each presiding question, as it gets pushed and pulled by the discussants,
raises new questions. Some of these are tacitly or explicitly abandoned by the discus-
sants, but some others become the target of a new discussion. When this happens,

Table 3.5 The main questions debated by Russell and Copleston

Segment Turns Description
I 1–10 Start of the debate
II 11–70 Discussion of the question whether there is a first cause of
everything or not
III 71–91 Discussion of the question whether the existence of an ultimate
ground of value is the best explanation for the mystical
experiences or not
IV 92–136 Discussion of the question whether the existence of an ultimate
ground of value is the best explanation for the form of
the moral law or not
V 137–138 Summing-up of the arguments and counterarguments
3.4 The Russell-Copleston Debate in a Nutshell 83

we can see that an overlap of questions takes place. Take for instance the following
two questions:
(36) whether everything has a cause or not
(37) whether it is (or is not) a presupposition of physics that
everything has a cause

These two questions arise at some point during the discussion of question (31),
namely whether there is a first cause of everything or not, yet (31)
is by no means identical to either (36) or (37). Nonetheless, these questions overlap
in the sense that, whilst philosophers may be focusing on one of the newly arisen
questions, they are perfectly aware that their answers to those questions have an
impact on the answer to the presiding question. So, they will in fact be discussing
two or more questions at the same time, one of them directly, whereas the others,
including the presiding question, are, as it were, not directly but obliquely answered.
This is what normally happens during a philosophical debate: one question raises
other questions, and, whilst philosophers may at times agree on the correct answer to
one or several of the newly arisen questions, this agreement will never cover all the
answers to the questions in dispute. In pragma-dialectical terms, all philosophical
disagreements are multiple (Leal 2019). We are going to observe this multiplicity
again and again, as we proceed with our analysis.
How many questions are raised in a philosophical debate? How many differences
of opinion? There is only way to answer those questions, which is by doing a careful
analysis of one such debate. The following analysis of the Russell-Copleston debate
identifies almost one hundred questions. Table 3.6 is intended to reproduce the way
in which questions are posed in the debate between Copleston and Russell, in which
they sometimes assimilate questions that seem different from each other. The location
number corresponds to the subdivisions contained in the analysis of Chaps. 4–8.
As can be imagined, not all of these questions are taken up or even recognized
explicitly, so that there is some margin of error as to the differences of opinion
between Russell and Copleston that they engender. But it will be fairly clear to
the reader of the following analysis that it is not the case that our discussants only
disagree about the question whether god exists or not. In fact, at some points
it becomes clear that their disagreements are much deeper, so deep in fact as to make
them look as though they are just talking at cross-purposes. Nonetheless, the points
of contention are rarely, if ever, purely verbal.
One last remark before going over to the analysis of the debate. The analysis takes
the form of a running commentary paragraph by paragraph, sentence by sentence,
or even clause by clause, according to what the text requires. To that extent, what
readers will find is not utterly different from the long tradition of commentary in
the history of Western philosophy, starting with the Neo-Platonic commentaries of
important texts (say, Plato’s dialogues, Aristotle’s treatises, or Euclid’s elements)
and continuing all the way down to the most recent philological commentaries of the
Greek and Roman philosophers. This means that I shall use any available information
84 3 Description of the Method Followed

Table 3.6 All the questions raised in the Russell-Copleston debate

No. Question Location
Q1 whether god exists or not 1.1
Q2 what the term ‘god’ means 1.2
Q3 whethersomething with the attributes ‘supreme’, 2.1
‘personal’, ‘transcendent’, and ‘creator of the
world’ exists or not
Q4 whether the non- existence of god can be proved 3.2
or not
Q5 whether the question of god’s existence is 5.1
important or not
Q6 whether the non- existence of god leads to 5.2
human self- assignment of purposes or not
Q7 whether the non- existence of god leads to 5.3
totalitarian regimes or not
Q8 whether the non- existence of god implies the 7.1
non- existence of absolute values or not
Q9 whether there is an absolute good or not 7.2
Q10 whether denying god’s existence is logically 8.1
equivalent to upholding the relativity of values
or not
Q11 what the meaning of the phrase ‘necessary being’ 12.2
is
Q12 whether the phrase ‘necessary being’ has a 12.2
meaning or not
Q13 whether mathematical logic may be used to 12.3
tackle philosophical questions such as the proofs
for god’s existence
Q14 what kind of propositions can be considered 12.4
necessary
Q15 what the status of those propositions is which 12.4
kant deemed synthetic a priori
Q16 whether the copleston conclusion (‘there is a 12.4
being which cannot not exist’) is either analytic
or else it is untrue (or meaningless)
Q17 whether there are necessary non- analytic 12.4
propositions
Q18 whether it is self- contradictory to deny that 12.5
god exists
Q19 whether leibniz’s distinction between truths of 12.6
reason and truths of fact is acceptable or not
Q20 hoe to reconcile human free will with god’s 13.3
omniscience
(continued)
3.4 The Russell-Copleston Debate in a Nutshell 85

Table 3.6 (continued)

No. Question Location
Q21 what the phrase ‘necessary proposition’ can mean 14.1
if not ‘analytic’
Q22 whether the copleston conditional (‘if there is a 15.1
contingent being, then there is a necessary
being’) is a necessary proposition
Q23 what the ‘logic’ might be within which phrases 16.3
like ‘necessary being’ and ‘contingent being’
have (or lack) a meaning
Q24 what (kind of logic) russell’s logic is 19.1
Q25 whether it is appropriate to base metaphysics on 19.6
logic or not
Q26 whether it is appropriate to identify metaphysics 19.6
with logic or not
Q27 whether there is or is not a necessary being 19.13
Q28 what the meaning of ‘existence’ is 20.3
Q29 whether existence is a predicate or not 20.3
Q30 whether the theory of descriptions can be used 20.3
to analyse and refute the ontological proof of
god’s existence or not
Q31 whether thr proposition ‘the cause of the world 21.3
exists’ has or lacks meaning
Q32 whether the cause of the world exists or not 21.5
Q33 whether the metaphysical argument at bottom 26.1
hides the ontological argument or not
Q34 whether ‘sufficient reason’ means the same as 32.1
‘cause’ or not
Q35 when, or under what circumstances, an 34.1
explanation is to be called ‘adequate’
Q36 whether the rubbing of a match against the 34.1
sandpaper of the matchbox is to be called an
adequate explanation of the making of a flame
with the match
Q37 whether complete explanations cannot be 39.1
achieved or rather whether they should not
even be attempted
Q38 whether the very question of (an explanation 39.1, 45.1
for) the existence of the universe is pointless /
whether it makes sense to raise the question
whether the existence of everything can be
causally explained
Q39 whether a question in which a handy but 41.1
meaningless word appear can be meaningful
(continued)
86 3 Description of the Method Followed

Table 3.6 (continued)

No. Question Location
Q40 whether (the existence of) the universe is 41.4
unintelligible
Q41 whether (the) existence (of anybody or anything) 41.4, 43.5
is at all intelligible / whether existence can be
at all explained
Q42 whether the world had a beginning or is eternal 43.2
Q43 give me a reason to think there is a cause of 46.1
the existence of all things
Q44 whether the universe has a cause or not / why 47.3, 51.1
there is something (anything at all) rather than
nothing
Q45 whether the metaphysical standpoint that the 55.2
world has no explanation can be derived from
pure logic or not
Q46 whether certain events in the world are 56.2
uncaused or not
Q47 whether the assertion that quantum transitions 57.1
in atoms are not caused is a temporary inference
which will be later revoked
Q48 whether it is possible to conceive an even in the 58.1
world that is uncaused
Q49 whether physical research can proceed without 59.2
searching for causes
Q50 whether every event has a cause 59.2
Q51 whether metaphysics can proceed without 59.5
searching for causes
Q52 what kind of argument the religious and moral 71.2
arguments are / whether the religious and moral
arguments have the same status as the
metaphysical argument or not
Q53 whether the best explanation of religious 71.5
(mystical) experience is that god exists
Q54 whether the mystical experience causally affects 77.4
the mystical transformation
Q55 whether the records of people hearing satanic 80.1
voices are similar to mystics asserting god
Q56 whether there are records of people hearing 80.1
satanic voices
Q57 whether we can compare mystical experiences to 80.3, 86.5
other phenomena, such as a hearing voices / such
as the reading of real or fictional account by
boys
(continued)
3.4 The Russell-Copleston Debate in a Nutshell 87

Table 3.6 (continued)

No. Question Location
Q58 whether a belief’s good moral effect upon a 83.1
person constitutes evidence for that belief’s
truth or not
Q59 whether the good moral effect exerted on a 83.1
believer by parts of a belief yields a presumption
of truth of those belief parts / whether the
extraordinary transformation following a
mystical experience is a presumption in favour of
the truth of the belief in god’s existence
Q60 whether the mystical transforation is evidence 83.2
of veracity and sanity
Q61 whether the lives of the mystical are what 84.5
copleston says they are
Q62 what sort of thing there is in an experience that 87.2
has a good influence on the experiencer
Q63 what the relation is between the existence of 89.2
god and the experience of his goodness as given
in the mystical experience
Q64 whether god is the sum total or the system of 92.1
what is good
Q65 whether god is the good, or the origin and 92.1
ground of everything that is good
Q66 whether the good is a matter of human feeling / 96.3, 112.4
whether human feeling is the origin and ground
of all things good / whether the good is a
matter of human feeling emerging from the
effects of actions / whether human feeling, as it
emerges from the effects of actions, is the origin
and ground of all things good
Q67 how people are justified in distinguishing good 97.1
and bad
Q68 what kind of distinction is that between good 97.1
and bad
Q69 what the faculty is by means of which human 99.1
beings can distinguish between good and bad
Q70 whether the behaviour of the commandant of 103.4
belsen is good for hitler and evil for russell
(continued)

to throw light on the meaning of what is said in the debate. However, my analysis
has four features that are unusual in commentaries.
The first feature does not need much comment after what is said in this chapter.
It is simply that I try to identify all questions explicitly asked by the discussants
88 3 Description of the Method Followed

Table 3.6 (continued)

No. Question Location
Q71 whether the goodness or badness of the 103.4, 107.2, 111.1
behaviour of the commandant of belsen only
depends on the feelings of people, given thet
different people have different feelings about it
/ whether its goodness or badness only depends
on the subjective criterion of feeling or also on
an objective criterion / whether it does so, as
compared to the behaviour of sir stafford cripps
and archbishop william temple
Q72 whether feelings can be mistaken or not 104.2
Q73 whether we can be right about our own feelings 105.1
or not
Q74 what sort of thing can or should receive the 106.2
predicate of goodness or badness
Q75 why we intellectually condemn the colour- blind 108.2
man
Q76 whether we intellectually condemn the 108.2
colour- blind man because he’s in the minority
Q77 in which way can argue with people who do not 112.3
share one’s feelings of good and bad
Q78 what ground there is to say that not everything 116.2
is yellow or that everything is
Q79 what ground there is to say that something is 116.3
bad or that it is good
Q80 whether the concept of moral obligation makes 117.1
sense or not
Q81 what the point is, in moral matters, to 118.1
distinguuish between theory and practice
Q82 whether ‘ought’ has an emotional connotation 119.1
and that is all there is to it
Q83 whether it is a fact that there is near- universal 121.4
consciousness of moral obligation or not
Q84 whether the fact of near- universal 121.4
consciousness of moral obligation can best be
explained by god’s existence
Q85 how the ground of goodness could possibly also 121.4
be the creator of the world
Q86 whether the moral law is absolute or not 121.6
(continued)
3.4 The Russell-Copleston Debate in a Nutshell 89

Table 3.6 (continued)

No. Question Location
Q87 whether the fact that one has a feeling about 124.1, 130.1
‘ought’ can best be explained by what parents
and nurses tell us when we are children /
whether the meaning of ‘ought’ is just ‘imagined
to be disapproved by someone’ (parents, nurses,
god)
Q88 whether the idea of the ought can be conveyed 125.2
in any other way or only by itself
Q89 whether saying that the idea of the ought 126.3
cannot be conveyed in other terms than itself
must necessarily assume that god is the
conveyor
Q90 whether there is a faculty of reason by which 127.3
we criticize the mores we have been taught
Q91 whether what we call moral obligation is 132.1
merely an effect of general operant conditioning
of humans on humans
Q92 whether russell considers kramer’s conduct 133.4
morally reprehensible
Q93 whether russell would ever act like kramer 133.4
Q92+93 whether russell would say that acting like the 135.2
nazis is wrong in itself, independently of any
utilitarian consideration
Q94 whether russell would act like kramer if he 133.4
would ever come to adopt a utilitarian view in
favour of nazi policies
Q95 whether russell’s position- - - a mixture of 133.4
emotivism, utilitarianism, and behaviourism with
some undertones of moral relativism- - - is at all
coherent and personally sustainable
Q96 how to use logic in philosophy 138.3.2
Q97 whether logic has to be used in philosophy 138.3.2
Q98 whether one is justified in making a fuss about 138.3.3
new logical discoveries

or less explicitly raised by what they say in the debate.10 Many texts in the tradi-
tion of commentaries do that occasionally, but I have attempted to do it fully and
systematically, even relentlessly. This follows directly from the proposal set forth at
the beginning of this chapter: standard pragma-dialectics is predicated upon the fact
that no discussion can take place if no question has arisen. If one follows through
with this fundamental insight, then philosophical argumentation will be seen as a

10Some readers may think that I have overdone my remit by identifying as many as ninety-eight
questions, but my own fear is rather that I have almost certainly missed at least a few.
90 3 Description of the Method Followed

proliferation of questions. Only in this way can the fact really be appreciated that the
disagreements of philosophers rarely if ever concern just one question, and very often
concern a great number of them. Of course, not all questions raised in a discussion
are actually tackled by the parties, but many more are than probably meets the eye.
Besides, the matter of the background information needed to interpret what philoso-
phers say (an absolutely crucial aspect of learned commentaries of philosophical
works) can be much more easily identified if the questions raised in a discussion are
made explicit.11 Having said that, I shall endeavour to add as few ‘missing’ premises
(or conclusions) to the arguments contained in the text as much as possible. To avoid
them completely is, in the nature of things, not always possible, but any added item
will have to be carefully justified on the basis of the text itself.
The second feature is that I try to identify and describe which stage of a critical
discussion any particular utterance belongs to. If the model of critical discussion is
a correct theory of the argumentative process, then there cannot be anything said in
a debate that cannot be pinned down as part of one of the stages and as performing
one of the operations appropriate to those stages (see Chap. 2, Sect. 2.3). This means
that in the present analysis there cannot be any utterance or part of an utterance that
may be characterized as clutter to be eliminated in order to get at what is being done
in the discussion. This has forced my hand whenever I was tempted to brush aside
some remark that seemed to be superfluous. Everything that is in the debate must
find a place in the ideal model, and everything in the ideal model must correspond
to something in the debate. As far as I have read the pragma-dialectical literature, an
endeavour such as the present one has not been undertaken before.
The third feature is that I try to reconstruct all arguments present in the text; and
there are far more of them than one may think at first sight. In contradistinction to the
other two features, I have, for this part of the analysis, tried to stay away as much as
possible from any heavy theoretical commitment. So, I have not used the techniques
of argument mapping (diagrams of the Whately-van den Gelder, Wigmore-Goodwin,
Beardsley-Copi-Walton, Toulmin-Marraud, or any other variety); nor have I used the
techniques of pragma-dialectical numbering; nor indeed any other method. Instead,
I have restricted myself to the most pedestrian form available, the one familiar to any
high-school or college student, namely the premise-conclusion set assumed and taken
for granted in any A-theoretical approach (see Chap. 2, Sect. 2.1). This minimum
procedure, a kind of least common denominator, allows for my argument reconstruc-
tions to be criticized in themselves, without having to appeal to, or throw doubt on,
any particular theory. This aspect of my analysis has the additional advantage, very
important in the context of this book, that readers may easily compare the arguments
set forth in my analysis with the much more sophisticated and theoretically far deeper
reconstructions which Marraud presents in Chap. 10.
The last feature is perhaps the most delicate. It concerns evaluation. I have always
been a bit suspicious about evaluating arguments from the outside, even though some
egregious examples in public life and the media seem urgently to call for it. This

11Any item of background information that is not strictly necessary for the reconstruction of the
arguments or the explication of the argumentative process, will be sent to footnotes.
3.4 The Russell-Copleston Debate in a Nutshell 91

urgency is after all the origin of the tradition of critical thinking, whose record in
improving standards among the population at large is, however, moot, to say the
least. When the arguments are propounded by philosophers, who not only invented
argument evaluation but have cultivated and applied it almost exclusively on each
other’s arguments, it would be presumptuous to try and do it for them. Russell and
Copleston evaluate each other all the time, so an analysis should describe their
procedures rather than replace them by the motley aggregate of proposals contained
within argumentation studies. In the particular case of standard pragma-dialectics,
the application of the rules of critical discussion presents an additional problem. The
first rule establishes the right to express opinions and doubts and the duty to let both
be freely expressed. Given the erotetic perspective adopted here (see Sect. 3.1 of
this chapter), this rule should also include a provision for freedom in asking and
raising questions of any sort. The problem with such a rule is that philosophers
have always been in the habit of rejecting other philosophers’ questions, opinions,
and doubts as meaningless. If we would apply that rule to philosophers, then an
important argumentative move would have to be excluded in principle. This would
distort the nature of philosophical argumentation in an unacceptable manner.

References

Barnes, J. (1969). Aristotle’s theory of demonstration. Phronesis 14 (2), 123–152.

Bencivenga, E. (2002). Free logics. In D. M. Gabbay and F. Guenthner (Eds.), Handbook of
philosophical logic (vol. 5, pp. 147–196). Dordrecht: Springer.
Cassell, A. K. (Ed.) (2004). The ‘Monarchia’ controversy. Washington: The Catholic University of
America Press.
Chalmers, D. J. (2011). Verbal disputes. The Philosophical Review 120 (4), 515–566.
Derrida, J. (1988). Limited inc. Evanston, IL: Northwestern University Press.
Derrida, J. (1971). Signature événement contexte. Paper given at the Congrès International des
Sociétés de Philosophie de Langue Française, Montréal. [Published in Marges de la philosophie
(pp. 367–393). Paris: Éditions de Minuit, 1972.]
Derrida, J. (1977). Limited inc a b c… Glyph 2, 162–254.
Eemeren, F. H. van, and Grootendorst, R. (1983). Speech acts in argumentative discussions.
Dordrecht: Foris.
Eemeren, F. H. van, and Snoeck Henkemans, A. F. (2017). Argumentation: Analysis and evaluation.
Revised edition. London: Routledge.
Ferejohn, M. (1991). The origins of Aristotelian science. New Haven: Yale University Press.
Hintikka, J. (1972). On the ingredients of an Aristotelian science. Noûs 6 (1), 55–69.
Irwin, Terence (1988). Aristotle’s first principles. Oxford: Clarendon Press.
Kahane, H. (1984). Logic and contemporary rhetoric: The use of reason in everyday life. 4th edition.
Belmont, CA: Wardsworth.
Kerbrat-Orecchioni, C. (2004). Introducing polylogue. Journal of Pragmatics 36 (1), 1–24.
Leal, F. (2019). On philosophical argumentation: Towards a pragma-dialectical solution of a puzzle.
Journal of Argumentation in Context 8 (2), 173–194.
Leal, F. (2021). On questions in standard pragma-dialectics. Quadripartita Ratio 6 (11), 41–54.
Lees, R. P. (1960). The grammar of English nominalizations. Bloomington, IN: Indiana University
Press.
92 3 Description of the Method Followed

Lewiński, M., and Aakhus, M. (2014). Argumentative polylogues in a dialectical framework: A

methodological inquiry. Argumentation 28 (2), 161–185.
Locke, J. (1689). An essay concerning humane understanding, in four books. London: The Basset.
Nelson, L. (1904). Die kritische Methode und das Verhältnis der Psychologie zur Philosophie: Ein
Kapitel aus der Methodenlehre. Abhandlungen der Fries’schen Schule (N.F.) 1, 1–88. [English
translation: The critical method and the relation of psychology to philosophy: An essay in method-
ology, in Socratic method and critical philosophy, translated by T. K. Brown III (pp. 105–157).
New Haven: Yale University Press, 1949.]
Nietzsche, F. (1886). Jenseits von Gut und Böse: Vorspiel einer Philosophie der Zukunft. Leipzig:
C. G. Naumann. [English translation: Beyond good and evil: Prelude to a philosophy of the future.
New York: Random House, 1966.]
Ortega y Gasset, J. (1932). Lectures on metaphysics, given at the University of Madrid. [First
publication: Unas lecciones de metafísica [Lessons in Metaphysics]. Madrid: Alianza Editorial,
1966.]
Popper, K. (1976). The myth of the framework. In J. C. Pitt and & M. Pera (Eds.), Rational changes
in science (pp. 35–62). Dordrecht: Reidel.
Porzig, W. (1942). Die Namen für Satzinhalte im Griechischen und im Indogermanischen. [The
names for propositional contents in Greek and Indo-European]. Berlin: Walter de Gruyter.
Romano, C. (2012). America the philosophical. New York. Random House.
Rorty, R. (1984). Deconstruction and circumvention. Critical Inquiry 11 (1), 1–23.
Ryle, G. (1949). The concept of mind. London: Hutchinson.
Sartre, J.-P. (1948). Qu’est-ce que la littérature ? Paris: Gallimard.
Searle, J. (1977). Reiterating the differences: A reply to Derrida. Glyph, núm. 1, pp. 172–197.
Shaw, P. (Ed., Transl.) (1996). Dante: Monarchy. Cambridge, UK: Cambridge University Press.
Williamson, T. (1995). Is knowing a state of mind? Mind 104 (415), 533–565.
Wittgenstein, L. (1921). Logisch-philosophische Abhandlung. [Logico-philosophical treatise.]
Annalen der Naturphilosophie 14, 185–262. [English translation: Tractatus logico-philosophicus.
London: Routledge and Kegan Paul, 1922, 1961.]
Chapter 4
Analysis of Segment I: Start of the Debate

Abstract The whole Russell-Copleston debate contains 138 turns or 69 turns per
debater. Segment I contains the first ten turns. The debate starts in a very orderly
fashion: in pragma-dialectical terms, we find here the confrontation and the opening
stages of a critical discussion. A few elementary starting points, both material and
procedural, are offered by Copleston and either accepted or rejected by Russell.

Segment I, Turn 1 (Copleston) Confrontation Stage / Opening Stage

[1.1] As we are going to discuss the existence of God, …

This first clause belongs to the Confrontation Stage of a critical discussion. That
there is a difference of opinion is clear from “we are going to discuss”, for people
only prepare to discuss what they disagree about. Moreover, the general issue about
which there is disagreement is described by the phrase ‘the existence of God’. Given
the Western tradition, it is easy to convert this description into the question whether
god exists or not. This question I shall refer to as Q1, even though this is only a
rough formulation for the real question in dispute. Now, the two main terms, ‘God’
and ‘exists’ (or ‘existence’) are, for philosophers, quite problematic and each has
a long and disputatious history. This problematic character of the two terms was
not necessarily present to the general audience of the broadcast, many of whose
members, not being philosophers, may be satisfied with a mediocre understanding
of those terms. It is thus clear that Q1 does not mean the same for a philosopher than
for a non-philosophical audience.
Moreover, the audience has tuned the radio in order to listen to a debate. So, the
audience knows that there is a disagreement between the parties of the debates, for
otherwise it would not be a debate. Still, for a difference of opinion to trigger a critical
discussion we need to know what the position of each party is. Some listeners will
have known, for instance, that Bertrand Russell had given, in 1927, a lecture on the
question why i am not a christian, subsequently made into a pamphlet; others
at least knew that he was often described in public as a notorious atheist. For them, it
was a safe bet that Russell was going to take the negative with regard to Q1. Paired
with the fact that Frederick Copleston must have been presented as a Jesuit priest,
these facts would have betrayed the game for most intelligent listeners: Copleston
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 93
F. Leal and H. Marraud, How Philosophers Argue, Argumentation Library 41,
https://doi.org/10.1007/978-3-030-85368-6_4
94 4 Analysis of Segment I: Start of the Debate

will surely take the affirmative as to Q1. None of this is explicit at this point, however;
and it will only be made explicit at Turn 3. To that extent, the Confrontation Stage is
set but the actors are still partially blurred.
On the other hand, “we are going to discuss”, also hints at the Opening Stage, for
it indicates a resolution actually to go beyond the fact of disagreement and actually
to have a discussion on the question.
[1.2] … it might perhaps be as well to come to some provisional agreement as
to what we understand by the term ‘God’.
The second clause belongs to the Opening Stage. Copleston is reaching out to Russell
in order to secure a material starting point, in this case a definition. To produce
definitions is quite usual in philosophy, especially when the term to be defined is
as problematic as ‘God’. Copleston’s move therefore implies that, in the case of
a term like ‘God’, as soon as Q1 is on the table, another question arises, namely
what the term ‘god’ means. This I shall call Q2. Whether the term ‘God’
refers to something and is not an empty name depends to a great extent on what
it means, so the two questions Q1 and Q2 overlap in that giving an answer to one
(whether god exists or not) is tied up with answering the other (what the
term ‘god’ means, or perhaps, more precisely: whether the term ‘god’ is
to be defined as copleston does). Quine (1960, §56) suggested the phrase
‘semantic ascent’ to denote a method to reach agreement on the meaning of prob-
lematic terms. This method would be a step towards the ultimate goal of reaching
agreement on a substantive issue. Notice that Copleston qualifies the agreement he
is looking for as provisional. The full force of this qualification will be brought out
at the end of this turn.
[1.3] I presume that we mean a supreme personal being, distinct from the world
and creator of the world.
At this point of the Opening Stage, Copleston might have asked Russell to define the
term ‘God’, so as to see whether the two philosophers might agree on the definition.
But Copleston proceeds instead to offer a definition himself for approval by Russell.
The description given as definiens by Copleston is utterly traditional: God is
(a) a supreme being, no other thing being above Him,
(b) a personal being, and not an impersonal force, as it were, like Nature,
(c) distinct from the world, i.e. not in the world, or, in philosophical parlance,
transcendent, and
(d) the creator of the world, which is a very specific relation between the
transcendent God and the world.
We shall have occasion to see that this very traditional definition embraces a whole
theology. Still, it is a very well-known theology, both to his adversary, Bertrand
Russell, who had received a Christian education, and presumably for the whole
audience of the broadcast. Christian theology goes far beyond these four attributes,
but it certainly includes all of them. So far, then, there is nothing untoward in the
definition offered by Copleston to Russell for his agreement.
4 Analysis of Segment I: Start of the Debate 95

[1.4] Would you agree—provisionally at least—to accept this statement as the

meaning of the term ‘God’?
The operation of looking for an agreement to a material starting point, so charac-
teristic of the Opening Stage, is realized through a speech act: the actual requesting
of agreement on the definition offered in 1.3. Notice that Copleston is not playing
hard: he allows Russell to withdraw his agreement later on. What he wants is just a
provisional agreement so as to be able to proceed.

Segment I, Turn 2 (Russell) Opening Stage

[2.1] Yes, I accept this definition.

Russell, as part of the Opening Stage, accepts the definition of the term ‘God’ offered
by Copleston as a material starting point of the discussion on Q1. There is thus no
disagreement about the answer to Q2 and therefore no discussion about it. The
question of the meaning of the term ‘God’ will not re-emerge until much later, at
Turn 92. Notice that one of the four attributes of God contained in Copleston’s
definition, namely (d) creator of the world, could be used to prove God’s existence.
The imaginary argument would be as follows:
Premise 1. The world exists.
Premise 2. God created the world.
Conclusion. God exists.

Premise 1 refers to a fact very difficult to deny, even for Russell. Premise 2 is just a
consequence of the definition which he has just agreed to. The conclusion seems to
follow from the premises.1
However, Russell would never admit such an argument, nor would Copleston
intend to offer it. It is therefore clear that Copleston is offering his definition as a
mere nominal definition, and that this is also how Russell is taking it: what is thereby
defined is just a name, a term, a word. Yet the question whether that name has an
actual reference or is empty remains open until it should be proved that what is so
named does exist. What is then the purpose of the nominal definition then? Only
this: to specify Q1 a bit more closely, so that the question whether god exists
or not means as much as the question whether something with all these
attributes exists or not. This I shall call Q3.
The operation of specification is vital in philosophical argumentation. From now
on, it is then agreed between Russell and Copleston that the debate is about Q3.
Thus, one function of the Opening Stage, namely, to clarify what the issue is, has
been fulfilled. Nonetheless, Q1 has not disappeared from view. Later on, Copleston

1 In fact, the very use of a definite description, ‘the creator of the world’, would already warrant
the conclusion in classical logic. Besides, the true conclusion of this imaginary argument could
only be that God existed sometime before and at the time He created the world, but not necessarily
thereafter. I owe this clarification to Hubert Marraud and Joaquín Galindo.
96 4 Analysis of Segment I: Start of the Debate

will surreptitiously add a further attribute to ‘God’, namely, ‘goodness’, which was
not in the original definition (see 9.1).

Segment I, Turn 3 (Copleston) Confrontation Stage / Opening Stage

[3.1] Well, my position is the affirmative position that such a being actually
exists, and that His existence can be proved philosophically.
With this sentence, we are back to the Confrontation Stage, for the difference of
opinion is now clearly expressed to the extent that Copleston assumes one possible
answer to Q3, namely the affirmative. Notice that the adjective ‘such’ in the phrase
‘such a being’ refers to the attributes he has enlisted in the definition given in 1.3.
Now, this sentence has two clauses. The first clause is indeed a statement of
Copleston’s standpoint (“my position is… that such a being actually exists”); and
so, it belongs to the Confrontation Stage. However, the second clause belongs to
the Opening Stage, because it determines the role Copleston intends to play in the
following moves: he assumes the role of a Protagonist in the debate, he carries the
burden of proof.
Notice that Copleston qualifies the proof he intends to give as philosophical.
Although this qualification is not commented on during the debate, for anyone having
a sufficient background knowledge of the disputes on the existence of God in the
Christian tradition will be aware that Copleston is thereby distinguishing his argument
from any argument taken from revelation. His argument will be purely philosophical
and not theological in the narrow sense of doctrine revealed in Holy Scripture.
[3.2] Perhaps you would tell me if your position is that of agnosticism or of
atheism. I mean, would you say that the non-existence of God can be proved?
In order to complete the task of role assignment in the Opening Stage, it has to be
established what role does Russell intend to play in the debate. Will he also carry
a burden of proof of the negative? Or will he abstain from any burden of proof and
just play the role of an Antagonist? Copleston asks Russell to be explicit about it.
First, Copleston uses the time-honoured distinction, introduced by Thomas Huxley,
between atheism and agnosticism (see Wace et al. 1889). Then, he makes it more
explicit, probably for the benefit of those listening to the broadcast, by means of
a straightforward question (“would you say that the non-existence of God can be
proved?”). For, if Russell would say yes to that question (let’s call it Q4), then he
would also carry a burden of proof for the negative. We would then have a mixed
difference of opinion, to whose resolution both parties would contribute in the same
way, by alternating the role of Protagonist and the role of Antagonist.
4 Analysis of Segment I: Start of the Debate 97

Segment I, Turn 4 (Russell) Opening Stage

[4.1] No, I should not say that; my position is agnostic.2

In keeping with the Opening Stage, Russell simply expresses his desire to play the
role of a pure Antagonist, and thus to reject any burden of proof. He acknowledges
Huxley’s distinction by calling this the position of an agnostic. Later on, we shall
have occasion to observe that this purely antagonistic position will be shifted towards
a far more protagonistic one; but, for the time being, Russell announces his intention
to rest content with punching holes in Copleston’s argumentation.
Russell here tacitly accepts Copleston’s definition of an atheist as somebody who
can prove that God does not exist, so that an agnostic would presumably be someone
who cannot prove either that God exists or that He does not exist. As an interesting
aside, it may be worthwhile saying that Russell subsequently admitted wavering
between declaring himself an atheist and an agnostic (Russell 1949), albeit without
rejecting either definition.
In terms of the distinction between the stances which I called Gamma (who ques-
tions the standpoint) and Delta (who questions the question itself), Russell would
clearly be adopting here a Gamma stance (see Chap. 3, Sect. 3.2). Later on, however,
he will move to a Delta stance (see 40.1 and 68.1).

Segment I, Turn 5 (Copleston) Opening Stage

[5.1] Would you, then, agree with me that the problem of God is a problem of
great importance?
Again, as part of the Opening Stage, Copleston is trying to secure agreement on a
further material point. He raises the question whether the question of god’s
existence is important or not. Let’s call it Q5. Copleston wants Russell to give
his agreement, so that all listeners see that the debate is not about a trivial question.
I should point out here that a crucial point in Huxley’s agnosticism was that the
question of God’s existence is of little or no importance. We can be lukewarm about
this and other religious questions, because nothing of importance follows one way
or the other. So, if Russell is an agnostic, he should say something about Q5. On
the other hand, Q5 is still a relatively vague question: important in what respect?
Anybody would find difficult to answer Q5 in that form. However, Copleston has
something more definite in mind, which he will presently put forward.

2 Russell (1949) on the distinction of atheism and agnosticism: “Here there comes a practical
question which has often troubled me. Whenever I go into a foreign country or a prison or any
similar place, they always ask me what is my religion. I never know whether I should say ‘Agnostic’
or whether I should say ‘Atheist’. It is a very difficult question and I daresay that some of you have
been troubled by it. As a philosopher, if I were speaking to a purely philosophic audience, I should
say that I ought to describe myself as an Agnostic, because I do not think that there is a conclusive
argument by which one prove that there is not a God. On the other hand, if I am to convey the right
impression to the ordinary man in the street I think I ought to say that I am an Atheist, because
when I say that I cannot prove that there is not a God, I ought to add equally that I cannot prove
that there are not the Homeric gods.”
98 4 Analysis of Segment I: Start of the Debate

But before going into that, notice that there is no discussion about Russell’s answer
to Q4. Somebody could challenge Russell’s position as insincere, for perhaps his
actual position was that of an atheist rather than a mere agnostic. There was no
shortage of evidence in Russell’s writings and lectures to support the challenge (see
comment on the next clause). However, Copleston accepts Russell at his word. On the
other hand, Copleston could even have insisted that, this being a debate, Russell was
obliged to take a definite position with regard to Q3. After all, debates are institutional
practices in which each party has to offer arguments, either for the affirmative or for
the negative. But Copleston accepts Russell’s role assignment without further ado,
perhaps knowing full well that, in due time, the old philosopher was bound to start
producing arguments for a far more forceful position than the assumed one of an
agnostic. This is certainly what actually happened during the debate, as we shall see.
[5.2] For example, would you agree that if God does not exist, human beings
and human history can have no other purpose than the purpose they choose to
give themselves, …
Still in the Opening Stage, we now get a more definite question (“would you
agree…?”—let’s call it Q6), which we need to unpack a little bit, though. Remember
that we are dealing here with a question framed in Christian terms. If the term ‘God’
means the Christian God, then that God, as our creator, has given us a purpose. That
is the doctrine meant here. Although Copleston, in the definition put forward in 1.3,
did not include this doctrine of Christian theology, it is very much in his mind. If the
question of God’s existence is dismissed as unimportant (and this was very much
part of the point in Huxley’s agnosticism), then the only way we as individuals and
our collective history may have meaning and purpose is by us taking our destiny
into our hands. Copleston wants Russell to admit that his agnostic position implies
something like that. In fact, at the end of the lecture “Why I am not a Christian”,
Russell said the following:
We want to stand upon our own feet and look fair and square at the world—its good facts,
its bad facts, its beauties, and its ugliness; see the world as it is and be not afraid of it.
Conquer the world by intelligence, and not merely by being slavishly subdued by the terror
that comes from it. The whole conception of God is a conception derived from the ancient
Oriental despotisms. It is a conception quite unworthy of free men. When you hear people
in church debasing themselves and saying that they are miserable sinners, and all the rest of
it, it seems contemptible and not worthy of self-respecting human beings. We ought to stand
up and look the world frankly in the face. [Russell 1927.]

This quote shows that Russell was an atheist rather than an agnostic, as I said
before; but it also shows that he drew the consequence which Copleston is at pains
to express in this clause. A couple of years later, and probably in the follow-up of the
BBC broadcast, Russell was invited to produce a paper on Q1, at the end of which
he again draws the consequence:
My conclusion is that there is no reason to believe any of the dogmas of traditional theology
and, further, that there is no reason to wish that they were true. Man, in so far as he is not
subject to natural forces, is free to work out his own destiny. The responsibility is his, and
so is the opportunity. [Russell 1952.]
4 Analysis of Segment I: Start of the Debate 99

Copleston is therefore only trying to make this consequence explicit, and to make
Russell agree to that.
It may be worthwhile remarking that Copleston begins this clause by the phrase
‘for example’. This means that the consequence drawn by Russell himself in the
passage quoted above is only one of several. Copleston is making explicit just one
consequence; and it is likely that he picked this one because he knew Russell had
expressed it himself.
In any case, Q6 could be framed as follows: whether the non-existence of
god leads to human self-assignment of purposes.
[5.3] … which—in practice—is likely to mean the purpose which those impose
who have the power to impose it?
This second clause goes much further than the first one. If Copleston is right, the
consequence of the non-existence of God I have been talking about, namely, that
“man is free to work out his own destiny”, to use Russell’s own words, in practice
tends to lead to the powerful of this world imposing their will on the rest of us.
Remember that this debate is taking place at the beginning of1948, less than three
years after the end of World War II and its horrors (an event we shall revisit later on,
starting with Turn 103). It is also the beginning of the Cold War between the West
and the Soviet Union. For all intelligent listeners, the idea of the powerful imposing
their will was a clear allusion to all totalitarian regimes. So, the question raised here,
let’s call it Q7, might be framed as: whether the non-existence of god leads
to totalitarian regimes or not. Copleston does not quite express it that way,
of course, but his formulation in that context invites this interpretation.
In any case, the example developed in 5.2 and 5.3 constitutes an argument, or at
least the sketch of one:
Premise. If God does not exist, then [certain momentous consequences follow].
Conclusion. Deciding whether God exists is important.

Segment I, Turn 6 (Russell) Opening Stage

[6.1] Roughly speaking, yes; though I should have to place some limitation on
your last clause.
As part of the Opening Stage, we have Russell agreeing with the answer to Q6 but
expressing some reservations as to the answer to Q7. In the light of the quotations from
two of Russell’s writings on religion given above, we can imagine his discomfort
at Copleston’s allusion to the horrors of totalitarianism. Russell was all his life a
believer in political freedom, so he probably had hope that it was possible for human
history to have a purpose neither given by God nor by dictators. This is only a surmise,
because Copleston does not ask Russell to explain the limitation he has in mind and
he himself does not insist on expressing it.
From what we know about Russell, it is quite likely that he had in mind something
like the following counterargument: The consequence that Copleston envisages is
100 4 Analysis of Segment I: Start of the Debate

not as bad as he suggests, because it won’t necessarily be the case that the powerful
will always manage to impose their (nefarious) purposes on us. If something like
this was Russell’s argument, then he could have been rejecting Copleston conclusion
that answering Q1 is very important.
Nevertheless, this is all speculative, because Russell in fact does not explain the
origin of his reticence, nor indeed an argument that could rebut Copleston. For the
purposes of the Opening Stage, it is enough to state that the material starting point
proposed for agreement is sufficiently rejected to pre-empt Copleston’s thesis of the
importance of Q1.

Segment I, Turn 7 (Copleston) Opening Stage

[7.1] Would you agree that if there is no God—no absolute being—there can be
no absolute values?
It is with this turn that Copleston finally comes out into the open. The agreement
he is seeking in this part of the Opening Stage is an agreement not so much about
something as mundane as the status of freedom under different political regimes as it
is about a definite philosophical position, moral or value relativism, which of course
is not just a highfalutin theory but has a certain deleterious effect on ordinary life
and human practices. Copleston suggests that this position and its effects stems from
thinking that God does not exist. Notice that now, to the four attributes mentioned in
the definition given by Copleston in 1.3, a new one is added:
(e) an absolute being.
This attribute is not explained by Copleston, but, based on the philosophical tradition,
we can say that he means a being which is utterly by itself and does not depend, for
His existence, on anything else. This is something that Copleston will articulate later
on. And in a sense, one could perhaps argue that, if this alleged God created the
world, then the world needs Him in order to exist, but not the other way around. To
that extent, we may accept that (e) after all is contained in the definition given in 1.3.
In any case, (e) must belong to the definition of God, since Copleston’s argument
will try to prove that An absolute being exists, and without that predicate he could
not derive the proposition That God exists.
On the other hand, there is something puzzling in the material starting point put
forward by Copleston in order for Russell to agree with it: the connection between
God and values does seem to be an attribute utterly outside Copleston’s definition.
This brings us to the next bit. But before that, note that a new question has been
raised (which will be confirmed in the next sentence): whether the non-of god
leads to totalitarian regimes or not. Let’s call this Q8.
4 Analysis of Segment I: Start of the Debate 101

[7.2] I mean, would you agree that, if there is no absolute good, the relativity of
values results?
The phrase ‘I mean’ makes explicit that Copleston is aware that, in his last sentence, he
has introduced a new attribute for God. This sentence makes the new Divine Attribute
fully explicit—by a legerdemain which replaces absolute being with absolute good.
The Christian God, in fact, is—theologically—not only, as set forth in Copleston’s
initial definition (1.3), the creator of the world, but He is also the “ultimate ground
of value”, as Copleston will later on express it himself (in Turn 135). Nonetheless,
Copleston was clear that one thing is to prove the existence of an absolute being and
another to prove that such a being is good (as he makes explicit in his summary at
137.4.5).
It is worth remarking that Russell does not seem to be aware of the abovesaid
fine point in theology; even less to know that the separation of these two attributes
(creating the world vs. being the absolute good or the ultimate ground of value) is the
origin of some of the most celebrated Christian heresies (most famously, Gnosticism,
Marcionism, and Manichaeism): one God created the world, but He is not good; and
another God, the real one as it were, is good, in fact, He is the hidden source of
all that is good, but He could not have created the world—precisely because of His
goodness. So, there are various theological questions which underlie this apparently
simple intervention by Copleston, which never become active questions in the debate.
When Russell later on introduces the Devil (one of his favourite characters) and
intimates that it is the Devil who seems in charge of the world rather than (a good)
God, he is just producing a late lay version of the old dualistic Christian heresy.
Underlying all this is the ancient question of theodicy (the question of the origin
of evil in the world and its compatibility with Divine goodness), that has plagued
Christian theology and Christian-inspired philosophy practically since the beginning.
However, this question never quite becomes a proper, explicit topic in this debate.
Apart from all these theological niceties, the fact of the matter is that Copleston,
by so framing his request for a further material starting point, has manoeuvred to
replace Q3, which seemed to be the presiding question after Copleston’s definition
(1.3) had been granted by Russell (2.1). The new question that should replace Q3
would have been whether there is an absolute good or not. Let’s call it
Q9. We shall see that Russell will tacitly reject Q9 and remain focused on Q1.

Segment I, Turn 8 (Russell) Opening Stage

[8.1] No, I think these questions are logically distinct.

In the Opening Stage, proposals of material starting points can be accepted, as in
Turn 2, accepted with restrictions, even if not spelt out in detail, like at Turn 6, or they
can be flatly rejected, as in this turn. Notice that Russell speaks here of questions,
two of them. One question must surely be Q1, the existence of God, here understood
by Copleston as the existence of an Absolute Good (even if this was not contained in
his definition of ‘God’ agreed to in Turns 1 and 2). The other question is, probably,
102 4 Analysis of Segment I: Start of the Debate

whether values are relative or not. This is a variant of Q9. Russell is

saying that answering Q1 is logically distinct from answering Q9. Coming from a
logician, this probably implies that all combinations are possible:

Combination Q1 Q9
Does God exist? Are values relative?
I Yes Yes
II Yes No
III No Yes
IV No No

Copleston wants Russell to agree that only combinations II and III are admissible,
i.e., logically, that a Yes to one of the two questions implies and is implied by a No
to the other question—a logical equivalence in fact. Russell rejects such a logical
equivalence, and proceeds to give an argument, thereby assuming the burden of proof
for a moment.
Notice that, by his answer, Russell has effectively raised the question: whether
denying god’s existence is logically equivalent to upholding the
relativity of values (let’s call it Q10 and note that it is stronger than Q8). This
is, so to speak, a meta-question, concerning not the content of a particular philosoph-
ical proposition but rather the logical dependence of a philosophical proposition on
another. Copleston has, with his attempt at securing a material starting point in the
Opening Stage, entered the turf of the logician, and the logician has answered Q10
in the negative.
[8.2] Take, for instance, G. E. Moore’s Principia Ethica, where he maintains that
there is a distinction of good and evil, that both of these are definite concepts.
Russell’s argument starts with a fact: Moore’s position in his famous book of 1903:
that good and evil are definite concepts, i.e. that if something is good, it cannot be
evil, and the other way around. There is no mixture between the two concepts. One
cannot say that goodness depends on the perspective, on the context, on the situation,
on the person who is making the judgment, for if it did, then the same thing could be
good in one condition and evil in another. This would blur the distinction between
the two concepts. There is no moral relativism in Moore. In other words, Moore is a
philosopher who would answer No to Q9.
Notice that it is Q1 which Russell keeps firmly in mind, and not Q9, as somewhat
insidiously thrown in by Copleston. This is important, because asking about the
relation between Q9 (whether there is an absolute good or not) and
Q1, Copleston’s implication would be difficult to reject, for that implication would
appear to be tautological.
[8.3] But he does not bring in the idea of God to support that contention.
If Copleston’s logical equivalence would hold, then Moore, having answered No
to Q9, should have answered Yes to Q1. But Moore is completely innocent of any
4 Analysis of Segment I: Start of the Debate 103

theological position in Principia Ethica (1903). In establishing his answer to Q9, he

leaves Q1 alone. Like Laplace, when he was apocryphally asked by Napoleon about
the role of God in his celestial mechanics, Moore would have said ‘I did not need
that hypothesis’.
Russell’s precise argument bears some repeating. Basically, it is this: one can
answer Q9 without having to answer Q1, and Moore is living proof of this. In other
words, he gives a factual counterexample. Copleston could have asked the further
question, whether moore was right in his ethics or not, but he wisely
abstained from this. Sometimes one has to leave questions that arise in the course of
a discussion if one wants to keep the discussion focused. This is very good practice.
We should, in any case, resist the temptation to think that Russell is giving an
argument of authority. He is not saying that he subscribes Moore’s ethics. He did
adhere to it at some point (Russell 1910), but by 1948 he had certainly outgrown it
(see Blackwell 1985). Russell’s argument does not presuppose that Moore is right.
He is just saying that there is a respectable (and presumably consistent) ethical
system in which, far from being forced to answer one question (Q1) in a certain way
when answering another (Q9), you can go ahead and answer the latter whilst utterly
ignoring the former.
In this part of the Opening Stage, Copleston, having obtained agreement to a
material starting point (the definition of God), and at least partial agreement to a
second one (if God does not exist, then ‘man’ must create his own values), completely
fails to obtain agreement to a third one (if God, conceived as an absolute good, does
not exist, then all values are relative).

Segment I, Turn 9 (Copleston) Opening Stage

[9.1] Well, suppose we leave the question of good till later, till we come to the
moral argument.
The Opening Stage has three tasks to fulfil: securing agreement to material starting
points, assigning the roles of Protagonist and Antagonist, and acquiescing on a proce-
dure for the discussion. We have observed the first task at work in Turns 1–2 (full
success), 5–6 (partial success), and 7–8 (no success). The second task was fulfilled
in Turns 3–4. In this turn, Copleston offers a proposal for organizing the discussion.
The first sentence is particularly interesting. Imagine for a moment that Russell, at
Turn 8, had been caught off his logical guard and had said Yes to Copleston’s crypto-
tautological equivalence, although this equivalence hid a swapping of concepts (see
Nelson 2016), as I explained in 7.2. If that had happened, then Copleston might have
used that concession later on. For, as we shall see soon (9.2), Copleston planned to
start with the metaphysical argument. For that argument, he needed something like his
definition of God, as set forth in 1.3, where no moral qualities are involved. Afterward,
he planned to go smoothly over to the moral argument (see 9.3). And for that other
argument, he needed something like the identification of God with Absolute Good,
which is conspicuously absent from the definition at 1.3. Copleston surreptitiously
embedded this swapping of concepts (see Nelson 2016) in his crypto-tautological
104 4 Analysis of Segment I: Start of the Debate

question. By securing, at this initial point in the debate, Russell’s acquiescence to the
equation of God = Absolute Good = Absolute (Non-Relative) Values, Copleston
might later on try to force Russell to admit that the third member of the equation
is something real, and so to the reality of the first member, which is what Q1 is
about. That plan would not have worked anyway, because Russell was a relativist of
sorts; but this seems to have been Copleston’s strategy all along. That it was is not
idle speculation on my part, but it is borne out by the clause “suppose we leave the
question of good till later till we come to the moral argument”. There was a definite
plan, even if it was bound to fail.
[9.2] I should like first to advance a metaphysical argument.
As part of the procedure which Copleston offers for Russell to agree to, the former
announces that his first line of argumentation will be metaphysical. This adjective
comes from Kant (a figure of the history of philosophy whom Copleston will invoke
on several occasions during the debate), who distinguishes three metaphysical argu-
ments, which he calls the ontological, the cosmological, and the physico-theological
proof. (The latter belongs to what is called, in English, ‘natural theology’, and it
is nowadays widely known as the teleological argument.) Of these three arguments
recognized by Kant, Copleston will choose the cosmological proof, of which he will
present a version inspired in Leibniz (see Turn 11).
[9.3] I would like to put the main weight on the metaphysical argument based
on Leibniz’s argument from contingency, and then later we might discuss the
moral argument.
Copleston announces that he has a second line of argumentation, which he calls
moral, but emphasizes that the first, metaphysical argument is the most important
one (the only one worthy of the name ‘proof’, see Turn 95). This is important for
the procedure proposed in this turn, in that Copleston is telling Russell that the
metaphysical (more precisely, cosmological) reasons he has for his standpoint (that
God exists) are much stronger than his moral reasons. Of course, Russell is not going
to be convinced, and Copleston knows that; but the cards are on the table.
[9.4] Suppose that I give a brief statement on the metaphysical argument and
then we go on to discuss it?
Having exposed the general procedure he would like to follow, Copleston now
proposes to begin with the first line of argumentation. He asks Russell whether
he agrees to the following order: First, Copleston expound his full argument in one
go, and then Russell responds to it.
Notice that the rules for discussion could have been framed differently. For
instance, Copleston might have proposed to begin by stating one of his premises,
and open discussion on it alone, independently of the other premise. He might even
have proposed to follow a Socratic procedure, as exhibited in Plato’s dialogues. In
that case, Copleston would ask a series of short, definite questions for Russell to
answer. Out of Russell’s answers a definite philosophical argument would slowly
emerge, and that would be Copleston’s metaphysical argument. And quite a few
4 Analysis of Segment I: Start of the Debate 105

other different rules might in principle be proposed; but Copleston proposes a very
particular procedure, which was quite typical of scholastic philosophy.
Again, Copleston might have proposed to start with a different argument, say, one
in Aquinas’ list. It is remarkable, in particular, that the argument from design, which
is far more popular than the argument from contingency, was in fact excluded from
discussion. The choice of argument can perhaps stem from a desire, on Copleston’s
part, to be as careful as possible at a time when the Roman Catholic Church was
much stricter than it is today.3
Anyway, it should be clear that Copleston is trying to secure agreement, not, as
before, on a material starting point, but on a procedural one. And this starting point
is by no means trivial.

Segment I, Turn 10 (Russell) Opening Stage

[10.1] That seems to me to be a very good plan.

Russell agrees to Copleston’s procedural starting point. This agreement will soon
enough cause a bit of trouble for Russell, as we shall see when he, at the beginning
of Turn 12, will say that Copleston has said so many things that “it’s not alto-
gether easy to know where to begin” (12.1). He might have anticipated this problem,
characteristic of the scholastic method, and more generally of the ‘long discourses’
(makrología) which Plato’s Socrates had already identified. In that case, Russell
might have suggested a different, slower, step-by-step procedure. It is clear that he
missed his chance to do so in time.
With Russell’s agreement to Copleston’s rules for discussion, we can say that all
the three tasks of the Opening Stages have been realized and fulfilled:
First, as to finding material starting points before the discussion, three of them have
been put forward by Copleston: (a) the definition of the term ‘God’ at Turn 1, which
Russell accepted; (b) the thesis of the paramount importance of discussing the ques-
tion of God’s existence, at Turn 5, where a negative answer would shift the determi-
nation of purposes from God to human beings—a consequence accepted by Russell
fully—so that the powerful, even dictators, would be impose those purposes—a
consequence Russell does not accept completely; and (c) the thesis that denying
God’s existence is tied up with moral relativism, which Russell rejects. Notice that
the material starting point (b) tries to smuggle a Divine Attribute which goes well
beyond the definition agreed in (a), but the discussion of this is postponed until the
discussion of the moral argument.
Secondly, as to assigning the roles of Protagonist and Antagonist: this is done
at Turn 3, by establishing that Copleston will be the sole Protagonist in the debate,
carrying the burden of proof; and that Russell will only assume the role of Antagonist.

3 Copleston (1993, p. 135): “As a Catholic priest, especially in Pre-Vatican II days, I was very
conscious of what my correligionists and the ecclesiastical authorities might think about the positions
which I adopted. I thought it incumbent on me to defend at any rate one traditional metaphysical
line of proof of God’s existence.”
106 4 Analysis of Segment I: Start of the Debate

The difference of opinion is thus defined to be non-mixed. At first, the difference

of opinion appears to be single, but there are already signs that more disagreements
will appear. This is borne in the sequel, so that the originally single difference of
opinion will very quickly become multiple. With regard to some of the other points
of contention, Russell will soon assume a role of Protagonist, so that the difference
of opinion will at times become mixed.
Thirdly and finally, as to agreeing to some definite rules for discussion: Copleston
has proposed something like the old scholastic method of first presenting a full
argument and then allowing the interlocutor the chance of trying to find one of several
holes in it. This procedure will always be followed when Copleston’s arguments are
first presented; but, during the debate, the rules will change somewhat, so that, at
some points, we shall have something approaching the Socratic method.
All this is proof that the Opening Stage is not a stable agreement, established once
and for all; but it is rather subject to revision, as indeed it was indicated by Copleston
when he, at Turn 1, suggested that his definition was provisional and it was intended
that, if Russell accepted it now, his agreement could be later withdrawn.

References

Blackwell, K. (1985). Russell’s Spinozistic ethics. London: George Allen & Unwin.
Copleston, F. C. (1993). Memoirs of a philosopher. Kansas City: Sheed & Ward.
Moore, G. E. (1903). Principia ethica. Cambridge, UK: Cambridge University Press.
Nelson, L. (2016). A theory of philosophical fallacies. Cham: Springer.
Russell, B. (1910). ‘The elements of ethics.’ In: B. Russell, Philosophical essays (Chapter 1).
London: Longmans, Green & Co.
Russell, B. (1927). ‘Why I am not a Christian.’ Lecture delivered on March 6, 1927, at Battersea
Town Hall, under the auspices of the South London Branch of the National Secular Society.
[Reprinted in: B. Russell, Why I am not a Christian, and other essays on religion and related
subjects (pp. 1–19). London: Routledge, 2004.]
Russell, B. (1949). ‘Am I an atheist or an agnostic?’ The Literary Guide and Rationalist Review 64
(7), 115–116. [Reprinted in: Slater, J. G. (Ed.), The collected papers of Bertrand Russell, vol. 11:
Last philosophical testament 1943–1968 (pp. 90–92). London: Routledge, 1997.]
Russell, B. (1952). ‘Is there a God?’ Paper requested for the London magazine Illustrated, but never
published there. [Printed for the first time in: Slater, J. G. (Ed.), The collected papers of Bertrand
Russell, vol. 11: Last philosophical testament 1943–1968 (pp. 543–548). London: Routledge,
1997.]
Quine, W. V. O. (1960). Word and object. Cambridge, MA: The MIT Press.
Wace, H., Huxley, T., Magee, W. C., Mallock, W. H., and Ward, H. (1889). Christianity and
agnosticism: A controversy. New York: Appleton.
Chapter 5
Analysis of Segment II: Discussion
of Copleston’s Metaphysical Argument

Abstract Segment II of the debate is the longest and tackles the most difficult
part of the whole debate: the cosmological argument for God’s existence, what
Copleston calls ‘the metaphysical argument’, which he considers a proper proof.
Segment II contains sixty turns (from 11 to 70), i.e. almost half of the debate. This
segment starts with a complicated version of the cosmological argument, although
other simpler formulations are later given. The discussion is inconclusive because
on several occasions an unbridgeable gap opens between the concepts and methods
of the two philosophers, who find necessary to discuss possible common starting
points, without much success.

Segment II, Turn 11 (Copleston) Argumentation Stage

[11.0] Well, for clarity’s sake, I will divide the argument into distinct stages.
The argument which Copleston will present is complex, so it is a good idea to explain
first what its structure is. The way he tries to do so is by means of a division of his
argument into stages. Before I try to explain what these stages are, I would like to
say that the verbal clues indicate clearly that there are five of them. This is why
the numbering, after this initial intervention (marked as 11.0 to mark its prefatory
character) will be 11.n.m marking the mth clause or sentence of the nth stage of Turn
11.
Let us now look briefly at the five stages of Copleston’s metaphysical argument.
The first and the second stages—introduced, respectively, by ‘First of all’ and
‘Now, secondly’—are preparatory.1 The first stage (11.1) presents the main factual
premise (11.1.1, that Everything in the world has a reason external to itself ) and,
given that its abstract metaphysical character calls for some clarification, this clari-
fication is given immediately (11.1.2). The second stage presents the second, defini-
tional premise (the world is defined as the totality of things in the world in 11.2.1).
This definition again calls for some clarification (11.2.2 through 11.2.4).

1 I owe this insight to the argument-dialectical analysis which Hubert Marraud will present in
Chapter 9. In the first draft of my analysis I made the mistake of interpreting 11.1 and 11.2 as
arguments. This is, for me, the fruit of adversarial collaboration.
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 107
F. Leal and H. Marraud, How Philosophers Argue, Argumentation Library 41,
https://doi.org/10.1007/978-3-030-85368-6_5
108 5 Analysis of Segment II: Discussion of Copleston’s …

The third, fourth, and fifth stages are themselves arguments, or rather partial
arguments, or sub-arguments, of Copleston’s metaphysical argument. The third stage
is the first sub-argument, which purports to prove, from the two premises presented
before, that The world has a reason external to itself . The fourth stage is the second
sub-argument, it has the structure of a dilemma, and it purports to show that, on each
branch of the dilemma, we have to come to the same conclusion: that The ultimate
reason of the world cannot have a reason external to itself . The fifth and last
stage is the third sub-argument, which carries the true punchline of the metaphysical
argument. It is a kind of transcendental argument: the condition of possibility of
explaining the world is the existence of the ultimate reason of the world. In that
quasi-Kantian sense, the existence of the ultimate reason of the world is necessary.
In other words, Copleston believes he can prove not only that God does exist but that
He must exist.
Note that this complicated formulation of what he calls ‘the metaphysical argu-
ment for God’s existence’ is not the only one. Later on, Copleston will produce
different versions, notably at Turns 53 and 55. For ease of reference, I shall call
this argument the First Metaphysical Argument. It is clear that, for Copleston, his
different formulations are all equivalent (see Chap. 2, Sect. 2.1).
[11.1.1] First of all, I should say, we know that there are at least some beings in
the world which do not contain in themselves the reason for their existence.
This is the first stage of the argument, clearly indicated by “First of all”. A proposition
is put forward with great force (“we know that”) and seemingly clear scope (“there
are at least some beings in the world which”). The phrase “we know that” indicates
that this proposition is to be understood as a matter of fact, that nobody in her right
mind would contest.
The predicate of the proposition, however, is somewhat opaque. If we express it
in the perspicuous manner of Russell’s propositional functions, we get: ‘x contains
in x the reason for the existence of x’. I shall refer to this as the Copleston Predicate,
given the importance it has in the subsequent development of the argument.
In the Copleston Predicate we have both a metaphor (“contain”) and an old meta-
physical phrase: “the reason for the existence of”, or ratio essendi, to use the tradi-
tional expression. The metaphor should perhaps not be made much of. It probably
does not mean much more than ‘have’. The idea seems to be that the reason of some-
thing’s existence can be either in it, presumably as part of it, or else outside of it, i.e.
there must be some individual different from x that is the reason for the existence of
x. If this is so, then 11.1.1 will have to be represented as a doubly quantified propo-
sition, for in it not only the existence of beings in the world would be asserted but
also the existence of other beings which are reasons for the existence of the former. I
shall have occasion to spell this out later on, when I comment on 11.3.1. As we shall
see, it will prove decisive.
Now, given that Copleston says that he wants to use Leibniz’s argument from
contingency, we can predict that, sooner or later, there will be some talk about ‘suffi-
cient reason’, and what that talk entails. Russell will indeed come to ask Copleston
whether ‘reason’ is the same as ‘cause’ (see Turn 32).
5 Analysis of Segment II: Discussion of Copleston’s … 109

[11.1.2] For example, I depend on my parents, and now on the air, and on food,
and so on.
Given the opacity of the Copleston predicate, it is necessary to clarify it. This is done
by means of a homely example. First, he adduces the fact that Copleston’s parents
engendered Copleston. One can either say that Copleston’s parents are the reasons
of his existence or that the fact that they engendered him is that reason. Secondly,
Copleston adds that he now (in his present existence) depends on air, food, and so
on. This is also a fact: without air, food, and so on, he would die. In other words,
thanks to his parents, Copleston came to exist, and thanks to food, air, and so on, he
keeps existing. Naturally, the same is true of Russell, and of every other human being,
indeed of every other living being on earth (mutatis mutandis, of course, for some
organisms do not have parents and some are anaerobic). So, this example, suitably
extended, makes perfectly clear that at least some beings in the world do not contain
in themselves the reason for their existence. Note that the beings in question are all
living beings.
Later on, Copleston will say that “no object of experience contains within itself
the reason of its existence” (11.2.2, 11.3.1), which is a much stronger statement. So, it
seems as though he all along had in mind a universal statement rather than a particular
one. We may therefore ask: what about non-living beings? Although Copleston does
not say anything about those, we may press the point in his favour and argue that non-
living stuff also exists because of certain forces that produced it in the past and keep
it together in the present. We could even say, somewhat anachronistically, that those
forces are, in the last instance, the four fundamental forces of nature (Davies, 1979).
The idea of force gives in fact an interesting twist to the old concept of ‘reason’.
One last thought. The forces of nature produce things, it is true, but they also
destroy things. So, by parity of reasoning, one could say that not only some things,
but all things do not contain in themselves the ‘reason’ for their destruction. This
would lead to the conclusion that there must be, outside the world, a ‘reason’ for the
destruction of the world; and God may perhaps be defined as the destructor of the
world with as much reason as He is defined as its creator (in Copleston’s definition,
1.3).
Some readers may think that I am overdoing this comment. I hope that they will
appreciate why this is needed when we come to 11.3.1. The point is that, once the
obscure metaphysical terms are clarified by the example, then anyone can recognize
11.1.1 as the expression of a fact well known to all of us.
[11.2.1] Now, secondly, the world is simply the real or imagined totality or
aggregate of individual objects.
This is the second stage, as indicated by “Now, secondly”. In contradistinction with
the first stage, no fact is advanced here but only a definition. (This definitional
character is expressed by “is simply”.)
Brief digression: it seems to be a consensus among argumentation theorists that
argumentation is a form of justification of a controversial proposition. If a propo-
sition is not controversial but rather purports to express a matter of fact or some
110 5 Analysis of Segment II: Discussion of Copleston’s …

other undeniable state of affairs, then there is room for clarification, explanation, and
the like, yet not for argumentation proper. The theoretical question arises whether a
definition leaves room for clarification, explanation, and so on, but not for disagree-
ment. Consider the following cases. Two lexicographers disagree about whether a
given definition renders the meaning of a word in a historically accurate way. Two
mathematicians may disagree about whether the definition is useful for the purposes
of a given axiomatic system. Two lawyers may disagree about whether the defini-
tion captures the reasoned conclusions of a past verdict. It follows that people can
disagree about the question whether a given definition is appropriate to the end it
is supposed to fulfil. However, they would not be disagreeing about the definition
itself, insofar as it is just a definition.
In this particular case, it is pretty clear that Copleston is merely suggesting that
the phrase ‘the world’ be taken to refer to “the real or imagined totality or aggregate
of individual objects”. Russell may comply with this stipulation or he may reject it,
although it may be difficult to foresee what the purpose of such a definition might be.
We shall see in due course that everything in Copleston’s First Argument depends
on this apparently harmless stipulation (see 11.3.1), which is why we shall refer to
it as the Copleston Definition.
Before going any further, it may be important to remark that the world can be
conceived as a set of objects or as a set of events. Wittgenstein is famous for
saying, right at the beginning of his first book, that the world is the latter not the
former (Tractatus 1, 1.1, 1.11, 2, 2.01). Naturally, the concept of a ‘state of affairs’,
which Wittgenstein uses, is wider in meaning than ‘event’; yet both events and states
of affairs are to individual objects as sentences are to names—a very consequen-
tial difference. Anyway, Copleston is quite cavalier regarding this distinction (and,
remarkably enough, Russell does not object). In this clause, as in 11.2.3, the world
is a set of objects, but later on, in 11.3.1, he will be explicitly indifferent to the
distinction. In fact, the somewhat vague word ‘being’ seems to comprise, for him,
both individual objects and events.
I insist on these logical minutiae, because later on (from Turn 16 onwards) the
question of whether ‘modern logic’ is more appropriate when trying to answer philo-
sophical questions like that of god’s existence than some other logic will be
explicitly raised. It is, therefore, a good thing to be alert to these details.
[11.2.2] none of which contain in themselves alone the reason for their existence.
First of all, notice that the Copleston Predicate reappears in this second clause of
11.2; but it is now slightly changed. At 11.1.1 we had ‘x contains in x the reason for
the existence of x’; here we have the addition of ‘alone’, a logically very important
matter. Copleston does not justify the change; but, being charitable, we may imagine
that he meant that from the beginning, only he omitted it in his earlier formulation.
After all, a living being keeps alive not only by food or air, which it takes from the
external environment, but he has in itself a certain organic structure which manipu-
lates, physically and chemically, that food and air in order to keep alive. So, keeping
alive is a matter of stuff that is external to the organism as well as stuff that is
internal to the organism. And the same is true, mutatis mutandis, of all non-living
5 Analysis of Segment II: Discussion of Copleston’s … 111

stuff. Notice that the adjectives ‘external’ and ‘internal’, although not present in so
many words in 11.2.2, come quite handy as clarifying the preposition ‘in’ (and in
fact the ‘containment’ metaphor). Later on, these adjectives will be made explicit
(see 11.3.1).
Now, the clause 11.2.2 hides more than meets the eye. It is a relative clause and,
as of all relative clauses, one can ask of it whether it is restrictive or not. If it is
restrictive, this means that Copleston is proposing to define ‘the world’ not just as
‘the totality of individual objects’ but rather as ‘the totality of all those and only
those individual objects which do not contain in themselves alone the reason of their
existence’. In that case, we would not have to ask the empirical question of whether
the ‘beings in the world’ do, as a matter of fact, contain in themselves alone the
reason for their existence, because they would not do so by definition. In that way,
the Copleston Definition would pre-empt the factual question, making it a conceptual
one. This is important, because later on (at Turn 31) Copleston will insist that his
First Metaphysical Argument is a posteriori, which is, of course, in tune with the
argumentation by example in 11.1.2. So, we better hold on to the other grammatical
interpretation of this second clause and consider it a non-restrictive relative clause.
In that case, ‘the world’ is indeed defined as ‘the totality of individual objects’, and
Copleston would additionally assert that none of those objects contain in itself the
reason for its existence.
If so, then the whole sentence has two clauses, the first of which introduces the
Copleston Definition (world = the totality of all individual things), whereas the
second clause puts forward an assertion, namely that each individual thing depends,
for its existence, on an external reason, at least in part. I beg the reader to keep this
in mind, for I shall come back to it very soon.
[11.2.3] There isn’t any world distinct from the objects which form it,
Clause 11.2.3 insists on the Copleston Definition of ‘world’, as given in 11.2.1, by
drawing a consequence of it: if the world is nothing but all the objects, then there is
no world apart from those objects. In a sense, what Copleston has in mind could be
rendered by saying that he wants to use ‘the world’ as a proxy for ‘everything’, but
without being ontologically committed to the existence of the set of all things as such.
It is this equation (world = everything) that will come very handy later on, when
Copleston enters the third stage of the First Metaphysical Argument (11.3.1). In fact,
I would like to suggest that we have here the main trick in Copleston’s toolbox.
By the way, something should strike us as quite remarkable. Copleston is perfectly
aware, like all English-speaking philosophers at the time, of Russell’s accomplish-
ments in mathematical logic and the theory of sets. Nonetheless, Copleston actually
dares to speak about sets (aggregates, classes, totalities) when debating Russell, prob-
ably one of the philosophers who has thought longer and harder than most about these
questions. However, it is equally remarkable that Russell will not pay any attention
to the ontological question thereby raised. Perhaps he was painfully aware, certainly
more so than Copleston, that he was being broadcast to ordinary listeners, who would
be overwhelmed by such abstract philosophical stuff. This hypothesis is borne by
his behaviour whenever things become too difficult for laypeople. In any case, if
112 5 Analysis of Segment II: Discussion of Copleston’s …

something is agreed upon among both mathematicians and philosophers, it is that

sets exist apart from their members. Without affirming their existence, mathematics
is not possible, and neither is natural science. This is at odds with Copleston’s asser-
tion that the world, which is after all a set, is not “distinct from the objects which
form it”.
[11.2.4] any more than the human race is something apart from the members.
Clause 11.2.4 appears to be a clarification, by means of an example, of 11.2.3, or
is it an argument? If the latter, then the argument uses an analogy, for the phrase
“the human race”, as here used, is equivalent to the collective noun ‘humankind’.
Copleston would then be reasoning like this: The world is not distinct from the
objects which form it, because humankind is not distinct from the objects which
form it, namely humans. Later on, Russell will remember this bit and use it wittily
to attack Copleston’s argument at Turn 51.
[11.3.1] Therefore, I should say, since objects or events exist, and since no object
of experience contains within itself the reason of its existence, the totality of
objects must have a reason external to itself.
This is the third stage of Copleston’s argument, as indicated by “Therefore, I should
say”. In contradistinction to the clarifications we have recognized in the first two
stages of Copleston’s First Metaphysical Argument, this complex sentence explicitly
takes the form of an argument (‘since p and q, r’, i.e. ‘p, q, therefore r’). So, we ought
to try to understand what the argument is. After some modifications to improve on
Copleston’s apparently careless formulation, we obtain the following:
(1) Premise 1. Objects exist.
Premise 2. No object contains within itself alone the reason of its existence.
Conclusion. The totality of objects has a reason external to itself.

Before going into argument (1), I ought to justify the modifications introduced in
this reconstruction.
I have taken the phrase “or events” out of Premise 1, so as to leave aside the
question of whether the world consists of objects or events, because Copleston does
not really seem to care about that question.
In consonance with this deletion, I have also taken the phrase “of experience” out
of Premise 2, even though Copleston would argue later on (at Turn 31) that the First
Metaphysical Argument is a posteriori, so that the deleted phrase should actually
be quite important to the argument. My justification is that, if I did not delete that
phrase, then I should have to add it also to the first premise, so as to have: ‘Objects
of experience exist’. However, this change would be quite problematic, because it
would force a change in the Copleston Definition of ‘world’ as ‘the totality of objects
of experience’. Yet it must have been clear to Copleston that not all individual objects
are, or indeed can be, objects of experience. Our experience of ‘beings in the world’ is,
on the contrary, always extremely limited with respect to the totality of them. Besides,
the phrase ‘of experience’ does not play any role in the conclusion (Copleston does
5 Analysis of Segment II: Discussion of Copleston’s … 113

not say: ‘the totality of objects of experience must have a reason external to itself’).
So, all things considered, it is much the safest to ignore the phrase ‘of experience’
altogether, as I try to reconstruct Copleston’s actual argument.
On the other hand, I have added the word ‘alone’ to Premise 2. Although this
word seemed to be a bit of an afterthought in 11.2.2, I have nevertheless shown in
my comment to that passage that the word ‘alone’ really belongs to the thrust of
Copleston’s argument and was elided in some parts of his argument only because
the formulation in the debate was in part improvised.
In the Conclusion, I have taken away the modal verb ‘must’. This modal verb,
when it turns up in a conclusion, can have two readings. On the one hand, it may
just express the fact that a conclusion has been drawn, that it actually follows ‘of
necessity’ (as Aristotle said a long time ago) from the premises. In that case, we
can safely delete it from the wording of the Conclusion. On the other hand, the
modal verb ‘must’ may also express the modal force of the proposition itself which
is concluded from the premises. In that case, the Conclusion would be itself a modal
proposition. Yet, seeing that none of the premises, as formulated by Copleston, are
modal, it would be impossible to draw a modal conclusion from them. So, again, it
is much the safest to delete the modal verb altogether.
Notice, by the way, that Copleston uses, for the first time in this argument, the
expression ‘reason external to’. In my comment of 11.2.2, I had occasion to introduce
the adjective ‘external’, in opposition to ‘internal’, as clarifying the Copleston Pred-
icate. The fact that Copleston now uses it as well, may serve as a further justification
of my procedure at that point.
Let us now have a closer look at argument (1). Its two premises are admittedly
quite plausible. Nobody is going to deny Premise 1. As for Premise 2, the comment
on 11.1.2 should be sufficient to accept that Copleston has a point here (there we saw
that a plausible argument could be constructed that would generalize Copleston’s
example of himself to all human beings, all organisms, and all physical objects).
All this is fine then. And yet, there is a problem, for it is not at all clear how the
Conclusion could follow from those premises as they stand, no matter how plausible
they are. Something has to give if we really want to have an argument that we can
accept, at least provisionally.
Better start from the obvious. Premise 2 is negative, whereas the Conclusion is
affirmative. Could we transform one of the two propositions so as to make it a match
for the other? Let’s try:
(2) Premise 1. Objects exist.
Premise 2. Every object has a reason external to itself.
Conclusion. The totality of objects has a reason external to itself.

I am here making full use of the propositional function ‘x has a reason external to x’
as an alternative expression for the negative of the Copleston Predicate:

[∼ (x contains in x the reason for the existence of x)] ≡ (x has a reason external to x)
114 5 Analysis of Segment II: Discussion of Copleston’s …

Now, we have all the elements to see through Copleston’s first explicit argument.
He wants to go from ‘F (every x)’ to ‘F (the totality of x’s)’, where F is either one
of the predicates in the above equivalence. This is, of course, in accordance with
the Copleston Definition, namely ‘the world’ = ‘everything’, the equation that I
commented upon before (11.2.3). If you accept the Copleston Definition, then two
things are clear. One is that argument (2) is correct. The other is that Premise 1 in (2)
is completely superfluous. It only adds an appearance of empirical robustness; but
the whole work is done by the Copleston Definition.
Nonetheless, the argument is not acceptable. The world could be made up of all the
objects, yes. Each single object might have reasons external to itself, yes. However,
the world—understood simply as ‘all objects’ or ‘everything’ or ‘the totality of
objects’ (a set that, in accordance with Copleston’s strictures, would not exist by
itself, independently of the objects that constitute it)—could perfectly well lack a
reason external to itself. For each single object might have a reason in another object
in such a way that all objects depend, for their existence, on all the others. Copleston
himself uses living beings as an example. Well, the idea of an ecosystem or even a
food chain, are good examples of such interdependence, which does not need any
reason which is external to the whole system or chain. Interdependence kills, as it
were, the need for system-external reasons. So, the conclusion does not follow, in
spite of the Copleston Definition.
So, the argument so far, as represented in (2), is, first, not empirical but conceptual,
resting as it does on a definition; and it is, secondly, not acceptable on account of the
exclusion of the possibility of interdependence.
It may be of some interest to see what the ‘modern logic’ which Copleston will
put under suspicion later on, would make of his argument. I suggest that Premise 2
in (2) could be represented as follows:
a. (∀x)(∃y) (y is the external reason of x)

Here, the variable x has as its scope all ‘beings in the world’ (physical, biological,
human), whereas the variable y could be either in this world or out of this world. The
Copleston Conclusion, in its turn, would look like this:
b. (∃y)(∀x) (y is the external reason of x)

The scope of the variables is the same as before, wider for y than for x. Now, it is true
that one can derive a from b, although of course a has much less information than
b. However, one cannot derive b from a, precisely because much more information
is in b than in a. But it is this latter, logically incorrect derivation that Copleston’s
argument requires. The British logician Peter Thomas Geach became famous for
pointing out the presence of this ‘formal fallacy’ in some philosophers, such as
Plato, Aristotle, Berkeley, and Spinoza (Geach 1972, pp. 1–13). Whether Geach’s
imputations are historically accurate or not, is a matter best discussed somewhere
else. All I want to claim here is that something like this ‘formal fallacy’ may lurk
in the argument formulated in 11.3.1. Some readers might regard this is otiose, but I
believe it is important in view of Copleston’s later indictment of ‘modern logic’.
5 Analysis of Segment II: Discussion of Copleston’s … 115

[11.3.2] That reason must be an existent being.

We are, again, confronted with a modal verb which needs interpretation. Let us
call R the external reason of the world as the set of all things. There are two readings
of 11.3.2:
(a) It follows [necessarily] that R exists.
(b) R exists necessarily.

In (a) the adverb ‘necessarily’ is strictly superfluous, just an emphatic way of

expressing the logical force of the verb ‘to follow’, i.e. the consequence relation
between the premises and the conclusion in (2). In contrast, the adverb ‘necessarily’
is central to what (b) asserts. Given that no modality is present in (2), it is better to
accept reading (a), which is just an immediate inference from the conclusion of (2):
if an existing object has a reason external to itself, then that reason also exists. I shall
refer to 11.3.2 in this reading as the Copleston Corollary.
[11.4.1] Well, this being is either itself the reason for its own existence, or it is
not.
We now come to the fourth stage, marked by “Well”. This is a whole argument in the
shape of a dilemma cornutus. If you take one horn, it leads you to a certain conclusion;
if you take the other horn, it leads you to exactly the same conclusion. Because
this kind of argument is so conceived that the two horns are seen as exhausting
the alternatives, then the inescapable conclusion is that proposition. You cannot go
anywhere else.
The phrase “this being” might perhaps be more clearly identified as R, the external
reason of the world. Copleston presents us with two exclusive and exhaustive alter-
natives, for which he uses one more time the Copleston Predicate. Either R has in R
the reason for R’s existence or R’s existence has a reason external to R.
[11.4.2] If it is, well and good.
The first horn of the dilemma is a neat shortcut to where Copleston wants to get: if R
has in R the reason for R’s existence, in other words, if the reason for R’s existence is
internal to R, then, if we assume that R is God, then God’s existence would have been
proved, given that it had already been proved that R exists (the Copleston Corollary,
11.3.2).
[11.4.3] If not, then we must proceed further.
The second horn of the dilemma is subtler. If you take that horn, then you are assumed
to be forced to find the reason of R outside of R. That would be the external reason
of the external reason of the world, or R2 .
Note a major assumption in Copleston’s way of thinking: that we are forced to
find a further reason. It looks very much like the endless why-question common in
some children.2 This is the Copleston Assumption.

2 Russell (1927, quoted from Russell 2004, p. 4): “Perhaps the simplest and easiest to understand is
the argument of the First Cause. (It is maintained that everything we see in this world has a cause,
116 5 Analysis of Segment II: Discussion of Copleston’s …

[11.4.4] But if we proceed to infinity in that sense, then there’s no explanation

of existence at all.
Here comes the most specious part of the sub-argument constituting the fourth stage
of Copleston’s metaphysical argument.
Remember that we are reasoning ex hypothesi, under the hypothesis that the reason
of R is outside R, and given that, according to Copleston, we must go on looking
for that reason, R2 as I have called it, then we must ask again whether the reason
for the existence of R2 is outside R2 . If not, then we are back to the result of the
first hypothesis, as in 11.4.2; but if not, then we are supposed to ask again about the
external reason for the existence of R2 , which now would be called the external reason
of the external reason of the external reason of the world, or R3 . This is supposed to
go on for some time, if needed, but it is also supposed to stop somewhere, say at Rω .
If we don’t stop at Rω , then we would have to admit that no explanation for existence
is at all possible.
For some reason, Copleston finds this conclusion abhorrent (although Russell
will happily embrace it, at Turn 50). There must be an explanation for existence. As
Leibniz used to say, pourquoi il y a quelque chose plutôt que rien?—why is there
something rather than nothing? That ultimate explanation of all existence, Rn , would
then be identified with God. If it is, then the Copleston Corollary would swiftly settle
that He is an existent being, i.e. that He exists.
Notice that here, for the first time, the concept of explanation is introduced by
Copleston. Later in the debate, the word ‘cause’ will be introduced, so that we have
to deal with a triad of terms: reason, explanation, and cause. It seems that Copleston
thought that all causes are reasons but not all reasons are causes. As for explanation,
it would be an epistemic relation between reasons and an inquirer.
[11.5.1] So, I should say, in order to explain existence, we must come to a being
which contains within itself the reason for its own existence—
We come finally to the fifth and last stage in Copleston’s argument, marked by “So, I
should say”. At first, this argument looks just like a summary of the result obtained in
the fourth stage: we want to explain existence, therefore, by the force of the dilemma

and as you go back in the chain of causes further and further you must come to a First Cause, and
to that First Cause you give the name of God). That argument, I suppose, does not carry very much
weight nowadays, because, in the first place, cause is not quite what it used to be. The philosophers
and the men of science have got going on cause, and it has not anything like the vitality it used to
have; but, apart from that, you can see that the argument that there must be a First Cause is one that
cannot have any validity. I may say that when I was a young man and was debating these questions
very seriously in my mind, I for a long time accepted the argument of the First Cause, until one
day, at the age of eighteen, I read John Stuart Mill’s Autobiography, and I there found this sentence:
‘My father taught me that the question, ‘Who made me?’ cannot be answered, since it immediately
suggests the further question, “Who made God?”’ That very simple sentence showed me, as I still
think, the fallacy in the argument of the First Cause. If everything must have a cause, then God must
have a cause. If there can be anything without a cause, it may just as well be the world as God, so
that there cannot be any validity in that argument.”
5 Analysis of Segment II: Discussion of Copleston’s … 117

cornutus, we must admit the existence of Rω . That would, of course, be God, for the
reasons described in my comment to 11.4.2.
[11.5.2] that is to say, which cannot not exist.
Then, somewhat surprisingly, Copleston says that not only does Rω exist, but that it
is impossible for Rω not to exist. However, the existence of Rω is settled, and can
only be settled, as a pure matter of fact. It is the Copleston Corollary that settles
it, and the Copleston Corollary is just a factual conclusion. As any conclusion, the
Copleston Corollary is necessitated by the premises; but it is just necessitated as
a conclusion. So, 11.3.2 is by no means a necessary proposition. Therefore, no
necessary proposition can be drawn from it, for instance, that it is impossible for
Rω —for the ultimate reason of all existence—not to exist.
Although to believe otherwise is wrong, the fact that Copleston takes this big cat
out of his small bag is an indication that Russell was onto something when he stated
later on that the ontological argument underlies the cosmological. He had after all
exhaustively explained the trick in his Leibniz book (Russell 1900, Chap. XV); and
the basic idea was, of course, already in Kant’s transcendental dialectic.
On the other hand, there seems to be a way out, which I hope will tickle the
ironic inclinations of readers. For let us read Copleston’s final sub-argument as a
transcendental argument in the grand Kantian manner:
(3) Premise 1. It is a fact that we explain things.
Premise 2. If Rω did not exist, then we could not explain things.
Conclusion. Rω must exist.

For reasons that will presently become clear, I shall refer to the sentence There is
a being which cannot not exist, namely Rω , as the Copleston Conclusion. It will
immediately attract Russell’s fire.
Note that Premise 1 in (3) is a factual premise, as belongs to a transcendental
argument. As standardly conceived, transcendental arguments start from the factual
premise that P is the case and go on to conclude that X must be the case too, since
X is a necessary condition for the possibility of Y. In Kant’s long argument of
the Critique of Pure Reason, the factual premise was nothing less than Euclidean
geometry and Newtonian physics, the two oft-repeated cases of a human inquiry that
had succeeded in entering ‘the secure path of a science’ (1787, 2nd edition, Preface,
1st paragraph). In Premise 1, we find the fact that we humans explain things, although
what Copleston had in mind was rather something like the Copleston Assumption,
namely that we humans are forced to keep on explaining things. Thus, a better
rendering of Copleston’s final sub-argument might perhaps be:
(4) Premise 1. It is a fact that we are forced to keep explaining things.
Premise 2. If Rω did not exist, then we could not finish explaining things.
Conclusion. Rω must exist.
118 5 Analysis of Segment II: Discussion of Copleston’s …

Something like this argument will come back later in the debate, when the two
discussants tackle the question of when to stop looking for causes or explanations
(see Turns 32–66).
Summarizing the whole First Metaphysical Argument, we can say that there are
three steps in that argument that do all the work:
(i) the way the phrase ‘the world’ is defined at 11.2.1, and especially the way it
is argumentatively used at 11.2.3 and 11.3.1, i.e. the Copleston Definition;
(ii) what I have called the Copleston Corollary at 11.3.2;
(iii) the instruction or requirement to iterate the pursuit of reasons at 11.4.3 and
11.4.4, which depends on the Copleston Assumption.

Each one of these steps is questionable for the reasons given in situ. On the other
hand, some readers might also object to the leap from the particular proposition at
11.1.2 to the general proposition at 11.2.2. However, I find that leap acceptable in
the light of the argumentation I provided in my comment to 11.1.2. In this particular
instance, Copleston might be accused of lack of explicitness but not of bad reasoning.

Segment II, Turn 12 (Russell) Opening Stage (Sub-Discussion, Procedural Starting

Point, Material Starting Point)

[12.1] This raises a great many points and it’s not altogether easy to know where
to begin,
Readers might be tempted to look at this turn as belonging to the Argumentation
Stage, namely as presenting doubts, objections, and counterarguments; but I shall
argue that it is better understood as an attempt, by Russell, to suggest going back to
the Opening Stage. There are so many questionable aspects, from a logical point of
view, in Copleston’s argument that Russell wants to put forward a procedural starting
point before continuing the discussion. At Turn 10, Russell had indeed accepted, as
‘a very good plan’, Copleston’s procedural starting point: “Suppose that I give a brief
statement on the metaphysical argument and then we go on to discuss it?” However,
the time has come to put some order on the proceedings.
It is easy to imagine that the public of the BBC Third Programme broadcast,
being mostly composed of laypeople, well-educated but unaccustomed to forms
of argumentation like the classical proofs of God’s existence, must have felt that
Copleston was talking well over their heads. Another question is whether Russell
could also have been taken by surprise.
I am quite sure he was not. It is true that Russell was not a specialist in the history
of philosophy, as Copleston—arguably the foremost English-speaking historian of
philosophy of the twentieth century—certainly was. Yet, in preparing a course of
lectures on Leibniz that was accidentally thrust upon him at the end of the nineteenth
century, Russell came to know and understand the Leibnizian philosophy better than
most. In his famous book on the topic, he has a whole chapter dedicated to the so-
called ‘proofs’ of God’s existence as developed by Leibniz (Russell 1900, Chap.
5 Analysis of Segment II: Discussion of Copleston’s … 119

XV). If one reads that chapter, it becomes clear that Russell was a true connoisseur
of this kind of argument, so that Copleston’s complicated piece of reasoning was not
new to him at all.
Russell’s problem, I suspect, was rather how to respond in the context of a BBC
broadcast. He will try to avoid technicalities as much as possible, but that will not
prove quite feasible. As none other than the Director General of the BBC commented
shortly after the broadcast, “the argument… was a bit too long for hearing” and
“perhaps a little too difficult for the broad run of even a Third Programme audience”
(Haley 1948).
[12.2] but I think that, perhaps, in answering your argument, the best point at
which to begin is the question of a necessary being.
Russell decides to tackle what might appear to some readers to be not only a relatively
secondary concept but indeed a term which Copleston had not used in the first place.
Copleston had in fact said quite a few things, but he had not used the phrase ‘necessary
being’ in his first presentation of the argument from contingency. Nonetheless, at
11.5.3, Copleston had concluded that Rω , the ultimate reason of everything, cannot
not exist. This grammatical predicate is rephrased by Russell as: ‘is a necessary
being’.
On the other hand, it is not completely clear what the question is which Russell
calls “the question of a necessary being”. We could perhaps formulate it at first in
terms of what the meaning of the phrase ‘necessary being’ is, which
we may call Q11. However, we will soon come to appreciate that Russell’s actual
question is rather whether the phrase ‘necessary being’ has a meaning
or not. Let’s call this Q12. This is a far more radical question, and it is the first hint
of the direction taken by Russell, from his original agnostic stance (‘I don’t know
whether God exists or not’) to a much more definite stance (‘the question whether
God exists is not meaningful’), a stance that is not really agnostic anymore. In any
case, we see that the discussion, from this point forward, will not just be about Q1,
but, in a kind of overlap, at the same time about Q12.
[12.3] The word “necessary”, I should maintain, can only be applied significantly
to propositions,
This is the first clause of a coordinated sentence. In it, Russell puts forward a logical
stricture: ‘necessary’ is not a predicate of objects but of propositions. Because it is a
stricture, we can say that Russell is not so much putting forward a standpoint to be
defended but rather suggesting a procedural starting point (in spite of the clause “I
should maintain”). The concrete form of Russell’s logical stricture takes the shape of
a syntactic rule for the use of a certain word; but it entails, more generally, a license
to use the hard-won insights and techniques of mathematical logic.
Notice that Russell is thereby going back to the Opening Stage, even though,
in contradistinction to Copleston, he does not put this forward as a request for his
partner to accept or reject, even though, as rules for discussion go, Russell’s proposal
is much more extensive and powerful than Copleston’s proposal at Turn 9. This lack
of an explicit request on the part of Russell, however, is probably due to the fact that,
120 5 Analysis of Segment II: Discussion of Copleston’s …

in 1948, the triumph of mathematical logic could be considered virtually complete,

at least in the Anglo-Saxon world. As one of the founders, foremost contributors, and
untiring champions of mathematical logic, not least for the handling of philosophical
questions, Russell may perhaps be forgiven for taking mathematical logic in this sense
for granted. (Cf. Chap. 8, 138.3.3.)
Nonetheless, a challenge is issued here and a new question, even if implicitly,
raised: whether mathematical logic may be used to tackle philo-
sophical questions such as the proofs for god’s existence. Call this
question Q13.
Copleston will not immediately recognize the starting point as such, with all its
implications, but he will eventually see the point and react to it by questioning the
authority of what he will call ‘modern logic’ (at Turn 17 and onwards).
In any case, Russell’s logical stricture entails that the Copleston Conclusion (see
11.5.3) is logically equivalent to saying that It is necessary that God (or the ulti-
mate reason of the world) exists. It is this modalized proposition that Russell wants
Copleston and the audience to pay attention to.
[12.4] and, in fact, only to such as are analytic, that is to say such as it is self-
contradictory to deny.
Copleston concluded that the existence of a certain being (no matter under what
name) is necessary, that this being exists necessarily. The question, raised by 12.4,
the second clause of the coordinated sentence, is what kind of propositions
can be considered necessary. Call this Q14. Russell answers that Only the
analytic propositions can be considered necessary. This is where a new difference
of opinion with Copleston will soon appear.
It is a striking fact that Russell, against his custom, enlists the term ‘analytic’,
a Kantian term which seldom occurs in his writings, and then usually only when
talking about Kant, whose relevant views he rejected. Nonetheless, he uses it often
in his Leibniz book, e.g. when he formulates the third of the five principles which
according to him underlies the whole of Leibniz’s philosophy: “True propositions not
asserting existence at particular times are necessary and analytic, but such as assert
existence at particular times are contingent and synthetic” (Russell 1900, Sect. 4).
This may have been the trigger for the introduction of the term ‘analytic’ in this
context. (For a focused discussion of ‘analytic’, see Russell 1900, Sects. 11–12).
Anyway, since the term ‘analytic’ has been introduced, I cannot go on without
some historical clarifications. First, I should remind the reader of Kant’s distinc-
tion between analytic and synthetic judgments or propositions (the latter being the
preferred term in accordance with the ‘linguistic turn’ in contemporary philosophy).
According to Kant’s distinction, analytic propositions merely serve to clarify knowl-
edge already attained, whereas synthetic propositions actually allow us to extend
knowledge, to find out new truths. Synthetic propositions are, therefore, epistemo-
logically far more important than analytic propositions. By the way, Kant sometimes
explained analytic propositions just as Russell does here, namely as propositions
5 Analysis of Segment II: Discussion of Copleston’s … 121

which it is self-contradictory to deny. Take, for example, the clarification I under-

took in my comment to 11.3.1, when I asserted the equivalence of the negative of the
Copleston Predicate with the concept of an external reason:

[∼ (x contains in x the reason for the existence of x)] ≡ (x has a reason external to x)

This equivalence is an analytic proposition: if the left member of the equivalence

is true of something, then the right member of the equivalence is also true of that.
We agreed that the left member of the equivalence is true of Copleston, and more
generally of all human beings. Therefore, on pain of contradicting ourselves, we have
to say that the right member of the equivalence is also true of Copleston and of all
human beings. But we have not learnt anything new by this. We have only clarified
what we meant.
Now, 12.4 goes beyond what mathematical logic demands. In order to understand,
I must complete the Kantian story which is the background here. The readers will soon
see that what may appear at first sight as a historical digression will not only illuminate
the meaning behind Russell’s first response to Copleston but also something of great
import for the following.
In Kant’s philosophy, the analytic-synthetic distinction was paired with another
one: the distinction between a priori and a posteriori propositions, the former being
independent of any experience and the latter empirical. This double distinction allows
Kant to refer to ordinary empirical propositions, as synthetic a posteriori. Those
propositions extend knowledge by means of experience, as based ultimately on
sense perception. Kant then separates synthetic a posteriori propositions from those
propositions which he believed were at the basis of all solid knowledge, namely the
synthetic a priori propositions. For Kant, in fact, synthetic a posteriori propositions
only worked in extending knowledge thanks to synthetic a priori ones. How this
is supposed to work, is the purpose of the whole argument in the Critique of Pure
Reason, but one upshot of this is that both analytic propositions and synthetic a priori
propositions are necessarily true.
Therefore, when Russell says that only analytic propositions are necessary or,
alternatively, that all necessary propositions are analytic, he is clearly rejecting the
Kantian double distinction. This distinction was a central bone of contention for
philosophers during the nineteenth century. The empiricist tendency, more at home
in Britain than in the European continent, was to reject Kant’s double distinction as
superfluous, by arguing that only empirical propositions could be synthetic and only
analytic propositions could be a priori. This was certainly Russell’s position as well.
Now, the first consequence of the rejection of Kant’s synthetic a priori propositions
is that the only necessity one can predicate of a proposition is analytic necessity, which
is what Russell is putting forward here. But there is a second, and philosophically
further-reaching, consequence, namely, that it raised the question of what the status
of those propositions is which kant deemed synthetic a priori. Let’s call this Q15.
122 5 Analysis of Segment II: Discussion of Copleston’s …

Now, we should not forget that Kant himself rejected some propositions which,
if they were true, could only be described as synthetic a priori, namely those of
traditional metaphysics, including such as appear in Copleston’s argument. So, he
was the first to suspect that a proposition that presents itself to the world as synthetic
a priori may be false or even meaningless. Kant’s detractors went just further than
him. For ease of reference, we can denote the set of all propositions than can be
described as synthetic a priori by the letter K. The by far most famous element of
K was the proposition that Every event has a cause. For Kant, all elements of K are
either metaphysical (and thus false or cognitively meaningless) or else universal and
necessary, in fact foundational for all solid knowledge attainable by human beings
(whose exemplars for Kant were Euclidean geometry and Newtonian mathematical
physics).
Kant’s detractors went much further than Kant and rejected some elements of K
which Kant wanted to retain, and not only those which he himself had attacked. For
instance, Russell rejected the proposition that Every event has a cause (see Russell
1912a; see later on Turns 32–68). However, there is wide agreement between Kant
and Russell as to the status of arguments of the sort Copleston has presented: they
both rejected them. It is not for nothing that Copleston will later on refer to Kant: he
knows that, as far as the traditional ‘proofs’ of God’s existence, Russell is more or
less a Kantian. Given this agreement, if Russell’s proposal at this point is correct, then
the question he’s raising is whether the copleston conclusion is analytic
or else whether it is untrue (or meaningless). Let’s call this Q16. Russell
was too much of a gentleman to put things in such a stark form, but this is what is
ultimately at stake, as it will transpire in the sequel.
For the time being, though, we need to keep in mind something else: if one
answers that the Copleston Conclusion is analytic, as Copleston would have to in
case he accepts Russell’s stricture that necessary = analytic, then the Copleston
Conclusion cannot be obtained, as Copleston says, from a premise which is factual
(cf. Russell 1900, Sect. 109). However, we established that Copleston derives it from
purely factual premises; and Copleston will confirm it in all explicitness later on (see
31.2). We shall soon see Russell pushing forward and asserting that no existential
proposition can be analytic (in which he is, again, thoroughly Kantian). But, for
the time being, he is, at least implicitly, rejecting the Copleston Conclusion as not
following from the premises, unless they are all analytic. If the premises are not
analytic, then the Copleston Conclusion does not follow.
In the background lurks Russell’s idea (taken from Kant) that cosmological argu-
ments, in spite of being described as a posteriori, always depend, for their conclu-
sions, on ontological arguments, which are a priori. This idea, however, will be not
be made explicit until later in the debate.
It is appropriate to insist that the sentence I am analysing here (‘all necessary
propositions are analytic’) goes further than Russell’s original logical stricture (‘only
propositions, not objects, can be necessary’). Consequently, the procedural starting
point associated to that development cannot be said to be purely logical. Some
eminent mathematical logicians, such as Frege, and possibly Hilbert and Gödel,
were far more friendly towards the idea of necessary and non-analytic propositions
5 Analysis of Segment II: Discussion of Copleston’s … 123

than Russell was. This is important because, later on, the question of the relation
between ‘modern logic’ and the whole of philosophy will be raised by Copleston (at
Turns 19 and 137). Nonetheless, the question of whether there are necessary
non- analytic propositions (Q17) will never be made in so many words in the
debate between Russell and Copleston, perhaps because it was considered too tech-
nical for a BBC broadcast, or perhaps because, for very different reasons, neither
Russell nor Copleston were full-blooded Kantians.
[12.5] I could only admit a necessary being if there were a being whose existence
it is self-contradictory to deny.
This sentence is just a consequence of Russell’s logico-philosophical procedural
starting point; but the way it is framed indicates that it should be considered as a
challenge to Copleston. To make this clear, we can formulate it as a question which
Copleston should answer: whether it is self- contradictory to deny that
god exists. Let’s call this Q18. It is clearer than ever that Russell is not really
interested in objecting to Copleston’s premises, or in rejecting the inference from
them to the conclusion, or in refuting the conclusion. He rather wants to propose a
method to analyse this or any other argument for God’s existence. In other words,
Russell’s move is back to the Opening Stage. If Russell’s method is applied, he
argues, then Copleston’s whole argument is much more deeply in trouble than if
Russell was rebutting a part of it.
[12.6] I should like to know whether you would accept Leibniz’s division of
propositions into truths of reason and truths of fact; the former, the truths of
reason, being necessary.
Fully in spirit with the proposed return to the Opening Stage, Russell is now issuing
a request for a starting point. In his Leibniz book, Russell had argued that Leibniz’s
philosophy was a wondrous logical construction, in which all parts had great consis-
tency. So, naturally, he suggests that, if Copleston wants to argue by using Leibniz,
then it is better to go to first principles, namely to Leibniz’s distinction between
vérités de raison and vérités de fait. Russell is challenging Copleston to accept or
reject Leibniz’s distinction. Should he accept it, then Russell could start arguing
from there. This would appear to be a material starting point. However, it is also part
of Russell’s general strategy to tackle any argument for God’s existence from the
standpoint of modern formal logic, so that the line dividing material from procedural
starting points is, in this instance, somewhat blurred.
Notice that, in 12.6, Russell is raising a question: Do you, Copleston, accept
Leibniz’s distinction between truths of reason and truths of fact? As opposed to the
procedural starting point described before (summarized as Q13), 12.6 is a direct
question, concerning a material starting point. This confirms that we are in the
Opening Stage. Call it Q19.
124 5 Analysis of Segment II: Discussion of Copleston’s …

Segment II, Turn 13 (Copleston) Opening Stage (Sub-Discussion, Material Starting

Point)

[13.1] Well, I certainly should not subscribe to what seems to be Leibniz’ idea
of truths of reason and truths of fact,
Copleston responds to Q19 by acknowledging that he is aware of the Leibnizian
distinction (“Well, …”) but making clear that he does not accept an independent
idea that is associated, in Leibniz’s thinking, with that distinction (“I should not
subscribe to…”). By thus objecting to that idea, he prevents Russell from using the
distinction as a starting point for further argument.
Consider, first, that the most relevant aspect of the distinction is that, as observed
by Russell, truths of reason are all analytic, whereas truths of fact are not. Let us
suppose that the distinction should be accepted by Copleston. In that case, Russell
could have argued that, for Leibniz, all truths of reason are necessary and the other
way around: all necessary propositions are truths of reason. It would follow, since
the Copleston Conclusion, as he himself puts it (11.5.3), is a necessary proposition,
that it then would have to be analytic. This admission is where Russell is leading
to; drawing such a consequence is the precise reason why Russell is bringing up
Leibniz’s distinction in the first place.
This strategic move on Russell’s part is not at all arbitrary, given that Copleston
himself say that he wants to put ‘the main weight on the metaphysical argument
based on Leibniz’s argument from contingency’ (see Turn 9). So, it was Copleston
who opened that flank. Since Copleston presumably wants to resist the consequence
of the analyticity of the Copleston Conclusion, he rejects at least one idea associated
with Leibniz’s distinction.
[13.2] since it would appear that for him there are in the long run only analytic
propositions. It would seem that for Leibniz truths of fact are ultimately
reducible to truths of reason, that is to say, to analytic propositions, at least
for an omniscient mind. I couldn’t agree with that.
We can see now that Copleston does not reject the Leibnizian distinction per se, but
the fact that Leibniz abandons it by accepting that in God’s mind all truths are truths
of reason and thus analytic propositions, so that the so-called truths of fact only exist
for us human beings.
Copleston’s move is an indication that at this point we have entered a sub-
discussion about Q19. To that question (let us put it so: whether one should
accept leibniz’s distinction between truths of reason and truths of
fact) Copleston is not quite accepting nor quite rejecting the distinction. Russell, by
the way, is himself not taking a stance in regard to that question either, but only, as
suggested above, preparing a possible ad hominem argumentation (i.e. an argument
based on Copleston’s own premises).
5 Analysis of Segment II: Discussion of Copleston’s … 125

[13.3] For one thing it would fail to meet the requirements of the experience of
freedom.
Copleston gives now an argument to reject Leibniz’s abandonment of his own distinc-
tion by reducing truths of fact to truths of reason from God’s perspective. Even though
this new argument has no role in the sequel, it may be interesting to describe roughly
what Copleston meant. Note that I am not reading the mind of Copleston but only
using the background information that gives us the history of philosophy.
For Leibniz, all propositions have the subject-predicate form (for an excellent
discussion of this Leibnizian doctrine, see Russell, 1900, Sect. 10). So, an analytic
proposition of the form ‘Copleston is…’ would contain in the predicate slot (signalled
by the ellipsis) each and every predicate that is true of Copleston, independently of
time. This entails that everything that Copleston will do from now on is already
contained in the predicate slot, never mind what Copleston himself believes about
the freedom of his choices. So, it could be argued that Copleston is not free; and
neither is any other human being. This is a consequence that has never been accepted
in Roman Catholic theology; so, Copleston cannot accept it, either.
The whole thing is, of course, a hoary old theological problem, which precedes,
by many centuries, anything that Leibniz said and wrote. We know that a very long
and complex debate has raged in theology and philosophy in Europe about how to
reconcile human free will with god’s omniscience (Q20). In any case,
Russell was aware of the difficulties inherent in the whole problem (see Russell,
1900, Sect. 118), so he left it well alone.
[13.4] I don’t want to uphold the whole philosophy of Leibniz. I’ve made use of
his argument from contingent to necessary being, basing the argument on the
principle of sufficient reason, simply because it seems to me a brief and clear
formulation of what is, in my opinion, the fundamental metaphysical argument
for God’s existence.
Like some philosophers do, Copleston wants to keep only one little piece taken from
Leibniz’s edifice whilst rejecting the rest.
This was probably not much to the liking of Russell, who thought that the main
merit of Leibniz’s philosophy was his overall coherence, so that this kind of selective
picking was out of order. But Russell will not go into that, probably because of the
general public, who would be lost in a discussion on the technical details of the
doctrine of an old German philosopher.
However that may be, Copleston knows that he must then justify why he is bringing
Leibniz into the debate. For that purposes, he adduces a reason of comfort: by using
Leibniz’s principle of sufficient reason, he can present the cosmological argument
with more brevity and clearness than he could do if he had used a different starting
point. This is fine as far as it gets, but notice that (1) the argument presented by
Copleston is neither brief nor entirely clear, as we have seen; (2) in that argument
Copleston does not name the principle of sufficient reason, thus preventing Russell
from inquiring about it, as he does later on, at Turn 32; (3) Copleston firmly asserts for
the first time that this kind of argument is the fundamental one for God’s existence,
126 5 Analysis of Segment II: Discussion of Copleston’s …

thereby devaluing other possible arguments. Later on, Copleston will go as far as to
say that it is the only argument that proves God’s existence, whereas the other two
he will use do not have this demonstrative character (see Turn 95).
In sum, Copleston responds to Russell’s proposed material starting point by
remaining comfortably within his specialty—the history of philosophy. This is a
field that is as far from the interests of a lay public as one could possibly imagine.

Segment II, Turn 14 (Russell) Opening Stage (Sub-Discussion, Material Starting

Point)

[14.1] But, to my mind, a necessary proposition has got to be analytic. I don’t

see what else it can mean.
Russell is unwilling to go into the history of philosophy and prefers to remain at
the level of the main logical (and methodological) point raised by him. Letting the
material starting point put forward at 12.6 (Leibniz’s distinction), he returns to the
procedural starting point, which is in any case far more consequential, argumen-
tatively speaking; but he expresses it as though it was merely a material starting
point.
Notice that the second sentence of 14.1 raises a new question: what the phrase
‘necessary proposition’ can mean if not ‘analytic’. Call it Q21 and notice
that it is further specification of Q11 and Q12. It is Q21 that really interests Russell,
so that the sub-discussion on Q19 is abandoned and a new one opened.
Now, as I explained in my comment to 12.4, Kant maintained that, if N is the set of
all necessary propositions, then some elements of N, but not all of them, are analytic.
Some other elements of N are synthetic a priori; and these are, according to Kant,
the basic building blocks which make possible for us to have empirical knowledge in
the first place. So, this is what ‘a necessary proposition’ could mean from a Kantian
point of view: if not analytic, then it is synthetic a priori. If Copleston were a Kantian,
then this is what he could have replied to Russell’s Q21. But he is not.
[14.2] And analytic propositions are always complex and logically somewhat
late.
Russell proceeds to explain to Copleston, and to the lay public who is listening to
the broadcast, what makes analytic propositions so peculiar.
One feature of analytic propositions is that they are always complex. If we keep
to the subject-predicate form, this means that either the subject or the predicate will
never be simple in an analytic proposition. To take again the example mentioned in
my comment on 13.3, try to imagine the number of words needed to fill the predicate
slot in the full analytic proposition ‘Copleston is …’, when you assume, with Leibniz,
that everything that was, is, and will be the case about Copleston is at once predicated
of him. Such a predicate will have thousands upon thousands of specifications. This
Leibnizian example is, of course, an extreme case; but in any analytic proposition of
the subject-predicate form you will be analysing the concept in the subject slot, so
5 Analysis of Segment II: Discussion of Copleston’s … 127

there will be some measure of complexity in the proposition. Notice that, if we accept
the ontological argument, then the proposition that God exists would be analytic, for
existence would be seen as part of the concept of God. This is the reason why Russell
wants to go in this direction, as will become clearer later on (at Turns 20 and 26).
The other feature of analytic propositions is that they are “logically somewhat
late”. Remember that Russell is a thoroughly analytic philosopher. All his life he
kept analysing concepts, propositions, arguments, theories; that is what he did, over
and over again. (Principia Mathematica, published in 1910, is just his most spec-
tacular achievement.) But analysis is not what people do first. It is a culturally late
problem; and many other problems must be solved before people start analysing.
Notice that this is, in embryo, already a powerful argument against the idea that the
proposition that God exists is analytic; for people believed in God’s existence long
before the concept of God was analysed, and in particular long before the ontological
argument was invented. However, Russell will not pursue that particular argument
(from lateness, as it were) in the sequel. Given that Copleston does not believe, either,
that the proposition that God exists is analytic, it would in any case be a waste of
time to follow that particular argumentative route.
[14.3] Rational animals are animals is an analytic proposition; but a proposition
such as This is an animal can never be analytic.3 In fact, all the propositions that
can be analytic are somewhat late in the build-up of propositions.
Russell illustrates, and thereby reinforces, what he asserted in 14.2. He uses homely
examples, as is appropriate for a lay audience. The statement This is an animal
sounds like something a child might be told and then repeat, to herself and other
children, in order to fix, in her memory, the new noun. In all of this, there is absolutely
no analysis of the concept of animal or any other concept. As opposed to this, the
concept of a human being as a rational animal already implies an analysis, that must
have come pretty late in human history; and with its help the proposition is formed:
Rational animals are animals, a proposition that could not be formed before the
analysis of the concept of a human being was carried out. The proposition thus
serves to remind us that, in spite of being rational, we humans are also animals. So,
its sole purpose is clarification.
Notice that the proposition Rational animals are animals is somewhat
unorthodox as an example for an analytic proposition. Analytic propositions, at
least those of the subject-predicate form, are usually conceived in such a way that
the predicate is longer than the subject, for it gives an analysis of the concept under-
lying the subject. This is not the case with Russell’s example, but I hope that the
explanation given above is satisfactory.
There is, however, something still stranger in 14.3. He says: “a proposition such
as This is an animal can never be analytic”. Readers would be excused if they
thought that it is not allowed to say that a given proposition can be analytic (or not),
since any proposition is (or is not) analytic. To say that a proposition can be analytic

3At this point, it is possible to detect, in the recording, an attempt at interruption by Copleston,
which Russell overrides by continuing to talk. See Chap. 1, footnote 8.
128 5 Analysis of Segment II: Discussion of Copleston’s …

seems to entail that a proposition is analytic from a certain point of view or at a

certain time and not analytic from another point of view or at another time. Is that
what Russell meant? The strangeness of his wording is only enhanced by the last
sentence: ‘all the propositions that can be analytic are somewhat late in the build-up
of propositions’. Russell must have meant, surely, that ‘all the propositions that are
analytic are somewhat late in the build-up of propositions’?
One possible explanation for the strange wording may be that Russell is here
conflating the realm of words and sentences with the realm of concepts and proposi-
tions. (He was often accused of that, starting with Frege.) Take the sentence ‘Salt is
sodium chloride’. Is this statement analytic or synthetic? The first chemist that discov-
ered that salt is sodium chloride clearly expressed, with his statement, a synthetic
proposition; for the knowledge was utterly new then. Yet at some later point in time
this knowledge had become so old, well-established, and widely known, that a student
of high-school chemistry or a lexicographer— let alone an analytic philosopher such
as Kripke (1972) or Putnam (1975)—might refer to it as an analytic proposition, in
which the predicate just expressed the very meaning of the subject. In that case, the
analytic-synthetic status of a sentence could be said to vary according to time, space,
speaker, and context. And insofar, one may be justified in saying things like ‘this
sentence can never be analytic’ or ‘all sentences that can be analytic are somewhat
late in the build-up of sentences’.
Of course, the niceties of such distinctions as between words and concepts or
sentences and propositions would have been of no interest to the lay audience. So,
we should not be troubled by this matter anymore. Let us only note that in this
turn Russell presents himself as offering a material starting point about the status of
certain propositions.

Segment II, Turn 15 (Copleston) Opening Stage (Sub-Discussion, Material Starting

Point)

[15.1] Take the proposition If there is a contingent being, then there is a necessary
being. I consider that that proposition, hypothetically expressed, is a necessary
proposition.
In order to respond to Q21 (what the phrase ‘necessary proposition’ can
mean if not ‘analytic’), Copleston first presents an example of an analytic
proposition. His example is a conditional, for this is what he means by the phrase
‘hypothetically expressed’. I shall call it the Copleston Conditional in view of the
importance it has in his thinking.
The Copleston Conditional is the ‘hypothetical expression’ of an argument of the
form: There is a contingent being; therefore, there is a necessary being. This is in
fact the most compressed form of Copleston’s metaphysical argument.
Although Copleston affirms that the Copleston Conditional is a necessary propo-
sition, the fact that we are still in the Opening Stage, amidst the sub-discussion on
Q21, we could also interpret Copleston’s argumentative move as raising the question
5 Analysis of Segment II: Discussion of Copleston’s … 129

whether the copleston conditional is a necessary proposition. Call

this Q22.
There is one problem, though. It is not quite clear whether the Copleston Condi-
tional is, as a whole, a necessary proposition, or if only its consequent is. If the
former, then the full expression of the Copleston Conditional should be: It is neces-
sary that, if there is a contingent being, then there is a necessary being. It would
be then a modal conditional. If the latter, then the correct formulation should be: If
there is a contingent being, then it is necessary that there is a necessary being. My
impression is that Copleston in fact conflates three logically different propositions:
(1) (∃x) C(x) ⊃ (∃y) N(y)
(2)  [(∃x) C(x) ⊃ (∃y) N(y)]
(3) (∃x) C(x) ⊃  (∃y) N(y)
This conflation seems on a par with the conflation between the deductive and
the modal meaning of ‘must’, which I pointed out in some of Copleston’s earlier
interventions (11.3.1, 11.5.3).
[15.2] If you are going to call every necessary proposition an analytic proposition,
then, in order to avoid a dispute in terminology, I would agree to call it analytic,
though I don’t consider it a tautological proposition.
Copleston presents himself as conciliatory. He says he does not want to partake of
a verbal dispute about Q21. Therefore, he accepts that every necessary proposition
is analytic and the other way around, as long as ‘analytic’ is not taken to mean the
same as ‘tautological’. Even though the set N (of necessary propositions) should be
coextensive with the set A (comprising all analytic propositions), Copleston proposes
that at least some elements of A (or N) are not tautological.
It is not quite clear what Copleston means by ‘tautological’ (for a discussion of the
chequered history of the word, see Dreben 1991). Let us consider three alternatives:
(1) ‘Tautological’ means (à la Wittgenstein) ‘logically true’. This is unlikely, in
view of the fact that Copleston consistently resents and rejects the intrusion
of mathematical logic in philosophy, as becomes clear later on (Turns 19, 55,
137).
(2) ‘Tautological’ means ‘true by definition’. There would seem to be something
to it, because we usually define ‘contingent’ as ‘not necessary’. On the other
hand, ‘not necessary’ seems to mean, in this definition, the same as ‘not caused’;
and this meaning is not clearly modal, as Copleston wishes it to be. So, this
meaning would not work for Copleston, either.
(3) ‘Tautological’ means ‘repetitive’. This may be the best interpretation, for
Copleston wants the Copleston Conditional to be informative, in fact its purpose
is to inform us of nothing less than God’s existence, starting from the humble
recognition that we and everything else in the world have not made ourselves.
130 5 Analysis of Segment II: Discussion of Copleston’s …

[15.3] But the proposition is a necessary proposition only on the supposition that
there is a contingent being.
This confirms the analysis presented at the end of my comment on 15.1. Copleston
wavers between considering the Copleston Conditional itself a necessary proposition
and considering that only its consequent as necessary. In the latter case, though, the
necessity would not be a modal property of the consequent, but a way of indicating
that it follows from the antecedent.
[15.4] That there is a contingent being actually existing has to be discovered by
experience, and the proposition that there is a contingent being is certainly not
an analytic proposition, though once you know, I should maintain, that there is
a contingent being, it follows of necessity that there is a necessary being.
A further confirmation that Copleston has expressed himself a bit sloppily. In any
case, his meaning is that you have two propositions, one empirical and obtained from
observation (‘there is a contingent being’), the other philosophical and obtained by
reasoning from the first proposition. For Copleston, then, there is an argument, his
metaphysical argument, which says: There is a contingent being; therefore, there
is a necessary being. This argument can be expressed ‘hypothetically’, i.e. as the
Copleston Conditional, and when we do express the argument in this way, then the
Copleston Conditional is not analytic in the sense of repetitive.
However, if this is what Copleston meant, then he is not justified in calling the
Copleston Conditional a necessary proposition in Russell’s sense; and the whole
sub-discussion is irrelevant to Q21, namely whether necessary propositions
could be non- analytic. So, the sub-discussion is not getting any traction at this
point, and it is difficult to pin the disagreement down.
Note that Copleston, in this turn, has not yet become aware of the deeper logical
point, namely the procedural starting point of taking mathematical logic as the
implicit arbiter of the questions at issue.

Segment II, Turn 16 (Russell) Opening Stage (Sub-Discussion, Procedural Starting

Point)

[16.1] The difficulty of this argument is that I don’t admit the idea of a necessary
being.
Russell abandons the sub-discussion on the analyticity of necessary propositions
(Q21), for he recognizes that Copleston is not so much talking about propositions as
about arguments. So, without further ado about the nature and form of the argument,
Russell insists that the Copleston Conclusion attributes necessity to an individual
and not to a proposition. This is what he won’t admit.
We have to pay attention, though. Russell’s point of order (to use a legal metaphor)
is not purely formal: it is not just about the fact that Copleston uses the phrase
‘necessary being’ (which, incidentally, he did not). One could dismiss the phrase as
such and say something like: It is necessary that a certain being exists, and still
5 Analysis of Segment II: Discussion of Copleston’s … 131

be open to Russell’s point of order. The issue at bottom is rather that existence (of
anything whatever) can never be necessary.
This allows us to bring the question of the deductive and the modal meaning of
‘must’ back to the discussion. Russell has nothing against the deductive meaning,
what is sometimes called ‘logical necessity’. In fact, he is famous for defining math-
ematics as “the subject in which we never know what we are talking about, nor
whether what we are saying is true” (Russell 1901; here quoted from its reprint in
Russell 1917, p. 75). The latter part of this witticism is a humorous way of saying that
all propositions in pure (not applied) mathematics are conditional: we never know
whether what we are saying is true, for it is only true if something else is true. If the
antecedents of a mathematical conditional (in the last instance, the axioms) are true,
then the consequent is also true—true by necessity, as we might add with Aristotle.
And this necessity, logical necessity, is all the necessity that one needs.4
This, however, raises a problem. There are many consequents in pure mathematics
which are existential, e.g. the proposition that certain algebraic equations have at least
one solution. So, existential propositions asserted with logical necessity, as following
from other propositions all the way back to axioms, seem to be unobjectionable for
Russell. If that is all the necessity that Copleston can claim, and if he does claim it,
then there should not be any problem for Russell accepting the Copleston Conclusion,
once the conflation of logical and modal necessity is dispelled.
[16.2] and I don’t admit that there is any particular meaning in calling other
beings contingent.
Russell’s protest against ‘contingent being’ is slightly more complicated than was the
case for ‘necessary being’, since the word ‘contingent’ has several meanings. Before
going into that, however, consider that, if there is no necessary being, then all beings
are contingent; but, if all beings are contingent, then the concept of ‘contingent’
has no use whatever. For concepts are there to make distinctions, and this concept,
according to Russell, would not make any. In that sense, we can argue that the
corresponding term has no meaning. This is borne out by Russell’s remark at Turn
138: “there [is no] meaning in calling things ‘contingent’ because there isn’t anything
else they could be”.
Notice that, if ‘contingent’ is supposed to work by separating contingent beings
from necessary beings, then Russell’s argument would entail that Copleston’s argu-
ment is not so much incorrect (or invalid or unsound) as complete nonsense. A further,
perhaps less welcome, consequence, however, would be that the word ‘necessary’
is as meaningless as ‘contingent’. But this is difficult to conciliate with Russell’s
assertion that the word ‘necessary’ only applies to propositions and that it applies to
all analytic ones. If so, then the word ‘necessary’ is not meaningless after all.

4 In this, Russell seems to agree with Leibniz, of whom he quotes the following (1900, §13): “As
for the eternal truths, we must observe that at bottom they are all conditional, and say in fact: Such
a thing posited, such another thing is.” The quotation comes from the Nouveaux Essais, 446: «Pour
ce qui est des vérités éternelles, il faut observer que dans le fond elles sont toutes conditionnelles et
disent en effet: telle chose posée, telle autre chose est.» Notice the similarity with Aristotle’s own
phrasing (Topics, 100a25-27; Prior Analytics, 24b18-20).
132 5 Analysis of Segment II: Discussion of Copleston’s …

I believe the confusion comes from the fact that the words used by Copleston have
different uses that are incorrectly brought together:
(1) If the word ‘necessary’ is taken to mean ‘logically necessary’, then it only
applies to propositions, namely to those that have been deduced from other
propositions. The meaning of ‘it is necessary that p’ is then identical to ‘p has
been correctly deduced from q’ (where q represents either an already proved
theorem or an axiom). One can also say that the deduced proposition p is
grounded on those propositions from which p has been deduced; and those
propositions can be called the ‘logical ground’ of p. In this sense, the word
‘contingent’ is not at all opposed to the word ‘necessary’ (unless we want to
say that all axioms are contingent).
(2) If the word ‘contingent’ means as much as ‘caused’, then it has nothing to
do with propositions but rather with states of affairs. Thus, a hot stone is
contingently hot because the sun made it hot. One particular case of state of
affairs that needs a cause is the coming-to-be of a human being (Copleston
owes his existence to his parents); and so does the fact that a human being,
having come to be, continues being for a while (Copleston owes his continuing
existence to air, food, and so on). Under this meaning, ‘necessary’ would
mean something which does not owe its existence to anything beyond itself. If
Copleston is right, such a being would also be the cause of everything else.
As long as these meanings are conflated, only confusion may result. We can
perhaps say that the trouble lies in what we may call, à la Wittgenstein, the grammar
of ‘necessary’ and ‘contingent’. If ‘contingent’ means ‘caused’, it has a particular
grammar, which is not the same for Russell as for Copleston, as we shall see in Turns
32–68). If ‘necessary’ means ‘logically necessary’ (as Russell insists), then it again
has a particular grammar, the grammar of mathematical logic, and it is not opposed to
‘contingent’ in the sense of ‘caused’, whose grammar is completely different. Finally,
if ‘necessary’ means ‘having its reason or explanation in itself’ (as Copleston would
have it), then Russell would reject the concept out of hand and, with the concept,
also the theological grammar within which alone it would make sense.
[16.3] These phrases don’t, for me, have a significance, except within a logic that
I reject.
This last remark confirms the idea that, all along, Russell was trying to propose math-
ematical logic, whose use in philosophy he had strikingly and successfully pioneered
(see especially Russell 1905), as a method by which one may make progress on the
question or questions under debate and perhaps even go some way towards resolving
the disagreement (even if that end result was, in the nature of things, not to be
expected). There is indeed an old adage, of somewhat obscure scholastic origin,
according to which no one can argue with or against someone who denies the prin-
ciples (contra principia negantem non est disputandum). Nelson (1904) rejected
this adage arguing that, on the contrary, philosophical argumentation is precisely all
about principles. Popper was in this respect, as in many others, a disciple of Nelson
(Popper 1976; see also Popper 1979, Appendix). And in fact, Russell and Copleston
5 Analysis of Segment II: Discussion of Copleston’s … 133

will argue about principles, starting with Turn 17, when Copleston follows Russell’s
cue (“within a logic that I reject”). However, in the end all they seem to be be able
to do is summarize their positions (Turns 137–138).
Although not wholly explicit, Russell is actually presenting an argument that
could be reconstructed as follows:
Premise 1. In the only acceptable logic, the word ‘necessary’ can only be applied to analytic
propositions.
Premise 2. In the only acceptable logic, the phrase ‘contingent being’ lacks meaning.
Conclusion. The phrases ‘necessary being’ and ‘contingent being’ are meaningless.

The two premises have been explicitly asserted by Russell, with the sole exception
of ‘in the only acceptable logic’; but this insertion is implied by the phrase “within a
logic that I reject”). As for the conclusion, Russell has also made it explicit by saying
“these phrases… don’t have significance”.
It is true that Russell only says “I don’t admit” (16.1, 16.2) and “for me” (16.3),
but it would be a mistake to think that he is just expressing a personal opinion,
unsupported by reasons, for he says explicitly that the offending phrases would only
have meaning or significance “within a logic I reject”. Even here, Russell’s phrasing
is personal (“I reject”), but it is not just a personal opinion, for this cannot be squared
with the previous assertion that ‘necessary’ can only be predicated of propositions
and indeed only of analytic propositions (Turn 14).
In fact, Russell is here raising a very important question: what the ‘logic’
might be within which phrases like ‘necessary being’ and ‘contingent
being’ have a meaning. Call this Q23. This question will hover over the rest of
the debate, although it will never be tackled in a straightforward manner. In any
case, Copleston will use Russell’s remark in order to suggest that he, Copleston, may
be right after all, only within a different ‘logic’; and that Russell is not justified in
imposing his logic. Nonetheless, neither discussant will take the time to explain to
each other, or to the audience, what that other ‘logic’ consists in, or what the nature,
function, or status of a ‘logic’ is when a philosophical question such as the existence
of God (Q1) is being discussed. It is possible that the forbiddingly technical character
of the issues involved may have deterred them from going into logical details.
It could, of course, be argued a contrario that the word ‘logic’ as used here by
Russell has not its technical meaning. Might Russell be using that word in a colloquial
sense? I think we can reject this interpretation as inconsistent with the facts. First of
all, Russell’s distinction between individual objects and propositions is technical, no
matter how lightly he refers to it (see especially 18.1). We have seen that Copleston
is much less severe in his use of terms; but, and this is the second reason, he will
immediately react to Russell’s use of the word ‘logic’, recognizing it as a technical
term, starting with Turn 17, and then he will refer both to the variety of formal logical
systems (19.1) and to the theory of descriptions (21.1–21.2), both highly technical
aspects of mathematical logic. He may have lost the audience by so doing, but it
was Russell who opened that particular door and Copleston is only following him,
as a good discussion partner should. It is true that Russell will consistently shy away
from almost all technical details in his replies, but this fact does not invalidate the
interpretation offered here, and it can be easily accounted in some other way.
134 5 Analysis of Segment II: Discussion of Copleston’s …

With this turn, Russell has made it clear where he stands: there is, at bottom, a
silent battle between a logic that he rejects (although he does not specify what logic
that is), and a logic that he considers crucial to decide on the question of a necessary
being. What is a stake is a procedural starting point of enormous significance.

Segment II, Turn 17 (Copleston) Opening Stage (Sub-Discussion, Procedural Starting

Point)

[17.1] Do you mean that you reject these terms because they won’t fit in with
what is called ‘modern logic’?
Finally, it dawns upon Copleston that this was the procedural starting point that
Russell had put forward at Turn 12. Notice that, although 17.1 is a formulation of
Q23, Copleston is thereby indicating that he finds ‘modern logic’ a questionable
procedural starting point.
The adjective ‘modern’ in the phrase ‘modern logic’ is Copleston’s not Russell’s.
It has an aftertaste of dismissal, which will become more pronounced as the debate
progresses.5 In contrast, logic is, for Russell, a unified subject, inaugurated by Aris-
totle but then progressing very slowly, and at times more or less stagnant, until the
advent of mathematical logic in the nineteenth century. So, even if one could qualify
the development made possible by the mathematical tools adumbrated by Leibniz,
initiated by Boole, and then extended by Frege, Peirce, Schröder, Peano, Whitehead,
and of course Russell himself, as ‘modern’, nonetheless the subject is not considered
by Russell to be different from what Aristotle began.
As for the noun ‘logic’ in the phrase ‘modern logic’, I would say that it is not
clear what Copleston meant by it. He never contributed anything to the subject, either
before or after the debate. Did he perhaps believe, with Kant, that “since Aristotle,
logic has not had to retrace any of his steps… [nor] been able to advance a single step,
and is thus to all appearance a closed and completed doctrine” (Kant 1787, Preface,
2nd paragraph)? This is unlikely, given that his next intervention indicate that he was
informed about modern logic. It is a shame that Q23 is never quite tackled properly
in the debate, so that we could find out what exactly he meant by ‘logic’.

Segment II, Turn 18 (Russell) Opening Stage (Sub-Discussion, Procedural Starting

Point)

[18.1] Well, I can’t find anything that they could mean. The word ‘necessary’,
it seems to me, is a useless word, except as applied to analytic propositions, not
to things.
Although again with a ‘personal’ tone (“I can’t find”), Russell emphasizes his specific
technical logical point: the word ‘necessary’ has only one possible use. From the

5 I am not making this up. Copleston was a Jesuit, i.e. member of an order that was born to defend
the Roman Catholic faith against the reformers and protestants. The historical process involved is
what has given the Church in Rome its reputation of conservatism ever since. For a conservative
thinker, all innovations of ‘modernity’ are suspect.
5 Analysis of Segment II: Discussion of Copleston’s … 135

analysis of the discussion so far, it seems that we can say that this use has one
function: to separate the logical truths from all other truths. Russell also believed
that logical truths were foundational to the whole of mathematics; in fact, that logic,
as the body of all logical truths, was the first and most basic of all mathematical
disciplines. This position, called ‘logicism’, was and is very controversial; but the
logic itself is not.
Of course, it could be argued that, apart from the logical necessity of the propo-
sitions of logic, there may be other necessary propositions. This question has been
the subject of many philosophical discussions in the decades after this debate (see
Kment 2012). If one accepts these alleged necessary yet non-logical propositions,
then one could raise the question whether they are analytic; but this is not a question
which Copleston either raises or tackles. Modal logic is, of course, a notoriously
difficult and controversial subject; and certainly not a subject which would count
with Russell’s sympathy; yet it seems that something like modal logic may lurk
behind what Copleston is trying to say, even if rather clumsily and unsuccessfully.
In any case, it is worth observing that Russell actively avoids responding directly
and unambiguously to the request formulated by Copleston (“do you mean that you
reject these terms because they won’t fit in with what is called ‘modern logic’?”). That
request is Copleston’s attempt at getting at what he has justifiedly understood to be the
ground for Russell’s rejection of the phrase ‘necessary being’, the distinction between
‘necessary being’ and ‘contingent being’, and all that they imply. That ground is a
less than perfectly explicit procedural starting point suggested by Russell. We see
here the beginning of a curious struggle: Copleston wants Russell to admit that he
is setting up mathematical logic as an arbiter of the debate, in order to question its
authority; and Russell wants to present his rejection of one phrase as a particular
thing, with no further theoretical ramifications. This struggle, that is played out until
Turn 25 may be connected with Russell’s desire to keep on a nontechnical level in
view of the audience.

Segment II, Turn 19 (Copleston) Opening Stage (Sub-Discussion, Procedural Starting

Point)

[19.1] In the first place, what do you mean by “modern logic?” As far as I know,
there are somewhat differing systems.
Although the phrase ‘modern logic’ did not come from Russell himself and in fact
is singularly inappropriate with regard to Russell’s conception of the subject (see
comment to 17.1), the fact that Russell spoke of a ‘logic’ that he rejects, is enough
for Copleston to raise the question what kind of logic russell’s is (Q24).
We are still in the Opening Stage, but now it is clear that the sub-discussion is
not just about the particular questions Q11-12, Q14, Q17-19, Q21-22, but squarely
about the much more general Q13, namely, whether mathematical logic may
be used to tackle philosophical questions such as the proofs for
god’s existence (and perhaps also about Q23: what the ‘logic’ might be
136 5 Analysis of Segment II: Discussion of Copleston’s …

within which phrases like ‘necessary being’ and ‘contingent’ being

may have a meaning). This becomes the crucial pivot of the sub-discussion.
If Copleston had waited for an answer, a material and procedural starting point may
have been agreed on; but what he does instead is to throw doubt on Russell’s logic.
19.1 is the first such doubt. Copleston seems aware that, apart from what has come to
be referred to as ‘classical logic’, namely, two-valued extensional propositional and
predicate logic with identity, other formal systems had been proposed after Principia
Mathematica and before the debate took place, mainly modal logic, three-valued
logic, and intuitionistic logic (starting with Lewis 1914, 1918; Łukasiéwicz 1920;
Post 1921; Heyting 1930). Although phrased with due modesty (“as far as I know”),
Copleston’s intervention has an undertone of challenge: if there is not one logic (as
Russell might be taken to presume) but several logics, then why should we simply
accept Russell’s logic as the standard? Such a challenge would have been stronger,
if Copleston had adduced at least one such non-classical logic that might make his
complaint more specific. However, he does not do that but rests content with planting
a seed of doubt, especially perhaps in the ears of the audience, concerning Russell’s
assertions about the meaninglessness of ‘necessary’ and ‘contingent’ as applied to
‘beings’.
[19.2] In the second place, not all modern logicians surely would admit the
meaninglessness of metaphysics.
Copleston continues to throw doubt on Russell’s assertion, but again not in the form
of argument. Now, if in 19.1 the assumption that there may be more than one logic
was warranted by Russell’s talk of “a logic that I reject”, the assumption on which
Copleston’s doubt in 19.2 rests is that, for Russell, all of metaphysics is meaningless.
This assumption is utterly without basis and will very soon be rejected by Russell
(see Turn 20).
[19.3] We both know, at any rate, one very eminent modern thinker whose
knowledge of modern logic was profound, but who certainly did not think
that metaphysics are meaningless or, in particular, that the problem of God
is meaningless.
Copleston expresses here a third doubt, which was perhaps designed to be especially
painful to Russell. For 19.3 is clearly alluding to Alfred North Whitehead, the co-
author, with Russell, of Principia Mathematica, still the canonical presentation of
‘classical logic’ at the time of the debate. In his Process and Reality, originally
presented as the Gifford Lectures on Natural Theology for the years 1927–28 and
published as a book in 1929, both metaphysics and God’s existence had pride of
place. Whether a BBC audience would have been aware of this allusion, is rather
unlikely. But Russell could not have misunderstood both the allusion and Copleston’s
intention in bringing it up. Of course, Copleston’s allusion to Whitehead shares the
dubious assumption of Russell’s rejection of all metaphysics as meaningless that I
have just commented upon; so, it misses its target completely.
5 Analysis of Segment II: Discussion of Copleston’s … 137

Now, if we tried to construct an argument from this allusion to Whitehead, it

would look like this:
Premise 1. Not all modern logicians admit the meaninglessness of metaphysics.
Premise 2. Russell’s objection implies the meaninglessness of metaphysics.
Conclusion. Russell’s objection is not as strong as it may appear, given Russell’s authority
as a logician.

The rest of this turn could then be interpreted as so many arguments constructed in
the same way. I will not carry out such an interpretation but shall come back to the
question at the end of my last comment (19.12).
[19.4] Again, even if all modern logicians held that metaphysical terms are
meaningless, it would not follow that they were right.
Now we have Copleston’s fourth doubt, marked by ‘again’. 19.4 shares the general
assumption that Russell rejects all metaphysical terms; but it makes a further, and
even more dubious assumption, namely that Russell’s objection is based on the
authority of ‘modern’ logicians. It has another interesting feature. In terms of the old
rhetorical theory of stáseis (Dieter 1940; Nadeau 1959, Kennedy 1965, pp. 306–314),
Copleston is adopting the last resort of throwing doubt on ‘modern logic’ as such,
i.e. of recusing the authority of ‘modern logic’ as the court which can adjudicate in
matters such as the metaphysical argument presented by Copleston.
[19.5] The proposition that metaphysical terms are meaningless seems to me to
be a proposition based on an assumed philosophy.
Copleston now goes further in his attempt at questioning Russell’s point of order
by means of a fifth doubt. For Copleston now asserts that Russell is hiding a whole
philosophy under the pretext of using a philosophy-neutral ‘logic’; it is such an
unacceptable manoeuvre on the part of Russell which leads to his alleged rejection
of all metaphysical terms, part of which is the rejection of the phrase ‘necessary being’
in Copleston’s argument. This fifth, and most expansive, doubt expresses perhaps
with utmost clarity what is at bottom the trouble for Copleston: no philosophy can
be based on logic. To understand how deeply this argument cuts, we must remember
that Russell admired Leibniz precisely because he thought that Leibniz had had the
courage to formulate certain logical propositions (see Russell, 1900, §4) and then
base his whole ‘monadistic’ metaphysics on those propositions alone.6 In later years,
Russell would try to formulate, in several versions, his own relational metaphysics on

6 It may be useful to remind readers that, according to Russell’s thinking, all metaphysics prior to the
discovery of the logic of relations (more broadly, the classical predicate calculus with identity) was
either monistic or monadistic. The same logic underlies both kinds of metaphysics, namely a logic
for which all propositions are of the subject-predicate form, so that relational predicates are reduced
to non-relational ones. In a monistic metaphysics, such as Hegel’s, there is only one subject (the
totality), whereas in a monadistic metaphysics, such as Leibniz’s, there is a large, perhaps infinite
number, of subjects (the ‘monads’). For Russell then, it is the discovery of the calculus of relations
and the device of quantification which helps get rid of both kinds of traditional metaphysics and
allow for a new kind, exemplified by the logical atomism of Wittgenstein (1921) and Russell (1918,
138 5 Analysis of Segment II: Discussion of Copleston’s …

the basis of the mathematical logic he pioneered. To this extent, Copleston certainly
had a point. In the sequel, this question (whether it is appropriate to base
metaphysics on logic or not, Q25) will re-appear at 19.7, but it will at that
point unfortunately become confused with a completely different question (namely,
whether it is appropriate to identify metaphysics with logic or not,
Q26). Russell will later try to dispel the confusion.
[19.6] The dogmatic position behind it seems to be this: What will not go into my
machine is non-existent, or it is meaningless, or it is the expression of emotion.
Copleston now tries to go even further, and put forward a sixth doubt, by asserting
that Russell’s rejection of the phrase ‘necessary being’ is dogmatic and entails one of
three options: (a) the thing referred to by that phrase does not exist, or (b) the phrase
itself is meaningless, or (c) it is just an expression of emotion.
Copleston seems to be conflating different things.
As to (a), Russell has never asserted that God does not exist, for he said his position
is that of an agnostic. He probably thinks that the phrase ‘necessary being’, if that is
what God is, cannot refer to anything at all, but he has not yet asserted that.
As to (b), what Russell has asserted is that the phrase ‘necessary being’ is either
to be re-interpreted in terms of an analytic proposition or else meaningless, and not
that it is meaningless.
As to (c), Russell has never asserted that the phrase ‘necessary being’, or any other
metaphysical term, is just an expression of emotion. Copleston seems to attribute an
emotivism in metaphysics which belongs rather more to Alfred Ayer than to Russell.
It is true that Russell sometimes suspected that certain emotions were associated
with a given metaphysics, but he never believed that the content of metaphysics was
itself emotional. If anything, Russell rather believed the contrary, that the contents
of metaphysics was purely rational, because based on certain logical principles,
either explicit (like in Leibniz or himself) or implicit (in perhaps the majority of
metaphysicians).
By the way, the word ‘dogmatic’ is not here used in the philosophical technical
sense, as the opposite of ‘sceptic’, but rather in the more colloquial sense of ‘believing
things without argument’. What Russell, according to Copleston, believes without
presenting an argument in its favour is that logic, i.e. Russell’s logic, decides what
exists and what has meaning (we’d better leave emotion out of the discussion).
Copleston could hardly think that Russell had no arguments to back up his use of
logic to decide questions of existence and meaning; and he could not expect Russell
to go into the difficult questions involved here. So, it is a bit hard to understand why
Copleston charges Russell with dogmatism, except perhaps to score a point before
the audience. On the other hand, the underlying questions concerning the relationship
between logic and metaphysics are genuine questions.

1924; later reprinted in Russell 1972). In one sense, therefore, it is true that, if not the whole of
philosophy, at least the whole of metaphysics is based on logic.
5 Analysis of Segment II: Discussion of Copleston’s … 139

[19.7] I am simply trying to point out that anybody who says that a particular
system of modern logic is the sole criterion of meaning is saying something that
is over-dogmatic; he is dogmatically insisting that a part of philosophy is the
whole of philosophy.
This intervention seems at first sight to be merely a repetition of the sixth doubt; yet
there is a difference. The assertion here is not that Russell’s philosophy is based upon
his logic but something utterly different, namely that philosophy (or metaphysics)
is identical with logic (see last comment on 19.5). This is certainly unwarranted, as
Russell will say later on.7
[19.8] After all, a contingent being is a being which has not in itself the complete
reason for its existence; that’s what I mean by a contingent being.
Copleston now pursues a different route. He ceases to express doubts about Russell’s
point of order with regard to the phrases ‘necessary being’ and ‘contingent being’.
Instead he tries to explain what the meaning is of those phrases.
Copleston starts with a definition: ‘x is a contingent being if and only if x is a being
which has not in x the complete reason for the existence of x’. Here, we recognise
the Copleston Predicate yet again. There is a slight difference in phrasing, however,
for at this point, instead of saying that something has not in itself alone the reason
for its existence (11.2.2), Copleston says that it has not in itself the complete reason
for its existence. This change in phrasing will later on become quite important, when
Copleston and Russell will discuss the meaning of ‘sufficient reason’ as ‘adequate
explanation’ and as ‘total explanation’ (see Turns 32–35).
[19.9] You know, as well as I do, that the existence of neither of us can be
explained without reference to something or somebody outside us, our parents,
for example.
Here, the concept of explanation, that had appeared at 11.4.4–11.4.5, is re-introduced.
Otherwise, this is exactly the same sort of clarification we saw at 11.1.2, although at
that point Copleston talked of dependence, presumably causal dependence. The two
concepts are not at all identical, and this will cause some trouble later on.

7 Talk about the ‘parts’ of philosophy harks back to the classical Greek division into logic, physics,
and ethics. If we substitute ‘metaphysics’ for ‘physics’ (a title that was to be reserved to the modern
science by that name) and if, knowing that the classical division was more or less retained by Kant,
we take what Copleston says in this light, then he would be asserting that Russell’s logic contains
both metaphysics and ethics. To this Russell would probably reply, first, that ethics is a wholly
different field, about which he would have a couple of things to say later on, but to which logic has
little if anything to contribute; and secondly, that logic does not contain or englobe metaphysics
but only forces it to take into account the relational predicates which the tradition did not know
how to handle. Why didn’t Russell say this simply and without further ado, whereas he was willing
to talk about the difference between objects and propositions, and then to struggle with Copleston
about names and descriptions, both quite difficult and technical topics? I’m afraid I don’t know the
answer to that question. All I know is that, if he was trying to avoid technicalities in view of the
audience, then he was curiously selective in his avoidance.
140 5 Analysis of Segment II: Discussion of Copleston’s …

Table 5.1 Equivalent readings of ‘contingent’ and ‘necessary’ according to Copleston

I II
General term Contingent Necessary
Reading in terms of Has not in itself the complete Has in itself the complete
reasons reason for its existence reason for its existence
Modal reading Can be non-existent Must be existent
Causal reading Causally dependent on Causally independent
other beings from other beings
Reading in terms of Explained by other beings Not explained by anything else
explanation

[19.10] A necessary being, on the other hand, means a being that must and
cannot not exist.
Copleston produces now a second definition: ‘x is a necessary being if and only if x
is a being that must and cannot not exist’. In this way, the two metaphysical terms,
‘contingent’ and ‘necessary’, seem to cover all ground, at least if we accept that all
the expressions contained in each of the columns of Table 5.1 are equivalent.
Copleston expresses himself in such a way that he implies this equivalence.
He alternates between reasons, modality, causality, and explanation, as if the
corresponding concepts had the same extension, if not perhaps the same content.
[19.11] You may say that there is no such being, but you will find it hard to
convince me that you do not understand the terms I am using.
Given than Russell says that he cannot find anything else that ‘necessary’ could
mean apart from logical necessity (14.1), Copleston proposes that, by attending
to the explicit definitions given (and the equivalences in Table 5.1), Russell will
now understand Copleston’s terms and, therefore, will not be able to declare them
meaningless. This seems to be a meta-argument against Russell’s point of order:
Premise 1 (Definition). A necessary being means a being that must and cannot not exist
Premise 2. Russell may assert that there is no necessary being, but he must surely understand
the definition.
Conclusion. Russell cannot say that the proposition There is a necessary being is
meaningless.

Copleston, of course, admits that Russell is entitled to say that there is no necessary
being as defined, for this is a separate question (namely, Q1: whether god exists
or not, once ‘God’ is identified with ‘necessary being’). Now, does this meta-
argument constitute an answer to Q21 (what the phrase ‘necessary propo-
sition’ can mean if not ‘analytic’)? If Russell’s challenge is to explain what
a necessary proposition could be if not an analytic one, then Copleston does not
seem to satisfy that challenge. In fact, the only known philosophical answer would
be to say, with Kant, that a non-analytic necessary proposition is a synthetic a priori
5 Analysis of Segment II: Discussion of Copleston’s … 141

proposition; but for some reason, best known to himself, Copleston does not want to
give that answer.
[19.12] If you do not understand them, then how can you be entitled to say that
such a being does not exist, if that is what you do say?
Copleston dares Russell to answer straight away the question: whether there is
or is not a necessary being. Let’s call it Q27. He expresses his suspicion that
Russell will answer Q27 in the negative. However, if Russell does exactly that, then he
will be admitting that he understands the phrase ‘necessary being’, and it would then
be a bit harder to maintain that the phrase is meaningless. This is a straightforward
challenge. If this was poker, Copleston’s move would be a case of calling Russell’s
bluff.
Notice, however, that the dialectical trap being set by Copleston is not well
designed. For Russell has never said that he does not understand that phrase, only that
he does not see how ‘necessary’ means anything else than ‘analytic’, which means
for him, according to my earlier comments, ‘logically necessary’.
There is nevertheless a source of possible misunderstanding at this point. When
Russell says that ‘necessary’ can only mean ‘analytic’, then he is at the same time
asserting that these two predicates only apply to propositions and not to individual
objects (or ‘beings’, as Copleston likes to say). But mark: this is not a problem of form
or expression. We have already seen that the propositional function ‘x is a necessary
being’ means as much as ‘it is necessary that x exists’. This latter, Russell says, can
only be interpreted as ‘it is analytic that x exists’, which for Russell means: ‘it is
logically necessary that x exists’. But once we follow through with that, we leave
the realm of cosmological arguments (e.g. Copleston argument from contingency,
presented in Turn 11) and gone over to the realm of ontological arguments. This
fundamental point will come back in due time (see 20.2).
Summing up this whole turn, Copleston has been offering several reasons to doubt
Russell’s procedural starting point and, with it, the authority of what he dubs ‘modern
logic’ to arbitrate in philosophical questions.8 One could reformulate those reasons
as full-fledged arguments in formal detail; but this would be too heavy-handed and
would necessitate supplying doubtful ‘missing premises’. It is enough to see that
Copleston is not yet quite rejecting the tribunal of ‘modern logic’ but seeding doubts
about it, probably for the benefit of the listeners.

8 This is not the first, or the only, time that this question was raised. Perhaps the most famous relevant
document is Carnap (1931), in which he castigated Heidegger for producing logical nonsense. For his
part, Heidegger had rejected, in university lectures held in 1928, the authority of mathematical logic
for metaphysics by proposing an interpretation of Leibniz different from that of Russell (1900) and
Couturat (1901). It is rather unlikely that Carnap was aware of those lectures, since they were only
published posthumously half a century later (see Heidegger, 1978). For an interesting discussion of
the Carnap-Heidegger relationship, readers may want to consult Friedman (2000, Chapter 2), even
if he, strangely enough, does not refer to the lectures, by then already published and translated.
142 5 Analysis of Segment II: Discussion of Copleston’s …

Segment II, Turn 20 (Russell) Opening Stage (Sub-Discussion, Procedural and

Material Starting Points)

[20.1] Well, there are points here that I don’t propose to go into at length. I don’t
maintain the meaninglessness of metaphysics in general at all. I maintain the
meaninglessness of certain particular terms—not on any general ground, but
simply because I have not been able to see an interpretation of those particular
terms. It is not a general dogma; it is a particular thing. But those points I will
leave for the moment.
Russell does not stoop to reply to Copleston’s charge, that his point of order was
based on authority (19.1–19.4). As for Copleston’s main assumption, namely, that
Russell considered all metaphysics without meaning (19.2–19.6), he rejects it out of
hand. He never said that. In fact, nor was there anything in Russell’s many writings
that would authorize such an assumption. As I said before, one might accuse Russell
of basing his different metaphysical proposals on mathematical logic, and even of
producing “a different system of [metaphysics] every few years” (Broad 1924, p. 79),
but most surely not of considering metaphysics generally meaningless. So, quite
rightly, Russell expresses his intention to leave the whole matter of the first part of
Copleston’s intervention (19.1–19.6) to one side, although one issue (concerning the
relation between logic and philosophy, voiced in 19.7) will recur later at the end of
the debate (see Turns 137–138).
[20.2] And I will say that what you have been saying brings us back, it seems to
me, to the ontological argument—that there is a being whose essence involves
existence, so that his existence is analytic.
As to the second part of Copleston’s intervention (19.8–19.12), Russell finally makes
fully explicit what he has been all along driving at: if ‘necessary’ only applies to
propositions, and then only to analytic propositions, and if an analytic proposition
is one that is logically necessary, i.e. logically derived from other true propositions,
then the only way in which one can make sense of any argument to the effect that it
is necessary for something to exist, is the ontological argument. For only then would
the existential proposition asserting that God exists be necessary in the only sense
that Russell admits, namely, analytically or logically true. Russell defended already
in his Leibniz book that, if there should be a valid argument for God’s existence, then
that argument could only be the ontological argument, and that all other arguments
presuppose it. (Russell learnt that from Kant, of course, but enriched Kant’s insight
with his logical acumen.)
[20.3] That seems to me to be impossible, and it raises, of course, the question
what one means by existence, and as to this, I think a subject named can never
be significantly said to exist but only a subject described; and that existence, in
fact, quite definitely is not a predicate.
The problem with the ontological argument is, of course, that, even if we should
consider it valid, it is yet not sound. In other words, one might admit that, from
5 Analysis of Segment II: Discussion of Copleston’s … 143

(a) the definition of God as a being possessing all perfections, and from (b) the
assumption that existence is a perfection, it would follow that God exists; and yet at
the same time one could object either to (a) or to (b).
In a very terse manner, Russell sketches two arguments against the soundness of
the ontological proof. The first one he owes to Kant: existence is not a predicate;
this would be an objection to (b). The other argument is Russell’s own: the theory
of descriptions says that, in order to be able to say that something exists, we cannot
use a name, such as ‘God’, but we need a description. This seems to be not an
objection to one of the premises of the ontological argument but rather a refutation
of its conclusion: ‘God exists’ is something that cannot be asserted.
By this argumentative reply, Russell is raising three new questions: what the
meaning of ‘existence’ is (Q28), more specifically whether existence is a
predicate (Q29), and whether the theory of descriptions can be used
to analyse and refute the ontological proof of god’s existence
(Q30). Copleston, as I have said and as will soon transpire, was no Kantian, so he
rejected Kant’s and Russell’s answers to Q28 and Q29. But he will only tackle Q30
in his reply.
Before going into that, it is important to remark that the whole discussion, from
Turn 11 (where Copleston presents his argument) up to this turn, must have been
increasingly hard to understand for a lay audience. Consider all the things that the
lay audience would have had to know in order to understand the last bits of Russell’s
intervention: (1) that one has to distinguish between constants and variables; (2) that
a well-formed formula in predicate logic with quantifiers needs variables to be bound
by the quantifier; (3) that constants cannot be bound by a quantifier; and, finally, (4)
that names are constants whereas descriptions are at bottom well-formed formulas
with both an existential and a universal quantifier. Assuming all that, it is clear that,
at least in classical predicate logic, one is not allowed to say what the ontological
argument, or indeed what any argument for God’s existence for that matter, wants to
say. In a sense, Russell seems to try to shift the conversation to a different subject: not
so much whether the phrase ‘necessary being’ and the sentence ‘there is a necessary
being’ has or lacks meaning, but rather whether Copleston’s metaphysical argument
can only work through the intervention of the ontological argument (which was
Kant’s conclusion in the Critique of Pure Reason as well as Russell’s conclusion
in his Leibniz book). If Copleston would ever admit that, then Russell would know
how to demolish Copleston’s argument. In other words, Russell is trying to set a
dialectical trap for his partner.
The proposed shift towards the theory of descriptions will be taken up by Cople-
ston in the following turn. But before going into that, and although this goes beyond
what the discussants actually say, it may be of interest to note that, in modal quanti-
fied logic, a subject that was just being born at the time of the debate (Barcan 1946a,
1946b, 1947) and has remained controversial ever since, one is allowed to say that
It is necessary for something to exist. In fact, a few decades after the debate the
first attempts at giving a proof, in quantified modal logic, for the existence of God
would be given (Plantinga 1978). The proof is, of course, highly controversial; but
this is another matter. I mention it only because it bears on Q23 (what the ‘logic’
144 5 Analysis of Segment II: Discussion of Copleston’s …

might be within which phrases like ‘necessary being’ and ‘contingent

being’ have or lack a meaning).
Note that Russell again avoids being fully explicit about the procedural starting
point (Q23 through Q26) and keeps replacing it by a merely material starting point,
which is besides presented as a purely personal opinion (“that seems to me…, and
as to this, I think…”). That material starting point is, of course, there, but it obscures
the more important procedural one.

Segment II, Turn 21 (Copleston) Opening Stage (Procedural and Material Starting
Points)

[21.1] Well, you say, I believe, that it is bad grammar, or rather bad syntax, to
say, for example, T. S. Eliot exists.
Copleston shows himself unwilling to go into Russell’s trap, by accepting the idea
that an ontological argument (in which God’s existence is analytic) lurks behind
Copleston’s metaphysical argument. He is aware all the same that he has to go into
the logical details outlined by Russell when he distinguished between names and
descriptions, or as he, trying perhaps to be more intelligible to the audience, prefers
to say: ‘things named’ vs. ‘things described’.
Copleston knows something about the theory of descriptions. Referring to the
celebrated poet Thomas Stearns Eliot, who still lived at the time of the debate and
who was in fact a friend of Russell’s albeit also a critic of his religious writings,
Copleston uses the name of the poet to refer to him. According to what Russell
himself has said, the sentence ‘T. S. Eliot exists’ is badly constructed from the point
of view of mathematical logic. In this sentence, the verb ‘to exist’ is used as a
predicate, and a name, ‘T. S. Eliot’, is used as its subject.
The point of this is not, of course, that we cannot say, in English, that T. S. Eliot
exists, for plainly we can, and the sentence makes sense. The point is rather that, in
the severely regimented syntax of classical predicate logic, it would be impossible
to form the equivalent sentence. In that syntax, existence is expressed by means of
a special operator, called an existential quantifier, whose use needs a variable and a
predicate:

(1) (∃x)F(x)
There is an individual x such that F is true of x

Here, the letter ‘F’ is a placeholder for an arbitrary predicate, e.g. ‘smokes’. So,
you can say things like There is someone who smokes, or, more briefly, Someone
smokes. If you want to predicate ‘F’ of T. S. Eliot, you can do so:

(2) F (a)
F is true of a
5 Analysis of Segment II: Discussion of Copleston’s … 145

Here, ‘a’ is a constant; and one can predicate something of a constant as well as of
a variable; yet only the variable, but not the constant, can accompany the quantifier.
The difference between constants and variables may be explained by saying that
a constant refers to a definite (existing) individual whereas a variable has no such
reference. (Russell used to say that a variable has an indefinite or an ambiguous
reference.)
In (1), it is said that the existential quantifier ‘∃’ binds the variable x, so that ‘F’
can be predicated of x. So, basically, you can say T. S. Eliot smokes and Someone
smokes, but you cannot say that T. S. Eliot exists. Now, can one say that Someone
exists? There are two answers to that question. One can say that it is not possible,
because the existential quantifier accompanied by the variable does not constitute a
sentence. This would probably be Russell’s position if one had asked him. The other
possibility, devised by the Finnish logician and philosopher Jaakko Hintikka much
later (1962, p. 6; see also 1966), would be to take advantage that constants do refer
to definite existing individuals and to write:
(3) (∃x)x = a
There is an individual x which is the same individual as a

The precise translation of the partial symbol ‘(∃x)’ is, of course, ‘there is at least
one individual x such that’. For ease of expression, I shall use ‘there is an’ instead
of ‘there is at least one’.
Before this turn ends, we shall see that Copleston’s humble reciting of the lesson
of the theory of descriptions is strategic on his part. He is in fact preparing a trap for
Russell. This is part of Copleston’s dialectical prowess, which again is part of his
scholastic training.
Notice that Copleston here is accepting, for the sake of argument, the procedural
starting point suggested by Russell. He is arguing with the resources of Russell’s
theory of descriptions, which belongs to his mathematical logic.
[21.2] One ought to say, for example, The author of “Murder in the Cathedral”
exists.
Copleston produces now a description of T. S. Eliot, as opposed to a name. The
description uses the fact that “Murder in the Cathedral” is a play written by T. S.
Eliot as its sole author. The sentence given by Copleston as an example would be
logically parsed by Russell as follows:
(4) (∃x)[F(x) & (∀y)[F(y) ⊃ y = x]]
There is an individual x such F(x) and, for every individual y, if F(y), then y
and x are the same individual

Here, the predicate ‘F’ stands for ‘is the author of “Murder in the Cathedral”’.
By this intervention, Copleston shows that he understands at some level Russell’s
theory of descriptions: although descriptions look superficially like names, they can
only be understood as part of quantified sentences like (4). We can now reconstruct
146 5 Analysis of Segment II: Discussion of Copleston’s …

what Russell meant at 20.3: existence is not a predicate but a quantifier which binds
a variable within a sentence that has a predicate. In this way, ‘a subject named can
never be significantly said to exist but only a subject described’, as Russell said.
[21.3] Are you going to say that the proposition The cause of the world exists is
without meaning?
Copleston now constructs a sentence according to the strictures of the theory of
descriptions. Notice that this English sentence still looks on the surface as though
the verb ‘to exist’ were a predicate. However, if you put Copleston’s sentence into
the symbols of the predicate calculus, this sentence would look exactly like (4).
Substituting the schematic predicate letter ‘F’ appropriately, one would have: ‘There
is an individual x such that x is cause-of-the-world and, for every individual y, if y is
cause-of-the-world, then y and x are the same individual’, or more simply: ‘There is
only one individual x such x is cause-of-the-world’. (The hyphens are there to make
clear that ‘cause-of-the-world’ is now a predicate and not part of a description.)
By this adroitly constructed sentence, Copleston is offering a material starting
point, expressed as a request, which Russell can hardly reject. The question expressed
by that request is: ‘Does the proposition The cause of the world exists lack meaning?’
Let’s call it Q31.
[21.4] You may say that the world has no cause; but I fail to see how you can
say that the proposition that The cause of the world exists is meaningless.
Copleston pushes his point more deeply by answering Q31 in Russell’s place, making
it clear that Russell cannot answer it in any other way on his own principles. Copleston
certainly allows Russell to deny the proposition expressed by his sentence, but he
does not allow him to reject it as lacking meaning. This is a very smart argumentative
move on Copleston’s part. Through it he has managed, at least temporarily, to disarm
Russell’s rejection of the phrase ‘necessary being’.
Note that 21.4 is constructed in parallel to 19.11. On that occasion, the trap for
Russell was not well designed, but this time around it is indeed.
[21.5] Put it in the form of a question: Has the world a cause? or Does a cause of
the world exist? Most people surely would understand the question, even if they
don’t agree about the answer.
Pressing his advantage, Copleston further elaborates his point. If Russell can deny
but not reject Copleston’s sentence, then neither can he reject the question: ‘Has the
world a cause?’ or ‘Does a cause of the world exist?’ This question, or rather the
question whether the cause of the world exists, is the one which matters
to Copleston. Let us call it Q32, and let us note that it has, in Copleston’s mind, the
same meaning as Q1 (whether god exists). So, we are back to the beginning of
the debate.
Notice that it is not important how Russell answers Q32 (or Q1 for that matter, as
long as the questions are considered to have the same meaning). What is important,
as material starting point, is to concede that Q32 is meaningful. (Whether Q1 is also
meaningful depends, of course on whether it is synonymous with Q32.)
5 Analysis of Segment II: Discussion of Copleston’s … 147

The dialectical trap is now set. It is a very neat case of argumentation ad hominem,
starting from Russell’s own premises.9

Segment II, Turn 22 (Russell) Opening Stage (Material Starting Point)

[22.1] Well, certainly, the question Does the cause of the world exist? is a question
that has meaning.
Russell is forced to admit, at this point, that Copleston’s Q32 has meaning. He
cannot refuse what follows from his own premises. (Note, incidentally, that Russell
substituted the definite article “the” for Copleston’s indefinite “a”.)
In due course, at Turn 68, we shall see that Russell will withdraw the concession
offered here and declare Q32 meaningless. At that point, Russell’s argument will
have changed from an objection based on the theory of descriptions (see 20.3 and
comment) to an objection based on the appropriate application of the concept of
causality, another of Russell’s favourite topics.
[22.2] But, if you say: Yes, God is the cause of the world, where you’re using “God”
as a proper name, then God exists will not be a statement that has meaning. That
is the position that I’m maintaining.
This is a rather poor attempt at saving face on the part of Russell. One can assert
that God exists is meaningless from the point of view of the theory of descriptions,
because it is. However, once the meaningfulness of The cause of the world exists is
admitted, nothing in mathematical logic forbids defining ‘God’ as ‘the cause of the
world’ (as in fact was done by Copleston and conceded by Russell earlier); and once
this definition is admitted, it does not matter whether we say ‘God exists’ or not.
[22.3] Because, therefore, it will follow that it cannot be an analytic proposition,
ever, to say that this or that exists.
This is a pretty obscure step in Russell’s reasoning. One can hear in the recording
that he is trying to regroup.
First, consider the proposition which is here presented as a conclusion, namely It
cannot be an analytic proposition to say that this or that exists.
One could certainly argue that, in mathematical logic, one cannot say that ‘this’ or
‘that’ exists, insofar as the demonstrative pronouns are supposed to behave in a way
logically similar to proper names, i.e. as constants. If an analytic proposition means
for Russell a logically necessary proposition, then Russell’s conclusion would mean
as much as The proposition that this or that exists is never logically necessary,
although one would need an argument for that. That argument might be that all
existential propositions are synthetic, perhaps even empirical, knowing Russell’s

9When I say ‘argumentation ad hominem’, I of course mean the original dialectical sense of the
phrase, namely, as an argument which uses premises conceded by one’s interlocutor, and not in the
modern distorsion of that original sense, as a ‘fallacy’ (Nuchelmans 1993; see also van Eemeren
and Grootendorst 1993).
148 5 Analysis of Segment II: Discussion of Copleston’s …

empiricist proclivities. But the fact of the matter is that Russell gives us no argument,
empiricist or otherwise; or rather, he says that it follows from what he has said before
(“Because, therefore, …”, a somewhat weird construction). But it is very difficult to
see how it follows.
As I said, my impression is that, having fallen into Copleston’s very able dialectical
trap, Russell is trying to get out of the trap, but he is not succeeding.
[22.4] For example, suppose you take as your subject “the existent round
square”, it would look like an analytic proposition that the existent round square
exists, but it doesn’t exist.
Now, Russell attempts a different argument, based on his old chestnut against
Meinong (Russell 1905, p. 483). Take the phrase “the existent round square”. In
it appears the pseudo-predicate ‘to exist’, so that it would be self-contradictory to
deny that the existent round square exists. Since we know, however, that the said
phrase already contains a contradiction (“round square”), we also know that the
things that phrase purports to denote does not exist.
All this is perfectly sound, but it is not clear what it has to do with Q32. It is
clearly not a refuting counterexample, for, if I can show you one allegedly analytical
existential proposition that is false, it surely does not follow that all such proposi-
tions are also false. Another possibility is that Russell is suggesting that ‘God’ is a
self-contradictory term, like “the existent round square”. Such an assertion has been
made of the Christian God in philosophical theology (see e.g. Kenny 1979). However,
Russell does not say that. In fact, his argument purports to apply to all analytic propo-
sitions, not only to those whose subject is self-contradictory. No analytic proposition
can be existential—that is what Russell is asserting. But this is not an argument in
favour of that.
My earlier impression is thus confirmed. Russell is trying to gain time whilst
he finds what to respond to Copleston, but he signally fails. Nonetheless, Russell is
offering a material starting point in the shape of his “existent round square” example.
Copleston will, chivalrously enough, take it up.

Segment II, Turn 23 (Copleston) Opening Stage (Material Starting Point)

[23.1] No, it doesn’t. Then surely you can’t say it doesn’t exist unless you have
a conception of what existence is. As to the phrase, “existent round square,” I
should say that it has no meaning at all.
Copleston kindly acquiesces to the last bit of Russell’s intervention. Still, he tries to
gain some traction for his side of the debate by adding that, in his example, Russell
is using existence as a concept, therefore a predicate. The problem, in the sentence
put forward by Russell, lies in the subject not the predicate. It is that subject which
lacks meaning.
This intervention is a beautiful example of chivalry. Copleston has seen that
Russell has not been able to respond to the dialectical trap ably set in Turn 21.
5 Analysis of Segment II: Discussion of Copleston’s … 149

However, Copleston declines to attack his ‘wounded’ adversary, but instead offers
him a hand.

Segment II, Turn 24 (Russell) Opening Stage (Material Starting Point)

[24.1] I quite agree. Then I should say the same thing in another context in
reference to a “necessary being”.
Russell gratefully takes the hand that Copleston gave him and is back on his feet. If
there is a problem with the phrase ‘existent round square’, then the same problem
occurs in the phrase ‘necessary being’. Both would be meaningless.
Notice that we are back with Q23 (what the ‘logic’ might be within
which phrases like ‘necessary being’ and ‘contingent being’ have a
meaning). As he made it already clear at 16.3, Russell does not admit any such logic.

Segment II, Turn 25 (Copleston) Concluding Stage (Meta-Argumentation)

[25.1] Well, we seem to have arrived at an impasse. To say that a necessary being
is a being that must exist and cannot not exist has for me a definite meaning.
For you it has no meaning.
Copleston summarizes the situation by suggesting that a material starting point, in the
shape of the proposition The phrase ‘necessary being’ is meaningful, is presupposed
in his argument but is denied by Russell. Under the circumstance, no discussion can
take place. Copleston seems to argue as follows:
Premise 1. To say that a necessary being is a being that must exist and cannot not exist has
for Copleston a definite meaning.
Premise 2. To say that a necessary being is a being that must exist and cannot not exist has
for Russell no meaning.
Conclusion. Copleston and Russell should stop the debate on God’s existence.

In fact, Copleston is trying to bring the discussion to a conclusion by giving up on

the debate as such. He does not even try to argue that the phrase ‘necessary being’
has meaning. If Russell had given no argument for his position, then Copleston’s
renunciation might be to some extent justified, or at least understandable. However,
Russell has given an argument, namely, (1) that ‘necessary’ only applies to proposi-
tions, (2) that it only applies to analytic propositions, and (3) that, even if one would
convert the phrase into a sentence such as ‘It is necessary that something exists’,
the corresponding proposition would have to be analytic, something which it cannot
be (for no analytic proposition can assert the existence of anything). Copleston has
not given any counterargument to Russell’s argument. He has only asserted that the
Copleston Conclusion is not analytic. Copleston is thus giving up too soon.
150 5 Analysis of Segment II: Discussion of Copleston’s …

Segment II, Turn 26 (Russell) Opening Stage (Meta-Argumentation)

[26.1] Well, we can press the point a little, I think. A being that must exist
and cannot not exist, would surely, according to you, be a being whose essence
involves existence.
Russell will not allow Copleston to escape so easily. He suggests returning to the
Opening Stage by pressing his old point, but this time without using the (to a Jesuit
odious) phrase ‘ontological argument’. He does that by taking up Copleston’s concept
of a necessary being. Notice that, for that purpose, Russell uses a reinforced version
of Copleston’s own phrase, “a being that must exist and cannot not exist” (deployed
in the Copleston Conclusion at 11.5.3, and again at 19.10). But, says Russell, that
phrase can only refer to “a being whose essence involves existence”. Notice that
the latter phrase belongs to any discussion of the ontological proof, for this is an
argument that tries to prove God’s existence from the very concept of God, from a
definition of ‘God’, or from God’s ‘essence’. The underlying meta-argument could
be represented as follows:
Premise. Copleston conceives of God as a being that must exist and cannot not exist.
Conclusion 1. Copleston conceives of God as a being whose essence involves existence.
Conclusion 2. Copleston is, whether he wants or not, trying to prove God’s existence by an
ontological argument.

Russell is, therefore, offering a material starting point, as implied by “surely”; and a
starting point which is presented as included in Copleston’s own premises (“according
to you”). Nonetheless, this material starting point is craftily designed to trap Cople-
ston, again, into admitting that the phrase ‘necessary being’ can only make sense
within a version of the ontological argument.
Notice that, at this point, a new question has been raised: whether the meta-
physical argument at bottom hides the ontological argument. Let us
call this Q33 and note that this is the question that matters most to Russell, as is also
clear from his critique of Leibniz’s proofs of God’s existence.

Segment II, Turn 27 (Copleston) Opening Stage (Meta-Argumentation)

[27.1] Yes, a being the essence of which is to exist. But I should not be willing to
argue the existence of God simply from the idea of His essence, because I don’t
think we have any clear intuition of God’s essence as yet. I think we have to
argue from the world of experience to God.
Copleston deftly escapes from Russell’s trap. He does admit that his idea of God is as
Russell says: he indeed conceives of God as “a being whose essence is to exist”. To
that extent, Copleston’s accepts Russell’s “according to you”. Nonetheless, Copleston
rejects Russell’s implication, namely, that his metaphysical argument is ontological
at bottom. He even gives a reason for not using the ontological argument, namely,
that human beings have no clear intuition of God’s essence. To quote Saint Paul,
5 Analysis of Segment II: Discussion of Copleston’s … 151

“now [while on earth] we see through a glass, darkly; but then [once in heaven] we
shall see face to face” (I Cor. 13:12). Given that human being are now in the dark
and suspect but do not clearly see that God’s essence is to exist, they must restrict
themselves to a humbler argument, starting from ‘the world of experience’. The
underlying meta-argument is then as follows:
Premise. Human beings have no clear intuition of God’s essence.
Conclusion 1. Human beings cannot prove God’s existence from God’s essence.
Conclusion 2. Human beings can only try to prove God’s existence from the world of
experience.

Copleston thus remains within the orthodox Thomistic tradition of rejecting Anselm’s
ontological argument.

Segment II, Turn 28 (Russell) Opening Stage (Meta-Argumentation)

[28.1] Yes, I quite see the distinction. But, at the same time, for a being with
sufficient knowledge, it would be true to say: Here is this being whose essence
involves existence.
Russell, of course, accepts that the two modes of argumentation are different but at
the same time tries to salvage his point concerning the ontological argument. He goes
about it in a somewhat haphazard way. The issue is not that a being with sufficient
knowledge is such that it would be true to say, of it, that his essence involves existence;
but rather that a being with such epistemological superiority over us human beings
would know that God’s essence involves His existence. Russell seems to suggest that,
if we can say that, then the proposition known by God can be believed by human
beings.
Once the confused expression is amended, it seems that Russell is presenting two
arguments. The first, first-order argument would be like this:
(1) Premise. A being with sufficient knowledge knows that God’s essence involves
existence.
Conclusion. God exists.

The fact that, for Copleston, the premise in (1) is true should make him see that he
is all the time, albeit unawares, pursuing the ontological argument. Now, if he does,
then a meta-argument is implied, which goes like this:
(2) Premise 1. Copleston admits upholding argument (1).
Premise 2. Argument (1) is equivalent to the ontological argument.
Conclusion. Copleston upholds the ontological argument.

The meta-argument (2) contains the first-order argument (1). Together, they
constitute, again, a dialectical trap.
152 5 Analysis of Segment II: Discussion of Copleston’s …

Segment II, Turn 29 (Copleston) Opening Stage (Meta-Argumentation)

[29.1] Yes, certainly: if anybody saw God, he would see that God must exist.
Copleston sees what Russell is driving at, and in a sense accepts the point, but he puts
it hypothetically (“if…”). By doing that, he deftly excludes himself from actually
upholding the ontological argument. The meta-argument seems to be as follows:
Premise 1. Only a being who has a clear intuition of God would be able to uphold the
ontological argument.
Premise 2. Copleston is no such being.
Conclusion. Copleston does not, and could not, uphold the ontological argument.

In this way, Copleston accepts Russell’s starting point, i.e. the proposition that A
being with sufficient knowledge knows that God’s essence involves existence, used
as a premise in argument (1) at 28.1, yet rejects the inference in that argument, and
so rejects meta-argument (2).

Segment II, Turn 30 (Russell) Opening Stage (Meta-Argumentation)

[30.1] So that, I mean, there is a being whose essence involves existence, although
we don’t know that essence. We only know there is such a being.
Russell tries to prevent Copleston from escaping the consequences of his own admis-
sion (the conditional in 29.1). He even goes as far as to generalize it by actually
including himself in it through the first-person plural (‘we’), although of course only
Copleston is the only intended target of the meta-argument in 29. This could be spelt
out as follows:
Premise. Copleston knows that there is a being whose essence involves existence, even if
he has no clear intuition of that essence.
Conclusion. Copleston upholds the ontological argument.

What makes this argument work seems to be the assumption of a variety of epistemic
closure, the highly controversial idea that, if we know that p, then we know all the
consequences of p (see e.g. Dretske vs. Hawthorne in Steup and Sosa, 2005, Chap. 1).

Segment II, Turn 31 (Copleston) Opening Stage (Meta-Argumentation)

[31.1] Yes, I should add we don’t know the essence a priori.

Copleston’s intervention at this point is perhaps the most subtle in the whole debate.
First, he secures a starting point, taking advantage of the first-person plural general-
ization introduced, for rhetorical effect, by Russell in 30.1. It is an epistemological
starting point: the essence of God (with whose help Russell wants to derive God’s
5 Analysis of Segment II: Discussion of Copleston’s … 153

existence) is not known to us human beings a priori, never mind whether it may be
known to a superior being.
[31.2] It is only through a posteriori—through our experience of the world that
we come to a knowledge of the existence of that being.
This sentence, like Russell’s at 28.1, is a bit garbled. Copleston first starts to say
something like It is only through a posteriori [sc. knowledge], and then corrects
himself and says It is only through our experience of the world that we come to a
knowledge of the existence of that being, namely a being with such an essence. So,
it is by imperfect but possible thinking tools such as the metaphysical argument that
we can come to know that such a being as God exists (and not through impossible
thinking tools such as the sheer concept of God in the ontological argument).
Over twenty years before Kripke (1972), Copleston is here proposing something
like a necessary a posteriori proposition. We discover through experience that water
is H2 O, yet what we have discovered is an essence, something that is necessarily
true. Nobody could have a clear intuition beforehand into the essence of water; if
somebody could, a priori, see water for what it is, he would also see that water
must be H2 O. But we human beings do not. If Copleston is right, his metaphysical
argument is a similar voyage of discovery, at the end of which we find that God has
an essence that involves, or consist in, existing.
[31.3] And then one argues, the essence and existence must be identical.
It is only then, after the proposition of God’s existence has been secured through
the metaphysical argument that we can argue further, saying: Such a being exists,
therefore, its essence is to exist.
[31.4] Because if God’s essence and God’s existence were not identical, then
some sufficient reason for this existence would have to be found beyond God.
Copleston gives now an argument for asserting that, after going through the meta-
physical argument, God’s existence and God’s essence are identical, namely that, if
this were not the case, then it would be necessary to go on, beyond God, to find a
sufficient reason for His existence.
This intervention, to sum it all up, contains an impressive meta-argument, which
might perhaps be reconstructed as follows:
Premise 1. Copleston does not know God’s essence (as involving existence) a priori.
Premise 2. Copleston only knows that God’s essence involves existence through the
metaphysical argument, which is a posteriori.
Conclusion. Copleston can know that God’s essence involves existence without upholding
the ontological argument.

Through this meta-argument, Copleston has very effectively answered Q33 in the
negative—and so escaped the dialectical trap prepared in 26.1 and set in 28.1. Having
recognized that he has not won and cannot win the meta-argumentative struggle that
occupies Turns 26–31, Russell will never try again to argue in favour of an affirmative
154 5 Analysis of Segment II: Discussion of Copleston’s …

answer to Q33 (whether the metaphysical argument at bottom hides

the ontological argument or not). The rest of the discussion about the
metaphysical argument will, from now on, take a completely different direction.

Segment II, Turn 32 (Russell) Opening Stage (Material Starting Point)

[32.1] So, it all turns on this question of sufficient reason. And, I must say, you
haven’t defined “sufficient reason” in a way that I can understand. What do you
mean by “sufficient reason”? You don’t mean cause?
Russell abandons the line of argumentation he had been pursuing since the motion
to return to the Opening Stage at Turn 26, namely, to reduce Copleston’s meta-
physical argument to the ontological one. This implies leaving aside the question of
the meaning of the phrase ‘necessary being’ (which had proved a cul de sac, 25.1)
and instead questioning the meaning of the phrase ‘sufficient reason’, a Leibnizian
concept with which Russell was thoroughly familiar (see Russell 1900, Sect. 14).
Still, Russell knows that he cannot assume that Copleston shares all of Leibniz’s
insights and assumptions (Copleston explicitly said so at 13.4). But at least he can
suggest one of the meanings of ‘sufficient reason’: cause.
The request issued to Copleston corresponds to the general question whether
‘sufficient reason’ means the same as ‘cause’ or not. Let’s call this
Q34. Copleston’s answer will deliver a convenient material starting point to support
the ensuing discussion. Although the second sentence of this intervention speaks of
definition, the third and fourth sentences show that Russell is merely issuing a request
for clarification, and it is a specific request (a ‘closed question’), in that he offers a
particular word, ‘cause’, as his preferred starting point.
Notice that causes are ordinarily understood as things, events, or processes, in
any case as objects of ontological study, whereas explanations are rather understood
as logical or epistemological objects. The word ‘reason’ oscillates between the two
areas of philosophical study, with some risk of conflating different concepts and even
different methods.10

Segment II, Turn 33 (Copleston) Opening Stage (Material Starting Point)

[33.1] Not necessarily. Cause is a kind of sufficient reason. Only contingent being
can have a cause. God is His own sufficient reason; but He is not cause of Himself.
Copleston partially rejects Russell’s material starting point in 32.1. He seems to
agree with Leibniz that ‘sufficient reason’ applies to contingent beings, and in that
particular case its meaning is the same as ‘cause’; but it also applies to God, and then
it is not the same as ‘cause’.

10 As the reader may remember, the German philosophers Arthur Schopenhauer was famous for
insisting on this point in his dissertation (Schopenhauer 1813, revised edition 1847).
5 Analysis of Segment II: Discussion of Copleston’s … 155

In the history of philosophy, some thinkers have conceived God as self-caused (the
most famous being Spinoza), but this has always seemed to Christian theologians
to imply pantheism, which they abhor. This is why Copleston, a Jesuit father and a
scholastic, rejects the idea of self-causation. Nonetheless, he wants to keep the idea
that “God is His own sufficient reason”. So, he must maintain a distinction between
two kinds of sufficient reason.
The underlying argument would look more or less like this:
Premise 1. In the case of contingent beings, a sufficient reason for their existence is the same
as a cause for their existence.
Premise 2. In the case of God, a necessary being, a sufficient reason for its existence cannot
be a cause.
Conclusion. We must distinguish between two kinds of sufficient reason: causes for
contingent beings and a different kind, exclusive for God.

This distinction, of course, requires Copleston to say something more either about
the special kind of sufficient reason that applies to God or else at least, as required
by Russell’s request, about the general meaning of ‘sufficient reason’. He will opt
for the latter.
[33.2] By sufficient reason, in the full sense, I mean an explanation adequate for
the existence of some particular being.
Copleston tries to clarify the phrase ‘sufficient reason’ by the phrase ‘adequate expla-
nation’. He implies that an adequate explanation may be causal, but it does not have
to be causal. In fact, a causal explanation for God would not be adequate for Cople-
ston. The adjective ‘adequate’ often implies that, what is adequate for one purpose
is not for another. Depending on what sort of thing we want to explain, we have to
pick the right one, the one that is appropriate, or adequate, to the nature of that thing.
Note that, faithful to his own argument for God’s existence, both ‘reason’ and
‘explanation’ are explicitly said to be relational: something is said to be a sufficient
reason or an adequate explanation for the existence of something (either the same
thing or something else).

Segment II, Turn 34 (Russell) Opening Stage (Material Starting Point)

[34.1] But when is an explanation adequate? Suppose I am about to make a

flame with a match. You may say that the adequate explanation of thatis that I
rub it on the box.
Naturally enough, Russell raises the general question when, or under what
circumstances, an explanation is to be called ‘adequate’. Let’s call
this Q35. Probably in order to prevent Copleston to deal in abstractions, he suggests a
humble example. Notice that the example is not formally framed in terms of existence
but rather in terms of a process. The process chosen for illustration is the making of
a flame with a match. The ordinary explanation for that process is the rubbing of the
156 5 Analysis of Segment II: Discussion of Copleston’s …

match against the sandpaper of the matchbox. The question is then whether the
rubbing of a match against the sandpaper of the matchbox is to be
called an adequate explanation of the making of a flame with the
match. Let’s call this Q36 and note that, as a closed question, it is a specification of
Q35.
It is not easy to know whether this request for a very humble starting point for
further discussion is part of Russell’s argumentative strategy. He may just have been
fishing for a concrete thing to anchor the discussion; but he may also have thought that
Copleston had only two choices. If he said Yes, then Russell might argue that such
a humble explanation was incommensurable with the idea of explaining the whole
universe. If he said No, then he may be brought to admit that he is dealing with an
impossible task. However that may be, Copleston will take the latter direction, so
that the whole discussion from this point onward will undergo a profound change,
turning into a discussion of the scope and limits of the law of causality.

Segment II, Turn 35 (Copleston) Opening Stage (Material Starting Point)

[35.1] Well, for practical purposes; but, theoretically, that’s only a partial expla-
nation. An adequate explanation must ultimately be a total explanation, to which
nothing further can be added.
Copleston rejects Russell’s example as purely practical and so useless for the theoret-
ical purposes of the debate. Now fully in epistemological mood (see my last comment
on 32.1), Copleston distinguishes between partial and total explanations, and he then
asserts that an adequate explanation cannot be partial but has to be total. In this way,
Q36 is answered in the negative, and an answer to Q35 is also given by implication:
an explanation is adequate only when it is total, i.e. when nothing can be added to
that explanation to complete it.
In the context of the debate’s presiding question (Q1), this means that God is an
adequate, i.e. a total explanation both of Himself and of the world. The latter part
of this answer could certainly not be otherwise if the definition of God, proposed by
Copleston (1.3) and accepted by Russell (2.1), includes the notion that he created
the world.
The material starting point concerning Q35 is still very much in the air.

Segment II, Turn 36 (Russell) Opening Stage (Meta-Argumentation)

[36.1] Then I can only say you’re looking for something which can’t be got, and
which one ought not to expect to get.
Russell is pretty emphatic with his opinion that total explanations cannot be obtained
and even that one should not expect to obtain them. Although Russell does not give
any argument for that opinion, it may be considered reasonable if we are talking
about scientific explanations, which are recognised by both scientists and laypeople
5 Analysis of Segment II: Discussion of Copleston’s … 157

to be always partial. We can therefore suspect that, underneath Russell’s words, there
lurks the following argument:
Premise 1. The only explanations available, or at least worth considering, are those given
by science.
Conclusion 1. No explanation is total, and we should not aspire to total explanations that
might be on offer.
Conclusion 2. We should not look for an explanation of the world (or of everything).
Conclusion 3. There is no point in discussing God’s existence in Copleston’s terms.

It is, however, pretty clear that Copleston is not talking about scientific explanations.
His argument is, after all, explicitly labelled by him as ‘metaphysical’.
Russell’s position is, therefore, confrontational. He is giving up on his own ques-
tion Q35 (when an explanation is to be called ‘adequate’). Instead he
puts two very different views of the concept of explanation against each other, in such
a way that no proper material starting point appears to be possible. In other words,
Russell is, by this intervention, effectively closing the discussion with Copleston. A
similar situation is thus obtained as the one I commented upon at 25.1.
Notice two things. Russell is making a normative point: “one ought not to expect to
get” a complete explanation of anything. It is a forceful point. Besides, this is the first
passage in which Russell’s scientism manifests itself, i.e. the idea that science is the
only arbiter of what may counts as knowledge. To this extent, Russell’s intervention
may readily be interpreted as an implicit procedural starting point: when discussing
causes, the only competent tribunal is science, particularly natural science. Since it is
only insinuated, however, we cannot take it as a proper procedural point, especially if
we consider that Russell knew or suspected that a Jesuit father would hardly concur
with a scientistic perspective.

Segment II, Turn 37 (Copleston) Opening Stage (Meta-Argumentation)

[37.1] To say that one has not found it is one thing; to say that one should not
look for it seems to me rather dogmatic.
It is now Copleston’s turn to reject the closing of the discussion. He does not go
into the implied argument about scientific explanations. Applied to Russell’s humble
example (34.1), Copleston is saying that the fact that we cannot yet explain the making
of the flame in a complete manner does not entail that we should not keep looking for
one. This seems to be eminently reasonable. In fact, what we find, even in science,
is that human beings are all the time pushing forward towards ever more complete
explanations of things, events, and processes. For practical purposes, as Copleston
says, one may rest content with knowing that, if one rubs the match against the
sandpaper, one will get fire. But the question why surely appeals to everybody. Each
time we pause in our attempts to explain things, we know that it is not the true end
158 5 Analysis of Segment II: Discussion of Copleston’s …

of the story. In a generalized form (as in 36.1), Copleston is arguing that the fact that
we have thus far not found any complete explanation for any particular process in
the world does not entail that we should not keep looking for one.
His meta-argument against Russell’s could be represented as follows:
Premise. It is dogmatic to deduce, from the fact that all available explanations have hitherto
been incomplete, that we should abandon the search for complete explanations.
Conclusion. We should not abandon the search for complete explanations.

Notice that Copleston has quite understood Russell’s normative point. However,
he finds that point ‘dogmatic’, i.e. unsupported. Consequently, it is unacceptable, at
least without argument, to forbid the search for complete explanations.

Segment II, Turn 38 (Russell) Opening Stage (Meta-Argumentation)

[38.1] Well, I don’t know. I mean, the explanation of one thing is another thing
which makes the other thing dependent on yet another, and you have Our this
phrase between inverted commas: ‘to grasp this sorry scheme of things entire’
to do what you want, and that we can’t do.
Russell does not acquiesce to Copleston’s reasoning. He tries again to express what
he means at 36.1. As far as I can make sense of Russell’s somewhat convoluted
phrasing, he seems to argue as follows:
Premise. The explanation of a thing x is another thing y, but to make the explanation of x
complete, we have to explain y also, through yet another thing z, and this has to go on in
order to complete the explanation.
Conclusion. In order to explain anything x, we have to explain everything (the universe).

This argument is fully explicit. Its conclusion is expressed poetically by the striking
use of Edward FitzGerald’s translation of Omar Khayyam’s Rubaiyat (“to grasp this
sorry scheme of things entire”). If Copleston is right, argues Russell, then one would
have to explain the whole universe in order to explain anything at all. This is where the
importance of Russell humble example at 34.1 kicks in. For it would seem obvious
that we can explain the making of the flame without having to explain everything.
It would seem that Russell is attempting a reductio ad absurdum. However, if one
concedes, which is only reasonable, that all scientific explanations are partial and
provisional, how could that be an argument against pushing further the frontier of
knowledge and research?
A couple of considerations might illuminate the dialectical situation here. First,
we should remember that, in many human endeavours, it makes sense to go beyond
the limit set to any given point. Thus, in the Olympic games, athletes are always
trying to beat the sport records. The same is true of research. There is, as it were,
a kind of ‘proximal infinity’ in human affairs, such that we are always trying to go
further. No matter how cognitively limited we are, Copleston seems to argue, it would
be too far to actually assert that we are constitutionally unsuited to either ask for or
5 Analysis of Segment II: Discussion of Copleston’s … 159

Table 5.2 Four kinds of

Explanations Partial Total
explanation
Of something in particular 1 2
Of everything (the universe) 4 3

give complete explanations (for interesting discussions in philosophy, see McGinn

1993, Chalmers 2015).
On the other hand, even if Russell’s argument does not work against what I have
called ‘proximal infinity’ (the breaking of limits), it is actually intended against a
leap into actual infinity. And we should beware that an explanation of everything
(i.e. of the universe) is exactly what Copleston is trying to achieve. However, he is
proposing neither a practical nor a scientific explanation, but a philosophical and
theological one. Therefore, even if Russell’s argument is powerful, is is not quite
engaging Copleston. The ensuing conundrum is probably due to the fact that, at this
point, neither discussant is speaking in a fully clear and distinct manner.
Finally, there is a subtle shift in meaning in Turns 34–38, which brings about the
ambiguities shown in Table 5.2.
It is quite clear that neither Copleston nor Russell have anything against Option 1
in Table 5.2: explaining the making of the flame (34.1) is one such example and there
is nothing objectionable in such explanations. In 36.1, Russell had rejected Option 2
as something that cannot be obtained by us mortals. But the upper row, that comprises
Options 1 and 2, concern explanations of particular things (events, processes, states
of affairs) and so is not really at stake in this debate. Let us, therefore, look at the
lower row, comprising explanations of everything:
As for Option 3, it is clear from the preceding argumentation, that this is the
option that Copleston is really interested in, for he has asserted that the metaphysical
argument intimates that in God we find a total explanation of everything or of the
universe. He has also made clear that we do not know the content of the explanation,
only that there is one.
As for Option 4, it does not seem to come into the discussion, at least explicitly.
Nonetheless, Russell would presumably accept that physical cosmology is on the way
of providing a partial explanation of the universe, an explanation that will perhaps
always remain partial, if we accept that science is not in the business of providing
total explanations of anything. This much seems to be implied both by what Russell
says in the debate and by what his general view of science is. Whether this is fully
compatible with Einstein’s and other physicists’ dream of a theory of everything
(Durham and Rickles 2017), is another matter (see also Weinberg 1993).
160 5 Analysis of Segment II: Discussion of Copleston’s …

Segment II, Turn 39 (Copleston) Opening Stage (Request for Clarification)

[39.1] But are you going to say that we can’t, or we shouldn’t, even raise the
question of the existence of the whole of this ‘sorry scheme of things’, of the
whole universe?
Copleston gets to the point by putting forward a direct request about Russell’s
assumption that we cannot explain everything, or that there is no possible expla-
nation for the universe. Unfortunately, without the complete recording of the debate,
the transcribed text of Copleston’s intervention at this point is ambiguous between
two interpretations:
On the one hand, Copleston may be trying to clarify the meaning of Russell’s
assumption, by distinguishing between the normative and the factual perspective.
He would then, in other words, be asking whether complete explanations
cannot be achieved or rather whether they should not even be
attempted. This multiple-choice question may be referred to as Q37. The first
option of Q37 is suggested in 38.1, the second in 36.1.
On the other hand, Copleston may rather be trying to clarify the meaning of
Russell’s assumption but not with regard to the difference between the normative and
the factual perspective. In that case, Copleston would just be asking whether the
very question of (an explanation for) the existence of the universe
is pointless. Let’s call this Q38. Notice that, in Copleston’s phrasing, Q38 is
abridged: he says: ‘the question of the existence of the universe’ whilst meaning,
more fully, ‘the question of an explanation for the existence of the universe’. This is
important in view of Russell’s immediate reply, as we shall see.
Whatever the meaning intended by Copleston might have been, Russell obviously
understood his request according to the second interpretation.
Notice that, in this intervention, Copleston is confirming that Option 3 of Table
5.2 in my comment to 38.1 is what he wants to discuss, so that Russell’s example
of matches and flames is now seen to be beside the point. Nonetheless, the idea that
Copleston has never tried to give a scientific explanation of the universe, but only a
philosophical one (see 3.1), is for some reason not made explicit.

Segment II, Turn 40 (Russell) Opening Stage (Sub-Discussion)

[40.1] Yes. I don’t think there’s any maning in it at all.

Russell takes the request as an occasion to clarify his meaning, and in so doing
produces, rather suddenly, a retraction of a previously accepted material starting
point. Copleston had asked him explicitly (at 21.5) whether the question ‘Has the
world a cause?’ or ‘Does a cause of the world exist?’ has meaning (Q31); and Russell
had answered in the affirmative (22.1). Now he is saying that it does not. It is here that
Russell starts to move from an initial Gamma stance towards Copleston’s standpoint
that God (defined as the cause of the world) exists, which he adopted at 4.1, to the
5 Analysis of Segment II: Discussion of Copleston’s … 161

final Delta stance that he will adopt at 68.1. For the distinction between Gamma and
Delta stances, see Chapter 3, Sect. 3.2.11
[40.2] I think the word ‘universe’ is a handy word in some connections, but I
don’t think it stands for anything that has a meaning.
Russell now gives an argument for his assertion that Q32 (whether the world
has a cause) has no meaning. Before going into the argument, notice that Russell’s
phrasing of his reason is somewhat convoluted: The word ‘universe’ does not stand
for anything that has a meaning. I propose that what Russell means is either:
The word ‘universe’ does not stand for anything, or: The word ‘universe’ has
no meaning, or perhaps he intended to say that it has neither.
Anyway, according to the former interpretation, Russell is saying that there is
no object for the word ‘universe’ to refer. However, remember that Copleston had
already asserted as much without a murmur of protest from Russell, even though
Copleston’s assertion is not without problem. Compare 11.2.3. Consider that Russell,
who famously had discovered a contradiction at the roots of set theory, could certainly
have argued that ‘universe’ is too big a set to exist. However, he does not put forward
any such or similar argument during the whole debate.
As for the latter interpretation, we have to take the word ‘meaning’ as contrasting,
in Frege’s manner, to ‘reference’. In that case, Russell is saying that the word
‘universe’ does not have any content, that we have no concept of the universe. This
would be very much a Kantian notion.
Of course, whether the sense or the reference of the word ‘universe’ is intended,
Russell would be justified in either case to conclude that the question of the
explanation of the universe would be empty:
Premise. The word ‘universe’ has no sense (or no reference).
Conclusion. The question of the explanation of the existence of the universe is meaningless.

This is a very powerful argument; and yet, in the sequel, Copleston will insist on
speaking of the universe, as though the word had, if not reference, at least some
sense.
It is important that, in the way Russell presents his argument, he compares two
reasons: (1) the word ‘universe’ is handy in some connections, (2) the word ‘universe’
has no meaning—no sense, no reference or neither sense nor reference. Reason (1)
would support the negative in answer to Q38 (whether the very question of
an explanation for the existence of the universe is pointless); but
reason (2) would support the affirmative. By the word ‘but’, Russell indicates that
reason (2) is much stronger than reason (1). The word ‘universe’ is handy enough,
but we should not take it too seriously.

11 Alpha believes that God exists and Beta that God does not exist. Gamma, in turn, holds that there
is no way to prove that God exists or that God does not exist. Finally, Delta contests the relevance
of the question.
162 5 Analysis of Segment II: Discussion of Copleston’s …

One last remark. The adjective ‘handy’ is a colloquial word which in this context
means as much as ‘useful’. This would imply for some readers that Russell is some-
thing of a nominalist, given that nominalists hold that words are just useful but have
no real meanings. However, Russell, although an empiricist, was also a noted oppo-
nent of the nominalist negative attitude towards real meanings (see Russell 1912b,
Chap. IX). So, we should not overinterpret or overgeneralize what he says here.
In any case, it is possible that Russell is aware of his retraction and so is trying
to ‘save face’: if earlier on he got along with Copleston’s talk of the world or the
universe, it was only because that word is handy, precisely in the way in which
Copleston introduced it, as a harmless proxy for ‘everything’; but once the arguments
becomes serious and we are actually talking about what is the cause of, or how we
can explain, the whole thing in one fell swoop, then we ought to be more careful with
what we say and what we accept as meaningful. After all, at the time when Russell
accepted Q31 as meaningful (22.1), the stress was laid on the fact that Copleston had
replaced a name (‘God’) with a description (‘the cause of the world’) in the context
of saying that the thing described exists. Within those argumentative boundaries,
Q31 is arguably meaningful. On that occasion, however, the predicate was taken as
unified and indivisible: ‘being a cause-of-the-world’ is, in other words, a monadic
predicate. Thus, if we represent that predicate with the letter ‘F’, then we have.
(1) (∃x)(F(x)&(∀y)(F(y) ⊃ y = x))
There is an individual x such x is cause-of-the-world, and, for every individual
y, if y caused-the-world, then y and x are the same individual.

What Russell would have conceded at 22.1 would only have been, under this inter-
pretation, that (1) is a well-formed sentence. However, at this other moment of the
discussion, ‘F’ is replaced by a dyadic predicate, i.e. a relation between one thing
(God) and another thing (the world). If we represent this other predicate with the
letter ‘R’, then we have rather something like.
(2) (∃x)(∃y)(R(x, y)&(∀z)(∀w)(R(z, w) ⊃ ((z = x)&(w = y))))
There is an individual x and an individual y such that x causes y, and, for every
pair of individuals <z, w>, if z causes w, then <z, w> and <x, y> are the same
pair of individuals.

Of course, (2) is as well-formed as (1), and in that restricted sense, (2) is as meaningful
as (1). However, the question is now whether one can speak of the world as an
individual. This is what Russell would be rejecting now. If one has enough trouble
with God as an individual, one has perhaps even more trouble with the world as an
individual. (This was already a Kantian point.)
Now all this is technical stuff within the syntax of the predicate calculus, and it
is clear that Russell has not said any of this in this turn. All I am suggesting is that
his position may be re-interpreted in this light.
In any case, it is clear that we have entered a new sub-discussion on the very
meaning of Copleston’s question.
5 Analysis of Segment II: Discussion of Copleston’s … 163

Segment II, Turn 41 (Copleston) Opening Stage (Sub-Discussion)

[41.1] If the word is meaningless, it can’t be so very handy.

It is not quite clear what role this sentence has. On its surface, it appears as though
Copleston disputes Russell’s argumentative comparison, suggesting that ‘being’
handy’ is hardly separable from ‘having meaning’. This would then seem to be
an objection: if the word ‘universe’ is handy, as Russell admits, then it must be
meaningful, but if it is meaningful, then Russell must uphold the position expressed
at 22.1 (viz., the question Does the cause of the world exist? has meaning). But note
that the adjective ‘handy’ is quite vague. It can probably only mean that the word is
useful as a kind of abbreviation, which is exactly how Copleston, correctly or not,
is trying to use it from the very beginning. The question that arises is not so much
whether a word can be handy and meaningless at the same time
as rather whether a question in which a handy but meaningless word
appears can be meaningful. Let’s call it Q39. Copleston seems to indicate that
he would say Yes to that question.
[41.2] In any case, I don’t say that the universe is something different from the
objects which compose it. I indicated that in my brief summary of the proof.
This is quite true: Copleston refers to what he said at 11.2.1 and 11.2.3. Russell never
demurred. Now, I have shown that Copleston uses the alleged equivalence between
‘universe’ and ‘everything’ to produce a dubious argument (see my comment on
11.3.1); but, given Russell’s silent acquiescence, this is irrelevant.
Returning to the handy-meaningful distinction: if one can use ‘universe’ as a
comfortable abbreviation for ‘everything’, then it is clear that that word is handy.
But what about its meaning? By using words like ‘universe’ as a mere abbreviation,
is Copleston accepting that it has no reference? Suppose he is. Does it follow that it
has no meaning? Even non-referring terms such as ‘unicorn’ or ‘Sherlock Holmes’
are widely considered meaningful. It is rather unfortunate that we cannot answer any
of these questions, given that the discussants do not take them up.
[41.3] What I’m doing is to look for the reason, in this case the cause, of
the objects, the real or imagined totality of which constitute what we call the
universe.
Having said that the universe is not “something different from the objects that
compose it” (41.2), and having in particular expressed a decidedly nominalistic posi-
tion (note here, in 41.3, the stress in “what we call the universe”, clearly perceivable
in the extant recording), Copleston argues as follows:
164 5 Analysis of Segment II: Discussion of Copleston’s …

Premise 1. Russell objects that it does not make sense to ask about the explanation for the
existence of the universe.
Premise 2. Copleston is not asking about the reason for the existence of some strange new
and separate object, called ‘universe’; he is instead asking about the reason for the existence
of anything and everything.
Conclusion. Russell’s objection does not touch Copleston’s question.

Notice that Copleston is, again, using the equation world = everything, but now as
a way to disarm Russell’s argument at 40.2.
[41.4] You say, I think, that the universe, or my existence, if you prefer, or any
other existence, is unintelligible?
Copleston issues a request for clarification of Russell’s standpoint, asking whether
(the existence of) the universe is unintelligible. Call this Q40. Notice
that Russell’s argument at 40.2 belongs to the meta-language. He is speaking about
the meaning, or rather the lack of meaning, of a word, and from that he infers the lack
of meaning of a question. Up to this point in Copleston’s intervention, he has also
been arguing meta-linguistically. But now he suddenly shifts to the object-language:
he is now talking not about the word ‘universe’ but about the universe itself, or rather
its existence. Copleston seems to assume that saying that the word ‘universe’ lacks
meaning is logically equivalent to saying that the universe itself is unintelligible. He
does not justify the assumption, nor is it clear why he makes it (but see 43.1).
Now, when he uses the word ‘universe’, Copleston is allegedly merely using
an abbreviation to refer to any old thing. So, he feels justified in substituting any
particular thing, so that his question for Russell now is whether existence is at
all intelligible. Call this Q41. The argument may be represented as follows:
Premise 1. For Russell, the word ‘universe’ lacks meaning.
Premise 2. The word ‘universe’ is just an abbreviation for everything.
Conclusion. For Russell, everything is unintelligible.

Mark Copleston’s argumentative strategy at this point. Affirming that the universe—
or rather the existence of the universe—is unintelligible is one thing, but affirming
that the existence of Copleston, or anybody else, or indeed of anything else, is unin-
telligible is a very different thing. The former might be defended (say, by a meta-
linguistic argument purporting to show the lack of meaning of the word ‘universe’);
but the latter sounds very dubious. Copleston has, therefore, managed to go from
the meaninglessness of the word ‘universe’ to the unintelligibility of anything and
everything. He cannot, of course, believe that Russell means the conclusion of the
above argument. So, he is trying to get a material starting point from which to discuss.
When he fails, he will go back to it later on (see 43.5).
Finally, it is worth remarking that Copleston never complains that Russell had
committed, at 22.1, to the proposition that the question of the cause of the world
had meaning. This is in accord with the methodological hint, at 1.4, that all agreed-
upon starting points are accepted provisionally and so can be withdrawn later on in
5 Analysis of Segment II: Discussion of Copleston’s … 165

the debate. In certain dialogue games, such as medieval obligationes or the game
described by Aristotle in his Topics (see Moraux 1968; Yrjönsuuri 1993), withdrawal
of concessions is forbidden, but Copleston seems to play rather by the rules of the
Socratic dialogue, where the answerer can always withdraw a proposition if he is in
trouble.

Segment II, Turn 42 (Russell) Opening Stage (Sub-Discussion)

[42.1] First may I take up the point that if a word is meaningless it can’t be
handy. That sounds well but isn’t in fact correct. Take, say, such a word as ‘the’
or ‘than’. You can’t point to any object that those words mean, but they are
nevertheless very useful words; I should say the same of ‘universe’.
Russell responds first to 41.1. His argument is by analogy:
Premise 1. Grammatical function words such as the definite article ‘the’, or the comparative
particle ‘than’, are useful but lack any reference.
Premise 2. The word ‘universe’ is like the definite article ‘the’, or like the comparative
particle ‘than’.
Conclusion. The word ‘universe’ is useful but lack any reference.

In this particular argument, more indicated than expressed by Russell, he seems

to choose the referential sense of ‘meaning’ (see comment on 40.2), which is the
less powerful sense in view of Copleston’s view of ‘universe’ as a kind of handy
abbreviation (see comment on 41.1). It is a very weak argument, unless Premise 2 is
elaborated to show that ‘universe’ behaves in a way relevantly similar to a grammat-
ical function word. But the elaboration might make the grammatical behaviour of
‘universe’ perilously close to Copleston’s proposal to equate it with ‘everything’. Of
course, if the metaphysical argument could be logically parsed, Russell might still
defeat it (in a way perhaps not dissimilar to my analysis of 11.3.1).
[42.2] But leaving that point, you ask whether I consider that the universe is
unintelligible. I shouldn’t say unintelligible: I think it is without explanation.
Intelligible, to my mind, is a different thing. Intelligible has to do with the thing
itself, intrinsically, and not with its relations.
In order to understand Russell’s point, we must remember that his whole way of
doing philosophy was marked by his discovery of the ‘logic of relations’. To say
that the universe is intelligible or unintelligible, we have to think this as a monadic
predicate, whereas to say that it has or doesn’t have an explanation corresponds to
a dyadic predicate (what is explained is explained by something else). If we use
the letters ‘F’ and ‘R’ as before, we get four alternatives, as in Table 5.3. Russell
is upholding the contents of the lower cell on the right, which is different from the
upper cell in the same column. This somewhat pedantic distinction is well taken,
although a bit disappointing as an answer to Q40. If, when he said that Copleston’s
question is meaningless, he did not mean that the universe is unintelligible, we still
166 5 Analysis of Segment II: Discussion of Copleston’s …

Table 5.3 Distinction

The universe Affirmative Negative
between ‘is intelligible’ and
‘has an explanation’ is intelligible F (x) ~ F (x)
has an explanation R (x, y) ~ R (x, y)

don’t know whether Russell considers the universe intelligible or not. On the other
hand, we should remember that Copleston may be conflating the object level and the
meta-level. Saying that the word ‘universe’ is intelligible seems to be more or less
equivalent to saying that it is meaningful but different from saying that the universe
itself is intelligible (see comment on 41.4).

Segment II, Turn 43 (Copleston) Opening Stage (Sub-Discussion)

[43.1] Well, my point is that what we call the world is intrinsically unintelligible,
apart from the existence of God.
The motive behind the shift from the meta-language favoured by Russell to the
object-language introduced by Copleston at 41.1 becomes apparent. Copleston does
not want to speak about words (like ‘universe’ or ‘world’) but about the universe itself
(or about the total aggregate of things, or, as he will presently say, the total series of
events). It is this, the universe itself, which is unintelligible—its existence cannot be
understood—unless God exists. For, of course, it is also about God Himself that he
wants to talk about, and not the word or the name ‘God’. In other words, he wants
to get back to the metaphysical argument itself, forsaking talk about language and
words and meanings.
[43.2] You see, I don’t believe that the infinity of the series of events—I mean a
horizontal series, so to speak—if such an infinity could be proved, would be in
the slightest degree relevant to the situation.
In a surprising move, Copleston starts talking about ‘the series of events’, by which
he means the total series of events (this is a new description of the universe), and then
to assert that, if somebody would argue that that series is infinite, that argument would
have no force (“it would not be in the slightest degree relevant to the situation”).
To understand why Copleston should bring this up, we need some background infor-
mation. There is a very old philosophical discussion about the question whether
the world had a beginning or is eternal (Q42). The Aristotelian posi-
tion was that the world is eternal; but this was unacceptable to Christian theologians
who believed that the world was created by God, so there had to be a beginning of
the world. Now, eternity implies an infinite series of events. So, naturally, Christian
philosophers wanted to refute that the universe could be seen as an infinite series of
events, for this would preclude creation. It turns out that refuting this is as hard as
proving it, so some philosophers, notably Leibniz, argued that, even if the series of
events should be infinite, the argument for God’s existence would remain valid. That
5 Analysis of Segment II: Discussion of Copleston’s … 167

is why Copleston puts the question on the table, only to dismiss it as irrelevant to
his metaphysical argument. As for Russell, he was familiar with all that, which is
why he does not say anything about this aspect of the argument (see Russell 1900,
Sect. 109).
[43.3] If you add up chocolates, you get chocolates after all and not a sheep.
If you add up chocolates to infinity, you presumably get an infinite number of
chocolates. So, if you add up contingent beings to infinity, you still get contingent
beings, not a necessary being.
Copleston now produces an argument for believing that infinity of the series of events
is irrelevant to his argument. It is an argument by analogy:
Premise 1. If you add up chocolates, you get chocolates after all and not a sheep, never mind
if the number of chocolates is finite or infinite
Premise 2. Chocolates are to a sheep as contingent beings are to a necessary being.
Conclusion. If you add up contingent beings to infinity, you still get contingent beings, not
a necessary being.

Adding up is one way to create a series: the first chocolate followed by the second
chocolate, and so forth to the nth chocolate. So, a further analogy is intended, as we
shall see presently.
[43.4] An infinite series of contingent beings will be, to my way of thinking, as
unable to cause itself as one contingent being.
As we saw in the comment to 11.2.1, Copleston is not very fastidious about the
distinction between things and events, and his preferred word ‘being’ is a kind of
proxy for both, a dummy term. That explains why he goes from a series of events (at
43.2) to a series of chocolates (at 43.3), and now to a series of beings. Copleston has
in mind an argument by analogy:
Premise 1. One contingent being cannot cause itself.
Premise 2. One contingent being is similar to an infinite series of contingent beings.
Conclusion. An infinite series of contingent beings cannot cause itself.

The careful reader will see, in Premise 2, the old idea that certain set-like expressions
(such as ‘world’, ‘universe’, here ‘series of beings’) are a handy abbreviation for
everything. We have a convenient means to convert many things into one, and we
can transfer any predicate which we use for each and every one of the many things
to the one set-like thing. This convenience is precisely what our old equation (world
= everything) was supposed to help us do (see comment on 11.2.3).
[43.5] However, you say, I think, that it is illegitimate to raise the question of
what will explain the existence of any particular object?
Copleston comes back to the argument underlying 41.4:
168 5 Analysis of Segment II: Discussion of Copleston’s …

Premise 1. For Russell, the word ‘universe’ lacks meaning.

Premise 2. The word ‘universe’ is just an abbreviation for everything.
Conclusion. For Russell, everything is unintelligible.

In that turn, this led from Q40 (whether the existence of the universe is
unintelligible), which is a least somewhat related to something that Russell actu-
ally asserted (that the word ‘universe’ has no proper meaning), to Q41(whether the
existence of anybody or anything is at all intelligible), the question
which Copleston had produced by means of the equation world = everything.
On this occasion, however, Copleston leaves talk of intelligibility for talk of
explanation. So, the conclusion of the above argument would be: For Russell, nothing
has an explanation, which would seem a reductio ad absurdum. Under the equation
Russell would have to reject the question whether existence can be at all
explained (a variant of Q41, say Q41’). This is a more clearly phrased proposal of
the same material starting point offered at 41.4.

Segment II, Turn 44 (Russell) Opening Stage (Sub-Discussion)

[44.1] It’s quite all right if you mean by explaining it simply finding a cause for
it.
Russell responds to Q41’ with robust common sense, thus acquiescing in Copleston’s
material starting point. He is, of course, placing himself squarely inside the Option 1
cell of Table 5.2 in my comment to 38.1: partial explanations of particular things—like
the making of the flame at 34.1.
Russell’s acquiescence is, however, not quite genuine, for he again wants to restrict
‘explanation’ to causal explanation, in spite of the fact that Copleston had argued that
not all explanation is causal at 33.1. Again, in asking Q41’, Copleston is probably
thinking of total explanations, in spite of the fact that Russell had said there isn’t
any. If the two philosophers are aware that they are talking at cross-purposes, they
certainly don’t seem to mind.
To sum it up, they both use the verb ‘to explain’ as if they meant the same; but it
has become clear that Copleston means ‘to explain totally, but not always causally’,
whereas Russell means ‘to explain partially, and always causally’. They seem to
move in circles, dancing around each other.
In particular, it seems that Russell, having his strategy of reduction deflated, has
become passive and reactive. He has never abandoned the role of antagonist, but
from turn 26, he had been looking for a refutation of Copleston’s position, based
on the reduction of his metaphysical (or cosmological) argument to the ontological
argument. Having failed to achieve such a refutation by Copleston’s deft defence at
turn 31, he started assuming the role of questioner. This was for a while ineffective,
5 Analysis of Segment II: Discussion of Copleston’s … 169

from turn 32 until now, when both discussants give the appearance of dancing around
each other without putting their cards on the table.

Segment II, Turn 45 (Copleston) Opening Stage (Sub-Discussion)

[45.1] Well, why stop at one particular object? Why shouldn’t one raise the
question of the cause of the existence of all particular objects?
Copleston takes again the initiative and tries to push Russell into a corner. If Russell
(as was easily predicted) would accept that it makes sense to look for an explanation
of one thing (never mind now whether total or partial), then why should he not also
accept that it makes sense to look for an explanation of all things? The underlying
argument is as follows:
Premise. It is possible to raise the question of what causes one particular object.
Conclusio. It is possible to raise the question of what causes all particular objects.

Notice that this is a significant variant of Q40 (whether the existence of the
universe is unintelligible), and in fact it is none other than a variant of Q38
(whether the very question of an explanation for the existence
of the universe is pointless). This variant, which we may call Q38’, should
be rephrased as follows: whether it makes sense to raise the question
whether the existence of everything can be causally explained.
Notice the differences:
(1) The word ‘universe’ in Q40 and Q38 is replaced by ‘everything’ or all things’
in Q38’, as per Copleston’s equation.
(2) Talk of intelligibility in Q40 and talk of explanation in Q38 is replaced by talk
of cause, as per Russell’s insistence.
(3) Q38’ does not ask whether the universe has causes but only whether it makes
sense to ask whether the universe has causes. The ontological question has
been replaced by an epistemological or perhaps a semantic one, as in Q38, in
contradistinction to Q40.
Copleston is, therefore, requesting, through Q38’, a material starting point, and
indeed a crucial one. Russell had acquiesced in something like that at 22.1 and then
retracted his acquiescence at 40.1 (although see last part of comment on 40.2). This
time around, however, Copleston is careful to avoid using the word ‘universe’ and
perfectly willing to talk about ‘causes’ instead of ‘reasons’. Remember that Russell
had given the meaninglessness of the word ‘universe’ as the reason to position
himself in the negative regarding Q38, and he had expressed his wish to talk about
‘causes’ (at Turn 32). Therefore, if Russell wants to say No to Q38’, it would be
reasonable to expect from him a different reason than quibbles on words—or at least
what Copleston seems to view as quibbles on words.
To repeat, we have moved from ‘asking about the reason for the existence of
the universe’ to ‘asking about the cause for the existence of all things’. These
170 5 Analysis of Segment II: Discussion of Copleston’s …

apparently small modifications may be seen as Copleston’s friendly attempt to meet

Russell halfway. As such, it also highlights the real abyss between the worldview
and vocabulary of the two discussants, and the difficulties to bridge that abyss.

Segment II, Turn 46 (Russell) Opening Stage (Sub-Discussion)

[46.1] Because I see no reason to think there is any.

Although formulated as an assertion, Russell’s intervention can perhaps be inter-
preted as a request: give me a reason to think there is a cause of the
existence of all things. Let’s call this Q43. This interpretation will be confirmed
later on, as Russell makes the request fully explicit (see Turn 54).
[46.2] The whole concept of cause is one we derive from our observation of
particular things; I see no reason whatsoever to suppose that the total has any
cause whatsoever.
Russell, ever the argumentative philosopher, is not yet quite decided to play the role
of questioner and putting forward a request to Copleston. Instead, he puts forward an
argument for saying No to Q38’ instead of waiting for a response from his discussion
partner to the veiled request at 46.1. Russell’s quite emphatic argument for not seeing
a reason to think there is a cause of all things is quite remarkable:
Premise 1. Human beings derive the concept of cause from their observation of particular
things.
Conclusion. Human beings have no reason to ask for the cause of all things.

Notice, first, that Russell, by the phrasing he chooses, is tacitly conceding that the
phrase ‘the total’ (here equivalent to ‘universe’ or ‘world’) is readily substitutable
for phrases like ‘all things’ or ‘everything’. The fact, already noted, that Russell does
not resist the equation that is so central to Copleston’s argumentation explains why
the latter keeps thinking that they are in agreement about this point. He will soon be
disappointed (Turn 52).
A final remark on the content of the argument. On the face of it, it looks like a
combination of Hume and Kant. From Hume, Russell has the idea that the concept of
cause is not a priori; from Kant, the idea that a concept that is applied to individual
‘beings’ (to use Copleston’s dummy term), as they are given in experience, should
not be used beyond the realm of possible experience. To be slightly facetious, the
Universe is on a different level than chocolates.
5 Analysis of Segment II: Discussion of Copleston’s … 171

Segment II, Turn 47 (Copleston) Opening Stage (Sub-Discussion)

[47.1] Well, to say that there isn’t any cause is not the same thing as saying that
we shouldn’t look for a cause.
Copleston starts by making explicit the distinction between an ontological answer
(there is no cause of all things) and an epistemological answer (it makes no sense to
ask whether there is). Notice that Copleston is here, for the third time, insisting on
the justification underlying his metaphysical argument (compare Turns 37 and 41).
He presupposes an equivalence: We are entitled, and even obliged, to try to answer
the question of the existence of the universe if and only if that question makes
sense. He thinks the question makes sense, and that is his justification for searching.
And he correctly assumes that the reason for Russell’s rejection of the metaphysical
argument is his assertion that the question does not make sense (40.1); but he wants a
reason for that assertion. In other words, he wants to start a sub-discussion in which
Russell plays the role of protagonist and defends his position that the said question
has no meaning.
So far, Russell has given two arguments. One is that ‘universe’ is useful but
meaningless (Turn 40). This obstacle has been, at least on the surface, removed. The
second argument is that the concept of cause should not be applied to ‘the total’ of
things (Turn 46).
[47.2] The statement that there isn’t any cause should come, if it comes at all,
at the end of the enquiry, not the beginning.
Next, Copleston gives a methodological argument to back the ontological-
epistemological distinction and to throw some doubt onto Russell’s second argument
at 46.2, which amounts to a prohibition issued to stop inquiry, or at least a warning.
Now, in Kant the prohibition or warning is based on a long argument to the effect
that (1) searching for a cause of the universe has been tried and failed to attain
consensus among metaphysicians; (2) searching for causes is limited to a certain
kind of objects, which does not include ‘the total’, i.e. the universe, for this is not a
possible object of experience; and (3) the error of trying to extend the search of causes
beyond its only domain of application can be described and explained satisfactorily as
an error in reasoning. This line of defence, however, would be contrary to Russell’s
rejection of Kantian philosophy. But, by producing the argument at 46.2 he has
opened that flank.
Notice that Kant has not yet been mentioned up to this point, but his presence in
the background is undeniable and will soon be made explicit (see Turn 51).
[47.3] In any case, if the total has no cause, then to my way of thinking it must
be its own cause, which seems to me impossible.
Copleston could have waited for a reply to the above invitation to engage in a sub-
discussion where he played the role of antagonist; but his argumentativeness gets in
the way, and he cannot resist giving his own reasons. 47.1 and 47.2 insinuated that, at
bottom, Russell believes that the universe is not caused by anything. Given, however,
172 5 Analysis of Segment II: Discussion of Copleston’s …

that Russell has not asserted that, Copleston had retained there a tentative stance,
as it were, on the purely epistemological and methodological level. In contrast, 47.3
goes beyond that tentative stance and assumes a negative answer to the question
whether the universe has a cause (Q44). His argument is then:
Premise 1. If the total has no cause, then it must be its own cause.
Premise 2. The total cannot be its own cause.
Conclusion. The total must have a cause.

It is understood, of course, that the cause of ‘all things’ must be external to them,
and Copleston has a name for that cause—God.
[47.4] Moreover, the statement that the world is simply there, if in answer to a
question, presupposes that the question has meaning.
Copleston does not rest content with this argument but gives another one that
concludes in an affirmative answer to Q38’, in which he presents Russell with the
statement that contemporary philosophers and militant atheists think came from
Russell himself: The world is simply there, or, as Russell will rephrase it (50.1), The
universe is just there.
Premise 1. The statement that the world is simply there is an answer to the question whether
the world has a cause.
Premise 2. Anybody who argues like Russell is asserting that the world is simply there.
Conclusion 1. Anybody who argues like Russell is answering the question whether the
world has a cause.
Premise 3. When someone answers a question, then that question has meaning for them.
Conclusion 2. Anybody who argues like Russell cannot assert that the question whether
the world has a cause has no meaning.

Even though Russell has carefully refrained to say, loud and clear, that the world has
no cause, this argument is designed to be a bait to get Russell out of the closet. It
will eventually succeed in doing so.

Segment II, Turn 48 (Russell) Opening Stage (Sub-Discussion)

[48.1] No, it doesn’t need to be its own cause;

Russell at first does not bite. He limits himself to responding to 47.3, by denying the
conditional in Copleston’s argument (Premise 1 in my reconstruction).
[48.2] what I’m saying is that the concept of cause is not applicable to the total.
Russell re-iterates the conclusion of his quasi-Kantian conclusion at 46.2.
5 Analysis of Segment II: Discussion of Copleston’s … 173

Segment II, Turn 49 (Copleston) Opening Stage (Sub-Discussion)

49.1] Then you would agree with Sartre that the universe is what he calls
“gratuitous”?
Copleston insists on his bait, this time by asking Russell a direct question. If one
considers that Sartre had no particularly good reputation in the English-speaking
world, then we can suspect an argument smacking of the guilt-by-association
manoeuvre:
Premise. Asserting that the world is simply there is equivalent to asserting, with Sartre, that
the universe is gratuitous.
Conclusion. Do you still want to say that the world is simply there?

Notice that Copleston is not engaging with Russell’s argument at Turn 48. For that
argument does not assert that the world is simply there (or has no cause), but only
that talk of causes in regard to the universe is beside the point. Copleston, however,
probably thought that this quasi-Kantian argument in a non-Kantian thinker like
Russell was a case of sitting on the fence; and he wants to bring Russell down. (By
the way, one could add to our list of question the one raised by Copleston here, namely,
whether the universe is ‘gratuitous’, whatever that may mean. Given that
Russell, as we shall see, treats this question as a red herring or at least as a hot potato,
I shall leave it unnumbered.)

Segment II, Turn 50 (Russell)Opening Stage (Sub-Discussion)

[50.1] Well, the word “gratuitous” suggests that it might be something else. I
should say that the universe is just there, and that’s all.
Copleston succeeds in getting Russell off the fence. But Russell’s reply is very
ingenious. He seems to argue as follows:
Premise 1. The universe is. just there, and that’s all.
Premise 2. Asserting, with Sartre, that the universe is gratuitous implies that it might be
something else.
Conclusion. I have nothing in common with Sartre.

Russell thus manages to extricate himself out of Copleston’s trap. On the one hand,
he seems to concede what Copleston wants, for he repeats Copleston’s statement that
The universe is just there. On the other hand, his reply very economically restates
his quasi-Kantian position that the word ‘cause’ is not applicable to the universe.
174 5 Analysis of Segment II: Discussion of Copleston’s …

Segment II, Turn 51 (Copleston) Opening Stage (Sub-Discussion)

[51.1] Well, I can’t see how you can rule out the legitimacy of asking the question
how the total, or anything at all comes to be there. Why something rather than
nothing? That is the question.
Copleston finally engages Russell’s argument. He acknowledges that Russell’s posi-
tion is not to deny that the ‘total’ has a cause but only to rule out the legitimacy
of asking what that cause may be (or, as he puts, how ‘anything at all comes to be
there’). Given that Russell has given an argument for that, he proceeds to tackle it.
With a Shakespearean undertone, Copleston raises the famous Leibnizian question
why there is something (anything at all) rather than nothing. This
is just a dramatic formulation of Q44. We may call it Q44’. It will never come up
again in the debate.
[51.2] The fact that we gain our knowledge of causality empirically, from partic-
ular causes, does not rule out the possibility of asking what the cause of the series
is.
Copleston first rejects the inferential step in the argument contained in 46.2. By so
doing, he has to give an argument for this rejection. And he does.
[51.3] If the word “cause” were meaningless or if it could be shown that Kant’s
view of the matter were correct, the question would be illegitimate, I agree;
but you don’t seem to hold that the word “cause” is meaningless, and I do not
suppose you are a Kantian.
Copleston’s counterargument takes the form of a dilemma:
Premise 1. I would agree that the question about the cause of the world is illegitimate if
either the word ’cause’ were meaningless or else if Kant’s view of the matter were correct.
Premise 2. Russell doesn’t seem to hold that the word ‘cause’ is meaningless.
Premise 3. It is not likely that Russell is a Kantian.
Conclusion. He cannot really believe that it is illegitimate to raise the question how anything
at all comes to be there.

Let’s consider the first horn of the dilemma. If the word ‘cause’ should be considered
without meaning (as formerly was the case of the word ‘universe’), then it would not
make sense to ask for the cause of the universe (or indeed of anything else). On the
other horn, if Russell shared Kant’s view of the matter (a long argument that I sketched
in my comment on 47.2), then, again, it would not make sense to apply the concept
of cause to the universe (for the concept of cause can only be applied to possible
objects of experience, whereas the universe is not a possible object of experience).
If these two arguments against raising the question that matters to Copleston are the
only possible ones, as Copleston suggests, then it is perfectly reasonable to say that
Russell has no reason to reject the question.
5 Analysis of Segment II: Discussion of Copleston’s … 175

It may be of interest to remark that Russell had publicly attacked the notion
of cause as meaningless (Russell, 1912a) and that he had always rejected Kant’s
philosophy. So, Copleston has good grounds for his dilemma.

Segment II, Turn 52 (Russell) Argumentation Stage (Objection to First Argument)

[52.1] I can illustrate what seems to me your fallacy. Every man who exists has
a mother, and it seems to me your argument is: Therefore, the human race must
have a mother. But obviously the human race hasn’t a mother: that’s a different
logical sphere.
At this point, Russell abandons all the sub-discussions emerging in the very long
Opening Stage, which he himself initiated at Turn 12 and that took no less than
forty turns altogether. He seems to have run short of quibbles about the meaning
of Copleston’s words and he comes back, in one sudden leap, to the metaphysical
argument itself. We are thus back in the Argumentation Stage.
Having been cornered by Copleston’s dilemma, Russell discovers an argumen-
tative strategy directed to one of the weaknesses in the metaphysical argument: the
transition from ‘everything’ to ‘the universe’ (see comments on 11.2.3 and 11.3.1).
What can be predicated of particular things can rarely be predicated, without much
further ado, of a set of particular things, let alone of the set of all particular things
(which should have been problematic for the discoverer of the problems caused by
certain big sets). The argument humorously uses Coplestons’s own examples (see
11.1.2 and 11.2.4):
Premise 1. The argument from ‘every man has a mother’ to ‘the human race has a mother’
is incorrect.
Premise 2. Copleston’s metaphysical argument is like the argument from ‘every man has a
mother’ to ‘the human race has a mother’.
Conclusion. Copleston’s metaphysical argument is incorrect.

One problem with Russell’s argument is that it uses an analogy; and analogies tend
to be weak arguments, crying to be taken down.

Segment II, Turn 53 (Copleston) Argumentation Stage (Response to Objection,

Second Argument)

[53.1] Well, I can’t really see a parity. If I were saying Every object has a phenom-
enal cause; therefore, the whole series has a phenomenal cause, there would be
a parity. But I’m not saying that; I’m saying, Every object has a phenomenal
cause—if you insist on the infinity of the series—but the series of phenomenal
causes is an insufficient explanation of the series; therefore, the series has not a
phenomenal cause but a transcendent cause.
176 5 Analysis of Segment II: Discussion of Copleston’s …

Predictably, Copleston questions the analogy. For the analogy to work, Copleston
metaphysical argument should be:
(1) Premise 1. Every object has a phenomenal cause.
Premise 2. The series of phenomenal objects is nothing but these objects taken
together.
Conclusion. The series of phenomenal objects has a phenomenal cause.

In this argument, Premise 2 renders Copleston’s equation at 11.2.1 and 11.2.3.

However, Copleston says, his argument is rather the following:

(2) Premise 1. Every object has a phenomenal cause.

Premise 2. The series of phenomenal causes is an insufficient explanation of the series.
Conclusion. The series of phenomenal causes has not a phenomenal cause but a
transcendent cause.

Notice that argument (2) is interestingly different from the argument originally given
at Turn 11. Copleston refers in this turn not to the series of ‘beings’ (objects or events)
but to the series of causes. Let’s call (2) Copleston’s Second Metaphysical Argument.
A big difference relative to the First Metaphysical Argument is that the equation
‘world = everything’ (see comment on 11.3.1) does not make an explicit appearance
in the Second. Nonetheless, the argument as reconstructed is not wholly transparent,
and a close analysis may reveal the hidden work of the equation. After all, Premise
1 of argument (2) does not speak of a series, as does Premise 2 and the Conclusion.
So, there has to be a transition from the phenomenal cause of ‘every object’ to the
series of phenomenal causes.
Now, the series of phenomenal causes either has a first phenomenal cause or not.
If it does, then by definition that first phenomenal cause has no phenomenal cause,
which would be against Premise 1. In fact, the first cause itself (or perhaps the series
itself with a first cause at the beginning) would seem to be a sufficient explanation
of the series, which would defeat the purpose of argument (2). In the interest of
consistency, therefore, we must say that the series cannot have a first phenomenal
cause. In other words, the series of phenomenal causes has to be infinite—at least in
one direction, starting from the present towards the past, for we can for the time being
leave open the question of the end of the world. The traditional idea of the infinity of
the series would naturally bring with it all the pandemonium of the historical debate,
from Aristotle down to Kant’s antinomies, creating no end of trouble.
Finally, Copleston’s argument presupposes that it makes sense to speak of the
universe as a series in the first place. However, the idea of a system in equilibrium,
like a food chain, an ecosystems, or indeed Newton’s solar system, would seem to
indicate that a series, i.e. a linear ordering of causes, might not be the right model for
the universe or any interesting subset of the universe (see my comment on 11.3.1).
The reader may, therefore, be justified in thinking that (2), Copleston’s Second
Metaphysical Argument has sufficiently weak premises for Russell to object to. But,
5 Analysis of Segment II: Discussion of Copleston’s … 177

as in the case of the First Metaphysical Argument, Russell will prefer to focus on the
inferential step itself.

Segment II, Turn 54 (Russell) Argumentation Stage (Objection to Second Argument)

[54.1] Well, that’s always assuming that not only every particular thing in the
world, but the world as a whole must have a cause. For that assumption I see no
ground whatever. If you’ll give me a ground, I’ll listen to it.
We must first observe that Russell does not acknowledge the difference between
Copleston’s Second and his First. He just re-iterates his idea that Copleston is oper-
ating under the assumption that causality is applicable to the whole universe, some-
thing he rejects. In other word, Russell’s criticism points to a hidden assumption,
an unexpressed Toulminian warrant as it were, that allows the transition from the
premises to the conclusion in Copleston’s above argument.
To question this warrant is, of course, the old Kantian objection all over again, yet
without any argumentative apparatus, Kantian or otherwise, something that Cople-
ston had indirectly requested with his counterargument at 51.3. It is true, on the
other hand, that Russell has been assigned the role of an antagonist. So, he is
arguably dispensed from having to produce a counterargument and content himself
with issuing a request to Copleston: ‘You produce an argument for your assumption
(or warrant)’. This request was dimly discernible at 46.1, but at that point Russell
succumbed to the temptation to argue instead.

Segment II, Turn 55 (Copleston) Argumentation Stage (Responde to Objection, Third

Argument), Opening Stage (Role Assignment)

[55.1] Well, the series of events is either caused or it’s not caused. If it is caused,
there must obviously be a cause outside the series. If it’s not caused, then it’s
sufficient to itself, and, if it’s sufficient to itself, it is what I call necessary. But it
can’t be necessary since each member is contingent, and we’ve agreed that the
total has no reality apart from the members; therefore, it can’t be necessary.
Therefore, it can’t be uncaused. Therefore, it must have a cause.
Copleston complies with Russell’s request and produces the Third Metaphysical
Argument, an alternative version of the arguments in Turns 11 and 54. We may
analyse it as follows:
Premise 1. The series of events is either caused or uncaused.
Premise 2. If the series of events is caused, then there is a cause outside the series.
Premise 3. If the series of events is uncaused, then it’s sufficient to itself.
Premise 4. If the series of events is sufficient to itself, then it is necessary.
178 5 Analysis of Segment II: Discussion of Copleston’s …

Premise 5. Each member of the series is contingent.

Premise 6. The series of events has no reality apart from its members.
Conclusion 1 (from premises 4, 5, 6). The series of events is not necessary.
Conclusion 2 (from conclusion 1 and premise 4). The series of events is not sufficient to
itself.
Conclusion 3 (from conclusion 2 and premise 3). The series of events is not uncaused.
Conclusion 4 (from conclusion 3 and premises 1, 2). The series of events has an external
cause.

Note that, in this reconstruction, I have avoided the modal verbs ‘cannot’ and ‘must’,
for they seem merely to express the logical force of deduction (see comments on
11.3.1, 11.3.2, 15.1, 15.3, 16.1). Another important difference between the Third
Metaphysical Argument and the Second is that our old equation ‘world = everything’
(11.3.1) appears as Premise 6, which allows for Conclusion 1, and on and on until
we reach Conclusion 4.
It should be said that, when Copleston formulates Premise 6, and thus the equation,
he explicitly asserts that this was agreed upon by Russell (“we’ve agreed that the
total has no reality apart from the members”). And indeed, if silence is consent, then
Russell seems to be in agreement for a long stretch of the debate (say, from Turn
12 to Turn 50). And even when he finally makes an objection, namely at 52.1, he
withdraws it as soon as Copleston rejects his analogy, and he takes instead refuge
into his quasi-Kantian objection.
[55.2] And I should like to observe in passing that the statement The world is
simply there and is inexplicable can’t be got out of logical analysis.
After presenting the Third Metaphysical Argument, Copleston could have waited
to see whether Russell would present a new objection and then produce a Fourth
Argument, and so on. Instead, Copleston challenges Russell to put his cards on the
table. If Russell says that The world is simply there, as Copleston had invited him
to admit publicly and explicitly at 47.4, then he is no longer an antagonist but has
his own standpoint, and with it the obligation to present an argument. As yet, he has
not given one, except to say that the concept of cause is inapplicable to the universe.
Now, this argument exists but is rather Kantian. So, if Russell is no Kantian, then
the only other source that may be relevant is mathematical logic itself. This is why
Copleston counters that such a metaphysical heavy proposition as that the world
has no explanation cannot be obtained from logical analysis alone. It is clear that
Copleston is issuing to Russell a challenge to take position on the question whether
the metaphysical standpoint that the world has no explanation can
be derived from pure logic. Call this Q45. It is related to previous questions:
Q13 (12.3), Q23-26 (16.3, 19.1, 19.6), and Q30 (20.3); and it will become a major
protest against Russell’s mode of argumentation.
Copleston’s methodological objection could be spelt out as the following
argument:
5 Analysis of Segment II: Discussion of Copleston’s … 179

Premise 1. Russell seems to think that from logical analysis alone on can derive the
proposition that the world is simply there and has no explanation.
Premise 2. That proposition cannot be derived from logical analysis alone.
Conclusion. Russell’s position is untenable, and he should produce a proper argument in
favour of his proposition.

This is a delicate situation. Copleston is trying to force Russell to own his standpoint
and to defend it as a Protagonist. We are now not discussing the arguments of the
parties, but their dialectical roles.

Segment II, Turn 56 (Russell) Argumentation Stage (Objection)

[56.1] I don’t want to seem arrogant, but it does seem to me that I can conceive
things that you say the human mind can’t conceive.
This is, on the face of it, an exceedingly strange remark. Copleston has never said
that he was unable to conceive something. The only explanation for Russell’s unex-
pected argumentative move is the assumption that Copleston’s request to give an
argument for the proposition that The world is simply there can only stem from his
inability to conceive uncaused events. Note that Russell does not register Copleston’s
methodological objection at 55.2. What is he then trying to do instead?
[56.2] As for things not having a cause, the physicists assure us that individual
quantum transitions in atoms have no cause.
The cat is out of the bag. Copleston had tried, at 55.2 to make Russell argue in favour
of The world is simply there and has no explanation. Copleston hopes to show that,
if Russell produces an argument, it would come from pure logical analysis. Instead
of doing what he is requested to do, Russell takes advantage of his role of antagonist
in order to lure him into the open as using Kant’s principle of causality (All events
have a cause). This will give the debate a wholly new turn.
The way Russell goes about it is by referring to the at the time of debate still
relatively recent creation of quantum theory, in which, according to the specialists, it
is impossible to assign a cause to sub-atomic events, such as the ‘quantum leaps’ of
electrons or the emission of particles by a radioactive element (for a philosophically
interesting discussion, see Heisenberg, 1971, Chap. 10; cf. Crull and Bacciagaluppi
2017). The argument seems to go like this:
Premise. Physicists assure us that individual quantum transitions in atoms have no cause.
Conclusion. Some events in the world have no cause.

It is rather an argument of authority, at least as Russell presents it; and besides, it

sounds rather irrelevant, at least as a reaction to Copleston’s Turn 55, where what
is at issue is the extra-mundane cause of the world and not, as usual with Russell
(remember Table 5.2 in my comment to 38.1), the cause of certain particular events in
180 5 Analysis of Segment II: Discussion of Copleston’s …

the world. How is one supposed to go from that to the rejection of Copleston’s ques-
tions and arguments, remains quite unclear, unless we consider 56.2 as a belated
reaction to Premise 1 in the Second Metaphysical Argument (‘every object has
a phenomenal cause’, 53.1) In any case, the question arises whether certain
events in the world are uncaused. Let’s call this Q46.

Segment II, Turn 57 (Copleston) Argumentation Stage (Response to Objection)

[57.1] Well, I wonder now whether that isn’t simply a temporary inference.
Copleston does not reject the quantum theory example as irrelevant but takes it in
his stride by raising a question (“I wonder”) which is different from Q46, namely
whether the assertion that quantum transitions in atoms are not
caused is a temporary inference which will be later revoked. Let’s
call it Q47. If we take the affirmative, as Copleston seems to think, then no firm
answer to Q46 is possible yet.
The reader may remember that, at the time of this debate, there was still some
controversy about Q47. Einstein, for instance, had still some hope that unknown,
hidden variables could causally produce the sub-atomic events that seemed to be
uncaused; and, just a few years after the broadcast, the first hidden-variable quantum
theory would be published (Bohm 1952). Whatever the consensus of physicists in the
twenty-first century may be, we must think back to what could plausibly be argued
in 1948.

Segment II, Turn 58 (Russell) Argumentation Stage (Renewed Objection)

[58.1] It may be, but it does show that physicists’ minds can conceive it.
Russell does not dispute Copleston’s idea and refrains from taking position in regard
to Q46. Instead, he suggests that, at 56.1, he was not proposing the argument I
reconstructed in my comment but rather the following one:
Premise. Physicists assure us that individual quantum transitions in atoms have no cause.
Conclusion. Some people can conceive events in the world which have no cause.

This argument still has a certain air of irrelevance, in that Russell is not thereby
engaging Copleston’s talk of an extra-mundane cause of all events in the world. Even
if the argument given by Copleston would contain a premise on the inconceivability
of an uncaused world (which it does not), the uncaused nature of quantum events
would be unconnected with Copleston’s point.
The question raised by Russell, whether relevant or not, is different from both Q46
and Q47: whether it is possible to conceive an event in the world
that is uncaused. Let’s call it Q48. Russell upholds the affirmative, whereas
Copleston does not seem to have an opinion one way or the other, although Russell
5 Analysis of Segment II: Discussion of Copleston’s … 181

thinks he upholds the negative, possibly because he thinks that Copleston is, in this
respect, a Kantian.

Segment II, Turn 59 (Copleston) Argumentation Stage (Response to Renewed

Objection)

[59.1] Yes, I agree, some scientists—physicists—are willing to allow for indeter-

mination within a restricted field. But very many scientists are not so willing. I
think that Professor Dingle, of London University, maintains that the Heisen-
berg uncertainty principle tells us something about the success (or the lack of it)
of the present atomic theory in correlating observations, but not about nature
in itself, and many physicists would accept this view.
Copleston does not go into Russell’s imaginary charge of inconceivability and prefers
to reply by insisting on the fact that there exists a controversy among physicists on
causal indeterminacy at the sub-atomic level. He illustrates the opposition by quoting
a physicist of those times who wrote widely on quantum theory. The argument seems
to be:
Premise. Professor Dingle thinks that the quantum theory is not about how nature is but only
about how successful or unsuccessful we are on correlating observations.
Conclusion. There is no agreement among physicists as to the question whether quantum
events are causally determinate or not.

From the conclusion to this argument it follows, as a corollary, that no definite answer
to Q46 can be given.
[59.2] In any case, I don’t see how physicists can fail to accept the theory in
practice, even if they don’t do so in theory. I cannot see how science could
be conducted on any other assumption than that of order and intelligibility in
nature.
With this part of Copleston’s intervention, it becomes possible to identify what is
worrying him. Even though Russell’s interventions at Turns 56 and 58 are somewhat
alien to the whole march of the discussion up to and including Turn 55, Copleston
finds a way to interpret it as relevant. At this point, the argument goes like this:
Premise. For science to be properly conducted, it is necessary to assume that nature is ordered
and intelligible, and this includes, crucially, the unremitting search for causes.
Conclusio. Scientists cannot abandon, in their scientific practice, the unremitting search for
causes.

In this argument, it is, first of all, quite remarkable to hear, again, Kant’s voice, for
it was he who insisted, as no other philosopher before, on the tight link connecting
the assumptions of order and intelligibility (‘ideas of reason’), the law of causality
(one of the ‘principles of the understanding’), and the very possibility of science. Is
182 5 Analysis of Segment II: Discussion of Copleston’s …

this what Russell was trying to get Copleston to admit when he started talking about
conceivability?
The other important aspect of this argument is that Copleston knows that what
I have called ‘the unremitting search for causes’ is also a crucial presupposition of
the metaphysical argument. That is why he is defending it from Russell’s quantum
theory example.
The question raised here, call it Q49, is whether physical research can
proceed without searching for causes. Copleston is very definite in taking
an affirmative position here, whereas Russell is not thus far quite straightforward
about his position. Q49 is very close to a further question, which does not get fully
expressed, namely whether every event has a cause, to use Kant’s formula-
tion of the principle of causality. Let’s call it Q50. Here the important thing to keep
in mind is that, according to Kant, this principle only applies to objects of experience
in the world but not to the world itself; to every particular event in the world but not
to the whole series.
[59.3] The physicist presupposes, at least tacitly, that there is some sense in
investigating nature and looking for the causes of events, just as the detective
presupposes that there is some sense in looking for the cause of a murder.
Copleston now extends the case from physical science to criminal investigation. This
is not an analogy, as though Copleston was producing something like the following
argument:
Premise 1. Criminal investigators need to presuppose that it makes sense to look for causes
in order to do their work.
Premise 2. Physical research is like criminal investigation.
Conclusion. Physicists need to presuppose that it makes sense to look for causes in order to
do their work.

The argument, if it is one, can better be reconstructed as follows:

Premise 1. Physicists need to presuppose that it makes sense to look for causes in order to
do their work.
Premise 2. Criminal investigators need to presuppose that it makes sense to look for causes
in order to do their work.
Conclusion. If two fields as different as physical research and criminal investigation both
need that presupposition, then causality is not lightly to be questioned.

Reasons for this reconstruction: (1) Premise 1 was already proved at 59.2; (2)
although the conclusion is not expressed right away, something like that is behind
the next part of this intervention.
[59.4] The metaphysician assumes that there is sense in looking for the reason
or cause of phenomena,
Now it is possible to see in perspective why criminal investigators (a homely but
effective example) were brought up. It is almost an incomplete induction hinting at
a general conclusion, something like: Everyone needs to presuppose that it makes
sense to look for causes, almost as if no thinking about objects and events can proceed
without it.
5 Analysis of Segment II: Discussion of Copleston’s … 183

It is interesting to note that Russell in 1912 had already argued that ‘the word
“cause” is so inextricably bound up with misleading associations as to make its
complete extrusion from the philosophical vocabulary desirable” and that an inquiry
was needed about ‘what principle, if any, is employed in science in place of the
supposed “law of causality” which philosophers imagine to be employed’ (Russell,
1912a, p. 1). This paper is a resounding No to Q50, but it is not repeated during
the debate, where Russell expresses himself much more cautiously. In any case, I
am bringing this up only to convey that Russell’s negative attitude towards causal
talk thus precedes by quite a long time the emergence of quantum theory and its
doubts concerning the applicability of causality to sub-atomic events, which Russell
adduced at 56.2. Nonetheless, Russell had accepted causes up to this point; in fact, it
was him who introduced them for the first time at 32.1, whereas Copleston preferred
to talk about reasons. It almost looks as though Russell has no argumentative strategy
to begin with and is just making it up as the discussion goes.
[59.5] and, not being a Kantian, I consider that the metaphysician is as justified
in his assumption as the physicist.
Copleston is most emphatic about his difference with Kant. No other modern philoso-
pher had made more to put forward the law of causality as an indispensable presup-
position of physical science than Kant, but he had indeed rejected its use by the
metaphysician, because the metaphysicians tried to apply it beyond any possible
experience, to transcendent ‘objects’ such as the universe. I noted before that this
Kantian position is assumed by Russell, even though he was a very sharp critic of
Kant’s philosophy.
This all raises a crucial question for Copleston, namely whether metaphysics
can proceed without searching for causes, or Q51, parallel to Q49. Taking
the two questions together, it seems possible to distinguish between (1) Copleston,
who assumes the need of causal talk in both physics and metaphysics; (2) Kant, who
accepts it in physics but rejects it in metaphysics; and (3) Russell, who might not
only reject it in metaphysics but also in physics, and yet is not straightforward about
it. Copleston, who must have been aware that (3) is Russell’s position, desperately
tries to make Russell take a stand, so as to have an orderly discussion. This desire
is especially justified, as the whole matter was put on the table by Russell, not by
Copleston.
[59.6] When Sartre, for example, says the world is gratuitous, I think that he
has not sufficiently considered what is implied by “gratuitous”.
For some reason, Copleston is very much troubled by what he takes to be Sartre’s
position. It may be of interest to note that Sartre’s idea of gratuitousness is not
metaphysical in Copleston’s sense. One can, of course, interpret the famous passage
from Nausea (1938) as supporting the idea that the universe ‘is just there’, yet the
feeling Sartre is conveying in that passage is not at all different from the feeling of a
Pascal. Now, that feeling is ‘metaphysical’ in the sense that it makes the Leibnizian
question (Q44’, which is equivalent to Q44) poignantly meaningful. Nonetheless,
184 5 Analysis of Segment II: Discussion of Copleston’s …

the meaningfulness of the question is, like everything in Sartre, purely personal
and subjective: it has no ‘cosmic’ dimension. This is borne out by the frequent talk
of gratuitousness in Being and Nothingness (1943). So Copleston’s argument (the
implications of ‘gratuitousness’ are awful, therefore Sartre cannot have meant what
he said) seems to be a misunderstanding. We should rather say that Sartre’s question
is appreciably different from Copleston’s.

Segment II, Turn 60 (Russell) Argumentation Stage

[60.1] I think there seems to me a certain unwarrantable extension here. A

physicist looks for causes; that does not necessarily imply that there are causes
everywhere.
Russell ignores Q51 and takes a noncommittal position on Q49, by putting forward a
standpoint, namely: Even if physicists look for causes, it does not follow that causes
are everywhere.
[60.2] A man may look for gold without assuming that there is gold everywhere;
if he finds gold, well and good; if he doesn’t, he’s had bad luck. The same is true
when the physicists look for causes
It is not easy to pin down what Russell is saying here. He seems to be giving a reason
for the standpoint offered at 60.1. The argument seems to be based on an analogy:
Premise 1. People who look for gold don’t assume that gold is everywhere.
Premise 2. Physicists who look for causes are similar to people who look for gold.
Conclusion 1. Physicists don’t assume that causes are everywhere.
Conclusion 2. Causes are not everywhere.

Take first the argument that ends in Conclusion 1. It is a strange argument. First,
people do not look for gold everywhere but always somewhere. In that respect, they
may be compared to physicists who also look for causes in some definite way, e.g.
by performing a definite experiment. Of course, gold-diggers assume and even know
that they may be looking in the wrong place. Analogously, physicists know that their
particular experiment may fail to reveal the cause they are looking for. However, all
this reasoning is curiously irrelevant, for the assumption that interests Copleston is
not that physicists may fail to find the cause of an event, but rather that, given an
event, they should look for its cause because it must have one. If a gold-digger fails
to find gold where he is looking, then he starts looking somewhere else, but he never
thinks: ‘Maybe there is no gold anywhere.’ The same is true of physicists. Indeed,
this is why there was such a huge controversy around quantum theory and the law
of causality.
However, Conclusion 2, which seems to be an authority-based corollary of
Conclusion 1, is much stronger. In fact, it is so strong that it is not as such expressed
but only insinuated. If Russell would fully express it, then it would amount to a
negative answer to Q50. As it is, it only concerns Q49.
5 Analysis of Segment II: Discussion of Copleston’s … 185

[60.3] As for Sartre, I don’t profess to know what he means, and I shouldn’t
like to be thought to interpret him; but for my part, I do think the notion of the
world having an explanation is a mistake. I don’t see why one should expect it
to have <one> , and I think what you say about what the scientist assumes is
an over-statement.
Russell does not want to go into an exegesis of Sartre, which is probably quite wise.
In his History of Western Philosophy, published less than two years before the BBC
broadcast, he does not mention either him or any other philosopher associated to
him. Apart from this, he re-asserts his position: it is a mistake to believe that the
universe has an explanation (No to Q44) and it is an exaggeration to believe that
physicists assume that every event has a cause (which would be a No to Q50). No
new argument is offered at this point.

Segment II, Turn 61 (Copleston) Argumentation Stage

[61.1] Well, it seems to me that the scientist does make some such assumption.
When he experiments to find out some particular truth, behind that experiment
lies the assumption that the universe is not simply discontinuous. There is the
possibility of finding out a truth by experiment. The experiment may be a bad
one, it may lead to no result, or not to the result that he wants, but at any rate
there is the possibility, through experiment, of finding out the truth that he
assumes. And that seems to me to assume an ordered and intelligible universe.
Copleston is now a mirror image of Russell. He just re-states his position, as Russell
re-states his (the former Yes, the latter No to Q49). No further argument is given. Of
course, there is an argument in this passage, but it is the same one we know from 59.2.
The positions of the discussants have begun to harden and the debate is degenerating.

Segment II, Turn 62 (Russell) Argumentation Stage

[62.1] I think you’re generalizing more than is necessary. Undoubtedly the scien-
tist assumes that this sort of thing is likely to be found and will often be found.
He does not assume that it will be found, and that’s a very important matter in
modern physics.
Again, Russell re-states his position; but he does not offer any new argument. The
situation is the same as it was at 60.2. He offers a description of what he thinks
happens in physics. The only remarkable aspect in this re-formulation if his position
is that, whereas at 60.2 he defended a general position (as he did in his famous 1912
186 5 Analysis of Segment II: Discussion of Copleston’s …

anti-cause paper), he now believes himself to be confirmed by ‘modern physics’, i.e.

by quantum theory, but without recognising that there was a controversy as to the
real status of causality in physics.

Segment II, Turn 63 (Copleston) Argumentation Stage

[63.1] Well, I think he does assume or is bound to assume it tacitly, in practice.

It may be that, to quote Professor Haldane,

when I light the gas under the kettle, some of the water molecules will fly off as vapor, and
there is no way of finding out which will do so,

but it doesn’t follow necessarily that the idea of chance must be introduced except
in relation to our knowledge.
After looking in all books and papers by the celebrated English biologist John
B. S. Haldane which are available on the internet, I have been unable to locate this
passage or a similar one. It probably does not matter, because it is easy to recognise
the source of the example: statistical mechanics, which was designed to explain
the thermodynamic behaviour of huge collections of particles, e.g. water molecules
being heated. The argument that Copleston seems to convey here is:
Premise 1. Statistical mechanics introduces chance only in relation to our knowledge.
Premise 2. Statistical mechanics is like quantum theory.
Conclusion. Quantum theory introduces chance only in relation to our knowledge.

The problem with Copleston’s example, and with this argument, is that statistical
mechanics was devised precisely as a statistical substitute in order to avoid the
difficulties of calculating the behaviour of so many individual particles.
In statistical mechanics, chance is indeed introduced only in relation to our knowl-
edge. Each individual particle is acted upon by some definite force, but we prefer to
calculate the bulk of the interactions. Our lack of knowledge is thus a practical affair,
not a matter of principle, as allegedly in quantum theory. (I say allegedly, because,
even though today’s consensus is that chance is real and not just knowledge-related,
there have always been discordant voices, and this was even more so at the time of the
broadcast). In any case, statistical mechanics is not like quantum theory. Therefore,
Copleston’s example is irrelevant to the physical theory that Russell means. It is,
however, an irrelevance following upon another, for Russell has never tried to argue
that quantum theory has anything to do with the question that Copleston is trying to
answer.
The question arising at this point is whether statistical mechanics is sufficiently
like quantum theory to allow for a proper answer to Q47.
5 Analysis of Segment II: Discussion of Copleston’s … 187

Segment II, Turn 64 (Russell) Argumentation Stage

[64.1] No, it doesn’t—at least if I may believe what he says.

Russell seems to concede to Copleston his point, without registering any difference
between statistical mechanics and quantum theory. But this concession does not
amount to a definite answer to Q47. This is in harmony with the hesitancy observable
at 58.1.
[64.2] He’s finding out quite a lot of things… the scientist is finding out quite a
lot of things that are happening in the world, which are, at first, beginnings of
causal chains: first causes which haven’t in themselves got causes. He does not
assume that everything has a cause.
Russell improvised, and somewhat halting, reply is very much in the spirit of empiri-
cism. Scientists are humble: they find causes now and then. When they take an event
and find its cause, then that cause has not itself a cause yet—until scientists try to
find the cause of the cause. But this does not imply that everything has a cause. But
why, could Copleston argue, does he feel entitled to look for a cause if not because
he assumes that everything has a cause?

Segment II, Turn 65 (Copleston) Argumentation Stage

[65.1] Surely that’s a first cause within a certain selected field. It’s a relatively
first cause.
Copleston is not happy with the expression “first cause”, which is so important in
traditional metaphysics (prima causa). So, he relativises it. Note, however, that this
relativization does not contradict what Russell himself had said at 64.2.

Segment II, Turn 66 (Russell) Argumentation Stage

[66.1] I don’t think he’d say so.

Russell is here contradicting himself. He himself said, at 64.2, that “the scientist is
finding out quite a lot of things that are happening in the world, which are, at first,
beginnings of causal chains: first causes which haven’t in themselves got causes”, so
these are first causes only relatively, until research goes on looking for their causes.
They are beginning of causal chains, as far as research has gone. Thus, Koch’s bacillus
causes tuberculosis, and that’s good enough, but we may, of course, assume that there
is a cause for Koch’s bacillus, only that research is for the time being left aside. It
is a relatively first cause. And that is what Russell himself said. So how would the
scientist not say that it is?
188 5 Analysis of Segment II: Discussion of Copleston’s …

[66.2] If there’s a world in which most events, but not all, have causes, he will
then be able to depict the probabilities and uncertainties by assuming that this
particular event you’re interested in probably has a cause. And since in any case
you won’t get more than probability that’s good enough.
This is not an argument that supports saying that there are only probabilities and
not causes. In fact, Russell himself says that the scientist (‘he’) assumes ‘that this
particular event… probably has a cause’. That the scientist now and then rests content
with probabilities does not mean that ‘he’ is not assuming causality.

Segment II, Turn 67 (Copleston) Argumentation Stage, Opening Stage

[67.1] It may be that the scientist doesn’t hope to obtain more than probability,
but in raising the question he assumes that the question of explanation has a
meaning.
Within the sub-discussion concerning Q49, Copleston repeats the same point: resting
content with probabilities does not mean that the question of explanation has no
meaning nor that it does not make sense to look for a cause.
[67.2] But your general point then, Lord Russell, is that it’s illegitimate even
to ask the question of the cause of the world?
Copleston sees that Russell’s position has hardened and no further argument will be
given. So, Copleston asks Russell point-blank whether his position is that asking the
question of the cause of the world has no meaning. This was Q31, a question raised at
21.3 by Copleston, answered in the affirmative by Russell at 22.1, and then retracted
at 40.1. This is now put forward, for the first time, as a material starting point which
is indispensable to go on discussing.

Segment II, Turn 68 (Russell) Opening Stage

[68.1] Yes, that’s my position.

The retraction noted at 40.1 is here confirmed. Russell definitely says No to Q31
and so completes the move from a Gamma (there is no way to prove that God exists
or that God does not exist) to a Delta stance (the question whether God exists is
illegitimate).

Segment II, Turn 69 (Copleston) Opening Stage

[69.1] Well, if it’s a question that for you has no meaning, it’s of course very
difficult to discuss it, isn’t it?
5 Analysis of Segment II: Discussion of Copleston’s … 189

This intervention acknowledges that the discussion of Q1, as it has developed around
Copleston’s metaphysical argument, cannot be continued if acquiescence to Q31 is
denied. The meta-argument goes like this:
Premise 1. A discussion about the existence of God (described as the cause of the world)
cannot be carried out unless it is conceded that it is legitimate to ask the question of the cause
of the world.
Premise 2. Russell does not concede that it is legitimate to ask the question of the cause of
the world.
Conclusion. Copleston and Russell cannot continue discussing about the existence of God
(described as the cause of the world).

Note that Copleston expresses this momentous conclusion in the shape of a typically
British understatement, as though it was just a matter of great difficulty instead of
sheer impossibility.

Segment II, Turn 70 (Russell) Opening Stage (Material Starting Point, Procedural
Starting Point)

[70.1] Yes, it is very difficult.

Russell accepts that Q31 has to be affirmed to have a meaningful discussion. The
argument set forth in my comment to 69.1 is thereby accepted.
[70.2] What do you say—shall we pass on to some other issue?
Russell proposes a procedural starting point: to leave the discussion of Q1 by way
of the metaphysical argument for God’s existence and to go on to the discussion of
Q1 in relation to the other arguments announced by Copleston.

Segment II, Turn 71 (Copleston) Opening Stage (Procedural Starting Point)

[71.1] Let’s.
Copleston expresses his agreement with 70.2.

References

Barcan, R. C. (1946a). A functional calculus of first order based on strict implication. Journal of
Symbolic Logic 11 (1), 1-16.
Barcan, R. C. (1946b). The deduction theorem in a functional calculus of first order based on strict
implication. Journal of Symbolic Logic 11 (4), 115-118.
Barcan, R. C. (1947). The identity of individuals in a strict functional calculus of second order.
Journal of Symbolic Logic 12 (1), 12-15.
190 5 Analysis of Segment II: Discussion of Copleston’s …

Bohm, D. (1952). A suggested interpretation of the quantum theory in terms of ‘hidden’ variables.
Physical Review 85 (2), 166-193.
Broad, C. D. (1924). Critical and speculative philosophy. In J. H. Muirhead (Ed.), Contemporary
British philosophy: Personal statements (vol. I, pp. 77-100). London: George Allen & Unwin.
Carnap, R. (1931). Überwindung der Metaphysik durch logische Analyse der Sprache. Erkenntnis
2, 219–241. [English translation: “The elimination of metaphysics through logical analysis of
language”, in A. J. Ayer (Ed.), Logical positivism (pp. 60–81). New York: Free Press, 1959.]
Couturat, L. (1901). La logique de Leibniz, d’après des documents inédits. [The logic of Leibniz,
according to some unpublished documents.] Paris: Félix Alcan.
Crull, E., and Bacciagaluppi, G. (Eds.) (2017). Grete Hermann: Between physics and philosophy.
Cham: Springer.
Davies, P. C. W. (1979). The forces of nature. Cambridge, UK: Cambridge University Press.
Dieter, O. A. L. (1940). Stasis. Speech Monographs17, 345-369.
Dreben, B. (1991). Tautology: How not to use a word. Synthese 87, 23-49.
Durham, I. T., and Rickles, D. (2017). Information and interaction: Eddington, Wheeler, and the
limits of knowledge. Cham: Springer.
Eemeren, F. H. van, and Grootendorst, R. (1993). The history of the argumentum ad hominem since
the seventeenth century. In E. C. W. Krabbe, R. J. Dalitz, & A. Smit (Eds.), Empirical logic and
public debate. Essays in Honour of Else M. Barth (pp. 49–68). Amsterdam: Rodopi.
Friedman, M. (2000). A parting of the ways: Carnap, Cassirer, and Heidegger. Chicago: Open
Court.
Geach, P. T. (1972). Logic matters. Oxford: Basil Blackwell.
Haley, W. (1948). ‘Copleston/Russell discussion.’ Memorandum of the Director General, Russell
Archives, McMaster University, Hamilton, Ontario.
Heidegger, M. (1978). Metaphysische Anfangsgründe der Logik im Ausgang von Leibniz, edited
by K. Held. Frankfurt am Main: Vittorio Klostermann. [English translation: The metaphysical
foundations of logic. Bloomington, IN: Indiana University Press, 1984.]
Heisenberg, W. (1969). Der Teil und das Ganze: Gespräche im Umkreis der Atomphysik. Munich:
Piper. [English translation: Physics and beyond: Encounters and conversations. New York: Harper
& Row.]
Heyting, A. (1930). Die formalen Regeln der intuitionistischen Logik. [The formal rules of intu-
itionistic logic.] Sitzungsberichte der Preussischen Akademie der Wissenschaften: Physikalisch-
mathematische Klasse, 42–56.
Hintikka, J. (1962). Cogito, ergo sum: Inference or performance? Philosophical Review 71 (1),
3-32.
Hintikka, J. (1966). The logic of existence and necessity I: Existence. The Monist 50 (1), 55-76.
Kant, I. (1787). Critik der reinen Vernunft. [Critique of pure reason.] 2nd edition. Riga: Johann
Friedrich Hartknoch.
Kenny, A. (1979). The God of the philosophers. Oxford: Clarendon Press.
Kment, B. (2017). Varieties of modality. Stanford Encyclopedia of Philosophy, https://plato.sta
nford.edu/entries/modality-varieties/.
Kripke, S. (1972). Naming and necessity. In D. Davidson and G. Harman (Eds.), Semantics of
natural language (pp. 253-355). Dordrecht: Reidel.
Lewis, C. I. (1914). The calculus of strict implication. Mind (N.S.) 23 (90)240–247.
Lewis, C. I. (1918). A survey of symbolic logic. Berkeley: University of California Press.
Łukasiewicz J. (1920). O logice trójwartościowej. Ruch filozoficzny 5, 170–171. [English translation:
On three-valued logic, in L. Borkowski (Ed.), Selected works by Jan Łukasiewicz (pp. 87–88).
Amsterdam: North–Holland, 1970.]
Moraux, P. (1968). La joute dialectique d’après le huitième livre des Topiques. In G. E. L. Owen
(Ed.), Aristotle on dialectic: The ‘Topics’ (pp. 277-311). Oxford: Clarendon Press.
Nadeau, R. (1959). Classical systems of stases in Greek: Hermagoras to Hermogenes. Greek,
Roman, and Byzantine Studies 2, 53-71.
References 191

Nelson, L. (1904). Die kritische Methode und das Verhältnis der Psychologie zur Philosophie: Ein
Kapitel aus der Methodenlehre. Abhandlungen der Fries’schen Schule (N.F.) 1, 1–88. [English
translation: The critical method and the relation of psychology to philosophy: An essay in method-
ology, in Socratic method and critical philosophy, translated by T. K. Brown III (pp. 105–157).
New Haven: Yale University Press, 1949.]
Nuchelmans, G. (1993). On the fourfold root of the argumentum ad hominem. In E. C. W. Krabbe,
R. J. Dalitz, & A. Smit (Eds.), Empirical logic and public debate. Essays in Honour of Else M.
Barth (pp. 37–47). Amsterdam: Rodopi.
Plantinga, A. (1978). The nature of necessity. Oxford: Clarendon Press.
Popper, K. (1976). The myth of the framework. In J. C. Pitt and & M. Pera (Eds.), Rational changes
in science (pp. 35–62). Dordrecht: Reidel.
Popper, K. (1979). Die beiden Grundprobleme der Erkenntnistheorie. Tübingen: Mohr Siebeck.
[English translation: The two fundamental problems of the theory of knowledge. London:
Routledge, 2009.]
Post, E. L. (1921) Introduction to a general theory of elementary propositions. American Journal
Mathematics 43, 163-185.
Putnam, H. (1975). The meaning of ‘meaning’. In K. Gunderson (Ed.), Language, mind, and
knowledge [Minnesota studies in the philosophy of science, Vol. 7] pp. 131–193). Minneapolis:
University of Minnesota Press.
Russell, B. (1900). A critical exposition of the philosophy of Leibniz, with an appendix of leading
passages. Cambridge, UK: Cambridge University Press.
Russell, B. (1901). Mysticism and logic. Hibbert Journal12, 780-803.
Russell, B. (1905). On denoting. Mind (N.S.) 14 (56), 479–493.
Russell, B. (1912a). On the notion of cause. Proceedings of the Aristotelian Society (N.S.) 13, 1–26.
Russell, B. (1912b). The problems of philosophy. London: Williams and Norgate.
Russell, B. (1917). Mysticism and logic, and other essays. London: George Allen & Unwin.
Russell, B. (1918). The philosophy of logical atomism. The Monist
Russell, B. (1924). Logical atomism. In J. H. Muirhead (Ed.), Contemporary British philosophy:
Personal statements (vol. I, pp. 357-383). London: George Allen & Unwin.
Russell, B. (1927). ‘Why I am not a Christian.’ Lecture delivered at Battersea Town Hall, under the
auspices of the South London Branch of the National Secular Society.
Russell, B. (1972). The philosophy of logical atomism. London: Fontana.
Russell, B. (2004). Why I am not a Christian, and other essays on religion and related subjects.
London: Routledge.
Schopenhauer, Arthur (1813). Über die vierfache Wurzel des Satzes vom zureichenden Grunde.
Rudolfstadt: Hof-, Buch- und Kunsthandlung. [Revised edition: Frankfurt am Main: Johann
Christian Hermann, 1847. English translation: On the fourfold root of the principle of sufficient
reason, and other writings. New York: Cambridge University Press, 2015.]
Steup, M., and Sosa, E. (Eds.) (2005). Contemporary debates in epistemology. Malden, MA:
Blackwell.
Weinberg, S. (1993). Dreams of a final theory: The search for the fundamental laws of nature. New
York: Random House.
Whitehead, A. N., and Russell, B. (1910). Principia mathematica, vol I. Cambridge, UK: Cambridge
University Press.
Wittgenstein, L. (1921). Logisch-philosophische Abhandlung. Annalen der Naturphilosophie 14
(3–4), pp. 185–262. [Renamed Tractatus logico-philosophicus by G. E. Moore in 1922, twice
translated into English, and printed many times.]
Yrjönsuuri, M. (1993). Aristotle’s Topics and medieval obligational disputations. Synthese 96 (1),
59-82.
Chapter 6
Analysis of Segment III: Discussion
of Copleston’s Religious Argument

Abstract Segment III of the debate is quite brief (Turns 71-91) and is dedicated
to the argument from religious or mystical experience to God’s existence. Neither
this argument nor the argument from moral experience are presented by Copleston as
proofs but only as inferences to the best explanation. Although Copleston starts with a
reasonably clear example of what he calls religious experience, Russell keeps substi-
tuting different kinds of experience which appear to be quite unlike that intended by
Copleston. The debate is never really engaged in the original terms and slowly shifts
towards the moral argument.

Segment III, Turn 71 (Copleston) Opening Stage (Procedural Starting Point, Material
Starting Point: Clarification), Argumentation Stage

[71.2] Well, perhaps I might say a word about religious experience, and then we
can go on to moral experience.
Copleston begins by proposing an order of discussion (a procedural starting point):
first, the argument for God’s existence from religious (later on, especially mystical)
experience; and then, the argument from moral experience. These arguments are
quite simple in principle:
(1) Premise. There is religious (mystical) experience.
Conclusion. God exists.
(2) Premise. There is moral experience.
Conclusion. God exists.

The first question that arises here is: what kind of argument these are,
or rather, more specifically, whether they have the same status as the
metaphysical argument or not. We can refer to it as Q52.
[71.3] I don’t regard religious experience as a strict proof of the existence of
God, so the character of the discussion changes somewhat, but I think it’s true
to say that the best explanation of it is the existence of God.
Copleston immediately answers Q52 in regard to the religious experience: this argu-
ment is not a proof or a demonstration, but what is often called an inference to the
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 193
F. Leal and H. Marraud, How Philosophers Argue, Argumentation Library 41,
https://doi.org/10.1007/978-3-030-85368-6_6
194 6 Analysis of Segment III: Discussion of Copleston’s …

best explanation. Later on, he will say that the same is true of the argument from
moral experience (see 121.4).
Although the matter is controversial, it is often said that this means we are dealing
here with abductive arguments, or abductions. We shall come back to that when the
argument is more clearly delineated (see comment on 77.4). In any case, the careful
specification of the kind of argument Copleston wants to use is also a procedural
starting point.
[71.4] By religious experience I don’t mean simply feeling good. I mean a loving,
but unclear, awareness of some object which irresistibly seems to the experi-
encer as something transcending the self, something transcending all the normal
objects of experience, something which cannot be pictured or conceptualized,
but of the reality of which doubt is impossible—at least during the experience.
Following a very good dialectical method, Copleston tries first to describe what he
means by ‘religious experience’. It is not a matter of mere feeling. It is an awareness
of an object. This is philosophically important because it can be argued that mere
feelings do not necessarily have an object, are not intentional states of mind or
propositional attitudes. Copleston wants to assert that religious experiences do have
an object. Of course, having an object does not imply that the object exists. But it is
a first step.
The object of religious awareness has certain attributes: it transcends the self, it
transcends all normal objects of experience, it cannot be put into pictures, it cannot
be put into concepts. Besides, the experience is accompanied by absolute certainty
that its object is real. This is what Copleston means. When considering the argument
(1) of 71.2, I beg readers to keep it in mind as the meaning of ‘religious experience’.
We shall see that Russell will not object to it, but he will nonetheless ignore it at his
own pleasure.
On the other hand, Copleston also says that that powerful experience of a Presence
is tinged with emotion, in particular with love—mystical love. However, that emotion
is not just ‘feeling good’. I insist on this aspect of Copleston’s description of religious
experience, because love will appear several times during the discussion, although
not really in the shape of mystical love, which is what Copleston has in mind (see in
particular comment on 88.2).
All this amounts to a substantial material starting point.
[71.5] I should claim that cannot be explained adequately, and without residue,
simply subjectively. The actual basic experience at any rate is most easily
explained on the hypothesis that there is actually some objective cause of that
experience.
Copleston explains, secondly, what allows for the transition from the Premise to
the Conclusion in argument (1) of 71.2. He produces two propositions as quasi-
Toulminian warrants. The first is that an experience such as is described at 71.4
cannot be explained adequately and completely by saying it is only subjective. The
second is that the easiest way to explain the experience described is to say that a real
object causes the experience. This is the argument. We may represent it as follows:
6 Analysis of Segment III: Discussion of Copleston’s … 195

Premise 1. There are religious (mystical) experiences (as defined).

Premise 2. The best explanation for religious (mystical) experiences is that God exists.
Conclusion. God exists.

The question raised here is whether the best explanation of religious

(mystical) experience is that god exists. Let’s call it Q53. Copleston clearly
thinks that it is the best explanation (for only God could cause such experiences).
Russell would not immediately deny Premise 2, although in the end he will, by
suggesting other causal explanations which he will eventually say are better than the
existence of God.
Before going on, I want to emphasize that the experience that we are charged
with explaining is not what anybody (even if he is Russell) decides to call ‘religious
experience’ but rather what Copleston has described under that term. His argument
depends on which phenomenon he is trying to explain.

Segment III, Turn 72 (Russell) Argumentation Stage (Meta-Argumentation)

[72.1] I should reply to that line of argument that the whole argument from our
own mental states to something outside us, is a very tricky affair. Even where
we all admit its validity, we only feel justified in doing so, I think, because of
the consensus of mankind. If there’s a crowd in a room and there’s a clock in a
room, they can all see the clock. The fact that they can all see it tends to make
them think that it’s not an hallucination: whereas these religious experiences
do tend to be very private.
Russell’s reply is puzzling, to say the least. Although his argument is not completely
clear as it stands, this is probably the gist of it:
Premise 1. People are justified in believing their perceptions to be of real objects outside
them when other people confirm them.
Premise 2. Religious experience is private, so that other people cannot confirm them.
Conclusion. Religious experience cannot easily be justified as being of real objects
outside them.

The problem with this argument is, of course, that Russell is thereby ignoring Cople-
ston’s clarification of the concept. For Copleston has said that religious experience,
as he wants it understood, is experience of an object transcending ‘all the normal
objects of experience’. So, the object of religious experience cannot be compared to a
clock or any other perceptually accessible object. Russell does not reject Copleston’s
material starting point (his description of the phenomenon to be explained), and yet
he implicitly brushes it aside. Russell is, of course, free to treat religious experi-
ence as a sensory hallucination, but then he would have to produce some alternative
description of the religious experience.
Another troublesome aspect of Russell’s argument is that it leaves aside all those
perceptions that (a) are veridical yet (b) face the opposition of people who do not see
the same thing. Consider the famous experiment by Solomon Asch (1951), in which
196 6 Analysis of Segment III: Discussion of Copleston’s …

some subjects upheld their own perceptual judgment, even though the experimenter’s
confederates declared an opposite one, and other subjects preferred to conform to
the majoritarian false perceptual judgment. Again, some perceptions are trained, so
that people without training will not confirm them (Fleck 1935, 1947; Feyerabend
1975, Chapters 8–10). I do not insist on this, because perceptions are ultimately
irrelevant to the issue: Copleston has explicitly excluded them from his description
of the phenomenon he calls ‘religious experience’ (71.4), and he will insist on this,
although it will take Russell until Turn 80 to absorb this point.
In any case, Russell’s counterargument is presented as meta-argumentation
directed against all arguments from experience to reality (“that line of argument
from our own mental states to something outside us”, the subject of Russell 1915).
In a sense, one could argue that Russell is proving too much.

Segment III, Turn 73 (Copleston) Opening Stage (Clarification) Argumentation Stage

(Counter-Argumentation)

[73.1] Yes, they do.

Copleston admits that religious experiences are private, not shared with nor commu-
nicable to others. This is just confirming a part of what he had said himself at
71.4.
[73.2] I’m speaking strictly of mystical experience proper, and I certainly don’t
include, by the way, what are called visions. I mean simply the experience, and
I quite admit it’s indefinable, of the transcendent object or of what seems to be
a transcendent object.
Copleston kindly reminds Russell of his description of the phenomenon he calls
‘religious experience’ at 71.4. To make his meaning easier to grasp, he first replaces
‘religious’ with ‘mystical’ (which is why I took the liberty of putting that word in
parentheses in my earlier comments); and then he explicitly excludes visions, i.e.
perceptual-like experiences.
It is interesting that Copleston feels obliged to admit, as if this was some-
thing special, that mystical experience is indefinable, i.e. that it defies verbaliza-
tion. However, it could at least be argued that all experience is ineffable. Certainly,
perceptual experience has at least a component that seems to be beyond verbalization.
It could, of course, be argued that adjectives such as ‘proper’ (here and also 75.2)
are tricky words in philosophy. It is a bad habit, when you define a concept and your
interlocutor produces a counter-example, to reject it out of hand as not being a proper
(true, genuine, adequate, and so on) case of the concept defined. But Russell never
uses that objection.
[73.3] I remember Julian Huxley in some lecture saying that religious experi-
ence, or mystical experience, is as much a real experience as falling in love or
appreciating poetry and art. Well, I believe that, when we appreciate poetry
6 Analysis of Segment III: Discussion of Copleston’s … 197

and art, we appreciate definite poems or a definite work of art. If we fall in love,
well, we fall in love with somebody and not with nobody.
Copleston appeals to a contemporary scientist, the biologist Julian Huxley (1887–
1975), in an argument that seems to go like this:
Premise 1. When we appreciate poetry and art, we appreciate definite poems or a definite
work of art.
Premise 2. If we fall in love, we fall in love with somebody.
Premise 3. Mystical experience is as much a real experience as falling in love or
appreciating poetry and art.
Conclusion. Mystical experience is about something definite and real.

This is far from being a knock-down argument. Its value lies probably in the fact
that appreciating art or poetry is not a case of perception; and neither is falling in
love. Notice that this argument seems to be multiple (or convergent), in that aesthetic
appreciation and falling in love seem to be separate and independent phenomena
which are here compared to mystical experience.
Also notice that, by comparing mystical experience with other, very different
experiences, Copleston has opened a flank for attack. Russell will charge with alacrity.

Segment III, Turn 74 (Russell) Argumentation Stage (Counter-Argumentation)

[74.1] May I interrupt for a moment here.

This is one of the rare interruptions in a very gentlemanly debate. Note how politely
it is announced.
[74.2] That is by no means always the case.
Russell is denying Copleston’s last assertion only: “If we fall in love, we fall in love
with somebody and not with nobody.”
[74.3] Japanese novelists never consider that they have achieved a success unless
large numbers of real people commit suicide for love of the imaginary heroine.
Russell, somewhat facetiously and perhaps apocryphally, presents the following
argument to support his rejection of Premise 2 in the argument at 73.3):
(1) Premise 1. People in Japan are known to fall in love with the heroines of novels.
Premise 2. The heroines of novels are imaginary, not real.
Conclusion. Not all falling is love is falling in love with somebody.

This objection only partially touches on the argument at 73.3, at least if we accept
that it is multiple. Russell is tackling the experience of falling in love but ignoring the
experience of aesthetic appreciation. Besides, it could be argued that ‘falling in love’
with a novel’s heroine has very little to do with falling in love with a real woman,
even if the lover may be imagining his beloved to have attributes that she does not
have.
198 6 Analysis of Segment III: Discussion of Copleston’s …

Russell’s argument could eventually be extended to the main issue as follows:

(2) Premise 1. Falling in love does not imply the existence of the loved being.
Premise 2. Mystical experience is like falling in love.
Conclusion. Mystical experience does not imply the existence of the object of that
experience.

Notice, however, that Russell does not put forward argument (2) but only argument
(1). He is being extremely cautious. Also note that Copleston had not said that
mystical experience is like falling in love, but only that it “as much a real experience
as falling in love”. So, cautiousness is very well advised.

Segment III, Turn 75 (Copleston) Opening Stage (Material Starting Point)

[75.1] Well, I must take your word for these goings on in Japan. I haven’t
committed suicide, I’m glad to say, but I have been strongly influenced in the
taking of two important steps in my life by two biographies.
First, Copleston gives a very British, good-humoured reply: maybe in exotic Japan
this sort of thing has happened; and if he himself had had the experience of killing
himself for such an imaginary love, he could follow Russell’s argument better. In any
case, trying to grasp the point of argument (1) at 74.3, Copleston suggests a different
example which is within his range of experiences. This move seems to assume that,
since discussants are tackling experiences, they’d better use real ones.
In any case, if there is an argument here, it is more suggested than explicitly set
forth. If we tried to reconstruct it, it would probably go like this:
Premise 1. Copleston has been strongly influenced in the taking of two important steps
in his life by two biographies.
Premise 2. The characters of biographies are at least in part imaginary and not quite real.
Conclusion. Copleston has been strongly influenced by people whose attributes may be
doubted, at least in part.

Notice that Copleston has extended the scope from ‘committing suicide because
of falling in love with x’ to ‘being influenced in taking important decisions by x’,
where x may be completely imaginary (like in the alleged Japanese novels) or partly
imaginary (like often in biographies). So, he is not only conceding Russell’s point,
as far as it goes, but making it less exotic.
[75.2] However, I must say I see little resemblance between the real influence
of those books on me and the mystical experience proper, so far, that is, as an
outsider can obtain an idea of that experience.
Secondly, he rejects Russell’s objection to the argument at 73.3 by saying that it
does not apply to what he means by a mystical experience. The adversative particle
‘however’ indicates that this is a rebuttal of the objection (traditionally, this would
be labelled a case of ignoratio elenchi). We can now see that the ‘argument’ at
75.1 was actually intended as a concession designed to introduce the real point: that
6 Analysis of Segment III: Discussion of Copleston’s … 199

Russell is not talking about religious experience as described by Copleston at the

beginning. This intervention is then a return to the Opening Stage in order to restate
that description as a material starting point for the discussion.
Note in any case that Copleston is conceding that he does not know the mystical
experience from the inside, as it were, i.e. that he has never had one himself. Note
also that Russell, who had in various writings alluded to mystical experiences as
something not altogether alien to him, does not bring them to bear on the discussion.
This is a curious situation, to say the least.

Segment III, Turn 76 (Russell) Opening Stage (Material and Procedural Starting
Point)

[76.1] Well, I mean we wouldn’t regard God as being on the same level as the
characters in a work of fiction. You’ll admit there’s a distinction here?
Russell admits the flippancy of his last intervention, involuntarily brought about, it
is true, by Copleston’s appeal to Julian Huxley. Russell himself recognizes then that
God is not at all like the character of a novel. And so, he suggests being serious about
the issue at hand and agree on the said distinction.
In a sense, this intervention is a call to order, suggesting, as a procedural starting
point, namely, to avoid comparing God to a fictional character.

Segment III, Turn 77 (Copleston) Opening Stage (Material and Procedural Starting
Point). Argumentation Stage

[77.1] I certainly should.

Copleston agrees on the distinction, and thus both to the material and the procedural
starting point put forward by Russell.
[77.2] But what I’d say is that the best explanation seems to be the not purely
subjectivist explanation.
Copleston re-iterates that his argument (see comment on 71.5, soon to be amended
as set forth in my comment on 77.4) is an inference to the best explanation: mystical
experiences cannot be explained as merely subjective.
[77.3] Of course, a subjectivist explanation is possible in the case of certain
people in whom there is little relation between the experience and life, in the
case of deluded people and hallucinated people, and so on.
Copleston now introduces a crucial distinction: in deluded and hallucinated people,
their experiences (hearing voices, having visions, and the like) have no relation to
their lives. So, those experiences can easily be explained by other causes than the
existence of their objects.
200 6 Analysis of Segment III: Discussion of Copleston’s …

[77.4] But when you get what one might call the pure type, say St. Francis of
Assisi, when you get an experience that results in an overflow of dynamic and
creative love, the best explanation of that, it seems to me, is the actual existence
of an objective cause of the experience.
As opposed to deluded and hallucinated people, mystics have experiences that trans-
form them into very extraordinary persons, full of a special kind of love (characterized
by Copleston as ‘dynamic’ and ‘creative’). It is that sort (which Copleston calls the
‘pure type’ of a religious experience) that can best be explained by postulating that
there is an objective cause. He does not say it explicitly, but he implies that that cause
is God, so that the transition from premise to conclusion in argument (1) of 71.2 is
warranted. The argument is now something like this:
Premise 1. Mystical experiences are best exemplified by the experience of Francis of
Assisi.
Premise 2. The experience of Francis of Assisi crucially includes an extraordinary
transformation, characterised by an overflow of dynamic and creative love.
Conclusion 1. Mystical experiences produce extraordinary transformations, charac-
terised by an overflow of dynamic and creative love.
Premise 3. The best explanation for extraordinary transformations characterised by an
overflow of dynamic and creative love is that God exists.
Conclusion 2. God exists.

Copleston unfortunately rests content with suggesting this argument without quite
making it explicit. He also refrains from describing more clearly what he calls ‘an
overflow of dynamic and creative love’ (although the reference to the biography of
St. Francis of Assisi is intended to be sufficient for Russell, and perhaps for the
audience of the BBC Third Programme, to understand what he means). This lack of
explicitness is to a great extent responsible for Russell’s frequent lapses into ignoratio
elenchi.
It may be worth remarking that Copleston’s argument, as now reconstructed,
closely corresponds to Peirce’s definition of abduction as a piece of reasoning that
starts with the observation of “a surprising fact”—in this case, the extraordinary
transformation of Francis of Assisi—and then hypothesizes that, if God existed, then
that surprising fact, a transformation like that, “would be a matter of course”. The
conclusion is that “there is reason to suspect” that the hypothesis, God’s existence,
is true.1
Finally, note that, in this argument, Copleston does not assert a causal connec-
tion between the experience of the mystic and the transformation of his or her
life, what I shall call ‘the mystical transformation’. This question of causality
(say, whether the mystical experience causally effects the mystical
transformation, call this Q54) will soon return with a vengeance (see 83.1).

1 According to Peirce (1903, Lecture VII, §2), “the form of inference… is this: The surprising fact,
C, is observed. But if A were true, C would be a matter of course. Hence, there is reason to suspect
that A is true.”.
6 Analysis of Segment III: Discussion of Copleston’s … 201

Segment III, Turn 78 (Russell) Opening Stage (Procedural Starting Point),

Argumentation Stage (Counterargumentation)

[78.1] Well, I’m not contending in a dogmatic way that there is not a God. What
I’m contending is that we don’t know that there is.
Russell is ostensibly cautious. He is not definitely saying No to Q53 (whether the
best explanation of religious/mystical experience is that god exists).
He seems to want to distinguish between knowledge of God’s existence (which we
don’t have) and something else. He is again playing the card of the agnostic, i.e. of the
Gamma stance, even though he had already shifted to the less hesitant Delta stance
at 68.1, by accepting that it is illegitimate even to ask the question of the cause of
the world. However, if we consider that Copleston explicitly said that the argument
from religious experience is not a proof but only an inference to the best explanation
(or even, as I suggest, an abduction), then it is not quite clear that Copleston, on the
basis of that argument, is saying that We can know that God exists, but perhaps only,
in Peirce’s expression, that We have reason to believe that God exists (see comment
on 77.4).
[78.2] I can only take what is recorded as I should take other records, and I do
find that a very great many things are reported—and I am sure you would not
accept things about demons and devils and what not—and they’re reported in
exactly the same tone of voice and with exactly the same conviction.
Russell says that he will treat all records equally. This could then be interpreted as a
second procedural starting point on his part: let’s agree to treat all records of a certain
group of experiences in the same way. It is much more consequential than his first
procedural starting point at 76.1.
Now, what this group of experiences should be characterised as, Russell does
not say. We may perhaps talk of strange, eerie, uncanny, supernatural experiences.
However, that is exactly the problem. Copleston has made clear that the records of
mystical experiences are very different from all others, let alone from those Russell
mentions. Anyway, if what Russell says here should announce a possible argument,
it would seem to go like this:
Premise 1. Experiences of demons and devils are reported in exactly the same tone of
voice and with exactly the same conviction as mystical experiences.
Premise 2. Experiences that are reported in exactly the same tone of voice and with
exactly the same conviction should be given the same credence.
Conclusion. Mystical experiences should be given the same credence as experiences of
demons and devils.

Russell goes even further, for he suggests that Copleston himself “would not
accept things about demons and devils and what not”. So, the implicit corollary of
the above argument would be that Copleston should not accept the records of mystical
experiences, either. This comes pretty close to saying No to Q53, but it does not quite
get there.
202 6 Analysis of Segment III: Discussion of Copleston’s …

Before going on, let me emphasize again that, although Russell claims to treat all
records on the same basis, he carefully avoids including the one record that Copleston
claims as ‘the pure type’ of religious experience, namely, the record of the experience
and life of St. Francis of Assisi.
[78.3] And the mystic, if his vision is veridical, may be said to know that there
are devils. But I don’t know that there are.
Emboldened by his own suggestion of devils, Russell makes a more daring step. This
could be represented as the following argument:
Premise 1. The mystical in his vision surely also sees devils.
Premise 2. It is very much to be doubted that devils exist.
Conclusion. The mystical vision is largely unbelievable.

Again, Russell does not quite state such a conclusion. It is rather an innuendo.
Underlying this suggestion lies probably Russell’s impish pleasure in reminding
Copleston that in the Christian faith, and in Roman Catholicism especially, devils
are deemed to exist. So, this part of Russell’s intervention is not only an innuendo but
a provocation to boot. However, Russell makes a crucial mistake: he speaks of the
mystical experience as consisting of visions, something that Copleston had explicitly
excluded from his description (see 71.4, completed at 73.2).

Segment III, Turn 79 (Copleston) Opening Stage (Material Starting Point).

Argumentation (Counterargumentation)

[79.1] But surely in the case of the devils there have been people speaking mainly
of visions, appearances, angels or demons and so on. I should rule out the visual
appearances, because I think they can be explained apart from the existence of
the object which is supposed to be seen.
Copleston re-iterates his original material starting point: a description according to
which visions are not part of mystical experiences properly so called (73.2). He even
gives a reason for the starting point:
Premise 1. Visions can be explained apart from the existence of the object which is
supposed to be seen.
Premise 2. Visions are not part of mystical experiences.
Conclusion. Causal explanations of visions should not be applied to mystical experiences.

The conclusion of this argument is not explicit in Copleston but can be taken as
intended by him.
6 Analysis of Segment III: Discussion of Copleston’s … 203

Segment III, Turn 80 (Russell) Opening Stage (Material and Procedural Starting
Points). Argumentation Stage (Meta-Argumentation)

[80.1] But don’t you think there are abundant recorded cases of people who
believe that they’ve heard Satan speaking to them in their hearts, in just the
same way as the mystics assert God?
Russell starts with a material starting point of his own, as an alternative to Cople-
ston’s. It is formulated as a question: whether the records of people hearing
satanic voices are similar to mystics asserting god. We may call this
Q55. However, because of the way the discussion will continue, it is better to separate
this question from one whose affirmative answer is presupposed: whether there
are records of people hearing satanic voices. Let’s call this Q56.
Note that Q55 does not only try to obtain a purely material starting point, but it
is also a specification of the procedural starting point put forward at 78.2: Russell
is trying to compare, on an equal footing, the reports of hearing voices and the
mystical reports. He does that without explicitly rejecting Copleston’s starting point
that mystical experiences are not sensory in nature.
Now, the verb ‘to assert’ as describing what happens in the experience of a mystic
is a bit surprising (“in just the same way as the mystics assert God”). Russell is
clearly comparing the ‘hearing of voices coming from Satan’ to ‘asserting God’; but
‘hearing’ is not, strictly speaking, on the same level as ‘asserting’. It is almost as
though Russell was avoiding the use of a particular psychological predicate compa-
rable to hearing voices. He does not want to talk of ‘seeing God’, because it has
finally dawned upon him that Copleston has explicitly excluded visions from his
description of mystical experience (73.2). This might explain the curious choice of
the verb ‘to assert’.
[80.2] And I’m not talking now of an external vision, I’m talking of a purely
mental experience.
Russell is here trying to comply with Copleston’s stricture, by saying explicitly that
he is not considering the mystical experience as a vision, more generally a perceptual-
like experience (see comment on 73.2). However, Russell’s concession sounds a bit
forced and perfunctory.
[80.3] That seems to be an experience of the same sort as mystics’ experience of
God, and I don’t see that from what mystics tell us you can get any argument
for God which is not equally an argument for Satan.
As the proverb goes, Russell wants to eat his cake and to keep it. He is saying that
hearing voices is perfectly comparable to (“of the same sort as”) whatever happens in
a mystical experience without saying that it is a case of having visions. His argument
might be represented as follows:
(1) Premise 1. Hearing voices of Satan is perfectly explainable in subjective terms.
Premise 2. Hearing voices of Satan is of the same sort as mystics’ experience of God.
Conclusion. Mystics’ experience of God is perfectly explainable in subjective terms.
204 6 Analysis of Segment III: Discussion of Copleston’s …

This argument is directed against Copleston’s argument at 77.4. In fact, Russell goes
further and produces a meta-argument, an argument about arguments:
(2) Premise 1. Copleston argues from mystics’ experience of God to the existence of
God.
Premise 2. Hearing voices of Satan is of the same sort as mystics’ experience of God.
Conclusion. Copleston should argue from reports of people hearing voices from Satan
to the existence of Satan.

Russell’s argumentative strategy, based on the above-mentioned procedural starting

point of treating all reports of a certain kind, seems to be intended as a reductio ad
absurdum. He thinks that no reasonable person in this day and age would accept the
existence of Satan, so, by parity of reasoning, neither should anybody admit that God
exists. The problem, of course, is that the words I have just italicized (‘of a certain
kind’) presuppose what has to be proved, namely, that hearing voices is indeed “of
the same sort as” mystical experiences (Premise 2 in the two arguments above).
In any case, the idea that mystical experiences are comparable to hearing voices is
so important that I propose to call it Russell’s Principle. The question raised by it is
whether we can compare mystical experiences to other phenomena,
such as hearing voices. Let’s call this Q57. Russell simply assumes an affirma-
tive answer to Q57, whereas Copleston will insist in the sequel on differences among
the two phenomena (or records) which incline towards a negative answer.

Segment III, Turn 81 (Copleston) Opening Stage (Material Starting Point, Procedural
Starting Point). Argumentation Stage (Counter-Argumentation,
Meta-Argumentation)

[81.1] I quite agree, of course, that people have imagined or thought they have
heard or seen Satan.
Copleston says Yes to Q56. This is why it was necessary, in my comment on 80.1,
to separate this question from Q55, which, as we shall soon confirm, Copleston will
not say Yes to (see 81.3).
[81.2] And I have no wish in passing to deny the existence of Satan.
In case Russell wanted to embarrass Copleston by bringing up Satan, he was disap-
pointed. Copleston, as Jesuit father, does not and will not deny Satan’s existence,
which has been part of the Roman Catholic catechism for a very long time. Nonethe-
less, it is quite probable that Copleston holds that the existence of Satan is a pure
article of faith and not, like God’s existence, both an article of faith and a proposition
that can be reached by the use of reason and a variety of arguments.
[81.3] But I do not think that people have claimed to have experienced Satan in
the precise way in which mystics claim to have experienced God.
Copleston now says No to Q55 and rejects Premise 2 of the two arguments recon-
structed in my comment to 80.3. To Russell’s expression “of the same sort as”, Cople-
ston now opposes “not in the precise way in which”. By this opposition, Copleston is
6 Analysis of Segment III: Discussion of Copleston’s … 205

effectively putting forward a procedural starting point: let us be precise when we talk
about mystical experiences; let us not compare what is not comparable. This is the
exact opposite to Russell’s second procedural starting point (at 78.2). This struggle
between starting points is a sign that the Opening Stage is running in parallel with
the Argumentation Stage.
[81.4] Take the case of a non-Christian, Plotinus. He admits the experience is
something inexpressible, the object is an object of love, and therefore, not an
object that causes horror and disgust.
To prove that mystical experiences are different from hearing Satanic voices, Cople-
ston goes pretty far back, to the experience of a non-Christian philosopher. The
argument can be represented as follows:
Premise 1. Plotinus had mystical experiences and he said that their object was an object
of love.
Conclusion 1. Mystical experiences do not concern an object that causes horror and
disgust, like hearing Satanic voices.
Conclusion 2. Mystical experiences are radically different from hearing Satanic voices.

Conclusion 2 of this argument clearly counters Premise 2 in the two Russellian

arguments set forth in my comment to 80.3, and is thus intended to refute Russell’s
second procedural starting point at 78.2.
[81.5] And the effect of that experience is, I should say, borne out, or I mean the
validity of the experience is borne out in the records of the life of Plotinus. At any
rate it is more reasonable to suppose that he had that experience, if we’re willing
to accept Porphyry’s account of Plotinus’ general kindness and benevolence.
Copleston goes further. As he says already in his Francis of Assisi example (77.4),
the mystical experience transforms the life of the mystic. Copleston argues, from
Porphyry’s biography of Plotinus, that something similar happened also in the case
of a non-Christian. We can reconstruct the argument as follows:
Premise 1. Plotinus, a mystic, is reported to have been a loving, kind, benevolent person.
Conclusion 1. Mystical experiences transform the experiencers into loving, kind,
benevolent persons.
Premise 2. Hearing Satanic voices are not followed by this kind of transformation (if
anything the contrary is the case).
Conclusion 2. Mystical experiences are radically different from hearing Satanic voices.

Again, this insinuated rather than explicit argument can be seen as a refutation both
of Premise 2 in Russell’s two arguments (see comment on 80.3) and of his second
procedural starting point (see comment on 78.2).
Note that Russell could in principle impugn the causal chain from having a
mystical experience to becoming a loving, kind, benevolent person (from Premise 1
to Conclusion 1). Copleston has certainly not provided an argument for that. Turn
83 will confirm the importance of this observation.
206 6 Analysis of Segment III: Discussion of Copleston’s …

Segment III, Turn 82 (Russell) Opening Stage (Material Starting Point,

Meta-Argumentation)

[82.1] The fact that a belief has a good moral effect upon a man is no evidence
whatsoever in favour of its truth.
Interestingly enough, Russell does not dispute the truth of the mystical transformation
claimed by Copleston in his two examples (Francis of Assisi and Plotinus). However,
he disputes the validity of arguing therefrom (if taken as a fact) that the object of the
mystical experience does exist. Russell seems to be arguing like this:
Premise 1. The fact that a belief has a good moral effect upon a man is no evidence
whatsoever in favour of its truth.
Premise 2. Copleston argues that the mystical experiences of Francis of Assisi and
Plotinus had a good moral effect upon them.
Conclusion. From the extraordinary transformation of the lives of the mystics it does not
follow that God exists.

Russell seems finally to be prepared to engage Copleston’s argument from religious

experience. Notice, however, that some aspects of the latter have been subtly changed:
For the mystical experience, as defined by Copleston at 71.4, Russell substitutes
“belief in God”; and for the mystical transformation of the mystic by his or her
experience, as exemplified by Copleston at 77.4 and 81.4, Russell substitutes “a
good effect upon a man”. Premise 1 functions as a material starting point, a kind of
Toulminian warrant, that is supposed to authorize, in the above meta-argument, the
transition from Premise 2 to the Conclusion. Yet it is highly doubtful that Premise 2
can actually fulfil that function when its content has been weakened and bowdlerized
to such an extent. In any case, I shall refer to it as Russell’s Warrant.
Whether we agree or not with Copleston, he is certainly not talking merely of
‘beliefs’ and ‘good moral effects’ in the normal run of life. Beliefs and their good
moral effects can indeed be frequent, ordinary, and uninteresting; but Copleston is not
talking about those. He is, on the contrary, referring to very rare, out-of-the-ordinary
events, which have caught the attention of many people as striking, remarkable, and
in need of explanation. Somehow, this point does not quite enter into Russell’s head.
However that may be, it is clear that Russell’s Warrant seems to be a kind of
epistemological principle that is intended to put a restraint on Copleston’s religious
argument.

Segment III, Turn 83 (Copleston) Opening Stage (Material Starting Point,

Meta-Argumentation)

[83.1] No, but if it could actually be proved that the belief was actually respon-
sible for a good effect on a man’s life, I should consider it a presumption in
favour of some truth, at any rate, of the positive part of the belief, not of its
entire validity.
Unfortunately for the quality of the debate, Copleston accepts the bland generalities
and abstractions of Russell’s Warrant (the material starting point acting as Premise 1
6 Analysis of Segment III: Discussion of Copleston’s … 207

in the argument at 82.1), whilst he nevertheless introduces an important modification.

This intervention is altogether a very important moment of the discussion, for it is here
that Copleston’s argument from religious (i.e. mystical) experience starts to morph
into the argument from moral experience. Consider, one by one, all the remarkable
aspects of this sentence.
First, by accepting Russell’s Warrant, the debate is leaving the relatively concrete
case of the mystical experience and its relation to God’s existence, which is the
proper issue in dispute, for the much more abstract case of beliefs (in general) having
good effects (in general) on believers (in general). So, mystics, their extraordinary
experiences, and the even more extraordinary transformations of their lives leave the
stage to be replaced by the humdrum and the common. Is this perhaps an indication
that Copleston is not, after all, quite at home discussing mystical experiences?
Secondly, if nonetheless we keep in mind that, behind the generalities imposed
by Russell’s Warrant, there is an overlap with the point of Copleston’s religious
argument, then we can see that Copleston is clearly aware of not having provided
evidence for the causal chain leading from mystical experiences to the mystical
transformation. Notice how striking the repetition of the modifier in the conditional
is (‘if it could actually be proved that the belief was actually responsible…’). The
conditional considerably diminishes the force of Copleston’s argument at 77.4 as well
as the original argument at 71.5 (see comments on those sections for a reconstruction).
Those arguments are now made to depend on a causal chain for which no evidence
so far has been provided.
Thirdly, by talking about a presumption of truth, Copleston is once more making
clear that his argument from religious experience is not a proof in the strict sense
of the term, but something far weaker. By the way, three different causal chains
may get entangled in the debate: (a) God causing the mystical experience, (b) the
mystical experience causing the mystical transformation, and (c) God causing the
mystical transformation. At 71.5, as the argument began, (a) was the crucial point.
At this point, although Copleston, following Russell, is talking in generalities, he
nevertheless implies that evidence for (b) would count as a presumption of truth for
(a), and thus for the belief in God’s existence. This indicates that (c) is also part of the
argument. We should therefore be aware of the differences between the three causal
chains.
Fourthly, Copleston distinguishes between the positive part of a belief that has “a
good effect on a man’s life” and other parts of that belief, which are perhaps less
“positive”. The underlying idea seems to be that beliefs, independently of their truth
value, can also be classified by what we may call their moral value, i.e. the effect, for
good or bad, on the believer. If you, for instance, believe that people should keep their
promises, that may have a good effect on your life, and in fact on the life of many
other people. The same presumably applies to beliefs in the existence of Zeus, Allah,
or the spirits of the ancestors that dwell in the neighbouring woods. Again, all those
beliefs may occasionally have a bad effect, for instance, if a hitman keeps his promise
or if a suicide bomber acts on faith. Copleston suggests that part of such beliefs is
‘positive’ (has a good effect) and another part is either ‘neutral’ or ‘negative’. This is
also crucial for the religious argument, in that the mystical transformation is deemed
208 6 Analysis of Segment III: Discussion of Copleston’s …

by Copleston to have been not only ‘positive’ but hugely so. This is, by the way,
something for which, again, no evidence is provided.
All these observations have two upshots.
One is a quite radical amendment of Russell’s Warrant. According to Copleston,
we should not say: The fact that a belief has a good moral effect upon a man is no
evidence whatsoever in favour of its truth. We should rather say: The fact that parts
of a belief have a good moral effect upon a man is a presumption in favour of the
truth of those parts. For Copleston’s amendment to make sense, we have to assume
two things: (i) that the religious argument is in part already a moral argument; and
(ii) that the truth in this domain cannot be separated from its effects on human life.
Point (i) will preside the turning point around Turns 90–91; point (ii) will come back
to haunt the notable discussion of Turns 103–116.
The second upshot is a curious inversion of reasoning in regard to Russell’s implicit
argument at 81.2. The reconstructed argument, also more implicit than explicit, could
be represented as follows:
Premise 1. The fact that parts of a belief have a good moral effect upon a person is a
presumption in favour of the truth of those parts.
Premise 2. Francis of Assisi and Plotinus, and in general mystics, are extraordinarily
transformed for the better by their mystical experiences of God.
Conclusion. The extraordinary transformation of the lives of the mystics yields a
presumption of truth in favour of God’s existence.

The question raised by Russell’s Warrant is whether a belief’s good moral

effect upon a person constitutes evidence for that belief’s truth
or not. If we call this Q58, then we may call Q59 the one raised by Copleston:
whether the good moral effect exerted on a believer by parts of a
belief yields a presumption of truth of those belief parts, or rather, if
we take Copleston’s original argument seriously, whether the extraordinary
transformation following a mystical experience is a presumption in
favour of the truth of the belief in god’s existence. One of the diffi-
culties in assessing disagreements in philosophical discussions is the fact illustrated
by comparing all these questions. Russell proposed a material starting point in the
shape of what I have called Russell’s Warrant (which is a negative answer to Q58).
Copleston accepted it but with a modification (thus answering Yes to Q59); and
behind that modification there is a vastly different inquiry.
The question of agreement and disagreement between the discussants has become
blurred. Is Copleston’s modification illusory, giving way to what would amount to
a merely verbal dispute? It is possible; but it is also possible that the difference in
formulation hints at a larger and deeper philosophical difference. In the sequel, we
shall indeed see a strange dancing duet around these questions. We shall see Russell
defending the idea that Effects is all there is to moral beliefs and rejecting the idea
that Moral beliefs have truth value; and we shall see Copleston defending the idea
that It is the truth value of moral beliefs which accounts for their effects, so that
those effects are a pointer to the truth value of moral beliefs.
6 Analysis of Segment III: Discussion of Copleston’s … 209

[83.2] But in any case, I am using the character of the life as evidence in favour
of the mystic’s veracity and sanity rather than as a proof of the truth of his
beliefs.
In a surprising move, Copleston becomes explicit as to what his argument actually
is. As opposed to my reconstruction of 83.1, this is what he is now saying:
Premise. The character of life of Francis of Assisi, Plotinus, and other mystics, as it
was recorded after their mystical experiences, is evidence that they were truthful in their
reports and that they were not crazy.
Conclusion. One should not consider the experiences of mystics as comparable to the
hearing of Satanic voices.

After courteously acknowledging Russell’s Warrant, and even accepting it with a

modification, it becomes clear that Copleston somehow wants to go back to his
argument, which is not about beliefs in general but about mystical experiences, and
it is not about good moral effects but about extraordinary transformations of the
mystics’ lives. In accordance with this wish not to get diverted by bland generalities,
he proposes a new rebuttal of Premise 2 of the arguments reconstructed under 80.3.
The question raised by this return to Copleston’s original argument from reli-
gious would be: whether the mystical transformation is evidence of
veracity and sanity. This is Q60; and if the differences between Q58 and Q59
are relatively minor, those between Q58 and Q60 are abysmal.

Segment III, Turn 84 (Russell) Opening Stage (Material Starting Point)

[84.1] But even that I don’t think is any evidence.

First, Russell says even that is not evidence. What does he mean by “even that”?
Given the immediate context put forward in 83.2, it would seem that we should
understand the reference of “that” as the proposition: The character of the life of the
mystics after their experiences is evidence that they were telling the truth and that
they were not crazy. It seems Russell denies this. Let us now attend to his reasons.
[84.2] I’ve had experiences myself that have altered my character profoundly.
All right, but are they mystical experiences? and are the transformation he underwent
a mystical transformation, akin to those of Francis of Assisi or Plotinus? That is rather
the question.
[84.3] And I thought, at the time at any rate, that it was altered for the good.
The answer is: apparently not. Russell’s transformation of character is something
different from that of the mystics. He is not even sure now that the transformation
was altogether for the good. So, Russell seems not to be talking about the same topic
as Copleston.
210 6 Analysis of Segment III: Discussion of Copleston’s …

[84.4] Those experiences were important, but they did not involve the existence
of something outside me,
By saying this, Russell admits that the experiences associated to the transformation
is different from those of the mystics. He is talking of something else and he knows
it.
[84.5] and I don’t think that if I’d thought they did, the fact that they had a
wholesome effect would have been any evidence that I was right.
Russell’s reply is somewhat convoluted, but he seems to be arguing like this:
Premise 1. Russell has had experiences (not mystical) which had a wholesome effect on
his character.
Premise 2. Those experiences did not involve the existence of something outside Russell
(again, not mystical).
Premise 3. If those experiences would have involved the existence of something outside
Russell (i.e. if Russell had believed that they did, if he had had a mystical experience),
the wholesome effect on Russell’s character would not be evidence that his belief that
they involved something outside him was right.
Conclusion. Russell’s experiences are comparable to the experiences of the mystic.
The reader may rightly say that this argument is preposterous. We know it is; but
consider that my reconstruction of Russell’s argument has been done under the
assumption that Russell wants to talk about the same question as Copleston.
It is pretty clear that he does not, first of all, because he insists on talking about
truth (being right about one’s belief that there is something outside him), whereas
Copleston says he is talking not of truth but of truthfulness (the mystic is neither a
liar nor a loony).2 Another reason to think that Russell wants to talk of something
else, is that he insists on “wholesome effects”, which is a very inadequate way of
talking of the lives of the mystics, at least if we take Copleston’s way of talking as a
point of comparison.
Of course, one could dispute whether the lives of the mystics are what
copleston says they are. We may call this Q61. Maybe there is nothing like
what I have called ‘the mystical transformation’, so that Copleston’s arguments at
77.4, 81.5, and 83.2 are without merit. The point, however, is that Russell has not
disputed Copleston’s account about those topics. We can say that Q61 certainly arises
from what Copleston has said, but neither discussant has elaborated on the question.
In sum, Russell is not talking about mystical experience or the mystical transfor-
mation, which is the subject of Copleston’s religious argument but rather about his

2 Copleston (1993, p. 136: “Russell seemed to think that I was arguing from the quality of the lives
and characters of certain mystical writers to the truth of their claims. But that was not what I had in
mind. The point which I wished to make was simply that knowledge of a person’s life and character
may be sufficient to warrant the conclusion that he or she sincerely believed what was said, and
that the person was not making claims which he or she knew or strongly suspected to be bogus, out
of a spirit of self-exaltation and a desire for notoriety. In the text of the broadcast there is indeed
mention of this point, but, if I remember correctly, a fuller explanation of my meaning was one of
the items which I struck out when reducing my material to the amount required by parity of space.”
See Chapter 1, Sect. 1.1.
6 Analysis of Segment III: Discussion of Copleston’s … 211

own experiences and his own transformations, which he declares not to be mystical.
This is a glaring case of ignoratio elenchi. Moreover, he is, in Premise 3, simply
presupposing what he should prove, namely, that his experiences are comparable to
those of the mystics (see Q57 at 80.3). No wonder his conclusion begs the question.

Segment III, Turn 85 (Copleston) Opening Stage (Material Starting Point)

[85.1] No, but I think that the good effect would attest your veracity in describing
your experience.
Copleston insists on his argument at 83.2. If Russell had had a mystical experience
(which he admits he had not), i.e. an experience that includes the awareness of
‘something outside him’, and if his life had been correspondingly transformed, then,
Copleston argues, that transformation of Russell’s life would attest to the veracity of
his belief. Maybe the belief is wrong, and there was not something outside Russell,
as he believed at the time, but at least we could say that he was being truthful.
Although Copleston does not elaborate on this, we can surmise that his argument
must be something like this:
Premise 1. We know (or have good reasons to believe) that someone is a religious
charlatan, a religious liar or a religious loony because they never undergo, in their actual
lives, a radical and extraordinary transformation of the sort we can observe in those we
consider genuine mystics.
Premise 2. The cases of religious charlatans, religious liars, and religious loonies are
very numerous.
Premise 3. The mystical transformation is an exceedingly rare phenomenon.
Conclusion. The mystical transformation is pretty good evidence that the person who
reports a mystical experience is telling the truth about what he experienced.

It would have been very interesting if Russell had raised questions that would
have led Copleston to produce an argument like this; but, again, Russell just seems
not to be interested in what Copleston has to say.
[85.2] Please remember that I’m not saying that a mystic’s meditation or
interpretation of his experience should be immune from discussion or criticism.
Copleston issues a very important caveat, and moreover a caveat which is in very
good standing within the Roman Catholic Church, at least as a modern institution.
Thus, all reports of sainthood are routinely treated to pretty stringent examinations of
evidence; and so are reports of demonic possession. Both topics have even reached
popular culture. Copleston is thus presenting a plea to Russell not to assume that
belief in mystical experience is a bad case of credulity. This plea reinforces Premise
3 in the above argument.
212 6 Analysis of Segment III: Discussion of Copleston’s …

Segment III, Turn 86 (Russell) Opening Stage (Material Starting Point)

Argumentation Stage (Counterargumentation)

[86.1] Obviously the character of a young man may be—and often is—
immensely affected for good by reading about some great man in history,
Russell ignores completely Copleston’s last remark (85.2), which was designed to get
back on track about the subject under discussion, namely mystical experience, and
keeps going as though the discussion was on another subject entirely. In fact, Russell
narrows the topic considerably, probably under the effect of his own reminiscences:
the good effects on the character of boys (cf. 86.3) reading about great men in
history. The way he formulates is as though he was granting a concession to some
material starting point put forward by Copleston, which is most certainly not the case.
Consequently, we have to take it as Russell’s own suggestion of a possible starting
point.
[86.2] and it may happen that the great man is a myth and doesn’t exist,
A further specification of Russell’s starting point: in some cases, the great man that
the boy reads about in a book either did not exist or was not as depicted in that book. (It
comes to the same thing, at least according to Russell’s own theory of descriptions.)
[86.3] but the boy is just as much affected for good as if he did.
A further specification: the effect on the boy is independent of the truth in the account
given by the author of the book.
[86.4] There have been such people. Plutarch’s Lives—take Lycurgus as an
example, who certainly did not exist, but you might be very much influenced
by reading <Plutarch’s life of> Lycurgus under the impression that he had
previously existed.
Russell now supports his full material starting point by, first, asserting that there
have been such fictional great men that have exerted good effects on boys. As an
example, Russell offers Lycurgus as an example of such a great man. Russell is certain
(perhaps too certain) that Lycurgus did not exist, and then suggests the possibility
of somebody, not an actual boy, though, who was influenced, presumably for the
good, as a consequence of believing, wrongly, that the great man did exist. (He
does not say so, but perhaps Russell’s character was improved by reading Plutarch’s
account.) Notice that, where Copleston uses actual historical characters (Francis of
Assisi, Plotinus), Russell tends to be quite happy with fictional ones. One can impugn
Copleston’s account of the mystics, but it is difficult to impugn imaginary characters.
[86.5] You would then be influenced by an object that you’d loved, but it wouldn’t
be an existing object.
Russell finally draws the consequence. We might reconstruct his argument as follows:
Premise 1. Anyone could possibly be influenced for the good by reading Plutarch’s
account of Lycurgus.
6 Analysis of Segment III: Discussion of Copleston’s … 213

Premise 2. Lycurgus did not exist.

Conclusion 1. Anyone could possibly be influenced for the good by reading fictional
accounts of non-existent beings.
Premise 3. The influence for the good would be realized by inspiring love for the non-
existent being.
Conclusion 2. Anyone could be brought to love a non-existent being.
Premise 4. The object of a mystical experience may be a non-existent being.
Conclusion 3. The mystic could be brought to love a non-existent being through his or
her mystical experience.

Of course, Russell does not deploy the whole argument here reconstructed. He goes
as far as Conclusion 2; but, if we assume that he is still talking about the subject of
Copleston’s religious argument, then we can say that Russell is suggesting Conclu-
sion 3. In this case, we should not forget that the argument would be assuming a
variant of Russell’s Principle (see comment on 80.3), according to which we can
compare mystical experiences with youthful readings of inspirational books.
Note, however, that Russell’s Principle is constantly being modified, widened,
and made ever more inane. When he suggested that principle for the first time, he
said literally:
I can only take what is recorded as I should take other records, and I do find that a very great
many things are reported—and I am sure you would not accept things about demons and
devils and what not—and they’re reported in exactly the same tone of voice and with exactly
the same conviction. [78.2, italics added]

This principle faced resistance from Copleston which Russell swept away without
any consideration; but in this form it at least had something to do with the mystical
experiences Copleston is talking about. A bit later on, Russell reinforces the same
idea
But don’t you think there are abundant recorded cases of people who believe that they’ve
heard Satan speaking to them in their hearts, in just the same way as the mystics assert God?
[80.1, italics added]

Still later on, Russell goes much further and unaccountably extends his principle to
his own experiences:
I’ve had experiences myself that have altered my character profoundly. And I thought, at the
time at any rate, that it was altered for the good. Those experiences were important, but they
did not involve the existence of something outside me, and I don’t think that if I’d thought
they did, the fact that they had a wholesome effect would have been any evidence that I was
right. [84.2-84.5]

Finally, in the present intervention, things get completely diluted into boys reading
fictional or half-fictional biographies of great men. Russell’s principle is so flexible
that it seems to say that Mystical experiences are comparable to whatever Russell
thinks they are. Copleston obviously thinks they are not, yet no discussion ensues
about this issue (I speak of an ‘issue’ or question, because the comparison with the
reading, by boys, of real or fictional accounts of the lives of ‘great men’ is a variant
of Q57).
214 6 Analysis of Segment III: Discussion of Copleston’s …

Segment III, Turn 87 (Copleston) Opening Stage (Material Starting Point),

Argumentation Stage (Meta-Argumentation)

[87.1] I agree with you on that, of course—that a man may be influenced by a

character in fiction.
Invariably courteous, Copleston accepts Russell’s material starting point, but he will
try to bring the discussion back on track.
[87.2] Without going into the question of what it is precisely that influences him
(I should say a real value),
Copleston points to the moral argument that will come later. He thereby raises the
question about what sort of thing there is in an experience that has
a good influence on the experiencer. This we may refer to as Q62. For
Copleston, we are dealing in all such cases with something real, a value which is not
a figment of people’s imaginations or a mere convention. This question will become
very important in Segment IV of the debate. However, Copleston very properly says
this is not the place to discuss it.
[87.3] I think that the situation of that man and <the situation of> of the mystic
are different.
Once more, Copleston has to remind Russell what the discussion is supposed to be
about, by pointing that mystical experiences are not “of the same sort as” the situation
of someone reading Plutarch and becoming enthusiastic about Lycurgus.
[87.4] After all, the man who is influenced by Lycurgus hasn’t got the irresistible
impression that he’s experienced in some way the ultimate reality.
Copleston gives a reason to support his last contention. That reason has, of course, to
do with his own description of the mystical experience, placed at the start of Segment
III of the debate (71.4). Being overwhelmed by Plutarch’s Life of Lycurgus, however
wonderful and beneficial an experience, is not a mystical experience. There is here
a meta-argument directed against Russell’s alleged counterexample, which might be
thus reconstructed:
Premise 1. Mystical experiences involve something conceived to be an ultimate reality.
Premise 2. The experience of somebody who, having read Plutarch’s account, is full of
admiration for Lycurgus is not a mystical experience.
Conclusion. Russell cannot draw any consequence from the experience of a reader of
Plutarch to the case of mystical experience.

Again, Russell could protest against Copleston’s description of religious experience

or discuss the justified scope of Russell’s Principles, but there is a sense in which,
by these examples, he is talking past Copleston. Russell may be right in everything
he asserts, but his argumentative moves seem strangely unsupported.
6 Analysis of Segment III: Discussion of Copleston’s … 215

Segment III, Turn 88 (Russell) Opening Stage (Material Starting Point)

[88.1] I don’t think you’ve quite got my point about these historical characters—
these unhistorical characters in history.
Curiously enough, it is Russell who maintains that Copleston does not follow him,
whereas the analysis up to this point indicates that it is exactly the other way around.
I say this, of course, without in any way prejudging who is ultimately right about the
subject under discussion.
[88.2] I’m not assuming what you call an effect on the reason. I’m assuming that
the young man reading about this person, and believing him to be real, loves
him—which is quite easy to happen, and yet he’s loving a phantom.
Now, this is an interesting point that has five features we should distinguish:
(1) It rejects talk of an “an effect on reason”, even though Copleston did not include that
in his description of the mystical experience, where he expressly talks about “loving
awareness” of something transcendent.
(2) It reinforces Russell’s particular emphasis on love, i.e. the kind of love he expresses
at 86.5.
(3) It indicates that Russell is interested in the emotional aspects of the experiences he is
talking about, and in particular the love of the admired person (Lycurgus, Churchill,
or whomever).
(4) It thereby comes perilously close to the phenomenon of love for clearly fictional
characters that was brought up and dismissed earlier on by Russell himself (74.3,
76.1).
(5) This love is emphatically not mystical love, and its object is not the mystical Presence
or ultimate reality Copleston wants to talk about (see comment on 7.14).

So, all in all, it seems that Russell is stuck with the argument reconstructed in my
comment of 86.5, in spite of Copleston’s protests at 87.4. I have the impression that
Russell believes that all things mystical that Copleston puts forward (mystical love,
mystical presence, the mystical experience itself, and the mystical transformation)
are hoaxes and fakes, but somehow he does not bring himself to say that, and instead
he constantly replaces those by red herrings that do not help the discussion.

Segment III, Turn 89 (Copleston) Opening Stage (Material Starting Point)

[89.1] In one sense he’s loving a phantom; that’s perfectly true: in the sense, I
mean, that he’s loving X or Y who doesn’t exist.
Again, Copleston accepts the point that, in those cases mentioned by Russell, there
is a kind of love for something that does not exist. If this is the material starting point
that Russell is interested in, then Copleston admits it. The question is whether it is
relevant to the religious argument.
216 6 Analysis of Segment III: Discussion of Copleston’s …

[89.2] But at the same time, it is not, I think, the phantom as such that the young
man loves; he perceives a real value, an idea, which he recognizes as objectively
valid, and that’s what excites his love.
For some reason, Copleston decides to get deeper into the question of what is the
object of love that, according to Russell’s material starting point, may sometimes be
imaginary (‘a phantom’). Copleston says that it is something really existing, a real
value, an objectively valid idea.
So, Copleston proposes here a different material starting point: whenever people
have a reading experience of the sort put forward by Russell, that awakens love in
the readers and has a good effect on their lives and characters, then the object of love,
and the cause of the change, is not the concrete person described in the book (real
or imaginary) but something far more abstract, a value, an idea, and that something
is real not imaginary. It now becomes clearer what Copleston meant when he spoke
(at 83.1) of “the positive part of a belief”. It is that part which expresses the real
value and the objectively valid idea. So, Lycurgus may not have existed as the person
described by Plutarch, but his goodness does exist, and it is that goodness which, so
to speak, fires the heart and soul of the reader.
Consider the following variant of Russell’s argument reconstructed in my
comment on 86.5:
Premise 1. Anyone could possibly be influenced for the good by reading Plutarch’s
account of Lycurgus.
Premise 2. It is the goodness of Lycurgus, as portrayed by Plutarch, that exert the good
influence on the reader.
Conclusion. No one could possibly be influenced for the good by reading about qualities
that are not real (even if the person portrayed as having them does not exist).

This might seem to undercut the rest of Russell’s argument at 86.5, which led to the
unexpressed conclusion that the mystical experience might not be about an existent
being. The question arising from this would be what the relation is between
the existence of god and the existence of his goodness as given
in the mystical experience. Call this Q63. This is in fact an old chestnut of
traditional Christian theology; and it is easy to see that it is quite close to Q62 (see
comment on 87.2).

Segment III, Turn 90 (Russell) Opening Stage (Material Starting Point)

[90.1] Well, in the same sense we had before about the characters in fiction.
Russell is quite happy with the last part of Copleston’s intervention. This seems to
equate fiction and legendary biographies in their effects on some people; and so, by a
kind of slippery slope, it would bring both close to the mystical experience. In a sense,
Russell’s Principle, extended all the way to fiction, would have been vindicated.
6 Analysis of Segment III: Discussion of Copleston’s … 217

Segment III, Turn 91 (Copleston) Opening Stage (Material Starting Point)

[91.2] Yes, in one sense the man’s loving a phantom—perfectly true. But, in
another sense, he’s loving what he perceives to be a value.
Copleston agrees with Russell, but he insists on his distinction between the values
depicted in a story and the fictional bearers of those values. The bearers do not exist,
but the values do. The argument suggested in 89.2 is reinforced.
There is no solution of continuity between the religious and the moral argument;
but this intervention is a kind of bridge between the two arguments. The topic of
mystical experience, never really engaged with by Russell, is not mentioned again
in the debate.

References

Asch, S. E. (1951). Effects of group pressure on the modification and distortion of judgments. In
H. Guetzkow (Ed.), Groups, leadership and men (pp. 177-190). Pittsburgh: Carnegie Press.
Copleston, F. C. (1993). Memoirs of a philosopher. Kansas City: Sheed & Ward.
Feyerabend, P. K. (1975). Against method. London: Verso.
Fleck, L. (1935). O obserwacji naukowej i postrzeganiu w ogóle. Przegl˛ad Filozoficzny [Philosoph-
ical Review] 38, 57–76. [English translation: Scientific observation and perception in general. In:
R. S. Cohen and T. Schnelle (Eds.), Cognition and fact: Materials on Ludwik Fleck (pp. 59–78).
Dordrecht: Reidel, 1986.]
Fleck, L. (1947). Patrzeć, widzieć, wiedzieć. Problemy 1947 (2), 74–84. [English translation: To
look, to see, to know. In: R. S. Cohen and T. Schnelle (Eds.), Cognition and fact: Materials on
Ludwik Fleck (pp. 129–152). Dordrecht: Reidel, 1986.]
Peirce, C. S. (1903). Harvard lectures on pragmatism. [Unpublished in Peirce’s lifetime. There are
many reprints. Available at https://www.textlog.de/7664-2.html.]
Russell, B. (1915). Our knowledge of the external world—as a field for scientific method in
philosophy. Chicago: Open Court.
Chapter 7
Analysis of Segment IV: Discussion
of Copleston’s Moral Argument

Abstract Segment IV of the debate is relatively long (Turns 92 to 136) but by far
the easiest to follow for a lay public. It handles the argument from moral experi-
ence to God’s existence. Very soon a big gap opens between fundamentally different
conceptions of human moral life: Copleston defends a variety of the natural law tradi-
tion with some Kantian overtones, whereas Russell argues from a science-inspired
mixture of emotivism, utilitarianism, and behaviourism. Although noticeable points
of contact appear now and then, none of the debaters take advantage of them and
so the debate on the moral argument is as inconclusive as that on the metaphysical
argument.

Segment IV, Turn 92 (Russell). Opening Stage (Material Starting Point)

[92.1] But aren’t you now saying in effect: I mean by God whatever is good or
the sum total of what is good—the system of what is good, and, therefore, when
a young man loves anything that is good, he is loving God?
With one leap, Russell suggests that, behind all that talk about real values, there
lurks in Copleston an argument for God that is based on the love of the good. The
argument he suspects Copleston of having can be represented as follows:
Premise. God is the sum total or the system of what is good.
Conclusion. Everybody who has come to love something good (by reading or in some other
way) is loving God.

The material starting point for this argument takes the shape of a question: whether
god is the sum total or the system of what is good. Let’s call this Q64.
Notice that if the question had been whether god is the good, or the origin
and ground of everything that is good (call it Q65), then Copleston would
probably say Yes. In traditional Christian theology (incidentally, also in Neo-Platonic
philosophy), God is indeed The Good, for at this highest of theological levels, Being
= Goodness = Truth. But the way Russell phrases the question (as Q64 rather
than Q65), saying Yes would amount to a heresy (or in Neo-Platonism as a crude
misunderstanding).
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 219
F. Leal and H. Marraud, How Philosophers Argue, Argumentation Library 41,
https://doi.org/10.1007/978-3-030-85368-6_7
220 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

We are now entering the moral argument: (2) in 71.2. So, Copleston is playing
the role of Protagonist and Russell that of Antagonist. The Confrontation Stage is
presupposed, and Russell is trying to confirm whether Copleston may have offered
a major material starting point, or rather two. One possible material starting point is
expressed in set-theoretical terms: God as the set of all things good, or the Premise
in the argument above. The other possible material starting point is embedded in
the premise: to love anything good is to love God. Note that the question of God’s
existence (Q1) is tied up with the intentional act of loving Him—a point that was
introduced by Copleston in 71.4, and kept recurring in different form throughout
Segment III.
[92.2] Is that what you’re saying? Because, if so, it wants a bit of arguing.
There are three moves here.
Russell asks, first, whether Copleston agrees with the premise of the above argu-
ment (i.e. with the affirmative answer to Q64), i.e. with the first candidate for a
material starting point on Copleston’s part.
Russell asks, secondly and somewhat more insidiously, whether Copleston agrees
with its conclusion. In the latter point lies a possible dialectical trap, given that any
regular Christian theologian would assert that God is the source of all good. To
that extent, the proposition put forward as conclusion in the reconstructed argument
above is a much better candidate for a material starting point, although Copleston
would never use it with an agnostic or atheist like Russell.
Russell adds, thirdly and finally, that Copleston should produce a reason, although
it is not completely clear whether for the premise, for the conclusion, or for both.
This is a direct challenge: whatever your material starting point is, it calls for an
argument.

Segment IV, Turn 93 (Copleston) Opening Stage (Material and Procedural Starting
Points)

[93.1] I don’t say, of course, that God is the sum-total or system of what is good
in the pantheistic sense. I’m not a pantheist.
If we remember that God, for Copleston, is the extra-mundane cause of the world, and
if we accept that the good things Russell has been talking about (say, great courage
or great wisdom) are in the world, and if we remember that, always according to
Copleston, the aggregate of things in the world is not something different from those
things (Copleston’s Equation), then he cannot agree that God is the sum total of
all good things. That would make God immanent to the world. This idea of God’s
immanence is the chief objection, in Christian theology, against pantheism. As a
Christian thinker, Copleston has to reject that. So, his answer to Q64 is No. Russell’s
first candidate for a material starting point is rejected by Copleston—and he gives a
reason: he is no pantheist.
7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument 221

[93.2] But I do think that all goodness reflects God in some way and proceeds
from Him, so that in a sense the man who loves what is truly good, loves God
even if he doesn’t advert to God.
As I suggested in my comment on 92.2, it is orthodox for a Christian thinker to think
of God as the Good, or as the origin and ground of everything that is good, both in the
world and outside of it. So, a variant of the argument at 92.1 with which Copleston
would agree is as follows:
Premise. God is the Good (or: God is the origin and ground of everything that is good).
Conclusion. Everybody who has come to love something good (by reading or in some other
way), is loving God.

The premise of this argument is an affirmative answer to Q65. Of course, this premise
would also need an argument. We’ll see that the Conclusion will provoke rejection
in Russell at 96.3. In any case, can we say that the premise is really Copleston’s
material starting point for this discussion?
[93.3] But still, I agree that the validity of such an interpretation of a man’s
conduct depends on the recognition of God’s existence, obviously.
In order to be able to provide an argument in favour of the premise in the above argu-
ment, one should first admit that God exists. By saying that, Copleston is assuming
that, before you can prove of an object that it is so and so, you have to prove that the
object exists. We shall refer to this as Copleston’s Principle—a logical and method-
ological one, and so a procedural starting point. So, we see that the premise in the
argument above is not put forward as a material starting point by Copleston.

Segment IV, Turn 94 (Russell) Opening Stage (Procedural Starting Point)

[94.1] Yes, but that’s a point to be proved.

Russell agrees with Copleston and concludes that, in order to discuss Q65, it has first
to be proved that God exists, which is what the debate is all about (Q1). Copleston’s
procedural starting point is accepted.

Segment IV, Turn 95 (Copleston) Opening Stage (Procedural Starting Point)

[95.1] Quite so, but I regard the metaphysical argument as probative, but there
we differ.
Copleston acquiesces but says that the metaphysical argument is intended to be a
proof of God’s existence, and he is prevented from building on that proof, in order
to prove that God is the Good (or the origin and ground of all good), by the fact that
Russell does not admit that proof, by the fact that, more specifically, he maintains
that the inquiry about the cause of the world, and even the very phrase ‘the cause of
the world’, lack all meaning.
222 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

By the way, when I refer to a proof , it goes without saying that I am only repro-
ducing Copleston’s wording and thinking. After Kant, it is very hard to accept the
traditional arguments for God’s existence as proofs. Nonetheless, Copleston is a
Jesuit and a scholastic, so he thinks of them in that way.

Segment IV, Turn 96 (Russell) Opening Stage (Role Assignment), Confrontation Stage

[96.1] You see, I feel that some things are good and that other things are bad.
For some reason, Russell abandons the role of Antagonist, which he had, more or
less consistently, adopted up to this point in the debate, and instead he puts forward
a standpoint of his own. Notice how he starts with the words “I feel”, which should
not be interpreted according to contemporary fashion, as a kind of replacement for ‘I
believe’ or ‘I think’. The words “I feel” are not an indicator that a standpoint is being
put forward. It is rather an expression of the way in which Russell conceives of his
relation to good and bad. He is thus not saying: I believe that some things are good
and that other things are bad, which would be a pretty bland proposition (who does
not believe that?). Russell is rather saying: The kind of relation that obtains between
a person and goodness or badness is a relation of feeling, thus not an intellectual
or cognitive relation but an emotional one.
The theory, according to which morality is a matter of feeling, and not of thinking
or knowing, is a very old one in British philosophy. It found its classical expression
in Shaftesbury and in the Scottish Enlightenment (Hutcheson, Smith, Hume). Kant
stood out as the most emphatic denier of the tenets of moral sense, opposing to it
his conception of practical reason. Russell inclined all his life to some version of
moral sense or moral feeling, whereas Copleston, as a Catholic thinker, upholds some
version of natural law and is thus closer to Kant, even if definitely not a Kantian by
any means.
[96.2] I love the things that are good, that I think are good, and I hate the things
that I think are bad.
The interpretation given above of the meaning of Russell’s standpoint is confirmed
by this sentence. Note that, if our relation to the good is a matter of feeling, not of
knowing, then what is good for a person will not be good for another. It will depend
on how each person feels. This is why Russell, when specifying the kind of feeling
that he has towards the good (love) and the bad (hatred), he adds, as an afterthought,
“the things that I think are good”, even though, strictly speaking, this is not a matter
of thinking. A clearer and more precise version of Russell’s standpoint would then
be this: I love certain things and hate other things; those that I love I call ‘good’,
those that I hate I call ‘bad’. In a sense, this is all the meaning that the words ‘good’
and ‘bad’ have for Russell.
The opposition between love and hate as the feelings of good and bad will become
increasingly important as the debate on the moral argument proceeds.
7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument 223

[96.3] I don’t say that these things are good because they participate in the
Divine goodness.
Naturally, when somebody maintains such a standpoint, he cannot accept the incom-
patible proposition: Good things are good because they have their origin and ground
in God. Notice that Russell introduces Platonic or Neo-Platonic jargon and speaks of
‘participation’. Thus, he clearly rejects the conclusion of the argument reconstructed
in my comment on 93.2. The standpoint at 96.1 is now seen to be the reason for a
new standpoint, defined in opposition to Copleston:
Premise. Good or bad is a matter of feeling. (The good is something a person loves, the bad
something a person hates.)
Conclusion. It is not the case that everybody who has come to love something good (by
reading or in some other way), is loving God. (He or she just loves it.)

We are now in the Confrontation Stage, but the point of disagreement has undergone
a dramatic change.
Notice that this disagreement and this confrontation takes place in outright defi-
ance to Copleston’s Principle (enunciated at 93.3 and accepted by Russell at 94.1).
Even though the ostensible question of the debate is Q1, whether god exists or
not, and even though it has been agreed between the discussants that Q65 cannot
be tackled before Q1 is answered in the affirmative, curiously enough it is now Q65,
whether is the origin and ground of all things good, that has become
the target of the debate. Or rather, given that Copleston has now the role of Antago-
nist, given Russell’s willingness to defend an incompatible position, we should say
that the question at issue is now whether the good is a matter of human
feeling, or even whether human feeling is the origin and ground of
all things good. Let’s call this Q66. This will be the topic of discussion from now
on and until the final summing up in Turns 137 and 138. God will re-appear briefly
at Turns 121 and 135 but will never again be at the centre of the debate. The shift in
the subject matter of debate could be interpreted as a confirmation that Russell did
not really want to talk about God’s existence in the first place.

Segment IV, Turn 97 (Copleston) Opening Stage (Role Assignment), Argumentation

Stage (Request for Justification)

[97.1] Yes, but what’s your justification for distinguishing between good and
bad or how do you view the distinction between them?
Copleston does not express disagreement, although we know from 93.2 that there
is a difference of opinion. He seizes the fact that the pragma-dialectical roles have
been reversed and quickly adopts the role of the Antagonist. In a typical move for an
Antagonist, he issues a request at the new protagonist, asking for a justification for
the standpoint expressed in Turn 96. Note that two new questions have arisen: how
people are justified in distinguishing good and bad (Q67) and what
224 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

kind of distinction is that between good and bad (Q68). Both questions
open a sub-discussion.

Segment IV, Turn 98 (Russell) Argumentation Stage (Sub-Discussion)

[98.1] I don’t have any justification any more than I have when I distinguish
between blue and yellow.
Russell says that he has no justification. In order not to appear unreasonable, he
suggests an analogy:
Premise 1. People cannot justify why they distinguish between blue and yellow.
Premise 2. Distinguishing between blue and yellow is like distinguishing between good and
bad.
Conclusion. People cannot justify why they distinguish between good and bad.

[98.2] What is my justification for distinguishing between blue and yellow? I

can see they are different.
Russell, whilst saying that he cannot justify his distinction of good and yellow,
nonetheless gives an explanation: he can see, he is not blind or colour-blind. A
person distinguishes between blue and yellow because she can do so.

Segment IV, Turn 99 (Copleston) Argumentation Stage (Sub-Discussion)

[99.1] Well, that is an excellent justification, I agree. You distinguish blue and
yellow by seeing them, so you distinguish good and bad by what faculty?
Copleston is not particularly interested in the implicit distinction between justifying
and explaining assumed by Russell, so he just expresses his agreement with the fact
that a person distinguishes between blue and yellow because she can see the difference
between those two colours. However, being a skilled dialectician, he immediately, in
the best Socratic manner, seizes his opportunity to put forward a new request directed
at the analogy: if one distinguished between blue and yellow by the faculty of sight,
then the question arises what the faculty is by means of which human
beings can distinguish between good and bad (Q69). This manoeuvre
is traditionally designed to get us beyond the realm of sensation into the realm of
intellect. But this will not work with Russell, who is prepared to bite the bullet.

Segment IV, Turn 100 (Russell) Argumentation Stage (Sub-Discussion)

[100.1] By my feelings.
Russell bites the bullet: he will be an emotivist all the way down. There is, for him,
no higher, cognitive faculty by which we humans know the good from the bad, or the
7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument 225

right from the wrong, but it is all a matter of feeling. He is upholding the standpoint
offered at 96.1 and 96.2.

Segment IV, Turn 101 (Copleston) Argumentation Stage (Sub-Discussion)

[101.1] By your feelings. Well, that’s what I was asking. You think that good and
evil have reference simply to feeling?
Copleston confirms that this is Russell’s standpoint: The faculty by which we distin-
guish between good and evil is feeling—or, to use a more traditional expression,
moral sense. Copleston’s point in repeating it is just to make it a matter of record.
This is also true for the next question, which is also rhetorical in the proper sense
of the word: it is all about making a position perfectly clear and public. It is this
relation of feeling, this ability or faculty which we call ‘feeling’ the only one involved
in distinguishing good from bad. Copleston is probably as confident as Socrates ever
was that such a position can be refuted. Yet, before starting his case, he needs to have
it loud and clear, beyond any shadow of a doubt that Russell indeed maintains that
position, that is, at the end of the day, that he gives an affirmative answer to Q66.

Segment IV, Turn 102 (Russell) Argumentation Stage (Sub-Discussion) Opening

Stage (Procedural Starting Point)

[102.1] Well, why does one type of object look yellow and another look blue?
Russell pays back with a rhetorical question of his own, designed to have the same
effect. He insists on the analogy at 98.1 and prepares to use it in order to make it
clear that his position is not outrageous. Note, however, that in this reply a subtle
rhetorical change has taken place. Copleston’s original standpoint is active: humans
are the active subject and colours are the passive objects; it is humans who do the
distinguishing upon the colours, as it were. Russell’s new formulation is passive:
colours are now the active subjects that do something to humans: the fact that coloured
objects look blue or yellow is conceived in terms of a certain effect those colours
have upon our eyes. This apparently harmless reformulation allows putting science
back on the table, a favourite move of Russell’s.
[102.2] I can more or less give an answer to that, thanks to the physicists,
Russell suggests that, if Copleston would insist on a more elaborate answer (which
he will not), namely, more elaborate than saying that we distinguish between colours
by sight, then a far more elaborate scientific answer could be provided. He does not
mean a neuroscientific account, of course, as we would think in our day and age,
because no such thing was readily available in 1948; but he means physics, concretely
optics. In order to be able to put optics on the table, Russell had to invert the relation,
as I explained in my last comment. It is light, falling upon a coloured object and
being partly absorbed and partly reflected towards our eyes, that makes us see blue
or yellow (so that, colloquially, it looks yellow or blue). Note that this subtle change
226 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

of direction is exactly what Copleston is not asking: he is interested in the internal

capacity of humans to recognize colours and to know and tell them from one another.
Once more, Russell is evading the issue or, more positively, trying to impose on his
interlocutor a certain outlook and a certain procedure.
[102.3] and as to why I think one sort of thing good and another evil, probably
there is an answer of the same sort, but it hasn’t been gone into in the same way
and I couldn’t give it you.
Now, Russell is suggesting a possible ‘scientific’ answer in which we are not really
the active subjects of distinguishing good or bad but rather the passive objects of
something happening outside of us and impacting our feelings. The same way things
look blue or yellow to our eyes, so do things feel good or bad, right or wrong, to
us; the light makes us see those colours, and quite probably something akin to light
makes us feel the good and the bad, the right and the wrong. The argument would
go like this:
Premise 1. Certain optical phenomena make us distinguish blue from yellow.
Premise 2. Distinguishing between blue and yellow is like distinguishing between good and
bad.
Conclusion 1. Some thus far unknown physical phenomenon makes us distinguish between
good and bad.
Conclusion 2. Nobody can expect a proper (scientific) answer to the question how we
distinguish between good and bad.

For contemporary readers, oxytocin is a fashionable candidate (see e.g. Churchland

2011, 2019),1 and good old genes are an unfailing staple. The only problem is that
this is not what Copleston is looking for. This sub-discussion is again a case in which
we can see the abyss opposing the traditional approach of Copleston in terms of the
distinction between sense and intellect (sensibilia like colours vs. intelligibilia like
good and bad, right and wrong) to the scientific, or scientistic, outlook of Russell—
what is known nowadays as ‘naturalism’.
Turn 102, taken together, could be therefore interpreted as an attempt to fix a
procedural starting point: if we are to discuss the explanation of our sensations
and feelings, we’d better appeal to scientific evidence and theories. The point is
so important for the subsequent discussion that I propose to speak of the Russell
Analogy and to work out what it implies:
First of all, in the following I am going to use the verb ‘to strike’ as an appropriate
acknowledgment of the optical Analogy: as the light reflecting back from a coloured
body strikes the eye of the observer in such a way that the observer sees the colour
yellow, so would a certain person, event, or action strike a subject capable of feeling

1 Churchland argues that moral values are rooted in a behavior common to all mammals: the caring
for offspring. A key part of the story is oxytocin, a body-and-brain molecule that, by decreasing the
stress response, allows humans to develop the trust in one another necessary for the development
of close-knit ties, social institutions, and morality.
7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument 227

Table 7.1 Three versus five terms in the Russell Analogy*

Phenomena Term 1 Term 2 Term 3 Term 4 Term 5
compared
Yellow The coloured The light that is The medium The The observer’s
versus blue body partly absorbed traversed by perceiving organ of sight,
and partly the light subject the eye, upon
reflected from which the
the body with a reflected light
certain falls
wavelength and
frequency
Good versus The person, The feeling The subject’s
bad event, or subject ability or
action that faculty of
inspires love feeling*
in the subject
* Taken from Russell’s reply at 100.1

(moral sense) in such a way that the subject feels love (or hatred) for the person,
event, or action, i.e. she would find that person, event, or action to be good (or bad).
Secondly, if we look hard at the Russell Analogy, then we can see that it is
incomplete. In the case of colours, we have at a minimum five terms, whereas Russell
gives us only three for the case of good and bad. See Table 7.1.
The reader can see that we do not know what the missing terms are for the moral
realm, and Copleston does not ask, perhaps because Russell does not mention them
explicitly in his formulation of the analogy, even though this is crucial for the analogy
to work if optics is indeed the relevant physical science.
I may seem to belabour the point to an unacceptable extreme, but later on we
shall come to the question of mistakes in moral feeling, on being wrong about good
and bad, so it is certainly relevant whether the trouble comes from the organ itself or
rather from some other term in the analogy, e.g. the medium that distorts the light or
something else (see comment on 105.2).

Segment IV, Turn 103 (Copleston) Opening Stage (Procedural and Material Starting
Points) Argumentation Stage (Counter-Argumentation)

[103.1] Well, let’s take the behaviour of the Commandant of Belsen.

Copleston resists Russell’s attempt at bringing the discussion to ‘physics’ and
proposes a different procedure for answering Q69. Very skilfully, he proposes to
consider a recent historical case, fresh in the memories of everyone listening to the
broadcast: the case of Josef Kramer, first ‘commandant’ of the Auschwitz-Birkenau
concentration camp and afterwards of the Bergen-Belsen camp, where his vicious
cruelty won for him, before the British public, the label ‘Beast of Belsen’ (Shep-
hard 2005). His trial in Lüneburg, lasting from September to November 1945, fully
228 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

revealed the horrors of his behaviour both in Auschwitz and in Belsen, and he was
duly hanged in December of that year, a little bit over two years before the BBC
broadcast (Phillips 1949).
[103.2] That appears to you as undesirable and evil and to me too.
Carefully weaving his case, Copleston first ascertains that Kramer’s behaviour is
something that appears “undesirable and evil” both to Copleston and to Russell (i.e.
that strikes them so). This is a kind of material starting point, whose purpose is to
make sure there is no difference in moral judgment (and, if Russell is right, in moral
feeling) between the two interlocutors.
[103.3] To Adolf Hitler we suppose it appeared as something good and desirable.
The second step in Copleston argumentation is to bring the person up who, more than
anyone else, can be considered the architect of what happened at Belsen. Copleston
proposes to consider what the moral judgment and feeling of Hitler would be if
he was asked about Kramer’s behaviour. Naturally enough, Copleston suggests that
Hitler would be opposite in judgment and feeling to both Russell and Copleston (i.e.
it would have struck him in the opposite way). This is a second material starting
point.
[103.4] I suppose you’d have to admit that for Hitler it was good and for you it
is evil.
The third, and decisive step, is the rhetorical question designed to make Russell
admit, loud and clear, that for Hitler the actions committed by Kramer were good,
not evil as for Russell and Copleston. The question thus raised is whether the
behaviour of the commandant of belsen is good for hitler and
evil for russell. Let’s call this Q70. Copleston is transparently interested in
drawing the relativistic implication contained in Russell’s standpoint that good and
bad, right and wrong, “have reference simply to feeling” (99.1), which amounts to
an affirmative answer to Q66. We can formulate this as a further question: whether
the goodness or badness of the behaviour of the commandant of
belsen depends only on the feelings of people, given that different
people have different feelings about it. This would be Q71. Notice that
Q71 is a more specific question than Q66 in two respects: (a) it is not universal,
concerning all things good but only the behaviour of a particular man, Josef Kramer;
(b) moral relativism belongs to the background of the question at issue.
Some readers might believe that I am overinterpreting Copleston. It is true that his
request only expresses Q70 and not Q71. Yet we shall see that Russell understood
perfectly what Copleston meant.
7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument 229

Segment IV, Turn 104 (Russell) Argumentation Stage

[104.1] No, I shouldn’t quite go so far as that.

It is not clear what exactly Russell would not quite go so far as to assert. In one
sense, he has to admit that Hitler feels differently than he himself does, both about
the behaviour of Kramer and about practically everything else. But if one adds the
idea that feelings is all there is to good or evil, and feelings vary as widely as between
Russell and Hitler, then one should conclude that good or evil are relative things.
That conclusion is going too far for Russell.
If this interpretation of Russell’s reply is correct, then my interpretation of 103.4,
namely that, by asking Q70, Copleston was actually asking Q71, is thereby confirmed.
[104.2] I mean, I think people can make mistakes in that, as they can in other
things. If you have jaundice, you see things yellow that are not yellow. You’re
making a mistake.
Russell gives now a reason why the conclusion suggested by Copleston goes too far:
in matters of good and evil, people make mistakes. The argument is again based on
the Russell Analogy (with colours):
Premise 1. People with jaundice see yellow when no yellow is there.
Premise 2. When people see yellow when no yellow is there, then they are making a mistake.
Conclusion 1. People can make mistakes in colour perception.
Premise 3. Seeing that something is yellow is like feeling that something is good (or bad).
Conclusion 2. People can make mistakes in their feelings about good and bad.

The suggestion here is that Kramer or Hitler had a kind of moral jaundice: they
mistook evil for good. This is an interesting addition to the Russell Analogy, for
the introduction of jaundice implies that something is wrong in the subject or in the
organ or faculty by which he sees yellow where there is no yellow.
Russell’s sketch of a theory of mistakes is quite interesting in itself, because it
implies that feelings are a form of cognitive processing, for they can be right or
wrong, true or false. The question whether feelings can be mistaken (Q72)
is quite old, having been aired already in Plato’s Philebus (from 36C on). Most
philosophers tend to say No to Q72, but Russell is apparently saying Yes, at least as
far as the feelings of good and bad are concerned. To that extent, his theory seems
close to Franz Brentano’s, who defined something as good if loving it is right, i.e.
not mistaken (1889; for a discussion see Nelson 2016, p. 130).

Segment IV, Turn 105 (Copleston) Argumentation Stage (Challenge)

[105.1] Yes, one can make mistakes, but can you make a mistake if it’s simply a
question of reference to a feeling or emotion?
There are at least two different ways in which a philosopher can interpret the idea
that feelings can be mistaken. To be more concrete, I shall speak of love and hate,
230 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

given that Russell uses these two words to refer, respectively, to the feeling of good
and the feeling of evil (see comment on 104.2):
(1) x is mistaken about loving y ≡ x thinks x loves y & x does not love y
x is mistaken about hating y ≡ x thinks x hates y & x does not hate y
(2) x is mistaken about loving y ≡ x loves y & y is not worthy of love
x is mistaken about hating y ≡ x hates y & y is not worthy of hate

Interpretation (1) is what Plato was considering in his Philebus, and it is a well-
known problem in the epistemology of feeling: whether we can be wrong
about our own feelings (Q73). As we shall presently see, this is precisely the
interpretation that guides Copleston in his present request, which he issues as part
of his role as an Antagonist.
Interpretation (2) is quite different, and it is arguable that, if Russell should take
that path, Copleston would probably use it against the moral relativism he suspects
in Russell’s position. In (1), we have feelings and truth values of propositions about
feeling; in (2), not a truth value but a different kind of value, perhaps a moral value,
is posited, a value which, incidentally, Copleston will probably find quite acceptable
(good ≡ worthy of love).
[105.2] Surely Hitler would be the only possible judge of what appealed to his
emotions.
Copleston is taking (1) as the correct interpretation of what Russell may mean when
he suggests that Hitler made a mistake by considering Kramer’s behaviour “good
and desirable” (103.3).
Before going on, however, we may want to consider Russell’s sketch of his position
at 102.3. This might be an alternative interpretation of what he means by being
mistaken:

(3) y is good ≡ (∀x)(if F(x), then y L-strikes x)

y is bad ≡ (∀x)(if ~ F(x), then y H-strikes x)

A couple of explanations on this bit of formalization are in order.

First, I introduce two predicates, ‘to L-strike’ and ‘to H-strike’, which mean ‘to
strike with love (or hatred)’ or ‘to make someone love (or hate) something’. This is
in accordance with the convention suggested in my comment to 102.3, a convention
whose intention is to take seriously the Russell Analogy.
Secondly, I introduce a dummy predicate ‘F’ which is supposed to ground
Russell’s theory of mistakes: the cause for the mistake would not reside in a quality
of the object of the feeling but rather in a quality of its subject (or, perhaps, more
specifically, in a quality of the subject’s faculty of feeling). Given that Russell insists
on the analogy with colours, and given that he appealed, in an elaboration of that
analogy, to the case of jaundice (104.2), we may start by saying that, in the case of
colours, F≡ healthy (say, not being affected by jaundice in the narrower case of the
colour yellow). Continuing with the analogy, therefore, we may, in the case of good
7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument 231

and bad, suggest something like F≡ in one’s right mind. Could something like this
be what Russell means? (For other possibilities, see comment on 102.3.)

Segment IV, Turn 106 (Russell) Argumentation Stage (Response)

[106.1] It would be quite right to say that it appealed to his emotions,

It seems at first as though Russell agrees with Copleston. Note, however, that Russell
does not retain Copleston’s predicate ‘to judge’, which seems a kind of action of the
judging subject. Russell prefers a more passive predicate: it is Kramer’s behaviour
that does something to Hitler, namely, it appeals to his emotions, that is, in my
terminology, it L-strikes Hitler.
[106.2] but you can say various things about that, among others that, if that sort
of thing makes that sort of appeal to Hitler’s emotions, then Hitler makes quite
a different appeal to my emotions.
There seem to be two ways to interpret this somewhat convoluted passage, which
seems to offer a material starting point:
(i) Kramer’s behaviour L-strikes Hitler, and this fact H-strikes Russell. In Cople-
ston’s way of putting things, Russell judges that it is bad that Hitler judges that
Kramer’s behaviour is good. In this interpretation, the object of moral judgment
or moral feeling is a state of affairs.2 This interpretation has the disadvantage
that it is clearly not what Russell said, although it might be what he meant. In
its favour, we can say that it is the sort of theory that Russell cultivated when he
famously struggled with a relational theory of judgment (for this controversial
part of Russell’s philosophy, see Griffin 1985; Wahl 1986; Landini 1991).
(ii) Kramer’s behaviour L-strikes Hitler, and because of that Hitler H-strikes
Russell. In Copleston’s terminology, Russell judges that Hitler is bad because
Hitler judges that Kramer’s behaviour is good. In this interpretation, the object
of moral judgment or moral feeling is a person. This interpretation has the
advantage that it is what Russell actually says.
It is a notorious fact that moral talk easily shifts from qualifying acts as morally
good (right, just, correct, admirable, or whatnot) to extending the qualification to
cover states of affairs of various sorts associated with the acting (including results
and consequences of the acting), then to the agents themselves, their characters,
their habits, their intentions, and so on, and even to the rules, principles, maxims,
agendas, and whatnot presumably underlying or fostering the acting. So, one question
that arises from this passage is what sort of thing can or should receive

2 It is a second-order state of affairs. Leaving the quotes aside for comfort: if p is the proposition
that Kramer behaved in such and such a way, then L (p, a) is the state of affairs that p inspires ‘love’
in Hitler (here denoted by a), and then, of course, H ([L (p, a)], b) is the state of affairs that L (p,
a) inspires ‘hatred’ in Russell (here denoted by b). In other words, then, interpretation (i), namely
what Russell might have meant, is: if L (p, a), then H ([L (p, a)], b). In contrast, interpretation (ii),
or what Russell said, would be, more simply: if L (p, a), then H (a, b).
232 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

the predicate of goodness or badness. Call this Q74. It is not explicit as a

question and it will not be taken up in the sequel.

Segment IV, Turn 107 (Copleston) Argumentation Stage (Request for Clarification)

[107.1] Granted.
Copleston agrees with Russell’s material starting point, whatever it is (see discussion
of 106.2 on the two interpretations that are possible).
[107.2] But there’s no objective criterion outside feeling, then, for condemning
the conduct of the Commandant of Belsen, in your view?
Copleston insists on Q71, but now he uses the phrase “objective criterion”. This gives
us a variant which deserves spelling out: whether the goodness or badness
of the behaviour of the commandant of belsen depends only on the
subjective criterion of feeling or whether it depends besides on an
objective criterion. This passage is a further confirmation that Copleston did
indeed intend Q71 and not just Q70 at 103.4.
Just in passing, Copleston’s formulation indicates that he is well aware, and gladly
accepts, that certain states of affairs, actions, people, and so on have an effect on our
feelings and emotions; but he wants to separate this domain of (mental) phenomena
from the alleged objective aspects pertaining to our moral judgments.

Segment IV, Turn 108 (Russell) Argumentation Stage

[108.1] No more than there is for the colour-blind person who’s in exactly the
same state.
Russell first holds on to his analogy (see Turns 98, 102, 104). This means that,
whatever should be the case for the colour-blind person, that is also the case for the
‘morally blind’ person (such as, presumably, Hitler or Kramer).
[108.2] Why do we intellectually condemn the colour-blind man? Isn’t it because
he’s in the minority?
Russell puts forward two rhetorical questions. First, he asks an open question:
‘Why do we intellectually condemn the colour-blind man?’ (Q75). Observe that
Q75 is, curiously enough, formulated in terms of ‘intellectual condemnation’. Is this
Russell’s way to say that, in the analogous case of good and bad, a moral condemna-
tion is issued instead of an intellectual one, and a moral condemnation is a feeling-
based, and emotional condemnation? In the light of everything that Russell has been
saying, this is quite likely.
There may be a good number of answers to Q75. However, Russell narrows down
the possible range of answers to just one, by asking a closed question and thus hinting
at what he considers the right answer: ‘Isn’t it because he’s in the minority?’ This
7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument 233

would then be Q76. Russell’s argument, contained in the nutshell of his rhetorical
question, may be represented as follows:
Premise 1. We intellectually condemn the colour-blind man because he is in the minority.
Premise 2. Colour-blindness is like a mistake in the realm of good and bad.
Conclusion. We morally condemn people who make mistakes in the realm of good and bad
because they are in the minority.

Applied to the concrete case of Q71, this conclusion amounts to saying that our moral
condemnation of Hitler (or Kramer) is based on a statistical fact: there are many more
people who hate what these people did and stood for than there are people who love
it. Russell is indeed biting the bullet.
Now, we should remember that Russell very much defers to science in all things.
And for post-Galilean, quantitative science, colours do not exist as such—they are
secondary qualities (Russell would perhaps say: predicates), dependent on our organs
(eyes and visual cortex). What exists are only the primary qualities, those that can
be captured by the ever more sophisticated quantitative methods of science. In the
case of colours, what exists is a certain position in the spectrum. Yellow is just the
sensation evoked in our organs by a wavelength of between 590 and 560 nm and a
frequency of between 510 and 540 terahertz. When a coloured body reflects light
within those parameters, then our eyes and brains yield the sensation of yellow.
There is one element missing in this picture, though. For light to have that effect
on our organs, these should be properly constituted (anatomically) and they should
function properly (physiologically). Wavelength and wave frequency by themselves
are not enough. Now, proper constitution and proper functioning presuppose criteria
(see Plantinga 1993). Those criteria are simply brushed aside by Russell in favour
of pure statistical counting.

Segment IV, Turn 109 (Copleston) Argumentation Stage (Meta-Argumentation)

[109.1] I would say because he is lacking in a thing which normally belongs to

human nature.
This time, Copleston finally catches on. The missing link in the Russell Analogy (see
comment on 102.3 and 105.2) becomes explicit by the use of the adverb ‘normally’.
Note, however, that this adverb has two different connotations. One is precisely
the statistical one invoked by Russell in 108.2. The other is the anatomical and
physiological (ultimately, medical) criterion of proper constitution and functioning—
a topic dear to Aristotle and so to Copleston as well. That this is the way Copleston is
thinking is clear from his introduction of the loaded philosophical concept of human
nature. His argument would then be:
234 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

Premise 1. We intellectually condemn the colour-blind man because he is lacking in a thing

which normally belongs to human nature.
Premise 2. Colour-blindness is like a mistake in the realm of good and bad.
Conclusion. We morally condemn people who make mistakes in the realm of good and bad
because they are lacking in a thing which normally belongs to human nature.

Note that Copleston, in this argument, is admitting the Russell Analogy. This admis-
sion may be just for the sake of argument or it may be because it has dawned on him
that an Aristotelian reading of that analogy is possible and may yet lead to rebutting
Russell.
We can conceive of Copleston’s move as a meta-argument, indicating that the
above way of constructing Russell’s argument is better than the one at 108.2. It all
depends, of course, on the interpretation of ‘normal’.

Segment IV, Turn 110 (Russell) Argumentation Stage

[110.1] Yes, but if he were in the majority, we shouldn’t say that.

Russell concedes Copleston’s point at 109.1, but obviously taking ‘normally’ in its
merely statistical sense. The argument:
Premise 1. If the majority of people were colour-blind, then we would not say that they are
lacking in a thing which normally belongs to human nature.
Premise 2. Colour-blindness is like a mistake in the realm of good and bad.
Conclusion. If the majority of people felt like those who we now consider morally mistaken,
then we would not say that they are lacking in a thing which normally belongs to human
nature.

In this argument, it is clear that the philosophical concept of human nature is not
taken seriously, but rather as philosophically unimportant, a mere phrase. There is, in
Russell, no such thing as human nature (in the sense of Aristotle or Aquinas). What
he has instead is no more than a statistical concept.

Segment IV, Turn 111 (Copleston) Argumentation Stage

[111.1] Then you’d say that there’s no criterion outside feeling that will enable
one to distinguish between the behaviour of the Commandant of Belsen and the
behaviour, say, of Sir Stafford Cripps or the Archbishop of Canterbury?
Copleston chooses not to argue the philosophical point about ‘normal’ and simply
returns to Q71 in the variant formulated at 107.2. The only thing that enriches that
variant is the addition of a term of comparison: Sir Stafford Cripps, a British politician
well known for being a Christian, a Socialist, and a sworn enemy of Fascism and
7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument 235

Nazism; and William Temple, the celebrated anti-Nazi Archbishop of Canterbury

during the war years 1942–1944. These examples are carefully chosen, because
Cripps and Temple were considered at the time outstanding instances of integrity
and courage against Nazi Germany. Therefore, the variant of Q71 could perhaps be
phrased like this: whether the goodness or badness of the behaviour
of the commandant of belsen, as compared to the behaviour of sir
stafford cripps and archbishop william temple, depends only on the
subjective criterion of feeling or whether it depends, apart from
feeling, on an objective criterion. Copleston is being as specific as one
could desire.

Segment IV, Turn 112 (Russell) Argumentation Stage

[112.1] The feeling is a little too simplified.

Russell seems to retract from what he himself says at 106.2. There, he says that
Hitler or his feelings towards Kramer’s behaviour had a negative appeal on Russell’s
feelings. By saying that in the context of discussion in which he says it, he implies
that it is this feeling of his that is the only criterion of badness. In this passage,
however, he says that this is not enough, that it is too simple.
[112.2] You’ve got to take account of the effects of actions and your feelings
toward those effects.
Russell now introduces a new object for the moral feeling, namely, the effects of
actions. Very well then, it is the effects of Kramer’s behaviour, and perhaps the
effects of Hitler’s approval of it, which Russell hates, or which H-strikes Russell.
Please observe that, by this argumentative move, Russell is anticipating the neo-
utilitarian position he will fully embrace at Turn 120. (In philosophical shop talk, he
is adding moral consequentialism to moral emotivism.)
[112.3] You see, you can have an argument about it if you can say that certain
sorts of occurrences are the sort you like and certain others the sort you don’t
like. Then you have to take account of the effects of actions.
Russell now tackles one of the most important issues contained in Q71: in which
way one can argue with people who do not share one’s feelings
of good and bad. Let’s call this Q77. Russell answers to it by enlisting two
conditions:

(1) One explains to other people which sorts of occurrences one likes and which
sort of occurrences one does not like.
(2) One asks other people to consider the effects of actions.

Applied to the concrete cases invoked by Copleston, Russell would be saying that, if
somebody has feelings opposed to his in regard to Hitler, Kramer, Cripps or Temple,
then he would first explain what he likes (e.g. integrity and courage in opposition to
236 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

fascism) and what he does not like (e.g. cruelty in concentration camps); and then he
would urge consideration of the effects of opposing fascism and the effects of torture
and wanton maiming and killing. We shall refer to this as Russell’s Argumentative
Strategy. We have thus the following meta-argument:
Premise In arguments about good and bad, it is possible to argue in accordance to Russell’s
Argumentative Strategy.
Conclusion. In arguments about good and bad, it is neither necessary to invoke special moral
qualities inhering in human actions and characters nor to base one’s arguments on traditional
concepts such as human nature.

In other words, for Russell there is no other reasonable way to argue in favour of
one’s moral feelings. We can therefore say that, in order to defuse the charge of moral
relativism contained in Q71, Russell proposes a method of argument that goes beyond
the mere expression of personal feelings. It is only the simplified version which makes
Copleston believe that Russell is incapable of arguing reasonably about goodness or
badness; but the more complex version will correct that impression.
[112.4] You can very well say that the effects of the actions of the Commandant
of Belsen were painful and unpleasant.
In arguments performed according to Russell’s Argumentative Strategy (as sketched
in 112.3), it is appropriate to argue that Kramer’s conduct has painful and unpleasant
effects. Note that the wording chosen by Russell is characteristic of British under-
statement. Torture, maiming, and outright killing are, of course, really much more
than just painful (as a toothache would be) or unpleasant (as dentists getting their
hands in people’s mouths would be for those people). Yet Russell’s meaning is clear,
if we take British rhetoric habits in our stride.
From all this it follows that a variant of Q66 has been raised: whether the good
is a matter of human feeling emerging from the effects of actions,
or even whether human feeling, as it emerges from the effects of
actions, is the origin and ground of all things good. This is, as we
now see, the full and specific meaning intended by Russell.

Segment IV, Turn 113 (Copleston) Argumentation Stage

[113.1] They certainly were, I agree, very painful and unpleasant to all the people
in the camp.
Copleston understand the under-statement and accepts what is intended by it.
However, he ignores the underlying message about how to argue when facing
disagreement as to feelings of what is good and bad. Copleston does not seem to
register Russell’s Argumentative Strategy, so he does not engage with Russell’s meta-
argument. Instead, Copleston keeps thinking only of Russell’s feelings as opposed
7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument 237

to Kramer’s or Hitler’s. So, naturally, Copleston insists on the fundamental fact

that pain and unpleasantness, as feelings, belong exclusively to the inmates of the
concentration camps. His argument seems to be this:
Premise. Pain and unpleasantness are feelings experienced by the camp inmates and by no
one else.
Conclusion. If feeling is all there is to good and bad, then all we can say is that Kramer’s
behaviour was bad for camp inmates and for no one else.

Given that Copleston firmly believes that Kramer’s behaviour is bad, independently
of anyone’s feelings, he is still struggling with that moral relativism which he believes
Russell’s position to entail. Copleston is therefore missing the point.

Segment IV, Turn 114 (Russell) Argumentation Stage

[114.1] Yes, but not only to the people in the camp, but to outsiders contemplating
them also.
Of course, Russell counters that he is not just talking of the feelings of the camp
inmates, the victims of Nazi cruelty, but also of outsiders who contemplate the
suffering (for instance, we may add, the feelings of the judges presiding in the Belsen
trial, see Phillips 1949). His argument could be rendered as follows:
Premise 1. Pain and unpleasantness are feelings experienced both by the camp inmates and
by Russell as well as other outsiders contemplating all that suffering.
Premise 2. Feeling is all there is to good and bad.
Conclusion. Kramer’s behaviour was bad both for camp inmates and it is bad for Russell as
well as other outsiders contemplating all that suffering.

Of course, one could counter that it should be possible to say that Kramer’s behaviour
is bad absolutely, but this is exactly what Russell rejects (see Turn 122). It may be
of interest to quote from a little-known paper, written by Russell in 1922 against the
absolute conception of the good:
Since people disagree in their judgments of good and bad to just the same extent to which
they differ in their feelings of approval and disapproval, the objectivity secured by ethical
predicates is only theoretic, and does nothing to mitigate ethical disputes in practice [Pigden
1999, p. 124].

In that paper, Russell gives this as one (out of five) reasons to reject an absolute
conception of the good. Still, in order to have a critical discussion, Russell thinks it
suffices to follow his Argumentative Strategy (112.3).
238 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

Segment IV, Turn 115 (Copleston) Argumentation Stage

[115.1] Yes, quite true in imagination.

Copleston accepts that outsiders contemplating the suffering inflicted by Kramer on
the inmates of the Bergen-Belsen camp will have feelings of disapproval, albeit they
will not have the suffering itself except in imagination.
[115.2] But that’s my point.
Nonetheless, Copleston thinks that the fact of disapproval by as many outsiders as
Russell may postulate is not enough.
[115.3] I don’t approve of them, and I know you don’t approve of them, but I
don’t see what ground you have for not approving of them, because, after all,
to the Commandant of Belsen himself they’re pleasant, those actions.
The disapproval of outsiders, either Russell or anyone else, needs a ground or reason.
This seems to be part of an argument like this:
Premise 1. Commandant Kramer found his actions pleasant.
Premise 2. Goodness or badness is just a matter of feeling.
Conclusion. Kramer’s actions are good (for him).

Never mind whether Kramer’s actions are bad for (i.e. are disapproved by) Russell
or any other outsider contemplating them, the fact of the matter (if Premise 2 is
correct) is that they are good for Kramer, or for Hitler or for many Nazi outsiders
contemplating the very same actions. Somehow, only if a reason, apart from feeling,
could be given for the badness of Kramer’s action, could we say that they are bad.
We can see that Copleston is still following his one track, without trying to under-
stand Russell’s Argumentative Strategy. He is therefore not engaging the argument
of the Protagonist, something that belongs to his duties as Antagonist.

Segment IV, Turn 116 (Russell) Argumentation Stage

[116.1] Yes, but you see I don’t need any more ground in that case than I do in
the case of colour perception.
Russell accepts that for Kramer his actions are pleasant and therefore good for
Kramer. But this is not a matter of giving grounds or reasons: the Russell Analogy
kicks in and precludes grounds or reasons, in the sense of justifications (see Turns
97–99) from being able to be produced by Kramer, Russell or any other party.
[116.2] There are some people who think everything is yellow, there are people
suffering from jaundice, and I don’t agree with these people. I can’t prove that
the things are not yellow, there isn’t any proof.
7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument 239

Russell’s argument, up to this point, says:

Premise 1. Some people, suffering from jaundice, think everything is yellow.
Premise 2. Russell does not think everything is yellow.
Premise 3. There is no more ground to say that everything is yellow (as a person afflicted
by jaundice) than there is to say that not everything is yellow (as a normal person).
Conclusion 1. There is no point in asking Russell to give a ground to say that not everything
is yellow or to prove it.

Notice that, in the typical manner of empiricists, a certain perceptual question (Q78:
what ground there is to say that not everything is yellow or that
everything is) is deemed to lack meaning.
[116.3] But most people agree with me that they’re not yellow, and most people
agree with me that the Commandant of Belsen was making mistakes.
By means of Russell’s analogy, the argument at 116.2 can be completed as follows:
Premise 4. Colour perception is like feeling that something is good or bad.
Conclusion 2. There is no point in asking Russell to provide a ground to say that certain
things (like Kramer’s deeds) are bad or to prove that they are

As with the argument at 116.2, the parallel moral question (Q79: what ground
there is to say that something is bad or that it is good) has no more
meaning than the perceptual one.
To be consistent, Russell should, of course, say that, there is no point, either,
in asking a person with jaundice to provide a ground for saying that everything is
yellow (let alone a proof of that proposition). Analogously, there would also be no
point in asking Kramer or Hitler to provide a ground for saying or thinking that
robbing, manhandling, imprisoning, and killing Jews is good (let alone a proof of
that proposition). The only problem with that answer is that Hitler and other Nazis
did think and say, loud and clear, that they had very good grounds and arguments
to do what they did to the Jews of Europe. They kept giving those grounds and
arguments, orally and in writing, again and again (see Weikart 2009). It is a shame
that Copleston does not press Russell a bit more on such an important point.

Segment IV, Turn 117 (Copleston) Argumentation Stage

[117.1] Well, do you accept any moral obligation?

Faced with Russell’s resistance, Copleston decides to take a different tack. He keeps
the role of an Antagonist and continues his interrogation. However, like an attorney
who sees that a certain line of questioning is not getting anywhere with a witness,
he changes it. Leaving the question of the exclusiveness of feeling as a criterion
for moral judgment (Q71) aside for the moment, he introduces a new concept and,
knowing by now how prone Russell is to disqualify philosophical terms as meaning-
less, Copleston cautiously asks whether Russell would accept such a concept. In view
240 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

of the discussion so far, we could then formulate Copleston’s question as follows:

whether the concept of moral obligation makes sense or not. Let’s
call this Q80.
It is important to keep in mind that the concept of moral obligation is, in modern
philosophy, usually identified with Kantian ethics. However, it is much older. In
particular, it is an essential component of the natural law tradition. Being a Jesuit
trained in scholastic philosophy, it is more than likely that Copleston is thinking of
traditional natural law rather than Kantian ethics when he introduces the concept
of moral obligation into the discussion. (It could, of course, be argued that Kant
belongs to the natural law tradition, even if a somewhat rebellious latecomer, see e.g.
Schneewind 1993).
At bottom, it is this connection with natural law, which in the Christian tradition
means the law handed down by God Himself, which underlies the question raised
here by Copleston (see below 121.1).

Segment IV, Turn 118 (Russell) Argumentation Stage

[118.1] Well, I should have to answer at considerable length to answer that.

Practically speaking—yes. Theoretically speaking, I should have to define moral
obligation rather carefully.
Russell is no less cautious than Copleston. Remember that he has already made a
major retraction (cf. 22.1, 40.1) and, arguably, also a minor one (106.2, 112.1). He
probably wants to tread more carefully.
Notice the curious dualism of theory and practice in 118.1. In practice, Russell
accepts moral obligation: in his life, Russell accepts having moral obligations as
much as the next decent person. In theory, however, Russell has some reservations:
in order to accept the concept as part of a theoretical account, he would need to
define it carefully, probably so carefully as to deprive it of its ordinary meaning.
One interesting question arising at this point is: what the point is, in moral
matters, to distinguish between theory and practice. This is one of
the oldest questions in Western moral philosophy, put forward at the beginning of
Aristotle’s Nicomachean Ethics. Let us call it Q81. It will not be taken up in the
discussion.

Segment IV, Turn 119 (Copleston) Argumentation Stage

[119.1] Well, do you think that the word ‘ought’ simply has an emotional
connotation?
Copleston tries now with a narrower question. Instead of speaking, quite abstractly, of
moral obligation, which is a general concept, he now speaks of a word, a modal verb,
‘ought’, and of its possible meaning. This confirms that Copleston is worried about
the possibility of Russell denying meaning to a phrase like ‘moral obligation’ or a
verb like ‘ought’. Copleston guardedly speaks of connotation, a traditional semantic
7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument 241

concept which cannot be rejected, even for suspicious concepts. Thus, a word like
‘God’ can certainly be said to be empty, that it does not denote any existent individual;
but it cannot be denied that it connotes something, at the very least certain emotions,
even quite strong ones. So, the question arising here is: whether ‘ought’ has an
emotional connotation and that is all there is to it. Call this Q82.
Given that Russell has said that feeling is all there is to questions of good and bad,
then it is easy to draw the consequence that, no matter what he thinks about the deno-
tation or lack of denotation of a phrase like ‘moral obligation’ or a verb like ‘ought’,
he will at least say that they have emotional connotations. The question is whether
their meaning goes beyond that. Hence the adverb “simply” in this intervention.

Segment IV, Turn 120 (Russell) Argumentation Stage

[120.1] No, I don’t think that,

Russell seems to admit here that ‘ought’ (and ‘moral obligation’) do have some
meaning beyond an emotional connotation. Russell seems to be harking back to
112.1, where he insists that talk of feeling is not enough.
[120.2] because you see, as I was saying a moment ago, one has to take account
of the effects,
As he explained in Turn 112, Russell explains that his position is not the lazy rela-
tivistic position of ‘everyone has his or her own feelings’. He is rather proposing an
argumentative strategy, which is a very pertinent response to Copleston’s objection
to the facile moral relativism he attributes to Russell. And, as I said before (comment
on 113.1), Copleston does not engage with that response. Now, if we put Russell’s
insistence on this point (“as I was saying a moment ago”) in relation to Q82, then
his answer seems to be: The word ‘ought’ does not only have an emotional conno-
tation. Whether the effects one has to take into account are part of the connotation
(or even the denotation) of ‘ought’ and ‘moral obligation’ is far less clear.
[120.3] and I think right conduct is that which would probably produce the
greatest possible balance in intrinsic value of all the acts possible in the
circumstances,
Russell now introduces a new concept, right conduct, which is presumably related
to ought as per a definition like this:
(1) x’s conduct is right ≡ x does what x ought to do
Now, Russell gives a definition of his own, which is a variant of the old utilitarian
definition of a good action. Utilitarianism has been objected to many times, and
it seems that Russell is trying to avoid some, or even all, of those objections, by
introducing four important concepts:
(i) We often cannot predict what the consequences of our action might turn out
to be: hence the adverb “probably”.
242 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

(ii) Pleasure as the index of utility is replaced by “intrinsic value”.

(iii) The greatest amount of utility for the greatest number of sentient agents
becomes “the greatest possible balance” (of intrinsic value).
(iv) Russell emphasizes “the [particular] circumstances” in which an agent stands
before making a decision (what neo-classical economists refer to when they
talk about people acting ‘on the margin’).
Each one of these concepts would merit some discussion, which I shall forgo here.
Apart from them, Russell also makes a controversial move by blurring the well-known
distinction between the two commonly recognized methods of ethics: deontology
and consequentialism. He blurs the distinction by equating right conduct (in his
definition) and moral obligation. It should be remembered, as I said in my comment
to 117.1, that Copleston is not making a Kantian but a natural law move when he
introduces the concept of moral obligation. Natural law not only is incompatible with
utilitarianism, but utilitarianism was born as a direct and frontal attack on the natural
law tradition.
It follows from this analysis that Copleston, in his role as Antagonist, would have
enough objections to make to this part of Russell’s intervention.
[120.4] and you’ve got to take account of the probable effects of your action in
considering what is right.
This is just a corollary of Russell’s consequentialist position. Note that old utilitar-
ianism assigned a great role to the very basic feeling of pleasure and pain. The two
discussants have certainly talked about those (112.4, 113.1), yet they have also used
the somewhat differently aligned concepts of love and hate or hatred (92.1, 93.2,
96.2).
At this point, the two main components of Russell’s moral outlook should be clear:
feelings and effects—more precisely, the feelings emerging in people anticipating
the probable effects, or contemplating the actual effects, of actions, their own or
other people’s. Anything that is said about good and bad, right and wrong, moral and
immoral, obligation and prohibition, has no meaning except in relation to that.
In other words, Russell is saying to Copleston that his position is no moral
relativism but rather what we may call ‘utilitarian emotivism’.

Segment IV, Turn 121 (Copleston) Opening Stage (Role Assignment), Argumentation
Stage

[121.1] Well, I brought in moral obligation because I think that one can approach
the question of God’s existence in that way.
Copleston reveals his intention in introducing the concept of moral obligation at
117.1. After all, the general question underlying and triggering this debate is still
Q1: whether god exists or not (1.1). So, Copleston is re-assuming his role as
Protagonist and begins to construct his case.
7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument 243

[121.2] The vast majority of the human race will make, and always have made,
some distinction between right and wrong.
This is a classical move within the natural law tradition, the consensus of mankind.
Interestingly enough, Copleston does not speak of universal but only of near-universal
consensus. Note that the near-universal consensus Copleston is talking about does
not refer to the content of the right-wrong distinction but only to the form, i.e. to
the fact that there is such a distinction. This point will become crucial later on (see
129.2).
Copleston’s cautiousness may have been conditioned by the all too recent horrors
of World War Two and especially the shocking discovery of the Nazi concentration
camps. It is difficult to say with any certainty. In any case, even in this cautious form,
Copleston’s consensus argument is quite an argument: if we accept such a consensus
as a fact, it certainly cries for explanation.
[121.3] The vast majority, I think, has some consciousness of an obligation in
the moral sphere.
Copleston makes a further step, going from the distinction between right and wrong to
the consciousness of moral obligation. The former could be just a matter of convention
and practice; the latter cannot be reduced to any of those. The need to explain that
is even greater than the mere distinction of right and wrong.
[121.4] It’s my opinion that the perception of values and the consciousness
of moral law and obligation are best explained through the hypothesis of a
transcendent ground of value and of an author of the moral law.
We have again an inference to the best explanation (see Turns 71 and 77). Note that
Copleston presents it with some diffidence (“it’s my opinion”). He is perfectly aware
that there may be other explanations. Russell will in fact try to give an alternative
one. The argument is then:
Premise 1. There is near-universal consciousness of moral obligation.
Premise 2. The best explanation for the near-universal consciousness of moral obligation is
that God exists.
Conclusion. God exists.

There are two questions involved in this argument. One is a factual question:
whether it is a fact that there is near- universal consciousness
of moral obligation or not. Let’s call this Q83. The other is an explana-
tory question: whether the fact of near- universal consciousness of
moral obligation can best be explained by god’s existence. This is
Q84. Premise 1 of the above argument says Yes to Q83 and Premise 2 says Yes to
Q84. As for Russell, he will not directly say No to Q83, although he will re-interpret
the import of the alleged fact; and he will definitely take the negative in regard to
Q84.
It is important to notice that, in 121.4, Copleston gives of God, first, an impersonal
description (“a transcendent ground of value”), and then a personal one (“an author
244 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

of the moral law”). This is a common feature of theological arguments. The concept
of God is a kind of ‘limit concept’ or ‘border concept’ (Grenzbegriff , cf. Husserl
1913, §79, fn. 1), which theologians are the first to say is enormously difficult, if not
indeed impossible, to apprehend. Notice that, by that description, we are now quite
far from the original definition as the creator of the world (1.3). The old theological
question of how the ground of goodness could possibly also be the
creator of the world (Q85) is here left hanging (see comment on 7.2). Again,
none of the discussants will take it up.
[121.5] I do mean by “author of the moral law” an arbitrary author of the moral
law.
This is a crucial move. When Copleston admits that the moral law, as conceived by
him, is just the divine will, then the old question, raised in Plato’s Euthyphro, (i.e. the
question whether the deity commands the good because it is good or
rather, the other way around, whether the good is good because the deity
commands it) is given the latter answer. The word “ground” at 121.4 is meant in
its most literal sense.
[121.6] I think, in fact, that those modern atheists who have argued in a converse
way There is no God; therefore, there are no absolute values and no absolute law
are quite logical.
Finally, having equated the absolute good of moral obligation with God’s will,
the only remaining thing for Copleston to recognize is his fundamental agreement
with those people who say, with Doestoevsky: If God does not exist, everything is
permitted.3 This agreement is, by the way, the underlying motivation of Elizabeth
Anscombe’s famous moratorium on all moral philosophy, which was issued only
a decade after the BBC broadcast and directly targeted at the concept of a moral
obligation without divine command (Anscombe 1958).
The question raised here is whether the moral law is absolute or not.
Let’s call it Q86. It is clear that Copleston says Yes to Q86. Moreover, the moral
law is absolute because it is God’s will, and the will of God certainly is absolute if
anything is.

Segment IV, Turn 122 (Russell) Subdiscussion (Confrontation Stage, Argumentation

Stage)

[122.1] I don’t like the word “absolute”.

This is a surprising statement. One would have thought that likes and dislikes are out
of place in a debate. Still, it is revealing of Russell’s attitude.

3 In Demons (1871, Part III, Chapter 6), Brothers Karamazov (1880, Part I, Chapter 7), Letter

to Nicholay Lukich Ozmidov, February 1878). See Scanlan (2002), pp. 16–30; McGrath (2004),
pp. 145–149.
7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument 245

[122.2] I don’t think there is anything absolute whatever.

This seems to be a proper assertion and not a mere expression of feeling, in contrast
with 122.1. As such, it belongs to a subdiscussion: the difference of opinion has been
shifted from God’s existence (the initial question, Q1) to the existence of absolutes,
one of which is Copleston’s moral law (Q86). Within this subdiscussion, we can say
that Russell is now putting forward a forceful standpoint.
Before going on, it may be useful to make clear that such a standpoint on the part
of Russell is not without problems.
One trouble is that what Russell says is said in an absolute tone of voice (“what-
ever”). So, there is something of a paradox here, as though he was saying: There is
absolutely no absolute or There is no absolute and that is absolute.
Besides, Russell had earlier on implied that ‘contingent’ was not a useful concept
because nothing exists in any other way. So, if nothing is absolute, the concept has
no use; but then, the concept of ‘relative’, which is absolutely (!) central to Russell’s
philosophy, should not be any more useful (see comment on 16.2).
Finally, Russell would surely accept that the laws of logic, to which he sought all
his life to reduce mathematics, are absolute. As far as I know, he never contemplated
the possibility of other logics.
All this only goes to show that Russell is probably expressing an emotion, again,
as in 122.1, only with different words. If so, then it would, again, not quite be a
standpoint.
[122.3] The moral law, for example, is always changing.
Given that neither 122.1 nor 122.2 express a standpoint, it is to be assumed that this
is the standpoint. In the light of Q86 and Copleston’s answer to it (121.6), Russell’s
assertion should be understood as meaning: The moral law is not absolute, i.e. as
contradicting Copleston. Another argument in favour of taking 122.3 as a standpoint
rather than an argument for a standpoint, is that it is a rather lame reason for 122.2:
even if it was accepted that the moral law is not absolute, it would hardly follow that
nothing else is.
Now, it is clear that the moral law that Russell wants to talk about has nothing to
do with Copleston’s concept of moral law. Of course, Copleston could be completely
wrong in saying what he says; it could be that his concept of moral law is incoherent,
meaningless, empty; but in order to engage Copleston’s argument, one has to target
his concepts not somebody else’s. When Russell says, in Segment II of this debate,
that it does not make sense to raise the question of the cause of the world, he did
engage Copleston’s argument; for, even if a philosopher rejects a presupposition of
another philosopher’s argument, she should try to speak the same language. This
is not happening here. From now on, we should beware that two different concepts
are at stake. In order to differentiate them, I propose to leave the term ‘moral law’
for Copleston’s concept and use the term ‘human mores’ for Russell’s concept. His
assertion is therefore: The human mores are always changing. This assertion does
not contradict Copleston, it changes the subject.
246 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

Now, if Russell is indeed putting forward a standpoint here, and he believes it is

opposed to Copleston, that would mean that we are in the Confrontation Stage of a
subdiscussion, and the difference of opinion is mixed or bilateral: each of the two
parties has a standpoint.
[122.4] At one period in the development of the human race, almost everybody
thought cannibalism was a duty.
Russell gives an example that justifies his saying that the human mores are always
changing. By putting forward that example, he seems to be arguing; and if he is
arguing, then he is presenting himself as a Protagonist and assuming the burden of
proof.
Let us now consider the example as an argument. Russell says that cannibalism
was a near-universal duty in some remote past. We can represent his argument as
follows:
(1) Premise. At one period in the development of the human race, almost everybody
thought cannibalism was a duty.
Conclusion. The human mores are always changing.

The anthropology expressed in the premise of Russell’s argument is highly dubious.

Russell certainly does not attempt to buttress such an outrageous assertion by citing
any particular source in its favour. However, we can let that pass. The point here is
that he is assuming that, if a certain group of people considers a certain conduct or
rule of conduct as duty, then it is a duty. This is not at all what Copleston, rightly
or wrongly, has in mind when he upholds that some values are absolute. If they are,
then those values cannot depend on what people think, so that one can say that some
people cherish this value and other people don’t, and that’s all there is to it.
Consider now the following argument:
(2) Premise. At one period in the development of the human race, almost everybody
thought cannibalism was a duty.
Conclusion. There is no such thing as the moral law.

This other argument would indeed engage Copleston’s standpoint and argument
at Turn 121. But Russell does not really present (2). If he did, then he would seem
to be upholding moral relativism, as Copleston suspected all along.
Is Russell then a relativist after all? The least one must say is that there is a clear
tension between, on the one hand, rejecting all absolutes, including moral ones, as
Russell does in this turn, and on the other hand, upholding utilitarianism, as he does
in Turn 120. The only way out is to say that Russell is talking about two different
topics: (a) human mores, about which he says they are always changing, and (b)
some unobjectionable version of the objective criterion of utilitarian maximization,
as sketched in Turn 120.
If this is the right way to understand what Russell is up to, then he seems to
be saying: ‘When people like Copleston start talking about the moral law, the only
phenomenon worth discussing is human mores, and the right position is (a). In
contrast, people like Copleston never talk about anything remotely like (b), which
7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument 247

would indeed be a proper subject of discussion yet quite far removed from their talk
about the moral law.’
If something like this is what’s going on here, then what we have is the clash
between two incompatible worldviews, say, the old tradition of natural law (Cople-
ston) versus modern utilitarianism (Russell). If so, then why don’t Russell and Cople-
ston say so clearly once for all, so that at least the issues become clearer? Do they
perchance think that the listeners to the BBC Third Programme broadcast wouldn’t
understand?

Segment IV, Turn 123 (Copleston) Subdiscussion (Argumentation Stage)

[123.1] Well, I don’t see that differences in particular moral judgments are any
conclusive argument against the universality of the moral law.
Copleston presents the classical defence of the universality of the moral law against
the argument of differences in mores. Ever since Plato argued against Protagoras, it
has often been said that the argument of differences in mores is, as Copleston says,
not conclusive.
[123.2] Let’s assume for the moment that there are absolute moral values. Even
on that hypothesis it’s only to be expected that different individuals and different
groups should enjoy varying degrees of insight into those values.
Copleston now presents his argument in a hypothetical form. He asks Russell to
imagine that there are absolute moral values (i.e. that there is a moral law in his
sense, not in the sense of human mores). His argument goes like this:
Premise. If there are absolute moral values, different individuals and different groups would
enjoy varying degrees of insight into those values (i.e. they would have different mores).
Conclusion. The argument from differences in mores against the universality (or near
universality) of the moral law is not acceptable.

Note that Copleston’s original argument started from the existence of a moral law in
the consciousness of almost everybody (note, again, the cautious formulation noted
in my comment to 121.2) and proceeded to God’s existence as the best explanation
for that fact. So, Copleston is taking Russell’s argument as a counterargument to
his—as starting the subdiscussion about the absoluteness of the moral law (Q86).
It is worth observing that there is a remarkable agreement between Russell and
Copleston which, unfortunately, neither of them takes up in order to build upon it. The
long discussion from Turn 104 to Turn 110 is all about the possibility, in the moral
realm, of ‘making mistakes’ (to use Russell’s words) or of ‘enjoying varying degrees
of insight’ (to use Copleston’s). Nonetheless, Copleston misses here the opportunity
to connect what he says here with Russell’s talk of mistakes. Russell will miss that
opportunity as well (at 136.2).
248 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

Segment IV, Turn 124 (Russell) Argumentation Stage, Confrontation Stage

[124.1] I’m inclined to think that ‘ought’—the feeling that one has about
‘ought’—is an echo of what has been told one by one’s parents or one’s nurses.
Russell now presents his alternative explanation for the near universality of the
consciousness of a moral law, or rather, to use his own words, for the feeling that
one has about ‘ought’. So, we don’t need God to explain that feeling. It is enough
to consider parents and nurses. (In that sense, it appeals to the universality or near
universality of parents, nurses, and generally caretakers, to use the contemporary
term.) His argument goes like this:
Premise 1. The feeling that one has about ‘ought’ is a fact to be explained.
Premise 2. The best explanation for the factual feeling that one has about ‘ought’ is that
parents and nurses tell us something when we are children.
Conclusion. The factual feeling that one has about ‘ought’ does not need to postulate God’s
existence.

This argument is intended to disarm Copleston’s argument as reconstructed in my

comment to 121.4. The only remaining trouble is, of course, whether Copleston’s near
universal “consciousness of moral obligation” is the same sort of thing as Russell’s
“feeling that one has about ought”. Copleston’s concept seems, at the very least, to
be much weightier than Russell’s. Perhaps what Russell is doing here is to suggest, in
a muted and polite way, that Copleston’s concept is just a mythological, religious, or
metaphysical elaboration of a much more elementary feeling acquired in the nursery.
However, Russell does not actually say that, but rather he is, again, changing the
meanings in such a way that we lose track and do not know anymore whether the
two discussants are engaging with each other or talking at cross-purposes.
In any case, 121.4 raises a question parallel to Copleston Q84, namely, whether
the fact that one has a feeling about ‘ought’ can best be explained
by what parents and nurses tell us when we are children. This I
shall call Q87. It is pretty clear that there are not two best explanations of the same
fact. So, if we assume, for argument’s sake, that the discussants are not talking
at cross-purposes, then an affirmative answer to Q87 would constitute a negative
answer to Q84, and the other way around. In that case, we would have a difference
of opinion in the Confrontation Stage of another subdiscussion, one about what the
best explanation of X is, where X stands for ‘ought’ in either Copleston’s or Russell’s
sense. Given that each discussant upholds a definite standpoint, always assuming that
the two standpoints are incompatible (a moot point), then the difference of opinion
is mixed.
Observe that, in the reconstruction above, we have said that parents and nurses
tell children something, but what they tell them is not further specified, for the simple
reason that Russell himself does not specify what the object of the ‘telling’ is. Going
by the double meaning of the verb ‘to tell’ in ordinary English in the context of
raising a child, it seems safe to assume, as I shall in the sequel, that parents and
7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument 249

nurses do not only tell them stories but also tell them what to do. Now, ‘telling’ in
the latter sense is a kind of command, only a human instead of a divine command.

Segment IV, Turn 125 (Copleston) Subdiscussion (Confrontation Stage,

Argumentation Stage)

[125.1] Well, I wonder if you can explain away the idea of the Ought merely in
terms of nurses and parents.
By resisting Russell’s answer to Q87, Copleston confirms that, from his perspective,
there is something like a mixed difference of opinion about the same issue. The
way Copleston formulates it reveals that he is aware of having the role of antagonist
with respect to Russell’s standpoint. This is a striking confirmation of one of the
tenets of pragma-dialectics: each one of the discussants is both a protagonist of his
standpoint (i.e. he has the dialectical duty to offer relevant arguments in its favour)
and an antagonist of the other party’s standpoint (i.e. he has the dialectical duty to
raise relevant questions about it).
One interesting feature of 125.1 is the phrasal verb ‘to explain away’. This verb
usually conveys the idea that, under cover of explaining a recognized fact, an account
is given which actually denies that it is a fact. By using that particular phrasal verb,
Copleston is expressing his awareness of what I indicated in my comment to 122.3,
namely, that Russell does not really accept the moral law in Copleston’s sense but
replaces it by a different concept.
So, we have a very interesting situation here, which is probably quite common in
philosophy (see Nelson 2016). Philosopher A asserts that p, and philosopher B asserts
that q. Both A and B believe, or seem to believe, that p and q are incompatible, even
perhaps that p and q contradict each other. Whether there is or not incompatibility
or even contradiction, one thing is clear: A rejects q and B rejects p. So, A and B
are both protagonists of a standpoint and both are antagonists of the other party’s
standpoint. This is sufficient to declare the difference of opinion mixed, even if in
the end it should turn up that they are not really talking about the same thing.
[125.2] I really don’t see how it can be conveyed to anybody in other terms than
itself.
Copleston now gives an argument for rejecting Russell’s position. The argument may
be represented as follows:
Premise. The idea of the Ought cannot be conveyed to anybody in other terms than itself.4
Conclusion. The idea of the Ought cannot be explained in terms of nurses and parents.

This argument has three interesting features.

4 Although it could be argued that the proposition that the idea of the Ought cannot be conveyed
in other terms than itself is reminiscent of Moore’s undefinability of ‘good’, it is unlikely that
Copleston took that from him. There is a whole tradition from Plato to Kant that is a more likely
source.
250 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

First, Copleston uses the definite article with the modal verb ‘ought’. By so doing,
he gives that verb a certain dignity which is absent in Russell’s expression “the feeling
that one has about ‘ought’” (124.1). Copleston’s phrase becomes a synonym for the
concept of moral obligation, and it is a way of distancing himself from Russell’s way
of talking. To mark it clearly, I write “the Ought”.
Secondly, Copleston does not speak of the feeling of ‘ought’. This is an indication
that he does not accept that such a feeling is the same as his moral consciousness.
By the way, the word he uses instead is an interesting choice: ‘idea’, which has clear
Kantian undertones. Few things are in philosophy as opposed as the feeling of X and
the idea of X.
Thirdly, assuming something like Russell’s nursery scenario, Copleston is
asserting this: If the conveyor of the idea of the Ought is a parent or a nurse
and the conveyee is a child, then that idea cannot be conveyed except in terms of
itself . There are, therefore, two possible cases: (a) parents and nurses, when they tell
children what to do, are not conveying the idea of the Ought but doing something
else; (b) parents and nurses do, at least occasionally, actually convey to children the
idea of the Ought, but then this conveying is not just a case of telling them what to
do.
When we combine these two features, we could say that, according to Copleston’s
argument, parents and nurses, by telling them what to do, produce in children a certain
feeling, which we may call, with Russell, “the feeling that one has about ‘ought’”,
but they do not thereby convey the idea of the Ought. Again, if by any chance parents
and nurses ever convey that idea, that idea is not a feeling and it is not conveyed just
by telling children what to do.
Note that it is possible that Copleston is wrong and there is no such thing as the
idea of the Ought, so that the premise of his argument falls by the side; but this is
what Copleston means, so that, to knock it down, Russell would have to produce an
argument.
In any case, Copleston’s argumentation raises a new philosophical question:
whether the idea of the Ought can be conveyed in any other way
or only by itself. Let’s call this Q88.
[125.3] It seems to be that, if there is a moral order bearing upon the human
conscience, then that moral order is unintelligible apart from the existence of
God.
Copleston now phrases his point in a different way, by talking about a moral order
bearing upon the human conscience. This is quite likely another way of talking about
the idea of the Ought. Copleston reveals his awareness that such a moral order could
be questioned, for he produces a conditional (“if there is a moral order…”); he is thus
arguing under a hypothesis, asserting that such a hypothesized moral order cannot
be understood except by God’s existence. He is therefore re-stating the argument
offered at Turn 121, not offering a new one.
7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument 251

Segment IV, Turn 126 (Russell) Subdiscussion (Argumentation Stage)

[126.1] Then you have to say one or other of two things.

Russell seems finally aware of what the argument is, and he prepares himself to refute
it by setting forth two non-exclusive alternatives, each as outrageous as the other.
In this turn, Russell is clearly playing the role of antagonist, not so much by
asking questions but by producing a counterargument. (On the other hand, the fact
that Russell employs a disjunction seems to be a sign that he is asking after all,
for he seems at least to challenge Copleston to take one of the horns of the alleged
dilemma.)
[126.2] Either God only speaks to a very small percentage of mankind—which
happens to include yourself—
The first option is that, under Copleston’s hypothesis, the idea of the Ought, or the
moral order bearing upon the human conscience, is conveyed by God to very few
people. Copleston would be among the blessed with God’s confidence.
This option is obviously sarcastic. Copleston does not say that the conveying of
which he is talking is a kind of direct communication from God to humans, let alone to
some privileged part of humankind, and certainly not to Copleston. Russell assumes
this may be what Copleston means, and this option is intended to cast ridicule upon
his interlocutor.
[126.3] or He deliberately says things that are not true in talking to the
consciences of savages.
The second option is that God’s direct communication, in the case of “the consciences
of savages”, is a deliberate act of deception.
This option is awfully condescending and Eurocentric. Russell does not say who
the savages are, perhaps his imaginary dutiful cannibals of time immemorial (122.4),
in any case not British citizens such as himself for sure. This implies that, for instance,
eating other people is wrong for Russell as much as it is wrong for God, yet for some
unfathomable reason God tells savages that it is their duty.
His full argument seems to go like this:
Premise 1. Anybody who asserts that the idea of the Ought cannot be conveyed to anybody
in other terms than itself must necessarily assume that God is the conveyor of the idea.
Premise 2. If God is the conveyor of the idea of the Ought, He either communicates that
idea correctly to a privileged elite or He communicates it incorrectly to a bunch of savages
or both.
Premise 3. It cannot be true that God communicates the idea of the Ought correctly to a
privileged elite nor that He communicates it incorrectly to a bunch of savages nor that He
does both things.
Conclusion 1. It cannot be true that God is the conveyor of the idea of the Ought.
Conclusion 2. Anybody who asserts that the idea of the Ought cannot be conveyed to anybody
in other terms than itself is in error.
252 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

In this turn, Russell does not deploy the conclusions of the above argument in an
explicit manner, but his intention to refute what he takes to be Copleston’s argument
is pretty clear. The only problem is, of course, that Copleston does not say nor imply
what Russell attributes to him in Premise 1. In any case, we should highlight the new
question raised here: whether saying that the idea of the ought can
only be conveyed by assuming that god is the conveyor. This would
be Q89. We see that Russell takes the affirmative on Q89; and we shall soon see that
Copleston upholds the negative.
Finally, it is important to remark, once more, that Copleston could be wrong to
say the things that he does say; but nothing is gained by substituting those things he
says by things he does not say.
Segment IV, Turn 127 (Copleston) Argumentation Stage (Subdiscussion and Main
Discussion)

[127.1] Well, you see, I’m not suggesting that God actually dictates moral
precepts to the conscience.
Copleston starts his intervention by pointing out that Russell has misunderstood him
in putting forward Premise 1. If that premise is not true, the argument offered at
126.3 is not an appropriate counterargument. Copleston clearly says No to Q89.
[127.2] The human being’s ideas of the content of the moral law depends entirely
to a large extent on education and environment,
Copleston concedes part of what Russell implied, although without any particular
emphasis on feelings, which may imply that, for Copleston, this is secondary.
The sentence in 127.2 is, however a bit strange, in that Copleston uses two adver-
bial modifiers, “entirely” and “to a large extent”, which do not harmonize very well.
Without the benefit of the recording, we cannot say whether the second adverbial
phrase is a correction of the first one. It certainly squares better with the rest of his
utterance if we understand Copleston to mean that the moral law certainly depends
to a large extent, yet by no means entirely, on education and environment—to which,
of course, belong the parents and nurses highlighted by Russell.
Note that Copleston speaks here, again, of ‘ideas’ in the plural (“ideas of the
content of the moral law”), which contrast with the solemn (and quasi-Kantian)
singular idea (of the Ought) mentioned in 125.1.
[127.3] and a man has to use his reason in assessing the validity of the actual
moral ideas of his social group.
In a decisive step, Copleston adds that those collective ideas which largely (though
not entirely) depend on education and environment can be, and often are, examined
and eventually modified by the individual. This introduces an important change in
Russell’s “what parents and nurses tell children”, which, not being properly qualified,
seems to imply that the moral ideas instilled in children are fixed forever.
It is of particular importance that Copleston introduces here the idea of reason
as a faculty capable of the abovesaid examination and modification of given moral
7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument 253

ideas. The concept of reason as such a critical faculty has a long history, from the
Pre-Socratics all the way down to Kant. It is as opposed to mere feeling as one could
possibly imagine. Russell may not believe in such a faculty; it is his good philosoph-
ical right; but a philosopher like him, who all his life kept arguing—reasoning—
against the collective moral ideas of his time (from war to marriage), should know
that it is not enough to substitute feelings for reasons without argument. The ques-
tion raised here, call it Q90, is whether there is a faculty of reason by
which we criticize the mores we have been taught. In fact, Russell’s
Argumentative Strategy (112.3) is precisely a way in which we reason (or use one’s
reason) to go beyond the moral teachings of our elders and teachers. Yet, for some
reason, Copleston does not see that there is here a possible rapprochement between
his views and Russell’s. As for the latter, he will be equally blind to a rapprochement
with Copleston.5 Of course, in the end there may be still a difference of opinion
concerning the right method of modifying the code we receive from the society in
which we are born and raised; and as Copleston may never accept utilitarianism in
any form, so may Russell never accept natural law; but they may yet meet halfway
and agree that humans are not condemned to follow blindly our moral upbringing.
[127.4] But the possibility of criticizing the accepted moral code presupposes
that there is an objective standard, that there is an ideal moral order, which
imposes itself, I mean the obligatory character of which can be recognized.
Copleston puts forward a kind of transcendental argument. In order to examine and
modify the moral ideas of our group, an objective standard, an ideal moral order has
to be presupposed. This is the same thing as the idea of the Ought which Copleston
referred to at 125.1, or the concept of moral obligation which he introduced at 117.1
and then used, at 121.1, as the explanandum in an inference for God’s existence.
Following up on the possible rapprochement indicated in my comment on 127.3,
Russell might at this point say: ‘Yes, there is an objective standard, namely some
form of utilitarian maximization; and we could discuss about what form is less objec-
tionable.’ If Russell said something like that, then they might make some progress
in the discussion. But we shall see not only that Russell does not make that step, but
also Copleston makes no effort to build a bridge.
[127.5] I think that the recognition of this ideal moral order is part of the
recognition of contingency.
Copleston makes now a connection with the concept of contingency that was so
central to his metaphysical argument in Segment II of the debate.
[127.6] It implies the existence of a real foundation—of God.
And so, Copleston’s argument ends. It could be reconstructed as follows:
Premise 1. People can and do use their faculty of reason in order to criticize the ideas about
the content of the moral law that they receive from their group.

5 See the long quotation from Copleston’s autobiography in Chapter 1, Section 1.1, footnote 7.
254 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

Premise 2. Such criticism could not take place unless an objective standard is presupposed
whose obligatory character is recognized (by the faculty of reason).
Premise 3. The rational recognition of the obligatory character of the objective standard
that underlies criticism of the collective moral ideas instilled in people is at the same time
recognition of one’s contingency.
Conclusion 1. Recognition of one’s contingency implies the existence of God.
Conclusion 2. Rational criticism of the mores of one’s group implies the existence of God.
Conclusion 3. God exists.

In this argument, something like Premise 2 of Russell’s argument offered at 124.1

(‘the best explanation for the factual feeling that one has about ought is that parents
and nurses tell us something when we are children’) is absorbed into Premise 1 of the
above argument. Seen from Copleston’s perspective, what Russell puts at the centre
is just a small element of a very complex situation. Nonetheless, Russell could have
been brought to admit Premises 1 and 2, even if not Premise 3. The difference of
opinion might have been made more precise and so more fruitful.
Note that Conclusion 3 is not made explicit, but Copleston’s whole line of argu-
mentation has to finish at that point in order to link up with the main question of the
debate (Q1). This is also what Copleston asserts as his intention at Turn 121. So, we
are now back into the main discussion, which is about Q1.

Segment IV, Turn 128 (Russell) Argumentation Stage (Subdiscussion and Main
Discussion)

[128.1] But the lawgiver has always been, it seems to me, one’s parents or
someone like <them> .
Russell ignores Copleston’s argument and in particular ignores Q90. Although there
are many questions and doubts about both Copleston’s argument in 127.6 as well as
about the question of using reason to criticize mores, Russell only insists, again, on
the facts of education and environment. These facts Copleston has already admitted,
but he has put them into a larger context than Russell had done. The most important
part of that larger context is precisely the fact that people often criticize the mores
taught them (whether by means of a special faculty of reason or in some other way).
It is particularly remarkable that Russell does not take up the question of criticism of
social mores, given that this is probably a constant motive of his ethical and political
writings, from the earliest to the latest text (Pigden 1999). He could also have built
a bridge to his utilitarianism, given that he always thought of it as what reason had
to offer to replace ordinary moral conscience.
Incidentally, the use of the adversative conjunction ‘but’ seems to indicate that
Russell considers the fact that parents, or someone like parents, are the lawgivers is
stronger than any of the reasons offered by Copleston in his arguments.
7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument 255

[128.2] There are plenty of terrestrial lawgivers to account for it, and that would
explain why people’s consciences are so amazingly different in different times
and places.
The reason that Russell now adduces is not explicitly directed to anything that
Copleston says. His argument seems to be as follows:
Premise 1. There are different mores among human groups.
Premise 2. There are plenty of terrestrial lawgivers.
Conclusion 1. The fact in Premise 2 explain best the fact in Premise 1.
Conclusion 2. There is no need for a celestial lawgiver.

This argument is an explanation of the fact of the diversity of mores by the diversity
of lawgivers. Neither fact in this argument has been questioned by Copleston. This
argument down to Conclusion 1 shows one discussant, Copleston, trying to explain
one fact (namely, that people are able to criticize the particular mores of their group)
and another discussant, Russell, trying to explain a different fact (namely, that mores
are different). If we are not to assume that this turn is completely irrelevant to what
Copleston says, we need to add Conclusion 2, which is just hinted at by the use
of the adjective ‘terrestrial’. By that word, Russell indicates he has understood that
Copleston wants to go back to Q1, and he produces a counterargument.

Segment IV, Turn 129 (Copleston) Argumentation Stage (Main Discussion)

[129.1] It helps to explain differences in the perception of particular moral

values, which otherwise are inexplicable. It will help to explain changes in the
matter of the moral law, in the content of the precepts as accepted by this or
that nation, or this or that individual.
Copleston fully and explicitly accepts Russell’s argument, as reconstructed in my
comment to 128.2, but only down to Conclusion 1, which is, I remind the reader,
what Russell actually said. As to what he meant (Conclusion 2), this does not escape
Copleston.
[129.2] But the form of it, what Kant calls the categorical imperative, the Ought, I
really don’t see how that can possibly be conveyed to anybody by nurse or parent
because there aren’t any possible terms, so far as I can see, with which it can be
explained.
By the conjunction ‘but’, Copleston now indicates that Russell cannot go on to
Conclusion 2 without taking into account the form of the moral law, as opposed to its
variable content. He briefly re-iterates his argument at 125.2. We have an argument,
a meta-argument, like this:
Premise. A variety of terrestrial lawgivers can best explain the diversity in the content of
the mores of a group, but it is no proper explanation of the form of the moral law.
Conclusion. Russell’s full argument at 128.2 does not defeat Copleston’s argument at 127.6.
256 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

It is extremely interesting that Copleston, who has said that he is not a Kantian (at
59.5), nonetheless here exhibits a partial Kantianism. If we remember that the moral
argument for God’s existence is a crucial part of Kant’s philosophy, this becomes
easy to understand (for more on the history of the moral argument and the role played
by Kant in it, see Baggett and Walls 2019).
[129.3] It can’t be defined in other terms than itself, because, once you’ve defined
it in other terms than itself, you’ve explained it away. It’s no longer a moral
Ought. It’s something else.
Copleston offers, again, his reason for holding that whatever the mores designed by
terrestrial lawgivers are, this cannot account for the form of the moral law. So, the
full meta-argument put forward by Copleston seems to be as follows:
Premise. The moral law (its obligatory character, its form) cannot be explained in other
terms than itself (e.g. in terms of the content of whatever mores happen to be taught in a
group).
Conclusion 1. A variety of terrestrial lawgivers can best explain the diversity in the content
of the mores of a group, but it is no proper explanation of the form of the moral law.
Conclusion 2. Russell’s full argument at 128.2 does not defeat Copleston’s argument at
127.6.

This counterargument prohibits any argument from mores to moral obligation and
is thus designed to undercut any argumentative strategy of the sort Russell keeps
deploying against Copleston.

Segment IV, Turn 130 (Russell) Argumentation Stage (Main Discussion)

[130.1] Well, I think the sense of ‘ought’ is the effect of somebody’s imag-
ined disapproval, it may be God’s imagined disapproval, but it’s somebody’s
imagined disapproval. And I think that is what is meant by ‘ought’.
Russell counters that ‘ought’ means ‘imagined to be disapproved by someone’, never
mind who does the disapproving. This ‘someone’ could be one’s parents, one’s
nurses, one’s teachers, the authorities, the priests, or even an imaginary, supernat-
ural character such as God. Russell is proposing a meta-argument which would
answer Q87, if in a more generalized form (say, whether the fact that one
has a feeling about ‘ought’ can best be explained by someone’s
disapproval of certain things in certain circumstances):
Premise. The moral law (its obligatory character, its form) can perfectly well be explained
in other terms than itself, namely in terms of a sense of disapproval by someone real or
imaginary.
Conclusion 1. A variety of terrestrial lawgivers can best explain the form of the moral law.
Conclusion 2. Russell’s full argument at 128.2 does defeat Copleston’s argument at 127.6
after all.
7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument 257

The premise of Russell’s argument implies either that there is such a thing as moral
obligation or the form of the moral law, but it can be explained in the way indicated,
or else that there is not such a thing. The distinction proposed by Copleston between
an idea and a feeling is thereby rejected, albeit without making it explicit in any way.
The positions are hardening again. The possibility of meeting each other halfway
is almost gone.

Segment IV, Turn 131 (Copleston) Argumentation Stage (Counterargumentation)

[131.1] It seems to me to be external customs and taboos and things of that sort
which can most easily be explained simply through environment and education;
but all that seems to me to belong to what I call the matter of the law, the content.
The idea of the Ought, as such, can never be conveyed to a man by the tribal
chief or by anybody else, because there are no other terms in which it could be
conveyed. It seems to me entirely…
Copleston makes a start to restate his position. He may well be on his way to produce
a further reason for it, or perhaps just a plausible story of how the idea of the Ought
is born in human consciousness, when Russell interrupts him.

Segment IV, Turn 132 (Russell) Argumentation Stage (Main Discussion)

[132.1] But I don’t see any reason to say that—I mean we all know about condi-
tioned reflexes. We know that an animal, if punished habitually for a certain sort
of act, after a time will refrain. I don’t think the animal refrains <by> arguing
within himself Master will be angry if I do this. He has a feeling that that’s not
the thing to do. That’s what we can do with ourselves and nothing more.
Whatever Copleston’s story was about to be told, Russell offers an alternative one. In
Russell’s story, people are animals who behave according to the operant conditioning
developed by behaviourists a decade before the broadcast (Skinner 1938).
Copleston’s story is part of an argument that I have already reconstructed, as far
as it has been expressed, in my comment on 127.6. Russell’s story is partly new, so
it would be appropriate to give it its place in another reconstruction:
Premise 1. An animal, if punished habitually for a certain sort of act, after a time will refrain.
Premise 2. The animal does not refrain by arguing within himself Master will be angry if I
do this—it just has a feeling that that’s not the thing to do.
Premise 3. Humans are exactly like animals.
Conclusion 1. What humans do when they punish animals is the same as what they can do
with themselves and nothing more.
Conclusion 2. What animals do when they are punished by humans is the same as when
humans are punished by humans.
Conclusion 3. Moral obligation is a feeling produced by a process of operant conditioning
carried out by some people on other people.
Conclusion 4. There is no need to invoke any extraordinary agency to explain this fact.
258 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

It seems that we are at a final impasse or close to it. Copleston tells the story of human
individuals using their faculty of reason in order to examine and modify the mores of
their respective groups, perhaps even converging towards the same or a similar moral
content. Russell tells the story of a peculiar race of animals who not only train other
animals (like dogs or horses) but also train each other. Russell’s feeling that one has
about ‘ought’ is thereby explained according to the particular form of behaviourism
that was quite fashionable at the time. (The fact that animals do not seem to engage
in some form of utilitarian reasoning, as Russell claims he himself does, is ignored.)
As he has done before, Russell only makes explicit one conclusion (the first in
the argumentative chain above); but all the others are needed to make the argument
relevant to the current discussion. They are implied by Russell and understood by
Copleston.
The heavy question raised by Russell’s line of argumentation is: whether what
we call moral obligation is merely an effect of general operant
conditioning of humans on humans. Let’s call this Q91. Russell says Yes,
Copleston will say No. Note by the way that Russell is again talking about feelings
and nothing more (“[the animal] has a feeling that that’s not the thing to do”). The
expression “the thing to do” is hardly the form of a moral obligation or a categorical
imperative. This is also ignored by Russell but immediately taken up by Copleston
in the next turn.

Segment IV, Turn 133 (Copleston) Argumentation Stage (Main Discussion)

[133.1] I see no reason to suppose that an animal has a consciousness of moral

obligation; and we certainly don’t regard an animal as morally responsible for
his acts of disobedience. But a man has a consciousness of obligation and of
moral values.
Copleston accepts Premises 1 and 2 of the argument offered at 132.1, but utterly
rejects Russell’s analogy between humans and animals (Premise 3). Copleston asserts
instead that consciousness of obligation, consciousness of moral values, and moral
responsibility are not predicates we use of animals but only of humans. By this
assertion, Copleston is swiftly rejecting Russell’s story of moral obligation as operant
conditioning by selective punishment which constitutes the chain of conclusions of
the argument at 132.1.
[133.2] I see no reason to suppose that one could condition all men as one can
condition an animal, and I don’t suppose you’d really want to do so even if one
could.
Copleston goes further than rejecting Russell’s argument and now explicitly rejects
Russell’s story. It is the very possibility of conditioning a human being in the way we
condition animals that is denied. Moreover, even if, per impossibile, all-round moral
operant conditioning of humans was feasible, Copleston declares it not desirable.
So, we have the following meta-argument:
7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument 259

Premise 1. Animals do not have consciousness nor moral obligation.

Premise 2. Animals are not responsible for their acts of disobedience.
Premise 3. People do have a consciousness of obligation and moral values.
Conclusion 1. Humans are not like animals with regard to moral matter.
Premise 4. General operant conditioning of humans is not possible.
Premise 5. Even if general operant conditioning of humans were possible, it would not be
desirable.
Conclusion 2. Moral obligation cannot, and should not, be produced by a process of operant
conditioning carried out by some people on other people.
Premise 6. Moral obligation is a fact.
Conclusion 3. We need an explanation for that fact that is utterly different from operant
conditioning.

Copleston’s use of the second person (“I don’t suppose you’d really want to [condition
all men as one can condition an animal] even if one could”) is interesting, for it implies
that even Russell should and would not accept his sort of explanation. For Russell
does not only contend that his sketched conditioning process is a fact, but he seems to
see the fact with great equanimity, and not with disgust. This is extremely important,
because it goes to show that we humans can contemplate the possibility of global
operant conditioning of moral obligation with disgust (as Copleston does here, and
as he claims Russell would if he thought more closely about it), whereas it is pretty
clear that animals, who are the subjects of operant conditioning, are not capable of
doing so. However, Copleston does not press this argument.
It may be of some interest to recall that, just a few years after the broadcast,
it would be experimentally proved that operant conditioning cannot, as a matter
of fact, produce, in animals, any desired effect (Garcia, Kimeldorf and Koelling
1955). Skinner’s proposal to realize what the founder of behaviourism, John Watson,
had dreamed of, namely that humans could be conditioned to do anything, beyond
freedom and dignity, was scientifically revealed to be just one more scientistic fantasy.
[133.3] If ‘behaviourism’ were true, there would be no objective moral
distinction between the emperor Nero and St. Francis of Assisi.
Copleston makes clear that he understands that the doctrine of behaviourism is at the
basis of Russell’s story. He adds a further reason against it as part of the following
argument:
Premise 1. If behaviourism were true, there would be no objective moral distinction between
Nero and Francis of Assisi.
Premise 2. There is an objective moral distinction between Nero and Francisco of Assisi.
Conclusion 1. Behaviourism is not true.
Conclusion 2. Moral obligation is not produced by a process of operant conditioning carried
out by some people on other people.
Premise 3. Moral obligation is a fact.
Conclusion 3. We need an explanation for that fact that is utterly different from operant
conditioning.
260 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

This is a clear rejection of Russell’s line of argumentation by means of an argument

that is presented as independent of the argument at 133.2.
[133.4] I can’t help feeling, Lord Russell, you know, that you regard the conduct
of the Commandant of Belsen as morally reprehensible, and that you yourself
would never under any circumstances act in that way, even if you thought, or
had reason to think, that possibly the balance of the happiness of the human
race might be increased through some people being treated in that abominable
manner.
Copleston has not forgotten another component of Russell’s worldview, namely util-
itarianism. So he comes back to it. Now, Russell has consistently argued from a
scientific, if not indeed scientistic, point of view. He has proposed a just-so story of
how what Copleston calls moral obligation is a feeling produced by a conditioning
process. It is a ‘view from nowhere’. In this last part of his intervention, Cople-
ston uses a tactic which is characteristic of a humanistic (not scientific, or at least
non-scientistic) perspective, namely to ask Russell to adopt a view from somewhere.
Copleston asks Russell point-blank what he thinks about Kramer’s behaviour in
the concentration camps (a topic that suddenly re-appears) and even how he would
have acted if in Kramer’s position and with Kramer’s thoughts about the effects of
Nazi policies. This is as opposed to Russell’s ‘view from nowhere’ as possible. The
suggested argument seems to go like this:
Premise 1. Russell regards Kramer’s behaviour as morally reprehensible.
Premise 2. Russell would never have acted the way Kramer did, even if thinking like a Nazi
about the balance of human happiness.
Conclusion. Russell does not really believe what he says about moral matters.

Note three interesting features of this argument:

First of all, as mentioned before, the argument is personal. This is not about how
humans in general acquire their feelings or obligations, let alone how animals are
trained, or whether certain actions are likely to produce certain effects on the general
population. It is about Russell himself, as he stands now before Copleston, a human
being before another.
Secondly, in Premise 1, Copleston carefully avoids talking about Russell’s feel-
ings, what Russell’s parents or nurses told him, how he was conditioned to feel
and do certain things, briefly all the elements of Russell’s story and argumentation.
Copleston’s phrasing implies that he is only concerned with Russell’s judgments as
they concern (objective) moral reprehensibility.
Thirdly, in Premise 2, Copleston includes Russell’s utilitarian ideas about balance
of effects. This implies that, no matter how utilitarian Russell might be, he would
not draw the consequences the Nazis did.
Copleston’s argument is thoroughly Socratic. It is an intensely moral and personal
ad hominem argument in the purest sense of that word: Copleston is using Russell’s
own premises in combination with what he knows about the personal morality and
7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument 261

actions of this man. And, as so often in Plato’s Socratic dialogues he is inviting his
interlocutor to reflect upon the incoherence between his talk and his real life.
It is true that the external form in which Copleston speaks is not the interrogative
sentence. But he is really urging several questions for Russell to answer, namely,
whether russell considers kramer’s conduct morally reprehensible
(Q92), whether russell would ever act like kramer (Q93), and, most
specifically, whether russell would act like kramer if he would ever
come to adopt a utilitarian view in favour of nazi policies (Q94).
All these questions seem to imply a larger, philosophical one: whether Russell’s
position—this mixture of emotivism, utilitarianism, and behaviourism with some
undertones of moral relativism—is at all coherent and personally sustainable. This
implicit question we may call Q95.

Segment IV, Turn 134 (Russell) Argumentation Stage

[134.1] No. I wouldn’t imitate the conduct of a mad dog. The fact that I wouldn’t
do it doesn’t really bear on this question we’re discussing.
When humanistic thinkers urge the sort of personal question we have seen at 133.4,
the typical reaction of scientistic thinkers is quite predictable, and we see it here. The
implied argument is as follows:
Premise 1. It is true that, personally, Russell would never act like a Nazi.
Premise 2. Russell and Copleston are discussing an impersonal question.
Conclusion. Copleston’s argument is irrelevant.

Observe, however, that Russell only actually responds to Q93 (see comment on
133.4). He avoids admitting that he would regard Kramer’s conduct as morally
reprehensible (Q92), which would involve using Copleston’s vocabulary; and he also
avoids saying anything about his conduct under a form of Nazi utilitarianism (Q94),
which would involve applying his own confessed view under a compromising circum-
stance. Naturally enough, he expresses no recognition of the underlying, implicit
general question (Q95).
Russell’s predictable response is a familiar sign that we are fast approaching a
dead end. Two worldviews with their vocabularies and narratives are facing each
other without any attempt at resolving their disagreement, more intent in making
clear what their stance is in a perennial struggle (Leal 2019) rather than trying to find
out some common ground and some degree of coalescence (Gilbert 1997).
Segment IV, Turn 135 (Copleston) Argumentation Stage (Argumentation ad hominem
in the classical sense)

[135.1] No, but if you were making a utilitarian explanation of right and wrong
in terms of consequences, it might be held, and I suppose some of the Nazis of
262 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

the better type would have held, that although it’s lamentable to have to act in
this way, yet the balance in the long run leads to greater happiness.
In this turn, Copleston, seeing that no further progress in the debate is feasible,
proceeds to restate his moral argument.
First, he freely admits that no purely personal attitude can be taken to answer the
impersonal questions under debate:
• Q1, whether god exists or not (1.1);
• Q65, whether god is the origin and ground of all things good
(92.1);
• Q66, whether human feeling, as it emerges from the effects of
actions, is the origin and ground of all things good (112.4; cf. the
earlier, incomplete version at 96.3); and
• Q91, whether what we call moral obligation is merely an effect
of general operant conditioning of humans on humans (133.2)
Copleston assumes the role of protagonist for Q1 and Russell that of antagonist;
but each discussant plays both roles in relation to the other questions. In fact, Q65 and
Q66 can be seen as incompatible answers. As for Q91, it is an addition or complement
to Q66.
In spite of admitting the limits of personal stance in regard to impersonal questions,
Copleston makes a last attempt at restating Q94 by adding a very important statement
of fact: At least some Nazis reasoned in a utilitarian fashion and took the policies
of killing and mayhem as a means to the benefit of humankind. Obviously, this
statement divides the Nazis into two groups: the evil Nazis and the utilitarian Nazis.
That Russell rejects the evil Nazis is clear, but Copleston wants Russell publicly to
reject the utilitarian Nazis as well.
For the better understanding of this move, we should keep in mind Copleston’s
dialectical training. He cannot carry out an ad rem argument by showing that the
best explanation for the fact of moral consciousness is God’s existence, for Russell
has a wholly different story with different facts and different causal chains. As the
only dialectical alternative, therefore, Copleston will try an ad hominem argument
by extracting, in classical Socratic fashion, certain admissions that lead to the same
result.
I do not think that Copleston is guilty of the so-called ad Hitlerum fallacy (Strauss
1953, pp. 42–43). Consider the following argument:
Premise 1. Russell is a moral utilitarian.
Premise 2. The better Nazis (the not purely evil ones) were moral utilitarians.
Conclusion. Russell is wrong in upholding utilitarianism.

If Copleston would have argued like this, then we would be justified in saying that
the conclusion does not follow from the premises. But this is not his argument. He
mentions the fact of Nazi utilitarianism only to establish that their actions could
hardly be condemned by using the utilitarian method. In this way, the relevance of
7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument 263

Q94 is made plausible and the path to Copleston’s real ad hominem argument to
follow is cleared.
[135.2] I don’t think you’d say that, would you? I think you’d say that sort of
action is wrong—and in itself, quite apart from whether the general balance of
happiness is increased or not.
Copleston now oversteps his own admission and again asks Russell a direct personal
question, which skilfully fuses Q92 and Q93: whether russell would say that acting
like the nazis is wrong in itself, independently of any utilitarian consideration. Let’s
call this Q92 + 93.
[135.3] Then, if you’re prepared to say that, then I think you must have some
criterion of right and wrong, that is outside of feeling, at any rate.
Assuming an affirmative answer to Q92 + 93, Copleston now deploys again a quasi-
transcendental argument to the effect that, in order to condemn Nazis actions in a non-
utilitarian way, Russell must have a criterion of right and wrong that is independent
of feeling. Copleston’s mention of feeling would prove to be a wrong move, as we
shall see. It would have been better, dialectically, to stick to utilitarianism at this
point.
[135.4] To me, that admission would ultimately result in the admission of an
ultimate ground of value in God.
In a very condensed manner, Copleston says that such a criterion will argumentatively
lead to God’s existence. The full ad hominem argument looks then like this:
Premise. Russell would say that acting like the Nazis is wrong in itself, independently of
any utilitarian consideration.
Conclusion 1. Russell has to have some criterion of right and wrong, that is, outside of
feeling.
Conclusion 2. Russell would ultimately have to admit that God exists.

Obviously, there is a huge leap from Conclusion 1 to Conclusion 2, but the important
part of Copleston’s argument is the step from Premise to Conclusion 1.

Segment IV, Turn 136 (Russell) Argumentation Stage (Counterargumentation)

[136.1] I think we are perhaps getting into confusion. It is not direct feeling
about the act by which I should judge, but rather a feeling as to the effects.
Russell starts by restating the argument presented at Turn 112. It is true that Cople-
ston did not really respond to it, which is probably why he was not careful in his
formulation of Conclusion 1 at 135.4. Nonetheless, Russell is perfectly aware that
this is not the point.
[136.2] And I can’t admit any circumstances in which certain kinds of behaviour,
such as you have been discussing, would do good. I can’t imagine circumstances
264 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

in which they would have a beneficial effect. I think the persons who think they
do are deceiving themselves.
Russell now rejects not so much the fact that some Nazis justified their actions by
some form of utilitarian reasoning. But he defends the idea that they were wrong in
so doing. The argument can be reconstructed as follows:
Premise. Behaviour such as that of Kramer cannot do good under any circumstance.
Conclusion 1. The Nazis who justified their actions by utilitarian reasoning made a mistake.
Conclusion 2. There is no need to renounce utilitarian reasoning on account of the faulty
reasoning of the Nazis.

Although an argument to defend the premise of this argument is needed, Russell’s

defence is appropriate. Curiously enough, it could serve as a rejection of the ad
Hitlerum argument which, although not Copleston’s, can be imagined to have come
to the minds of some people listening to the broadcast.
Note, finally, that Conclusion 2 (implicit in Russell’s intervention) is a kind of
response to Q95: all things considered, Russell considers his ethical theory consistent
enough, so that there is no need to give it up. On the other hand, Russell could have
referred to Copleston’s statement, at 123.2, that Different individuals and different
groups should enjoy varying degrees of insight into those values. This is another
missed opportunity to build on a shared proposition.
[136.3] But if there were circumstances in which they would have a beneficial
effect, then I might be obliged, however reluctantly, to say Well, I don’t like these
things, but I will acquiesce in them, just as I acquiesce in the Criminal Law,
although I profoundly dislike punishment.
The last turn is quite surprising, for Russell unexpectedly bites the bullet after all
and admits that he would accept the actions of the Nazis if circumstances demanded
it. His explicit argument is based on an analogy:
Premise 1. Russell acquiesces in the Criminal Law, although he profoundly dislikes
punishement.
Premise 2. The Criminal Law is comparable to Nazi policies if there were circumstances in
which they would have a beneficial effect.
Conclusion 1. If there were circumstances in which actions such as Kramer’s at Bergen-
Belsen would have a beneficial effect, then Russell might be obliged reluctantly to acquiesce
in them.
Conclusion 2. There is no need to renounce utilitarian reasoning on account of the Nazis.

Note that both Premise 1 and Conclusion 1 reinforce the standpoint and argument
presented by Russell at Turn 112. His ethical position does not primarily concern
feelings about actions, but rather the effects of those actions and then the feelings
awoken by those effects. Note also that the word ‘obliged’, in “I might be obliged”,
has nothing to do with moral obligation, but rather with logical consistency. As I said
7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument 265

before, Russell is a utilitarian all the way through, including its most unpalatable
consequences.6
In any case, Russell’s argument precludes discussion of Copleston’s position in
such a striking way that it becomes obvious that Russell does not think that a resolu-
tion of disagreement is possible—perhaps even that he does not want to resolve this
particular disagreement.

References

Anscombe, G. E. M. (1958). Modern moral philosophy. Philosophy 33 (124), 1-19.

Baggett, D., and Walls, J. L. (2019). The moral argument: A history. New York: Oxford University
Press.
Brentano, F. (1889). Vom Ursprung sittlicher Erkenntnis. Leipzig: Duncker & Humblot. [English
translation: The origin of the knowledge of right and wrong. Westminster: Archibald Constable
& Co., 1902].
Churchland, P. S. (2011). Braintrust: What neuroscience tells us about morality. Princeton: Princeton
University Press.
Churchland, P. S. (2019). Conscience: The origins of moral intuition. New York: Norton.
Copleston, F. C. (1993). Memoirs of a philosopher. Kansas City: Sheed & Ward.
Garcia, J., Kimeldorf, D. J., and Koelling, R. A. (1955). Conditioned aversion to saccharin resulting
from exposure to gamma radiation. Science 122 (3160), 157-158.
Gilbert, M. A. (1997). Coalescent argumentation. Mahwah; NJ: Lawrence Erlbaum.
Griffin, N. (1985). Russell’s multiple relation theory of judgment. Philosophical Studies 47 (2),
213-247.
Husserl, E. (1913). Ideen zu einer reinen Phänomenologie und phänomenologischen Philosophie.
Jahrbuch für Philosophie und phänomenologischen Forschung 1 (1), 1–323. [English translation:
Ideas pertaining to a pure phenomenology and to a phenomenological philosophy. The Hague:
Martinus Nijhoff, 1983.]
Landini, G. (1991). A new interpretation of Russell’s multiple relation theory of judgment. History
and Philosophy of Logic 12 (1), 37-69.
Leal, F. (2019). On philosophical argumentation: Towards a pragma-dialectical solution of a puzzle.
Journal of Argumentation in Context 8 (2), 173-194.
McGrath., A. (2004). The twilight of atheism: The rise and fall of disbelief in the modern world.
New York: Doubleday.
Nelson, L. (2016). A theory of philosophical fallacies. Cham: Springer.

6 Copleston’s recollection of the free discussion done in preparation of the actual debate (see Chap. 1,
Sect. 1.1) confirms Russell’s stubborness: “I suggested to Russell that he was really convinced that
the policy of genocide pursued by the Nazi government in regard to the Jews was wrong in an
absolute sense, and that even if, ex hypothesi, it could be shown that the human race would be
benefited at some future date by the extermination of the Jews at Auschwitz and elsewhere, he
would still condemn the extermination as wrong. Russell agreed of course that he felt in this way.
But he found some difficulty, he admitted, in squaring the implications of this admission with his
professed ethical theory. He even said something like this. ‘I find myself in a dilemma. On the one
hand I certainly want to condemn the Nazis’ behavior towards the Jews as wrong in itself. On the
other hand my ethical theory does not allow me to say this.’ When, however, it came to finalizing
the text for broadcasting, Russell remarked that he could not say in public that he was in a dilemma,
and he modified his admissions.” (Copleston 1993, pp. 136–137.).
266 7 Analysis of Segment IV: Discussion of Copleston’s Moral Argument

Phillips, R. (Ed.) (1949). War crimes trials, vol. II: The Belsen trial. London: William Hodge and
Company.
Pigden, C. (Ed.) (1999). Russell on ethics. London: Routledge.
Plantinga, A. (1993). Warrant and proper function. New York: Oxford University Press.
Scanlan, J. P. (2002). Dostoevsky the thinker. Ithaca: Cornell University Press.
Schneewind, J. B. (1993). Kant and natural law ethics. Ethics 104 (1), 53-74.
Shephard, B. (2005). After daybreak: The liberation of Bergen-Belsen, 1945. New York: Schocken.
Skinner, B. F. (1938). The behavior of organisms: An experimental analysis. New York: Appleton.
Strauss, L. (1953). Natural right and history. Chicago: The University of Chicago Press.
Wahl, R. (1986). Bertrand Russell’s theory of judgment. Synthese 68, 383-407.
Weikart, R. (2009). Hitler’s ethics: The Nazi pursuit of evolutionary progress. New York: Palgrave
Macmillan.
Chapter 8
Analysis of Segment V: Summing-Up
of the Arguments

Abstract Segment V of the debate contains very articulate summaries of the main
standpoints and arguments on both sides. Copleston restates his three arguments but
dwells especially on the metaphysical argument, which he considers the strongest. He
also objects to the intrusion of formal mathematical logic in a metaphysical issue.
Russell for his part rejects this objection and repeats the main arguments against
Copleston’s position.

Segment V, Turn 137 (Copleston) Concluding Stage

[137.1] Well, perhaps it’s time I summed up my position.

As is usual in philosophical discussions, the concluding stage is not the place where
the disagreements are resolved, for they never (or almost never) are. However, there
is a kind of closure in that the positions have become clear through the process of
argumentation. This is Copleston’s turn to summarize his various positions, those
of his opponent Russell, as he sees them, as well as the back and forth of argument
and counterargument that has taken place in the debate. A full summary would, of
course, have to include all the raised questions, separating them into three classes:
• Questions on which there was agreement from the start.
• Questions on which there was disagreement to begin with yet it was resolved in
due course.
• Questions on which there was and still is disagreement.
It is quite difficult to be sure how many questions are raised in a philosophical
debate which has at least the degree of complexity of this one; but I have at least
identified 95 of them so far (a yet finer analysis might easily discover at least a few
more). It may also be difficult to be sure which those questions are. In a masterful but
brief summary such as this, Copleston will only be able to highlight those questions
which he finds of special importance.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 267
F. Leal and H. Marraud, How Philosophers Argue, Argumentation Library 41,
https://doi.org/10.1007/978-3-030-85368-6_8
268 8 Analysis of Segment V: Summing-Up of the Arguments

[137.2] I’ve argued two things. First, that the existence of God can be philosophi-
cally proved by a metaphysical argument. Secondly, that it is only the existence of
God that will make sense of man’s moral experience and of religious experience.
Although Copleston offered a great deal of arguments, he separates those two he
considers to be the main ones. One is a philosophical proof of God’s existence by
means of a metaphysical argument. The other is not a proof, but a piece of reasoning
in which God’s existence is presented as the best explanation for the facts of general
moral experience and particular mystical experience.
Both these arguments purport to reach, as their conclusion, an affirmative answer
to the question that presides over the debate: Q1, whether god exists. In both
cases, God is shown to be an explanation of the facts; but the metaphysical argument
shows that no other explanation is possible for the fact that things exist, whereas the
other arguments conclude that, although other possible explanations for the fact of
moral and religious experience can be given, none is better than God. In Copleston’s
eyes, this makes the metaphysical argument a proof , in a quasi-mathematical way,
whereas the other arguments are mere inferences to the best explanation.
[137.3] Personally, I think that your way of accounting for man’s moral judg-
ments leads inevitably to a contradiction between what your theory demands
and your own spontaneous judgments. Moreover, your theory explains moral
obligation away, and explaining away is not explanation.
Copleston first compares his inference to the best explanation of moral experience
with Russell’s account. Note that he thereby accepts that at least two theories of moral
experience are possible, his and Russell’s. (There may be others, of course, but only
Russell’s is available for examination as an alternative to his own.) The arguments
supporting these two theories are of the same general nature, aimed at producing
a theory for the same facts. Copleston has two objections: (a) that Russell’s theory
contradicts “his own spontaneous judgments”, and (b) that it does not really explain
the facts but ‘explains them away’. These objections cannot and, therefore, do not
prove Russell’s theory wrong. Let us compare their relative scope and strength:
Objection (a) is ad hominem in the classical sense: Russell’s “spontaneous judg-
ments” contradict his theory. Objection (b) is methodological: Russell’s theory does
not really explain the facts, but rather it replaces them by other facts (feelings about
effects), which it then proceeds to explain (through a mechanism of reinforcement
within an operant conditioning schedule). This is the methodological meaning of
‘explaining away’.
Objection (a) leaves room for conceding that Russell’s theory may be a true
account of the facts; but Russell cannot uphold it. Objection (b) leads to the conclusion
that it is by no means a theory of moral experience in the first place.
Objection (a) denies that it could be Russell’s theory, objection (b) that it could
be Russell’s theory.
Some readers may think that the description just given of objection (b) contradicts
an earlier statement of mine, namely, that Copleston fails to prove that Russell’s
account is unable to explain the moral facts. This is a delicate point. Explaining away
8 Analysis of Segment V: Summing-Up of the Arguments 269

is a legitimate strategy in philosophical and scientific controversies. In science, for

instance, Lavoisier’s discovery of oxygen did explain away phlogiston; but, precisely
because of that, Lavoisier cannot be said to have a theory of phlogiston (Conant 1964).
In contemporary philosophy of mind, Dennett (1991) has been accused by critics of
explaining away consciousness, whereas he insists that his theory explains all the real
facts that his opponents classify under that label (see, in particular, McGinn 1995).
The case of Russell and Copleston is certainly not closed as the one of phlogiston
but remains as open as the one around Dennett and his explanation of consciousness.
Finally, please observe that Copleston does not here try to present any objections to
Russell’s account of religious or mystical experience. Copleston may have thought
that he does not have as strong a case here as he has for the moral facts; or he
may have thought that, ultimately, his account of the facts of mystical experience
naturally leads and reduces to the moral argument. In the debate, we have seen that
the religious argument indeed yields to the religious argument without solution of
continuity: what appears in the experience of the mystic seems to be, for Copleston,
a special manifestation of goodness, whereas ordinary moral experience seems to be
the same general kind of thing, but in a form accessible to everyone.
[137.4.0] As regards the metaphysical argument,
The rest of Copleston’s summary is dedicated to his metaphysical argument, which
for him really is the most important aspect of this debate. His summary divides into
eight parts, which are accordingly numbered in the following comments.
[137.4.1] we are apparently in agreement that what we call the world consists
simply of contingent beings, that is, of beings no one of which can account for
its own existence.
First, he states the main point of agreement, which is factual. This fact is nothing
less than the existence of things, of everything and anything at all, of the world
or universe. Remember that God is for Copleston not just the best explanation for
everything and anything that exists but the only explanation possible.
It is perhaps unfortunate that Copleston decides to use the term ‘contingent’ to
qualify all beings in the world. Of course, he needs it to put those beings in contrast
with God as the non-contingent but necessary being. However, the use of that term
will provoke Russell to make short thrift of Copleston’s argument by simply denying
meaning to the term (138.1)—and so, effectively destroying what could have been,
and what Copleston always thought it was, a point of agreement.1
[137.4.2] You say that the series of events needs no explanation: I say that, if there
were no necessary being, no being which must exist and cannot not exist, nothing

1This is, incidentally, the only way in which (pace Chalmers 2011) there can be a verbal dispute in
philosophy: if philosophers are unwittingly led by words to deny agreement or to declare disagree-
ment. For, as soon as it dawns on them that a different formulation would induce agreement, they
should institute an inquiry which will show them whether they do really agree after all or what
exactly the disagreement is about apart from mere terminological preferences.
270 8 Analysis of Segment V: Summing-Up of the Arguments

would exist. The infinity of the series of contingent beings, even if proved, would
be irrelevant.
In the second place, Copleston states the main point of disagreement. Russell rejects
the very question of what the explanation of the world is (Q44; note
that the world is at this point called ‘the series of events’ and not ‘an aggregate of
individual objects’, as in 11.2.1). Copleston asserts not only that the question makes
sense but also that God is the only possible answer.
Copleston insists, in the third place, on bringing up an old philosophical and theo-
logical controversy as to whether the infinity of the world precludes an explanation
of it by God (Q42; see comment on 43.2). Given that Russell never raised this ques-
tion, one reason to bring this up is that some listeners may be familiar with it and its
possible use as an objection against the metaphysical argument. So, Copleston wants
it out of the way.
[137.4.4] Something does exist; therefore, there must be something which
accounts for this fact, a being which is outside the series of contingent beings.
In the fourth place, Copleston recapitulates the metaphysical argument in what must
be its shortest version (see also 15.1). This version emphasizes that this argument is
proposing an explanation for a fact, like the moral and the religious arguments, only
in this case the explanation is portrayed as the only possible one. Note that the name
of God does not appear in the conclusion, but He is clearly described as “a being
which is outside the series of contingent beings”, i.e. a being outside the world, or a
transcendent being.
[137.4.5] If you had admitted this, we could then have discussed whether that
being is personal, good, and so on.
In the fifth place, Copleston says that, if agreement was reached that some such being
must exist to account for the fact of existence, then a further discussion might have
been entered as to the other attributes of such a being. We now can see that Copleston
all along wanted to prove the existence of what he defined, right at the beginning of
the debate, as ‘the creator of the world’, whereas the attribute of personality, which
was part of his original definition (1.3), was not really part of the proof as intended.
[137.4.6] On the actual point discussed, whether there is or is not a necessary
being, I find myself, I think, in agreement with the great majority of classical
philosophers.
After presenting a summary of the argument for his position, Copleston now turns to
this position itself, about which he insists that there is agreement among most classical
philosophers. Notice that, given the organization of his summary, this agreement is
not presented as part of the argument but as a mere description of the standpoint
itself. So, even if the said agreement could perhaps be interpreted as an argument
by some members of the broadcast audience, it would be too facile to reject it as
fallacious. Standpoints are not fallacious. By the way, Russell’s lack of response to
8 Analysis of Segment V: Summing-Up of the Arguments 271

this part of Copleston’s summary indicates that he did not understand it as a so-called
argument ad verecundiam. (Compare 52.1, 134.1.)
[137.4.7] You maintain, I think, that existing beings are simply there, and that I
have no justification for raising the question of the explanation of their existence.
But I would like to point out that this position cannot be substantiated by logical
analysis; it expresses a philosophy which itself stands in need of proof. I think we
have reached an impasse because our ideas of philosophy are radically different;
it seems to me that, what I call a part of philosophy, that you call the whole,
insofar at least as philosophy is rational.
At 137.4.2, Copleston had stated a point of disagreement with Russell, who said
that the world just is and the question of the existence of the world is not justified
(or is meaningless, as it was expressed at the time). Here, Copleston argues against
Russell’s position by means of a methodological argument (see also 55.2); and he
states that the use of logic in Russell’s mode of argumentation depends on a particular
philosophy. This is the question of logic as the method of choice in philosophical
inquiry (Q45; cf. Turn 19).
The point of Copleston’s rebuttal is well taken, for a feature of Russell’s way of
philosophizing is indeed the use of the mathematical logic to which he had contributed
so much and of which he was a pioneer. For Russell, the laws of logic are paramount,
ruling over everything else, including mathematics. Whether he would go as far as,
say, Rudolf Carnap, who declared philosophy to be ripe for replacement by the ‘logic
of science’ (defined as ‘logical syntax of the language of science’: Carnap 1934, pp.
III-IV), is much less clear; Russell will certainly deny this (138.3), but one would
have hoped for more.
The way Copleston ends this section deserves some comment. He says that what
he considers one part of philosophy is the whole of philosophy for Russell, and then
he adds “insofar at least as philosophy is rational”. This may sound a bit strange, for
what else could philosophy be? The easiest explanation of Copleston’s meaning has to
invoke philosophy’s traditional division into three parts: rational philosophy, natural
philosophy, and moral philosophy—the English version of philosophia rationalis,
naturalis, moralis, which in its turn is just the Latin translation of the Stoics’ logic
physics ethics (λoγική ϕυσική ºθική). With this in mind, it is easy to see that, for
Copleston, logic (some form of logic) is just one part of philosophy, namely rational
philosophy; but there are two other parts, namely natural philosophy and moral
philosophy, which cannot be reduced to rational philosophy. This is particularly
important in regard of both his metaphysical argument and his moral argument,
which would belong, respectively, to natural and to moral philosophy (see above
Chap. 5, footnote 7).
[137.4.8] It seems to me, if you will pardon my saying so, that, besides your
own logical system, what you call ‘modern’ in opposition to antiquated logic (a
tendentious adjective), you maintain a philosophy which cannot be substanti-
ated by logical analysis. After all, the problem of God’s existence is an existential
272 8 Analysis of Segment V: Summing-Up of the Arguments

problem, whereas logical analysis does not deal directly with problems of exis-
tence. So, it seems to me, to declare that the terms involved in one set of problems
are meaningless because they are not required in dealing with another set of
problems, is to settle from the beginning the nature and extent of philosophy,
and that is itself a philosophical act which stands in need of justification.
Having said that, for him, logic is just one part of philosophy, but not the whole,
Copleston now emphasizes an important point in his position, namely the plurality
of logical systems. Readers proficient in formal logic may object that the logic which
Copleston upheld—let us call it ‘traditional logic’—is not a proper logic, at least not
in the sense in which we speak today of formal logic. This is an interesting point,
and I myself have often wondered what Copleston was really talking about when he
spoke of logic (for some ideas on that see Veatch 1969). In several of my comments
on Segment II of the debate, I tried hard to understand what he meant, but I am
not sure whether I managed to do so, because my reconstructions often lead me to
identify large logical errors, and this may be too harsh.
On the other hand, Copleston seems to have been aware that, apart from traditional
logic, there is a plurality of formal logical systems (he says as much in 19.1). This
point was not very widely accepted at the time of the debate and remained suspect
for some decades afterward. However, it is nowadays defended by an increasing
number of philosophers (see Gabbay and Guenthner 2001–2018, van Benthem et al.
2006; Beall and Restall 2006). This plurality reinforces the argument that Copleston
presents. This argument may be reconstructed as follows:
Premise 1. Logical analysis does not deal directly with problems of existence.
Conclusion 1. Logical analysis cannot decide the problem of God’s existence.
Conclusion 2. It is not acceptable to declare that the terms involved in the problem of
God’s existence are meaningless because they are not required in dealing with logical
problems.
Premise 2. Whoever declares that the terms involved in the problem of God’s existence
are meaningless because they are not required in dealing with logical problems is thereby
settling from the beginning the nature and extent of philosophy.
Premise 3. Settling from the beginning the nature and extent of philosophy is itself a
philosophical act which stands in need of justification.
Premise 4. Russell has not justified that philosophical act.
Conclusion 3. Russell is making a huge methodological mistake.

Obviously, this long argument is condensed, but the charge against Russell is serious.
Copleston is saying that logic (whether mathematical or not, whether Russell’s or
not) is a magnificent instrument in the service of philosophy, but it cannot pretend to
rule everything in philosophy.
It is interesting to note that Russell, during the debate, never speaks of “antiquated
logic”, nor indeed of “modern logic”. As for the former label, we know, however,
from the history of how it came to the debate (see Chap. 1), that Russell had said that
“The proofs of the existence of God, so far as known to me, rest on an antiquated
logic or an illegitimate objectification of human ethics” (Slater 1997, p. 522; italics
added). The latter label, “modern logic” may or may not have been one of Russell’s
8 Analysis of Segment V: Summing-Up of the Arguments 273

phrases during the preparations of the debate, but it is important that the adjective
‘modern’ has been used as derogatory term in some intellectual circles.

Segment V, Turn 138 (Russell) Concluding Stage

[138.1] Well, I should like to say just a few words by way of summary on my
side.
Note that, naturally enough, Russell organizes his summary in the form of a rebuttal
of Copleston’s points, although not all of them, nor always in the particular form in
which they are summarized by Copleston.
Russell’s summary tackles three points: first, the metaphysical argument as such or
rather one aspect of it; then, the methodological accusation that Copleston adds to
his metaphysical argument; finally, the moral argument.
[138.2.1] First, as to the metaphysical argument: I don’t admit the connotations
of such a term as ‘contingent’ or the possibility of explanation in your sense.
Russell rejects two things. He rejects, first, the use of the term ‘contingent’. He denies,
secondly, that an explanation in Copleston’s sense of the word is at all possible. One
can see the ‘analytical turn’ at its sharpest. Where Copleston is trying to argue in
favour of the existence of a very special being, Russell is busy analysing the uses of
certain words. Later on, when we come to Copleston’s methodological accusation,
I shall discuss whether Russell, by using the analytical procedure exhibited here, is
guilty as charged.
[138.2.2] I think the word ‘contingent’ inevitably suggests the possibility of
something that wouldn’t have this, what you might call, accidental character
of just being there, and I don’t think <that> is true except in the purely causal
sense.
Russell repeats his objection to the use of ‘contingent’ (16.2–3). To complete the
argument, we might add that, if everything is contingent, then the concept does not
serve its function of making distinctions (see comment on 16.2). When Russell says
that a contingent thing is something that is accidentally there, he uses another word
from the philosophical tradition (‘accident’, of Aristotelian lineage), but he adds
‘what you might call’, thereby indicating that he does not like that word, either.
Russell does not go so far as to say that there is no point in calling things accidental,
because he is obviously prepared to distinguish between things, or rather events, that
are caused and random events (which have no cause; see 56.2).
Now, we know that Russell placed grave doubts on the very concept of causality
elsewhere (Russell 1912), but here he seems to accept it. However, a perusal of
Table 1 (in comment on 19.10) may show that it was at least possible for Russell
to admit that ‘contingent’, in the sense of ‘caused’, does have a meaning, because
he recognizes it, at least here. If so, there might be some ground for coalescence
(Gilbert 1997). Nonetheless, by now we are long past that point.
274 8 Analysis of Segment V: Summing-Up of the Arguments

[138.2.3] You can sometimes give a causal explanation of one thing as being the
effect of something else, but that is merely referring one thing to another thing
and there’s no—to my mind—explanation in your sense of anything at all; nor
is there any meaning in calling things “contingent” because there isn’t anything
else they could be.
According to Russell, it makes sense to speak of caused events and even of causal
chains; but it does not make sense to say that there is a cause of all events, so that there
would be a single causal chain—with God at the beginning, as Copleston wishes.
[138.3.1] That’s what I should say about that, but I should like to say a few
words about your accusation that I regard logic as all philosophy. That is by no
means the case. I don’t by any means regard logic as all philosophy.
Russell is right to counter that he does not ‘regard logic as all philosophy’ and never
has. We can safely assume that Copleston also knows that, but this is hardly the point.
Consider 138.2.3. When all is said and done, the only explicit reason Russell gives
to deprecate the use of ‘contingent’ and thus disarm Copleston’s argument is: There
isn’t anything else things could be. Call this R (as an abbreviation of ‘Russell’s
Reason’). If we take R with Copleston’s statement that existential questions do not
belong to logic (137.4.8), then a measure of agreement is given. There are two
alternatives:
(a) R is a proposition of logic.
(b) R is a philosophical proposition outside of logic.
If Russell means (a), then Copleston rightly objects that Russell’s philosophy is
unduly dominated by logic. If Russell means (b), then R is a proper metaphysical
statement that says something definite about the world, and then Copleston would
again be right in asking for a defence of that metaphysics (see again 137.4.8). In
view of all this, merely asserting without argument, as Russell does here, that one
does not ‘regard logic as all philosophy’ is not a proper reply. Clearing up whether
Russell means (a) or (b) would have been a step towards understanding each other
and facing the issues.
[138.3.2] I think logic is an essential part of philosophy, and logic has to be used
in philosophy, and in that I think you and I are at one.
This is a very good step, for it shows what Russell and Copleston have in common:
they both agree that logic is essential to philosophy and that it has to be used in
philosophy. However, the question is how to use logic in philosophy. This
question, call it Q96, is close but by no means identical to Q13. That question in its
turn arose in Turns 12–19, but it was never answered properly, perhaps because of
the difficulty of the subject in the context of a broadcast. Now, thanks to Copleston’s
summary, it takes a more general form. Yet, as we have just seen, Q96 is brushed
aside, as though it was identical with Q97, whether logic has to be used in
philosophy, about which the discussants are in agreement.
8 Analysis of Segment V: Summing-Up of the Arguments 275

It has to be added that the question whether logic is a part of philosophy

received from Russell different answers in different times (for a good discussion, see
Landini (2010, pp. 370–375).
[138.3.3] When the logic that you use was new, namely in the time of Aristotle,
there had to be a great deal of fuss made about it. Aristotle made a lot of fuss
about that logic. Nowadays it’s become old and respectable, and you don’t have
to make so much fuss about it. The logic that I believe in is comparatively new,
and therefore I have to imitate Aristotle in making a fuss about it, but it’s not
that I think it’s all philosophy by any means. I don’t think so.
Russell deviates even more from Q96 and explains, very wittily, it is true, that math-
ematical logic is comparatively new, so that one is justified in ‘making a fuss about
it’.2 By this remark, Russell responds to Copleston’s complaint about the opposition
of modern to antiquated logic, while gently mocking the latter. This is a good rhetor-
ical coup, but it hardly answers Copleston’s question. The trivial question whether
one is justified in making a fuss about new logical discoveries (call
it Q98) is utterly out of place here.
Besides, the two main differences between the Aristotelian-inspired traditional
logic that Copleston was trained in and the mathematical logic Russell pioneered
are hardly relevant to the metaphysical argument. Neither the existential import of
propositions nor the ‘logic of relatives’ as such is involved, so it is hard to understand
why Russell substitutes a serious question (Q97) by a trivial one (Q98).
[138.3.4] I think it’s an important part of philosophy and, when I say that I
don’t find a meaning for this or that word, that is a position of detail, based
upon what I’ve found out about that particular word, from thinking about
it. It’s not a general position that all words that are used in metaphysics are
nonsense, or anything like that, which I don’t really hold.
Again, Russell manages to deviate the attention from the point. It is true that Cople-
ston, at Turn 19, had unjustifiedly complained that Russell rejects all metaphysical
terms, a charge that Russell had rebutted at Turn 20. But that was then, and Cople-
ston is not issuing a general accusation now, but a particular one, concerning the

2 As a historical aside, it is quite doubtful that Aristotle ‘made a great fuss’ about logic. First of
all, we should remember that Aristotle’s work on arguments has two distinct parts: ‘dialectics’ and
‘analytic’ (or ‘apodeictics’). Only the second part, apodeictics, corresponds to what Russell calls
‘logic’, and about his discoveries in that field Aristotle most definitely did not make ‘a great fuss’.
In fact, his actual research practice clearly indicates that he did not think logic in that sense was
of much practical use (see Smith 1993), in marked contrast with Russell’s aspirations. As for the
first part, when, as a young man, he made his first discoveries in dialectics, in the early treatise On
sophistical refutations, it is true that Aristotle did make ‘a great fuss’ about them, rightly asserting
that, in contrast with rhetoric, where there had been many teachers before and the doctrine had even
improved over time, ‘nothing existed at all’ in the case of the subject of that treatise, so that he was
the pioneer and readers should be indulgent with its shortcomings (183b 15-184b 8). Nonetheless, it
seems that, by his later years, Aristotle had a much more modest opinion of the value of dialectics
for philosophical and scientific research (see Zingano 2017).
276 8 Analysis of Segment V: Summing-Up of the Arguments

legitimate and illegitimate uses of logic (mathematical or otherwise) in philosophy

when issues of existence are at stake.
[138.4.1] As regards the moral argument, I do find that, when one studies anthro-
pology or history, there are people who think it their duty to perform acts which
I think abominable,
The adjective ‘abominable’ is a loaded moral term which nonetheless can be blandly
understood as just expressing a feeling. Note also that Russell only commits himself to
this judgment by saying “which I think abominable”, thereby excluding any suspicion
that Russell is for a moment entertaining that values could be objective.
Although it is at least a moot point to what extent Russell had actually read the
anthropological literature of his time, given that he does not offer here any sources and
given the general scarcity of proper references in his many works, it is certainly true
that at different times people have considered “their duty to perform acts” which he,
together with Copleston and, I am sure, all the readers of this book find abominable.
One does not need to go farther back than the Nazi atrocities, quite fresh at the time of
the broadcast and a marked object of remembrance in the debate. But this is exactly
the point: Copleston agrees with him that these acts are abominable, in fact he says
they are objectively abominable.
The fact, which I have had occasion repeatedly to point out, that both Cople-
ston and Russell believed in the power of argument to criticize moral codes (either
of Russell’s imaginary cannibals or of Copleston’s exceedingly real Nazis) is not
discussed here. Even if the methods of argumentative criticism each of these two
philosophers favoured were very different indeed, the fact itself should have been
given pride of place during the summary. That it did not, is a clear dialectical failing
on both sides.
[138.4.2] and I certainly can’t, therefore, attribute Divine origin to the matter of
moral obligation, which you, Father Copleston don’t ask me to;
Fortunately, Russell remembers that Copleston has explicitly rejected the idea that
the content of moral obligation is of divine origin, so he catches himself in time and
admits the point.
[138.4.3] but I think even the form of moral obligation, when it takes the form
of enjoining you to eat your father or what not, doesn’t seem to me to be such a
very beautiful and noble thing; and, therefore, I cannot attribute a Divine origin
to this sense of moral obligation, which I think is quite easily accounted for in
quite other ways.
Unfortunately, Russell immediately rejects Copleston’s distinction and misses once
more the opportunity of discussing it. He adds an example of pop anthropology that
is unsourced and unanalysed but is bound to have a great rhetorical impact on the lay
public. Then, in a very cavalier fashion, he just asserts that such imaginary or real
mores can be quite easily accounted for, we can only assume by using a combination
of behaviourism, emotivism, and utilitarianism. That should be extremely good news
for all poor hard-working anthropologists.
References 277

References

Beall, J. C., and Restall, G. (2006). Logical pluralism. Oxford: Clarendon Press.
Benthem, J. van, Heinzmann, G., Rebuschi, M., and Visser, H. (Eds.) (2006). The age of alternative
logics: Assessing philosophy of logic and mathematics today. Dordrecht: Springer.
Carnap, R. (1934). Logische syntax der Sprache. Vienna: Springer. [English translation: Logical
syntax of language. London: Routledge, 1937.]
Chalmers, D. J. (2011). Verbal disputes. The Philosophical Review 120 (4), 515-566.
Conant, J. B. (Ed.) (1964). The overthrow of the phlogiston theory: The chemical revolution of
1775-1789. Cambridge, MA: Harvard University Press.
Dennett, D. C. (1991). Consciousness explained. Boston: Little, Brown, and Company.
Gabbay, D. M., and Guenthner, F. (Eds.) (2001–2018). Handbook of philosophical logic, 2nd edition,
18 volumes. Dordrecht: Springer.
Gilbert, M. A. (1997). Coalescent argumentation. Mahwah; NJ: Lawrence Erlbaum.
Landini, G. (2010). Russell. Milton Park, UK: Routledge.
McGinn, C. (1995). Consciousness evaded: Comments on Dennett. Philosophical Perspectives 9,
241-249.
Russell, B. (1912). On the notion of cause. Proceedings of the Aristotelian Society (N.S.) 13, 1–26.
Slater, J. G. (1997). The collected papers of Bertrand Russell, vol. 11: Last philosophical testament
1943–1968. London: Routledge.
Smith, R. (1993). What use is Aristotle’s Organon? Proceedings of the Boston Area Colloquium in
Ancient Philosophy 9, 261-285.
Veatch, H. B. (1969). Two logics: The conflict between classical and neo-analytic philosophy.
Evanston, IL: Northwestern University Press.
Zingano, M. (2017). Ways of proving in Aristotle. In W. Wians and R. Polansky (Eds.), Reading
Aristotle: Argument and exposition (pp. 7–49). Leiden: Brill.
 Part II
An E-Theoretical Analysis of the Debate,
by Hubert Marraud
Chapter 9
Argument Dialectic

Abstract My analysis of the Copleston-Russell debate can be described as both

logical and dialectical. Our understanding of the main perspectives on argumen-
tation derives largely from Wenzel’s three Ps principle: the rhetorical perspective
understands and evaluates arguing as a natural process of persuasive communica-
tion; the dialectical perspective understands and evaluates arguing as a procedure or
cooperative method for making critical decisions; and the logical perspective under-
stands and evaluates arguments as products that people create. My analysis is logical
in that it focuses on intra- and inter- argumentative relations, and at the same time it is
dialectical in that it places at the core the relations of opposition between arguments.

9.1 Logic or the Theory of Argument

In what follows I will attempt an analysis of the Copleston-Russell debate that could
be described either as logical or as dialectical. I will explain this hesitation before I
undertake the analysis.
Argumentation theory emerged from the integration of rhetorical, logical and
dialectical perspectives. Today, our understanding of the main perspectives on argu-
mentation derives largely from Wenzel’s three Ps principle: the rhetorical perspective
understands and evaluates arguing as a natural process of persuasive communica-
tion; the dialectical perspective understands and evaluates arguing as a procedure or
cooperative method for making critical decisions; and the logical perspective under-
stands and evaluates arguments as products that people create. The tripartition of
the theory of argumentation into logic, dialectic and rhetoric may be simplistic and
even unfair, since it leaves out some important perspectives on argument, such as
the socio-institutional one emphasized by Luis Vega Reñón (2013) and the linguistic
one that is dominant in the French-speaking area. However, that does not diminish
its usefulness as a starting point.
Wenzel’s characterization identifies logic with the theory of arguments: the part
of the theory of argumentation that studies the products of the practice of argu-
mentation. Within the logic or theory of arguments two major subdivisions can be
drawn: the analytics or theory of argument and the critique or theory of appraisal.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 281
F. Leal and H. Marraud, How Philosophers Argue, Argumentation Library 41,
https://doi.org/10.1007/978-3-030-85368-6_9
282 9 Argument Dialectic

The former deals with the nature, structure and typology of arguments, while the
latter is entrusted with the search for standards and criteria for the critique and eval-
uation of arguments. Ralph Johnson (2000, p. 39) mentions the following questions
as pertaining to analytics:
1. What is the nature and purpose of argument?
2. What are the different types of argument?
3. How are we justified in reconstructing argument so as to include elements not
explicitly asserted by the arguer?
4. How is argument to be understood or defined?
5. What is the relationship between argument and inference?
And these other questions as pertaining to critique (Op.cit., p. 40):
1. What are the standards for the appraisal of arguments?
2. How do the personalities and beliefs of the arguer and the audience affect the
merits of an argument?
3. Is the truth of the premise too strong a condition to demand soundness of
argument?
4. Is there an important difference between evaluating an argument and criticizing
it?
5. What constitutes logical virtue in an argument? What are the qualities that make
an argument a good argument?
6. What constitutes profitable criticism of an argument?
Analytics is, in a certain sense, subordinated to critique, because, as Wenzel points
out, logic (that is, the theory of argument).
… is a retrospective viewpoint which is activated when someone adopts a critical stance
and “lays out” an argument for inspection and evaluation. In such a case, a fourth version
of “the” argument is created. Such versions of arguments, reconstructed for purposes of
examination, become the subject matter for logical evaluation. So, the special concerns of
the logical perspective are with techniques for representing an argument in a form amenable
to criticism and with standards for evaluation (Wenzel 2006, p. 17).

Quite interestingly, this declaration entails that the arguments, which Wenzel defines
as products of argumentation, are rather products of criticism and evaluation. The
logical question by excellence, Wenzel goes on, is “Shall we accept this claim on the
basis of the reasons put forward in support of it?”. Thus, Wenzel closely links the
theory of arguments to the theory of reasons. His characterization of logic presup-
poses a definition of arguing like this: arguing is to present to someone something
as a reason for something else. This reasons-based view contrasts with an inferential
view of logic, according to which the central question in logic is “Does the conclusion
follow logically from the premises?”. The main point of contention is that reasons,
unlike logical inference, is a weighted and context-dependent concept. Reasons are
weighted because there can be good reasons for and against something, reasons that
must be compared with each other before reaching a conclusion. Contexts differ in
terms of whether a certain consideration constitutes a reason at all, as well as in terms
9.1 Logic or the Theory of Argument 283

of the weight of the reason. But insofar as logical inference is concerned, argument
evaluation is context-independent.
I will focus neither on the series of actions performed by Russell and Copleston
during the debate nor on the procedure followed by them in doing so, but on the rela-
tionships between the arguments advanced by them. When the relationships between
the arguments are taken into account, it is found that “Although arguments occur in
dialogue, often a dialogue can best be seen as one large argument” (Walton 1996,
p. 41). Thus, my analysis can be properly termed “logical”.

9.2 Arguer Dialectics Versus Argument Dialectics

Wenzel’s conception of dialectic is procedural, in that it focuses on the conditions

and conventional rules governing argumentative exchanges between a proponent
and an opponent who argue and counter-argue in turns: “Of all the meanings of the
term ‘dialectic’, the one I employ here takes dialectic to be a method, a system or a
procedure for regulating discussions among people” (2006, p. 14). This is also the
conception of dialectics in the pragma-dialectical model.
We believe this interaction [the discussion between a protagonist and an antagonist] to be an
essential feature of dialectical process of convincing. However, it will only be able to lead to
a resolution of the dispute at the centre of the discussion if the discussion itself is adequately
regimented. This means that in a dialectical theory of argumentation it will be necessary to
propose rules for the conduct of argumentative discussions. (van Eemeren and Grootendorst
1984, p. 17).

But dialectics can also be understood otherwise. Zeno of Elea, credited with the
invention of dialectics, presents dialectical reasoning as a monological method of
counterargumentation. Likewise, Plato’s dialectics is a method of inquiry, not a
procedure directed at the reasonable resolution of a difference of opinion. More
recently, Nicholas Rescher has advocated for the transition from rational controversy
to rational inquiry, from procedures of controversy and disputation to a cognitive
methodology of inquiry:
The object now is not to refute the contentions of an opposing spokesman, but to appraise
the rational credentials of a thesis. The process of reasoned exchange is reoriented from a
bilateral adversary procedure of controversy and disputation to the unilateral enterprise of
a “discussion” carried on with oneself, in foro interno, within a self-contained course of
reflective thought. (Rescher 1977, p. 46).

Dialectics thus becomes a method for assessing arguments by weighing them against
other arguments. In other words, arguer dialectics becomes argument dialectics.
I contend that, so far as argumentation is concerned, there are two main and
complementary conceptions of dialectics. If there is some concept permanently
bound to dialectic throughout its whole history, from Zeno to van Eemeren, through
Hegel and Marx, it is that of opposition (and resolution of an opposition). On the
one hand, dialectics can be conceived of as the art of controversy or debate, with the
284 9 Argument Dialectic

confrontation of opinions and hence of arguers. The focus of dialectic thus under-
stood is the conventional rules and procedures governing such confrontations. If the
rationale of these rules and procedures is that they allow for the reasonable resolu-
tion of a difference in opinion, dialectics will be bound to the quest of some kind
of consensus. This is what I call ‘arguer dialectics’. Pragma-dialectics is a good
illustration of arguer dialectics.
On the other hand, dialectics can also mean the study of the oppositions between
arguments, the theory of counterargument, as we might say. This conception of
dialectics rests on a contextual, comparative and non-qualitative conception of cogent
argument. Such comparative concepts have been historically linked to the notion
of argument strength, and therefore dialectic can also be defined as the study of
argument strength. Argument dialectics deals then with standards and criteria for
comparing and assessing the relative strength of arguments. Given that argument
dialectics deals with the relationships between arguments as products, it is logical
in character. Hegel’s dialectics is to some extent an example of argument dialectics
since the opposite sides in the dialectical process are definitions of concepts, not
persons playing the roles of proponent and opponent.
Plato’s dialectics is more difficult to classify. Although it is true that Plato uses
the term ‘dialectic’ to refer to a method of argument that involves some sort of
contradictory process between a group of participants, these are characters, not real
persons. On the one hand, it can be thought that characters are used to recreate real
persons, in which case Plato’s dialectics would be a dramatization of a conversational
method, and thus an example of arguer dialectics. On the other hand, the opposite
characters can be conceived as personifications of opposite lines of argument about
some controversial issue, in which case it would be better classified as argument
dialectics.
Something like the sketched distinction between arguer dialectics and argument
dialectics can be found in some dictionaries. The Merriam-Webster Dictionary, for
instance, defines dialectic (singular) as “discussion and reasoning by dialogue as
a method of intellectual investigation”, and dialectics (plural) as “any systematic
reasoning, exposition, or argument that juxtaposes opposed or contradictory ideas
and usually seeks to resolve their conflict”. Thus, according to these definitions,
“dialectic” refers to arguer dialectics and “dialectics” to argument dialectics.
I intend to analyze the relationships of comparison, reinforcement and opposition
among the arguments deployed in the debate between Copleston and Russell, as
presented by them. My analysis can therefore be conceptualized quite properly as an
exercise in the dialectic of arguments.

9.3 Dialectical Rules Versus Logical Norms

Wenzel himself describes the dialectical opposition sometimes as an opposition

between arguments (e.g., inherently dialectical situations “encourage the critical
testing of positions one against the other in the give and take of the debate” op.cit.,
9.3 Dialectical Rules Versus Logical Norms 285

p. 13) and sometimes as an opposition between arguers (“The notion of one speaker
serving as a check on another brings us to the dialectical perspective and its prac-
tical purposes”, op.cit., p. 14). To see the difference let us consider the notion of
debate by Ehninger & Brockriede, that Wenzel presents as a model of the dialec-
tical method. Debate is a method of critical discussion because it implements six
directives (Ehninger and Brockriede 2008, p. 15).
1. The competing views are subjected to a full and fair competition to assess their
relative worth.
2. This competition consists of two phases. First, each view is set forth in its
own right, together with the most convincing supporting proofs. Second, each
view is tested by seeing how well it withstands the strongest objections and
counterarguments an informed opponent raises against it.
3. Decision is delayed until both sides have been presented and subjected to testing
4. The decision is rendered by an external adjudicating agency.
5. The adjudicating agency weighs the competing arguments and produces a
decision critically.
6. The participants agree in advance to abide by such a decision.
These six rules allocate roles and obligations to the participants and regulate their
interventions in the debate. Despite their generality these are conventional rules, very
different from the empirical and non-conventional rhetorical precepts. They belong
therefore to what I have called ‘arguer dialectics’. However, Ehninger & Brockriede’s
directives don’t say anything on the method used by the external adjudicating agency
for weighing and assessing the competing arguments. This method belongs to the
realm of what I have called ‘argument dialectics’ and, unlike the six directives above,
they are not perceived as conventional by the participants.1
The distinction and imbrication of arguer dialectics and argument dialectics can
be also found in the pragma-dialectical code for critical discussion (van Eemeren
et al. 2002, pp. 182–183). The argument scheme rule and the validity rule refer to
standards and criteria for argument evaluation and they order the participants to apply
these standards and criteria, and thus to behave “logically”.
• Argument scheme rule: a standpoint may not be regarded as conclusively defended
if the defense does not take place by means of an appropriate argument scheme
that is correctly applied.

1 In the same vein, Rescher insists upon the contrast between the unconventional rules of rational
inquiry and the conventional rules of rational debate or controversy. “We must return to the difference
between the natural—the “purely rational”—mechanisms for assessing presumption and plausibility
and the merely conventional mechanisms, illustrated by certain essentially arbitrary and unreasoned
devices of law (e.g., statutorily determined presumptions) and of disputation (e.g., conventionally
canonical proof-tests), etc. All such merely artificial devices of probative procedure are abrogated
in inquiry, where “pursuit of knowledge” is itself the only relevant task. Here the evidential rules
of knowledge-oriented controversy apply without distorting constraints or restraints. When we
make the transition from controversy to inquiry, it is purely rational controversy, with its natural
(nonconventional) ground rules, that constitutes the paradigm.” (Rescher 1977, pp. 47–48).
286 9 Argument Dialectic

• Validity rule: the reasoning in the argumentation must be logically valid or must
be capable of being made valid by making explicit one or more unexpressed
premises.
Accordingly, these two rules could be unified as the single rule:
• Parties must assess argumentations advanced in the course of the critical
discussion according to current logical standards and criteria.
The precept of using logical standards and criteria is a rule of action, not a logical rule
(i.e., a function which takes premises and returns a conclusion). In fact, logical stan-
dards and criteria cannot be part of the pragma-dialectic decalogue, which is a set of
rules of action. While it is true that the pragma-dialectical rules for critical discussion
are justified both by their problem-solving effectiveness and by their intersubjective
acceptability (van Eemeren and Grootendorst 2004, p. 132), it is nonetheless true
that the pragma-dialectical rules for critical discussion present logical standards and
criteria as something “external” to the critical discussion itself, as something logic
prescribes to the participants.
The claim that weighing rules, and more generally the rules that govern the rela-
tionships among arguments, are logical rules collides with Wenzel’s idea of logic,
and hence with his way of understanding the relations between logic and dialectic.
According to Wenzel logic deals with the internal structure of arguments, with
the relation between premises and conclusion. This position is espoused by many
contemporary argumentation scholars, such as James Freeman:
The goal of a logical analysis and evaluation of an argument is to determine whether the
premises constitute good reasons for accepting the conclusion, good in the sense of trans-
ferring the acceptability of the premises, given that the premises are acceptable, to the
conclusion. The unit of analysis, then, is the premise-conclusion nexus. (Freeman 2011,
p. 109).

But the task of the adjudicating agency is to compare clusters of arguments with
each other in order to make a decision, and such a comparison is not so much a
matter of properties of arguments as a matter of relationships between arguments.
For Freeman, adjudication, as described by Ehninger & Brockriede is an exercise in
dialectic:
A dialectical analysis of argumentation, then, will focus on a different unit, a whole argu-
mentation, possibly containing multiple arguments, where a logical analysis will take an
individual argument as its unit of analysis. (Freeman 2011, p. 110)

When Wenzel grounds the three perspectives on argumentation in the triad process,
procedure and product, he says that the conceptions of argument as rhetorical process
and logical product are already established and what needs justification is the addi-
tion of dialectic as a different and definite perspective on argument. To that end
he “intends to show how Habermas’ treatment of epistemological problems in his
pursuit of a critical social theory may help us to understand the nature, significance
and promise, of a dialectical perspective in the study of argumentation” (Wenzel 1979,
p. 83). It is therefore not surprising to find the tripartition rhetoric-logic-dialectic in
9.3 Dialectical Rules Versus Logical Norms 287

Habermas’ The Theory of Communicative Action (1981). Habermas holds that argu-
mentation theory should respond to questions such as these: How can validity claims
that have become problematic be supported by good reasons? How can reasons be
criticized in turn? What makes some arguments, and thus some reasons, which are
related to validity claims in a relevant way, stronger or weaker than other arguments?
These questions concern three different and complementary aspects of argumentative
speech: process, procedure and product. Like Wenzel, Habermas proposes to differ-
entiate three perspectives on argument in correlation to these three analytic aspects.
The chosen perspective, rhetoric, dialectic or logic, will lead to the discovery of a
different structure of argumentation; respectively:
[…] the structures of an ideal speech situation, immunized against repression and inequality
in a special way; then the structures of a ritualized competition for the better arguments;
finally the structures that determines the construction of individual arguments and their
interrelations. (Habermas 1981, p. 26; emphasis added)

Hence Habermas, unlike Wenzel and Freeman, assigns to logic not only the study of
the internal structure of arguments but also the study of their interrelations, so that
argument dialectics becomes part of logic. The repeated use of the expression “the
force of the better argument” in The Theory of Communicative Action also shows
that Habermas has a comparative concept of the notion of cogent argument.
Snoeck Henkemans too, from different assumptions, has emphasized the necessity
of analyzing the interrelations between arguments in order to evaluate an argumen-
tation: “An overall judgment of the quality of a complex argument requires not just
a clear picture of individual arguments, but also insight into the relations among
these arguments” (2000, p. 447). By “complex argument” she means a combination
of single arguments with a common conclusion, and she specifies that a ‘single’ or
‘individual’ argument is the equivalent of a ‘reason’.

9.4 The Scope of Logical Rules

In their natural state reasons and arguments rarely stand alone. Analytically a dialogue
(i.e. an exchange of reasons) begins with a question “Why…” and continues with a
reason “Because…”. If the dialogue does not end there, with the acceptance of the
given reason, it can be continued with the presentation of a new reason, more or less
explicitly related to the former one. From this moment on, each additional reason is
presented as a reason that is chained, coincidental, opposite or analogous to some
previous reason. In this way the participants weave together an argumentative web
that constitutes the object of study of the dialectic of arguments. Even when some
participant tries to answer directly the initial question, her contribution is interpreted
by the other participants in connection with the reasons already adduced. I call
“macro-argument” this fabric of reasons.
Dialectical relevance refers to the web of reasons woven by the participants
during a dialogue. Dialectical relevance properly refers to how each movement in the
288 9 Argument Dialectic

dialogue must fit with the previous and following movements, where a movement
is a sequence of locutions advanced by a participant at a given point in the dialogue
sequence (Walton and Krabbe 1995, p. 71). But, for our purpose, we can restrict
ourselves to the multilinear composition of arguments delineated in that sequence
of moves. What I want to stress here is that whenever in a dialogue anyone offers
an argument for consideration, she does so by connecting it, more or less explicitly,
with prior arguments. As shown below, the answer to the logical question par excel-
lence: “Shall we accept this claim on the basis of the reasons put forward in support
of it?” depends both on the connection of the premises to the conclusion, and on
the connection of the argument itself with the concurring arguments. Thus, a proper
logical analysis of an argument must pay attention to both intra-argumentative and
inter-argumentative relations.
An example is worth a thousand words. Let us imagine that a theist argues for
miracles by appeal to witness testimony:

The central Christian miracles are attested by the

testimony of sincere and able eyewitnesses
Therefore
Miracles happen

Then the opponent attempts to refute this argument in a Humean fashion with a
second argument:

A miracle is a violation of the laws of nature. A law of nature is a

regularity no exception to which has previously been experienced
Therefore
Miracles cannot happen

In doing so, the opponent commits himself to the following statements:

(1) It is true that a miracle is a violation of the laws of nature.
(2) It is also true that a law of nature is a regularity no exception to which has
previously been experienced.
(3) The facts that a miracle is a violation of the laws of nature and a law of nature
is a regularity to which no exception has previously been experienced taken
together constitute a reason to believe that miracles cannot happen.
(4) The facts that a miracle is a violation of the laws of nature and a law of nature
is a regularity to which no exception has previously been experienced is a
stronger reason for Miracles cannot happen than the considerations adduced
by the theist are for Miracles can happen.
(1) and (2) are conceptual elucidations, (3) is a logical assumption concerning the
internal structure of the argument, and (4), I maintain, is also a logical assumption
about the relationship between the two competing arguments. To make it clear, (4)
involves a relationship between both arguments as objects, not between the actions,
9.4 The Scope of Logical Rules 289

positions or moves of both arguers. Were (4) rejected, the consideration that a miracle
is a violation of a regularity to which no exception has previously been experienced
would not provide a sufficient reason to accept the claim that miracles cannot occur.
Generally speaking, a condition for a conclusion to be credible is that the argument
supporting it is not overcome by a stronger argument for a contradictory conclusion.
Quite interestingly reconstructions of Hume’s argument against miracles are full
of such balancing principles as “The argument for a miracle, from testimony, is at
best a strong but somewhat weaker argument from experience” (cfr. McGrew 2019).
The logical character of statement (4) is confirmed by the non-conventional character
of reasons weighing. Proponent and opponent can conventionally agree on who will
speak first, who has the burden of proof, or how they are going to use the term
“Miracle”, but which argument is stronger is not the kind of question that can be
settled by agreement.

9.5 Logic, Dialectic and the Macrostructure of Argument

Sometimes the difference between the dialectical approach and the logical approach
is explained in terms of a contrast between the macrostructure (or macro-level) and
the microstructure (or microlevel) of arguments. Thus Wenzel, comparing logic with
dialectic, writes: “Logic is also concerned with decision-making, but on what we
might call a micro-level. The practical purpose of logic is to apply appropriate
criteria to judge the merits of particular arguments” (Wenzel 2006, p. 15). And
James Freeman, as we have seen, argues that where a logical analysis takes as its
unit of analysis the individual argument, a dialectical analysis of argumentation
focuses on the whole argumentation, possibly containing multiple arguments that
need not form one single unified argument developed over the course of the discus-
sion (Freeman 2011, p. 109). What I have called “intra-argumentative relations”
belong to the microstructure and what I have called “inter-argumentative relations”
belong to the macrostructure.
By the microstructure of an argument, we mean its logical form as studied in deductive or
inductive logic. Specifically in formal deductive logic, microstructural analysis reveals how
the constituent statements of an argument are built up from simple or atomic components
by means of truthfunctional connectives, quantifiers, and in some cases other operators such
as adverbial modifiers and modal or propositional attitude connectives. In inductive logic,
microstructural analysis may classify an argument as instancing inductive enumeration,
argument by analogy, or one of Mill’s methods. Microstructural analysis thus concerns the
internal structure of the constituent statements of an argument. By contrast, the macrostruc-
ture of an argument concerns how its component statements (together perhaps with other
elements) fit together as wholes to allegedly lend support to some claim or claims (Freeman
2011, p. 1).

Thus, on this account logic and dialectic appear as divisions of the theory of argu-
ment with different subject matters: intra-argumentative and inter-argumentative rela-
tionships, respectively. This account of the difference between logic and dialectic
290 9 Argument Dialectic

assumes that the merits of an individual argument can be judged in isolation from
the web of reasons and arguments, the macroarguent, woven in the course of the
dialogue in which it appears.2 To argue, we have said, is to present something as
a reason for something else. It seems therefore natural to define a logically good
argument as the one that presents a good reason for its conclusion. A “good reason”
can mean either a pro tanto reason or an all things considered reason. A pro tanto
reason is a consideration that counts in favor of some position, and thus has genuine
weight, but nonetheless may be outweighed by other considerations. A pro tanto
reason for something that is not outweighed is an all things considered reason. One
might say, according to the precedent view, that logic identifies pro tanto reasons in
argument while dialectic does so with all things considered reasons.
But even to determine whether an argument offers a pro tanto reason one has to
place it with respect to other arguments. Going back to the discussion on miracles,
let us imagine that the theist had countered in turn that the fact that a miracle is a
violation of a regularity to which no exception has previously been experienced is
not a pro tanto reason to believe that miracles cannot occur, since the step from the
premises to the conclusion overlooks the possibility that a regularity to which no
exception has previously been experienced be also a regularity of which no instance
has previously been experienced (or that has not been instanced very often). Now
these possibilities, the theist points out, would undercut the inference (Hájek 2008,
p. 91).
I call ‘rebuttal’ this kind of counterargument (vid. Chap. 10). The premises of
a rebutting counterargument point at some circumstance in which allegedly the
adduced consideration no longer counts a reason for the claim -i.e. is not a prima
facie reason for this claim.
Here too there would be an opposition between two arguments, although of a
different kind from the balance of reasons considered on 8.4. Here we are faced with
the argument.

A miracle is a violation of the laws of nature. A law of nature is a

regularity to which no exception has previously been experienced
Therefore
There is as compelling a proof from experience as can possibly be
imagined against a miracle

And the argument

2Given the close relation of arguments and reasons, this assumption is in flat contradiction with
holism. In the theory of reasons holism is the claim that contexts differ about whether a certain
consideration constitutes a reason at all, as well as on the weight of the reason.
9.5 Logic, Dialectic and the Macrostructure of Argument 291

It is possible that a regularity of which no exception has been previously experienced be

also a regularity of which no instance has been previously experienced.
Therefore
The fact that a miracle is a violation of a regularity to which no exception has previously
been experienced is not a por tanto reason to believe that miracles cannot occur

Weighing is the mechanism that brings us from a pro tanto reason to an all things
considered reason. But what is at stake here is whether there is a pro tanto reason at
all. If the counterargument is cogent, and some exceptional circumstance occurs, the
fact that physicists assure us that individual quantum transitions in atoms have no
cause is not, in this context, a pro tanto reason to believe that the series of events may
have no cause. Therefore, any comparison of the relative strengths of both arguments
would be out of order.
I have tried to show that in order to appraise the merits on this particular occa-
sion of the argument that a miracle is a violation of the laws of nature, one has
to appraise the merits of the related argument that a regularity to which no excep-
tion has previously been experienced is also a regularity of which no instance has
previously been experienced. Thus, the logical evaluation of any argument leads to
the logical evaluation of other related arguments. If I am right, then the idea that
logic deals with intra-argumentative relations while (argument) dialectics deals with
inter-argumentative relations has to be rejected.

9.6 Criticism and Evaluation

Perhaps the distinction between logic and dialectic could be redrawn resorting to the
distinction between argument evaluation and argument criticism, so that logic would
become the theory of the evaluation and dialectics the theory of criticism.
As we have seen, the theory of argument is usually divided into analytics or theory
of analysis and critique or theory of appraisal. Ralph Johnson proposes a further
distinction in the theory of appraisal into evaluation and criticism. On Johnson’s
view evaluation refers to both the process and the result of the process of assessing
an argument in terms of a set of criteria with the purpose of establishing the value of
the product. Throughout Manifest Rationality Johnson says several times that formal
logic, the paradigm of logicality, is a theory of evaluation (although not the one he
embraces) (Op.cit., pp. 119, 179, 223 and 367). Criticism, on the other hand, is the
articulated and reasoned evaluation of something communicated to the proponent
with the purpose of helping her to improve the product. Johnson points out that:
“Criticism goes beyond evaluation in that it must consider the strengths as well as the
weaknesses of the product and is intended for the one who produced the argument as
a vehicle whereby the argument may be improved. Thus, it may be said that criticism
is part of a dialectical process, whereas evaluation is not” (Johnson 2000, p. 219).
Hence in Manifest Rationalty argument evaluation and criticism are associated with
logic and dialectic, respectively.
292 9 Argument Dialectic

Johnson proceeds, then, to distinguish between the theory of argument and the
theory of criticism (Op.cit., p. 220). The aim of a theory of evaluation is to develop
appropriate criteria to evaluate arguments as good or bad, strong or weak. Thus,
the theory of evaluation answers the question “What qualities or properties make an
argument a good one?” (Op.cit., p. 180), the kind of question logic is intended to deal
with. By contrast the aim of a theory of criticism is to lay down the guidelines to carry
the process of criticism and to indicate how to integrate and weigh the criteria used
in the evaluation of arguments. These guidelines are intended to guide the behavior
of the participants in an argumentative exchange, and thus they fall into the realm
of arguer dialectics. Here “weighing” also refers to the actions performed by the
participants, not to the principles that justify the assignment of greater weight to
an argument than to another. This is made clear in the principle of discrimination,
the only principle of criticism that has to do with weighing: when criticizing an
argument, the critic must (1) assess both the strengths and the weaknesses of the
argument, (2) focus on the most important problems, and (3) structure the critique as
to give greatest emphasis to those criticisms which are major (Johnson 2000, p. 241).
These, too, are rules for arguer dialectics, not for argument dialectics.
Lilian Bermejo Luque takes up the distinction between evaluation and criticism
and explains it in terms of models:
[There is a distinction] between models to assess the argument in the sense of determining
its correctness, appropriateness, etc.-which we might call evaluation-models and models for
assessing argumentation in the sense of explaining what’s right or wrong in it what we might
call models for critical argumentation (Bermejo Luque 2014, p. 63; my translation).

In other words, according to Bermejo Luque, evaluation consists in the application

of standards to determine whether an argument satisfies some regulative conditions,
while criticism “consists of pointing out the defects of the argumentation, its weak-
nesses and possible counter-examples, which requires producing new argumentation
that justifies our negative judgments and our answers and objections.” (Op.cit., p. 66;
my translation). As a result, she says, the aim of evaluation is to determine the value
of an argument and the aim of criticism is to produce new arguments in order to show
the weakness of a given argument.
The picture that emerges from Bermejo Luque’s account is that evaluation and
criticism are different forms of argument appraisal. In evaluation arguments are
individually appraised by applying criteria or standards, while in criticism they are
appraised throughout the contrast with other arguments. This is in line with the idea,
already found in the pragma-dialectical code of conduct, that logical criteria and
standards are somehow external to the critical discussion.
It might be said that logic, as the theory of evaluation, is the normative theory
of argumentative practices, while argument dialectics, as the theory of criticism, is
the theory of practices that are normative on their own (Cfr. Olmos 2019a). The
assertion of the normative character of logic, understood as its power to legislate on
the validity and invalidity of arguments, clashes with the very idea of argumentative
practices. A practice can be defined as an exchange of socially significant actions
governed by rules that participants recognize. In turn argumentative practices are
9.6 Criticism and Evaluation 293

those exchanges that consist, wholly or in a significant part, in asking for, giving
and appraising reasons. The logician’s retrospective evaluation is in opposition to
the direct evaluation of those who participate in a debate. Participants do more than
exchanging arguments in favor or against a thesis, accepting or questioning premises
or unveil alleged fallacies. The practice of arguing is intrinsically evaluative. Beyond
offering reasons for consideration, the mastery of the art of arguing involves the
ability to balance and weigh competing arguments (think of adverbs like “but” or
“even”), and the ability to justify and explain the resulting weighing. Any description
of people’s argumentative practices would be incomplete if it ignores people’s own
normative ideas about argument appraisal. As says Marianne Doury: “A quick look
at argumentative practice makes it obvious that such spontaneous theories have a
normative component, which helps the arguers to elaborate their case and to evaluate
their opponent’s argument according to some standards” (Doury 2013, p. 492).
Robert C. Pinto contends that a distinctive mark of dialectic is the rejection of any
rule or standard for argument evaluation external to the argumentative exchange.
One cannot appraise an argument from a position one takes up outside the context of the
dialectical interchange in which that argument occurs. One cannot appraise an argument
in the role or office of neutral judge. Appraising an argument requires one to step into the
dialectical interchange, become party to it, become a participant in it. Informal logic, insofar
as it seeks to be an art of argument appraisal, would turn out to be the very art of arguing
itself. Plato had a name for it. He called it the art of dialectic (2001, pp. 8-9).

Charles Hamblin, to whom we owe the contemporary recovery of dialectic, puts it

differently:
Logicians are, of course, allowed to express their sentiments but there is something repugnant
about the idea that Logic is a vehicle for the expression of the logician’s own judgements
of acceptance and rejection of statements and arguments. The logician does not stand above
and outside practical argumentation or, necessarily, pass judgement on it. He is not a judge
or a court of appeal, and there is no such judge or court: he is, at best, a trained advocate. It
follows that it is not the logician’s particular job to declare the truth of any statement, or the
validity of any argument. (1970, p. 244).

Hamblin is rejecting here the traditional view of the logician as an onlooker of

debates or critical discussions. Hamblin claims that there is no normativity out of
the argumentative practices themselves. Given what Hamblin means by dialectical
criteria for argument evaluation (e.g. “The passage from premisses to conclusion
must be of an accepted kind”, op.cit. p. 245), argument dialectics may be defined
as the study of standards, criteria and procedures for the evaluation of arguments
involved in argumentative practices. In this sense logic, or the theory of argument,
consists in the theory of a normative practice rather than the practice of a normative
theory.
Argument dialectics is the result of placing logic within argumentative practices,
rejecting any form of normativity external to them. Argument dialectics is logic in
nature insofar as it is concerned with the structures determined by the construc-
tion of individual arguments and their interrelations. However, it is quite different
from formal and informal logic. According to Johnson and Blair “Informal logic is
294 9 Argument Dialectic

best understood as the normative study of argument. It is the area of logic which
seeks to develop standards, criteria and procedures for the interpretation, evalua-
tion and construction of arguments and argumentations used in natural language”.
(Johnson and Blair 1987, p. 148). Argument dialectics can be the defined as the
study and systematization of the norms and principles used by people in argumen-
tative exchanges to determine whether, in the given context, a certain argument
offers a pro tanto reason and, if so, its relative strength.3 Thus argument dialectics
endorses Woods’ principle of dialectification, according to which the question of
whether or not an argument possesses a certain logical property must be answered
in a dialectically reasonable discussion (Woods 2000, pp. 154–155).
Logic, either formal or informal, and argument dialectics are not two distinct disci-
plines with different subject matters. They are two competing, and even antithetical,
conceptions of the standards and criteria for good and bad argumentation.

References

Bermejo-Luque, Lilian (2014): Falacias y argumentación. Madrid: Plaza y Valdés.

Doury, Marianne (2013). “The Virtues of Argumentation from an Amoral Analyst’s Perspective”.
Informal Logic 33/4, pp. 486-509.
Eemeren, Frans H. van, & Rob Grootendorst (1984). Speech acts in argumentative discussions: A
theoretical model for the analysis of discussions directed towards solving conflicts of opinion.
Dordrecht: Floris Publications.
Eeemeren, Frans H. van, Rob Grootendorst & A. Francisca Snoeck Henkemans (2002). Argumen-
tation. Analysis, Evaluation, Presentation. Mahwah, NJ: Lawrence Erlbaum.
Ehninger, Douglas & Wayne Brockriede (2008 [1963]): Decision by Debate. New York: Idebate
Press.
Freeman, James B. (2011). Argument Structure: Representation and Theory. Dordrecht: Springer.
Habermas, Jürgen (1981): The Theory of Communicative Action. Boston: Beacon Press, 1984.
Hájek, Alan (2008).2008, “Are Miracles Chimerical?” In Jonathan Kvanvig, ed., Oxford Studies in
Philosophy of Religion (Volume 1), pp. 82–104. Oxford: Oxford University Press.
Hamblin, Charles (1970). Fallacies. London: Methuen.
Johnson, Ralph H. (2000). Manifest Rationality. A Pragmatic Theory of Argument. Mahwah, NJ:
Lawrence Erlbaum.
Johnson, Ralph H. & J. Anthony Blair & (1987): “The current state of informal logic”. Informal
Logic 9, pp. 147-151.
McGrew, Timothy (2019): “Miracles”, The Stanford Encyclopedia of Philosophy (Spring 2019
Edition), Edward N. Zalta (ed.), URL = <https://plato.stanford.edu/archives/spr2019/entries/mir
acles/>.
Olmos Gómez, Paula (2019). Normatividad argumentativa: “naturalización” vs. “socialización”.
Available at http://www.eafit.edu.co/escuelas/humanidades/departamentos-academicos/depart
amento-humanidades/debate-critico/Documents/Olmos_Normatividad_argumentativa_Naturli
zaci%C3%B3n_Socializaci%C3%B3n.pdf Accessed 14/06/2020.
Pinto, Robert (2001) Argument, Inference and Dialectic. Dordrecht: Kluwer.
Rescher, Nicholas (1977). Dialectics. A Controversy-Oriented Approach to the Theory of Knowl-
edge. Albany: State University of New York Press.

3A second difference with formal and informal logic is that arguments’ dialectics analyzes cogency
as a comparative, not qualitative, concept.
References 295

Snoeck Henkemans, A. Francisca (2000): “State-of-Art: The Structure of Argumentation”.

Argumentation 14, pp.447-473.
Vega Reñón, Luis (2013). La fauna de las falacias. Madrid: Trotta.
Walton, Douglas N., & Krabbe, Erik C.W. (1995). Commitment in Dialogue. Albany: State
University of New York Press.
Walton, Douglas N. (1996). Argument structure, a pragmatic theory. Toronto: University of Toronto
Press.
Wenzel, Joseph (1979): “Jürgen Habermas and the Dialectical Perspective on Argumentation”.
Journal of the American Forensic Association 16, 83–94.
Wenzel, Joseph (2006 [1990]): “Three Perspectives on Argument. Rhetoric, Dialectic, Logic”. In
Trapp, R. & Schuetz, J.H., Perspectives on Argumentation: Essays in Honor of Wayne Brockriede,
pp. 9–26. New York: Idebate Press.
Woods, John (2000): “How Philosophical is Informal Logic?”. Informal Logic Vol. 20, No.2,
pp. 139-167.
Chapter 10
Argumentation Structures
and Operations

Abstract This chapter and the following two deal with argumentative operations that
weave together arguments to produce complex arguments. An argumentative opera-
tion is a process by which two or more arguments are integrated into a single, more
complex, argument. Many argumentative connectors are commonly used as signs
of argumentative operations. An argumentative operation yields an argumentative
structure: the arrangement of parts in an argument, either simple or complex. Thus,
e.g. the operation of concatenation yields a subordinative argument. The successive
interventions of the participants in an argumentative exchange can be seen as appli-
cations of these operations, by means of which participants interactively weave a
macro-argument in the course of an argumentative exchange.

10.1 Argumentative Operations

This chapter and the following two deal with argumentative operations that weave
together arguments to produce complex arguments. An argumentative operation is a
process by which two or more arguments are integrated into a single, more complex,
argument. Many argumentative connectors are commonly used as signs of argumen-
tative operations. Thus, it could be said that in such cases the connectors act as
operators, i.e. as symbols that denote an argumentative operation. An argumentative
operation yields an argumentative structure: the arrangement of parts in an argument,
either simple or complex. Thus, e.g. the operation of concatenation yields a subordi-
native argument. The successive interventions of the participants in an argumentative
exchange can be seen as applications of these operations, and the macro-argument
to which I referred in 9.4 is thus the complex argument constructed in the course of
an argumentative exchange.
Two well-known concepts related to that of argumentative operations are those of
argumentative indicator and argumentative connector. An argumentative indicator
is a word or expression that can be used to indicate that a particular dialogical
move is being made (van Eemeren et al. 2007, p. 1). Since argumentative indicators
reveal that a dialogical move is being made, they pertain to arguer dialectics. By
contrast, argumentative operations pertain to argument dialectics. Remember that

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 297
F. Leal and H. Marraud, How Philosophers Argue, Argumentation Library 41,
https://doi.org/10.1007/978-3-030-85368-6_10
298 10 Argumentation Structures and Operations

Fig. 10.1 Basic

argumentative operations Operation Structure
and structures
Concatenation Subordinative argument
Backing Backed argument
Analogy Parallel argumentation
Conjunction Conjunctive argument
Disjunction Disjunctive argument
Objection Objection
Rebuttal Rebuttal
Refutation Refutation

argument dialectics is concerned with the structures determined by the construction

of individual arguments and their interrelations. Finally, an argumentative connector
is a morpheme that articulates two linguistic units giving them an argumentative
value (Moeschler 1985, p. 62); hence the notion of argumentative connector falls
into the domain of linguistic analysis.
Figure 10.1 provides a summary of the argumentative operations and structures
we will analyze in the sequel.
Any argument structure has an explicit argumentative orientation, positive, nega-
tive or neutral. Since arguing is to present to someone something as a reason for
something else, an argument structure contains a set of reasons. Roughly, an argu-
ment structure S is positively oriented towards a claim C if S favors the conclusion
that C, all reasons in S considered. An argument structure S is negatively oriented
towards C if S favors the opposite conclusion that not C, all reasons in S considered.
Finally, an argument structure S is neutrally oriented towards a claim C if (1) S
contains some substructure S’ either positively or negatively oriented towards C, and
(2) S favors neither the conclusion that C nor the conclusion that not C, all reasons
in S considered.

10.2 Simple Arguments

Let us quickly review some basic notions before embarking on the description and
analysis of argumentative operations and structures.
To argue is to exhibit, offer for examination or lay before someone’s cognizance
something as a reason for another thing. To that end the arguer may use a number of
conventional devices, among which the arrangement of statements within the text,
10.2 Simple Arguments 299

punctuation marks, pauses and intonation schemes, connectors and argumentative

operators, and the use of explicitly argumentative vocabulary. To realize the occur-
rence of an argument is to notice that something is presented as a reason for something
else, a process based on these conventionalized clues.
A simple argument is the minimum autonomous unit of argumentation, composed
of a reason and a claim. The statements that, taken together, express a reason are
called “premises”, while the conclusion is the statement that expresses the claim.
When arguments are understood as textual products, they are said to be formed by
linguistic units, which play the roles of premises and conclusion. When arguments are
understood as abstract entities, they are said to be formed by units that correspond to
the contents of beliefs, intentions and attitudes. Premises and conclusion are the terms
of a relation: a statement can act as a premise for some statement and a conclusion
for another statement.
First quarter earnings will provide a taste of what to expect in the second quarter. Discussion
will likely be limited to preserving company strength and survival, along with strategies for
the current environment. Expect any hints of post-coronavirus thinking to be tempered by
conditional statements.
Therefore, valuations and forecasts based on 1st quarter earnings and company
outlooks/strategies will be unreliable.
(John S. Tobey, “Stock Market: Run-Up Looks Ready To Pause Or Reverse”. Forbes
9/04/2020; my italics. Retrieved at https://www.forbes.com/sites/johntobey/2020/04/09/
stock-market-run-up-looks-ready-to-pause-or-reverse/#60f0966965b7).

1st quarter earnings will provide a taste of what to expect in the second quarter.
Discussion will likely be limited to preserving company strength and survival,
along with strategies for the current environment. Expect any hints of post-
coronavirus thinking to be tempered by conditional statements
Therefore
Valuations and forecasts based on 1st quarter earnings and company
outlooks/strategies will be unreliable

People who experience exposure to secondhand smoke can also develop lung cancer
Therefore
Maintaining a smoke-free home can help non-smokers reduce their risk of the condition

The connector “therefore” gives this paragraph its argumentative skeleton, attributing
to the linguistic segment that precedes it the role of premises and to the one that
succeeds it that of a conclusion.1 To diagram a simple argument we will place its

1 Note that, according to the preceding definitions, the argumentative connector “therefore” does
not work here as an operator denoting an argumentative operation, since it brings together two
sentences into a simple argument.
300 10 Argumentation Structures and Operations

premises and its conclusion in two parallel rectangles of the same length, joined by a
consecutive connector like “therefore”, with the premises above and the conclusion
below.
In the next example an explicit argumentative vocabulary (‘For this reason’) is
used to designate the status of text elements.
It is important to note that people who experience exposure to secondhand smoke can also
develop lung cancer. For this reason, maintaining a smoke-free home can help non-smokers
reduce their risk of the condition.
Mathieu Rees, “Lung cancer prevention: 5 ways to lower the risk”. Medical New Today,
30/03/2021; my italics. Accessed 01/04/2021, https://www.medicalnewstoday.com/articles/
lung-cancer-prevention

“This” refers to the antecedent part of the text, which is thus presented as a reason
why maintaining a smoke-free home can help reduce the risk of cancer.
A subtle but quite common way of presenting an argument is to put the offered reason
right after the claim, with a period, a comma or a semicolon in between.
The scale of the economic damage is breathtaking. In one recent poll, more than half of all
Americans under the age of 45 said that they had lost their jobs or suffered a loss of hours.
(‘We’re Going Down, Down, Down, Down, Down’. The New York Times 10/04/2020.
Retrieved at https://www.nytimes.com/2020/04/10/opinion/coronavirus-unemployment-
small-business.html?action=click&module=Opinion&pgtype=Homepage

Here the fact that over 50% of all Americans under the age of 45 said that they had
lost their jobs or suffered a loss of hours is presented as an evidence that the scale of
the economic damage from the coronavirus pandemic is breathtaking.

In one recent poll, more than half of all Americans under the age of 45 said that
they had lost their jobs or suffered a loss of hours.
Therefore
The scale of the economic damage from the coronavirus pandemic is breathtaking

10.3 Concatenation

Chaining or concatenation is an operation that allows two arguments to be combined

into a single complex argument when the conclusion of one of them figures among
the premises of the other argument. The result of performing a concatenation is a
subordinative argument, where two arguments are chained together by virtue of the
conclusion of the one also functioning as a premise in the other.
Crimes are short term, circumscribed events that presuppose a peculiar set of necessary
conditions (e.g. activity, opportunity, adversaries, victims, goods). Criminality, in contrast,
10.3 Concatenation 301

refers to stable differences across individuals in the propensity to commit criminal (or equiv-
alent) acts. Accordingly, criminality is only one element in the causal configuration leading
to a criminal act, and criminal acts are, at best, imperfect measures of criminality. It follows
that the frequency with which individuals participate in criminal events may vary over time
and place without implying change in their criminality. […]
(Michael Gottfredson and Travis Hirsch, “The distinction between crime and criminality”,
p. 190; my italics. In Travis Hirsh, The Craft of Criminology. Selected Papers, ed. John H.
Laub, 187-202. Piscataway NJ: Transaction Publishers 2015).

Abstracting from content and focusing on structural elements, we get the pattern: A.
B. Accordingly C. It follows further that D. In order to interpret this passage, we have
to determine the value and scope of the structural devices. The first two sentences
are mutually independent characterizations of crime and criminality. Since charac-
terizations are usually part of the ground of a reason, and neither is subordinated to
the other, I take them to be co-premises working together to express a single reason.
If I am right, then, that period can be replaced by conjunction “and”. The adverb
“accordingly” indicates that the preceding conjunction is presented as a reason for
the claim “criminality is only one element in the causal configuration leading to
a criminal act, and criminal acts are, at best, imperfect measures of criminality”.
Hence, we have a first argument which can be diagrammed as follows:

Crimes are short term, circumscribed events that presuppose a peculiar set
of necessary conditions (e.g. activity, opportunity, adversaries, victims,
goods). Criminality, in contrast, refers to stable differences across
individuals in the propensity to commit criminal (or equivalent) acts
Accordingly,
Criminality is only one element in the causal configuration leading to a
criminal act, and criminal acts are, at best, imperfect measures of
criminality

The locution “it follows” plays a double function. First, it functions as a consecu-
tive connector linking the linguistic units that flank it, as premise and conclusion,
respectively. Thus, we have a second argument:

Criminality is only one element in the causal configuration leading to a

criminal act, and criminal acts are, at best, imperfect measures of criminality
Therefore
The frequency with which individuals participate in criminal events may
vary over time and place without implying change in their criminality

Second, “it follows” makes the final statement (“the frequency with which individuals
participate in criminal events may vary over time and place without implying change
302 10 Argumentation Structures and Operations

in their criminality”) dependent on the whole preceding fragment. In this sense, “it
follows further” works here as an operator for concatenation.
To sum up, the passage contains two arguments with a common element which is
the conclusion of the first argument and the premise of the second, so that chaining
them together produces a subordinative argument:

Crimes are short term, circumscribed events that presuppose a peculiar set of
necessary conditions (e.g. activity, opportunity, adversaries, victims, goods).
Criminality, in contrast, refers to stable differences across individuals in the
propensity to commit criminal (or equivalent) acts
Accordingly,
Criminality is only one element in the causal configuration leading to a criminal
act, and criminal acts are, at best, imperfect measures of criminality
Therefore
The frequency with which individuals participate in criminal events may vary
over time and place without implying change in their criminality

Thus, a subordinative argument is made up of two subarguments combined so

that the conclusion of one is at the same time one of the premises of the other. A
subordinative argument is a composite argument, with its basic premises and its
main conclusion. A basic premise in a subordinative argument is a statement that is a
premise of some argument but not the conclusion of any subargument. Likewise, the
main conclusion in a subordinative argument is a statement that is the conclusion of
some subargument but not a premise of any subargument. A subordinative argument
is positively oriented towards its main conclusion. In the example above, “Crimes
are short term, circumscribed events that presuppose a peculiar set of necessary
conditions” and “Criminality refers to stable differences across individuals in the
propensity to commit criminal acts” are the basic premises, and “The frequency
with which individuals participate in criminal events may vary over time and place
without implying change in their criminality” is the main conclusion. The remaining
statement, “Criminality is only one element in the causal configuration leading to a
criminal act, and criminal acts are, at best, imperfect measures of criminality”, which
is the conclusion of one argument and the premise of another, is an intermediate
conclusion.
It can be added, from the point of view of arguer dialectics, that concatenation is
often the response to the questioning of a premise and is therefore associated with
objection (Cfr. Chap. 11).

10.4 Warrants and Backings

Whoever offers an argument must also be willing, if required, to explain or justify

the assumption that its premises present a reason for its conclusion. If needed, this
10.4 Warrants and Backings 303

assumption can be made explicit by means of the associated conditional If premises,

then conclusion. One way to satisfy that requirement is to provide a warrant. A
warrant thus answers questions such as: “How do you go from the premises to the
conclusion?”, “What do the premises have to do with the conclusion”, or “Why is that
a reason?”. Warrants, therefore, have to do with the nature of the connection between
premises and conclusion. Those questions can be used to ask for a justification or
to ask for an explanation of the proposed reasoning, depending on whether it is
being questioned or not. Thus, the warrant can serve as either a justification or an
explanation of the relationship between the given premises and the conclusion.
The addition of a warrant increases the complexity of an argument without turning
it into a composite argument, since warrants are conditional statements, not argu-
ments. Warrants express that in virtue of which something is a reason for something
else, they “correspond to the practical standards or canons of argument” (Toulmin
2003, p. 91). This is different from the consideration that constitutes the reason,
which is what the premises express.
Premises and conclusion, on one side, and warrants, on the other, are not
constituents of an argument on equal footing. Premises and conclusion are the
minimal constituents of an argument; warrant is an optional constituent. To present
a consideration as a reason for something else it is not necessary to specify why it is
supposed to be a reason for it.
In any case, warrants are statements and statements can be the conclusion of an
argument. I call “backing” the operation that combines two arguments when the
conclusion of one of them is the warrant of the other. The outcome is like a subordi-
native argument, but the shared element is now the conclusion of one argument and
the warrant of the other. So backing is an operation that gives rise to a special kind
of chaining, that I will term “backed argument”. Let us consider an example.
If [renowned cricket commentator Henry] Blofeld is not [Freddie] Calthorpe’s [the amateur
all-rounder who captained England in a four Test series in West Indies in 1930] nephew, the
question remains: how are they related, if at all? Blofeld’s middle name [Calthorpe] suggests
that he is descended from someone with the surname Calthorpe—not necessarily that he is
related to anyone with that surname. […] Two people with the surname Calthorpe probably
both had ancestors who lived in that village—in which case it’s likely that they are related.
If Henry Blofeld is related to someone with the surname Calthorpe, he is probably related to
anyone with that surname; however, it’s possible that the connection may be several centuries
earlier. Since, as a rule of thumb, the number of records available for researching British
genealogy decreases the further back in time one progresses, even if Blofeld and Calthorpe
are related, it may not be possible to prove it.
(Michael Jones: “Henry Blofeld: Nephew of an England captain?”. Cricket Country,
November 28, 2014. https://www.cricketcountry.com/articles/henry-blofeld-nephew-of-an-
england-captain-217019).

Michael Jones’ argumentation is an attempt to answer the question at the beginning

of the passage, which can be summarized as: “How are Henry Blofed and Freddie
Calthorpe related to each other? Since it is a complex argument, we are going to
break it down into subarguments. Jones’ answer, at the end of the paragraph, is
that even if Blofeld and Calthorpe are related, it may not be possible to prove it.
304 10 Argumentation Structures and Operations

The connector “Since” introduces the warrant connecting this conclusion with the
antecedent premises “Henry Blofeld is probably related to Freddie Calthorpe” and
“It’s possible that the connection may be several centuries earlier”. Moreover, that
the sentence “the number of records available for researching British genealogy
decreases the further back in time one progresses” is said to be a rule of thumb
confirms that it functions as a warrant. To diagram warrants, we shall write it just
before the connector indicating the step the warrant is intended to justify:

Henry Blofeld is probably related to Freddie

Calthorpe. It’s possible that the connection may
be several centuries earlier
The number of records available for
researching British genealogy decreases Therefore
the further back in time one progresses:
Even if Blofeld and Calthorpe are related, it may
not be possible to prove it

Why are we entitled to add the premise “Henry Blofeld is probably related to Freddie
Calthorpe”? As it is well known, Toulmin claims that “data are appealed to explicitly,
warrants implicitly.” (2003, p. 92). As a matter of fact, this is not altogether true.
What happens rather is that, for linguistic economy’s sake, whenever all the premises
are made explicit, the warrant is left implicit. Mutatis mutandis, whenever the warrant
is made explicit, some of the premises is left implicit: from the explicit premises and
warrant the addressee can easily infer the missing premise. Likewise, when all the
premises and the warrant are made explicit there is no need to do the same with the
conclusion, as happens in:

Blofeld’s middle name, Calthorpe, suggests that he is

descended from someone with the surname Calthorpe
If Henry Blofeld is related to someone
with the surname Calthorpe, he is
Therefore
probably related to anyone with that
surname:
Henry Blofeld is probably related to Freddie
Calthorpe

Since the conclusion of the second argument is one of the premises of the first, they
can be chained together to produce a subordinative argument:
10.4 Warrants and Backings 305

Blofeld’s middle name, Calthorpe,

suggests that he is descended from
someone with the surname Calthorpe
If Henry Blofeld is related
to someone with the
surname Calthorpe, he is Therefore
probably related to anyone
with that surname:
It’s possible that the
Henry Blofeld is probably related to
connection may be
Freddie Calthorpe
several centuries earlier
The number of records
available for researching
British genealogy Therefore
decreases the further back
in time one progresses:
Even if Blofeld and Calthorpe are related, it may not be possible
to prove it

We still have to analyze the sentence “Two people with the surname Calthorpe prob-
ably both had ancestors who lived in that village—in which case it’s likely that they
are related”. The phrase “in which case” gives it the structure of an argument:

Two people with the surname Calthorpe probably both had ancestors who lived in that village
Therefore
It’s likely that two people with the surname Calthorpe were related

The conclusion of this argument is a categorical statement that precedes the condi-
tional statement “If Henry Blofeld is related to someone with the surname Calthorpe,
he is probably related to anyone with that surname”, which suggests that they are
related as backing and warrant, respectively, since “statements of warrants, we saw,
are hypothetical, bridgelike statements, but the backing for warrants can be expressed
in the form of categorical statements of fact quite as well as can the data appealed to
in direct support of our conclusions.” (Toulmin 2003, pp. 97–98). Thus, we have a
backed argument:
306 10 Argumentation Structures and Operations

It’s likely that two people with the

surname Calthorpe were related
Blofeld’s middle name, Calthorpe, suggests that he is
Therefore
descended from someone with the surname Calthorpe
If Henry Blofeld is related to someone
with the surname Calthorpe, he is
Therefore
probably related to anyone with that
surname:
Henry Blofeld is probably related to Freddie
Calthorpe

Putting all these pieces together, we get the full diagram of Jones’ argument:

Two people with the

surname Calthorpe
probably both had
ancestors who lived in
that village
Therefore
It’s likely that two people
with the surname
Calthorpe were related
Blofeld’s middle name, Calthorpe,
Therefore suggests that he is descended from
someone with the surname Calthorpe
If Henry Blofeld is related
to someone with the
surname Calthorpe, he is Therefore
probably related to anyone
with that surname:
It’s possible that the
Henry Blofeld is probably related to
connection may be
Freddie Calthorpe
several centuries earlier
The number of records
available for researching
British genealogy Therefore
decreases the further back
in time one progresses:
Even if Blofeld and Calthorpe are related, it may not be possible
to prove it

I use ‘backing’ as a name for an argumentative operation that produces a backed

argument. This sense should not be confused with the sense popularized by Toulmin.
In Toulmin’s model, backing is an element in the layout of an argument whose
function is to establish the warrant. The relationship between Toulmin’s backing and
warrant is similar to that between grounds and claims:
10.4 Warrants and Backings 307

But the backing of the warrants we invoke need not be made explicit—at any rate to begin
with: the warrants may be conceded without challenge, and their backing left understood.
Indeed, if we demanded the credentials of all warrants at sight and never let one pass unchal-
lenged, argument could scarcely begin. Jones puts forward an argument invoking warrant
W1 , and Smith challenges that warrant; Jones is obliged, as a lemma, to produce another
argument in the hope of establishing the acceptability of the first warrant, but in the course
of this lemma employs a second warrant W2 ; Smith challenges the credentials of this second
warrant in turn; and so the game goes on. Some warrants must be accepted provisionally
without further challenge, if argument is to open to us in the field in question (Toulmin 2003,
p. 98).

Thus, Toulmin’s “backing” is the reason adduced to support the warrant in a

backed argumentation. “Backing” in my sense is the operation one performs when
one justifies a warrant with a “backing” in Toulmin’s sense. The result of performing
this operation is a backed argument, such as:

It’s likely that two people with the

surname Calthorpe were related
Blofeld’s middle name, Calthorpe, suggests that he is
Therefore
descended from someone with the surname Calthorpe
If Henry Blofeld is related to someone
with the surname Calthorpe, he is
Therefore
probably related to anyone with that
surname:
Henry Blofeld is probably related to Freddie
Calthorpe

Here “Blofeld’s middle name, Calthorpe, suggests that he is descended from

someone with the surname Calthorpe, therefore Henry Blofeld is probably related
to Freddie Calthorpe” is the warranted argument, and “It’s likely that two people
with the surname Calthorpe were related, therefore If Henry Blofeld is related to
someone with the surname Calthorpe, he is probably related to anyone with that
surname” is the backing argument. A backed argument is positively oriented towards
the conclusion of its warranted subargument (“Henry Blofeld is probably related to
Freddie Calthorpe” in the example above).
Giving and backing warrants occur when the bearing of the premises on the
conclusion is challenged. For this reason, warrant and backing are related to plain
rebuttal (see Sect. 11.5)).

10.5 Analogous Arguments

The function of the warrant, as just explained, is to show that the premises are properly
connected to the conclusion. This can also be shown comparing the disputed argument
with another one that, supposedly, has the same internal structure—meaning that the
308 10 Argumentation Structures and Operations

premises of the first one are to its conclusion as the premises of the second one are to
its conclusion. Arguments with the same internal structure are said to be analogous.
If argument A is analogous to argument B, the relationship between the premises and
the conclusion is similar in both cases. Whatever property is conferred upon a given
argument by virtue of its internal structure, is also conferred upon any argument
sharing that structure. Thus, if the inference proposed by one of them is taken to
be legitimate, coherence requires the same consideration is given to the inference
proposed by the other. In other words, if an argument is analogous to an inferentially
sound argument, then it is also inferentially sound.
By an inferential property I mean a property pertaining to the internal structure of
an argument. In an argument by analogy (or by parity of reasoning)2 it is argued that
some inferential property of an argument (the source) can be transferred to another
argument (the target) because they are analogous to each other.
You can get a large audience together for a strip-tease act—that is, to watch a girl undress
on the stage. Now suppose you come to a country where you could fill a theatre by simply
bringing a covered plate on to the stage and then slowly lifting the cover so as to let everyone
see, just before the lights vent out, that it contained a mutton chop or a bit of bacon, would
you not think that in that country something had gone wrong with the appetite for food? And
would not anyone who had grown up in a different world think there was something equally
queer about the state of the sex instinct among us?
C.S. Lewis, Mere Christianity, p. 96. San Francisco: Harper Collins.3

Lewis begins by stating a fact: one can get a large audience together to watch a
girl undress on the stage, and then introduces a supposition. Whoever introduces an
assumption does so in order to invite the addressee to explore its consequences. In
this way Lewis is trying to get the addressee to concede that if such strip-tease acts
with food were popular, we would conclude that something had gone wrong with
the appetite for food —i.e., that the popularity of such acts is a reason to believe that
something is wrong with food. But in such a case, Lewis goes on, the popularity of
strip-tease is a reason to believe that something is wrong with sex.
Thus Lewis contends that the arguments below are parallels (although the first is
a suppositional argument while the second is a linear argument)4 :

Suppose that in some country you could fill a theatre by simply bringing a covered
plate on to the stage and then slowly lifting the cover so as to let everyone see, just
before the lights vent out, that it contained a mutton chop or a bit of bacon
Therefore
In that country something had gone wrong with the appetite for food

and

2 Woods and Hudak (1989, p. 127).

3 Cited by Govier (2010, pp. 350–351):
4 In a suppositional argument some of the premises is or is equivalent to an supposition with the

form ‘Suppose P’; otherwise, if all the premises are asserted, the argument is linear.
10.5 Analogous Arguments 309

You can get a large audience together for a strip-tease act—that is, to watch
a girl undress on the stage
Therefore
There is something equally queer about the state of the sex instinct among us

If so, the two arguments “stand or fall together and that they do so because they
are relevantly at parity, that they possess similar deep structures by virtue of which
they coincide in logical form” (Woods and Hudak 1989, p. 127).
That two arguments are parallel means that they rest on analogous inferential
presuppositions or invoke analogous reasons. Since inferential presuppositions can
be made explicit by means of the associated conditional, resorting to conditionals a
parallel argumentation can be represented as a simple argument:

If such striptease acts with food were popular, then something

had gone wrong with the appetite for food
Therefore
If one can get a large audience together to watch a girl undress
on the stage, then something is wrong with the sex instinct

Once established that fact that one can get a large audience together to watch a girl
undress on the stage is a reason to believe that something is wrong with sex, given
the starting point, we must conclude that there is something queer about the state of
the sex instinct among us. In the following diagram the step-authorizing function of
parallel argumentation is clearly displayed:

If such striptease acts with food were popular,

then something had gone wrong with the
appetite for food
One can get a large audience together to
Therefore
watch a girl undress on the stage
If you can get a large audience together to
watch a girl undress on the stage, then Therefore
something is wrong with the sex instinct:
There is something equally queer about
the state of the sex instinct among us

The criticism Lewis is trying to anticipate and defuse with this comparison is
that the popularity of striptease is not a reason to believe that something is wrong
with sexual appetite. More often argument by analogy is used to rebut argument, on
the assumption that any argument analogous to an argument in which the premises
do not bear on the conclusion will have the same flaw. This sort or rebuttal, called
“counter-analogy”, will be considered later on, in Sect. 11.5.
310 10 Argumentation Structures and Operations

References

Eeemeren, Frans H. van, Grootendorst, Rob, & Snoeck Henkemans, A. Francisca (2002).
Argumentation. Analysis, Evaluation, Presentation. Mahwah, NJ: Lawrence Erlbaum.
Eemeren, Frans H., Houtlosser, Peter, & Snoeck-Henkemans, A. Francisca (2007). Argumentative
Indicators in Discourse. Dordrecht: Springer.
Govier, Trudy (2010). A Practical Study of Argument. Seventh edition. Belmont CA: Wadsworth.
Moeschler, Jacques (1985). Argumentation et conversation. Paris: Hatier.
Toulmin, Stephen E. (2003 [1958]). The Uses of Argument, Cambridge: Cambridge University
Press.
Woods, John & Hudak, Brent (1989). “By Parity of Reason”. Informal Logic XI.3, pp. 125–139.
Chapter 11
Counterarguments

Abstract The concept of counterargument lies at the heart of argument dialectics.

The importance attached to the relations between arguments —especially those of
opposition— establishes a clear distinction between argument dialectics and logic,
whether formal or informal. My first aim in this chapter is to give a suitable definition
of counterargument; my second aim is to establish a typology of counterarguments.
Before we start, it is important to note that in the sequel we will consider only logical
counterarguments: arguments criticizing another argument on account of its logical
properties. Hence ways of attacking an argument for being ineffective for a given
audience or for procedural matters will not be considered here.

11.1 Counterarguments and Reasons

The concept of counterargument lies at the heart of argument dialectics. The impor-
tance attached to the relations between arguments—especially those of opposition—
establishes a clear distinction between argument dialectics and logic, whether formal
or informal. My first aim in this chapter is to give a suitable definition of counter-
argument; my second aim is to establish a typology of counterarguments. Before
we start, it is important to note that in the sequel we will take into account only
logical counterarguments: arguments criticizing another argument on account of its
logical properties. Hence ways of attacking an argument for being ineffective for a
given audience or for procedural matters will not be considered here. In addition, my
classification of counterarguments differs from other well-known typologies—such
as Walton (2013) or Finocchiaro (2013)—by being based on Toulmin’s model.
According to the Merriam-Webster Dictionary to counterargue is to give (reasons,
statements, or facts) in opposition to an argument or in support of an opposing argu-
ment. Thus, a counterargument offers a reason to reject another argument. Therefore,
there is a clear difference between a con argument, a reason against a claim, and a
counterargument, a reason against the cogency of an argument.
This week’s tragedy in Nova Scotia was unimaginable. A mass shooter took the lives of
22 great Canadians, including first responders and public servants helping us through the
COVID-19 crisis. And, as we heard from the RCMP Friday, many of the victims were heroes
trying to save others.
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 311
F. Leal and H. Marraud, How Philosophers Argue, Argumentation Library 41,
https://doi.org/10.1007/978-3-030-85368-6_11
312 11 Counterarguments

Given these facts, I understand the protective instinct of some readers who want The Globe
to stop naming the gunman.

The argument against publishing the name of mass shooters or terrorist acts that claim lives
is that the criminals responsible for the acts want notoriety, and we should not cave to this;
rather, we should remember the names of those whose lives were so cruelly taken.

(Sylvia Stead, “Nova Scotia gunman must be named, but keep the focus on the heroes and
victims”. The Globe and Mail 25/04/2020. Retrieved at https://www.theglobeandmail.com/
public-editor/article-nova-scotia-gunman-must-be-named-but-keep-the-focus-on-the-her
oes-and/).

This is an argument against the claim that it is permissible for newspapers and
magazines to publish the names of mass shooters or terrorists. A reason is given to
reject the claim that it is permissible, or more precisely to endorse the opposite claim
that it is not permissible.

(1) The criminals responsible for the mass shooting or terrorist

acts want notoriety and we should not cave to this
Therefore
Newspapers and magazines should not publish the names
of mass shooters and terrorists

By contrast, a counterargument is presented in the following passage.

ETCS [the European Train Control System] comes at a politically sensitive time as arguments
still rage over the viability and cost benefits of HS2 [High Speed 2, a partly planned high
speed railway]. Champions of the new line [HS2] have recently shifted their argument away
from pure speed benefits to capacity enhancement. They claim that without HS2 the existing
network will grind to a halt, but there will be some who will say that ETCS could create
extra capacity and therefore undermines that argument.

“Changing the signals on Britain’s railways”. BBC News 25/04/2025. Retrieved at https://
www.bbc.com/news/magazine-32446717

First, an argument is mentioned:

(2) Without the HS2 the existing network will grind to a halt
Therefore
HS2 must be build

And afterwards, a second argument is brought into play to undermine this argument:
11.2 Main Kinds of Counterarguments 313

(3) ETCS could create extra capacity

Therefore
Even without the HS2 the existing network will not grind to a halt

This is a counterargument, since the fact that ETCS could create extra capacity is
presented as a reason against the premise of the first argument. It should be note, by
the way, that the role of the adverb “even” in the conclusion of (3) is to make explicit
that (3) is being used as a counterargument.
Of course, whether something is a con argument or a counterargument depends on
how it is being used. Suppose Brown were to argue that newspapers and magazines
must publish the names of mass shooters and terrorists for public scorn, and that
Smith then replied that the criminals responsible for the acts want notoriety, and we
should not cave to this. In this context (1) would work as a counterargument against
Brown’s argument.

11.2 Main Kinds of Counterarguments

When someone rejects an argument presented to her, there are four possible moves.
1. She might reject the argument as a whole, alleging that there must be something
wrong in it.
2. She might challenge some of the premises: “Where do you get that from?”.
3. She might question instead whether the reason offered is actually a reason for
that conclusion: “What does that have to do with the conclusion?”
4. Finally, she might try to show that the conclusion should still be rejected, offering
an opposite reason: “Yes, but…”.
Consequently, there are four main strategies for logically attacking an argument: chal-
lenging the argument as a whole, challenging some of its premises, challenging the
relationship between the premises and the conclusion, or challenging its conclusion.
Counterarguing requires more than challenging some of the parts of an argument:
one has to give reasons for rejection. Thus one can either meta-argue that the offered
argument is not cogent or argue that some of the premises are not usable, that the
premises are not adequately connected to the conclusion, or that the conclusion is
nonetheless false or unacceptable. Accordingly, I shall distinguish four main kinds
of counterargument, which I term, respectively, dismissal, objection, rebuttal and
refutation. These four kinds are successively described below.
314 11 Counterarguments

11.3 Dismissal

An argument can be logically rejected as a whole for a variety of reasons; for instance,
it can be adduced that argument A has been rejected by most philosophers, that A
is not a scientific argument, that A leads to an unreasonable conclusion, etc. The
rationality of such holistic counterarguments has been advocated by Daniel Cohen
(2001). Here is an example:
[…] as we saw before, states and corporations differ in one crucial respect. The shareholders
of a corporation voluntarily take on the obligations of the corporation when they purchase
shares; indeed, the corporation’s obligations are reflected as a discount in the price of a share.
People who are born into citizenship of a state do not consent in a similar manner to take on
the obligations that others have acquired in the name of the state. Although Locke argued
that people give implicit consent to their government by not emigrating, no one takes this
argument seriously anymore. Consent requires more than the ability to choose an extremely
disagreeable alternative. (Posner, 2003: 1906).

For present purposes Locke‘s argument can be summarized thus:

People give implicit consent to their government by not emigrating

Therefore
People have the duty to live up to the state's obligations

Posner counterargues that no one takes this argument seriously anymore. The implicit
conclusion here is that Locke’s argument must be unsound. Hence a dismissal is an
argument whose conclusion is an assertion to the effect that a given argument is not
cogent. Since a meta-argument is “an argument about one or more arguments or about
argumentation in general” (Finocchiaro 2013, p. 1), dismissals are meta-arguments.

No one takes seriously anymore Locke’s argument that people have the
duty to live up to the state's obligations
Therefore
Locke’s argument that people have the duty to live up to the state's
obligations is not cogent

Some fallacy charges can also be seen as dismissals of some kind, as shows the
following example.
A judge has granted a conditional discharge to a Eureka, Pictou Co. woman [Emily Anne
MacIntosh] who stole more than $25,000 from the special care home where she worked as
an administrator earning $99,000 annually. […] Her lawyer tried to argue the offence “is
rendered somehow less serious as ‘there is no evidence of the government going insolvent’
or as ‘nobody missed out on lunch,’” [Pictou provincial court Judge Del] Atwood said. “In
11.3 Dismissal 315

my view, this is a fallacy of false simplicity.” (Lambie Judge grants discharge to Pictou
County woman in theft from special care home. The News 22/11/2018. Retrieved February
7, 2019 from https://www.ngnews.ca/news/local/judge-grants-discharge-to-pictou-county-
woman-in-theft-from-special-care-home-261878/).

The lawyer offers two similar arguments that can be reconstructed as follows:

There is no evidence of the government going

insolvent due to MacIntosh’s action
Therefore
MacIntosh’s offence is less serious than it appeared

Nobody missed out on lunch due to MacIntosh’s action

Therefore
MacIntosh’s offence is less serious than it appeared

These arguments are then dismissed by Lambie as fallacious: the argument that since
there is no evidence of the government going insolvent, MacIntosh’s offence is less
serious than it appeared is a fallacy of false simplicity, and so is not cogent, and the
same goes for the other argument.
Lambie’s first dismissal can be diagrammed as a meta-argument thus:

The argument that since there is no evidence of the government going

insolvent, MacIntosh’s offence is less serious than it appeared is a fallacy
of false simplicity
Therefore
The argument that since there is no evidence of the government going
insolvent, MacIntosh’s offence is less serious than it appeared is not cogent

To represent the incompatibility of two arguments I use the preposition against. I

place “against” between the incompatible elements of the two arguments. In the
present case these are the conclusion of the dismissing argument and the dismissed
argument as a whole, that we enclose in a box to indicate this incompatibility.
316 11 Counterarguments

The argument that since there is no

evidence of the government going
insolvent, MacIntosh’s offence is less
serious than it appeared is a fallacy of
false simplicity
Therefore
There is no evidence of the government
The argument that since there is no going insolvent due to MacIntosh’s action
evidence of the government going
against Therefore
insolvent, MacIntosh’s offence is less
serious than it appeared is not cogent MacIntosh’s offence is less serious than it
appeared

The resulting argument structure is neutrally oriented towards the claim that MacIn-
tosh’s offence is less serious than it appeared, the conclusion of the dismissed
argument.

11.4 Objection

An argument A is an objection to an argument B when A’s conclusion is incompatible

with some of the premises of B. This is obviously the case when they are mutually
contrary or contradictory, but in argumentation there are also less stringent forms of
incompatibility. Here is an illustration of this wider sense of incompatibility:
A. This patient has a streptococcal infection, so presumably this patient needs penicillin
treatment.

CA. Infection diagnosis is only based on symptoms, no clinical tests have been performed.

The implication of the reply is not that the premise in the original argument is false,
but that since no clinical tests have been performed, the speaker is not entitled to
assert it. This can be diagrammed thus, with the conclusion of the objection and the
contentious premise in the same line, united by preposition against:

Infection diagnosis is only based on symptoms and

no clinical tests have been performed
Therefore
It cannot be asserted that the patient has a This patient has a
Against
streptococcal infection streptococcal infection
Therefore
This patient needs penicillin
treatment
11.4 Objection 317

The resulting structure is neutrally oriented towards the claim that his patient needs
penicillin treatment.

11.5 Rebuttal

I am using “rebuttal” in a sense close to Toulmin (2003). To avoid possible confu-

sions, it should be noted that Pollock (2010) uses “rebutting defeaters” and “under-
cutting defeaters” for something similar to what I call “refutations” and “rebuttals”,
respectively.
Given an argument P therefore C, questions like “Why do you think P is a reason
for C?”, “How do you get from P to C?” or simply “Is it true that if P, then C?”
can be answered in either of two basic ways: (a) Resorting to an analogy: step
from P to C is like (is analogous to) the already accepted step from P’ to C’; or
(b) Providing a warrant, i.e. a general, hypothetical statement, which can act as a
bridge, and authorizes the sort of step to which our particular argument commits us
(Toulmin 2003, p.91).1 Mutatis mutandis, the inference proposed by an argument can
be rejected by arguing either that it is analogous to another already rejected inference
(counteranalogy or rebuttal by parallel argument), or that its warrant is problematic
(warrant rebuttal). There are other ways in which the inferential presupposition of an
argument can be attacked. A direct way to rebut the argument P therefore C would
be to argue that it is false that if P then C without recourse to analogy. I call it “direct
rebuttal”. As for its effects, rebuttal is similar to objection in that both are neutrally
oriented towards the conclusion of the rebutted argument.
Let us illustrate first rebuttal by parallel argument (counteranalogy) with an
example from Pies (How assisted suicide affects physicians. Undark Magazine
01/09/2018. Retrieved February 7, 2019 from https://undark.org/article/assisted-sui
cide-physician-health/).
A. As a physician and medical ethicist, I am opposed to any form of physician assistance
with a patient’s suicide. The Hippocratic Oath—arguably, the most important foundational
document in medical ethics— clearly states: “I will neither give a deadly drug to anybody if
asked for it, nor will I make a suggestion to this effect.”

CA. There was a time and place for the Hippocratic Oath to work, it doesn’t mean it’s always
appropriate for all situations far into the future. That would be like the idiots who argue that
we should always be allowed to keep guns just because it’s in the Constitution.

The corresponding diagram looks like this:

1See van Laar (2015) for a discussion of the different ways in which the adequacy of the connection
between the premises and the conclusion could be defended.
318 11 Counterarguments

The fact that it is in the Constitution

doesn’t mean that we should always be
allowed to keep guns
The Hippocratic Oath
clearly states: “I will
Both arguments neither give a deadly
Therefore
are analogous: drug to anybody if asked
for it, nor will I make a
suggestion to this effect.”
The fact that any form of physician
assistance with a patient’s suicide is
banned by the Hippocratic Oath doesn’t against Therefore
mean that physicians and medical
ethicists should oppose to it now:
Physicians and medical
ethicists should oppose
any form of physician
assistance with a
patient’s suicide

Let us turn to warrant rebuttal. A warrant can be held to be problematic by one

of two reasons: either the warrant is generally invalid, so that conclusions like C
cannot be inferred from premises like P through W (Plain rebuttal), or the warrant
cannot be applied in the case at hand, so that C cannot be inferred from P through
W (Exception). There are weaker forms of these (Disclaimers), since the validity
or applicability of the warrant may be controversial, so that C can only be inferred
with reservations from P, “at your own risk”, so to speak. Thus we have four kinds
of direct rebuttals: plain rebuttal, exception, warrant disclaimer and disclaimer by
possible exception.
Plain rebuttal.
… the political leader who has spent his entire career on bravura expressions of crime-
fighting and corruption-fighting [Philippines President Rodrigo Dutertre] suggests that “get-
ting angry” [with China for preventing Filipino fishermen from fishing in or around the
Scarborough shoal] is mere theater, because we do not have the resources to match those of
China. This is an argument that must be debunked; it is not true that if we get angry with
China because it has bullied us, we are only putting on airs.

The arbitral decision is itself proof that even the small and the bullied can put up a fight.

“Can’t win?”. Editorial in Philippine Daily Inquirer 14/10/2016. Accessed 02/04/2021.

https://opinion.inquirer.net/98186/cant-win.

The editorialist credits President Dutertre with the following argument:

Philippines does not have the resources to match those of China

Therefore
“Getting angry” with China for preventing Filipino fishermen from
fishing in or around the Scarborough shoal would be mere theater
11.5 Rebuttal 319

And then he goes on to argue that the inferential presupposition of the argument (if
we get angry with China because it has bullied us, we are only putting on airs) is
false:

The arbitral decision [of

the Permanent Court of
Arbitration in The Hague
on the suit lodged by the
Philippines against
China] is itself proof that
even the small and the
bullied can put up a fight
Therefore
Even the small and the
against
bullied can put up a fight
Philippines does not have the
resources to match those of
China
If someone does not
have the resources to
Therefore
confront another, he can
only feign to do so:
“Getting angry” with China for
preventing Filipino fishermen
from fishing in or around the
Scarborough shoal is mere
theater

Exception.
A. Suing a foreign government is generally prohibited by the Foreign Sovereign Immunities
Act of 1976; so the lawsuit by the state of Missouri that seeks damages from China over its
handling of the coronavirus pandemic has no legal basis at all.

C. The claims of the suit fit within the “commercial activities” exception to the Foreign
Sovereign Immunities Act, which denies foreign governments immunity from lawsuits
involving those governments’ commercial activities in the U.S.2

This is a quite clear example since the warrant of the attacked argument (Suing a
foreign government is generally prohibited by the Foreign Sovereign Immunities Act
of 1976) is stated explicitly, and the circumstance that allegedly rebuts its applicability
(the lawsuit involves China’s government’s commercial activities in the U.S.) is
explicitly identified as an exception to that rule.

2 Taken from “Chinese Foreign Ministry Spokesman Dismisses Missouri Coronavirus Lawsuit:
‘No Factual and Legal Basis At All’”, by Zachary Evans. National Review 23/04/2020.
Retrieved at https://www.nationalreview.com/news/chinese-foreign-ministry-spokesman-dismis
ses-missouri-coronavirus-lawsuit-no-factual-and-legal-basis-at-all/.
320 11 Counterarguments

The Missouri lawsuit

involves China’s
government’s commercial
activities in the U.S:
Therefore
The claims of the suit fit
within the “commercial
Against
The Missouri lawsuit accuses
various Chinese government
entities of negligence in their
handling of the coronavirus
pandemic
Suing a foreign government
is generally prohibited by

activities” exception to the Foreign Sovereign Therefore

the Foreign Sovereign Immunities Act of 1976:
Immunities Act
The Missouri lawsuit has no legal
basis at all

Once again, rebuttal by exception is neutrally oriented towards the conclusion of the
rebutted argument.
Warrant disclaimer.3
A. Primed for a November 7th release and coming from British powerhouse stable Working
Title, the new film from James Marsh is a likelihood for a fall festival bow. Its biopic structure
and potential crowd pleaser appeal hints at Toronto rather than Venice.

CA. Well, perhaps. But that’s a pretty unreliable rule of thumb and changing all the time.

That is just a rule of thumb, and

changing all the time
“The Theory of Everything” is
a likelihood for a fall festival
So bow. It has a biopic structure
and a potential crowd pleaser
appeal
Biopics with a
The rule that biopics with a
potential crowd
potential crowd pleaser appeal
pleaser appeal are
are more likely to be rewarded Against So
more likely to be
in Toronto than in Venice is not
rewarded in Toronto
totally reliable
than in Venice:
“The Theory of Everything”
is a likelihood for Toronto
rather than for Venice

3Taken from “The Fall Festival 50: Our Wishlist For The Venice, Telluride And Toronto Film Festi-
vals”. IndieWire 30/06/2014. Retrieved at https://www.indiewire.com/2014/06/the-fall-festival-50-
our-wishlist-for-the-venice-telluride-and-toronto-film-festivals-84321/.
11.5 Rebuttal 321

As an argumentative structure, warrant disclaimer is positively oriented towards

the conclusion of the undermined argument (“The Theory of Everything” is a
likelihood for Toronto rather than for Venice, in the case above).
Disclaimer by possible exception.4
A. Gandhi’s name would linger in the minds of humanity for several more centuries because
of his deep commitment and firm conviction towards a world free of nuclear warfare, peaceful
co-existence, law, order, human happiness and stability.

CA. This would be so unless we are graced with another incarnation of a Christ, Buddha or
Krishna.

Suppose that we were graced with

another incarnation of a Christ,
Buddha or Krishna
In that case
Gandhi would not linger in the
minds of humanity for several
more centuries, despite his deep
Against
Gandhi excelled by his deep
commitment and firm conviction
towards a world free of nuclear warfare,
peaceful co-existence, law, order, human
happiness and stabillity
If someone excells by
his/her deep
commitment and firm

commitment and firm conviction conviction towards

Therefore
in favor of a world free of nuclear human welfare, then
war, peaceful coexistence, law, he/she lingers in the
order, human happiness and minds of humanity for
stability. centuries:
Gandhi’s name would linger in the
minds of humanity for several more
centuries

To meet the argument, a merely possible circumstance that limits the strength of the
argument is pointed out; were it a real circumstance or at least a plausible one, a
rebuttal by exception would result. The difference lies in that rebuttal by exception is
neutrally oriented towards the conclusion of the rebutted argument, while disclaimer
by possible exception is positively oriented towards the same claim.

11.6 Refutation

As explained in Chap. 9, a prima facie reason is something that appears to be (i.e.

is presented as being) a reason, but may actually not be a reason at all. It follows
from the definition of arguing that any argument offers a prima facie reason for its

4Taken from “Mahatma Gandhi’s statue damaged in Patna” by Indo-Asian New Service. India.com
October 6, 2015. Retrieved at https://www.india.com/news/india/mahatma-gandhis-statue-dam
aged-in-patna-605633/.
322 11 Counterarguments

conclusion. Objecting to and rebutting are different ways of arguing that the alleged
reason is not actually a reason. Prima facie reasons should not be confused with pro
tanto reasons. A pro tanto reason is a consideration that counts in favor of some
position, and thus has genuine weight, but nonetheless may be outweighed by other
considerations (Kagan 1989, p.17n). A refutation is an attempt to show that a pro
tanto reason is outweighed by other consideration, so that it is inconclusive. Thus, it is
part of the pragmatics of refutation that its target be an argument that is acknowledged
to present a pro tanto reason.
There are three kinds of refutation, associated with the use and meaning of these
three the phrases: A but B, A but also B, and Although B, A.
In many cases when someone utters A but B she means that (1) she accepts A, (2)
she accepts B, (3) A is a reason for some conclusion C, (4) B is a reason for some
conclusion C’ incompatible with C, and (5) in the situation of utterance, B outweighs
A. When A but B is used in this way, the addressee is invited to infer C’ from the joint
consideration of A and B. In these cases, I will say that the argument B, therefore C’
is a contradicting refutation of the argument A, therefore C.
A. This patient has a streptococcal infection, so presumably this patient needs penicillin
treatment.

CA. But this patient is allergic to penicillin.

Conflating both arguments into a single argumentation, we get:

This patient has a streptococcal infection
Therefore
This patient needs penicillin treatment
but This patient is allergic to penicillin
< Therefore
No penicillin treatment should be
given to this patient

Contradicting refutation is positively oriented towards the conclusion of the stronger

argument and negatively oriented towards the conclusion of the weaker argument.
The utterance in the appropriate circumstances of A but also B carries commit-
ments similar to those in the previous case. Yet the implicature is now that A and B
are similar in force, so that they cancel each other out. When A but also B is a reply
to A, therefore C, the former is an attempt to suspend the inference of C from A. I
will speak then of cancelling refutation. Cancelling refutation is neutrally oriented
both towards the conclusion of the two balanced arguments.
A. Biofuels bring a number of environmental benefits, so production of biofuels should be
promoted.

CA. However, it is also true that their production can have some adverse effects on the
environment.
11.6 Refutation 323

Biofuels bring a number of Production of biofuels can have some

But also
environmental benefits adverse effects on the environment
Therefore ≃ Therefore
Production of biofuels should Production of biofuels should not be
be promoted promoted

The difference between but and although in their relevant uses has to do with
the suggested weighing. While but presents the second term as a stronger reason,
although flags the weaker term. When someone utters Although B, A in the appro-
priate circumstances, she conveys the indication that B is insufficient to defeat the
argument A, therefore C, so that, after weighing up A and B, the conclusion still
remains C. Some authors contend that even if B lacks the force to defeat A, it still
lowers the strength of the argument A, therefore C. Borrowing the term from Pollock
(2010, pp. 11–12), in such a case I will say that B, therefore C’ acts as a diminisher
for A, therefore C. Here is an example of diminisher5 :
A. It is impossible not to refer briefly to the Brahimi report and some of its main recommen-
dations. The report clearly contains guidelines that, even almost 10 years after they were
developed, need to be taken into consideration.

CA. Although the Brahimi report does not answer all of our questions.

The Brahimi report clearly contains

guidelines that, even almost 10 years The Brahimi report does not
Although
after they were developed, need to be answer all of our questions
taken into consideration
Therefore > Therefore
It is impossible not to refer briefly to the It is not necessary to refer to
Brahimi report and some of its main the Brahimi report in every
recommendations circumstance

Diminution is positively oriented towards the conclusion of the stronger argument

and negatively oriented towards the conclusion of the diminisher.

11.7 Weighing

Any refutation involves the weighing of pros and cons, since the conclusion is reached
from the joint examination of positive and negative considerations. This feature
relates refutation to Wellman’s third pattern of conduction.
The third pattern of conduction is that form of argument in which some conclusion is drawn
from both positive and negative considerations. In this pattern reasons against the conclusion
are included as well as reasons for it. For example "in spite of a certain dissonance, that piece

5United Nations, S/PV.6075, p.24. Retrieved at https://documents-dds-ny.un.org/doc/UNDOC/

PRO/N09/217/47/PDF/N0921747.pdf?OpenElement.
324 11 Counterarguments

of music is beautiful because of its dynamic quality and its final resolution" or "although
your lawn needs cutting, you ought to take your son to the movies because the picture is
ideal for children and will be gone by tomorrow." (Wellman, 1971, pp. 57).

Weighing is characteristic of refutation as a distinctive kind of counterargument.

While objection and rebuttal look for flaws or faults in the argument, refutation
starts from the acceptance of the argument under examination, as the analysis of
argumentative connectors like “but” and “although” shows. There is nothing wrong
with a refuted argument except that there is a stronger argument to the contrary.
The distinction of different kinds of counterarguments is important to limit the
scope of weighing. Some authors seem to consider weighing as an all pervasive
phenomenon in argumentation. Robert Pinto, for example, writes: “In short, we can
identify ‘exceptions’ to a qualified generalization only if we are already able to
compare the strength of arguments licensed by that generalization to certain other
arguments.” (2011, p. 117). Although this issue deserves a more detailed discussion,
the unbounded spreading of weighing promoted by such argumentation theorist as
Pinto is the result of mixing together the degree of acceptability of the premises and
the robustness or strength of the link between premises and conclusion. To borrow
an example overused in AI, there is no point in comparing (if it is possible at all) the
strength of the argument Tweety is a bird therefore Tweety flies since birds typically
fly with the strength of the rebutting argument Tweety was watched in Peninsula
Valdés, where there are important colonies of penguins, so probably it is a penguin.
In order to assess the later the sensible thing is to compare its strength with that of
Tweety was watched in Punta Loma, so it is more likely a rock shag than a penguin.
Of course, the possibility exists that the force of two arguments with compat-
ible premises be incomparable. So far as arguments are concerned, force or strength
defines in the best case a partial order. It might seem that whenever two such argu-
ments are conjoined it should be prudent to suspend judgment on the issue at stake.
This skeptical position can be contrasted with a credulous one, according to which
both conclusions are tenable, and in order to decide which argument is in force in
some particular argumentative situation one has to resort to procedural rules. I take
the terms “skeptical” and “credulous” from related discussions in AI (Prakken &
Vreeswijk 2002, § 4.1).
When someone refutes an argument, the opponent may ask him to justify why he
thinks the counterargument is stronger than the argument. This countermove gives
rise to specific sort of meta-argument, which I dub weighing meta-argument. Let us
consider an example.
A state lawmaker in Utah wants police to stop this practice [uploading DNA from crime
scenes to genetic databases or by obtaining warrants to search genealogy websites to help to
find criminals]. Legislation proposed by Rep. Craig Hall, a Republican, would prevent mass
searches of consumer DNA databases, which Hall referred to as “fishing expeditions.”

“We understand that law enforcement wants to use these tools, but the ends don’t justify the
means,” Hall said. “We don’t need a surveillance state to catch the bad guys.”
11.7 Weighing 325

Hall said he believes that law enforcement searches of DNA databases violate the particu-
larity requirement of the Fourth Amendment, which courts have interpreted as requiring law
enforcement to obtain search warrants and describe in detail to a judge the evidence they
plan to gather when invading someone’s privacy.

Emma Coleman, “One State May Become the First to Ban Law Enforcement Use of
Genealogy Databases”. Route Fifty 21/01/2020. Retrieved at https://www.routefifty.com/
public-safety/2020/01/utah-dna-databases/162544/

Craig Hall is balancing two arguments when he says that the ends don’t justify the
means. On one hand there is the argument from means to ends that the use of genetic
databases helps police to find criminals and therefore it should be allowed by law. On
the other hand there is the argument from values that the law enforcement searches of
DNA databases invade citizens’ privacy, therefore they should not be allowed by law.
So, we have reasons for and reasons against allowing it. Which should be given more
weight? Hall uses the principle that the ends don’t justify the means to support giving
greater weight to the second argument, so the conclusion that would be reached after
weighing up the two arguments is that law enforcement searches of DNA should be
banned. In this way, Hall constructs a weighting meta-argument whose conclusion is
that the argument from values has more strength than the competing argument from
means to ends.

The argument that the use of genetic databases helps police to find
criminals is a means to ends arguments, and the argument that law
enforcement searches of DNA databases invade citizens’ privacy is
an argument from values
The ends don’t justify
Therefore
the means:
Use of genetic databases Law enforcement searches
helps police to find of DNA databases invade
is weaker
criminals citizens’ privacy
than
Therefore Therefore
Police should not be
Police should be allowed
allowed to use genetic
to use genetic databases
databases

The conclusion of the whole argument is the comparative assessment between the
competing arguments, represented by the structure placed in the thick line box.
326 11 Counterarguments

11.8 Rebuttals, Refutations and Standards of Proof

It might be thought that sufficiency is a comparative notion to be explained in terms

of weighing. However, sometimes the stronger argument can also be rejected for
failure to comply with the standards of proof of the given dialogue. These standards
of proof have to do with acceptable risks and vary from one type of dialogue to the
other. Although perhaps this could be described as a situation in which the argument
has not the required weight, such cases cannot be classified as refutations for no
comparison of the strength of two opposite arguments is involved. Here is a putative
example (drawn from Collinson 2007)6 :
A. The wingbeat frequency of the bird in the Luneau video was measured at 8.6 beats s-1 ,
similar to that inferred from archival sound recording of a single Ivory-billed Woodpecker,
but claimed to be outside the range of Pileated Woodpeckers. So the woodpecker in flight
recorded by Luneau is an Ivory-billed Woodpecker.

CA. However, to regard the Luneau video by itself as presenting anything other than an
unidentified woodpecker falls below the standards of proof normally required for scientific
publication: the images are not good enough.

This is not a contradicting refutation since no reason is being offered to believe that
the woodpecker in flight recorded by Luneau is not an Ivory-billed Woodpecker. The
point is rather that given that the images are not good enough, they don’t provide,
according to the current standards of proof for scientific publications, a genuine
reason to accept that the woodpecker in flight is an Ivory-billed Woodpecker. So the
allegation that an argument fails to comply with the current standards of proof counts
as a case of rebuttal.

11.9 Counterargument Structures

The previous discussion brings to light that counterargumentation produces complex

argument structures, depicted by the previous diagrams. An argument structure can
be defined as a combination of arguments with an explicit argumentative orienta-
tion (positive, negative or neutral) towards some claim. The point is that when one
counterargues one combines the criticized and the criticizing argument into a single
argumentation. The join consideration of an argument A so C and an objection is
neutrally oriented towards C,7 as does its joint consideration with a plain rebuttal or
a cancelling refutation (marked by “but also”). By contrast, the conjoining of A so C

6 Collinson, J. Martin (2007). Video analysis of the escape flight of Pileated Woodpecker Dryocopus
pileatus: does the Ivory-billed Woodpecker Campephilus principalis persist in continental North
America? PMC. Retrieved February 3, 2019 from https://www.ncbi.nlm.nih.gov/pmc/articles/PMC
1838407/.
7 An objection to A, so C is neutrally oriented towards C insofar as it neutralizes the positive

orientation of A towards C. In such a case neither C nor not C can be concluded on this basis alone.
11.9 Counterargument Structures 327

Table 11.1 Orientation of

Argument Counterargument Joint orientation towards
counterargument structures
C
A so C Objection Neutral
Rebuttal by parallel Neutral
argument
Plain rebuttal Neutral
Exception Neutral
Disclaimer Positive
Contradicting Negative
refutation
Cancelling refutation Neutral
Diminisher Positive

with a contradicting refutation is negatively oriented towards C, while its conjoining

with a diminisher is positively oriented towards C (Table 11.1).
The recognition of counterargumentative structures is essential for understanding
an argumentative exchange as a joint construction of a macroargument.
This idea takes us directly from arguer dialectics to argument dialectics. Arguer
dialectics is dynamic while argument dialectics is static; static diagrams are to argu-
ment dialectics what dynamic profiles of dialogue are to arguer dialectics.8 Argument
dialectics is part of the theory of argument and, given that logic is concerned with
arguments as products, it is part of logic. Freeman argues that where a logical analysis
takes as its unit of analysis the individual argument, a dialectical analysis of argumen-
tation focuses on the whole argumentation, possibly containing multiple arguments
that need not form one single unified argument developed over the course of the
discussion (Freeman 2011, p. 109). But taking account the way counterarguments
logically relate to their target argument, it is also possible to analyze such a relation
as an additional kind of (counter)argumentative structure. It is like seeing just one
frame of a movie. Thus the standard catalogue of argument structures, which includes
subordinate, multiple and coordinate argumentation, needs to be supplemented with
counterargument structures: objection, rebuttal and refutation. Moreover, I claim that
there is a close relation between both kinds of structures.

11.10 The Order of Counterargumentation

In real argumentative practices argument criticism is seemingly conducted according

to a procedure that seeks to get the maximum effectiveness at the lowest cognitive
cost. For this reason it can be observed that priority is given to objection over rebuttal

8“A profile of dialogue is a connected sequence of moves and countermoves in a conversational

exchange of a type that is goal-directed and can be represented in a normative model of dialogue.”
(Walton 1999, p.53).
328 11 Counterarguments

and refutation, and then to rebuttal over refutation. Refutation can result in a costly
process of weighing pros and cons, the outcome of which is uncertain. Leibniz
said that the possession of a balance of reasons would be an even more important
achievement than the “fabulous science of producing gold”.
‘But’ is a typical counter-argumentation connector. As explained earlier, when
someone utters A but B she often means that (1) she accepts A, (2) she accepts B,
(3) A is a reason for some conclusion C, (4) B is a reason for some conclusion C’
incompatible with C, and (5) in the situation of utterance, B outweighs A. Thus,
in such cases the speaker admits that A is a pro tanto reason for C, although she
concludes C’. To admit that A is a pro tanto reason for C is tantamount to saying
that no objections or rebuttals are contemplated against argument A therefore C. This
shows that objection and challenge are prior to rebuttal.
To understand why rebuttal is a strategy more onerous than refutation, one just has
to think about the different functions that premises and warrants perform. Premises
are presented as data or facts while warrants are rules. Accordingly, premises are
appraised as true or false and warrants as more or less reliable. Toulmin (2003,
p. 98) puts it as follow: a general statement like “Philosophers in their intellectual
plenitude are unmarried” (according to Pierre Riffard 2004, 70% of philosophers
were unmarried when they published their masterpieces) can be used in two different
ways:
– as a datum, and then it can be expressed in the form “It has been observed that
philosophers were unmarried when they were in their intellectual plenitude”;
– as a warrant, and then it can be paraphrased “If a philosopher is in his intellectual
plenitude, he may be presumed to be unmarried”.
To demonstrate the falsity of a universal statement is easier than to demonstrate
that it is not a reliable guide to making inferences, if only because a false statement
is false in any circumstance, while a guide to inference can be reliable in some
circumstances and unreliable in others.
A pragmatic effect that confirms the effective operation of this cost-saving proce-
dure ordering counter-argumentation is that when someone attempts in the first place
to rebut or to refute an argument, this can be taken as a sign of her acceptance of the
premises.

References

Cohen, Daniel H. (2001). “Evaluating arguments and making meta-arguments”. Informal Logic
21/2, pp. 73-84.
Finocchiaro, Maurice (2013): Meta-argumentation. An Approach to Logic and Argumentation
Theory. London: College Publications.
Freeman, James B. (2011). Argument Structure: Representation and Theory. Dordrecht: Springer.
Laar, Jan-Albert van (2015). “Connection premises: Their character, criticism, and defence”. In
Cornelia Ilie & Giuliana Garzone, eds., Argumentation in real and virtual environments: Cross-
disciplinary perspectives, pp. 39–55. Amsterdam and Philadelphia: John Benjamins.
References 329

Pinto, Robert (2011). “Weighing Evidence in the Context of Conductive Reasoning”. In Blair, J.
Anthony y Johnson, Ralph H. (eds.), Conductive Argument: An Overlooked Type of Defeasible
Reasoning, pp. 104–126. London: College Publications, 2011.
Pollock, John L. (2010). “Defeasible Reasoning and Degrees of Justification”. Argument &
Computation, 1(1), pp. 7-22.
Posner (2003). E. A. (2003). Do States have a moral obligation to obey international law? Stanford
Law Review, 1906, 55(5), 1901–1919.
Prakken, Henry & Vreeswijk, Gerard (2002). “Logics for Defeasible Reasoning”. In Dov M.
Gabbay & Franz Guenthner, eds., Handbook of Philosophical Logic, (2nd edition), Vol. 4,
219–318. Dordrecht: Kluwer Academic Publishers.
Riffard, P. A. (2004). Les philosophes: vie intime. Paris: Presses Universitaires de France (PUF).
Toulmin, Stephen E. (2003 [1958]). The Uses of Argument, Cambridge: Cambridge University
Press.
Wellman, C. (1971). Challenge and Response: Justification in Ethics. Carbondale IL: Southern
Illinois University Press.
Walton, D. N. (1999). “Profiles of dialogue for evaluating arguments from ignorance”. Argumen-
tation 13, pp. 53–71.
Walton, D. N. (2013). Methods of argumentation. New York: Cambridge University Press.
Chapter 12
Co-Oriented Reasons and Modifiers

Abstract ‘Co-oriented reasons’ are reasons that can be adduced to support the same
claim, and ‘co-orientation’ is an operation combining two arguments with a common
conclusion. Co-orientation connectors include ‘also’, ‘even’, ‘on the one hand’, ‘on
the other hand’, among others. There are two main forms of co-orientation: conjunc-
tion and disjunction. Argument conjunction and disjunction are operations that join
two arguments to form a compound argument with the same conclusion. The differ-
ence between them can be apprehended by considering their resistance to counter-
argument. Reasons taken in disjunction stand of fall independently of one another,
whereas in conjunction reasons add up. Consequently, disjunction may be a reply
either to an objection or a rebuttal of an argument, while conjunction may be a reply
to a refutation.

12.1 Co-Oriented Reasons

We call “co-oriented reasons” those reasons that can be adduced to support the same
claim, and “co-orientation” an operation combining two arguments with a common
conclusion. Co-orientation connectors include “also”, “even" and “likewise”, “on
the one hand”, “on the other hand”, among others.
But with some policies, certain groups, such as health care professionals, may get higher
priority. This group includes not only workers who treat patients directly, but also support
staff, such as those who decontaminate rooms where a patient has been. “These are people
that we are putting at high risk for getting COVID,” [bioethicist Paul] Wolpe [PhD, director
of the Center for Ethics at Emory University in Atlanta] says.

There are two reasons for such policies. First is reciprocity. “This is what we owe them for
putting themselves in harm’s way so profoundly,” Wolpe says.

A second reason, he says: “It makes sense in a tie [between patients] to privilege a health
care worker because if we can get them better quickly, we can get them back onto the front
lines of helping us beat this.” Doing so would help stem shortages of such workers, he says.

Katherine Kam, “How Do You Choose Who Gets a Ventilator?”. Web MD 8/04/2020. https://
www.webmd.com/lung/news/20200408/how-do-you-choose-who-gets-a-ventilator

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 331
F. Leal and H. Marraud, How Philosophers Argue, Argumentation Library 41,
https://doi.org/10.1007/978-3-030-85368-6_12
332 12 Co-Oriented Reasons and Modifiers

Wolpe says that there are two reasons to support the claim that certain groups, such as
health care professionals, may get higher priority access to ventilators. The first is that
we owe them this priority for putting themselves in harm’s way so profoundly; the
second that we need health professionals to help us to beat COVID pandemic. Wolpe
indicates the warrant of the former argument when he says that it is a reciprocity
argument. The warrant of the latter argument is explicitly stated: if we can get them
better quickly, we can get them back onto the front lines of helping us beat this.
Therefore, these are co-oriented reasons, and “first” and “second” work here as
co-orientation connectors denoting an argumentative operation that combines the
argument.

Health-care professionals and support staff take

risks to care for others
Priority access to ventilators is what we owe
them for putting themselves in harm’s way Therefore
so profoundly:
Health-care professionals and support staff must
get priority access to ventilators

with the argument.

We need health professionals and support staff to

help us to beat COVID pandemic
If we can get health care worker better
quickly, we can get them back onto the front Therefore
lines of helping us beat this:
Health-care professionals and support staff must
get priority access to ventilators

into a single argument:

We need health
Health-care
professionals and
professionals and
moreover support staff to help us
support staff take risks
to beat COVID
to care for others
pandemic
Priority access to If we can get health care
ventilators is what worker better quickly,
we owe them for we can get them back
Therefore Therefore
putting themselves onto the front lines of
in harm’s way so helping us beat this:
profoundly:
Health-care Health-care
professionals and professionals and
support staff must get support staff must get
priority access to priority access to
ventilators ventilators
12.1 Co-Oriented Reasons 333

To simplify the co-orientation diagram, the conclusion can be written only once,
keeping the premises and the inferential steps (indicated by “therefore”) of the two
arguments separate, as shown below:

We need health
Health-care
professionals and
professionals and
moreover support staff to help us
support staff take risks
to beat COVID
to care for others
pandemic
Priority access to If we can get health care
ventilators is what worker better quickly,
we owe them for we can get them back
Therefore Therefore
putting themselves onto the front lines of
in harm’s way so helping us beat this:
profoundly:
Health-care professionals must get priority access to ventilators

This raises the question of how to distinguish single-reason-many-premises argu-

ments from many-reasons arguments. A premise is a statement that expresses, alone
or with others, a reason for something, whereas A single reason is the smallest amount
of information that by itself lends some measure of credence to a position.” (Blair
2012, p. 148). Premises working together to express a single reason are known as
copremises. Argumentative connectors like “moreover” or “besides” provide a clue
because they are used to combine different reasons and not just different premises.1

(1) This cake contains chocolate and Mary is allergic to chocolate, so she should
refrain from tasting it.
In a normal context, this a single reason two premises argument; this is why
the following paraphrase sounds weird:
(2) This cake contains chocolate, besides Mary is allergic to chocolate, so she
should refrain from tasting it*.
By contrast “besides” can be soundly inserted into a many-reasons argument
like.
(3) This coat costs more than I intended to spend, and it is not exactly what I was
looking for; so, I should not buy it yielding
(4) This coat costs more than I intended to spend, besides it is not exactly what I
was looking for; so, I should not buy it.

1 Actually, things are a bit more complicated than this. David Hitchcock points out that “There is also
an interpretive difficulty in determining whether an additional supporting reason introduced by a
bridging term like ‘besides’ or ‘moreover’ or ‘further’ is a new argument or merely an independently
relevant part of a single argument (Hitchcock 2017, p. 24). My full answer is that ‘besides’ and
the like can introduce either a new argument or a modifier, whose insertion in the given argument
produces a modified reason.
334 12 Co-Oriented Reasons and Modifiers

12.2 Conjunction and Disjunction of Arguments

There are two main forms of co-orientation: conjunction and disjunction. In logic
conjunction and disjunction are sentential operations that join two sentences to form a
compound sentence. Likewise, argument conjunction and disjunction are operations
that join two arguments with a common conclusion to form a compound argument
with the same conclusion. The difference between them can be apprehended by
considering their resistance to counterargument.
(a) If the counterargument is an objection or a rebuttal, a conjunction of two argu-
ments resists it if and only if both subarguments can be successfully defended
against that counterargument.
(b) If the counterargument is an objection or a rebuttal, a disjunction of two argu-
ments resists it if and only if at least one of the two subarguments can be
successfully defended against that counterargument.
This is analogous to the truth functional behavior of sentential conjunction
and disjunction, since the conjunction of two sentences is true if and only if
both sentences are true, whereas the disjunction of two sentences is true if and
only if at least one of these sentences is true.2
(c) A disjunction resists a refutation if and only if at least one of the two
subarguments withstands it.
(d) By contrast, a conjunction can resist a refutation even if none of the conjoined
argument separately does it.
To sum up, reasons taken in disjunction stand of fall independently of one another,
whereas in conjunction reasons add up.3 Consequently, disjunction is a natural
response both to an objection or a rebuttal of an argument, while conjunction is
naturally in order when it comes to defuse a refutation. In this sense disjunctive argu-
mentation is associated with objection and rebuttal and conjunctive argumentation
with refutation.
Although it is often difficult to determine whether we are dealing with a conjunc-
tion or a disjunction of arguments, the text may contain explicit indications in this
regard. Van Eemeren, Houtlosser and Snoeck-Henkemans (2007, pp. 200–219) make
an extensive list of indicators of disjunctive argumentation (for them, multiple argu-
mentation) and of conjunctive argumentation (for them, coordinated argumentation).
Among the first are: “if only because”, “anyway”, “anyhow”, “even if”, “whatever
the case may be”, “in any case”, “leaving aside that”, “not to mention that”, “also
because”, “at least as important”, “in the first place …, in the second place…”,
“to begin with”, “further”, “moreover”, etc. And among the second: “an additional

2 In set-theoretical terms, it could be said that the set of objections (rebuttals) of a conjunctive
argument is the union of the sets of objections (rebuttals) of the conjoined arguments. Likewise, the
set of objections (rebuttals) of a disjunctive argument is the intersection of the sets of objections
(rebuttals) of the disjoined arguments.
3 However, this metaphor should not be taken too far, since reasons are not quantities to be added

or subtracted, multiplied and divided.

12.2 Conjunction and Disjunction of Arguments 335

reason is”, “less importantly”, “the more so since”, “more importantly”, “and also”,
“decides”, “in addition to”, “together with”, “moreover”, “as well as”, “while”, “and
yet”, etc. Van Eemeren, Houtlosser and Snoeck-Henkemans are careful to warn that
“With the exception of very explicit cases, the indicators never provide certainty
about the structure of the argumentation.” (Op.cit., p.221), and their lists have been
convincingly criticized (see, e.g. Hitchcock 2003). I think that this issue deserves
a more systematic empirical study. Although the study of argumentative connec-
tors and discourse markers is a very active area of research in linguistics, however
such categories as conjunctive (or coordinative) argumentation and disjunctive (or
multiple) argumentation are not in the toolkit of linguists.
The difficulty in finding reliable textual criteria for conjunctive and disjunctive
argumentation is compatible with the existence of more or less clear-cut examples
of these structures.
Almost all Nietzsche scholars agree that the doctrine of eternal recurrence advances the view
that, in fact, everything eternally recurs, and that we ought to behave as if it does. Clearly the
first part is a descriptive hypothesis, the second is prescriptive. […] The empirical hypothesis
should be assimilated to, and understood as occasioned by, the prescriptive dimension of
the doctrine of eternal recurrence. The principal textual reason for suggesting this change in
emphasis is that Nietzsche elected to publish the non-cosmological arguments while, at the
same time, withholding the empirical formulation. A complementary reason for suggesting
this procedure is that the unpublished empirical arguments are no less ambiguous and subject
to interpretation, in my judgement, than are the non-cosmological formulations. The method-
ology advance, then, takes the published position as the principal source of evidence and the
Nachlass as an aid to clarifying that position.

(Bernd Magnus, Heidegger’s Metahistory of Philosophy: Amor Fati, Being and Truth, p. 7.
The Hague: Martinus Nijhoff 1979).

Magnus gives two reasons of unequal strength to subordinate the empirical dimension
of eternal recurrence to the prescriptive one. The principal reason is that Nietzsche
decided to publish the non-cosmological arguments while hiding the empirical argu-
ments. The complementary reason is that the unpublished empirical arguments are
no less ambiguous and in need of interpretation than are the non-cosmological ones.
Hence the complementary reason serves to fill out or complete the principal one
in order to achieve the strength required to support the claim that the prescriptive
dimension of eternal recurrence is prior to the empirical one. Thus we are faced here
with a salient example of conjunctive argumentative.

Nietzsche elected to publish the The unpublished empirical arguments

non-cosmological arguments while, are no less ambiguous and subject to
and
at the same time, withholding the interpretation, in my judgement, than are
empirical formulation the non-cosmological formulations
Therefore Therefore
The empirical hypothesis should be assimilated to, and understood as occasioned by,
the prescriptive dimension of the doctrine of eternal recurrence
336 12 Co-Oriented Reasons and Modifiers

Let us consider next an example of disjunctive argumentation.

His [Alan Dershowitz’s] claim—central to Trump’s defense—is that “purely noncriminal
conduct, including abuse of power and obstruction of Congress, are outside the range of
impeachable offenses.” Although he says a “technical” crime is not required, he means only
that criminal conduct can still be impeachable even if it cannot be prosecuted for some
jurisdictional or procedural reason. […] Perhaps recognizing the weakness of his position,
Dershowitz rests heavily on an alternative argument: Even if criminal-like conduct is not
required for impeachment and removal, the articles of impeachment that charge abuse of
power and obstruction of Congress are deficient because they employ “vague, open-ended
and entirely subjective” criteria.

Justin Amash, “Dershowitz’s view would destroy our constitutional system”. The Washington
Post 31/01/2020. Retrieved at https://www.washingtonpost.com/outlook/2020/01/31/justin-
amash-dershowitzs-view-would-destroy-our-constitutional-system/

According to Amash’s interpretation, Dershowitz advances two reasons why Trump

should not be impeached. A first clue for thinking that his is a disjunctive argumen-
tation is that the second is presented as “an alternative argument” (i.e., as they are
intended to be alternative defenses of the same standpoint, as pragma-dialecticians
like to say). This clue is reinforced by the subsequent occurrence of “even if”, one
of the non-univocal indicators of multiple argumentation listed by van Eemeren,
Houtlosser and Snoeck-Henkemans.

The articles of impeachment that charge

Purely noncriminal conduct,
abuse of power and obstruction of
including abuse of power and
or Congress are deficient because they
obstruction of Congress, are outside
employ vague, open-ended and entirely
the range of impeachable offenses
subjective criteria
Therefore Therefore
Trump cannot be impeached for abuse of power

12.3 Modifiers

When it comes to analyze how groups of statements can support a thesis, one must
also pay attention to presuppositions and factors that can modify a previously stated
reason. In the theory of reasons holism is the claim that contexts differ in terms of
whether a certain consideration constitutes a reason at all, as well as in terms of
the weight, and possibly even orientation, of the reason. Therefore, any reason —
and hence any argument— has to be evaluated in the context of all relevant reasons
that apply in a given situation. To explain how reasons can vary across contexts,
Ralph Bader (2016), drawing upon the work of Jonathan Dancy (2004), introduces a
distinction in the necessitation base (that which explains and necessitates the reason)
between the source or the ground of a reason, the conditions of the reasons and the
12.3 Modifiers 337

modifiers of the reason. Bader gives two non-equivalent definitions of the source or
ground of a reason:
– “that in virtue of which something is a reason — the source of the reason”;
– “the source or ground of a reason is to be identified with the consideration that
constitutes the reason” (Bader 2016, pp. 27 and 31).

The source or ground in the first sense is the warrant in the Toulmin model, while
in the second it corresponds to the data. Henceforth I will reserve the term “ground
of a reason” to designate the consideration that constitutes the reason, using “warrant
of a reason” to designate that which makes something a reason for something else.
The conditions of the reason are that on condition of which something is a reason.
There are two kinds of conditions: enablers and disablers. If the conditions are satis-
fied, which can consist in the presence of enablers or the absence of disablers, then
the ground does constitute a reason. Otherwise, if the relevant enablers are absent or
disablers are present, then it will fail to do so. Finally, the modifiers of the reason are
considerations that affect the weight of a reason. There are also two kind of modifiers:
intensifiers and attenuators. Intensifiers are facts that make the weight of some reason
greater without themselves being reasons, while attenuators are facts that make the
weight of some reason weaker without themselves being reasons. Transvaluators are
an extreme kind of attenuators that bring about a switch of polarity, so that while A
counts as a reason for C, A together with the transvaluator B provides a reason for
not C. I will not discuss here transvaluators.4 However, here is, perhaps, an example
of transvaluation.
Not everyone is convinced that [American novelist Jonathan] Franzen is worthy of his blanket
media coverage, however: […] At this year’s Edinburgh Book Festival, authors David Shields
and Eli Horowitz criticised Franzen for his backward-looking, 19th-century writing style.
Both agreed that in a multimedia world, the “antediluvian” novel needs to be invigorated by
experimentation.

But his rivals’ argument disregards what many television viewers, for example, adore –
dramas that resemble classic fiction in their length, breadth and depth: The Wire, The
Sopranos. Franzen’s 19th-century-style novels have a far better chance than his detractors’
of surviving into the 22nd.

Tim Walker, “Jonathan Franzen: Greatness, one sentence at a time”. The Independent
28/08/2010. Retrieved at https://www.independent.co.uk/news/people/profiles/jonathan-fra
nzen-greatness-one-sentence-at-a-time-2064056.html

David Shields and Eli Horowitz’s argument can be represented for present purposes
as follows:

4 I contend, against Bader, that transvaluators cannot be a special kind of attenuators. It is part of
the identity of a reason to be a reason for something. When we take a consideration as a reason for
two different things, it would be weird to say that we are examining the same reason. The fact that
something is very expensive is a reason to consider it a luxury product and a reason to decide not to
buy it, but it would be abusive to say that in both cases we are dealing with the same reason, even
if it is obviously the same consideration.
338 12 Co-Oriented Reasons and Modifiers

Franzen’s writing-style is backward-looking, 19th-century

Therefore
Franzen is not worthy of his blanket media coverage

Walker accuses them of disregarding that many television viewers adore dramas
that resemble classic fiction in their length, breadth and depth, as shown by the
success of TV series like The Wire and The Sopranos. This fact is not by itself
a reason to believe that Franzen is not worthy of his blanket media coverage. But
coupled with the fact that Franzen’s writing-style is backward-looking, 19th-century
provides a reason to think that Franzen is worthy of his blanket media coverage.

Franzen’s writing-style is backward-looking, 19th-century. Many television viewers

adore dramas that resemble classic fiction in their length, breadth and depth
Therefore
Franzen is not worthy of his blanket media coverage

If this analysis is right, the fact that many television viewers adore dramas that
resemble classic fiction in their length, breadth and depth acts as a transvaluator with
respect to the critics’ argument.
1. I promised to do it.
2. My promise was not given under duress.
3. I am able to do it.
4. Doing it would not be too costly for me.
5. So: I ought to do it.
6. Since: We ought to keep our promises.

1 is the ground for claim 5. If, given the argument 1 so 5 someone asks: “How do
you get there?” a possible answer would be 6. Hence 6 is the warrant explaining why
1 is relevant to 5. Neither 2 nor 3 are by themselves reasons for doing the action. 2
is an enabling condition consisting in the absence of a disabler, and 3 is an enabling
condition consisting in the presence of an enabler, and hence of a different type as 2.
Nor is the mere fact that doing it would not be too costly for me a reason for doing
that action. Even if it were costly for me doing what I promised, I would still have
a reason for doing that, a reason given by my promise to do so. But 4 does make a
rational difference, all the same. What it does is to intensify the reason given me by
1. Instead of two reasons, according to Bader, what we have here is one ground or
source 1 and one intensifier 4.
Given the close relationship of reasons to argument, the Bader-Dancy model of
the necessitation base of a reason can be converted into a holistic model of argument,
alternative to both the traditional premises-conclusion model and the Toulmin model.
Modifiers are the main novelty of this holistic model.
Conditions are somehow present in Toulmin’s model, as shown below. Conditions
come in two varieties: enablers and disablers. An enabler is a fact that must obtain
for some other fact to count as a reason. Disablers, in turn, are facts that, when
12.3 Modifiers 339

they obtain, prevent some other fact from constituting a reason. Disabling conditions
are closely related to Stephen Toulmin’s conditions of exception or rebuttal, as the
following passage makes it clear.
The special force of the qualifier used in this second type of case (presumably) is directly
connected with the idea of rebuttals. It registers the fact that the inference is warranted-
that the claim is directly supported by the grounds-only in the absence of some particular
exceptional condition, which would undercut (i.e., withdraw the authority of the warrant for)
the inference. (Toulmin et al. 1984, p. 96).

Let us consider another example (similar to the one in Toulmin et al. 1984, pp. 96–98).

1. This patient has a streptococcal infection.

2. So: this patient needs penicillin treatment.
3. Since penicillin may be safely and effectively prescribed for streptococcal
infections.
4. But this patient is allergic to penicillin.

1 is the ground or source of the reason, 2 is the claim, 3 is the warrant making
the ground 1 relevant to the claim 2, and 4 is a disabler or condition of rebuttal, as
explicitly pointed in Toulmin, Rieke and Janik (1984). On the holistic account, when
the patient is allergic to penicillin, the fact that she has a streptococcal infection is no
longer a reason for penicillin prescription. The Toulminian explanation is that since
it is the warrant what makes the datum or ground a reason for the claim, if some
exceptional condition withdraws the authority of the warrant, no reason remains.
Although it is more difficult to find something analogous to enabling conditions
on the Toulmin model, Toulmin, Rieke & Janik describe the process through which
disabling conditions give way to enabling conditions:
… if the situation develops to a point at which no standing presumption can be securely
established in the first place, there will be nothing left to rebut. Instead, we shall then have
to work with two alternative parallel arguments and apply one or the other of them in any
particular case, depending upon which of the alternative conditions holds good: "On the one
hand, if the patient is not allergic to it, penicillin may be safely and effectively prescribed for
upper respiratory infections. On the other hand, in cases of penicillin sensitivity, some other
broad-spectrum antibiotic should be prescribed, such as tetracycline." (Toulmin, Rieke &
Janik 1984, p.99).

That is, where the “exceptions” are not truly exceptional, we cannot present the
conclusions of our arguments as being “presumably” sound, subject only to a possible
rebuttal. Instead we do better to restate our warrants, explicitly, as holding good only
on condition that certain specific conditions are satisfied. (Ibid.).
Thus, when exceptions become frequent, the above argument must be reformu-
lated as follows, incorporating an enabling condition 4:
1. This patient has a streptococcal infection.
2. So: this patient needs penicillin treatment.
3. Since penicillin may be safely and effectively prescribed for streptococcal
infections.
340 12 Co-Oriented Reasons and Modifiers

4. Given that this patient is not allergic to penicillin.

Though the notion of modifier belongs to the theory of reasons, we can find some
inklings of it in argumentation theory, in particular in Pinto and Blair’s notion of
complementary premises and Mark Vorobej’s notion of supplementation. Douglas
Walton (1996, p. 133–134) tells us that Pinto and Blair, in unpublished work cited
by Snoeck Henkemans (1992), define a complementary argument as one in which
“Some premises complete others”. To illustrate what Pinto and Blair mean by “some
premises completing others”, Walton (op.cit., p. 134) cites the following argument:
1. I promised my girlfriend I’d take her to see the latest Woody Allen movie tonight.
2. She’ll be really disappointed if I don’t go to that movie with her.
3. I don’t have any excuse for not doing so.
4. So: I guess I should take her to see that movie tonight.
Pinto and Blair point out that 1 provides by itself a reason for the conclusion that
I should take her to see that movie tonight. Therefore, it is not a standard single-
reason-many-premises argument. By contrast, neither 2 nor 3 directly support the
conclusion, and so it is not a many-reasons argument either. Thus, they conclude,
this is a different argument structure: a complementary argument. In the terminology
of the holistic model, 1 is the ground or source of the reason, 2 is an intensifier
and 3 is an enabler. In Blair and Pinto’s words, as reported by Walton, premises 2
and 3 “complete” premise 1, increasing the amount of support. Hence modifiers and
conditions are complementary premises, and a complementary argument seems to
be an argument in which some modifier or condition occurs.
Vorobej’s uses supplementation instead of complementation, but the basic idea
remains the same. For the simplest case of a two premises argument, Vorobej (1995,
p. 292; see also 2006) gives the following definition of supplementation:
A premise P supplements a premise Q iff.
(a) P is not relevant to C,
(b) Q is relevant to C, and
(c) {P,Q} offers an additional reason R in support of C, which Q alone does not
provide.
Vorobej’s example is:
1. I’ve seen all the ducks on the pond.
2. All the ducks that I’ve seen on the pond are yellow.
3. So: All the ducks on the pond are yellow.
The fact that I have seen all the ducks on the pond is not a reason for itself to think
that all the ducks on the pond are yellow; on the contrary the fact that all the ducks
that I’ve seen on the pond are yellow is a consideration that counts in favor of the
claim that all the ducks on the pond are yellow. In a standard single reason-many-
premises argument, the premises only provide a reason for the conclusion when they
are all taken together. Thus this is neither a standard single-reason-many-premises
argument nor a many-reasons argument. In Vorobej terminology this is a “hybrid
12.3 Modifiers 341

argument” (Vorobej 1995, p. 293). In the terminology of the holistic model, 1 is the
ground while the supplementary premise 2 turns out to be a modifier. In general,
since modifiers by themselves are not reasons, they are supplementary premises, and
thus any argument offering a modified reason will count as a hybrid argument.

12.4 Modified Arguments

I call “modification” the operation that consists in adding a modifier to the premises of
a previously given argument. In this section I will address the following questions:
Are modifiers premises of a special sort, as Vorobej, Pinto and Blair contend? Is
modification simply another name for complementation (Pinto and Blair) or supple-
mentation (Vorobej)? Or is it a distinct form of co-orientation? If modifiers were
some sort of premises, modification would be an operation that applied to a single-
reason-single-premise argument yields a single-reason-many-premises argument. If,
on the contrary, modification were a kind of co-orientation, it would be a function that
applied to a single-reason-single-premise argument yields a many-reasons argument.
When modification occurs, Bader says, there are two distinct reasons: the unmodi-
fied reason and the modified reason (Bader 2016, p. 15). On the plausible assumption
that two arguments offering different reasons for the same claim are indeed different,
Bader’s account implies that whenever modification occurs a new argument comes
into existence. Thus, we will have to distinguish the “unmodified” argument,

A John promised to do it
Therefore
John ought to do it

from the “modified” argument:

B John promised to do it; doing it would not be too costly for John
Therefore
John ought to do it

Are A and B two different arguments? On the assumption that extensionally argu-
ments are individuated in terms of their constituents alone, I can think of three
possible answers to that question.

(a) The intensifier “Doing it would not be too costly for me” is a premise of B but
not a premise of A; therefore, we are confronted with two different arguments.
342 12 Co-Oriented Reasons and Modifiers

(b) A and B are the same argument, inasmuch as the intensifier “Doing it would
not be too costly for me is an implicit element” (a hidden premise) in A.
(c) A and B are one and the same argument, since the intensifier “Doing it would
not be too costly for me” is not properly speaking a constituent of B.

Bader’s talk of modified and unmodified reasons seems to favor interpretation (a).
This answer is not without difficulty. What is the relation between the relative weights
of A and B? Given that intensifiers “are facts that make the weight of some reason
greater without themselves being reasons” (Lord & Maguire 2016, p. 11), it follows
from (a) that the second argument is stronger than the first. Actually Bader emphasizes
that “the modified reason and the unmodified reason are of the same type and that
they are accordingly commensurable”, and that “in the case of modification there
is a distinction between the modified and the unmodified reason, which ensures
that when a 0-multiplieris applied the unmodified reason remains and it is only the
modified reason that is reduced to zero” (Bader 2016, pp. 41 and 37, respectively).
This suggests that the modified and the unmodified argument form some sort of
disjunction of reasons of different strength:

John promised to do it; doing it

John promised to do it or
would not be too costly for John
Therefore Therefore
John ought to do it John ought to do it

However, that would be a mistake. If we take holism seriously, arguments are to

be evaluated in context. In a situation in which doing the promised thing is not too
costly for me, both arguments have the same force or strength; otherwise, the second
argument would have an unacceptable premise, in which case strength comparison
would be pointless. We need to be careful to distinguish intensifiers-1, i.e. facts that
strengthen a reason, from identifiers-2, sentences expressing intensifiers-1. It is the
fact that doing the promised would not be too costly for me what makes greater the
weight of A, not the addition of the identifier-2 “Doing it would not be too costly for
me”.
The need to distinguishing intensifiers-1 from identifiers-2 is even clearer when
we consider attenuators. Here is another example from Dancy (2004: 42):
1. She is in trouble and needs help.
2. It is all her own fault, and she got in this situation through trying to spite someone
else.
3. But still: I must help her.
Here 1 is the ground, 3 the claim and 2 is an attenuator. Correspondingly we have
two distinct arguments:

C. She is in trouble and needs help; so I must help her

12.4 Modified Arguments 343

D. She is in trouble and needs help. It is all her own fault, and she got in this
situation through trying to spite someone else; but still, I must help her
The idea that inclusion of attenuator in D would make this argument weaker than C
clashes with the intuition that, in a given situation, the stronger argument is the one
that should guide our decision. I assume that, in a situation in which it is a fact that it
she got in this situation through trying to spite someone else, the guiding argument
will be D, supposedly the weaker argument.
Bader writes on the differences between the unmodified reason and the modified
reason:
These reasons can play different roles. The unmodified reason can be evaluated by abstracting
from the context, whereas the modified reason is the one that we identify when assessing the
consideration in question within the context in which it is to be found. (Bader 2016: 40)-

It seems now that the single reason I promised to do it can be evaluated either
abstracting from the context or within the context in which it occurs. The intensifier
“Doing it would not be too costly for me” is therefore a contextual factor in the
appraisal of this unique reason. Since “contexts consist in the extrinsic circumstances
in which a consideration is to be found” (Bader 2016: 27), such contextual factors are
not part of the reason itself, and the talk of the modified reason and the unmodified
reason should not be taken literally. These refer, rather, to different ways of assessing
the same reason.
The fact that modifiers refer to extrinsic circumstances or contextual factors in the
evaluation of a reason or argument also justifies to dismiss answer (b), that modifiers
are hidden premises. Moreover, an obvious shortcoming of the strategy of treating
contextual factors relevant to the assessment of an argument as hidden premises is
that it may be impossible to specify all the relevant factors that determine in any
situation whether a consideration is a reason and, if so, what its strength is. This
strategy will lead us to fill out the argument with the following hidden premises:

John promised to do it. John was not telling a joke or acting in a play . His
promise was not given under duress. John is able to do it. Doing it would not
be too costly for John. The promisee prefers John’s doing it to his not doing it
Therefore
John ought to do it

And with a little ingenuity it is possible to add further premises (e.g., John was
not drugged or drunk), so that the number of premises of an argument becomes
indefinitely large.5

5 “Naturally it would be a waste of time and energy to bother recording these wholly general
presuppositions explicitly in every actual case by writing them into the particular warrants on
which our practical argumentation relies. Nor may it in fact be practicable to enumerate all these
general assumptions exhaustively in advance of encountering the very rare exceptions that bring
them to light.” (Toulmin, Rieke & Janik 1984, p. 100).
344 12 Co-Oriented Reasons and Modifiers

Thus, the only available option is (c):

(c) A and B are one and the same argument, since the intensifier “Doing it would not be too
costly for me” is not properly speaking an element of A.

Modifiers and (conditions) are not constituents of an argument but contextual factors
that play a relevant role in the appraisal of the reason offered by the argument.
Modifiers, in particular, are weighing factors that are in order when there is a threat
of refutation. Modifiers are often involved in weighing meta-arguments, and therefore
rather than constituents of an argument, they are elements in a comparison between
arguments.
Let us consider an example from Robert Alexy (2005) to illustrate the working
of modifiers. A German satirical magazine, Titanic, described a paraplegic reserve
officer, first, as a “born murderer” and then, in a later edition, as a “cripple.” A
German court ruled against Titanic and ordered the magazine to pay damages to the
officer. Titanic brought a constitutional complaint. The Federal Constitutional Court
undertook a balance between the freedom of expression of the magazine and the
officer’s general right to personality. The Federal Constitutional Court considered
the intensity of infringement with the freedom of expression in the Titanic judgment
as serious. Concerning the epithet “born murderer”, the Court considered the impor-
tance of satisfying the right to personality of the officer as moderate, given its highly
satirical context, while the description of the officer as “cripple” was considered as
a serious. This gave rise to a stalemate, with the consequence that Titanic’s constitu-
tional complaint was not successful in so far as it related to damages stemming from
the description “cripple.”.
Focusing only in the first part, the Federal Constitutional Court was faced with
two arguments.
• Describing the officer as a “born murderer” Titanic has infringed upon his right
to personality; therefore Titanic must be condemned.
• Describing the officer as a “born murderer” Titanic has exercised its freedom of
expression; therefore Titanic must be acquitted.
The Federal Constitutional Court held that the second argument outweighs the
first, and supported this claim with two modifiers:
(a) The intensity of infringement with the freedom of expression in the Titanic
judgment is serious;
(b) The intensity of infringement with the right to personality of the epithet “born
murderer” is moderate.
(a) works as an intensifier while (b) works as an attenuator; notice however that
if in the first case the intensity of infringement were considered light, B would work
instead as an intensifier. This stresses the contextual character of modifiers.
12.4 Modified Arguments 345

The intensity of infringement with the freedom of expression in the

Titanic judgment is serious. The intensity of infringement with the
right to personality of the epithet “born murderer” is moderate.
Therefore
Describing the officer as a Describing the officer as
“born murderer” Titanic is weaker a “born murderer”
has infringed upon his than Titanic has exercised its
right to personality freedom of expression
Therefore Therefore
Titanic must be
Titanic must be acquitted
condemned

References

Alexy, Robert (2005). “Balancing, constitutional review, and representation”. International Journal
of Constitutional Law 3(4): 572–581.
Bader, Ralph (2016). “Conditions, Modifiers and Holism”. In Lord, Errol & Maguire, Barry, eds.,
2016, pp. 27-55.
Blair, J. Anthony (2012). Groundwork in the Theory of Argumentation. Dordrecht, Heidelberg,
London and New York: Springer.
Dancy, Jonathan (2004). Ethics without Principles. Oxford: Oxford University Press.
Eemeren, Frans H. van, Houtlosser, Peter, & Snoeck Henkemans, A.Francisca (2007). Argumenta-
tive Indicators in Discourse. A Pragma-dialectical Study. Dordrecht: Springer.
Hitchcock, David (2003). “Commentary on Snoeck Henkemans” (May 14, 2003). OSSA Confer-
ence Archive. Paper 84. http://scholar.uwindsor.ca/ossaarchive/OSSA5/papersandcommentar
ies/84
Hitchcock, David (2017). On Reasoning and Argument: Essays in Informal Logic and on Critical
Thinking. Dordrecht, Heidelberg, London and New York: Springer.
Lord, Errol & Maguire, Barry, eds. (2016). Weighing Reasons. New York: Oxford University Press.
Snoeck-Henkemans, A. Francisca (1992). Analysing Complex Argumentation. Amsterdam:
SICSAT.
Toulmin, Stephen E., Rieke, Richard & Janik, Allan (1984 [1978]). An Introduction to Reasoning.
New York: McMillan.
Vorobej, Mark (1995), “Hybrid Arguments”.Informal Logic 17, 289–296.
Vorobej, Mark (2006). A Theory of Argument. Cambridge: Cambridge University Press.
Walton, Douglas N. (1996). Argument Structure: A Pragmatic Theory. Toronto, ON: University of
Toronto Press.
Chapter 13
Intertwined Structures

Abstract Argumentative structures are photographs of argumentative processes that

capture the macro-arguments that participants construct interactively in the course
of the argumentative exchange. We might say, then, that argumentative structures are
the static counterpart of the dynamic dialectical profiles, as defined by van Eemeren
(2010, p. 98). Since dialectical profiles are made of moves and countermoves, oppo-
sitional relationships between arguments are the key to understanding argumentative
structures. Raising an objection, rebutting an argument and refuting an argument are
moves in an argumentative exchange. Every countermove amounts to the application
of an operation that produces a new argumentative structure. In this way, an abstract
structure formed by arguments and counter-arguments can be extracted from the
argumentative process.

13.1 Argumentative Structures and Dialectical Profiles

Argumentative structures, as defined in Chap. 10, are photographs of argumenta-

tive processes that capture the intrincate macroarguments that participants construct
interactively in the course of the argumentative exchange. We might say, then, that
argumentative structures are the static counterpart of the dynamic dialectical profiles,
as defined by van Eemeren:
A dialectical profile is defined as a sequential pattern of the moves the participants in a
critical discussion are entitled to make – and in one way or another have to make – to realize
a particular dialectical aim at a particular stage or sub-stage of the resolution process (van
Eemeren 2010, p. 98).

Since dialectical profiles are made of moves and countermoves, oppositional rela-
tionships between arguments are the key to understanding argumentative structures.
Raising an objection, rebutting an argument and refuting an argument are moves
in an argumentative exchange. These moves correspond to the three main kinds of
oppositional operations described in the preceding chapter: objection, rebuttal and
refutation (letting aside dismissal). Every countermove amounts to the application
of an operation that produces a new argumentative structure. In this way, an abstract

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 347
F. Leal and H. Marraud, How Philosophers Argue, Argumentation Library 41,
https://doi.org/10.1007/978-3-030-85368-6_13
348 13 Intertwined Structures

Move Operation / Countermove Operation Structure

Structure
Counterargue Objection, Offering an Disjunction Disjunctive
(raising an Rebuttal, alternative reason argumentation
objection; rebutting Refutation
an argument;
refuting an
argument)
Raising an objection Objection Supporting the Concatenation Subordinative
contentious premise argumentation
Rebutting an Rebuttal Backing the warrant Backing Underpinned
argument argumentation
Refuting an Refutation Reinforce with Conjunction Coordinative
argument another reason argumentation

Fig. 13.1 Moves and structures

structure formed by arguments and counter-arguments can be extracted from the argu-
mentative process. Figure 13.1 summarizes the game of moves and countermoves
and argument structures.
Concatenation and backing were described in Chap. 10, the different kinds of
counter-argumentation in Chap. 11 and the remaining operations of conjunction and
22 disjunction in Chap. 12.
We can illustrate the intertwining of argumentative operations and moves with
a couple of variations of a well-known example from The Uses of Argument. The
exchange starts with a simple argument:
(1) Proponent: Petersen is a Swede, so he is presumably not a Roman Catholic.
Then the dialogue runs as follows:
(2) Opponent: But Spanish is his native tongue, so he’s probably not a Swede.
(3) Proponent: In any case he has a Swedish passport.
Move (2) is an attempt to raise an objection against argument (1) that generates the
following structure:

Spanish is Petersen’s native tongue

Therefore
Petersen is probably not a Swede against Petersen is a Swede
Therefore
Petersen is presumably not a Roman Catholic

Move (3) could be described either is an attempt to support the challenged premise
that Petersen is a Swede applying concatenation or as an attempt to refute the
objection.
13.1 Argumentative Structures and Dialectical Profiles 349

Spanish is Petersen’s native tongue but Petersen has a Swedish passport

Therefore Therefore
Petersen is probably not a Swede against Petersen is a Swede
Therefore
Petersen is probably not a Roman
Catholic

On the right side of the diagram we have a subordinative argument, at the top a
refutation and at the bottom an objection. According to the preceding account, the
refutation on the top is positively oriented towards the conclusion that Petersen is a
Swede, and that allows to chain that refutation with argument (1), to recover the initial
conclusion that Petersen is probably not a Roman Catholic. Thus the conclusion we
finally reach from this complex argument constructed through the interaction of
proponent and opponent is that Petersen is probably not a Roman Catholic.
We turn now to another variation of Petersen’s example. The starting point is the
same as before:
(1) Proponent: Petersen is a Swede, so he is presumably not a Roman Catholic.
But now the opponent asks for a justification of the step from the premise
to the conclusion:
(2) Opponent: How do you get there?
The proponent reacts making the warrant explicit:
(3) Proponent: A Swede can be taken almost certainly not to be a Roman Catholic.
Then the opponent requires the opponent to provide that warrant with
backing:
(4) Opponent. Is that true?
(5) Proponent: The proportion of Swedes who are Roman Catholics is less than
2%, according to the tables in Whitaker’s Almanack, 1958 edition.
The dialogue ends with an attempt by the opponent to rebut the backing
argument:
(6) Opponent: These data are obviously outdated and hence unreliable.
Collecting all these pieces together we get the following diagram:
350 13 Intertwined Structures

These data are

obviously outdated
The proportion of Swedes
who are Roman Catholics is
Therefore less than 2%, according to
the tables in Whitaker’s
Almanack, 1958 edition
Whitaker’s Almanack,
These data are
against 1958 edition, is a Therefore Petersen is a Swede
unreliable
reliable source:
A Swede can be taken
almost certainly not to be a Therefore
Roman Catholic:
Petersen is
presumably not a
Roman Catholic

On the right half of the diagram we have a backed argument oriented towards the
conclusion that Petersen is presumably not a Roman Catholic. This backed argument
is rebutted by the argument in the left top corner. Since a rebuttal (by exception)
is neutrally oriented towards the conclusion of the rebutted argument, the whole
argument is neutrally oriented towards the conclusion that Petersen is presumably
not a Roman Catholic.

13.2 A User’s Guide of Argument Structure Diagrams

(The following diagrams are conceived to be drawn in Word using the Word Tables
toolkit or the like).
REVIEWING TOULMIN DIAGRAMS.
Simple argument:

PREMISES
Therefore
(QUALIFIER) CONCLUSION

Simple argument with warrant:

PREMISES
WARRANT Therefore
(QUALIFIER) CONCLUSION
13.2 A user’s Guide of Argument Structure Diagrams 351

Simple argument with warrant and backing:

BACKING
Therefore PREMISES
WARRANT Therefore
(QUALIFIER) CONCLUSION

BEYOND TOULMIN MODEL.

INTERARGUMENTATIVE RELATIONSHIPS.
concatenation: combination of two arguments in which the conclusion of one of
them is one of the premises of the other.
Structure: subordinative argument.

P1,…,Pn + Q, Q1,…,Qm becomes P1,…,Pn

Therefore Therefore Therefore
Q R Q Q1,…,Qm
Therefore
R

backing: combination of two arguments in which the conclusion of one of them (the
backing argument) is a warrant of the other (the warranted argument).
Structure: backed argument.

P1,…,Pn
Therefore Q1,…,Qm
R Therefore
S

Co-orientation: combination of two arguments with the same conclusion.

P1,…,Pn + Q1,…,Qm becomes P1,…,Pn moreover Q1,…,Qm

Therefore Therefore Therefore Therefore
R R R
352 13 Intertwined Structures

two main kind of co-orientation: conjunction and disjunction of arguments.

Conjunction: combination of two arguments with the same conclusion that reinforce
each other.
Structure: conjunctive argument.

P1,…,Pn + Q1,…,Qm becomes P1,…,Pn and Q1,…,Qm

Therefore Therefore Therefore Therefore
R R R

Disjunction: combination of two arguments with the same conclusion that are
presented as alternative defenses of it.
Structure: disjunctive argument.

P1,…,Pn + Q1,…,Qm becomes P1,…,Pn or Q1,…,Qm

Therefore Therefore Therefore Therefore
R R R

Weighing: comparison or balance of the probative force of two arguments. We use

the signs < , ≤ and ≈ to indicate the sense of the comparison; e.g.

P1,…,Pn < Q1,…,Qm

Therefore Therefore
R R

Weighing of arguments with mutually consistent premises and mutually inconsistent

conclusions is called “refutation” (see below).
COUNTERARGUMENTS.
Objection: an argument A is an objection to an argument B when A’s conclusion is
presented as being inconsistent with B’s premises.
13.2 A user’s Guide of Argument Structure Diagrams 353

Structure: objection.

R,…,R
Therefore
S against P1,…,Pn
Therefore
Q

Rebuttal: an argument A is a rebuttal of an argument B when A’s conclusion is

presented as being inconsistent with the assumption that B’s premises provide a
reason for B’s conclusion.
There are three main kind of rebuttal.
Plain rebuttal: the warrant W is generally invalid, so that conclusions like Q cannot
be inferred from premises like P1,…,Pn through W.
Structure: plain rebuttal.

R,…,R
Therefore P1,…,Pn
W is not a valid rule against W: Therefore
Q

Exception: the warrant cannot be applied in the case at hand, due to exceptional
circumstances R1,…,Rm, so that C cannot be inferred from P through W.
Structure: exception.

R1,…,Rm
Therefore P1,…,Pn
W is not applicable in this particular case against W: Therefore
Q

Counteranalogy: a sort of negative analogy, in which, on the basis of the similarity of

two suppositional arguments, the step from the premises to the conclusion is rejected,
without reference to any particular warrant.
354 13 Intertwined Structures

Structure: rebuttal by comparison.

It is not the case that if Q then R

Therefore T
It is not the case that if T then U against Therefore
U

Refutation. A pro tanto reason is a consideration that counts in favor of some position,
and thus has genuine weight, but nonetheless may be outweighed by other consid-
erations. A refutation is an attempt to show that a pro tanto reason is outweighed by
other consideration, so that it is inconclusive.
An argument A is a refutation of an argument B when (1) both are taken to offer
pro tanto reasons, (2) A’s conclusion is inconsistent with B’s conclusion, and (3)
A (the refuting argument) is presented as being at least as strong as B (the refuted
argument).1
There are two kinds of refutation, depending on the balance of the probative force
of the involved arguments.
Contradicting refutation. The refuting argument is presented as stronger than the
refuted argument. We indicate this using the connector but or the sign < , with the
refuting argument on the right and the refuted argument on the left.
Structure: contradicting refutation.

R1,…,Rm but P1,…,Pn

Therefore Therefore
Q Not Q

R1,…,Rm < P1,…,Pn

Therefore Therefore
Q Not Q

Alternatively, the connector although and the sign > can be used reversing the order
of the refuting argument and the refuted argument.

1 To simplify things in the diagrams below we assume that the conclusions of the confronted
arguments are contradictory to each other, although it is enough that they are incompatible in a
wider sense.
13.2 A user’s Guide of Argument Structure Diagrams 355

P1,…Pn although R1,…,Rm

Therefore Therefore
Not Q Q

P1,…,Pn > R1,…,Rm

Therefore Therefore
Not Q Q

Cancelling refutation. The refuting argument is presented as being “on a par” with
the refuted argument. We use the connector but also and the sign ≈, with the refuted
argument on the left and the refuting argument on the right.
Structure: cancelling refutation.

R1,…,Rm But also P1,…,Pn

Therefore Therefore
Q Not Q

R1,…,Rm ≈ P1,…,Pn
Therefore Therefore
Q Not Q

Weighing (meta)arguments: an argument concluding either that an argument is

stronger than another or that they are on a par. When the premises of both arguments
are mutually consistent, the conclusion of a weighing argument is a refutation.
Structure: weighing meta-argument.

S
Therefore
R1,…,Rm < P1,…,Pn
Therefore Therefore
Q Not Q
356 13 Intertwined Structures

S
Therefore
R1,…,Rm ≤ P1,…,Pn
Therefore Therefore
Q Not Q

S
Therefore
R1,…,Rm ≈ P1,…,Pn
Therefore Therefore
Q Not Q

HYPOTHETICAL OR SUPPOSITIONAL ARGUMENTS.

Hypothetical or suppositional arguments are arguments in which some of the
premises are unasserted propositions of the form “Suppose that…” or the like -
i.e. suppositions. Suppositional arguments are often conveyed using premises in
subjunctive mood and a conclusion in conditional mood, as in “Suppose he were to
recommend the drastic medicine of fee for service in place of any commissions or
bonuses, there would be myriad of difficult implementation challenges”.
First kind:

P1,…,Pn, Suppose that Q1,…, Suppose that Qm

In this case
R

Second kind:

P1,…,Pn
Suppose that W: In this case
R

Suppositional arguments of the first kind can be rewritten as.

P1,…,Pn
In this case
If Q1 y…y Qm, then R
13.2 A user’s Guide of Argument Structure Diagrams 357

When all the premises are suppositions, the hypothetical argument may be replaced
by the conditional statement If Q1 y…y Qm, the R.
Uses of suppositional arguments.
Suppositional arguments appear as premises in some argument patterns. To make
it easy to identity suppositional arguments and their role, they are shaded in the
following diagrams.
Reductio ad absurdum (ad falsum, ad ridiculum, etc.)

P1,…,Pn, Suppose that Q

In this case
R R is absurd, false, ridiculous, etc.
Therefore
Q is absurd, false, ridiculous, etc.

P1,…,Pn
Suppose that Q: In this case
R R is absurd, false, ridiculous, etc.
Therefore
Rule Q is absurd, false, ridiculous, etc.

Possibilitation.

P1,…,Pn, Suppose that Q

In this case
R Q is possible, plausible, credible, etc.
Therefore
R is possible, plausible, credible, etc.
358 13 Intertwined Structures

P1,…,Pn
Suppose that Q: In this case
R R is possible, plausible, credible, etc.
Therefore
Rule Q is possible, plausible, credible, etc.

Analogy.
In analogical arguments suppositional arguments play a double role as premises as
well as conclusion.
Structure: parallel argument.

P1,…,Pn, Suppose that Q

In this case
R
Therefore
S1,…,Sm, Suppose that T
In this case
U

When, all the premises are suppositions, this amounts to.

Suppose that Q
In this case
R
Therefore
Suppose that T
In this case
U

And can be conditionally rewritten as:

13.2 A user’s Guide of Argument Structure Diagrams 359

If Q then R
Therefore
If T then U

In this form, analogy often appears in the position of a backed warrant, so that its
conclusion acts as warrant:

If Q then R
Therefore T
If T then U Therefore
U

Reference

Eemeren, Frans H. van (2010). Strategic Maneuvering in Argumentative Discourse. Amsterdam

and Philadelphia: John Benjamins.
Chapter 14
An Argument Dialectical Analysis
of the Russell-Copleston Debate

Abstract I will present an argument dialectical analysis of the Russell-Copleston

debate. As explained in chap. 9 this will be a logical analysis. Hence, I will focus
neither on the series of actions performed by Russell and Copleston during the
debate nor on the procedure followed by them in doing so, but on the relationships
between the arguments advanced by them. I will be guided by the logical question
par excellence “Shall we accept this claim on the basis of the reasons put forward in
support of it?”, as addressed by the protagonists of the debate through a succession of
arguments and counter-arguments. What I want to capture are the macro-arguments
that Russell and Copleston construct successively and interactively in the course of
the debate -what might be termed, perhaps, the logical structure of the debate.

In what follows I will attempt an argument dialectical analysis of the Russell-

Copleston debate. As explained in Chap. 9 this will be a logical analysis. Hence,
I will focus neither on the series of actions performed by Russell and Copleston
during the debate nor on the procedure followed by them in doing so, but on the rela-
tionships between the arguments advanced by them. I will be guided by the logical
question par excellence “Shall we accept this claim on the basis of the reasons put
forward in support of it?”, as addressed by the protagonists of the debate through
a succession of arguments and counter-arguments. What I want to capture are the
macro-arguments that Russell and Copleston construct successively and interactively
in the course of the debate -what might be termed, perhaps, the logical structure of
the debate.
Mine will be a partial analysis of the debate since I will not be concerned with
every speech act that plays a role in investigating the acceptability of standpoints.
Besides advancing and countering an argument, debaters may raise a question, ask
for clarification, clarify something, allocate the burden of proof, agree on the rules
of the exchange, etc.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 361
F. Leal and H. Marraud, How Philosophers Argue, Argumentation Library 41,
https://doi.org/10.1007/978-3-030-85368-6_14
362 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

14.1 The Argument from Contingency

[11] C: Well, for clarity’s sake, I will divide the argument into distinct stages.
First of all, I should say, we know that there are at least some beings in the
world which do not contain in themselves the reason for their existence. For
example, I depend on my parents, and now on the air, and on food, and so on.
Now, secondly, the world is simply the real or imagined totality or aggregate
of individual objects, none of which contain in themselves alone the reason for
their existence. There isn’t any world distinct from the objects which form it, any
more than the human race is something apart from the members. Therefore,
I should say, since objects or events exist, and since no object of experience
contains within itself the reason of its existence, the totality of objects must have
a reason external to itself. That reason must be an existent being. Well, this
being is either itself the reason for its own existence, or it is not. If it is, well and
good. If not, then we must proceed further. But if we proceed to infinity in that
sense, then there’s no explanation of existence at all. So, I should say, in order
to explain existence, we must come to a being which contains within itself the
reason for its own existence—that is to say, which cannot not exist.
Although it is not entirely clear what Copleston means by “stages,” he seems to
use the expressions “first of all” and “secondly” to list and distinguish the supporting
elements (premises and warrants) from the supported elements (conclusions). The
connector “Therefore” marks the step from these supporting elements to the conclu-
sion. Moreover, after “therefore” Copleston recapitulates his argument with the
construction “since…, …”. The first supporting element appears both before “there-
fore” and after “since”, making it clear that it is a premise. On the contrary the second
supporting element only appears before “therefore”, which leads me to think that it
is a warrant or an element related to the warrant. The warrant has to be a general-
ization that covers the associated conditional “If no object of experience contains
within itself a total explanation of its existence, the world as a whole must have a
total explanation external to itself”. This allows the first option to be discarded, and
points to the backing of the implicit warrant1 :

1Van Laar (2015, p. 54) argues that a reasonable dialogue allows for the possibility of defending
one’s argumentative connection in a fully particularist manner, without appealing to a general
warrant.
14.1 The Argument from Contingency 363

The world is the real or imagined totality or

aggregate of individual objects
Objects or events exist and no
Therefore object of experience contains
within itself reason of its existence
If no object of experience contains within
itself a total explanation of its existence, the
Therefore
world as a whole must have a total explanation
external to itself:
The world must have a reason
external to itself

I think that the phrases “For example, I depend on my parents, and now on the air, and
on food, and so on” and “There isn’t any world distinct from the objects which form it,
any more than the human race is something apart from the members” are clarifications
of what he means, and not justificatory reasons for what he says. Alternatively this
phrase could be seen as a case of argumentation by example, and it certainly supports
saying that ‘at least some beings in the world have not in themselves the reason for
their existence’, at least if the fact that Copleston’s parents engendered Copleston
counts as ‘the reason for their existence’. When he adds that he now depends on
air, food, and so on, Copleston refers to the fact that, without air, food, and so on,
he would die. In other words, thanks to his parents, Copleston came to exist, and
deprived of food, air, and so on, he would cease to exist. Naturally, the same is true
of Russell, and of every other human being, indeed of every other living being on
earth (mutatis mutandis, of course, for some organisms do not have parents and some
organisms are anaerobic). So, this example, suitably extended, indeed proves that at
least some beings in the world do not contain in themselves the reason for their
existence, namely all living beings.

I depend on my parents, and now on the air, and on food, and so on

Therefore
We know that there are at least some beings in the world which do
not contain in themselves the reason for their existence

Now, later on, Copleston will say that “no object of experience contains within itself
the reason of its existence”, which is a much stronger statement.
The reasoning continues from this conclusion until the final conclusion, marked
by “so”, that “we must come to a being which contains within itself the reason for its
own existence”. As with the former argument, Copleston recapitulates this argument
in the last sentence: “in order to explain existence, we must come to a being which
contains within itself the reason for its own existence.”
364 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

The world must have a reason external to itself

Therefore
We must come to a being which contains within itself the
reason for its own existence (i.e. which cannot not exist)

This final conclusion is stated twice. First as “That reason must be an existent being”,
and later as “we must come to a being which contains within itself the reason for its
own existence.”
Let us now look at the second part of the paragraph [11].
Well, this being is either itself the reason for its own existence, or it is not. If it is, well and
good. If not, then we must proceed further. But if we proceed to infinity in that sense, then
there’s no explanation of existence at all. So, I should say, in order to explain existence, we
must come to a being which contains within itself the reason for its own existence—that is
to say, which cannot not exist.

It seems clear that (1) the discourse marker ‘well’ connects this part with the previous
one, and (2) that this passage contains an argument, as indicated by the occurrence
of the connectors ‘If… then’, ‘but’, and ‘so’. I will analyze that argument first. Its
conclusion is clearly indicated by the connector ‘so’: “in order to explain existence,
we must come to a being which contains within itself the reason for its own existence”.
‘But’ points to a contradiction and that makes me think of some form of reductio ad
absurdum. A reductio ad absurdum contains a suppositional (sub)argument that can
also be expressed by means of a conditional. In the passage we find the conditional
“If not, then we must proceed further” which, according to the previous explanation,
can be rendered as a suppositional argument: “Suppose that the external reason for
the world were not the reason for its own existence; in this case we would have
to proceed further”. If so, the conclusion of that suppositional argument (“there’s
no explanation of existence at all”) contradicts some previous concession, which I
identify with the intermediate conclusion (“The world must have a reason external
to itself”) of the previous argument. The reductio ad absurdum is then as follows:

Suppose that the external reason for the world were not the
reason for its own existence
In this case
We would have to proceed further (= we would have to
proceed to infinity in that sense)
In that case
The world must have a reason
There would be no explanation of existence at all
external to itself
Therefore
In order to explain existence, we must come to a being which contains within itself the
reason for its own existence—that is to say, which cannot not exist.

The discourse marker ‘well’ has received a lot of attention in the specialized literature;
one of the earliest descriptions of ‘well’ as a discourse marker is found in Lakoff
14.1 The Argument from Contingency 365

(1973). Lakoff analyzes ‘Well’ as a response marker, that may be used when “the
speaker senses some sort of insufficiency in his answer” (1973:463). Svartvik agrees
with Lakoff’s analysis of ‘well’ in answers, but he identifies other two majors uses
of ‘well’: as a qualifier and as a frame. As a frame, normally occurring non-initially,
‘well’ introduces explanations, clarifications, etc. (Svartvik 1980: 174). This is a
reason to think that the conclusion of the above reductio works here as a warrant,
since “the warrant is, in a sense, incidental and explanatory, its task being simply
to register explicitly the legitimacy of the step involved” (Toulmin 2003, p. 92;
my italics). Thus I’m inclined to think that this explanation is a warrant-establishing
argument (Op.cit., pp. 111–113). In addition, its conclusion has the form of a warrant:
“Given data such as objects or events exist, one may take it that there is a being which
cannot not exist”.
Putting this warrant-establishing argument and the warranted argument together,
we get the following diagram of the argumentation in [11]:

Suppose that the external reason for

the world were not the reason for its
own existence
In this case
We would have to proceed further
(= we would have to proceed to
infinity in that sense)
In that case
The world must have a There would be no explanation of
reason external to itself existence at all
The world must have a reason external
Therefore
to itself
In order to explain existence, we must come to a being which
contains within itself the reason for its own existence—that is Therefore
to say, which cannot not exist:
There is a being which contains within
itself the reason for its own existence
(i.e. which cannot not exist)

[12] R: This raises a great many points and it’s not altogether easy to know where
to begin, but I think that, perhaps, in answering your argument, the best point
at which to begin is the question of a necessary being. The word “necessary”, I
should maintain, can only be applied significantly to propositions, and, in fact,
only to such as are analytic, that is to say such as it is self-contradictory to deny.
I could only admit a necessary being if there were a being whose existence it
is self-contradictory to deny. I should like to know whether you would accept
Leibniz’s division of propositions into truths of reason and truths of fact; the
former, the truths of reason, being necessary.
For Russell’s remarks to be a relevant criticism, the final conclusion of Copleston’s
argument, “We must come to a being which cannot not exist”, must be rephrased as
“We must come to a necessary being” (even if Copleston has not used the expression
“necessary being”). If so, Russell argues as follows:
366 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

The word “necessary” can only be applied significantly to

propositions such as it is self-contradictory to deny
Therefore
We could only admit a necessary being if there were a
being whose existence it is self-contradictory to deny

Then Russell proposes for the sake of discussion to reformulate the final step of
Copleston’s argument as follows:

The world must have a reason (explanation)

external to itself
It is self-contradictory to deny
the existence of the external Therefore
reason of the world:
We must come to a being which cannot not exist

To indicate that this is the only version of Copleston’s argument that he is willing to
consider, without making it his own, Russell uses a subjunctive conditional (I could
only admit … if there were …).
[13] C: Well, I certainly should not subscribe to what seems to be Leibniz’ idea
of truths of reason and truths of fact, since it would appear that for him there are
in the long run only analytic propositions. It would seem that for Leibniz truths
of fact are ultimately reducible to truths of reason. That is to say, to analytic
propositions, at least for an omniscient mind. I couldn’t agree with that. For one
thing it would fail to meet the requirements of the experience of freedom. I don’t
want to uphold the whole philosophy of Leibniz. I’ve made use of his argument
from contingent to necessary being, basing the argument on the principle of
sufficient reason, simply because it seems to me a brief and clear formulation
of what is, in my opinion, the fundamental metaphysical argument for God’s
existence.

[14] R: But, to my mind, “a necessary proposition” has got to be analytic. I don’t

see what else it can mean. And analytic propositions are always complex and
logically somewhat late. Rational animals are animals is an analytic proposition;
but a proposition such as This is an animal can never be analytic. In fact, all
the propositions that < are > analytic are somewhat late in the build-up of
propositions.
In Turn 13 Copleston rejects the equivalence of “There is a necessary being” with
“There is an analytical proposition that asserts the existence of a being”. Against
his denial, Russell argues, on the basis of his inability to conceive otherwise, that a
necessary proposition (i.e. a proposition asserting that something is necessary) has
got to be analytic.
14.1 The Argument from Contingency 367

I don’t see what else “a necessary proposition” can mean

Therefore

To my mind, “a necessary proposition” has got to be analytic

Here semicolon is used to connect conclusion and premise, in that order.

Despite Copleston’s protests, Russell still seems to consider that the propositions
“There is a necessary being” and “There is a necessary proposition that asserts
the existence of a being” interchangeable and proceeds accordingly. Throughout all
these reinterpretations and reformulations, in the 14th turn Russell seeks to refute
the formulation of Copleston’s argument in the 12th turn:

Any proposition
asserting the existence
of a being is simple
An analytic
proposition is Therefore
always complex:
There is a being which
No analytical
contains within itself
but proposition asserts the
the reason for its own
existence of a being
existence
It is self-contradictory to
To my mind, “a
deny the existence of any
necessary
being which contains Therefore Therefore
proposition” has got
within itself the reason for
to be analytic:
its own existence:
There is a necessary There can’t be a
being necessary being

The key to interpreting this turn is the connector “but”. This connector is used here
to construct a complex argument by joining two arguments with opposite conclu-
sions. The argument on the left (columns 1–2 of the above diagram) is provided
by Copleston on Turn 12 (and deformed or reformulated later by Russell). With
“but” Russell (1) connects his intervention with that of Copleston, and (2) makes
Copleston’s argument his own. Thus, Russell does not conclude that there can’t be
a necessary being directly from his new argument (columns 3–4), but through the
balance of the two opposite arguments. The “but” connector often, but not always,
introduces a weighting of arguments and indicates that the argument that follows it
is considered by the arguer stronger than the one that precedes it, and thus imposes
its conclusion.
The successive steps in this subordinative refutation are identified through the
corresponding warrants, using the connector “and” to mark bottom-up the progress
368 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

from one argument to another.2 The phrases “Rational animals are animals” and
“This is an animal” serve as illustrative examples, and hence their function is to
clarify the assertion that analytic propositions are always complex, not to justify it.
Examples can perform, alternatively, a justificatory role. If this were the case here,
these examples would be presented as a justificatory reason for the assertion just
before the full-stop.

“Rational animals are

animals” is an analytic
proposition; but a
proposition such as
“This is an animal” can
never be analytic

Any proposition
asserting the
Therefore
existence of a being
is simple
An analytic proposition
Therefore
is always complex:
There is a being which No analytical
contains within itself proposition asserts
but
the reason for its own the existence of a
existence being
It is self-contradictory to
deny the existence of any To my mind, “a
being which contains Therefore necessary proposition” Therefore
within itself the reason has got to be analytic:
for its own existence:
There is a necessary There can’t be a
being necessary being

Going back to the first diagram, the only sentences from Russell’s counterargument
that explicitly appear in the text are those in the third column. Assuming that there
is an argument in the passage, “but” relates it to Copleston’s previous argument
and indicates that the final conclusion of Russell’s argument in 14 is the opposite of
Copleston’s argument in 12: “There can’t be a necessary being”. This justifies the last
cell in the 4th column. For the sake of simplicity, in the remaining I will formulate
Copleston’s conclusion in the formal mode, as Russell claims: “’God exists’ is a
necessary proposition”.

2 Although Toulmin says that “data are appealed to explicitly, warrants implicitly” (2003, p. 92),
according to my experience, what is rare is to make explicit the warrant when the premises and
conclusion are explicit, since in normal circumstances, the addressee will be able to guess which is
the warrant. For the same reason, it is unusual to make explicit the premise when the warrant and
conclusion are explicit.
14.1 The Argument from Contingency 369

The question is how is reached that conclusion, not explicit but clearly indicated
by the connector “but”, from the two sentences that do appear: “An analytical propo-
sition is always complex”, and “To my mind, ‘a necessary proposition’ has got to be
analytical”. The disposition of the text and the place where “but” has been inserted
point that the opposite conclusion to that of Copleston’s previous argument is reached
somehow from the sentence that follows, namely “a necessary proposition has got
to be analytical”. A possible interpretation would take this as a premise:

“a necessary proposition” has got to be analytic

Therefore
“God exists” is not a necessary proposition

The inferential step is thoroughly incomprehensible, but let us examine how the
sentence “’a necessary proposition’ has got to be analytic”, which is connected to
the previous one by “and” (if we omit the emphatic remark “I don’t see what else it
can mean”), would fit in this diagram. The connector “and” is multifunctional, and
can be used (a) to add a co-premise to an already stated one:

“a necessary proposition” has got to be analytic,

and an analytic proposition is always complex
Therefore
“God exists” is not a necessary proposition

(b) to add a second reason to an already given one, forming a conjunction of reasons
or arguments:

“a necessary proposition” has got to be analytic and an analytic proposition is always complex
Therefore Therefore
“God exists” is not a necessary proposition “God exists” is not a necessary proposition

Finally (c) “and” sometimes has a progressive meaning, and marks stages in the
course of a reasoning, and the passage therefore could be interpreted alternatively as
a subordinative argument:

“a necessary proposition” has got to be analytic

Therefore
an analytic proposition is always complex
Therefore
“God exists” is not a necessary proposition
370 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

None of these reconstructions seem plausible, basically because both steps from the
premises to the conclusion seem unfathomable.
If “a necessary proposition has got to be analytical” leads somehow to the conclu-
sion “There can’t be a necessary being” and it is not a premise, the alternative is
that it be a warrant. This is based on an empirical principle, mentioned in 10.4,
according to which when the premises and the conclusion are explicit, the warrant
remains implicit, and when the conclusion and the warrant are explicit, the premise
remains implicit. The pragmatic reason is that usually, given two of these elements,
the addressee can easily infer the third.

To my mind, “a necessary
Therefore
proposition” has got to be analytic:
“God exists” is not a necessary
proposition

This arrangement invites to infer that the implicit premise is “´God exists’ is not an
analytical proposition”. As for the remaining phrase and the “and” that precedes it,
it could be inferred by a similar reasoning that it is a concatenation in which “an
analytical proposition is always complex” either the premise of the first link3 :

an analytic proposition is always complex

Therefore
“God exists” is not an analytic proposition
To my mind, “a necessary
Therefore
proposition” has got to be analytic:
“God exists” is not a necessary proposition

Or the warrant of this link:

An analytic proposition is always complex: Therefore

“God exists” is not an analytic proposition
To my mind, “a necessary proposition” has
Therefore
got to be analytic:
“God exists” is not a necessary proposition

3 Assuming, of course that “and” has a progressive sense and that the sequence full stop + and
indicates that the order of reasoning is the inverse of the disposition of statements in the text.
(Usually, when full stop is used to separate the premise from the conclusion, the conclusion is
placed before the premise).
14.1 The Argument from Contingency 371

The form of “an analytical proposition is always complex” corresponds to that of a

warrant, and if it is so, it is clear that the missing premise is something of the type of
“’God exists’ is not a complex proposition”. On the contrary, the first interpretation
introduces, once again, an incomprehensible inferential step.

"’God exists’ is not a complex proposition"

An analytic proposition is always complex: Therefore
“God exists” is not an analytic proposition
To my mind, “a necessary proposition” has
Therefore
got to be analytic:
“God exists” is not a necessary proposition

[15] C: Take the proposition If there is a contingent being, then there is a necessary
being. I consider that that proposition, hypothetically expressed, is a necessary
proposition. If you are going to call every necessary proposition an analytic
proposition, then, in order to avoid a dispute in terminology, I would agree to
call it analytic, though I don’t consider it a tautological proposition. But the
proposition is a necessary proposition only on the supposition that there is a
contingent being. That there is a contingent being actually existing has to be
discovered by experience, and the proposition that there is a contingent being is
certainly not an analytic proposition, though once you know, I should maintain,
that there is a contingent being, it follows of necessity that there is a necessary
being.
Copleston rejects the proposed equivalence among “necessary proposition”, “ana-
lytical proposition” and “tautological proposition”, and rephrases the final step of
his argument making explicit its warrant:

There is a contingent being actually existing

If there is a contingent being,
Therefore
then there is a necessary being:
There is a necessary being

It is a kind of condensed or simplified version of the original argument, focused on

what Copleston takes to be the key aspect to dismiss as irrelevant Russell’s criticism.
The purpose of the reformulation is therefore explanatory.
Copleston declares that the warrant of this argument is a necessary proposi-
tion, and so the conclusion necessarily follows from the premise. Hence the adverb
“necessarily” works here as a qualifier, in Toulmin’s sense:
Warrants are of different kinds, and may confer different degrees of force on the conclusions
they justify. Some warrants authorise us to accept a claim unequivocally, given the appropriate
data—these warrants entitle us in suitable cases to qualify our conclusion with the adverb
‘necessarily’; others authorise us to make the step from data to conclusion either tentatively, or
else subject to conditions, exceptions, or qualifications—in these cases other modal qualifiers,
such as ‘probably’ and ‘presumably’, are in place. It may not be sufficient, therefore, simply
to specify our data, warrant and claim: we may need to add some explicit reference to the
372 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

degree of force which our data confer on our claim in virtue of our warrant. In a word, we
may have to put in a qualifier. (Toulmin 2003, p.93)

We may write the qualifier above the conclusion and immediately after the connector
“Therefore”:

There is a contingent being actually existing

If there is a contingent being,
Therefore necessarily
then there is a necessary being:
There is a necessary being

[16] R: The difficulty of this argument is that I don’t admit the idea of a necessary
being and I don’t admit that there is any particular meaning in calling other
beings contingent. These phrases don’t, for me, have a significance, except within
a logic that I reject.
Russell dismisses the preceding argument as containing a meaningless statement.
Remember that a dismissal is a meta-argument that presents a reason to reject an
argument as a whole (Cfr. 10.3).

I don’t admit the idea of a necessary being

and I don’t admit that there is any particular
meaning in calling other beings contingent
Therefore
There is a contingent being actually
existing
This argument is nonsense against
Therefore
There is a necessary being

[17] C: Do you mean that you reject these terms because they won’t fit in with
what is called “modern logic”?
Copleston asks Russell to justify the premise of his dismissal.
[18] R: Well, I can’t find anything that they could mean. The word “necessary”,
it seems to me, is a useless word, except as applied to analytic propositions, not
to things.
[19] C: In the first place, what do you mean by “modern logic?” As far as I know,
there are somewhat differing systems. In the second place, not all modern logi-
cians surely would admit the meaninglessness of metaphysics. We both know, at
any rate, one very eminent modern thinker whose knowledge of modern logic
was profound, but who certainly did not think that metaphysics are meaning-
less or, in particular, that the problem of God is meaningless. Again, even if all
modern logicians held that metaphysical terms are meaningless, it would not
follow that they were right. The proposition that metaphysical terms are mean-
ingless seems to me to be a proposition based on an assumed philosophy. The
14.1 The Argument from Contingency 373

dogmatic position behind it seems to be this: What will not go into my machine is
non-existent, or it is meaningless, or it is the expression of emotion. I am simply
trying to point out that anybody who says that a particular system of modern
logic is the sole criterion of meaning is saying something that is over-dogmatic;
he is dogmatically insisting that a part of philosophy is the whole of philosophy.
After all, a contingent being is a being which has not in itself the complete reason
for its existence; that’s what I mean by a contingent being. You know, as well
as I do, that the existence of neither of us can be explained without reference to
something or somebody outside us, our parents, for example. A necessary being,
on the other hand, means a being that must and cannot not exist. You may say
that there is no such being, but you will find it hard to convince me that you do
not understand the terms I am using. If you do not understand them, then how
can you be entitled to say that such a being does not exist, if that is what you do
say?
Due to the lack of an explicit answer to his question in Turn 16, Copleston attributes
the following argument to Russell:

Modern logic rejects the terms “contingent” and “necessary”

Therefore
The terms “contingent” and “necessary” are meaningless words

To be dialectically relevant, the conclusion of this argument has to be taken as equiv-

alent to the premise in Russell’s dismissal — “I don’t admit the idea of a necessary
being and I don’t admit that there is any particular meaning in calling other beings
contingent.”
Copleston treats as equivalent the statements “Modern logic rejects the terms
‘contingent’ and ‘necessary’” and “Modern logic says that metaphysics is meaning-
less”. Whatever the interpretation of the premise, Russell’s is an argument based on
expert opinion (cfr. Walton et al. 2008, pp. 13–14 and 310).

Modern logicians say that metaphysics

is meaningless
The opinions of modern logicians in the domain
Therefore
of (meta)philosophy are generally reliable:
Metaphysics is meaningless

Copleston then raises several critical questions appropriate to this type of argument.
He uses the connectors “in the first place”, “in the second place” and “again” to
separate the critical questions from each other.
The first question (“what do you mean by modern logic?”), and the remark that
follows it (“as far as I know, there are somewhat differing systems”), are relevant
374 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

for those arguments that appeal to the opinion of collectives such as scientists,
researchers, etc. and points to an objection. However, Copleston does not develop
this objection and leaves it as a question, in an embryonic form, so to speak. His
second critical question appears in every analysis of the scheme of argument from
expert opinion. Walton, Reed and Macagno (2008, p. 310) call it “the Consistency
Question”:
CQ5. Consistency Question: Is A consistent with what other experts assert?

On this occasion Copleston gives the critical question the form of an argument (with
a veiled allusion to Whitehead):

We both know, at any rate, one very eminent modern thinker whose knowledge of
modern logic was profound, but who certainly did not think that metaphysics are
meaningless or, in particular, that the problem of God is meaningless
Therefore
Not all modern logicians surely would admit the meaninglessness of metaphysics

This critical question points to a rebuttal by exception of the argument attributed to

Russell:

We both know, at any rate, one

very eminent modern thinker
whose knowledge of modern logic
was profound, but who certainly
did not think that metaphysics are
meaningless
Therefore
Not all modern logicians surely
would admit the meaninglessness
of metaphysics
Modern logic says that the
word “necessary” is a
useless word, except as
applied to analytic
propositions, not to things
The opinions of modern
logicians in the domain of
against Therefore
(meta)philosophy are
generally reliable:
The word “necessary” is a
useless word, except as
applied to analytic
propositions, not to things

The last critical question, introduced by “again”, also points to a rebuttal, but now to
a plain rebuttal:
14.1 The Argument from Contingency 375

Modern logic says that the

word “necessary” is a
useless word, except as
applied to analytic
propositions, not to things
Even if all modern logicians held The opinions of modern
that metaphysical terms are logicians in the domain of
against Therefore
meaningless, it would not follow (meta)philosophy are
that they were right generally reliable:
The word “necessary” is a
useless word, except as
applied to analytic
propositions, not to things

This plain rebuttal is akin to the traditional charge against the ad populum fallacy:
“Beliefs, attitude and values need to be closely examined separately from the groups
that support them. No matter who or how many people hold a certain belief, their
advocacy does not prove its correctness” (van Vleet 2011, p. 20).
Once rebutted this argument, Copleston goes on conclude, in the absence of any
supporting reason, that “The proposition that metaphysical terms are meaningless
seems to me to be a proposition based on an assumed philosophy”.
Copleston counter-arguments that the experts being called upon (“a particular
system of modern logic”) are only a part of the relevant expert community (“the
whole of philosophy”), and therefore taking a particular system of modern logic as
the sole criterion would be dogmatic. This is a variant of Walton, Reed and Macagno’s
critical question CQ5, which points to expert consensus as a necessary condition for
this kind of argument, and therefore another rebuttal by exception.

A particular system of modern logic

is only a part of philosophy, is not
the whole of philosophy
Modern logic says that the
word “necessary” is a
Therefore useless word, except as
applied to analytic
propositions, not to things
Anybody who says that a particular The opinions of modern
system of modern logic is the sole logicians in the domain
against Therefore
criterion of meaning is saying of (meta)philosophy are
something that is over-dogmatic generally reliable:
The word “necessary” is a
useless word, except as
applied to analytic
propositions, not to things

Copleston goes on to argue that Russell does in fact understand the terms “necessary”
and “contingent”. He gives two reasons for this: those terms can be defined, and
Russell claims that there is no such thing as a necessary being.
The connector “after all” presents what follows it as a fact of common knowledge,
and thus a suitable premise:
376 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

A “contingent” being means a being which has not in itself the

complete reason for its existence. A “necessary” being, on the
other hand means a being that must and cannot not exist
Therefore
The terms “contingent” and “necessary” are meaningful

This could be an analytical argument (i.e., an argument based on the meaning of

words or concepts), given that to define is to “to set forth or explain what a word or
expression means; to declare the signification of a word” (OED). Copleston’s second
counterargument (marked by “but”) is a charge of pragmatic inconsistency: Russell
position is pragmatically inconsistent for he does not practice what he preaches.

Russell says that a necessary being does not exist

Therefore
Russell understands the words “necessary being”

“But” is used to mark the rejection of Russell’s assertion that he does not understand
“necessary being”. Thus Copleston claims that Russell understands the words “nec-
essary being”. The full stop separates Copleston’s claim from the reason alleged to
support it. As it has been said, when the conclusion is explicit, it is enough to provide
either the premise or the warrant to present a supporting reason. Here Copleston
chooses the second option:

Russell says that a necessary being does not exist

If you do not understand the words
“necessary being”, then you are not
Therefore
entitled to say that such a being does not
exist:
Russell understands the words “necessary being”

However in 1.4 Russell declares that his position is agnosticism, in blatant

contradiction of what Copleston attributes to him in the premise.
[20] R: Well, there are points here that I don’t propose to go into at length. I don’t
maintain the meaninglessness of metaphysics in general at all. I maintain the
meaninglessness of certain particular terms—not on any general ground, but
simply because I have not been able to see an interpretation of those particular
terms. It is not a general dogma, it is a particular thing. But those points I will
leave for the moment. And I will say that what you have been saying brings us
back, it seems to me, to the ontological argument—that there is a being whose
essence involves existence, so that his existence is analytic. That seems to me to
be impossible, and it raises, of course, the question what one means by existence,
and as to this, I think a subject named can never be significantly said to exist
14.1 The Argument from Contingency 377

but only a subject described; and that existence, in fact, quite definitely is not a
predicate.
Russell rejects Copleston’s reconstructions (19) of his argument (14) concluding that
a necessary proposition has got to be analytic on the basis of the authority of modern
logic. These reconstructions are:

Modern logic rejects the terms “contingent” and “necessary”

Therefore
The terms “contingent” and “necessary” are meaningless words

and

Modern logicians say that metaphysics is meaningless

Therefore
Metaphysics is meaningless

Russell reiterates his initial formulation in 14 that supports the claim that the terms
“contingent” and “necessary” are meaningless appealing to an ad ignorantiam (“I
have not been able to see an interpretation of those particular terms”, “I don’t see
what else ‘a necessary proposition’ can mean”). He quite explicitly says (“but simply
because”) that the sole reason he has for maintaining that these terms are meaningless
is his inability to find an appropriate interpretation of them.
The phrase “But those points I will leave out for the moment” announces a change
of topic, and Russell moves from the argument from contingency to the ontological
argument, that he casts as follows:

There is a being whose essence involves existence

Therefore
There is a being whose existence is analytic

Russell rejects the conclusion of the ontological argument, calling it “impossible”,

and then makes two assertions: (1) a subject named can never be significantly said to
exist but only a subject described, and (2) existence, in fact, quite definitely is not a
predicate. The principle of relevance invites us to think that these two assertions are
intended to justify his rejection of the ontological argument, but it is not clear what
the relationship between them is, and whether the reference of the pronoun “that” in
“That seems to me to be impossible” is the conclusion of the ontological argument
or the argument as a whole.
[21] C: Well, you say, I believe, that it is bad grammar, or rather bad syntax, to
say, for example, T. S. Eliot exists. One ought to say, for example, The author of
“Murder in the Cathedral” exists. Are you going to say that the proposition The
cause of the world exists is without meaning? You may say that the world has no
378 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

cause; but I fail to see how you can say that the proposition that The cause of
the world exists is meaningless. Put it in the form of a question: Has the world a
cause? or Does a cause of the world exist? Most people surely would understand
the question, even if they don’t agree about the answer.
Copleston understands that Russell, in his previous turn, was arguing that “The
cause of the world exists’” is without meaning, and, if I interpret his example
correctly, that he did so from his first assertion (“A subject named can never be
significantly said to exist but only a subject described”).

In “The cause of the world exists” a named subject is said to

exist
A subject named can never
be significantly said to exist Therefore
but only a subject described:
“The cause of the world exists” is without meaning

Allusion to problems of grammar or syntax suggests that Copleston interprets

Russell’s speech as an attempt to dismiss the ontological argument.
First Copleston objects that “the cause of the world” is not a name, but a descrip-
tion. Then he goes on to argue that the contentious proposition has meaning, and he
does so using a kind of argument from popular opinion:

Most people surely would understand the questions “Has

the world a cause?” or “Does a cause of the world exist?”
Therefore
“The cause of the world exists” is meaningful

This argument from popular opinion comes after Copleston’s objection, that cancels
the argument that a subject named can never be significantly said to exist. So the argu-
ment from popular opinion is not directly opposed to Russell’s. Properly speaking
the argument from popular opinion is not used as a refutation of the argument that
a subject named can never be significantly said to exist, even if their conclusions
are contradictory to each other. When the argument from popular opinion appears,
the field is already clear. Thus there is no need to compare the strength of the two
contradictory arguments.
[22] R: Well, certainly, the question Does the cause of the world exist? is a question
that has meaning. But, if you say: Yes, God is the cause of the world, where you’re
using “God” as a proper name, then God exists will not be a statement that has
meaning. That is the position that I’m maintaining. Because, therefore, it will
follow that it cannot be an analytic proposition, ever, to say that this or that exists.
For example, suppose you take as your subject “the existent round square”, it
would look like an analytic proposition that the existent round square exists, but
it doesn’t exist.
14.1 The Argument from Contingency 379

Russell concedes the objection and says that “God” is a proper name, as used in
“Yes, God is the cause of the world”, and consequently (“then”) the proposition
“God exists” is meaningless. Russell thus dismisses an argument that Copleston has
never advanced:

“God” is used as a proper name in “Yes, God

is the cause of the world”
Therefore
God is the cause of the world
“God exists” has not meaning against Therefore
God exists

(By the way, if the objection were removed, Copleston’s second counterargument
would become a contradicting refutation).
With the expression “Because, therefore, it will follow that” Russell returns to his
critique of the ontological argument [20]: it is impossible that the existence of a being
be analytic.

It cannot be an analytic proposition to say that this or that exists

Therefore
It is impossible that the existence of a being be analytic

Russell’s intervention concludes with a Gaunilonian example4 : “’the existent

round- square exists’ look like an analytical proposition but the round-square doesn’t
exist”, whose function is to support the premise of the previous argument.

‘the existent round square exists’ looks like an analytic proposition

that the existent round square exists, but it doesn’t exist
Therefore
It cannot be an analytic proposition to say that this or that exists
Therefore
It is impossible that the existence of a being be analytic

This is intended to be a refutation of the statement of the ontological argument

mentioned in Turn 20.

4 “There is an objection to the Ontological Argument which is almost as old as the argument itself.
In his critique of St. Anselm’s Proslogion, Gaunilo claims that by using the same argument that
Anselm uses to prove the existence of God, one can prove the existence of a perfect island.” (Mann
1975, p. 417).
380 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

[23] C: No, it doesn’t. Then surely you can’t say it doesn’t exist unless you have
a conception of what existence is. As to the phrase, “existent round-square,” I
should say that it has no meaning at all.

[24] R: I quite agree. Then I should say the same thing in another context in
reference to a “necessary being”.
[25] C: Well, we seem to have arrived at an impasse. To say that a necessary being
is a being that must exist and cannot not exist has for me a definite meaning.
For you it has no meaning.
Copleston notes that they have reached an impasse. Russell dismisses the argu-
ment from contingency, arguing that expressions such as “being necessary” have
no meaning. Russell’s arguments (basically an argumentum ad ignorantiam) do not
convince Copleston, who otherwise fails to convince Russell that the expression
“necessary being” is significant.
[26] R: Well, we can press the point a little, I think. A being that must exist
and cannot not exist, would surely, according to you, be a being whose essence
involves existence.
Russell tries to get out of the impasse, leaving aside the question of the meaning
or lack of meaning of statements such as “the necessary being exists”. For that he
tries to obtain some concessions from Copleston to formulate the argument from
contingency without using the concept of being necessary.
[27] C: Yes, a being the essence of which is to exist. But I should not be willing
to argue the existence of God simply from the idea of His essence, because I
don’t think we have any clear intuition of God’s essence as yet. I think we have
to argue from the world of experience to God.
Copleston rejects the argument implicitly suggested in Turn 26, viz.

God’s essence involves existence

Therefore
God exists

His reason to do it (as indicated by “because”) is that humans don’t have any clear
intuition of God’s essence. At first sight, this sounds like an objection whose target
is the usability of the premise: humans are not entitled to assert that God’s essence
involves existence since they can’t have a clear intuition of God’s essence.
14.1 The Argument from Contingency 381

I don’t think we have any clear

intuition of God’s essence as yet
Therefore
We cannot assert that God’s
against God’s essence involves existence
essence involves existence
Therefore
God exists

[28] R: Yes, I quite see the distinction. But, at the same time, for a being with
sufficient knowledge, it would be true to say: Here is this being whose essence
involves existence.
[29] C: Yes, certainly: if anybody saw God, he would see that God must exist.
[30] R: So that, I mean, there is a being whose essence involves existence,
although we don’t know that essence. We only know there is such a being.
This exchange makes it clear that the point in 27 is not that the argument suggested by
Russell in 26 is not sound because his premise is not acceptable, but that it does not
constitute a proof (in the sense of a deductive argument that does not depend on any
empirical assumption) of the existence of God. Indeed, it would be paradoxical to
claim that I know that for a being with sufficient knowledge God’s essence involves
existence and to deny at the same time that I know that God’s essence involves
existence.
Russell’s introduction of an imaginary being with sufficient knowledge serves to
set aside the status of the premise, and focus instead on its relation to the conclusion,
i.e. on the deductive validity of the argument. 28 could then be paraphrased as: if we
knew that God’s essence contained his existence, could we infer his existence? In 29
Copleston concedes that a being with sufficient knowledge would see that God must
exist something completely different from the claim that such a being would infer
God’s existence from his knowledge of God’s essence. Thus I guess that Copleston
is rather offering a meta-argument:

I don’t think we have any clear intuition of God’s essence as yet

Therefore
The existence of God cannot be proved from the idea of His essence

This can be seen as a meta-argument, insofar as the conclusion amounts to: The
argument “God’s essence involves existence; therefore, God exists” is not a proof. In
his last sentence, Russell insists that we can know that there is a being whose essence
contains its existence without knowing its essence.
[31] C: Yes, I should add we don’t know the essence a priori. It is only true a
posteriori, <it is only> through our experience of the world that we come to a
382 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

knowledge of the existence of that being. And then one argues, the essence and
existence must be identical. Because if God’s essence and God’s existence were
not identical, then some sufficient reason for this existence would have to be
found beyond God.
In the first sentence of this passage Copleston clarifies his previous meta-
argument:

It is only through our experience of the world

that we come to a knowledge of the existence of
that being whose essence involves existence
Therefore
It is only true a posteriori that God’s essence
involves existence,
In a proof the truth of every
Therefore
premise must be known a priori:
The argument “God’s essence involves
existence; therefore, God exists” is not a proof

Once rejected the ontological argument, Copleston reformulates the metaphysical

argument starting from our empirical knowledge of the existence of a being that
contains within itself the reason of its existence. It can be argued from this premise
that the essence and existence of this being must be identical. Finally, Copleston
explains the legitimacy of the proposed step —thus “because” introduces here the
warrant of the argument.

We know through our experience of the world that

there is a being contains within itself the reason of its
existence
If God’s essence and God’s existence
were not identical, then some
Therefore (and then one argues)
sufficient reason for this existence
would have to be found beyond God:
There is a being whose essence involves existence

Most probably the premise of this argument alludes to the first sub-argument in the
subordinative argument of 11:

Objects exist and no object of experience contains within itself

the reason of its existence
therefore
The world must have a reason external to itself (i.e., there is a
being contains within itself the reason of its existence)
14.1 The Argument from Contingency 383

If I am right, this allows us to construct Copleston’s argument as a subordinative

argument:

Objects exist and no object of experience contains

within itself the reason of its existence
Therefore
The world must have a reason external to itself (i.e.,
there is a being that contains within itself the reason
of its existence
If God’s essence and God’s existence
were not identical, then some sufficient
Therefore
reason for this existence would have to
be found beyond God:
There is a being whose essence involves existence

[32] R: So, it all turns on this question of sufficient reason. And, I must say, you
haven’t defined “sufficient reason” in a way that I can understand. What do you
mean by “sufficient reason”? You don’t mean cause?
[33] C: Not necessarily. Cause is a kind of sufficient reason. Only contingent
being can have a cause. God is His own sufficient reason; but He is not cause of
Himself. By sufficient reason, in the full sense, I mean an explanation adequate
for the existence of some particular being.
[34] R: But when is an explanation adequate? Suppose I am about to make a
flame with a match. You may say that the adequate explanation of that is that I
rub it on the box.
[35] C: Well, for practical purposes—but theoretically, that is only a partial
explanation. An adequate explanation must ultimately be a total explanation,
to which nothing further can be added.
Russell asks for clarification of what is meant by “sufficient reason” in the above
formulation of the argument from contingency. After Copleston’s clarification that
“sufficient reason” means “total explanation”, the argument goes like this:

Objects exist and no object of experience contains

within itself a total explanation of its existence
Therefore
The world must have a total explanation external to
itself (i.e., there is a being that contains within itself
a total explanation of its existence
If God’s essence and God’s existence
were not identical, then some a total
Therefore
explanation for this existence would
have to be found beyond God:
There is a being whose essence involves existence
384 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

[36] R: Then I can only say you’re looking for something which can’t be got,
and which one ought not to expect to get.
[37] C: To say that one has not found it is one thing; to say that one should not
look for it seems to me rather dogmatic.
[38] R: Well, I don’t know. I mean, the explanation of one thing is another thing
which makes the other thing dependent on yet another, and you have to grasp
this sorry scheme of things entire to do what you want, and that we can’t do.
The phrase “you have to grasp this sorry scheme of things entire”, that subsequent
turns make clear that it refers to the universe as a whole, suggests that Russell
is attacking the first chained argument of 32–35. What Copleston wants with this
argument is to go from the premise that no object of experience contains within itself
a total explanation of its existence to the conclusion that the world must have a total
explanation external to itself. Thus, when Russell say: “and that we can’t do”, he is
rejecting this step, which comes down to a rebuttal of the argument.

The explanation of one thing

is another thing which
makes the other thing
dependent on yet another,
and you have to grasp this
sorry scheme of things
Therefore
The world cannot be
against
grasped as a whole
Objects exist and no object
of experience contains
within itself a total
explanation of its
existence
If no object of experience
contains within itself a total
explanation of its existence, the
Therefore
world as a whole must have a
total explanation external to
itself:
The world must have a
total explanation external
to itself (i.e., there is a
being that contains within
itself a total explanation of
its existence

[39] C: But are you going to say that we can’t, or we shouldn’t, even raise the
question of the existence of the whole of this ‘sorry scheme of things’, of the
whole universe?
Copleston asks Russell to clarify whether his rejection that the world can be grasped
as a whole implies that the question of the existence of such a whole cannot even be
raised.
14.1 The Argument from Contingency 385

[40] R: Yes. I don’t think there’s any meaning in it at all. I think the word
“universe” is a handy word in some connections, but I don’t think it stands for
anything that has a meaning.
The insistence on the meaning (or rather the reference) of the word “universe” leads
one to think that Russell is dismissing the first link of Copleston’s subordinative
argument (taking “world” to mean universe):

Objects exist and no object of experience contains within itself a total

explanation of its existence
Therefore
The world must have a total explanation external to itself (i.e., there is
a being that contains within itself a total explanation of its existence

This a dismissal of just the same kind as the one in Turn 16, since it is alleged that
the argument cannot be sound because its conclusion contains a meaningless term.

I don’t think that the word “universe”

stands for anything that has a meaning
Therefore
The argument that concludes that the
universe must have a total explanation
external to itself is meaningless
Objects exist and no object of
experience contains within itself a
total explanation of its existence
against
Therefore
The universe must have a total
explanation external to itself

[41] C: If the word is meaningless, it can’t be so very handy. In any case, I don’t
say that the universe is something different from the objects which compose it (I
indicated that in my brief summary of the proof). What I’m doing is to look for
the reason, in this case the cause, of the objects, the real or imagined totality of
which constitute what we call the universe. You say, I think, that the universe,
or my existence, if you prefer, or any other existence, is unintelligible?
To begin with Copleston argues from Russell’s admission that it is a handy word,
that the word “universe” has meaning, thus raising an objection to the argument in
Turn 40.
386 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

The word “universe” is

a handy word
If the word “universe”
is meaningless, it
Therefore
can’t be so very
handy:
The word “universe” I don’t think that the word “universe”
against
can’t be meaningless stands for anything that has a meaning
Therefore
The argument that concludes that the
universe must have a total explanation
external to itself is meaningless

“In any case” suggests a second objection, taken in disjunction with the preceding
objection: The word “universe” stands for the totality of existing objects (and hence,
it stands for anything that has a meaning).

The universe is not

something different from the
objects which compose it
Therefore
The word “universe” can’t be I don’t think that the word “universe” stands for
against
meaningless anything that has a meaning
Therefore
The argument that concludes that the universe
must have a total explanation external to itself is
meaningless

Having rejected Russell’s dismissal of his argument in 38 that the world must have
a total explanation external to itself, Copleston tries to clarify what his argument is,
insisting on the words “reason” and “cause”:

Objects exist and no object of experience contains within

itself the reason, or cause, of its existence
Therefore
Objects must have a reason or cause external to itself

Copleston then asks Russell if he rejects the conclusion of this argument, taking it
for granted that that which has no cause or reason for being is unintelligible.
[42] R: First may I take up the point that if a word is meaningless it can’t be
handy. That sounds well but isn’t in fact correct. Take, say, such a word as
“the” or “than”. You can’t point to any object that those words mean, but they
are nevertheless very useful words; I should say the same of “universe”. But
leaving that point, you ask whether I consider that the universe is unintelligible.
I shouldn’t say unintelligible: I think it is without explanation. Intelligible, to my
mind, is a different thing. Intelligible has to do with the thing itself, intrinsically,
and not with its relations.
14.1 The Argument from Contingency 387

First Russell rebuts Copleston’s first objection in Turn 41 using an argument from
example:

You can’t point to

any object that
words like as
“the” or “than”
mean, but they
are very useful
words
The word
therefore “universe” is
a handy word
A word can be If a word is
both meaningless against meaningless it Therefore
and handy can’t be handy:
The word I don’t think that the
“universe” word “universe” stands
against
can’t be for anything that has a
meaningless meaning
Therefore
The argument that
concludes that the
universe must have a
total explanation external
to itself is unsound

Although this is a consistent interpretation of Russell’s words, the phrase “I should

say the same of ‘universe’” seems more like a refutation (using an argument from
similarity):

You can’t point to any

object that words like as
The word “universe”
but “the” or “than” mean,
is a handy word
but they are very useful
words
If a word is The word “universe” is
meaningless it Therefore similar to the words Therefore
can’t be handy: “the” and “that”:
You can’t point to any
The word “universe” object that the word
can’t be meaningless “universe” means, but it
is a handy word

Second, to answer Copleston’s question, Russell distinguishes between unintelli-

gible and unexplained, and asserts that the universe is unexplained. This distinction,
however, plays no role in the sequel.
[43] C: Well, my point is that what we call the world is intrinsically unintelligible,
apart from the existence of God. You see, I don’t believe that the infinity of the
388 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

series of events—I mean a horizontal series, so to speak—if such an infinity

could be proved, would be in the slightest degree relevant to the situation. If you
add up chocolates, you get chocolates after all and not a sheep. If you add up
chocolates to infinity, you presumably get an infinite number of chocolates. So,
if you add up contingent beings to infinity, you still get contingent beings, not
a necessary being. An infinite series of contingent beings will be, to my way of
thinking, as unable to cause itself as one contingent being. However, you say, I
think, that it is illegitimate to raise the question of what will explain the existence
of any particular object?
Copleston says that unless the world is intrinsically unintelligible, God exists, and
he tries to show that the possibility of an infinite series of events is not a reason to
challenge this assertion (“would be in the slightest degree relevant”). To that end he
uses an analogy with chocolates and sheeps:

If you add up chocolates, you get chocolates after all and not a sheep. If you add up
chocolates to infinity, you presumably get an infinite number of chocolates
Therefore (“so”)
If you add up contingent beings to infinity, you still get contingent beings, not a
necessary being

It may be inferred from their position in the sequence of moves that these remarks
refer to Copleston’s argument in 11 and 31:

Objects or events exist and no object of experience

contains within itself reason of its existence
If we must proceed to infinity,
there’s no explanation of existence Therefore
at all:
The world must have a reason (explanation)
external to itself

The conditional nature of the statement “unless the world is intrinsically unintelli-
gible, God exists” refers to the warrant which connects the premise “Objects exist
and no object of experience contains within itself the reason of its existence” with
the conclusion “There is a being whose essence involves existence”.
Copleston’s analogy anticipates and defuses a consideration (“the series of events
could be infinite”) that might be adduced to rebut the above argument. This consid-
eration—Copleston argues—is irrelevant for an infinite series of contingent beings
will be as unable to cause itself as one contingent being. Thus the defused rebuttal
could be something like this:
14.1 The Argument from Contingency 389

The series of events could be infinite

Therefore
An infinite series of contingent beings will be able to cause itself

Now Copleston tries to refute this rebuttal with the analogy of sheeps and chocolates.

but If you add up chocolates, you get chocolates after all

The series of events could be
and not a sheep. If you add up chocolates to infinity,
infinite
you presumably get an infinite number of chocolates
Therefore Therefore
An infinite series of contingent If you add up contingent beings to infinity, you still
beings will be able to cause itself get contingent beings, not a necessary being

All this leads us to the following complex diagram:

If you add up chocolates,

you get chocolates after
all and not a sheep. If The series of
you add up chocolates to > events could be
infinity, you presumably infinite
get an infinite number of
chocolates
Objects or events
exist and no object
of experience
Therefore Therefore
contains within
itself reason of its
existence
If we must
If you add up contingent An infinite series
proceed to
beings to infinity, you of contingent
infinity, there’s
still get contingent beings will be against Therefore
no explanation
beings, not a necessary able to cause
of existence at
being itself
all:
The world must
have a reason
(explanation)
external to itself

The final question (“Do you say that it is illegitimate to raise the question of what
will explain the existence of any particular object?”) apparently challenges the step
from the premise “Objects exist and no object of experience contains within itself the
reason of its existence” to “The world must have a reason external to itself”, insofar
as this step rests on the assumption that there is a reason for the existence of every
object of experience.
390 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

[44] R: It’s quite all right if you mean by explaining it simply finding a cause
for it.
[45] C: Well, why stop at one particular object? Why shouldn’t one raise the
question of the cause of the existence of all particular objects?
The exchange of questions suggests that they continue to debate the intelligibility
of statements such as “The world must have a total external explanation to itself”.
If so, Copleston would be suggesting that if it makes sense to ask about the cause
of the existence of a particular object, it makes sense to ask about the cause of the
existence of all objects.

It makes sense to ask for the cause of the existence of particular objects
Therefore
One can raise the question of the cause of the existence of all particular objects

(This looks like a fallacy of composition; see also [52] below).

[46] R: Because I see no reason to think there is any. The whole concept of
cause is one we derive from our observation of particular things; I see no reason
whatsoever to suppose that the total has any cause whatsoever
This looks like as a plain rebuttal of the argument suggested by Copleston’s questions
in 45:

The whole concept of cause is one we

derive from our observation of
particular things
Therefore
I see no reason whatsoever to suppose
that the total has any cause whatsoever
It makes sense to ask for the cause of the
existence of particular objects
against
Therefore
One can raise the question of the cause
of the existence of all particular objects

[47] C: Well, to say that there isn’t any cause is not the same thing as saying that
we shouldn’t look for a cause. The statement that there isn’t any cause should
come, if it comes at all, at the end of the enquiry, not the beginning. In any case,
if the total has no cause, then to my way of thinking it must be its own cause,
which seems to me impossible. Moreover, the statement that the world is simply
there, if in answer to a question, presupposes that the question has meaning.
Copleston misunderstands Russell. What Russell maintains that there is no point in
talking about a cause of totality, and Copleston attributes to him the thesis that totality
14.1 The Argument from Contingency 391

has no cause. To counter the thesis he attributes to Russell, Copleston argues that (1) in
such a case the totality should be its own cause, something that he finds impossible,
and (2) asserting that the world is simply there presupposes that the question of
whether the world has a cause makes sense. The first argument is introduced by “In
any case” and the second by “Moreover”. In both cases the arguments are identified
through their warrants (conditional statements).”

It is impossible that It makes sense to

the total be its own Moreover answer that the world
cause is simply there
If the total has no The statement that the
cause, then it must world is simply there,
be its own cause: if in answer to a
Therefore Therefore
question, presupposes
that the question has
meaning:
It makes sense to ask
The total has a cause whether the world
has a cause

The argument on the left is directed against the thesis that the world doesn’t have
a cause, while the second is directed against the thesis that the question “Does the
world have a cause?” is meaningful. The fact that their conclusions, although related,
differ significantly from each other suggests that this is a disjunctive argument.
[48] R: No, it doesn’t need to be its own cause; what I’m saying is that the
concept of cause is not applicable to the total.

[49] C: Then you would agree with Sartre that the universe is what he calls
“gratuitous”?

[50] R: Well, the word “gratuitous” suggests that it might be something else. I
should say that the universe is just there, and that’s all.
Russell tries to explain to Copleston why he misunderstood him. In 48 he rejects
the argument based on the assertion (1) in 47, and in 50, showing his disagreement
with Sartre’s way of speaking, he rejects the argument based on the second assertion
in 47, which could be used, apparently, against whoever claims that the world is
“gratuitous”.
[51] C: Well, I can’t see how you can rule out the legitimacy of asking the
question how the total, or anything at all comes to be there. Why something
rather than nothing? That is the question. The fact that we gain our knowledge
of causality empirically, from particular causes, does not rule out the possibility
of asking what the cause of the series is. If the word “cause” were meaningless
or if it could be shown that Kant’s view of the matter were correct, the question
392 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

would be illegitimate I agree; but you don’t seem to hold that the word “cause”
is meaningless, and I do not suppose you are a Kantian.
Copleston asks Russell to justify his rejection of the legitimacy of “the question how
the total, or anything at all comes to be there”. When he points out that “The fact
that we gain our knowledge of causality empirically, from particular causes, does not
rule out the possibility of asking what the cause of the series is”, he is interpreting
Russell’s ‘ rebuttal in 46 as a substantive argument leading to the conclusion that the
question of the cause of the existence of all particular objects is meaningless.

The whole concept of cause is one we derive

from our observation of particular things
Therefore
The question “Why something rather than
nothing?” is illegitimate

Copleston goes then to reject the step from the premise to the conclusion, claiming
that “the fact that we gain our knowledge of causality empirically, from particular
causes, does not rule out the possibility of asking what the cause of the series is”.
[52] R: I can illustrate what seems to me your fallacy. Every man who exists has
a mother, and it seems to me your argument is: Therefore, the human race must
have a mother. But obviously the human race hasn’t a mother: that’s a different
logical sphere.
Russell does not accept Copleston’s challenge and attempts a new rebuttal, now
by parallel argument, to the argument in 44.

It is not the case that, since every man

who exists has a mother, the human race
must have a mother
It makes sense to ask for the
Therefore cause of the existence of
particular objects
It is not the case that, since every
particular object who exists has a cause, against Therefore
the world as a whole must have a cause
One can raise the question of the
cause of the existence of all
particular objects

The fallacy allegedly committed by Copleston would be to have taken an unwarranted

inferential step. As explained in Chap. 11, the warrant is that in virtue of which
something is a reason; thus, if there is no warrant connecting the premises to the
conclusion, these don’t give any reason for it.
14.1 The Argument from Contingency 393

[53] C: Well, I can’t really see a parity. If I were saying Every object has a phenom-
enal cause; therefore, the whole series has a phenomenal cause, there would be
a parity. But I’m not saying that; I’m saying, Every object has a phenomenal
cause—if you insist on the infinity of the series—but the series of phenomenal
causes is an insufficient explanation of the series; therefore, the series has not a
phenomenal cause but a transcendent cause.
Copleston rejects the relevance of the analogy proposed by Russell (“I can’t really
see any parity”) by resorting to a distinction between phenomenal and transcendental
causes. This rejection amounts to a plain rebuttal of the argument:

It is not the case that, since every man who exists has a
mother, the human race must have a mother
Therefore
It is not the case that, since every particular object who
exists has a cause, the world as a whole must have a cause

Russell’s analogy entitles him to draw the conclusion that it is not the case that, since
every particular object who exists has a phenomenal cause, the world as a whole
must have a phenomenal cause. However, his argument would run as follows:

Every object has a phenomenal cause. The series of phenomenal causes

is an insufficient explanation of the series
Therefore
The series of phenomenal causes has a transcendent cause

The absence of a comparable distinction between two types of mother would undo
the analogy invoked.
[54] R: Well, that’s always assuming that not only every particular thing in the
world, but the world as a whole must have a cause. For that assumption I see no
ground whatever. If you’ll give me a ground, I’ll listen to it.
If “that” refers to the Copleston’s reformulation in 53, Russell is saying that this
argument assumes that the world as a whole must have a cause, which probably refers
to the step from the premises to the conclusion.

Every object has a phenomenal cause. The

series of phenomenal causes is an insufficient
explanation of the series
The series as a whole must have a cause: Therefore
The series of phenomenal causes has a
transcendent cause

[55] C: Well, the series of events is either caused or it’s not caused. If it is caused,
there must obviously be a cause outside the series. If it’s not caused, then it’s
394 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

sufficient to itself, and, if it’s sufficient to itself, it is what I call necessary. But it
can’t be necessary since each member is contingent, and we’ve agreed that the
total has no reality apart from the members; therefore, it can’t be necessary.
Therefore, it can’t be uncaused. Therefore, it must have a cause. And I should
like to observe in passing that the statement The world is simply there and is
inexplicable can’t be got out of logical analysis.
Copleston tries to justify the questioned premise developing a dilemma: the series
of events is either caused or it’s not caused; if it is caused, there must obviously be
a cause outside the series. Otherwise:

Suppose that the

series of events it’s
not caused
If it’s not caused, then it’s
In this case
sufficient to itself:
Each member in
The series of events
but the series of events
is sufficient to itself
is contingent
If it’s sufficient to itself, it The total has no
is what I call necessary: Therefore reality apart from Therefore
the members:
The series of events The series of events
is necessary can’t be necessary
Therefore
The series of events can’t be uncaused

[56] R: I don’t want to seem arrogant, but it does seem to me that I can conceive
things that you say the human mind can’t conceive. As for things not having
a cause, the physicists assure us that individual quantum transitions in atoms
have no cause.
In the second part of this passage Russell develops the following argument:

Physicists assure us that individual quantum transitions in atoms do not have a cause
Therefore
Some contingent things do not have a cause

It is not clear how this argument is related to the previous argument in 55. From the
warrants in the argument on the left it follows by transitivity that if something is
not caused, then it is necessary, and then by contraposition that if something is not
necessary (i.e., is contingent), then it has a cause. Hence, Russell’s argument can be
seen as a rebuttal of the following “contraction” of the subordinative argument in 55:
14.1 The Argument from Contingency 395

Physicists assure us that

individual quantum transitions
in atoms do not have a cause
Suppose that the series
Therefore
of events it’s not caused
If something is not
Some contingent things do not
have a cause
against caused, then it is Therefore
necessary:
The series of events is
necessary

[57] C: Well, I wonder now whether that isn’t simply a temporary inference.
Copleston notes that physicists are not sure that individual quantum transitions in
atoms do not have a cause, which presumably weakens Russell’s counterargument.
(This could be an example of disclaimer by exception).
[58] R: It may be, but it does show that physicists’ minds can conceive it.
Russell concedes the previous point, noting that in order to challenge this warrant
there need not be an actual counterexample, it is sufficient that someone can conceive
one.

Some physicists tell us that

individual quantum transitions in
atoms have no cause
Suppose that the series
Therefore
of events it’s not caused
if something is not
Some physicists can conceive that
contingent events have no cause
against caused, then it is Therefore
necessary:
The series of events is
necessary

[59] C: Yes, I agree, some scientists—physicists—are willing to allow for inde-

termination within a restricted field. But very many scientists are not so willing.
I think that Professor Dingle, of London University, maintains that the Heisen-
berg uncertainty principle tells us something about the success (or the lack of it)
of the present atomic theory in correlating observations, but not about nature
in itself, and many physicists would accept this view. In any case, I don’t see
how physicists can fail to accept the theory in practice, even if they don’t do so
in theory. I cannot see how science could be conducted on any other assumption
than that of order and intelligibility in nature. The physicist presupposes, at
least tacitly, that there is some sense in investigating nature and looking for the
causes of events, just as the detective presupposes that there is some sense in
looking for the cause of a murder. The metaphysician assumes that there is sense
in looking for the reason or cause of phenomena, and, not being a Kantian, I
consider that the metaphysician is as justified in his assumption as the physicist.
396 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

When Sartre, for example, says the world is gratuitous, I think that he has not
sufficiently considered what is implied by “gratuitous”.
Now it’s Copleston who grants Russell’s premise. First he tries to undermine the argu-
ment from authority used by Russell. The argument is inconclusive because Professor
Dingle maintains that Heisenberg’s principle of indeterminacy is not applicable to
nature itself, and therefore does not say that there are events that are uncaused, and
therefore cannot be used as evidence that this is conceivable.

Professor Dingle, of London University,

maintains that the Heisenberg uncertainty
principle tells us nothing about nature in itself,
and many physicists would accept this view
Therefore
It is possible that the Heisenberg uncertainty
principle tells us nothing about nature in itself
Some physicists tell us that
individual quantum transitions in
atoms have no cause
against
Therefore
Some physicists can conceive that
contingent events have no cause

Next, he changes his strategy (“in any case”), moving from rebuttal to refutation of
Russell’s rebuttal in 56.

Some physicists tell us that The physicist presupposes, at least tacitly,

individual quantum transitions in
atoms have no cause
Therefore
Some physicists can conceive that
contingent events have no cause
but that there is some sense in investigating
nature and looking for the causes of events
Therefore
The physicist can’t conceive of uncaused
events

[60] R: I think there seems to me a certain unwarrantable extension here. A

physicist looks for causes; that does not necessarily imply that there are causes
everywhere. A man may look for gold without assuming that there is gold every-
where; if he finds gold, well and good; if he doesn’t, he’s had bad luck. The same
is true when the physicists look for causes. As for Sartre, I don’t profess to know
what he means, and I shouldn’t like to be thought to interpret him; but for my
part, I do think the notion of the world having an explanation is a mistake. I
don’t see why one should expect it to have < one > , and I think what you say
about what the scientist assumes is an over-statement.
Once again, Russell offers a rebuttal (there is an “unwarrantable extension”) by
parallel argument. The target is Copleston’s refutation in 59.
14.1 The Argument from Contingency 397

A man may look for gold without

assuming that there is gold
everywhere
The physicist presupposes, at least tacitly, that
Likewise there is some sense in investigating nature and
looking for the causes of events
A physicist may look for causes
without assuming that there are against Therefore
causes everywhere
The physicist can’t conceive uncaused events

[61] C: Well, it seems to me that the scientist does make some such assumption.
When he experiments to find out some particular truth, behind that experiment
lies the assumption that the universe is not simply discontinuous. There is the
possibility of finding out a truth by experiment. The experiment may be a bad
one, it may lead to no result, or not to the result that he wants, but at any rate
there is the possibility, through experiment, of finding out the truth that he
assumes. And that seems to me to assume an ordered and intelligible universe.
The claim is stated twice: once at the beginning of the passage, with the supporting
reason immediately following after a full stop, and then at the end, preceded by “And
that seems to”, which functions here as a marker of recapitulation.

When the scientist experiments to find out some particular truth, behind that
experiment lies the assumption that the universe is not simply discontinuous
Therefore
The scientist assumes an ordered and intelligible universe

There are two possible interpretations of this passage. The first is that Copleston
simply formulates his argument more perspicuously, fearing that Russell has misun-
derstood or misinterpreted it. Perhaps he feels that this is the reason why Russell
thinks that the example of individual quantum transitions in atoms ruins his argu-
ment. The second is that Copleston intends to give some support to the premise that
the physicist presupposes that there is some sense in investigating nature and looking
for the causes of events, mistakenly believing that it has been attacked by Russell.
398 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

When the scientist experiments to find out some particular truth, behind that
experiment lies the assumption that the universe is not simply discontinuous
Therefore
The physicist presupposes, at least tacitly, that there is some sense in
investigating nature and looking for the causes of events
Therefore
The physicist can’t conceive uncaused events

[62] R: I think you’re generalizing more than is necessary. Undoubtedly the

scientist assumes that this sort of thing is likely to be found and will often be
found. He does not assume that it will be found, and that’s a very important
matter in modern physics.
Russell reiterates his previous rebuttal.
[63]C: Well, I think he does assume or is bound to assume it tacitly, in practice.
It may be that, to quote Professor Haldane, “when I light the gas under the
kettle, some of the water molecules will fly off as vapor, and there is no way of
finding out which will do so”, but it doesn’t follow necessarily that the idea of
chance must be introduced except in relation to our knowledge.
Copleston returns to the rebuttal in 56, alleging that “It may be that there is no way
of finding out the cause of individual quantum transitions in atoms” does not follow
necessarily from “The physicists assure us that individual quantum transitions in
atoms have no cause”. In order for this remark to be relevant, the rebutted rebuttal
must be rephrased in epistemic terms:

Physicists assure us that there is no

way of finding out the cause of
individual quantum transitions in
atoms
Suppose that the
Therefore series of events it’s
not caused
Some contingent things do not have a If it’s not caused,
against Therefore
cause then it’s necessary:
The series of events
is necessary

[64] R: No, it doesn’t—at least if I may believe what he says. He’s finding out quite
a lot of things… the scientist is finding out quite a lot of things that are happening
in the world, which are, at first, beginnings of causal chains: first causes which
haven’t in themselves got causes. He does not assume that everything has a
cause.
14.1 The Argument from Contingency 399

Russell concedes that the inference in the rebuttal in 63 is unwarranted, as Copleston

holds, but he insists that the physicist does not assume that everything has a cause.
Indeed, it seems that Russell does not identify this rebuttal with his counterargument
in 56. In any case, he offers a new argument against the claim that the scientist
assumes an ordered and intelligible universe:

The scientist is finding out quite a lot of things that are happening in the world, which are,
at first, beginnings of causal chains: first causes which haven’t in themselves got causes
Therefore
The physicist does not assume that everything has a cause

[65] C: Surely that’s a first cause within a certain selected field. It’s a relatively
first cause.
Copleston nuances the premise of the preceding argument by Russell, saying that the
first causes found by scientists are only relatively so, within a specific field. Since
they both agree (see also 66 below) that relative first causes do not allow for the
conclusion that some contingent things have no cause, the nuance seems to count
as a rebuttal to the argument of 64. That is, Copleston uses dissociation and splits
up a notion considered by Russell to form a unitary concept into two new notions:
relative and absolute first causes, so to speak. Thus the argument in 64 becomes:

The scientist is finding out relative first causes which haven’t in themselves got
relative first causes
Therefore
The physicist does not assume that everything has an absolute first cause

The warrant of the original argument - if scientists discover things in the world that
are uncaused (absolute) first causes, then scientists do not assume that everything
has an absolute first cause- does not cover this argument, which is thus rebutted.
[66] R: I don’t think he’d say so. If there’s a world in which most events, but not
all, have causes, he will then be able to depict the probabilities and uncertainties
by assuming that this particular event you’re interested in probably has a cause.
And since in any case you won’t get more than probability that’s good enough.
Russell rejects Copleston’s nuance and argues that the practice of physics does not
presuppose that all events do have a cause, a variation of their previous arguments.
400 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

If there’s a world in which most events, but not all, have causes, the physicist will
then be able to depict the probabilities and uncertainties by assuming that this
particular event you’re interested in probably has a cause. And since in any case
you won’t get more than probability that’s good enough
Therefore
Physics could be conducted without assuming that everything has a cause

[67] C: It may be that the scientist doesn’t hope to obtain more than probability,
but in raising the question he assumes that the question of explanation has a
meaning. But your general point then, Lord Russell, is that it’s illegitimate even
to ask the question of the cause of the world?
[68] R: Yes, that’s my position.
[69] C: Well, if it’s a question that for you has no meaning, it’s of course very
difficult to discuss it, isn’t it?
[70] R: Yes, it is very difficult. What do you say—shall we pass on to some other
issue?
Copleston slightly rephrases his argument:

The physicist presupposes, at least tacitly, that there is some

sense in investigating nature and looking for the causes of events
Therefore
The physicist presupposes that the question of explanation (of the
world as a whole?) has meaning

He then acknowledges that they are in a deep disagreement. Engaging in an argumen-

tative exchange, presupposes a background of shared commitments. Robert Fogelin
defines deep disagreements as follows:
But we get a very different sort of disagreement when it proceeds from a clash in under-
lying principles. Under these circumstances, the parties may be unbiased, free of prejudice,
consistent, coherent, precise and rigorous, yet still disagree. And disagree profoundly, not
just marginally. Now when I speak about underlying principles, I am thinking about what
others (Putnam) have called framework propositions or what Wittgenstein was inclined to call
rules. We get a deep disagreement when the argument is generated by a clash of framework
propositions. (Fogelin 2005, p.8)

Fogelin holds that deep disagreements cannot be resolved through the use of argu-
ment, for they undercut the conditions essential to arguing and in this case Russell
and Copleston seem to agree on. Consequently, they decide to pass on to some other
issue.
Although the usual examples of deep disagreements concern practical issues -
such as abortion or reverse discrimination- the question of the cause of the world
belongs to the domain of theoretical reason.
14.2 Religious Experience 401

14.2 Religious Experience

[71] C: Let’s. Well, perhaps I might say a word about religious experience,
and then we can go on to moral experience. I don’t regard religious experience
as a strict proof of the existence of God, so the character of the discussion
changes somewhat, but I think it’s true to say that the best explanation of it is
the existence of God. By religious experience I don’t mean simply feeling good.
I mean a loving, but unclear, awareness of some object which irresistibly seems
to the experiencer as something transcending the self, something transcending
all the normal objects of experience, something which cannot be pictured or
conceptualized, but of the reality of which doubt is impossible—at least during
the experience. I should claim that cannot be explained adequately, and without
residue, simply subjectively. The actual basic experience at any rate is most
easily explained on the hypothesis that there is actually some objective cause of
that experience.
Copleston proposes two new issues to be successively addressed: religious
experience and moral experience.
Copleston distinguishes strict proof from argument. The idea seems to be that in
a strict proof the premises deductively entail the conclusion, while in an argument
they favor the conclusion, so to speak. This distinction is reminiscent of Perelman’s
opposition of demonstration and argument. “Demonstration,” says Perelman, “is a
calculation made in accordance with the rules that have been laid down before hand”
(Perelman 1979, p.10),while argumentation is a tool to “induce or increase the mind’s
adherence to the theses presented for its assent.“ (Perelman & Olbrechts-Tyteca 1969,
p.14).
When Copleston warns that religious experience is not strictly proof of God’s
existence, he is implying that the argument from contingency is a proof (as he says
in 95).
Copleston begins with an argument from religious experience, which he explic-
itly presents as abductive. This is coherent with the previous distinction, since a
widespread classification divides inferences and arguments into deductive, inductive
and abductive. Sometimes the differences among these three types of inferences are
explained in terms of modal qualifiers:
– In a deductive inference the truth of the conclusion follows necessarily from the
truth of the premises.
– In an inductive inference the truth of the conclusion follows probably from the
truth of the premises.
– In an abductive inference the truth of the conclusion follows plausibly from the
truth of the premises
Although abduction is most commonly conceived as is a distinctive type of inference
or reasoning, here we are interested in abduction mainly as an argument scheme.
Olmos offers the following definition:
402 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

Presenting an abductive argument means, thus, to offer at least a minimally justifying reason
for a hypothesis (which is a doubtful content, what is at stake) adducing the factive quality
of the data, for which the hypothesis is deemed to be explanatory. This means that it is
suggested that the hypothesis could be used (in a different act and setting, allegedly once
accepted) as an explanans or explanatory reason for the acknowledged data. (Olmos 2019b,
p.7)

And she gives the following basic diagram for abductive arguments:

Data: Shared (usually empirical)

data
Warrant: Hypothesis could explain data Therefore
Claim:Hypothesis

Copleston’s argument from religious experience fits this pattern perfectly:

Some people experience a loving, but unclear, awareness

of some object which irresistibly seems to them as
something transcending the self, something transcending
all the normal objects of experience, something which
cannot be pictured or conceptualized, but of the reality of
which doubt is impossible
This experience is most easily
explained on the hypotheses that
Therefore
there is actually some objective
cause of that experience:
God exists

[72] R: I should reply to that line of argument that the whole argument from our
own mental states to something outside us, is a very tricky affair. Even where
we all admit its validity, we only feel justified in doing so, I think, because of
the consensus of mankind. If there’s a crowd in a room and there’s a clock in a
room, they can all see the clock. The fact that they can all see it tends to make
them think that it’s not an hallucination: whereas these religious experiences
do tend to be very private.
Olmos says explicitly that the premises in an abductive argument must be shared data.
Russell objection is just that religious experiences are private, so that their contents
cannot be considered as shared data. Russell classifies Copleston’s argument as an
“argument from our own mental states to something outside us,” and points out that
any such argument presupposes the consensus of mankind, a condition that fails in
the present case due to the private character of religious experiences. We are therefore
faced with a rebuttal by exception: such experiences are most easily explained on the
hypothesis that there is actually some objective cause of that experience, unless they
are private. Despite the brevity of the reply, Russell develops a relatively complex
argument since it includes the backing of the warrant of the rebutting argument.
14.2 Religious Experience 403

If there’s a crowd
in a room and
there’s a clock in a
room, they can all
see the clock. The
fact that they can
all see it tends to
make them think
that it’s not an
hallucination
These religious
therefore experiences do tend
to be very private
It is only because
Some people experience a
they are shared
loving, but unclear, awareness
that these
of some object which
experiences are
Therefore irresistibly seems to them as
explained on the
something transcending the
hypotheses that
self, … the reality of which
they have some
doubt is impossible
objective cause:
These religious This experience is
experiences are not most easily
explained on the explained on the
against

hypothesi that there hypotheses that there Therefore

is actually some is actually some
objective cause of objective cause of
that experience that experience:
God exists

[73] C: Yes, they do. I’m speaking strictly of mystical experience proper, and
I certainly don’t include, by the way, what are called visions. I mean simply
the experience, and I quite admit it’s indefinable, of the transcendent object or
of what seems to be a transcendent object. I remember Julian Huxley in some
lecture saying that religious experience, or mystical experience, is as much a real
experience as falling in love or appreciating poetry and art. Well, I believe that,
when we appreciate poetry and art, we appreciate definite poems or a definite
work of art. If we fall in love, well, we fall in love with somebody and not with
nobody.
Copleston distinguishes visions from mystical experiences proper and seems to limit
the private character to the former. To argue that the religious experiences to which
his argument appeals are not visions but genuine experiences, and therefore immune
to Russell’s criticism, he uses an argument by double analogy:
404 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

When we appreciate poetry and art we

appreciate definite poems or a definite work
of art. If we fall in love, well, we fall in
love with somebody and not with nobody
Religious experience, or mystical experience, is
as much a real experience as falling in love or Therefore
appreciating poetry and art:
When we have a mystical experience
proper, we have a mystical experience of
the transcendent object or of what seems to
be a transcendent object

[74] R: May I interrupt for a moment here. That is by no means always the case.
Japanese novelists never consider that they have achieved a success unless large
numbers of real people commit suicide for love of the imaginary heroine.
Russell raises a (partial) objection to Copleston’s analogy:

Japanese novelists never consider that

they have achieved a success unless
large numbers of real people commit
suicide for love of the imaginary heroine
Therefore
It is not always the case that if we fall in If we fall in love, well, we fall in love
love, well, we fall in love with against with somebody and not with nobody
somebody and not with nobody
Therefore
When we have a mystical experience
proper, we have a mystical experience
of the transcendent object or of what
seems to be a transcendent object

The use of the phrase “May I interrupt for a moment here” to ask for permission
to raise an objection, without waiting for Copleston to complete his argumentation
confirms to a certain extent that, as I have argued in 11.10, objection precedes any
other form of counter-argumentation.
[75] C: Well, I must take your word for these goings on in Japan. I haven’t
committed suicide, I’m glad to say, but I have been strongly influenced in the
taking of two important steps in my life by two biographies. However, I must
say I see little resemblance between the real influence of those books on me and
the mystical experience proper, so far, that is, as an outsider can obtain an idea
of that experience.
Copleston offers an argument by similarity, presumably aimed at blocking the infer-
ence, from the conclusion of Russell’s objection in 74, “It is not always the case that
if we fall in love, well, we fall in love with somebody and not with nobody”, to the
14.2 Religious Experience 405

conclusion “It is by no means always the case that we have a mystical experience
proper, we have a mystical experience of a transcendent object”. That is, Copleston
does not answer Russell’s objection to his second argument, but anticipates a possible
rebuttal to his first argument; namely,

It is not always the case that if we fall in

love, well, we fall in love with somebody and
not with nobody
Religious or mystical, experience is an
Therefore
experience of the same kind as falling in love:
It is not always the case that we have a
mystical experience proper, we have a
mystical experience of a transcendent object

The danger of the opponent twisting an analogy to refute its proponent is analyzed
by Perelman and Olbrechts-Tyteca in The New Rhetoric:
This method of refutation assumes that one has always the right to extend an analogy beyond
its original statement and that, if, as a result of being extended, it works against its author or
becomes inadequate, it was already inadequate at the start (§.85).

Copleston therefore anticipates this possible argument and rejects it:

Love of readers of Japanese novels for imaginary heroines is like the influence
of two biographies in Copleston’s life. There is little resemblance between the
real influence of those books on Copleston and the mystic experience proper
Therefore
There is little resemblance between falling in love with imaginary heroines of
Japanese novels and the mystic experience proper

[76] R: Well, I mean we wouldn’t regard God as being on the same level as the
characters in a work of fiction. You’ll admit there’s a distinction here?
Russell’s reply is at first sight disconcerting, since the comparison of God with
a fictional character, if it has been proposed by someone, has been proposed by
Russell himself. Perhaps it should be understood that, in order to accept God as
an adequate explanation of religious experience, it must be established that God is
something very different from a fictional character (which Russell has argued may
have a certain causal effectiveness).
[77] C: I certainly should. But what I’d say is that the best explanation seems to
be the not purely subjectivist explanation. Of course, a subjectivist explanation
is possible in the case of certain people in whom there is little relation between
the experience and life, in the case of deluded people and hallucinated people,
and so on. But when you get what one might call the pure type, say St. Francis
406 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

of Assisi, when you get an experience that results in an overflow of dynamic and
creative love, the best explanation of that, it seems to me, is the actual existence
of an objective cause of the experience.
After the digression in 74–76, Copleston goes back to 71. To develop his argument,
he tries to justify why the existence of God is a better explanation of mystical or
religious experiences than a subjectivist explanation (such as the account that could
be given for an impression made on a subject by a work of fiction). Copleston admits
that an objectivist explanation is not preferable “in the case of certain people in whom
there is little relation between the experience and life, in the case of deluded people
and hallucinated people,” although he maintains that it is “when you get an experi-
ence that results in an overflow of dynamic and creative love”. That is, Copleston
seems to propose an intensifier or weighting factor that can be expressed as follows:
under normal conditions of perception, an objectivist explanation is preferable to a
subjectivist explanation. This principle corresponds to the only undercutter for the
argument from perception by Walton, Reed and Macagno: «Are the circumstances
such that having a ϕ image is not a reliable indicator of ϕ?» (2008, p. 345).
Copleston’s answer raises the question of whether Russell can be understood as
suggesting an alternative, subjectivist explanation of religious experience, and thus
a refutation of the argument in 71, which makes weighting necessary. For example,
Copleston may think that Russell was arguing like this:

Some people experience a loving, but unclear,

awareness of some object which irresistibly seems
to them as something transcending the self,
something transcending all the normal objects of
experience, something which cannot be pictured
or conceptualized, but of the reality of which
doubt is impossible
This experience can be explained along
the lines of the influence of fictional Therefore
characters in the life or persons:
God doesn’t exist

This hypothesis fits in with the way the conversation evolves, as we will see below.
[78] R: Well, I’m not contending in a dogmatic way that there is not a God.
What I’m contending is that we don’t know that there is. I can only take what
is recorded as I should take other records, and I do find that a very great many
things are reported—and I am sure you would not accept things about demons
and devils and what not—and they’re reported in exactly the same tone of voice
and with exactly the same conviction. And the mystic, if his vision is veridical,
may be said to know that there are devils. But I don’t know that there are.
Russell begins his intervention with a statement in which he rejects the subjectivist
argument that according to the interpretation suggested above Copleston would
14.2 Religious Experience 407

attribute to him. His argumentative strategy is not cancelling refutation (i.e., it is

not aimed at establishing the non-existence of God), but rebuttal (showing that the
conclusion God exists cannot reasonably be inferred from Copleston’s premises). To
make it clear, Russell draws upon a rebuttal by parallel argument.

You would not accept mystic experiences

about demons and devils as a reason to believe
that there are demons and devils
They’re reported in exactly the same tone of
Therefore
voice and with exactly the same conviction:
You must not accept mystic experiences about
God as a reason to believe that God exists

[79] C: But surely in the case of the devils there have been people speaking
mainly of visions, appearance, angels or demons and so on. I should rule out
the visual appearances, because I think they can be explained apart from the
existence of the object which is supposed to be seen.
As might be expected, Copleston rejects the analogy, pointing out that in the case of
angels and demons there are, above all, visions, and that visions are not explained as
mystical experiences, something he had already said in 73.
[80] R: But don’t you think there are abundant recorded cases of people who
believe that they’ve heard Satan speaking to them in their hearts, in just the same
way as the mystics assert God? And I’m not talking now of an external vision,
I’m talking of a purely mental experience. That seems to be an experience of the
same sort as mystics’ experience of God, and I don’t see that from what mystics
tell us you can get any argument for God which is not equally an argument for
Satan.
Russell insists on the warrant of his argument of 78 (i.e. that Satanic experiences and
God experiences are similar in the relevant respects), a warrant that Copleston has
just rejected, separating the backing from the warrant.5 The result is the following
meta-argument:

People who believe that they’ve heard Satan speaking to them in their hearts
have an experience of the same sort as mystics’ experience of God
Therefore
Any argument for God from what mystics tell us is equally an argument for Satan

Putting all these pieces together, we have the following diagram:

5Toulmin (2003 [1958], pp.101–103) explains how the same argument can be expanded either
making its warrant explicit or stating the backing for the warrant, which is itself left unstated.
408 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

People who believe that they’ve heard

Satan speaking to them in their hearts have
an experience of the same sort as mystics’
experience of God
You would not accept mystic experiences
Therefore about demons and devils as a reason to believe
that there are demons and devils
Any argument for God from what mystics
Therefore
tell us is equally an argument for Satan:
You must not accept mystic experiences about
God as a reason to believe that God exists

[81] C: I quite agree, of course, that people have imagined or thought they have
heard or seen Satan. And I have no wish in passing to deny the existence of
Satan. But I do not think that people have claimed to have experienced Satan in
the precise way in which mystics claim to have experienced God. Take the case of
a non-Christian, Plotinus. He admits the experience is something inexpressible,
the object is an object of love, and therefore, not an object that causes horror and
disgust. And the effect of that experience is, I should say, borne out, or I mean
the validity of the experience is borne out in the records of the life of Plotinus.
At any rate it is more reasonable to suppose that he had that experience, if
we’re willing to accept Porphyry’s account of Plotinus’ general kindness and
benevolence.
Copleston clarifies that he accepts both the premise and the conclusion of what would
be the source of the counter-analogy in 78:

People have imagined or thought they have heard or seen Satan

Therefore
Satan exists

What he denies (“But I do not think”) is that these experiences count as a reason
in favor of the existence of Satan: “I do not think that people have claimed to have
experienced Satan in the precise way in which mystics claim to have experienced
God”. Thus, Copleston holds this argument is significantly different from the target
of the counter-analogy:

Some people experience a loving, but unclear, awareness of some object

which irresistibly seems to them as something transcending the self, …
Therefore
God exists

Copleston slightly modifies his argument, moving from a general argument from
perception to an argument from example, appealing to Plotinus’ religious experiences
14.2 Religious Experience 409

(this change is marked by “Take the case of”). He argues first that those who claim
to have had an experience of Satan describe an experience very different from that
of the mystics who claim to have experienced God.

Plotinus felt an inexpressible experience of an

object of love
Therefore
Plotinus did not felt an inexpressible Those who claim to have had
experience of an object that causes horror and experienced Satan felt an experience of
disgust an object that causes horror and disgust
Therefore
Those who claim to have had an experience of Satan describe an experience very
different from that of the mystics who claim to have experienced God

This is intended to be an attack to the backing of the warrant of Russell’s argument in

80. Copleston then goes on to clarify that the validity of the experience he is talking
about is related to the distinction between visions and proper experiences, mentioned
in 73. A valid experience is a genuine experience, and its validity consists in the
fact that it is an experience of something, so that, if an experience is valid, one can
reasonably infer the existence of its object. The main novelty is that Copleston seems
to claim that its effects allow to distinguish a valid experience from a hallucination.

Porphyry reports that Plotinus’

general kindness and
benevolence were effects of
his inexpressible experience of
an object of love
Therefore
Plotinus felt an inexpressible Plotinus’ general kindness and
experience of an object of benevolence were effects of
love his experience
It is more reasonable to
suppose that any experience
Therefore
with such effects is a valid
experience:
Plotinus experienced an object of love

[82] R: The fact that a belief has a good moral effect upon a man is no evidence
whatsoever in favour of its truth.
The warrant of Copleston’s argument in 81 can be formulated in more detail as
follows: the effects on the subject of a valid experience entitle us to infer that it is
an experience of an object. Russell identifies those effects as positive moral effects,
and then rejects the warrant on the grounds that the morally positive effects of an
experience are not indicative of its truth. The truth of an experience, as Russell speaks,
410 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

seems to consist, in part, in the fact that it is an experience of an external object (that
it is, so to speak, an intentional experience).
[83] C: No, but if it could actually be proved that the belief was actually respon-
sible for a good effect on a man’s life, I should consider it a presumption in
favour of some truth, at any rate, of the positive part of the belief, not of its
entire validity. But in any case, I am using the character of the life as evidence
in favour of the mystic’s veracity and sanity rather than as a proof of the truth
of his beliefs.
Copleston’s reply has two parts, separated by “but in any case”, which marks a
change in his argumentative strategy. Copleston first confirms the warrant of his
previous argument, which he now makes fully explicit: the fact a belief is actually
responsible for a good effect on a man’s life counts as a presumption in favor of its
truth. He further introduces a distinction between the positive part of a belief and
its full validity, which he does not explain. My guess is that the positive part is the
existence of the object of the belief (x exists) and full validity requires that the object
in question possess the properties that the belief attributes to it (x is so and so).
Given that Russell does not accept that the moral effects of a belief have to do with
its truth, Copleston prefers to argue differently, rephrasing his argument. According
to his clarifications, he did not intend to establish the warrant of his argument, but
to anticipate possible counterarguments that would question the mystical truthful-
ness and mental health of those with religious experience. This movement involves
interpreting the argument as an argument from testimony, since a typical exception to
this type of arguments is the lack of veracity of the witness. The argument in 81 thus
becomes:

Plotinus felt an inexpressible experience of an object of love

Plotinus is a reliable witness: Therefore
Plotinus experienced an object of love

Here are two possible rebuttals by exception to this argument:

The veracity of Plotinus is dubious Plotinus was insane

Therefore Therefore
Plotinus is not a reliable witness Plotinus is not a reliable witness

Copleston rejects them because they are both based on false premises:

Plotinus was kind and benevolent

Therefore
Plotinus was truthful and sane
14.2 Religious Experience 411

[84] R: But even that I don’t think is any evidence. I’ve had experiences myself
that have altered my character profoundly. And I thought, at the time at any rate,
that it was altered for the good. Those experiences were important, but they did
not involve the existence of something outside me, and I don’t think that if I’d
thought they did, the fact that they had a wholesome effect would have been any
evidence that I was right.
Although Russell claims to accept Copleston’s new approach, he does not actually
endorse the argument, as he continues to attack the warrant of the argument in 82,
which he now challenges by appeal to his own experiences:

I’ve had experiences myself that have altered my character profoundly. And I thought
at the time at any rate that it was altered for the good. Those experiences were
important, but they did not involve the existence of something outside me, and I don’t
think that if I’d thought they did, the fact that they had a wholesome effect would have
been any evidence that I was right
Therefore
The fact that an experience has a wholesome effect is not evidence of its truth

[85] C: No, but I think that the good effect would attest your veracity in
describing your experience. Please remember that I’m not saying that a
mystic’s meditation or interpretation of his experience should be immune from
discussion or criticism.
Copleston accepts the conclusion of Russell’s argument, while pointing out that his
rebuttal is not in line with the reformulation proposed in 83.
[86] R: Obviously the character of a young man may be—and often is—
immensely affected for good by reading about some great man in history, and it
may happen that the great man is a myth and doesn’t exist, but the boy is just
as much affected for good as if he did. There have been such people. Plutarch’s
Lives—take Lycurgus as an example, who certainly did not exist, but you might
be very much influenced by reading < Plutarch’s life of > Lycurgus under the
impression that he had previously existed. You would then be influenced by an
object that you’d loved, but it wouldn’t be an existing object.
Russell continues to disregard Copleston’s clarifications, insisting that a person can
be influenced by his beliefs about a fictional character whom he takes for real, so the
fact that his religious experiences made Plotinus kind and benevolent is not a reason
to believe that God exists. Russell rebuts the warrant of the argument in 81 using a
hypothetical argument from an example:
412 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

Plutarch’s Lives take Lycurgus as an example,

who certainly did not exist, but you might be
very much influenced by reading Lycurgus
under the impression that he had previously
existed.
Therefore
You would then be influenced by an object
that you’d loved, but it wouldn’t be an
existing object:
Plotinus experience of an object of
Therefore
love made him kind and benevolent

You may be influenced by an object that you

against Therefore
love, but doesn’t exist

This object of love (God) exists

[87] C: I agree with you on that, of course—that a man may be influenced by

a character in fiction. Without going into the question of what it is precisely
that influences him (I should say a real value), I think that the situation of
that man and <the situation of> of the mystic are different. After all, the man
who is influenced by Lycurgus hasn’t got the irresistible impression that he’s
experienced in some way the ultimate reality.
Copleston interprets Russell’s rebuttal as a counter-analogy, not as an argument from
example. Furthermore, to invalidate it, he returns to his initial argument:

You might be very much influenced by reading Lycurgus under the impression
that he had previously existed, although he is a legendary character
So likewise
You might be very much influenced by mystic experiences about God under
the impression that he exists, even if it were a legendary character

Copleston contends that this counter-analogy does not apply to its initial argument,
which ran as follows:

The mystic experiences a loving, but unclear, awareness of

some object which irresistibly seems to them as something
transcending the self, something transcending all the normal
objects of experience, something which cannot be pictured or
conceptualized, but of the reality of which doubt is impossible
God’s existence is an
adequate explanation of Therefore
these religious experiences:
God exists
14.2 Religious Experience 413

Now the emphasis is on the impossibility of doubting the reality of such an experience,
which relates Copleston’s position to the idea that the criterion of truth for perceptions
is their clarity and evidence (as is the Stoic concept of cataleptic representation).
Lycurgus-like counterexamples do not affect this argument, because for this to be
the case the premise would have to be something like this: “The man who is influenced
by Lycurgus has the irresistible impression that he’s influenced by a real character”.
This remark allows us to see how Copleston understands the warrant of his argument.
“God’s existence is an adequate explanation of these religious experiences” would
be a particular case of the more general principle “The existence of the object is an
adequate explanation of any feeling of awareness of an object the reality of which
doubt is impossible”.
[88] R: I don’t think you’ve quite got my point about these historical charac-
ters—these unhistorical characters in history. I’m not assuming what you call
an effect on the reason. I’m assuming that the young man reading about this
person, and believing him to be real, loves him—which is quite easy to happen,
and yet he’s loving a phantom.

[89] C: In one sense he’s loving a phantom; that’s perfectly true: in the sense,
I mean, that he’s loving X or Y who doesn’t exist. But at the same time, it is
not, I think, the phantom as such that the young man loves; he perceives a real
value, an idea, which he recognizes as objectively valid, and that’s what excites
his love.

[90] R: Well, in the same sense we had before about the characters in fiction.

[91] C: Yes, in one sense the man’s loving a phantom—perfectly true. But, in
another sense, he’s loving what he perceives to be a value.
The discussion reaches, again, an impasse: Russell reaffirms that the cases compared
are analogous, while Copleston insists that they are not. Copleston claims that the
young person who loves an imaginary being loves something he perceives as real,
and that it is not a person, but “a real value, an idea which he recognizes as objectively
valid”. This mention of values gives way to the discussion of the moral argument.

14.3 The Moral Argument

[92] R: But aren’t you now saying in effect: I mean by God whatever is good or
the sum total of what is good—the system of what is good, and, therefore, when
a young man loves anything that is good, he is loving God. Is that what you’re
saying? Because, if so, it wants a bit of arguing.

[93] C: I don’t say, of course, that God is the sum-total or system of what is good
in the pantheistic sense. I’m not a pantheist. But I do think that all goodness
414 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

reflects God in some way and proceeds from Him, so that in a sense the man who
loves what is truly good, loves God even if he doesn’t advert to God. But still, I
agree that the validity of such an interpretation of a man’s conduct depends on
the recognition of God’s existence, obviously.

[94] R: Yes, but that’s a point to be proved.

[95] C: Quite so, but I regard the metaphysical argument as probative, but there
we differ.
[96] R: You see, I feel that some things are good and that other things are bad. I
love the things that are good, that I think are good, and I hate the things that I think
are bad. I don’t say that these things are good because they participate in the Divine
goodness.
[97] C: Yes, but what’s your justification for distinguishing between good and
bad or how do you view the distinction between them?

[98] R: I don’t have any justification any more than I have when I distinguish
between blue and yellow. What is my justification for distinguishing between
blue and yellow? I can see they are different.

[99] C: Well, that is an excellent justification, I agree. You distinguish blue and
yellow by seeing them, so you distinguish good and bad by what faculty?

[100] R: By my feelings.

[101] C: By your feelings. Well, that’s what I was asking. You think that good
and evil have reference simply to feeling?

[102] R: Well, why does one type of object look yellow and another look blue? I
can more or less give an answer to that, thanks to the physicists, and as to why
I think one sort of thing good and another evil, probably there is an answer of
the same sort, but it hasn’t been gone into in the same way and I couldn’t give
it you.
This exchange establishes Russell’s thesis that will be examined next: what
distinguishes good things from bad things is our feelings about them.
[103] C: Well, let’s take the behaviour of the Commandant of Belsen. That
appears to you as undesirable and evil and to me too. To Adolf Hitler we suppose
it appeared as something good and desirable. I suppose you’d have to admit that
for Hitler it was good and for you it is evil.
Copleston begins to develop a reductio ad Hitlerum, using as a warrant the rule
conceded by Russell in his previous interventions.
14.3 The Moral Argument 415

The behavior of the Commandant of Belsen appears

to you as undesirable and evil. To Adolf Hitler we
suppose it appeared as something good and desirable.
We distinguish good and
Therefore
bad by our feelings:
The behavior of the Commandant of Belsen for
Hitler was good and for you was evil

[104] R: No, I shouldn’t quite go so far as that. I mean, I think people can make
mistakes in that, as they can in other things. If you have jaundice, you see things
yellow that are not yellow. You’re making a mistake.
Russell rejects the conclusion of Copleston’s argument, namely that the behavior of
the Commandant of Belsen’s was good for Hitler, arguing that Hitler was wrong to
judge the behavior of the Commandant of Belsen as good and desirable. Certainly
the thesis that we distinguish good and bad by our feelings do not entail that we can’t
be wrong when we feel something is good or bad.
[105] C: Yes, one can make mistakes, but can you make a mistake if it’s simply
a question of reference to a feeling or emotion? Surely Hitler would be the only
possible judge of what appealed to his emotions.
Copleston replies that one cannot judge right and wrong based on feelings alone. In
argumentative terms, what he points out is that the rule “good or bad is what we feel
as good or bad” leaves no room for exceptions of the type “unless we are wrong”.
[106] R: It would be quite right to say that it appealed to his emotions, but you
can say various things about that, among others that, if that sort of thing makes
that sort of appeal to Hitler’s emotions, then Hitler makes quite a different
appeal to my emotions.

[107] C: Granted. But there’s no objective criterion outside feeling then for
condemning the conduct of the Commandant of Belsen, in your view?

[108] R: No more than there is for the color-blind person who’s in exactly the
same state. Why do we intellectually condemn the color-blind man? Isn’t it
because he’s in the minority?
The rule “good or bad is what we feel as good or bad” admits two readings: an
individualistic reading, according to which what is good or bad for each of us is what
each of us feels as such, and a collectivistic reading, according to which what is good
or bad for us as a whole is what we as a group feel as such. Copleston’s reading is
the first, Russell’s the second.
Russell justifies the assertion that Hitler was wrong to regard the behavior of the
Belsen commander as good by arguing that most people regard the conduct of the
concentration camp commander as bad. This involves the collectivistic reading of the
warrant of moral judgments. To support this interpretation of the rule, he compares
416 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

it to this one: something is of a certain color if and only if it looks this color for the
majority of us. That’s why the color-blind person is wrong.
[109] C: I would say because he is lacking in a thing which normally belongs to
human nature.

[110] R: Yes, but if he were in the majority, we shouldn’t say that.

[111] C: Then you’d say that there’s no criterion outside feeling that will enable
one to distinguish between the behaviour of the Commandant of Belsen and the
behaviour, say, of Sir Stafford Cripps or the Archbishop of Canterbury.

[112] R: The feeling is a little too simplified. You’ve got to take account of the
effects of actions and your feelings toward those effects. You see, you can have
an argument about it if you can say that certain sorts of occurrences are the
sort you like and certain others the sort you don’t like. Then you have to take
account of the effects of actions. You can very well say that the effects of the
actions of the Commandant of Belsen were painful and unpleasant.
In response to Copleston’s questions, Russell polishes up his warrant of moral judg-
ments a little more, which would now look like this: an action is good (bad) if and
only if most moral agents feel that this action or its effects are good (bad).
[113] C: They certainly were, I agree, very painful and unpleasant to all the
people in the camp.

[114] R: Yes, but not only to the people in the camp, but to outsiders
contemplating them also.

[115] C: Yes, quite true in imagination. But that’s my point. I don’t approve
of them, and I know you don’t approve of them, but I don’t see what ground
you have for not approving of them, because, after all, to the Commandant of
Belsen himself they’re pleasant, those actions.
Copleston objects that Russell’s moral rule does not allow him to answer the question:
“Why do you think the behavior of the Commandant of Belsen was bad? That is, in
the form of an argument:

The rule x is good/bad iff the majority feels that x,

or its effects, are good/bad makes it impossible to
answer questions like “What ground do you have
for approving/condemning x?”
A suitable moral rule must answer such
questions as “What ground do you have Therefore
for approving/condemning x?”:
The rule x is good/bad iff the majority feels that x,
or its effects, are good/bad must be rejected
14.3 The Moral Argument 417

[116] R: Yes, but you see I don’t need any more ground in that case than I do
in the case of colour perception. There are some people who think everything is
yellow, there are people suffering from jaundice, and I don’t agree with these
people. I can’t prove that the things are not yellow, there isn’t any proof. But
most people agree with me that they’re not yellow, and most people agree with
me that the Commandant of Belsen was making mistakes.
Russell appeals to an analogy with color judgments to deny that a moral rule should
allow such questions to be answered - that is, he rebuts Copleston’s argument in 115.

In the case of The rule x is good/bad iff

color judgments, I
don’t need any
more ground than
collective feeling
Therefore
In the case of moral
judgments, I don’t
need any more against “What ground do you have for
ground than
collective feeling
the majority feels that x, or
its effects, are good/bad
makes it impossible to
answer questions like “What
ground do you have for
approving/condemning x?”
A suitable moral rule must
answer such questions as
Therefore
approving/condemning x?”:
The rule x is good/bad iff
the majority feels that x, or
its effects, are good/bad
must be rejected

[117] C: Well, do you accept any moral obligation?

[118] R: Well, I should have to answer at considerable length to answer that.

Practically speaking—yes. Theoretically speaking, I should have to define moral
obligation rather carefully.

[119] C: Well, do you think that the word “ought” simply has an emotional
connotation?

[120] R: No, I don’t think that, because you see, as I was saying a moment ago,
one has to take account of the effects, and I think right conduct is that which
would probably produce the greatest possible balance in intrinsic value of all
the acts possible in the circumstances, and you’ve got to take account of the
probable effects of your action in considering what is right.
Copleston insists on his argument against Russell’s emotivism, moving from the
ground of moral judgments to that of moral obligations. To show that his theory
leaves room for moral obligation, Russell proposes the following definition of “right
conduct”: right conduct is that which would probably produce the greatest possible
418 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

balance in intrinsic value of all the acts possible in the circumstances, taking an
utilitarian turn.

Right conduct is that which would probably produce the greatest possible
balance in intrinsic value of all the acts possible in the circumstances
Therefore
The rule x is good/bad iff the majority feels that x, or its effects, are
good/bad can account for moral obligation

[121] C: Well, I brought in moral obligation because I think that one can
approach the question of God’s existence in that way. The vast majority of
the human race will make, and always have made, some distinction between
right and wrong. The vast majority, I think, has some consciousness of an obli-
gation in the moral sphere. It’s my opinion that the perception of values and
the consciousness of moral law and obligation are best explained through the
hypothesis of a transcendent ground of value and of an author of the moral law.
I do mean by “author of the moral law” an arbitrary author of the moral law.
I think, in fact, that those modern atheists who have argued in a converse way
There is no God; therefore, there are no absolute values and no absolute law are
quite logical.
Copleston frames the moral argument as an abductive argument, as he had already
done with the argument based on religious experience. To expose the argument he
follows the pattern Thesis. Ground. Warrant.

The vast majority of the human race has

some consciousness of an obligation in
the moral sphere
The perception of values and the consciousness of
moral law and obligation are best explained
Therefore
through the hypothesis of a transcendent ground of
value and of an author of the moral law:
There is a transcendent ground of value
and an author of the moral law

The premise is stated in two slightly different ways, giving rise to two different
versions of the argument: the one in the above diagram and.
14.3 The Moral Argument 419

The vast majority of the human race will

make, and always have made, some
distinction between right and wrong
The perception of values and the consciousness of
moral law and obligation are best explained
Therefore
through the hypothesis of a transcendent ground of
value and of an author of the moral law:
There is a transcendent ground of value
and an author of the moral law

[122] R: I don’t like the word “absolute”. I don’t think there is anything absolute
whatever. The moral law, for example, is always changing. At one period in the
development of the human race, almost everybody thought cannibalism was a
duty.
In fact Copleston does not use in 121 the word “absolute” that Russell dislikes. Russell
objects, appealing to the example of cannibalism, that the moral law is continually
changing. To be relevant for the preceding argument, this is to be taken as an alleged
reason against the claim that the vast majority human race has consciousness of
absolute moral obligations.

At one period in the development of the

human race, almost everybody thought
cannibalism was a duty. Today
cannibalism is morally rejected
Therefore
The moral law, for example, is always
changing
Therefore
The vast majority of the human race against The vast majority of the human race
does not have some consciousness of an has some consciousness of an absolute
absolute obligation in the moral sphere obligation in the moral sphere
Therefore
There is a transcendent ground of
value and an author of the moral law

[123] C: Well, I don’t see that differences in particular moral judgments are any
conclusive argument6 against the universality of the moral law. Let’s assume
for the moment that there are absolute moral values. Even on that hypothesis
it’s only to be expected that different individuals and different groups should
enjoy varying degrees of insight into those values.

6 “Argument” is used here in the sense of “a reason urged in support of a proposition”, and not in
the common logical sense of “a connected series of statements or reasons intended to establish a
position” (OED).
420 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

Copleston contends that changes in particular moral judgments are not a conclusive
reason against the moral law. This seems to be a plain rebuttal of the last step of the
subordinative argument in 122.

The moral law, for example, is

Even if there where absolute values, it’s
only to be expected that different
individuals and different groups should
enjoy varying degrees of insight into
those values
always changing
against Therefore
The vast majority of the human
race does not have some
consciousness of an absolute
obligation in the moral sphere

[124] R: I’m inclined to think that “ought”—the feeling that one has about
“ought”—is an echo of what has been told one by one’s parents or one’s nurses.
Russell is now trying to rebut the abductive argument in 121, arguing that the fact
that “The vast majority of the human race will make, and always have made, some
distinction between right and wrong” is (better) explained by the circumstance that
“ought is an echo of what has been told one by one’s parents or one’s nurses”.

I’m inclined to think

that “ought”—the
feeling that one has
about “ought”—is an
echo of what has been
told one by one’s
parents or one’s nurses
Therefore
The hypothesis of a
transcendent ground of
value and of an author
of the moral law is not
necessary to account against
for the perception of
values and the
consciousness of moral
law and obligation
The vast majority of the
human race will make,
and always have made,
some distinction between
right and wrong
The perception of values and
the consciousness of moral
law and obligation are best
explained through the
hypothesis of a transcendent Therefore
ground of value and of an
author of the moral law:
There is a transcendent
ground of value and an
author of the moral law
14.3 The Moral Argument 421

[125] C: Well, I wonder if you can explain away the idea of the “ought” merely in
terms of nurses and parents. I really don’t see how it can be conveyed to anybody
in other terms than itself. It seems to be that, if there is a moral order bearing
upon the human conscience, that that moral order is unintelligible apart from
the existence of God.
Copleston rejects the alternative explanation proposed by Russell and insists that the
moral conscience is unintelligible apart from the existence of God.

I really don’t see how the idea of the “ought” can be conveyed to
anybody in other terms than itself
Therefore
The idea of the “ought” cannot be explained away merely in
terms of nurses and parents

[126] R: Then you have to say one or other of two things. Either God only speaks
to a very small percentage of mankind—which happens to include yourself—or
He deliberately says things that are not true in talking to the consciences of
savages.
This is meant to be a reduction a reductio ad absurdum of the warrant of Copleston’s
abductive argument in 121.

The conscience of moral law

and obligation of savages is
quite different from ours
That God actually dictates
moral precepts to the
conscience of humans Therefore
explains their consciousness
of moral law and obligation:
Either God only speaks to a It is absurd to suggest that
very small percentage of either God only speaks to a
mankind—which happens to very small percentage of
include yourself—or He mankind—which happens to
deliberately says things are include yourself—or He
not true in talking to the deliberately says things are
consciences of savages not true in talking to the
consciences of savages
Therefore
The perception of values and the consciousness of moral law and obligation are not well-
explained the hypothesis of a transcendent ground of value and of an author of the moral law
422 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

Reductio ad absurdum is a well-known pattern of structural meta-argumentation, in

which a conclusion is reached on the basis that it is possible to infer an untenable
conclusion from a certain assumption. The premises of a reductio are an argument
(usually a suppositional argument7 ) and a set of assertions. I shall refer to the argu-
ment which provides one of the premises as “the embedded argument”. In the above
reductio the embedded argument is.

The conscience of moral law and obligation of

savages is quite different from ours
That God actually dictates moral
precepts to the conscience explains
Therefore
the consciousness of moral law and
obligation:
Either God only speaks to a very small percentage
of mankind—which happens to include yourself—
or He deliberately says things are not true in talking
to the consciences of savages

There are two main forms of reductio. If the allegedly untenable conclusion is inferred
using an accepted inference rule, it is concluded that some supposition that appears as
a premise in the embedded argument should be rejected. Suppositions are unasserted
premises typically introduced by phrases such as “let us suppose”, “let us say”, “sup-
pose”, “imagine”, “let us suppose for the sake of argument”, etc. If, on the contrary,
premises are uncontroversial, it is concluded that the rule that allows an untenable
consequence to be derived from them is not valid, as in the above argument. We there-
fore distinguish the reductio ad absurdum of a supposition (which can be either an
objection or a refutation) and the reductio ad absurdum of a warrant or inference rule
(which is a kind of rebuttal). Russell’s argument at 126 is a reductio of the latter type.
[127] C: Well, you see, I’m not suggesting that God actually dictates moral
precepts to the conscience. The human being’s ideas of the content of the moral
law depends entirely to a large extent on education and environment, and a
man has to use his reason in assessing the validity of the actual moral ideas
of his social group. But the possibility of criticizing the accepted moral code
presupposes that there is an objective standard, that there is an ideal moral
order, which imposes itself, I mean the obligatory character of which can be
recognized. I think that the recognition of this ideal moral order is part of the
recognition of contingency. It implies the existence of a real foundation —of
God.
Copleston rejects the warrant of the embedded argument in the former reductio.
Russell attributes to Copleston the idea that humans are aware of moral law and
obligation because God actually dictates moral precepts to their conscience. Since

7 Roughly speaking, suppositional arguments are arguments beginning with the phrase ‘Suppose
that X were the case.’ (Fisher 2004, Chap. 8).
14.3 The Moral Argument 423

this is a reason to believe that the consciousness of moral law and obligation are
explained by the hypothesis of an author of the moral law, this could the backing of
the warrant of Copleston’s abductive argument in 121. Copleston argues first against
this assertion:

The human being's ideas of the content of the moral law depends entirely to
a large extent on education and environment, and a man has to use his
reason in assessing the validity of the actual moral ideas of his social group
Therefore
God does not dictate moral precepts to the conscience

And then goes on to rehearse the abductive moral argument:

Humans use their reason in assessing the validity

of the actual moral ideas of their social group
The possibility of criticizing the
accepted moral code presupposes that
there is an objective standard, that there
Therefore
is an ideal moral order, which imposes
itself, I mean the obligatory character of
which can be recognized:
There is an objective standard, there is an ideal
moral order, which imposes itself, I mean the
obligatory character of which can be recognized
(i.e. the vast majority of the human race has some
consciousness of an absolute obligation in the
moral sphere)
The recognition of an ideal moral order
which imposes itself is best explained
through the hypothesis of a transcendent Therefore
ground of value and of an author of the
moral law:
There is a real foundation of value and the moral
law ―God

In the first part of this subordinative argument Copleston supports the premise of his
abductive argument in 121 that Russell had attacked in 123. The remaining part can
be interpreted as a development of his abductive argument, an attempt to specify how
the hypothesis of a transcendent ground of value and of an author of the moral law
is the best explanation of the recognition of the ideal moral order. On this reading
“I think that the recognition of this ideal moral order is part of the recognition of
contingency” would be an incidental comment (as suggested by “I think”) while the
424 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

pronoun “it” would refer to the recognition of this ideal moral order. An alternative
reading takes “it” to refer to the recognition of contingency and the quoted sentence
as a warrant, inserting an intermediate step at the end of the argument:

There is an objective standard, there is an ideal moral

order, which imposes itself, I mean the obligatory
character of which can be recognized (i.e. the vast
majority of the human race has some consciousness
of an absolute obligation in the moral sphere)
I think that the recognition of this
ideal moral order is part of the Therefore
recognition of contingency:
The vast majority of the human race has some
consciousness of its contingency
Consciousness of contingency implies
Therefore
the existence of a real foundation:
There is a real foundation of value and the moral law
―God

However, it seems strange to use a connector of opinion like “I think” to propose a

warrant.
[128] R: But the lawgiver has always been, it seems to me, one’s parents or
someone like < them > . There are plenty of terrestrial lawgivers to account for
it, and that would explain why people’s consciences are so amazingly different
in different times and places.
Russell is adamant that no explanation of humanity’s moral conscience is needed
beyond the plethora of human legislators.
[129] C: It helps to explain differences in the perception of particular moral
values, which otherwise are inexplicable. It will help to explain changes in the
matter of the moral law, in the content of the precepts as accepted by this or
that nation, or this or that individual. But the form of it, what Kant calls the
categorical imperative, the “ought”, I really don’t see how that can possibly
be conveyed to anybody by nurse or parent because there aren’t any possible
terms, so far as I can see, with which it can be explained. It can’t be defined in
other terms than itself, because, once you’ve defined it in other terms than itself,
you’ve explained it away. It’s no longer a moral “ought”. It’s something else.
14.3 The Moral Argument 425

Copleston concedes that the multitude of human legislators helps to explain different
perceptions of moral values and changes in accepted precepts, but he insists that it
cannot explain the categorical imperative, the “ought”, against Russell’s claim in 124.
His argument begins with “but”, stating the thesis and using “because” to indicate
the successive reasons in his argumentation.

Once you’ve defined the “ought” in other terms than itself, you’ve
explained it away. It’s no longer a moral “ought”. It’s something else
Therefore
The “ought” can’t be defined in other terms than itself
Therefore
The form of the moral law, what Kant calls the categorical
imperative, the “ought”, I really don’t see how that can possibly be
conveyed to anybody by nurse or parent

[130] R: Well, I think the sense of “ought” is the effect of somebody’s imag-
ined disapproval, it may be God’s imagined disapproval, but it’s somebody’s
imagined disapproval. And I think that is what is meant by “ought”.
This observation does not respond to Copleston’s immediately preceding interven-
tion, and can be understood as a reiteration of Russell’s rebuttal in 124 of the argument
from the universal consciousness of moral law and obligation to the existence of a
transcendent ground of value and an author of the moral law.
[131] C: It seems to me to be external customs and taboos and things of that sort
which can most easily be explained simply through environment and education;
but all that seems to me to belong to what I call the matter of the law, the content.
The idea of the “ought”, as such, can never be conveyed to a man by the tribal
chief or by anybody else, because there are no other terms in which it could be
conveyed. It seems to me entirely…
Copleston does two things in his reply. First, as indicated by the connector “but”, he
rejects as irrelevant Russell’s purported rebuttal in 124. That the feeling that one has
about “ought” is an echo of what has been told one by one’s parents or one’s nurses
can only explain external customs and taboos, but not the matter or content of the
law.
426 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

I’m inclined to think

that “ought”—the
feeling that one has
about “ought”—is an
echo of what has been
told one by one’s
parents or one’s nurses
That the feeling
that one has
about “ought” is
an echo of what The vast majority
has been told of the human race
one by one’s will make, and
against

parents or one’s always have

Therefore
nurses can only made, some
explain external distinction
customs and between right and
taboos, but not wrong
the matter or
content of the
law
The perception of
The hypothesis of a values and the
transcendent ground of consciousness of
value and of an author moral law and
of the moral law is not obligation are best
against

necessary to account for explained through Therefore

the perception of values the hypothesis of a
and the consciousness transcendent ground
of moral law and of value and of an
obligation author of the moral
law:
There is a
transcendent
ground of value
and an author of
the moral law

Second, as indicated by the connector “because”, Copleston reiterates his argument

in 125, in slightly different terms:

There are no other terms in which the idea of the

“ought”, as such, could be conveyed
Therefore
The idea of the “ought”, as such, can never be conveyed
to a man by the tribal chief or by anybody else

[132] R: But I don’t see any reason to say that—I mean we all know about condi-
tioned reflexes. We know that an animal, if punished habitually for a certain
sort of act, after a time will refrain. I don’t think the animal refrains < by >
arguing within himself Master will be angry if I do this. He has a feeling that
14.3 The Moral Argument 427

that’s not the thing to do. That’s what we can do with ourselves and nothing
more.
The pronoun “that” can refer either to the premise of the foregoing argument or to
its conclusion. If it were the latter, since Copleston has just given a reason to believe
it, Russell would have to offer a rebuttal of that argument. Therefore, I take “that” to
refer to the premise “There are no other terms in which the idea of the “ought”, as
such, could be conveyed”, and Russell’s response as an objection.
Russell tries to show that the categorical imperative and duty can be explained in
purely social terms, using an analogy:

We know that an animal, if punished habitually for a certain sort of act, after a time
will refrain. I don’t think the animal refrains by arguing within himself, “Master
will be angry if I do this.” He has a feeling that that’s not the thing to do
Likewise
That’s what we can do with ourselves and nothing more.
Therefore
The idea of the “ought” could be explained in terms of conditioned reflexes (i.e.
there are other terms in which the idea of the “ought”, as such, could be conveyed).

[133] C: I see no reason to suppose that an animal has a consciousness or moral

obligation; and we certainly don’t regard an animal as morally responsible for
his acts of disobedience. But a man has a consciousness of obligation and of
moral values. I see no reason to suppose that one could condition all men as one
can condition an animal, and I don’t suppose you’d really want to do so even
if one could. If “behaviourism” were true, there would be no objective moral
distinction between the emperor Nero and St. Francis of Assisi. I can’t help
feeling, Lord Russell, you know, that you regard the conduct of the Commandant
of Belsen as morally reprehensible, and that you yourself would never under
any circumstances act in that way, even if you thought, or had reason to think,
that possibly the balance of the happiness of the human race might be increased
through some people being treated in that abominable manner.
Copleston rejects the analogy in 132. What needs to be explained is the moral
conscience of human beings; therefore, Russell’s analogy, to be relevant, would
have to be rather like.

An animal consciousness or moral obligation is explained by the fact that he

develops a feeling that an act is not the thing to do if punished habitually for it
Likewise
Human consciousness of obligation and of moral values may be explained along
the same lines
428 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

Copleston first argues that animals and humans differ significantly since the former,
unlike the latter, have no moral conscience and are not responsible for their actions.
This is tantamount to a rejection of the warrant of Russell’s argument in 132.

I see no reason to suppose that an

animal has a consciousness or
moral obligation; and we certainly
don’t regard an animal as morally
responsible for his acts of
disobedience. But a man has a
consciousness of obligation and of
moral values:
We know that an animal, if
punished habitually for a certain
sort of act, after a time will refrain.
I don’t think the animal refrains
Therefore
from arguing within himself,
“Master will be angry if I do this.”
He has a feeling that that’s not the
thing to do

Human behavior can

I see no reason to suppose that one
against

be explained in terms
could condition all men as one can Likewise
of conditioning as is
condition an animal
animal behavior:

That’s what we can do with

ourselves and nothing more.

Thus Copleston argues first that the behaviorist principle in Russell’s analogy is
groundless. Taking a further step, Copleston then argues that it is false, that moral
conscience cannot be explained in behavioral terms.

You regard the conduct of the Commandant of Belsen

as morally reprehensible, and you yourself would never
under any circumstances act in that way, even if you
thought, or had reason to think, that possibly the
balance of the happiness of the human race might be
increased through some people being treated in that
abominable manner
Therefore
There is an objective moral distinction between the
emperor Nero and St. Francis of Assisi.
If “behaviorism” were true, there
would be no objective moral
Therefore
distinction between the emperor
Nero and St. Francis of Assisi:
Behaviorism is false (i.e. the thesis that human and
animal behavior can be explained in terms of
conditioning).
14.3 The Moral Argument 429

[134] R: No. I wouldn’t imitate the conduct of a mad dog. The fact that I wouldn’t
do it doesn’t really bear on this question we’re discussing.
Russell concedes the basic premise of Copleston’s second argument, but considers
it irrelevant to the question under discussion; he therefore denies that there are any
objective moral distinctions, in Copleston’s sense.
[135] C: No, but if you were making a utilitarian explanation of right and wrong
in terms of consequences, it might be held, and I suppose some of the Nazis of
the better type would have held, that although it’s lamentable to have to act in
this way, yet the balance in the long run leads to greater happiness. I don’t think
you’d say that, would you? I think you’d say that sort of action is wrong—and
in itself, quite apart from whether the general balance of happiness is increased
or not. Then, if you’re prepared to say that, then I think you must have some
criterion of right and wrong, that is outside of feeling, at any rate. To me, that
admission would ultimately result in the admission of an ultimate ground of
value in God.
Copleston first attempts to refute Russell, arguing that the fact that Russell would
condemn Nazism even if its policies could be shown to increase happiness in the
long run is an indication that Russell has some objective criteria for discerning right
and wrong, and then he suggests that the denial of Russell’s emotivist thesis leads to
admit the existence of God.

It might be held that although it’s lamentable to have to act as the Nazis did, yet the
balance in the long run leads to greater happiness. You’d say that sort of action is
wrong—and in itself, quite apart from whether the general balance of happiness is
increased or not
Therefore (then)
You must have some criterion of right and wrong, that is outside of feeling, at any rate
Therefore
You must admit an ultimate ground of value in God

[136] R: I think we are perhaps getting into confusion. It is not direct feeling
about the act by which I should judge, but rather a feeling as to the effects.
And I can’t admit any circumstances in which certain kinds of behaviour, such
as you have been discussing, would do good. I can’t imagine circumstances in
which they would have a beneficial effect. I think the persons who think they do
are deceiving themselves. But if there were circumstances in which they would
have a beneficial effect, then I might be obliged, however reluctantly, to say Well,
I don’t like these things, but I will acquiesce in them, just as I acquiesce in the
Criminal Law, although I profoundly dislike punishment.
Russell rejects the premise of Copleston’s argument: he would not condemn Nazism
if there were circumstances in which its policies had a beneficial effect. He also
430 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

points out that his judgment would not be guided by his feelings about Nazism, but
by his feelings about its effects.

14.4 Concluding Stage

[137] C: Well, perhaps it’s time I summed up my position. I’ve argued two things.
First, that the existence of God can be philosophically proved by a metaphysical
argument. Secondly, that it is only the existence of God that will make sense of
man’s moral experience and of religious experience. Personally, I think that your
way of accounting for man’s moral judgments leads inevitably to a contradic-
tion between what your theory demands and your own spontaneous judgments.
Moreover, your theory explains moral obligation away, and explaining away is
not explanation.
As regards the metaphysical argument, we are apparently in agreement that what
we call the world consists simply of contingent beings. That is, of beings no one of
which can account for its own existence. You say that the series of events needs no
explanation: I say that, if there were no necessary being—no being which must exist
and cannot not-exist—, nothing would exist. The infinity of the series of contingent
beings, even if proved, would be irrelevant. Something does exist; therefore, there
must be something which accounts for this fact, a being which is outside the series
of contingent beings. If you had admitted this, we could then have discussed whether
that being is personal, good, and so on. On the actual point discussed, whether there
is or is not a necessary being, I find myself, I think, in agreement with the great
majority of classical philosophers.
Copleston had presented both the argument from religious experience (71) and the
argument from moral experience (121) as abductive arguments. Now he argues that
Russell has been unable to provide a better explanation of moral experience, since (1)
Russell’s account of moral judgments is in contradiction with his own spontaneous
judgments, and (2) his theory explains moral obligation away, and explaining away
is not explanation.
Copleston then presents a condensed version of the metaphysical argument of 11:
14.4 Concluding Stage 431

(1) What we call the world consists simply

of contingent beings; that is, of beings no
one of which can account for its own
existence
(2) If something does exist, then there
must be something which accounts for
Therefore
this fact, a being which is outside the
series of contingent beings:
(3) There is a being which is outside
the series of contingent beings, a
necessary being

Copleston begins by pointing out that Russell concedes the premise (1) while disputes
the warrant (2). He then argues that if they had agreed on the conclusion (3), they
could have discussed whether that necessary being is personal, good, and so on.
Finally, Copleston suggests something akin to an argument from authority:

On the actual point discussed, whether there is or is not a necessary being, I find
myself, I think, in agreement with the great majority of classical philosophers.
Therefore
There is a necessary being

[137] (cont.) You maintain, I think, that existing beings are simply there, and that
I have no justification for raising the question of the explanation of their existence.
But I would like to point out that this position cannot be substantiated by logical
analysis; it expresses a philosophy which itself stands in need of proof. I think we
have reached an impasse because our ideas of philosophy are radically different; it
seems to me that, what I call a part of philosophy, that you call the whole, insofar at
least as philosophy is rational.
It seems to me, if you will pardon my saying so, that, besides your own logical
system, what you call “modern” in opposition to antiquated logic (a tendentious
adjective), you maintain a philosophy which cannot be substantiated by logical anal-
ysis. After all, the problem of God’s existence is an existential problem, whereas
logical analysis does not deal directly with problems of existence. So, it seems to
me, to declare that the terms involved in one set of problems are meaningless because
they are not required in dealing with another set of problems, is to settle from the
beginning the nature and extent of philosophy, and that is itself a philosophical act
which stands in need of justification.
Copleston focuses on the reasons allegedly put forward by Russell for rejecting the
inference of (3) from (1) through (2). These reasons would be “logical” and would
lead to the dismissal of the metaphysical argument as a whole, in the technical sense
of Turn 16.
432 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

The word “necessary” and “contingent” are meaningless, except as

applied to analytic propositions, not to things. The metaphysical
argument asserts that there are contingent and necessary beings
Therefore
The metaphysical argument is meaningless

Copleston dismisses Russell’s dismissal using the following meta-argument:

The metaphysical argument deals with the

problem of God’s existence. God’s existence
is an existential problem
Logical analysis does not deal
Therefore
directly with problems of existence:
The metaphysical argument cannot be
dismissed on purely logical grounds

It is interesting to note how the disagreement between Russell and Copleston that in
67 had prevented the discussion from continuing produces an argumentative ascent,
from argumentation to meta-argumentation.
[138] R: Well, I should like to say just a few words by way of summary on my
side. First, as to the metaphysical argument: I don’t admit the connotations
of such a term as “contingent” or the possibility of explanation in your sense.
I think the word “contingent” inevitably suggests the possibility of something
that wouldn’t have this, what you might call, accidental character of just being
there, and I don’t think < that > is true except in the purely causal sense. You can
sometimes give a causal explanation of one thing as being the effect of something
else, but that is merely referring one thing to another thing and there’s no—to
my mind—explanation in your sense of anything at all; nor is there any meaning
in calling things “contingent” because there isn’t anything else they could be.
That’s what I should say about that, but I should like to say a few words about your
accusation that I regard logic as all philosophy. That is by no means the case. I don’t
by any means regard logic as all philosophy. I think logic is an essential part of
philosophy, and logic has to be used in philosophy, and in that I think you and I are at
one. When the logic that you use was new, namely in the time of Aristotle, there had
to be a great deal of fuss made about it. Aristotle made a lot of fuss about that logic.
Nowadays it’s become old and respectable, and you don’t have to make so much
fuss about it. The logic that I believe in is comparatively new, and therefore I have to
imitate Aristotle in making a fuss about it, but it’s not that I think it’s all philosophy
by any means. I don’t think so. I think it’s an important part of philosophy and, when
I say that I don’t find a meaning for this or that word, that is a position of detail,
based upon what I’ve found out about that particular word, from thinking about it.
14.4 Concluding Stage 433

It’s not a general position that all words that are used in metaphysics are nonsense,
or anything like that, which I don’t really hold.
Russell confirms that, indeed, he dismisses Copleston’s metaphysical argument by
his use of terms such as “contingent” or “necessary” and tries to explain his position.
Russell then responds to Copleston’s meta-argument. For his reply to be relevant,
the warrant must be understood as follows:

The metaphysical argument deals with the problem of

God’s existence. God’s existence is an existential problem
Metaphysics, not Logic, deals
Therefore
with problems of existence:
The metaphysical argument cannot be dismissed on purely
logical grounds

Russell argues that his argument does not presuppose that metaphysics, or its vocab-
ulary, is meaningless and that all philosophical questions should be treated as logical
issues. He therefore acknowledges the general validity of this warrant (“It’s not a
general position”); what he denies is that it can be applied to this particular case
(“based upon what I’ve found out about that particular Word”).
[138] (cont.) As regards the moral argument, I do find that, when one studies anthro-
pology or history, there are people who think it their duty to perform acts which
I think abominable, and I certainly can’t, therefore, attribute Divine origin to the
matter of moral obligation, which you, Father Copleston don’t ask me to; but I think
even the form of moral obligation, when it takes the form of enjoining you to eat your
father or what not, doesn’t seem to me to be such a very beautiful and noble thing;
and, therefore, I cannot attribute a Divine origin to this sense of moral obligation,
which I think is quite easily accounted for in quite other ways.
Russell, unlike Copleston, concludes by referring to the moral argument in 121:

The vast majority of the human race has

some consciousness of an absolute
obligation in the moral sphere
The perception of absolute values and the
consciousness of absolute moral law and
obligation are best explained through the Therefore
hypothesis of a transcendent ground of value and
of an author of the moral law:
There is a transcendent ground of value
and of an author of the moral law

To this end, Russell offers a double refutation of the conclusion of this argument:
434 14 An Argument Dialectical Analysis of the Russell-Copleston Debate

There are people who think it their
duty to perform acts which I think
even
abominable
Therefore
I certainly can’t attribute Divine origin
to the matter of moral obligation
When the form of moral obligation takes
the form of enjoining you to eat your father
or what not, doesn’t seem to me to be such
a very beautiful and noble thing
Therefore
I cannot attribute a Divine origin to this
sense of moral obligation

And he supplements this double refutation with a rebuttal of the warrant of the
abductive moral argument:

The sense of moral obligation is quite easily accounted for in quite other ways
Therefore
There is no need to attribute a Divine origin to this sense of moral obligation

A curious feature of this counter-argumentation is that Russell presents first the

refutation and then the rebuttal, when normally we proceed in reverse order to avoid
the complication of weighing different reasons. That is, usually we first get rid of the
opponent’s argument, overturning it with a rebuttal (or an objection, as did Copleston
in Turn 21), and then we give reasons for the opposite conclusion. Russell’s approach
leads me to believe that here, despite the above analysis, the statement “I think moral
obligation is quite easily accounted for in quite other ways” is not intended to argue,
but simply to express an opinion.

References

Fogelin, Robert (1985). “The logic of deep disagreements”. Informal Logic 7(1), pp. 3-11.
Freeman, James B. (2011). Argument Structure: Representation and Theory. Dordrecht: Springer.
Laar, Jan-Albert van (2015). “Connection premises: Their character, criticism, and defence”. In
Cornelia Ilie & Giuliana Garzone, eds., Argumentation in real and virtual environments: Cross-
disciplinary perspectives, pp. 39–55. Amsterdam and Filadelfia: John Benjamins.
Lakoff, Robin (1973). “Questionable answers and answerable questions”. In: Braj B. Kahru, Rober t
B.Lees, Yacov Malkiel, Angelina Pietrangeli and Sol Saporta (eds.). Issues in Linguistics. Papers
in Honor of Henry and Renée Kahane, pp. 453–467. Urbana: University of Illinois Press.
Mann, William E. (1976). “The Perfect Island”. Mind new Series, Vol. 85, No. 339, pp. 417-421.
Olmos Gómez, Paula (2019). “Abduction and comparative weighing of explanatory hypotheses.
An argumentative approach”. Logic Journal of the IGPL, jzz038, https://doi.org/10.1093/jigpal/
jzz038
Olmos Gómez, Paula (2021). “Metaphilosophy and Argument: The Case of the Justification of
Abduction”. Informal Logic vol. 41, n.2, pp. 131–164. https://informallogic.ca/index.php/inf
ormal_logic/article/view/6249
Perelman, Chaïm (1979). The new rhetoric and the humanities: Essays on rhetoric and its
applications. Dordrecht: D. Reidel.
References 435

Perelman, Chaïm & Olbrechts-Tyteca, Lucie (1969 [1958]). The new rhetoric: A treatise on
argumentation. Notre Dame: University of Notre Dame Press.
Svartvik, Jan (1980). “Well in conversation”. In: Sidney Greenbaum, Geoffrey N. Leech, Jan
Svartvik (eds.). Studies in English linguistics for Randolph Quirk, pp. 167–177.
Toulmin, Stephen E. (2003 [1958]). The Uses of Argument, Cambridge: Cambridge University
Press.
Vleet, Jacob E. van (2011). Informal Logical Fallacies. A Brief Guide. Lanham, Maryland:
University Press of America.
Walton, Douglas N., Reed, Chris, & Macagno, Fabrizio (2008). Argumentation Schemes. New York:
Cambridge University Press.
Chapter 15
Final Cross-Examination

Abstract The present book closes with an exercise of cross-examination in which

each one of the authors comments on the kind of analysis done by the other author.
This is at the same time an occasion for the reader to compare one theory of the
argumentative process (the pragma-dialectics of the Amsterdam School) with one
theory of argument exchange (the ‘argument dialectics’ of the Madrid School) as
applied to a philosophical debate. One question arising from this is what kind of
analysis would emerge from the other kinds of theories as outlined in Chap. 2 of the
book.

In this final chapter, each of us will try to summarize his view of what the experience
of analyzing the Russell-Copleston debate has been and how he sees the relationship
between the two theoretical and methodological frameworks now that the work is
done. Although this chapter may have shortcomings, we believe it is an exercise in
the debating procedure devised by Daniel Kahneman as adversarial collaboration. To
avoid repetition, we invite the reader to peruse again the description of the method
that we give in Chap. 1, Sect. 1.3.

15.1 Comments on Part I (Hubert Marraud)

First of all, Fernando’s erotetic pragma-dialectical analysis and my argument dialec-

tical analysis differ in aim and scope. The focus of the erotetic pragma-dialectical
analysis is the Copleston-Russell debate as an exercise in argumentation. Argumen-
tation, according to the well-known pragma-dialectical definition, “is a verbal, social,
and rational activity aimed at convincing a reasonable critic of the acceptability of
a standpoint by putting forward a constellation of propositions justifying or refuting
the proposition expressed in the standpoint” (van Eemeren and Grootendorst 2004,
p. 1). Therefore, the subject of an erotetic pragma-dialectical analysis is the behavior
of the participants, and, since pragma-dialectics is a specific version of normative
pragmatics, its aim is to determine whether and to what extent participants behave
reasonably. On the other hand, the gist of argument dialectics lies in the opposition

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 437
F. Leal and H. Marraud, How Philosophers Argue, Argumentation Library 41,
https://doi.org/10.1007/978-3-030-85368-6_15
438 15 Final Cross-Examination

relationships between arguments that allow participants to interactively construct a

macro-argument. In other words, although both are dialectical, they are not dialec-
tical in the same sense. In the terminology of Chap. 9, pragma-dialectics belongs to
arguer dialectics, not to argument dialectics.
Erotetic pragma-dialectics deals with dialogues:
A dialogue (to use our generic term) or a conversation (to use an everyday term) is defined
as a goal-directed conventional framework in which two speech partners reason together in
an orderly way, according to the rules of politeness or normal expectations of cooperative
argumentation for the type of exchange they are engaged in. (Walton 1998, p. 3.)

It is true that the concept of dialogue is not a pragma-dialectical concept. But it

is sufficiently similar to the pragma-dialectical concepts of communicative activity
type and genre for the previous observations to be valid for Fernando’s erotetic
pragma-dialectical approach:
Communicative activity types are conventionalized practices whose conventionalization
serves, through the implementation of certain “genres” of communicative activity the institu-
tional needs prevailing in a certain domain of communicative activity. The genres of commu-
nicative activity that may be implemented (alone or combined) in the various communicative
activity types vary from adjudication and deliberation to mediation, negotiation, consultation,
disputation, promotion, communion, and others. The institutional mission that a commu-
nicative activity type is meant to fulfill in a certain communicative domain is accomplished
by realizing the relevant institutional point through the use of the appropriate genre(s) of
communicative activity. (van Eemeren 2010, p. 139.)

A dialogue can thus be conceived of as an exchange of reasons. Since, according

to a well-known definition, a reason is an answer to a question (Crosswhite 1996,
p.78) or an answer to a ‘why-question’ (Skow 2016; Engel 2017), a dialogue can
also be conceived of as a succession of questions and answers (and, since roles
can alternate, of counter-questions), as it in fact is in the erotetic pragma-dialectical
approach. (Fernando identifies 98 questions in the debate between Copleston and
Russell). This determines the segmentation that Fernando makes of the turns, almost
sentence by sentence.
According to the model of critical discussion, four stages can be distinguished
in an argumentative dialogue: the confrontation stage, the opening stage, the argu-
mentation stage, and the concluding stage. Since Fernando is interested in the whole
dialogue, is analysis encompasses these four stages. As for me, I focus on the state
of argumentation, in which participants jointly construct a macro-argument. This
requires paying attention to connectors and other devices used to bind together the
successive arguments. As a consequence, the unit of analysis cannot be the sentence,
but the paragraph. This is a substantial difference between our analyses.
Partly because of his understanding of debate as a succession of questions and
answers, partly because of his choice of the premise-conclusion model, Fernando
pays little attention to argumentative operations and structures (except for chaining
and subordinate argumentation). Participants propose reasons, offering generally
simple arguments, which other participants accept or reject. In turn, addressees may
produce arguments to justify their acceptance or rejection of the proposed arguments.
In this way, what binds together the participants’ arguments is what they do with
15.1 Comments on Part I (Hubert Marraud) 439

them. That explains the large number of “dialectical meta-arguments” in Fernando’s

analysis (I have found 55 of such arguments). They are meta-arguments in that their
components (premises or conclusion) include an assertion or an assumption about
other arguments. They are dialectical because they talk about what the participants
do or should do about the arguments. Here is a sample of these dialectical meta-
arguments.

1 Premise 1. Not all modern logicians admit the meaninglessness of metaphysics

Premise 2. Russell’s objection implies the meaninglessness of metaphysics
Conclusion. Russell’s objection is not as strong as it may appear, given Russell’s authority as
a logician. [19.3]
2 Premise 1. To say that a necessary being is a being that must exist and cannot not exist has
for Copleston a definite meaning
Premise 2. To say that a necessary being is a being that must exist and cannot not exist has
for Russell no meaning
Conclusion. Copleston and Russell should stop the debate on God’s existence. [25.1]
3 Premise 1. Copleston conceives of God as a being that must exist and cannot not exist
Conclusion 1. Copleston conceives of God as a being whose essence involves existence
Conclusion 2. Copleston is, whether he wants or not, trying to prove God’s existence by an
ontological argument. [26.1]
4 Premise 1. Russell objects that it does not make sense to ask about the explanation for the
existence of the universe
Premise 2. Copleston is not asking about the reason for the existence of some strange new
and separate object, called ‘universe’; he is instead asking about the reason for the existence
of a anything and everything
Conclusion. Russell’s objection does not touch Copleston’s question. [41.3]
5 Premise 1. It is possible to raise the question of what causes one particular object
Conclusion. It is possible to raise the question of what causes all particular objects. [45.1]
6 Premise 1. Russell seems to think that the unexplainability of the world can be derived from
logical analysis alone
Premise 2. The unexplainability of the world cannot be derived from logical analysis alone
Conclusion. Russell’s position is untenable, and he should produce a proper argument in
favour of the unexplainability of the world. [55.2]
7 Premise 1. Copleston argues from mystics’ experience of God to the existence of God
Premise 2. Hearing voices of Satan is of the same sort as mystics’ experience of God
Conclusion. Copleston should argue from reports of people hearing voices from Satan to the
existence of Satan. [80.3]
8 Premise 4. Colour perception is like feeling that something is good or bad
Conclusion 2. There is no point in asking Russell to give a reason to say that certain things
(like Kramer’s deeds) are bad. [116.3]
9 Premise. A variety of terrestrial lawgivers can best explain the diversity in the content of the
mores of a group, but it is no proper explanation of the form of the moral law
Conclusion. Russell’s full argument at 128.2 does not defeat Copleston’s argument at 127.6.
[129.2]
10 Premise. Russell would say that acting like the Nazis is wrong in itself, independently of any
utilitarian consideration
Conclusion 2. Russell would ultimately have to admit that God exists. [135.4]
440 15 Final Cross-Examination

In most of these arguments (2, 3, 5, 6 ,7 8 and 10) the conclusion refers to how the
participants should act or what their commitments are, what makes of them dialectic
arguments, understanding dialectics as the art of conversation. They are arguments
about the behavior, commitments and entitlements of the participants in the debate.
The same is true for the remaining arguments (1, 4 and 9), but in these cases it is
more difficult to grasp.
What is interesting about 1 is that what in my analysis corresponds to an inter-
argumentative relationship (rebuttal) is expressed here as a judgment about a move of
one of the participants. A rebuttal may be either an argument that rebuts another argu-
ment, or the act of rebutting. Argument dialectics deals with the former, arguer dialec-
tics, and hence pragma-dialectics, with the latter. Compare Fernando’s reconstruction
of 19.3 with mine:

We both know, at any rate, one very

eminent modern thinker whose
knowledge of modern logic was
profound, but who certainly did not
think that metaphysics are meaningless
Therefore
Not all modern logicians surely would
admit the meaninglessness of
metaphysics
Modern logic says that the
word “necessary” is a
useless word, except as
applied to analytic
propositions, not to thingse
The opinions of modern logicians in
against the domain of (meta)philosophy are Therefore
generally reliable:
The word “necessary” is a
useless word, except as
applied to analytic
propositions, not to things

That’s even clearer in argument 4 of the previous list, whose premises and conclu-
sion are about Copleston and Russell’s actions and their effects. Fernando uses
“objection” in the conclusion in the sense of an expression of disapproval or rejec-
tion, rather than in the sense of an argument opposed to another argument. This is
made clear by the occurrence of “Russell objects” in the first premise. Likewise, the
conclusion of 9 is stated in terms of the effects of one of Russell’s moves, a coun-
terargument, on one of Copleston’s moves, a positive argument this one. In fact, and
this is very revealing, it could be said that in 9 Fernando is explaining why Russell’s
counterargument in 128 does not affect Copleston’s positive argument in 127.
The point of view adopted by Fernando fits quite well with what Finocchiaro
understand by argument evaluation:
we may say that an evaluation of an argument consists of a new argument whose conclusion
attributes some merit or flaw to the original argument, whose reasons are meant to justify
such an attribution, and where it is understood that normally such a justification can only be
more or less strong rather than completely right or completely wrong. (Finocchiaro 2013,
p. 23).
15.1 Comments on Part I (Hubert Marraud) 441

The evaluation of arguments, thus conceived, leads, more or less directly, to meta-
argumentation, since “the process of evaluation consists not only of advancing an
evaluative claim about the ground-level argument, but also of justifying that claim.”
(Op.cit., p.1).
Argument evaluation can be made from different perspectives; to cite the best
known, it can be made from a rhetorical, dialectical or logical perspective. From
the first, the question is whether the argument is an effective means to promote the
achievement of the proponent’s ends (generally to persuade the addressee); from the
second, whether in proposing it the arguer complies with the rules of the exchange;
and from the third whether the reason expressed is a good reason for the thesis it
supposedly supports.
Dialectical criticism takes the form of an argument in which the conclusion refers
to the use of an argument and the warrant refers to the conventional rules of dialogue.
Thus, in argument 6 in the above sample:
Premise 1. Russell seems to think that the unexplainability of the world can be derived from
logical analysis alone.

Premise 2. The unexplainability of the world cannot be derived from logical analysis alone.

Conclusion. Russell’s position is untenable, and he should produce a proper argument in

favour of the unexplainability of the world.

The passage from the premises to the conclusion seems to appeal to something akin
to the closure rule of the pragma-dialectical decalogue (Inconclusive defenses of
standpoints may not lead to maintaining these standpoints, and conclusive defenses
of standpoints may not lead to maintaining expressions of doubt concerning these
standpoints).1
In contrast, rhetorical criticism takes the form of a meta-argument in which the
conclusion refers to the use of an argument (as in a dialectical criticism) and the
warrant to the empirical rules of persuasion. Argument 2 in the above sample can be
seen as a rhetorical meta-argument.
Premise 1. To say that a necessary being is a being that must exist and cannot not exist has
for Copleston a definite meaning.

Premise 2. To say that a necessary being is a being that must exist and cannot not exist has
for Russell no meaning.

Conclusion. Copleston and Russell should stop the debate on God’s existence

1 Due to the presence of “logical” rules in the pragma-dialectical decalogue, such as the validity
rule (Reasoning that in an argumentation is presented as formally conclusive may not be invalid
in a logical sense) or the argument scheme rule (Standpoints may not be regarded as conclusively
defended by argumentation that is not presented as based on formally conclusive reasoning if the
defense does not take place by means of appropriate argument schemes that are applied correctly) a
dialectical critique can have logical roots. But the conclusion of such meta-arguments is not that this
or that argument is invalid or inconclusive, but that Copleston (or Russell) cannot use that argument
or has not conclusively defended his standpoint.
442 15 Final Cross-Examination

According to pragma-dialectics participants in a critical discussion attempt to

reasonably resolve a difference of opinion (dialectical objective) by persuading the
other part of one’s position (rhetorical objective). If so, it could be alleged that if the
parties understand key concepts in diametrically opposed senses, they are unlikely
to achieve their rhetorical goal, and therefore there is no point in arguing.
As I have said, I only intend to analyze Copleston’s and Russell’s argumentation
understood as an interactively constructed macro-argument. The core of my anal-
ysis is counter-argumentation, something very different from evaluation, in Finoc-
chiaro’s sense, because offering an argument in opposition to another argument is
quite different from arguing about another argument.
Arguably, most of the time Copleston and Russell counter-argue while Fernando,
as the above sample shows, evaluates their arguments, meta-argues. Fernando’s
account of the Copleston-Russell debate also reflects the fact that participants also
occasionally argue about how their considerations fit into the framework of the discus-
sion. Of course, these “procedural” assertions, when they do not correspond to inter-
argumentative relationships, fall outside the macro-argument that is the subject of my
argument-dialectical analysis. This explains the striking difference in the number of
arguments identified in the debate by both approaches, which is significantly higher
in the case of Fernando’s analysis. Sometimes this is because I consider mere illustra-
tions or explanatory reasons what for Fernando are arguments from example or justi-
fying reasons (for example in 11 or 35), but more often it is due to the reconstruction
of dialectical meta-arguments.
Other important differences between Fernando’s erotetic pragma-dialectical anal-
ysis and my argument dialectical analysis concern the choice of a model of argument.
By “argument model” I mean a specification of the parts and structure of a simple
argument, and by “simple argument” I mean one that has no parts that are themselves
arguments, i.e. one that has no subarguments. Fernando’s pragmatic erotetic analysis
assumes the traditional premise-conclusion model, while I opt for my own version
of Toulmin’s model, distinguishing in consonance between premises, warrant and
conclusion. This seems at first sight a difference in style: nothing would prevent
us from using Toulmin’s model in an erotetic pragma-dialectical account, although
some adjustments would be necessary. However, the traditional premise-conclusion
model is atomistic while my version of Toulmin’s model is holistic. The distinction
between atomism and holism, in the sense which is relevant here, comes from the
theory of reasons. As defined in 9.5, To argue is to present to someone. something
as a reason for or against something else. Now the central question in the theory of
argument is “Shall we accept this claim on the basis of the reasons put forward in
support of it?” (Wenzel 2006, p.16). By logical properties I mean those properties
of arguments that are relevant to answer Wenzel’s question and can be described
without referring either to the effects on the audience or to the rules of argumenta-
tive exchanges. For the atomist the logical properties of an argument are completely
determined by the properties of its parts and the relationships among them, while
for the holist they are also dependent on contextual elements that are not part of the
argument. In this sense, it could be said that for the atomist logical properties are
intrinsic properties of arguments, while for the holist they can be extrinsic properties.
15.1 Comments on Part I (Hubert Marraud) 443

The choice between one or the other model of argument has consequences for the
identification and addition of non-explicit elements. In general, analysis shaped by
the premise-conclusion model are more likely to “discover” non-explicit or hidden
elements. The traditional premises-conclusion model endorses what Don Levi calls
the PC Requirement for Premises-Conclusion), the atomistic principle that “every-
thing about the rhetorical context that is relevant for determining what is being argued
be incorporated into its restatement as a PC sequence.” (Levi 1995, p. 80).
The problem of implicit or hidden premises is a specific problem of argumentation
theory arising from adherence to an atomistic principle such as the PC Requirement,
which should not be confused with the more general pragmatic problem, studied
in philosophy of language and linguistics, of the distinction between what is said
and what is communicated, which has given rise to concepts such as implicature.
The reason for adding implicit premises is not the discrepancy between the literal
meaning of what is said and the interpretation of the speaker, but that the explicit
premises do not seem sufficient to determine the logical value of the argument.
My idea of warrant is holistic. One of the key differences between a premise and
a warrant is that a non-explicit warrant is not a missing component of the argument,
but something that can be asked for. Just as one can ask for a justification of one
of the premises, one can demand a justification of the passage from these to the
conclusion. The first can lead to a chain of arguments, the second to the specification
of a warrant. Knowledge of the warrant is not necessary to identify the argument,
but to evaluate it. 2
Given my adherence to holism, my analyses of Russell’s and Copleston’s argu-
ments will displease any atomist, who will consider them incomplete because they
do not make explicit all the assumptions on which their validity depends. Indeed,
the very notion of the totality of assumptions on which the validity of an argument
depends is strange, if not incomprehensible, from a holistic perspective. What I have
intended is to bring out all the contextual considerations that Copleston and Russell
consider relevant in their dialogue for the cogency of the arguments proposed and
examined.
A second difference between the two analyses, related to the choice of argument
model, is the use of formal logic, that points to a deeper difference. The concept of
valid argument, as usually defined, presupposes the premise-conclusion approach,
and therefore is not easily combined with a model like Toulmin’s (and is certainly
incompatible with any holistic model). Moreover, Toulmin’s model arises from the
estimation of “the standards and values of practical reasoning (developed with an
eye to what I called ‘substantial’ considerations)” versus “the abstract and formal
criteria relied on in mathematical logic and much of twentieth-century epistemology
(Toulmin 2003, p. xi). Fernando occasionally uses the concept of valid argument,
although he also mentions other forms of inference, or the concept of consequence
(as in “the conclusion does not follow, in spite of Copleston’s definition” [11.3.1]),
something that belongs to an inferential approach to argument. The point is not the

2 While this is an issue that deserves more detailed consideration, this difference may be important
in clarifying the idea of a “version” of an argument.
444 15 Final Cross-Examination

opposition between different forms of logical inference, but the contrast between an
inferential and a reasons-based approach to argument. On the inferential account, in a
good argument the conclusion may be inferred or follows (abductively, deductively,
inductively,…) from the premises, while on a reasons-based account, in a good
argument the premises express a good reason for the conclusion.
Although the distinction between reasons-based theories and inference-based
theories of arguments is of my own making, the need to make some similar distinc-
tion has been defended, among others, by authors like Gilbert Harman (2002) and
John Woods (2016), although the terminological variation from one author to another
may be confusing. The difference between inferentialism and reasonism can also be
explained in terms of the different senses of ’conclusion’. That word has an infer-
entialist sense when speaking of the conclusion of an argument (e.g., an argument
consists of a set of premises and a conclusion) and a reasonist sense when speaking
of the conclusion of an argumentative dialogue or exchange (e.g., after deliberation,
they came to the conclusion that...).
In inferential approaches the conclusion is first and foremost the conclusion of
an argument. If the concept of the conclusion of an argument is taken as basic, a
dialogue can be conceived as a process of selection between competing arguments,
so that the conclusion of the dialogue is the conclusion of the selected argument. In
reasonist approaches the conclusion is the provisional conclusion of a stage of the
exchange, resulting from the joint consideration of the considerations available at that
stage (what is called in chapter 9 the orientation of an argumentative structure), and
the conclusion of an argument is defined in terms of the effect of the considerations
adduced on the conclusion that can be drawn at the time they are adduced.
Another elective difference is that Fernando’s analysis is philosophical, in a sense
that, unfortunately, mine is not. Fernando places Copleston and Russell’s contri-
butions in a broader context, with references to other philosophers, such as Kant,
Wittgenstein, Geach, Hintikka, or Kripke, and other doctrines. That is, Fernando’s
analysis is an analysis made by a philosopher of a philosophical discussion. What is
more, Fernando is a philosopher who, from time to time, dialogues with Copleston
and Russell, as the following passage shows:
Although Copleston does not say anything about those, we may press the point in his favour
and argue that all non-living stuff also exists because of certain forces that produced it in
the past and keep it together in the present. (We could even, somewhat anachronistically,
that those forces are, in the last instance, the four fundamental forces of nature; see Davies
1986). The idea of force gives in fact an interesting twist to the old concept of ‘reason’.”
[11.1.2]

Mine is a more self-contained analysis, something that in all probability is a

comparative weakness.
Despite all these differences, I think that both approaches are complementary. The
point is that the macro-argument I’m interested in is the logical structure that under-
lies the discussion. To support this claim I will focus on what we might consider to
be the argument-dialectical counterpart of the concluding stage of a critical discus-
sion. In the concluding stage the parties assess the extent to which the difference
of opinion has been resolved and in whose favor. If the protagonist withdraws the
15.1 Comments on Part I (Hubert Marraud) 445

standpoint, the difference of opinion is resolved in favor of the antagonist; if the antag-
onist clears up her doubts, it is resolved in favor of the protagonist. (van Eemeren
et al. 2002, p. 25). From the point of view of argument dialectics, in the argumenta-
tion stage participants try to interactively weave a macro-argument. Sometimes they
succeed and sometimes they don’t (they may be unable to assemble all the pieces
together or they may each assemble a different macro-argument, as it happens in the
Russell and Copleston debate). If they succeed, the macro-argument reflects, so to
speak, the reasonable resolution of the conflict, since a macro-argument has a certain
argumentative orientation.
The foregoing suggests a procedure to determine whether the participants have
reasonably resolved their difference of opinion. This procedure is based on two
questions:
• Have the participants interactively constructed a macro-argument?
• If so, what is the orientation of the macro-argument (positive, negative or neutral)
towards the issue under discussion?
These questions allow us to distinguish various types of failure. First, the partici-
pants may have been unable to construct a macro-argument. Second, if they have been
able to do so and if the difference of opinion is mixed, then the success of the discus-
sion would require that the interactively constructed macro-argument be oriented
positively or negatively towards the controversial claim. Therefore, the discussion
is also unsuccessful whenever the resulting macro-argument is not at all oriented
towards the controversial issue or has a neutral orientation towards it.
In line with these considerations, a reasonable resolution of a difference of opinion
would be a resolution based on the joint assessment, along the discussion, of the
concurring arguments’ merits, as reflected in the orientation of the resulting macro-
argument. Reasonableness requires that the participants’ final verdict coincides with
this orientation, not being a matter of mere agreement.
The logical dimension is necessary to make sense of the idea that in a reasonable
discussion a difference of opinion is solved on the merits of the adduced arguments.
At the same time, in order not to betray the dialectical character of the model, the
logical dimension cannot be external to the discussion itself — a distinctive mark of
dialectic is the rejection of any external standard for argument evaluation. In argument
dialectics this is achieved through the notion of an interactively built macro-argument.

15.2 Comments on Part II (Fernando Leal)

For reasons beyond my control, Hubert finished the above comments before I had
the chance of writing mine. I shall take advantage of that happenstance to refer to
its contents in the following. But let me start with four basic ideas that guide my
approach to argumentation. Following the example of pragma-dialecticians’ talk of
‘commandments’, let me present those ideas in the form of a creed.
446 15 Final Cross-Examination

(1) First of all, I believe that argumentation is everywhere. People cannot interact
without asking for, producing, reacting to, assessing, and rejecting or accepting
reasons, and with them, of course, all those things, said, implied, done, planned,
or envisaged, in support of which reasons were asked for in the first place. We
do this all the time, and to such an extent that the ancient definition of a human
being (animal rationale, ζùoν λóγoν χoν) should best be translated as ‘the
argumentative animal’. In fact, I believe that many things that we do with words
have an argumentative character, even if, on the face of it, they seem to belong
to different categories, say, telling stories, developing similes, inventing new
descriptions, raising questions.
(2) Having said that, and in view of the fact that all these constant, obsessively
persistent argumentative activities have different levels of performance and are
more or less obedient to different sets of rules and institutions, I believe, in the
second place, that whatever theory we try to construct will unavoidably feed,
depend, and be based on whatever examples we happen to consider interesting.
No theory can be stronger than the instances of argumentation it accounts for.
This conviction is what attracted me to Hubert’s theory in the first place. In
contrast to many other admirable theorists in our field, who are able to write
extensively and cogently without the support of examples of argumentation,
Hubert needs to ground everything he says on concrete cases which he analyzes
with rare ability in order to come up with his extraordinary insights. I doubt
whether there is any other researcher in argumentation studies who comes close
to having the database of argumentative texts that he has accumulated over the
years. This wonderful empiricism is something that I share with him.
(3) I believe, thirdly, that the kinds of argument that emerge from human interaction
are distributed along a gradient or continuum. On one extreme, we have the
extremely long and complex arguments that we find in mathematics, the law,
history, and the various technologies, sciences and humanities. Our civilization
wholly depends on those long chains of reasons that argumentative animals
slowly and laboriously, often collectively, organize in order to prove theorems,
solve problems, establish legal opinions and verdicts, explain big structures and
large processes, create engines, machines, and gadgets, confirm and disconfirm
models, hypotheses, and theories, criticize works of literature and art, or devise
new and striking worldviews. On the other extreme, we find, practically in every
human interaction, what we may call mere argumentative gestures, i.e. not quite
arguments as such, but rather something like indications of the general direction
of the argumentative activity going on in us as we struggle to be reasonable
in the face of an opponent. In the middle between these two extremes we find
the simple arguments which Hubert mentions above, i.e. arguments no parts of
which are themselves arguments. Those simple arguments are, quite naturally,
the delight of all those logical minds who found their way to argumentation
studies once the limitations of formal logic dawned on them.
(4) Finally, I believe that, given the huge odds against us human beings with regard
to mutual comprehension, that is, the high probability of misunderstanding each
other when trying to communicate, the best, if not indeed the only chance, we
15.2 Comments on Part II (Fernando Leal) 447

have to construe correctly what our discussion partners mean, is to have a

firm grasp of what they are talking about. In the argumentation literature, this
factor is described variously as background knowledge, background beliefs, or
background information. As far as I can see, there is no proper theory of that
factor at all. We know it is there, but we find it hard to model appropriately
how it works. Every time we analyze and interpret, evaluate or criticize an
argument, we try to pin down or even guess that background, for otherwise
we cannot even start to understand the argument, no matter how many purely
formal clues there may be in the text before us. This pinning down or guessing
is sometimes done in full awareness of the problems, sometimes we do it
haphazardly, unsystematically, even perfunctorily; but we have to do it, one
way or another.

Although I may have failed in articulating clearly and convincingly these four
ideas, it is quite likely that what I have presented here as a creed will be, totally or
in part, controversial. When I started working on the field of argumentation, they
were not clear to me; but they have become increasingly so over the years, to the
point where I think that taking them seriously is necessary, at least if one is trying
to analyze a piece of argumentation as intricate as the debate between Russell and
Copleston. Let me now try to express what I have learned both from analyzing that
debate and from the comparison with Hubert’s analysis.
In Sect. 15.1 of this chapter, Hubert has described certain differences between
our analyses. I can only confirm both his description and his diagnosis. In fact, these
differences between us are the reason why we did not manage to achieve a unified
analysis, as we both hoped when we started working, independently, on the Russell-
Copleston debate. In terms of the ideal model of critical discussion, Hubert is mainly
interested in the Argumentation Stage, in which
protagonists advance their arguments for their standpoints that are intended to systemati-
cally overcome the antagonist’s doubts or to refute the critical reactions given by the antag-
onist. The antagonists investigate whether they consider the argumentation that is advanced
acceptable. If they consider the argumentation, or parts of it, not completely convincing, they
provide further reactions, which are followed by further argumentation by the protagonist,
and so on. In this way, the structure of the argumentation a protagonist puts forward in the
discourse can become very complicated: This structure may, in fact, vary from extremely
simple to extremely complex. Although in practice, as a rule, parts of the argumentation stage
remain implicit, there is only an argumentative discourse if it is clear that argumentation is,
in some way or other, advanced. It is crucial for the resolution of a difference of opinion
that argumentation is not only advanced, but also critically evaluated. Without both these
activities taking place, there can be no question of a critical discussion. [van Eemeren and
Grootendorst 2004, p. 61.]

Pragma-dialecticians have synthesized insights from linguistic pragmatics and

informal logic to supply a method to identify the structure of the arguments presented
by the protagonist. In the case of mixed differences of opinion, of course, both parties
do produce arguments, whose structure has to be identified. And, if we accept mixed
polylogues, then we shall have more than two protagonists, each presenting her own
arguments.
448 15 Final Cross-Examination

To this picture, Hubert adds two things. First, he allows for counterarguments,
i.e. actual arguments produced by the antagonist as part of her ‘critical reac-
tions’. The protagonist, faced with such counterarguments, may produce counter-
counterarguments, and so on. Secondly and crucially, those counterarguments build
upon previous arguments in such a way that we can witness the joint construction of
what Hubert calls macro-arguments. One of his new thoughts, emerging from this
collaboration, is that, when the critical discussion has been successful, those macro-
arguments can be the best representation of a proper Concluding Stage, in which the
results of the joint work of the arguers is precisely seen as the resolution of the initial
disagreement.
These two extraordinarily fruitful concepts—counterargument and macro-
argument—seemed to me, when our collaboration started, eminently compatible with
pragma-dialectics; and I still think they are. However, other aspects of my method of
analysis are, at the present stage, obstacles that impede an easy harmonization with
this simple solution.
The most obvious obstacle has to do with article (1) of my creed above. I just don’t
think that arguments are only presented in the Argumentation Stage. In principle, I see
no reason why discussion partners might not ask for and produce arguments in any of
the four stages of a critical discussion. As van Eemeren says in an important footnote
added to a reprint of van Eemeren and Houtlosser (2007): “In pragma-dialectics we
reserve the term subdiscussion for critical exchanges about material premises and the
term meta-dialogue for critical exchanges about procedural premises” (van Eemeren
2015, p. 638, n. 1). These subdiscussions and meta-dialogues require argumentation,
which nonetheless is part of the Opening Stage. Consider Table 15.1, which summa-
rizes my analysis of Segments I and II, the latter dedicated to Copleston’s argument
from contingency.
The details can be found in my analysis, but the reader can see that there is quite
a lot of argumentation going on in the recurring Opening Stages—a fact that is
quite typical of philosophical discussions. By the way, the verb ‘to win’ applied to
Copleston twice in Table 15.1 is obviously just shorthand for the fact that Russell’s
arguments are successfully countered by Copleston, so that Russell modifies his
argumentative strategy.
In fact, in some philosophical discussions there may be argumentation already
at the level of the Confrontation Stage.3 As the Russell-Copleston debate begins,
Copleston starts by suggesting, as a material starting point, a definition of the concept
of God, which has been a notorious subject of controversy in Western theology and
philosophy. This is a very skillful move on Copleston’s part, and Russell could have
started a discussion right there, so that we would have had a subdiscussion about
the meaning of the term “God” even before entering the main discussion about the
question whether god exists or not. It is clear that, if two philosophers cannot

3The Confrontation Stage is absent from Table 15.1, because it was implicit as part of the commu-
nicative event that is a debate. From time to time, the main difference of opinion (as to whether
god exists or not) is alluded to; and the same is true of other differences of opinion, tied up to
subdiscussions, which emerge from the main discussion.
15.2 Comments on Part II (Fernando Leal) 449

Table 15.1 Summary of Segments I-II of the Russell-Copleston debate

Turns Stage Observations
1–10 Opening Role assignment as well as material and procedural starting points
agreed on by both parties
11 Argumentation Presentation of the first version of the argument from contingency by
Copleston
12–24 Opening Return to the Opening Stage by a major material and procedural
starting point put forward by Russell. Subdiscussions about the
meaning, or lack of it, of the term “necessary being” as well as the
sentence “God exists”. The subdiscussion on “necessary being” is left
undecided, but the subdiscussion on “God exists” is ‘won’ by
Copleston. At Turn 20, Russell’s first attempt at reduction of the
argument from contingency to the ontological argument does not
succeed
25 Concluding Copleston tries, for the first time, to close the discussion in view of
unresolved disagreement on the meaning of “necessary being”
26–31 Opening Return to the Opening Stage by Russell’s second attempt to reduce the
argument from contingency to the ontological argument. A
subdiscussion about this reduction takes place that is, again, ‘won’ by
Copleston
32–51 Opening Russell requests clarification of the term “sufficient reason”. This leads
to a subdiscussion on the meaning of two terms, “cause” and
“explanation”, and to another subdiscussion on the proper rules to
follow when looking for causes and pursuing explanations. There is an
impasse
52–67 Argumentation Russell finally engages the argument from contingency. Copleston
produces two different versions of the argument, which Russell objects
to
67 Opening Copleston tries to re-assign roles and shift the burden of proof to
Russell but fails to do so
68–71 Concluding Copleston tries, for the second time, to close the discussion in view of
Russell’s rejection, not of the argument from contingency, but of the
very question which presides over the debate

agree on the meaning of the term, they could hardly discuss the question of existence.
As it is, such a discussion, at the Confrontation Stage, does not happen there, although
it could be argued that, later on, the question of the meaning of ‘god’ does indeed
come up.
Finally, there may be argumentation at the Concluding Stage. Hubert has already
suggested that, in a critical discussion, the full macro-argument jointly constructed
may be seen as the Conclusion, for the discussion would then be resolved on the
relative strength of the arguments presented and established by the joint work of
both parties. On the other hand, philosophy is known for unresolved and unresolvable
disagreements, at least as long as we expect convergence on any given standpoint
(Woods 1988; Chalmers 2015). However, if we are more modest in our expectations
and make do with an agreement on what standpoints are arguable (Leal 2019), then
450 15 Final Cross-Examination

we may understand the special kind of Concluding Stage that is at all possible in
philosophy. Anyway, even at the Concluding Stage, we may find argumentations, as
is certainly the case in the last two turns of the Russell-Copleston debate.
If it is accepted, therefore, that argumentation does not only take place in the
Argumentation Stage of a critical discussion, then it is difficult to accept, with Hubert,
that “From the point of view of argument dialectics, in the argumentation stage
participants try to interactively weave a macro-argument,” for this would imply that
no such weaving can take place anywhere else in the discussion. This difficulty
is probably tied up with my finding many more arguments in a text than Hubert
does. The fact that I do so depends on articles (1) and (3) in my creed, namely, that
argumentation is everywhere and that arguments fall somewhere along a continuum.
Let me explain.
Hubert’s methods of analysis can, I think, be used to great profit in the disentangle-
ment of the very long and complex arguments that are part of our judicial decisions,
our experimental set-ups, our axiomatic systems, and our construction of sophisti-
cated scientific theories. All that argumentative work is collective and depends on the
polishing and working-out of arguments done over generations. Think, for instance,
of the theory of relativity. It is well known that its discovery was part of a network in
which physicists all over the world were working on the same or similar problems.
Einstein became the first to reach the summit, but somebody was bound to do so
sooner or later. And once he wrote his famous paper, the theory did not stand still,
but has been well honed and enriched by many people, so that, as Walton says, no
student is expected to learn the theory from Einstein’s original paper, given that its
“doctrines…have been… improved upon and are better expressed, more lucidly and
unitedly presented in a good college physics text” (Walton 2008, p. 159).
The same is true of judges, engineers, or indeed philosophers. They all rehearse
arguments again and again, modify them, make them more concise, extend and apply
them, in a long sequence of collectively constructed macro-arguments. This is the
pole of long and complex arguments, the stuff of political constitutions, law codes,
technical procedures, scientific theories, philosophical systems. Although Hubert’s
methods have so far been applied to relatively short texts, they seem to me to be most
promising when collective argumentation across time is at stake.
On the other hand, when two concrete people, like Russell and Copleston in their
debate, meet and try to have a reasonable discussion, even if knowing that none of
them will be persuaded by what the other party says, many more things happen:
One can certainly find single arguments, like those presented by Copleston, and
dignified by prestigious, tradition-laden names such as The Argument from Contin-
gency or The Cosmological Argument. These are indeed excellent examples of argu-
ments that were formulated hundreds and even thousands of years ago, which have
been reformulated many times by many people working together and against each
other, which would not exist if it were not because of so much collective work. But
the two parties in dispute are heirs, they have been handed down those arguments and
have each their own take on them: Copleston cherishes them as capable of revealing
great truths, Russell has long come to the conclusion that they are failed arguments,
full of sophistry and even scandalous (Russell 1900, §112). However, neither of them
15.2 Comments on Part II (Fernando Leal) 451

is known for having improved upon or developed this or that aspect of the famous
arguments in support of God’s existence that are the subject matter of their debate.
But, as they work their way through those arguments, in Copleston’s presentation of
them, they manage to produce many mini-arguments, local and limited, which any
analysis has to recover. Part of the work of recovery can be done, and done very well,
by Hubert’s methods, namely all those that have a clear structure. In fact, Hubert
can recover not only the structure of all those mini-arguments, but also the structure
of the larger arguments of which they are revealed to be part thanks to his analysis.
However, another part of the work of recovery escapes through the microscopic holes
of the fine sieve that Hubert uses to catch arguments. They are just so small, often
no more than argumentative gestures, hints at possibly full arguments that are never
quite worked out as arguments. Yet I contend that they are there, produced by the
incessant activity of arguing that is our lot as argumentative animals.
As an example, Take Turn 25 of Segment II of the debate, in which Hubert does
not see an argument. As background information, let me remind the reader that
Copleston presents his argument from contingency at Turn 11, immediately after
the initial Opening Stage has been successfully carried out and some important
material and procedural starting points have been secured. Does Russell respond to
the argument in one of the many ways we are used to? Does he find one or several of
Copleston’s premises unacceptable or irrelevant to the conclusion? Does he find that
they are not sufficient for the conclusion? No, he does not. Instead he starts quibbling
about a phrase (“necessary being”) which was not even used at all by Copleston. For
reasons that I cannot go into here, I interpret this as a return to the Opening Stage
in which certain preliminary issues should be discussed, which are in part material
(whether the word ‘necessary’ can be applied to individual objects
or only to propositions) and in part procedural (whether the strictures
of russell’s mathematical logic can or should be used to decide a
metaphysical question).
At first, Copleston is at a loss, because the move back to the Opening Stage is
not clear to him, but at some point a revealing sentence on Russell’s part seems to
strike him: “These phrases don’t, for me, have a significance, except within a logic
that I reject” (Turn 16). At that point, Copleston recognizes that all the quibbling is
indeed a matter of logic—of formal mathematical logic to boot. And so, he tries to
find out more about it. There is some more quibbling, and an extremely interesting
discussion about the theory of descriptions, until it becomes clear to Copleston that
Russell will not be moved by any argument to accept that the phrase “necessary being”
has any meaning at all. So, Copleston tries to close Segment II (on the cosmological
argument) and go over to Segment III (on the moral argument). He says: “Well, we
seem to have arrived at an impasse. To say that a necessary being is a being that must
exist and cannot not exist has for me a definite meaning. For you it has no meaning.”
This move for closure is not accepted by Russell, but what is interesting to me is
that Hubert does not see any argument here (see his analysis). I myself think that
there is here, if not a full-blooded argument, at least an argumentative gesture. It is
a move for closure that is based on reasons. Whether those reasons justify closing
the discussion on the cosmological argument can be disputed; in fact, Russell does
452 15 Final Cross-Examination

not agree and proposes to go on. He believes that an argumentative strategy that he
tried to use before (at Turn 20), namely, the reduction of the cosmological to the
ontological argument may be a way to refute Copleston more effectively than the
logical quibblings that led nowhere. He fails in that, but that is another story.
To bring this discussion to an end, I very much believe that the other difference
highlighted by Hubert between our two approaches is greatly responsible for my
finding more arguments than he does and, in general, for analyzing the debate in the
way I do. This difference concerns the fact that my “analysis is an analysis made
by a philosopher of a philosophical discussion”. This is true. And it means that I
bring to bear all philosophical background information that I possibly can in order
to understand what is going on in this debate. This corresponds to article (4) of my
creed. Sometimes, I have the impression that argumentation theorists believe that
their superior analytic skills allow them to pass judgment on the structure and worth
of the arguments they examine. But my position is that the only way in which one can
pass such judgments is by actually knowing what the arguers are talking about. If one
wants to evaluate an economic argument, one should better know some economics;
and the failings of a legal argument presented by an attorney can only be correctly
spotted by a person with lots of legal knowledge. The study of argumentation is a
wonderful field, full of ideas and insights that deserve to be better known; but it does
not substitute the know-how of the expert arguer. To think otherwise is to incur, once
more, in the arrogance of the ancient Greek sophists and rhetors that Plato rightly
castigated.

References

Chalmers, D. J. (2015). Why isn’t there more progress in philosophy? Philosophy 90 (1), 3-31.
Crosswhite, J. (1996). The rhetoric of reason. Madison, WI: University of Wisconsin Press.
Davies, P. C. W. (1979). The forces of nature. Cambridge, UK: Cambridge University Press.
Engel, N. (2017). Normative reality: Reasons fundamentalism, irreducibility, and metaethical
noncentralism. Ph. D. Thesis, Columbia University. Retrieved from https://doi.org/10.7916/D80
P14RT.
Eemeren, F. H. van (2010). Strategic maneuvering in argumentative discourse. Amsterdam: John
Benjamins.
Eemeren, F. H. van (2015). Reasonableness and effectiveness in argumentative discourse: Fifty
contributions to the development of pragma-dialectics, Cham: Springer.
Eemeren, F. H. van, and Grootendorst, R. (2004). A systematic theory of argumentation: The pragma-
dialectical approach. New York: Cambridge University Press.
Eemeren, F. H. van, and Houtlosser, P. (2007). Countering fallacious moves. Argumentation 21,
243-252.
Eemeren, F. H. van, Grootendorst, R., and Snoeck Henkemans, A. F. (2002). Argumentation:
Analysis, evaluation, presentation. Mahwah, NJ: Lawrence Erlbaum.
Leal, F. (2019). On philosophical argumentation: Towards a pragma-dialectical solution of a puzzle.
Journal of Argumentation in Context 8 (2), 173-194.
Russell, B. (1900). A critical exposition of the philosophy of Leibniz, with an appendix of leading
passages. Cambridge, UK: Cambridge University Press.
Skow, B. (2016). Reasons why. New York: Oxford University Press.
References 453

Toulmin, S. E. (2003) The uses of argument. Updated edition. New York: Cambridge University
Press.
Walton, D. N. (1998). The new dialectic: Conversational contexts of argument. Toronto: University
of Toronto Press.
Woods, J. (1988). Is philosophy progressive? Argumentation 2, 157-174.
Appendix
Text of the Russell-Copleston Debate

Note on the text.

Russell-Copleston debate divided by turns (C = Copleston, R = Russell; 138 turns,
69 turns each)

The only reliable source for the text of the Russell-Copleston debate is the recording
of the broadcast itself. We have written to the BBC archives to request access to it,
but, probably due to the coronavirus emergency, we have not received any response.
The Russell archives at MacMaster University in Canada have kindly sent us an
edited recording of about 15 min of the broadcast. This recording was part of the
Open University course “The Problems of Philosophy” and is available on YouTube.
We have used it here, marking with ‘Quine corners’ the beginning () and end () of
the fragments. Through careful extrapolation of the facts as audible in the recording,
we have studied and amended the critical edition made by Professor Slater in The
Collected Papers of Bertrand Russell (vol. 11, paper 68, pp. 521–541; critical notes,
pp. 815–820), which unfortunately is based only on extant transcriptions and editions,
and not on the recording. We were very conservative in our changes. First, we restored
the colloquial contractions (it’s, I’ve, and so on) which Slater (2011) took over from
the version printed in Morris (1956), for this allows better to distinguish the empha-
sized auxiliaries (“it is”), which have some importance as argumentative indicators.
Secondly, we restored the very frequent use of “Well” to start a response, given that
this particle has a definite argumentative value. Thirdly, we added the oral emphases
audible in the recording by use of italics, and in a few cases also in passages where it
was quite likely that they also took place during the broadcast. Fourthly, we amended
passages which were obscure due to the partly improvised character of the inter-
ventions. These amendments we mark with angle brackets (< > ), as is usual in
philological critical editions. Fifthly, we used bold italics to highlight propositions
and arguments which the discussants propose for consideration. In the case of terms
(nouns and phrases), we merely used inverted commas. All other inverted commas

© The Editor(s) (if applicable) and The Author(s), under exclusive license 455
to Springer Nature Switzerland AG 2022
F. Leal and H. Marraud, How Philosophers Argue, Argumentation Library 41,
https://doi.org/10.1007/978-3-030-85368-6
456 Appendix: Text of the Russell-Copleston Debate

in the extant editions were deleted. Finally, on two occasions, Copleston says some-
thing he didn’t intend (“this reason” at Turn 11; “caused” at Turn 55) but corrects
himself immediately. We did not include the mistaken words in this text to avoid
confusions. This leaves only three sentences where our departure from Slater (2011)
merits some justification, two in Turn 14 and one in Turn 135. The justification is
given in footnotes on site.
1. C: As we are going to discuss the existence of God, it might perhaps be as
well to come to some provisional agreement as to what we understand by
the term “God”. I presume that we mean a supreme personal being, distinct
from the world and creator of the world. Would you agree—provisionally at
least—to accept this statement as the meaning of the term “God”?
2. R: Yes, I accept this definition.
3. C: Well, my position is the affirmative position that such a being actually
exists, and that His existence can be proved philosophically. Perhaps you
would tell me if your position is that of agnosticism or of atheism. I mean,
would you say that the non-existence of God can be proved?
4. R: No, I should not say that; my position is agnostic.
5. C: Would you, then, agree with me that the problem of God is a problem of
great importance? For example, would you agree that if God does not exist,
human beings and human history can have no other purpose than the purpose
they choose to give themselves, which—in practice—is likely to mean the
purpose which those impose who have the power to impose it?
6. R: Roughly speaking, yes; though I should have to place some limitation on
your last clause.
7. C: Would you agree that if there is no God—no absolute being—there can
be no absolute values? I mean, would you agree that, if there is no absolute
good, the relativity of values results?
8. R: No, I think these questions are logically distinct. Take, for instance, G. E.
Moore’s Principia Ethica, where he maintains that there is a distinction of
good and evil, that both of these are definite concepts. But he does not bring
in the idea of God to support that contention.
9. C: Well, suppose we leave the question of good till later, till we come to
the moral argument. I should like first to advance a metaphysical argument.
I would like to put the main weight on the metaphysical argument based on
Leibniz’s argument from contingency, and then later we might discuss the
moral argument. Suppose that I give a brief statement on the metaphysical
argument and then we go on to discuss it?
10. R: That seems to me to be a very good plan.
11. C: Well, for clarity’s sake, I will divide the argument into distinct stages.
First of all, I should say, we know that there are at least some beings in the
world which do not contain in themselves the reason for their existence. For
example, I depend on my parents, and now on the air, and on food, and so on.
Now, secondly, the world is simply the real or imagined totality or aggregate
of individual objects, none of which contain in themselves alone the reason
Appendix: Text of the Russell-Copleston Debate 457

for their existence. There isn’t any world distinct from the objects which
form it, any more than the human race is something apart from the members.
Therefore, I should say, since objects or events exist, and since no object of
experience contains within itself the reason of its existence, the totality of
objects must have a reason external to itself. That reason must be an existent
being. Well, this being is either itself the reason for its own existence, or it
is not. If it is, well and good. If not, then we must proceed further. But if we
proceed to infinity in that sense, then there’s no explanation of existence at
all. So, I should say, in order to explain existence, we must come to a being
which contains within itself the reason for its own existence—that is to say,
which cannot not exist.
12. R: This raises a great many points and it’s not altogether easy to know where
to begin, but I think that, perhaps, in answering your argument, the best point
at which to begin is the question of a necessary being. The word “necessary”,
I should maintain, can only be applied significantly to propositions, and, in
fact, only to such as are analytic, that is to say such as it is self-contradictory
to deny. I could only admit a necessary being if there were a being whose
existence it is self-contradictory to deny. I should like to know whether you
would accept Leibniz’s division of propositions into truths of reason and
truths of fact; the former, the truths of reason, being necessary.
13. C: Well, I certainly should not subscribe to what seems to be Leibniz’ idea
of truths of reason and truths of fact, since it would appear that for him there
are in the long run only analytic propositions. It would seem that for Leibniz
truths of fact are ultimately reducible to truths of reason. That is to say, to
analytic propositions, at least for an omniscient mind. I couldn’t agree with
that. For one thing it would fail to meet the requirements of the experience of
freedom. I don’t want to uphold the whole philosophy of Leibniz. I’ve made
use of his argument from contingent to necessary being, basing the argument
on the principle of sufficient reason, simply because it seems to me a brief and
clear formulation of what is, in my opinion, the fundamental metaphysical
argument for God’s existence.
14. R: But, to my mind, “a necessary proposition” has got to be analytic. I don’t
see what else it can mean. And analytic propositions are always complex and
logically somewhat late. Rational animals are animals is an analytic propo-
sition1 ; but a proposition such as This is an animal can never be analytic. In
fact, all the propositions that can be analytic are somewhat late in the build-up
of propositions.2

1 For some strange reason, all editions have “irrational animals” instead of “rational animals”, which
is what one can clearly hear in the recording. The proposition Rational animals are animals makes,
in any case, better sense, in that it corresponds perfectly to Russell’s idea that analytic propositions
are “somewhat late”.
2 Russell says “all propositions that can be analytic” but probably just meant “all propositions that

are analytic”. However, whereas a sentence can be (interpreted as) analytic or as non-analytic,
i.e. as representing an analytic or a non-analytic proposition, propositions either are or are not
458 Appendix: Text of the Russell-Copleston Debate

15. C: Take the proposition If there is a contingent being, then there is a neces-
sary being. I consider that that proposition, hypothetically expressed, is a
necessary proposition. If you are going to call every necessary proposition an
analytic proposition, then, in order to avoid a dispute in terminology, I would
agree to call it analytic, though I don’t consider it a tautological proposition.
But the proposition is a necessary proposition only on the supposition that
there is a contingent being. That there is a contingent being actually existing
has to be discovered by experience, and the proposition that there is a contin-
gent being is certainly not an analytic proposition, though once you know, I
should maintain, that there is a contingent being, it follows of necessity that
there is a necessary being.
16. R: The difficulty of this argument is that I don’t admit the idea of a necessary
being and I don’t admit that there is any particular meaning in calling other
beings contingent. These phrases don’t, for me, have a significance, except
within a logic that I reject.
17. C: Do you mean that you reject these terms because they won’t fit in with
what is called “modern logic”?
18. R: Well, I can’t find anything that they could mean. The word “necessary”,
it seems to me, is a useless word, except as applied to analytic propositions,
not to things.
19. C: In the first place, what do you mean by “modern logic?” As far as I know,
there are somewhat differing systems. In the second place, not all modern logi-
cians surely would admit the meaninglessness of metaphysics. We both know,
at any rate, one very eminent modern thinker whose knowledge of modern
logic was profound, but who certainly did not think that metaphysics are
meaningless or, in particular, that the problem of God is meaningless. Again,
even if all modern logicians held that metaphysical terms are meaningless,
it would not follow that they were right. The proposition that metaphysical
terms are meaningless seems to me to be a proposition based on an assumed
philosophy. The dogmatic position behind it seems to be this: What will not
go into my machine is non-existent, or it is meaningless, or it is the expres-
sion of emotion. I am simply trying to point out that anybody who says that
a particular system of modern logic is the sole criterion of meaning is saying
something that is over-dogmatic; he is dogmatically insisting that a part of
philosophy is the whole of philosophy. After all,  a contingent being is a
being which has not in itself the complete reason for its existence; that’s what
I mean by a contingent being. You know, as well as I do, that the existence of
neither of us can be explained without reference to something or somebody
outside us, our parents, for example. A necessary being, on the other hand,
means a being that must and cannot not exist. You may say that there is no

analytic, without a possibility of changing. Russell was notoriously sloppy with regard to semantic
differences, which might explain this slip.
Appendix: Text of the Russell-Copleston Debate 459

such being, but you will find it hard to convince me that you do not under-
stand the terms I am using. If you do not understand them, then how can you
be entitled to say that such a being does not exist, if that is what you do say?
20. R: Well, there are points here that I don’t propose to go into at length. I don’t
maintain the meaninglessness of metaphysics in general at all. I maintain the
meaninglessness of certain particular terms—not on any general ground, but
simply because I have not been able to see an interpretation of those particular
terms. It is not a general dogma, it is a particular thing. But those points I
will leave for the moment. And I will say that what you have been saying
brings us back, it seems to me, to the ontological argument—that there is
a being whose essence involves existence, so that his existence is analytic.
That seems to me to be impossible, and it raises, of course, the question what
one means by existence, and as to this, I think a subject named can never be
significantly said to exist but only a subject described; and that existence, in
fact, quite definitely is not a predicate.
21. C: Well, you say, I believe, that it is bad grammar, or rather bad syntax, to say,
for example, T. S. Eliot exists. One ought to say, for example, The author of
“Murder in the Cathedral” exists. Are you going to say that the proposition
The cause of the world exists is without meaning? You may say that the world
has no cause; but I fail to see how you can say that the proposition that The
cause of the world exists is meaningless. Put it in the form of a question: Has
the world a cause? or Does a cause of the world exist? Most people surely
would understand the question, even if they don’t agree about the answer.
22. R: Well, certainly, the question Does the cause of the world exist? is a
question that has meaning. But, if you say: Yes, God is the cause of the
world, where you’re using “God” as a proper name, then God exists will not
be a statement that has meaning. That is the position that I’m maintaining.
Because, therefore, it will follow that it cannot be an analytic proposition,
ever, to say that this or that exists. For example, suppose you take as your
subject “the existent round square”, it would look like an analytic proposition
that the existent round square exists, but it doesn’t exist.
23. C: No, it doesn’t. Then surely you can’t say it doesn’t exist unless you have
a conception of what existence is. As to the phrase, “existent round-square,”
I should say that it has no meaning at all.
24. R: I quite agree. Then I should say the same thing in another context in
reference to a “necessary being”.
25. C: Well, we seem to have arrived at an impasse. To say that a necessary being
is a being that must exist and cannot not exist has for me a definite meaning.
For you it has no meaning.
26. R: Well, we can press the point a little, I think. A being that must exist and
cannot not exist, would surely, according to you, be a being whose essence
involves existence.
27. C: Yes, a being the essence of which is to exist. But I should not be willing
to argue the existence of God simply from the idea of His essence, because
460 Appendix: Text of the Russell-Copleston Debate

I don’t think we have any clear intuition of God’s essence as yet. I think we
have to argue from the world of experience to God.
28. R: Yes, I quite see the distinction. But, at the same time, for a being with
sufficient knowledge, it would be true to say: Here is this being whose essence
involves existence.
29. C: Yes, certainly: if anybody saw God, he would see that God must exist.
30. R: So that, I mean, there is a being whose essence involves existence, although
we don’t know that essence. We only know there is such a being.
31. C: Yes, I should add we don’t know the essence a priori. It is only true a
posteriori, <it is only> through our experience of the world that we come to
a knowledge of the existence of that being. And then one argues, the essence
and existence must be identical. Because if God’s essence and God’s existence
were not identical, then some sufficient reason for this existence would have
to be found beyond God.
32. R: So, it all turns on this question of sufficient reason. And, I must say, you
haven’t defined “sufficient reason” in a way that I can understand. What do
you mean by “sufficient reason”? You don’t mean cause?
33. C: Not necessarily. Cause is a kind of sufficient reason. Only contingent
being can have a cause. God is His own sufficient reason; but He is not cause
of Himself. By sufficient reason, in the full sense, I mean an explanation
adequate for the existence of some particular being.
34. R: But when is an explanation adequate? Suppose I am about to make a flame
with a match. You may say that the adequate explanation of that is that I rub
it on the box.
35. C: Well, for practical purposes; but, theoretically, that’s only a partial expla-
nation. An adequate explanation must ultimately be a total explanation, to
which nothing further can be added.
36. R: Then I can only say you’re looking for something which can’t be got, and
which one ought not to expect to get.
37. C: To say that one has not found it is one thing; to say that one should not
look for it seems to me rather dogmatic.
38. R: Well, I don’t know. I mean, the explanation of one thing is another thing
which makes the other thing dependent on yet another, and you have to grasp
this sorry scheme of things entire to do what you want, and that we can’t do.
39. C: But are you going to say that we can’t, or we shouldn’t, even raise the
question of the existence of the whole of this ‘sorry scheme of things’, of the
whole universe?
40. R: Yes. I don’t think there’s any meaning in it at all. I think the word “universe”
is a handy word in some connections, but I don’t think it stands for anything
that has a meaning.
41. C: If the word is meaningless, it can’t be so very handy. In any case, I don’t say
that the universe is something different from the objects which compose it (I
indicated that in my brief summary of the proof). What I’m doing is to look for
the reason, in this case the cause, of the objects, the real or imagined totality of
Appendix: Text of the Russell-Copleston Debate 461

which constitute what we call the universe. You say, I think, that the universe,
or my existence, if you prefer, or any other existence, is unintelligible?
42. R: First may I take up the point that if a word is meaningless it can’t be
handy. That sounds well but isn’t in fact correct. Take, say, such a word as
“the” or “than”. You can’t point to any object that those words mean, but
they are nevertheless very useful words; I should say the same of “universe”.
But leaving that point, you ask whether I consider that the universe is unin-
telligible. I shouldn’t say unintelligible: I think it is without explanation.
Intelligible, to my mind, is a different thing. Intelligible has to do with the
thing itself, intrinsically, and not with its relations.
43. C: Well, my point is that what we call the world is intrinsically unintelligible,
apart from the existence of God. You see, I don’t believe that the infinity of the
series of events—I mean a horizontal series, so to speak—if such an infinity
could be proved, would be in the slightest degree relevant to the situation. If
you add up chocolates, you get chocolates after all and not a sheep. If you add
up chocolates to infinity, you presumably get an infinite number of chocolates.
So, if you add up contingent beings to infinity, you still get contingent beings,
not a necessary being. An infinite series of contingent beings will be, to my
way of thinking, as unable to cause itself as one contingent being. However,
you say, I think, that it is illegitimate to raise the question of what will explain
the existence of any particular object?
44. R: It’s quite all right if you mean by explaining it simply finding a cause for
it.
45. C: Well, why stop at one particular object? Why shouldn’t one raise the
question of the cause of the existence of all particular objects?
46. R: Because I see no reason to think there is any. The whole concept of cause
is one we derive from our observation of particular things; I see no reason
whatsoever to suppose that the total has any cause whatsoever.
47. C: Well, to say that there isn’t any cause is not the same thing as saying that
we shouldn’t look for a cause. The statement that there isn’t any cause should
come, if it comes at all, at the end of the enquiry, not the beginning. In any
case, if the total has no cause, then to my way of thinking it must be its own
cause, which seems to me impossible. Moreover, the statement that the world
is simply there, if in answer to a question, presupposes that the question has
meaning.
48. R: No, it doesn’t need to be its own cause; what I’m saying is that the concept
of cause is not applicable to the total.
49. C: Then you would agree with Sartre that the universe is what he calls
“gratuitous”?
50. R: Well, the word “gratuitous” suggests that it might be something else. I
should say that the universe is just there, and that’s all.
51. C: Well, I can’t see how you can rule out the legitimacy of asking the question
how the total, or anything at all comes to be there. Why something rather than
nothing? That is the question. The fact that we gain our knowledge of causality
empirically, from particular causes, does not rule out the possibility of asking
462 Appendix: Text of the Russell-Copleston Debate

what the cause of the series is. If the word “cause” were meaningless or if
it could be shown that Kant’s view of the matter were correct, the question
would be illegitimate I agree; but you don’t seem to hold that the word “cause”
is meaningless, and I do not suppose you are a Kantian.
52. R: I can illustrate what seems to me your fallacy. Every man who exists has
a mother, and it seems to me your argument is: Therefore, the human race
must have a mother. But obviously the human race hasn’t a mother: that’s a
different logical sphere.
53. C: Well, I can’t really see a parity. If I were saying Every object has a
phenomenal cause; therefore, the whole series has a phenomenal cause,
there would be a parity. But I’m not saying that; I’m saying, Every object
has a phenomenal cause—if you insist on the infinity of the series—but the
series of phenomenal causes is an insufficient explanation of the series;
therefore, the series has not a phenomenal cause but a transcendent cause.
54. R: Well, that’s always assuming that not only every particular thing in the
world, but the world as a whole must have a cause. For that assumption I see
no ground whatever. If you’ll give me a ground, I’ll listen to it.
55. C: Well, the series of events is either caused or it’s not caused. If it is caused,
there must obviously be a cause outside the series. If it’s not caused, then it’s
sufficient to itself, and, if it’s sufficient to itself, it is what I call necessary.
But it can’t be necessary since each member is contingent, and we’ve agreed
that the total has no reality apart from the members; therefore, it can’t be
necessary.  Therefore, it can’t be uncaused. Therefore, it must have a cause.
 And I should like to observe in passing that the statement The world is
simply there and is inexplicable can’t be got out of logical analysis.
56. R: I don’t want to seem arrogant, but it does seem to me that I can conceive
things that you say the human mind can’t conceive. As for things not having
a cause, the physicists assure us that individual quantum transitions in atoms
have no cause.
57. C: Well, I wonder now whether that isn’t simply a temporary inference.
58. R: It may be, but it does show that physicists’ minds can conceive it.
59. C: Yes, I agree, some scientists—physicists—are willing to allow for indeter-
mination within a restricted field. But very many scientists are not so willing. I
think that Professor Dingle, of London University, maintains that the Heisen-
berg uncertainty principle tells us something about the success (or the lack
of it) of the present atomic theory in correlating observations, but not about
nature in itself, and many physicists would accept this view. In any case, I
don’t see how physicists can fail to accept the theory in practice, even if they
don’t do so in theory. I cannot see how science could be conducted on any
other assumption than that of order and intelligibility in nature. The physicist
presupposes, at least tacitly, that there is some sense in investigating nature
and looking for the causes of events, just as the detective presupposes that
there is some sense in looking for the cause of a murder. The metaphysician
assumes that there is sense in looking for the reason or cause of phenomena,
and, not being a Kantian, I consider that the metaphysician is as justified in
Appendix: Text of the Russell-Copleston Debate 463

his assumption as the physicist. When Sartre, for example, says the world is
gratuitous, I think that he has not sufficiently considered what is implied by
“gratuitous”.
60. R: I think there seems to me a certain unwarrantable extension here. A physi-
cist looks for causes; that does not necessarily imply that there are causes
everywhere. A man may look for gold without assuming that there is gold
everywhere; if he finds gold, well and good; if he doesn’t, he’s had bad luck.
The same is true when the physicists look for causes. As for Sartre, I don’t
profess to know what he means, and I shouldn’t like to be thought to interpret
him; but for my part, I do think the notion of the world having an explanation
is a mistake. I don’t see why one should expect it to have <one>,  and I think
what you say about what the scientist assumes is an over-statement.
61. C: Well, it seems to me that the scientist does make some such assumption.
When he experiments to find out some particular truth, behind that experiment
lies the assumption that the universe is not simply discontinuous. There is
the possibility of finding out a truth by experiment. The experiment may be
a bad one, it may lead to no result, or not to the result that he wants, but at
any rate there is the possibility, through experiment, of finding out the truth
that he assumes. And that seems to me to assume an ordered and intelligible
universe.
62. R: I think you’re generalizing more than is necessary. Undoubtedly the scien-
tist assumes that this sort of thing is likely to be found and will often be found.
He does not assume that it will be found, and that’s a very important matter
in modern physics.
63. C: Well, I think he does assume or is bound to assume it tacitly, in practice.
It may be that, to quote Professor Haldane, when I light the gas under the
kettle, some of the water molecules will fly off as vapor, and there is no way
of finding out which will do so, but it doesn’t follow necessarily that the idea
of chance must be introduced except in relation to our knowledge.
64. R: No, it doesn’t—at least if I may believe what he says. He’s finding out
quite a lot of things… the scientist is finding out quite a lot of things that
are happening in the world, which are, at first, beginnings of causal chains:
first causes which haven’t in themselves got causes. He does not assume that
everything has a cause.
65. C: Surely that’s a first cause within a certain selected field. It’s a relatively
first cause.
66. R: I don’t think he’d say so. If there’s a world in which most events, but not all,
have causes, he will then be able to depict the probabilities and uncertainties
by assuming that this particular event you’re interested in probably has a
cause. And since in any case you won’t get more than probability that’s good
enough.
67. C: It may be that the scientist doesn’t hope to obtain more than probability,
but in raising the question he assumes that the question of explanation has a
meaning. But your general point then, Lord Russell, is that it’s illegitimate
even to ask the question of the cause of the world?
464 Appendix: Text of the Russell-Copleston Debate

68. R: Yes, that’s my position.

69. C: Well, if it’s a question that for you has no meaning, it’s of course very
difficult to discuss it, isn’t it?
70. R: Yes, it is very difficult. What do you say—shall we pass on to some other
issue?
71. C: Let’s. Well, perhaps I might say a word about religious experience, and
then we can go on to moral experience. I don’t regard religious experience as a
strict proof of the existence of God, so the character of the discussion changes
somewhat, but I think it’s true to say that the best explanation of it is the
existence of God. By religious experience I don’t mean simply feeling good. I
mean a loving, but unclear, awareness of some object which irresistibly seems
to the experiencer as something transcending the self, something transcending
all the normal objects of experience, something which cannot be pictured
or conceptualized, but of the reality of which doubt is impossible—at least
during the experience. I should claim that cannot be explained adequately,
and without residue, simply subjectively. The actual basic experience at any
rate is most easily explained on the hypothesis that there is actually some
objective cause of that experience.
72. R: I should reply to that line of argument that the whole argument from our
own mental states to something outside us, is a very tricky affair. Even where
we all admit its validity, we only feel justified in doing so, I think, because of
the consensus of mankind. If there’s a crowd in a room and there’s a clock in a
room, they can all see the clock. The fact that they can all see it tends to make
them think that it’s not an hallucination: whereas these religious experiences
do tend to be very private.
73. C: Yes, they do. I’m speaking strictly of mystical experience proper, and I
certainly don’t include, by the way, what are called visions. I mean simply the
experience, and I quite admit it’s indefinable, of the transcendent object or of
what seems to be a transcendent object. I remember Julian Huxley in some
lecture saying that religious experience, or mystical experience, is as much a
real experience as falling in love or appreciating poetry and art. Well, I believe
that, when we appreciate poetry and art, we appreciate definite poems or a
definite work of art. If we fall in love, well, we fall in love with somebody
and not with nobody.
74. R: May I interrupt for a moment here. That is by no means always the case.
Japanese novelists never consider that they have achieved a success unless
large numbers of real people commit suicide for love of the imaginary heroine.
75. C: Well, I must take your word for these goings on in Japan. I haven’t
committed suicide, I’m glad to say, but I have been strongly influenced in the
taking of two important steps in my life by two biographies. However, I must
say I see little resemblance between the real influence of those books on me
and the mystical experience proper, so far, that is, as an outsider can obtain
an idea of that experience.
76. R: Well, I mean we wouldn’t regard God as being on the same level as the
characters in a work of fiction. You’ll admit there’s a distinction here?
Appendix: Text of the Russell-Copleston Debate 465

77. C: I certainly should. But what I’d say is that the best explanation seems to be
the not purely subjectivist explanation. Of course, a subjectivist explanation is
possible in the case of certain people in whom there is little relation between
the experience and life, in the case of deluded people and hallucinated people,
and so on. But when you get what one might call the pure type, say St. Francis
of Assisi, when you get an experience that results in an overflow of dynamic
and creative love, the best explanation of that, it seems to me, is the actual
existence of an objective cause of the experience.
78. R: Well, I’m not contending in a dogmatic way that there is not a God. What
I’m contending is that we don’t know that there is. I can only take what is
recorded as I should take other records, and I do find that a very great many
things are reported—and I am sure you would not accept things about demons
and devils and what not—and they’re reported in exactly the same tone of
voice and with exactly the same conviction. And the mystic, if his vision is
veridical, may be said to know that there are devils. But I don’t know that
there are.
79. C: But surely in the case of the devils there have been people speaking mainly
of visions, appearance, angels or demons and so on. I should rule out the visual
appearances, because I think they can be explained apart from the existence
of the object which is supposed to be seen.
80. R: But don’t you think there are abundant recorded cases of people who
believe that they’ve heard Satan speaking to them in their hearts, in just
the same way as the mystics assert God? And I’m not talking now of an
external vision, I’m talking of a purely mental experience. That seems to be
an experience of the same sort as mystics’ experience of God, and I don’t see
that from what mystics tell us you can get any argument for God which is not
equally an argument for Satan.
81. C: I quite agree, of course, that people have imagined or thought they have
heard of seen Satan. And I have no wish in passing to deny the existence of
Satan. But I do not think that people have claimed to have experienced Satan
in the precise way in which mystics claim to have experienced God. Take
the case of a non-Christian, Plotinus. He admits the experience is something
inexpressible, the object is an object of love, and therefore, not an object that
causes horror and disgust. And the effect of that experience is, I should say,
borne out, or I mean the validity of the experience is borne out in the records
of the life of Plotinus. At any rate it is more reasonable to suppose that he had
that experience, if we’re willing to accept Porphyry’s account of Plotinus’
general kindness and benevolence.
82. R: The fact that a belief has a good moral effect upon a man is no evidence
whatsoever in favour of its truth.
83. C: No, but if it could actually be proved that the belief was actually responsible
for a good effect on a man’s life, I should consider it a presumption in favour
of some truth, at any rate, of the positive part of the belief, not of its entire
validity. But in any case, I am using the character of the life as evidence in
466 Appendix: Text of the Russell-Copleston Debate

favour of the mystic’s veracity and sanity rather than as a proof of the truth
of his beliefs.
84. R: But even that I don’t think is any evidence. I’ve had experiences myself
that have altered my character profoundly. And I thought, at the time at any
rate, that it was altered for the good. Those experiences were important, but
they did not involve the existence of something outside me, and I don’t think
that if I’d thought they did, the fact that they had a wholesome effect would
have been any evidence that I was right.
85. C: No, but I think that the good effect would attest your veracity in describing
your experience. Please remember that I’m not saying that a mystic’s medi-
tation or interpretation of his experience should be immune from discussion
or criticism.
86. R: Obviously the character of a young man may be—and often is—
immensely affected for good by reading about some great man in history,
and it may happen that the great man is a myth and doesn’t exist, but the
boy is just as much affected for good as if he did. There have been such
people. Plutarch’s Lives—take Lycurgus as an example, who certainly did
not exist, but you might be very much influenced by reading <Plutarch’s life
of> Lycurgus under the impression that he had previously existed. You would
then be influenced by an object that you’d loved, but it wouldn’t be an existing
object.
87. C: I agree with you on that, of course—that a man may be influenced by a
character in fiction. Without going into the question of what it is precisely
that influences him (I should say a real value), I think that the situation of
that man and <the situation of> of the mystic are different. After all, the man
who is influenced by Lycurgus hasn’t got the irresistible impression that he’s
experienced in some way the ultimate reality.
88. R: I don’t think you’ve quite got my point about these historical charac-
ters—these unhistorical characters in history. I’m not assuming what you
call an effect on the reason. I’m assuming that the young man reading about
this person, and believing him to be real, loves him—which is quite easy to
happen, and yet he’s loving a phantom.
89. C: In one sense he’s loving a phantom; that’s perfectly true: in the sense, I
mean, that he’s loving X or Y who doesn’t exist. But at the same time, it is
not, I think, the phantom as such that the young man loves; he perceives a
real value, an idea, which he recognizes as objectively valid, and that’s what
excites his love.
90. R: Well, in the same sense we had before about the characters in fiction.
91. C: Yes, in one sense the man’s loving a phantom—perfectly true. But, in
another sense, he’s loving what he perceives to be a value.
92. R: But aren’t you now saying in effect: I mean by God whatever is good or
the sum total of what is good—the system of what is good, and, therefore,
when a young man loves anything that is good, he is loving God. Is that what
you’re saying? Because, if so, it wants a bit of arguing.
Appendix: Text of the Russell-Copleston Debate 467

93. C: I don’t say, of course, that God is the sum-total or system of what is good
in the pantheistic sense. I’m not a pantheist. But I do think that all goodness
reflects God in some way and proceeds from Him, so that in a sense the man
who loves what is truly good, loves God even if he doesn’t advert to God.
But still, I agree that the validity of such an interpretation of a man’s conduct
depends on the recognition of God’s existence, obviously.
94. R: Yes, but that’s a point to be proved.
95. C: Quite so, but I regard the metaphysical argument as probative, but there
we differ.
96. R: You see, I feel that some things are good and that other things are bad. I
love the things that are good, that I think are good, and I hate the things that I
think are bad. I don’t say that these things are good because they participate
in the Divine goodness.
97. C: Yes, but what’s your justification for distinguishing between good and bad
or how do you view the distinction between them?
98. R: I don’t have any justification any more than I have when I distinguish
between blue and yellow. What is my justification for distinguishing between
blue and yellow? I can see they are different.
99. C: Well, that is an excellent justification, I agree. You distinguish blue and
yellow by seeing them, so you distinguish good and bad by what faculty?
100. R: By my feelings.
101. C: By your feelings. Well, that’s what I was asking. You think that good and
evil have reference simply to feeling?
102. R: Well, why does one type of object look yellow and another look blue? I
can more or less give an answer to that, thanks to the physicists, and as to why
I think one sort of thing good and another evil, probably there is an answer
of the same sort, but it hasn’t been gone into in the same way and I couldn’t
give it you.
103. C: Well, let’s take the behaviour of the Commandant of Belsen. That appears
to you as undesirable and evil and to me too. To Adolf Hitler we suppose
it appeared as something good and desirable. I suppose you’d have to admit
that for Hitler it was good and for you it is evil.
104. R: No, I shouldn’t quite go so far as that. I mean, I think people can make
mistakes in that, as they can in other things. If you have jaundice, you see
things yellow that are not yellow. You’re making a mistake.
105. C: Yes, one can make mistakes, but can you make a mistake if it’s simply
a question of reference to a feeling or emotion? Surely Hitler would be the
only possible judge of what appealed to his emotions.
106. R: It would be quite right to say that it appealed to his emotions, but you can
say various things about that, among others that, if that sort of thing makes
that sort of appeal to Hitler’s emotions, then Hitler makes quite a different
appeal to my emotions.
107. C: Granted. But there’s no objective criterion outside feeling, then, for
condemning the conduct of the Commandant of Belsen, in your view?
468 Appendix: Text of the Russell-Copleston Debate

108. R: No more than there is for the colour-blind person who’s in exactly the
same state. Why do we intellectually condemn the colour-blind man? Isn’t it
because he’s in the minority?
109. C: I would say because he is lacking in a thing which normally belongs to
human nature.
110. R: Yes, but if he were in the majority, we shouldn’t say that.
111. C: Then you’d say that there’s no criterion outside feeling that will enable
one to distinguish between the behaviour of the Commandant of Belsen and
the behaviour, say, of Sir Stafford Cripps or the Archbishop of Canterbury.
112. R: The feeling is a little too simplified. You’ve got to take account of the
effects of actions and your feelings toward those effects. You see, you can
have an argument about it if you can say that certain sorts of occurrences are
the sort you like and certain others the sort you don’t like. Then you have to
take account of the effects of actions. You can very well say that the effects
of the actions of the Commandant of Belsen were painful and unpleasant.
113. C: They certainly were, I agree, very painful and unpleasant to all the people
in the camp.
114. R: Yes, but not only to the people in the camp, but to outsiders contemplating
them also.
115. C: Yes, quite true in imagination. But that’s my point. I don’t approve of
them, and I know you don’t approve of them, but I don’t see what ground
you have for not approving of them, because, after all, to the Commandant
of Belsen himself they’re pleasant, those actions.
116. R: Yes, but you see I don’t need any more ground in that case than I do in
the case of colour perception. There are some people who think everything is
yellow, there are people suffering from jaundice, and I don’t agree with these
people. I can’t prove that the things are not yellow, there isn’t any proof. But
most people agree with me that they’re not yellow, and most people agree
with me that the Commandant of Belsen was making mistakes.
117. C: Well, do you accept any moral obligation?
118. R: Well, I should have to answer at considerable length to answer that. Prac-
tically speaking—yes. Theoretically speaking, I should have to define moral
obligation rather carefully.
119. C: Well, do you think that the word “ought” simply has an emotional
connotation?
120. R: No, I don’t think that, because you see, as I was saying a moment ago,
one has to take account of the effects, and I think right conduct is that which
would probably produce the greatest possible balance in intrinsic value of all
the acts possible in the circumstances, and you’ve got to take account of the
probable effects of your action in considering what is right.
121. C: Well, I brought in moral obligation because I think that one can approach
the question of God’s existence in that way. The vast majority of the human
race will make, and always have made, some distinction between right and
wrong. The vast majority, I think, has some consciousness of an obligation
in the moral sphere. It’s my opinion that the perception of values and the
Appendix: Text of the Russell-Copleston Debate 469

consciousness of moral law and obligation are best explained through the
hypothesis of a transcendent ground of value and of an author of the moral
law. I do mean by “author of the moral law” an arbitrary author of the moral
law. I think, in fact, that those modern atheists who have argued in a converse
way There is no God; therefore, there are no absolute values and no absolute
law are quite logical.
122. R: I don’t like the word “absolute”. I don’t think there is anything absolute
whatever. The moral law, for example, is always changing. At one period in
the development of the human race, almost everybody thought cannibalism
was a duty.
123. C: Well, I don’t see that differences in particular moral judgments are any
conclusive argument against the universality of the moral law. Let’s assume
for the moment that there are absolute moral values. Even on that hypothesis
it’s only to be expected that different individuals and different groups should
enjoy varying degrees of insight into those values.
124. R: I’m inclined to think that “ought”—the feeling that one has about
“ought”—is an echo of what has been told one by one’s parents or one’s
nurses.
125. C: Well, I wonder if you can explain away the idea of the Ought merely in
terms of nurses and parents. I really don’t see how it can be conveyed to
anybody in other terms than itself. It seems to be that, if there is a moral order
bearing upon the human conscience, that that moral order is unintelligible
apart from the existence of God.
126. R: Then you have to say one or other of two things. Either God only speaks to
a very small percentage of mankind—which happens to include yourself—or
He deliberately says things that are not true in talking to the consciences of
savages.
127. C: Well, you see, I’m not suggesting that God actually dictates moral precepts
to the conscience. The human being’s ideas of the content of the moral law
depends entirely to a large extent on education and environment, and a man
has to use his reason in assessing the validity of the actual moral ideas of
his social group. But the possibility of criticizing the accepted moral code
presupposes that there is an objective standard, that there is an ideal moral
order, which imposes itself, I mean the obligatory character of which can be
recognized. I think that the recognition of this ideal moral order is part of the
recognition of contingency. It implies the existence of a real foundation—of
God.
128. R: But the lawgiver has always been, it seems to me, one’s parents or someone
like <them> . There are plenty of terrestrial lawgivers to account for it, and
that would explain why people’s consciences are so amazingly different in
different times and places.
129. C: It helps to explain differences in the perception of particular moral values,
which otherwise are inexplicable. It will help to explain changes in the matter
of the moral law, in the content of the precepts as accepted by this or that
nation, or this or that individual. But the form of it, what Kant calls the
470 Appendix: Text of the Russell-Copleston Debate

categorical imperative, the Ought, I really don’t see how that can possibly
be conveyed to anybody by nurse or parent because there aren’t any possible
terms, so far as I can see, with which it can be explained. It can’t be defined
in other terms than itself, because, once you’ve defined it in other terms than
itself, you’ve explained it away. It’s no longer a moral Ought. It’s something
else.
130. R: Well, I think the sense of “ought” is the effect of somebody’s imag-
ined disapproval, it may be God’s imagined disapproval, but it’s somebody’s
imagined disapproval. And I think that is what is meant by “ought”.
131. C: It seems to me to be external customs and taboos and things of that
sort which can most easily be explained simply through environment and
education; but all that seems to me to belong to what I call the matter of the
law, the content. The idea of the Ought, as such, can never be conveyed to a
man by the tribal chief or by anybody else, because there are no other terms
in which it could be conveyed. It seems to me entirely…
132. R: But I don’t see any reason to say that—I mean we all know about condi-
tioned reflexes. We know that an animal, if punished habitually for a certain
sort of act, after a time will refrain. I don’t think the animal refrains <by>
arguing within himself Master will be angry if I do this. He has a feeling that
that’s not the thing to do. That’s what we can do with ourselves and nothing
more.
133. C: I see no reason to suppose that an animal has a consciousness or moral
obligation; and we certainly don’t regard an animal as morally responsible
for his acts of disobedience. But a man has a consciousness of obligation
and of moral values. I see no reason to suppose that one could condition
all men as one can condition an animal, and I don’t suppose you’d really
want to do so even if one could. If “behaviourism” were true, there would
be no objective moral distinction between the emperor Nero and St. Francis
of Assisi. I can’t help feeling, Lord Russell, you know, that you regard the
conduct of the Commandant of Belsen as morally reprehensible, and that you
yourself would never under any circumstances act in that way, even if you
thought, or had reason to think, that possibly the balance of the happiness of
the human race might be increased through some people being treated in that
abominable manner.
134. R: No. I wouldn’t imitate the conduct of a mad dog. The fact that I wouldn’t
do it doesn’t really bear on this question we’re discussing.
135. C: No, but if you were making a utilitarian explanation of right and wrong in
terms of consequences, it might be held, and I suppose some of the Nazis of the
better type would have held, that although it’s lamentable to have to act in this
way, yet the balance in the long run leads to greater happiness. I don’t think
you’d say that, would you? I think you’d say that sort of action is wrong—
and in itself, quite apart from whether the general balance of happiness is
increased or not. Then, if you’re prepared to say that, then I think you must
have some criterion of right and wrong, that is outside of feeling, at any rate.
Appendix: Text of the Russell-Copleston Debate 471

To me, that admission would ultimately result in the admission of an ultimate

ground of value in God.3
136. R: I think we are perhaps getting into confusion. It is not direct feeling about
the act by which I should judge, but rather a feeling as to the effects. And I
can’t admit any circumstances in which certain kinds of behaviour, such as
you have been discussing, would do good. I can’t imagine circumstances in
which they would have a beneficial effect. I think the persons who think they
do are deceiving themselves. But if there were circumstances in which they
would have a beneficial effect, then I might be obliged, however reluctantly,
to say Well, I don’t like these things, but I will acquiesce in them, just as I
acquiesce in the Criminal Law, although I profoundly dislike punishment.
137. C: Well, perhaps it’s time I summed up my position. I’ve argued two things.
First, that the existence of God can be philosophically proved by a metaphys-
ical argument. Secondly, that it is only the existence of God that will make
sense of man’s moral experience and of religious experience. Personally, I
think that your way of accounting for man’s moral judgments leads inevitably
to a contradiction between what your theory demands and your own spon-
taneous judgments. Moreover, your theory explains moral obligation away,
and explaining away is not explanation.
As regards the metaphysical argument, we are apparently in agreement that
what we call the world consists simply of contingent beings. That is, of beings
no one of which can account for its own existence. You say that the series of
events needs no explanation: I say that, if there were no necessary being—
no being which must exist and cannot not-exist—, nothing would exist. The
infinity of the series of contingent beings, even if proved, would be irrelevant.
Something does exist; therefore, there must be something which accounts for
this fact, a being which is outside the series of contingent beings. If you had
admitted this, we could then have discussed whether that being is personal,
good, and so on. On the actual point discussed, whether there is or is not a
necessary being, I find myself, I think, in agreement with the great majority
of classical philosophers.
You maintain, I think, that existing beings are simply there, and that I have no
justification for raising the question of the explanation of their existence. But
I would like to point out that this position cannot be substantiated by logical
analysis; it expresses a philosophy which itself stands in need of proof. I think
we have reached an impasse because our ideas of philosophy are radically
different; it seems to me that, what I call a part of philosophy, that you call
the whole, insofar at least as philosophy is rational.

3 Slater (2011) takes over the reading in Morris (1956): “that admission would eventually result in
the admission of an ultimate source of value in God”, replacing “eventually” for “ultimately” and
“source” for “ground”. The first substitution was done presumably to avoid repetition; but elegance
is not always important in philosophical discussions. The second substitution is utterly unwarranted,
given that Copleston is all the time talking about “grounding” and “foundation”: it is his metaphor,
and we should respect it.
472 Appendix: Text of the Russell-Copleston Debate

It seems to me, if you will pardon my saying so, that, besides your own logical
system, what you call “modern” in opposition to antiquated logic (a tenden-
tious adjective), you maintain a philosophy which cannot be substantiated by
logical analysis. After all, the problem of God’s existence is an existential
problem, whereas logical analysis does not deal directly with problems of
existence. So, it seems to me, to declare that the terms involved in one set
of problems are meaningless because they are not required in dealing with
another set of problems, is to settle from the beginning the nature and extent
of philosophy, and that is itself a philosophical act which stands in need of
justification.
138. R: Well, I should like to say just a few words by way of summary on my side.
First, as to the metaphysical argument: I don’t admit the connotations of such
a term as “contingent” or the possibility of explanation in your sense. I think
the word “contingent” inevitably suggests the possibility of something that
wouldn’t have this, what you might call, accidental character of just being
there, and I don’t think <that> is true except in the purely causal sense. You
can sometimes give a causal explanation of one thing as being the effect of
something else, but that is merely referring one thing to another thing and
there’s no—to my mind—explanation in your sense of anything at all; nor is
there any meaning in calling things “contingent” because there isn’t anything
else they could be.
That’s what I should say about that, but I should like to say a few words about
your accusation that I regard logic as all philosophy. That is by no means
the case. I don’t by any means regard logic as all philosophy. I think logic is
an essential part of philosophy, and logic has to be used in philosophy, and
in that I think you and I are at one. When the logic that you use was new,
namely in the time of Aristotle, there had to be a great deal of fuss made
about it. Aristotle made a lot of fuss about that logic. Nowadays it’s become
old and respectable, and you don’t have to make so much fuss about it. The
logic that I believe in is comparatively new, and therefore I have to imitate
Aristotle in making a fuss about it, but it’s not that I think it’s all philosophy
by any means. I don’t think so. I think it’s an important part of philosophy
and, when I say that I don’t find a meaning for this or that word, that is a
position of detail, based upon what I’ve found out about that particular word,
from thinking about it. It’s not a general position that all words that are used
in metaphysics are nonsense, or anything like that, which I don’t really hold.
As regards the moral argument, I do find that, when one studies anthropology
or history, there are people who think it their duty to perform acts which I
think abominable, and I certainly can’t, therefore, attribute Divine origin to
the matter of moral obligation, which you, Father Copleston don’t ask me
to; but I think even the form of moral obligation, when it takes the form of
enjoining you to eat your father or what not, doesn’t seem to me to be such
a very beautiful and noble thing; and, therefore, I cannot attribute a Divine
origin to this sense of moral obligation, which I think is quite easily accounted
for in quite other ways.