THE APORETIC TRADITION IN ANCIENT

PHILOSOPHY

Ancient philosophers from an otherwise diverse range of traditions

were connected by their shared use of aporia – translated as
‘puzzlement rooted in conflicts of reasons’ – as a core tool in
philosophical enquiry. The essays in this volume provide the first
comprehensive study of aporetic methodology among numerous
major figures and influential schools, including the Presocratics,
Plato, Aristotle, Plutarch, Alexander of Aphrodisias, Academic
sceptics, Pyrrhonian sceptics, Plotinus and Damascius. They explore
the differences and similarities in these philosophers’ approaches to
the source, structure and aim of aporia; their views on its function
and value; and ideas about the proper means of generating such a
state among thinkers who were often otherwise opposed in their
overall philosophical orientation. Discussing issues of method,
dialectic and knowledge, this volume will appeal to those interested
in ancient philosophy and in philosophical enquiry more generally.

GEORGE KARAMANOLIS is Associate Professor of Philosophy at

the University of Vienna. His publications include Plato and Aristotle
in Agreement? Platonists on Aristotle from Antiochus to Porphyry
(2006) and The Philosophy of Early Christianity (2013).

VA S I L I S P O L I T I S is Associate Professor of Philosophy at Trinity

College Dublin. He is the author of The Structure of Enquiry in
Plato’s Early Dialogues (2015) as well as numerous journal articles
on philosophical enquiry.
THE APORETIC T RADI T ION IN
ANCIENT PHILOSOPH Y
Edited by

George Karamanolis
University of Vienna

Vasilis Politis
Trinity College Dublin
 University Printing House, Cambridge CB2 8BS, United Kingdom

One Liberty Plaza, 20th Floor, New York, NY 10006, USA

477 Williamstown Road, Port Melbourne, VIC 3207, Australia

314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre,

New Delhi – 110025, India

79 Anson Road, #06–04/06, Singapore 079906

Cambridge University Press is part of the University of Cambridge.

It furthers the University’s mission by disseminating knowledge in the

pursuit of education, learning, and research at the highest
international levels of excellence.

www.cambridge.org

Information on this title: www.cambridge.org/9781107110151

DOI: 10.1017/9781316274293

© Cambridge University Press 2018

This publication is in copyright. Subject to statutory exception and to the

provisions of relevant collective licensing agreements, no reproduction of
any part may take place without the written permission of Cambridge
University Press.
 First published 2018

Printed in the United Kingdom by Clays, St Ives plc

A catalogue record for this publication is available from the British

Library.

ISBN 978-1-107-11015-1 Hardback

Cambridge University Press has no responsibility for the persistence or

accuracy of URLs for external or third-party internet websites referred to in
this publication and does not guarantee that any content on such websites
is, or will remain, accurate or appropriate.
 Contents
List of Contributors

Introduction
George Karamanolis and Vasilis Politis

1 Contradiction and Aporia in Early Greek Philosophy

John Palmer

2 Socrates and the Benefits of Puzzlement

Jan Szaif

3 Aporia and Sceptical Argument in Plato’s Early Dialogues

Vasilis Politis

4 Aporia in Plato’s Parmenides

Verity Harte

5 Aporia in Plato’s Theaetetus and Sophist

Lesley Brown

6 Aporia and Dialectical Method in Aristotle

Christof Rapp

7 Aporia in Aristotle’s Metaphysics Beta

Friedemann Buddensiek

8 Uses of Aporiai in Aristotle’s Generation of Animals

Jessica Gelber

9 Aporia and the New Academy

James Allen

10 Aporetic Elements in Plutarch’s Philosophy

John Dillon

11 Aporia and Enquiry in Ancient Pyrrhonism

Luca Castagnoli

12 Aporia and Exegesis: Alexander of Aphrodisias

Inna Kupreeva

13 The Aporetic Character of Plotinus’ Philosophy

George Karamanolis

14 Aporia and the Limits of Reason and of Language in

Damascius
Damian Caluori

Bibliography
Index Locorum
Subject Index
 Contributors

JAMES ALLEN is Professor of Philosophy at the University of

Toronto.

LESLEY BROWN is Emeritus Fellow of Philosophy at Oxford

University.

FRIEDEMANN BUDDENSIEK is Professor of Ancient

Philosophy at Goethe University Frankfurt.

DAMIAN CALUORI is Associate Professor of Philosophy at

Trinity University in San Antonio, Texas.

is Associate Professor of Ancient

L U C A C A S TA G N O L I
Philosophy at the University of Oxford and Stavros Niarchos
Foundation Clarendon Fellow in Ancient Greek Philosophy at
Oriel College, Oxford.

JOHN DILLON is Emeritus Regius Professor of Greek at Trinity

College Dublin.

JESSICA GELBER is Assistant Professor of Philosophy at the

University of Pittsburgh.

VERITY HARTE is Professor of Philosophy and Classics at Yale

University.

GEORGE KARAMANOLIS is Associate Professor of Philosophy

at the University of Vienna.

is Senior Lecturer in Philosophy at the

I N N A K U P R E E VA
University of Edinburgh.
J O H N PA L M E R is Professor of Philosophy at the University of
Florida.

is Associate Professor of Philosophy at

VA S I L I S P O L I T I S
Trinity College Dublin.

CHRISTOF RAPP is Professor of Philosophy at Ludwig

Maximilian University Munich.

JAN SZAIF is Professor of Philosophy at the University of

California, Davis.
 Introduction
◈

George Karamanolis and Vasilis Politis

What is an aporetic philosopher? It is hardly a philosopher

distinguished by his or her mental condition of perplexity, puzzlement
or confusion, much as this is what the term aporia signifies in its
principal use in philosophy. We are all now and then subject to such
a condition, but we do not all make use of it as do aporetic
philosophers. Aporia signifies above all a certain state of mind, which
it is as difficult to describe as it is easy to recognise when one is in it.
Being in a state of aporia is a characteristic way of being perplexed,
which philosophers have, from the beginning, used a variety of
metaphors and images to describe. ‘It feels like being tied
(intellectually tied, tied in one’s mind)’, is a famous Aristotelian
image. ‘It is like being without means and without resource
(intellectually, that is)’, is a common metaphor going back to an
original everyday use of the term aporos, to mean “penurious” and
“needy”. ‘It feels like being numb, numb in mind and tongue’, is a
memorable Platonic image. ‘It is a state of speechlessness and
inarticulateness’, is another Platonic description. ‘It feels like an
unstable, vertiginous state in which things won’t stay fixed and are
thoroughly shaken’, or ‘like being tempest-tossed’, are yet other
Platonic metaphors. To conclude with what is perhaps the first
philosophical image of aporia, ‘It is like being unable to reach
through to a much-desired place’. This image goes back to another
original everyday use of the term aporon, to mean ‘un-passable’, ‘un-
traversable’. This image is immortalised by Heraclitus in a very early
philosophical statement regarding aporia: ‘Unless one hopes for that
which is not to be hoped for (anelpiston), one shall not find it (ouk
exeurēsei). For it is hard to search for (anexereunēton) and to reach
through to (aporon)’ (fragment DK18).
Is it distinctive of being an aporetic philosopher that one should
take the generation of this mental condition to be a major part of
doing philosophy? This, we believe, may justly be considered a basic
mark of an aporetic philosopher. It means that an aporetic
philosopher is a person who self-consciously assigns a certain
function and significance to this mental condition, aporia, if the
condition is generated in a peculiar way and through a peculiar
intellectual enquiry – that is, the search for wisdom and, in that
sense, philosophy (philosophia). However much ancient
philosophers may disagree about the character and nature of
philosophy, its means and its ends, they are all, even relativists such
as Protagoras, agreed that philosophy is an enquiry – and a
supremely important enquiry – aspiring, whether confidently or not
so confidently, to a supreme intellectual state.
What is it to generate this state of mind, aporia, in a peculiarly
philosophical way and in general through intellectual enquiry? As the
contributors to this volume will demonstrate, about each and all of
our candidate aporetic philosophers, it is to think that there is
something distinctive which this state of mind, aporia, is about and
by which it is caused; and this object and cause of the state of aporia
is a certain form or forms of question: a question that, in one way or
another, presents an intellectual problem or difficulty. (We shall
presently consider what forms this question may take, according to
our candidate aporetic philosophers.) The question which is the
object and cause of this mental state, aporia, is itself properly called
an aporia. This, as the contributors show, means that there are two
basic, and related, uses of, the term aporia: one to mean the state of
mind, and another to mean the object and cause of this state of
mind. The contributors will commonly distinguish between the two
uses by referring to the first as the subjective use and to the second
as the objective use. We may observe that to speak of an objective
use of the term aporia is to mean that the state of mind, aporia, is
object-directed, its ‘object’ being a certain form or forms of question.
This means that the mental state of aporia, as understood by our
candidate aporetic philosophers, is a cognitive state; cognitive in the
sense of object-directed.
Our mark of an aporetic philosopher has the virtue of being
flexible and allowing for a variety of ways of being aporetically
disposed, depending on a variety of questions. These are questions
that the contributors to this volume will take up in various and diverse
ways. The following is a selection, without a claim to
comprehensiveness or suggestion that each contributor is
addressing but a single question: it is important to bear in mind that
the volume is structured chronologically, not thematically, with each
contributor addressing one or more philosophers and one or more
works.

What is considered the proper intellectual means (singular or plural) of

generating this mental state of aporia?
What is considered the place and the function of aporia in philosophical
enquiry?
Is being in this state of aporia important only for philosophical enquiry,
or also for enquiry in natural science?
Cannot the generation of this state of aporia just as much be used
simply to trip up and confuse people?
What is considered the ethical benefits, or the ethical harms, of being
in a state of aporia?
Is there something especially productive and creative in being in this
state?
What is the relation between aporia-involving argument and dialectical,
disputative, and in general refutative argument?
Is it necessary to go through being in this state of aporia if one aspires
to knowledge?
Is being in this state of aporia reason to question that it is possible to
attain knowledge?
Can the fact, or appearance, that the same thing has opposite qualities
generate aporia? How?
What is the place of hypotheses in a method of aporia-inducing
argument?
Can one be in this state of aporia about things that are familiar to us
and that we take for granted in how we speak and think and act?
Is there a preferred means of getting out of the state of aporia?
If aporia is a troubled state to be in, is it necessary to get out of this
state to attain intellectual tranquility, or is being in aporia compatible with
tranquility?
Is the commitment to aporia, as a method of philosophical enquiry,
compatible with a commitment to systematic philosophy and speculative
 theory?
May this state of aporia indicate that there are limits not only to what
we can know but also to what we can think of and speak of?

The aim of the present collection of essays is to trace a continuous

aporetic tradition through a millennium of philosophy in antiquity,
from Heraclitus and Zeno, through Plato and Aristotle, and up to
Plotinus and Damascius, and to examine different and potentially
opposed ways of thinking that aporia occupies a major place in
philosophical enquiry. The volume explores potentially shared
commitments – relating especially to the source, the structure, and
the aim of philosophical enquiry – of philosophers who may
otherwise be wide apart in temper and convictions.
The topic of the present volume – the place of aporia in ancient
philosophy – is quite novel, or as novel as a topic in ancient
philosophy can be. Very useful groundwork has been done on the
meaning and the uses of the term aporia and its cognates, up to
Aristotle.1 And considerable work has been done on Aristotle’s
Metaphysics Beta, in which Aristotle gives a central place to aporiai
and aporia-based argument.2 There has not been an attempt to
trace the function that ancient philosophers assign to the state of
aporia, or the proper means of generating this state, in philosophical
or in scientific3 enquiry.4 Some of the lacunae that mark the scant
and uneven attention to the topic to date are surprising, such as the
lack of an investigation of the role of aporia in Pyrrhonian, or indeed
in Academic, scepticism.5
 Such attention as the topic of the present volume has received
is marked by some notable and questionable assumptions. First,
there is a tendency to understand aporia-involving argument in Plato
exclusively in terms of elenctic and refutative argument. But the
equation of aporia-involving argument with elenctic or refutative
argument is questionable – in Plato, in Aristotle, and in general.6
Secondly, this tendency has, it appears, stood in the way of
recognising what it is that the New or Sceptical Academy was
picking up on in Plato; namely, that argument based on an aporia
can take the form of a two-sided question with apparently good
reasons on both sides.7 Thirdly, there is a general and deep-set
tendency to assume that an aporetic philosopher cannot at the same
time be committed to systematic views and speculative theories:
hence that we need to choose between thinking of a philosopher,
such as Plato or Aristotle or Plotinus, as aporetic and thinking of
them as ‘dogmatic’ (i.e. committed to dogmata, positive beliefs). But
this assumption is open to question.8
 On the Mark of Aporetic Philosophy
If this is our basic mark of an aporetic philosopher – one who
considers the generation of the state of aporia to be a major part of
philosophical activity – then being an aporetic philosopher is
compatible not only with being committed to the search for
knowledge – this much would hardly exclude any ancient
philosopher save for such extreme eristics and contradiction-
mongers as those parodied by Plato in the Euthydemus9 – but also
with believing that it is possible to attain knowledge and even that
one may have attained knowledge. Does this not render our mark of
aporetic philosophy objectionably broad? We need a broad mark, if
we are to look for a common dimension, tendency or thrust –
something like a single, continuous tradition – among such diverse
philosophers as, on the one hand, Aristotle, Plotinus and Plato
(supposing that there may be doctrines present in Plato’s dialogues),
and, on the other hand, Pyrrhonian sceptics. Is such a broad mark a
drawback and a fault? Consider the obvious alternative, which is to
propose that an aporetic philosopher is one who considers the
generation of aporia to be, not simply a major part of, but the
principal and ultimate aim of philosophical activity. This would
exclude everyone except sceptics. It is, we think, objectionably
narrow. It stands in the way of recognising that whereas Plato, or
Aristotle in the Metaphysics, and Pyrrhonian sceptics end up in very
different places regarding the attainability and attainment of
knowledge and the desirability of speculative theory and systematic
philosophy, they share basic commitments regarding the source and
the structure, if not the aim, of philosophical enquiry.10 The point is
that sharing in a single method, namely, aporia-based enquiry, and a
single aim, namely, the attainment of knowledge through the
resolution of aporiai, is compatible with different and even opposite
outcomes, depending on whether or not the philosopher in question
thinks there is a general reason to doubt that such aporiai are
capable of being resolved.
If it did not exclude any philosophers, our mark of aporetic
philosophy would be objectionable. How can a philosopher positively
deny that the generation of the state of aporia is a major part of
philosophical activity? Consider a philosopher who is committed not
only to the search for knowledge and the attainability of knowledge
but also to the existence of a criterion for knowledge. A criterion of
knowledge is a cognitive experience and impression that is such as
to guarantee its truth and to exclude the possibility of the subject of
this experience being mistaken. Such a philosopher can admit that
ridding oneself of a state of aporia, and, consequently, engaging as
far as is necessary with the difficulties and problems that are
responsible for such a state, is part of the preparation for the search
for knowledge: a search which will be involved, rather, with the
exercise of a supposed natural faculty for infallible knowledge. He or
she will have to deny that the generation of aporia is part of the
search for knowledge proper, since, as he or she believes, the
search for knowledge does not stand in a substantial relation to the
state of aporia and indeed has the means of bypassing it. Among the
ancients, the Stoics appear to have thought of knowledge along
these lines, and they, at any rate, are not on any account justly
characterised as aporetic philosophers.11 Indeed, their stance on this
matter may well have been a deliberate reaction against Plato and
Aristotle.
 What Is the Proper Intellectual Means of
Generating Aporia?
Aristotle is perhaps the first to have reflected on the meaning of the
term aporia as this term is deliberately used by philosophers, when
he says that it designates in the first instance a certain mental state
but that philosophers – including Aristotle – use it also for that which
generates the mental state (see Topics VI. 6, 145b16–20; he has just
characterised an aporia as being ‘an equality [i.e. equality in
strength, or apparent strength] of opposite reasonings’, 145b1–2).12
Aristotle appears to be right in this observation, which, as we have
noted, identifies an important feature of the use of the term aporia;
namely, that the object to which this state of mind, aporia, is directed
is likewise properly called an aporia. This is an aporia in the sense of
a certain form, or certain forms, of question, and a question that
presents an intellectual difficulty or problem. An early instance of this
twofold use of the term aporia is clearly recognisable in Plato, when,
in the Protagoras (324d–e), and twice in quick succession, he uses
the phrase hē aporia hēn su aporeis, in the sense of ‘the problem (or
“puzzle”, or “difficulty” or “question”) about which you are in a state of
puzzlement’ (or, ‘which you are puzzling over’).
What, according to the philosophers taken up in the present
volume, is the proper intellectual means of generating the mental
state of aporia? And is there in these philosophers, according to the
findings of the contributors, a single means of generating the state of
aporia? On any account, determining what is the proper means of
generating an aporia requires establishing what those objects are at
which this state of mind, aporia, is directed. If we follow Aristotle, we
will expect an affirmative answer, for he supposes a single way of
generating the state of aporia, that is, by means of what he calls ‘an
equality of opposite reasonings’. He means the advancing of
competing apparently good reasons on both sides of a two-sided,
whether-or-not question. (He calls such a question a problēma.)
Aristotle’s answer is very important, and it contains two related
elements, which, we think, deserve to be considered separately.
First, there is the reference simply to contradiction, and so the very
notion of contradiction plays a role. Secondly, there is the reference
to compelling contradiction, that is, a contradiction both sides of
which are supported by apparently good reasons.
It is remarkable that, according to the contributors to the present
volume, this account is largely correct, though not without
exceptions. This means that there is among these philosophers, and
spanning a millennium of philosophy in antiquity, very considerable
agreement about the proper means of generating aporia in
philosophy. Generally, though not exclusively, aporia is generated
either through generating a contradiction or through generating a
compelling contradiction. This finding – for it appears to be a major
finding of the present project – is as important as it is remarkable. It
also allays a worry that one may have as to whether the topic of the
present volume – the place of aporia in ancient philosophy – may not
be too broad and diffuse to sustain investigation. As Harte (72)
observes, ‘Were we to identify anything capable of inducing aporia in
the form of an intellectual condition as an aporia in the sense of a
puzzle, the term aporia understood as puzzle would, I submit,
become so broad in its compass as to become uninteresting.’13
We should note, finally, a question regarding aporia that will
emerge as prominent in, and perhaps controversial among, the
contributions. It is the question of the association, especially since
Aristotle and the method of dialectical argument of the Topics, of
aporia-based argument with dialectical argument. Dialectical
argument, as it is understood by Aristotle in the Topics, is a method
of argument that, by using endoxa, that is, opinions that are credible
owing to the number or the expertise of those who hold them,
aspires to the power of disputing, against an adversary, equally and
indifferently on either side of any two-sided question. The critical
question is this: How close should we understand this association, of
aporia-based argument with dialectical argument, to be? The closer
we think it is, the more readily we shall associate aporia-based
argument with elenctic and refutative argument; because, clearly,
dialectical argument is a form of refutative argument. And we have
noted that the association of aporia-based argument with refutative
argument is questionable if taken too far. Whereas dialectical
argument is adversarial and involves competing persons or parties,
aporia-based argument need not be adversarial; it can be
cooperative. Aporia-based argument can be conducted by a single
person by himself or herself.14 And, most important, whereas the
basic reliance on endoxa may be proper in the case of dialectical
and disputative argument, it may be questioned in the case of
aporia-based argument, especially if such argument is thought to be
an essential part of the search for knowledge.15
1 See Motte and Rutten 2001; also Erler 1987.

2 See especially Madigan’s 1999 commentary and the Symposium

Aristotelicum collection (Crubellier and Laks 2009); and, of course,
Aubenque’s 1961b paper and in his 1966 classic. See also Politis 2002 and
Politis and Su 2017.

3 Gelber’s contribution to this volume (Chapter 8) addresses the role of

aporia in Aristotle’s biological works.

4 With the possible exception of Matthews 1999a; which, however, is limited

in its compass and largely intended for a general audience.

5 The contributions to this volume of Allen (Chapter 9) and of Castagnoli

(Chapter 11) redress this omission. Castagnoli (206) argues that the
omission is in large part responsible for the fact that ‘Since antiquity
interpreters have described Pyrrhonism as a philosophical approach
somehow “alien”, radically different in its motivations, nature, and conception
of the philosophical endeavour from all other ancient philosophies, those
philosophies that the Pyrrhonists lumped together as “dogmatic”.’ By
attending to the role that the Pyrrhonists assign to aporia in philosophical
enquiry, he argues for a very different account, which brings closer to each
other the Pyrrhonists and Plato and Aristotle. Opsomer (1998) is a clear
exception to our claim of a lack of an investigation of aporia in Academic
scepticism.

6 Politis’ contribution to this volume (Chapter 3) argues against this tendency

(see also Politis 2006, 2008, 2012b and 2015); as does Castagnoli, who
argues (especially contra Woodruff 1988 and 2010), against any
‘straightforward equation of “aporetic” and “refutative”’ (214–15). This, of
course, is not to deny that elenctic or refutative argument can result in
aporia and so be aporetic; it is to deny that all aporia-involving argument is
elenctic or refutative. Szaif’s contribution (Chapter 2) is instructive on this
point.

7 See the contributions of Allen, Castagnoli and Politis (Chapters 9, 11 and

3, respectively).
8 For this issue, and different views on it, see the contributions of Allen,
Dillon, Karamanolis, Castagnoli and Politis (Chapters 9, 10, 13, 11 and 3,
respectively).

9 For the distinction between aporetic and eristic philosophy, see Szaif’s
contribution (Chapter 2).

10 For this point, see Castagnoli’s contribution (Chapter 11).

11 See Allen’s contribution (Chapter 9).

12 See Rapp’s contribution to this volume (Chapter 6).

13 See also Rapp’s contribution, when he urges a need to ‘push back

against the tendency to view all instances of questions, queries, enquiries,
procedural remarks as somehow qualifying as aporiai, leading to an excess
of aporiai in Aristotle’s work’ (Chapter 6, 112).

14 With regard to Plato and the early dialogues, Politis (2015: esp. chs. 5–6)
has argued that it is essential to aporia-based argument that the reasons on
either side of a two-sided question should appear good not only to different
people but also to one and the same person.

15 For helpful comments on this Introduction, we are grateful to Lesley

Brown, Kate Kiernan, Peter Larsen, Pauline Sabrier and Jun Su. We are
grateful also to the anonymous CUP reader. We are extremely grateful to
Peter Larsen for assisting with the manuscript and with the indices. Finally,
this volume originates in a conference held in Trinity College Dublin in the
autumn of 2014, and in this regard we would like to thank everyone then
present and to acknowledge our gratitude to: the Department of Philosophy
at Trinity College Dublin, and The Trinity Plato Centre.
 Chapter 1
Contradiction and Aporia in Early
Greek Philosophy
◈

John Palmer

An aporia is, essentially, a point of impasse where there is

puzzlement or perplexity about how to proceed. Aporetic reasoning
is reasoning that leads to this sort of impasse, and an aporia-based
method would be one that centrally employs such reasoning. One
might describe aporia, more basically, as a point where one does not
know how to respond to what is said. In the Platonic dialogues
dubbed ‘aporetic’, for instance, Socrates brings his interlocutors to
the point where they no longer know what to say. Now, it should be
obvious that there was a good deal of aporetic reasoning prior to
Socrates. It should also be obvious that the form of such reasoning
is immaterial so long as it leads to aporia. In particular, the reasoning
that leads to an aporia need not take the form of a dilemma.
Instances of reasoning generating genuine dilemmas – with two
equally unpalatable alternatives presented as exhausting the
possibilities – are actually rather rare in early Greek philosophy.
Moreover, there need not in fact be any reasoning or argumentation
as such to lead an auditor to a point where it is unclear how to
proceed or what to say. Logical paradoxes such as Eubulides’ liar do
not rely on argumentation at all but on the exploitation of certain
logical problems to generate an aporia. All that is required in this
instance is the simple question: ‘Is what a man says true or false
when he says he is lying?’ It is hard to know how to answer this
question because any simple response snares one in contradiction.
Likewise, among the remains of Heraclitus’ book are a number of
provocative statements that induce a certain puzzlement without any
argument as such. Consider, for example, ‘The path up and down is
one and the same’ (Heracl. 22B60 DK). One does not know quite
what to say about this, for the description appears to harbour a
contradiction: there is a path up and a path down, yet there is also a
single path. The apparent contradiction between the multiplicity and
unity of the same object calls for some explanation and resolution.
What is perhaps most important to the generation of aporia is
the production of at least the appearance of contradiction, by one
means or another. The appearance of contradiction is intolerable
because contradiction is itself impossible: the same thing cannot be
at once F and not-F in the same respect. When led to accept
contradictories even though one knows they cannot both be true,
there are two main ways to respond. One can, appreciating the
difficulty of the issue, say nothing, or one can, appreciating the
difficulty of the issue, try to say something. These possibilities
correspond to the two broad purposes the fabricator of the apparent
contradiction may have: he may want the auditor to realise that the
question is so fraught with difficulty that it is best to say nothing, or
he may want the auditor to persist in trying to say something useful
even while appreciating the difficulty of the question. That one can
respond to an aporia, and that someone may intend for one to
respond, in these distinct ways reflects a basic division in the uses
and purposes of aporetic reasoning. In its broadly negative uses, the
aim of aporetic reasoning is aporia. In its broadly positive uses, the
aim of aporetic reasoning is escape from aporia. There will be further
variations within each category. For instance, the negative ends for
which aporetic reasoning may be employed include simple
confutation and, more positively, the promotion of a sceptical
attitude. Also, significantly, when the appearance of contradiction
leaves one puzzled about how to proceed and uncertain as to what
to say, one need not respond as the fabricator of the aporia
intended. In particular, one can be stimulated to make positive
progress by a piece of aporetic reasoning developed for basically
negative purposes.
I want to focus on the uses of contradiction to generate aporia in
Heraclitus, the Eleatics Zeno and Melissus, and the sophists
Protagoras and Gorgias. Heraclitus merits attention here because he
is unique among the early Greek philosophers in inducing aporia
with the positive aim of provoking his audience to a deeper
understanding of the world’s workings. Both Zeno and Melissus, by
contrast, reason in ways designed to contradict common sense and
ordinary experience. Their essentially negative aporetic reasoning
sets the trend for the uses of contradiction among the sophists,
though with Protagoras and Gorgias contradiction and aporetic
reasoning are employed in novel and sophisticated ways.
 Heraclitus
Timon of Phlius called Heraclitus ainiktēs or ‘riddler’ (D.L. 9.6), and
the epithet skoteinos or obscurus commonly attaches to him in the
later tradition. Heraclitus appears intent on provoking his audience to
understanding by making deliberately puzzling or paradoxical
statements. Although he describes himself as ‘distinguishing each
thing according to its nature and telling how it is’, at the same time
he says people generally fail to understand the logos, this being both
his own discourse and the principle of the natural order it describes
(22B1 DK, cf. 22B19, 22B34). So there is some justification for the
tradition’s view that Heraclitus was obscure, though he would have
said the apparent obscurity of his writings simply mirrors the evident
obscurity of things. ‘Nature likes to hide’ (22B123 DK), he says. His
attitude towards the general level of human understanding is like
Socrates’ without the irony: ‘The multitude do not understand the sort
of things they encounter, nor do they know by learning, though they
seem to themselves to do so’, he says (22B17 DK, cf. 22B28a,
22B40, 22B57, 22B104). Although Heraclitus differs from Socrates in
professing to know the kinds of things most people only think they
know (cf. 22B41 DK, 22B50 etc.), they both in a general way seek to
dispel their auditors’ false conceit of wisdom. Socrates exposes
latent contradictions among his interlocutors’ beliefs so that they
might abandon their misplaced confidence regarding their
understanding of ethical matters. Heraclitus provokes his audience
to deeper understanding of the world’s workings with declarations
that induce puzzlement in a variety of ways.
Sometimes, as with the river fragment (22B12 DK), he employs
a striking image to serve as one term in an unspecified analogy that
leaves one puzzling over what the image is supposed to convey. The
road fragment (22B60 DK) bears witness to Heraclitus’ penchant for
inducing puzzlement with a statement that appears to harbour a
contradiction. Other fragments show this to be one of his preferred
devices: ‘They do not understand how drawn apart it is brought
together with itself: a back-stretched harmony like a bow’s and a
lyre’s’ (22B51 DK).1 ‘Combinations, wholes and not wholes, brought
together drawn apart, concordant discordant, and from all things one
and from one thing all’ (22B10 DK). ‘Sea, water most pure and most
polluted: for fish, drinkable and sustaining, but for humans,
undrinkable and destructive’ (22B61 DK). ‘God, day night, winter
summer, war peace, hunger satiety – he undergoes alteration just as
fire, when mixed with spices, is called by each one’s aroma’ (22B204
DK). ‘Invisible harmony is stronger than visible’ (22B207). Heraclitus
employs prima facie contradictions, along with other devices, as a
way of provoking his audience to question their understanding of the
world’s workings. His enigmatic utterances are designed not merely
to lead his audience into impasse. He offers an understanding of the
ultimate principles governing the world’s workings to those who
relinquish the false conceit of their own wisdom. Not all aporiai are
meant to be final. Many if not most of the ancient philosophers who
employed aporetic arguments did so as a way of framing problems.
When contradictions are employed to generate aporiai in this way,
the contradictions are supposed to be only apparent – their
fabricators intend for them to be resolved. Such seems to be the
case with the contradictions employed by Heraclitus. Sometimes he
actually indicates himself how the contradiction he has introduced is
to be resolved. When he says in 22B88 DK that the same thing is
present living and dead, waking and sleeping, young and old, he
then explains how so: ‘for (gar) these once changed are those and
those once changed are these’. The puzzlement does not
necessarily cease with Heraclitus’ explanation. Nevertheless, it
should be clear that he means for the contradictions his discourse
makes manifest to be resolved by the deeper understanding he also
aims to provide.2
 Zeno and Melissus
Zeno of Elea deployed his own arsenal of contradictions to provoke
his audience to question their understanding of how the world works.
The most famous of his ingenious paradoxes purport to show that
motion is impossible by showing that common-sense assumptions
regarding its occurrence lead to problems. For instance, if a tortoise
starts ahead of Achilles in a race, in the time it takes Achilles to get
to where the tortoise started, the tortoise will have moved some
distance ahead. And in the time it takes Achilles to get there, the
tortoise will again have moved some distance ahead. And the
tortoise will always have moved some distance ahead during the
period of time it takes Achilles to get to where it was at the beginning
of that period, so that the tortoise will always be ahead and will never
be overtaken by Achilles. One of the remarkable features of this
argument is the simplicity of its conceptual apparatus, reflected in
this reconstruction and evident in Aristotle’s testimony: ‘Second is
the [argument] called “Achilles”: this is that the slowest runner never
will be overtaken by the fastest; for it is necessary for the one
chasing to come first to where the one fleeing started from, so that it
is necessary for the slower runner always to be ahead some’ (Ph.
6.9.239b14–18). The argument employs the common-sense
assumption that a first runner and a lagging runner both cover some
distance while the lagging runner gets to where the first runner
started in order to generate a conclusion that flatly contradicts
common sense. The result is an impasse or aporia, where one does
not know what to say in reply. The aporia is generated by the
contradiction between the conclusion of Zeno’s reasoning and the
belief grounded in one’s experience of races and moving objects.
One wants to say that of course Achilles can overtake the tortoise.
There is a striking depiction of this response to Zeno in the interior of
a red-figure drinking cup discovered in the Etrurian city of Falerii and
dated to the mid-fifth century BC, where we see a heroic figure
racing nimbly ahead of a large tortoise.3 The painter’s response is
amusing though not particularly satisfying, for until one identifies
where Zeno’s reasoning goes wrong, the contradiction he has
generated persists.
Zeno likewise argued in various ways that the common-sense
assumption that there are many things leads to contradiction. The
arguments against plurality that we know of are more elaborate than
the arguments against motion reported by Aristotle. Consider the
antinomy of limited and unlimited. We are better informed about this
argument than about any other argument by Zeno thanks to
Simplicius’ quotation in his commentary on Aristotle’s Physics of
what must be the greater portion of the original. Simplicius means to
rebut Porphyry’s opinion that the argument from dichotomy Aristotle
mentions as motivating the early atomists belongs to Parmenides:

And why speak at length when in fact the argument is given in Zeno’s
very treatise? For in showing that if there are many things they are limited
and unlimited, Zeno writes word for word as follows: ‘If there are many
things, it is necessary that they be just so many as they are and neither
greater than themselves nor fewer. But if they are just as many as they
are, they will be limited. If there are many things, the things that are are
 unlimited; for there are always others between these entities, and again
others between those. And thus the things that are are unlimited.’ And in
this way he demonstrated their numerical infinity by means of the
dichotomy.

(Zeno 29B3 DK ap. Simp. in Ph. 140.27–34 Diels)

All that appears to be lacking is the conclusion that there are not
many things because they cannot be both limited and unlimited. This
could have come before or after the text quoted by Simplicius, or
some more general statement to the effect that saying there are
many things commits one to asserting contradictories could have
prefaced a series of arguments. This possibility is suggested by the
way Simplicius introduces his account of Zeno’s antimony of large
and small with the general remark that each of the arguments in his
treatise was designed to show that one who says there are many
things winds up saying opposites (Simp. in Ph. 139.5–7 Diels).
The following reconstruction aims to adhere closely to Zeno’s
words while making their reasoning a bit clearer. The general goal is
to show both that if there are many things, then there must be finitely
many things, and if there are many things, then there must be
infinitely many things. The assumption that there are many things is
thus supposed to have been shown to lead to the contradiction that
things are both finitely many and infinitely many. The particular
argument for the first arm of the antinomy seems to be simply: If
there are many things, they must be just so many as they are. If the
many things are just so many as they are, they must be finitely
many. Therefore, if there are many things, there must be finitely
many things. Simplicius somewhat loosely describes the antinomy’s
second arm as demonstrating numerical infinity through dichotomy.
In fact, the argument depends on a postulate specifying a necessary
condition upon two things being distinct, rather than on division per
se, and it may be reconstructed as follows: If there are many things,
they must be distinct, that is, separate from one another. Postulate:
Any two things will be distinct or separate from one another only if
there is some other thing between them. Two representative things,
x1 and x2, will be distinct only if there is some other thing, x3,
between them. In turn, x1 and x3 will be distinct only if there is some
other thing, x4, between them. Since the postulate can be repeatedly
applied in this manner unlimited times, between any two distinct
things there will be limitlessly many other things. Therefore, if there
are many things, then there must be limitlessly many things.
Gregory Vlastos describes this argument as ‘beautiful in its
simplicity’. Jonathan Barnes regards it as ‘merely simpliste’.4
Whatever judgement one passes on its substance, one has to
acknowledge that the form of Zeno’s reasoning is audaciously
original. There are some intricately structured arguments in
Parmenides 28B8 DK, of course, but nothing quite like the pattern of
reasoning whereby Zeno argues against his targeted claim by
showing how it leads to contradiction. Zeno may therefore fairly be
credited with inventing the technique of reductio ad absurdum. His
achievement is only augmented by the way he recurs to the same
pattern in other arguments for which we still have evidence to the
effect that if there are many things, they must be both like and unlike,
which is impossible (Pl. Prm. 127e1–4), and that if there are many
things, they must be both so large as to be unlimited in magnitude
and so small as to have no magnitude at all (Zeno 29B1 and 29B2
DK ap. Simp. in Ph. 139.7–15 and 140.34–1.8). This latter argument
is actually a super reductio, in that it purports to show not only that
the assumption that there are many things leads to contradiction but
also purports to reduce each of the incompatible consequences to
absurdity. It is the one true dilemma among Zeno’s arguments. The
antinomy of limited and unlimited does not present two equally
unpalatable alternatives. What is unacceptable is the contradiction
that things, if many, are both finitely and infinitely many. The
repetition of the basic pattern of argumentation suggests that Zeno
had some grasp of the argument’s form and appreciated its general
power. The only qualification necessary if we are to credit him with
the invention of the reductio technique is that it is not clear that Zeno
meant to establish positively that there are not many things by
showing that the claim that there are many things leads to
contradiction. The technique of Zenonian reductio is not the
technique of indirect proof. It appears, instead, to be a technique for
inducing aporia.
It may or may not be a mere coincidence that Zeno’s arguments
against plurality all take the form of antinomies while none of his
arguments against motion do so. In any case, these arguments
generate contradiction and aporia in distinct ways. The arguments
against plurality present two lines of argument to generate explicit
contradiction: if there are many things, they are both limited and
unlimited, both infinitely large and vanishingly small, and both like
and unlike. These contradictions are supposed to call into question
the assumption that there are many things. The paradoxes of
motion, by contrast, generate an implicit contradiction between the
ordinary experience of motion’s occurrence and the rational
considerations Zeno deploys against it. Of course, there is a similarly
implicit contradiction in the arguments against plurality, in that the
rational considerations not only lead to contradictory conclusions but
taken together contradict the ordinary experience of there being
many things. Both his opposition of logos to logos in the antinomies
and the broader opposition he generates between logos and
perceptual experience would have a long history. Furthermore,
unlike Heraclitus, Zeno at no point suggests how the contradictions
he presents might be resolved. Although later philosophers and
mathematicians, from antiquity to our own era, have developed
responses in the course of their own enquiries into space, time,
motion, and infinity, it seems unlikely that Zeno meant his paradoxes
to stimulate enquiry by framing a set of problems. Zeno’s purposes
appear to have been generally negative rather than positive, and in
this respect he set the trend in the use of contradiction in the rest of
early Greek philosophy.
It has often been supposed that Zeno’s arguments against
plurality and motion were meant to maintain in a different form the
position of Parmenides. Socrates says as much in Plato’s
Parmenides when he accuses Zeno of trying to conceal the fact that,
in saying that things are not many, he is really just saying the same
thing as Parmenides, who said that things are one (Pl. Prm. 128a6–
b6). But Plato has Zeno correct Socrates on this point: Zeno says
that his book was instead meant to provide indirect support for
Parmenides’ teaching against those who supposed its
consequences were ridiculous by arguing that their own presumption
that there are many things leads to even more absurd results (Prm.
128c6–d6). Plato’s Zeno does not countenance Socrates’ view that
his arguments against plurality reached the same conclusion as
Parmenides by different means. Likewise, the historical Zeno should
not be regarded as a defender of a view – namely, that only one
thing exists – that should not be ascribed to the historical
Parmenides. So Jonathan Barnes states:

Zeno was not a systematic Eleatic solemnly defending Parmenides

against philosophical attack by a profound and interconnected set of
reductive argumentations. Many men had mocked Parmenides: Zeno
mocked the mockers. His logoi were designed to reveal the inanities and
ineptitudes inherent in the ordinary belief in a plural world; he wanted to
startle, to amaze, to disconcert. He did not have the serious metaphysical
purpose of supporting an Eleatic monism.5

I would add that Parmenides himself does not belong to the early
history of aporetic reasoning because his arguments are not
designed to leave us in aporia. He aims instead to show that it is
possible to achieve an understanding that does not wander in the
way human understanding typically does when focused on the
mutable entities apprehended via the senses. A more stable form of
understanding is possible when we try to focus our minds on what is
and cannot not be and consider what such an entity must be like just
in virtue of its necessary mode of being.6
Melissus of Samos, by contrast, certainly does belong to the
early history of aporetic reasoning. Like Zeno, he developed
arguments contradicting the common-sense presumption of plurality
and change rooted in perceptual experience. Unlike Zeno, however,
Melissus’ arguments exploit difficulties in the logic of being in a
manner not unusual in the wake of Parmenides. Melissus’ treatise
contains two major arguments: one in 30B1–7 DK for the thesis that
‘one thing only is’, an argument which he calls his ‘greatest proof’,
and a second in 30B8 DK against the view that many things are. In
the first, he argues that whatever is, is ungenerated, sempiternal,
spatially unlimited, unique, homogeneous; it is subject to neither
alteration nor rearrangement, it suffers neither pain nor anguish, and
it is full, unmoving, neither dense nor rare, and nowhere divided. He
begins his argument as follows: ‘Whatever was always was and
always will be. For if it came to be, it is necessary that prior to its
coming to be there is nothing; if then nothing there was, in no way
could anything come to be from nothing’ (30B1 DK). The first
sentence of 30B2 DK, which may have followed directly upon these
words, completes the argument: ‘Since then it did not come to be, it
is and always was and always will be’. Melissus appears to be
referring here to the totality of what was, is, and will be rather than to
each individual entity in the set of all entities. In this way he can rely
on the principle that there is no genesis ex nihilo to generate the
conclusion that whatever is always was and will be. (If he meant only
each entity in the set of entities, the principle would not secure the
conclusion.) By the end of his argument, he has effectively ruled the
individual entities belonging to the set of all entities out of existence,
for he has argued that there is only a single, limitless, unchanging,
and completely undifferentiated entity. He also moves to restrict use
of ‘being’ to entities that are not subject to change: if whatever is
always was and always will be, then whatever has not always been
and will not always be – that is, whatever is subject to change –
cannot be something that ‘is’. The restriction of use of ‘being’ to what
is always is crucial for Melissus’ arguments in 30B7 that what is
cannot suffer diminution, growth, rearrangement, pain, or distress.
He stresses that all these varieties of change involve some sort of
becoming or perishing of what is.
The notion that whatever is cannot be subject to becoming and
whatever becomes cannot properly be said to ‘be’ becomes central
to Melissus’ second argument against the common-sense view that
many things are. He says:

If many things were, these would have to be just as I say the One is. For
if earth is and water and air and fire and iron and gold, and the living and
the dead, and black and white and other things such as people say are
real, if indeed these things are, and we see and hear correctly, each must
be just such as it first seemed to us, and it must not change or become
different, but each thing must always be just as it is.
(30B8.2)

People are prone to say that all manner of things ‘are’, but since this
verb, according to Melissus, properly applies only to things that are
(what they are) always and invariably, if we are right to say that the
various objects of our experience ‘are’, then they must perpetually be
just as we encounter them, and they cannot be subject to change or
alteration. The passage is not concerned with the mere existence of
earth, water, air, and the rest, but with the question of whether any of
these things can properly be said to ‘be’, that is, whether any of
these things really are, where this is taken by Melissus to amount to
their only, or ever and immutably, being (what they are). In short,
Melissus denies that entities subject to change can properly or
strictly be said to ‘be’. This is not immediately equivalent, however, to
denying that entities subject to change do not exist. He proceeds to
draw out the contradiction between our experience of the mutability
of things and what would be entailed by saying that such things ‘are’:
‘while we say that many things “are” and so eternal and having their
own characters and strength, it seems to us that all things become
different and change from how they appear on any particular
occasion’ (30B8.4). On the one hand, if the things people speak of
as being real are in fact so, then each of them must always be just
as it is (≈ 30B8.2), and yet experience shows that even those things
that seem strong and permanent do not continue being what they
once appeared to be (≈ 30B8.3). Melissus then resolves the
contradiction by rejecting the hypothesis that numerous things ‘are’,
a hypothesis based on the impressions of stability that lead people to
speak of various things as ‘being’ or ‘being real’. ‘Therefore it is
clear’, he says, ‘that we have not seen correctly and that those many
things do not correctly seem to be: for they would not change if they
were real, but each would be just such as it appeared to be; for
nothing is stronger than real being’ (30B8.5).7
Plato, Aristotle, and Isocrates in the fourth century all look back
upon Zeno and Melissus as forerunners of the ‘antilogic’ and eristic
disputation prevalent among the sophists.8 Isocrates does not
hesitate to group Zeno and Melissus together with Gorgias and other
sophists flourishing in the era of Protagoras as all having produced
exasperating treatises that advocate the most outrageous claims
imaginable (Isoc. Orat. 10.2–3). Aristotle saw Zeno as a
controversialist and paradox-monger whose arguments were
nevertheless both sophisticated enough to qualify him as the
inventor of dialectic and were important for forcing clarification of
concepts fundamental to natural science. Aristotle’s view of Zeno
basically accords with Plato’s portrayal of him as a master of the art
of contradiction.9 Aristotle’s view of Melissus is more negative, since
he sees his main argument as relying upon an obvious equivocation
(Arist. Ph. 1.2.185a5–12,1.3.186a10–22). The influence of Zeno and,
to a lesser extent, Melissus on the techniques of argumentation
promulgated among the sophists seems undeniable. Protagoras’
development of the techniques of antilogic, rooted in his claim that
there are two opposed arguments on every matter (D.L. 9.51),
seems likely to have been inspired by Zeno’s novel forms of
argumentation as well as by his advocacy of the most
counterintuitive of theses. The influence of both Zeno and Melissus
is especially clear, moreover, in Gorgias’ treatise, ‘On Nature, or On
What Is Not’, not only in its penchant for antithetical argument and
reductio but also in its use of premises drawn straight from Zeno and
Melissus themselves (as at [Arist.] MXG 979a23, b25, b37). More
generally, though, Protagoras and Gorgias can both be seen as
challenging the opposition between the deliverances of reason and
the senses exploited by Zeno and Melissus.
 Protagoras
Diogenes Laertius reports that Protagoras was the first to claim that
on every matter there are two accounts opposed to one another
(D.L. 9.51 = 80A1 DK, cf. Clem. Strom. 6.65 = 80A20 DK). Seneca
ascribes to Protagoras the richer claim that it is possible to argue on
either side of every issue with equal force, even regarding this very
issue, whether every issue is arguable on either side (Sen. Ep.
88.43). The claim that there are two opposed and equally forceful or
plausible cases on every matter would be essential to rhetorical
training – where the goal was to enable a speaker to win his point
regardless of its merits. A speaker who could do this was described
as able to make the weaker logos the stronger, and this ability was
associated particularly with sophistic rhetoric. As an example of the
type of argumentation this involved, Aristotle describes how the early
rhetorician Corax of Syracuse employed the commonplace that what
is improbable is probable given the probability of improbable things
happening since they do in fact happen:

It is of this commonplace that Corax’s Art of Rhetoric is composed: ‘If the

accused is not open to the charge – for instance if a weakling is tried for
violent assault – the defence is that he was not likely to do such a thing.
But if he is open to the charge – i.e. if he is a strong man – the defence is
still that he was not likely to do such a thing, since he could be sure that
people would think he was likely to do it’.

(Arist. Rh. 2.24.1402a17–20, Revised Oxford Translation)

Aristotle continues, after his diagnosis of the error here, by saying
that this is what is meant by making the weaker argument the
stronger and by adding that people were properly contemptuous of
Protagoras’ profession of such an ability (1402a24–6). One can
easily imagine that the two books of Antilogiai attributed to
Protagoras in Diogenes’ catalogue (D.L. 9.55) consisted of
techniques such as Corax’s commonplace for responding or
‘speaking against’ the kinds of claims likely to be made by an
opponent in public debate. There is in any case no evidence
attesting to the specific character and content of the Antilogiai, so
one should be wary of supposing that it consisted of fully developed
antinomies. The character of the Antilogiai is plausibly indicated by
the way Theaetetus in the Sophist understands the Eleatic Visitor’s
reference to various publications on individual arts designed to equip
speakers with the resources to contradict any particular craftsperson
as a reference to Protagoras’ treatises on wrestling and the other
arts (Pl. Sph. 232d5–e1).
Protagoras is also committed, after a fashion, to non-
contradiction. Plato has Socrates in the Euthydemus attribute to him
the view that contradiction is impossible (Euthyd. 286b8–c3, cf. D.L.
9.53). Unlike Plato and Aristotle, however, Protagoras does not deny
the possibility of contradiction because he thinks that only one of a
pair of contradictories can be true. On the contrary, Protagoras’ view
that contradiction is impossible is part and parcel of the broader
relativism encapsulated in his famous measure doctrine: ‘Man is the
measure of all things, of those that are that they are, and of those
that are not that they are not’ (Prot. 80B1 DK ap. D.L. 9.51, cf. Pl.
Tht. 152a2–4, S.E. PH 1.216, M. 7.60). Prior to his own quotation of
this famous pronouncement Sextus Empiricus notes that some
counted Protagoras among the philosophers rejecting the criterion
‘since he says that all impressions and beliefs are true and that truth
is a relative matter because whatever appears or seems to someone
actually is the case for him’ (S.E. M. 7.60, cf. D.L. 9.51). Plato has
Socrates interpret the measure doctrine in much the same way, after
quoting it in the Theaetetus: ‘So he is saying, then, that whatever
way things appear to me, that’s the way they are for me, and
whatever way they appear to you, that’s the way they are for you –
as you and I are each a man?’ (Pl. Tht. 152a6–8). Thus Protagoras
can plausibly be understood as committed to non-contradiction as a
corollary of his measure doctrine. If things seem one way to me and
another to you, there is no contradiction here, but things are for you
as they seem to you and for me as they seem to me. Likewise, if
things seem to me one way at one time and another way at another
time, there is no contradiction, but things are for me at each time as
they seem to me at each time.10 Protagoras’ view that on every
matter there are two accounts opposed to one another might seem
to sit uneasily with his view that antilogia or contradiction is
impossible. However, Protagoras can hold both views consistently –
and let us not worry just now about whether he would have thought it
necessary to be consistent – if the two opposed logoi are not
opposed in such a way that one must be true if the other is false.
Protagoras is reported to have held that all things are true (D.L. 9.51)
or that all appearances and opinions are true (S.E. M. 7.60), and it in
fact makes good sense to see a commitment to the truth of all
appearances as another corollary of the measure doctrine.
Protagoras might have left it at that and held that in each case
things are for each individual as he or she finds them and that if they
seem one way to one person and another way to another there is
nothing more to be said about the matter. He seems not to have left
it at that, however, but to have drawn some more general
conclusions. He announces, for instance, in the opening of his
treatise On the gods, ‘Regarding the gods I have no way of knowing,
neither that they are, nor that they are not, nor what sort of form they
have – for numerous are the hindrances to knowing, including the
non-evidence and the brevity of human life’ (Eus. PE 14.3.7, cf. D.L.
9.51). In a similar vein, he is quoted in Didymus the Blind’s
commentary on the Psalms as saying, ‘I appear to you who are
present to be seated, but to one who is absent I do not appear to be
seated; it is non-evident whether I am seated or I am not seated’
(Prot. ap. Didymus, in Psalmos Pt. III, p.  380, 222.21–2 Gronewald
and Gesché).11 What is especially interesting here is how Protagoras
in each case forces an impasse by generating a higher order
appearance. If Protagoras appears to you to be seated, then,
according to the measure doctrine, it is true for you that Protagoras
is seated, and you might have thought that this would be the end of
the matter as far as you are concerned. However, on this as on
every issue, there are two opposed logoi, pro and contra. Protagoras
balances the case for supposing he is seated with the fact that he
does not appear seated to someone not there to observe him, and,
again according to the Measure Doctrine, if Protagoras does not
appear to someone absent to be seated, then it is not true for
someone absent that Protagoras is seated. There is no contradiction
here, for contradiction is impossible. Instead, it is true that
Protagoras is seated for you, and it is not true that Protagoras is
seated for someone absent. When Protagoras asks you to entertain
this state of affairs, a change is supposed to come about in how
things seem to you, a change which results in its no longer being
clear to you that Protagoras is seated. It does not cease to appear to
you that Protagoras is seated – if this were supposed to be the case,
then the argument would collapse – yet it is non-evident whether
Protagoras is seated. The reason, apparently, is that appreciating
that to someone else it does not appear that Protagoras is seated
leads you to question the privileging of your own experience as a
guide to how thing in fact are. If you wonder how it can seem to you
simultaneously that Protagoras is seated and that it is unclear
whether Protagoras is seated or not, you might come to appreciate
that Protagoras is developing an opposition between your perceptual
and rational appearances.
One wants to be cautious about reading into Protagoras the sort
of stratagems developed in the later sceptical tradition. A measure of
assurance that this unpacking of the little bit of reasoning preserved
in Didymus is not anachronistic is provided by Aristotle’s account of
some of the considerations that led Democritus to hold that either
there is no truth in appearances or it is adēlon or non-evident to us:

And again, many of the other animals receive impressions contrary to

ours; and even to the senses of each individual, things do not always
 seem the same. Which, then, of these impressions are true and which
are false is not obvious; for the one set is no truer than the other, but both
are alike. And this is why Democritus, at any rate, says that either there is
no truth or to us at least it is not evident.

(Arist. Metaph. 4.5.1009b7–12, Revised Oxford Translation)

Numerous Democritean fragments confirm that his tendency was to

deny the truth of appearances because he associated truth with the
reality of atoms and void. Most strikingly, he says, ‘in actuality we
know nothing – for truth is in the abyss’ (Democr. 68B117 DK).
Instead of concluding from the diversity of appearances that none
are true, as Democritus did, Protagoras concludes that all are true.
He draws the contrary conclusion because he has abandoned the
project in which Democritus was engaged of giving an account of
how things really are that transcends these contradictory
appearances. Democritus distinguishes between the obscure form of
judgement afforded by the senses and the genuine form afforded by
the intellect (Democr. 68B11 ap. S.E. M. 7.138). The intellect
enables us to transcend the conflicting appearances to reach an
understanding of the true nature of things. For Protagoras, however,
the intellect is only a source of more conflicting appearances.
Later philosophers beginning with Plato and Aristotle would find
Protagoras’ position unstable. Part of the problem is that its
foundations – the measure doctrine, its two corollaries of non-
contradiction and the truth of all appearances, and the related claim
that there are two opposed and equally forceful logoi on every issue
– are all themselves paradoxical. Unlike Zeno, who reasons from
common-sense presumptions to paradoxical conclusions,
Protagoras argues from principles that are already paradoxical.
Unless one is prepared to accept the truth of the measure doctrine,
one will hardly be inclined to accept that contradiction is impossible
and that all appearances are true. The common-sense opinions that
contradiction is possible and that not all appearances are true are
motivation for rejecting the measure doctrine. One way of doing so is
suggested by the line of argument Protagoras himself employs in the
quotation preserved by Didymus. If the Measure Doctrine appears
true to Protagoras, he must nevertheless acknowledge that there are
others to whom it does not appear true, so that via the Measure
Doctrine itself, the doctrine is true for himself and not true for others.
He should consequently acknowledge that it is non-evident whether
the Measure Doctrine is true or not. It may continue to seem to him,
as it did at first, that the Measure Doctrine is true even as he
acknowledges that it is unclear whether it is true or not, but this
attitude certainly seems unstable and even irrational.
After Protagoras, the paradoxical claims that contradiction and
falsehood are impossible survive as eristic devices for silencing
one’s opponents in debate.12 They are no longer grounded in his
particular relativist stance but are based on more purely logical
considerations. The result is a particularly aggressive means of
driving a debate to impasse. Despite the destructive purposes of
those who employed it, the paradoxical rejection of falsehood and
contradiction on logical grounds stimulated philosophical efforts to
develop a theory of truth and falsity capable of resolving the
paradoxes. In this respect, the sophistic arguments against the
possibility of falsehood and contradiction are much like Zeno’s
arguments against plurality and motion.
 Gorgias
In his tour de force, ‘On Nature, or On What Is Not’, Gorgias erected
an elaborate structure of argument on behalf of three successive
claims: nothing is; even if something is, it is unknowable; and even if
something is and is knowable, it cannot be revealed or
communicated to others.13 While Gorgias has too often been taken
as arguing in the first part of his treatise that nothing exists,14 the
description in the pseudo-Aristotelian De Melisso Xenophane Gorgia
of the treatise’s doxographical preface signals clearly enough
Gorgias’ aim in arguing that nothing is:

And to show that it is not he collects the statements made by others who,
in speaking on the things that are (peri tōn ontōn), apparently
contradicted each other – some showing that they are one and not many,
while others show that they are many and not one, and some showing
that they are ungenerated, while others show that they are generated –
he draws his inferences against both camps.

([Arist.] MXG 979a13–18)

The reports of the arguments of the treatise’s first division in the

sources indicate that Gorgias likely classified views regarding ta onta
according to a third opposition – mutable vs. immutable – in addition
to the oppositions one vs. many and ungenerated vs. generated
mentioned in the MXG report. Jaap Mansfeld has argued decisively
that ta onta here ‘are not in the first place the phenomenal things, but
the speculative theoretical constructs of the Presocratic
philosophers’.15 That is to say, ta onta here are only and specifically
those entities that occupy the fundamental positions in the
Presocratic ontologies, and thus Gorgias’ purpose is to argue
systematically against the various efforts by his predecessors to
establish the existence of such fundamental entities. We may go
ahead and call such entities ‘substances’. Gorgias’ doxographical
classification is the forerunner of similar classifications by Plato in
the Sophist and by Aristotle in Physics 1.2.
Having isolated the contradictions among his predecessors’
varying accounts of ta onta, Gorgias set about demolishing them all
by arguing that each option in each opposition proves untenable.
The arguments of the treatise’s first division are thus structured as a
set of genuine dilemmas. For example, if something is, he argues, it
must be either ungenerated or generated, yet each of these two
options is supposed to prove impossible. The MXG reports that he
argued ‘via the axioms of Melissus’ that what is ungenerated must
be unlimited, that what is unlimited could not be anywhere, and then
‘via Zeno’s argument concerning place’ that if it is nowhere it is
nothing; but if it is generated, it would have to come to be either from
what is or from what is not, neither of which, he argued, is possible
(MXG 979b21–34). Gorgias’ conclusion in the MXG paraphrase is as
follows: ‘If then it is necessary, if something is, for it to be either
ungenerated or generated, and these are both impossible, it is also
impossible for something to be’ (MXG 979b34–5). Here we are less
concerned with the details of the argument than with its form and
function. To this end, consider the influence Gorgias’ treatise
apparently had on Socrates. Xenophon explains that Socrates
refrained from discussing natural philosophy and that he marvelled
at the blindness of his predecessors in not seeing that human beings
cannot solve the riddles of the cosmos – as evidenced by the great
diversity of opinion among those who concerned themselves with the
nature of the universe: ‘Some hold that what is (to on) is one, others
that it is unlimited in number; some that all things are ever changing,
others that nothing can ever undergo change; some that all things
come to be and perish, others that nothing can ever come to be or
perish’ (Xen. M. 1.1.14). These are the same oppositions as the
ones that emerge from Gorgias’ doxographical preface and that
structure the arguments of his treatise’s first division. For Socrates,
though, the divergence among the theories regarding the number
and nature of the world’s substantial entities led him to adopt a
sceptical stance on these issues and to refrain from enquiring about
them. Gorgias goes further. The doxographical schema displaying
the contradictory views of those who had engaged in the enquiry peri
phuseōs does not in itself license a sceptical conclusion but gives
systematic structure to a set of dilemmatic arguments purporting to
show that every option here leads to impossible results.
Although Gorgias’ arguments are Zenonian in their complex
structure, they target not common sense but theoretical views.
Gorgias’ effort to demonstrate that the enquiry peri phuseōs leads
only to impasse or aporia may plausibly be understood as validating
operation at the level of everyday appearances. Such validation he
may have seen as underpinning his rhetorical practice in something
like the way Protagoras’ practice was supported by his measure
doctrine. Be that as it may, none of Gorgias’ arguments in the first
division are as powerful or brilliant as Zeno’s better arguments, and
there is no evidence that later philosophers felt compelled to respond
to them in detail. Even so, Plato does respond to Gorgias, in a
general way, in the Parmenides. After highlighting various problems
with the young Socrates’ conception of forms, Parmenides echoes
Gorgias’ three principal claims in saying that the problems just
presented might lead one to contend that the forms or hai ideai tōn
ontōn do not exist, that even if they do, they are necessarily
unknowable, and that it takes a truly remarkable ability not only to
gain an understanding of these entities but also to teach these things
to another (Pl. Prm. 135a3–b2). The First Deduction in the
subsequent dialectical exercise, moreover, is reminiscent of the first
division of Gorgias’ treatise in its general argumentative strategy and
at points in its particular argumentation, and it is no surprise that it
reaches a similarly negative conclusion.16 Even so, how seriously
Plato took the arguments in the first division of Gorgias’ treatise is
unclear.
Plato would appear to have taken more seriously the position
articulated in the treatise’s subsequent divisions, for he would have
appreciated that, despite their negative purpose, they raise serious
problems for understanding the relations between language, thought,
and the world. Gorgias’ arguments in the second division for the
general claim that if something is, it is unknowable and unintelligible
all turn on the subsidiary claim, ‘If ta phronoumena are not onta, then
to on is not thought’ (S.E. M. 7.77, 78).17 The three arguments
reported by Sextus as belonging to this division are all designed to
secure by means of reductio the truth of the antecedent that ta
phronoumena are not onta and thus the truth of the consequent that
to on is not thought.18 The first two arguments run as follows. If ta
phronoumena are onta, then all things thought are, no matter how
one thinks them, but this is counterindicated (apemphainon) given
that merely thinking, for example, that a person is flying does not
mean that a person is flying; therefore, ta phronoumena are not onta
(S.E. M. 7.78–9). If ta phronoumena are onta, then ta mē onta will
not be thought, given that opposites belong to opposites; but this is
absurd (atopon) given that many things that are not, such as Scylla,
are thought; therefore, ta phronoumena are not onta (S.E. M. 7.80).
At play here are some of the problems concerning the relation
between thinking and being found in the sophistic denials of the
possibility of falsehood, which Plato would sort out in the Sophist.
The arguments of the third division of Gorgias’ treatise – where
he argued that even if something is and is knowable, it cannot be
communicated to another – pose various problems for semantic
theory, some of which Plato would seem to have taken quite
seriously. The MXG indicates that the various arguments here turn
on two basic points, the difference or gap between things and their
linguistic signifiers, and the difficulties in supposing that one person
can have the same conception as another (MXG 980b18–20).
Insofar as he takes the view that there is no knowledge of particulars
as such but that they are grasped by perception, Plato might actually
be supposed to have accepted Gorgias’ point when he argues that
neither speech nor any other sign for an object is capable of getting
a person to conceive of an object unless he is already perceiving it
and that speech is generally incapable of making external objects
manifest as this role belongs instead to perception (MXG 980a20–b9,
S.E. M. 7.83–6). Plato most definitely resisted Gorgias’ arguments
turning on the other point, even though he well appreciated the
challenge they pose. Here Gorgias argued, in the first place, that an
auditor cannot have in his mind the same thing that is in the
speaker’s mind, on the grounds that it is not possible for the same
thing to be at the same time in a number of separate entities
because what is one would thus be two (MXG 980b9–12). He went
on to argue that even if the same thing could be in numerous entities
simultaneously, nothing would prevent it from appearing differently to
them given that they are themselves different and differently
situated, since even the same person does not perceive the same
thing in the same way at different times (MXG 979b12–18). Plato
certainly recognised the depth of the challenge posed by these
arguments, although this is not the place to delve into his response.
What is important for our purposes is that, as with Zeno’s
arguments that motion is impossible, Gorgias’ arguments in support
of this claim that, if something is and is intelligible, it cannot be
communicated to another proved a spur to philosophical progress
despite their essentially negative purpose. It would also be left to the
later tradition to find more positive uses for aporetic reasoning, for
perhaps the most striking feature of its uses among the early Greek
philosophers, with the exception of Heraclitus, is their tendency to
employ the instruments of contradiction and aporia primarily to
negative and destructive ends.
1 See the textual note at Kirk, Raven and Schofield 1983: 192, for a defence
of the readings xumpheretai and palintonos.

2 Aristotle’s association of Heraclitus with violation of the Law of Non-

Contradiction on the grounds that some people thought that he
simultaneously supposed the same thing to be and not to be (Arist. Metaph.
4.3.1005b17–25) has prompted some modern interpreters to worry about
Heraclitus’ toleration for contradiction, but it is clear enough from fragments
such as 22B61 and 22B88 that the contradictions he points up are meant to
be merely apparent and ultimately explicable. See further Barnes 1982a:
69–75, Barnes 1983, Mackenzie 1988a, and Granger 2004.

3 Rome, Museum Villa Giulia National Museum, inv. 3591. See Hoffman
2004 and the accompanying plate.

4 Barnes 1982a: 252, cf. Vlastos 1967: 371.

5 Barnes 1982a: 236. See further Palmer 2009: 189–205, for the view of
Zeno presented here.

6 Palmer 2009: chs. 2–4, develops this view in detail.

7 See further Palmer 2009: 205–24, for the view of Melissus presented here.
For a different view, see Makin 2005.

8 George Kerferd has argued both that the patronage of Pericles and his
keen interest in the intellectual developments of his day must have been
critically important to the sophistic movement and that Zeno’s paradoxes
were a profound influence on the development of the sophistic method of
antilogic, which he sees as ‘perhaps the most characteristic feature of the
thought of the whole period’ (Kerferd 1981: 18–23, 59ff., 85).

9 Plato consistently associates Zeno with the rise of eristic disputation and
especially the specific brand of argument known as antilogic. See Pl. Prm.
128d–e, Phdr. 261d6–8 (cf. Plu. Per. 4.5), Sph. 216a–b. Aristotle by his own
criteria would have regarded Zeno’s arguments as more eristic than properly
dialectical, for he clearly believes that some of Zeno’s assumptions have
only a specious plausibility. See Arist. Top. 1.1.100a29–30, b22–5,
8.8.160b7–9, SE 24.179b17–21, Ph. 6.2.233a21–31, Metaph. B.4.1001b13–
16.

10 The view that contradiction is impossible is also attributed to Prodicus in

the commentary on Ecclesiastes by Didymus the Blind, on the different
grounds that if two people (purportedly) contradict one another, both cannot
be speaking about the same things but only the person who says what is
true and describes things as they are is actually speaking about them. See
Binder and Liesenborghs 1976 for the text and discussion.

11 The text was first published in Gronewald 1968 and subsequently, with
slight variations, in Gronewald and Gesché 1969. For discussion and more
deflationary interpretations, see Mejer 1972, Mansfeld 1981: 51–2 and
Woodruff 1985.

12 See Palmer 1999: 124–34.

13 Although Gorgias’ original version has not survived, we are comparatively

well informed about its contents thanks to two later summaries, one in
[Arist.] MXG 5 and 6 and one in S.E. M. 7.65–87. Although Diels and Kranz
did not include the MXG summary in Die Fragmente der Vorsokratiker, it has
come to be recognised as the superior source. Comparison of the two
sources shows that the summary in Sextus has rearranged the material in
the first part of the treatise to make it conform to a trilemmatic rather than a
dilemmatic structure. See Palmer 1999, Appendix I. For the text of Gorgias’
treatise, see Buchheim 1989.

14 Or that ‘the verb “to be” cannot be used of phenomena either positively or
negatively without contradiction resulting’ (Kerferd 1955: 14, et passim,
developing the proposal of Calogero 1932: 197).

15 Mansfeld 1985: 248–9, contra Kerferd 1955.

16 See Palmer 1999: 111–16, for detailed demonstration of Gorgias’

influence on the Parmenides’ First Deduction.
17 Sextus gives every appearance of quoting Gorgias, or at least closely
paraphrasing. He repeats these words verbatim, first interspersed with the
phrase phēsin ho Gorgias (M. 7.77) and then with to introducing the whole
sentence (M. 7.78).

18 The arguments of this section are unfortunately not well preserved in the
MXG, where the text breaks off and has suffered corruption. One
nevertheless can still discern in it sufficient similarities with the arguments
given by Sextus to confirm that he is not recasting them to the extent he (or
his source) has done in the first section.
 Chapter 2
Socrates and the Benefits of
Puzzlement
◈

Jan Szaif

Focusing on some of Plato’s early and transitional dialogues, this

essay aims to enable a better understanding of the philosophical and
ethical significance of aporia as a cognitive state of mind induced by
Socrates in his interlocutors. After commenting (1) on what the
aporetic outcome is supposed to tell us about Socrates’ interlocutors,
I am going to discuss (2) the interlocutors’ reactions to their
predicament and (3) the difference between Socratic refutations and
the combative (‘eristic’) style of refutation. The final section (4) will
explore the potential intellectual and ethical benefits of the aporia-
inducing discourse for Socrates’ interlocutors. I won’t discuss here to
what extent (if at all) the Socrates character in the aporetic dialogues
is himself genuinely puzzled, as this would require a more extensive
treatment of the epistemology of Plato’s early works and of the
function of the Socrates persona in them.
 The primary textual reference for my investigation is what I am
going to call the ‘core group’ of aporetic dialogues (Laches,
Charmides, Euthyphro, Protagoras, and the first part of the Meno,
70a–80d).1 These conversations seem to be coherent in style,
content, and purpose. Apart from their close thematic connections,
they also maintain a characteristic balance between personal
examination and engagement with philosophical problems that we
don’t find in the Lysis or the Hippias major. In addition, I shall look
more closely at the role of aporia in the Euthydemus since this
dialogue exemplifies, and contrasts, two different kinds of aporia-
inducing argumentation: Socratic versus sophistic or eristic
argumentation. From the Euthydemus, together with the comments
on aporia in the middle part of the Meno, we obtain some crucial
evidence on how Plato wanted the philosophical significance of
aporia to be understood.2
In the dialogues typically viewed as aporetic, Socrates’
interlocutors don’t always admit that they have been reduced to a
state of aporia. Accordingly, I am using a criterion for the literary form
of an aporetic dialogue that relies on Socrates’ assessment of the
outcome: A Socratic dialogue shall count as aporetic if and only if, as
by Socrates’ admission, it has failed to produce an answer that can
stand (at least) as a preliminary result. For the core group of aporetic
dialogues, I am going to assume an ‘aporetic reading’, to be
distinguished from two alternative styles of interpretation that can be
dubbed ‘doctrinal’ and ‘sceptical’. For reasons of space, I cannot
here justify the merits of my proposed criterion and of the aporetic
reading of such dialogues. Suffice it to say that this style of reading
tries to do justice to their literary form by approaching these
dialogues as philosophical challenges that undercut the trust in our
habitual concepts and judgements, but also provide us with cues and
directions for further enquiry – not yet with definitive solutions, but
with starting points and seeds, as it were, for future solutions. The
‘aporetic reading’ assumes, in other words, that these dialogues are
meant to lead their readers onto a path of philosophical enquiry by
making us, first, aware of our cognitive (and, by implication, ethical)
deficits and, second, by engaging our explorative thinking with the
right kinds of questions and arguments. In this specific sense, one
might say, with V. Goldschmidt, that their intent is ‘to form rather than
inform’.3
 1
In Plato’s early and transitional works, the philosophically relevant
meaning of the noun aporia and its cognates relates to the situation
of someone who proves unable to complete a philosophical task or is
(as yet unsuccessfully) struggling with a philosophical problem. The
relevant connotation is that of an intellectual impasse, or state of
perplexity, or also that of a lack of resources for fulfilling the task at
hand. The characteristic sign of this condition is that the person has
been rendered speechless.4
In order better to understand the philosophical and ethical
implications of this kind of predicament, I’ll begin with some
comments on the role of the articulation and elenctic testability
requirement employed in Socratic conversations. This requirement is
an epistemic criterion applied to the interlocutors’ understanding of
concepts.5 It contains the assumption that the kind of knowledge we
have whenever we clearly understand a concept-word (e.g.
‘courage’) includes the ability to articulate this knowledge through a
definitional account and to defend the account under elenctic testing
(cross-examination, elenchos). The Laches and the Charmides
provide us with good examples for how this requirement is
introduced in Socratic conversations.
After its long introductory part, the Laches presents us with two
elenctic conversations concerning the nature of courage, one with
Laches and the other with Nicias. Laches is confident that they all
know and understand what virtue is and that he would also be able
to say what it is (190c). When asked the same kind of question with
respect to a part of virtue, viz., courage, he is again certain that an
answer won’t be difficult for him (190e). Accordingly, Laches not only
believes that he knows but also thinks that such knowledge comes
with the ability to articulate it. This admission is not at all harmless.
From a modern perspective, we might be inclined to view our
understanding of a general term like ‘courage’ as the ability correctly
to apply a semantic rule (or set of interconnected rules). Knowledge
of linguistic rules is a form of knowing-how, and such knowing-how
need not include the ability to articulate the rule. Native speakers are
able to employ numerous syntactic rules without thinking of them
and usually without being able to articulate them in the form of
general rules. The same may hold for the semantic rules that provide
terms with descriptive meaning. Think of Wittgenstein’s theory of
linguistic understanding as mastery of ‘language games’.
Alternatively, one might stipulate that the understanding of concepts
and corresponding concept-words is a form of intuitive knowledge
defying informative articulation. An example would be Husserl’s
theory of intuiting essences (Wesensschau).
Yet Socrates’ interlocutors accept the assumptions underlying
the articulation requirement without hesitation. They are thus
surprised and embarrassed when they realise that they can’t
articulate their understanding without contradicting themselves under
elenctic testing. The phenomenology of this experience is perhaps
best described in the Laches. Having expressed his confidence that
he will be able to say what courage is, Laches’ first attempt falls
short since he tries to answer the question by citing some typical
example. After some instruction by Socrates, he gives an answer
that comes somewhat closer to fulfilling at least the formal
requirements for a correct definiens, suggesting that courage is
‘some kind of perseverance’ (karteria tis).6 This proposal is then
subjected to an elenchos (192c–3e), in which Socrates first directs
the argument towards the result that the perseverance in question
must be combined with some form of practical intelligence or
knowledge (phronēsis), but then reverses the direction of the
argumentation so that it now has the outcome that a (potentially
risky) action is done more courageously if unsupported by
knowledge. This apparent contradiction resulting from arguments pro
and contra leaves Laches in a state of perplexity. He expresses his
puzzlement in a famous reply (194ab) in which he talks about his
inexperience with this kind of argumentation and the ambition
(philonikia) that has now seized him, yet also emphasises how much
his inability to articulate his understanding of courage annoys him.
He is still convinced that he does understand (noein) what courage
is.7 What troubles him is the breakdown of the connection between
his (alleged) understanding and his ability to articulate it, leaving him
unable to ‘apprehend’ courage through an account (sullabein tō[i]
logō[i]). The way in which he describes his predicament tells us that
he views it only as a temporary condition, caused by his lack of
experience with arguments.8 We, the readers, to be sure, are meant
to see that he does not have knowledge, or full understanding, of the
nature of courage. His belief that he already knows what courage is
is an illusion. Yet for Laches himself, as for many other Socratic
interlocutors, it is hard to admit to themselves that they lack full
understanding of such basic ethical concepts as courage or
virtue/excellence. In the case of Laches, having an understanding of
courage or bravery is also clearly part of his self-image as a military
man.
When Socrates solicits Nicias’ help, he describes his and
Laches’ predicament by using the word aporia together with its
cognates aporein and aporon (194bc). He compares their aporetic
state to the situation of people who are ‘storm-tossed’
(cheimazomenoi) in their words or discussion (en logō[i]): a
metaphor that suggests the image of a boat tossed around in a
winter storm. Nicias is asked to release them (ekluein) from their
aporia by way of articulating his understanding (ha noeis) in an
account that would provide a stable ground (tō[i] logō[i] bebaiōsai).
Through this imagery, aporia is characterised as a situation which is
both constrained and insecure: We can’t advance in the desired
direction (like people in a winter storm that drives their boat away
from their destination), and our thoughts are unstable, or ‘tossed
around’, under the impact of changing appearances in the absence
of a reliable account.
Based on the articulation and testability requirement, failure
under elenctic examination implies that the person examined is not
qualified as a teacher of, or adviser on, virtue. This is the specific
angle of the Laches: Socrates examines whether Laches and Nicias
would be qualified to give advice to young people on how best to
acquire virtue. Their claim to virtue is not questioned, at least not
openly. In the Charmides, Socrates goes a step further, since he now
connects the inability to articulate the essence of a virtue with the
absence of genuine virtue in the interlocutor. The focus is here on a
putative recipient of moral education, and the guiding concern is to
establish whether young Charmides already possesses the virtue in
question, viz., sōphrosune (soundness of mind, moderation).9
Socrates claims that if Charmides has sōphrosune, he should also
have some view or belief (doxa) regarding ‘what it is and what it is
like’. For, so he says, if it is ‘in you’, it should also cause some
perception of it (aisthēsis tis) and hence some belief about it. Since,
moreover, Charmides knows Greek, it should then not be difficult for
him to state how he perceives and thinks of (hoti soi phainetai)
sōphrosune, provided he has it (159a1–8). In a later passage,
160de, Socrates even urges Charmides: ‘Look into yourself’, clearly
suggesting that the virtue in question, if present, could be detected
by some form of introspection.
The claim that the presence of a virtue in the soul must be
accompanied by some introspective perception is rather
questionable.10 It is not even clear what it could mean to ‘perceive a
virtue in one’s soul’, since a virtue is a disposition rather than a
mental happening. Charmides might just have thoughtlessly agreed
(which also happens on other occasions in Socratic conversations),
or he might have followed a hunch about how people’s inner
experience of the workings of a virtue in the soul provides them with
a perception of this virtue.11 Yet whatever he may have been
thinking, from Socrates’ angle there is usually some way of making
sense of a seemingly counterintuitive yet important assumption in a
Socratic conversation. This also applies to this case, as I will now
argue.
 When Socrates suggests that the presence of sōphrosune in
Charmides’ soul ought to be accompanied by some perception of,
and hence belief about, the nature of this virtue, he speaks very
cautiously of doxa.12 However, from the remainder of the dialogue it
is clear that articulating just some belief (doxa) about sōphrosune
won’t count as evidence for Charmides’ possession of sōphrosune.
After all, he does articulate beliefs when questioned by Socrates; but
his (and Critias’) proposals fail, and at the end of the dialogue
Charmides seems to acknowledge that his soul is still in need of
healing and does not yet possess genuine sōphrosune.13
The underlying methodological assumption is that the
possession of genuine virtue entails the ability to articulate beliefs
that don’t become unhinged under examination. It implies an
intellectualist understanding of virtue: if virtue were not identical with,
or inclusive of, some form of wisdom cognitively transparent to itself,
then Socrates could not use his test of Charmides’ understanding of
what sōphrosune is as a means to finding out if he has this virtue.14
If we combine the articulation and testability requirement with the
intellectualist assumption of the self-transparency of virtue, then
failure under elenctic examination and the resulting aporetic state of
mind turn out to be indicators not only of a cognitive but also of an
ethical deficit.15
Let’s step back for a moment. Are these intellectualist
assumptions too implausible to have any credibility? I don’t think so.
This Socratic and Platonic project is driven by the idea that human
self-improvement, and hence the realisation of ‘virtue’ or excellence
in the soul, is a function of developing our understanding of the
things that matter most, including our understanding of our
intellectual potential as humans. This should seem a very important
project, perhaps doomed to fail, but at least well worth the attempt.
 2
Interlocutors can be put in a state of confusion through trick
arguments. Yet we wouldn’t want to recognise the success of such
trickery as a reliable indicator of a fundamental cognitive and ethical
deficit in the interlocutor. This presents a problem for our
assessment of Socratic conversations in Plato, since Socrates’
refutations contain numerous examples of non-sequiturs due to
vagueness, ambiguity, (potentially controversial) missing premises
etc. It is difficult to draw a sharp line, based on logical criteria,
between the Socratic elenchos and sophistic argumentation. Yet our
recognition of the pedagogical or psychagogic value of the Socratic
elenchos and of the resulting state of aporia presupposes that there
is some significant difference between aporia produced by Socratic
dialectic and puzzlement induced by combative (eristic) debating
techniques. If not in logical validity, wherein then lies the difference?
16

Let’s first look at how Socrates’ interlocutors assess the

situation.17 Generally speaking, age, self-image, and the
interlocutors’ relation to Socrates seem to be key factors determining
their reaction to Socratic examination. Laches, for instance, who
enjoys high social standing as a general but is not a professional
debater or self-professed intellectual, thinks that his lack of
experience with argumentative discourse is the cause of his
perplexity, yet he wouldn’t go so far as to admit ignorance of virtue.
There are other interlocutors, especially among the self-professed
intellectuals, who put the blame squarely on Socrates’ questioning
technique. At the conclusion of the Euthyphro’s first elenctic
conversation, Euthyphro uses the same phrase as Laches (eipein ha
noō) to vent his frustration about his inability to successfully
formulate and defend his (alleged) understanding of the virtue of
piety, lamenting the fact that their accounts prove elusive and don’t
stay put (11b). Yet when Socrates teasingly compares Euthyphro’s
proposals to the mythical sculptures of Daedalus (which were known
to run away from their owners), Euthyphro replies that Socrates is
the Daedalus who makes proposals run away; in other words, he
blames Socrates’ knack for disputation, not his own ignorance, for
the impasse (11b–e, cf. 15bc). Euthyphro maintains his pretence of
expertise throughout the dialogue and at the end aborts the
conversation under a lame pretext.
Meno, too, tries to deflect the blame for his failure onto Socrates
(79e–80a). After his most substantive attempt at defining virtue has
failed, the conceited young nobleman and former student of Gorgias
retorts that he has already heard others say about Socrates that he
is in a state of aporia and causes aporia in others. Meno then
proceeds to describe the effect of Socrates’ questioning in colourful
words as a form of bewitching or sorcery (80a2 f: goēteueis me kai
pharmatteis kai atechnōs katepa[i]deis, cf. b6) and to compare it to
stepping on a torpedo fish. Through this image, he suggests that
contact with Socrates has a paralyzing effect on mind and tongue,
analogous to how contact with a torpedo fish numbs the body. As a
result, he, Meno, who has given so many beautiful speeches about
virtue (80b), is now not even able to answer this (allegedly) so
simple question:18 what virtue is. The metaphor of sorcery evokes
the idea of harmful manipulation, while the image of the stinging
torpedo fish characterises aporia purely negatively as incapacitation.
Like Laches, Meno sees what has happened to him as a breakdown
of the natural connection between his understanding and the ability
to articulate, but while Laches blames his inexperience, Meno
accuses Socrates of manipulative questioning.
Socrates responds that this image fits him only if the torpedo
fish paralyzes others while it is itself paralyzed (autē narkōsa … poiei
narkan, 80c6 f ). For he, Socrates, does not induce aporia in others
without himself being in such a state. Superficially, this agrees with
what Meno has said about Socrates, since Meno also mentioned
Socrates’ own state of aporia. Yet Socrates does not accept the
charge that he harms his interlocutors by making them share his
aporia.19 Meno, by contrast, suspects malevolence on Socrates’s
part. He completely lacks Socrates’ appreciation of the potential
benefits of philosophical aporia. The situation only annoys him (as
we can also tell from his intimidating remark in 80b4–7).20
In order to regain the upper hand, Meno conjures up the
paradox later named after him (80d), a move which is unique among
the aporetic dialogues since it represents an argumentative
counterattack by an interlocutor who has just been reduced to a
state of aporia. What is, from Meno’s point of view, the purpose of
inserting this paradox at this particular juncture of their conversation?
Since Meno views his impasse as a temporary mental paralysis,
which will pass because it is only an effect of manipulation, he can
still think of himself as knowledgeable about virtue (aretē). So, from
his angle, the paradox of enquiry is not actually a problem for him.
He does not need to enquire what virtue is, as if this basic concept
were something unknown to him.21 If Socrates, on the other hand,
should really be ignorant of virtue, then his search would be
pointless, since he wouldn’t even know what he’d be looking for.
With this move, Meno can maintain both his claim to knowledge of
virtue and undermine the professed goal of Socratic conversations.
Students of the Meno tend to read Meno’s reaction in this passage
as a genuine realisation of his ignorance.22 Yet I think this misreads
both Meno’s character and the dynamic of his interaction with
Socrates in this passage.
If Socrates were really some kind of mischievous Daedalus-like
sculptor of untrustworthy arguments, or a conversational torpedo fish
causing mental paralysis, his conversations could, at best, be
considered impressive displays of a combative skill. Socrates would,
in other words, turn out to be a practitioner of combative debating,
often referred to in Platonic dialogues as eristikē or agonistikē (also
antilogikē, the art of counterargumentation23). We know from the
Apology, but also from Aristophanes’ Clouds, that Socrates was in
fact widely perceived as a teacher of harmful debating skills.
Even in such an apparently civil conversation as the Laches,
suspicion and resentment regarding the presumed motives of the
interlocutor/opponent can rise to the surface. In the elenchos of
Nicias, there is a moment early on at which Laches inserts himself
and temporarily assumes the role of a questioner. When it turns out
that Nicias is quite capable of managing his answers so as to avoid
refutation by Laches, the latter accuses Nicias of trying to win the
debate in words only (196ab). We can tell that Laches is still
embarrassed by his own ‘defeat’ and has developed an agonistic
attitude towards Nicias (cf. 195ab). Referring to Nicias’ alleged
verbal trickery, he claims that Nicias, instead of honestly admitting
his aporia, tries to wriggle himself out from the grip of Laches’
refutation by ‘twisting this way and that way’ in his answers
(strephetai anō kai katō),24 just so he can conceal his impasse
(epikruptomenos tēn autou aporian). Note a characteristic change of
metaphor: When Laches reached the point of crisis in his
conversation with Socrates, he compared the concept he was trying
to articulate to a quarry they were hunting for, and the definition to
the act of catching and subduing it (194b2–4); now Laches is
describing the situation as an agōn between him and Nicias: a
contest that will be won by whoever manages to catch and subdue
his opponent, the other person. Why this change of attitude? Laches
does not think of Socrates as a rival, but this is how he sees Nicias
(naturally so, since both are highly regarded military experts
competing for esteem and commissions in the same polis).
Accordingly, the comparison between his and Nicias’ performance
becomes a matter of prestige and face-saving, which redirects his
ambition (philonikia, 194a8) from the hunt for the definition to
outwitting his rival.25 These and other examples indicate that there is
always a latent suspicion about the true motives of elenctic dialectic:
that its real aim is not to clarify the subject matter, or test our
knowledge, but to defeat and embarrass the interlocutor (cf. Charm.
166c5f ).26
 3
What does, in the face of such misgivings, distinguish Socratic
aporia-inducing conversation from an eristic debating style? Plato
never provides an abstract analysis of how sophistic trick-arguments
differ from genuine philosophical objections and puzzles. Yet there is
one dialogue, the Euthydemus, which is specifically dedicated to
showing us by example what the difference is. The eristic
showpieces proudly presented by the two martial-arts experts turned
professional debaters, Euthydemus and Dionysodorus, not only
stand in stark contrast to Socrates’ performance of a protreptic
conversation in the same dialogue, but they also allow us to see how
Socratic refutations in the aporetic elenctic dialogues differ from
sophistic debating.
The sophists’ arguments are ‘inescapables’ (aphukta, 276e5, cf.
275e) because they lead to refutation whichever way the respondent
chooses to answer, whereas Socratic refutation engages the
interlocutor’s beliefs and thus turns into a personal examination
(Lach. 187e–8a). Sophistic refutations are short set pieces that are
easy to memorise and easy to mimic, as we can see from the fact
that Ctesippus, just from watching, quickly learns how to defeat the
sophists with their own tricks (298b–d, 300cd). Socratic examination,
by contrast, is a skill hard to reproduce, as attested by the failure of
those interlocutors who try to assume the role of a questioner (e.g.
Polus in Gorg. 462b–3d, Laches in Lach. 195b–6a). The trick-
arguments of the two brothers exploit various kinds of ambiguity
(ambiguous words, incomplete phrases, scope ambiguities), which is
why they try to suppress qualified answers beyond a mere ‘yes’ or
‘no’. While insisting on short answers is also part of Socrates’
tactics,27 he does not place such an extreme restriction on his
respondents.
Yet beyond these widely acknowledged points of difference,
there are also certain similarities, especially from the angle of the
interlocutors’ experience. Socrates, too, knows how to render his
interlocutors speechless, and his interlocutors get the impression
that whatever they say, in the end Socrates will be able to overpower
them. Socrates is also, as mentioned, no stranger to exploiting
ambiguities and vagueness, or other sources of fallacy, in his
refutations. Even his protreptic conversation in the Euthydemus
includes fallacious reasoning.28 Since he shows himself able to
discern and expose such trickery in arguments proffered by the two
sophists (e.g. 277d–8b, 295b–6d), we have to suspect that he is
aware of illicit ambiguities and non-sequiturs in his own inferences
and that he exploits them knowingly.29 Nor will the criterion of
philosophical relevance allow us to distinguish sophistry from
Socratic puzzles. Despite the frivolous way in which the two brothers
present their trick-arguments and paradoxes, at least some of them
have a bearing on genuine philosophical questions (e.g. 283e–4c,
285e–6b, 300e–01c), while the study of others could help to advance
the readers’ understanding of logical syntax. Meno’s paradox of
enquiry, or the various versions of the argument against the
possibility of saying what is false,30 are set pieces of eristic
argumentation, but in Plato’s hands they morph into major
philosophical challenges. Yet for the two sophists – and here, I think,
we are arriving at the most crucial distinction – the philosophical
potential of such eristic puzzles is completely irrelevant because they
have no genuine educational or philosophical agenda. To confuse
their opponents and render them speechless is the end goal. It is
what success in debate means to them. Hence, what separates
them from Socrates is, first and foremost, the ethos underlying the
refutational practice.31 Recall how Laches used a fighting metaphor
when he wanted to scold Nicias for his alleged dishonesty in the
debate. The Euthydemus has many examples of metaphors taken
from the domain of fighting, the final eristic display ending in a
metaphorical knockdown that has Socrates lie speechless on the
ground (303a). From the vantage point of such argumentative
agonistic, the state of aporia is a situation in which an opponent has
been overpowered, constrained and, finally, ‘knocked out of his
mind’. In contrast to this, aporia induced by Socrates, as exemplified
by the outcome of the protreptic conversation between Socrates and
Clinias, is a medium for getting interlocutors seriously engaged with
philosophy for the sake of their own cognitive and ethical self-
improvement.
It stands to reason that this goal also informs how Socrates
presents his own role as critical questioner in the aporetic dialogues.
Socratic refutation claims to result from agreed-upon premises in a
joint investigation. On the one hand, these ‘agreements’ serve to
commit the interlocutors to the premises that will lead to their
refutation. Yet, on the other, emphasis on the shared character of the
premises also allows Socrates to include himself in the aporia: he
was not knowledgeable at the outset, and since the proposal on the
table has been refuted on the basis of agreed-upon premises, he,
Socrates, is still at a loss and shares his interlocutors’ predicament.
No winners or losers! – Or, at least, that’s how Socrates likes to
frame it, for instance in the Laches (200e), after both Laches and
Nicias have succumbed. We, as readers, and also Socrates’
interlocutors, have reason to believe that his pretence of ignorance is
not fully sincere and that the agreed-upon premises don’t necessarily
reflect Socrates’ own beliefs. Yet presenting himself (ironically) as an
ignorant person, out to find knowledge in others, assists his
therapeutic approach. Avoiding both top-down instruction32 and
combative refutation, he tries to make his interlocutors aware of their
deficiency in a non-humiliating manner and to turn33 them onto a
path of intellectual advancement that employs their own intellectual
resources.
Such ‘turning-around’, resulting in a lasting commitment to
philosophy, paired with gratitude34 for Socrates, the Athenian ‘gadfly’
(Apol. 30e), is, however, not the usual outcome of the aporetic
dialogues. Contemporary ancient readers would have been aware of
this, since the biographies of the key interlocutors were well enough
known. What explains this failure? Some of the interlocutors are
intellectuals who view themselves as skilled debaters or expert
speech-makers; they won’t own up to their puzzlement because they
have too much at stake in terms of their reputation and, as we would
put it, self-image. Others are people of political influence who share
the widespread anti-intellectual resentment (for instance, Anytus in
the Meno). They view Socrates as just another sophist. Both
categories of interlocutors don’t consider elenctic examination as the
path towards truth, the former because they advertise argumentative
techniques as a tool for social power and agonistic exploits, the latter
because they perceive such discourse as a waste of time or, worse,
as a threat to the established political order and accepted values.
Then there are also those young interlocutors whose characters and
lifestyles could still be shaped into a philosophical mould: adolescent
boys such as Charmides or Clinias. Yet their lives are exposed to
many other influences that counteract the work of a philosophical
educator. A vivid description of such conflicting emotional pulls is
given by Alcibiades in his encomium on Socrates in the Symposium
(215e–16c, 218ab), while Republic VI describes in general terms the
detrimental social influences that operate on the development of a
talented young person (489e–95b).
The Platonic Socrates cannot have been so naïve as to believe
that he could change a Protagoras or reform a Meno. So why did he
bother to talk to them, or why did Plato bother to recreate such
seemingly fruitless conversations? I surmise that such sample
refutations were meant to benefit the readers (or, at the fictional
level, Socrates’ audience), first, by raising awareness of the shaky
foundations of our accepted notions and our trust in experts, and,
secondly, by engaging us with questions and puzzles that could
serve as starting points for further enquiry. Finally, these dialogues
also warn us about how the potential benefits of aporia can be
wasted if interlocutors don’t ‘hang on’ and don’t interpret their
situation correctly.
 4
Notwithstanding these sad observations on how little the
interlocutors actually benefit from conversing with Socrates, we can
gain a more positive outlook if we reflect on how they could benefit if
they weren’t in denial about their condition. To this end, let’s return to
the Meno. We left its conversation right after Meno’s paradox of
enquiry. As argued above, Meno does not think that this paradox is a
problem for him. For he does not think that he lacks understanding of
the topic of their conversation. He takes his perplexity to be a result
of Socratic debating ‘witchcraft’. The paradox’s bite is directed
against the Socratic project of philosophising through a stage of
aporia: if the argument of the paradox stands, there would be no way
out of the aporia.
In his response (80d–6c), Socrates explains his own
understanding of aporia and search, introducing the theory of
learning as recollection. To illustrate his points, he performs a
sample conversation with a slave-boy about a simple problem of
geometry. The first phase of the conversation represents the
elenchos and the resulting state of aporia; yet the demonstration
does not stop there but turns into a constructive mode of question
and answer that guides the boy to the discovery of the true solution.
Through this example, the Platonic Socrates tells us that aporia,
while essential to philosophical progress, needn’t be its final
outcome since humans have a potential for the (re)discovery of the
truth and can realise it, gradually, at least if they encounter the right
guide.35
We are interested here in the description of the aporetic phase
(Men. 84a–c). The task set up for the slave-boy is to identify the line
on which a square twice the size of a given square can be
constructed. After his first two attempts have been refuted, he is at a
loss (aporein). At this point, Socrates inserts some comments
addressed to Meno about the kind of progress the slave-boy has
already achieved: In the beginning, he didn’t know the correct
answer but thought he knew it; now he has realised that he doesn’t
know. Socrates adds that the boy is also in a better condition with
respect to the object of enquiry (beltion echei peri to pragma),
compared to his previous condition.
These comments contain two claims which, though intimately
connected, ought to be distinguished: First, thanks to the refutation
and resulting state of aporia, the boy is no longer subject to a
cognitive illusion about himself, but aware of his ignorance.
Ignorance paired with the conceit of knowledge is an irrational
cognitive stance, whereas ignorance aware of itself has achieved a
form of self-knowledge. Yet, secondly, according to Socrates, the
boy is also better off with respect to the object. How so? Isn’t the boy
as ignorant as before with respect to the object? Socrates probably
has in mind that previously the boy formed false opinions about how
to answer the question, whereas now he is no longer judging (truly or
falsely) but just keen on finding out. Such withholding of judgement
with respect to an object, if you have no secure ground yet for
judgement, is a much better cognitive attitude than judging
prematurely and, hence, oftentimes falsely. Moreover, the boy now
also realises that certain facile solutions that looked plausible at first
have no merit. This gives him a better appreciation of the complexity
of the problem.
The idea that the aporetic state itself amounts to some important
cognitive progress even with respect to the object of investigation is
reasserted in a later text, the sixth division in the Sophist, which
describes the elenctic craft in ways strongly reminiscent of the
Socrates figure in earlier aporetic dialogues. Since this text falls
outside the scope of this essay, let me just highlight how it confirms
the point in question: Conceit of knowledge is characterised in the
Sophist as the root cause of false judgements (dianoia[i] sphallesthai
229c6) and groundless certainties (megalai kai sklērai doxai,
230c1f).36 The ‘kathartic’ use of elenctic examination can bring a
change since it enables us to distinguish between what we know and
what we don’t know, thus purging us of the causes of error (230cd).
The guiding idea here seems to be that those who become aware of
their ignorance withhold judgement and thus avoid erroneous
judgement.
In the Meno (84a–c), Socrates goes on to emphasise that as a
result of his aporia, the slave-boy now has a desire to find the right
solution. We readers are meant to understand that, in general, we
are complacent and without interest in genuine research as long as
we remain attached to our conceit of knowledge. Yet once we find
ourselves in a state of aporia (and assuming that we are not in denial
about it), we begin ‘to long for’ (epothēsen, 84c6) the discovery of
the truth. This verb is presumably a rather strong expression here,
as it can also be used in the context of erotic longing (e.g. Phdr.
251e2, 255d8). In this connection, it is also significant that the very
word ‘philo-sophia’ is composed of a cognitive and an appetitive
term. A passage in the Lysis (218ab) comments on this aspect. Its
upshot is that the philosophical state of mind is a condition in which
one is affected by something bad (viz., ignorance) without ‘being’
bad, i.e. without being alienated from the good so as no longer to
strive for it. This ‘philo-sophical’ condition, which has us seek for the
good of wisdom, is made possible by the awareness of our
ignorance (218b1). The same idea is also expressed in the
Symposium (204a), with an explicit link to the topic of erōs as the
basic desiderative force in our souls. In light of the Meno passage,
one may add that it is through aporetic argumentation that we realise
our ignorance and that aporia is, hence, a necessary condition for
the philosophical desire to come to life.37
If interlocutors realise their condition but don’t commit
themselves to a philosophical lifestyle, this can leave them with a
lasting sense of shame or guilt. Shame is mentioned in the Sophist
as an effect of refutation (230d1f) and described very vividly in
Alcibiades’ speech (Symp. 215e–16c). Alcibiades’ shame is of a very
different kind than that experienced by Thrasymachus (Rep. 350cd).
The latter’s shame is really a form of embarrassment relating to his
public persona and self-image: by being refuted he has lost face
before an audience of supporters and respectable citizens. This kind
of shame does not jolt him out of his accustomed ways. Alicibiades’
shame, on the other hand, goes much deeper and resembles what
we would call guilty conscience. He feels guilty and ashamed
whenever the presence of Socrates reminds him of the fact that he
isn’t able to abide by a philosophical mode of life, although he knows
that this would be the right path. He just can’t tear himself away from
the lures of political power and high social esteem. Awareness of his
imperfection was not enough to fully turn him around; it left him only
with a sense of unease and regret.
At the end of section I, I mentioned the reasons for why the
state of aporia, when it occurs in reference to an ethical notion,
indicates not only a cognitive but also an ethical deficit. What we
have now learned about the potential emotional effects of Socratic
refutation in terms of desire and shame underlines the practical side
of the Socratic elenchos. The Apology and, indeed, Alcibiades’
speech in the Symposium (215e–17a, 218e, 221e–2a) lay much
emphasis on the quality of Socratic conversations as pieces of moral
exhortation.38 This feature of Socratic elenctic is, to be sure, far
more palpable in the case of constructive elenctic discourse on how
we ought to live, best represented by the Gorgias. Yet the aporetic
dialogues fulfil the same function, if in a subtler way. In the Socratic
perspective, rendering you aware of your ignorance concerning key
ethical notions amounts to moral exhortation, since it makes you
realise that you don’t yet have a secure grip on the correct ethical
standards. Cognisant of your condition, you will now seek
improvement through further engagement in joint philosophical
enquiry (Prot. 348cd). Or, at least, so would the interlocutor react
who does not cling to a misleading self-image or to inferior
attachments, but surrenders to Socrates as a ‘physician of the soul’
(Gorg. 521e–2b).39
 What is more, the person who strives after, and rationally
investigates, ethical standards that accord with the truth does not
only seek improvement; in the Socratic perspective, such
philosophical engagement already represents a greatly improved
condition. Although the best condition would be a firm and
comprehensive understanding of the ethical truths and their
psychological and metaphysical foundations, the ability to sustain a
sense of philosophical puzzlement and to engage in genuine
philosophical research is, as it were, second best. It is so not only
because it entails a great cognitive advancement over ignorance
unaware of itself, but also on account of the fact that a life dedicated
to improving one’s understanding of the essence and effects of virtue
has a distinctly more virtuous orientation than a life that pursues
conventional goals such as pleasure or social advancement.40

1 For the Protagoras, the aporetic form is not generally agreed (cf. Politis
2012a: 212f ). Yet recall how in the Gorgias (which is one of the examples in
Plato of constructive, rather than aporetic, elenctic) Socrates is emphatic
about the value of the dialogue’s argument as the only trustworthy guide in
our lives (527de). In the Protagoras (360e–1d), by contrast, he distances
himself from the argument and its results by stressing, first, the need for a
prior clarification of the essence of virtue and by pointing out, second, that
his and Protagoras’ positions in the debate have shifted from one opposite
to the other and that such a ‘topsy-turvy’ course of argumentation should
inspire little trust. There is, moreover, a usually overlooked reference to the
state of aporia in the concluding passage (361cd), when Socrates recalls the
foolish approach of Epimetheus that put Epimetheus in a state of aporein
(321c).

2 The elenctic manifesto in the Apology also provides some relevant

background, but it does not use the terminology of aporia to describe the
result of refutation. Socrates only mentions his puzzlement (aporein, 21b7)
concerning the intended meaning of the oracle’s statement.
3 Goldschmidt 1947: 3: ‘Le dialogue veut former plutôt qu’informer.’
(Goldschmidt is here characterising Plato’s dialogues in general, whereas I
am talking about the early aporetic dialogues in particular.)

4 The noun aporia and the corresponding verb aporein (‘to be in a state of
aporia’) are derived from the root of the noun poros (passage, pathway,
way/means of achieving). Their meaning also seems influenced by their
association with cognates such as porizein (‘provide’) and its antonym
euporia/euporein in the meaning of ‘plenty’, ‘abundance’ (cf. LSJ). To be in a
state of aporia can, hence, mean that one is somehow caught in an impasse
without perceiving a way out. It can also signify a lack of resources. (See
also Mackenzie 1988b: 16–20, who emphasises the ironical uses of these
expressions.) With Politis (2006: 89, 2007: 269–72), one may call problem-
specific aporia zetetic (e.g. Charm. 167b, 169cd; Men. 75c; Prot. 324d and
e, 326e) and distinguish it from kathartic aporia, occurring typically at the
end of an aporetic conversation (cf. Lach. 194bc, 200e; Men. 80cd). The use
of the term aporia as a label for philosophical problems or puzzles is not
unambiguously identifiable in Plato’s early works, but the metonymy from
the mental state to the philosophical problem causing this state is rather
natural.

5 The importance of this kind of requirement for Socratic enquiry has been
widely recognised, but its relation to how the interlocutors interpret the
aporetic outcome is, it seems to me, not yet sufficiently understood. It is
commonly assumed that Socrates’ interlocutors realise their ignorance when
they find themselves ‘to be at a loss’, whereas I am going to argue that
those who think of themselves as knowledgeable only admit, at best, to a
temporary breakdown between their understanding and their ability to
articulate.

6 This proposal is still not formally correct because the qualification ‘some
kind of’ does not belong in a well-formed definiens.

7 This is clearly implied in what he says in this passage: ‘aganaktō ei houtōsi

ha noō mē hoios t’ eimi eipein. Noein men gar emoige dokō peri andreias
hoti estin, … ’ (194a8–b2). He still speaks of his noein in the present tense
and not as a past illusion. Sprague’s widely used translation shifts from ‘to
think’ for the first occurrence of noein to ‘to know’ for the second occurrence
(and similar solutions can be found in other translations). Yet the second
occurrence (followed by a gar) takes up the first. In both instances, noein
must mean the same, and in the second instance it clearly means
something stronger than just ‘to think’. (The adverb houtōsi qualifies the
entire clause after ei.)

8 The fact that he sees his inability to articulate his understanding as

temporary is expressed by the aorist ‘hopē[i] me arti diephugen’ (‘how it just
now escaped me’ 194b2f ), which is to be contrasted with the present tense
form describing his understanding (noein; see note 7 above). For the widely
held view I am arguing against (i.e. genuine realisation of his ignorance),
see, for instance, Sprague 1973: 7f; for the contrary view (no realisation of
ignorance) Benson 2000: 30.

9 Cf. Irwin 1995: 21.

10 Objections are easy to come by. For instance, someone can have a talent
without knowing of it; and even in the case of mental states whose presence
entails perception, it doesn’t follow that we would also be able to grasp the
essence and provide a definition: Pains are perceived, or felt, but this does
not yet reveal their essence, whatever that may be.

11 Socrates’ interlocutors seem to agree that virtues are, in some sense,

‘powers’ (compare Lach. 192b5–8 and Prot. 330ab, 349bc, 359a with the
discussion of sōphrosune as a self-referential cognitive power in Charm.
167b–9a). Charmides might therefore understand the introspective
experience of sōphrosune as an experience of the workings of a power in
one’s own soul.

12 Cf. Tuozzo 2011: 146f.

13 This outcome is somewhat mitigated by Socrates’ self-deprecating

remarks in 175e–6a. Yet compare Charmides’ remark in 176b1f with 157a3–
6, 158b5–c2, d8–e1.

14 Socrates’ implied intellectualism also agrees with the thematic direction of

the Charmides (and other dialogues of the same group). Its high point is a
long discussion of virtue as some form of knowledge of knowledge and not-
knowledge, likely motivated by Plato’s engagement with the Socratic
intellectualist claim that virtue is some form of wisdom; cf. Irwin 1995: 39f.
From such an intellectualist vantage point, it stands to reason that acting
from wisdom (and hence acting virtuously in the fullest sense) includes a
correct understanding of what it means to be wise and act wisely, and this in
turn can be taken to imply that virtue has to be cognitively transparent to
itself – in other words, that it should include a correct understanding of its
own nature.

15 This reasoning applies, mutatis mutandis, also to the moderate

intellectualism of the Republic, on account of the fact that it views wisdom as
a necessary component of virtue. (My ‘aporetic’ reading of the core group of
aporetic dialogues means that I don’t read them as indicating a doctrinal
commitment to rigorous intellectualism that Plato would later have given up.)

16 I am disagreeing here with Benson 2000: 86–90, who argues that the
Socratic elenchos cannot use fallacious arguments since that would
invalidate its goal of demonstrating inconsistency among the beliefs held by
the interlocutor. In my interpretation, the goal is to show that the interlocutor
has a muddled and imprecise understanding of the relevant concepts. This
is why arguments that exploit vagueness and ambiguity can serve the
purpose of the elenchos. If the interlocutors’ concepts were clear, they would
see through these ambiguities.

17 Cf. Cotton’s survey (2014: 67–75) of the emotional reactions of the

Socratic interlocutors in the early dialogues.

18 Cf. Men. 71e1, 71b9–c2.

19 Compare Politis 2007 for Socrates’ angle on this situation.

20 My interpretation of the personal dynamic between Socrates and Meno at

80ab is quite close to Klein’s (1965: 88f ); yet, as far as I can see, Klein does
not state a position on whether, for Meno, his current impasse means that
he is ignorant about what virtue is. I disagree with Matthews’ claim (1999a:
51f ) that the metaphor of the torpedo fish provides ‘the canonical
expression of Socratic perplexity’ (cf. n. 35 below); rather, it expresses
Meno’s twisted interpretation of the situation. Cf. Brague 1978: 113–18 on
the meaning of the metaphors used in this passage; Scott 2006: 69–74, who
acknowledges that the torpedo fish image serves to criticise Socrates for his
allegedly manipulative style of refutation.
21 In his short speech in 80ab, Meno draws a rhetorical contrast between his
current speechlessness and the many fine speeches about virtue he has
given in the past to a great number of people. The qualifying clause ‘hōs ge
emautō[i] edokoun’ in 80b3 can be read either as an admission that he was
wrong when he thought that he spoke well, or as a narcissistic rhetorical
evocation of the great impression he had of himself on these occasions,
without the suggestion that he was wrong then. How we read this clause
depends on how we assess Meno’s character and the rhetorical tendency of
this passage. His character flaws alluded to (vanity, self-indulgence; 80bc,
86d), and the way in which he uses disparaging, yet also self-justifying,
metaphors to characterise Socrates’ skill and character (and even his
looks!), make it much more likely that he isn’t yet ready to avow his
ignorance to himself. Besides, his criteria for assessing if a speech is ‘fine’
are surely primarily rhetorical and relate to its effect on the audience. His
current impasse under questioning by Socrates gives him no reason to
discount previous proofs of his rhetorical bravura. – Once Socrates switches
from the mode of elenctic examination to an expository mode, patiently
demonstrating to Meno that he has, in fact, all truths in his soul and only
needs to recollect them (an idea that must be flattering to Meno), he
becomes much more amenable to the idea of a joint search (86b) and
assumes a more cooperative stance for the remainder of the dialogue.

22 For instance Fine 2014: 69f; more cautiously Sharples 1985 ad loc.

23 Cf. Szaif 1996: 260–70.

24 The parallel formulation in Euthd. 302b suggests a metaphor from the

domain of fishing, hunting, or fighting: twisting in all directions after having
been caught in a net; the parallel in Ion 541e7–8 recalls the story of how
Odysseus caught Proteus.

25 See also Beversluis 2000: 111–34 on the dramatic aspects of the Laches.

26 See also Gorg. 489bc: Callicles blaming Socrates for quibbling over
verbal ambiguities; Rep. 336c: Thrasymachus characterising the exchange
between Socrates and Polemarchus as driven by pride or ambition
(philotimia); Prot. 335a: Protagoras characterising their exchange as a
verbal contest (agōn logōn); Tht. 167d–8b: Protagoras lecturing Socrates on
how to conduct a philosophical conversation through question and answer in
a fair and constructive manner.

27 Compare Protagoras’ protest in Prot. 334de; Gorgias’ reminder in Gorg.

449b9f.

28 For instance in the argument for the identity of wisdom and good fortune
(eutuchia) in 279d–80b; see also Chance 1992: 67.

29 For the general idea cf. Sprague 1962, who argues that ‘fallacy is often
part of the elenchos, of the dialectical shock-treatment administered by
Socrates’ (87), cautioning that ‘reluctance to pass an adverse moral
judgement upon Plato (or perhaps […] Socrates)’ has had the result that
‘Plato’s competence as a logician has failed to be evaluated correctly’ (81).
Vlastos’ well-known argument for Socratic sincerity (1991: 132–56) is based
primarily on the constructive elenctic of the Gorgias, which, on account of its
apologetic objective, operates under different constraints (cf. Benson 2000:
80–5).

30 Compare Euthd. 283e7–4a8, Crat. 429d4–6 with Tht. 188c–9b, Soph.

237c–e, which in the Sophist serves as the starting point for the ontological
enquiry into not-being.

31 On this thesis, cf. Beversluis’ (adversarial) comments in 2000: 40f, and

Nehamas 1990.

32 Cf. M. Frede 1992a on how Plato’s conception of what it means to be

educated philosophically informs his use of the dialogue form (and of
aporetic dialogues in particular).

33 On aporia and ‘turning-around’ compare the Republic’s simile of the cave:

The aporetic state is likened there to the condition of someone who is
dazzled by a brighter light. The captive has already been unshackled and
‘turned around’ so as to look in ‘a more correct direction’ (515cd, cf. 518cd),
but he cannot yet cope with this brighter light (by which we are meant to
understand the greater intrinsic clarity of the objects of Socratic
investigation).
34 Cf. Tht. 167e–8b, Soph. 230bc on the adequate reaction to refutations
that make the interlocutor aware of his muddles.

35 Matthews 1999a: 67–75 argues that the description of the aporetic

situation in the slave-boy passage indicates a reassessment by Plato of the
role of elenctic conversation and aporia: Socrates abandons the conception
of aporia as a shared condition reflecting the deeply puzzling nature of
philosophical problems, which is captured in the image of the ‘self-stinging
stingray’. According to the new conception introduced in the Meno,
perplexity is but an intermediate stage affecting the disciple, not the
‘knowing guide’. – Here are my objections: First, the aporetic dialogues are
fully compatible with the idea that perplexity ought to be a temporary state, a
starting point for further philosophical research. Second, the fact that
Socrates guides the slave-boy to the solution and is, in this case, not himself
puzzled does not prove that Plato, or the Platonic Socrates, now believes in
relatively straightforward solutions to basic philosophical problems, whereas
previously he had an aporetic conception of philosophy. The primary goal of
the slave-boy passage is to vindicate the thesis that knowledge acquisition
is a form of recollection. A simple mathematical problem, with an
uncontroversial solution known to the other participants (and the readers), is
the best kind of example for this purpose. We can still assume that Plato
thought of the complete and definitive resolution of fundamental
philosophical puzzles as a much more difficult (and still unfulfilled)
enterprise. Regarding the notion of the competent philosophical guide, cf.
Symp. 210a6f. (In light of 209a–e, such a guide should probably be
understood not as a perfect knower but as an advanced searcher, not unlike
Socrates in this respect.)

36 The pseudo-Platonic Alcibiades I, 116e–18b belabours the same point.

37 The allegorical account in Symp. 203b–04c, according to which erōs is

born from Poros and aporia-stricken Penia (Want), also illustrates the point,
as one of the contributors to this volume has pointed out to me.

38 Cf. Tsouna 2015: 15–17 on Alcibiades’ speech and the character of his
association with Socrates.

39 In Gorg. 521e–2b, Socrates ascribes to himself the role of the only

genuine pursuer of the political art and then goes on to compare his
situation to that of a physician who has to defend himself against the
accusations of a cook before a jury of children. The idea of the analogy is
that the true task of the political art is to improve the souls of the citizens,
just like a physician improves his patients’ bodily condition. Socrates argues
that the physician causes, in some sense of the word, aporia in his patients
and that Socrates’ own logoi do so as well. In both cases, the ‘treatment’
benefits those who receive it. Yet since these ‘patients’ are unreasonable,
they fail to see the benefit.

40 This Socratic perspective (cf. Apol. 38a) also agrees with the remarks in
Phd. 68c–9d on how a philosophical lifestyle dedicated to the quest for truth
or wisdom is the foundation for genuine moral virtues, and in Rep. 485a–7a
on how love of truth, being the defining trait of a ‘philosophical’ character, is
trailed by all other virtuous character predispositions (cf. Szaif 2004: 186–
202). Gonzales 2002: 175–82 goes even further when he argues that the
pursuit of virtue and wisdom is already a way of having virtue and wisdom –
the way is also the goal (179–81). According to my interpretation here, the
core aporetic dialogues promote philosophical puzzlement as a turnaround
and start, yet while the genuine engagement with philosophical puzzles
does already have a healing effect on the soul, it is not as such the goal.
 Chapter 3
Aporia and Sceptical Argument in
Plato’s Early Dialogues
◈

Vasilis Politis
 Introduction
In the first, and major, part of this essay (Section 1) I argue that there
is a particular argumentative element in a number of Plato’s early
dialogues which has the following sceptical consequence:

The Sceptical Claim in Early Plato

For a variety of questions regarding which we do not in general believe

that we cannot know the answer, there is good reason to think that, first,
we cannot know the answer, and, secondly, we cannot have more reason
to believe, or be more justified in believing, that one answer is true than in
believing that a contrary or contradictory answer is true.

This scepticism is directed not only against knowledge, but also

against belief.
The argumentative element in Plato that I have in mind, as
motivating this sceptical claim, is what I call aporia-based argument.
I have argued elsewhere, and shall briefly rehearse here, that
aporia-based argument is a major element in the method of
argument and enquiry in a number of Plato’s dialogues generally
considered early (see esp. Politis 2015). I mean by aporia, that
which is articulated by a whether-or-not question, or in general a
two-sided question, such that it appears to one and the same person
that there are good reasons on both sides and this person does not
in the least know, and recognises that she does not in the least
know, how to resolve this conflict of reasons. I have argued
elsewhere (see e.g. Politis 2006) that we find this use of the term
aporia and its cognates in several places in these dialogues, an
exemplary passage being Protagoras 324d2–e2.
I do not, of course, mean to assert that this is the only use of the
term aporia and its cognates in Plato’s early dialogues; or that
argument involving aporia, in the sense of aporia just mentioned, is
the only aporia-involving argument in these dialogues. For, of
course, I do not mean to deny that there is also in these dialogues a
notion of aporia which (as I have argued elsewhere, esp. Politis
2006) is recognisably and significantly different, namely, aporia as
the mental state that results from the failure of an enquiry, especially
when this failure is due to the discovery of a contradiction in a set of
beliefs. This is the notion of aporia with which we have long been
accustomed – and not least through Vlastos’ account of the Socratic
elenchos – and which may be termed elenctic or refutative aporia. It
is not part of my aim in this paper to consider the differences
between these different uses and functions of aporia-involving
argument in Plato’s early dialogues, but some differences are worth
noting here. First, whereas elenctic or refutative aporia is naturally
situated at the end of an enquiry, argument based in aporia, in the
sense of being based in a two-sided question with apparently good
reasons on both sides, is situated at the beginning of an enquiry,
which, indeed, it serves to set in motion and to structure. Secondly,
whereas there is a readily available way of restoring consistency in a
belief-set upon recognising a contradiction in it, that is, by simply
giving up one or more of the beliefs, this remedy is not readily
available if the conflicting beliefs appear to one to be, or indeed
appear in general to be, well supported. Thirdly, whereas
demonstrating an inconsistency in a belief-set does not contribute to
considering the inherent credibility of the individual beliefs in it – this
indeed is why the Vlastosian account of the Socratic elenchos
concluded that this mode of argument is largely ad hominem1 – to
engage with the apparently good reasons for the propositions on
each side of a two-sided question is very much to consider the
inherent credibility of these propositions.
In general, an inference from aporia-based argument to a
sceptical conclusion is of questionable validity. But I shall argue that,
for particular reasons, Plato is, validly and cogently, committed to
this inference. It follows that, if there is not, in the same dialogues, a
countervailing argumentative element, then Plato will be a sceptic
with regard to a variety of important questions.2 I believe there is a
countervailing argumentative element in these dialogues, which we
may, accordingly, call anti-sceptical. What I have in mind is the
demand for definitions. I have argued elsewhere (see Politis 2012b
and 2015) that a major function of the demand for definitions in a
number of early Platonic dialogues is to provide a means of
answering questions that articulate aporiai. We may, accordingly,
formulate this anti-sceptical claim as follows:

The Anti-Sceptical Claim in Early Plato

If an aporia is such that there is good reason to think that it cannot be

resolved, then the one and only way of resolving it is by establishing the
definitions of the concepts that figure in it.
In the second part of the paper (Section 2) I argue that the sceptical
claim in Early Plato is not compatible with a certain account,
defended by a number of critics over the past fifty years, of Plato’s
early epistemology; that is, the epistemology associated with the
demand for definitions and the thesis that knowledge is based in
definitions. The account of Plato’s early epistemology that I have in
mind says that:

A Common Account of Plato’s Early Epistemology

PP1. There is a distinction in the early dialogues between knowledge

proper and a lesser cognitive state that is still reliable; and
PP2. The thesis which says that knowledge must be based in
definitions is intended by Plato as a thesis only about knowledge proper,
and not also about reliable belief.

When this account of Plato’s early epistemology was introduced by

critics, in the late 1960s and early 1970s, it was introduced for the
purpose of answering Peter Geach’s 1966 charge. Geach argued
that, according to Plato, one cannot tell whether an object, O, is an
example of a quality, F, unless one already knows the definition of F;
and he concluded that this has the unacceptable consequence that it
is impossible to search for the definition of F. The critics’ response to
Geach, based on this account of Plato’s early epistemology, is
illustrated in Gerasimos Santas’ 1972 paper, with a statement that is
representative of all subsequent responses to Geach based on this
account:

They [Socrates and the interlocutors] may not have knowledge of the
examples in some Platonic sense of ‘knowledge’ … but they are not
 totally ignorant of examples either; they can tell, and they can believe
(140).

Setting to one side Geach’s charge and the question of how to

respond to it (I shall come back to this in Section 3.2), this statement
perfectly illustrates the account. What the account intends is that, for
Plato, it is possible to believe that something is the case, and to do
so with some justification (or credibility, or reliability), without having
full and proper knowledge – without, as Santas says, having
knowledge ‘in some Platonic sense of “knowledge”’.
I want to argue that, if the argument of the first part of the paper
is correct, and there is in early Plato aporia-based argument with a
sceptical consequence that is directed not only against knowledge
but also against belief, then PP2, and consequently this account of
Plato’s early epistemology (which is the conjunction of PP1 and
PP2), is mistaken.
Finally, in the third part of the paper (Section 3) I consider, and
reply to, some objections to this criticism of a common account of
Plato’s early epistemology.
The argument of this paper will, I hope, be of interest to anyone
interested in: Plato’s early epistemology; Plato’s method of argument
and enquiry; the question whether there is a sceptical side to early
Plato; and the question whether, if there is a sceptical side to early
Plato, the scepticism is directed only against knowledge or also
against belief. The argument should be of relevance also for our
understanding of the New or Sceptical Academy, that is, the
sceptical turn that Plato’s school took under its director Arcesilaus
some eighty years after Plato’s death; and especially the question of
what the New Academy may have based itself on in the work of
Plato.3 Furthermore, it should be of relevance for our understanding
of a later director of the New Academy, Carneades. For Carneades
appears to have made a particular point of arguing, on grounds he
would have considered consistent with Plato’s thought, that whereas
knowledge proper may be beyond us, we can have credible belief;
and in this his position appears similar to the position ascribed to
Plato by the modern account that, if the argument of this paper is
correct, is mistaken.4
I do not believe the question whether PP1+2 provides a good
account of Plato’s early epistemology is of purely scholarly interest.
Reflecting on this question affects our image of Socrates as a
ceaseless and radical seeker of knowledge and wisdom, in life no
less than in discussion. The account of Plato’s early epistemology
that I am arguing against is, it seems to me, associated with a
particular image of Socrates. This image is exemplified in the
following account, defended by John Cooper, of the distinctively
Socratic search for knowledge and wisdom:

[according to Socrates] the final good condition of the soul – wisdom or

knowledge – is apparently unattainable by human beings, but progress
and improvement is always possible.
(Cooper 2007: 30)

This image of Socrates – it is the image of an optimistic if humble

rationalist – relies on, and it is a natural outgrowth of, the account of
Plato’s early epistemology that this paper argues against. In
particular, it relies on the view that, for Plato, we can be confident
that we are making progress in rendering our beliefs better justified,
and so can be confident that we are making progress towards
knowledge, while being in doubt about the attainability of the
knowledge that we are after. If our argument of this paper is correct,
then this view is questionable. What is true is something that, to my
mind, is not only different and incompatible, but antithetical. For
Plato there are cases, and cases that are of very particular and
special interest to the philosopher, where belief must, even to be
minimally justified (or reliable, or credible), be based on knowledge,
and knowledge of the most demanding kind.
 A Preliminary Point
Let me begin with a worry about the way in which this account of
Plato’s early epistemology (i.e. PP1+2) has been defended by critics
since it was introduced in the late 1960s and early 1970s. To my
knowledge, none of the critics who have defended this account have
considered the twofold question:

Whether there is aporia-based argument in Plato’s early dialogues; and, if

there is, whether this is compatible with this account (i.e. PP1+2) of
Plato’s early epistemology.

Critics who have defended the account either have not given
consideration to this question or have expressed confidence that
there is not aporia-based argument in Plato’s early dialogues. The
reason why I think this omission is worrisome, is not that I am
assuming an answer to either of these questions. I have defended
an affirmative answer to the former question on a number of
occasions (since Politis 2006; 2008, 2012a, 2012b and 2015); and in
the present paper I shall defend a negative answer to the latter
question. My point is that, unless and until one takes up and
addresses this twofold question, this account of Plato’s early
epistemology is not properly available.
One reason for this worry is that there is an old and venerable
tradition – we generally refer to it as the New or Sceptical Academy
– which prompts this twofold question and in which this question, or
at any rate closely related questions, loom large. In a classic paper
on the issue of Plato’s scepticism, Julia Annas writes:
 The two most interesting arguments [for the view that Plato is a sceptic]
are the two that Cicero and Anonymous share. Of these the more
surprising is the argument that Plato is a sceptic because he often argues
to establish both sides of an issue. What is in question is a familiar
sceptical strategy. The sceptic picks on the interlocutor’s rash assertion
that something is F. He argues convincingly against its being F. Then he
argues equally convincingly for its being F. The interlocutor is thus
brought to a state of ‘equipollence’ (isostheneia): every ground for holding
it to be F is matched by an equally strong ground for holding it to be not-
F.
(Annas 1992: 65–6)

Annas is confident that this is not at all a credible reading of Plato

and the early dialogues. Practically all she says on the matter is this:
‘But how could anyone ascribe this mode of arguing to Plato?’ (66) I
confess I do not follow. If such critics as Cicero and Anonymous, and
the tradition they are summing up and, to a greater or lesser extent,
representing, think that there is dilemma-based (and hence aporia-
based) argument in Plato, and if they associate this with a sceptical
thrust in Plato, then the one thing we need to do, if we want to
consider whether this sceptical reading of Plato and the early
dialogues is credible, is examine whether there is dilemma-based
argument in the early dialogues, and, if there is, whether it is
associated with a sceptical thrust in them.5
 1 The Inference from Aporia-Based Argument
to a Sceptical Conclusion
We may begin with the inference from the supposition that there are
two-sided questions that articulate aporiai, to the, notably sceptical,
conclusion that there is good reason to think that we cannot know
the answer to such questions and cannot have more reason to
believe, or be more justified in believing, that one answer is true than
in believing that a contrary or contradictory answer is true. I want to
argue that, whereas, in general, such an inference is of questionable
validity, Plato is, for particular reasons, validly and cogently
committed to it.
In general, such an inference is of questionable validity. An
aporia may be difficult to resolve; it may be as difficult to resolve as
anything. It does not follow that it cannot be resolved, or that there is
good reason to think that it cannot be resolved. Heraclitus is
instructive on this point, when he says that we may hold out hope,
may even hope against hope, of finding that which is aporon and far
from easy to search for – Try harder!, as the common exhortation
has it. What Heraclitus says is this: ‘Unless one hopes for that which
is not to be hoped for (anelpiston), one shall not find it (ouk
exeurēsei). For it is hard to search for (anexereunēton) and to reach
through to (aporon)’ (fragment DK18). This, I would like to think, fits
Plato remarkably well. There is the idea that truth is not apparent or
evident but requires searching; and the idea that searching is hard,
and that it is hard because it is hard to reach through to that which
we are ultimately searching for. If anything, Plato goes further,
because he associates this aporetic dimension and moment of
enquiry with a sceptical thrust of enquiry – in doing so he makes
especially apposite Heraclitus’ appeal to the need for hope.
Why think Plato is committed, validly and cogently, to the
inference from aporia to a sceptical conclusion? Not because he
thinks that such aporiai cannot be resolved; certainly not, since he
goes out of his way, heroically as it were, to argue that there is a way
of resolving them. The one and only means of resolving such aporiai,
he argues, is to search for the definition proper and the essence of
the concepts that figure in the aporia. For example, the dialogue
Protagoras is, up until the last page, dedicated to demonstrating that
the question whether or not virtue can be taught articulates an aporia
and one that is especially intractable and difficult to resolve. Then,
on the last page, it is argued that the one and only way of resolving
this aporia and establishing whether or not virtue can be taught, is by
setting out to search for, and ultimately establish, that which virtue is.
(See Politis 2012a.)
The reason why I think that Plato is, validly and cogently,
committed to the inference from aporia to a sceptical conclusion, is
that I think that he thinks that there is good reason to think that such
aporiai cannot be resolved. The following inference is valid, and
evidently so:

If a question, Q, articulates an aporia, and if there is good reason to think

that this aporia cannot be resolved, then there is good reason to think
that we cannot know the answer to this question, Q, and cannot have
 more reason to believe, or be more justified in believing, that one answer
is true than in believing that a contrary or contradictory answer is true.

The question, therefore, is whether Plato thinks that there is good

reason to think that such aporiai cannot be resolved.
The Protagoras provides evidence for thinking that Plato thinks
that there is good reason to think that certain aporiai cannot be
resolved. The aporiai here include those articulated by the question
whether or not virtue can be taught, as well as the following threefold
question: whether there are several virtues or only one; whether, if
there are several virtues, they substantially resemble each other;
and whether, if there are several virtues, it is possible to possess one
without possessing all. Up until the last page, the dialogue is
dedicated to demonstrating that these questions articulate aporiai
and ones that remain intractable and difficult to resolve even on
extensive and dedicated enquiry. Then, on the last page, it is argued
that the one and only way of resolving such aporiai, is by setting out
to search for, and ultimately establish, that which virtue is.
Perhaps it is an overstatement to say that Socrates argues for
this claim; it might be more accurate to say that he simply asserts it.
But I think there is an important piece of argument in this passage.
At 361c2–d2 Socrates asserts that establishing that which virtue is,
is necessary for establishing whether or not virtue can be taught.
This, he asserts, is necessary for overcoming the utterly unstable
outcome that, as he has just explained, is the present outcome of
their arguments in the dialogue. The present outcome of their
arguments in the dialogue is utterly unstable (cf. panta tauta
kathorōn anō katō tarattomena deinōs, 361c2–3), because they, he
and Protagoras, have argued on both sides of the question whether
or not virtue can be taught; they have done so each of them
individually and against themselves, and not only against each other.
Does this passage contain an argument for the assertion which
says that establishing that which virtue is, is necessary for
establishing whether or not virtue can be taught? Socrates provides
an argument, when he says that, if we do not undertake a renewed
enquiry into whether or not virtue can be taught, and if this renewed
enquiry is not based on the supposition that the answer to the
question whether or not virtue can be taught must be established on
the basis of establishing that which virtue is, then we will be at risk
that our enquiry into whether or not virtue can be taught will utterly
fail not only this once – its present outcome has turned out utterly
unstable – but ‘multiple times’ (pollakis, 361c7).
The implication and force of this argument is this:

A. We cannot be confident that, however extensively we may enquire into

whether or not virtue can be taught, we will find more reason to think that
it can be taught than to think that it cannot be taught, unless we suppose
that the answer to the question whether or not virtue can be taught must
be established on the basis of establishing that which virtue is.

Consider now the first part of this claim:

A-1. We cannot be confident that, however extensively we enquire into

whether or not virtue can be taught, we will find more reason to think that
it can be taught than to think that it cannot be taught, ….

I think it will be admitted that this implies the claim that:

 B. There is good, but inconclusive, reason to think that the aporia
articulated by the question whether or not virtue can be taught cannot be
resolved.

Clearly, A implies that there is good reason to think that this aporia
cannot be resolved. Does A determine whether the reason is
conclusive or inconclusive, or does it leave this open? Clearly, if we
formulate A as we have done, that is, as having the epistemic force
‘We cannot be confident that p’, then the reason in B will be an
inconclusive reason. The alternative would be to formulate A as
having the stronger epistemic force, that is, as saying that:

A-strong. We can be confident that, however extensively we enquire into

whether or not virtue can be taught, we will not find more reason to think
that it can be taught than to think that it cannot be taught.

In that case, the reason in B would be a conclusive reason to think

that the aporia articulated by the question whether or not virtue can
be taught cannot be resolved. But A-strong is ruled out as a reading
of Plato, because it is incompatible with the second part of Plato’s
claim; it says that:

A-2. … unless we suppose that the answer to the question whether or not
virtue can be taught must be established on the basis of establishing that
which virtue is.

For whereas A-2 implies that it may be possible to resolve the

conflict of reasons, or aporia, A-strong implies that this is not
possible.
 May we conclude that, on the basis of the Protagoras at any
rate, we have good reason to ascribe to Plato the view that:

B. With regard to some aporiai, there is good, but inconclusive, reason to

think that they cannot be resolved?

Not yet. It depends on what we understand to be the force of the

central claim at the end of the dialogue, which is that:

D. The one and only way of establishing whether or not virtue can be
taught is by establishing that which virtue is.

(This claim is, of course, a particular formulation of the thesis that

(some) knowledge must be based on knowledge of definitions.)
Clearly, if D is understood in such a way as to imply that:

E-strong. We can be confident that even the most intractable aporia can
be resolved

Then D is incompatible with B. For whereas B says that there is

good, but inconclusive, reason to think that an aporia cannot be
resolved, E-strong implies that there is conclusive reason to think
that an aporia can be resolved. But I think it will be admitted that D
need not be understood in such a way as to imply E-strong; and that
this is not how D is understood by Plato. As it is understood by Plato,
what D implies is, rather:

E-weak. We can be confident that even the most intractable aporia may
be capable of being resolved.
We can be confident that even the most intractable aporia may be
capable of being resolved, because we have a general way and
means of seeking for a resolution of such aporiai, and because we
know what the resolution would look like, and what it would be based
on, if we found it. The general way and means of seeking for a
resolution of such aporiai is to search for the relevant definitions and
essences.
But we cannot be confident, of any particular aporia, that it can
be resolved; not, that is, unless and until we have actually resolved
it, and done so, precisely, by having found the relevant definition or
definitions. This is because what Plato’s way and means of aporia-
resolution provides, is a target to aim at; and a target such that, IF
we can attain it, then we shall be able to resolve the aporia. I think it
will be admitted that, for Plato, this is a big ‘if’. Plato does not provide
a general reason to think that, for any ti esti question that we may
ask and any definition that we may request, we can be confident that
we can find the ti esti and the definition. On the contrary, it is well
familiar that in none of these dialogues are the enquirers presented
as ultimately finding the definition that they are searching for, and
that in many of the dialogues they are, on the contrary, presented as
failing to find it. This strongly suggests that Plato does not think that
we can be confident, in advance of particular enquiry, that we can
find any and every definition that we may set out to search for, or can
resolve any and every aporia that we may come up against.
I conclude that, on the basis of the Protagoras at any rate, we
have good reason to ascribe to Plato the view that:
 B. With regard to some aporiai, there is good, but inconclusive, reason to
think that they cannot be resolved.

We may conclude that, certainly on the evidence of the Protagoras,

Plato is committed to the inference from the supposition that there
are two-sided questions that articulate aporiai, to the, notably
sceptical, conclusion which says that there is good reason to think
that we cannot know the answer to such questions and cannot have
more reason to believe, or be more justified in believing, that one
answer is true than in believing that a contrary answer is true.
Which aporiai are they? Presumably, they are those aporiai that
require, for their resolution, the establishment of the relevant
definitions. At any rate, it should be clear that, according to Plato:

F. If an aporia is such that there is good reason to think that it cannot be

resolved, then the one and only way of resolving it is by establishing the
definitions of the concepts that figure in it.

Do these conclusions generalise from the Protagoras to other early

dialogues? I am inclined to think that they do, and I have argued for
this elsewhere (see Politis, 2015, chs. 4, 6 and 7). We may conclude
that the aporia-based argument in Plato’s early dialogues implies
that there is a substantial sceptical thrust in these dialogues, and
that this scepticism is directed not only against knowledge but also
against belief.
 2 Consequences for a Certain Account of
Plato’s Early Epistemology
Currently accepted by a number of critics is the following account of
Plato’s early epistemology:

PP1. There is a distinction in the dialogues between knowledge proper

and a lesser cognitive state that is still reliable; and
PP2. The thesis which says that knowledge must be based in
definitions is intended by Plato as a thesis only about knowledge, and not
also about reliable belief.

This account has two notable consequences: that, for Plato, we can
have reliable belief without having knowledge; and that, for Plato, we
do not need to know definitions for having reliable beliefs.
If we ask, What is the supposed distinction in Plato, between
knowledge and a lesser cognitive state that is still reliable?, I do not
think a single answer can easily be given. For it depends on how the
critic that defends this account of Plato’s early epistemology
conceives of the distinction, and there are many and important
differences in how critics conceive of it.6 If our task were to consider
whether, and if so, how, Plato is committed to a distinction between
knowledge proper and lesser cognitive state that is still reliable, then
we would need to consider the various different ways in which critics
conceive of this distinction. This is a task addressed to PP1. But this
is not our present task, which is, rather, addressed to PP2. Our task
is to demonstrate that, if our conclusion so far is correct (as
defended in Section 1; it says that ‘The aporia-based argument in
Plato’s early dialogues implies that there is a substantial sceptical
thrust in these dialogues, and that this scepticism is directed not only
against knowledge but also against belief’), then PP2 is false; and it
is false irrespective of how the distinction is conceived between
knowledge proper and a lesser cognitive state that is still reliable.
If there are aporiai in Plato’s early dialogues, and if they are
considered by Plato to be so strong that it is not even clear whether
or not they can be resolved, and if a function of the claim that
knowledge must be based in definitions is to provide a way and a
means of resolving such aporiai, then two important things follow:
first, what these aporiai render questionable is not only knowledge,
on a particular and especially demanding conception of knowledge,
but belief that claims any justification, or credibility, or reliability; and,
secondly, and contra PP2 and hence contra PP1+2, the thesis which
says that knowledge must be based in definitions is intended by
Plato as a thesis not only about knowledge, but also about belief that
claims any justification, or credibility, or reliability.
This is a stark choice, between, on the one hand, the aporia-
based account of Plato’s method of argument in the early dialogues
and, on the other hand, this account (i.e. PP1+2) of Plato’s early
epistemology. From the point of view of the aporia-based account of
Plato’s method of argument, PP2, and hence this account of Plato’s
early epistemology, must be rejected. On the other hand, from the
point of view of that account of Plato’s early epistemology, something
in the aporia-based account of Plato’s method of argument in the
early dialogues must be rejected; this being either the premise of the
presence of aporia-based argument in these dialogues or the
inference to the presence of a strong sceptical claim, that is, a
scepticism that is directed not only against knowledge but also
against justified (or credible, or reliable) belief.
I am not claiming that, and it is not a consequence of what I
have argued that, for Plato, all belief must, to be minimally justified
(or reliable, or credible), be based in definitions. What follows from
what I have argued is that, for Plato, some belief must, to be
minimally justified (or reliable, or credible), be based in definitions.
What belief? Evidently, it is the belief that is rendered questionable
by certain, sufficiently strong, aporiai. This is all that is needed to
refute PP2, and hence refute this account of Plato’s early
epistemology (i.e. PP1+2). For, as PP2 has been understood by its
proponents, it implies that: NO belief needs to be based in definitions
to have some justification (or reliability, or credibility).
Recall that this account of Plato’s early epistemology has two
notable consequences: that, for Plato, we can have reliable belief
without having knowledge; and that, for Plato, we do not need to
know definitions for having reliable beliefs. If our argument has been
correct, then each and both of these important claims, about Plato’s
early epistemology, are questionable. What is true is something that,
to my mind, is not only different and incompatible, but antithetical.
For Plato, there are cases, and cases that are of very particular and
special interest to the philosopher, where belief must, even to be
minimally justified (or reliable, or credible), be based on knowledge –
and knowledge of the most demanding kind, that is, knowledge
based in definitions.
 3 Queries and Responses
1. What, one may ask, about PP1? For present purposes, we may
leave PP1 be. Let it be the case that there is a distinction in the early
dialogues between knowledge proper and a lesser cognitive state
that is still reliable. For present purposes there is no reason to
question PP1. PP1 does not imply PP2. And to call into question this
account of Plato’s early epistemology, which is the conjunction of
PP1 and PP2, it is sufficient to question PP2. Whether there is such
a distinction in the early dialogues, between knowledge and a lesser
cognitive state that is still reliable, is a good, but difficult, question;
and it is not part of our present remit.7 For it concerns PP1, whereas
our argument is directed against PP2.
2. What, one may ask, about Geach’s charge? When this
account of Plato’s early epistemology (i.e. PP1+2) was introduced by
critics, in the late 1960s and early 1970s, it was introduced for the
purpose of answering Geach’s charge of 1966; and the account has
been invoked by critics at regular intervals over the past fifty years,
and invoked for this purpose. If we reject the account, then how can
we answer Geach?
Let me remind the reader of Geach’s charge, and how this
account of Plato’s early epistemology (i.e. PP1+2) provides a
response to it. Geach argued that, according to Plato, one cannot tell
whether an object, O, is an example of a quality, F, unless one
already knows the definition of F; and he concluded that this has the
unacceptable consequence that it is impossible to search for the
definition of F. The argument is formulated as follows by Geach:

If the parties to a discussion are agreed, broadly speaking, about the

application of a term [i.e. the examples], then they can set out to find a
criterion [i.e. the definition] for applying it that shall yield the agreed
application. On the other hand, if they are agreed on the criterion for
applying the term, then they can see whether this criterion justifies
predicating ‘T’ of a given example. But if there is no initial agreement
either on examples of things that are certainly T or on criteria for
predicating ‘T’, then the discussion is bound to be abortive.

(Geach 1966: 372)

It is an assumption in Geach’s argument, that searching for a

definition of T requires appealing to examples of things that are T;
and examples in which one is entitled to have some confidence that
they genuinely are examples of things that are T. But, Geach argued,
any such entitlement is undercut by Plato’s claim that knowledge
regarding T, including knowledge of any particular thing that it is T,
must be based on the definition of T; and hence requires that one is
already in possession of the knowledge of the definition.
Critics that introduced this account of Plato’s early epistemology
(i.e. PP1+2) for the purpose of answering Geach, shared with him
the assumption that searching for a definition of T requires appealing
to examples of things that are T, and examples in which one is
entitled to have some confidence that they genuinely are examples
of things that are T. But they denied Geach’s claim that any such
entitlement is undercut by Plato’s claim that knowledge regarding T,
including knowledge of any particular thing that it is T, must be based
on the definition of T. They did so, precisely, by invoking PP2. On
this view, we are entitled to appeal to examples of things that are T,
in the process of and for the purpose of searching for the definition of
T; and this is because we can have reliable belief, of a particular
thing, that it genuinely is an example of a thing that is T, even in the
absence of knowledge proper – which, according to Plato, we cannot
have unless and until we have established the definition of T.
If we reject PP2, then how can we answer Geach? I confess I
do not know how we can answer Geach without PP2, if it were
supposed, against PP2, that, for Plato, ALL belief must, to be
minimally justified or reliable, be based in definitions.8 But this
supposition, we saw, is not part of our argument against PP2. What
is part of our argument against PP2 is that, for Plato, certain beliefs
must, to be minimally justified or reliable or credible, be based in
definitions; namely, those beliefs that are rendered questionable by
sufficiently strong aporiai. It follows that we need not suppose (for
present purposes at any rate) that, for Plato, there is an absolute ban
on the appeal to examples of things that are T, in the process of and
for the purpose of searching for the definition of T. We need only
suppose that there is such a ban, if the examples in question are
rendered questionable by a sufficiently strong aporia.
What of this crux case? It is the case in which we are searching
for the definition of T, and any putative example of a thing that is T is
rendered questionable by an aporia regarding T. The Protagoras
offers a good illustration. Suppose we are searching for that which
virtue is. And suppose there is an aporia articulated by the question
whether or not virtue can be taught. Suppose, finally, this aporia is
sufficiently strong to render questionable not only of any putative
example of a teacher of virtue, such as Pericles, Protagoras or the
proverbial parent, whether he or she genuinely is such an example,
but also of any putative example of a virtuous person whether he or
she genuinely is such an example. How, in this case, can we search
for the definition of T, if we suppose that, for Plato, any claim not only
to knowing of a particular thing that it is an example of a thing that is
T, but also to having justified or reliable belief of a particular thing
that it is an example of a thing that is T, requires being in possession
of the knowledge of the definition of T?
We may suppose that Plato may think that, even in this crux
case, searching for the definition of T is possible, if we suppose that
he denies a supposition that both Geach and his critics share. This is
the supposition that searching for a definition of T requires appealing
to examples of things that are T, and examples in which one is
entitled to have some confidence that they genuinely are examples
of things that are T. We need not suppose that Plato denies this
supposition, in general and in all cases. All we need suppose is that
he denies it in the crux case. Furthermore, we need not suppose
that, in the crux case, the credibility that putative examples of things
that are T are thought to have, prior to and independently of the
aporia, is positively annulled by the aporia. What the aporia does is
render questionable this credibility, it does not annul it. This is
because the reasons that make up an aporia are, precisely,
inconclusive reasons. A person who recognises that an aporia
regarding T (e.g. the aporia whether or not virtue can be taught)
renders questionable any putative examples of things that are T,
does not, thereby, incur an obligation to put out of her mind her prior
beliefs about what things are regarded, prior to the aporia, as
examples of things that are T (e.g. teachers of virtue). The obligation
she incurs is to recognise that she cannot assign any credibility to
these examples independently of her engaging with the aporia.
This leaves us with a last question. What, in the crux case, will
the search for the definition of T base itself on, if not the appeal to
examples of things that are T? This, I admit, is a large and difficult
question. Part of the answer, at any rate, ought to be staring us in
the face; namely that, the one thing this search will base itself on in
this crux case is the engagement with, and attempt to resolve, the
aporia that lies at the root of the case. It will base itself on the
engagement with this aporia, both in its own right and in relation to
the apparent status, prior to and independently of the aporia, of the
examples it renders questionable.
3. What, one may ask, about Gorgias 508e6–09b1? This
passage has been invoked by critics in direct support of that account
of Plato’s early epistemology (i.e. PP1+2). Does it need to be read
like that? And is there a reading of it that is consistent with the
rejection of PP2, and with the rejection of PP2 for reasons based on
the presence of aporia-based argument in these dialogues? Here is
what Socrates says:

These things (tauta), which emerged to us as being thus and so (hēmin

houtō phanenta) earlier in our previous discussions (logoi) are, I’d say,
held down and bound by arguments (or reasons, logoi) of iron and
adamant, even if it’s rather rude to say so. So it would seem, anyhow.
And if you or someone more vigorous than you do not undo them, then
 anyone who says anything other than what I’m now saying cannot be
speaking well. And yet for my part, my statement (logos) is always the
same, namely, that I don’t know how these matters stand (tauta ouk oida
hopōs echei); but no one I’ve ever met, as in this case, can say anything
else without being ridiculous. And so, once again, I affirm that this is how
these matters stand (tithēmi tauta houtōs echein).
(Trans. Zeyl, revised)

Our present task is not a full interpretation of this passage (for this,
see Politis 2015, ch. 7.4). The question I want address is this. It is
indeed clear how this passage is read by the proponents of PP1+2.
It is read as saying that when Socrates says that he does not know
certain things (tauta ouk oida hopōs echei), what he means is that he
does not have full and proper knowledge of them, that is, definition-
based knowledge; and when he says that it has emerged that the
same things are thus and so, and that this has emerged through,
and on the basis of, iron and adamant arguments/reasons, this
shows that he thinks that one can have good reasons to believe that
p, even reasons of ‘iron and adamant’, without knowing the
definitions required to know these things fully and properly. The
question is whether this passage can be read with plausibility, if, as
we have done, one rejects PP2.
I confess that, if this passage is read in isolation, both from the
context in the Gorgias and from the context of the method of
argument and enquiry in a number of early dialogues, then I would
find it hard to argue that there is a different, and more plausible,
reading. If, on the other hand, we read this passage against the
background of an aporia-based account of Plato’s method of
argument in the early dialogues (including the Gorgias), then there
is, I think, an alternative reading that recommends itself. The point is
that, if a particular whether-or-not question articulates an aporia,
then it is possible, indeed in certain cases appropriate, when
considering the matter from the one side and from the reasons in its
favour, to say that there are iron and adamant reasons for p, and
also, when considering the matter from the other side and from the
reasons in its favour, to say that there are iron and adamant reasons
for not-p (or, for q, where it is supposed that p and q are
incompatible propositions).
It may be objected that there is surely something incongruent
about one’s saying that there are iron and adamant reasons for p, if
one thinks that there may well be iron and adamant reasons against
p. The appearance of incongruity is, I submit, due to a certain
misconception. If it is supposed that the claim that there are iron and
adamant reasons for p implies that the reasons against p, if there are
such, are of lesser force, then indeed there is an incongruity, indeed
inconsistency. But there need not be such an implication. If the
question, whether or not p, articulates an aporia, and an aporia of
sufficient strength, then there is no such implication. We may put this
point by saying that, if the question whether or not p articulates an
aporia, and an aporia of sufficient strength, then claiming great force
of the reasons on the one side does not imply a comparison with the
force of the reasons on the other.
Whether this reading of Gorgias 508e–09a recommends itself,
in the context of the passage in the dialogue, is a difficult question.
Let me, for present purposes, do no more than point out that,
certainly, such a reading recommends itself in the case of one
dialogue, the Protagoras and its ending; and arguably also of the
ending of the first book of the Republic. In both cases, Socrates has
out-argued the interlocutor, and the interlocutor has admitted that he
has been out-argued: Protagoras in the one case (see Protagoras
360e3–5), Thrasymachus in the other (see Republic I. 353e12). In
both cases, the reader is given the impression, I suppose
deliberately on Plato’s part, that Socrates really has offered very
strong reasons for his case – ‘reasons of iron and adamant’. In both
cases, Socrates directly goes on to rescind from and subvert this
impression, and does so by claiming that the issue between them
has not been resolved satisfactorily, because it has not been
resolved on the basis of the knowledge of the relevant definitions. In
the Protagoras, this moment of subversion is associated by Socrates
with the claim that he himself has offered what appear to be very
strong reasons – no less iron and adamant, as we might say – also
on the opposite side.
By way of conclusion, I want to suggest that, if we can recognise
this pattern of argument within a single person, Socrates, then we
can recognise a similar pattern across two people, be it Socrates
and Protagoras, Socrates and Thrasymachus, or Socrates and
Callicles. Making this analogy would, it is true, be based on the
supposition that, for Plato, the intrapersonal case, that is, the case of
a conflict of reasons within a single person and a single mind, is
basic, and the interpersonal case, that is, the case of a conflict of
reasons between different people, derivative.9 This is not the place
to take up this supposition (see Politis 2015, chs. 5 and 6). But I
cannot help being reminded of that remarkable device in the Hippias
Major, namely, the shadowy character of Socrates’ double, the
doppelganger. He is characterised as the one person who is not only
on most intimate terms with Socrates, but most intent on refuting
him. If we make this analogy, from the intrapersonal to the
interpersonal case, then we shall, in effect, be conceiving of these
characters, Socrates and his interlocutors, as competing reason-
giving and reason-based voices within Plato’s mind.

1 The view that Socrates’ arguments are purely or primarily ad hominem is

held by a number of critics. Thus Frede 1992b: xvi–xvii, who, after invoking
Vlastos’ account of the method of argument in these dialogues, says of the
arguments in the Protagoras: ‘The arguments do not so much refute a thesis
or establish its contradictory, as they refute a person by showing him to be
inconsistent and confused’; and ‘it is not the thesis, strictly speaking, but the
respondent who is refuted’ (xvii).

2 This account of a sceptical thrust in early Plato’s should not be confused

with the account defended by Forster 2006, with which it is, in fact,
incompatible. Forster bases his argument for a sceptical conclusion in early
Plato on the fact that attempts at providing definitions of ethical terms, and in
general of ta megista, are consistently presented as unsuccessful. I shall
argue, on the contrary, that the demand for definitions is part of an anti-
sceptical thrust in these dialogues. In general, I am doubtful of the view,
which is part of Forster’s argument and which says that Socrates’ profession
of ignorance, or his apparent inability to attain positive knowledge, amounts
to a sceptical thrust in those Platonic dialogues in which Socrates is thus
represented. For, as Allen drives home in Chapter 9 of this volume (185), ‘It
is one thing to say, … I do not know anything except this itself (that I do not
know anything, scil. else) … [i]t is quite another to think, as Cicero appears
to say that Socrates did, that nothing can be known, with one exception’.

3 See James Allen’s essay in Chapter 9 of this volume.

4 On Carneades, see esp. Dillon’s essay in Chapter 10 of this volume.

5 I note the salutary corrective provided by Allen in Chapter 9 of this volume:
‘[T]he New Academics championed a form of argument, in utramque partem
disputatio, as the best method for both enquiry and teaching. Like Aristotle
and, especially, Socrates, whom they justly regarded as pioneers in its use
….’ (178–9; emphasis added) Castagnoli likewise stresses the Platonic
provenance of aporia in the sense of arguing on both sides of a question.

6 See Anderson 1969: esp. 464. This article, which I have never seen
mentioned, is, as far as I am aware, the first example of this type of
response to Geach’s charge; i.e. the response which says that, for Plato,
knowing a definition is necessary only for ‘philosophical’ knowledge. The
same response is defended by Santas 1972: 136–41; see esp. 140, quoted
above. The accounts by Irwin 1977: 39–41, Vlastos 1994a [originally 1985]:
48ff.; 1994b [originally 1990], and Fine 1992, 2008 are now standard; and
though they differ amongst themselves, they all appeal to the distinction
between knowledge, in the sense of belief that is by its nature true, and
belief that, though it may be true and may have some justification, is not by
its nature true. Prior’s account (1998) is yet different, in that it associates this
response to Geach with a distinction in Plato between explanatory/scientific
knowledge and non-explanatory/non-scientific knowledge. Woodruff’s
account (1987, 1990) and likewise Reeve’s (1989) are different, again, in
that they associate this response to Geach with a distinction in Plato
between expert knowledge and non-expert knowledge.

7 On this question I would like to recommend Kanayama 2011.

8 How large is this concession, on my part, to the account of Plato’s early

epistemology that I am arguing against? This is a good question, though not
one I can take up here. But let me clarify what I intend the concession to be.
For I do not mean that I am confident that we cannot answer Geach, if we
suppose that, for Plato, all belief must, to be minimally justified or reliable,
be based in definitions. All I mean is that am I am not confident that we can
answer Geach, if we suppose this. I regard it as an open question whether,
as the upshot of the argument of the present paper, PP2 needs to be
significantly modified (from ‘NO belief must, to be minimally justified or
reliable, be based in definitions’ to ‘NOT ALL belief must, to be minimally
justified or reliable, be based in definitions’) or, on the contrary, it needs to
be rejected and replaced by its contrary (i.e. ‘ALL belief must, to be
minimally justified or reliable, be based in definitions’).
9 I recognise that this is a controversial supposition. If I understand their
view correctly, Castelnérac and Marion (2009) argue for the opposite priority:
interpersonal conflict of reasons is primary; intrapersonal conflict of reason
is ‘merely’ (67) an ‘internalisation’ (73, original emphasis) of the
interpersonal cases.
 Chapter 4
Aporia in Plato’s Parmenides
◈

Verity Harte
 1 Introduction
‘The Parmenides’, says Richard Robinson, ‘comes nearest of all
Plato’s works to being wholly methodological’ (1942: 178). Robinson
seems to me right to be struck by the extent of the Parmenides’
focus on method and, more particularly, by the way in which this
focus unifies the dialogue (1942: 176). Indeed, I would go further
than Robinson in seeing this unity of focus extend back to the
opening conversation of the dialogue proper, the conversation
between Socrates and Zeno, all too often treated merely as a
convenient excuse for Socrates’ introduction of forms.1 Attention to
this unity across the (reported) dialogue and the careful structure it
reveals involving the three reported conversations and their relations
to one another is one pay off, I shall argue, of attention to the
dialogue’s use of ‘aporia’ and its cognates.
There is need to be cautious, however, in how one approaches
this topic. The dialogue is sometimes characterised as partly or
wholly aporetic. Examples include Owen 1970, for whom the latter
part of the dialogue constitutes ‘the first systematic exercise in the
logic of aporematic and not demonstrative argument’ (89), and Allen
1997, who describes ‘the final result [of the dialogue as a whole] [as]
perfection of aporetic structure’ (111).2 But this characterisation of
the dialogue is not explicitly tied to consideration of the dialogue’s
own linguistic use of ‘aporia’ and cognates. Indeed, this topic – the
dialogue’s use of the relevant vocabulary and its own understanding
of aporia – seems to have been largely neglected.
 This point about the dialogue’s use of the vocabulary bears
some emphasis: Allen, for example, states that ‘Parmenides
consistently refers to his criticisms, in fact, as aporiai, perplexities,
rather than as refutations (e.g. at 129e, 130b, c, 135a)’ (1997: 110).
Though offered exempli gratia, Allen’s four sample citations in fact
constitute two thirds of the dialogue’s total usage of the vocabulary
of ‘aporia’ and cognates. Of Allen’s four, two (129e6 and 130c33) are
in the mouth of Socrates, not Parmenides, the first before
Parmenides has even begun to speak in the dialogue. The third
(130c7) is in the mouth of Parmenides, but simply picks up and
echoes Socrates’ immediately preceding use, which refers not to a
criticism put forward by Parmenides but to Socrates’ state of mind.
The fourth, 135a3, also in the mouth of Parmenides, refers not to his
own interrogative examination of forms, but to the putative response
of some third party hearing ‘someone mark off each form as
something itself’ (135a2–3). My own view is that at most one of the
dialogue’s six uses of ‘aporia’ and cognates could be said to refer to
one or more of Parmenides’ interrogative examinations of forms as
an aporia; ironically, the use in question is one of the six that Allen
does not cite here. Parmenides’ criticisms may yet be aporiai. But if
they are, there is no evidence in the dialogue to this effect in the
form of his ‘consistently’ referring to them as being so.
The relative scarcity of the dialogue’s use of ‘aporia’ and
cognates no doubt goes some way to explain the apparent lack of
interest in the use of the vocabulary in the dialogue, judged by the
(no doubt, imperfect) measure of its occurrence in the indices of
books devoted to the dialogue.4 But the occurrences, though few, do
turn up at – and contribute to marking – significant junctures in the
dialogue. Reflection on the vocabulary in the context of the passages
in which it occurs may thus be used to illuminate certain aspects of
the structure of the dialogue. Or so I shall argue, after first examining
the use of the relevant terms in the six places in which the
vocabulary occurs.
 2 The Passages
As is widely recognised, the term ‘aporia’, when used in an
intellectual context, can refer to a condition, being puzzled, or to the
kind of argumentative device that might provoke such a condition, a
puzzle. LSJ (sv aporeō 2) cite three Platonic passages for a use of
the cognate verb ‘aporeō’ to mean the act of producing such a
puzzle (Prt 324d, Sph 243b and Lg 799c), though of these three only
the occurrence in the Protagoras (Burnet’s 324e1–2) seems a clear
such case, underlined by the use of ‘aporia’ in the sense of ‘puzzle’
as cognate accusative; the Sophist occurrence (Duke et al.’s 243b8)
seems to me to point instead to the occurrence of the intellectual
condition, as the verb commonly may.
The Parmenides uses both the noun ‘aporia’ and the verb
‘aporeō’. Of the six uses of one or other, at least one use of the noun
clearly has the sense of puzzle and at least one the sense of
puzzlement; at least one use of the verb clearly has the sense of
being puzzled. The remaining three uses are less clear-cut.
 The Clear Cases

1 129e6
Towards the end of his opening conversation with Zeno, Socrates
summarises his response to the work of Zeno he has heard him
read. In their conversation, the work’s purpose has been clarified as
an attempt on Zeno’s part to support Parmenides by showing that
still more absurd consequences follow from Parmenides’ opponents’
supposition that things are many than those opponents take to follow
from Parmenides’ supposition that the all is one. The kind of absurd
consequences Zeno takes to follow from the supposition of many
things is illustrated at 127e2 by the claim that the very same things
are both like and unlike, which, Zeno says, is impossible. Socrates,
however, denies that this is impossible, or even problematic, given
certain assumptions.

So, if in the case of things of this sort – stones and sticks and the like –
someone attempts to show the same thing many and one, we shall say
he has shown something many and one, but not that the one is many nor
the many one; that he does not say anything marvelous, but just what all
may agree; if, by contrast, someone should first distinguish separately
forms, themselves by themselves, of the things I recently talked of, such
as likeness and unlikeness, plurality and the one, rest and change, and
all such, and then show these to be capable of mixing with and
separating amongst themselves, I would be marvelously amazed, Zeno,
he said. Whilst I think you have laboured at these matters very bravely, I
would, as I say, be much more amazed in this way, if someone were able
to exhibit this same aporia woven in manifold ways into the forms
 themselves, occurring in things grasped by reasoning in just the way you
recounted its occurrence amongst things seen.

(129d2–130a2)5

The aporia that Socrates refers to here is clearly a puzzle or

problem, in the sense of the content of something (reasonably)
puzzling, and not a state of mind. This is clear, amongst other
reasons, from the aporia being said to be interwoven ‘into the forms
themselves’; one can see how this might be true of a problematic or
puzzling condition, but not of a state of mind. No less clearly, the
aporia referred to here is a repeatable type; the aporia that a person
might show interwoven into the forms would be the ‘same aporia’
Zeno has recounted amongst things seen.
Though the aporia here is not itself a state of mind, its
occurrence is associated with a state of mind, here captured
especially in the language of ‘marvel’ or ‘wonder’ (thauma), but also
by other terms such as amazement (agamai). This state of mind is
context-sensitive: Socrates would be ‘marvelously amazed’ (129e3)
if the aporia were shown to occur amongst forms; there is, in his
view, nothing to be amazed at, nothing ‘marvelous’ (129d5) at Zeno’s
demonstration of its occurrence amongst sensibles.
It is difficult to be certain precisely what Socrates intends by the
aporia he mentions. Plausibly, however, he is referring specifically to
the apparent contradiction of the same things being both like and
unlike and others like it. That is, the aporia or puzzle which Zeno has
shown to occur amongst sensibles and which Socrates would find
amazing were it shown to occur amongst forms, involves the
compresent opposite condition of some F being G, perhaps in
combination with some G being F, where F and G are opposites
such as like and unlike, one and many and so on. This
characterisation of the aporia Socrates refers to is supported by the
fact that, in the run up to the passage quoted, as well as in the
passage itself, the candidate object of marvel or amazement is
repeatedly framed in terms of this compresent opposite condition.
For example, Socrates says that to show that the same things are
like and unlike each other in virtue of their participation in the forms,
likeness and unlikeness, would not be marvelous (thaumaston,
129b1), whereas to show that the likes themselves become unlike or
the unlikes like would be a wonder (teras, 129b2).
In sum, in his initial use of the term in the Parmenides, Socrates
talks of an aporia in the sense of a puzzle, exemplified by showing
compresent opposites – or a twin pair of such compresent opposites
– as belonging to objects in some domain, a situation with the
potential for inducing marvel and amazement.

2 130c3 and 130c7

If Socrates’ first use of ‘aporia’ in the Parmenides clearly refers to a
puzzle and not to an intellectual condition, his second no less clearly
points to the mental condition – his own – of being at a loss. It is
followed by a use of the verb ‘aporeō’, which must in context have
exactly the same meaning.
The passage occurs early in the conversation between
Parmenides and Socrates that follows the initial conversation
between Socrates and Zeno. As Socrates had interrogated Zeno on
aspects of the work he had read, so Parmenides interrogates
Socrates on a key point in Socrates’ explanation of why Zeno’s
demonstration of the occurrence amongst sensibles of the aporia
characterised above does not amaze him, his distinction of forms of
likeness, unlikeness and so on from like-named sensibles that
participate in them. The second and third of the dialogue’s six
occurrences of ‘aporia’ and cognates turn up in the course of
Parmenides’ initial questions as to the forms that Socrates takes
there to be beyond those introduced or implied in the course of his
conversation with Zeno.

What about this? [Do you think there is] a form of man separate from us
and from all who are such as we are, some form itself of man or of fire or
indeed of water?
Parmenides, he [Socrates] said, I have often been in aporia about
these, whether one must speak as one does regarding those or in a
different way.
And are you also at a loss (aporeis) about the following things too,
things which might seem ridiculous indeed, such as hair and clay and dirt
or any other thing that is of least value and most trivial, whether one
should say there is a separate form of each, being different in turn from
the things we grasp with our hands or whether one should not?
Not at all, said Socrates. Rather, these things indeed are these very
things we see; to think that there is a form of these things would surely go
too far in absurdity. And yet there have been times when they troubled
me lest the same should be the case regarding everything. And then
whenever I settle on this point, I turn tail and flee, fearing lest I should fall
into some abyss of nonsense and be destroyed. So, arriving at this point,
 I spend my time focusing on those things that we just now said to have
forms.
(130c1–d9)

Clearly enough, Socrates’ use of the noun and Parmenides’ use of

the verb must be treated in parallel. Equally clearly, Socrates is using
the noun ‘aporia’ to describe some mental condition of his, a
condition of some uncertainty. He finds himself in this condition
regarding the question of whether there are such forms as a form of
man, fire or water. He says that he is not in a similar condition
regarding the question of whether there are such forms as a form of
hair, clay or dirt, though in what he goes on to say we see him waver
on this.
Does the use of aporetic language in this passage indicate that
Parmenides’ questions about forms here and in what follows are to
be thought of as (raising) aporiai, as Allen may think given his
citation of this passage in characterising them so? It seems to me
clear it does not indicate this (whether or not there are independent
reasons to think of them so). First, it is not Parmenides’ act of
questioning Socrates as to whether there are forms of man or fire or
water that induces Socrates’ aporia, his condition of finding himself
at a loss. Socrates describes a condition in which he says he has
frequently found himself, in the past no less than the present.
Second, the question ‘Is there a form of man, of fire, of water and the
like?’ is not evidently in any way related to the specific, compresent
opposite form of puzzle that Socrates had earlier identified as an
aporia. Indeed, while raising a substantive question of philosophical
interest, it is not evidently raising a puzzle in any interesting sense of
the term. Were we to identify anything capable of inducing aporia in
the form of an intellectual condition as an aporia in the sense of a
puzzle, the term aporia understood as puzzle would, I submit,
become so broad in its compass as to become uninteresting.
It is consistent with the, possibly deflationary reading of this
passage that I have offered that its use of ‘aporia’ and ‘aporeō’
prefigures the mental condition that Socrates will find himself in as a
result of each and all of Parmenides’ subsequent questions about his
position on forms. My point is simply that one cannot read off from
this that the dialogue means us to think of Parmenides’ questions as
(raising) aporiai in some specific and substantive sense. Does the
dialogue ever characterise them as doing so? The answer to this
question depends on a decision about the dialogue’s remaining three
occurrences of ‘aporia’ and cognates, the less clear cases.
 The Less Clear Cases

3 133a8 and 133b1

The last three occurrences of ‘aporia’ and cognates are all closely
related to one another, as we shall see, but the two that concern me
now occur in the same passage and in such a way as to require their
being taken closely together. They occur at what seems an important
moment of transition in the course of Parmenides’ interrogation of
Socrates’ views about forms.
From the beginning of that interrogation, Parmenides has been
focused on Socrates’ views, what Socrates thinks. For example,
Parmenides’ very first question is addressed to Socrates in the
second person singular as follows:

Tell me, did you yourself distinguish in the manner you’re saying certain
forms themselves separately, on the one hand, and, on the other in turn,
the things that have a share of them separately? And does it seem to you
that likeness itself exists separately from the likeness that we have, and
one and many and all the things you just heard from Zeno?

(130b1–5)

And this insistent focus on what Socrates thinks or says has

continued through a series of questions largely, though not
exclusively focused on the relation of participation that Socrates
takes to hold between sensibles and forms. See, for example, the
second person singulars in the following passages: 130e5 (‘does it
seem to you … ?’), 131c9 (‘are you willing, Socrates … ?’), 132a1 (‘I
suppose you think … ’), 132c10 (’does it seem to you … ?’).
The dialogue’s fourth and fifth occurrence of ‘aporia’ and
cognates, two occurrences of the noun, occur in a passage where
the distinctive move about which Parmenides first questioned
Socrates (130b1–5, quoted previously) is highlighted in the
production of aporia of a scale that Parmenides thinks escapes
Socrates. That distinctive move – marking off each form as
something separate or distinct – is still attributed to Socrates; see the
second person singular at 133b2. At the same time, the move is
generalised – the aporia arises if ‘someone’ (anyone, tis, 133a8)
makes this move.

So, do you see, Socrates, he [Parmenides] said how extensive is the

aporia if someone marks [things] off as forms that are themselves by
themselves? Indeed.
Further, know well, he said, that you do not scarcely at all yet grasp the
scale of the aporia, if you posit one form each for the things that are,
each time marking it off.

(133a8–b2)

In elaborating his point in what follows, Parmenides also marks a

difference in his own relation to the discussion. His subsequent
critical questioning is framed as an argument to a specific conclusion
announced at the start – that forms prove unknowable. But the
argument is not in Parmenides’ own voice; instead he gives voice to
the position of an unnamed objector (‘someone’, tis, at 133b4,
translated below). Indeed, far from presenting himself as the author
of the critical line of questioning he elaborates, Parmenides seems to
place himself on the side of the defender of forms. He does so both
when, famously, at the end of the lengthy subsequent discussion he
concludes by stating the difficulty this or another unnamed objector
would find themselves in if they responded to what has been said by
rejecting the existence of separate forms (135b5–c3), and also, at
the beginning of this discussion, when he characterises the
objector’s conclusion that forms are unknowable as based on the
proposition that forms are ‘such as we say they must be’ (133b5–6,
translated below).
In this way, the passage in which the fourth and fifth
occurrences of ‘aporia’ and cognates occur comes at a moment of
transition, even though the critical examination of forms is not yet
complete. Parmenides’ interrogation of Socrates’ proposal of
separate forms undergoes a marked transition from an individualised
interpersonal conversation between Parmenides and Socrates in
which Socrates is the author of a theory of forms critically examined
by Parmenides to a de-individualised conversation in which
Parmenides gives voice to the argument of an unnamed objector
and Socrates answers, not for himself only, but for anyone who
makes the move to mark off separate forms. That this is the
perspective from which Socrates is made to answer is made explicit
in the first stage of the subsequent consideration of the objector’s
position.

I suppose, Socrates, that you and anyone else who supposes that there
is some fundamental being (ousian) of each thing by itself would agree,
 first, that none of these is in us.
(133c3–5, my emphasis)

This framework of having Parmenides and Socrates come together,

as it were, to reflect on the nature of an unnamed objector’s
reasoning makes an important difference to how we should think
about the occurrences of aporia in this context. Here are those
occurrences again, in a slightly broader context.

So, do you see, Socrates, he [Parmenides] said how extensive is the

aporia if someone marks [things] off as forms that are themselves by
themselves? Indeed.
Further, know well, he said, that you do not scarcely at all yet grasp the
scale of the aporia, if you posit one form each for the things that are,
each time marking it off.
How so? He said.
There are many others, he said, but the greatest is this: if someone
should say forms aren’t such as to be known if they are such as we say
they must be, one would not be able to display to the one who says this
that he is mistaken, unless the one disputing should happen to be
experienced in many things and not without natural talent, and willing to
follow the one making the display as they undertake very many distant
labours; rather, the one who compels them to be unknown would be
unconvinced.6
(133a8–c1)

Evidently enough, the two occurrences of aporia here require similar

treatment. But the point is complicated by the fact that they appear to
look in different directions. The first, at 133a8, closes out the
sequence of prior conversation focused on the nature of the
participation relation which concludes negatively immediately before
this passage as follows:

Then it is not in virtue of likeness that other things have a share of forms,
but some other means must be sought by which they may have a share.
It looks like it.

(133a5–7)

The second occurrence, by contrast, looks forward and will be

elaborated in what follows. The objection to forms elaborated in what
follows is sometimes characterised on the basis of this passage as
‘the greatest difficulty’.7 The difficulty in question is almost universally
characterised as consisting in the argument by which Parmenides
elaborates the view of the unnamed objector according to which
forms are unknowable (at least by human beings, as it turns out).8
But attention to the way in which Parmenides elaborates said
greatest difficulty at 133b4–c1 (translated above) suggests
something rather different.
The greatest difficulty is that should some unnamed objector
raise the objection detailed in what follows, according to which
forms, posited as separate and given Socrates’ and Parmenides’
characterisation of them, prove unknowable, it will not be possible for
anyone to demonstrate to that objector that they are mistaken; it will
not be possible unless that objector should happen to be a person of
wide experience, natural talent and considerable patience for the
labours ahead.9 Just as Parmenides later concludes the elaboration
of the objector’s position by drawing attention to the serious
consequences that would follow if someone should respond by
rejecting the existence of separate forms, so Parmenides here
seems to assume that the objector is mistaken, no matter how
difficult the circumstances may make a demonstration of this fact.
What does this understanding of the greatest difficulty tell us
about the passage’s uses of ‘aporia’? When scholars refer to the
argument that follows as ‘the greatest difficulty’, ‘difficulty’ seems
intended as a translation of ‘aporia’ and an indication that the term is
understood in its meaning of puzzle rather than puzzlement. This
passage, though curiously not amongst those specifically cited by
Allen in making his claim to this effect, is the strongest (prima facie)
evidence for the dialogue using the noun ‘aporia’ in its sense of
puzzle as a way of characterising one or more of Parmenides’ critical
examinations of forms as raising aporiai. At least in its second
occurrence here, however, I have argued, the noun ‘aporia’,
whatever its meaning, does not refer to any puzzle raised by a line of
questioning about forms. It refers to the situation in which the
defender of forms finds him or herself in attempting to respond to the
line of questioning of a particular objector, especially in light of the
intellectual qualities of that objector. At most, then, the first,
backward-looking occurrence of the noun ‘aporia’ could be
understood to characterise the lines of critical questioning that
Parmenides has spelt out as, jointly or severally, raising aporiai.
However, if this is the point of Parmenides’ question to Socrates
at 133a8–9 (the first occurrence of the noun here), it is oddly put: he
asks Socrates whether he realises how extensive or how big is the
aporia, singular, if a person marks off separate forms. But it is not at
all clear there is some single puzzle of significant scale that has
been set before Socrates in the preceding conversation. Parmenides
has raised a series of different questions and, though several have
focused on the nature of the participation relation between forms and
sensibles, other aspects of Socrates’ view of forms have been in
focus as well. Nor is it obvious that the several lines of question that
Parmenides has pursued all turn on a single type of difficulty or
puzzle.10
It is easier to understand the first occurrence of the noun in our
passage, not in the sense of puzzle, but in the sense of puzzlement.
Parmenides asks Socrates whether he sees how extensive is the
difficulty for one who marks off separate forms; that is, he asks him
whether he sees the extent to which such a person will find himself
at a loss. Socrates says he does indeed, and no wonder: he has just
found himself to be considerably at a loss, as the result of
Parmenides’ interrogation of his position, having begun by making
the relevant move of marking off separate forms, emphasised in the
first question that Parmenides put to him (133c3–5, translated
previously).
Nevertheless, what follows the second occurrence of the noun
in the passage pulls against this reading of the first occurrence. After
Parmenides suggests that, in fact, Socrates does not yet realise the
extent of the aporia resulting from marking off forms, he suggests a
plurality of possible illustrations, of which he chooses the greatest.
One naturally individuates not puzzlements, but puzzles. Of course,
the neuter plurals in 133b4 (‘There are many others (polla … alla),
he said, but the greatest is this (megiston … tode) …’) are not to be
filled out by supplying the feminine noun ‘aporia’, but the sense that
Parmenides is articulating the aporia he has in view combines with
the plural to make the common view that Parmenides is elaborating
the greatest difficulty, that is, the greatest aporia in the sense of
puzzle.
On balance, the two occurrences of ‘aporia’ in this passage may
be best understood as pointing to the source of the intellectual
condition of being at a loss rather than that condition. Nevertheless,
their use to this effect is a far cry from the rather precise use of the
noun in the sense of puzzle that we met in the conversation between
Socrates and Zeno at the start of the dialogue. The vagueness is
understandable: the point of the passage here is to show that
Socrates does not in fact as yet understand the extent of the aporia
facing him, and this is in part, no doubt, because he does not yet
clearly understand its source or nature. Being vague in this way, the
use of the term seems wholly parasitic on the state of mind it induces
even if it does not refer to that state of mind directly. We might
loosely paraphrase the passage as one in which Parmenides asks
Socrates whether he has a grip on how extensive is the source of
intellectual difficulty for one who posits separate forms and then, in
suggesting that he does not, goes on to gesture towards the
existence of many such sources of which the greatest is the one he
spells out.
Understood in this way, we might see the exchange regarding
the extent of the aporia arising for one who posits forms as
reorienting the focus of Socrates’ attention in much the way in which,
I have argued, the passage in which it is situated reframes the shape
and character of the conversation about forms between Parmenides
and Zeno. The passage begins with Socrates’ recognition of himself
as being in aporia as a result of Parmenides’ critical questioning in
just the way Socrates earlier described having frequently found
himself in aporia when reflecting on such questions as the scope of
forms by himself. But in correcting Socrates’ claim that he does
indeed understand the extent of the difficulty the proponent of forms
is facing, Parmenides redirects his understanding of the source of
that difficulty, away from the problem of Socrates as individual
responding to a line of critical questioning in defence of his theory
and towards the difficulty of persuading an objector to forms that
they are mistaken.
The first use of the noun ‘aporia’ in the dialogue showed that the
dialogue could use the term to pick out a puzzle in a reasonably
formal way, where that puzzle takes a specifiable form. Thus far,
however, this remains the only such occurrence of the term in the
dialogue and the dialogue has not, as yet, provided an example in
which an argument or line of questioning put forward by Parmenides
– or by the unnamed objector of the ‘greatest difficulty’ – is explicitly
identified as raising an aporia in this sense.

4 135a3
The sixth and final occurrence of ‘aporia’ and cognates in the
dialogue, an occurrence of the verb ‘aporeō’, occurs in a passage at
the end of Parmenides’ exposition of the greatest difficulty and looks
directly back to the passage we have just considered.
 These [consequences], indeed, Socrates, Parmenides said, and very
many others still in addition to these the forms necessarily admit, if there
are forms themselves of the things there are and someone marks off
each as something itself. The result is that the one who hears this is at a
loss (aporein) and makes the objection that these don’t exist and that, if
they should exist as much as you like, there is much necessity that they
be unknown to human nature, and when he says these things he seems
to have a point and, as we were saying just now, is amazingly difficult to
convince. It would call for a man of great natural talent to be able to
understand that there is some kind of each thing and being itself by itself;
and for one still more amazing to discover and be capable of teaching
another having elucidated all these matters adequately.
(134e9–b2)

I have translated the verb as pointing to the state of mind of being at

a loss, and not to the act of raising a puzzle. In the context, the work
that might be done by the latter interpretation seems covered by the
second verb (conjoined by te … kai, so not a candidate for
epexegesis), here translated ‘makes the objection’ (amphisbētein),
which objection is immediately spelt out.
The passage looks back to 133b4–c1, translated and discussed
previously, and is closely tied to it conceptually and linguistically.11
The intellectual difficulty or aporia that is mentioned, however, is
located in an importantly different place. The aporia that Parmenides
elaborated in the greatest difficulty passage, I argued, was a
difficulty for the proponent of forms in persuading a particular sort of
objector that he, the objector, is mistaken. Here, the aporia belongs
to that objector. The objector encounters someone making the same
distinctive move that was earlier highlighted – saying that there are
forms and marking each off by itself (horieitai, 135a2)12 – they also
encounter the kinds of claims about forms that have resulted from
Parmenides’ lines of critical questioning. The result is that they are at
a loss. In this, their position resembles Socrates’ position earlier,
except that the objector is not himself the person who has put
forward the proposal of separate forms. Being at a loss – perhaps as
a result of being at a loss – the objector makes their objection: they
deny that there are forms or, at the very least, that any forms there
may be can be known by human beings.
As before, Parmenides not only stresses how difficult such an
objector is to convince, but explains this at least in part in terms of
the need for talent on the part of the objector. This I take to be the
point of his remark at 135a7–b1: ‘It would call for a man of great
natural talent to be able to understand that there is some kind of
each thing and being itself by itself’. While the reference to
understanding (mathein) could be taken generally, it seems clear in
context, particularly when balanced against the next remark, which
focuses on teaching (didaxai, 135b2), that Parmenides is focused on
a pedagogical context in which the objector takes the role of student
to a defender of forms. Parmenides then adds a new thought: as he
said before, talent is needed in the student, the one at a loss in the
face of the difficulties encountered following the marking off of forms
who responds by raising the objection that there are no forms or
none that can be known by us; still more talent, however, is needed
in the defender of forms who is ‘to discover and be capable of
teaching another having elucidated all these matters adequately’
(135b1–3). Lest one think all this talent might be expended on a wild
goose chase, Parmenides goes on to point to the serious problems
arising for anyone who, in the face of all this, proposes simply to
refuse the move of marking off forms (135b5–c3).
Looking at the dialogue’s six uses of ‘aporia’ and cognates – all
the uses that there are – the linguistic results are these: The noun
‘aporia’ is used in the dialogue in a precise way to refer to a specific
puzzle, but just once in this way. The noun is also used less
precisely to refer to whatever is the source of the mental condition of
being at a loss. The noun is also used to refer directly to that mental
condition. The verb ‘aporeō’ seems to be used consistently in the
dialogue to refer to the occurrence of the condition and not for the
act of raising a puzzle. Despite Allen’s claims to the contrary, there is
no clear example – in my view, there is no example – of the dialogue
explicitly characterising Parmenides’ criticisms of forms – or the one
attributed to the unnamed objector – as (raising) aporiai.
 3 The Dialogue’s Structure Seen through the
Lens of the Passages
I have argued that the dialogue’s use of ‘aporia’ and cognates does
not support the specific claim of Allen and others that Parmenides’
criticisms of forms are aporiai. More broadly, when the dialogue uses
this terminology, it is not to characterise the dialogue or portions
thereof as ‘aporetic’ as opposed to characterising specific moments
of aporia arising within it, largely, though not exclusively in the sense
of an intellectual difficulty. Nevertheless, I shall now argue, attention
to where the vocabulary of ‘aporia’ is used does give some indication
of what we – if not the dialogue – might refer to as an aporetic
structure. More generally, putting the passages in which the
terminology is used together and thinking about their relation to each
other gives an interesting perspective on the structure of the
Parmenides and reveals a progressively recurring pattern occurring
within it which continues through the end of the dialogue though the
aporetic vocabulary has by then entirely disappeared. At the same
time, attention to what happens in response to the aporia that
Socrates points to in the first occurrence of the vocabulary, the only
clear occasion of its use to pick out a formal puzzle, offers some
indications of ways one might respond when one finds oneself facing
such an aporia, patterns of response that generalise to the less
specific cases of finding oneself in aporia.
I begin with the latter point. At Parmenides 129e6, I have
argued, Socrates uses the noun ‘aporia’ in a relatively formal way to
refer to a precise and repeatable form of puzzle, which I identified as
the apparent contradiction resulting from the attribution of
compresent opposites to items in some relevant domain. An aporia,
so understood, should be distinguished from the referent of another,
methodologically important term, ‘hupothesis’ or ‘supposition’.13
‘Hupothesis’ is used twice in the immediate context, at 127d6–7 and
at 128d5, and several times elsewhere in the dialogue.14
The first passage, at the very beginning of Socrates’
conversation with Zeno, does not provide clear evidence as to what
the supposition is in context. Better evidence comes from the second
occurrence, at 128d5. Socrates has described Zeno as saying the
same thing as Parmenides (‘the all is one’, 128a8–b1), but disguising
this fact by putting it in a different way (‘there are not many’, 128b1–
2), thereby attempting to baffle and impress us. In describing what,
despite his acuity, Socrates has missed about his actual ambitions
for his treatise, Zeno describes himself as having aimed to pay back
the opponents of Parmenides who are characterised as holding a
certain hupothesis. This hupothesis is immediately specified as ‘if
there are many’ (128d5–6).
This is consistent with the earlier occurrence of ‘hupothesis’.
Having requested and heard the ‘first hupothesis of the first
argument’ repeated, Socrates had characterised Zeno’s position as
follows:

Is this what you are saying, Zeno? If the things that are are many, it is
necessary that these very things be both like and unlike; but this is
impossible, for unlike things cannot be like nor like things unlike. Isn’t this
what you say?
 (127e1–4)

Given the evidence of the second occurrence of hupothesis, this first

mention of the ‘first hupothesis’ may be understood as referring to
the hupothesis ‘if the things that are are many’ or, as subsequently
abbreviated, ‘if there are many’.
It is not clear whether the hupothesis attributed to Parmenides’
opponents takes the specifically conditional form ‘if there are many’
or consists in the thesis ‘there are many’ framed by Zeno as a
conditional. Evidence for the latter comes from Zeno’s contrast of
those opponents’ hupothesis to the hupothesis associated with
Parmenides, ‘the [hupothesis] of there being one’.15 This suggests
that a hupothesis is a thesis or claim of some sort, one that is not
merely asserted, but taken up in an argumentative context for the
derivation of certain conclusions and which may thus be framed as a
conditional.16
In taking up the hupothesis of Parmenides’ opponents, what
Zeno aimed to do – he reports – was to defend Parmenides by
showing those opponents that the same or greater absurdities are
consequent upon their hupothesis as they took to be consequent
upon that of Parmenides. Neither set of absurdities are specified, but
it is not too much of a stretch to imagine that the ‘impossibility’ that
Socrates earlier identified as key to Zeno’s establishment of the
denial that there are many is at least an example of such an
absurdity. This is the ‘impossibility’ of unlike things being like and like
things being unlike, which Zeno takes to be a consequence of
Parmenides’ opponents’ hupothesis of many things (127e3–4 and
127e6–8). Since this ‘impossibility’ is no less an example of the
apparently contradictory compresent opposite condition that
Socrates shortly identifies as an aporia in the sense of a formal
puzzle, this gives a clear indication of one possible relation between
a hupothesis and an aporia in this sense of formal puzzle. Such an
aporia is one kind of consequence that may be produced by the
argument from a hupothesis. Doing so is, in context, set in a
dialectical framework in which competing hupotheseis are under
consideration.
What dialectical function might the production of such an aporia
serve and how might one respond to it? The passage spends
considerable time examining Zeno’s purpose in producing an aporia:
first, Socrates characterises it; then Zeno himself does so, in part as
a correction to Socrates’ portrait. Both characterisations suggest that
the production of an aporia from a hupothesis can be used in a
dialectical situation to cause difficulties for the holder of that
hupothesis.
Socrates’ characterisation suggests that Zeno’s production of
the aporia takes the form of an argument against the hupothesis,
designed to establish its negation: ‘it is impossible that many are’
(127e7).17 However, while Zeno endorses Socrates’ account of the
intent of the work as a whole (128a2–3), his own clarification of his
ambitions for the work does not make explicit a goal of maintaining a
specific thesis that there are not many things; he frames his work as
a reflection of his youthful competitive spirit (philonikia, 128d7, e2)
and as aimed at ‘gainsaying’ (antilegein, 128d2) Parmenides’
opponents. They had ‘ridiculed’ (kōmō[i]dein, 128d1) Parmenides by
arguing that many absurd or ridiculous consequences (geloia,
128d1) stem from his hupothesis that there is one. Zeno responds by
showing that still more absurd or ridiculous consequences
(geloiotera, 128d5) stem from their counter hupothesis.
Although the opponents of Parmenides to whom Zeno responds
in producing the aporia Socrates characterises are not explicitly said
to have produced aporiai derived from Parmenides’ hupothesis, they
are said to have pointed to apparently ridiculous and contrary claims
deriving from it, arguments to which Zeno responds in kind. Zeno is
thus not only a producer of aporia but also one of the dialogue’s
examples of a respondent to aporia. His response indicates one
thing a respondent to aporia might do: a respondent who finds
themselves facing an aporia in a dialectical context with an opponent
who holds a competing hupothesis may respond by gainsaying that
opponent by countering and going one better than them in the
production of aporia. Call this a strategy of proliferation.
Socrates’ response to Zeno (128e5–130a2) illustrates a different
strategy of response to an aporia. Call it a strategy of
neutralisation.18 Socrates’ specific neutralisation strategy might
seem surprising in light of alternative ways one might respond to an
argument. He does not evidently identify any flaw in the arguments
that Zeno has put forward to draw the aporetic conclusion from the
hupothesis. We are given no real insight into how Zeno’s arguments
worked: how, for example, Zeno drew from the hupothesis that there
are many things the conclusion that things that are like are unlike
and unlike things like. Nor does Socrates evidently reject the
argument’s conclusion(s). Instead he denies that such conclusion(s)
should trouble us, given certain assumptions. First, he assumes that
Zeno’s aporetic conclusion(s) apply to perceptible things. Second, he
assumes – and invites Zeno to agree in assuming – that there exist
certain non-perceptible forms. Third, he assumes – and invites Zeno
to agree in assuming – that like-named perceptible objects stand in a
relation of participating in or having a share of these forms. Socrates
says that the fact of perceptible objects standing in this relation to
each of a pair of opposing forms demystifies, in a manner he does
not spell out, the apparent contradiction in which the aporia consists,
provided that this very same aporia is not shown to occur amongst
forms as a result of a similar compresence of opposites.
Socrates’ response to Zeno – his introduction of forms and
various claims about their relation to perceptible objects – is the
obvious trigger for Parmenides’ entry into the conversation of the
dialogue. Parmenides raises a series of questions regarding forms
and their relations to perceptible objects. However, if this is the only
role we give to this conversation, we miss an opportunity to see the
ways in which the structure mapped out in the opening conversation
between Socrates and Zeno, connected to the consideration of a
hupothesis, the production of an aporia, here in the sense of formal
puzzle, and possible strategies of response to that aporia, maps out
a structure that is at least partly repeated and extended in the
conversations that follow in the remainder of the dialogue.
This is one reason it is important to replace the usual division of
the framed dialogue into two, uneven parts, and instead to see the
framed dialogue as made up of three, uneven conversations: the
conversation between Socrates and Zeno, the conversation between
Parmenides and Socrates, and the ‘conversation’ – admittedly
stretching the term a bit – between Aristotle and Parmenides. Each
conversation centrally features at least one hupothesis and its
examination.
The opening conversation between Socrates and Zeno, despite
its brevity, in fact features two – or even three – hupotheseis: there is
the hupothesis attributed to the opponents of Parmenides, the
supposition of there being many things; there is also the position that
Socrates attributes to Zeno as the intended conclusion of Zeno’s
critical response to the opponents of Parmenides, that it is
impossible that there be many things, taken to be a disguised form of
Parmenides’ hupothesis, the supposition of one being. Socrates’
critical examination of the latter Zenonian hupothesis takes the form
of an attempt to neutralise the aporia that Zeno takes to arise from
Parmenides’ opponents’ hupothesis. Socrates’ neutralisation
strategy turns on his own assumption of the existence of forms.
Though Socrates’ assumption of forms is nowhere explicitly
described as a hupothesis, it is attributed to him using a cognate
verb, when, towards the end of their conversation, Parmenides says:
‘Further, know well, he said, that you do not scarcely at all yet grasp
the scale of the aporia, if you posit (thēseis) one form each for the
things that are, each time marking it off’ (133a11–b2). It is also
explicitly framed in the conditional form associated with the earlier
formulations of hupotheseis, for example, at 135e1–3: ‘if there are
forms themselves of the things there are and someone marks off
each as something itself’. Structurally, the supposition that forms
exist plays the role of hupothesis in the conversation between
Socrates and Parmenides; Socrates – the critical examiner of the
hupothesis of the previous conversation – is now the author of the
hupothesis under critical examination by a new conversation partner,
Parmenides.
This same pattern – the critical examiner of the previous
conversation is the author of the hupothesis under examination in
the next conversation – is repeated in the third and final conversation
of the dialogue, between Parmenides and Aristotle. Parmenides is
explicit that his demonstration of the proposed method of training will
begin from his own hupothesis (137b3), though its manner of
investigating that hupothesis will involve the specific – and much
debated – structure mandated by his preceding account of that
method of training.19 Parmenides’ hupothesis is examined in a
conversation between Parmenides and Aristotle, but it is clearly
Parmenides himself who conducts the examination of his own
hupothesis.
Each of the dialogue’s three conversations thus involves the
examination of a hupothesis, which examinations form a relay, as the
examiner of one conversation becomes the author of the hupothesis
of the following conversation. In the first conversation, Zeno’s
reported examination of the hupothesis of Parmenides’ opponents
takes the form of the production of an aporia in the sense of a formal
puzzle. In the second conversation, Parmenides’ examination of
Socrates’ hupothesis of forms is not, I have argued, explicitly said to
take the form of the production of any formal aporiai. It is said to
result in various forms of aporia in the sense of intellectual difficulty.
There is the intellectual difficulty that Socrates finds himself in being
unable to answer Parmenides’ critical questions of him. There is the
intellectual difficulty of the unnamed objector who thinks that forms, if
they exist, must be unknowable. And there is the intellectual difficulty
that Parmenides describes as facing the defender of forms in
response to such an objector.
The third conversation between Parmenides and Aristotle is not
explicitly said to involve the production of aporia in the sense of a
formal puzzle. Nor is it said to, and nor does it evidently produce any
aporia in the sense of an intellectual difficulty in its participants,
Parmenides and Aristotle. Aristotle famously responds to
Parmenides’ summary of the extended series of arguments’
combined, apparently conflicting conclusions, with the simple
answer: ‘very true’ (166c5).
Nevertheless, this third conversation takes the dialogue circling
back to its first conversation and that conversation’s formal aporia, in
two ways. First, the third conversation is framed as an examination
of Parmenides’ hupothesis, which, in the first conversation, was the
target of ridicule by those opponents whose own hupothesis was
Zeno’s target. Second, the examination of Parmenides’ hupothesis in
the third conversation between Parmenides and Aristotle produces
results that mirror the structure of the aporia that Zeno produced
from Parmenides’ opponents’ hupothesis in his defence of
Parmenides’ hupothesis. A series of compresent opposites are
ascribed to the one and to the others or apparently contradictory
opposites are denied of the one and of the others on the basis of that
Parmenidean hupothesis; and, for good measure, the same set of
results is produced on the basis of the contradictory of that
hupothesis also.
In this way, the third conversation circles back, in the most
expansive way possible, to the challenge that Socrates posed in
neutralising the aporia that Zeno had earlier produced and in the
form that Zeno produced it. Socrates said that he would be amazed
if someone could show this same aporia interwoven amongst forms
themselves. While the entities referred to in the third conversation
between Parmenides and Aristotle are not officially said to be forms,
they are explicitly said to have the specific characteristics Socrates
ascribed to forms in his neutralisation of Zeno’s aporia. Parmenides’
comments at 135ed8–e4 look back to distinctions to which Socrates
had appealed in his earlier conversation with Zeno, for example, at
129e5–130a2.
Parmenides’ reproduction of the Zenonian aporia in the form
that Socrates had said he would be amazed to see it occur may be
thought an extension of the earlier Zenonian strategy of proliferation.
However, while readers of the Parmenides have very frequently
been amazed at this part of the dialogue, no such amazement is
registered in the course of the conversation between Aristotle and
Parmenides. This juxtaposition of prefigured amazement – which the
reader has been cued to imagine befalling Socrates silently listening
– and the lack of any such explicit amazement is, I now argue, part
of an overall goal suggested by the structure of the dialogue to
promote proliferation as a methodologically productive strategy. The
references to aporia in the sense of intellectual difficulty in the
dialogue’s second conversation can be seen in this light.
 In the opening conversation between Socrates and Zeno, I have
argued, we find exemplified two different strategies of response to
the production of an aporia in the sense of a formal puzzle arising
from a hupothesis: Zeno’s strategy of proliferation and Socrates’
strategy of neutralisation. Parmenides’ subsequent critical
examination of Socrates’ hupothesis of forms – the key to Socrates’
neutralisation of Zeno’s aporia – is not framed in terms of the
production of formal aporiai, but it induces various kinds of aporia in
the sense of intellectual puzzlement. It illustrates different responses
to such aporia as well.
Socrates continues to show a disposition towards neutralisation.
At least two of his critical interventions in the course of Parmenides’
critical questioning may be seen in this vein: (1) his proposal that
each form may indeed be, as a whole, in each spatially separated
perceptible participant, by comparison with the multi-location of a
day (131b3–6); (2) his proposal that there will be no endless
recurrence of forms of the same type, if forms are thoughts occurring
only in souls (132b3–6). Indeed, one might think, Parmenides’
questions continue until the point that Socrates gives up the attempt
to respond to his questions and acknowledges that he is in aporia,
intellectually at a loss (133a10).
At this point, where we find a cluster of three connected uses of
‘aporia’ and cognates, the three less-clear uses considered above,
the dialogue introduces, and, I argue, strongly discourages a third
possible strategy available to the proponent of a hupothesis who
finds himself in aporia, whether or not as the result of the production
of a formal puzzle: this is the strategy of surrender by abandonment
of the hupothesis. The strategy of surrender is attributed to the
unnamed objector of 133b4–6, who, faced with Socrates’ and
Parmenides’ characterisation of forms, denies that forms can be
known, and who, after the elaboration of the argument to show why
forms cannot be known, is said to object that there are no forms
(135a3–5).
Surrender by abandonment of his hupothesis of forms is one of
the obvious courses available to Socrates in the situation in which he
finds himself at this point in the conversation. It is, however, a course
from which Parmenides strongly discourages him in at least two
ways. First, he explains the intellectual difficulty in which Socrates
finds himself not as the intellectual difficulty of this objector, but as
the intellectual difficulty of explaining to such an objector that he is
mistaken on the assumption he is. Second, he makes explicit that no
less, indeed, arguably more serious difficulties threaten anyone who
does adopt the strategy of surrender by abandonment at this point.
Such a person is characterised at 135b5–c2 as having nowhere to
turn their thought and as having destroyed the power of dialectic or
even of conversation itself (dialegesthai, c2).20
The dialogue’s attitude to neutralisation seems less clear than
its attitude to surrender by abandonment. It is surely an available
option for dealing with puzzles and for situations in which one finds
oneself in aporia. But the dialogue’s overall structure points to the
fact that neutralisation generally involves the introduction of another,
potentially as vulnerable hupothesis and/or fails to succeed. Taken
as a whole, it is not neutralisation but proliferation that the dialogue
seems most clearly to favour.21 Parmenides points to the actual
intellectual difficulty in which Socrates finds himself as one of
explaining to the objector to forms the ways in which that objector is
mistaken; and the method of training he recommends to him is
exemplified in the examination of his own hupothesis that, as I have
argued, doubles down on the original Zenonian strategy of
proliferation while at the same suggesting there is nothing to be
amazed at in its results.
 4 Conclusions
In the preceding discussion, I have offered an interpretation of the
Parmenides’ relatively infrequent use of ‘aporia’ and cognates and
an indication of the methodological structure highlighted by attention
to the structure and location of this terminology’s occurrence. I now
suggest, though cannot fully defend, a number of different
conclusions that these results might be used to support, some
connected to the interpretation of the Parmenides, others to broader
questions about the history of aporia.
First, some dialogue-centred conclusions: To the extent that the
dialogue can be seen as offering, indirectly, a recommendation of the
strategy of proliferation, the dialogue’s methodological indications
might be taken as an endorsement of Zeno, the strategy’s initial
exponent in the dialogue. This runs counter to the view of Diès and
others22 that the dialogue is dismissive of Zeno as part of a bid to
elevate the stature of Parmenides. The view that the dialogue is
dismissive of Zeno may have indirectly contributed to the relative
neglect of the material in the opening conversation between
Socrates and Zeno beyond its introduction of forms. But this opening
conversation is in fact pivotal to the methodological import of the
dialogue; and the dialogue suggests as much when it has
Parmenides, prior to the start of his investigation of his own
hupothesis, explicitly identify the method of training he will thereby
exemplify as ‘the very one [Socrates] just heard from Zeno’ (135d8),
albeit with some important, signalled modifications.23
 Proliferation might be intended as destructive or productive of a
hupothesis. But the dialogue’s attitude towards the strategy of
surrender by abandonment, along with its presentation of
proliferation strongly suggests the strategy is regarded as being in
some way productive. In particular, the methodological structure
marked out, together with the interpretation I have offered of the
aporetic vocabulary in and around the ‘greatest difficulty’ passage,
provides some indirect argument that the dialogue expects the
hupothesis of separate forms to be maintained. This places my view
squarely on the side of those who hold that Plato does not abandon
forms. But it does not settle what would count as the hupothesis of
separate forms being maintained, nor indicate how specific lines of
questioning should be addressed, nor what, if any, role might be
played by Parmenides’ exemplification of the recommended method
of training in meeting this goal. Such questions fall outside the scope
of the present paper.
My lack of consideration of such questions of substance
regarding the dialogue is partly a function of scope (and space), and
partly a function of the paper’s focus on the methodological
character of the dialogue observed by Robinson, from whom I
began. Without wanting to suggest that the dialogue’s substantive
philosophical results are exclusively methodological, I have sought to
spell out in some detail some central aspects of its methodological
thinking.
Turning from narrow interpretative questions about the dialogue
to broader questions about the history of aporia, two conclusions
suggest themselves, one concerning aporia in Plato, the other, its
future beyond Plato. It is not news to say that, in his critical
examination of Socrates’ hupothesis of forms, Plato turns the tables
on his character, Socrates, by having Parmenides expose his
present inability fully to articulate or defend his own hupothesis and
reduce him to aporia. But careful attention to the source of aporia
that Parmenides directs Socrates’ focus towards in the ‘greatest
difficulty’ passage highlights a kind of de-individualisation of the
classic Socratic encounter.24
Such de-individualisation may itself be part of the Zenonian
strategy of proliferation. Socrates’ understanding of Zeno’s project
and Zeno’s clarification bring out that Zeno is the person who does
not strictly have a hupothesis of his own, but a variant of
Parmenides’, whom he supports. At the same time, Parmenides’
exemplification and extension of the Zenonian method of arguing
takes his own hupothesis as simply one amongst many in a lengthy
examination that is effectively conducted by himself. It is tempting to
connect this de-individualising move with the fact that Socrates
recedes from prominence in some, though by no means all of the
dialogues that might be agreed to be written post-Parmenides and
with the interest of several later dialogues in exploring positions not
specifically maintained or defended by any of the characters in the
dialogue in question.25
Finally, the methodological structure of the dialogue and the
strategy of proliferation that that structure appears to recommend
may be seen as advancing the philosophical fruitfulness of a method
that will later be associated explicitly with the production of aporiai in
the sense of specific, substantive puzzles, even if – I have argued –
the Parmenides does not itself generally use its aporetic vocabulary
in this way.26

1 Allen 1997: 75 is one notable exception, dividing the dialogue into not two,
but three unevenly sized parts.

2 Compare too Migliori 1990, for whom the dialogue as a whole is a ‘vera
piramide di aporie’ (121).

3 On the line numbering of Burnet 1901, there is no occurrence of ‘aporia’ or

cognate in 130b, but two in 130c. I assume these are the examples Allen
has in mind for ‘130b, c’.

4 No mention in the indices of five recent books in English on the dialogue:

Allen 1997, Miller 1986, Meinwald 1991, Sayre 1996, Rickless 2007. The
situation is similar in other modern languages: no mention in the index of
von Kutschera 1995; one of ‘apories’ in Brisson 1999, which follows Allen
1997 in describing Parmenides as referring to his interrogative questions as
aporiai, citing the same four passages. Scolnicov 2001 has multiple entries
under ‘aporia’. Like Allen, he routinely talks of Parmenides’ questions as
aporiai (e.g. Scolnicov 2001: 3, 7, 19), but despite extensive focus on the
notion of aporia his discussion is not linked to the dialogue’s use of the
vocabulary.

5 Except where noted, translations of the Parmenides are my own and follow
the text of Burnet 1901. I have profited from consulting and comparing the
translations of others.

6 With Burnet 1901 and all the MSS, I read all’ apithanos at 133b9–c1. von
Wilamowitz-Moellendorff 1920: v2, 226 n1 proposed alla pithanos;
unhelpfully, he suggests that no explanation is needed for the choice to read
it in this way. He is followed by Diès 1956, Moreschini 1966 and Brisson
1999: 260 with n92. Brisson’s note suggests the reading is favoured as
making clear that Parmenides takes the objector’s criticism to be defeasible.
But this seems no less true on the reading I have adopted. Cornford 1939:
95 n1 points to the parallel at 135a7 in support of the MSS reading.
7 See, for example, Gill 1996: 20 with n26, Peterson 1981 and Duncombe
2013.

8 An exception is Tamsin De Waal in her doctoral thesis (De Waal 2009).

9 I owe to De Waal 2009 the observation of the important role played by the
mystery objector in this passage and the observation that the stress in
133b7–c1 is on the need for that objector to have talent etc. The point is
worth underlining: Brisson, though he translates the passage (1990: 103) in
such a way as to make the objector in need of talent etc., nonetheless
summarises the passage (1990: 30) as though it were the defender of forms
in response who is here said to be in need of talent. In understanding ‘the
one who is disputing’ (ho amphisbētōn, 133b8), the person said to be in
need of talent, as the objector I am not guided only by this expression, which
could, after all, refer to the activity of the person in dispute with the objector.
However, given the men … de construction at 133b7–9, the ‘one who is
disputing’ must be the subject of etheloi, 133b8, the person who has to be
‘willing to follow the one making the display’ of the mistake of the objector to
forms as being mistaken, and this subject of etheloi must be the objector.
Additional support comes from the linguistic and other parallels between this
passage and 134e9–b2 (which I discuss later; see n11).

10 This point is consistent with thinking the various lines of question meet
with a single solution, a reading along the lines of Meinwald 1991, for
example.

11 ‘Amphisbētein’ at 135a3–4 looks back to ‘ho amphisbētōn’ at 133b8 and

further confirms my understanding of the person to whom that expression
refers (cf. note 9 above). The conclusion at 135a4–5 that forms must be
unknowable to human nature looks back to the conclusion at 133b4–6 that
forms won’t be fit to be known and ‘agnōsta’ at 135a5 looks back to its
occurrence to state the same conclusion at 133c1. The characterisation of
the person who makes this objection as ‘amazingly difficult to convince’
(135a6–7) looks back to the characterisation of the objector as liable to be
‘unconvinced’ (133c1) and the repetition of the point is explicitly noted (‘as
we were saying just now’, 135a6).

12 Compare ‘diorizētai’ and ‘aphorizomenos’ at 133a9 and b2; and ‘diērēsai’

with repeated ‘chōris’ at 130b1–5; looking ahead, note ‘horieitai’ at 135b7.
13 I pick this translation (a Latinate equivalent of the Greek term) for want of
a better and not to make a decision on the Greek term’s precise meaning, a
contested matter in general for Plato. For discussion and additional
bibliographical information, see Benson 2006. Scolnicov 2001 discusses the
use of ‘hupothesis’ in the Parmenides in the context of his discussion of
aporia. Rodriguez 2016 makes a convincing case against the relative
neglect of the Parmenides in scholarship on Plato’s use of ‘hupothesis’. I am
indebted to his work on its use in the dialogue.

14 Rodriguez 2016 demonstrates that ‘hupothesis’ is used more often – in

both absolute and relative frequency – in the Parmenides than in any other
Platonic dialogue.

15 The noun ‘hupothesis’ is left implicit at 128d6, but it is evident from the
context that this is the feminine noun to be supplied.

16 In English and Greek ‘hupothesis’ may refer to the act of supposing or to

the product of such an act. I am uncertain how clearly these are
distinguished in context. But the fact that we go on to focus on
consequences of a hupothesis, where these seem to be formal implications,
suggests the overriding focus is more on what is supposed than the act of
supposing itself.

17 The modality may simply reflect the fact that this is an intended proof of
the negation, not the modal status of the negative claim.

18 I choose ‘neutralisation’ as opposed to ‘resolution’, since the latter is a

success term and I am concerned to evaluate the form of different response-
strategies in a way that is at least initially independent of their argumentative
result. Resolution might be a species of neutralisation.

19 For detailed discussion of his description of the method of training, see

Meinwald 1991: 28–75, Sayre 2005 and Rodriguez 2016.

20 ‘Dialectic’ and ‘conversation’ are each possible translations of

‘dialegesthai’ here. Compare Gill and Ryan’s translation (Gill 1996, ad loc).
21 My focus is on what one might think of as the dialogue’s recommendation
for one’s initial response to finding oneself in aporia as the result of a
hupothesis. This does not mean that proliferation is an end in itself. Endless
proliferation might be regarded as counterproductive, though there is no
explicit sign of such a view in the Parmenides.

22 See Diès 1956: 3–4, 14–19; followed on the relative merits of Parmenides
and Zeno by, for example, Cornford 1939: 63. Allen 1997: 78 disputes the
view. Migliori 1990: 121 observes that Zeno’s methodology receives implied
endorsement in the dialogue.

23 Rodriguez 2016 rightly emphasises this point and also considers the
nature and significance of the modifications.

24 While this thought needs further articulation, I have in mind a move away
from arguments starting from a view held by some specific individual and in
which the benefit of the examination of the view might be thought in some
way tied to the individual who holds the view in question.

25 This practice is highlighted and characterised as ‘the new dialectic’ by

Brown 1998: 182. Compare too the ‘missing persons’ whose role in several
late Platonic dialogues is illuminatingly discussed in McCabe 2000.

26 My thanks to all the participants in Dublin for helpful comments and

discussion, to David Charles for written comments on an earlier draft, and to
Evan Rodriguez for much helpful discussion of methodology in the
Parmenides while I was advising his doctoral research.
 Chapter 5
Aporia in Plato’s Theaetetus and Sophist
◈

Lesley Brown

The chief aim of this essay is to examine the development of Plato’s

use of philosophical puzzles to guide his enquiries. Labelled aporiai,
they are prominent in Sophist, but already found in Theaetetus.
Section 2 identifies common features in such puzzles, and explores
how in Theaetetus they are presented but left unsolved. In both
dialogues the young Theaetetus is characterised as an ideal
interlocutor, quick to appreciate a philosophical puzzle, and to
respond appropriately. By these means Plato links the otherwise
very disparate dialogues: Theaetetus, a formally aporetic attempt to
define knowledge conducted by Socrates, and Sophist, whose new
protagonist, the Stranger from Elea, confidently announces results
both in the Outer Part’s search for the sophist and in solving the
problems of the Middle Part.1 Section 3 traces how the Sophist’s
Middle Part is explicitly structured around a series of philosophical
puzzles, and notes the plentiful terminology of aporia that signposts
this. Plato shows his readers the philosophical payoffs of a serious
attempt to diagnose the source of a given aporia: herein (I suggest)
lies the real difference between the sophist and the philosopher.
But first Section 1 explores the famous image in Theaetetus of
Socrates as a midwife, where Plato offers what I read as a new
approach to the respondent’s subjective aporia.
 1 Aporia as Labour Pains, Socrates as Midwife
I argue that with the midwife image Plato subtly transforms the
notion of a respondent’s subjective aporia and its benefits.
The Theaetetus resembles some earlier dialogues such as
Laches, Charmides and Euthyphro in having a formally aporetic
structure, as defined in Szaif’s essay. Socrates encourages the
young Theaetetus to answer the question ‘What is knowledge?’, and
by the end several proposed answers have been discussed and
rejected. In his closing remarks Socrates muses on the likely upshot
for Theaetetus: (a) he may in future conceive better theories thanks
to his unsuccessful enquiring, but (b) ‘if you stay barren, you’ll be
less burdensome to those who associate with you, and gentler,
because you’ll have the sense not to think you know things which in
fact you don’t know’ (210c). These remarks on the benefits of what
has been called elenctic aporia remind readers of similar remarks at
Prot 348c–d and Meno 86b–c. A further echo is found in the
discussion of the elenctic sophist at Soph 230b–c, discussed at the
end of this section. Both passages (Tht 210c, Soph 230b–c) recall
the state of subjective aporia that an interlocutor experiences as a
result of being cross-examined and their beliefs being found wanting.
But remark (b) above – though it would be appropriate to self-
confident respondents such as Meno and Euthyphro – is
incongruous used of Theaetetus. The youth has been portrayed from
the start as combining intellectual curiosity with extreme diffidence,
never thinking he knows things he doesn’t in fact know. What’s more,
Socrates had earlier diagnosed in him a rather different kind of
subjective aporia, one I call labour pain aporia.
Socrates’ famous comparison of himself to a midwife, able to
ease the intellectual labour pains of young men, comes in response
to Theaetetus’ reaction when asked by Socrates to sum up all kinds
of knowledge in a single account. Here’s the boy’s reply:

I’ve often set myself to think about it, when I’ve heard reports of your
questions. But I can’t convince myself that I have anything adequate to
say on my own account; and I haven’t been able to hear anyone else
saying the sort of thing you’re asking for. On the other hand I can’t stop
worrying about it either.
(148e1–6, tr. McDowell)

Though he uses no aporia term, Theaetetus is admitting that he is

perplexed, but without the familiar recriminations blaming Socrates,
and without any prolonged personal exposure to Socratic
questioning – more on this below. Socrates has already declared
that he himself is perplexed (aporō) about what knowledge is
(145e8), and will go on to emphasise that he is barren of wisdom
(150c4), another topos familiar from the early dialogues.
I divide Socrates’ response as follows: (1) you’re in labour
(ōdineis), Theaetetus, because you’re pregnant (enkumōn) (148e6–
7); (2) it’s a secret that I practise midwifery as my mother does; don’t
tell others (149a3–7); (3) though they and you have heard that I’m
very odd and that I make people aporein (suffer perplexity a8–9).2
Socrates elaborates on point (2), his hitherto secret midwife skills.
Female midwives can bring on labour with incantations, can help
women give birth, can procure miscarriage and act as matchmaker.
Socrates can do all that with young men who are pregnant and can
also test the resultant mental progeny to see if they are genuine or
so-called wind-eggs – as he will test the suggestion that knowledge
is perception (160e5ff). His young associates, like women in
childbirth, ‘feel the pains of labour and are full of aporia (aporias
apopimplantai) night and day … and my art can bring on that pain
and end it’(151a7–8) – the pain that Socrates has in effect equated
with (some kind of) aporia. Socrates promises to use his midwifery
skills on the pregnant Theaetetus, urging him not to be angry if
subsequent testing proves the offspring to be ‘an image and not true’
(151c4–5).
It is widely recognised that point (3): ‘men say I make people
aporein’ is designed to recall Meno’s remark to the same effect
(Meno 80a1–2, cf. 84b). It recalls episodes producing elenctic
aporia, the state of perplexity felt by previously self-confident
respondents after Socrates has examined and found inconsistencies
in their confident beliefs on a subject. But a rather different picture of
intellectual aporia is introduced by points (1) and (2): Socrates as
midwife to young men suffering what I’ve called ‘labour pain aporia’.
How should we understand this new role Socrates proclaims? As
Burnyeat notes, Robinson thought the midwife image a subterfuge of
(the later) Plato by which he suggests that what he had earlier
portrayed as Socrates’ negative or destructive procedure of elenchus
could after all yield positive results.3 But in reply we can note that all
along it was a hope that the elenchus could yield positive results,
and even when there is a negative outcome, when elenctic aporia
has been reached, potential good consequences should follow (see
Szaif’s chapter). Another suggestion, found in the Anonymous
Commentator and Cornford, is that the image of young men giving
birth under Socrates’ guidance to mental progeny is a reference to
recollection as the sequel to aporia (demonstrated in the
examination of Meno’s slave).4 However it should be interpreted, the
image of Socrates as a midwife, able to bring on and dispel
intellectual birth pangs, is a striking and novel one.
As noted, the phrase ‘You’ve heard I make people aporein’ is
designed to remind the reader of earlier dialogues, just like the
closing remarks of Theaetetus quoted above. But the kind of aporia
Socrates now characterises as intellectual labour pains is rather
different from the elenctic aporia that an interlocutor suffers after
cross-examination of his beliefs. This can be seen from the following
matrix, where → represents temporal succession.

Elenctic aporia: a self-confident respondent thinks he knows about a

subject → Socrates tests his beliefs and finds them wanting → the
respondent is reduced to aporia, but is the better for it in his increased
self-knowledge and willingness to enquire afresh.
Labour pain aporia: the respondent, in conversation with Socrates, is
full of inchoate ideas and is suffering the aporia of mental pregnancy and
labour → Socrates helps to deliver the offspring (at which point the
aporia/labour pains cease, cf. 151a8 quoted above) → Socrates and the
respondent together test the offspring.

In short, in the elenctic case, Socrates’ testing and finding

inconsistencies in beliefs precedes and produces the respondent’s
aporia; in the labour pain case, aporia – in part brought on by
Socrates – precedes testing, which happens only after the labour
pains (that is, the aporia) have ceased. In each case Socrates is
(part or whole) cause of the aporia, and in each case it has some
benefit to the respondent, but these similarities mask some major
differences.5
A further nod to elenctic aporia is found in the Sophist’s sixth
definition of sophistry as a kind of cleansing, which seems designed
to recall earlier depictions of Socratic examination.6 The branch of
sophistry designated the elenctic branch of education is the art of
questioning those who wrongly think themselves knowledgeable on
a subject and showing their beliefs to be inconsistent. This
questioning and refutation removes the stubborn beliefs; the subjects
are made gentle towards others (pros tous allous hēmerountai) and
incur shame; this makes room for the benefit of future learning.
Being cleansed is thinking you know just what you do know and
nothing more (Soph 230c–d). The echoes of earlier descriptions of
and comments on Socratic examination are evident, and we seem to
have an allusion to Tht 210 c (quoted above), suggesting Theaetetus
will be ‘less burdensome and gentler (hēmerōteros)’ in future.7 While
it is clear that both passages – the closing remarks of Theaetetus,
and the sixth definition in the Sophist – are designed to remind the
reader of the practice and effects of Socrates’ examinations on
respondents such as Meno, Euthyphro and others, it may perhaps
be significant that the term aporia is not used in either passage. We
may hazard a reason: with the powerful midwife image, Plato has
introduced a new conception of subjective aporia as the labour pains
of one who is ‘pregnant’ with some theory or other, and who needs
the help of a Socratic midwife to deliver it fully – arguably a more
appealing notion of intellectual difficulty and of the role of Socrates
than ‘elenctic aporia’, the perplexity of one who has been humbled
by Socrates’ cross-examination.8
2 Philosophical Puzzles in Theaetetus
 2.1
In the following sections on Theaetetus and Sophist, I explore Plato’s
use of philosophical puzzles that share some common features.
Here I focus on two stretches of Theaetetus where such puzzles are
explored but (arguably) not solved: the Dice and Size puzzles of
154cff. and the longer problems of false judgement. At the end of the
false judgement section Plato uses the term aporia to designate a
specific puzzle (one that has dominated the discussion): a use that in
earlier dialogues is rare but found occasionally – such as when
Protagoras refers to ‘the remaining aporia’ posed by Socrates.9 Such
puzzles are the foundation of the Middle Part of the Sophist (with
frequent use of the term aporia for a specified puzzle), where we find
them developed at length and finally solved.
The puzzles I examine share features I label as follows:
F1: Not whether but how. Zeno’s problems about motion did
not cast doubt on whether there really is motion, but rather offered
seemingly valid arguments for the impossibility of motion. The same
applies to the puzzles I have selected, especially to the problems of
false judgement. These are not intended to cast doubt on whether
people can make false judgements (that is, have false beliefs); that’s
taken for granted (187e5–8). The difficulty is how to avoid the
seemingly sound arguments that rule out false judgement; that is,
how to properly understand false judgement.
F2 the puzzle is a familiar one and
F3 is one that can be treated in both an eristic and a
philosophically serious manner.10
 2.2
I first examine briefly the Dice and Size puzzles (154cff.), in which
we find the above features and also a key moment when Socrates
hails Theaetetus’ amazement (thaumazein) at thinking about them
(155c–d). They come during the exposition of the theory of
perception, and I must pass over the interesting question of just what
moral they have for that theory.11 Here’s the Dice puzzle.
‘Take six dice. If you put four beside them, we’ll say they are
more than the four, but if we juxtapose them with twelve, we’ll say
they are fewer and half as many.’ Next Socrates imagines
‘Protagoras, or someone else’, asking
Q. Is there any way in which something can become larger or
more numerous, other than by undergoing increase? (154c8–10), to
which Theaetetus replies:

If I answer by saying what I think with a view to this present question,

Socrates, I’ll say that there isn’t. But if I answer with a view to the one
before, I’ll be on my guard against contradicting myself and say that there
is.

Thus Plato shows how Theaetetus has grasped the problem. The
question, Q., seems to demand the answer ‘no’. But the Dice
example, suitably adapted, seems to suggest a way that six dice, on
being compared first with twelve and then with four dice, can
become more numerous without undergoing increase. Now at one
level this problem is easy to resolve, once we specify: larger or more
numerous than what? Than it used to be? Or, than another set of
dice? In fact Socrates neither offers a resolution nor explains how
the passage is relevant to the theory of perception. But the sequel is
of great interest, when Socrates distinguishes between two ways of
approaching such a problem, (1) engaging in a battle as sophists do,
and (2) ‘inspecting our thoughts themselves, in relation to one
another, to see exactly what they are, and whether we find they
harmonise with one another or fail to do so’(154e1–5). In other
words, the puzzle can be discussed in both an eristic and a
philosophically serious manner (Feature 3 above). Socrates then
sets out three principles, each of them plausible but together making
trouble for the Dice case, and even more for the next scenario,
whereby the elderly Socrates, without growing or shrinking, can be
larger than the teenage Theaetetus now but shorter a year later ‘not
because I’ve lost any of my size but because you have grown’
(155c1). Should we say that next year he is smaller but without
having become smaller? Or that he has become smaller but hasn’t
changed? This nice little problem is a good example of something
that could be dismissed as a mere sophism, easily diagnosed by
insisting on disambiguating the comparatives ‘larger/more numerous’
by specifying larger than what. But it also raises deep philosophical
issues, such as that of so-called Cambridge change (relevant to
problems elsewhere, such as in the Sophist’s Gigantomachia).
As well as Feature 3, we find F2 (Familiar) when Socrates
surmises that that the lad has some experience (ouk apeiros) of
‘these sorts of thing’ (155c6–7). And it’s obvious that these are
puzzles Not whether but how, since the question is: how should
such scenarios be conceptualised, not whether they obtain. Also
notable are Theaetetus’ initial hesitation and the way Socrates offers
a model of how to develop the problem in an enlightening way when
he outlined three plausible principles that together gave trouble for
the phenomenon under consideration. And in an important
exchange, when Theaetetus reports his giddiness and amazement in
thinking about ‘what these things are’, the occasion is ripe for
Socrates’ famous comment that being amazed is a mark of a
philosophical nature, and is the only starting point (archē) of
philosophy, a claim that will find an echo in Aristotle (Met. 982b12–
13), who links it with aporia. Theaetetus’ response to the puzzle has
shown the precise nature of the amazement.
Rather to our surprise, we soon find Socrates using sophistical
arguments against the Protagorean thesis (163–4, 165b–e),
reproaching himself (164c8–d2), and even allowing the personified
Protagoras to give him a lecture on the difference between
controversy and dialectic (167d5–b3).12 At 168c2 ‘Protagoras’ urges
Socrates to avoid the verbal tricks with which many debaters give
difficulties (aporiai) to each other. The theme of the difference
between eristic and truly philosophical questioning is kept before the
reader’s mind.
 2.3 The Problems of False Judgement
The famous section on the possibility of false judgement in
Theaetetus Part 2 is a far more sustained discussion of philosophical
problems, with plenty of aporia language. Though officially unsolved,
the puzzles will provide a stimulus for important philosophical
discussion. Socrates first offers two arguments purporting to show
false judgment to be impossible: from ‘knowing and not knowing’,
and from ‘being and not being’ respectively. Most of the discussion
turns on the first puzzle; I return to the second briefly at the end of
this section. An attempted way round the second problem,
characterising false judgement as ‘other-judgement’, runs into a
further puzzle recalling the first, after which Socrates offers two
analogies of the mind – as a wax tablet and as an aviary – before
declaring that they were wrong to leave knowledge on one side and
look for false judgement first. In other words, he declares that they
have failed to resolve the problem of false judgement (though some
authors have held that Plato indicates to the reader the correct
solution – a matter to which I return).
The false judgement stretch begins and ends with references to
aporia. Here’s how Socrates introduces the topic:

It’s rather bothering me now, and it often has before, so that I’ve got into
a lot of difficulty (aporia pollē) by myself and with others. I can’t say what
exactly this experience is with us, and how it comes into being in us.
Theaetetus: What experience? Socrates: Judging something false.
(187d1–6)
Already here we find the first two features I identified above: the
problem is a Familiar one, and is introduced with the feature I
labelled Not whether but how. Socrates is perplexed about what
false judging is and how it comes about, but not about whether there
is such a thing as false judgement, as he goes on to make clear: ‘Do
we say from time to time there is false judgement and one of us
judges what’s true and the other what’s false; this being naturally
(phusei) the case?’(187e5–7). Already in refuting Protagoras’ thesis
(170a–1d and 177c–9b) Socrates had in effect argued for the
possibility of false belief or judgement.
A final and important reference to aporia comes late in the
discussion of false judgement, at the end of the comparison of the
mind to an aviary. Theaetetus suggests that the mental aviary might
contain pieces of unknowing as well as of knowledge, to avoid the
problem entailed by explaining error in terms of mistaking one
knowledge-bird for another. In reply Socrates remarks that with this
suggestion ‘We’ve come a long way round and now we are back at
our first difficulty again’ (epi tēn prōtēn paresmen aporian 200a11–
12). The ‘first aporia’ that has reappeared is the original argument of
188a–d ‘from knowing and not knowing’.13 It considered four cases
of a person P mistaking X for Y: viz. when P knows both X and Y, or
only X, or only Y, or neither. Any false judgement must involve one of
these four cases, but is impossible in all four, they agree.14 In effect,
the argument assumes that if I know a thing, I can’t mistake it for
something else I know, or for something I don’t know; but if I don’t
know it, I can’t make a judgement about it at all. That initial problem
reappears in different guises throughout the discussion of false
judgement, accompanied by the language of aporia.15 Now Socrates
invokes it one last time to refute Theaetetus’ suggested ‘pieces of
unknowing’. He imagines an expert in refutation rehearsing the
original problem (188a–d) in new terms: does someone know a
piece of knowledge and a piece of unknowing, and think one is the
other? Or does he know neither, or just one? (The implication is that
none of these cases is possible.) It is a masterly touch to make
Socrates exploit the argument yet again at the close of the
discussion, correctly labelling it the ‘first aporia’, that is, the one he
first used to challenge the possibility of false judgement.
Why did Plato include this digression on false judgement, with
its parade of problems and apparently failed solutions? The question
has long troubled critics. As I noted, Plato makes his speakers
assume that false judgement is possible, so readers can infer that he
recognises that the arguments dismissing it are unsound. But does
Plato know, and if so does he indicate, which premises or inferences
are to be rejected? One line of interpretation, favoured by Zeller and
some contemporary critics, answers as follows.16 The false
assumption underlying the first argument of 188a–d (the ‘four-case’
argument) is – according to this reading – precisely Theaetetus’
second suggested definition of knowledge, whereby it is equated
with true judgement. Only if you identify knowledge with true
judgement do you get entangled in the main difficulty, according to
this line. It has the great advantage of giving a clear role to the
otherwise unexplained digression into false judgement: the
assumption that Plato wants the reader to spot as the troublemaker
is precisely the wrong view of knowledge currently championed by
Theaetetus, that knowledge is simply true judgement. And it credits
Plato with a diagnosis of the problems, one a discerning reader can
spot.
An alternative response is to admire the skill with which Plato
has constructed this suite of problems, even while doubting whether
he has fully identified the source of the difficulties. The puzzles –
especially the initial one that recurs throughout the discussion – raise
deep and difficult questions about the nature of thought or
judgement: what kind of cognitive grasp is required for me to make a
judgement about something? Perhaps Plato’s own assumptions
make it hard for him to fully escape the puzzle in all its forms.17 Or
again, perhaps we should take the hint offered when Socrates claims
to offer a poros, a path or way out of the difficulties (191a3–6): a
relaxation of the rule outlawing knowing things you don’t know. This
led to the development of the wax tablet image of the mind,
recognised by many as containing the grains of a solution to the
difficulties about false judgement, since it illustrates ways in which
knowledge (here identified with memory) can permit mistakes.18
Whether or not Plato has a solution to the puzzle of false judgement
based on ‘knowing and not knowing’, he has used it to stimulate
valuable enquiry, most notably in the wax tablet model of the mind.
The second puzzle (‘from being and not being’) makes only a
brief appearance in the long section on false judgement. It is a
variant of a familiar eristic one making difficulties for the
characterisation of false saying as saying what is not: a difficulty
found in earlier dialogues and prominent in Sophist.19 Since one who
judges, judges one thing, hence something that is, it follows that to
judge what is not is to judge nothing, hence not to make a judgement
at all (189a6–13). So ‘one can’t judge what is not, either about things
that are or just by itself’: this undermines the familiar characterisation
of false judgement as judging what is not. Though perhaps Plato
offers a hint of a diagnosis in the above formulation, it is only in
Sophist that he presents an explicit solution to this old sophism (cf.
3.4 below).20 The solution will throw light on the nature of saying
(and of judging), specifically on their subject-predicate structure: a
valuable spin-off from an eristic’s puzzle.
3 The Sophist. Using Aporiai to Reach Results:
Philosophy’s Difference from Sophistry
 3.1
My focus is the so-called Middle Part of the Sophist, but first some
remarks about the Outer Part. The difficulty of distinguishing
philosopher from sophist is given prominence early on,21 when
Socrates wonders if the Stranger from Elea is a theos elenctikos, a
god here to examine us. In reply Theodorus insists that the Stranger
is no eristic sophist but a philosopher. Cue the topic for discussion:
sophist, statesman, philosopher – are these three, two or one? With
the help of the method of division, the Stranger starts to hunt the first
of these, the sophist, and he offers no fewer than seven accounts,
each purporting to define the sophist. Critical opinion is divided on
whether Plato intends all seven, or only the last, or none at all, to be
the correct answer to the question ‘what is a sophist?’.22
One point is uncontroversial. Plato uses the series of definitions
to underline how close philosophy is to sophistry, following the initial
hint (216d) that philosophers sometimes take on the guise of
sophists. We have already seen (Sec 1) how the sixth definition –
signalled as an outlier – presents sophistry as a kind of elenctic skill,
whose description brings to mind Socrates as the practitioner of this
art – a so-called noble sophistry. The fifth definition of the sophist as
a controversialist (antilogikos) also contained reminiscences of
Socratic practice. Already before the Middle Part readers have had
many indications that it can be difficult to distinguish sophist from
philosopher.
The Middle Part, the interlude between the stretches defining
the sophist, is the philosophical heart of the dialogue, where issues
of not being, being, communion of kinds and false statement are
discussed. I argue that it is in this apparent interlude (and not in the
official definitions of the sophist) that Plato enables his readers to
grasp the difference between philosophy and sophistic/eristic
practices. The Middle Part shows, rather than tells, the reader the
real difference: it lies in how each of them treats aporiai,
philosophical puzzles of a kind found already in Theaetetus, but with
a far more prominent role in Sophist.
The seventh attempt to define the sophist, this time as a
purveyor of images and falsehoods, inaugurates a section where the
Stranger expounds problems (attributed to an imaginary sophist)
about the notion of not being implicated in any attempt to speak of
images or falsehood. Its abundant terminology of aporia is studied
below.23 It culminates with the Stranger resolving to show – against
Parmenides’ dictum – that what is not is in some respect, and that
what is is not in a way (241d6–7). Then he develops a second suite
of problems, this time about the notion of being, after which the
Stranger makes the surprising claim that being has proved to have
an equal share of aporia with not-being (250e6–7). A further problem
about being – the prohibition issued by so-called Late-learners on
certain common locutions – is followed by moves towards solutions.
In a well-signposted constructive stretch, the Stranger investigates
five megista genē or greatest kinds, exploring systematically how
they combine with one another. He confidently announces important
conclusions both here and in the following difficult passage which
explains not being in terms of difference (257b–8e), and challenges
dissenters not to exploit superficial contradictions (259c–d). But our
wily sophist has one further difficulty to throw in the way of the
enquirers: can there be a combination of not being with speech – in
other words, can one say what is not? In the accounts the Stranger
goes on to give, first of what a logos is and then of how ‘Theaetetus
flies’ can be a genuine logos and yet say something false, Plato
finally puts to rest the old problem about how one can ‘say what is
not’, that is, about how a statement can be false but still meaningful.
This summary reveals the architecture of the whole stretch:
announcing aporiai, ascribing them to the sophist, displaying them in
a suite of arguments, then delivering and heralding solutions.
 3.2 The Aporiai about Not Being: Plato Outdoes the
Sophists
The amazing (thaumastos) sophist has got into a perplexing area
(aporon eidos 236d2), says the Stranger, by denying that there are
images and falsehoods. Thus Plato signals this denial as a sophistic
device, something we know, concerning falsehood, from
Euthydemus inter alia.24 Many of the arguments the Stranger
presents are said to be ones ‘the sophist’ will develop.25 Immediately
we learn that the problem is how to speak about these things and not
get caught in contradiction.

Str. The fact is, my friend, we’re involved in an extremely difficult enquiry.
This appearing and seeming but not being, and saying things but not true
ones – all these matters are fraught with difficulty (mesta aporias), just as
they always have been. To know how one should express oneself in
saying or judging that there really are falsehoods without getting
caught up in contradiction by such an utterance: that’s extremely
difficult, Theaetetus.26 Tht. Why’s that? Str. That statement <logos>
dares to lay down that what is not is; otherwise there could be no such
thing as falsehood.

(236d9ff.)

Noting the phrase in bold, we can add a fourth Feature to the three
already listed. This problem, like some later ones, is said to turn on
(F4) how to speak correctly in order to avoid contradiction.27 And
we can also check off features identified earlier. The puzzles are
Familiar ones (F2), which have been the stock-in-trade of eristic
sophists but will form the basis for serious philosophy (F3 Eristic
and serious). Lastly, there is no real doubt about whether images
and false speaking exist; after all, the Stranger introduced their
denial as a subterfuge of the sophist. Hence this puzzle satisfies F1
Not whether but how?.
A suite of problems about not being follows, all carefully
signposted. The initial question is: To what can one apply ‘not
being’? It can’t be applied to anything that is (on), nor to something
(ti) … but one who doesn’t say something (or, who says not
something) says nothing. In response to the Stranger’s conclusion
‘Must we say that when someone tries to utter “what is not”, he is not
saying at all?’ Theaetetus exclaims: ‘Then our argument would have
reached the top-end of aporia.’28 Not so, replies the Stranger.
‘There’s more to come, the greatest and first of aporiai’ heralds a
second problem. Not being, since it can’t have any being ascribed to
it, cannot be said to be one or many. So both singular and plural
locutions (‘mē on’ and ‘mē onta’) are ruled out: one can’t correctly
utter or say or think of what is not, just by itself; it’s unthinkable,
inexpressible, unutterable and unexplainable. Here’s the sequel:

Then I was wrong, wasn’t I, when I said just now I was about to tell you
the greatest aporia concerning it. Here we’ve got another and even
bigger one to tell.. …. Don’t you see from the very words I’ve used that
what is not reduces to aporia even the person who’s out to refute it? It’s
like this: whenever someone tries to refute it he’s forced to contradict
himself in what he says about it.
(238d5–8)
How? Because someone who parades that argument disobeys their
own ban by attributing being several times when concluding that
‘what is not is unthinkable, inexpressible …’29 Note how here Plato
uses the terminology of aporia to individuate and rank puzzles (‘the
greatest’; ‘no, another even bigger one’ etc.)
This third and ‘even bigger’ aporia adds a delightful twist: the
refutandum, not being, is personified, and itself reduces to aporia the
would-be refuter who made the problematic claim. Twice more the
Stranger remarks on the difficulty of speaking correctly about not
being. ‘If one is to speak correctly he mustn’t demarcate it (sc. not
being) as one or as many, or call it ‘it’ (auto) because that would be
addressing it as a single thing’ (239a8–10). Then the Stranger
admits to being defeated in the past and now (F2), and tells the
enquirers not to look in him for ‘correct speaking (orthologian)
about what is not’ 239b1–5).
Correct speaking continues to dominate the fourth and fifth
puzzles, about images and about false speaking and judging. The
imaginary sophist challenges our right to speak of images (239d1–4)
or of judging falsely, by developing arguments to show that an
account of each involves combining – in language, that is – what is
with what is not, something they’d agreed to be impossible. To define
an image is shown to amount to calling it something that ‘isn’t really
but really is’ (240a12).30 Then false judgement and false saying are
defined as judging or saying that what is is not and that what is not
is.31 The sophist will not allow this designation (241a3–4);
Theaetetus recognises how the sophist will accuse them of ‘often’
going against what they earlier agreed was impossible: attaching
what is to what is not.
It is not difficult to see that the problem in defining an image is
sophistic, created by omitting the complements. On different grounds
we can query the alleged difficulty raised by the complex definition of
false statement. There is nothing amiss in saying that a false
statement says that what is not is (or, if negative, that what is is not);
indeed this formula for falsehood, with its way of ‘attaching’
(proshaptein) what is to what is not, is an insightful one. Plato
perhaps hints at this through the responses of Theaetetus (for
instance at 240e3–4), while still letting the Stranger represent the
sophist as seizing indiscriminately on the formula and making a
further aporia out of it.
Aporia language continues in the back bookend to the puzzles
about not being, where the Stranger notes how many and ready-to-
hand (euporoi) are the sophist’s objections and difficulties
(antilēpseis kai aporiai 241b). To defend themselves they must
subject to torture (basanizein) the logos of Parmenides, and insist by
force that what is not is in some respect, and that what is is not in a
way.32 Plato thus indicates that the reader should expect moves
towards disarming the problems of not being, challenging the basic
idea that what is not cannot in any way be said to be. But not until
further puzzles (about being) have been developed.
 3.3 The Problems about Being and Their Diagnosis
As with the not-being puzzles, the passage in which the Stranger
raises difficulties for being is very clearly signposted. Plato inserts
copious references to aporia at both ends, in highlighting the
surprising claim that being gives rise to aporia just as not being had
done. In the stretch about being, aporia terms generally pick out the
subjective state of perplexity, rather than an objective puzzle. After
the amusing doxography outlining the approaches to being of
various Pre-Socratic schools (242c–3a) the Stranger asks if they
really understand the sayings of these theorists, earlier identified as
‘Parmenides and anyone who tried to delimit how many and of what
kind are things that are’ (242c4–6). He reminds Theaetetus of ‘that
which was problematic just now’ (to nun aporoumenon) viz. not
being, adding a new point: in his youth he thought he understood it
(243b7–10). Perhaps – he remarks – we’re in the same state about
being, viz. saying we’re well off (euporein) and we understand it,
though aware we are in trouble over not being (243  c). This
prediction is kept before the reader’s mind throughout the section
offering puzzles about being.
How does the Stranger justify the surprising claim at the end of
the discussion of theories of being, that the enquirers are in no less
aporia about being than they were about not being (250e1–2)? The
problem of not being was in part that any attempt to speak of it led to
contradiction, to saying somehow that it is. But it’s not obvious that
such a charge can be brought against talk about being. So just what
is the aporia about being? Here we must be selective and follow
Plato’s signposts carefully. While several difficulties are raised, I
follow Plato’s indication of the dominant aporia.
The key signpost comes at the end of the Gigantomachia, which
had explored and criticised the two theories of being (theories of
what exists or is real): those of the materialist Giants and the
‘Friends of Forms’. Now that section discusses ontology seriously,
using arguments free from tricks or fallacy. Critics are divided on how
to understand the Stranger’s suggested definition of being in terms
of the power (dunamis) to affect or be affected, and on whether Plato
accepts it and exploits it in the constructive section to follow.33 But it
is certainly not heralded as offering a solution to the problems about
being. The Gigantomachia culminates with a new and final
characterization of being as follows: ‘being and the all is what is
unchanged and what is changed’. Though this conclusion satisfies
Theaetetus, it prompts the Stranger to remark that they are about to
recognise ‘the aporia of the enquiry about it (sc. being)’ (249d10–11).
Thus Plato indicates that what follows is central to the difficulty.
It transpires that the problem derives from a misunderstanding of
certain locutions to the effect that this and that are, or are such and
such. The Stranger now proceeds to undermine that final
characterization of being with an argument that seems plainly
designed to be fallacious, though there is controversy about
precisely where the fallacy is supposed to lie. Plato inserts a clue
early on, when the Stranger says they will interrogate those who
propose the final characterization by asking them the same
questions as they asked of the dualists (250a), who said that all
things are hot and cold, or some such pair. Earlier the dualists had
been asked (243e1–2): ‘what is this that you apply to the pair when
you say that both and each of them are? How are we to understand
this “are” of yours? (ti to einai touto hupolabōmen humōn?)’ Note
that the question asks how to understand an expression. The
Stranger now uses a similar form of words (250ab): you say change
and stability are most opposed to each other, yet you say that both of
them, in the same way, are.34 When you say they both are, are you
saying they change? (no), or do you mean (sēmaineis) that they stay
stable? (no), so you reckon being as a third thing (250b8). (Compare
the conclusion extracted from those who said the all is hot or cold:
they must admit that being is a third thing.) From this innocuous
concession – that to be does not mean to change or to stay stable –
the Stranger will derive an impossible conclusion, via some highly
dubious reasoning. The conclusion (250d2–4) is that being is neither
changing nor stable – impossible since everything must be one or
the other.
The strong hints are that this fallacious argument (like that
against the dualists) depends on a misunderstanding of what is
being said in claims such as ‘the hot and the cold are’ or ‘change
and stability are’. Three times the Stranger has asked what is meant
by certain claims that this or that is or are (243b2–7, 243d8–e7,
250a11–b7). We will find confirmation of these hints in the sequel,
but first we must note how emphatically the Stranger declares the
initial prediction of 243 c fulfilled. He does so using four aporia words
in ten lines (250d7–e6), (i) recalling that they were in total aporia
about not being; (ii) asking if they are in any less aporia about being;
(iii) suggesting that they regard being as diēporēmenon – something
whose difficulty has been completely stated; and (iv) expressing the
hope that that, since being and not being have an equal share of
aporia, clarifying either one will make the other clearer too.
At various points in the following constructive discussion we find
confirmation that ‘the aporia of the enquiry’ about being predicted at
249d10 is a misunderstanding of certain ways of speaking. I can give
only the briefest of outlines to justify this. Straight after expressing
the hope of clarifying being and not being together the Stranger
introduces another problem that concerns certain ways of speaking.
The so-called Late-learners forbid us to say certain things, such as
making the innocuous claim that the man is good, on the grounds
that we thereby make one thing many (251a–b). Note that this
problem too possesses the features I identified earlier.35 One
diagnosis is that they do not allow predication, thinking that a
sentence such as ‘the man is good’ could at best state an identity;
alternatively they allow only per se or definitional predication and
forbid ordinary predication.36 Then, after the programme of
investigating the ways certain ‘Greatest Kinds’ combine has been
announced and delivered, we find Plato once again putting the
spotlight on the correct understanding of things said. In a systematic
investigation of how Change combines with the other four kinds, the
Stranger shows how various apparently contradictory claims are not
really contradictory once we understand ‘that we were not speaking
in the same way’ when we say that ‘Change is the same and not the
same’ (256a10–b4) or that ‘Change is not being [i.e. is different from
being] and being, since it shares in being’ (256d8–9).
 This is not the place to evaluate the various interpretations of
what Plato saw as the error in the Late-learners’ thinking, nor of what
precise distinctions he is drawing our attention to in the Communion
of Kinds passage.37 What the interpretations have in common is that
each finds Plato insisting on disambiguating sentences – ones
claiming that something is, or is such-and-such – so that a claim
that’s true and unproblematic on one reading is false on an
alternative construal.
I turn to a final important signpost, where the Stranger repeats
(with some variation) the conclusions from the passage of the
communion of kinds, and warns against revelling in such
contradictions without troubling to understand in what way something
is being said. ‘To rejoice in bringing up such contradictions in one’s
arguments is no genuine refutation (alēthinos elenchos)’ but (in
Cornford’s rendering) ‘the callow offspring of a too recent contact
with reality’ (259d). The phrase ‘no genuine refutation’ reminds us of
the difference between eristic (as practised by the brothers in
Euthydemus) and serious philosophy, and the remark about ‘a too
recent contact with reality’ recalls an earlier claim about how
sophists bedazzle young men (234d) as well as the label ‘Late-
learners’ for those who revel in the paradox about correct speaking.
With those remarks Plato underlines how the Stranger, with his
patient investigation and explanation of apparent contradictions
(255e–6e, resumed at 259a–b), has given the philosopher’s answer
to the paradox-mongers whose arguments he first aped, then
demolished by his careful reasoning. To oversimplify, the chief aporia
about being – the one specially remarked on at 249d10–11 – was
(like the Late-learners’ problem) a misunderstanding of predicative
statements. Rather to our surprise, the chief aporia was not about
the meaning of existence claims, or the nature of reality, though the
Gigantomachia episode did explore that in an entirely enlightening
and non-eristic way. The exploration of the Communion of Kinds has
shown how to diagnose the problem of misunderstanding different
kinds of statements and how to solve it.
 3.4 How Are the Problems of Not Being Solved?
To examine this adequately is beyond the scope of my essay.38 My
tasks instead are to follow the signposts Plato inserts into the
conversation, and to note the undoubted success in solving one
major puzzle about not being: the account of false statement.
I take up the thread after the Stranger has declared that they
have done what Parmenides had prohibited (saying that what is not
is), and more: they have dared to say what the form of not being is
(258c–d). He thereby refers to his explication of not being in terms of
a part of the different, an obscure and disputed account.39 But
fortunately far greater clarity is to come, and Plato goes out of his
way to signal the change of tack as they turn to the problem of false
statement.
Note the very firm hand with which Plato steers the remaining
discussion. Despite his bold claim to have found the form of not
being, the Stranger announces that a further problem remains to be
solved, that of showing what logos is and whether it can combine
with not being – for if not, there will be no false judgement or false
logos ‘because judging or saying what is not – that, I think, is what
falsehood is in thought and in statements’ (260c1–4). This is an
explicit recall of 236e1ff: the difficulty of saying or thinking that there
really is falsehood without getting caught in contradiction, with its
implicit reference to the well-known description of saying something
false as ‘saying what is not’.40 Over the next exchanges Plato inserts
careful reminders of the original problem raised by the imaginary
sophist. Note especially 260c9–d3: ‘we said the sophist had taken
refuge … and denied that there is any falsehood, since one can’t
think or say what is not’. This is the second reminder in ten lines of
that aporia-producing formula.
The signposting continues as the enquirers emphasise that the
difficulty of accounting for falsehood still remains, and therewith the
‘capture’ of the imaginary sophist. Using a fresh label, Theaetetus
grumbles that the sophist is laden with problems (problēmata) –
defences he can throw up – of which the latest is showing that
falsehood exists in both statement and judgement (cf. 241b1). This
last problem – initially presented as a dodge used by the sophist, but
also of considerable philosophical importance – will be triumphantly
solved in the most celebrated section of the dialogue (261d–4b),
which starts with the Stranger giving an account of what logos is,
before offering formulae for a true and a false logos. Many details of
the solution are contested, in particular how to interpret the two
formulae for a false statement that get rid of ‘what is not’ in favour of
‘what is different’.41 But it’s generally recognised that the key to the
solution lies in the account of a logos (that is, a statement), showing
that it’s not a mere name, or something that functions like a name,
but an essentially articulated form of words, containing functionally
different parts, labelled name and verb, though the key distinction is
between subject and predicate. The Stranger stresses that a logos
(or its speaker) doesn’t ‘only name but achieves something by
weaving together verbs with names’ (262d2–4). This leaves room for
what the earlier puzzles had made to seem impossible, viz. that a
logos such as ‘Theaetetus is flying’ is a genuine statement, is about
something and yet says something false. Next the Stanger swiftly
vindicates false judgement, on the grounds that a judgement (doxa)
is a silent statement, and then appearance (‘a judgement mixed with
perception’). Being thus akin to statement, they too – judgement and
appearance – can have false instances as well (263d–4b). They’ve
found false statement and false judgement sooner than they
expected (264b): a triumphant announcement of the solution of a
major difficulty.
Has Plato thereby solved all the problems of false judgement left
over from Theaetetus? Here’s my verdict. The second of the two
initial aporiai in Theaetetus, the logical problem spawned by the
locution ‘to judge what is not’, has indeed been solved with the help
of a) the celebrated dissolution of the problem that a false logos
‘says what is not’ and b) the analysis of judging as silent saying. But
perhaps the epistemological problem remains, of how someone can
believe a falsehood: the focus of the Theaetetus’ first aporia which
also governed much of the remainder of the false judgement
stretch.42 Note that while Plato’s choice of sample false logos –
‘Theaetetus is flying’ – serves well for his demonstration of a
genuinely meaningful and yet false logos, it gives no help with the
Theaetetus’ problem of how someone could believe such a
falsehood!
 Conclusion
It is well understood that the philosophical analysis found in Sophist
takes the form of systematic posing of philosophical puzzles to which
solutions are then offered, allowing us to diagnose the source of the
difficulty. This paper (Sections 2 and 3) has examined how skilfully
Plato signposts his uses of such aporiai. It has shown how such
puzzles are already found, though not solved, in Theaetetus and has
explored features shared by both dialogues’ aporiai. Arguably the
most important is that the puzzles are the common province of eristic
sophists and philosophers. By various means Plato has underlined
the similarities and differences between these approaches,
illuminating the real distinction between eristic sophistry and true
philosophy.43

1 I follow Szaif’s classification of a formally aporetic dialogue, Chapter 2,

Section 2. Like other formally aporetic dialogues, Tht has been the subject
of many doctrinal readings, cf. Sedley 2004.

2 Burnyeat 1977b: 7 notes how carefully Plato here distinguishes what is

common knowledge about Socrates from what is a hitherto unrevealed
secret. He infers that Plato is declaring the midwife image non-historical.

3 Burnyeat 1977b: 11 n14 (= id 2012: 30 n14) cites Robinson 1953: 83–4.

4 Difficulties for this tempting idea include (i) the fact that in the Theaetetus’
aviary model of the mind, it is declared empty of knowledge-birds at birth
(197e2–3), and (ii) that while the slave experiment suggests that all are
capable of recollection, here Socrates insists that only some of the young
men he associates with are pregnant (151b – he passes the barren ones on
to Prodicus). See Sedley (2004) 28–30 for further discussion.
5 My distinction has some features in common with that of Politis 2006
between what he labels cathartic and zetetic aporia. However, for Politis, so-
called zetetic aporia is puzzlement over a particular problem (2006: 105–7),
whereas what I have called labour pain aporia is not explicitly so linked.

6 Discussed by Szaif in this volume (Chapter 2).

7 A.A. Long 1998: 30 notes how the language of Soph 230a recalls that of
Socrates’ verdict (Tht 210c) on what will happen to Theaetetus if he
continues to be barren, but not how incongruous that verdict is.

8 The Digression on philosophy and rhetoric contains many aporia terms (all
but the first are forms of the verb aporein) used for an unwelcome and
ridiculous state of perplexity. Three occurrences (174c5, d1, and 175b6)
ironically ascribe it to the philosopher, while at 175d5 it describes the truly
ridiculous aporia of the rhetorician.

9 cf. Politis 2012a, esp. 221 and n15; and Szaif in this volume (Chapter 2).

10 In Section 3 I add a fourth feature, more prominent in Soph: the puzzles

concern correct speaking.

11 At 155d5–6 Socrates hints at a moral for the theory of perception; cf.

McDowell 1974: 133–7.

12 For detailed examination of the sophistical nature of the arguments here

see Crivelli 1996, Sec II.

13 For eristic versions of an argument from knowing and not knowing cf.
Euthd 293bc.

14 I pass over the question how this argument can be said to rule out false
judgement in general, rather than simply false identity-judgements. For
discussion see Sedley 2004: 120–5.

15 See for instance 190c–e, with aporoumen at e9.

16 See Burnyeat 1990: 70–119 who favours this line and discusses a similar
interpretation in Fine 1979: 70–80. Benson 1992 offers a subtle and complex
variant of such a reading.

17 See the difficulty Socrates raises at 209b about what’s entailed by having
a belief about Theaetetus.

18 cf. among many treatments McDowell 1973 215–6; Sedley 2004: 134–40.

19 Arguments turning on the equivalence between false saying and saying

what is not are found at Euthyd 283e7ff, where the sophist Euthydemus
uses it to bamboozle Ctesippus, and at Crat 429d1–6, where it is attributed
to ‘many contemporary and past’ debaters.

20 cf. Sedley 2004: 125–7, who notes that ‘judging what is not about the
things that are’ – as found in the passage quoted above – prefigures the key
role in Soph of the phrase ‘about Theaetetus’ in the dissolution of the
problem of false statement as saying what is not.

21 Frede 1996: 146–8 notes this, and throughout has insights to which my
treatment is indebted.

22 Notomi 1999 argues for the seventh as the correct definition; Brown 2010
argues that none is.

23 Ballériaux 2001 notes that the Sophist has the highest concentration of
aporia terminology of all Platonic dialogues, and that almost all of them
occur in the stretch from 236–51.

24 cf. n19 above.

25 239c9–d4, 239e1–40a6, 240c3–5, 241a3–b3, cf. Palmer 1999: 143.

26 This rendering follows that of Frede 1996: 144; Crivelli 2012: 29 takes it
differently. I do not translate the phanai conjectured in OCT 1995 at e4.
27 cf. Frede 1996: 143–5, with his summary at 145: ‘It is these aporiai which
give us a clue about what we have to say, and how what we say has to be
understood.’

28 telos aporias: Crivelli 2012: 34, n.23 notes a pun on telos as completion
and as culmination.

29 First in using the verb to be, and lastly in using the singular form of the
adjectives. The text and meaning of the intermediate remark (239a3) is
disputed, cf. Crivelli 2012: 45.

30 Suppose a painting of a carpenter (Rep 598b): the image isn’t really <a
carpenter>, but really is <an image>. Once the missing complements are
restored, the problem dissolves, but the abbreviated formula imports the air
of a contradiction.

31 This is the complex definition of falsehood. Also present is the simple

definition: false judgement is judging what is not (240d9); see 3.4 for its later
appearances and role.

32 With Palmer 1999: Chapter 6, I suppose Plato is critiquing a sophistic

misusing of Parmenides’ dictum, not its true understanding.

33 For discussion see Brown 1998. Leigh 2010 and Delcomminette 2014:
539 take the definition of being as dunamis still to apply in the Communion
of Kinds passage. Crivelli dissents (2012: 90).

34 While the final characterisation was in terms of unchanging and changing

things (hosa akinēta kai kekinēmena), the Stranger now speaks of change
and stability (kinēsis kai stasis) as being.

35 It’s not about whether one can call something many names, but how to
understand such a locution (Feature 1). Its assignment to ‘Late-learners’
reminds us of the eristic expertise of the brothers who came late to that
profession (Euthd 272b). Euthd 301ab has a related puzzle about
predication, known as ‘Dionysodorus’ ox’.
36 Brown 2008: 440–3 discusses rival interpretations; Frede 1992c: 400
favours an interpretation invoking per se predication, cf. Crivelli 2012: 107–
8.

37 For a survey of interpretations see Brown 2008: 443–51 (who favours a

distinction between identifications and predicative sentences) and Crivelli
2012: 149–66 (who distinguishes ordinary from definitional predications).

38 Crivelli 2012 is a meticulous study of the complex issues.

39 Examined in Brown 2012.

40 cf. end of Section 2 above.

41 See Crivelli 2012: Chapter 6 for a detailed exploration of options.

42 See n16 for the view that Plato does indirectly indicate the solution to the
epistemological problem.

43 cf. Szaif’s Chapter 2, Section 3 for a similar claim about the Euthydemus.
Grateful thanks to Jacob Fink, Gail Fine, Jan Szaif and the editors for helpful
comments.
 Chapter 6
Aporia and Dialectical Method in
Aristotle
◈

Christof Rapp

The Aristotelian treatise that is dedicated to Aristotle’s dialectic, the

Topics, aims to formulate a method by which we can construct
deductions from accepted premises for any problem that is proposed
(Topics I.1, 100a18–20). The use of such a method also enables us
to deduce contrary conclusions (Rhetoric I.1, 1355a33–5). Intuitively,
a situation in which we are faced with contrary conclusions, both of
which are derived from reasonable grounds, fulfils the most
important criterion of what we call ‘puzzlement’ or ‘aporia’. It
therefore seems that, in an Aristotelian framework, the dialectical
method and the occurrence of aporiai are mutually connected in a
significant way and that, hence, any assessment of the role of
aporiai in Aristotle’s philosophical method, should take dialectic and
the Topics into account. Against that background, it is somehow
surprising that the term aporia does not play a more prominent role
in the eight books of the Topics. We can say at least that the notion
of aporia is not among the key concepts of dialectic’s terminological
inventory. The following essay will first sketch in general terms the
relation between aporiai on the one hand and the dialectical method
on the other. Then it will look more closely into some passages from
the Topics which actually use the notion of aporia. In some of the
more pertinent passages that use aporia or its cognates, there is a
remarkable connection with the definition of the dialectical problem
(dialektikon problēma). It will be argued, first, that in the Topics the
word aporia in its most significant use is meant to qualify what
Aristotle defines as a ‘dialectical problem’ and, second, that the
problem in dialectical theory and practice more or less corresponds
to what Aristotle calls an aporia in philosophical or scientific contexts.
If this is roughly right, we can take the characteristics of the
dialectical problem to provide boundaries on the concept of an
aporia. Such a constraint enables us to push back against the
tendency to view all instances of questions, queries, enquiries,
procedural remarks as somehow qualifying as aporiai, leading to an
excess of aporiai in Aristotle’s work. Still, it would be unrealistic to
expect a completely precise or uniform notion of aporia. The picture I
would like to suggest is rather this: In his dialectic Aristotle takes up
the current and common notion of an aporia and associates it with
the more technical notion of a dialectical problem. This technical use
of problem, together with his understanding of certain dialectical
techniques attached to it will have a certain identifiable impact on
Aristotle’s understanding of what it is to be an aporia and on his use
of aporiai within the process of philosophical research.
 1 Dialectic and the Method of Diaporēsai
In the second chapter of the first book of the Topics, Aristotle speaks
about why the treatise or the method presented in this treatise is
useful. One respect in which it is useful is that the method enables
us to work through the difficulties (diaporēsai) on either side of a
philosophical or scientific topic. Since the practice of diaporēsai –
however it might be spelled out in detail – certainly implies an
occupation with aporiai, it is clear that there is an intimate connection
between the dialectical method and either the formulation or the
handling of aporiai. It is worthwhile, then, to take a closer look at this
particular use of dialectic. Prior to this, however, we should say a few
words about the context of the first two chapters of the Topics, where
Aristotle delineates the dialectical method and addresses possible
uses of this method.
 1.1 The Dialectical Method as Announced in Topics I.1
The Topics straightforwardly opens with a remark about the purpose
of the treatise. It reads:

[Text 1] The goal of this study is to find a method with which we shall be
able to construct deductions from acceptable premises concerning any
problem that is proposed (sullogizesthai peri pantos tou protethentos
problēmatos ex endoxōn) and – when submitting to argument ourselves –
will not say anything inconsistent.1

Immediately after the quoted passage he says that one must grasp
what the dialectical sullogismos is, since this – the dialectical
sullogismos – is what the treatise is looking for. Taken together,
these two pieces of information provide a good account of what this
treatise is about. First of all, the announced method is meant to
instruct us how to formulate sullogismoi, namely sullogismoi taken
from premises that are endoxa, i.e. accepted, acceptable, reputable
or whatever opinions. Since the quoted passage is directly followed
by the remark about dialectical sullogismoi and since a few lines
later the dialectical sullogismos is actually defined by the nature of its
premises, which, in opposition to the established true and primary
premises of the scientific (apodeictic) sullogismos, are qualified as
accepted, acceptable, reputable etc., it is clear that the announced
method is meant to aim at the construction of dialectical sullogismoi.
In addition to that, the quoted passage indicates at least one
more important point, namely that the construction of sullogismoi in
the sense of actively drawing conclusions from accepted premises
concerns only one side of the practice that the announced method is
meant to teach; for there seems to be another, more passive way of
engaging in the same practice, which Aristotle describes as logon
hupechein, i.e. as upholding an argument, sustaining an argument or
submitting oneself to an argument. It seems, then, that right from the
start, the method is sketched with a view to two interlocutors,
contestants or competitors, one of whom tries to draw conclusions
(probably a conclusion showing that the opponent’s claims are
inconsistent), while the other one tries not to concede anything that
allows the opponent to draw such a conclusion (probably a
conclusion that amounts to the refutation of a claim the defendant
made). It therefore seems that the method by which we can either
construct dialectical sullogismoi or avoid being refuted by dialectical
sullogismoi is primarily designed with a view to a sort of dialogical
disputation or examination that involves an attacking or examining
party on the one hand and a defending party on the other.
And, indeed, this first impression is confirmed by many details in
the first and, above all, in the eighth book of the Topics, where
Aristotle describes an argumentative procedure or disputation that is
going on between a questioner (who is supposed to construct
sullogismoi in order to refute the opponent) and an answerer (who is
supposed not to admit anything that commits him to accepting a
conclusion that is inconsistent with other things he had asserted).
Also, we learn in the course of these books that at the beginning of
such a conversation the answerer has to choose which side of a
dialectical problem (i.e. which of a pair of contradictory propositions)
he is going to defend. The questioner in this peculiar question-and-
answer examination is expected to ask only yes or no questions,
while the answerer is expected to either accept or to reject them –
provided that they are understandably formulated. By accepting or
rejecting the proposed assertions the answerer commits himself to a
set of propositions. These commitments, or some of them, can be
used by the questioner as premises for the sullogismos he wants to
formulate. This is why Aristotle goes so far as to define the
dialectical premises as questions, thus indicating that he takes this
question-and-answer examination as the essential context of the
dialectical sullogismos. Correspondingly, the method that is unfolded
in the Topics, and which is centred on how the dialectical
sullogismos comes about, is primarily thought to teach us how to
attack or defend claims of any content vis-à-vis real life opponents.
The great advantage of the dialectical method, as it is sketched
in the first lines of the Topics, seems to be that it can be applied to
‘any problem that is proposed’. Above all, this means that the
dialectical method is applicable regardless of the domain or scientific
field from which the problem is taken. The dialectical method, hence,
cuts across the distinctions between Aristotelian disciplines; as we
will see in one of the following sections, this is facilitated by the
nature of the topoi that dialectic deploys. It is different from scientific
reasoning also in that it does not argue from established principles
that are peculiar to one or the other field of knowledge, but only from
endoxa, which are only accepted, acceptable or reputable, but not
necessarily true (and certainly not established-as-true) premises.
Also, as we will see, a dialectical problem asks whether something is
the case or not. That means that, if the dialectical method applies to
dialectical problems, it not only helps us with theses with any content
or from any field of knowledge, but also does so regardless of which
side of the contradiction one wishes to defend. More or less the
same result can be reached by recalling that the announced method
is meant to help not only the one who wishes to test and to challenge
a thesis by drawing conclusions from accepted opinions, but also the
one who is trying to defend the same claim that the examiner is
trying to refute. This again implies that the examiner and the
defendant aim to establish or defend contradictory views, while the
dialectical method is meant to support both of them alike.
 1.2 The Third Use of the Dialectical Method according to
Topics I.2
In Topics I.2, Aristotle mentions three uses of dialectic: (i) the
gymnastic use (gumnasia), (ii) the encounter with the many and,
finally, (iii) the ‘philosophical sciences’. Aristotle actually presents two
different ways in which dialectic can be used in this latter context,
namely (iii.a) because one can go through the aporiai on both sides
(that’s the one we are ultimately interested in) and (iii.b) because
dialectic helps us with the well-known phenomenon that no particular
science can establish or demonstrate its own principles; dialectic, by
contrast, can discuss such principles on the basis of accepted
opinions. This is the use of dialectic many interpreters are primarily
interested in. This is not the place, however, to enter into the
corresponding discussion. It may suffice to say that nothing in the
text indicates that dialectic itself would actually establish such
principles. Aristotle rather emphasises the contrast between
reasoning on the basis of principles (which, of course, is not possible
if these principles themselves are at issue) and reasoning on the
basis of accepted opinions; if one accepts this contradistinction and
the general picture of apodeictic sciences in Aristotle, then it is clear
that reasoning on the basis of accepted opinions provides a way of
rationally assessing possible candidates for the rank of a scientific
principle – but not more than that.
At any rate, uses (iii.a) and (iii.b) are widely seen as the crucial
link between dialectic and ‘serious’ philosophical research – as
opposed to mere training (gumnasia) and encounter with the many.
The dialectical method is useful for the philosophical sciences in that
the philosophers can use certain aspects of this method in order to
raise aporiai and in order to discuss principles of the different
sciences on the basis of accepted opinions – it is not too difficult to
see which aspects of the complex dialectical method are more
significant for these two purposes, and which are less so (‘part-
whole-account’). Alternatively or in addition, one could say that
philosophers who have been trained in what Aristotle takes to be
dialectic proper (‘competence-account’) can use their dialectical
competence for raising aporiai and for discussing scientific principles
on the basis of accepted opinions. Both the whole-part-account and
the competence-account make clear how and why these
achievements are crucially linked with the dialectical method,
although, strictly speaking, we are not committed to saying that
whoever raises aporiai in philosophical contexts or whoever
discusses the principles of various sciences on the basis of accepted
opinions thus uses the whole of the dialectical method or engages in
the activity that is definitory of dialectic proper. Still, it is clear how
these achievements are derived from the dialectical method or the
dialectical competence respectively.
We are now in a position to proceed to the discussion of (iii.a),
the usefulness of dialectic with respect to the raising of difficulties in
philosophical sciences. The brief passage reads:

[Text 2] It is useful in relation to the philosophical sciences because if we

have the ability to go through the difficulties (diaporēsai) on either side we
 shall more readily discern the true as well as the false in any subject.2

The crucial ingredients of this remark are (a) the reference to a

specific ability that seems to derive from the exercise of the dialectic
method, (b) the description of it as the ability to go through aporiai on
either side, (c) the characterisation of the advantage it brings about
as the ability to more readily discern what is true and false in any
subject with which we have to deal.
The key concept in this passage is diaporēsai. In general,
diaporēsai or diaporein need not be more than a powerful, intensive
form of aporein, i.e. to be at loss, to be in the grip of a difficulty.3
However, it is possible that the word has attained a more technical
meaning in the context of Aristotelian methodology. According to
such a technical meaning, if there is one, the word would seem to
refer to the raising or stating of possible difficulties rather than to the
process of solving them.4 Accordingly, the process or procedure of
diaporein is typically connected with a plurality of distinct aporiai, so
that the prefix dia – if it is not simply intensifying – seems to indicate
that one goes through the pertinent aporiai in an exhaustive way5
(rather than that one goes through a single aporia from its initial
statement to the final conclusion). It seems to be implied that in the
process of diaporēsai or diaporein we must approach a given subject
by way of a series of ‘p or non-p’-questions (or, more generously,
questions implying not just contradictory but any sort of contrary
alternatives, such as ‘one or many’, ‘numerically one or qualitatively
one’, ‘with or without beginning’, ‘eternal or corruptible’ etc.). The
plurality of aporiai implied by the diaporēsai-procedure might be a
plurality of such questions or (perhaps in addition) a plurality of
difficulties associated with each one of them. Aristotle stresses, in
Text 2 and elsewhere, that the difficulties have to be raised on both
sides, which may mean either that one should raise difficulties that
concern alternative p and others that concern alternative not-p or,
simply, that one should raise a difficulty that concerns (perhaps by
providing good reasons for holding) both p not-p. Either way, it
seems to be crucial for the diaporēsai-procedure that it explores a
given subject by way of raising difficulties and that it presents
credentials or arguments for contrary views on the subject as a
whole or on the particular questions by which the subject is
approached.
The main advantage gained by this process is, according to Text
2, that we will more readily or more easily see or discern what is true
or false in the treated subject. What Aristotle seems to describe here
is an epistemic state. Being in this peculiar epistemic state implies
neither that we have reached reliable knowledge about the subject in
question, nor that we have reached any definitive result about it. The
epistemic state described rather seems to consist in a synoptic view
of the competing theoretical options and their respective advantages
and drawbacks. Such a state might plausibly be thought to result
from the diaporēsai-procedure: if, for example, the difficulties for
option p pile up and seem to be insurmountable, whereas there are
hardly any serious troubles for option non-p, one will ‘discern what is
true or false’ in that the diaporetic assessment indicates that p is
more likely to be false, whereas non-p deserves to be further
developed and to be liberated from remaining difficulties and
obstacles. Having achieved such a synoptic view is a considerable
advantage for further enquiry. It enables us to avoid wasting time on
relatively unpromising alternatives and helps us identify and focus on
the less problematic ones. It follows from the logic of the diaporēsai-
procedure that the enquirers detect difficulties not only for the
obviously false or less promising option, but also for the more
promising path, i.e. the path that will be pursued in the further course
of the enquiry. With respect to this less problematic option the raised
difficulties clearly define what we have to do – which obstacles we
have to remove, which differentiations we have to impose, which
qualifications we have to provide etc. – in order to reach a
satisfactory solution.6 In addition, it could be argued that it is in virtue
of the diaporēsai-procedure that we can foresee and preclude
objections to the suggested solution. And at least in some cases the
avoidance of the difficulties that regularly occur in a certain field may
already form an essential part of the justification of a suggested
solution.
All these manifest benefits flow from the ability to raise aporiai.
And it seems to be presupposed in Text 2 that this ability does not
involve more than a certain proficiency in the dialectical method and
some amount of practice in dialectical disputations. We should be
able, then, to identify the parts of the dialectical method that are
likely to enhance our ability to raise aporiai. Since aporiai may simply
derive from incompatible reputable views about the same subject,7 it
is obvious that the dialectical expertise in endoxa, accepted or
reputable opinions, is required. This, however, only seems to be the
simplest case. More frequently, it seems to be necessary to argue
that a difficulty for a given subject arises given certain plausible
assumptions; that means that we have to derive certain conclusions
from accepted opinions. Also, it may happen that there are no
contrary views and no obvious arguments for the contrary position
available. In all such cases it is the dialectician’s expertise in
sullogismoi that is required, alongside the dialectical system of topoi,
which provides an almost inexhaustible reservoir of different ways in
which sullogismoi can come about. Most notably, it seems to be the
dialectician’s expertise in attacking or defending either side of a
problem that puts the enquirers in a position to raise difficulties –
independently of their own theoretical preferences. In the Aristotelian
system of philosophical disciplines it is only the dialecticians that fully
abstract from their own views and opinions and train themselves to
argue for any given thesis. The same attitude is also crucial for the
procedure called ‘diaporēsai’; for without the willingness and ability
to raise difficulties, without a view to an envisaged result or to
preferred starting points, the collecting of aporiai would lose its
philosophical motivation and substance.
 2 ‘Aporia’ in Aristotle’s Topics
Given the close affinity between the dialectical method and the
formulation of aporiai in Aristotle’s philosophical-scientific treatises,
one might expect the treatise in which Aristotle’s original approach to
dialectic is systematically unfolded to include rich discussions of
aporiai. But in fact, the notion of aporia is relatively rare in Aristotle’s
Topics and does not play any role within the terminological inventory
of dialectic. A TLG search for apor (α – π – ο – ρ) resulted in less
than twenty hits within the uncontested parts of the Topics. One
obvious reason for this result might be that the Topics does not
present itself as carrying out a proper enquiry in the way other
philosophical treatises do. There are, hence, almost no hints in the
text of difficulties or puzzles that have to be solved. A majority of the
occurrences of ‘aporia’ and its cognates in the text of the Topics
have a merely procedural meaning in the sense that they refer to the
procedure of the dialectical disputants or that one of them wonders
how he should proceed or is at loss about what to do next etc.8
Three more uses come from a passage9 which presents a reported
definition of aporia in order to illustrate a certain failure in defining
things. Then we have the famous occurrence of diaporēsai (quoted
in Text 2) that we discussed in the first part of the paper, while most
of the remaining occurrences are connected with the discussion of
what Aristotle calls a ‘dialectical problem’. In the following sections I
would, hence, like to explore the connection between dialectical
problems and aporiai. Before proceeding to these passages it is
worthwhile, however, to consider Aristotle’s comments on the
reported definition of aporia.
 2.1 A Reported Definition of Aporia
In Book VI, Chapter 6 of the Topics Aristotle discusses various kinds
of fallacious definitions. In this context he comes up with a definition
of aporia that seems to be a reasonable starting point. He states in
general terms that some people make mistakes by ignoring the fact
that every disposition or affection is naturally formed in that of which
it is an affection or disposition, just as knowledge is formed in the
soul, being a disposition of soul. Then he mentions the – allegedly
mistaken – definition that aporia is equality between contrary
thoughts or reasonings (hē aporia isotēs enantiōn logismōn).10 This
definition is mistaken, he says, because aporia is not a feature of the
reasoning or thoughts (logismoi) themselves. Also, this account
mistakes the cause (equality) for its effect (the state of aporia), and
this does not belong to the thoughts (logismoi). Aristotle illustrates
this again by saying:

[Text 3] Likewise also equality between contrary reasonings (hē tōn

enantiōn isotēs logismōn) would seem to be a cause of aporia; for it is
when we reflect on both sides of a question and find everything alike to
be in keeping with either course that we are puzzled which of the two we
are to do.11

What Aristotle criticises is not the association of aporia with equality

or equal strength or persuasiveness of contrary reasonings, but only
the confusion of cause and effect. It remains true, hence, to
associate aporiai with equally strong reasonings or arguments for
both sides of a question, but it all depends on making clear that
aporia is a state of the soul resulting from such considerations.
Although Aristotle is by no means committed to the examples he
discusses in the course of the Topics it seems that he endorses at
least the latter claim, i.e. that an aporia, if anything, is the
psychological state resulting from such arguments or, more precisely
from the equal or similar strength of such arguments, for he reaffirms
this claim in the discussion of the mistaken definition.
That aporiai have to do with competing arguments can be
attested by several other passages, although the equal or similar
strength of these arguments will nowhere be highlighted as clearly
as in the current passage. That the aporia is primarily thought to be a
psychological state might be surprising at first sight. However, I will
make a suggestion later on, about how this psychological meaning
can be applied to a situation of argumentative deadlock.
 2.2 ‘Aporia’ and the Definition of a Dialectical Problem
Having defined what a deduction and a dialectical argument is and
having preliminarily spoken about the endoxa, the reputable
opinions, the first book of the Topics raises the question what the
dialectical arguments are made of and what they are about. It quickly
turns out that the things arguments are made of are equal in number
to, and the same as, the things which the dialectical arguments are
about. How come? Arguments are made from premises, he explains,
and they concern problems. That premises and problems are equal
in number derives from the fact that they differ only in form:

[Text 4] A problem is different from a premise in its form. For stated in

this way ‘Is it the case that two-footed terrestrial animal is the definition of
man?’, or ‘Is it the case that animal is the genus of man?’ it is a premise;
but stated in this way: ‘Whether two-footed terrestrial animal is the
definition of man or not’ it becomes a problem (and similarly in other
cases). Consequently it stands to reason that problems and premises are
equal in number, since you may make a problem out of any premise by
changing its form.12

It might be confusing here that a premise is introduced as a

question. Strictly speaking, it is not a question like ‘Are human
beings biped?’ that can be used as a premise, but only a question
together with an affirmative answer (with a negative answer the
question would provide the contradictory premise). What is more
remarkable in our context is the purely formal definition of the
dialectical problem. If a problem arises by just adding an ‘or not?’ to
a premise (or, strictly speaking, to a yes/no question) everything
could be turned into a problem. If this were the last word about
dialectical problems, it could not serve as a basis for the
consideration of aporiai, or else it would render the aporia a triviality.
However, it seems that in this particular passage Aristotle is only
interested in the interchangeability of premises and problems, as he
wants to show that all dialectical problems ultimately manifest one of
the four types of predication (the four ‘predicables’), which again can
be shown to be exemplified by each and every premise.
It is no surprise, hence, that in later chapters of the first book,
namely in Chapters 10 and 11, the seemingly uninteresting case of
dialectical problems is taken up again. Chapter 10 makes the
following straightforward claim:

[Text 5] First, then, it should be determined what a dialectical premise

and a dialectical problem are. For not every premise or every problem
should be counted as dialectical: no one in his right mind would hold out
as a premise what nobody thinks or make a problem of what is evident to
everyone or to most people, since the latter contains no puzzle, while
nobody would concede the former (ta men gar ouk echei aporian, ta d’
oudeis an theiē).13

The progress of this passage in comparison to the former purely

formal definition (quoted in Text 4) seems to lie in the idea that not all
premises and not all corresponding problems are dialectical
premises or dialectical problems respectively. Rather a dialectical
premise, or so it seems, is supposed to contain a view that is held, at
least, by a few people. The dialectical problem is not any alternative
of the form ‘Is p the case or not?’; rather a dialectical problem must
have or must involve an aporia. I take it, hence, that the dialectical
problem is defined by involving an aporia and that this is the main
function of the notion of aporia in this part of the Topics, i.e. to
distinguish a real dialectical problem from the purely formal notion of
a problem. The problems that the dialecticians consider are
expected to include an aporia. Now, since the two notions of aporia
and problēma seem to be quite close to each other, one might
wonder how the former could be a suitable ingredient in the definition
of the latter. One possible explanation can be derived from the
previously considered definition of the aporia. If an aporia is indeed
thought to be the psychological state of puzzlement brought about by
equal or equally strong reasonings, the dialectical problem would
seem to be a problem that actually leads to such a state of
puzzlement. If the alternative ‘Is p the case or not?’ does not lead to
puzzlement, it is not a dialectical problem – no problem worthwhile to
be considered in a dialectical disputation.
We have good reasons, then, to look more closely into the
notion of a dialectical problem. The passage quoted in Text 5 is only
a preliminary account of dialectical premises and dialectical
problems. In the subsequent passages of Topics I.10–11 Aristotle
goes on to give a fuller account of these two notions. Since it will turn
out that the notion of a dialectical problem is in some sense
dependent on the notion of the dialectical premise, we have to deal
with these dialectical premises first.
 2.3 Dialectical Premises
Aristotle defines the dialectical premise as follows:

[Text 6] A dialectical premise is the asking of something acceptable to

everyone, most people, or the wise (that is, either all of them, most of
them, or the most famous), provided it is not contrary to opinion (mē
paradoxos); for anyone would concede what the wise think, so long as it
is not contrary to the opinions of the many (ean mē enantion tais tōn
pollōn doxais).14

A dialectical premise is a proposition presented to the dialectical

opponent that is accepted by or acceptable to one of the following
groups: to everyone, to most people, or to all wise men or most of
them or to particular wise men who have the best reputation.
Remarkably, the idea that opinions can have a good reputation
because they are held by the wise is qualified by the proviso that it
must not be paradoxical.
Obviously, this definition of the dialectical premise presupposes
aspects of what we know as ‘being endoxon’ – we should therefore
add the definition of endoxa from the beginning of the Topics:

[Text 7] Those are endoxa on the other hand, which seem so to

everyone, or to most people, or to the wise – to all of them, or to most, or
to the most famous and esteemed.15

Text 9 is taken from the first chapter of the Topics. In this context, it is
just meant to capture the generic difference between scientific and
dialectical arguments, as the former are to be derived from premises
that are established as true, while the later are derived from
premises that are only reputable, but not established as true or
known to be true.
There are at least two ways of thinking about the definition of
endoxa and the corresponding definition of the dialectical premise.
Either one assumes that Aristotle wishes to define what is the single
endoxon-position regarding each question, or one assumes that
Aristotle not only defines what is endoxon, but at the same time
distinguishes what is endoxon-in-relation-to-all, what is endoxon-in-
relation-to-most, what is endoxon-in-relation-to-all-experts, what is
endoxon-in-relation-to-most-experts and what is endoxon-in-relation-
to-a-few-but-most-renowned-experts.16 For the first reading it is
important that there is a diminishing authority,17 for only those
opinions (doxai, ta dokounta) qualify as endoxa that are accepted
either by all or, if there is no such generally agreed upon view, what
is accepted by most or, if there is nothing like that, what is accepted
by the experts; and only if there is no standard expert view will the
endoxon position on this question be determined by individual
experts, however not by a random selection of individual experts, but
only by the views of the most renowned or most reputed experts.
Supporters of this reading usually refer to the proviso expressed in
Text 8 that the view of individual experts must not be contrary to
opinion.18 In this proviso they find the idea that Aristotle wishes to
identify the one view that deserves to be called endoxon, while not
allowing of a plurality of endoxa on a given question. However, the
proviso that excludes views that are contrary to opinion is part of the
definition of the dialectical premise, not of what it means to be
endoxon;19 and as for the use of premises in dialectical disputation,
it is quite clear that the questioner cannot expect to get assent to a
view that is contrary to opinion.
On the second reading at least some of the competing accepted
opinions about a certain problem are classified in accordance with
the group by which they are held. An argument for the second,
relational, reading is provided by the following passage from the
Rhetoric:

[Text 8] Since the persuasive is persuasive to someone …, and since no

art examines the particular – for example, the art of medicine does not
specify what is healthful for Socrates or Callias but for persons of a
certain sort (this is a matter of art while particulars are limitless and not
knowable) – neither does rhetoric theorise about each opinion – what
may seem so to Socrates or Hippias – but about what seems true to
people of a certain sort, as is also true with dialectic (alla to toioisdi,
kathaper kai hē dialektikē).20

In short, the passage says that just as rhetoric has to consider what
is persuasive in relation to groups of a certain type, dialectic must
distinguish what is endoxon in relation to several groups. Now, since
there is no other context in the Topics that distinguishes groups of
addressees – e.g. what is endoxon for sportsmen, what is endoxon
for musicians, what is endoxon for retired people and so forth – the
only possible reference in the Topics is the definition of endoxon and
the dialectical premise given in Texts 6 and 7, which on the relational
reading turns out to include a subdivision of types of endoxa.
Similarly, a passage in the Topics about the collection of endoxa
clearly seems to presuppose that there are different classes of
endoxa:

[Text 9] Now, premises are to be collected in as many ways as were

defined in connection with premises: making ready for use the opinions
either of everyone, or of the majority, or of the wise (and of the latter, the
opinions of all, or the majority, or the most famous), or the contraries of
opinions which appear to be so, and whatever opinions are derived from
the arts.21

Text 9 quite clearly states with respect to the collection of accepted

opinions that they should be sorted and classified in accordance with
the several groups that accept them. And since the groups
mentioned are the same as the ones that were mentioned in Texts 6
and 7, it is quite plausible to assume that this standard subdivision of
groups is generally meant to introduce a relational reading of the
endoxa.
In adjudicating between these two readings, it might be useful to
briefly consider the role that the endoxa play in the dialectical
question-and-answer disputation. The questioner is supposed to
present propositions as questions for which he needs to get the
assent of the answerer. For that purpose it would be useless (as Text
5 explicitly says) to hold out premises that virtually nobody would
accept. Hence, as a general rule, the questioner should only present
questions that are accepted by one of the relevant groups and are,
hence, acceptable for the answerer too. At the same time, it is only
important that the propositions used are indeed endoxon – no matter
by whom they are actually accepted. In order to formulate this basic
requirement for dialectical premises there is no need to indicate
which proposition is accepted by whom and hence one can stick to
an across-the-board-reading as opposed to a relativised reading.
However, among the opinions that, in principle, have a good
reputation (by being accepted or supported by one the groups
mentioned), some have a better and some have a worse reputation.
And in order to rank them in accordance with the degree of
reputation, one must consider by whom (by all, by the most, by the
experts …) they are accepted. This means that one has to adopt a
relativised way of considering the endoxa. Clearly, the reputation of a
proposition that it is literally accepted by all relevant groups
(including the many and the experts) would be ranked higher than
the reputation of a view that is only held by a few experts. In
addition, the relativised way of considering the endoxa becomes
indispensable if one starts distinguishing, as Aristotle does, between
propositions that are endoxon without qualification (haplōs) and
those that are endoxon for certain people.22 It is clear, then, that the
dialectical context of endoxa makes it indispensable to take into
account by whom a certain proposition or premise is accepted,
although it is often sufficient to speak of reputable propositions
across the board, i.e. without relativising them to different groups.
Also, if it is possible to rank endoxa in accordance with their
reputation, it is also possible to introduce comparatives and
superlatives with regard to being-endoxon, as is done in Book VIII of
the Topics. If a thesis is accepted by literally all relevant groups, it is
‘more endoxon’ or even ‘most endoxon’. What is more endoxon is
more likely to be accepted by the opponent than what is less
endoxon. And it would impair the quality of a dialectical argument, if
the questioner tries to deduce a more reputable conclusion from a
less reputable one,23 just because in these cases the premises could
not lend any further support to the conclusion. In this context of
comparing and ranking endoxa one might thus move from the factual
acceptance of a particular proposition by a particular group to
something like its acceptability – i.e. the acceptability that derives
from a proposition’s having a better or worse reputation – without
thereby implying that there is a tension between the interest in the
most acceptable view and the relativised treatment of endoxa.
Why is it, then, that the dispute between the two camps
concerning endoxa is carried out with fierceness? The main point of
contention concerns the implication of the relativised view that there
will be conflicting endoxa; while the relativised camp sees no
problem in allowing such conflict, the non-relativised camp wishes to
preclude it. Why should one want to preclude such conflict? It
presents a problem for those who expect the proposition deemed
endoxon in a given dispute to indicate what is true and the dialectical
selection of endoxa to express ‘a serious concern for truth’.

2.3.1 Are There Conflicting Endoxa?

If a proposition can be qualified as endoxon through being accepted
either by all or by most or by all experts or by certain experts,
nothing prevents the possibility of conflicting endoxa on one and the
same question – although there is often no need to specify by which
group an endoxon is accepted and although the dialectician would
not use a proposition that is held only by the wise but contradicts the
view of the many as a dialectical premise. Why should we think,
then, that there is a problem with conflicting endoxa? In Top. I.11,
104b1–524 Aristotle speaks frankly about conflicting views held by
the wise, the many etc.: Berti, who presupposes that there could be
no conflict between endoxa, takes this to be an instance of conflict
between mere opinions (doxai, dokounta).25 However, if the opinions
are held, e.g. by the wise, why shouldn’t they deserve to be called
endoxa? After all, they have the credential of being accepted by the
wise, as required in the definition of endoxa. Similarly, the following
text is an unambiguous indication that, for Aristotle, arguments
deriving from endoxa – not only from doxai – can be used for
contrary conclusions:

[Text 10] For the deductions are taken from accepted opinions, and
(such) opinions often contradict each other.26

Berti again criticises the revised Oxford translation for inserting the
‘such’, as he takes it that the opinions in the second clause are
different from the endoxa in the first clause.27 However, the claim
that opinions often contradict one another is meant to explain why
deductions taken from endoxa can issue in contradictions, so that
the claim in the second clause must apply to endoxa as well.
Here is a more serious concern. In Top. VIII.5 Aristotle
formulates rules for the answerer regarding the acceptance of
premises. He points out that the answerer should only concede
premises that are more reputable than the conclusion. If the
questioner wants to establish a conclusion that is itself adoxon, i.e.
of no reputation at all, then the answerer may concede any opinion
of some repute. If, by contrast, the questioner wants to establish a
conclusion that is endoxon, the answerer is to admit only those
premises that are even more endoxon. It is presupposed in this
context that the conclusion that the answerer wants to establish is
contrary to the thesis of the answerer; and this opposition between
the two theses also applies to their degree of reputation:

[Text 11] If the thesis is unacceptable, then, the conclusion must become
acceptable, and if acceptable unacceptable (for the questioner always
concludes the opposite of the thesis).28

It seems, then, that the opposite of an endoxon is always an adoxon,

something that is not accepted or even unacceptable.29 If so, it is
impossible that there should be opposing endoxa, for what is
opposed to any endoxon would be adoxon. This passage actually
conflicts with Text 5 and Text 12, for in these two texts from Book I of
the Topics it is clearly said that a dialectical problem would not come
about if the things are evident – and things would be evident if one of
the two opposing views in a dialectical problem were wholeheartedly
acceptable and the other stubbornly unacceptable. It seems that we
get contrary views about the role of endoxa if we emphasise one of
these two passages (Top. I.10–11 versus VIII.5) at the cost of the
other, and it is to be expected that scholars favour the passage that
fits better with their general point of view. However, there are quite
independent reasons for weakening the impact of the VIII.5 passage:
Unlike the pertinent passages in Top. I, the VIII.5 passage does not
occur within a definitory context, but just presents an executive rule. I
suggest that VIII.5 might put the opposition between endoxon and
adoxon too schematically, first, for the sake of brevity and, second,
for the purpose of distinguishing both of these from theses which
cannot be said to be either endoxon or adoxon. More importantly,
these rules are taken from a context that stresses that there are
several degrees of being endoxon or adoxon, so that these two
categories might have been taken as a shorthand for being more or
less endoxon; this would not threaten the main lesson of the
passage, that the conceded premises should be more endoxon than
the conclusion. Finally, if we suppose that endoxon should be
understood in a relativised way, it is true that the opposite of what is
endoxon-for-x is adoxon-for-x, but this does not exclude the
possibility that the opposite of what is endoxon-for-x is endoxon-for-
y. From this perspective, Aristotle’s division here between being
endoxon and being adoxon might be seen as a simplified way of
speaking that, nevertheless, does not jeopardise the results of Top.
I.10–11.
Let us finally mention the philosophical rationale behind the
resistance to a relativised understanding of endoxa. Scholars like
Berti and Bolton wish to challenge Brunschwig’s dictum that dialectic
involves ‘no serious concern for the truth’30. The concern for the truth
that they claim to detect in dialectic must be reflected in the
dialectician’s choice of the endoxa.31 Berti is hence interested in
what he calls ‘the truth-value’ or ‘the high truth-value’ of endoxa.32
However, no part of the dialectical method we considered so far
requires that the endoxa that the dialectician deploys be true or
almost true. On the contrary, dialectical reasoning is defined by its
exemption from the truth-requirement of the premises. If, in certain
contexts, Aristotelian dialectic is concerned with propositions that are
most endoxon, this does not strictly imply that they are privileged or
preferable with a view to truth, but only means that they are
preferable in terms of their reputation to propositions that are less
endoxon. This is all that dialectical disputation requires. But isn’t it
the case that better attested opinions are more suitable for leading
us to truth? Perhaps. However, to make this point we would need
certain additional assumptions (for example, that opinions in general
are apt to hit the truth, that opinions with good reputations are
indications of what is true), and such assumptions, even if they can
be found in Aristotle, are alien to the account of dialectical
disputation and its method. There is no single passage in which the
dialectical disputants are advised to assess the premises they
choose with a view to truth. At this point, Bolton and Berti could reply
that all this might be true for the gymnastic use of dialectic, but not
for its non-gymnastic purposes.33 To this I would reply, first, that
these non-gymnastic uses are not at issue in the Topics and,
second, that it has already been shown how the dialectical method is
useful in its non-gymnastic applications: by training in dialectical
method and engaging in many question-and-answer disputations, we
acquire a proficiency in assessing opinions held by different groups
of people and in seeing the assets and drawbacks of different
accepted opinions and their implications, so that we become better
able to discuss potential scientific principles without yet knowing
whether they are true or not.
 2.3.2 Interim Conclusion concerning Dialectical Premises
Dialectical premises are questions presented to an opponent; the
propositions included in these questions are expected to be
endoxon, i.e. accepted by all or most or the experts, and if they are
only accepted by single experts they must not be paradoxical, i.e.
contrary to common opinion. The answerer can be expected to
accept premises that are endoxon, but he is not expected to
concede premises that are less endoxon than the conclusion the
questioner tries to derive. Hence, dialectical premises might be more
or less endoxon in that they have a better or worse reputation. On
certain topics there are no views that are endoxon, while on others
there might be more than one view that is endoxon – if these views
are accepted by different groups, e.g. the one by the many, the other
by the experts. If certain views are only accepted by individual
experts, but not by common opinion, they might be paradoxical (in
which case they cannot be used as dialectical premises); however, it
might also happen that the many (and so common opinion) simply
have no stance on a particular matter. Views held by certain
acknowledged experts may also conflict with views held by other
acknowledged experts. Sometimes dialecticians even argue on the
basis of views that qualify as endoxa only in relation to a particular
expert or philosopher.
 2.4 The Definition of the Dialectical Problem
Here is, finally, Aristotle’s definition of the dialectical problem.

[Text 12] A dialectical problem is a point of speculation (theōrēma)

directed either towards choice and avoidance or to truth and knowledge
(either on its own or as working in conjunction with something else of this
sort), about which people either have no opinion, or the public think the
opposite of the wise, or the wise think the opposite of the public, or each
of these groups has opposed opinions within itself.34

If we bracket for a moment the remark that people have no opinion

at all, the definition says that dialectical problems come about if there
are conflicting views favoured by different groups of people. He
explicitly mentions conflicts between the wise and the many, conflicts
between several supposedly wise people, and conflicts within the
group called ‘the many’. Aristotle even attempts to give a basic
typology of these conflicts: they come about if (i) the many hold
opinions opposite to the wise or (ii) the wise hold opinions opposite
to the many or (iii) each of these groups have divided views within
itself; the latter possibility may imply that (iii.a) either the many are
divided on a certain topic or (iii.b) the wise are. Clearly, this
classification builds on what has been said about endoxa and about
dialectical premises. It presupposes that certain views can have a
good reputation within one group, but not in another.35 At first
glance, it may seem awkward that Aristotle distinguishes between
cases (i) and (ii). Does it really matter whether the wise disagree with
the many or the many with the wise? The solution lies in the fact that
the definition of the dialectical problem presupposes the definition of
the dialectical premise. In Text 4 Aristotle had said that the dialectical
premise and the dialectical problem differ only in form, while in Text 5
he pointed out that not every premise and not every problem are
dialectical, for no one would hold out a premise that is not accepted
by anybody, while no one would make a problem of what is evident
to everybody. It seems to be assumed, then, that a dialectical
problem includes a dialectical premise (or at any rate, the proposition
expressed by such a premise). Further, it is assumed that, although
the problem includes a presumably accepted view, there is a
problem because this view is not evident to everybody, e.g. because
it is not entirely uncontested. The difference between cases (i) and
(ii), then, seems to be worth mentioning for Aristotle because one of
them assumes an underlying dialectical premise which is accepted
by the many but contested by the wise, while the other assumes an
underlying dialectical premise that is accepted by the wise but
contested by the many. At any rate the difference between the two
cases makes sense only if Aristotle assumes an underlying default
position and, since the definition of the dialectical problem is meant
to build on the definition of a dialectical premise, this default position
is the one expressed by the presupposed dialectical premise.
A dialectical problem does not necessarily come about merely
by affixing ‘or not?’ to a dialectical premise. Since the dialectical
premise includes a proposition with a certain reputation, the
dialectical problem derives from such a premise only if the
alternative ‘or not?’ can also claim to have a certain reputation,
however preliminary or contestable. Unlike in the general
characterisation of dialectical problems in Text 5, Aristotle does not
mention here an aporia explicitly, but it seems to be understood that
this tension between a view that is accepted by one of the relevant
groups and the existence of some warrant or credential for the
opposite view gives rise to puzzles.
If a dialectical problem requires some reputation for both
alternatives, what should we say about the situation that Aristotle
mentions first in his definition, namely, that in which people have no
view at all? The quoted translation by Smith translates the Greek
phrase ‘peri hou … oudeterōs doxazousin’ as ‘about which … people
have no opinion’. The ‘people’ here are, strictly speaking, the two
groups mentioned afterwards, i.e. the many and the wise. ‘Having no
opinion’ is not entirely clear, as it could mean that they have not
formed an opinion on this particular point at all. The Greek wording
rather seems to require that the many or the wise accept neither side
of the alternative expressed in the problem.36 This means that there
are accepted views on a particular question, but that these accepted
views exclude both alternatives.
Aristotle said in Text 5 that not all problems are dialectical but
only those that are not evident to everybody (meaning those whose
solution is not evident to everybody). A problem would be evident, if
there were credentials or warrants (in the form of reputable opinions)
only for one side of the alternative. However, the solution of a
problem is no longer evident if there are conflicting credential or
warrants (in the form of reputable opinions) of both sides or if the
available reputable opinions seem to exclude both alternatives.
 In the lines following Text 12, Aristotle first elaborates on the
points mentioned in the definition and then takes up the
characterisation of dialectical problems again by referring them to an
aporia:

[Text 13] Those are also (dialectical) problems concerning which there
are contrary deductions (for there is an aporia whether it is so or not,
because there are persuasive arguments about both sides) …37

This broadens our understanding of a dialectical problem; it comes

about when there are not only conflicting endoxa (or endoxa
excluding both sides of a contradiction), but also persuasive
arguments in favour of each of these propositions. Again, Aristotle
characterises the dialectical problem by the aporetic state it brings
about – this time the aporetic state is due to the conflict of similarly
convincing arguments – conflicting in that they argue for conflicting
conclusions. What these arguments have in common with the
endoxa mentioned in the previous section is the fact that both are
thought to provide a sort of warrant or credential for a certain point of
view – the one through the fact that a certain view is held by one of
the groups we regard as relevant and reputable, the other by
providing a line of reasoning to the effect that a certain view follows
by necessity – even if we are not yet thereby persuaded of this
particular view. The conflicting provisional warrants are puzzling; the
puzzlement is brought about by the implicit idea of competing
arguments or by contrary conclusion deriving from seemingly
persuasive arguments.
 2.4.1 Interim Conclusion concerning Dialectical Problems
My interim conclusion concerning dialectical problems is obvious,
then. Dialectical problems are characterised or defined as those
problems that involve an aporia, and in particular, an aporia brought
about by the equal or, at any rate, comparable strength of opposing
provisional warrants – whether these are endoxa held by different
groups of people or arguments that one could put forward for
opposing conclusions.
 2.5 Still More Dialectical Problems
For reasons that should be obvious by now, a dialectician should be
equipped with a large repertoire of dialectical premises; apart from
the propositions that are accepted by all or most or the wise etc.,
Aristotle recommends using (a) things which are similar to what is
accepted, (b) the contraries of things which appear to be accepted if
one puts them forward by negation and (c) opinions that derive from
the established arts.38 For example, if it is accepted that the
knowledge of contraries is the same, then it is also likely to be
accepted that the perception of contraries is the same. Aristotle does
not explicitly comment on whether people are just misled by the
similarity into thinking that the second proposition is also accepted, if
the former is, or whether the similarity is actually thought to provide a
good reason for accepting the second proposition. Is the similarity a
means of deception or does it provide a legitimate way to increase
the dialectician’s repertoire of dialectical premises? After all, the
Topics also lists topoi based on similarity39 and the bulk of the topoi
are certainly not meant to instruct in the formulation of deceptive
arguments. As an example of the negation of a contrary, Aristotle
mentions that if it is accepted that one must do good to one’s friends,
it also seems to be accepted that one must not do harm to them; for
doing harm to friends is the contrary of doing good,40 so that the
negation of this contrary should be equally acceptable. Indeed, the
Topics introduces topoi working on the same sort of assumption;41 in
the contexts of these topoi Aristotle points out that doing good to
one’s friend and harming them is contrary, because it derives from
contrary character traits, so that the negation ‘not harming friends’
could be said to belong to the same character as ‘doing good to
them’, so that, in turn, both assumptions should be acceptable for
the same sort of people.
At any rate, it is clear that, owing to the close relation between
dialectical premises and dialectical problems, a greater number of
dialectical premises may also increase the number of dialectical
problems. From this point of view, a dialectical problem may not only
derive from propositions that are explicitly accepted, but also from
propositions that are similar to the accepted ones or from the
negations of their contraries etc.
An additional source of problems is the phenomenon Aristotle
calls ‘thesis’ – defined as a view held by a particular philosopher that
is paradoxical, e.g. Antisthenes’ thesis that contradiction is
impossible, the Heraclitean thesis that everything moves, or
Melissus’ view that what is is one.42 Just as dialectical problems in
general can derive either from conflicting opinions or from contrary
deductions, these paradoxical positions come in two forms: either
there is simply a view or opinion that is contrary to common opinion
or there is also an argument that runs counter to our expectation or
opinion,43 as in the case of sophistical arguments. By definition,
each thesis amounts to a problem, as there is disagreement
between the majority view and the claims of particular philosophers
(while, of course, not every problem amounts to a thesis).44
There are many occasions, then, for dialectical problems to
come about. It would be an oversimplification to think that problems
only derive from manifestly opposing views. Sometimes they derive
from conflicting arguments, sometimes they may derive from
conflicting views that are similar to accepted views, from the
negation of the contrary of an accepted view, from theses or
sophistical arguments contrary to common opinion etc. There are,
then, several ways in which dialectical problems can be constructed,
and the dialectical method seems to aim in part at the construction of
such problems, if they are not already manifest.
Still, one should not enquire into every problem:

[Text 14] One ought not to enquire into every problem or every thesis, but
only those which someone might be puzzled about who was in need of
arguments (all’ hēn aporēseien an tis tōn logou deomenōn), not
punishment or perception.45

Although this quotation is taken from Topics I.11, a chapter that is

completely dedicated to the dialectical problem and although it has
already been said that only problems involving an aporia are
dialectical ones, Aristotle repeats here that one should not enquire
into each and every problem, but only in those by which someone
might be puzzled who is in need of arguments – as opposed to
people who could overcome their puzzlement by the use of their
sense organs or people who should be punished, if they wonder, e.g.
whether to honour their parents or not. One might read this as a
qualification of how and why someone is puzzled in the pertinent
sense: a dialectical problem leads to puzzlement that, in principle,
can be overcome by argumentation and does not indicate any other
sort of defect.
 3 Conclusion
In the first main section of this text we have shown how the
dialectical method is apt to facilitate the procedure of diaporēsai, i.e.
of working oneself through the problems connected with a certain
theme or topic. In the second main section we focused on the
definition of the dialectical problem; such problems come about if
there are conflicting reputable views or arguments about a certain
question. In several passages Aristotle explicitly mentions the state
of aporia that is brought about by such conflicting warrants or
credentials. Indeed, within the inventory of dialectic, the dialectical
problem came closest to what we know as Aristotelian aporiai. In
certain passages aporia seems to refer to the psychological state of
perplexity – although one can speculate that it is probably not the
psychological state as such that Aristotle is interested in, but what it
indicates, i.e. a conflict of some kind in the accepted views or in the
arguments concerning certain philosophical questions. Outside of
dialectic the notion of a dialectical problem naturally plays no or no
significant role. However, it seems that the notion of aporia not only
plays a significant role in Aristotle’s philosophical or scientific works,
but also bears the characteristics that are specified for the dialectical
problem. Instead of using ‘aporia’ as an indicator for dialectical
problems, it seems, Aristotle came to use ‘aporia’ straightforwardly
for the constellation of opposing warrants or credentials. Thus, the
consideration of the dialectical problem might shed some light on
Aristotle’s use of the notion of aporia. At the beginning of
Metaphysics Beta, where aporiai play the most prominent role,
Aristotle characterises aporiai in a way quite similar to what we
heard about the dialectical problem:

[Text 15] … about which one ought to be puzzled first: these are the
things concerning which people have held diverging views and, apart
from those, any that may have been overlooked (kan ei ti chōris toutōn
tugchanei pareōramenon).46

The things about which we have to be puzzled first are those

concerning which people have held diverging views, just like the
dialectical problems arise through diverging opinions or arguments. If
there are no manifest opinions on the one or the other point (if there
are, as Aristotle says, things that ‘have been overlooked’), the
philosopher might point out what the aporia consists in – most
notably by using the means that we discussed for the dialectical
problem.

1 Top. I.1, 100a18–21; translation by R. Smith.

2 Top. I.2, 101a34–6; translation by R. Smith.

3 See LSJ on diaporeō, II.2: ‘commonly only a stronger form of aporeō, raise
an aporia, start a difficulty’.

4 This can be derived from the fact that Aristotle quite regularly says that

one should raise difficulties first (cf. Met. III.1, 995a29. 34. 35. b2; EN VII.1,
1145b3f.), i.e. before one gets, as it were, to the real business, thus
indicating that the process of diaporein, strictly speaking, does not involve
the statement of the solution to the difficulties.

5 Exhaustiveness is stressed e.g. in Met. III.1, 994a34.

6 This aspect of structuring an enquiry and of defining its target by the
diaporēsai-procedure is stressed at the beginning of the ‘aporia-book’ Met.
III.1.

7 See e.g. Met. III.1, 995a25f.: tauta d’ estin hosa te peri autōn allōs
hupeilēphasi tines.

8 See e.g. Top. 108b19–22, 112a14–15, 121b29–30, 129b35–130a4,

158b16–23, 159a4–6.

9 Top. 145a37–b20.

10 Top. VI.6, 145b1–2.

11 Top. VIII. 6, 145b16–20; translation based on Pickard-Cambridge.

12 Top. I. 4, 101b28–36; translation by R. Smith.

13 Top. I. 10, 104a3–8; translation by R. Smith.

14 Top. I. 10, 104a8–12; translation by R. Smith.

15 Top. I.1, 100b21–3; translation by R. Smith.

16 This latter understanding is defended, e.g. by Smith 1997 and 1999. This
seems to be a natural reading of Text 7: Aristotle determines that dialectical
arguments are taken from premises that are not first and true, but only
endoxon, and then goes on to explicate that this notion can be related to
different groups.

17 See Brunschwig 2000: 116.

18 One might find, as Berti 2005: 179 does, the same proviso expressed in

Topics I.14, 105a37, if one emends the text correspondingly, but (i) this is not
the most plausible reading of the line, and (ii) it also refers to dialectical
premises.
19 Even Primavesi 1996: 47 n. 72, falsely assigns this proviso to the
endoxon. To my mind there can be paradoxical endoxa, namely views held
by certain philosophers (see my discussion of the thesis in Section 2.5),
although they should not be used as dialectical premises.

20 Rhet. I.2, 1356b28–35; translation by G. Kennedy.

21 Top. I.14, 105a34–b1; translation by R. Smith.

22 See e.g. Top. VIII.5, 159b23–33.

23 This is, I take it, the meaning of ‘deducing well’ (kalōs) in Top. VIII.5,

159b8.

24 See below, Text 12.

25 Berti 2005: 179.

26 Rhet. II.25, 1402a33–4.

27 Berti 2005: 186.

28 Top. VIII.5, 159b4–6; translated by Smith.

29 Of course, this is also mentioned by Berti 2005 as a crown witness.

30 Brunschwig 1984/85: 40.

31 See Bolton 1990, sections VII–VIII.

32 Berti 2005, passim.

33 Here is what Bolton actually says 1990: 188–9: ‘If dialectic involves “no
serious concern for truth” […] then it is hard to see how it could be rational to
use dialectic to perform the nongymnastic functions.’
34 Top. I. 11, 104b1–5. Translation by R. Smith, slightly adapted. The word
theōrema is difficult to translate in this context; I take it to mean the object of
scrutiny, enquiry or speculation. The words hoi polloi tois sophois ē are
absent from the Parisinus Coislinianus 170 and from Alexander’s
paraphrase; however, they are contained in the majority of manuscripts and
in Alexander’s citation.

35 See the discussion in Section 2.3 and, in particular, Section 2.3.1, above.

36 Pickard-Cambridge’s translation ‘hold no opinion either way’ comes

closer to ‘oudeterōs doxazousin’. See Wagner and Rapp 2004: ‘wovon
entweder keine von beiden (Antworten) für richtig gehalten wird … ’

37 Top. I. 11, 104b12–4, translation by R. Smith.

38 Top. I.10, 104a12–5.

39 Top. II. 10, 114b25–34.

40 Top. I. 10, 104a20–5.

41 Top. II. 7, 112b27–113a19.

42 Top. I. 11, 104b19–24.

43 Top. I. 11, 104b24–8.

44 Top. I. 11, 104b29–34.

45 Top. I. 11, 105a3–5, translation by R. Smith.

46 Met. III.1, 995a25–7.

 Chapter 7
Aporia in Aristotle’s Metaphysics Beta
◈

Friedemann Buddensiek
 1 Introduction
At the beginning of Metaphysics Beta Aristotle presents a list of
aporiai which he then carefully elaborates in the main part of B.1
These aporiai concern the science of the first principles and the first
principles themselves. They consist of pairs of important, though
incompatible views concerning that science and the principles –
views that others have held as well as other views that might arise.
The aporiai are raised, as it seems, because the awareness and
discussion of them as well as the solution to them is a means to
accessing those principles: The aporiai  are mentioned and
presented not only because they concern that science and its
principles, but because their discussion (their ‘diaporein’) is helpful or
necessary for establishing that science and for finding and grasping
the principles. This – being helpful or necessary – is the reason for
their presentation.
The elaboration of the aporiai in Met. B is preceded by a brief,
yet presumably programmatic introduction, which presents reasons
for presenting the aporiai.2 I will focus on this introduction and on
some questions it raises. In particular, I want to pursue the question
of why the aporiai need to be raised (the ‘necessity question’), as
Aristotle seems to believe. This question is crucial for our
understanding of the rationale and motivation behind Met. B, but it
could seem to be an odd question as well: Whatever their exact
relation to dialectic is, it might seem to be obvious that aporiai form a
crucial part of the only way to the principles of first science: how else
should progress be possible? So it seems that in dealing with the
necessity question, we are merely stating the obvious. Still, merely
stating that aporiai need to be raised and discussed is not the same
as saying why this should be necessary.3
I want first to make a few remarks on the objects (2) and on the
subject (3) of the aporiai in B as well as on the (lack of) systematicity
in the layout of the aporiai (4). I will then comment on the formal
structure of the question that constitutes an aporia (5) and on the
reason why the aporiai arise (6). In the last section I want to discuss
the question of why aporiai thus understood need to be raised for
establishing the science we are searching for in metaphysics (7). I
will try to show that raising, discussing and answering them is
necessary for acquiring the capacity which consists in the grasp of
the principles of the science we are searching for and, thus, for
becoming sophos in the proper sense outlined in Met. A 1–2.
 2 The Science Being Sought for – the Objects
of Aporiai
Met. B starts with mentioning the epizētoumenē epistēmē – the
‘science being sought for’. This is the science for the principles of
which we need to go through the aporiai. The aporiai and their
discussions are not serving just some preparatory purpose. Instead,
they are dealing with the science itself as well as with its
fundamental objects. Discussing and solving aporiai means to
develop the science. This requires that we know what this science is
concerned with.
According to Met. A, the object of the science we are searching
are first causes and principles (see A 1, 981b27–9, 982a1–3, A 2,
982a4–6, b7–10). For Aristotle, it is clear (1) that sophia is an
epistēmē concerning aitiai and archai (see e.g. A 1, 982a1–3, A 2,
982a14–17, B 2, 996b8–10), (2) that we are searching (zētoumen,
A 1, 982a4) for this epistēmē and (3), because of this, that we have
to pursue the question of which aitiai and archai this epistēmē is
addressed to. Aristotle mentions the zētoumenē epistēmē again in
his résumé of the task at hand: ‘it has been said what is the nature of
the epistēmē being searched for’ (A 2, 983a21–3) and connects this
epistēmē with the aítia ‘ex archēs’: knowledge is knowledge of the
‘starting’ aítia, the ‘original causes’ (A 3, 983a24–6).
A ‘useful’ (prourgou, A 3, 983b4) discussion of the use his
predecessors made of principles and causes confirms that there are
no other kinds of causes than those established in the Physics (A 3,
983b5–6, A 7, 988a18–23, b16–19, A 10, 993a11–17), even though
the predecessors have touched upon (988a23, b18) them only in an
obscure way (amudrōs, 988a23). Having confirmed those principles,
Aristotle announces that he will go through (dierchesthai) the
possible aporiai connected with the principles, as they have been
understood by his predecessors (A 7, 988b20 f.).
The final lines of Met. A (A 10, 993a25–7) apparently lead to
Met. B.4 They again announce a presentation or discussion of
questions concerning the topics dealt with in A (A 10, 993a24–7).
Within B itself not every aporia is explicitly concerned with aitiai and
archai. However, (a) many aporiai deal with principles or aitiai; (b)
the science we are searching for would need to be a unified science;
(c) an Aristotelian science does depend on principles (and on a
clarification of what they are); and (d) a science of being – as it may
seem to be indicated in some aporiai – should not be a rival to a
science depending on archai, but should be integrated within such a
science. In fact, B itself seems to take it for granted that the search
for our science is a search for the aítia and archai: this is not
disputed, but simply presupposed when Aristotle asks what science
it is we are searching for, if it should be the case that there are
several sciences that are addressed to archai and aítia (see B 2,
996b1–3).
 3 The Subject of an Aporia
There may be questions about the subjects of the aporiai: Aristotle
does not present anyone who actually sees himself in an aporia of
the kind he presents nor does he say that he himself finds or once
had found himself in those aporiai. Many philosophers will hold views
that imply problems or are even untenable, but most of them will
knowingly hold just one of two opposing or contradictory views and
many enough of them will not see the problems implied by this view.
Various interpretations or questions might come to mind: (a) Is
an aporia – as we find it in Met. B – an epistemic state the readers
(or Aristotle’s audience) should find themselves in – for didactical
reasons – if they want to advance to an understanding of Aristotle’s
first science?5 (b)  Is an aporia an ‘impersonal’ problem, that is, a
problem that consists of incompatible views each of which may be
held by someone, but which are not being held in combination by
any one person in particular – so that an aporia represents
historically the current (construed) state of the discussion?6 (c)  Do
the aporiai represent Aristotle’s own state of mind (while he was
writing B)?7 (d) Is an aporia a problem anyone finds himself in, who
has not solved this problem himself, even if he may not be aware of
this problem?
In the last section, I will try to show that stating, going through
and solving aporiai is necessary for each one of us, if we want to
make progress and to succeed in the case of metaphysics. Raising
and going through aporiai is not just a didactical operation nor does it
merely serve the purpose of facilitating progress. An aporia is also
not simply an impersonal problem that someone could put together.
While any aporia that has not yet been stated (and solved) could be
regarded as such an impersonal problem, we need to state, to go
through and to solve aporiai for ourselves, if we want to make
progress. This holds for humankind as well as for any individual on
his or her way to the first science, though the progress will differ in
either case, for instance, because progress of humankind concerns
the development of a body of knowledge, while progress of an
individual concerns the development of a capacity.
 4 The Starting Point for Aporiai and Their
Lack of Systematicity
While Aristotle does begin by referring to views held by others that
require our attention (see B 1, 995a25–6),8 he does not present it as
a point of particular importance that we should begin our search for
the first science with a discussion of endoxa. He does not make the
point (as he does elsewhere) that endoxa may contain some truth
which would help us in elaborating the point in question and which
we therefore have to take account of. That his focus is not on
endoxa as such also becomes clear from his remark that some of
the issues that need to be discussed have been the object of
different views held by others, whereas some other of those issues
have not been discussed by others, but have been overlooked
(995a25–7).9
This may be surprising, if we notice the great extent to which
Aristotle will refer to beliefs held by other philosophers in the bulk of
B. It will be less surprising, though, if we remember that Aristotle is
concerned with finding the truth about principles. He does not
exclude or dismiss whatever may be retained from views held by his
predecessors. But he also needs to determine what the truth in those
views is and, moreover, how we can gain new ground, that is, how
we can extend our knowledge. We are searching for the first
science, which means we are searching for the truth about principles
and first causes. Raising the aporiai – whether they have been
raised by others or not – going through them and solving them is the
only way to do this. Relying on endoxa as such would not suffice.
A further question regards the aspects of the first science and of
the principles that form the basis of this science – that is, those
aspects on which we are to focus our search and with regard to
which we are to frame and raise aporiai in order to make progress.
Aristotle does not give a systematic outline of the aspects with which
the aporiai should be concerned nor does he indicate what would be
a complete account of the aporiai to be raised. Does he think we
could be confident that we will have covered all important questions
and problems pertinent to the epistēmē we are searching for (as B 1,
995a25–7 suggests), although we have not properly discussed how
many problems concerning principles there are and how they relate
to each other? If Aristotle is interested in an exhaustive account of
the aporiai,10 he does not discuss their number or the criteria for
their selection.
He may be confident that his predecessors – in their desire to
know and equipped by nature, in principle, with the right means – got
something right, when they developed their views concerning
aspects of the first science in some way or other, and this may be a
reason why he is concerned with their views, even if there is a lot in
them that is not true or not precise. He might even be confident that,
over time, somehow all aspects will be touched – given that our
natural desire to know is not in vain. But if there is no systematic
account of aspects to be dealt with in our search for the first science,
we can never be sure to have successfully completed this search.11
There could always be some further aspect which we have
overlooked so far or with regard to which we have not raised all the
required aporiai. We should not be surprised then, if the discussion
of those aporiai also led to a revision of previous solutions – given
that those discussions all contribute to the one science we are
searching for.
Some further remarks with regard to the systematicity of the
layout of the aporiai in B are in order: (a)  There is a disagreement
about the actual number of aporiai, and this is also due to the larger
question of how to individuate aporiai. (b) There is a mismatch
between the preliminary list of aporiai in B 1 and their exposition in B
2–6: not every aporia listed in B 1 is dealt with in B 2–6 (see B 1,
995b20–7, 996a11 f.) and vice versa (see B 6, 1002b12–32). (c) The
order of the list of aporiai in B 1 differs from the order of their
discussion in B 2–6. (d) Aristotle indicates a priority among the
aporiai with regard to their difficulty, but also with regard to the
necessity to discuss them, though he does not present criteria for
either priority.12
 5 The Formal Structure of the Aporia-
Questions
What is the structure of the questions that form the aporiai? They are
not simple questions like ‘how many sciences have to study the
aitiai?’, ‘what kinds of substances exist?’. Instead, they are
alternative questions like ‘is it the task of one or of more sciences to
investigate all genera of aitiai?’, ‘do only perceptible substances exist
or do others exist alongside these as well?’. That is, the structure of
the questions is: ‘is A the case or is B the case?’. In most cases, the
members of the disjunction are evidently incompatible (an exception
could be 996a1–2, which poses the question of whether the archai
are determinate in number or in eidos). Aristotle expresses this
either in an explicit or in an implicit way: In some cases, he presents
the question in the form of ‘is A the case or is non-A the case?’
(contradictory form; see 995b31–3, 996a12–14, see also 996a4–9).
In the majority of cases, however, he presents it in the ‘A or B?’-form
(contrary form). In this latter case, non-A may usually be inferred
from B – for instance, in cases like ‘one or many?’ (B is either
equivalent to non-A or it implies non-A).13
However, this does not yet necessarily give us an aporia. For an
aporia we need a problem that is (prima facie) insurmountable.
There are two main ways to construe such a problem:
(a) In the first case of an aporia – which is the standard case –
the problem arises from the fact that there are two opposing views
neither of which is prima facie tenable (there is no obvious way out;
let us call this the ‘negative aporia’, since both sides are to be
refuted). In order to generate a problem in this case the opposing
views have to cover (on the face of it, at least) the whole ground –
that is, the disjunction has to be complete.
(b) In the second, less frequent case of an aporia, there are
(two) incompatible views or disjuncts that are both seemingly well
defendable – let us call this the ‘positive aporia’. The problem here is
that we cannot have it both ways. In this case we do not need to
stress the completeness of the disjunction.14
In the case of negative aporiai, the point is not just that either
side has its own problems (that is: the point is not that A has its
problems and that B has its problems): that is only part of the story.
The point is rather that, at least prima facie, there are only two ways
and that neither of them is viable. This has to be the reason and the
core of the aporia: A and B together are not viable. An aporia of this
kind presents a double or two-sided problem. (It is not the case that
the aporia comes to be because of problems connected with A or
because of problems connected with B.) It is necessary to see the
problems on both sides together – not as isolated problems.
Otherwise, we might in a rush turn to the other side, not taking into
account the problems that are waiting for us there.
Similarly with the positive aporia: here as well we have to take
both sides together – otherwise, there would be no aporia. If we
were to consider one side isolated from the other, we might not even
become aware of the problems connected with this side. However,
whereas in the case of the negative aporia A and B together are not
viable, in the case of the positive aporia A and B are not viable
together. This ‘non-viability’ is a central feature of the structure of an
aporia.
Admittedly, this is a rather simplifying account. According to
Madigan’s description and count (1999: xviii), (a) about a fifth of the
arguments in B are constructive in their aim and procedure; (b)
almost a tenth are constructive only in their aim, but refutative in their
procedure, in that they establish one of the two theses by refuting
the opposite thesis; (c) there is one argument that is refutative in aim
and constructive in procedure; and (d) the bulk of all seventy-seven
arguments (on Madigan’s count) are those arguments that are
refutative in aim and procedure.
A refutation of a claim does not necessarily give sufficient
informative reasons for the opposite claim. If the claims are
contradictory, it may happen that the refutation of one of them only
provides the information that the opposite claim must be true, and
even this may be correct, only if we understand the opposite claim in
the right way (this may be a problem for (b) to (d)). If, on the other
side, we proceed in a constructive way and provide reasons in
favour of one of the claims, we still do not know for sure, whether we
have provided sufficient informative reasons for the claim understood
in the right way, and our conviction that we have may be shattered
by similar arguments for the contrary claim (this may be a problem
for (a)). Nevertheless, this constructive procedure may in some
cases still be somewhat more informative, since it provides some
constructive reasons which will have to be taken into account.
The structure of the question allows progress insofar as it (i)
covers the logical options or important options concerning a feature
of the principle (or of the science building on it) without already
knowing what this feature amounts to (how it may be properly
described) and without committing oneself to a certain, unavoidably
still empty view concerning that feature. It allows progress
furthermore in that it (ii) covers options that together will not stand
and in that it (iii) instigates our desire to know: we cannot avoid
taking a view on the issue in question, if by nature we desire to
know. And since together the options are not acceptable, we have to
develop our thought on the issue in question in such a way that it
avoids the difficulties that arise from either side of the aporia or from
both sides taken together. Finally, (iv) the arguments for or against
either side indicate aspects we will have to consider in our attempt to
come to a proper understanding and description of the principle.
Without aporetic questions of this kind there will not be the carefully
framed answers we need.
 6 Lack of Understanding as a Reason for
Aporiai
A quick answer to the question of why an aporia arises points to the
contradiction among A and B (or their implications) and the missing
viability of A and B taken together. But this is not a satisfying
explanation: it does not give the reason why aporiai arise.
According to Aristotle, previous or contemporary philosophers
have not seen the issue at stake clearly (or, in some cases, they
have not seen the issue at stake at all). None of them has
considered all kinds of causes and of the causes they have
considered they had at best a vague, obscure understanding. The
use they have made of their causes was either not sufficient or not
consistent (A 4, 985a16–18, b19 f., A 5, 987a22–5, A  7, 988a20–3,
b14–16, A 10, 993a13–17). They stumbled, made wrong decisions,
allotted explanatory tasks to causes of the wrong kind or wrongly
understood. They have underestimated the task at hand and
overestimated the explanatory power of their principles. In this, they
were like the inexperienced person who may occasionally strike
lucky blows, but who does not rise to the situation at hand (cf. A 4,
985a13–16).
They may, for instance, have held the view that there are just
two kinds of aitiai, which they then have taken to provide the full
explanation of any physical phenomenon. Or they may have shared
a non-sufficient understanding of, say, the material cause, may have
been unaware of this insufficiency and may have entertained the
material cause thus understood in their theory. The wrong
understanding of the principle will have been due to the wrong
understanding of what the principle is supposed to explain (since the
understanding of the principle depends on an understanding of what
the principle is supposed to explain).15 A necessary condition for
views constituting an aporia is the misjudgement concerning the
range, explanatory function, explanatory power and implications of
the principle.
The aporiai arise, since the views (in connection with arguments
against or in support of them) arise which lie at the bottom of the
aporiai. These views, in turn, arise due to features of the object with
which they are concerned. The fact that there are conflicting views
must be due then, in part, at least, to the elusiveness of features of
this object and to special difficulties that arise, if we try to grasp
those features appropriately.16 The way those features are taken
account of will also be shaped and influenced by other theoretical
preferences and interests of those whose views are concerned with
those features. But the spectrum or range of views worthy of
discussion will be limited and determined to a high degree by
features of the objects themselves.
For instance, we might think that it is due to structures of the
world that a certain aporia about ousia arises: on one side – if we
take just one of the better known problems – ousia has to account
for the fact that there is knowledge about features of the world
(namely insofar as the first fundamental structure has to be the
object the grasp of which will be the basis for any other knowledge) –
and as such, ousia  has to be definable and has to have universal
features. On the other side, ousia has to account for the fact that
there are individuals or things determined in a certain way (see the
last aporia in B).
Now even in this case it would seem doubtful that there are
conflicting structures in the world that lead to this aporia.17 How
could there be such extra-mental conflicts? Structures in the world
are just what they are – whether they are universal or particular (or
neither). But as this aporia shows, these structures (principles,
elements, ousiai) are such that they may be seen from different
perspectives – perspectives that seem to exclude each other, if they
are not properly understood. These structures lend themselves to be
seen in either way. They lend themselves easily to assumptions,
inferences or explanations that in turn cause serious difficulties.
We may follow our unobservant, ill-founded preconception of
what the features in question amount to and what actually may be
derived from them. We also may generalise those features we have
picked out – and may thus – without being aware of it – generate
problems for our account of ousiai. Or we may be aware of all the
relevant features (of ousiai), may have a certain understanding of
them, may think that those features all matter, but are not
compatible, and may be aware that we are in an aporia. This second
option is some advance over the first, as it includes some awareness
of the relevant features as well as of the aporiai  which are due to
those features (though we may not yet be aware of the fact that it
was our insufficient understanding of those features that caused our
aporia). Unless we improve our understanding of those features, we
will not make any progress. In order to untie the knot we have to
figure out what is wrong with our understanding of the elements of
this aporia.
 7 Necessary or Merely Helpful? The Function
of the Aporiai
So far we have been talking about why aporiai arise. But why should
it be necessary for them to arise or to raise them – as B 1 suggests?
At the very end of Book A, Aristotle invited us to return again to the
points with regard to which aporiai might be raised concerning aitiai:
for perhaps, he suggests, we will thus make some progress
(euporein) towards our later aporiai (A 10, 993a25–7). If these lines
should refer to B as the text that contains those points to which we
are invited to return, it seems to be a rather mild demand on
pursuing aporiai.
However, in B Aristotle starts by saying that it is necessary for
the science we are searching for that we first go through those
issues about which one has to raise difficulties first (aporein, B 1,
995a24–5). He goes on to say that it is conducive (prourgou)18 to
progress for those who wish to make progress (euporein) that they
go through the aporiai well (diaporein kalōs) (995a27–8): for the later
solution of difficulties (euporia) is an untying (lusis) of those
difficulties previously raised (aporoumena), untying, however, is not
possible (ouk estin) for those who are unaware of the bonds
(995a28–30). Those who are in difficulties resemble those in bonds:
for in both cases it is impossible (adunaton) to move forward
(995a31–3). One must (dei) have contemplated all difficulties
(duschereiai) beforehand (995a33–4): someone who is searching
(for the science) and has not gone through the difficulties first, will
resemble those who do not know where to go and who do not know
whether they have found what they are searching for (995a34–b1).
For the goal (telos) will be obvious only to those, who have raised
aporiai beforehand (995b1–2). Furthermore, one is necessarily in a
better condition to judge, once one has heard the conflicting
arguments just as the opposite sides in court (995b2–4).
Aristotle seems to be saying in the first part of B 1 as well as in
the main part of B (see B 1, 995b13–14, B 4, 999a24–6, 1001a4–5)
that it is indeed necessary to go through the aporiai in order to make
progress. Does this mean that this is necessary for those who are
bound? Or does Aristotle want to say that everyone is necessarily
bound at first? The answer to this question would make quite a
difference to our understanding of the function of aporiai (at least in
Met. B). It seems quite unlikely that Aristotle would have introduced
the matter in the way he did in B 1, if he had only meant to say that
we have to deal with aporiai, only if we are bound – as if, at the
beginning of their journey, not everybody were bound by the
shortcomings of their yet undeveloped knowledge.
According to Politis, the process of searching in metaphysics is
both ‘essentially aporetic’ and ‘aims at objective truth’ (2003: 162;
173). ‘Essentially aporetic’ means that ‘a search must be conducted
by seeking to identify and resolve particular aporiai’ (ibid.: 146; 151,
162). Only if we proceed by means of aporiai (by going through
aporiai), will we make progress and finally reach the end. But it is not
only a necessary condition, but also a sufficient condition: that
means, if and only if we proceed by means of aporiai, will we make
progress (see ibid.: 152). This way is viable, and it is the only way.
Since the object of metaphysics is not vague, bizarre or aporetic in
itself, metaphysics is searching for something that there is, when it is
searching for the basic structures or principles of being – there is no
ambiguity or vagueness in the object. The crucial question is, then:
How can the search for such an object (or, for truths about such an
object) be essentially aporetic (see 162)? According to Politis, ‘a
search is both essentially aporetic and aims at objective truth if, and
only if, a) the search involves aporiai that are themselves in a
particular sense objective, and b) these aporiai are necessary to the
subject-matter of the search’ (165). For an aporia to be essential to
the search for a certain object or truth, the object itself has to be
such that it can be grasped only by means of (recognition of) aporiai.
It would not do to call the procedure of metaphysics ‘essentially
aporetic’, if the aporiai were only due to our own capacity of thinking
(or the lack thereof) (166 f.). Both sides – the object, our thinking –
have to be involved. Politis suggests that we may understand the
objective side in such a way that the object (of metaphysics)
presents itself to us in different ways (or, ‘perspectives’, if we may
borrow the analogy from sense perception; see 167) that are
seemingly incompatible. The object itself is of a certain kind, so that
we tend to confuse or confound the kinds of being – we may miss or
may not be aware of the distinctions required for a proper
understanding.
I would like to take up this approach and to pursue the question
concerning the necessity of aporiai a bit further with regard to the
aspect of what it takes on our side to make progress. For the aporia
to arise its object has to be of a certain kind and our thinking has to
refer to it in a certain way. But this may be simply a necessary
condition for an aporia to arise. It does not yet mean that it is
necessary for metaphysical research to proceed via aporiai. Perhaps
we can explain why aporiai arise – and if they arise we have to solve
them in order to proceed. But why should they be a necessary part
of metaphysical research?
The metaphors or comparisons at the beginning of B 1 (the
knot, the clueless wanderer, the judge) will not by themselves
provide an answer to the question of necessity. They tell us that the
person bound cannot proceed, that the clueless wanderer somehow
can proceed, but that he does so with no orientation as to his
direction or goal, and that the person who has been listening to both
sides will make better judgements. Nothing of this tells us why being
bound and going through aporiai should be necessary.
As for the clueless wanderer, the question comes down not just
to the question of how we find principles, but to the question of how
we find reasons for choosing this or that route and for accepting this
or that claim (about principles) as the end of our way. In the case of
principles of the first science, which we cannot deduce from some
further principles, we are stuck with a discussion of suggestions by
previous philosophers – a discussion that implies a discussion of
implications of their assumptions about principles, their consistency
and explanatory potential. This discussion will help us in
demarcating the path, it will teach us what we have to take into
account – positively or negatively. As for the end, a discussion of the
problems of either side, that is, a discussion of the complex problem
constituted by the problems of either side will show us the area for
the end of our journey, if not the end itself. Nevertheless, this again
does not explain why there should not be any other route to the
principles.
As for the judge, the person who listens to the opposing parties
and the reasons they present for their own or against the other side
will be in a better position to judge. She will weigh the arguments
and assess them according to their conclusiveness and the
plausibility of their presuppositions. She will need input of her own:
she will not trust either side without further reason. Still, she is in a
better position to weigh pro and contra arguments and to see how
her decision relates to the case at hand. Her decision will be better
founded. However, why should the correct judgement depend on
aporiai and the arguments brought forward by or against either side?
For an answer to the ‘necessity-question’ we should further
focus on the relation between the object (the principle) with its
somewhat hidden features and the subject (the researcher) with his
naturally feeble cognitive abilities. Given our nature, our first
approach in grasping things operates by perception. As Aristotle
himself indicates this in A 1–2, it is a long way from there to proper
recognition of what there is. And while by nature we desire to know,
we do not open up to the principles easily, as we can only work with
our capacities at each step to the extent we have developed them.
With regard to the object in question – the principles of the first
science – we seem to be bound by our nature for a long period of
our journey.
However, even though this reference to our initial lack of
comprehension may prepare for an argument for the ‘intrinsic’
necessity of diaporein, it is not yet clear, again, how it would support
a stronger claim than the claim that it is merely most helpful to go
through the aporiai. So far, it seems that neither a case of lucky
guesses (that is, a case in which one hits upon the principle more or
less by chance) – improbable as it is – is ruled out nor a case of
learning merely by being indoctrinated (where the indoctrination may
include reasons, but does not make use of aporiai on its way).
The usefulness is underlined in Top. I 2, where Aristotle
highlights the fact that the philosophical sciences will benefit from his
pragmateia, since it will enable us to go through the difficulties
(diaporein) on both sides and detect what is true and what is false in
them (101a34–6). It will also be helpful for the first tenets of any
science, which cannot be deduced from principles proper to that
science. Instead, we may access those principles by going through
the endoxa pertinent to them (101a36–b4). But this, again, does not
mean that we could access the principles only via aporiai.19
In order to deal with the ‘necessity-question’, let us take a closer
look at the lucky-guesses case. Why is it not just very unlikely, but
impossible to get a grasp on the principles of the science we are
searching via lucky guesses? For someone to set out for the
principles, something must have first sparked his interest and
something must have given him an idea of what the principles he
arrives at by guessing can, and need to be able to, explain: he needs
to have the conceptual means to pre-phrase the guessed principles,
just as he needs to have a preconception of what is in need of
explanation. Even these minimal requirements are already
demanding. Due to our nature the beginning of our endeavour is
determined by the epistemic gap between our at best fragmentary
understanding – constituted by the frame of our outlook and most
likely taking its point of departure on the side of perception – and the
object to be properly grasped. This gap is due to the difference in
complexity between our insufficiently complex recognition in its pre-
epistemic state and the object with its complex, seemingly
contradictory features. Were our recognition right from start as
complex as the object, we might immediately grasp the object. And if
the object were less complex – only as complex as our recognition in
the beginning – the same would hold. Closing the gap always
means, and requires from us, to get closer to the object, that is, to
gain scientific knowledge and to develop the required capacity – in
short, to become sophos (see A 2, 982a8–14, a21–8, A 3, 983b1–3,
984a18 f., b8–11).
The true sophos needs to know the principles of the first
science. For this, he needs to know why these – framed in this way –
and not some other items or alleged entities, are the principles, and
this includes knowledge of how the principles are principles of those
things of which they are principles. What kind of knowledge does this
include in the case of the first science? The true sophos must know
what follows from the principles, that is, what their explanatory power
is and should be. In many cases, the principles need to fulfil
seemingly conflicting explanatory demands (for instance, with regard
to universality on one side, causality of a kind on the other side). The
true sophos needs to understand what these demands are. He then
has to understand how principles can fulfil those demands. He has to
understand principles as solutions to explanatory needs. He will
understand this only, if he has worked through the different demands
concerning the principles that seem to be in conflict with each other.
Scrutinising these demands means, again, to determine their
implications as well as their explanatory power (and the limits of
this). The sophos-to-be will come to see why the principle properly
framed will succeed in answering the demands and why it will not fail
to do so.20
To acquire the capacity appropriate for this means to acquire a
complete set of distinctions, implications, presuppositions, reasons
and arguments, including reasons and arguments in favour of certain
assumptions as well as reasons and arguments opposing differing
assumptions (see also Cael. II 10, 279b6–7: the proofs for one side
are aporiai for the opposite side). This set will be the content of the
capacity we acquire as the science we are searching for. The
relevant difference between the actualisation of this capacity and the
mere utterance of words, as in the lucky-guesses case, is not that
the former is somewhat more complex than the latter. The difference
is rather that only the former entails an understanding of how the set
of distinctions etc. relates to the basic structures of the world it is
supposed to capture. Developing the capacity is not learning to say
something that is true, but to actually grasp the truth.
Accordingly, the point of bringing in aporiai is not that they help
us to formulate the set. The point is that they are the means, and are
needed, to carve out the access to reality with its complex relation
between principles and things or properties that depend on those
principles. They are the means to figure out how the principles have
to be demarcated: They teach us how to frame the principles
sufficiently broadly for their task without letting them overstep their
boundaries. At the same time, going through them (diaporein)
teaches us to speak properly about principles in that it teaches us
the proper meaning of words.
By nature we desire to know – as by nature we have still to
develop the appropriate capacity of knowledge. At the beginning, we
are bound: we do not, and cannot, know yet how and where to
proceed. In order to advance, we have to recognise that we are
bound. Raising the aporiai and going through them will help us, as
we resemble the, at first clueless, wanderer, to overcome being
bound, to determine our way and its proper end. And on our way
ahead for the science we are searching for it will enable us more and
more to judge appropriately in the case of still existing aporiai.
Why is raising the aporiai and diaporein necessary? It is
necessary for developing the capacity the content of which consists
in the set of distinctions, implications, presuppositions, reasons and
arguments concerning the principles of the science we are searching
for. It is necessary as the means to frame the principles in the
appropriate way and in a meaningful way.
This is not to say that raising aporiai and diaporein will do all the
work with regard to the science for which we are searching. There
may be parts of the process where, for instance, epagōgē plays a
certain role in establishing parts of the science. The ways of how
some preliminary knowledge comes in would require closer attention
as well.21 Furthermore, not all aporiai will have the same weight or
will show the same urgency: for instance, some aporiai may be, as
aporiai, too unbalanced, some may relate to a minor problem only,
some may be of mostly historical interest. In addition, we would need
to see the systematicity in the display of aporiai that would guarantee
that we have covered every important problem pertinent to our
search for the first science.
But even if there are other parts of the procedure of developing
the science we are searching for and even if there are more aporiai
than we have come across so far, raising aporiai and going through
them will still be at least an essential part of our way to the first
science: it is essential for forming the capacity that corresponds to
the science we are searching for.
 8 Conclusion
Why is raising aporiai necessary? By nature we desire to know. This
means we ultimately desire to grasp the principles and first causes of
the (first) science. However, we have yet to develop the ability to
grasp these principles, that is, the capacity for knowledge: we have
to tread on unknown ground towards an end we do not know yet.
The only points of departure available to us are our capacities as we
have them. Unavoidably, we will look for principles and first causes
in the frame of our present outlook and will favour principles in
accordance with it. Since the principles are elusive and lend
themselves to be seen from different perspectives, our outlook will
not be sufficiently suitable for a grasp of the basic structures of
reality in the beginning, and hence we have to extend it. The only
way of doing this is to scrutinise the logical options with regard to
principles or aspects of the principles which are in need of
clarification (for instance, are there ousiai besides the perceptible
ousiai or not?). One has to figure out, then, the arguments for and
against either option and the explanatory contributions and gaps
implied by them. There is no other means for carving out the
principles, that is, for opening our growing understanding to the basic
structures of reality and for learning how they are to be described, if
the description is to represent those basic structures. It will enable us
to see why the principles thus described fulfil, and do not fail in, their
explanatory task. We will have found our way through the seemingly
conflicting demands, which we could not fend off as long as we have
not shaped our view appropriately by answering the aporiai. Raising
aporiai and discussing and answering them is necessary for shaping
our view on the basic structures of reality – that is, for acquiring the
capacity which consist in the grasp of the principles of the science
we are searching for and, thus, for becoming sophos in the proper
sense.

1 On the word ‘aporia’ in Met. B, its cognates, different meanings and the
problems of translation see Laks (2009: 25f.), Madigan (1999: xix–xxii) and
Owens (1978: 214–8); on linguistic issues see also Stevens (2001). I should
like to thank very much the participants of the Dublin conference, especially
its organisers, as well as the Frankfurt colloquium on ancient philosophy for
very helpful questions and suggestions.

2 This introduction does not offer an effective method of enquiry, but rather
‘define[s] what the relation of an impasse to its solution should in principle
be’ (Laks 2009: 46).

3 A strong interpretation of the necessity of aporiai has been suggested by

Halper (2009: 213). He takes the view (referring to B 1, 995a24f.) ‘that the
aporiai are objectively necessary for any thorough treatment of the subject
matter of metaphysics’: ‘working through the aporiai is necessary in order to
acquire metaphysical knowledge’. The ‘use Aristotle makes of the aporiai in
his metaphysics reflects the kind of demonstration that the subject matter of
metaphysics admits’ (2009: 213). On Politis’ account see below.

4 On these lines see Laks (2009: 28–30) and Cooper (2012: 351–4).
According to Cooper, the difficulties mentioned first are those presented in
B, while the later difficulties mentioned in 993a27 are ‘those that arise
consistently everywhere in the books subsequent to B’ (2012: 352). For a
connection of A 7, 988b21, A 10, 993a25f. and B 1, 995a24f., see Owens
(1978: 213 n. 10). For further remarks on B and views on its relation to other
parts of the Metaphysics, see Madigan (1999: xxii–xxxviii) and Bell 2004:
127–30. On the connection between A and B see also Politis (2003: 147f.).
5 According to Mansion, 1955: 162, Aristotle had already solved the aporiai
for himself before he wrote B and presents them in B for didactical reasons
only, taking as a point of departure for his presentation a state of mind
similar to his own before he had solved the aporiai. Owens (1978: 253f.)
suggests that Aristotle wanted to address people with a specific preliminary
knowledge. Halper (1992: 151, 169, 171) suggests that Aristotle addresses
specific Platonist views in B – he does not present aporiai he finds himself in
– but does so not for historical reasons, but because the very content of
those views is intrinsically connected to main features of metaphysics.
Against a merely didactical reading see also Laks (2009: 37).

6 According to Reale (1980: 64), the aporiai derive from ‘the opposition of
the doctrines of the “Naturalist” philosophers, on the one hand, and of the
“Pythagoreans and Platonist” on the other’ (see also ibid. 91–4).

7 I will not pursue an answer to (c). Leszl infers from his investigation of all
aporiai, that ‘Aristotle does not as yet possess a clear conception of a
universal science of being’ (1975: 141). The text ‘does not represent a very
advanced stage in Aristotle’s metaphysical reflection’ (ibid. 142). According
to Cleary (1995: 201), ‘the science in question has yet to be discovered’
(see also 213 n. 46), with reference to B 1, 995a24–7. In contrast see Reale
(1980: 96): ‘in writing Book B Aristotle was in full possession of the solution
of the aporias’.

8 This is highlighted by Madigan (1999: xvi), who – following the tradition –

characterises B as a whole as ‘dialectical’ (ibid. xiii): many aporiai are based
on endoxa, though the arguments of B also ‘draw on a variety of non-
endoxic premises’ (xvii). Lee 2010: 265 calls the procedure of B ‘dialektische
Aporetik’ (‘dialectical aporetics’). According to Owens 1978: 253, the
opinions used by Aristotle for the presentation of the aporiai ‘have been
reinterpreted, sometimes slightly, more often to a considerable degree if not
entirely in the setting of the Aristotelian physics and logic’.

9 An explicit example is aporia B 4, 1000a5–1001a3, which concerns the

question of whether the archai are the same for perishable and imperishable
things. Obviously, his contemporaries and predecessors did discuss those
archai, but apparently they did not pursue (paraleleiptai, 1000a5) the
question of the aporia which Aristotle considers to be the crucial question.
10 On this see also Laks 2009: 35f.; on omissions see Owens 1978: 253f.,
according to whom the ‘aporiae of Book B … can hardly be considered as
an ‘antinomical discussion of the main problems of the First Philosophy’’
(254). On the lack of systematicity see also Leszl 1975: 119f., 139f., 142.

11 On the danger of arbitrariness that comes with relying on actual endoxa

see Halper 2009: 211.

12 For the difficulty see B 1, 996a4–9, a15–17, B 2, 997a33, B 3, 998a20–1,

B 4, 999a24–6, 1000a5, 1001a4–5; for the necessity see B 1, 995b13–14
with B 3, 998a20–1, B 1, 995b31–6 with B 4, 999a24–6, 1001a4–5.

13 On the difference between contrariety and contradictory opposition in the

structure of the aporiai see Crubellier and Laks 2009: 8f. Some aporiai are a
bit more complex. Aristotle may present these cases not as complex aporiai,
but as a short series of basic aporiai (see B 2, 996b33–7a15 and 997a16–
25; see also B 3, 998a23–b13 and 998b17–9a23). In at least one case we
need to combine passages in order to actually construe a proper aporia: see
B 2, 997a34–8a19, which presents problems for the assumption of ousiai
besides the perceptible ousiai (either Forms or metaxy-ousiai) and B 4,
999a24–b17, which presents problems for the assumption that there are no
such para-ousiai. The solution to an aporia may show that one side of it has
been closer to the truth than the other or may show that both have been
equally far away from the truth (Top. VI 6, 145b1–2, b16–20 which seems to
contradict this claim does in fact not stress that an aporia consists of
opposite claims of equal weight, but that the equal weight of a pair of
opposite claims may generate an aporia).

14 Madigan 1999: xviii provides a short description of the variety of

arguments and gives a proportion of ‘constructive’ and ‘refutative’ arguments
(and distinguishes also between the procedure and the aim of the respective
arguments).

15 See for this Frede 2004: 13f., 17.

16 See Politis 2003: 167: The ‘endoxa and phainomena … that give rise to
intrinsically objective aporiai, must be understood as signifying different
ways in which things present themselves to us’ (see also ibid. 168).
17 Aubenque has suggested that Aristotle’s metaphysics is aporetic in the
sense that the aporiai concerning the question of ‘what is being’ do not have
any other solution than searching itself (see 1966: 508). The aporia is not
grounded in our ignorance, but in the pragma itself (1961a: 322f.).
Nevertheless, Aubenque seems to admit that gods could advance to first
philosophy and grasp the highest principles (see 1961a: 331). If they can, it
would seem to be due to a specifically human intellectual inability that we
cannot grasp those principles. However, it is hard to see why Aristotle would
think (a) that our desire to know is essentially in vain, (b) that the principles
of the science we are searching for are somehow inconsistent – as if
inconsistency could be a feature of reality, and (c) that the results we find in
the central books or in Λ should not be seen to be proper results (from
Aristotle’s perspective) and should not give the impression that Aristotle was
confident about the possibility of a consistent description of the principles.

18 According to Meno 84b–c, aporein – awareness of one’s ignorance of

how to proceed – is conducive (prourgou) to the discovery of how something
is, in that it shows us that we do not know yet the answer – which then
incites us our continuing search. On the various ‘degree[s] of necessity’ in B
1, 995a24–b4, and on ‘prourgou’ see Laks 2009: 41.

19 While according to Madigan 1999: xvi, the Topics passage shows that
‘one task of dialectic is to establish first principles, Smith 1997: 53 doubts
that this is ‘strong evidence for a view of dialectic as that which establishes
first principles’. The passage from the Topics ‘says neither that dialectic
establishes nor that it discovers these starting-points’ (ibid.).

20 Let us imagine how Aristotle might (in some cases) react to an aporetic
question such as: ‘A or non-A’; or to an aporetic question such as: ‘A or B
(with B implying non-A)?’ He might react by saying: ‘well, it depends’; or: ‘in
some way … in another way not … ’ (see for instance, Λ 4, 1070b10–11).
What does it depend on? It depends on how we determine the features in
question. A proper grasp of this ‘depends’-qualifier will prevent us from
jumping rashly from the refutation of A to the assumption of non-A.

21 Cf. A 9, 992b26–33, Θ 8, 1049b27–50a2. On the requirement for this

preliminary knowledge see Laks 2009: 45, with reference also to APo. I 1,
71a24–30, and especially Politis 2003: 152–8.
 Chapter 8
Uses of Aporiai in Aristotle’s Generation
of Animals
◈

Jessica Gelber

Given what Aristotle says about aporiai in Metaphysics B, one might

expect them to play an important methodological role in his biology.1
In Metaphysics B, we learn that going through aporiai enables us to
identify the ‘knots’ binding us, to know the direction in which we
should proceed, to recognise it when we have found what we are
seeking, and to be competent judges of competing accounts.2 These
achievements would seem, at any rate, to be equally necessary for
both first- and second-order causal enquiries. Moreover, one might
expect not only for there to be something significant in Aristotle’s use
of aporiai in biology, but also for the aporiai in this empirically
grounded domain to share some distinctive features. One might
suppose, for example, that the puzzles would be generated by a lack
of observable evidence as opposed to being due to some conceptual
difficulty.3
 It is difficult, however, to draw any general conclusions about
Aristotle’s use of aporiai in the biological treatises. Although these
treatises contain many instances of aporia language – i.e. uses of
the noun and its verbal cognate – there seems to be no single
purpose for which Aristotle raises them, no common structure, and
no shared source of puzzlement. In what follows I will attempt to
demonstrate this negative thesis by looking closely at the use of
aporia terminology in Generation of Animals. For ease of
presentation, I will follow Aristotle in referring to what he calls an
aporia by that name. One upshot of this discussion, however, is that
caution may be warranted in assuming that there is any single kind
of puzzle or difficulty being picked out by the use aporia language.
 1 For What Purpose Is an Aporia Raised?
If one thought that the raising and resolving of aporiai were an
integral feature of Aristotle’s methodology in biology, a natural way to
yoke them together might be by reference to the purpose for which
they are raised. Unfortunately, as I will show in this section, there is
no one purpose for which they are raised.
(i) In some cases, Aristotle states that some phenomenon ‘has
much aporia’ where that announcement introduces a survey of
existing views that are all equally problematic. Such ‘puzzles’ appear
to be raised in order to establish the superiority of Aristotle’s own
theory, by showing that it avoids problems that confront the
alternatives. These aporiai perform what we might call a ‘refutative’
function.
A prime example of a refutative aporia is Aristotle’s discussion
of bee generation in GA III.10. Aristotle’s own tentatively proposed
account4 is that bees generate without copulating, just as certain fish
do.5 This is the only possible explanation, he thinks, given what has
been observed. He demonstrates this by going through the problems
with all of the prevailing alternative accounts.
Bees cannot gather their young from outside the hive, as some
say.6 If they did, these young bees would have to be either (a)
spontaneously generated, or (b) produced by another kind of animal.
Aristotle says that it is absurd, however, to think they are
spontaneously generated. For, he reasons, in this case we should
have found spontaneously generated bees that are not collected
remaining in the places where other bees would have collected
them. However, this is not something we observe. Aristotle also
rejects (b), the idea that bees could collect offspring that are
produced by another kind of animal. For, animals generate offspring
that are the same in kind as themselves, and so this would violate a
principle that holds of animals, generally.7 Finally, against both (a)
and (b), he notes that while it makes perfect sense for bees to collect
honey, since honey is their food, it is absurd to think that bees would
collect anything that is not food.
It cannot be that bees are produced by the copulation of drones
and worker bees, as others claim, either. Generation by copulation
requires both male and female principles. So drones and worker
bees would have to exhibit sexual differentiation. But it does not
seem as though worker bees can be either male or female: Worker
bees cannot be female, because they have stingers, and nature
does not assign weapons to females. Worker bees cannot be male,
since worker bees tend to the young, and in Aristotle’s view, males
never do that.8 In fact, generation by copulation seems to be
impossible not only on the assumption that two different kinds of
bees, such as workers and drones, copulate with each other to
generate all of the kinds, but also on the assumption that the
members of the same kinds of bees copulate with each other to
produce another individual like themselves. This is apparent,
Aristotle claims, since young drones are produced when no drone is
present, and worker bees are only produced when king bees are
also around.
 The only viable option, Aristotle concludes, is that worker bees
are hermaphroditic and generate drones without copulation, and that
king bees generate both worker bees and kings. Although it is a
bizarre theory, it is the only one that makes sense of all of the
empirical evidence and coheres with his zoological principles.
Consequently, Aristotle thinks it is the best theory, given what we
have observed.
(ii) At other times, an aporia is presented not for the purpose of
rejecting other theories, but rather because doing so will advance
Aristotle’s own enquiry in some way. These aporiai move the
investigation forward by making clear that there is a genuine problem
or question needing to be addressed, and by making explicit what
the constraints on any adequate theory or explanation are. I will label
this a ‘zetetic’ use of aporia.
(iia) In some cases, Aristotle advances the enquiry by motivating
the introduction of something unobservable that would make sense
of the phenomenon being discussed. One such motivating, zetetic
aporia is raised in order to justify positing the presence of pneuma in
the male’s sperma.9 At GA II.2, 735a29–6a23, he says that someone
might be puzzled10 about the nature of sperma, since it does not
behave in the way that either watery or earthy substances do. Unlike
watery substances, it does not thicken (pachunetai) when cooled.
Unlike earthy substances, it does not thicken when boiled, as milk
does. But according to Meteorology IV.4, the moist and the dry
(referred to as ‘watery’ and ‘earthy’ in the GA passage) are the
passive potentials out of which all uniform bodies are made.11 So it
seems that it must be one or the other.
 Aristotle begins his resolution of this aporia by asking: ‘or have
we not distinguished all of the things that occur (sumbainonta)?’
(735b7–8). He points out that in addition to watery and earthy
moistures, there are also those made up of water and pneuma, for
example foam, oil, and lead ore. So, he now introduces a further
feature of his account: Semen is a compound of pneuma and water,
and pneuma is warm air.12 This accounts for the properties of
sperma, and more importantly, the presence of pneuma is now a
feature of the embryological account that he can appeal to in
explaining other phenomena.13
(iib) Another type of zetetic use of aporia occurs when Aristotle
wants to make some corrections or revisions. Some of these zetetic
aporiai are raised in order to correct what has been (wrongly) taken
to be the empirical facts. He says, for instance, that someone might
be puzzled about the growth of eggs (GA III.2, 752a24ff). Eggs do
not have an umbilical cord, he says, so they do not receive
nourishment in the way that animals that are born live do. They also
do not simply use themselves as nourishment in the way that larvae
do.14 How, then, does the nourishment get inside the egg in the first
place, when that egg is hard and lacks anything like the umbilical
cord?
Aristotle resolves this by wheeling in some empirical data that
has been overlooked (lanthanei): at the outset the egg is soft, though
it quickly hardens once released. If it were not soft, egg-laying would
be painful. That initial, soft egg has a very small umbilical-like ‘pipe’
protruding from the pointed end, by which it was attached to the
uterus. This is difficult to see, but Aristotle says it has been observed
in aborted chicken eggs.15
(iic) Other zetetic, corrective aporiai are raised when
qualifications or modifications to the explanation that Aristotle has
offered need to be made. For instance, in the course of explaining
what happens just after fertilisation in GA II.4, Aristotle says that
while all of the parts are ‘potentially’ present in the kuēma, which is
the first mixture of male and female spermatic residues,16 the first
part to be actually present is the heart. The heart has to be present
first, because it is the source (archē) of all the other parts, both
homoeomerous and heterogenous. The new fetus, while still
incomplete, and still only potentially an animal, needs nourishment
from elsewhere. Thus the heart sends off blood vessels (the
umbilical cord) to the uterus to procure the nourishment, just as
seeds in the ground send off roots and shoots. But about this,
Aristotle says that someone might be puzzled.

Someone might aporēseie, if on the one hand the blood is nourishment,

but on the other hand the heart – being bloody – comes to be first, but
nourishment [comes from] outside (thurathen),17 from where does the
first nourishment enter?

(740b2–5)

Aristotle has just said that the growing fetus gets its nutriment from
the uterus, to which the heart sends vessels. Now he points out that
this cannot be the way the heart gets its initial nourishment. So how
does the heart receive it? To answer this, he makes a qualification to
the theory:
 Or is this not true, that all [nourishment comes from] outside, but rather
just as in the seeds of plants there is some such thing that first appears
milky, so also in the matter of animals there is residue for the assemblage
straightway.

(740b5–8)

So, there is something not quite correct after all about the idea that
nourishment comes from outside the heart.
(iid) The previous two examples of zetetic aporiai are ones that
allow Aristotle to make a correction, either to his own theory or to
what is taken to be the empirical data about some phenomenon. A
further purpose for which zetetic aporiai arise is to motivate not
merely a correction but rather what we might think of as a radical
change in focus. An extended discussion in GA II.1 is a good
illustration of the type of use I have in mind. Aristotle’s purpose in
raising this aporia, as I understand it, is to prime his audience for a
surprising shift in the way we should conceive of agency – of what it
is to be an agent of some change – by painstakingly running through
the difficulties that will arise if such a shift is not made.18
At 733b32, Aristotle raises a question about what the agent of
embryonic development can be. Here he is concerned to identify the
agent that produces an embryo’s body parts out of the first mixture of
male and female spermatic residues, a mixture that he sometimes
refers to as the kuēma and sometimes the sperma. This agent must
be either something external or internal to this mixture, but there
does not appear to be any viable candidate. First, Aristotle rules out
the first possibility:
 For either something external or something present in the semen and the
sperma makes [the parts], and the latter is either a certain part of soul or
soul, or might be something having soul. Well, surely that something
external makes each of the viscera or other parts would seem absurd.
For it is neither possible for something not touching to move [anything],
nor possible for something to be acted upon by something not moving [it].

(GA II.1, 733b32–4a4)

According to Aristotle’s general account of change, there must be an

agent with an active potential to impose form and a patient with a
passive potential to receive form. But that is not sufficient: conditions
must obtain such that the change occurs, namely, that the agent and
patient be in contact. It is this general principle – that change
requires contact between the agent and patient – to which Aristotle
refers here.19 In his theory of animal generation, the male parent has
the active potential to impose form, but he never makes direct
contact with the matter that has the passive potential to receive that
form. So, it seems that there must be some part already present in
the kuēma that is the agent, since only something in the kuēma
could make contact with the matter. However, this option is also
ruled out.
Aristotle’s lengthy rejection of this second option involves
establishing that (1) if the agent were some part present in the
kuēma, it would have to be a part of the embryo that is being
formed.20 But (2) since there is no part of the body that does not
have soul, some ensouled part would have to be present from the
outset (734a14–16). Yet this, too, is impossible: (3) no part of the
embryo can be present in the seed, because seed is simply what the
parts come to be out of, and so must be produced prior to the parts
that come to be out of it (734b1–2).21 So, nothing can be made
simultaneously with seed such that it is both present in the seed from
the outset and also comes to be out of seed.
Aristotle resolves this by showing how we might reconceive of
the agent in a way that makes it possible for it to be something both
internal and external, after all.

We must try to resolve these. For perhaps something we said is not

unqualified,22 namely how in the world it is not possible to come to be by
something external. For in one way it is possible, but in another it is not.

(GA II.1, 734b4–7)

There is no agreement among interpreters about how to understand

the discussion that immediately follows.23 As I understand it, the
resolution involves showing that we were wrong to be looking for
some thing – ‘some this’ (tode ti, 734b18) – that could be making the
embryo’s parts. It is rather the movement being conveyed in the
semen that is making the parts.24

Surely, to say ‘sperma’ or ‘what sperma comes from’ makes no

difference, insofar as it has in it the movement which that one [sc. the
father] was moving. And it is possible for this to move this, and this to
move this, and to be like the spontaneous marvels. For in a way the
parts, although resting, retain a power. Whenever something external
moves the first of the parts, straightaway the following one comes to be in
actuality. Just as, then, in the spontaneous [cases], [a] in one way [the
mover] moves, not touching anything now, but having touched. And
similarly, also what sperma comes from or the one who made the sperma
did touch something, but is no longer touching. [b] And in another way the
 internal movement [moves], just as the house-building [makes] the
house. Well then, that there is something which makes, not being some
particular present in [the sperma], nor from the beginning completed, is
clear.’
(GA II.1, 734b7–19)

We need not worry that the father is no longer in contact with the
semen and directly moving it, for it is really the movement that the
father had set up (by concoction of the spermatic residues) that is
doing the work. And that movement in the semen can continue after
the father releases the semen, much like the ‘spontaneous marvels’
can continue to move after an external mover moves a part of it. So,
since the movement in the semen is making the parts, there is a
sense in which (tropon de tina, 734b16) the agent is something
internal; since the father set up at that movement, there is a sense in
which (tropon men tina, 734b13) the agent is something external.
(iii) Finally, sometimes Aristotle claims there is an aporia where
it is not clear that what he goes on to say is moving the enquiry
along in any way at all. In fact, in at least some cases, it appears that
he is simply pointing out that there is some phenomenon for which
we lack any explanation. These uses seem to be merely a way of
introducing the next topic to be discussed. An example of this sort of
use is the aporia about the uterine mole (mulē):

There is a puzzle, on account of what [uterine moles] have not come

about among other animals, unless something has entirely escaped
notice.
(776a8–9)
The uterine mole is an abnormality that occurs early on during
pregnancy (a ‘molar’ pregnancy) whereby a mass is formed inside
the uterus. Aristotle claims that the cause of this phenomenon is a
weakness of the heat relative to the matter that the heat needs to
concoct.25 But apparently, Aristotle reports, this phenomenon has
only been observed to occur in human women, and not in other
kinds of animals. So, there is a question, introduced by the aporia
terminology, about why this should be the case. Perhaps there is
something unique about human females that would explain this:

One must suppose (dei nomizei) that the cause is that the women alone
of other animals is husterikon and excessive in evacuations (katharseis)
and not able to concoct them. So whenever a kuēma is put together from
moisture that is difficult to concoct, then the so-called mulē comes about,
reasonably either especially or uniquely in women.

(776a9–14)

Although it is not entirely certain what it means for women to be

husterikon,26 it is this feature, specific to human women, that
accounts for the mola uteri only occurring in humans.27
From this survey of a variety of passages, it appears as though
there are a number of purposes for which Aristotle raises what he
calls an aporia. Some are raised in order to refute alternative
theories (refutative aporiai) by showing that they involve some
absurdity or conflict with the observed facts or zoological principles,
while others advance Aristotle’s own project (zetetic aporiai).
Moreover, the zetetic aporiai move the enquiry forward in several
ways: they either offer occasion to posit some unobservable entity, or
to add new empirical data, or to revise the details of Aristotle’s own
theory, or to change the conceptual framework being employed.
Finally, claiming that something has or involves an aporia sometimes
seems merely to be Aristotle’s way of introducing the next topic for
discussion, where he seems to mean that someone might not know
how to answer a question and thus be as yet ‘without passage’.
 2 Structure
In our previous discussion, we have seen that there are a variety of
purposes for which Aristotle raises aporiai. Now, we will see that an
aporia can take a number of different forms.
(i) Some of the aporiai are presented in the form of a dilemma.
In such cases, we are given what appear to be exhaustive options.
In some of these, the options are equally unappealing or impossible
(the ‘negative’ aporiai). The discussion of the agent of embryonic
development is an example of a negative aporia. Recall that it
initially seemed as though the agent must either be something
internal to the kuēma or something external to it, and it did not seem
that either could be the case. Aristotle resolved this by showing how,
in a way, the agent can be both: Since it is movement – not ‘some
this’ – that is the agent, and that movement is derived from the male
parent and conveyed to the kuēma, it is both something external (the
father’s movement) and something internal (the movement in the first
mixture of spermatic residues). Aporiai of this type have the form ‘Is
A or B the case?’
(ii) In other cases, we are given what looks to be an exhaustive
and exclusive dilemma, but no considerations for or against one of
the horns are given. For example, in an aporia about what occurs at
‘the beginning of generation’ in GA V.1, Aristotle begins by providing
two options.
 There is a puzzle concerning the beginning of generation: does
wakefulness obtain prior, or sleep?
(GA V.1, 778b23–5)

Without even considering the possibility that wakefulness is the prior

state, Aristotle proceeds to adduce reasons in favour of it being
sleep that is prior. It would be reasonable to think sleep is the earlier
stage, he says, since animals become more wakeful as age
advances, suggesting that they are proceeding towards wakefulness
from its opposite. Moreover, since animal generation is a transition
from not-being to being, it is plausible to think that an animal would
first be asleep, since that appears to be intermediate and ‘like a
boundary (methorion) between living and not living’ (778b29–30).
However, Aristotle is reluctant to call whatever the developing animal
does at the beginning ‘sleep’. Wakefulness is or at least involves the
exercise of sense perception, and its opposite – sleep – is the
absence sense perception. But it is not simply any absence of
perception that constitutes sleep, but only the absence of perception
in creatures for which perception can be present. Plants do not
sleep, and they cannot wake up from being asleep, and that is
because plants do not perceive.28 The state plants are in is
something else. So, too, at the earliest stage of development when
the animal lacks the ability to perceive, we must say that the state
they are in is not exactly sleep, but something like sleep.

If it is necessary for an animal to have perception, and whenever

perception has first come to be it is at that time first an animal, then it
 must be supposed that the condition from the beginning is not sleep but
like sleep, just as the sort that plants also have.

(GA V.1, 778b32–5)

Here, the aporia is initially about whether A or B is the case, but

Aristotle immediately moves to consider only B. This looks less like a
genuine dilemma than a question about whether and how B might be
the correct answer to the question that was originally posed.
(iii) Furthermore, some aporiai do not have the appearance of a
dilemma at all. Many times Aristotle says that there is an aporia
about some phenomenon, where what he apparently means is that
there is a question about it, such as how, what or why something
happens.
(iiia) Some of these questions are of the form ‘Why does F occur
among A’s but not among B’s?’ Typically, these are questions about
why something generally true of some larger kind is not true of some
subset of that kind. He uses the aporia language, for instance, when
he introduces the discussion of a difference between bird and fish
generation: bird eggs are separated from the uterus prior to being
‘complete’, whereas fish eggs remain attached to the uterus.
Aristotle says that ‘someone might aporēseien about why the
generation of birds and fish differ in this respect’ (754b20–1). Both
are egg-layers, but the particular ways in which they produce eggs is
not the same.29 So, Aristotle sometimes introduces with the aporia
language a question about why something true of A is not true of B,
where A and B are members of some wider kind, such as egg-layers
or live-bearers.30
 (iiib) Aristotle also uses the aporia language where what follows
is simply a question about what happens. For example, he begins
GA II.3 by asking what happens to the bodily part of semen when it
enters the uterus, given that that bodily part does not become part of
the developing embryo (736a24ff). And sometimes these are
questions about why something happens. He says that the reason
why the eyes appear largest at the beginning of generation but are
the last of the parts to be completed ‘has’ aporia (743b32ff). What
Aristotle says involves aporia here is simply a question about the
cause of some phenomenon that we do not yet know.
There does not appear to be any significant overlap among the
aporiai if we attend to the form in which they are presented, i.e.
whether as dilemmas or merely questions. So, just as we were not
able to draw any general conclusion about the purposes of Aristotle’s
biological aporiai, there is also no such conclusion to draw about
their structure.31
 3 Sources
Having been unable to discover any common purpose for which
Aristotle raises aporiai in biology, and having seen that they lack any
common form, one might wonder if instead there is some other
distinctive feature they share. Perhaps, one might think, the
biological aporiai are united in being generated by having a common
type of source, such as a lack of observational evidence. However,
there does not seem to be anything general to conclude about the
reasons – what I am calling the ‘sources’ – why the aporiai arise.
(i) Some involve a tension between Aristotle’s biological theory,
as it has been articulated so far, and some apparently recalcitrant
data or observations. The tension that gives rise to the aporia about
the heart’s nourishment (740b2–8) is of this sort. The theory
established the heart as the source of nourishment for the
developing embryo, but the heart, too, has to get its nourishment
from somewhere. Here the conflict is resolved by simply modifying
the theory: Perhaps there is nourishment in the heart from the
outside, so it is not true that all nourishment comes from somewhere
else.
(ii) Other aporiai arise because there is a conflict between the
details of Aristotle’s biological theory and his own deeply held,
general commitments. Attending to these is often illuminating, for
they reveal a natural scientist working within the constraints of a
grand metaphysical framework, self-consciously applying the
concepts and principles he argues for elsewhere, though at a fairly
abstract level. As I understand it, the aporia about the agent of
embryonic development has this type of source. Aristotle has argued
elsewhere that in general, change requires contact between agent
and patient. Here the aporia arises because it seemed difficult to see
how his embryological theory could accommodate this.
(iii) Further, there are cases said to involve aporia in which there
is a tension between some general principles and the empirical
evidence. For example, Aristotle announces at GA I.8, 718a35–7
that ‘someone might be puzzled about the facts concerning the uteri
in women, how they are’ since ‘many oppositions (hupenantiōseis)
belong to them’ (718a36–7). In some kinds, the uterus is higher up,
near the diaphragm, and in some it is lower, near the pudenda.
Five chapters earlier, (GA I.3, 716b12–13) Aristotle had already
said that ‘the things concerning’ (ta peri) the testes and uteri are not
the same in all blooded animals, and then proceeds to discuss
differences amongst the male sexual organs for the next five
chapters. He explains why some kinds have and some lack testes,
and why those which have them do so internally or externally. Thus
someone might rightly be puzzled about why variations amongst the
female sexual organs should be so puzzling.32
It appears as though this aporia arises because the variation in
the position and orientation of the uterus does not coincide with the
distinction between being a live-bearer and egg-layer. Some kinds of
live-bearing animals have the uterus high, and others have it low.
The same is true of the egg-laying creatures. However, the variations
found among male testes, too, cuts across the live-bearer/egg-layer
distinction, and that is not something Aristotle identified as puzzling.
Why might someone be puzzled about the variations in the position
of uteri in different animal kinds, but not about the variations in the
placement of testes?
Although Aristotle does not say so explicitly, I suspect that the
reason this is puzzling is that it is in tension with some familiar
principles of his natural science, such as the principle that nature
does nothing in vain, and that morphē follows function (and not the
reverse). In the case of the uterus, there seem to be variations in
morphē that do not track variations in its function, which is to provide
‘protection, shelter, and concoction’ (I.12, 719a33–4). Given this
function, one might expect that only a difference in what is being
produced (i.e. whether what is being protected, sheltered, and
concocted is an egg or a ‘complete’ animal) should require a
difference in the position of the uterus. But not all egg-layers have
the uterus positioned in the same way. The function of the testes is
to ‘make the movement of the spermatic residue more steady’
(717a30–1). Given that function, there is no reason to expect the
male organ’s presence or position to co-vary with the kind’s being an
egg-layer or live-bearer. This is why there is puzzlement about the
uterus but not about the male sex organs.
Aristotle resolves this by distinguishing different ways of being
egg-layers, and different ways of being live-bearers. Egg-layers are
divided into the kinds that produce ‘complete’ eggs (i.e. hard-shelled
ones, such as those of birds), and those that produce ‘incomplete’
eggs (i.e. soft ones, such as fish produce) that are ‘completed’
externally. For the hard shell to be formed, the uterus needs to be
near a source of heat, and this is what the higher placement
provides. If less heat is required, as is the case for the formation of
the soft-shelled, incomplete eggs, the uterus is positioned lower and
closer to the exit (peras), since that is more expedient. So, there is a
difference in function from which the differences in morphē follow,
after all.
(iv) Finally, sometimes the source of aporia is not any difficulty
or tension of the sort we have seen in the previous examples.
Sometimes the reason that there is aporia is that some fact has been
overlooked. For example, it has ‘escaped notice’ that bird eggs have
an umbilical-cord-like part at the beginning, and that missing
observation is what gives rise to the difficulty in explaining how the
egg receives its nourishment. And sometimes it is not that we lack
observations, but it is simply the fact that no explanation has yet
been offered that creates the impasse.
So, difficulties can be due to conflicts between some particular
explanation, whether Aristotle’s or someone else’s, and the empirical
evidence. An aporia can also involve tensions between some
particular explanation and some more general principles. Some of
the principles are specific zoological ones, such as the principles that
nature does not give weapons to females, or that males do not tend
to their offspring. Other principles are ones that hold of change,
generally, such as that change requires contact. Further, sometimes
the source of aporia is that the observed data do not cohere with the
general principles. Lastly, for some of the aporia, it is not obvious
that there is any genuine conflict or tension at all. These appear to
be cases where the aporia is simply due to an absence of evidence
or lack of any explanation, at least so far. Consequently, aporia in
Aristotle’s biology cannot be said to be generated by some specific
type of problem. There is no obvious way, at any rate, to unify the
aporiai in GA by their source.
 4 Conclusion
From a close look at the discussions in which Aristotle explicitly uses
the aporia terminology, it appears as though aporia in GA is
pollachōs legomenon. The purposes for which Aristotle raises them
vary a great deal, as do the structures the discussions exhibit, as
well as their sources. Moreover, it is possible to point to further ways
in which they differ. For example, Aristotle uses a variety of methods
to resolve them. Sometimes he brings in additional observations,
such as the fact that eggs do have something umbilical-cord-like,
even though it is hard to see.33 At other times, he modifies or
qualifies his theory in some way, as he does when discussing the
source of the heart’s nourishment. More interesting cases involve his
making some conceptual distinction, as he does when trying to
identify the agent that makes the embryo’s body parts.
The aporiai also differ quite a bit in terms of their seriousness.
Some, to be sure, are deeply problematic, and Aristotle seems to
think that their resolution is pressing. Others, however, do not seem
to be particularly worrisome to him at all, at least in the present
context. In fact, sometimes he seems merely to be pointing to a
potential problem for his theory, but one that he is not going to attend
to at this time. After raising the issue about heart’s nourishment, for
instance, Aristotle returns to discussing how the growing embryo
receives its nourishment through the umbilicus, and makes no
further mention of the heart’s nourishment.34
 Now, what can we conclude about Aristotle’s biological aporiai,
given that they seem to be so multiform and multipurpose? One
might reasonably harbour doubts that those puzzles that he calls
aporiai mark off some well-delimited group. Perhaps the concept of
an aporia may be more fluid for Aristotle than one might have
supposed. It may be simply a matter of being without passage, in
some vaguely defined sense.
Alternatively, one might maintain that he does in fact have some
clearly defined notion of an aporia, such as the state of puzzlement
arising from dialectical problemata,35 while thinking nevertheless that
such genuine aporiai are not tracked by the use of aporia language.
For, the multiplicity of purposes, structures and sources of the aporiai
we have looked at suggests that no common kind of puzzlement is
being indicated by the occurrence of the aporia terminology. At best,
the genuine aporiai would have to be merely a subset of the
difficulties Aristotle calls by that name. So, if he thinks an aporia is a
specific type of puzzlement, his use of aporia terminology will not be
sufficient for telling us when we are facing one. We also have reason
to doubt that the occurrence of aporia terminology is necessary for
telling us when there is a genuine aporia: There are at least a few
instances in Generation of Animals where one finds discussions that
are very similar in structure and purpose to some of those we have
examined. For instance, in his survey of the various arguments
adduced by the preformationist theory of generation36 in Book I,
chapters 17–18, there is the same systematic treatment of
preformationist arguments as one finds in passages explicitly flagged
as involving aporia. This is also true of his discussion of alternative
explanations of sexual differentiation in Book IV, Chapter 1. The
purpose for which he methodically runs through the arguments and
rival views, just as in the aporia about bee generation, is to show the
superiority of his own theory and the impossibility of any other. In
neither case does Aristotle employ aporia language.
Finally, we might conclude that aporia is not an especially
important part of Aristotle’s method in his biology. When studying the
heavens, perhaps aporiai are needed, since ‘we have but little to go
upon, and are placed at so great a distance from the facts in
question’ (de Caelo II.12, 292a15–7, Stocks trans.). The study of
eternally ungenerated and imperishable beings is more valuable and
more divine than that of generated and perishable beings, but it is
harder to study the former since about ‘those things on the basis of
which one would examine them and those things about them which
we long to know, the perceptual phenomena are altogether few’ (PA
I.5, 644b25–6, Lennox trans.). Where observable data is available,
on the other hand, perhaps the knots are readily apparent, and thus
running through the puzzles is not required. In the case of the study
of plants and animals, consequently, using aporia might thus be less
pressing: in this domain, we have more means (euporoumen mallon)
to attain knowledge, since we are fellow creatures (suntrophon) (PA
I.5, 644b28–9).

1 Michael Boylan 1982: 118 goes so far as to claim that Aristotle’s biological
methodology in Generation of Animals is based on the procedure of moving
from problems to their solutions, and indeed that the biological treatises are
organised around problems. According to Gareth Matthews 1999: 118–9,
Aristotle ‘normalises perplexity in the practice of philosophy’ and extends the
scope of this practice to deal with ‘puzzles in such sciences as cosmology
and biology’.

2 See Buddensiek’s essay in this volume (Chapter 7).

3 Terrence Irwin 1988: 32, for instance, claims that for ‘empirical puzzles’ the
‘source of our puzzles is empirical ignorance leaving us at a loss to say what
happens or why it happens’. Irwin also claims that in Aristotle’s ‘empirical
inquiries – those in which he surveys empirical appearances – Aristotle
characteristically raises these empirical puzzles’ (ibid. 42). If Generation of
Animals is considered by Irwin to be an empirical enquiry, then he is an
example of a scholar who thinks the aporiai in that treatise have a distinctive
character, namely, that they share a common source.

4 The proposal is only tentative, because Aristotle believes that this

explanation should be rejected if new facts are discovered, since ‘one must
believe in perception more than arguments (logōn)’ (GA III.10, 760b31–2).
The observable facts about bee generation, Aristotle says, have not been
sufficiently grasped. For a discussion of Aristotle’s argumentation in this
chapter of GA, see Joseph Karbowski 2014.

5 The way bees generate is unique, despite its similarity to how it goes for
some fish. The kinds of fish that generate with copulating produce other fish
of the same kind, whereas only the king bee produces other kings (760a4–
9).

6 Aristotle also notes that since no bee larvae are gathered from outside the
hive, it is also not possible for only some to be generated and others
gathered.

7 Aristotle knows about hybrids such as donkeys, however, although he

does not acknowledge such exceptions here.

8 For discussions of Aristotle’s use of such principles having to do with

males and females, e.g. that nature does not give weapons to females and
that males do not tend to their offspring, see G.E.R. Lloyd 1983.

9 Sperma in this discussion clearly refers to the male semen, although

Aristotle will sometimes use this term to refer to the female spermatic
residue, as well as the mixture of male and female residues – the ‘seed’.

10 It is not entirely clear whether the expression aporēseien an tis means

that ‘someone might be puzzled’ or that ‘someone might raise a puzzle’. The
same is true about the use of the noun aporia; it is not clear whether aporia
refers to a state of puzzlement or the difficulty that gives rise to puzzlement.
Consequently, it is not as obvious to me as it is to Gareth Matthews that for
Aristotle ‘an aporia is not a state of puzzlement at all; it is rather a puzzle or
conundrum or difficulty that produces, or is likely to produce, a state of
puzzlement’ (1999b: 130).

11 ‘We must discuss the forms of the passive potencies, the moist and the
dry. The passive principles of bodies are moist and dry, whereas other things
are combinations from these, and a body in its nature consists more of
whichever of the two is more – for instance, some things consist more of dry,
others of moist.’ (Meteor. IV.4, 381b23–5)

12 For the same reasons, Aristotle calls the nature of olive oil aporōtata: if it
were mostly water, it should be solidified (pēgnusthai) by cold, and if it were
mostly earthy, it should be solidified by fire (Meteorology IV.7, 383b20ff). As
it is, neither heat nor cold solidifies it, and both thicken it. His solution to the
puzzle is to point out that oil is full of air (aer). For a discussion of the ‘olive
oil aporia’, see James Lennox 2014: 288ff.

13 See Peck’s discussion in his Loeb edition of GA, Appendix B for an

overview of the role of pneuma in Aristotle’s accounts of movement, nutritive
soul activities and sensation.

14 As Aristotle explained in GA II.1, 732a32, it is only a part of the egg that

becomes the living organism, and the rest is used as the nourishment for
growth. See also the discussion in GA III.9.

15 cf. HA 559b16.

16 ‘Well then, among those animals not having separate males and females,
for these the sperma is like the kuēma. I mean by kuēma the first mixture
from the female and the male.’ (GA I.20, 728b32–4)
17 The idea that nourishment must come from something outside is not
explicitly mentioned by Aristotle in GA. As far as I know, the closest he
comes to saying this is the claim (at 724b34) that nourishment (trophē) is
manifestly epeisakton.

18 For a similar use of aporia, i.e. one that motivates a conceptual shift, see
de Caelo II.12, 292a19–22: ‘We may object that we have been thinking of
the stars as mere bodies, and as units with a serial order indeed but entirely
inanimate; but should rather conceive them as enjoying life and action. On
this view the facts cease to appear surprising.’ (Stocks trans., my emphasis)

19 Aristotle discusses contact in Generation and Corruption I.6–7 and

Physics III.2.

20 He does so by setting up a dilemma: If the agent were something existing

in the kuēma, it would have to be either a part of it, or something separate
from it (734a5–6). The second horn, i.e. that the agent be something
separate, is unreasonable (alogon) (734a5–7). For, if that were the case,
once the animal has been generated, then that separate part either remains
or perishes (734a7–8). It is evident that nothing remains in the new animal
or plant that is not also a part of it (734a8–9). So, it would have to be
something that perishes, which is also absurd (atopon) (734a10). Why? If it
perished, it would have to do so either after having made all of the parts or
after making some of them (734a9–10). If it perishes after making some of
them, then how are the other parts formed (734a10–11)? Aristotle’s
reasoning seems to be that if we said that this separate part makes some of
the internal organs, e.g. the heart and liver, and then perishes, it is unclear
how the rest of the parts and organs are formed. If we say that it is the heart
and liver that forms the rest of the parts, there is nothing to prevent these
from also perishing after they make the next organ. By parity of reasoning
(according to the ‘same logos’), the parts that make the other parts would
also perish, but they survive (734a11–13). Thus, the agent would have to be
a part in the kuēma, and not something separate (734a13–14).

21 I have omitted one epicycle in this long argument: The parts are produced
in succession, since the idea that they come to be simultaneously is ruled
out ‘by perception’ (734a20–1). Parts formed earlier cannot be agents of
later parts, because that conflicts with another general principle, namely, that
the agent must have the form in actuality of that which is being generated
(734a29–31). The idea that the heart, for example, has the form of the liver
is absurd (734a31–3).

22 This is the fallacy of secundum quid: ‘confusing what is general (haplōs)

and what is not general but some particular’, e.g. that ‘non-being’ exists,
since what-is-not is what-is-not (Rhetoric II.24, 1402a3–17). See also SE
166b37ff.

23 See Henry 2005 for a discussion the interpretations offered by Alexander

and Simplicius, as well as one of his own. This is one of two passages in GA
which draws on an analogy with the spontaneous marvels (automata tōn
thaumatōn), and it is unclear what kind of spontaneous marvels Aristotle has
in mind. It is clear enough that the father is being compared to the ‘first
external mover’ that gets the marvel moving, but beyond that the details are
all disputed. I will ignore this controversy here, since the point this example
is meant to illustrate has to do with the purpose for which Aristotle raises this
aporia. I think that everyone agrees that the point is to provide a model for
conceiving of the agent as something both external and internal.

24 Johannes Morsink 1982: 98 also takes the point of the automata analogy
to be to ‘teach us not to view the efficient cause as a thing or a tode ti but as
a movement (kinēsis – b17) or power (dunamis – b12)’.

25 Modern biologists attribute this phenomenon to an overproduction of the

tissues that are supposed to become the placenta. Interestingly, since the
placenta is what feeds the fetus, Aristotle was not far off (even if only by
accident) in thinking that the cause of uterine moles has to do with excess
nourishment.

26 It does not seem as though husterikon means merely ‘suffering in the

womb’, ‘hysterical’ or ‘of or belonging to the womb’, which are the meanings
listed in LSJ (online version, August 2015). It likely has rather to do with
overt menstruation, which is most apparent in humans. Other references to
husterikon in GA include husterikous membranes (717a5); husterikon part
(morion) (720b21); husterikon place (738b7). At HA 570a5, (cf. 566a11), the
husterikous passages are contrasted with seminal passages (thorikous).

27 A case might be made for thinking that Aristotle’s purpose in raising this
aporia about the uterine mole is to advance his enquiry, since it allows him
to point to a ‘fact’ about human women not previously mentioned. Thus one
might prefer to consider his purpose here to be zetetic. I am proposing,
however, that this use of aporia be placed in a distinct category. Unlike the
introduction of pneuma or revisions to Aristotle’s theory of the sort we have
seen, the claim that human women alone are husterikon does not play any
further role in Aristotle theory, as far as I can tell.

28 cf. de Somno I, 454b25–9: ‘We say that sleep is in a certain way a

motionlessness (akinēsia) of perception, and like a restraint (desmos), and
that the release and relief is wakefulness. But no plant is able to partake in
either of these affections; for without perception, neither sleep nor
wakefulness obtain.’

29 Besides the fact that birds and fish are both egg-layers, it is not clear why
someone would expect bird and fish generation to be the same. Aristotle
has discussed many other differences between birds and fish earlier in GA.
And Aristotle does not spend any time explaining why this difference, in
particular, should be puzzling. Rather, he immediately proceeds to simply
give the explanation.

30 Another aporia having this structure concerns a peculiarity of selachia

(757a14ff). Female selachia are not seen discharging their kuēmata, and the
males are not observed emitting their milt. Both of these phenomena,
however, are seen in other live-bearers. Here the question being raised is
why selachia do not exhibit all of the characteristics of other live-bearers.

31 Is there some other feature of their structure that might lead us to some
general conclusion? One might wonder, for example, whether there is some
important difference that is being tracked by occurrences of aporia versus
aporēseien an tis. Unfortunately, however, this seems not to be the case.
For, Aristotle uses both the noun and verbal forms when discussing the
same phenomenon on several occasions. Both aporia and aporēseien an tis
are used in the discussions of, for example, the behaviour of semen
(735a29ff), why the female needs the male in order to generate (741a6ff),
and whether the cause of multiple births and redundant parts is the same
(770b30ff).

32 This seems to be reflected in, for example, Platt’s translation: ‘It is not
easy to state the facts about the uterus in female animals, for there are
many points of difference.’ Platt seems disinclined to treat this as a puzzle at
all, but rather a ‘difficulty’ in ‘stating facts’ about the uterus.

33 See also, e.g., his criticism of the idea that hyenas have two pudenda
(757a2–13). ‘Cursory’ or ‘casual’ (ek parodou) observation has produced
this false belief.

34 The same is true of the aporia in GA V.1 about whether sleep or

wakefulness is prior. After concluding that it is not sleep but something ‘like
sleep’ that is the prior state, he immediately goes on to the next topic.

35 As discussed by Rapp in his contribution to this volume (Chapter 6).

36 According to the preformationist theories, the seed from which the

organism develops is composed of parts from each parent’s body.
 Chapter 9
Aporia and the New Academy
◈

James Allen

The verb and the noun, aporein and aporia respectively, mean,
among other things, to be at a loss and the condition of someone
who is. Translators resort to terms like ‘perplexity’ or ‘puzzlement’.
This condition and the awareness that one is in it occupy a
prominent place in Plato’s depiction of Socrates’ philosophical
activity. Unacknowledged ignorance, the false conviction that one
knows, is an impediment that must be replaced with the recognition
that one is at a loss before enquiry that may lead to true
understanding can begin. Plato also uses the term aporia of a
problem or difficulty, in which sense it is pervasive in Aristotle, for
whom systematically working through the difficulties in a domain
(diaporein, proaporein) is an indispensable means to the discovery
and grasp of the truths obtaining in it.
The question whether and how the same ideas were used by
Plato and Aristotle’s ancient philosophical successors is a natural
one. A number of reasons suggest that the philosophers whom we
call ‘Academic sceptics’, though they did not use the term ‘sceptic’ of
themselves, are especially promising figures to consider.
The Academy was founded by Plato. Abundant evidence attests
to the high regard in which members of the Academy also held
Aristotle. Indeed the Academy seems to have been an early home
for the idea that the two philosophers are in agreement on important
matters.1 Arcesilaus (316–242 BCE), the fifth successor of Plato as
head of the Academy began what came to be seen in retrospect as a
‘sceptical turn’ in the Academy’s philosophising – following an
ancient tradition, we speak of him as the founder of the ‘New
Academy’. The examination and criticism of the positions of other
schools, above all the Stoics and principally regarding questions
about the nature and possibility of knowledge, became the chief
occupation of the school. This activity, refined and elaborated and
taking different, competing and sometimes constructive forms,
persisted until the school’s dissolution in the first century BCE.
The resemblance between the way in which Arcesilaus and his
Academic successors subjected the views of the Stoics and others
to scrutiny and the way in which Socrates engaged his interlocutors
in a critical examination of theirs, an activity part of whose point is
often to make the interlocutor aware that he is at a loss and induce
feelings or perplexity commensurate with his ignorance, has often
been remarked. And as we shall soon see, the Academics
themselves acknowledged an affinity with Socrates.
What is more, the other ancient philosophers whom we call
‘sceptics’, the Pyrrhonists, made extensive use of the language of
aporia. Together with ‘sceptic’, zetetic (investigative), ephectic
(suspensive), aporetic was one of the descriptive appellations they
applied to themselves and their school (S.E. PH I 7; DL IX 69–70; cf.
Aulus Gellius 11.5.6). Although the school took the name
‘Pyrrhonian’ from Pyrrho of Elis (365–270 BCE), an older
contemporary of Arcesilaus, there are excellent reasons to believe
that it was in fact founded (or re-founded) in the first century BCE by
Aenesidemus, who was a member of the Academy before he
became disillusioned with it.2 Much of the evidence comes from a
summary by the ninth century CE Byzantine patriarch, Photius, of
Aenesidemus’ lost work, the Pyrrhonian Discourses.3 Drawing on
this evidence, another contributor to this volume argues plausibly
that aporia-language may have been more prominent in
Pyrrhonism’s early Aenesideman phase – so possibly part of its
Academic inheritance.4
The study of the Academy during its sceptical phase presents
special difficulties, however. Like Socrates, Arcesilaus was a non-
writer who exercised an influence on his students and
contemporaries through face-to-face conversation and teaching. The
same is true of Carneades, the most accomplished and influential of
his successors. Other members of the school did write (cf. Galen, De
optimo docendi genere CMG V 1,1, 98, 1–4 Barigazzi). Clitomachus,
Carneades’ student and eventual successor, in particular, was a
prolific author (D.L. IV 67), but none of his works nor those of any
other New Academic author writing in Greek have survived except in
scanty fragments. We are fortunate, then, to have most of the
philosophical works of Cicero (106–43 BCE), who was an adherent
of the Academy.
 Though preserved in an unusual and incomplete form, his so-
called Academica, which is chiefly occupied with epistemological
controversies that dominated the school’s last phase, is an
especially valuable source. We have the second of the first edition’s
two books (the Academica priora or Lucullus), and part of the first of
the four books that composed the second, revised edition (the
Academica posteriora or Varro).5 Both set out and rebut charges
leveled by Antiochus of Ascalon. Like Cicero, Antiochus was a
student of Philo of Larissa (c. 110–c. 79). Unlike him, he became
convinced that the Academy’s sceptical turn was a mistake both
philosophically and as an attempt to remain true to the legacy of
Socrates and Plato. He urged a return to philosophical system-
building, and by dubbing the group that formed around him the ‘Old
Academy’, asserted a claim to continuity with Plato’s immediate
successors. To make matters more confusing, many of the doctrines
he now defended, especially in epistemology, were, or were closely
related to, Stoic views (Luc. 69, 132, 138; S.E. PH I 235). Varro and
Lucullus are spokesmen for Antiochus. In the surviving portions of
both editions of the Academica, Cicero assigned himself the part of
the New Academy’s defender, a task that he had shared in the first
edition with Catulus, who represented a version of the New
Academic philosophy that was in some ways different from his own.
Cicero sometimes pauses to dwell on important Greek terms
and how best to render them into Latin. Though aporia and aporein
are not mentioned in this way, they are among the Greek terms with
which he peppers his letters to Atticus (VII. XI. 3, VII.XII.4, XIII.XIII.2,
XV.IV.2, XVI.VIII.2).6 The only plain use of the language of aporia in
connection with the New Academy is found in The dissension of the
Academics from Plato by the second century Platonist philosopher,
Numenius of Apamea, who relates an involved comic anecdote
about how Lacydes, Arcesilaus’ successor, was led to embrace the
Academic philosophy of ‘inapprehensibility’ (akatalēpsia) by tricks
played on him by his slaves (fr. 26).7 Undetected by him, they
repeatedly circumvent the measures he takes to safeguard the
household stores from theft, which are as a result forever being
inexplicably depleted. Numenius’ use of aporia – Lacydes does not
suffer from a dearth of slaves – and especially the verb aporein (two
occurrences) – he is puzzled or perplexed – could well play on the
Academics’ own use of this language.
 1
Plato’s dialogues sometimes provide an illuminating picture of an
important development by presenting it as though it had occurred in
the course of a single conversation – we shall see something like
this in Cicero as well. Asked to explain what virtue is by Socrates –
something that Socrates says he is not in a position to do himself –
Meno, in the dialogue of the same name, is at first confident; he
does not, as he puts it, lack for things to say (aporia eipein) (72a).8
Perhaps the pre- or non-philosophical use of aporia that is most
relevant here is to be at a loss for things to say, speeches, parts of
the same, arguments, answers to questions. This is the condition
that an orator or controversialist must at all events avoid and which,
in different ways, rhetorical, dialectical and eristical instructors seek
to remedy by ensuring that their pupils are in the opposite condition
of euporia, i.e. well provided with the same. When his efforts to
display the skill to which he lays claim come to naught in the face of
Socratic questioning, Meno gives vent to his frustration. Even before
meeting Socrates, he says, he had heard that Socrates was in
aporia himself and a cause of it in others (79e).
There are cases in which reducing interlocutors to silence or
incoherence does not reflect well on the persons who do it or badly
on those to whom it is done. Those on whom Euthydemus and
Dionysodorus practice their eristical skills earn our sympathy, and
Aristotle takes a dim view of the argumentative techniques by which
controversialists reduce their opponents to committing solecisms or
stuttering (S.E. 3, 165b15–22). And he holds that one may be unable
to defend oneself from sophistical arguments without thereby
showing that one is ignorant (S.E. 8, 169b27–9).
Socrates’ examination of the slave boy is different, however. By
showing that the boy’s confidence that he can solve a geometrical
problem is unfounded, Socrates illustrates the stages through which
Meno must pass himself. Like Oscar Wilde answering the accusation
that outside agitators stir up unrest in the lower orders, Socrates
accepts the charge that he is in aporia and a cause of it in others,
but insists it is to his credit (84b). Socrates benefits his interlocutors
in two ways by bringing about or revealing aporia in them. He
unmasks the illusions that impede enquiry; and the puzzles which he
raises are the ones by tackling and solving which, and doing so by
themselves, his interlocutors advance towards knowledge and
understanding.
The moral in the immediate context of the dialogue is that there
is hope for the kind of joint enquiry into the nature of virtue that
Socrates invites Meno to undertake with him, though they both begin
from a position of ignorance. The broader effect is to vindicate
Socratic dialectic, the rigorous questioning by Socrates, who
professes ignorance, of interlocutors, who profess to know, which
typically induces feelings of puzzlement in them as they are shown
to lack the understanding that they took themselves to have. The
example also shows that, even when one party to a discussion is
more knowledgeable than the other, as Socrates is about geometry,
he does not teach by conveying knowledge, but by confronting the
student with, and guiding him through, the puzzles that he must
solve for himself. Discovering for oneself and learning from another
are standardly treated as alternatives to each other (cf. Prot. 320b,
Phaedo 99c; cf. 85c), but the thrust of the Meno is to diminish the
distance between them. In a way, there is no such thing as teaching;
in a way, there is (84cd; 87bc).
 2
For Aristotle as for Plato, enquiry has a chance to succeed only if is
based on a full and fair appreciation of the problems or difficulties to
which the questions that are its point of departure give rise.9 Though
the metaphor behind the term ‘solution’ (lusis) is dead for us,
Aristotle maintains that the discovery and grasp of the truth sought in
enquiry, is the solution of the aporiai, i.e. the loosing or releasing of
the intellect from the bond that impedes understanding (Metaph. B 1,
995a28–32). If Aristotle’s treatises give us a clue about his method
of teaching, then he too will have favoured an aporematic method of
instruction, requiring the student to retrace a path through the
difficulties first trodden by successful enquirers.
Some of the problems with which the enquirer and the student
must come to grips will be discovered by the enquirer and his
colleagues as they pursue their own researches; others may be
difficulties generally recognised by workers in the field. But Aristotle
lays special emphasis on one source, differences of opinion (though
he acknowledges that some issues requiring attention may have
been overlooked; Metaph. B1, 995a25–7). In many areas at least,
the difficulties by wrestling with which enquirers advance towards the
truth, the range of positions that it makes sense to choose from or
modify, the considerations to which a successful solution must do
justice, the objections that must be disarmed and so on can be found
in good part by attending to views that already have defenders.
 The art of dialectic that Aristotle expounds in the Topics and
Sophistical Refutations provides a method for uncovering difficulties
for a thesis by drawing on reputable opinions or endoxa. Much
scholarly ink has been spilt enquiring whether, where and with what
qualifications the methods of enquiry practiced in the treatises is
dialectical in character. Dialectical and philosophical enquiry are, to
be sure, not one and the same. Unlike philosophy, dialectic is
restricted to endoxa (Metaph. B 1, 995b23–4). And the face-to-face
and competitive character of dialectical practice requires certain
skills which the philosopher can dispense with as he pursues his
enquiries alone or in a different and more cooperative spirit (Top. VIII
1, 155b7–16). But that dialectical and philosophical forms of enquiry
as Aristotle understands them overlap in important ways cannot be
denied.
Dialectical arguments are between two parties: an answerer
whose task is to defend a thesis and a questioner whose job is to
argue against the thesis from the answerer’s responses to his
questions. Detachment is typically required of both parties. The
thesis defended by the answerer need not be one to which he is
committed, and he regulates his answers by consulting not his own
convictions, but endoxa.10 The questioner likewise chooses his
questions with an eye on the same assumptions, which he need not
share. They are prepared to exchange parts (Top. VIII 5, 151b33–5).
Dialectical skill is a faculty for arguing on both sides of the question
(Rhet. I 1355a33–6). Properly used, by drawing out what can be said
on each side, this power can contribute to the discovery and grasp of
the truth (Top. VIII 14, 163b9–16; cf. I 2, 101a34–6; Rhet. I 1,
1355a36–8). One passage in the Topics describes this as raising
aporiai (diaporein) on both sides (I 2, 101a35), and Aristotle
compares the benefits of considering competing arguments in
philosophy to those of hearing both sides in a court of law (Metaph.
B 1, 995b2–4).
 3
Accounts of the New Academy assign a place of importance to a set
of practices, in utramque partem disputatio, argument on both sides
of the question, whose Socratic and Aristotelian provenance the
Academy emphasised.11 There were, it seems, several varieties of
the practice, not all of which require one party to present both sides
of the argument. Arcesilaus is said to have avoided setting out his
own views, if any, but to have encouraged his conversational partner
to present his – the partner’s – views, against which he – Arcesilaus
– would then argue, in this way following the example of Socrates
(De oratore III 67, DND I 11, Fin. II 2). Carneades did the same,
though the account of his lectures for and against justice,
supposedly delivered during an embassy to Rome in 155 BCE,
shows that he could undertake to argue on both sides himself.
Cicero treats the New Academics’ practices of argument in general
as instances of arguing on both sides of the question (Luc. 60, Fin V
10). Even when the Academic presents only one side, both sides are
covered, one presented by the first party and heard by the second,
the other presented by the second and heard by the first (cf. Luc. 7).
The Academics assigned the practice a place of importance in
both teaching and enquiry. In the preface to the Lucullus, Cicero sets
out the New Academy’s objections to the didactic practices of other
schools (Luc. 8). Their students are drilled from an early age in the
school’s doctrines; they are like shipwrecked sailors, clinging
desperately to the rock on which chance has deposited them. By
contrast, from first to last the Academic philosopher preserves his
powers of judgement intact, free from the obligation to uphold school
orthodoxy; it is open to him to pursue the truth unfettered by dogma.
Later in the dialogue, after challenging the New Academics to say to
what truths their practice of argument for and against everything had
led them – implying that there are none – Lucullus quotes the stock
Academic answer: ‘so that students or auditors will be led by reason
and not by authority’ (Luc. 60; cf. Div. II 150, TD IV 7; cf. Galen De
optimo docendi genere 93, 15–17; cf. 100, 2–3 CMG V 1, 1
Barigazzi).
Additional evidence about the Academics is furnished by Stoic
objections to them. Zeno of Citium, likely with the Academy in view
and using the same courtroom analogy that Aristotle had, maintained
that it is wrong to insist that judgement be passed only after hearing
both sides; there is no need to listen to the opposing speaker if the
first has proved his case (Plutarch Stoic. Repugnan. 1034e = SVF I
78 = FDS 40). Though he was not altogether against it, Chrysippus,
maintained that the technique of arguing on opposite sides of a
question should be used sparingly; its unrestricted use suits the
purpose of those ‘who aim to bring about suspension of judgement
about everything’, but not that of those ‘whose object is to impart the
knowledge by grasping which students will live well’ (Stoic repugnan.
1035 F = SVF II 127 = FDS 351; cf. 1037B). The aim of the latter is
best served if they present the opposing case in the manner of a
juridical advocate, enough to prepare the way for the demolition of its
plausibility and no more.
 To summarise our results so far, the new Academics
championed a form of argument, in utramque partem disputatio, as
the best method for both enquiry and teaching. Like Aristotle and,
especially, Socrates, whom they justly regarded as pioneers in its
use, they reduced the distance between enquiry and study. To
achieve knowledge and understanding, enquirers must come to grips
with the difficulties raised by these arguments with an open mind,
free of dogmatic commitments, not swayed by authority and willing to
follow the argument wherever it may lead. Learners must do the
same, and a teacher can at most assist them by bringing the
difficulties to light and compelling his students to grapple with them.
Some of these themes are brought together by Cicero in the
praefatio to the Lucullus, speaking as the author before the
beginning of the dialogue proper (Luc. 7).

[the New Academy’s] case is straightforward, because we want to

discover truth without any contention, and we search for it conscientiously
and enthusiastically. To be sure, knowledge is always surrounded with
difficulties,12 and the obscurity of the things and weakness of our
judgements is such that one can see why the earliest and most learned
philosophers lost confidence in their ability to discover what they desired.
Still they didn’t give up, and we won’t abandon our enthusiasm for
investigation owing to exhaustion. Nor do our arguments have any
purpose other than to draw out or ‘formulate’ the truth or its closest
approximation by arguing for and listening to both sides.
(trans. Brittain modified)

These inspiring sentiments are echoed in other passages that

describe the discovery and grasp of the truth as the Academy’s goal
and their practice of argument in utramque partem as a means to it.
(Luc. 60, 66, 76; Plutarch Stoic. Repugnan. 1037C)
 4
There is a problem, however. Not a few of our ancient witnesses,
including Cicero himself, also maintain that the New Academy held
and defended a position, with variations, of its own. Apart from being
paradoxical in its own right, that position is peculiarly hard to
reconcile with the picture of an Academy committed to open-minded
enquiry.
It will help us to appreciate the strangeness of the New
Academy’s position – a term that I shall use for lack of a better – if
we begin by regarding it from the doxographer’s perspective. We
survey the extraordinary range of views that philosophers of different
schools have entertained – their placita – about a list of questions,
e.g. whether the world is eternal; of how many elements it is made
and what they are; whether the gods are providential; what the end
of goods or the criterion of truth is and so on. Very roughly speaking,
an actual philosophical position arises when to a system of placita
belief, conviction, endorsement or something of the kind, supported
by arguments, explanations and the like, is added. And the belief or
conviction of the philosophers who hold the position typically implies
the further belief on their part that they know, or are well on the way
to knowing, the placita at issue. Their placita are at the very least
what they expect we shall know when the – attainable – goal of
knowing how things are is attained (cf. Luc. 115).
The Academy did not lack for placita of its own. Clitomachus,
cited approvingly by Cicero’s character in the Lucullus, says the wise
man will hold decreta (Luc. 110), the term that renders dogma (cf.
27, 29). The same Cicero says that he shares a decretum with
Lucullus (133); that the Antiochean cause will lie in ruins once the
whole sententia of Carneades has been expounded (98); that one
sententia sets the Academics apart from all other philosophers (70;
cf. 29); he speaks of what Arcesilaus thinks or believes (visa est
Arcesilae, censet Arcesilas, 77, 66). And Cicero freely uses placere
in his description of Carneades’ and Clitomachus’ views (99, 104).
According to our sources, the New Academics were attached
above all to two theses, though some dispensed with the second.13
Very roughly speaking they were:

1) That nothing can be known (more precisely that nothing can

be apprehended, i.e. is perceptum, cognitum, comprehensum in
Cicero’s Latin or that all things are akatalēpta in the New
Academics’ Greek).14

2) That one should suspend judgement about, or withhold

assent from, everything (in Greek epochē).

Let us call these theses, the positions that comprise them and the
arguments advanced in their support, ‘sceptical’.
We are familiar with paradoxes like the liar, propositions that can
be true only if they are false and false only if they are true (cf. Luc.
95–8). The problem with the sceptical placitum that nothing can be
known (and the other, that judgement should be suspended about all
matters) is not the same. It could be true, in which case it would not
be false; or false, in which case it would not be true. Paradox arises
when we try to put it into a doxographical framework, by imagining it
combined with assent, belief or endorsement and supported by
argument, as it must be if a sceptical position is to have more than a
notional existence. If it were true that nothing can be known, it would
be impossible to know that it was, and it would seem, no one who
understood what it meant could affirm it with any confidence. The
problem does not go away if we exempt it – the first of the placita –
from its own scope by taking it to mean that nothing else can be
known. This would solve the logical problem: it now becomes
possible to envisage someone who knew that nothing else could be
known without contradiction. But the epistemological problem
remains. The one thing known would be a freak, stranded in
complete isolation.15 In fact we meet it as the conclusion of a
formidable accumulation of – sceptical – arguments and it would be
known, if it were known, on the strength of the grounds assembled in
these arguments – which grounds, according to the conclusion that
they appear to entail, could not be known. Unsurprisingly there was
long ancient tradition of attributing to the Academics the paradoxical
claim to know that nothing can be known (nova scientia, nihil scire:
Seneca Ep. 88, 44).16
Problems arise even if we view the Academics’ sceptical
positions from a closer perspective. To be an open-minded enquirer
in relation to a domain, one would think, is to regard questions in it
as still open, not settled and still in need of further study. To be one
without qualification would be to maintain the same attitude towards
questions in all domains not excepting those to which the sceptical
placita belong. Dogmas, decreta, placita should be things about
which one no longer needs to enquire, matters which one regards
precisely as settled. Viewing matters from a slightly different angle,
we may well ask with Lucullus (in a passage on which I have already
touched), how if, out of a commitment to open-minded enquiry and
intellectual independence on the part of their students, the
Academics refuse to impose their authority, they can nevertheless
lend their authority to the sceptical thesis that nothing can be known
(Luc. 60). And if the Academics were somehow entitled to make an
exception for it, why did this not give them a powerful reason to give
up enquiring about other matters, the end of enquiry being to know?
Sextus Empiricus, makes the fact that they continue to enquire
the distinctive mark of the Pyrrhonists; the dogmatists do not enquire
because they take themselves to have discovered the truth, the
Academics because they hold that it cannot be apprehended (PH I
1–4).
 5
The tension between the picture the New Academic as an open-
minded enquirer, on the one hand, and as something like a sceptic
by conviction, if you will, on the other, can be seen at its plainest in
the opposition between Lucullus’ overwrought depiction of the New
Academics as convinced sceptics intent on plunging everything into
darkness, depriving us of sight and so on (Luc. 31, 33, 54, 62, cf.
102–3. 110). But it can also be found in a subtler form in the words of
Cicero’s own character in the Academica. My aim in what follows is
to demonstrate the existence of this tension and then to explore the
side that justifies placing the New Academy in the aporematic
tradition. The questions this tension raises about the history of the
Academy and Cicero’s relation to his sources I shall largely put aside
as a problem for another day.
In the Varro Cicero’s character explains Arcesilaus’ sceptical
turn and the motives behind it in this way (44–5):

It was not out of stubbornness or contentiousness, as we accept, that

Arcesilaus began this whole struggle with Zeno, but because he was
moved by the obscurity in things that had led Socrates to his confession
of ignorance, and before Socrates, Democritus, Anaxagoras,
Empedocles and pretty much all the ancients, who said nothing can be
cognised, nothing perceived, nothing known, that our senses are narrow,
our intellects weak, the course of our lives short and, as Democritus put
it, truth was buried in the depths, all things in the grip of opinion and
custom, nothing left for truth, and finally they said everything is shrouded
in shadow. Accordingly Arcesilaus denied that there was anything that
 could be known (sciri), not even this itself which Socrates had left for
himself, viz. that he knew he knew nothing. He thought all things were so
concealed and that there was nothing that could be discerned or
understood. For which reasons [he thought] that no one should profess or
affirm anything nor to approve it with assent, but should hold back always
and from every lapse to keep rashness in check, which is conspicuous
when either something false or unperceived is approved, and [he thinks]
that nothing is more base than for assent and approbation to outrun
cognition and perception. He did what was consistent with this reasoning,
so that by arguing against the views of all he led most people away from
their own, since when equal weights of argument on opposing sides were
discovered, assent was more easily suspended.

By contrast with the passages on which I have dwelt so far, where

Cicero presents in utramque partem disputatio as a means to the
discovery and grasp of the truth, it is said here to have been used by
Arcesilaus as means to spread the sceptical gospel, by revealing to
believers the error of their ways, so promoting his position, viz. that
nothing can be known (and that we should suspend judgement in
consequence).17 In this passage at least, the thesis that nothing can
be known seems to occupy a place apart; unlike other contentions, it
is not one for and against which arguments are brought, but one lent
support by opposing arguments about other matters.
What is more, ‘equal weights of arguments’ cannot fail to remind
us of Pyrrhonian isostheneia and the use of conflicting arguments to
bring about and sustain this state. And it invites the suspicion that
Arcesilaus is, as the Pyrrhonians can seem to be, content to remain
in aporia, and that he regards the activity of wrestling with aporiai
more as a means of attaining and remaining in it than as a means to
the discovery and grasp of the truth. One way of the putting the
problem, then, is this. Was the Academics’ commitment to argument
and aporia more like that of Socrates and Aristotle or that of the
Pyrrhonists’? Or rather more like one side of Pyrrhonism, the
interpretation of which faces a similar difficulty.18
The passage that we have just been examining responds to an
earlier sketch of the Academy’s history by Varro (15 ff.), which
likewise aims to vindicate the claim of the speaker’s party to be the
true present-day representatives of the authentic Academic tradition.
Varro’s Antiochean version begins with Socrates, who is credited
with two innovations: first with turning philosophy away from the
mysteries of nature and towards common life, the virtues and vices
and things good and bad; second, with a profession of ignorance –
here the claim not to know anything apart from this itself, that he
does not know anything (16). Though treated with great respect,
Socrates occupies a place apart in Varro’s history; he is an innovator
who breaks with the past, but who, apart from the central place
henceforth to be occupied in philosophy by ethics, exerts no lasting
influence. It is left to Plato to set philosophy on its proper course, one
which, with room for additions and corrections by Aristotle and Zeno
of Citium, it has followed successfully ever since – or would have
were it not for the New Academic turn that began with Arcesilaus,
likewise an isolated figure who, apart perhaps from one aspect of
Socrates, had no antecedents and whose ‘defection’ (discidium,
defectio) therefore stands in special need of defence and
explanation (43; cf. Luc. 14–15).
 This Cicero’s alternative history at Varro 45 supplies, but
apparently at the cost of constructing a succession of philosophers
who seem to be united by their unswerving commitment to
something like a sceptical orthodoxy, the conviction that ‘nothing can
be cognised, nothing perceived, nothing known’.19 Referring back to
the counterpart of this speech in the lost first book of the first edition,
Lucullus faults Arcesilaus for hiding behind the authority of the
ancients who (allegedly) denied that anything can be known or
perceived (Luc. 15).20
Cicero’s character turns to the subject in the Lucullus at 72–6 in
a passage that he represents as an appeal to ancient authorities and
an explanation for Arcesilaus’ innovations (76). It responds to
remarks by Lucullus (13–15), which should in turn be a response to
a speech in the now lost part of the first edition that was the basis for
Varro 45.21 Lucullus’ object is, once again, to isolate Arcesilaus and
his Academic followers (cf. Luc. 61). To this end, he questions –
fairly enough – whether it is right to regard the ancient philosophers
as sceptics simply because of their occasional expressions of doubt
or frustration (Luc. 14). He removes Socrates from the list of those to
whom the Academy can appeal as authorities, explaining his
profession of ignorance as irony (15).22 And Lucullus charges that
the Academic party’s appeal to ancient philosophical authority
resembles the efforts of seditious persons to hide their designs
behind a pretended affinity with illustrious heroes from the Roman
past (13).
Cicero’s response to Lucullus in defence of the Academy’s
appeal to history differs in tone and emphasis from his remarks at
Varro 45. I should like to suggest that it fulfills the expectation
aroused by the praefatio at Luc. 7 for a different kind of sceptical
history and a different understanding of Arcesilaus’ philosophical
innovations, one that stresses open-minded enquiry rather than
sceptical conviction.23
To Anaxagoras, Democritus and Empedocles, already
mentioned in the Varro, Cicero first adds Xenophanes, Parmenides
(mentioned by Lucullus at Luc. 14) and Metrodorus of Chius, before
going on to cite as the authors of aculatea sophismata Stilpo,
Diodorus and Alexinus, and finally Chrysippus and the Cyrenaics.
Though some of these figures are credited with sceptical or
sceptical-seeming sentiments, others are not.24 Anaxagoras earns
his place here not, as in the Varro, because he said nothing can be
known, but rather because he said snow is black (cf. 100). The
Cyrenaics’ view that we apprehend only our own affections, which
resembles positions sometimes regarded in modern and
contemporary philosophy as sceptical, is here treated by Cicero as
another variety of dogmatism, albeit an unusual one that can be
opposed to more familiar forms (cf. 142). By raising difficulties, the
sophisms of Stilpo, Diodorus and Alexinus serve a similar purpose.
Chrysippus, though a Stoic, is mentioned here because the
arguments against the senses and common experience that he
collected with a view to refuting them shows that they merit
respectful attention (cf. Plutarch, De stoic. repugnan. 1036b–c).
To be sure, this way of interpreting the passage faces several
obstacles.
In particular, the reference to Socrates and Plato is
disappointingly brief and appears to depict Socrates in a more
dogmatically sceptical light than the Varro. It is one thing to say, as
Socrates is held by Varro to have said, I do not know anything
except this itself (that I do not know anything, scil. else) (Varro 16, cf.
44).25 This is non-committal about what it is possible for Socrates or
other people to know, and about what other people know, e.g. about
themselves. It is quite another to think, as Cicero appears to say
Socrates did, that nothing can be known, with one exception,

… Socrates thought that nothing could be known. He made just one

exception, that he knew that he knew nothing, and excluded nothing else.

(trans. Brittain)26

How are we to understand the one thing that is known and is an

exception to the rule that nothing can be known? Is it that Socrates
can know that he does not know (anything else)? Could other people
know this of Socrates? Of themselves? Or is the only thing that can
be known something that can be known only of Socrates and by
Socrates alone? This would be exceedingly odd, among other things
ruling out the possibility that Socrates could serve as a model for
others. To be sure, for one to know of oneself that one does not
know, or for a person to know of himself or herself that he or she
does not know would be less odd and might count as a single
exception to the general rule that nothing can be known, but this is
not what Cicero says and it would be, at best, a very strained
interpretation of Socrates’ famous confession of ignorance (cf. Plato,
Apology 21b).27
 There remains the fact that Cicero, contrasting the New
Academy’s views with Democritus’, says ‘we do not deny that
anything is true; we deny that anything can be apprehended (percipi
posse)’ (73). For reasons that will be plain soon, I think it is
noteworthy that Cicero here speaks of the impossibility of
‘apprehension’ not as his character in the Varro and Lucullus did of
knowledge or apprehension (cognosci … percipi … sciri Varro 44;
sciri … percipi Luc. 15).
The impression conveyed by Lucullus 72–6, then, is more that
of a heterogeneous set of difficulties and arguments that furnish
material for enquiry by raising problems for the self-confident
dogmatism of Lucullus and philosophers like him than a roll call of
convinced sceptics. And when Cicero then turns to Arcesilaus –
‘enough of authorities’ he says – responding to yet another
accusation of Lucullus’, viz. that Arcesilaus took up the debate with
Zeno of Citium in a spirit of contention and in order to disparage him
(obtrectare), he insists that it was rather because he wished to
discover the truth (Luc. 76–7; cf. 16). This contrasts sharply with the
reasons why ‘Arcesilaus began his whole struggle with Zeno’
specified at Varro 44–5, where Cicero says he was led by the same
considerations that moved numerous illustrious predecessors to
maintain that nothing can be known to draw the same conclusion.
Three features especially distinguish the – no doubt idealised –
depiction of Arcesilaus here from the corresponding passage in the
Varro. First, the order and the relative importance assigned to the
considerations that support the two sceptical decreta are reversed.
The emphasis in the Lucullus is on the proscription of opinion and
not the alleged impossibility of knowledge as the motive for the
suspension of judgement. Second, the conflict between Arcesilaus
and Zeno is treated here not as one episode among others, but as a
turning point. Third Cicero speaks here and in the broader context of
this passage of apprehension and of arguments that nothing can be
apprehended (percipi posse), rather than casually running such talk
together with talk of knowledge and its impossibility as Lucullus does
and his character had done at Varro 45 (percipi … sciri) (Varro 44–5;
Luc. 15).28
The proscription of (mere) opinion is not an inspiration of
Arcesilaus’ own, but an idea he takes from Zeno, to whom he is
grateful for it.

None of Zeno’s predecessors had ever explicitly formulated, or even

suggested, the view that a person could hold no opinions – and not just
that they could, but that doing so was necessary for the wise person.
Arcesilaus thought that this view was both true and honourable as well as
right for the wise person.
(Luc. 77; cf. 66, 113) (trans. Brittain)

Arcesilaus was undoubtedly pre-disposed to be receptive, not

however as a convinced sceptic, but, Cicero emphasises, as truth-
seeker. And it is in the same spirit, I suggest, that he tackles the
Stoics’ theory of knowledge; it is not simply more grist for the long-
convinced sceptic’s mill, but a promising view worthy of examination.
Cicero describes Arcesilaus pressing Zeno to say what would
happen if nothing were apprehensible by the wise person and then,
after Zeno affirms that there is something that can be apprehended
(quod percipi posset), to say what is apprehensible. Zeno replies
with the two-clause version of the definition of the cognitive
impression as ‘an impression (i) from what is (ii) stamped and
impressed in exact accordance with what is’, and when pressed
further, with the third clause (iii) ‘such that there could not be a false
impression like it’ (77).29
According to Cicero:

Arcesilaus agreed this was rightly added as neither a false impression or

a true impression like which there was a false one was apprehensible;
and it was incumbent on him in these disputes to show that there was no
true impression such that there could not be a false one of the same kind
[i.e. aparallaxia].

Despite appearances, this passage can be seen to lend support to

the so-called dialectical interpretation of the Academy.30 Its
proponents hold that the Academy’s members argued in a way that
corresponds to one species of Aristotelian dialectic. In it, the
answerer is no longer detached, but replies in accordance with his
own views (Top. VIII 5, 159a39–b1, cf. b26–7). The task of the
questioner is to refute the answerer by arguing from his – the latter’s
– own views. If he succeeds, he will have drawn attention to
problems internal to the answerer’s position. This form of dialectic
Aristotle sometimes calls peirastikē (S.E. 165b3, 8, 169b25, 11,
172a22), and it resembles Socratic dialectic as we know it from
some of Plato’s dialogues (cf. S.E. 34, 183b7). On this interpretation,
Arcesilaus, Carneades and their Academic emulators play a part
corresponding to that of the questioner. It is clearly an excellent way
of uncovering difficulties that the defender of a position must tackle
and solve and, just as plainly, it need not tell us anything about the
views of the questioner, who can be acting as something like an
advocatus diaboli.
The dialectical interpretation the New Academy emphasises the
extent to which sceptical theses and arguments draw on Stoic theory
and vocabulary. The Stoic definition of the cognitive or apprehensive
impression, which I have just cited, plays a critical part in the
Academic’s argument for inapprehensibility; it specifies the
conditions that the Academy argues cannot be met. And the
Academics emphasised especially that suspension of judgement is
required in the absence of cognitive impressions by a Stoic doctrine,
the proscription of opinion (cf. M VII 155 = LS § 41  C; Cicero Luc.
68, 78).
The attitude of detachment the dialectical interpretation
attributes to the New Academics stands in stark contrast to a naively
dogmatic form of scepticism that unreflectively takes over
substantive and potentially contentious elements of a philosophical
theory, e.g. the Stoics’; it then assembles sceptical arguments to
show that the conditions specified by the theory for, say,
apprehension cannot be satisfied, and concludes that nothing is
apprehensible. Taking the Stoic requirement that judgement should
be suspended in the absence of cognition, it goes on to endorse
universal suspension of judgement.
The zeal with which Arcesilaus appears to assist in the
formation of the Stoic theory of the cognitive impression might
suggest that he embraces the doctrine that emerges as a result, but I
suspect that the obligation incumbent on him is not that of upholding
a sceptical dogma, but rather a dialectical obligation under which he
placed himself by entering these disputes, namely to seek out and
direct his fire against the weak points in his opponents’ position,
including improved versions of it that his arguments have elicited. He
does not uphold, but merely argues for, inapprehensibility.
As we have seen, however, his detachment is not that of a mere
controversialist. His attitude towards the Stoic position also includes
a measure of sympathetic engagement. The Stoic theory seems to
be regarded by Arcesilaus, as Cicero understands him, neither as
the home to some indisputable truths, e.g. about the nature of
knowledge, nor simply as an object upon which to exercise their
dialectical facility. Rather it is an especially interesting specimen,
whose theses, together with the problems and difficulties brought to
light by the arguments directed against them, are somehow
exemplary. Raising and wrestling with them are means to
discovering the truth, the motive Cicero gave for Arcesilaus’ debate
with Zeno (76). The picture of Arcesilaus of which we get a glimpse
here, then, is of a philosopher enquiring in the aporetic spirit of
Socrates, Plato and Aristotle, using the method of in utramque
partem disputatio to this end, and not excluding questions about the
nature and possibility of knowledge from its scope.
The significance of Cicero’s choice of apprehension and of
arguments that nothing can be apprehended (percipi posse), rather
than knowledge and its impossibility seems to be this.31 In the
context of Stoic epistemology, apprehension is necessary but not
sufficient condition for knowledge. If apprehension is impossible, so
is knowledge, a condition defined by still tighter restrictions. Outside
that context, matters are less clear; the assertion that nothing can be
known may sound not like the denial that the strictest epistemic
conditions conceivable can be satisfied, but rather like a blanket
denial that standards entitling us ever assert anything at all can be
met, i.e. like the sceptical attitude apparently being attributed to the
ancients in the Varro (and by Lucullus to Arcesilaus). If the language
of apprehension is used strictly and with the Stoic framework in view,
then the possibility of something that might have qualified, outside
the Stoic framework, by lower or different standards, as knowledge,
even if Cicero and the New Academics would not call it that
themselves, or something that can serve in its place, is not excluded
together with the possibility of apprehension (strictly so-called).
Enquiry aimed at that something remains eminently worthwhile. And
Cicero often emphasises just how little we should be concerned if, as
the Academy’s arguments appear to show, apprehension is
impossible (Luc. 32, 34, 99, 102, 103, 105, 110, 146).32
 6
Cicero now effects a transition involving a shift in focus. The
connecting link is inapprehensibility for which we have just heard
Arcesilaus argued on the basis of aparallaxia. This is the one
contention, Cicero maintains, that has persisted to the present day;
the proscription of opinion, which had so captivated Arcesilaus, is
now relegated to a place of secondary importance – it is optional for
the purpose of the present argument.

For this, the view that the wise person will assent to nothing, did not
pertain to this controversy. For it was permitted ‘to apprehend nothing and
nevertheless to opine’, which is said to have been approved by
Carneades, though for my part, believing Clitomachus more than Philo
and Metrodorus, I take it to have been argued for rather than approved by
him.
(78)

It is implied that one thesis, that nothing is apprehensible, was not

merely argued for, but also approved.
Evidence for the existence of sceptical decreta is clearest in
connection with Philo and Clitomachus to the debate between whom
Cicero adverts here. Clitomachus approved both, Philo only the first.
It is a nice question what exactly the disagreement between Philo
and Clitomachus comes to. Though the latter rejects assent for the
wise person, he distinguishes between two kinds of withholding or
suspension, only one of which, that applying to assent, is required of
the wise person, and by implication two kinds of favourable reaction
to impressions, only one of which, the one for which he reserves the
term ‘assent’, is off limits (Luc. 104).33 Likewise the question of how
the decision by Cicero, who plainly owes much to his teacher Philo,
especially it seems his attitude to probability, to side with
Clitomachus on this point affected his version of the Academic
philosophy. I shall largely ignore them here and touch only very
briefly on another intriguing complication.
Late in his career, Philo experienced a change of mind, which
depending on how one views it, was either a complete betrayal of
everything he had ever stood for and an incoherent mess of a view
to boot or a natural development, perhaps no more than a
refinement or clarification of, the outlook that he had held all along
(Luc. 12, 18; S.E. PH I 235). His new ideas, set out in the so-called
Roman books, were roughly that the Academy’s arguments showed
not that nothing is apprehensible, but merely that matters are
inapprehensible so long as the Stoic account of apprehension is in
force. The implication is that knowledge, on a correspondingly
revised conception of it, is also within our reach.
Our purposes here will be best served if we attend to Philo’s
earlier view, which seems to have made the deepest impression on
Cicero (though he dissents from it on the crucial point about which
he agrees with Clitomachus). According to it, though there are no
cognitive impressions, the wise person will nevertheless assent, with
the proviso that he might be wrong, so forming opinions and violating
the Stoic proscription against the same. He will regulate his assent
by probability. Views that, after careful scrutiny with due
consideration given to objections and alternatives, are seen to be
probable deserve to be accepted, though one must never lose sight
of the fact – as it seems it is – that their truth cannot be guaranteed
with certainty. One of the views that deserve to be accepted in this
way is that nothing can be apprehended, and therefore if the
implication is accepted, that nothing can be known. But understood
in this way, the first New Academic thesis is not incompatible with
open-minded enquiry. Indeed the fact – as it seems that it probably is
– that highly probable views can be discovered by means of well and
properly conducted enquiry is a spur to more enquiry.
If, on the other hand, knowledge turns out to be possible on a
better understanding of knowledge, e.g. the understanding that Philo
championed in the Roman books or another which emerged from the
Academy’s in utramque partem disputes, this too would have been a
triumphant vindication of the Academy’s version of the aporematic
method – however designated – and a powerful inducement to
continue applying it to other questions.

1 Weische 1965, Karamanolis 2006.

2 Details of the argument are in Mansfeld 1995; but cf. Polito 2014:49.

3 Text and translation is in LS §71 C; Polito 2014: 62–137.

4 See Castagnoli in this volume (Chapter 11).

5 See Griffin 1997, Brittain 2006: xiii–xix.

6 Cicero refers to logical puzzles like the liar as inexplicabilia, which

corresponds to the Greek apora, but this is a Stoic usage (Luc. 95–97; DL
VII 44, 82, 175, 198).

7 Cited from des Places 1973.

8 Cf. Szaif in this volume (Chapter 2).

9 See Buddensiek in this volume (Chapter 7).

10 For complications, see Top. VIII 5, 159a38–b29.

11 Cicero, De oratore I 84, III 67, 80, 84, 107, DND I 11, II 168, Fin. II 1–2, V
10, Off. II 8, Varro 16, 46, Luc. 7, 60, 104, Tusc. II 9, Div. II 150; DL IV 28,
Plutarch Stoic. repugnan. 1037c, Galen De optimo docendi genere 93, 15–
17; cf. 100, 2–3 CMG V 1, 1 Barigazzi, Numenius fr. 26, 103–107 Des
Places = Eusebius Praep. ev. XIV 7, 15; Index Academicorum XIX 35 ff.

12 Where Cicero has difficultates his Greek authorities may have had
aporiai.

13 Cf. Striker 1980.

14 Which granted certain easily made, but not entirely uncontroversial,

assumptions implies that nothing can be known. Sometimes the distinction
seems to be ignored: [The ancient authorities said] nihil cognosci nihil
percipi nihil sciri posse (Varro 44); [The ancient authorities denied] quicquam
sciri aut percipi posse (Luc. 15). At other times it seems to be respected.

15 Cf. Luc. 110. (If the Academic wise person had the nota cognitionis in the
decretum, nothing can be known, he would use it elsewhere too.)

16 Cf. Luc. 28–9, 109. On these issues in the Lucullus, see Burnyeat 1997.

17 Cf. Plutarch Adv. Colotem 1121F–1122A.

18 Cf. Palmer 1990.

19 Nihil cognosci, nihil percipi, nihil sciri posse.

20 Qui negavissent quicquam sciri aut percipi posse.

21 Cf. Brittain and Palmer 2001.

22 The apparent discrepancy between this picture of Socrates and Varro’s

has been the occasion for puzzlement and speculation. Cf. Burnyeat 1997,
Sedley 2012, Brittain 2012.

23 Contra Brittain 2001: 175–8 and Cooper 2004: 86–8, who regard the two
historical accounts as in essential agreement, but in agreement with the
broader interpretative aims of both: with Brittain, that both histories are
meant to defend the innovations of the New Academy that began with
Arcesilaus as they were understood before Philo of Larissa’s Roman books;
and with Cooper, that the Varro’s account of Arcesilaus is unsatisfactory (‘As
Cicero presents him, he grossly contradicts himself’ 87 n. 11). If the two
histories are in tension with each other, what are we to make of the fact that
Cicero puts both of them in the mouth of his own character? It may be that
the history spoken in the Varro by Cicero was spoken by Catulus in the first
edition (So Hirzel 1895, vol. I 509 n. 4). To be sure, Lucullus’ observation
that ‘in citing the early physicists, what you are doing – here he addressed
me by name – seems to me to be exactly what seditious citizens when they
list a selection of famous men from the past (13)’ suggests that he is
responding to a speech of Cicero’s, but he uses the second person plural –
in effect ‘you people’; later when he urges Cicero to mend his ways, he uses
the singular (Luc. 61–2). If this hypothesis is right, was Cicero guilty of
carelessness when he reassigned Catulus’ part in this way or did he
overlook the different tenor of different sources? It is not impossible, but
there was scope for qualifications in the lost portions of the first book and
the lost books of the second edition.

24 The passage seems to present Xenophanes and Parmenides as general

or unrestricted sceptics, but need this be so?
illi [Xenophanes and Parmenides] … increpant eorum adrogantiam quasi
irati, qui cum sciri nihil possit audeant se scire dicere.
These philosophers, as though enraged, rebuke the arrogance of those
who dare say they know [,] when nothing can be known.
Are Xenophanes and Parmenides angry at people who claim to know
even though knowledge is impossible or at those people who claim to know
in circumstances in which nothing can be known (though there may be
others in which knowledge is possible)? Brittain’s comma before ‘when
nothing can be known’ tips the balance in favour or the former, perhaps
rightly, but the second cannot be excluded.

25 Reid 1885 ad loc.

26 … Socrati nihil sit visum sciri posse; excepit unum tantum, scire se nihil
se scire, nihil amplius (Luc. 74).

27 For different reasons, Burnyeat 1997: 297–300 suspects this passage; he

proposes ways of narrowing the scope of the thought that nothing can be
known attributed to Socrates and also suggests that the posse, without
which Socrates would be claiming simply not to know, was a slip of Cicero’s.

28 Cf. n. 14.

29 Cf., e.g. Frede 1983, Barnes 1989.

30 Couissin 1929, Striker 1980.

31 Cf. nn. 14, 20.

32 Cf. Allen 1997: 237–41.

33 On this passage, see Hirzel 1883: 168 n. 1.

 Chapter 10
Aporetic Elements in Plutarch’s
Philosophy
◈

John Dillon

Within the Platonist tradition, Plutarch would not normally be

regarded as being particularly close to the sceptical or aporetic wing
or tendency, but – as Jan Opsomer has well pointed out in his
extremely useful book In Search of the Truth1 – that impression
would not be entirely correct. For one thing, despite his own
adherence to positive dogma, Plutarch is one of the few later
Platonists to persist in regarding the philosophers of the New
Academy as an integral part of the Platonic tradition, and he likes to
strike sceptical attitudes, particularly when engaging in polemic with
the Stoics and, to some extent, with the Epicureans.2 The existence
of On the Unity of the Academy from Plato – now sadly lost, but
attested in the Catalogue of Lamprias (no. 63) – testifies to this
tendency of his and stands in sharp contrast to the strong attack on
the Sceptical Academy launched later in the second century by the
Neopythagoreanising Platonist Numenius of Apamea, in his work On
the Divergence of the Academics from Plato (Frr. 24–6 Des Places).
Indeed, even Plutarch’s composition of dialogues, starting as
they each do from an aporia of one sort or another (e.g. ‘What is the
meaning of the E at Delphi?’, On the E in Delphi); ‘What is the
reason for the decline of oracles?’ (The Decline of Oracles); ‘What
was the nature of Socrates’ daimonion?’ (On the Daimonion of
Socrates), ‘What is the reason for the face that appears on the
Moon?’ (On the Face in the Moon) – leading to the more general
question ‘What is the substance of the Moon?’ – can also be seen as
a manifestation of this tendency. He shows a deep interest in, and
sympathy for, the personality of Socrates and of his peculiar mode of
philosophising, both in his dialogue On the Daimonion of Socrates
and in the first Platonic Question, which involves an examination of
the rationale behind Socratic ‘midwifery’.
However, it must be admitted that, despite his reverence for
Socrates (which provides a continuity with that which animated the
leaders of the New Academic tradition, such as Arcesilaus and
Carneades), for Plutarch, an aporia – as particularly evidenced by
the nine books of Symposiac Questions, as well as the volume of
Questions about Nature (problēmata phusika) – is more in the
tradition of the Aristotelian problēma than of properly Socratic
aporetic questioning, as practised by the New Academy.3
 1
A good place to start our investigation, I think, would be the rather
programmatic first zētēma of Plutarch’s collection of Platonic
Questions, which poses the question, arising from a passage which
obviously much affected the New Academicians, Theaet. 149a–51d:
‘Why ever did the god, as stated in the Theaetetus (150c7–c8), bid
Socrates act as midwife to others, but prevent him from himself
begetting?’ This passage had already been dealt with by the
Anonymous Theaetetus Commentator (cols. 49–56) – where Anon.,
it must be said, is rather more concerned than Plutarch to counter
the New Academics’ embracing of Socrates as a true sceptic;
Plutarch, on the other hand, is prepared to accept Socrates’
declaration of ignorance as quite free of irony and fooling.4 He also
feels the need to defend Socrates’ apparently arrogant comparison
of himself to a god at 151c5–d3:

For a great many men, my excellent friend, have got into such a state of
mind towards me as practically to bite when I remove some silliness of
theirs; and they do not believe that I am doing this out of benevolence, for
they are a long way from knowing that no god is malevolent towards men,
and that neither do I do any such deed out of malevolence (oudeis theos
dusnous anthrōpois oud’ egō dusnoia[i] toiouton ouden drō), but that it is
quite illicit for me to admit falsehood and suppress truth.
(trans. Cherniss)

To take this latter passage first, it would seem to me probable that

Plato intends Socrates to mean merely that he is acting as the
servant of God (or more specifically, of the god Apollo)5 and that the
god would not have commissioned him to do any evil to his fellow
men, but this solution (which itself leaves Socrates sounding
somewhat pompous) does not seem to have occurred to any ancient
commentator. In attempting to defend Socrates against a charge of
pomposity and ‘talking big’ (megalēgoria), such as was levelled at
him happily by Epicureans such as Colotes, Plutarch offers the
solution that Socrates is really just referring to his own intellect as
the highest part of himself, which is characterised by Plato at the end
of the Timaeus (90a) as a daimōn given to each of us by God –
though Plutarch here instead quotes Menander (‘for our intelligence
is god’, Fr. 762 Kock) and Heraclitus (‘the character of a man is his
daimōn’, Fr. B114 D–K).
As to the former, and primary, point, Plutarch (after explaining,
at 999EF, that Socrates saw it as his mission to rescue young men
from the clutches of the Sophists, who were filling them with self-
conceit (oiēma) and sham wisdom (doxosophia), and encouraging
them to futile disputatiousness, by introducing his elenchus like a
purgative medicine, kathartikon pharmakon) goes on to explain why
this calls for the suppression, or even the effective absence, of any
partisan views of one’s own (1000A):

In the second place, while the exercise of judgement is beneficial,

begetting (sc. of doctrines) is an obstacle to it, for what loves is blinded
about the thing it loves,6 and nothing of one’s own is so beloved as is an
opinion or an argument by its parent. For the distribution of offspring that
is proverbially7 most just (hē gar legomenē teknōn dikaiotatē dianomē) is
most unjust when applied to arguments, for in the former case one must
 take what is one’s own, but in the latter what is best, even if it be
another’s. For this reason, the man who begets his own becomes a
poorer judge of others.

(trans. Cherniss)

His argument here would suit either the more properly sceptic view
of Socrates as claiming to know nothing, or the more broadly
Platonist view of Socrates as holding back on his opinions for
pedagogical reasons. From what follows, however, it appears that
Plutarch is more inclined to the sceptic position. He adduces the
comparison of someone whose hearing is obstructed by internal
ringing and buzzing (a sort of tinnitus, one might say) with one
whose judgement is obstructed by the possession of doctrines to
which he is devoted; he simply cannot give a fair hearing to other
points of view (1000BC). He continues (1000C):

Furthermore, if nothing is apprehensible (katalēpton) and knowable to

man, it was reasonable for God to have prevented Socrates from
begetting inane and false and baseless notions and to compel him to
refute (elenchein) the others who were forming such opinions. For the
discourse that liberates from the greatest of evils, deception and vanity,
was not a slight but a very great help – the gift God didn’t grant even to
the sons of Asclepius (Theogn. 432). For the treatment given by Socrates
was not of the body, but was a purgation of the ulcerous and corrupted
soul.

Plutarch here plainly has in mind (as duly noted by Cherniss ad loc.)
such a passage as Soph. 230c–1b, where the visitor from Elea is
commending the method of elenchus as the best purgative of the
corrupted soul. Interestingly, though, after all this emphasis on
Socrates’ freedom from opinions of his own, Plutarch ends his first
Platonic Question by asserting that Socrates’ real purpose in
subjecting young men to the elenchus was not simply to reduce
them to perplexity, but rather to provoke in them a reminiscence
(anamnēsis) of true reality, of which they have innate conceptions
(1000DE):

Consider too that, while the other things, poetry and mathematics and
rhetorical speeches and sophistic doctrines, which the daimonion
prevented Socrates from begetting, were worth no serious concern, what
Socrates held to be alone wisdom, that which he called ‘passion for the
divine and intelligible’,8 is for human beings a matter not of generation or
of discovery, but of reminiscence. For this reason Socrates was not
engaged in teaching anything, but by exciting perplexities (aporiai) as if
inducing the inception of labour-pains in young men, he would arouse
and quicken and help to deliver their innate conceptions (emphutoi
noēseis)9; and his name for this was obstetric skill (maiōtikē technē),
since it does not, as other men pretended to do, implant in those who
come upon it intelligence from without, but shows that they have it native
within themselves but undeveloped and confused and in need of nurture
and stabilisation.

The upshot of this whole zētēma, then, would seem to be to situate

Socrates judiciously within the Platonic tradition, recognising on the
one hand both the genuineness of his claim to ‘know nothing’, at
least of those areas of expertise professed by sophists and other
‘experts’, and on the other his concern, by proper maieutic
procedures, to lead his interlocutors to the ‘recollection’ of Forms, or
the activation of their ‘innate concepts’. In one respect, therefore, his
claim to ‘know nothing’ is straightforward and has two aspects: (1)
his abjuring of any expertise in what one might term ‘science and
technology’, such as that professed by sophists, and (2) his belief
that knowledge in the fullest sense (that is, a complete
comprehension of the causes and effects of everything) belongs to
God alone. However, on the other hand, his claim of ignorance, the
holding back of any views of his own, primarily on moral questions,
can be seen as a maieutic strategy to draw out, through the
elenchus, the concepts inhering in his interlocutors.10
 2
This, then, constitutes one good instance of Plutarch’s attitude to the
aporetic aspect of his Platonic heritage. I propose to focus on just
two others. First, as I have remarked earlier, one may expect to find
Plutarch at his most aporetic when attacking one or other of his two
main bugbears, the Stoics or the Epicureans – in either case, largely
because of their attacks on the Sceptical Academy of Arcesilaus and
Carneades.11 Of the five treatises in question, two against the Stoics
(On the Contradictions of the Stoics; On Common Notions, against
the Stoics)12 and three against the Epicureans (That a Pleasant Life
Is Impossible on the Principles of Epicurus; Is ‘Live Unknown’ a Wise
Precept?; Against Colotes, on Behalf of the Other Philosophers), I
will select some passages from the last one, since here Plutarch
deals explicitly with Socrates (1116E–19C), as part of a long list of
earlier philosophers, starting with Democritus, whom Colotes has
attacked.
Colotes’ primary target was in fact his contemporary
Arcesilaus,13 but for the purpose of undermining Arcesilaus’ position,
he chooses to place him in a sequence of philosophers who denied
the possibility of knowledge, among whom is Socrates. In his
polemic, Colotes is adopting a version of the Stoic imputation against
Arcesilaus of apraxia, the impossibility for the sceptic, if he withholds
assent to impressions, of acting at all – at least in any rational or
consistent way. Why should he not eat grass instead of bread? Why,
when he goes to eat, should he not try to stuff the food into his ear
rather than his mouth? Why should he not try to exit a room by
walking through the wall rather than the door? Why, when he wishes
to bathe, should he not head for the mountain rather than the baths?
These gibes may have had some force against anyone
maintaining total scepticism, as Arcesilaus appears to have done
(though possibly only for polemical anti-Stoic purposes), but
Plutarch, as becomes plain, is operating on the basis rather of the
nuanced sceptical stance of Carneades, where one may act on the
basis of perceived degrees of probability,14 though without ever
conceding the possibility of Stoic katalēpsis. Plutarch, admittedly,
nowhere sets out the elaborate three-stage scale of plausibility that
Carneades seems to have propounded, but it is pretty clear that it is
Carneades’ position that he is adopting.15 After some brisk satirical
ripostes to Colotes, he comes out with the following (1118B):

For it is only in doctrine and argument that these sages have the
advantage over the rest of us;16 to perceive with the senses and to
receive impressions when confronted with appearances happens to
everyone, since it is the work of causes that have nothing to do with
reasoning. The inductive argument by which we conclude that the senses
are not accurate or trustworthy does not deny that an object presents to
us a certain appearance, but forbids us, though we continue to make use
of the senses and take the appearance as our guide in what we do, to
trust them as entirely and infallibly true (to pisteuein hōs alēthesi pantē[i]
kai adiaptōtois ou didōsin autais). For we ask no more of them than
utilitarian service in the unavoidable essentials, since there is nothing
better available; but they do not provide perfect knowledge and
understanding of a thing that the philosophical soul longs to acquire.

(trans. Einarson-De Lacy)

This, it seems to me, serves as a very good statement of at least the
Carneadic position, which Plutarch seems to endorse. We do not
have to regard the senses as absolutely accurate or trustworthy in
order to make use of them as our guide in day-to-day living, without
taking them to be ‘entirely and infallibly true’. The very last sentence,
however, recalls us to the fact that there is, after all, a realm of true
being, knowledge and understanding (epistēmē kai gnōsis) of which
can be attained by the use of the mind, not the senses. We have no
evidence that Plutarch subscribed to the pious rumour, possibly
already promulgated by Philo of Larissa,17 but firmly attested both by
Numenius (Fr. 25. 75–83) and by Sextus Empiricus (PH I.234), that
Arcesilaus had only practised scepticism as a polemical tool against
the Stoics, and to test the acumen of his students, while
‘dogmatising’ in private, but his emphasis on Socrates’ desire to
stimulate anamnesis in his students seems to contain a suggestion
of this theory, or at least to be compatible with it.
However that may be, Plutarch, after dealing briefly with the
Megarian Stilpo, turns (1121Eff) to a defence of Arcesilaus himself,
who has been the real object of Colotes’ whole tirade. Significantly,
Plutarch, having accused Colotes of projecting back Arcesilaus’
maintenance of epochē onto not only Socrates and Plato, but also
such figures as Parmenides and Heraclitus, is quite prepared to
accept that Arcesilaus would have been justified in this (1122A). The
issue, after all, concerns the perception of the physical world about
whose uncertain epistemic status Plato is on the side of his sceptical
followers. One may adduce, for instance, such a passage as Phd.
65ab, where Socrates says to Simmias:
 Now take the acquisition of knowledge (phronēsis). Is the body a
hindrance or not, if one takes it into partnership to share an investigation?
What I mean is this. Is there any certainty in human sight and hearing, or
is it true, as the poets are always dinning into our ears, that we neither
hear nor see anything accurately? Yet if these senses are not clear or
accurate, the rest can hardly be so, because they are all inferior to the
first two. Don’t you agree? Certainly.

Such a passage can be connected with, for example, Socrates’

exposition of the ‘Protagorean’ theory of perception in the
Theaetetus to support a thoroughly ‘sceptical’ position on the
accuracy of sensory perceptions. What Plutarch is concerned with
countering here, however, is the Stoic and Epicurean gibe of apraxia
levelled, ‘like the raising of a Gorgon’s head’ (1122B), against
Arcesilaus’ position. There is no reason, he argues, to suppose that
impulse (hormē), without which action could not take place, requires
‘assent’ (sunkatathesis), in anything like the Stoic sense:

The soul has three motive forces (kinēmata): sensation, impulse, and
assent. Now the movement of sensation cannot be eliminated, even if we
wanted to; instead, upon encountering an object, we necessarily receive
an imprint and are affected. Impulse, aroused by sensation, moves us in
the shape of actions directed to suitable goals (pros ta oikeia)18: a kind of
casting weight (rhopē) has been put in the scale of our governing part
(hēgemonikon), and a directed movement (neusis) is set afoot. So those
who suspend judgement about everything do not eliminate this second
movement either, but follow their impulse, which leads them to the
suitable apparent object (pros to phainomenon oikeion). Then what is the
only thing they avoid? That only in which falsity and error can arise,
namely forming an opinion (doxazein) and thus falling rashly into assent
(sunkatathesis), although such an assent is a yielding to appearance that
 is due to weakness and is of no use whatever.19 For two things are
requisite for action (praxis): a presentation (phantasia) of something
suitable, and an impulse towards the suitable object thus presented to
appearance – neither of which conflicts with suspension of judgement
(epochē). For it is opinion (doxa) that the argument relieves us from, not
impulse or sensation. So once some suitable object is perceived, no
opinion is required to set us moving and keep us going in its direction; the
impulse comes directly, and is a movement initiated and pursued by the
soul.

(trans. Einarson-De Lacy, slightly altered)

Plutarch thus appears to identify himself firmly with the position of

Arcesilaus, as interpreted by himself, and to claim that position
essentially for Plato and for Socrates – with the proviso, of course,
that this relates exclusively to the physical world; our knowledge of
the Forms, the contents of the intelligible world, is sourced quite
differently. He proceeds to defend the New Academy for many pages
more, revealing in the process, I would suggest, a good deal of the
contents of his lost treatise, mentioned above, On the Unity of the
Academy from Plato. It is interesting, for instance, that he can
embrace the acceptance of such things as traditional religious
practices and belief in oracles, which would certainly have been
endorsed by Plato and Socrates (despite their attacks on unsuitable
popular beliefs about the gods), within the ambit of Academic
scepticism (1125Dff.). His rationale for this presumably is that such
phenomena as religious practices, since they are endorsed by
countless generations of men, fulfil the (Carneadic) condition of
being ‘thoroughly investigated (diexōdeumenē)’ and thus merit the
highest level of assent to their plausibility. At any rate, on religion he
comes out with the following (1125DE):

In your travels you may come upon cities without walls, writing, king,
houses or property, doing without currency, having no notion of a theatre
or gymnasium; but a city without holy places and gods, without any
observance of prayers, oaths, oracles, sacrifices for blessings received or
rites to avert evil, no traveller has seen or will ever see. No, I think a city
might rather be formed without the ground it stands on than a
government, once you remove all religion from under it, get itself
established or, once established, survive.

Oddly, he here concords with the Stoic argument advanced by

Chrysippus for the existence of gods, from the observation that all
known nations and tribes believe in them, but the argument works
just as well within the framework of Carneadic levels of plausibility.
 3
We may observe here, then, the lineaments of Plutarchan Platonism
that incorporates a moderate, or modulated, degree of scepticism,
embracing the tradition of the New Academy. This sceptical
tendency seems primarily to emerge, it must be said, in the context
of combating the Stoics and Epicureans, but it can surface in
interesting ways within Plutarch’s non-polemical works. I propose to
end this essay with a study of a rather curious and entertaining work,
Plutarch’s essay On the Principle of Cold.20
This short treatise, dedicated to his pupil Favorinus of Arles,
who himself professed to be a follower of the Sceptical Academy,21
is a somewhat enigmatic document, but gains considerably in stature
and interest, I think, if one regards it as essentially a light-hearted jeu
d’esprit, designed both to salute and (in a friendly spirit) to tease
Favorinus.
The topic is broached, initially, as an (Aristotelian-style) aporia
or problēma: ‘Is there an active principle (dunamis) of Cold, as there
is of Heat (in the form of fire), through the presence (parousia) of
which, and through participation (metochē) in which, everything else
becomes cold? Or is coldness rather a negation (sterēsis) of warmth,
as they say darkness is of light, and rest of motion?’
Now this touches on the much-discussed question within the
Platonic tradition – of which there is an extended account, for
example, in Proclus’ Commentary on the Parmenides (Book III, cols.
815–33) – namely, of what things there are Forms, though Proclus
does not there broach the question of Forms of opposites. The
existence of Forms of Evils is dismissed, as are those of Forms of
artificial objects and of individuals, but there is no discussion of such
opposites as black and white, light and dark, or hot and cold. So
Plutarch’s postulation of a Form of Cold is not necessarily
unorthodox.22
However, my concern is rather with the manner in which he
advances his various solutions to the problēma, since they are
couched in terms of increasing levels of pithanotēs. First (946A),
there is a brief listing of arguments in favour of cold being simply a
sterēsis; but these are quickly countered (946B–8A) by a series of
considerations in favour of its being a positive dunamis of some sort.
I give, as a sample, the beginning of Plutarch’s counterattack
(946BC):

First of all, must we not be wary of one point in this argument (sc. that
cold is simply the sterēsis of heat)? It eliminates many obvious forces
(dunameis) by considering them not to be qualities or properties, but
merely the negation of qualities or properties, weight being the negation
of lightness and hardness that of softness, black that of white, and bitter
that of sweet, and so in any other case where there is a natural
opposition of forces rather than a relation of positive and negative.
Another point is that all negation is inert and unproductive: blindness, for
instance, and deafness, silence or death. Here you have defections of
definite forms (eidē) and the annihilation of realities (ousiai), not things
that are of themselves natures (phuseis) or realities. It is the nature of
coldness, however, to produce affects and alterations (pathē kai
metabolai) in bodies that it enters no less than those caused by heat.
Many objects can be frozen solid, or become condensed, or made
viscous, by cold. Moreover, the property whereby coldness promotes rest
 and resists motion is not inert, but acts by pressure and resistance, being
endowed with a constrictive and preservative tension (tonos).

(trans. Cherniss and Helmbold, slightly emended)

We see here counterarguments being brought to bear against an

initial assertion, ‘cold is a mere absence of heat’, very much in the
manner of a Socratic elenchus. This position continues to be
established, with further plausible arguments, over the next few
chapters (3–7), until, in Chapter 8 (948AB), we are in a position to
advance to the next stage: if cold is a positive dunamis, while being
the opposite of heat, which is a dunamis of the element Fire, then to
which of the other three elements might it best be related?
This initiates a three-part enquiry, which once again could be
reconstructed as a Socratic dialogue – one thinks of something like
the successive candidates for knowledge in the Theaetetus. Plutarch
proposes to take in turn the three other elements – Air, Water and
Earth – each of which might be opposed, with varying degrees of
plausibility, to Fire, and examine their claims to be the principle of
cold. At the outset, in Chapter 9, he identifies the proponents of the
candidature of Air as the Stoics (thus putting them nicely in their
place as the proponents of the least plausible alternative), those of
Water as the rather motley combination of the Presocratic sage
Empedocles and the Peripatetic Strato of Lampsacus, while he
leaves Earth, as the third candidate, for the moment without a
champion.
All three candidates, of course, have something to be said for
them, but then again, the first two have much to be said against
them. The case of Air is proposed, with various persuasive
arguments, in Chapters 9–12, rounded off by this significant address
to Favorinus (949F):

So now, Favorinus, the argument that attributes the primal force of cold to
the air depends on such plausibilities (pithanotētes) as these.

We are then directed to Water, the case for which is set out in
Chapters 13–16, with many plausible arguments, at the end of which
Favorinus is once again invited to weigh up the probabilities, while
Plutarch moves him gently on to his final alternative (952CD):

Now you must pursue the subject by setting these arguments against
their predecessors (skopei dē kai tauta paraballōn ekeinois). For
Chrysippus, thinking that the air is primordially cold because it is also
dark, merely mentioned those who affirm that water is at a greater
distance from the aether than is air; and wishing to make them some
answer, he said, ‘If so, we might as well declare that even earth is
primordially cold because it is at the greatest distance from the aether’ –
tossing off this argument as if it were utterly inadmissible and absurd. But
I have a mind to maintain that earth too is not destitute of probable and
convincing arguments (eikotes kai pithanoi [logoi]).

Plutarch here conjoins, significantly, Plato’s key term for the account
of physical reality in the Timaeus (eikōs logos) with the favourite
term of Carneades’ epistemology. He then goes on to appropriate
Chrysippus’ identification of Air as dark and cold with greater
plausibility for Earth (952D), and for the next six chapters (17–22)
proceeds to develop a series of arguments in favour of the essential
coldness of Earth, ending with the following thoroughly Platonist
flourish:

We must, therefore, believe that the reason why the wise and learned
men of old held that there is no mingling between earthly and celestial
reality was not that they distinguished up and down by relative position,
as we do in the case of scales; but rather it was the difference in powers
that led them to assign such things as are hot and bright, swift and
buoyant, to the immortal and eternal nature, while darkness and cold and
slowness they considered the unhappy heritage of perishable and
submerged beings (phthitoi kai eneroi)23 Then too, the body of a living
creature, as long as it breathes and flourishes, does, as the poets say,
enjoy both warmth and life; but when these forsake it and it is abandoned
in the realm of earth alone, immediately frigidity and congelation seize
upon it, since warmth naturally resides in anything else rather than the
earthy.
(trans. Cherniss and Helmbold, slightly emended)

Having thus laid out in order the arguments in favour of Air, Water
and Earth, with a certain bias towards the claims of Earth, Plutarch
ends with what seems to me to be a teasing and light-hearted
flourish directed at his former pupil, recognising his championing of
the sceptical tradition of the New Academy and encouraging him to
stick to it (955A):

Compare these arguments, Favorinus, with the pronouncements of

others; and if they neither fall short or much exceed them in plausibility,
then say farewell to opinions (doxai), being convinced as you are that it is
more philosophic to suspend judgement (epechein) in face of things that
are unclear (adēla) than to grant assent (sunkatatithesthai) to them.
What we seem to have here, then, is a dialectical fugue played by
Plutarch for the delectation of Favorinus, and presumably any other
sympathetic readers who might come upon it, on a theme
concerning a representative problem related to the physical world,
showing how a Platonist with loyalties both to the New Academy and
to the Old might approach it. I choose it, not because it seems to me
to embody Plutarch’s most serious philosophising,24 but rather
because it demonstrates rather well his dexterity in manipulating the
full extent of the tradition which he saw himself as inheriting. He
would be prepared to accept that aporia lies at the core of true
philosophy, but he thinks this view is compatible with dogmatism as
regards first principles and with the acceptance of a realm of true
being superior, and largely antithetical, to the sublunary, material
world. This, I believe, he felt was also true of the chief figures within
the New Academic tradition, Arcesilaus and Carneades.

1 Opsomer 1998. I shall be much indebted to this excellent study in what

follows. A useful discussion also can be found in Brittain 2007.

2 I have been fairly reprimanded by Daniel Babut 2007: 67 n.17 for in the
past rather dismissing Plutarch’s New-Academic sympathies, as follows
(Dillon 1999, 305): ‘As for the New Academy, despite his retention of some
Academic sceptic traits as weapons against the Stoa, he reveals no affinity
for such figures as Arcesilaus and Carneades.’ I would certainly modify that
view now, upon more mature consideration, without wishing to retract it
entirely!

3 The distinction between the Platonic and Aristotelian uses of aporia is

nicely brought out by Gareth Matthews 1999a, in his penultimate chapter
(11), pp. 109ff. For Aristotle, in general, an aporia is not an occasion for
existential perplexity, but rather a particular problem to be solved by further
research.
4 Ou gar eirōneuomenos ge kai paizōn prosechrēsat’ an tō[i] tou theou
onomati – with a reference to Alcibiades’ remarks at Symp. 216e4–5.

5 Modern commentators, at least in the English-speaking world, are

strangely unconcerned about the meaning of this remark. F. M. Cornford
1939, John McDowell 1973 and Myles Burnyeat 1990, make no mention of
it. Only Lewis Campbell in his 1883 edition offers a rather coy interpretation,
supposing that Socrates is referring to ‘the presiding genius of his Art’ – by
which he may mean his daimonion, but that is not clear. I fear, however, that
I have not made a comprehensive check of editions.

6 A creative application of a dictum from Plato’s Laws, V 731e, where Plato

is exploring the causes of crime, as rooted in selfishness.

7 No relevant proverb seems to be elsewhere attested, but it must, I think,

have concerned the selection of one’s own offspring by parents in some
situation.

8 A reference here to Diotima’s description of the culmination of the ascent

to the Beautiful in the Symposium (210e–12a), combined with a glancing
allusion to the description of the wisdom of the Guardians in Rep. IV 429a.

9 An interesting employment here of an essentially Stoic concept, the koinai

ennoiai, or ‘common concepts’, but overlaid with a distinctively Platonic
element: the capacity for receiving these noēseis is innate, but their source
is independently existing Platonic Forms. Plutarch is by no means alone
here within the spectrum of Middle Platonism: Alcinous presents a very
similar position in Did, Chapter 4.6.

10 The suggestion that Socrates has a concept of intelligible reality, we may

note, is also made by Plutarch at Adversus Colotem 1114C.

11 Mauro Bonazzi 2003: 219–32, has a useful discussion of Plutarch’s

attitude towards Arcesilaus.

12 The summary of a third one exists: That the Stoics make more
paradoxical utterances than the poets.
13 Colotes of Lampsacus, a favoured pupil of Epicurus, seems to have
composed this work in the 260s BCE.

14 For the fullest account of Carneades’ three stages of pithanotēs, or

‘plausibility’, see Sextus Empiricus, AM VII 159–89. One moves from a basic
pithanē phantasia to one that ‘plausible and uncontradicted (pithanē kai
adiaspastos), and ultimately to one that is all that, but also ‘thoroughly
checked-out (diexōduemenē)’. On the basis of this last, one can act with
virtual certainty, without, however, conceding anything to the Stoics.

15 An alternative would be that he is adopting the position of Philo of

Larissa, the pupil of his pupil Metrodorus, and last head of the Sceptical
Academy. I do not rule this out, but I feel that Carneades’ position will do
very well.

16 He has just mentioned the dogma of Epicurus that ‘no one but the sage
(sophos) is unalterably convinced of anything’ (1117F–Fr. 222 Usener).

17 Augustine, in his Contra Academicos (3. 20. 43) attributes this rumour to
Cicero, in a lost part of his Academica (Fr. 210 Plasberg), which puts it back
to the early first century BCE. It is hard to see who Cicero could have picked
this up from other than Philo.

18 If Plutarch is reporting Arcesilaus accurately here, he would appear to be

using a Stoic technical term for his own purposes. The term oikeion, which
may be rendered ‘suitable’ or ‘akin’ to us, occurs repeatedly throughout this
passage, as an equivalent for ‘good’ (agathon).

19 Again, Plutarch/Arcesilaus would appear to be turning Stoic terminology

against themselves: for Zeno (SVF I 67–9), opinion is ‘weak’ or ‘false’
assent; for Arcesilaus here, all assent is a result of ‘weakness’ (astheneia).

20 In this I am much indebted to the most useful discussion in Boys-Stones

1997b.

21 Favorinus is an interesting figure in many ways, which need not be

covered here. Jan Opsomer 1993, devotes a good discussion to him, in
which he shows that, contrary to the intemperate allegations of Galen in his
treatise De Optima Doctrina, Favorinus was actually most probably a sceptic
of the Carneadic persuasion. Cf. also Charles Brittain 2007, who provides a
good discussion of Plutarch and of Numenius; and Mauro Bonazzi 2003:
158–70.

22 To be fair to Plutarch, he nowhere speaks of cold as being a Form, but

rather a dunamis and an ousia, which is compatible with its being merely an
essential quality of one of the elemental Forms.

23 A suggestion here, perhaps, with this latter term (which denotes properly
the ghosts of the dead or gods of the underworld) of the notion that the
sublunary world is the Hades of the poets, as Plutarch was on occasion
prepared to maintain (cf. De gen. Socr. 591AB).

24 In fact, as is argued persuasively by George Boys-Stones 1997b; 237,

Plutarch probably feels that he has made a good case for Earth, his
arguments being based on Plato’s theory of the basic triangles and
elemental bodies as laid out in Tim. 55d–6b. I feel, therefore, that this is an
aspect of the ironic and teasing quality of this treatise.
 Chapter 11
Aporia and Enquiry in Ancient
Pyrrhonism
◈

Luca Castagnoli
 Setting the Stage: Plato and Aristotle on
Aporia and Enquiry
When considering the relationship between aporia and philosophical
enquiry, one might think of aporia as the negative end point of an
unsuccessful enquiry, as a token of philosophical failure, or at least
as a stumbling block towards success. The two towering figures of
classical Greek philosophy, Plato and Aristotle,1 had far more
nuanced views on aporia, and its relationship with enquiry, which
could be summarised, roughly, as follows:

Although aporia can be the outcome of a certain enquiry, it is

not the definitive end of that enquiry.

Aporia is in fact the very origin of all enquiry, and thus of

philosophy itself, both for human beings collectively, and for
individuals.2

But aporia is not a mere preliminary to enquiry; the

observations, questions, reasons, arguments which are the
source of aporia, as a psychological state, or constitute a
certain aporia, intended as a particular puzzle or conundrum,3
shape and direct (further) philosophical enquiry.4

For these reasons, aporia is ultimately beneficial to us,

although it can be psychologically unsettling and painful.5

Since antiquity interpreters have described Pyrrhonism as a

philosophical approach somehow ‘alien’, radically different in its
motivations, nature, and conception of the philosophical endeavour
from all other ancient philosophies,6 those philosophies that the
Pyrrhonists lumped together as ‘dogmatic’. Of course scholars have
recognised that Pyrrhonism entertains an intimate dialectical rapport
with all dogmatic philosophies, from which it borrows,
noncommittally, concepts, terminology, arguments and
argumentative patterns to be retorted against dogmatism itself. But
there has been less sensitivity towards the ways in which, in their
self-characterisation, the Pyrrhonists adopted some traditional tropes
on the motivations, nature and aims of the philosophical enterprise.7
I will argue that the Pyrrhonian understanding of aporia and
enquiry, and especially of their interconnection, was no less nuanced
than Plato’s and Aristotle’s, and strikingly reminiscent of theirs. I will
also argue that the unique way in which those notions are interlinked
at the heart of Pyrrhonian philosophy raises, however, a unique set
of philosophical questions. I will spell out some of those questions,
and suggest how the Pyrrhonists could have attempted to answer
them. Sextus Empiricus’ brand of Pyrrhonism will be my main focus,
but I will make some brief references to earlier versions of
Pyrrhonism.
 The Pyrrhonists: ‘Enquirers’ and ‘Aporetics’
Two of the five ancient labels for the Pyrrhonian agōgē (‘way’),
skeptikē (‘examining’) and zētētikē (‘enquiring’), reflect the idea,
stressed by Sextus from the opening of his Outlines of Pyrrhonism,
that the Pyrrhonists are devoted to, and indeed defined by, an
ongoing activity of philosophical enquiry.8 Whereas the Dogmatists
are no longer enquiring because of their misguided belief that they
have already discovered the truth, and the Academics because of
their conclusion that finding the truth is impossible, the Sceptics, the
‘Enquirers’, suspend their judgement about the matter under
investigation and continue their zētēsis (PH 1.1–3).
But Pyrrhonism is also ‘aporetic’ (aporētikē):

T1 The sceptical way, then, is also called ‘enquiring’, from its activity in
enquiring and examining (zētētikē apo energeias tēs kata to zētein kai
skeptesthai); ‘suspensive’, from the affection that comes about in the
sceptic after the enquiry; ‘aporetic’, either from raising puzzles over and
enquiring into everything, as some say, or from having no means to either
assent or deny (kai aporētikē ētoi apo tou peri pantos aporein kai zētein,
ēōs enioi phasin, ē apo tou amēchanein pros sugkatathesin ē arnēsin);
and ‘Pyrrhonian’, from the fact that Pyrrho appears to us to have attached
himself to scepticism more systematically and conspicuously than those
before him.

(PH 1.7)9

I suggest that Sextus introduces here two senses of the verb aporein
related to the two possible uses of the noun aporia clearly attested
since Plato: the Pyrrhonists are aporetic both in the sense that (1)
they raise aporiai, i.e. difficulties or puzzles,10 about everything, or in
the sense that (2) they are in the psychological state of aporia
themselves, namely they feel resource-less (amēchanein) in
establishing what is true and what is false, unable to give their
assent to or deny anything. The disjunction need not be exclusive,
and the qualification hōs enioi phasin (‘as some say’) does not
indicate that Sextus is distancing himself from the first sense of
aporein11 and favouring the second, as confirmed by Sextus’ own
usage of the aporia vocabulary, which I will examine below.12
Although Sextus uses the labels ‘Sceptics’ and ‘Pyrrhonists’ much
more frequently, he refers to the ‘Aporetics’ several times in M (8.75,
78, 80, 99, 160, 278; 9.207, 303; 10.67, 68, 105, 246; 340; cf. also
7.30: ton aporētikōs philosophounta).13
The meaning of ‘aporetic’ is explained less transparently by
Diogenes Laertius:

T2 All these men were called ‘Pyrrhonists’ after their teacher, ‘aporetics’,
‘sceptics’ and even ‘suspensive’ and ‘enquirers’ after their doctrine, as it
were. … ‘aporetic’ from the fact that both the dogmatists and they
themselves are in aporia (aporētikē d’ apo tou tous dogmatikous aporein
kai autous de).14
(D.L. 9.69–70)

The MSS text appears problematic. Since the dogmatists do not

raise aporiai (nor would their rasing aporiai be a reason for the
Pyrrhonists to be called ‘aporetic’),15 the sentence could be taken,
loosely and charitably, to mean something like ‘from the fact that as
the result of it [sc. Pyrrhonian philosophy] both the dogmatists and
the Pyrrhonists themselves end up in aporia’. Alternatively, we could
postulate here a transitive use of the verb aporein with the meaning
of ‘to make [someone] fall into aporia’: ‘from the fact that they make
both the dogmatists and themselves fall into aporia’.16 Although such
a use is not recorded by ancient Greek dictionaries and thesauri, it
seems to occur at M 10.58: ‘it is possible to make fall into aporia also
in other ways those who thus define the concept of locomotion’
(paresti de kai heterōs aporein tous houtō tēn epinoian tēs
metabatikēs kinēseōs apodidontas). Several editors and
commentators have considered the MSS text corrupt, and intervened
in various ways.17 I suggest that, if we opt to emend the MSS text, a
more plausible conjecture is that poiein fell before or after aporein:18
‘“aporetic” from the fact that it makes both the dogmatists and
themselves [sc. the Pyrrhonists] be in aporia’.19
 Raising Aporiai in Sextus Empiricus
In this section I provide a selective survey of Sextus’ usage of the
language of aporia to refer to philosophical problems or puzzles, or
to the activity of raising them. Sextus typically employs the verb
diaporein for the Pyrrhonian procedure of going systematically
through the difficulties concerning (peri or en) a certain dogmatic
subject, or through the arguments against a certain dogmatic
position.20 Consider the following examples:

T3 Just as we offer the modes of suspension of judgement, so also some

set out modes in accordance with which we bring the Dogmatists to a halt
by raising aporiai concerning their particular causal explanations (en tais
kata meros aitiologias diaporountes).
(PH 1.180)

T4 Lest the Dogmatists should try to slander us because they are in

aporia as to how to produce substantial counter arguments, we shall
raise more general aporiai about active causes (koinoteron peri tou
energētikou aitiou diaporēsomen), having first tried to focus on the
concept of cause.
(PH 3.13)

Notice that in T4 the dogmatists are said to be in a state of aporia as

the result of the aporiai previously presented by the Sceptics about
the gods as active causes; the Sceptics will now raise a battery of
more general aporiai about causes, clearly with the purpose of
strengthening that state of aporia.21
 The adjective aporos is frequently used to indicate that
something (often a dogmatic concept or logos) is an object of aporia,
typically in the sense that (unsolved or unsoluble) aporiai have been,
or can be, raised about it.22 The passive form of the verb aporein
and its compounds can also be used in this way.23
As for the noun aporia, Sextus often uses it broadly to refer to
any argument or difficulty that can be raised against a certain
dogmatic doctrine or argument,24 but sometimes more narrowly, to
denote some specific philosophical puzzle (e.g. sōritikēn aporian, the
sorites [PH 3.80 and M 1.68, 70, 84]) or mode of suspension of
judgement.25 In particular, aporia is frequently associated (e.g. PH
2.9, 20, 183, 197, 199; 3.22, 242) with the fifth mode of Agrippa,
‘reciprocity’ or ‘circularity’ (ho diallēlos tropos), which Sextus calls
aporōtatos (‘most aporetic’) at M 8.445 and 9.47 and aporōteron
(‘more aporetic’) at 8.379:26

T5 The reciprocal mode occurs when what ought to be confirmatory of

the object under enquiry needs to be made convincing by the object
under enquiry; then, being unable to take either to establish the other, we
suspend judgement about both.

(PH 1.169)

We have seen that in T1 Sextus describes the condition of aporia as

a kind of epistemic ‘resourcelessness’. This is reminiscent of one of
the two fundamental nuances that the term aporia had in its
everyday use, namely the lack or scarcity of supplies or provisions,
of means or resources to do something (typically economic poverty
or need).27 I suggest that Sextus’ use of the language of aporia to
denote arguments or puzzles shows traces of the other original
nuance of the term, the ‘topological’ one: an a-poria is an ‘im-passe’,
literally the absence of a poros or poreia, of a way forward, in, out or
through something. For the language of aporia occurs time and
again in the presence of the typical Sextan dilemmatic, trilemmatic or
quadrilemmatic arguments in which all possible options are spelled
out, and discarded one by one.28 Consider this example:

T6 And teachers and learners … are objects of aporia (aporountai) also

on their own. Either the expert teaches the expert or the non-expert the
non-expert or the non-expert the expert or the expert the non-expert. Now
the expert does not teach the expert: for insofar as each is an expert
neither of them needs to learn. Nor does the non-expert teach the non-
expert – any more than the blind can lead the blind. Nor does the non-
expert teach the expert; for that would be ridiculous. It remains to say that
the expert teaches the non-expert; but this too is impossible. … But if
neither the expert teaches the expert nor the non-expert the non-expert
nor the non-expert the expert nor the expert the non-expert, and there is
no other option apart from these, then there are no teachers and no one
is taught.

(PH 3.259–65)

Whereas in the ‘reciprocal mode’ of Agrippa there is no way out of

the circle, T6’s quadrilemma produces an impasse because all
routes are blocked,29 and there is no way forward. Finally, aporia
also occurs when, in the presence of multiple routes, one has no
reason whatsoever to prefer one over the others (compare the
notorious Buridan’s ass). In cases of diaphōnia, ‘dispute’ or
‘disagreement’, for example, in the absence of reliable criteria of
truth, signs or demonstrative procedures,30 it becomes impossible to
adopt any of the conflicting positions because they all have the same
degree of credibility, and not because they are all equally unviable:31

T7 As for those who say that good things cannot be lost, we shall bring
them to suspension of judgement as a result of the aporia arising from
the dispute (ek tēs aporias tēs kata tēn diaphōnian).
(PH 3.238)

To summarise, Sextus’ aporiai often take one of the following forms:

Figure 11.1 Sextus’ aporiai

 Psychological Aporia in Sextus Empiricus and
the Earlier Pyrrhonian Tradition
In the previous section I surveyed Sextus’ use of the aporia
vocabulary to refer to philosophical difficulties, their objects, or the
activity of raising them. Sextus also uses the language of aporia,
less frequently, to denote a mental state.32 It is on the psychology of
aporia, so to speak, that I will focus in the rest of this essay. Sextus’
description of the condition of aporia in T1 is clearly related to his
definitions of suspension of judgement as a ‘standstill of the mind,
owing to which we neither deny nor affirm anything’ (PH 1.10:
epochē de esti stasis dianoia di’ hēn oute airomen ti oute tithemen)33
and of ‘non-assertion’ as the ‘affection owing to which we neither
affirm nor deny something’ (PH 1.192: aphasian pathos hēmeteron
di’ ho oute tithenai ti oute anairein phamen).34 Whereas both the
verb epechō and the noun aphasia appear in the list of the sceptical
‘formulae’ (phōnai) whose meaning is clarified at PH 1.187–205,
neither the verb aporō nor the noun aporia occur. This might confirm
that the use of the aporia vocabulary to refer to the sceptical frame of
mind was no longer prevalent at Sextus’ time. But it could equally
indicate that the Pyrrhonian usage was uncontroversial: after all,
Sextus’ aim in that section is to explain why certain formulae are not
expressions of dogmatic belief, and there is no reason to fear that
avowals of aporia might be misinterpreted as statements of belief.
To be in aporia is to be unable to either accept or reject a certain
proposition as true or false. But what is its cause? According to
Sextus aporia is the result of the apparent isostheneia,
‘equipollence’, i.e. equal strength (or weakness), of opposed
perceptions, reasons and arguments35 – as we have seen, some
such reasons and arguments are sometimes described as aporiai.
The aporetic condition will result, in turn, in epochē and aphasia.36
The existence of such an intimate connection between equipollence,
aporia and suspension of judgement confirms that, as some modern
interpreters have emphasised, Pyrrhonian aporia is a mental state
different from that of doubt, at least if ‘doubt’ is consistent with
(uncertain) belief – in this respect, the Pyrrhonists are firmly within
the ancient tradition of thinking about aporia, and differ radically from
Cartesian forms of sceptical doubting.37
More typically in the corpus, including in his initial account of
scepticism at PH 1.8, Sextus presents epochē as the direct effect of
equipollence: one’s inability to give one’s assent, i.e. one’s aporia,
and the fact that one ends up not giving one’s assent are often
conflated in the single concept of epochē, and the former is typically
omitted.38
Judging by Photius’ summary of Aenesidemus’ Arguments of
the Pyrrhonists, the language of aporia was more central in the
founder of the Pyrrhonian school in the first century BC.39 The label
aporētikoi is used to refer to the Pyrrhonists, at a time in which
skeptikoi was most probably not yet in currency (the first attested
occurrence is in Favorinus,40 first through second century AD):41

T8 The Academics are dogmatists: they affirm some things with

confidence and deny others unambiguously. The Pyrrhonists, on the
 other hand, are aporetic (aporētikoi) and free of all belief.

(Phot. Bibl. 212.169b38–41)

On the basis of our scant evidence we cannot trace the adoption of

the language of aporia back to Pyrrho42 or Timon; I suggest that
Aenesidemus’ appropriation of the originally Socratic/Platonic
language had a polemical thrust.43 The Academics, while debating
on whether Socrates and Plato had been aporetic,44 were guilty,
according to Aenesidemus, of not being consistently aporetic
themselves, in fact reverting back to dogmatism (resembling,
according to a well-known Aenesidemean verdict, ‘Stoics fighting
against Stoics’ [170a16–17]). I believe that this anti-Academic
polemic45 should also be connected to Aenesidemus’ strikingly
Socratic (cf. Pl. Apol. 21b4–5) portrayal of Pyrrho himself as
someone possessing the ‘wisdom’ of knowing that he lacks
knowledge:

T9 Neither the Pyrrhonists nor the others know the truth in things; but the
philosophers of other schools, as well as being ignorant in general, and
wearing themselves out uselessly and expending themselves in
ceaseless torments, are also ignorant of the very fact that they have
apprehended none of the things which they think they have apprehended.
But he who philosophises after the fashion of Pyrrho is happy not only in
general, but also, and especially, in the wisdom of knowing that he has
not apprehended anything firmly.
(Bibl. 212.169b21–9; transl. Long and Sedley, slightly revised)

Aporia was identified by Aenesidemus as an intellectual virtue, along

the lines of Plato’s account of Socrates’ ‘human wisdom’. I am not
suggesting any direct influence of Plato’s Socrates, or Plato’s texts,
over and above the appropriation of the aporia language46 when
Aenesidemus positioned his Pyrrhonism against contemporary
Academics. We should not forget, however, that Aenesidemus could
have been himself a member of the Academy before seceding and
founding his school.47
I can only briefly mention here the view that Aenesidemus’
Pyrrhonism was a brand of ‘aporetic’ scepticism radically different
from Sextus’ later ‘zetetic’ or ‘enquiring’ Pyrrhonism, and influenced
by Platonic principles. According to Woodruff (1988, 2010),
Aenesidemus’ Pyrrhonism was ‘aporetic’, that is purely negative and
refutative: it drew the negative conclusion that things are by nature
none of the ways they variably manifest themselves to us, on the
basis of the ultimately Platonic metaphysical principle that something
is F by nature only if it is F invariably and unqualifiedly.48 Apart from
other problems which this reading incurs,49 Woodruff’s
understanding of aporia, and his straightforward equation of
‘aporetic’ and ‘refutative’, appear problematic. Whereas the
psychological state of aporia can be the result of multiple refutations,
and some Pyrrhonian aporiai are arguments concluding the
falsehood of a certain position, the concepts of aporia, on the one
hand, and refutation as the demonstration of a negative conclusion,
on the other, should not be conflated. This is true not only of the
Platonic and Aristotelian tradition but, as I have argued, of Sextus
Empiricus too.50 As for the Platonic influence, Sextus Empiricus
gave a negative answer to the question of whether Plato himself was
‘aporetic’ or ‘sceptic’51 in a difficult passage which might suggest
Aenesidemus had done the same:

T10 As for Plato, some have said that he is dogmatic, others aporetic,
others partly aporetic and partly dogmatic … As to whether he is purely
sceptical, we deal with this at some length in our Commentaries. Here, in
an outline, we say, like Menodotus and Aenesidemus,52 who have been
the main proponents of this position, that when Plato makes assertions
about Forms or about the existence of providence or about a virtuous life
being preferable to a life of vice, then if he assents to these things as
really being so, he is dogmatic, and if he commits himself to them as
being more plausible, he has abandoned the distinctive character of
scepticism … And even if he makes some utterances sceptically when,
as they say, he is exercising, this will not make of him a sceptic.
(PH 1.221–3)

This rejection of Plato’s sceptical pedigree is consistent, of course,

with my suggestion that Plato’s portrayal of Socrates might have
influenced, if only indirectly, Aenesidemus’ adoption of the language
of aporia.53
 Aporia, Enquiry and Ataraxia
Aenesidemus’ reference to Pyrrho’s happiness in T9, as opposed to
the ‘torments’ of the other philosophers, brings us back to Sextus’
characterisation of the effects of aporia and epochē in human life. As
we have seen, aporia is strictly related to epochē; and a mental state
of psychological tranquillity, ataraxia, is supposed to follow epochē
(cf. e.g. PH 1.8 and T12 below). But the suggestion that aporia leads
to a desirable condition of ataraxia flies in the face of the common-
sense perception of aporia reflected in the ordinary connotations of
the Greek term: no one wants be in a state of impasse, stuck
somewhere, with no way out or forward; no one wants to feel
resourceless, whether materially or intellectually.54 Such conditions
are far more likely to engender distress and anxiety than tranquillity.
Plato had recognised that, however good it may be for our
souls, aporia can be a rather uncomfortable, and sometimes
distressing, experience. The numbing psychological effects of aporia
are famously depicted through the stingray analogy at Men. 79e7–
80d4; aporia is compared to labour pains at Tht. 151a5–b1; aporia is
described as a sort of vertigo or dizziness at Lysis 216c and
Theaetetus 175d, or as the sea-sickness of people ‘storm-tossed by
argument’ at Laches 194c and Theaetetus 191a; Socrates’
interlocutors are often ashamed of their own aporia (cf. e.g. Euthd.
275d5–6; Chrm. 169c–d), which easily becomes the object of
laughter (Theaetetus 175d), and thus they try to hide it (cf. e.g.
Chrm. 169d1; Laches 196b2, Apol. 23d); a no less frequent reaction
is that of anger towards the person who induced aporia in them (cf.
e.g. Men. 79e–80a). All these conditions and emotional reactions are
as far as one can imagine from Sextus’ ataraxia.
This is a dimension of aporia which was overlooked by Aristotle.
His focus was on the aporiai, the arguments or puzzles which cause
aporia in the mind, and how they should be used in philosophical
enquiry, and not on the contingent psychological aspects of the
pathos of aporia. In his optimistic outline of the genesis of
philosophical enquiry, the human mind did not offer any resistance to
the zetetic drive of aporia and wonder (cf. Metaph. 1.2.982b12–
3a21). Similarly, the ethical dimension of aporia as something
intrinsically valuable, insofar as it can become a first step towards
the purgation of conceit and the attainment of a higher degree of
self-knowledge, so central in Plato, was not part of Aristotle’s
analysis.55 On the contrary, the charge that the dogmatists are
affected by ‘conceit’ (oiēsis) and ‘rashness’ (propeteia) in their
pretension of knowledge is ubiquitous in the Pyrrhonian tradition
(starting as early as Pyrrho and Timon).56
Did Sextus wilfully ignore the pains of aporia? The picture is
actually more complex. When describing the genesis of scepticism,
Sextus explains how certain ‘naturally gifted men’ (megalophueis)
first came to enquiry:

T11 The causal principle of scepticism we say is the hope of becoming

tranquil. Naturally gifted men, troubled by the anomaly in things and in
aporia (aporountes) as to which of them they should rather assent to,
came to enquire what in things is true and what false, thinking that by
deciding these issues they would become tranquil.
 (PH 1.12)

Aporia is connected here to the undesirable psychological condition

of tarachē (tarassomenoi), and it is exactly to cure this condition and
reach the goal of a-taraxia that the would-be Sceptics embarked on
philosophical enquiry. Just as for Plato and Aristotle, according to
Sextus aporia is thus the origin of philosophy (or, at least, of
sceptical philosophy);57 but for Sextus it is the wish to cure the
mental distress associated with aporia (rather than, say, Aristotle’s
human innate and dispassionate desire to know) that prompts
enquiry.58 Sextus’ claim that the soon-to-be Sceptics are
megalophueis suggests that some special philosophical sensitivity to
the variety, discrepancy and contradiction in our experiences
(anōmalia) is necessary to be driven towards enquiry (although of
course everyone experiences conflicting appearances, and not only
the Sceptics, as Sextus himself comments at PH 1.210). What
separates the Sceptics from other philosophers is, as we have seen,
the outcome of the enquiry:

T12 A story told of the painter Apelles applies to the Sceptic. They say
that he was painting a horse and wanted to represent in his picture the
lather on the horse’s mouth; but he was so unsuccessful that he gave up,
took the sponge on which he had been wiping off the colours from his
brush, and flung it at the picture. And when it hit the picture, it produced a
representation of the horse’s lather. Now the Sceptics were hoping to
acquire tranquillity by deciding the anomaly in what appears and is
thought of, and being unable to do this they suspended judgement. But
when they suspended judgement, tranquillity followed as it were
fortuitously (tuchikōs), as a shadow follows a body.
 (PH 1.28–9)59

Despite engaging in systematic enquiry, the Sceptic is still unable to

solve the conflict of experiences, views and arguments which initially
troubled him: he is still in aporia (as we have seen, Sceptics are
‘aporetic’). As the result of his aporia, he thus suspends judgement,
reaching, unexpectedly,60 the very same goal, ataraxia, that he had
envisaged would be attained by ‘deciding the anomaly’ and resolving
his state of aporia. Similarly, Apelles achieved his goal, painting the
horse’s lather (= ataraxia), in an unexpected fashion, not by using
his brush (= deciding the conflict), but by throwing in his sponge (=
suspending judgement) as the result of his persistent inability to
paint the lather with his brush (= aporia).
Sextus’ account raises a host of problems for the consistency of
the Pyrrhonian approach:

(1) how is it possible that the condition of aporia, which was the
source of tarachē before the enquiry started, becomes a source
of ataraxia when the enquiry leads to suspension of judgement?
61 If anything, enquiry-resistant aporia should cause even more
distress, and suspension of judgement, if it corresponds to
Apelles’ throw of the sponge, could be seen as the
manifestation of a high degree of frustration.

Moreover,

(2) if the Pyrrhonist, like Apelles, reaches his goal (however

mysteriously: see (1) above) when he ‘throws in the sponge’,
how can Sextus justify his depiction of the Pyrrhonist as
 zētētikos and skeptikos? Why should the Pyrrhonist keep
enquiring if his present aporetic and suspensive state has
already secured his goal?62 Surely Pyrrhonism must be a
retirement from all enquiry and philosophy?63

A final objection, which cuts even deeper in the flesh of Sextus’

account, is that

(3) for a Pyrrhonist the commitment to continued enquiry should

be not only pointless (see (2) above), but damaging and
perverse, in light of Sextus’ argument that, independently of the
unsettling experience of the conflict of appearances and views,
having beliefs about how things are, and in particular beliefs
about values, is in itself psychologically harmful:64

T13 Those who hold the opinion that things are good or bad by nature
are perpetually troubled. When they lack what they believe to be good,
they take themselves to be persecuted by natural evils and they pursue
what (so they think) is good. And when they have acquired these things,
they experience more troubles; for they are elated beyond reason and
measure, and in fear of change they do anything so as not to lose what
they believe to be good. But those who make no determination about
what is good and bad by nature neither avoid nor pursue anything with
intensity; and hence they are tranquil.

(PH 1.27–8)

In light of this account of the connection between certain kinds of

belief and tarachē,65 not only could a resolution of the aporia via
enquiry never add to the Sceptic’s ataraxia: it would destroy it.
 Although the ancient critics advanced several objections to the
logical coherence and pragmatic possibility and desirability of a
Pyrrhonian outlook,66 to the best of my knowledge none of the three
questions spelled out above are attested in our sources. It is not
surprising, then, that Sextus nowhere engages with them; but those
modern readers who have raised analogous objections have often
doubted that convincing answers would have been available to the
Pyrrhonists anyway.67 I will now sketch a possible line of response,
on Sextus’ behalf, which attempts to reconcile the Pyrrhonist’s
avowal of aporia, desire for ataraxia, and commitment to zētēsis.

(1) When faced with conflicting experiences, views and

arguments concerning a certain matter, the would-be Pyrrhonist,
like any other human being, or at least those human beings
gifted with a certain degree of intellectual sensitivity and
curiosity, is confused, puzzles over the reasons for the
anōmalia, and tries to find out how things really stand. The
intellectual discomfort caused by the realisation that such a
conflict exists and by his prima facie inability to adjudicate it
leads him to enquire for the first time into the matter, with the
goal to discover the truth, and in this way reach peace of mind.
This enquiry will include careful examination and weighing of all
the ‘anomalous’ evidence which originally stirred his mind, but
also gathering and considering additional evidence, views and
arguments, and reflecting on the best means to solve the
impasse. If having thoroughly considered and examined the
subject-matter, he still finds himself in aporia, because all the
conflicting options appear equally strong, or perhaps equally
weak, he will not be able to do anything but suspend judgement
(the presupposition seems to be that you cannot just choose to
believe something if you have no more grounds to accept it than
to reject it). In most cases, suspension of judgement will involve
abandoning previously held beliefs which have turned out not to
be sufficiently grounded. But whereas his initial inability to
decide the matter was for the soon-to-be Pyrrhonist a source of
anxiety, because it was perceived to be the result of some
subjective deficiency and appeared menacing for his existing
beliefs, once a thorough investigation has been undertaken, the
Pyrrhonist, while still in aporia, reaches the peace of mind of
realising that aporia and suspension of judgement appear to be,
provisionally, the ‘right’ position to be in,68 the ‘reasonable’69
result of the zētēsis. This is not an uncomfortable numbness or
paralysis of the mind, but a state of mental equilibrium. (stasis
[PH 1.7];70 arrepsia [PH 1.190])
Compare my account with the situation of someone who, for
the first time, becomes aware that the beliefs concerning God
with which he has been raised are open to all kinds of
objections, and that many others have very different views about
God, and yet others do not even believe that God exists. This
person will feel confused, and will feel at a loss when he realises
that he cannot offer, to others and to himself, sufficient grounds
to accept his beliefs and reject the conflicting ones. He will begin
enquiring into the matter, to find out the truth about God. But
suppose that the more he investigates the matter, the more it
appears to him that there are no grounds for preferring any of
the competing views as more convincing, including his. He will
end up in a state of suspension of judgement as to whether God
exists, and if it exists, as to what kind of entity it is, and what
attitude human beings should have towards it. In other words,
he will adopt a form of agnosticism as the ‘right’ stance.71 But
this stance will bring him a peace of mind which the initial and
partial realisation of the threatening conflict could never have
brought. When it results from systematic enquiry, aporia will be
consistent with, and indeed conducive to, ataraxia.

(2) Is this state of mind consistent with commitment to the

continuation of enquiry, however? Ancient critics levelled against
the Pyrrhonists Meno-paradox-style charges to the effect that
someone who is in aporia and suspends judgement about X
must lack a grasp of X, and thus cannot even enquire about X,
undermining the Pyrrhonists’ self-depiction as Enquirers and
Sceptics. Sextus’ presentation of, and defence against, the Stoic
(PH 2.1–11) and Epicurean (M 7.337–6a) versions of the charge
have been the object of in-depth analysis in recent years,72 and
I will not discuss them here.
It was on the desirability, rather than the possibility, of
Pyrrhonian enquiry within the framework of aporia and epochē
as a tranquillity-inducing condition that my second question
focused. If, as Sextus claims, ataraxia was the goal all along,
and enquiry had been adopted by the Pyrrhonist merely a
means to it, further enquiry will be pointless at best (but actually
harmful: see question (3) above) since aporia and epochē have
already secured the prize. Against this objection, we should
reconsider the question of what counts as a zētēsis for Sextus.
The Pyrrhonist avows that, up to now, nothing has been able to
resolve his state of aporia about those matters he has
investigated; he never claims (unlike the Academics, according
to Sextus) that it is impossible that this should happen, or that all
matters have now been fully and conclusively investigated. Only
in the latter case would further enquiry be impossible. But the
continuing zētēsis of the mature Pyrrhonist need not be an
attempt to break the impasse; it can be an enquiry into a certain
matter which consists in further examination of all reasons pro
and contra any possible judgement concerning it, or an enquiry
into some matter which has not been (sufficiently) examined
before. In other words, the zētēsis of the mature Pyrrhonist
might be nothing but the continuous exercise of that ability of
setting out equipollent oppositions leading to epochē and
ataraxia that Sextus presents as the definition of the ‘sceptical
way’ at PH 1.8.73 This might explain why some reports (e.g.
D. L. 9.107) make of suspension of judgement – to which
ataraxia follows ‘like a shadow’ – the Pyrrhonian telos: for the
mature Sceptic the two have so far proved to coincide.74
The first part of Sextus’ explanation of the meaning of
aporētikē in T1 can be called upon to support this reading:
scepticism is called ‘aporetic’ because it ‘raises aporiai over and
investigates everything’, apo tou peri pantos aporein kai zētein,
where the kai can be taken to have epexegetic force. To enquire
into something is, for a Pyrrhonist, to puzzle over it.75 Consider
also the way in which the concepts of aporia and enquiry are
associated in the following passage:

T14 And if everything is clear there will be no such thing as

enquiring and raising aporiai (to zētein kai aporein) about
something. For a person enquires and raises aporiai (zētei gar tis
kai aporei) about a matter that is unclear to him, not about what is
apparent. But it is absurd to do away with enquiry and aporia
(zētēsin kai aporian); therefore not every appearance is true, and
not everything is true.

(M 7.393)76

Is it disingenuous on Sextus’ part to present the activity which

consists in raising and examining aporiai as a zētēsis?77 If we
keep in mind the central role that according to Plato and
Aristotle aporia plays in the origin, development and structuring
of philosophical enquiry, it would be more surprising if Sextus
had not thought of the Pyrrhonian aporetic activity as an
enquiry.78 But – one might ask – is such an enquiry an enquiry
after the truth, if the goal of the Pyrrhonian skepsis is no longer
the discovery of truth? To be sure, Sextus is careful not to say
that the mature Pyrrhonist is still enquiring into the truth in the
opening of PH and in T1 (unlike Diogenes in T2).79 But the
stress on truth as the only goal of a meaningful and genuine
philosophical enquiry can be misleading anyway. Both the
Dogmatists and the Pyrrhonists enquire into non-evident
matters: the former think that they have found the truth about
them, while the latter suspend judgement on the truth and
falsehood of all the conflicting views and arguments about them
which they have examined.80 They are thus enquiring into the
same matters, sharing the same objects of enquiry.81 But, one
might protest, the goal of the enquiry is different: the Dogmatists
aim at the truth about their objects of enquiry, whereas the
Pyrrhonists aim to suspend judgement about them. The
difference, however, is not as sharp as one might imagine:

(a) On the one hand, ‘aiming at the truth’ about X need not
consist uniquely in aiming to discover as many true
propositions about X as possible; it might well include refusing
assent to a number of propositions, not necessarily because
they are deemed to be false, but also because assent has
been revealed to be insufficiently warranted. After all,
suspension of judgement about some matters is the
recommended outcome of truth-directed enquiry in many
forms of ancient ‘dogmatism’, from the Presocratics
onwards.82

(b) On the other hand, although suspension of judgement can

be described as the goal of Pyrrhonian zētēsis, this does not
mean that such a zētēsis is ‘truth-averse’ or ‘truth-
irresponsible’. To begin with, the Pyrrhonist does not deny that
truth exists (but suspends judgement about the matter);
second, he objects against those (the Academics) who affirm
that truth cannot be found or known83 (but suspends
judgement on whether it can be found or known, and by what
 means); finally, he does not exclude that in the future he might
end up conceding his assent to some propositions as true,
should he encounter reasons and arguments which appear to
have no equipollent counterpart.84 In fact it is because of what
appears to him to be the case at the moment, that the
conflicting views and arguments have equal force in their
claims to truth, that the Pyrrhonist is suspending judgement.85

With these qualifications in mind, let us return to our theological

simile. If an agnostic keeps examining and discussing open-
mindedly all the conflicting views and arguments about God,
with others and/or by himself, it would be ungenerous to protest
that he is no longer really enquiring into God only because he
expects that being confirmed and strengthened in his
agnosticism would be for him a desirable reward of the enquiry,
and that others too might be better off if they became agnostic.
Provided that the agnostic does not believe that it is impossible
that he could ever be argued out of his agnosticism, there is a
logical space for him to engage in a genuine enquiry: perhaps
he is not seeking God, but he is still enquiring about God. In the
same way, the Pyrrhonist, who does not exclude the possibility
of ever finding reasons which will sway him away from epochē,
still engages in enquiry, in the sense that he keeps examining or
re-examining conflicting views, evidence, reasons, and
arguments, with the hope to fortify his epochē, because he has
discovered that, after all, this has been for him, up to now, the
best treatment against psychological distress.86 Continued
aporetic enquiry is thus a way of avoiding relapsing into the
troubles of dogmatism:

T15 [By using the formula ‘to every account let us oppose an equal
account’] they make this exhortation to the Sceptic to prevent him
from being seduced by the Dogmatist into abandoning his enquiry
and thus through rashness missing the tranquillity apparent to them,
which, as we suggested above, they deem to supervene on
suspension of judgement about everything.
(PH 1.205)

Provided that epochē is not pursued by the Pyrrhonist pig-

headedly, in spite of reason and truth, an enquiry aiming at it
need not be brushed off as bogus.

(3) My answer to the question of why enquiry is not pointless if

the aim of ataraxia has already been secured by aporia and
suspension of judgement also addresses the question of why
enquiry is not, in fact, potentially damaging for the Pyrrhonist
since, if successful, it might lead to certain beliefs about value
which are a source of distress and unhappiness. If enquiry has,
so far, proved to lead to the perpetuation and strengthening of
suspension of judgement through the pursuit of a systematic
aporetic programme, then the Pyrrhonist might expect that
future performance will make any assent to value beliefs, and
the tarachē it engenders, less likely to occur. As we have seen
in T15, it is forsaking enquiry that, on the contrary, risks making
us relapse into those pernicious beliefs, because new evidence,
reasons, arguments or influences of some other kind might push
us to rash assent.
It is a different question whether the Pyrrhonist should be
distressed by the very possibility that in the future, as the result
of enquiry (or despite it), he might solve his aporia. The very
possibility that some positive answers about what is good and
bad, to be sought and to be shunned, to be done or to be
avoided, might be found should be a source of some present
anxiety.87 Perhaps so, but certainly continuous performance of
sceptical enquiry into these matters will not be what fuels this
residual anxiety. If anything, it will keep it at bay, and
progressively reduce it: even if it remains possible that there is
some truth ‘out there’ which the Pyrrhonist ought to assent to
and which he will finally discover, his expectation that this might
happen will reduce more and more as his sceptical enquiry
unfailingly keeps leading him to aporia and suspension of
judgement.
 Conclusions
I have started my enquiry with a broad-brush sketch of what I
understand to be some distinctive traits of the interconnection
between aporia and enquiry in Plato and Aristotle. I hope to have
shown that those traits also appear in Sextus Empiricus:

For Sextus, just as for Plato and Aristotle, aporia is not the
end of enquiry (even if for him, unlike Plato and Aristotle, all
enquiries undertook so far by the Pyrrhonists have led to
aporia).

Aporia is presented by Sextus, just as by Plato and Aristotle,

as the very origin of enquiry, and thus of (sceptical)
philosophy as an activity of open-ended enquiry.

But, for Sextus just as for Plato and Aristotle, aporia is not just
a preliminary to enquiry; the conflicting observations,
questions, reasons, arguments which are the source of
aporia, and make up aporiai, constitute sceptical enquiry. The
oppositional and dilemmatic (or pluri-lemmatic) structure of a
large number of Pyrrhonian aporiai, and the connection
between aporia and equipollence, are also firmly within the
Platonic and Aristotelian tradition of structuring aporiai.88

Finally, just as for Plato and Aristotle, for Sextus aporia is

beneficial to us: although aporia is initially distressing, the
suspension of judgement to which the aporetic method
 constantly leads is conducive to a desirable condition of
intellectual equilibrium and tranquillity. Unlike Plato and
Aristotle, of course Sextus does not believe that aporia also
benefitted us by helping us to discover the truth; but the
aporetic condition is itself ‘right’ and beneficial as a healthy
source of psychological tranquillity and purgation of conceit
and rashness.

These similarities need not suggest any direct influence of

Plato’s and Aristotle’s texts on Sextus (or his Pyrrhonian sources);
what they reveal is that the conception of philosophical enquiry first
invented, shaped and developed by the two great figures of classical
Greek philosophy still informed the thought and practice even of the
most vociferous critics of traditional ‘dogmatic’ philosophy in late
antiquity, and not in a merely dialectical way. With his constant
emphasis on examination (skepsis)89 and enquiry (zētēsis), and on
their tight interconnection with aporia, Sextus appears to have been
influenced by what is an ultimately Platonic legacy: not a dogmatic
metaphysical principle, but a deeper and more pervasive lesson on
the nature and methods of the philosophical endeavour itself.

* Many thanks to audiences in Bordeaux, Dublin and Rome for their

feedback on earlier drafts of this essay, and to Diego Machuca and Daniel
Vazquez Hernandez for their written comments.

1 There is only limited evidence for the pre-Platonic usage of the language
of aporia (cf. Motte-Rutten 2001: 13–35).

2 Cf. e.g. Pl. Apol. 21a4–d8 (the Delphic oracle and the aporia it prompted in
Socrates were the origin of his elenctic enquiry); Rep. 7.524a6–b5 and
524e2–5a2 (conflicting appearances and aporia prompt intellectual enquiry);
Symp. 203b2–d8 (Penia, because of her aporia, becomes the mother of
philosophical Eros); Arist. Metaph. 1.2.982b12–3a21 (human wonder and
aporia are the origin of philosophy).

3 The difference between aporia as a psychological state of puzzlement and

aporia(i) as (a) puzzle(s) is sometimes referred to as the difference between
‘subjective’ and ‘objective’ aporia.

4 The ‘zetetic’ role of aporiai, and the aporetic structuring of enquiries, can
be identified in several Platonic dialogues, including early ones (cf. Politis
2006). That role is no less evident in Aristotle’s philosophical practice; for
Aristotle’s reflections cf. e.g. Top. 1.2.101a25–36 (the role of oppositional
aporiai in establishing what is true and what is false); Metaph. 3.1.995a24–
b4 (the role of raising and solving aporiai for successful metaphysical
enquiry; cf. Politis 2003); EN 7.1.1145b2–7 (saving the phenomena and
solving the aporiai is sufficient to prove one’s point in ethics); An.
1.2.403b20–4 (the preliminary inspection of aporiai and endoxa in the study
of the soul).

5 Cf. e.g. Pl. Men. 84a3–d2 (the benefits of aporia for Meno’s slave); 79e7–
80d4 (the numbing effects of aporia); Rep. 7.515c4–d7 (the pain of aporia of
the liberated slave in the Cave); Tht. 151a5–b1 (the labour pain and fertility
of aporia in Socratic midwifery). On the pains and benefits of aporia in early
Plato cf. Szaif’s essay in this volume (Chapter 2).

6 Or most of the other ancient philosophies: for the historical and theoretical
connections between Pyrrhonism and Academic scepticism cf. e.g. Striker
1996 and Ioppolo 2009.

7 For two recent exceptions cf. Vogt 2012 and Olfert 2015, in which the
Pyrrhonian concept of ‘investigation’ is fruitfully read against the background
of Platonic and Aristotelian reflections. For an investigation of the relation
between aporia and sceptical conclusions in early Plato cf. Politis’ essay in
this volume (Chapter 3).

8 For criticism of alternative adjectives such as ‘lifelong’, ‘infinite’, ‘perpetual’,

‘endless’ cf. Machuca 2011: 251.

9 Translations for PH are from Annas-Barnes, slightly modified.

10 For the occurrence of aporia in this sense already in early Plato cf. Politis
2006.

11 Contra Woodruff 1988: 142.

12 For the opposition between aporia as resourcelessness and mēchanē as

‘resource’ or ‘expedient’ cf. e.g. Pl. Crat. 409d3–4; 425d6; Symp. 203b2–d8.
For the mēchanē vocabulary in Sextus cf. e.g. PH 1.61.

13 But never in PH: I will consider below whether this should affect our
understanding of the development of ancient Pyrrhonism.

14 A similar definition of ‘aporetics’ appears in the Suidas s.v. Purrōneioi:

aporētikoi de apo tou tous dogmatikous aporein kai autous.

15 Barnes 1992: 4290 suggests that the MSS sentence can only mean ‘from
the fact that the Dogmatists themselves also raised puzzles’ and that this is
‘patent nonsense’. For a defence of this kind of translation cf. Couloubaritsis
1990: 11–15 (‘du fait que les Dogmatiques eux-mêmes posent aussi des
apories’) and Decleva Caizzi 1981 (‘dal fatto che sia i dogmatici, sia essi
stessi, sollevano aporie’).

16 This reading seems to underlie the translation in Vogt 2015: ‘because

they brought both those who put forward doctrines and themselves to a
state of perplexity’.

17 Marcovich adds peri pantos aporein, ei kai (on the basis of Sextus’ T1);
Barnes, followed by Brunschwig, also adds peri pantos aporein, but deletes
tous dogmatikous aporein kai autous as an intrusive marginal gloss.

18 Cf. Pl. Men. 79e7–8: autos te aporeis kai tous allous poieis aporein.

19 The transition ad sensum from the singular ‘Pyrrhonian philosophy’ to the

plural ‘themselves’ (autous) is not impossible, especially in light of the
occurrence of Purrōneioi earlier in T2; and notice that the Suidas version
(see note 14), by defining aporētikoi, does not incur the same problem.
20 For the similar Aristotelian usage cf. Crubellier-Laks 2009: 3.

21 Cf. also M 7.314 (criterion), 446 (truth); 9.12 (efficient causes), 358
(material causes), 414 (line and surface); 3.60 (lines and surfaces). At M
1.169 diaporein is used for ‘to raise the problem about X of whether p or q’;
cf. similarly M 1.205. For diaporein as ‘to raise the question’ (how something
could have happened) cf. M 5.91. Cf. also PH 3.16 for the noun diaporesis.
For aporein and other compounds (epi-, pro- anti-) with the sense of ‘to raise
difficulties (against)’, cf. PH 3.115, 270; M 7.388; 8.118, 140, 244, 481;
9.258, 330, 352, 358; 10.169, 215, 247; 11.219, 257; 1.30, 131, 205, 231;
2.89; 4.21; 5.94.

22 Cf. PH 1.184; 2.95, 115; 3.55, 102, 134; M 7.283, 287, 303, 364; 8.46, 52,
77, 118, 130, 393, 394, 402; 9.42 (bis), 194, 267, 330, 430, 436, 440; 10.74,
122, 139 (bis), 153, 169, 181, 205, 245, 292 (bis), 319; 11.232 (bis), 234,
239 (bis), 246; 1.18, 29, 30, 163, 170, 232; 2.90, 96; 3.48, 77, 82 (bis), 102,
115; 4.15; 5.65. For to aporon with the meaning of ‘puzzle’ or ‘difficulty’ cf. M
9.31, 311; 10.190.

23 Cf. PH 3.134, 259, 266; M 7.87, 343, 388; 8.124, 125, 198, 336; 9.2, 13,
267; 10.5, 16, 17 (tris), 189, 215, 237 (bis), 246, 247; 11.236; 1.33.

24 Cf. PH 2.61, 127, 225; 3.79, 142, 157, 176, 238, 258; M 7.262, 304, 308,
384, 435; 8.14, 31, 32, 35, 36, 55, 65, 87, 123, 437; 9.218, 330, 348, 350
(bis), 351, 365, 421, 433; 10.44, 45, 53, 61, 103, 107, 142, 144, 211, 213,
284, 291, 298 (bis), 337; 11.1, 89, 96, 167, 235; 1.7, 15, 35, 68, 74, 84, 108,
125, 131, 132, 160, 228, 232; 2.69, 100, 113; 3.1, 80, 98, 104; 4.20, 22, 31;
6.59.

25 Cf. PH 3.73: Agrippa’s mode of hypothesis is aporos. For the aporia

caused by diaphōnia cf. T7 below.
Cf. also D.L. 9.79, where Aenesidemus’ ten modes of suspension of
judgement are presented as catalogues of aporiai. On Aenesidemus’ modes
cf. Annas and Barnes 1985.

26 On Agrippa’s modes cf. Barnes 1990.

27 For this use of aporein cf. M 5.89.

28 Cf. e.g. M 7.378, 8.40, 470; 9.326, 356, 357, 368; 10.104; 11.243.

29 For Gorgias’ use of this argumentative strategy cf. e.g. Long 1984 and
Palmer’s essay in this volume (Chapter 1).

30 Cf. e.g. PH 3.139.

31 The sorites aporia might fall under this general pattern: for example, we
seem to have no reason to accept the proposition ‘while x grains of sand are
not a heap, x+1 grains of sand are a heap’ for some value of x, while
denying it for some other value. On the sorites cf. e.g. Barnes 1982b,
Burnyeat 1982, Bobzien 2002.
For the suggestion that the argument patterns exemplified by T6 and T7
counted as aporiai for Aristotle cf. Buddensiek’s essay in this volume
(Chapter 7).

32 For aporein as ‘to be in aporia’ cf. e.g. PH 1.179; 3.54; M 7.264; 10.86,
302 (bis). For the psychological sense of the noun aporia cf. e.g. PH 1.178;
3.139; M 7.410; 2.99.

33 Cf. also PH 1.196.

34 For the frequent association between aporia and the inability to say
anything in Plato cf. e.g. Men. 80b4; cf. also Politis 2006: 96; Motte-Rutten
2001: 44–5.

35 Sextus seems to be in agreement with the point made by Aristotle at Top.

4.6.145a38–b20 that aporia is the result of equipollence, and not
equipollence itself. On this passage cf. Rapp’s essay in this volume
(Chapter 6).

36 For the interconnection between equipollence, aporia and suspension of

judgement cf. e.g. M 2.99.

37 Cf. e.g. Mates 1996: 30; Striker 2001: 113. Aporia/aporein were
translated into Latin as dubitatio/dubitare since antiquity.
38 Cf. the account of the meaning of epechō at PH 1.196: To de epechō
paralambanomen anti tou ouk echō eipein tini chrē tōn prokeimenōn
pisteusai ē tini apistēsai, dēlountes hoti isa hēmin phainetai ta pragmata
pros pistin kai apistian.

39 For all the testimonies on Aenesidemus cf. Polito 2014.

40 Gell. 11.5.1.

41 zētētikoi is not attested either in our evidence for Aenesidemus. This is

not surprising if, as Tarrant 1985 suggests, zētēsis was the ideal of the
‘fourth’ Academy, contemporary and rival of Aenesidemus.

42 The language of Ps-Gal. Hist. Philos. 7 is anachronistic: ‘Sceptics are

Zeno of Elea, Anaxarchus of Abdera and Pyrrho, who is deemed to have
practised aporetics (tēn aporētikēn) in an excessive way’.

43 Cf. Hankinson 1995: 127.

44 Cf. e.g. Tarrant 1985; Annas 1992; Bonazzi 2003. It is difficult to trace the
use of the language of aporia directly back to the late Hellenistic debates on
the doctrinal continuity within the Academy; but T10 below might refer to
these debates.

45 Aenesidemus devoted the whole first book of his work to the difference
between Pyrrhonists and Academics. This would become a topos in the
imperial age: cf. Favorinus (Gell. 11.5.6), Plutarch (Lamprias’ catalogue, n.
64) and Sextus Empiricus (PH 1.220–35, including T10 below).

46 For other occurrences of the aporia language in Photius’ report cf. Bibl.
212.170a26–33; b3–8; b15–19.

47 Cf. Bibl. 212.169b.32–4; Decleva Caizzi 1992 for a deflationary reading of

the implications of the passage.

48 Cf. also Bett 1997 and 2000. On this account traces of Aenesidemean
‘aporetic’ Pyrrhonism are evident in Diogenes Laertius, Aristocles and
Sextus’ M, which would also explain the frequent reference to the aporētikoi
in M (unlike the possibly later PH).

49 Woodruff 2010 distinguishes two ways of employing the modes of

suspension of judgement: their use could amount either to a therapeutic,
epochē-inducing rhetorical strategy (211), or to an ‘aporetic’, negatively
demonstrative (i.e. refutative) tool, which raises the obvious problem of how
to construe the modes ‘consistently with sceptical hygiene’ (211). As
Woodruff acknowledges, the latter reading, which he favours, narrows the
scope of epochē down to avoiding positive claims about the nature of things,
while admitting all kinds of negative ‘aporetic’ claims on the same subject
and allowing all kinds of beliefs, both positive and negative, about how
things are (relatively and qualifiedly). Is this position ‘only weakly dogmatic’
(217), as Woodruff suggests, or blatantly incompatible with Sextan
Pyrrhonism in PH (cf. Castagnoli 2002, 2013 and 2018)?

50 At times Sextus mentions a ‘more aporetic’ (aporētikōteron) way of

conducting a sceptical enquiry (cf. M 7.28; 9.12; 6.4–5); this way is not
opposed, however, to a ‘suspensive way’, but to a more ‘dogmatic’ (6.4,
9.12) or ‘exegetic’ approach (7.28). The ‘more aporetic’ approach seems to
consist in raising aporiai against the foundational concepts in a certain area
of philosophical enquiry, as opposed to attacking, singulatim, specific
dogmatic tenets.

51 Sextus clearly uses the two terms interchangeably here; cf. also 1.225.

52 Reading, with Spinelli, kathaper hoi peri ton Mēnodoton kai Ainēsidēmon.
MSS: kata permēdoton kai Ainēsidēmon. Other editorial proposals include:
kata Mēnodoton kai Ainēsidēmon (Fabricius); kata tous peri Mēnodoton kai
Ainēsidēmon (Natorp, Mutschmann: ‘in accordance with Menodotus and
Aenesidemus’); kata tōn peri Mēnodoton kai Ainēsidēmon (Heintz and Mau:
‘against Menodotus and Aenesidemus’, accepted, among others, by Bury
and Tarrant 1985: 75–7, Mates 1996, Annas and Barnes 2000 – but this use
of kata + genitive would be a hapax in Sextus). On this passage cf. Spinelli
2000; Bonazzi 2003: 148–58.

53 The language of aporia occurs also within Sextus’ discussion of the

question of whether Arcesilaus was a Pyrrhonian sceptic (cf. PH 1.232–4).
After having emphasised the affinity between Arcesilaus and the
Pyrrhonists, Sextus reports an alternative view, according to which because
he used ‘aporetics’ (dia tēs aporētikēs) to test his associates Arcesilaus
appeared to be aporetic; but he was actually ‘Platonic’.

54 For the common-sense view that aporia is bad cf. Pl. Crat. 415c2–9.

55 Cf. Nightingale 2010.

56 Timon: cf. e.g. Aristocl. ap. Eus. PE 14.18.28–9; D. L. 9.64 (on Pyrrho),
69 (on Philo); Aenesidemus: see e.g. T9 above; Sextus Empiricus: see e.g.
PH 3.281.

57 For the broader interpretation cf. Grgić 2006: 147.

58 This therapeutic element is absent from Sextus’ similar account (M 1.6) of

why the sceptic approached the enquiry into the mathēmata (‘liberal arts’).

59 For a shorter, similar account cf. PH 1.26.

60 The adverb tuchikōs does not suggest that the result occurred ‘casually’
or ‘by chance’ in the sense that it was an unlikely or extraordinary
occurrence. Ataraxia follows suspension of judgement ‘unexpectedly’ in the
sense that it follows by a means different from the one originally chosen.
This is compatible with the idea that there is a constant connection between
suspension of judgement and tranquillity (‘as as a shadow follows a body’:
the simile is attributed to Timon and Aenesidemus at D.L. 9.107), although
of course the existence and nature of the connection cannot be an object of
dogmatic belief for the Sceptic (for a weak construal of the connection cf.
Machuca 2006: 116–7).

61 Cf. e.g. Barnes 2007: 328.

62 Ataraxia is presented not only as the goal of the would-be Sceptic, but as
the telos of Pyrrhonism. More precisely, ataraxia ‘in matters of opinion’ and
metriopatheia (‘moderation of affections’) in necessary and involuntary
matters (e.g. physical pains and pleasures) are identified as the twofold goal
at PH 1.25–30. For the question of whether, and in what sense, ataraxia can
be identified as the telos by a Pyrrhonist cf. e.g. Moller 2004 and Grgić
2006. For the view that the Pyrrhonist’s pursuit of ataraxia is not based on
the belief that ataraxia is good, and is not a defining aspect of Pyrrhonism,
cf. Machuca 2006.

63 Cf. e.g. Barnes 2007: 329: ‘Sextan scepticism is not a philosophy; it is a

retirement from philosophy’. Cf. analogously Striker 2001: 121: ‘contrary to
Sextus’ initial claim that the Sceptic goes on investigating, philosophical
investigations seem to be precisely what the Sceptic’s way of life is
designed to avoid’. Perin 2006 and 2010 stress that the Pyrrhonist is
engaged, throughout, in a search for truth, not only as a means (to attain
ataraxia) but as an end in itself (2006: 353; cf. also Machuca 2011: 253). But
this – he notices – is incompatible with the ‘value argument’ according to
which any belief about value will undermine ataraxia (see T13 below). Perin
thus concludes that the value argument (and, more generally, the
therapeutic goal) represents a deviant, anti-rationalistic, strand in
Scepticism. On the interpretation I will sketch below some kind of
commitment to the ‘demands of reason’ is compatible with, and indeed
necessary for, the therapeutic function of Pyrrhonism. For a critical review of
Perin’s arguments cf. Machuca 2013.

64 For the suggestion that the two sources of psychological trouble are
connected, since undecided conflict is distressing because we value truth
and knowledge, cf. Machuca 2013: 208–10.

65 For similar accounts cf. PH 3.235–8; M 11.110–61 (especially 112–14).

Vogt (2012: 124–6) downplays the importance of Sextus’ consideration that
beliefs about value are themselves a source of anxiety. For the argument
that ‘value nihilism’, i.e. the denial that there is anything good or bad, would
be as effective as suspension of judgement in protecting us from anxiety cf.,
however, Taylor 2014.

66 On the self-refutation charge cf. e.g. Burnyeat 1976; Castagnoli 2010; on

the apraxia charge cf. e.g. Striker 1980; Vogt 2010.

67 For an excellent status quaestionis on the answers offered by scholars to

(some versions of) the questions I have raised cf. Machuca 2011: 251–3.

68 For my use of the term ‘right’ here cf. PH 1.17: ‘But if one counts as a
school an approach which follows some account in accordance with
appearance, that account which shows how it is possible to live rightly
(where ‘rightly’ (orthōs) is taken not only with reference to virtue, but more
loosely) and extends to the ability to suspend judgement, in that case we
say that the Sceptics have a school’.

69 The Pyrrhonist’s yielding to the ‘demands of reason’ (to use Perin’s

(2010) jargon), or ‘responding to the value of truth’ (to use Vogt’s (2012)
jargon) need not be interpreted as manifestations of dogmatic commitment,
but can be subsumed under the Pyrrhonist’s acceptance of the
phainomenon as a criterion of action (cf. e.g. PH 1.23–4).

70 Clearly Sextus does not use stasis in its political sense of ‘division’,
‘discord’ or ‘civil conflict’.

71 Such an intellectual stance is not itself believed to be definitely correct by

the Pyrrhonist. It is also supposed to be compatible with the noncommittal
adoption of some religious stances or practices as part of the reliance on the
phainomenon as criterion of action (cf. e.g. PH 1.24; 3.2; M 9.49). On
Sextus Empiricus’ arguments about, and suggested attitudes towards, the
gods and religious belief, cf. e.g. Knuuttila and Sihvola 2000, Annas 2011;
Thorsrud 2011; Bett 2009 and 2015.

72 Cf. e.g. Brunschwig 1994; Grgić 2008; Fine 2010, 2011 and 2014; Vogt
2012: 140–57.

73 For this approach cf. e.g. Hiley 1987: 189–92; Palmer 2000; Grgić 2006;
Marchand 2010.

74 Cf. also PH 1.30: for some ‘notable sceptics’ suspension of judgement is

an additional goal.

75 My suggestion can be supplemented by Palmer’s (2000: 366–8) useful

survey of the Sextan usage of the zētēsis vocabulary, showing that it is often
associated with the ideas of ‘examination’ (skepsis) or ‘questioning’, rather
than ‘search for something’.

76 Cf. also M 8.156, 10.247. The logic of the paradox of enquiry at PH 2.1–5
also presupposes the equivalence of zētein and aporein.
77 For a positive answer to this question cf. e.g. Barnes 1990: 11: ‘Whatever
Sextus may say, the Pyrrhonists did not – in any normal sense – prosecute
philosophical and scientific researches’.
For the proposal of some key criteria to recognise something as a genuine
‘epistemic investigation’ cf. Olfert 2015: ‘a bonafide epistemic investigation
has an object; a motive or stimulus; some starting content; a method; and it
aims both at knowledge and truth, and at a discovery or epistemic advance
that defines the success or failure of the investigation’. Olfert offers a
nuanced discussion of the ways in which the Pyrrhonian programme, as
presented in Diogenes Laertius, meets these criteria.

78 The connection between enquiry and aporia emerges strongly also in two
passages concerning Epicurus’ discussion of the role of ‘preconceptions’:
Ep. Her. 37; S.E. M 11.21 (cf. also M 1.57).

79 Cf. Włodarczyk 2000: 57; Palmer 2000: 366. Palmer 2000 believes that it
would be disingenuous for Sextus to depict himself as someone who
searches for the truth, since the de facto undiscoverability of truth and
suspension of judgement should be guaranteed by the generality and
universal applicability of his arguments and modes. This, however,
problematically turns Sextus into a (closet) negative dogmatist.

80 And the Academics are supposed to believe that no truth can be found
about them.

81 From this point of view, truth is in fact just one of the many objects of
philosophical enquiry for Dogmatists and Pyrrhonists (cf. e.g. PH 2.80–96; M
7.38–45), on a par with, e.g. criteria, signs, causes, gods, time, the good.

82 Cf. Vogt 2012 for in-depth exploration of this idea in Plato and Hellenistic
philosophy.

83 Cf. PH 1.226.

84 Pace Grgić 2006.

85 For similar analyses cf. Vogt 2012: 119–39 (the Pyrrhonist is responsive
to ‘the value of truth’) and Olfert 2015. I am not sure, however, that the
mature Pyrrhonist’s enquiry is still best described as ‘truth-directed’ (contra
Vogt 2012: 128).

86 On this conception of what counts as an enquiry, the Dogmatists could

object that their belief of having found the truth about X does not prevent
them from enquiring further into X. Unless they believe that they have
already reached absolute and irrefutable certainty, they can (and should)
enquire further. Cf. the first horn of the paradox of enquiry at PH 2.1–11: if Y
‘apprehends’ X, in the sense that Y cannot be wrong about X (and is aware
of this: an internalist interpretation of katalēpsis must be presupposed here),
then Y cannot genuinely raise aporiai or enquire about X.

87 A related, and more general, problem is that the Pyrrhonist’s pursuit of

epochē and ataraxia, and avoidance of belief and tarachē, could themselves
be a source a tarachē. A possible answer is that the Pyrrhonian pursuit and
avoidance are not ‘intense’, because they are not associated with the
dogmatic belief that epochē and ataraxia are good, and assent and tarachē
are bad (cf. PH 1.232–3 for Sextus’ criticism of Arcesilaus’ firm belief that
suspension is ‘by nature’ good and assent is ‘by nature’ bad).

88 A tradition inspired, in turn, by Presocratic philosophers and sophists. Cf.

Palmer’s essay in this volume (chapter 1).

89 In Plato the aporia vocabulary is connected with the skepsis vocabulary

almost as frequently as with the zētēsis vocabulary (cf. e.g. Tht. 145d5–e9).
 Chapter 12
Aporia and Exegesis
◈

Alexander of Aphrodisias
Inna Kupreeva

Aporia features throughout Alexander’s work – in the Aristotelian

commentaries, opuscula, and school treatises. Much of Alexander’s
use of aporia is prompted by Aristotle’s texts he comments on, and
often aporetic frameworks are developed following upon the earlier
school discussions and philosophical polemic against other
philosophers – both practices going back to Aristotle himself.
In this essay it is not my goal to discuss aporia as a genre in
Alexander’s literary work.1 Instead I would like to probe into a
somewhat different area, that of Alexander’s thinking about the
aporia as a part of philosophical method. This is not an easy task,
since despite the ubiquity of aporetic contexts in Alexander’s work,
there is no single place where we can find the statement of his views
on this subject. Is there a specific role for aporia in Aristotle’s
scientific methodology, according to Alexander? I will present an
answer in the affirmative and try to show, using several important
texts, that for Alexander, aporia is a vehicle of dialectical method,
and this method itself has an important formative and auxiliary role in
sciences – helping to elucidate and clarify key concepts and
arguments, respond to objections, and bring out conceptual
problems. This approach informs Alexander’s exegesis of Aristotle’s
aporiai in Metaphysics Beta. I begin in section one with a survey of
Alexander’s Aristotelian background. In section two, I present
Alexander’s view on the methodological function of dialectic. In
section three, I show how this function is fulfilled by an aporia in
relation to first philosophy.
 1 Aristotelian Preliminaries: The Form and
Scope of Aporia
Arthur Madigan notes that in the commentary on Metaphysics Beta,

(T1) Alexander uses the term aporia in at least four senses: [i] a physical
impediment to a movement in a certain direction (the original sense); [ii]
a state of perplexity (the aporia in us); [iii] a problematic object or issue,
such as to give rise to perplexity (the aporia in the thing); [iv] a
philosophical discussion which seeks to clarify a problematic issue, and
to relieve perplexity, by arguing on both sides of the issue.2

In the Topics, Aristotle criticises the definition of aporia as equality of

opposite arguments3 as ill-formed, because it suggests that aporia is
a condition (pathos) of the arguments, whereas it is a condition of
the soul. Aristotle says that aporia is caused by the equality of
opposite arguments, ‘for when we are reasoning in utramque partem
and everything on each side seems to us similar, we have a difficulty
as to which way to act’.4
Alexander elaborates on this explanation as follows:

(T2) (1) But neither is it the case that the aporia is productive of the
contrary arguments, but rather the other way around. (2) For the aporia is
a kind of affection of thought which occurs due to the contrariety of
arguments. (3) For when we are considering and scrutinising two contrary
arguments as to which one seems more fitting, and it appears to us that
equality and similarity and being in both ways belongs to each of them,
then this kind of affection arises. (4) For instance, when [a question] has
been proposed whether the soul is immortal or mortal, and the arguments
undertaken for each case prove both [the positions] sought by the
 arguments, and with strong demonstrations, in that case an aporia arises,
which position should be sided with. (5) So, when everything [in the
proofs] appears strong and similar to such an extent that we have a
difficulty which side should rather be taken, there is an aporia.5

Here Alexander focuses on aporia as a psychological state [ii] and

distinguishes this state from that which causes it, as prompted by
Aristotle’s context, namely the discussion of definitions. Outside this
context, however, neither Aristotle nor Alexander aim to reduce the
aporia to a psychological state leaving out the question of its specific
cause. In Metaphysics Beta, Aristotle uses the terminology of aporia
to refer not only to the psychological state of perplexity, but also to its
specific cause, the underlying conceptual difficulty. Alexander’s
usage in the commentary follows that of Aristotle, and the description
of the cause of psychological aporia in (T2) is referred to as aporia in
the meaning [iii] of Madigan’s list, ‘a problematic object or issue’.
Aristotle outlines the progressive, dynamic structure of a
complete aporetic argument. This is what Aristotle, and Alexander,
also called ‘aporia’, in Madigan’s sense [iv]. Aristotle distinguishes
three key points within this structure. First there is an aporia proper:
the original perplexity, which includes both the state of the soul and
its cause, the difficulty with regard to the subject matter. Aristotle
compares the objective difficulty with a knot or an obstacle which
must be known by anyone who wants to make progress.6 Although
the second meaning of the Greek word aporia, ‘lack of resources’, is
not explicitly highlighted, it probably is understood, insofar as
conceptual resources ought to be sought in order to overcome the
obstacle. In the first book of Metaphysics Aristotle famously speaks
of the state of ‘wonder and perplexity’ as the starting point of a
philosophical investigation.7
The next structural point is described as diaporia. The term
might suggest a process of dwelling on the original aporia, but
Aristotle seems to have in mind a much more precise technical
procedure of identifying and presenting the logical form of the
aporetic argument, and spelling out the difficulties in this logical
framework.8 This framework typically includes the two competing
quasi-equipollent claims (thesis and antithesis), and two respective
series of arguments pro and contra.9 On both Aristotle’s and
Alexander’s view, the impression of equipollence produced by the
aporetic argument cannot reflect the truth of the matter in question
and must instead be taken as signalling a problem to be dealt with
by a philosopher. The examples of such a fully-fledged argument
form can be found in the fifteen aporiai of Metaphysics Beta and
throughout the corpus.10 The opposition of the arguments underlying
the aporia has to be sufficiently stable and well-founded, caused by
a true puzzle and not a result of a mere oversight or a simple
mistake that can be easily corrected.11 The ‘diaporetic’ analysis
helps clarify the structure of arguments on both sides of an aporia
and can uncover and diagnose errors of reasoning and conceptual
representation used. This logical and conceptual work is not
equivalent to a solution of aporia, but it creates conditions for a
solution by laying out the problems which may not be obvious on a
prima facie reading.
Finally, aporetic reasoning includes the stage of euporia, when a
solution, ‘passage’ or discovery of conceptual resources sought, has
been obtained. At this stage, the initial sense of surprise and
difficulty should disappear. As Aristotle says, it should become more
surprising if it turns out that things are different from the way they
are.12 There is no uniform method of attaining euporia, and there is
no single type of solution in Aristotelian science.13 Still there is a
robust expectation that the aporia will be solved once we find a way
of thinking about the object which will avoid all the shortcomings and
limitations of the two opposing positions. Hence the scientific role of
aporia is seen primarily as a conceptual framework which allows us
to study all the shortcomings and limitations, as well as all the more
promising elements in the aporetic arguments.
 2 Aporia and Scientific Method
According to Aristotle, the scientific method of reasoning is
demonstration or scientific deduction. It involves the application of a
valid deductive procedure14 to a properly defined subject genus in
order to derive the proper attributes of this genus on the basis of the
axioms.15 The premises of demonstration are true and primary,
immediate, better known than the terms of the conclusion, and have
an explanatory priority to the conclusion.16 The first principles of any
science are indemonstrable, i.e. they cannot be derived from any
more fundamental principles.17
Demonstrative reasoning is distinguished from dialectical
reasoning, which is based on the approved or reputable (endoxic)
premises entertained by the two participants of a dialectical
argument, the ‘questioner’ and the ‘answerer’.18 These endoxic
premises may or may not be true. The reasoning used by a
dialectician to arrive at a conclusion from endoxic premises is
deductive. Apart from deduction, Aristotle’s dialectic presupposes
the use of inductive reasoning,19 and although Aristotle’s discussion
of it in the Topics is tantalisingly terse, Alexander adopts and
develops it as a regular part of a dialectical method along with the
arguments based on endoxic premises (we shall see an example of
his use of both methods shortly below).
The aim of the ‘questioner’ is to get the ‘answerer’ to accept a
particular conclusion (for instance, a claim which will make the
answerer’s position inconsistent and thus refute his argument).20
The strategy of the answerer is to maintain the consistency of his
position as far as possible and not yield to a refutation,21 i.e. be
careful when granting agreements to the questioner’s proposed
claims (protaseis).22
The goal of demonstration is truth. The scientist has to ensure
that the starting points of his demonstration are true and appropriate
to the subject genus of his science. The dialectician, unlike the
scientist, is not restricted in his choice of premises: he can examine
any thesis in any discipline and argue for the opposite theses. The
fully-fledged aporetic structure, with two opposing arguments, can be
an illustration of a dialectical argument conducted on both sides,
without any truth-constraints for the premises.
The question of the place of dialectic in Aristotle’s methodology
of science is controversial. That it must have some place is
suggested by the overall structure of Aristotle’s argument in many
works where the study of the subject matter has as its starting point
the analysis of the difficulties which arise from authoritative endoxic
claims.23 This might suggest that dialectic does, after all, form a
regular part of Aristotle’s scientific methodology. But such an
inclusive understanding of the role of dialectic seems to clash with
the Organon view of scientific reasoning as strictly demonstrative. It
is not my goal to discuss the whole debate about the role of dialectic
and aporetic reasoning in Aristotelian science,24 but I shall try to
outline Alexander’s position.
On Alexander’s view, every science, including first philosophy, is
demonstrative and definitional. This pretty much rules out dialectic
as a scientific method proper.25 Still, Alexander takes seriously
Aristotle’s remarks in the Topics detailing the ways in which dialectic
is useful for philosophy.26 Dialectic makes it easier to see on which
side the truth is, ‘just as the judge comes to know what is right
through listening to both parties’,27 and the person who has argued
on both sides will not be led astray by what is persuasive, and is in
the best position to see the solution to the puzzles.28 The most
detailed and technical is the discussion Alexander devotes to the last
point: dialectic contributes towards the first principles.29

(T3) (1) What he adds is to say that dialectic is useful also with a view to
the principles in each science: (2) for no science can argue about its
proper principles, because if one would speak scientifically about these
and prove them, he has to prove them from first things – this is the nature
of scientific and demonstrative proof – but one does not have any such
first thing prior to the principles. (3) So those principles of sciences which
need to be provided with some confirmation must, because they cannot
be proved through what is true and primary, be proved and justified
through what is approved – and syllogising through this is a distinctive
property of dialectic. (4) Another distinctive property of it, as Aristotle will
go on to say, is to provide a confirmation for the point at issue through
induction; and principles come to be justified most through induction. (5)
So the scientist will speak of the principles proper to his science as a
dialectician or the dialectician will do this on his behalf.(6) And if dialectic
is useful with a view to the first things, the principles of each science, it
will be so, as Aristotle says, for philosophy and its principles as well,
providing its usefulness there too.

(trans. van Ophuijsen, slightly modified)30

Both the utility of dialectic (T3.1) and the indemonstrability of the first
principles (T3.2) are Aristotelian points. Alexander’s expression
‘which need to be provided some confirmation’ in (T3.3) may require
a disambiguation. In the Greek phrase tas oun deomenas tōn archōn
tōn kata tas epistēmas sustaseōs tinos the participle deomenas
could be understood as implying that all the first principles of science
are in need of some confirmation, since no confirmation can be
provided by the science itself, which has no further foundation to rely
on beyond the first principles themselves.31 The force of the partitive
genitive construction will be to isolate the proper indemonstrable
principles as the subclass whose characteristic feature is this need
for a certain dialectical foundation. On this reading, the role of
dialectic in science, as outlined in (T3.5), would be understood along
the lines suggested by Irwin’s interpretation of Aristotle: the ‘strong
dialectic’ would set a kind of scientific discourse supplementary to
demonstration, providing a second-order justification to the first
principles of science which cannot be demonstrated.32
There is another possibility, however, and I will argue that it is
the one that Alexander has in mind in his discussion of dialectic, both
here and in the Metaphysics Beta commentary. If we take the
participial construction in (T3.3) in a narrower sense, as meaning ‘in
cases where they need some kind of confirmation’, the need for
confirmation will be dictated by circumstances, such as the necessity
to respond to a dialectical objection. In this case the partitive
construction will be isolating not the proper indemonstrable principles
as a subclass of all the principles, but very specifically the principles
which happen to be in need of some corroboration, for instance,
when they are under attack by opponents or critics. It is in this case
that dialectic can be helpful in both defending the principles and at
the same time showing ‘the way’ towards them starting from the
endoxic premises. None of these helpful roles amounts to
establishing the principles.
The battery of examples that follows in Alexander’s commentary
seems to me to give support to this reading. Alexander gives two
kinds of example to show how dialectical reasoning can provide
confirmation to the principles that need it. The first example is
showing that there are some things in philosophy that require a
dialectical proof. It comes from Aristotle’s Physics 3.5, where
Aristotle argues against the existence of an infinite body.33
Alexander gives his own interpretation of Aristotle’s argument.

(T4) (1) Aristotle himself often when proving things in philosophy, adds
‘logically’ (logikōs) in the sense of ‘dialectically’, implying that there are
also things in philosophy that require this kind of proof. (2) An example of
such [proof] is as follows: (3) [P1] Every body is delimited by a surface.
(4) This is something approved, since it has been posited that a surface
is the limit of a body. (5) Aristotle used [this premise, viz. [P1]] in his
Physics to show that there is no unlimited body.34 By adding to this that
(6) [P2] Nothing which is delimited is unlimited he has deduced that (7)
[C] Therefore: no body is unlimited.35

(Trans. van Ophuijsen, slightly modified)

On Alexander’s interpretation at (T4.1), by ‘logically’ Aristotle means

‘dialectically’ understood here as ‘proceeding from the endoxic
premises’. Aristotle in Physics says nothing about this condition for
premises, and draws a distinction rather between the ‘logical’ and
‘physical’ arguments, along the lines of a familiar discussion of the
two definitions of anger in De anima.36 In fact, it seems that
Alexander struggles to explain why [P1] above is endoxic. His
solution in (T4.3) is to say that it derives from a common formula ‘a
surface is the limit of the body’, which is criticised by Aristotle himself
in Topics 6.4 as less scientific, because it defines things prior
‘without qualification’ through things posterior without qualification.37
Aristotle in Physics 3.5 has nothing to say about this derivative
endoxon. Premise [P2] is supplied by Alexander to derive the
conclusion (7)[C], namely that ‘no body is unlimited’.38
The dialectical context of this argument in Aristotle’s Physics is
defined by the Pythagorean theory of the separate infinite, which is
discussed immediately before this argument. The argument itself
thus can be construed as a necessary response to the opposite
argument, within a well-formed dialectical framework.39
Next follows a series of arguments showing how dialectic
discusses the first principles, for the geometrical definitions.
Geometry faces the objection that it is impossible for there to be
magnitudes with only two dimensions (surfaces), only one dimension
(lines), no dimensions at all (geometrical points), and it is impossible
for us to conceive of such magnitudes.

(T5) (1) That it is a dialectician’s task to speak about principles can be

made plain from the following. (2) The geometer posits as principles of
geometry that (a) surface is that which has length and width only, and
also posits that (b) a line is a length without width, and that (c) a point is
that which has no part. (3) Some people object to this, saying that (a) it is
not possible for a magnitude to have only two dimensions, (b) still less to
have only one, and (c) that there is no such thing as a point at all, since
(i) there is nothing that will neither diminish what it is taken from nor
increase what it is added to, as Zeno of Elea said, (ii) for one cannot
 even form an impression (phantasia) of what is without dimension. (4)
Now it is not possible to offer a geometrical proof that any of these are
real, (5) but the dialectician will have no difficulty in providing a
confirmation for them through things approved. (6) For having obtained
that [P1] Surface is the limit of body which is approved, and that [P2] A
limit is other than that which it is a limit of and [I1] having provided a
confirmation for this by induction, he deduces that [C1] Surface is other
than body, i.e. that what has three dimensions; and if it is other than that,
[C2] it cannot have three dimensions, since if it did it would be the same
as body, for having three dimensions is what body has its being in. (7)
However, [I2] surface is seen to have length and width; therefore it
cannot have depth; therefore it [C3] has just the two dimensions.40

The opening formula ‘to speak about principles’ (to peri archōn
legein) in (T5.1) is general enough to suggest that for Alexander
dialectic is a special science of the first principles. However the
argument that follows shows something rather different: the role of
dialectic consists in answering philosophical or sceptical objections
against the geometrical principles. Alexander’s exact sources for this
whole argument are difficult to track down. The principles listed in
(T5.2) are post-Aristotelian and correspond verbatim to the
Euclidean definitions.41 The complex objection of the critics of
geometry (T5.3) can be related to a long tradition going back from
Sextus Empiricus through the Epicureans, Stoics, possibly earlier
Pyrrhonists, to Protagoras, and the Eleatics.42 The objection points
up the inconsistency between the physical concept of magnitude and
the geometrical concepts of point, line, surface. We don’t have any
further information about the position of Alexander’s challenger: it
can be a dialectician, sceptic, or a corporealist of some sort. The
argument for the case of points (T5.3c) is spelled out as follows: (i)
that which cannot contribute to the increase or diminution does not
exist, for (ii) it is inconceivable because it lacks extension. The same
arguments mutatis mutandis are implied for lines and surfaces,
respectively. We shall consider Alexander’s argument in defence of
surfaces, focusing on its form and function.
In a nutshell, Alexander argues that the geometers’ concept of
surface as distinct from body is both sound and conceivable.
Alexander says in (T5.6) that a dialectician can obtain two premises:
one [P1] is a familiar ‘less scientific’ definition of a surface as a limit
of body, and another [P2] is an analytical statement that limit is other
than the body.
The reference to the inductive confirmation (I1) at this point is
instructive: it seems to be taking care of the ‘conceivability’ objection
similar to (T5.3c) for the case of surfaces. Alexander may be making
use of Aristotle’s defence of geometrical objects. Sextus reports that
Aristotle defended the geometrical definition of a line against a
criticism similar to our (T5.3b) in an argument ‘by privation’ (sterēsis)
with the help of an illustration from ordinary experience:

(T6) (1) Yet Aristotle […] affirms that the length without breadth they talk
of is not inconceivable but can come into our minds without any difficulty.
(2) He bases his argument on an obvious and clear example. (3) ‘Thus
we perceive the length of a wall, he says, without thinking simultaneously
of its breadth, and therefore it will be possible also to conceive of the
“length without any breadth” talked of by the Geometers, seeing that
“things evident (phainomena) are the vision of things non-evident”’; (4)
but he is in error, or perhaps humbugging us. (5) For whenever we
conceive the length of the wall without breadth, we do not conceive it as
 wholly without breadth but without the breadth which belongs to the wall.
And thus it is possible for us by combining the length of the wall with a
certain amount, however small, of breadth to form a conception of it; so
that in this case the length is perceived not without any breadth at all, as
the Mathematicians claim, but without this particular breadth. (6) But
Aristotle’s problem was to prove not that the length talked of by the
Geometers is devoid of a certain breadth, but that it is wholly deprived of
breadth; and this he has not proved.43

Alexander in (T5.6) seems to be using a similar strategy of arguing

from privation. However, he does not rely on induction alone for
establishing the privation, but supplies the universal premise [P2]: a
limit is other than that which it limits. The inductive argument is thus
given a role in the logical structure of a dialectical argument, where it
supports the endoxic premises.44 Thus, valid deduction from [P1]
and [P2] gives us a conclusion [C1] that surface is other than the
body. It is as sound as a proper conclusion supported by credible (if
not necessarily true) premises can be. Then it is easy to derive a
corollary [C2] that surface cannot have three dimensions, and using
it again as a premise in combination with the second inductive
argument [I2], viz. everyone can see that a surface has two
dimensions, it is possible to establish a valid conclusion [C3] that
surface has only two dimensions. Once again, the conclusion is
derived by a valid deductive procedure supported by induction. It is
not a truth of geometry, but it is useful for geometry because it helps
justify its theoretical project and defend it from criticisms. So dialectic
does not form a part of geometrical method, yet ‘has a way’ to the
principles of geometry. This is what dialectic does with regard to the
first principles of all sciences, including first philosophy.
Since the context of a dialectical argument is a debate against
the opponents it might be legitimate to ask: would Alexander be able
to win this dialectical joust against a Sceptic, or an Epicurean, or a
Stoic? Does Aristotelian dialectic compete effectively against
persuasive strategies developed by other philosophical schools?
Aristotle in Topics 1.3 says that the mastery of dialectical method is
similar to that of medicine and rhetoric: it is impossible to develop a
winning strategy that would suit all individual circumstances, but the
method presupposes that none of the winning opportunities that
depend on the dialectician have been omitted.45
Alexander builds on this explaining that dialectic belongs to the
so-called ‘stochastic’ arts, which ‘do not proceed by definite steps,
but also require an understanding appropriate to them with a view to
accommodating the circumstances and ordering what is said and
done in such a way that this order makes it practically effective’.46
This is how he details the task of dialectic following Aristotle:

(T7) ‘[Aristotle] says that our command of it will be complete when we

have not omitted any of the things that can be used in a dialectical
argument conducted in a plausible way over the set thesis. For it is not
required of the dialectician that the interlocutor should always be led into
a contradiction, just as it is not required of the orator always to persuade:
his task is to omit nothing that is persuasive with a view to making the
issue credible’.47
(trans. van Ophuijsen)
In non-stochastic arts, which operate in accordance with well-defined
methods, the function (ergon) of the art coincides with the production
of the end result (telos): the task of house-building is to build houses,
and the task of weaving is to produce woven fabrics. Houses and
fabrics are also their end results. In the stochastic arts, the end result
depends not just on following all the prescriptions of the art, but also
on external factors which are outside the control of the practitioner of
the art. Thus, the function of a physician is to do everything possible
to cure the patient, but not simply to cure the patient. Curing could
happen as a result of unskilled help, as a matter of luck: such case
would not count as an achievement of medicine. On the other hand,
valid rule-based efforts of an excellent doctor are sometimes
unsuccessful because of the nature of the case.48
This discussion may shed some additional light on Alexander’s
view of the role of dialectic in first philosophy. The task of ‘omitting
no possibility’ requires that the dialectician should have full
command of the method of dialectic and know how to exploit its
prescriptions for a winning strategy in a particular case. The aporetic
argument in utramque partem is particularly well-suited for this task,
since its very structure presupposes laying out all conceptual options
as fully as possible: A, not A, pro-A, contra-A, pro-not-A, contra-not-
A. This is why dialectic may somehow ‘hit’ on the truth, even if it is
not able to establish the truth in the way the scientific demonstration
can. Alexander’s discussion of aporiai in Metaphysics Beta can
illustrate this approach in more detail.
 3 Aporetic Method and Exegesis
In his Metaphysics commentary, Alexander explicitly connects the
utility of aporiai with the utility of dialectic discussed in the Topics:

(T8) These remarks about the need first of all to work through the aporiae
would also show the usefulness of dialectic for philosophy and for the
discovery of truth. For it is characteristic of dialectic to work through
aporiae and to argue on both sides [of a case]. So what was said in the
Topics [1.2], that dialectic is useful for philosophical enquiries, is true.

(trans. Madigan)49

The Topics account of dialectic informs Alexander’s interpretation of

the aporiai in the Metaphysics Beta as arguments largely from
endoxic premises, ‘logical’ and ‘dialectical’. In his closing summary
characterisation of the arguments in Beta he says:

(T9) The aporiae presented in Beta contain arguments from accepted

opinions and conducted on the level of plausibility. And indeed it is
impossible for people to argue for opposed positions, except by using
verbal50 arguments (logikais epicheirēsesi): for nor could the aporiae be
solved, if this were not the case.51

(trans. Madigan, slightly modified)

The claim that the aporiai cannot be solved unless such logical,
endoxic arguments are used, merits attention. Alexander does not
seem to be saying that the principles from which a solution can be
demonstrated are somehow established in a dialectical argument.
This would involve a much stronger view of dialectic than what we
have seen in the Topics commentary. But Alexander’s claim here
seems to be rather counterfactual: if, per impossibile, one could
demonstrate both the thesis and the antithesis of an aporia, then
such an ‘aporia’ would not have had any solution. Such an ‘aporia’
would amount to supporting the view that both A and not-A are
genuinely and demonstrably true, which is clearly an impossibility.
So in a way the demonstrative weakness of dialectical method may
prove to be a methodological asset, because it allows us to inspect
and sort through a wide range of arguments.
It has been noticed that in the Metaphysics Beta commentary,
Alexander on several occasions uses the words ‘dialectical’ and
‘logical’ in a special sense when referring to the parts of aporetic
arguments which do not look very strong (and sometimes also have
logical faults).52 This distinction between the good and bad
arguments has been presumed, in turn, to be based on a ‘proleptic’
reading of the aporiai by Alexander. Arthur Madigan observes in the
preface to his translation of Alexander’s Metaphysics commentary:

(T10) (1) Where a developmental theorist might read large parts of

Metaphysics 3 as indicating honest perplexity on the part of an Aristotle
who feels the force of opposed positions and strives to accommodate
truth in them, Alexander reads the book in the light of his knowledge of
Aristotle’s system, and so distinguishes, at least part of the time, the
arguments which are merely dialectical from the arguments which are
well founded. (2) At no point does Alexander suggest that Aristotle
himself is seriously perplexed. (3) Perhaps surprisingly, however,
Alexander does not volunteer information about how or where in
Metaphysics the aporiae are supposed to be solved.53
Thus it is suggested that Alexander in his interpretation imports the
elements of Aristotle’s ‘official’ doctrine such as ‘well-founded’
arguments which are different from ‘merely dialectical’. This might
even be thought consistent with his general strategy of systematic
exegesis, explaining Aristotelem ex Aristotele.54 But if this is how the
exegetical strategy works, one might raise a question about the
function of the aporiai in the commentary: are they ‘genuine’ puzzles
or rather circular arguments?
Let us consider as an example Alexander’s discussion of
Aristotle’s argument for the existence of form and matter as
constituents of a sensible substance in Aporia 8. Aristotle here
operates with some elements of his hylomorphic theory which with
hindsight might be developed into a full solution.55
The question discussed by Aristotle in this aporia, which he calls
‘the most difficult and the most necessary to consider’ is as follows:
is there, or is there not, anything apart from sensible particulars? The
solution he wants to support is that what exists apart from particulars
is not the genera or species, and not the separate entities at all, but
form and matter, the hylomorphic constituents of substance.56 We
shall look at the part of the argument which derives the existence of
form and matter from the existence of coming-to-be and change. I
present its structure below as a sequence of three arguments (while
being aware of other interpretations), because this is how Alexander
construes it.

(T11) (1) [Argument for the eternity of matter] (i) [If there is nothing
besides the particulars] there would not be anything eternal nor yet
 motionless (since all objects of sense perish and are subject to motion).
(ii) But if nothing is eternal, even coming to be is impossible: for that
which is coming to be must be something and so must that from which it
is coming to be; (iii) and the last of these must be ungenerated (if (iv) the
series comes to an end and (v) nothing can come to be out of non-being).
(2) [Argument for the limit] Furthermore, if coming to be and motion
exist, there must also be limit. For first: no motion is unlimited; rather
every motion has an end; and secondly: nothing can be in process of
coming to be if it is incapable of getting into being, and that which has
come to be must (at the first moment of having come to be) be.
(3) [Argument for the eternity of form] Furthermore, if matter exists
(because of its being ungenerated), it is yet more reasonable by far that
there exists essence/substance: that which the matter is coming to be.
For if there is neither essence/substance nor matter, there will be nothing
at all; but if that is impossible, there must be something besides the
concrete whole, namely the shape and the form.
(4) [A difficulty with this position]: But, on the other hand, if one
does posit this, there is a difficulty: in which cases shall one posit it, and
in which not? That it is impossible to do so in all cases is obvious. For we
would not suppose there to be a house besides the particular houses.57
(trans. Broadie)

The argument is summarised by Alexander as follows: ‘[He says this]

to prove that if there is not something eternal, neither will there be
becoming; and if there is no becoming, neither will there be things
generated; and if there are no things generated, neither will there be
sensibles. From which it follows that if only sensible things exist,
then even sensible things do not exist.’58 Alexander points out that
the ‘eternity’ requirement in (T11.1.i) is derived as a conclusion of
endoxic argument. In Aristotle’s system the eternity would not be
ruled out by the absence of anything other than sensible substances,
since the heaven is both sensible and eternal.59 However, Alexander
treats the subsequent steps in the argument as relatively
independent from this endoxic derivation.
This is how he sets out the first problem of eternity (= T11.1.ii):

(T12) (1) That if there is not something eternal neither will there be
becoming, Aristotle proves in the following way. (2) If something comes to
be, it is necessary that there be [i] something that [it] is coming to be, that
is, that which the thing coming to be is coming to be, and, [ii] different
from this, that from which it is coming to be. (3) For example, if a man is
coming to be, there must be and must be able to be, both [i] that which a
man is coming to be (for, if man were not already in existence, a man
could not come to be – so man, which it is said to come to be, must exist
as something) – and in addition [ii] that from which this man comes to be
(for everything that comes to be comes to be from what is unlike itself; for
if it were it, it could not be becoming it); this is the subject, matter.60

The two constituents of the process of change whose eternity will be

proved are called [i] ‘that which [a thing coming to be] is coming to
be’ and [ii] ‘subject, matter’. Alexander’s example does not spell out
the exact ontological status of ‘that which’ [i]: it could be form, but it
could also be an instance of a kind. Alexander’s example of man in
[T12.3.i] suggests that the coming to be requires the presence of an
instance of a kind ‘man’.61
Strikingly, Alexander understands matter as prime matter rather
than the last proximate matter.62 He is surely familiar with the
account of hylomorphic compound in Meta. Z 8–9, where the
ungenerated matter of the bronze sphere is bronze rather than the
liquid or the prime matter.63 Alexander would have no difficulty
supplying a suitable example for a living substance.64 But this more
nuanced view is consciously omitted. His reason, I think, is that in a
dialectical argument he envisions, any proximate matter can be
considered as a sensible compound which itself has been
generated. To avoid a regress, it is necessary to make a case for
matter isolated from form.
The eternity of matter is established by two arguments as
indicated by Aristotle in [T11.1.iii]: the reduction to the infinite
regress and the reduction to the generation ex nihilo.65 The latter
argument is explicitly said to be accepted as a ‘common opinion’ of
the students of nature.66 Alexander fails to see the case for form in
Aristotle’s second argument [T11.2]67 and takes ‘limit’ to refer to the
temporal point of completion of the process of coming to be. He
develops a tortuous interpretation supplying an additional premise
‘forgotten’ by Aristotle, namely that everything that has the end point
(= limit) must have a starting point (archē), thus turning this
argument into a third proof of the ultimate prime matter.68
Alexander introduces Aristotle’s proof of the eternity of form
[T11.3] as following upon the proof of the eternal ungenerated prime
matter:

(T13) (1) Having proven, then, that the primary subject must be
ungenerated, and that coming to be does not go on to infinity, Aristotle
now proves that the form, which comes to be in the matter, must be
eternal as well, (2) thereby proving and establishing that there will be
some unitary eternal substance.69 (3) For if there is a nature of matter,
then it is all the more reasonable for there to be this essence, which the
 matter receives; this is what he indicated by saying ‘whatever the matter
comes to be’ [999b14]. (4) By ‘essence’ he means ‘form’. For that
according to which each thing has being is essence. (5) For matter,
having received form, presents that which is coming to be from it as that
which has come to be, that is as that which it receives and that which it
becomes. (6) That it is reasonable, then, for the form too which the matter
receives to pre-exist, being eternal, Aristotle proves as follows. (7) Just
as it was impossible for anything to come to be if the subject did not exist,
so too it would be impossible for there to be becoming, if that which the
subject receives did not exist. (8) Aristotle says this in the words: ‘for if
neither the latter nor the former is to be, nothing will be there at all
[999b14–15] which is equivalent to ‘for unless both matter and form were
eternal, nothing at all could come to be’. (9) Aristotle makes it clear that
this is his meaning saying: ‘It is necessary that there exist something
distinct from the composite: the shape, the form’ [999b16], meaning by
the composite that which has come to be, which is conjoint and sensible.
… (11) He rightly assumes that, as matter [exists as eternal], there must
also exist some eternal form – not that the form which comes to be in the
matter must be this; it is rather the productive [form] which, if it is like the
form that is produced, would be in some manner pre-existent.’70
(trans. Madigan, slightly modified)

In (T13.3), Alexander says that the existence of form follows a fortiori

since the being of matter has been established independently, and
since it has been assumed that there is the coming to be. The small,
but important addition Alexander makes here (T13.4) is that essence
is that which each thing is. In the Topics commentary, Alexander
gives as an example of the indemonstrable principle: ‘Of each of the
things that are, the form is that according to which it is’.71 But in our
argument (T13.3–5) it is not used as a premise of demonstration.
This argument shows that the form must reasonably exist given the
coming to be and the matter.
The eternity of form is proved at the next step (T13.6–11). Again
Alexander signals that this conclusion is established as reasonable
(T13.6). It is reasonable again given what has been established
about matter (T13.7), and this time Alexander derives the eternity of
form from the eternity of matter, reinterpreting to this effect in (T13.8)
Aristotle’s rather weaker and more ambiguous wording to say
precisely that the coming to be would not be possible unless both
matter and form were eternal. We can see that at this point
Alexander interprets Aristotle’s phrase ‘if neither the latter nor the
former …’ as meaning ‘if not both the latter and the former …’ –
violating de Morgan’s law and making a conjunction of negations into
a negation of conjunction (T13.8). But this minor logical tour de force
is in Aristotle’s interest: otherwise just one hylomorphic component
(for instance, matter) would have been sufficient for the coming to be
of a compound. Alexander dwells on this point unusually long,
perhaps to make sure that the correct meaning comes across
despite what is suggested by Aristotle’s text.
Alexander’s final clarification in (T13.11) to the effect that it is
not the future enmattered form that possesses eternity, but the
productive form which already pre-exists, seems tantalisingly
incomplete. How is the eternity of the pre-existing productive form
established? Are we to think of some version of infinite regress of
forms which will require us to stop at the first pre-existent form?
More importantly, there is a question of the force of this claim in
Alexander’s interpretation of Aristotle. It could be taken simply as
another way of saying that every sublunary living being partakes of
eternity through the species, in line with Alexander’s earlier
formulations in this text, and in line with Aristotle’s principle
anthrōpos anthōpon genna[i]. Alternatively, the expression
‘productive form’ might suggest a stronger version of the theory of
form sketched out by Alexander as a part of his own substantialist
interpretation of Aristotle’s theory of form.72 Marwan Rashed has
plausibly suggested that the passage should be read in this latter
sense and understood as Alexander’s response to the earlier
nominalist, non-substantialist interpretation proposed by Boethus of
Sidon.73 Even so, this substantialist interpretation would be used
here to clarify the reasonable conclusion of a dialectical argument.
It is also clear that Alexander is far from dismissing the
antithesis of the position backed by the hylomorphic account in our
argument, i.e. the view that nothing exists apart from sensible
substances. This view serves as a platform from which to raise
further constructive puzzles about the draft hylomorphic
interpretation. These include the problem (T11.4) of a distinction
between the cases where there is an eternal form and those cases
where there patently is not any, as in the case of artefacts. Notably,
Alexander points out that this difficulty is valid both with regard to the
hylomorphic version of the thesis developed so far and against the
‘Ideas’ version (which has not been discussed in this case perhaps
to avoid the repetition of arguments that were used against it earlier,
in aporiai five and seven). He also elaborates on Aristotle’s next
puzzle which asks whether the eternal form (as established in T11.3)
is numerically one or multiplied according to the number of sensible
substances (999b20–3). Both prima facie answers seem implausible.
A good answer will require a more precise account of form’s
presence in matter, which is the subject of the final puzzle, and an
account of the way form and matter are combined in the composite
substance (999b23–4). Alexander points out that Aristotle deals with
this problem elsewhere, ‘enquiring what it is that unifies and holds
together the form in matter; there he says that it is the potential
character of matter which becomes the cause of [matter’s] grasping
the form and [of the form’s] remaining in matter while matter is
changing into that which, up to this point, it has been potentially; and
clearly this takes place with some pre-existing productive cause’.74
Thus, although Alexander’s commentary makes a few
references to Aristotle’s hylomorphic theory, none of these
references seem to affect the dialectical nature of the arguments in
the aporiai, nor is there any evidence of reliance of the arguments on
the ‘imported’ positions of Aristotle’s metaphysics established
elsewhere. Individual arguments may differ in their logical properties:
some may contain errors and be invalid, others can be sound, but
their common ‘dialectical’ status is defined by the fact that they are
all a part of an aporetic structure which is a vehicle of dialectical
method. Such a structure cannot contain demonstrative arguments.
So perhaps Alexander’s use of the words ‘dialectical’ and ‘logical’ in
the description of the arguments in Beta should not be taken as
simply dismissive (unless indicated by context). As we have seen in
the discussion of the Topics, according to Alexander, dialectic has its
positive role, its special usefulness for the scientific principles. This
view of dialectic is confirmed in Metaphysics Beta (T8) and borne out
by our case study, where aporia is seen as providing a conceptual
network in which numerous possible difficulties and implausibilities of
the hylomorphic theory are uncovered and negotiated. If aporia is
understood in this way, then there is no reason to think that
Alexander’s commentary does not reflect a genuine perplexity. And if
he does indeed include clarifications based on his own interpretation
of Aristotle’s hylomorphic theory, this may be further evidence that
he takes Aristotle’s aporia as a framework for the living exegetical
debate.

I am grateful to Laura Castelli, George Karamanolis and Stephen Menn for

their helpful comments on the draft versions; to Stephen Menn for a very
helpful discussion of the final draft; and to George Karamanolis and Vasilis
Politis for their support and patience. I am responsible for any remaining
errors. Where not indicated otherwise, translations are mine.

1 Much work has been done on this, see Bruns 1887, 1892; Madigan 1987;
Sharples 1987, 1990, 1992, 1994, 2004, 2008; Fazzo 2002. Much still
remains to be done.

2 Madigan 1992: 87 n3.

3 hē aporia isotēs enantiōn logismōn (Top. 6.6 (145b1–2)), on provenance

cf. Düring 1968: 212.

4 Top. 6.6, 145b16–20.

5 Alexander in Top. 458, 26 – 459, 3.

6 Meta. 3.1, 995a28–33.

7 Meta.1.2, 982b11–21.
8 For this understanding, see Aubenque 1961a; Laks 2009: 28–9; Crubellier
2009: 49.

9 It can be compared, mutatis mutandis, with the structure presupposed by

the method of hypothesis in Plato’s Meno (86E–87C) and the dialectical
method in Parmenides (135E–6D).

10 We can also find examples of ‘abbreviated’ aporetic arguments, with only

the most important opposing considerations presented explicitly. This is
more characteristic of the ‘empirical’ or ‘internal’ aporiai which arise with
respect to various positions within Aristotle’s theories which seem to be
contradicted by experience or other weighty considerations.

11 Irwin speaks of ‘objective’ aporiai, Irwin 1988: 41.

12 Meta.1.2, 983a11–21.

13 Pierre Aubenque gave a preliminary classification of different types of

euporia in Aristotle: (i) euporia is a solution proper which eliminates the
difficulty and replaces it with a positive theory (an example is the discussion
of akrasia in EN 7); (ii) euporia is a plausible hypothesis which is in principle
open to revision; (iii) euporia preserves some elements of truth that are
contained in both the thesis and antithesis; (iv) it is accepted from the start
that aporia does not have a definitive solution, and the solutions that are
accepted are provisional because such is the nature of the ‘eternal’ question
(Aubenque 1961b: 14–17).

14 For the meanings of sullogismos in Aristotle, see Barnes 1982b. Here we

can use the definition of Top. 1.1, where syllogism is defined along the lines
of the modern valid argument.

15 An. Post. 1.10, 76b11–16.

16 An. Post. 1.2, 71b19–23.

17 An. Post.1.2, 71b26–9.

18 The importance of this ‘double-sided’ structure of a dialectical argument is
brought out in Smith 1993.

19 Aristotle, Top. 1.12.

20 Top. 1.10, 104a8–12, see discussion in Smith 1993: 337–8.

21 Top. 1.1, 100a18–21.

22 Alexander’s explanation of Aristotle’s definition of dialectic in the Topics

commentary does full justice to the roles of questioner and answerer.
Alexander in Top. 3, 4–24.

23 See Owen 1961.

24 The literature is huge. For the argument for ‘strong’ dialectic as the
method of Aristotle’s first philosophy, see Irwin 1988, cf. Barnes 1991. For
the argument that demonstration is the method of first philosophy, see Bell
2004.

25 On Alexander’s interpretation of first philosophy as demonstrative

science, see Bonelli 2001.

26 Top. 1.2, 101a25–b5.

27 In Top. 29, 30–1.

28 In Top. 30, 5–16.

29 My interpretation here differs from that of Smith, who relies on

Alexander’s construal of the phrase in in An. Pr. 293, 6–10, but does not
seem to take into account his discussion of geometrical examples in the
Topics commentary (Smith 1993: 349–54).

30 In Top. 29, 23–30,9.

31 Van Ophuijsen’s translation renders sustasis throughout as ‘foundation’,
which may give additional weight to this reading.

32 Irwin 1988: 196–8.

33 Phys. 3.5, 204a34–b22, at b4–10.

34 Aristotle, Phys. 204b5–7: ei gar sti sōmatos logos to epipedō[i]

ōrismenon, ouk an eiē sōma apeiron, oute aisthēton.

35 In Top. 30, 12–18.

36 De anima 1.1, 403a27–b19.

37 Aristotle, Top. 6.4, 141b15–28. Aristotle notes that these definitions are
commonly used. Brunschwig cites as an example a definition of shape as a
limit of the solid in the Meno 76 A (Brunschwig 2007: 217 n2).

38 In Simplicius’ commentary the two interpretations are amalgamated, so

that the ‘logical’ argument is presented as ‘dialectical’, and the ‘physical’ as
demonstrative. The ‘logical’ argument is said to proceed from the endoxic
premises, but also to be the most common (476, 23–9). Although Simplicius
does not cite Alexander here, he mentions the Topics, and given the
dependence of his commentary on Alexander’s, one might wonder whether
Alexander is not his source for this interpretation of ‘logical’/‘physical’
distinction.

39 Aristotle, Physics 3.5, 204a8–34. Simplicius construes the argument as a

response to Pythagoreans (in Phys. 475, 11–19).

40 In Top. 30, 18–31,4.

41 Elem. Defs. I 1,2,5.

42 On Sextus and his Hellenistic sources, see Mueller 1982, Dye and Vitrac
2009. For Protagoras’ criticism of geometry, see Aristotle, Meta. 3.2,
997b35–8a6, Alexander in Meta. 200, 18–21.
43 Sextus Math. 3.57–8.

44 It is tempting to see Alexander’s argument as refining on Aristotle’s

response to the criticism of a geometrical definition, but establishing this
requires more space and has to be postponed for another occasion. Ian
Mueller notes the use of a similar argument in a later geometrical tradition
by Apollonius of Perga. Mueller 1982: 80. Apollonius apud Proclum in Eucl.
100, 6–10; [Hero] Deff. 16.9–16.

45 Top. 1.3, 101a5–10.

46 Alexander In Top. 32, 17–20 (trans. van Ophuijsen).

47 In Top. 32, 22–6.

48 Alexander in Top. 32,12–34,5.

49 In Meta. 173,27–174, 4.

50 The English translation by Madigan has ‘merely verbal’, but ‘merely’ is not
in the Greek, and as we have seen, Alexander tends to use ‘verbal’ as a
synonym of ‘dialectical’.

51 In Meta. 236, 26–9.

52 Alexander in Meta. 206, 12–13; 210, 20–1; 218,17, cf. Madigan 1992: 76
n4.

53 Madigan 1992: 79.

54 As explained by Donini 1994 [2011]: 226.

55 In fact, the argument was used by scholars as an example of Alexander’s

own interpretation of Aristotle’s theory of form, see n73 below.

56 For recent analysis of Aristotle’s aporia, see Broadie 2009.

57 Meta. 3.4, 999b4–28.

58 In Meta. 212, 25–7.

59 To the same effect, Broadie 2009: 142: ‘Aristotle is ignoring his own
heavens and stars’.

60 In Meta. 212, 27–35.

61 The same ambiguity is present in Alexander’s alternative summary of the

whole argument at 213, 19–23.

62 This is at odds with some modern interpretations of the arguments: cf.

Ross 1924, vol. 1: 240, Broadie 2009: 142–3.

63 Meta. Z 8, 1033a31–b10; Z 9, 1034b7–19.

64 Cf. his argument against the critics of Aristotle’s definition of the soul that
‘the body that has life potentially’ refers to the embryo. Alexander Quaest.
2.27.

65 In Meta. 213, 3–10.

66 In Meta. 213, 11–13. Koinē gar autē doxa tōn peri phuseōs eipontōn ti, to
mēden ek tou mē ontos gignesthai, kai phanerōs atopon kai adunaton to
houtō ti legein gignesthai.

67 Differently from Ross 1924, vol. 1, ad 999b12, but cf. Broadie 2009: 144–
5. However, Alexander uses the language of process and completion used
by Aristotle in (T11.2) in his discussion of the next argument concerning
form (T11.3), so maybe he is still aware of the force of this argument for the
argument for form. But he definitely does not want to identify form with the
limit of the process of coming to be, probably because this would endanger
its relative independence from this process and foundational priority to it in
this dialectical argument.

68 In Meta. 213, 26–214, 17.

69 I am inclined to mark this whole section (T13.2) as a possible gloss:
although it does not necessarily conflict with the rest of Alexander’s
argument, the adjective monadikos is a hapax in the extant corpus of
Alexander, but has fifteen occurrences in the commentary by Michael of
Ephesus.

70 214, 24–215, 18.

71 In Top. 1.1, 17, 3.

72 Amply attested in Alexander’s school treatises: Mantissa 5, Quaest. 1.3,

1.8, 1.11, 1.17, 1.26, Ellis 1994; Sharples 2004; Sharples 2005; Rashed
2007; Chiaradonna 2013.

73 Rashed 2007: 240–1.

74 Alexander, in Meta. 216, 8–11. There is no clarity on whether Alexander

has in mind one particular text, or whether he is referring to some parts of
Aristotle’s work more generally. In the Metaphysics, the definition of matter
as that which not being a tode ti in actuality is a tode ti potentially is found in
H1 (1042a17–18). In H2, we have a discussion of the types of combination
of matter which account for a variety of kinds of substance. Madigan ad loc.
refers to GC 1.3 and 1.4.
 Chapter 13
The Aporetic Character of Plotinus’
Philosophy
◈

George Karamanolis
 Introduction
In his introduction to Plotinus’ philosophy back in 1967, Hilary
Armstrong, a scholar who did a great deal to revive interest in
Plotinus, writes as follows:

Plotinus was a systematic and dogmatic philosopher, who had no doubt

that he knew the right answers to the great philosophical questions which
he treated: but he was not the sort of systematizer and dogmatist who
cannot tolerate queries, objections and interruptions. He had a Socratic
belief in the value of discussion, and once a discussion had started in his
school it had to go on to the end, till the difficulties raised had been
properly solved, however long it took.1

In what follows Armstrong makes clear that in his view Plotinus

engaged in discussion with members of his school in order to clarify
to them his answers to questions they were raising. Plotinus,
Armstrong claims, followed Socrates in valuing philosophical
discussion for pedagogical reasons, while as a philosopher Plotinus
was quite different from Socrates insofar as Socrates was essentially
aporetic and elenctic while Plotinus was dogmatic and systematic.
This still remains a very widespread view about Plotinus.
Despite the enormous revival of the study of Plotinus’ philosophy in
recent decades, he is still repeatedly presented as the exact
opposite of an aporetic philosopher, that is, as a philosopher who
seeks solutions and is determined to find them in all of his treatises.2
Plotinus allegedly does this by offering a comprehensive
interpretation of Plato, addressing the major philosophical problems
occurring in Plato with a view to solving them by means of elaborate
theories, which he structured into an organic whole.3 This
widespread view clearly is not only about Plotinus’ aims but also
about his methods. Plotinus’ search for answers allegedly accounts
for his focus on Plato’s mature dialogues, such as the Republic, the
Parmenides and the Timaeus, as well as for his synthetic approach
to the earlier dogmatic Platonic tradition, the so-called Middle
Platonists, and for his relative neglect of Plato’s earlier, aporetic,
dialogues and the sceptical Platonist tradition too.4 It is the aim of
this essay to show that Plotinus was systematically using aporetic
methodology and in so doing he profoundly appreciated Plato’s
aporetic dialogues and method of enquiry. I do not want to claim,
though, that Plotinus was exclusively aporetic; as will be seen, he
was for the most part striving for answers, which he sought
predominantly in Plato, since he believed that Plato was close to
truth on most philosophical issues. Yet Plotinus did not seek ready-
made answers in Plato; he rather systematically employed aporiai as
a means for searching for Plato’s doctrines.
It is true that some scholars do speak about an aporetic element
in Plotinus’ thought. Yet, they either credit it to his students, who
used to ask questions and challenged Plotinus,5 or, they maintain
that the aporetic element was not strong in Plotinus, as he was
determined to articulate answers to the aporiai he was raising.6 I will
argue that none of these views can be quite right. Not only was
Plotinus himself using aporetic methodology in his treatises, but
there is a number of features in Plotinus’ work that point to a strong,
systematic, and elaborate aporetic character of his philosophy as a
whole that is parallel to, and compatible with, his search for answers
mainly in Plato’s work.
Feature number one is that so many treatises of Plotinus start
with an aporia, which often leads to further aporiai. As a result we
end up with a series of aporiai, often outlined already in the first
chapter of a treatise. Let me give some examples. In the first Ennead
seven consecutive treatises have such a beginning, namely Enn.
I.1–7. All of them either begin with a series of aporiai (Enn. I.1, I.3,
I.4, I.7), or, alternatively, a series of aporiai follows after a general
statement (Enn. I.2, I.6), either a general statement of Plotinus that
sets out to give a sense of the philosophical field we are moving in,
such as that of beauty (Enn. I.6), or after a relevant statement of
Plato (Enn. I.2).7 In all these treatises it is fairly clear that the author
of the aporiai is Plotinus and not a student or a member of his circle,
much less Plato or some other philosopher. It may well be the case,
of course, that Plotinus’ students raised such aporiai in his circle of
study and that Plotinus may have been inspired by them. But this
does not change the fact that the articulation of aporiai in the
Enneads mentioned is Plotinus’ own, as they are characterised by
structure, depth, and sophistication, as I explain below.
The aporiai start in a simple and straightforward way and
progressively become more complex and sophisticated as a result of
philosophical analysis that involves one of the following moves: a)
weighing various possibilities or ways out of the aporia (Enn. I.1.1),
in which case they have a dilemmatic form, or b) analyzing the sense
of a term which can be applied to a variety of instances and this
causes aporia (e.g. in Enn. I.6 regarding the term kalon), or c) by
taking a first step towards answering the aporia, which amounts to
making an assumption or a first move that originally stems from
Plato but is such as to generate in turn further aporiai (e.g. Enn. I.2).
There are, then, many possible connections between the initial
aporia and the subsequent ones: a) a series of similar aporiai (I.6),
b) a tree model of aporiai, in which the initial one gives rise to further
ones (I.1, I.2, I.7), or c) both. Such is the case in Enn. I.2.1, where
the aporia about what fleeing (phugē) is, leads to assuming that it
consists in becoming like God, but since God has virtue, the aporia
is what kind of virtue God has, the so-called civic or the major,
theoretical virtues, and whether or not we should acquire both kinds
of virtues or only the best, and how this is possible at all. In this case
the answer to the aporia about what fleeing is (that it consists in
attaining virtue) gives rise to a tree of aporiai about the kinds of
virtue and about the relation between human virtue and virtue in the
intelligible realm.8
It is all these moves that I have in mind when I speak of the
structure, depth and sophistication of aporia, which clearly point to
Plotinus himself as the author of the aporia. This is what
differentiates an aporia from a mere question: the former is a
dialectical starting point with dilemmatic structure that serves as an
instrument for philosophical research and leads to further aporiai,
resulting from the investigation of the horns of the dilemma, a
strategy that we find in abundance in Plato and in Aristotle.
Questions on the other hand do not have such structure and are thus
easier to handle. In Porphyry’s short commentary on the Categories,
for instance, which has the form of question and answer (kata peusin
kai apokrisin), questions are pedagogical devices that are
instrumental in outlining a philosophical theory with clarity, that is,
step by step. The aporiai occurring in Plotinus’ work on the other
hand play an important role in shaping his entire enquiry.
Feature number two is that many of Plotinus’ treatises are
flagged as aporetic already in their titles, which, though assigned by
the editor of the Enneads, Porphyry, do justice to the nature of the
treatises. Enneads IV.3, IV.4, and IV.5 [treatises 27–29 in
chronological order] are clearly such cases. They are said to discuss
‘Aporiai about the soul’ (peri psuchēs aporiōn) and indeed the aporiai
there make up a series, in which one aporia leads up to the next and
that to the following one. What we have here is an aporia-based
argument with dilemmatic structure, as in early Plato9 – a feature
that is captured by Porphyry’s titles. For the most part Porphyry’s
titles are not conventional but rather aim to give us a sense of the
topic that Plotinus investigates (e.g. whether there are Forms of
particulars, Enn. V.7), or of the thesis that Plotinus maintains (e.g.
that the intelligibles are not outside the intellect, Enn. V.5). In the
case of Enneads IV.3–5, Porphyry’ titles characterise both the topic
and the structure of the treatises; the general topic is the soul and
the way of engaging with it is specifically aporetic, that is, Plotinus
proceeds by raising and considering a series of aporiai about the
soul. I will return to them in detail below.
Feature number three is that in his treatises, as we have them
after Porphyry’s editorial work, Plotinus often comes back to some of
the aporiai he has raised in other treatises earlier in his career, and
raises them again in similar or modified form and reconsiders them.
In Ennead VI.7, for instance, treatise number 38 in chronological
order, we find Plotinus asking what man in the intelligible realm is,
but perhaps, continues Plotinus, we need first to ask what man in the
physical realm is before answering the previous aporia (Enn.
VI.7.4.1–6),10 and then he addresses the latter one, which he
considers as more basic. However, Plotinus had raised such an
aporia already in Ennead IV.4 [28] (esp. 18–20). There he enquires
how we should understand the compound of soul and body that
constitutes man and how precisely the soul qualifies the body so as
to be of a living kind. He goes back to this and to similar aporiai
again in Ennead I.1 [53] (esp. 10–13). Here the aporia becomes
somewhat more complex. The question now is whether man – that
is, his essence – is identical with the compound or, on the contrary,
with the intelligible part of herself, the soul, and if this is the case,
how exactly this is the case. And further, what is the subject of
perceptions and affections – the soul, the ensouled body, or both in
some way (Enn. I.1.13)? The important point for us is that Plotinus,
like Plato, comes back in his work to the same or similar aporiai and
reconsiders them, which means that he is not entirely happy with his
earlier treatments of them.
Feature number four is one of the few that has already been
appreciated by the critics, namely that raising and discussing aporiai
was an established practice in Plotinus’ school. Porphyry tells us
specifically that he was raising aporiai and Plotinus was either
addressing them personally or he asked Amelius to do that (V.Plot.
13.10–17, 18.10–19). Even if that story is fiction, it is telling that
Plotinus was favouring aporiai in his seminars and he was
considering them as an essential philosophical tool; for Porphyry’s
point in the Vita Plotini apparently was to show that he was aligned
with, and even promoting, Plotinus’ own philosophical method.
Porphyry’s aim was not so much to create difficulties for Plotinus, let
alone to question Plotinus’ views and interpretations of Plato, but
rather to emulate Plotinus’ way of philosophising. It is because
Plotinus had adopted the aporetic method of enquiry, I suggest, that
he favoured raising aporiai in his seminar and why Porphyry raised
many of them in emulation of his teacher’s method. If this is the
case, then the aporetic character of many of Plotinus’ works is
reflected in the overall way of philosophising in his school.
The above facts require interpretation, however, which is what I
aim to do in what follows. What particularly requires interpretation is
Plotinus’ uses of aporia and the different senses of aporia that we
find in his work. I tend to distinguish between uses and senses of
aporia. Roughly speaking, the ‘use’ has to do with the application of
aporia within a dialectical context, with the form of an aporetic
enquiry, while the ‘sense’ is about whether aporia constitutes a
difficulty in the mind of the philosopher who carries out an enquiry,
an aporetic state of mind, or, alternatively, is a puzzle in a rather
objective way, an impasse concerning a certain topic which may or
may not be capable of being resolved. The aporiai in the former,
subjective, sense can be elenctic and critical or refutative, as in the
early Socratic dialogues, or more constructive, aiming to map out the
relevant philosophical territory, like the aporiai we find in Aristotle’s
Metaphysics B, for instance. As we will see, Plotinus often remains
non-committal to a final solution of such an aporia and we can detect
a certain sceptical thrust of the aporia. The aporetic state of mind
often becomes a sceptical state of mind. Plotinus seeks to resolve
the aporia, but he does not always succeed, at least not entirely.
Within the subjective sense I will distinguish several uses.
Concerning the objective sense of aporia as a puzzle that confronts
us Plotinus is not confident that such an aporia can be resolved (see
below section 3).
 1 Zetetic, Exegetical and Elenctic Aporiai
Let me start with Enneads IV.3[27], IV.4[28], and IV.5[29], which bear
the title ‘Aporiai about the soul’ (Peri psuchēs aporiōn), which is
translated as either ‘Difficulties about the Soul’ or ‘Problems about
the Soul’.11 As their chronology indicates, these works make up a
whole, a unified treatise that Porphyry divided for reasons of
convenience, and it is the whole that is characterised by an aporetic
structure.12 The beginning of the first treatise confirms this and is
telling about Plotinus’ overall aim.

Concerning the soul, the right course, I feel, would be to conduct our
enquiry in such a way as either to arrive at solutions to the relevant
problems, or, if remaining in a state of puzzlement on those points, to
regard this at least as a gain, that we know what in this area does not
admit of solution. On what subject, after all, would one more reasonably
spend one’s time in prolonged discussion and investigation than on this
one? (Enn. IV.3 [27] 1.1–6, Trans. Dillon-Blumenthal)

Plotinus tells us here that he means to focus on the issues about the
soul that cause puzzlement to him. His purpose, he says, is to find
solutions, but even if we do not find any, he continues, it would be a
gain for us simply to know what those puzzling issues are and also
what makes them puzzling. He adds that the soul is one of the most
worthwhile subjects to investigate and to discuss. Plotinus’ plan is
reminiscent on the one hand of the Socratic method that we find in
Plato’s early dialogues, namely of the aporia as to how to define
something and Socrates’ professed ignorance about this (although
Plotinus does not seek to define anything here), but it also reminds
us of Aristotle’s emphasis on the aporetic enquiry he embarks upon
in the beginning of Metaphysics B.13 The similarity between Plotinus’
cited passage and the beginning of Metaphysics B is indeed striking.
Like Aristotle, Plotinus underlines that we cannot reach conclusions
and thus attain euporia, resourcefulness, unless we first consider the
relevant aporiai carefully, so that we get clear about them, which is
partly a matter of finding what element is primarily or essentially
responsible for the aporia, that is, what is the cause of aporia, of our
puzzlement.
If we now turn to treatises IV.3 [27], IV.4 [28], and IV.5 [29]
themselves, we can see how they make up a whole and how they
carry out the aporetic plan that Plotinus announced in the passage
cited above. It is not only that they are all permeated by a series of
aporiai regarding the soul, as is also the case in Metaphysics B
regarding being; the case rather is that we find in them an aporetic
structure such that one aporia builds on the previous one, that is the
tree model of aporia I was describing earlier. More precisely, Plotinus
starts with a most general or not fully articulated aporia, and moves
to more specific or more articulated ones. The most general aporia is
how the soul comes to be in body (pōs en sōmati psuchē gignetai
zētōmen; IV.3.1.15). This is clearly a general enquiry that can be
about the way the soul enters the body, but also about how the soul
is present in the body and how it functions in it. Plotinus will spell out
all these possibilities and consider them one by one. It is interesting
to note that Plotinus addresses those philosophers, apparently
Platonists, who tend to give a quick answer to this general aporia, by
claiming that our souls stem from the world-soul and cite various
texts from Plato in support of their view. Plotinus sets out to show
that this view is beset with many and serious difficulties and needs to
be spelled out and considered carefully, which is what he does in the
following chapters. We should highlight here the contrast between
the attitude of Platonists like Longinus, for instance, who jump to
answers and ground them in passages from Plato and that of
Plotinus who is not satisfied with these answers and is not
impressed with their alleged grounding in Plato either.14 Plotinus
rather takes them as starting points of his enquiry that should be
investigated further with the aim of advancing a charitable
interpretation of them.
Plotinus sets out to enquire first about the relation between the
world-soul and the individual soul, which in turn leads him to enquire
about the sense in which something (the individual soul) is part of a
whole (the world-soul, IV.3.2). Plotinus suggests that it makes no
good sense to speak about parts in the case of the soul, as it is no
quantity or magnitude (IV.3.2.29–32), although, as he later admits
(IV.3.7.1–12), this in a certain sense is true. But the aporia about the
sense in which the individual soul is part of the world-soul is not the
only one. The further aporia is how a soul is about the entire world
and other souls about parts of the world (IV.3.2.57–58). Is the
relation, Plotinus wonders, similar to that between the soul being
present in one’s finger and being present in the entire living being
(IV.3.3.1–3). For the soul is one in this case and there are no parts of
it, which means that the world-soul would be present in all parts of
the world, in all living beings. This may well be the case if the soul is
nous, intellect, yet he notes that we see, on the other hand, that
there is a soul that is divided into bodies, and one wonders how this
actually happens, given that the soul as such is not divisible
(IV.3.4).15 Furthermore, one also wonders, Plotinus claims, how one
soul is yours, the other is of that person, and still the other mine
(IV.3.5). There follows a series of aporiai about the world-soul, its
cosmic role, and its role as a principle of souls, which Plotinus
considers from Platonic passages from the Philebus, the Timaeus
and the Phaedrus (IV.3.6–8). The central aporia here is how the
world-soul gives rise to individual souls.
The aporia that follows and can be considered as the second
major aporia after that concerning the role of the world-soul vis-à-vis
the individual souls, is how the soul enters the body (IV.3.9), that is,
how the incorporeal soul associates with the body and rules over it.
Plotinus considers various possibilities here and at the end he rules
out that the soul is present in the body as a quality in an object or as
form in matter (IV.3.20), a view defended by Aristotle and revived by
Alexander of Aphrodisias.16 In the following chapters Plotinus takes
up the aporia again, focusing on the manner (tropos; IV.3.21.7) in
which the soul is present in body, and he considers further options
(IV.3.21–22).17 Plotinus suggests now that the soul is present in the
body in the way light is in the air and in an illuminated object; just as
an object transforms light in different ways, so the body, he claims,
receives the power of the soul in different ways depending on the
bodily organ or part that receives it, that is, as sight, hearing, taste,
smell and so on (IV.3.22). Plotinus’ imagery suggests the way in
which he is going to resolve the aporia about the soul’s presence in
the body, namely by arguing that the soul is not in the body but
rather the body is in the soul (IV.3.22.7–11). For, he argues, it is
because the body participates in the soul that it becomes something,
namely living body.18
This is a dialectical move that allows Plotinus to go on in his
enquiry. For he still has to show how the soul makes the body living.
Plotinus moves to investigate how specific sense functions work,
especially memory (IV.3.25–32). Plotinus is particularly interested in
memory as an individuating feature of the soul, as an element that
accounts for the distinct personal character of the individual soul.19
In treatise IV.4 [27] Plotinus takes on further the enquiry about the
functions of the soul that affect the body, such as memory once
again, but also perception and cognition more generally and, in
addition, emotions and desires. And then in Ennead IV.5 [28] he
arrives at a very specific aporia about the sense-perceiving function
of the soul, namely whether vision is contingent on the existence of a
medium that mediates between the eye and the perceived object, as
the Peripatetics had suggested.20
What Plotinus is doing in these treatises that make up a unity,
as I have said, is to systematically explore the territory of what the
individual human soul is and how it is related to, and functions in, the
body. He apparently believes that this territory needs to be explored
anew from what he considers a Platonist point of view in its wide
complexity, if we are to acquire some understanding about the soul.
This driving motivation surfaces throughout these treatises, which at
regular intervals bring us back to the main question of how the soul
comes to be in body. Throughout this investigation Plotinus refers or
alludes to a number of Platonic passages, from the Philebus, the
Timaeus, the Phaedo and the Phaedrus, which relate to the question
regarding the soul and function of the soul, most clearly in Enn.
IV.3.7; none of them, however, resolve the aporiai Plotinus raises. In
the course of the investigation Plotinus critically reviews Platonist
and Peripatetic views but he eventually rejects them because they
raise more aporiai without resolving any.21
Plotinus does take a position regarding some of the aporiai that
he raises, as is the case with the way the soul relates to body.
Concerning others he takes a more dialectical approach. In Enn.
IV.5, on the function of vision, Plotinus outlines in detail a possible
Platonist response to the Peripatetic theory of a transparent medium,
which he considers but eventually rejects. It is important to note,
however, that Plotinus remains aporetic up to the end of the
investigation, even while endorsing some views regarding the issues
he examines. For some of the aporiai he discusses he seems to
remain actually sceptical as to how they can be resolved. Quite
telling of the sceptical thrust of these aporiai is that Plotinus comes
back to the same ones in later treatises. In Enn. V.3 [49] Plotinus
comes back to the enquiry about the soul and he raises the same
aporiai that we find in the last part of IV.3 and in IV.4, namely how
the soul perceives, experiences emotions and has memories. Of
course Plotinus has some unconditional views about the soul that he
does not challenge, such as the soul’s intelligible and unaffected
nature or the soul’s presence in the body, but much else remains
under investigation and a part of it eventually unresolved.
 With regard to the soul, then, Plotinus proceeds by means of
aporiai because in his view the entire territory is full of difficulty and
needs to be explored anew and because Plato had not investigated
it systematically. Those who did, like Aristotle, for instance, or the
Stoics, did not give satisfactory answers, while Plotinus’ Platonist
predecessors or contemporaries did not appreciate the difficulties
besetting Plato’s work regarding the soul that Plotinus’ series of
aporiai bring to the surface. Plotinus proceeds aporetically here
because he does not have any other secure basis for his
investigation, since Plato’s works alone pose more difficulties than
they resolve and they require careful interpretation. Plotinus clearly
is not merely asking questions here which he subsequently tries to
answer; as he does in treatises like Enn. I.3, On Dialectic, for
instance, where he asks what dialectic is and then he comes up with
a definition (Enn. I.3.4). Rather, he invests his energy in carefully
formulating aporiai, because, as he makes clear at the beginning of
IV.3, he believes that we gain understanding about the soul when we
specify what causes us puzzlement about it. Some of these aporiai
may have been raised in some form by members of his circle,22 but
Plotinus gives them structure and integrates them in his enquiry,
which is motivated by the view that the difficulties about the soul are
telling about its nature. In this sense Plotinus’ approach is similar to
that of Aristotle in Metaphysics B. For, like Aristotle, Plotinus thinks
that the aporiai are necessary for investigating the nature of X (soul
or being) and he is focused primarily on the best possible articulation
of aporiai, not on their solutions.
 One thing that makes the entire enquiry about the soul
particularly difficult is Plotinus’ leading idea, which he takes over
from the Timaeus (and which the Stoics adopted), according to
which we are part of a bigger living organism, the world, which has a
soul too, and that there is actually one soul that is present in all living
entities (cf. Enn. IV.9, if all souls are one). The fact that Enn. IV.2
starts with the discussion of the world-soul (IV.2.2) shows that in
Plotinus’ view the question of the soul-body relation should be
approached from this perspective. Plotinus repeats in many places in
his work that soul is an external cause that makes us, like many
other entities, living, that is, the soul, quite generally, is the cause of
a certain structure, that of the living organism (cf. Enn. IV.8.2).23 This
belief is one important source of puzzlement about the ways in which
soul is a cause of a wide variety of phenomena accounting for that
structure while it still is not in body strictly speaking. Plotinus
elaborates on the aporia of how the soul is not in the body but rather
informs the body and accounts for its living structure without being
present in it as such. However, this raises further difficulties as to
how exactly the living being, the compound in Plotinus’ terms, is to
be conceived, if the soul is not in the body but still operating in it.
This is the aporia that Plotinus raises and explores in Ennead I.1
[53].
If we look in closer detail, we see that this mature treatise of
Plotinus proceeds aporetically from beginning to end.24 The driving
aporia is what the subject of perceptions, emotions and affections
more generally is: is it the soul, the compound, or both? This aporia,
however, stems again from views about the soul that Plotinus
endorses, such as the view that the soul cannot be in the body in a
literal sense. The aporiai, then, that Plotinus raises and explores in
Enn. I.1 are once again conditional, as they result from the
assumption that the soul is not in the body but remains outside it yet
operates in it. This is a pattern that we also find in Enn. IV.3–5,
according to which a condition, p, say the soul’s affinity with the
world-soul, gives rise to a series of aporiai, p1, p2, p3. Plotinus
considers various candidates as solutions of the aporiai raised in
Enn. I.1, which were suggested by Peripatetics, Stoics, and fellow
Platonists, which he invariably rejects, until he arrives at his own
suggestion (Enn. I.1.7), according to which the soul operates in the
body through faculties that enable the body to carry out its living
functions.
This is not the resolution of the initial aporia, however. Plotinus
raises an aporia to his own suggestion: if that is the case, he says,
then the soul is not affected by perception. How is this possible, he
wonders, given that perception involves judgement? If the soul is not
involved at all, then how can we explain its care and sympathy for
the body that it enlivens? Plotinus moves on to claim that the soul is
present as an image (eidōlon) that acts through the body without
being present in it. Plotinus has outlined a similar theory already in
Ennead IV.4 [28], as mentioned earlier, where he speaks of the trace
of the soul (IV.4.28) or the shadow of the soul (IV.4.18), which
according to Plotinus operates in the body so that the body acquires
the structure of the living body and allows the operation of the soul
without its actual presence in the body.25 Why, then, does Plotinus
come back to the same issue in Enn. I.1? I suggest that it is because
Plotinus was not entirely satisfied with his earlier answer. He actually
lists the three candidates for causing bodily desires and affections –
the soul, the soul using a body, and the composite of soul and body
(Enn. I.1.1.1–4). The soul-trace that he favoured in Enn. IV.4 for
explaining how the body has desires, affections and so on, it does
not explain how exactly the body is organised so as to be living. In
Ennead I.1 Plotinus now seeks an answer by means of a scheme
that involves faculties (dunameis) and activities (energeiai) and he
considers the implications of this view. One striking aporia that can
be detected throughout the treatise and is voiced clearly at the end
of it, is, ‘who is doing this investigation?’ Is it the soul or the soul in
body, and if the latter how exactly does it achieve this? This reflects
a kind of existential concern.26 Plotinus voices a similar aporia in
VI.7 [38].4, where he enquires what the man is, a soul or a soul
using a body (VI.7.4.7–11). Plotinus eventually answers the aporia in
Enn. I.1, that it is we in so far as we are soul and in so far our soul
operates as an intellect. But the aporia still resonates. At any rate
Plotinus remains in this late treatise as thoroughly and engagingly
aporetic as in the earlier Enneads IV.3–5.
In all these treatises aporia is a form of dialectical reasoning, on
which Plotinus relies for the purpose of determining the subject
matter of his philosophical enquiry, namely the soul and especially
the human soul, as Aristotle does in the Metaphysics B. But Plotinus
differs from Aristotle first in that his aporiai make up a unified
argument about the soul, or at least a more unified argument than
Aristotle’s concerning being,27 and, second, in that the aporiai
Plotinus is formulating do not only aim to guide us further in defining
the philosophical puzzle we are confronted with, but eventually have
a clear epistemic implication; they lead us to dialectical or even
sceptical conclusions. In this sense Plotinus proceeds like Plato in
the early dialogues.28 Plotinus remains at least non-committal about
the right answer regarding the status and function of the soul. This
does not mean that Plotinus disavows knowledge in a manner
reminiscent of Socrates, because, as I said, his aporiai are
conditional, that is, they result from specific views that he takes as
starting points; but it does mean that Plotinus remains inquisitive
about some key issues. It is indicative, for instance, that Enn. I.1
ends with a number of aporiai similar to the initial ones and that Enn.
IV.5 ends up by rejecting the initial hypothesis of the transparent
medium without replacing it with another. I would call this kind of
aporia zetetic: it is an aporia that guides an enquiry.
There is a particular form of zetetic aporia. This occurs when
Plotinus investigates what something is in general (holōs), as he
says, for instance quality (Enn. II.6, esp. 1.15), time (Enn. III.7), love
(Enn. III.5), or how something happens to be the case, such as the
mixture of all (Enn. II.7), or how distant things appear smaller (Enn.
II.8). In a sense Plotinus’s enquiry is a revival of the Socratic ‘ti esti;’
question. Of course, he does not always strive to define something,
such as a concept, as Plato’s Socrates did. He often sets out get a
clear idea about what something is, namely a natural phenomenon,
such as about the appearance of distant things. When taking this
approach, Plotinus enters into dialogue with the views of other
philosophers and he criticises them by raising aporiai in response to
them. This is what he does in Ennead III.7 while enquiring about
what time and eternity is, for instance – he considers the relevant
views of the Stoics and Aristotle, respectively. Plotinus starts from a
state of aporia (aporountes, Enn. III.7.1.8), given the diversity of the
relevant doctrines of the ancients, but this enquiry leads Plotinus to
take up ultimately an interpretation of the relevant parts of Plato. In
the case of time, for instance, he does this with respect to the
Timaeus (e.g. III.7.6). But his enquiry is not about how Plato should
be interpreted, but rather about what time is and what sense it
makes to conceive of it in this or the other way. The views of Aristotle
and the Stoics are considered as unsatisfactory answers in light of
the difficulties that Plotinus raises for them, but once again the way
out of the impasse is not a mere appeal to a relevant Platonic
passage taken as Plato’s view, but rather a thesis that Plotinus
formulates after considering the possible answers and how they fare.
There is, however, one use of aporia in Plotinus that concerns
the interpretation of a specific part of Plato. We find Plotinus often
expressing an aporia about what Plato exactly meant in this or the
other passage, for instance in the Symposium about love (Enn.
III.5.2),29 or in the Phaedrus about the descent of the soul, in the
Republic about the ascent from the cave (Enn. IV.8.1), or how
Plato’s statements about the soul being present in earth should be
interpreted (Enn. IV.4.22.6–12),30 or how according to Plato soul and
body are related to each other (Enn. IV.8.1.23–8). This use of aporia,
even though it revolves around a specific text of Plato, is integrated
in a greater aporia set by the philosophical topic Plotinus is enquiring
about, such as love, the descent of soul, or the soul-body relation.
The aporia about the interpretation of a certain text of Plato arises at
a certain juncture of Plotinus’ work, where a clear sense of Plato’s
view would help the enquiry to move forward. But the appeal to Plato
raises further the aporia as to what Plato means or how we should
interpret Plato. Even when the evidence from Plato helps us to get a
clear sense of Plato’s view, this is hardly the end of the enquiry for
Plotinus, but only a small part of it.
This is clearly the case in Ennead IV.8 (On the descent of the
soul), where Plotinus starts with such an aporia about what Plato
means (IV.8.1) but this only helps to determine better the object of
enquiry (IV.8.2). Plotinus still needs to specify what the soul’s
descent into the body amounts to. The evidence from Plato
continues to play a role in the treatise. Plotinus needs to make sense
of Plato’s metaphors and images of the soul’s descent (e.g. IV.8.5)
while he is working towards a theory of the soul’s descent into body.
The aporia that is raised about the interpretation of Plato is part of a
wider, zetetic, aporia. It is also, however, distinct from it, because it
has a more narrow scope. I would label it therefore exegetical
aporia.
Now I want to turn to a different use of aporia. As mentioned in
the introduction, six treatises of the first Ennead begin with an
aporia. Consider for instance the beginning of Ennead I.4 [46], or I.5
[36], or I.7 [54]. In all these cases Plotinus takes up an Aristotelian
view or a Stoic view, and asks whether they can be right. In Enneads
I.4 and I.5, for instance, Plotinus takes up the Aristotelian view of
happiness and he raises a number of aporiai about it. Plotinus does
not enquire about Aristotle’s doctrine of happiness in general but
rather about specific implications of it, such as (in Enn. I.4) whether
we need to extend the possibility of happiness to living beings other
than humans, if indeed we proceed on the assumption, allegedly
Aristotle’s, that happiness (eudaimonia) and well-being (euzō[i]ia)
are identical;31 or whether happiness, though an occurrent state,
increases with time.32 The latter possibility arises, however, only if
one sides with Aristotle’s view of happiness according to which, at
least in the interpretation of Plotinus outlined in Ennead I.4,
happiness consists not only in virtue but also in pleasure and the so-
called external goods.
Such aporiai are partially constitutive of a critical review of
Aristotle’s ethical theory and they have a critical aim, namely to show
that Aristotle’s theory leads to impossible conclusions. This is the
case with Stoic views too, such as the Stoic view about rational life,
or about time. The critical scope of these aporiai does not mean,
however, that their purpose is exhausted in the refutation of rival
philosophical theories. The aim of the critical aporiai, as I would term
them, is rather to move forward the philosophical enquiry into the
relevant issue by considering the implications of a certain theory. In
the case of happiness this is quite clearly the case. Plotinus
proceeds finally to outline his own view about happiness, but this
takes place in a continuous dialogue with Aristotle’s theory and its
underlying assumptions. In Ennead I.7 Plotinus starts off from
Aristotle’s reasoning about the good for an entity and shows where
he goes wrong and what must be the right turn in Plotinus’ view.33 It
is not the soul’s good, Plotinus suggests, but the good as such, the
good haplōs, that we should seek, which permeates everything and
towards which everything strives (Enn. I.7.1.17–28), that is, the
Good of the Republic that Aristotle rejects in Nicomachean Ethics I.6
as a source of explanation of what counts as good.
In such cases the scope of aporia is more limited and concerns
a specific philosophical doctrine that Plotinus is subjecting to
scrutiny. This use of aporia is similar to the Socratic elenctic aporia,
which aims to show that the interlocutor’s view leads to perplexity or
to impossible conclusions. Like Socrates, Plotinus is testing the
validity of a certain view, but unlike Socrates, Plotinus uses the
results of this aporetic investigation for building up his own view.
Plotinus was actually confident from the start in these cases that the
aporiai he was considering could be resolved. This is a difference
from sceptics, especially Pyrrhonian ones, who believed than it may
not be possible to find any positive answer. As I said earlier, this
shows that Plotinus’ aporetic approach is not at odds with his
commitment to finding answers, especially in Plato, that is, with his
doctrinal interpretation of Plato, but quite compatible with it.
 2 An Ontological Aporia
We have seen so far three uses of aporia in Plotinus, namely a
zetetic, an exegetical and a critical or elenctic. There is finally
another use of aporia that is present in Plotinus’ work, and this is, I
think, peculiar to him. In several places in his work Plotinus enquires
whether something that exists in the sensible realm and is called X
or Y, is identical with what is also called X or Y in the intelligible
realm. We find Plotinus applying this aporia to a variety of subjects,
namely to man, to potentiality and actuality (Enn. II.5),34 to
substance (Enn. II.6.1, Enn. VI.1.2), to quality (Enn. II.6, Enn.
VI.1.10), to beauty (Enn. I.6). The aporia that is raised here is,
specifically and distinctively, an ontological one. Plotinus enquires
into whether we are dealing with two distinct kinds of X, a sensible
and an intelligible one, in spite of their superficial similarity, that is,
whether we have a homonymy, a P-series as A. Lloyd has
suggested.35 If this is the case and the two things X and X1, as it
were, do not fall under the same genus, as Plotinus believes, then a
number of further questions arise, such as how we should conceive
the sensible X (i.e. the X1), given that the intelligible X is the real one
and indeed the model of all Xs (X1, X2, Xn). This gives rise to the
further question of how the two kinds of entities, sensible and the
intelligible, relate to each other.
Plotinus employs this aporia already in his first treatise in
chronological order, in Ennead I.6, where he investigates the nature
of the beautiful or the fine (to kalon). Plotinus asks how it can be
possible that things in the intelligible realm are beautiful in the same
way as those in the sensible realm. How is it possible, he asks, that
material things are identical in some respect with intelligible things,
how can intelligible and sensible entities be similar in some respect
(Enn. I.6.2.11–13)? Plotinus suggests that sensible entities become
beautiful by participation in the intelligible ones, that is, the intelligible
Forms, since, as Plotinus claims, the Forms are that which are as
such, beautiful or beauty in its essence. This answer shows first that,
in Plotinus’ view, the two classes of things, sensible and intelligible,
are not beautiful in the same sense, and also, second, that they are
not unrelated either; rather, intelligible beauty is the cause of
sensible beauty. But it still remains open as to how the two items are
related and how justified we are to think of both as X, kalon.
Although this is a specifically Plotinian kind of enquiry it has its
roots in a well-known Platonic kind of enquiry. We find this, for
instance, in Hippias Major (esp. 287c–d, 288a, 289d), where
Socrates and Hippias discuss how beautiful things are beautiful
(kala) by sharing in what is beautiful (to kalon). One central issue
here is what the affinity is between that which is beautiful and the
many beautiful things that are so diverse, as a beautiful girl and a
beautiful monkey. Plotinus follows Plato in distinguishing between
the intelligible Form F, which is the cause of all Fs, and the many Fs,
which become such by participating in the Form of F.36 Plotinus’
aporia is not about the distinction between the Form and the
instances of F but rather about their similarity given the distinction.
More precisely, he is concerned to enquire about where exactly their
similarity lies and how the intelligible F is the cause of sensible Fs.
 Plotinus uses this aporia as a methodological tool in order first
to distinguish entities, such as man or beauty, or concepts, such as
potentiality and substance that do not belong to the same genus.
This distinction often has a critical point, namely to reject views of
philosophers like Aristotle, for instance, regarding substance or
quality, who applied these terms without distinction to both intelligible
and sensible entities, universals and individuals. Similar is the case
with the Stoics, who according to Plotinus also failed to distinguish
between intelligible and sensible beauty, which in Plotinus’ view
leads them to wrong conclusions about what beauty is. Their
conclusion, that beauty is a kind of symmetry of the parts, shows in
Plotinus’ view that they ignore what intelligible beauty is the cause of
(Enn. I.6.1.20–50). For Plotinus both Aristotle and Stoics make a
category mistake; they fail to distinguish two classes of entities,
which in Plotinus’ view are merely homonymous. By making that
mistake, Aristotelians and Stoics fail to realise what makes
something the thing it is, for instance beautiful. Plotinus first clears
the category mistake and then he proceeds to enquire into the
relation of the ontologically distinct entities. He is confident that he
can resolve this form of aporia and he typically comes up with a
theory according to which the ontologically superior entity is the
cause of the ontologically inferior one in a manner he specifies. In
the case of beauty Plotinus eventually maintains that the Beautiful is
an item of the intelligible realm that should be identified with the
Good, that is, the Form of the Good (Enn. I.6.9.42–4).
This form of ontological aporia, however, does not only have a
dialectical role to play in Plotinus’ enquiries, but once again also has
epistemic implications. Once we realise what the kalon really is, for
instance, the question arises of how we can get to know it. His
suggestion at the end of Ennead I.6 is that this knowledge
presupposes a certain state of mind, or a state of soul, as he says,
namely that the soul becomes similar to the object of cognition, the
beautiful. Only if one achieves this cognitive state, Plotinus claims,
can one be in a position to ascend to the beautiful. This Plotinus
eventually identifies with the good, which is considered to be the
source of the beautiful. But both the good and the beautiful are
beyond the sensibles (epekeina). The aporia about the way in which
the sensible and the intelligible kalon differ, brings us to a sceptical
state of mind with regard to the latter, which despite the progress
that Plotinus makes in the treatise is not entirely resolved. It remains
unclear at the end what exactly the good is. Plotinus has come out
nevertheless with an answer regarding the nature of the kalon, an
answer that rests on a certain interpretation of Plato.
 3 Plotinus’ Two Senses of Aporia
I have said earlier that I distinguish between uses and senses of
aporia in Plotinus. So far I have been discussing one sense of aporia
that occurs in variations in Plotinus’ work. In all these cases where
Plotinus uses aporia as a difficulty or a problem that is raised
dialectically in order to promote his enquiry and more precisely in
order to determine clearly the subject matter in question, such as the
nature of soul or beauty; it may be a difficulty about the thought of
Plato or that of Aristotle and the Stoics, or one that pertains to
Plotinus’ own thought about the matter. As we have seen, the aporiai
raised for the most part are conditional on accepting a certain view
that Plotinus endorses, such as that sensible and intelligible Fs are
different, or that the world-soul governs the individual souls, or that
the soul cannot be affected, but they can also be unconditional,
when the difficulty or the puzzlement concerns a text of Plato and the
aporia then is how it should be interpreted. In all these cases the
aporia is a state of mind, a cognitive state. Plotinus’ aim is to resolve
that puzzlement. The way in which he proceeds, however, is by
seeking to determine what precisely causes the aporia. Plotinus is
often sceptical about whether and how an aporia can be resolved.
The fact that he often articulates an answer does not mean that he
considers an aporia completely settled; he often either remains non-
committal to any position, or he comes back to the same aporia,
raises it again and takes it up once more. Precisely this strategy
shows that he is not always convinced that the aporia can be
resolved completely; pursuing it systematically and carefully is the
best he can do.
We find, however, in Plotinus also another sense of aporia.
Aporia in this sense marks an impasse, a puzzle, which we are at a
loss even about how we should handle it. The aporia in this sense is
not a subjective cognitive state of mind, but rather objective, as it
concerns a topic we all find puzzling. This aporia is of rather limited
application in Plotinus, as far as I can see. We find it, for instance,
when Plotinus talks about the One or sometimes also the Intellect. In
Ennead VI.9.3, for instance, Plotinus voices his aporia about the One
as follows:

What then could the One be, and what nature could it have? There is
nothing surprising in its being difficult to say, when it is not even easy to
say what Being or Form is; but there is knowledge in us based upon the
Forms. As the soul, however, goes towards the formless, being utterly
unable to comprehend it because it is not delimited and, so to speak,
stamped by a richly varied stamp, it slides away and is afraid that it may
have nothing at all.
(trans. Armstrong, modified)

Here Plotinus identifies a puzzle for which he has reasons to believe

that it is difficult to be resolved given the object of the enquiry on the
one hand and human nature on the other. Plotinus does not
completely rule out the possibility of finding a way of making
progress on this issue, but he acknowledges the difficulties that
besets this enterprise and which have to do with the human mind in
general, not with an individual thinker, like Plotinus himself, for
instance. Plotinus goes on to highlight the cause of the difficulty of
the aporia. (Enn. VI.9.4.)

The perplexity arises especially because our understanding of that One is

not by way of reasoned knowledge or of intellection, as with other
intelligible things, but by way of a presence superior to knowledge. The
soul experiences its falling away from being one and is not at all one
when it has reasoned knowledge of anything; for reasoned knowledge is
a rational process, and a rational process is many. The soul therefore
goes past the One and falls into number and multiplicity.
(trans. Armstrong, modified)

Plotinus now points more precisely to the cause of the difficulty in

enquiring about the One. This lies in the fact that our human soul
cannot grasp such an entity by means of epistēmē and noēsis, both
of which indicate a grasp by means of reasoning that would involve
the use of concepts and syllogising. This is an objective difficulty, so
to speak, that has to do with the particular object of enquiry, the One;
since the One has no determinate features of itself, the aporia lies on
the side of the One itself, not on the side of the thinker about the
One. In this sense the aporia strikes us as insurmountable. This
does not mean that this is the end of Plotinus’ enquiry, however. He
goes on to elaborate on this aporia, by demarcating the two realms,
as it were, of human souls and of the cause of everything, the One.
But while he does this, Plotinus establishes more solidly what is the
cause of aporia. Later Platonists like Damascius elaborate on this
kind of aporia, which first surfaces with clarity in Plotinus’ work.37
Both in Plotinus and later in Damascius there is a sceptical thrust to
this aporia; we need to admit the limitations of our knowledge and
our cognitive capacities more generally. Plotinus keeps reminding us
of our limitations in understanding when we enquire into the first
principles.38 In this respect the impact of the objective aporia is not
unlike that of the subjective one; for both generate ultimately a
sceptical or at least a non-committal state of mind.
 Conclusion
To conclude, I have distinguished two senses of aporia in Plotinus
that we find also in earlier ancient philosophy, the subjective sense
of aporia as cognitive state of mind of an individual thinker, and the
objective aporia, namely a puzzle with features that suggest that it
cannot be resolved. I have also distinguished within the first sense of
aporia four uses or forms, the zetetic, exegetical, elenctic or critical,
and ontological. The distinction concerns only the use and scope of
aporia and should not be overemphasised as they are not mutually
exclusive but rather complementary; Plotinus uses all these aporiai
as a methodological tool that would move forward his philosophical
enquiry. All these forms of aporia are similar in that they aim to
express Plotinus’ spirit of enquiry against the received wisdom of
previous philosophers and sometimes Plotinus’ own earlier thought.
Plotinus’ aporiai mark one stage of the investigation, as is the case
with the Socratic aporia in Plato’s early dialogues, which aims to
motivate further philosophical research and also to structure future
enquiry, yet the difference is that Plotinus does seek answers, unlike
Socrates in the early dialogues. The ontological aporia in particular is
often invoked to play that role. Through it Plotinus eliminates what he
considers to be category mistakes and prepares the way for a
specific kind of enquiry, namely one that would mark the intelligible
item as the cause of the sensible one in a way that needs to be
specified. Plotinus is reasonably confident that such aporiai can be
resolved and the sceptical state of mind can be overcome. The new
epistemic state, however, is not secure or certain knowledge but
better knowledge compared with that of the beginning of the enquiry.
This is why Plotinus, like Plato, goes back to the same issues and
sets up a new investigation.
It is, therefore, a mistake to consider Plotinus as merely
dogmatic in an unqualified sense. Plotinus does have views or
doctrines, for which he argues forcefully and at length, that is,
doctrines he finds in Plato. But the question is through what method
he arrives at them and what their epistemic status is. My suggestion
is that Plotinus is aporetic both in the sense that he employs aporiai
as a way of enquiring into Plato as well as into the concepts he
investigates and in the sense of remaining often dissatisfied with the
answers he gives, which makes him come back to the same topics
again and again, although he, unlike Pyrrhonian sceptics, for
instance, does not think that there is no answer. In both senses of
‘aporetic’ Plotinus is borrowing an important methodological tool for
research from the Platonist tradition that also includes Aristotle. In
this regard Plotinus continues and also unifies this tradition that was
often split with regard to the status of Plato’s philosophy. Plotinus’
philosophy is a synthesis not only of the views occurring in different
parts of Plato or of previous Platonist interpretations, but also of
Platonist methodology, which essentially consists of raising,
addressing, but also solving aporiai. Plotinus’ work is, then, a
Platonist synthesis that is crucially and importantly inspired also by
the aporetic Plato.

I would like to thank the participants of the conference in Trinity College

Dublin on the Aporetic Tradition on Ancient Philosophy for their feedback on
the presentation of my essay. I am particularly grateful to Jonathan Craig,
Chris Noble and Damian Caluori for invaluable comments and suggestions,
which improved my essay significantly.

1 Armstrong 1967: 211.

2 See e.g. Gatti 1996: 27, who claims that ‘he [Plotinus] sought in Plato not
aporias but solutions; nor a method, but a doctrine’. John Rist 1967: 24
moves along the same lines. See also Bréhier 1968: 5, 7, who states (16)
‘De là vient que la doctrine de Plotin ne s’est pas développée partie par
partie dans une suite de traits, mais que, un peu à la manière de Leibniz, il
expose presque dans chaque traité sa doctrine tout entire sous le point de
vue particulier du sujet qu’il à examiner.’

3 See, for instance, O’Meara 1993: 6, Corrigan 2005: 100, and also the
earlier studies of Heinemann 1921: 243–8 and Volkmann-Schluck 1957: 1–
11.

4 See e.g. Gatti 1996: 28.

5 As is the case with Bréhier 1968: 17, for instance, who acknowledges the
presence of aporia in Plotinus’ work but treats it as a question posed by
others.

6 Charrue 1978: 28–30.

7 The opening statement of Enn. I.6 is full with echoes from Plato’s Hippias
Major and the Symposium. I will come to Ennead I.6 and the definition of
kalon in Section IV below.

8 For a commentary on this treatise, see Kalligas 2014: 131–48.

9 See Politis’ essay in this volume (Chapter 3).

10 To deal with this difficulty [aporian], therefore, we must go back and take
up the question of who that man in the intelligible world is. But perhaps we
should first say exactly who this man here below is – in case we go looking
for that man on the supposition that we have got this one, though we do not
even know this accurately. (Enn. VI.7.4.1–6, Armstrong tr.)

11 See the new translation with commentary of Enn. IV.3 by Dillon and
Blumenthal (2015).

12 The treatise is dated around 265, which means that it comes from
Plotinus’ middle period, when Porphyry was present in his school. See Dillon
and Blumenthal 2015: 18.

13 Metaphysics B, 995a27–31. See Buddensiek’s essay in the present

volume (Chapter 7).

14 Plotinus refers critically to such Platonists in Enn. IV.3.1.14–37.

15 On this topic see Emilsson 2005.

16 See the apparatus fontium of the Henry-Schwyzer editio minor ad loc.

17 One of which is the suggestion made in Aristotle’s De anima (413a9) that

the soul is like a captain on a ship, which Plotinus finds unsatisfactory; see
Enn. IV.3.21.10–21.

18 On this claim of Plotinus, see further Caluori 2015: Chapter 8.

19 On the role and the function of memory in Plotinus as discussed in Enn.

IV.3–4, see King 2009: 106–93.

20 The last part of Enn. IV.4 and Enn. IV.5 have recently been translated and
commented on by Gurtler 2015.

21 See for instance Enn. IV.3.20.36–51, where the Peripatetic view is said to
give rise to a series of difficulties (aporiai).

22 The aporiai examined in Chapters 9–18 and 19–24 correspond to

questions that Porphyry asked (VPlot. 13). See Dillon and Blumenthal 2015:
19.
23 On this idea about the general character of the soul see Caluori 2015,
and below.

24 See D’Ancona 2006, who does not sufficiently appreciate the aporetic
structure of the treatise in my view.

25 On the notion of the trace of the soul, see Noble 2013.

26 What is it that has carried out this investigation? Is it ‘we’ or the soul? It is
‘we’ but by the soul. And what do we mean by ‘by the soul’? Did ‘we’
investigate by having soul? No, but in so far as we are soul (Enn. I.1
[53].13.1–3).

27 On the issue of the systematicity und unity of Aristotle’s aporiai in

Metaphysics B see Buddensiek’s essay in this volume (Chapter 7).

28 This is illustrated by Politis’ essay in this volume (Chapter 3).

29 See esp. Enn. III.5 [50] 2.4–14.

30 See also Enn. IV.2 [28] 22.5–13.

31 See Enn. I.4 [46] 1.1–4. We find this view in the Magna Moralia 1206b30–
7.

32 Does well-being increase with time, though it is understood always to

refer to our present state? (Enn. I.5 [36] 1.1–2).

33 ‘Would someone say that the good for any thing is other than the full
natural activity of its life? And if the thing is made up of many parts, would
that activity then be the proper, natural, and never failing of the best part of
it?’ (Enn. I.7 [54] 1.1–4). Cf. Aristotle, NE 1098b14–16.

34 “Now I should give the reasons for which I have made the earlier
statements, namely in order to find out how actuality [to energeia[i]] applies
to intelligibles, and whether all are only in actuality, or whether each of them
is in actuality and everything is actual and also if potentiality applies there
[sc. intelligible realm] as well.” (Enn. II.5 [25] 3.1–4).
35 Lloyd 1990.

36 Concerning the kalon for instance, we find this enquiry pursued (apart
from Hippias Major) in the Phaedo (100d) and the Symposium (201c, 204e,
210a–12A) and in Plotinus (apart from Enn. I.6) in Enn. II.9.17.20–21, Enn.
VI.7.32.38 and in Enn. V.8. See further O’Meara 1993: 88–99.

37 On the aporia in Damascius see Caluori’s essay in this volume (Chapter

14).

38 Cf. Enn. V.1.12.1–3, V.3.13.1–8.

 Chapter 14
Aporia and the Limits of Reason and of
Language in Damascius
◈

Damian Caluori

Aporiai occur in dialectical contexts, often in contexts of enquiry, as

they paradigmatically do in the works of Plato and Aristotle.1 This
has been demonstrated in other contributions to this volume. Since
Damascius (c.460 CE to after 537 CE) was one of the greatest
dialecticians of late ancient Platonism, second perhaps only to his
famous third century predecessor Plotinus, it is not surprising that
aporiai also play a crucial role in his writings. His major work, Aporiai
and Solutions Concerning First Principles (De Principiis), is of
particular interest if we want to study the use Damascius makes of
aporiai.2 While ‘Aporiai and Solutions’ was a traditional genre at his
time, and while the term ‘aporia’ in such titles often meant nothing
more than problems quite generally, it often has a technical sense in
Damascius that is familiar from the tradition: It designates an
impasse that we find ourselves in if we have equally good reasons to
believe both a certain claim and its negation.
 I shall argue that Damascius’ use of aporiai in the technical
sense is new: Damascian aporiai are neither starting points of
enquiry as in Plato and Aristotle, nor do they serve any sceptical
purposes as in Pyrrhonism. Rather, their functions consist in
revealing to us certain crucial epistemic limitations and, at the same
time, in pointing to a reality that is beyond the grasp of language and
reason. To show this, I will discuss the first aporia of Damascius’ De
Principiis.
 1 The First Aporia
The attempt to explain the whole of reality in terms of ultimate
transcendent principles is a hallmark of ancient Platonism that can
already be found in Plato. In the Timaeus, the sensible world is
explained in terms of an intelligible world of Forms, and in the similes
of the Republic, the Good appears to be a principle even beyond the
intelligible world of Forms. Against this background, Platonists like
Plotinus and Proclus attempted to explain the whole of reality
ultimately by means of one first or highest principle. Damascius aims
at showing that such Platonist attempts lead into the following
aporia:

Is what is called the one principle of everything beyond everything or part

of everything, like the summit of those things that stem from it? And do
we say that everything is with it [i.e. with what is called the one principle
of everything] or after it and [stemming] from it? (Princ. 1, 4–7)3

As is reasonable within a Platonist context, the two questions

implicitly presuppose that there is something that is said to be the
principle of the whole of reality, purporting to account both for what
the whole of reality is and for its existence. Given this, Damascius
asks whether this principle4 belongs to the whole of reality (that is,
whether the principle itself is part of the whole of reality) or whether it
is beyond the whole of reality and thus not part of the whole of
reality. The two options are mutually exclusive if we assume that the
principle cannot be both part of the whole of reality and beyond the
whole of reality (for example, by being partly within the whole of
reality and partly beyond the whole of reality). We will see that
Damascius shares this assumption.5 Formally, De Principiis thus
starts by asking which one of an exclusive disjunction of two
propositions is true. What turns the initial question into an aporia is
the fact that Damascius in what follows adduces reasons against
(and thus for) either side that purport to be of equal strength.
First let us assume that the principle is outside the whole of
reality.6 The problem for this view is obvious: if the whole of reality is
indeed the whole of reality then its principle cannot be outside it. For
otherwise, either there would be something outside what is
considered the whole of reality. This would mean that what is
considered the whole of reality would not include reality in its
entirety. Or the principle would not really exist, since it would not be
part of reality. A further reason, perhaps a little less obvious and
more controversial, can be illustrated by a couple of examples that
Damascius presents. We call a whole city, he claims, the ruler and
the ruled, and we call a whole genus or clan the founder and his
descendants (Princ. 2, 6–8). In giving the second example, he may
have had Porphyry’s Introduction in mind where the genus of the
Heraclids is said to be so called from their relation to Heracles.7
Heracles is the principle of the genus of the Heraclids as well as a
part of it. Both examples are thus supposed to illustrate that the
specific sort of whole that Damascius is considering must include its
principle since the principle is constitutive of the sort of whole that is
in question. This is so because the relation of the principle to that of
which it is the principle is constitutive of the whole. In the case of
both the city (as Damascius conceives of it) and the genus or clan,
we cannot meaningfully call it a whole if we leave out what its
principle is: the Heraclids qua Heraclids are only what they are in
relation to Heracles and the ruled qua ruled are only so in relation to
the rulers: Neither the Heraclids nor the ruled are a whole without
Heracles and the rulers, respectively.
Now we may have our doubts about whether the same holds
true of the relation between the whole of reality and its principle, that
is, we may well doubt that the whole of reality is only what it is in
relation to its principle and thus that the whole of reality is a whole of
the same sort as a genus or clan (in the way discussed). Yet it is a
distinguishing mark of Platonism to explain things by postulating the
real existence of principles. Moreover, at least in later Platonism, the
productive relation between a principle and that of which it is the
principle is seen as essential to both the principle and that of which it
is the principle in the way the productive relation between Heracles
and his descendants is essential to the Heraclids.8
Up to now I have discussed Damascius’ view that, from a
Platonist point of view, there are strong reasons to believe that the
principle of the whole of reality is part of the whole of reality and
cannot be outside the whole of reality. I will now turn to his reasons
in favour of the opposite view, to wit that the principle cannot be part
of the whole of reality but rather must be outside, or beyond, the
whole of reality.9
If the principle is part of the whole of reality, then the following
dilemma unfolds: the whole of reality (including its principle) either
has no principle or we end up in an infinite regress. Neither horn of
the dilemma is acceptable. We end up in an infinite regress if we
postulate a second principle outside the whole of reality (including
the first principle), which then again, given our presupposition, has to
be part of reality so that we have to postulate a third principle for the
whole of reality (including the first and second principle) which would
then, given our presupposition, have to be part of the whole of reality
etc. ad infinitum.10 What about the other horn of the dilemma? Why
not say that the whole of reality (including its principle) has no
principle outside itself? Citing Aristotle, Physics III 203b6, Damascius
claims that everything either is a principle or has a principle.
According to the presupposition currently under discussion, the
whole of reality does not have a principle. Therefore, it must be a
principle. But there is nothing, Damascius claims, of which the whole
of reality could be a principle. But for something to be a principle,
there must be something of which it is the principle. Therefore, the
whole of reality neither is nor has a principle, in violation of the
Aristotelian claim that everything either must be or must have a
principle.
So far so good, perhaps. But why not reject Aristotle’s claim?
Perhaps Damascius accepts it for the following reason: if there are
things which neither have nor are principles then there are things for
which there are no explanations – brute facts, in other words. And
Platonists do not want to claim that the whole of reality (what it is and
that it is) is just a brute fact.11 A version of this conviction was still
held by Leibniz: ‘Outside the world, … there is some one being that
is ruling … It is higher than the world and, so to speak, outside it
(extramundanum), and is thus the ultimate reason of things (ultima
ratio rerum).’12
We can conclude that, according to Damascius, there are strong
reasons to believe that the principle of the whole of reality is outside
the whole of reality while the earlier arguments have given us strong
reasons to believe that the principle of the whole of reality cannot be
outside the whole of reality. The first aporia of De Principiis is thus
complete.
 2 The Sense of ‘Aporia’
Let us now look at the sense in which it is an aporia. Aristotle
famously claims that ‘equality of contrary reasoning (logismoi) would
seem to be the cause of aporia. For when we reason on both sides
and it appears to us that everything can come about in accordance
with either, we are in aporia as to which of the two to take up’ (Ar.
Top. 145b16–20). Aristotle uses the word ‘aporia’ here to refer to a
mental state that results from the fact that our reasoning on both
sides of a question appears to us to be of equal strength. We may
call the resulting aporia ‘subjective’ in the sense that it is a state of
mind – namely a puzzlement. After all, both sides appear to us to be
of equal strength. However, we may also call something an aporia in
the objective sense of a puzzle or an impasse.13 It seems to me that
a Platonist is not only in a state of puzzlement (aporia in the
subjective sense) when confronted with Damascius’ first aporia
(although she will be that, too) but also that she ought to be because
Damascius’ first aporia is indeed an impasse – an aporia in the
objective sense. And the right attitude of a rational being in view of
an objective aporia is being in a state of aporia (in the subjective
sense) unless he can solve or dissolve the puzzle. In this way aporiai
(in the objective sense) have normative force. This will be important
in our context. For Damascius not only wants to confront us with an
intellectual puzzle but also put us – as rational beings – into a state
of aporia (in the subjective sense). This will be crucial for his use of
aporiai. In order to see why, we will have to establish the function of
the first aporia. But first, more needs to be said about the item that ‘is
called the one principle of everything’.
 3 The Item That Is Called the One Principle of
Everything
What is the item that is said to have the function of being the one
principle of everything? His predecessors provide Damascius with
three options.
First, according to the dominant strand of the Platonist tradition,
represented by, for example, Plotinus and Proclus, the one, the
subject matter of the first hypothesis of the second part of Plato’s
Parmenides, is the first principle of reality.14 This one becomes the
first principle by being identified with the Good beyond being from
Republic 509b.15 Having identified the One as a candidate for the
role of what is called the one principle of everything, we may ask: is
the One part of the whole of reality or is it beyond the whole of
reality?16 According to the arguments discussed above, since it is
the principle of the whole of reality, it would be part of the whole of
reality (in the same way that the ruler is part of the city). But if the
One is a principle inherent in the whole of reality, ‘then our reason is
looking for a further principle’, Damascius claims, that is prior to the
whole of reality.17
The second option is perhaps Porphyry’s.18 According to this
option, the One is both prior to the whole of reality and part of it in
such a way that the One is in itself prior to the whole of reality but,
qua principle, a part of the whole of reality.19 Damascius rejects this
proposal because, as he argues, it would introduce a duality into the
One (as principle and as non-principle), which is incompatible with its
simplicity as deduced from the first hypothesis of Plato’s
Parmenides.
After having rejected also this option, Damascius states:
‘Therefore, our soul divines (manteuetai) that there is a principle of
the whole of reality, in whatever way conceived, which is beyond the
whole of reality and unrelated to it.’20 Now we will have the greatest
difficulties conceiving of such a principle. As Damascius’ earlier
considerations have shown, it cannot be a principle at all because, if
it is a principle, it is as such related to the things whose principle it is
– a difficulty that Damascius has shown to lead to an infinite regress
or to the (unacceptable) postulation of a brute fact. Damascius is
well aware of this difficulty and accordingly continues immediately
after the above quotation in the following way, correcting the
divination of the soul mentioned in the above quotation: ‘Therefore, it
is not correct to call it a principle or a cause or first, or prior to the
whole of reality or beyond the whole of reality.’21 It is not correct to
call it such because each of these designations implies that it is
related to something else, namely to something whose principle or
cause it is, to something in relation to which it is prior or to something
beyond which it is.22
It is crucial for us to remind ourselves here that Damascius,
when presenting the first aporia in the initial two questions, spoke of
‘what is called the one principle of everything’ and not simply of the
one principle. The item we are looking for is called a principle
without, as it turns out now, being one.
This leads us to the third option, presumably originating in
Iamblichus.23 According to this option, there is an item even beyond
the One. Because of its utter independence of everything else,
scholars have sometimes called it the Absolute.24 Damascius
himself, for reasons we will come back to, prefers to call it the
Ineffable.
It is important to note at this point that Damascius’ distinction
between the One and the Ineffable does not solve the first aporia.
The aporia could only be solved if we found something that is a
principle and at the same time external and unrelated to the whole of
reality. But the first aporia shows precisely that we cannot
consistently postulate the existence of such an item. And indeed,
Damascius introduces the Ineffable explicitly as something that is not
a principle. Hence, the first aporia has not been solved. For this
reason, the question of what the function of the first aporia is
becomes all the more pressing. However, introducing the Ineffable
also provides us with a further problem, namely that we (being part
of the whole of reality) cannot properly speaking relate to it. Thus,
there is a question about our cognitive stance towards it. Let us look
at this further problem first.
 4 Plato, the One, and the Ineffable
The background of Damascius’ contention that we cannot relate to
the Ineffable can be found in Plato’s Parmenides and Sophist. I have
claimed above that Platonists identified the One with the one that is
the subject matter of the first hypothesis of the second part of Plato’s
Parmenides. According to the Parmenides, there is no name, no
logos, no knowledge nor perception and not even a belief of it and
that therefore it is not named, nor spoken of, nor the object of belief
or knowledge or perception (Pl. Parm. 142a3–7). The reason given
in the Parmenides is that the one does not participate in being
(ousia) and that it therefore in no way is (Parm. 141e9–2a1). The
idea seems to be that nothing can be predicated of or attributed to
the one (Parm. 141e10–2a3). For every true positive predication
implies that something applies to the one, which in turn implies that
the one participates in being (i.e. that it is what is predicated of it).
Since, if the one does not participate in being, there is nothing which
applies to the one (i.e. there is nothing that the one is), we cannot
truly say of anything that it applies to the one. Thus, we cannot truly
say anything positive of the one. But for the same reason, we cannot
believe or know or perceive anything of the one.
Since these claims also apply to Damascius’ Ineffable, some
scholars are of the view that Damascius believes the first hypothesis
to be about the Ineffable.25 If so, this passage would immediately
give us the background to Damascius’ view. But even if Damascius
thinks that the first hypothesis is about the One (as distinct from the
Ineffable), what we read in this passage is a fortiori true of the
Ineffable (as it were) if it is true of the One. Either way, Damascius
can infer from this passage that we cannot have any positive
cognitive attitude towards the Ineffable.26
Damascius also refers to Sophist 236d–41b,27 where the Eleatic
visitor and Theaetetus discuss non-being (to mēdamōs on (Soph.
237b7–8)). At Sophist 238c9–11, the Eleatic visitor concludes: ‘Do
you understand, then, that it is impossible to rightly utter or say or
think [anything] of non-being as such but that it [i.e. non-being] is
inconceivable (adianoēton), ineffable (arrhēton), unspeakable and
without logos (alogon)?’ This is the conclusion of an argument that
shows (or purports to show) that no expression can be used to refer
to non-beings, not even the expressions ‘non-being’ and ‘non-
beings’. For if we say the former of non-being, we apply unity and if
the latter, plurality. Both, the attribution of unity and of plurality, imply
being. By applying unity or plurality to non-being, we would therefore
also apply being to it, which is impossible. This argument is similar to
the one we have found in the Parmenides. It is relevant for our
purposes that it makes explicit that we can say of non-being not
even that it is one. Damascius takes this passage to be about the
Ineffable (which is not even one) even though, of course, the
argument from the Parmenides also implies that the one discussed
there is not one because, if it is one, it is something, namely one.28
 5 Silence, Ignorance and Speaking about the
Ineffable
These considerations lead the Eleatic visitor and Theaetetus to the
conclusion that we cannot say anything (ti) of what in no way is and
that, if someone cannot say anything, ‘he must quite necessarily say
absolutely nothing’ (Soph. 237e1–2). He must be silent with regard
to what is not, a silence that plays a prominent role in Damascius’
explorations of the Ineffable as well.
Damascius’ use of silence is often depicted as religiously
motivated, and the language he uses can certainly be interpreted in
this way. I do not necessarily wish to deny this. But in addition to the
reasons for silence inferred from the Sophist, I want to add that
silence also has a role to play in ancient dialectic. In dialectical
questioning the respondent is allowed to answer only with ‘yes’ and
‘no’; any other answer would break the rule of the dialectical
exercise.29 However, there are situations in which neither answer is
adequate, for example, if a conclusion is reached fallaciously. Gellius
gives the following sophistical example: “Do you, or do you not, have
what you have not lost? I demand that you answer “yes” or “no””.
Whichever way he briefly replies, he will be caught … Therefore, this
proviso also is commonly added to the rule, that one need not
answer catch-questions’ (Gell. NA 16.2.9–13).30 The Stoics used a
related additional dialectical rule famously when confronted with
Sorites-type arguments.31 While Sorites-questions are not
sophistical, they share with sophistical questions like the one
discussed by Gellius, that there are questions to which neither ‘yes’
nor ‘no’ is a rational answer. It is clear, for example, that one grain of
corn does not make a heap and thus that the correct answer to the
question ‘Is one grain of corn a heap?’ is ‘no’. It is equally clear that
a large number of grains (if piled up) do make a heap and that,
accordingly, the correct answer to the question ‘Are 10.000 grains of
corn a heap?’ is ‘yes’. Yet it is unclear at which number the
respondent should switch from ‘no’ to ‘yes’ when the questioner
increases the number of corns by one per question.32 Ancient
dialecticians, therefore, recommended falling silent before a critical
point has been reached, that is, as long as the answer is still clear.
Cicero, for example, reports that Chrysippus stops ‘some time before
coming to many’ (Cic. Acad. 2.93). Relatedly, we also learn that we
should stop as long as the answer is still clear and before we arrive
at what is obscure (Cic. Acad. 2.94). Moreover, the Stoics declare,
according to Sextus Empiricus, that, in the case of a Sorites, we
ought to stop ‘while the argument is proceeding’ (S.E. PH 2.253).
Determining the precise moment to stop is, of course, no easy
matter either, but this need not concern us. Important for us is the
dialectical appropriateness of this silence when we arrive at a point
where we are not rationally justified to answer ‘yes’ or ‘no’ (or
perhaps even before that point) and thus when reason arrives at
something it cannot decide – an impasse to which there is no
rational solution.
The first aporia in Damascius points us to the Ineffable but when
reason attempts to grasp the Ineffable, it reaches a point where
silence is dialectically appropriate because reason understands that
there is nothing it can say or think about the Ineffable. I thus suggest
that Damascius considers silence to be the correct rational attitude
towards the Ineffable, not only because of the awe-inspiring
greatness of the intended object (as it were), but also due to the
aporia (in the subjective sense) that the thinker rightly experiences
when attempting to grasp that object.
If this is indeed what Damascius thinks then we may wonder
why he has written the part of De Principiis about the Ineffable at all.
For the required silence would certainly also extend to writing.
Damascius is well aware of this problem: ‘But the Ineffable must be
honoured with complete silence and, even prior to this, with
complete ignorance (agnoia) … But let us examine this second point,
in what way it is said to be completely non-cognisable (agnōston); for
if this is true, how can we, dealing with it [i.e. the Ineffable], write all
these things?’ (Princ. 11, 15–19). According to this passage, there is
a problem of ignorance prior to the problem of silence. It seems clear
that we must be silent about the Ineffable because there is no way
for us to cognitively grasp it. But the claim that the Ineffable is non-
cognisable leads Damascius to a further problem: do we know that
the Ineffable is non-cognisable or not? If not, then we are not
justified in claiming that it is non-cognisable. For in this case, we do
not know whether or not it is non-cognisable. However, if we know
that it is non-cognisable, then, it appears, we know something about
it, namely that it is non-cognisable (Princ. 12, 3–6). But if so, then it
is not absolutely non-cognisable. Thus, either we know that it is non-
cognisable or not. If not, we have no reason to claim that it is non-
cognisable. If yes, we have no reason to claim that it is non-
cognisable. In this way, it appears that we cannot rightly say that it is
non-cognisable in spite of our earlier arguments.
To further explain this problem, Damascius refers to a passage
from the Theaetetus where Socrates and Theaetetus discuss the
possibility of false belief (pseudē doxa). At Theaetetus 188c2–3
Socrates claims: ‘But certainly no one somehow believes that what
he knows is that which he does not know nor [does he believe] that
what he does not know is what he knows.’ And Theaetetus agrees
finding such a belief quite incredible (teras). Damascius interprets
this as saying: ‘For no one will say that what he knows is or is not
what he does not know’ (Princ. 12, 9–10). We will not believe not
only that what we know is what we do not know but also that what
we know is not what we do not know. Damascius’ problem is this:
how can we deny something we know of something we do not
know? Damascius thus doubts, for example, that we can know that
something of which we know nothing does not have a certain
property we know. Accordingly, he asks: ‘how, then, can we deny
what we know in some way of what we do not know in any way?’
(Princ. 12, 11–13). More specifically: how can we claim that
something of which we know nothing, is not knowable? In order to
illustrate this problem, Damascius makes the following comparison:
A person born blind has no knowledge whatsoever of colour. Now
she claims that colour is not hot on the ground that she knows heat
by means of the sense of touch. She concludes that colour is not
touchable. Yet if so, she now seems to know something about
colour, namely that it is not touchable. Damascius rejects this
conclusion: when the person born blind makes the above claim, she
does not, contrary to first impressions, make a claim about colour at
all but rather about her own ignorance (Princ. 12, 13–19). By saying
that colour is not touchable, the person born blind does not speak
about colour at all, in Damascius’ interpretation, but rather about her
incapacity to cognitively access colour: she cannot touch it. The
person born blind cannot know that colour possesses or lacks
certain properties because colour is cognitively totally inaccessible to
her. Accordingly, for the person born blind to say that colour is not
hot is not a claim about colour’s lacking a certain property but a
confession of this person’s ignorance.
Similarly, when we say that the Ineffable is non-cognisable, we
do not deny of it that it is cognisable, that is, we do not claim that it
lacks the property of being cognisable. This would be impossible for
us because we do not know the Ineffable in any way and thus do not
know any of its properties. For any property, we do not know whether
the Ineffable lacks it. What it means for us to say ‘the Ineffable is not
cognisable’ is that we do not know the Ineffable. Thus, by saying
this, we speak about ourselves: ‘And in particular if we say that it is
not cognisable, we do not report any of its properties (ouk autou ti)
but rather concede our own affection (pathos) in relation to it’ (Princ.
12, 19–21). This solves the above problem: We can truly say that the
Ineffable is not cognisable without attributing to, or denying of it, any
property and thus without being committed to the claim that it is in
some way cognisable.
Now this may all be very well, but why should we believe that
beliefs dealing with the Ineffable only concern our own affections
(pathē) instead of the Ineffable itself? Why should we believe, for
example, that the belief ‘The ineffable is ineffable’ is about ourselves
instead of being about what clearly appears to be what the belief is
about, namely the Ineffable?
 6 The Reversal and a Limit of Reason and
Language
In a discussion of an opponent who claims that, surely, by saying all
those things about the Ineffable, we have beliefs about the Ineffable,
at least the belief that it is not an object of belief (adoxaston),
Damascius answers: ‘The logos incurs reversal [peritrepetai], as he
[Plato] says, and in truth we do not have any belief [about it]’ (Princ.
I.16.5–6).33
Damascius uses the notion of a reversal or turning around
(peritropē) quite a few times, and there is hardly any doubt that he
has inherited it from the Pyrrhonian sceptics, as Rappe has shown in
her classic paper on this issue.34 Note that we even find the whole
expression that Damascius uses, namely ‘the logos incurs reversal’,
in Sextus Empiricus (PH 2.88; 2.91). Reversal is a technical term for
self-refutation.35 In his book on this topic, Castagnoli distinguishes
different sorts of self-refutation in ancient dialectic. He calls the sort
of self-refutation that Damascius in our context is using ‘operational
self-refutation’ (a term he borrows from Mackie) and defines it thus:
‘A proposition p is operationally self-refuting when, although it could
be true, there is no way of coherently presenting it, since the very act
of asserting p also entails a commitment to something else which is
in conflict with p, and thus to a contradiction’ (Castagnoli 2010: 205).
This nicely fits with Damascius’ discussion of the Ineffable.36 When
saying anything, positively or negatively, about the Ineffable, we
thereby commit ourselves to a claim that contradicts our initial claim.
For example, if we say ‘the Ineffable is ineffable’ then, by saying this,
we commit ourselves to the view that we can say something about
the Ineffable and this contradicts the content of what we say.
Generalising this point, it means that it is not possible for us to
coherently say anything about the Ineffable.
Instead, the reversal leads us back to ourselves, as in particular
Hoffmann (1997) has emphasised. When attempting to discuss and
to think about the Ineffable, we are actually speaking and thinking
about ourselves. Damascius follows the Pyrrhonists – at least
terminologically – when claiming that, at the occasion of a reversal,
we speak about our own affections (pathē).37
The reversal not only reveals our own state of ignorance but
also the limits of reason: ‘But this complete reversal of our logoi and
our thoughts is the proof envisioned by us of what we are saying.
And what will the limit of logos [language, reason] be but a silence
without means38 and the confession that one knows nothing of those
things of which it is not right to acquire knowledge since they are
inaccessible?’ (Princ. 21, 18–22). We have already discussed the
silence appropriate if one is in a state of aporia (in the subjective
sense). It is easy to see how silence is appropriate when the limit of
logos has been reached if we understand logos as speech. However,
on the basis of the results just achieved, we can say a little more
about it. The limit of logos is also a limit of reason, reached when
reason rightly finds itself in a state of aporia (in the subjective sense)
when dealing with an aporia (in the objective sense) that cannot be
rationally solved. Accordingly, the ignorance, which is our affection
when confronted with an aporia of this sort, is an ignorance that is
not due to happenstance (because we happen not to know this or
that) but an ignorance that reason necessarily finds itself in when
faced by an aporia of this type.
Note, however, that the fact that we cannot coherently speak
about the Ineffable does not imply that the Ineffable does not exist or
that it is not, as a matter of fact, ineffable. It only means that we
cannot coherently say or think or believe that it is. In other words, it
shows that there is a cognitive gap between us and the Ineffable and
that there may be metaphysical truths that are not rationally
accessible to us.39 According to Damascius, there are indeed such
metaphysical truths as can be seen from a passage where he
considers someone who believes that there is nothing beyond the
One and who therefore denies the existence of the Ineffable.
Damascius states: ‘If someone speaks in this way [namely saying
that there is nothing beyond the One], we will forgive him his
puzzlement (aporia) … but, starting from what is better known to us,
we must excite the ineffable pangs of labour in us towards an
ineffable (I do not know how to say) awareness (sunaisthēsis) of this
sublime truth’ (Princ. 6, 11–16). Damascius is hesitant in naming the
attitude towards this truth and does not know how to call it – for
reasons that are by now familiar. But he refers to a sublime truth that
is not accessible to us.
This shows, what should be uncontroversial anyway, that
Damascius is not a sceptic even though he uses sceptical terms and
argument forms. Rather, he is a dogmatist, albeit a subtle one. For,
metaphysically speaking, he does not doubt the existence of the
Ineffable. It is just that we cannot coherently speak and think about
it. The limits of logos may thus be the limits of my world, to adapt a
famous phrase by Wittgenstein (if we understand ‘my world’ as the
world in so far as it is rationally accessible to me), but not the limits
of the world. For Damascius, there exist things beyond the reach of
reason and language.40
Two more problems need to be addressed. What are the pangs
of labour that he is referring to in the above quotation? And how is it
possible for Damascius to be so certain that there are such rationally
inaccessible truths given that they are rationally inaccessible?
Should not our conclusion be that the insoluble aporia shows that
there are no such truths?
 7 Pangs of Labour, Aporiai, and Experience
Damascius, as we have seen in the passage cited above, states that
‘we must excite the ineffable pangs of labour (ōdines) in us towards
an ineffable (I do not know how to say) awareness of this sublime
truth’.41 What are these pangs of labour and why must we excite
them?
The image of pangs of labour here used originates from the
Theaetetus where Socrates’ young interlocutor states: ‘But be
assured, Socrates, that I have often tried to enquire (skepsasthai)
into this, when I heard the reports of the questions that you asked.
But I cannot convince myself that I have any sufficient answer, nor
can I find anyone else giving the kind of answer on which you insist;
and yet I cannot stop caring for it’ (Tht. 148e1–5). Socrates explains
that what Theaetetus experiences are pangs of labour and later
describes exciting pangs of labour as a skill of midwives (Tht.
149c8–d3). Socrates famously describes his own skill as midwifery
concerned with souls. What is the state Theaetetus is in? He has
questions that concern him and that he deeply cares and worries
about without, however, finding any sufficient answer; in other words,
he is in a state of aporia. The relation of pangs of labour and aporiai
is explicitly established at Theaetetus 151a6–7 where Socrates tells
us: ‘Now those who associate with me experience in this matter the
same as women in childbirth. They suffer pangs of labor (ōdinousi)
and are full of aporiai night and day and much more so than the
women. And my skill can excite these labour pangs and cause them
to cease.’
People thus experience pangs of labour in their souls when they
are in a state of aporia and Socrates’ maieutic skill can excite those
pangs. When Damascius thus says that we ‘must excite’ those
labour pangs, he not only puts himself into the tradition of Socratic
maieutic but also means that we must engage dialectically with
aporiai, an example of which we have been studying in this paper.
But there remains the question of why we must engage with aporiai
that not only cannot be solved but also engagement with which
subjects us to self-refutation. The answer to this question, it seems
to me, is also the answer to the question of what the use of
Damascius’ first aporia is. I suggest that it actually has two related
functions.
When thinking about this aporia (in the objective sense) then,
Damascius argues, we are brought into a state of aporia (in the
subjective sense) and experience the labour pangs that midwife
Damascius (or the reader’s internal midwife) excites in us. I have
argued that there is no solution to this aporia. Hence, there is no
birth (not even of wind-eggs), no progress, no enquiry that would
lead to knowledge. However, what the initial aporia purports to give
us is an awareness of the Ineffable as Damascius claimed in the
above quotation. This awareness is neither knowledge nor belief.
Thus, the aporia points to something beyond the reach of reason
and language. This pointing, I suggest, is the first function of the first
aporia.
 But how can the aporia point to the Ineffable by making us
aware of it, and what does this awareness consist in if it is not a
rational state at all (‘rational’ in the sense of being a state in which
we grasp something by means of reason)? I suggest that it is a state
of experience. By properly engaging with the aporia, we experience
a mental state of awareness that is not rational. This does not imply
that this state is irrational. Indeed, there are many experiences that
are accessible to us only first-personally, e.g. the experience of
colour, and for which it may be hard or even impossible to give a
rational account from a third person point of view, as Nagel has
famously argued.42 This is not a sign of epistemic deficiency, it
seems to me, as long as these experiences are in principle also
available to other people. Whether there are such experiences with
regard to the Ineffable is an empirical question. Their occurrence
perhaps seemed more plausible in the cultural context of late
antiquity than it does nowadays.
No matter whether or not the aporia provides us with some such
awareness, as Damascius claims it does, there is still no solution to
it: the Ineffable itself remains rationally inaccessible. However,
whenever we try to grasp it, we are subject to a reversal. This leads
us to the second function of the first aporia: it shows us a limit of
reason. By showing us a limit of reason, it helps us to understand
what we are as rational beings. It thus also helps us to understand
ourselves and thus to follow the Delphic command.

1 I would like to thank George Karamanolis, John Dillon, Regina Füchslin,

and Peter Turner for their helpful comments on earlier drafts of this paper.
2 Important studies of De Principiis include: Westerink and Combès 1986
Introduction; Galpérine 1987 Introduction; Dillon 1996; Hoffmann 1997;
Rappe 2000; Ahbel-Rappe 2010 (an English translation of De Principiis);
O’Meara 2013.

3 All references to De Principiis are to the edition of Westerink and Combès

1986. I refer to it by page and line number and, unless otherwise indicated,
to the first volume. Translations are my own.

4 The qualification ‘what is called’ that we find in the quotation will become
important later in the paper. To simplify the current discussion, I will here
merely use expressions like ‘the principle’ or ‘this principle’ without
qualification.

5 He does so, roughly, because he thinks that the first principle, being
completely simple, must be partless. Compare Plato, Soph. 245a8–9 where
the Eleatic visitor claims that the one must be without parts.

6 Princ. 1, 7– 2, 8.

7 Porph., Intr. 1, 17–22 Busse. See also Plot. Enn. VI 1, 3, 3–4.

8 See e.g. Procl. ET Prop. 7.

9 Princ. 2, 9–20.

10 This is structurally similar to the famous Third Human Being argument.

11 Note that this presupposes that even if every part of the whole of reality
either is or has a principle, we still need, Damascius argues, a further
explanation (and thus principle) for the whole of reality as a whole.

12 Leibniz, De rerum originatione radicali, vol. VII p. 302 Gerhardt.

13 For the distinction between puzzle and puzzlement see Matthews 1999a:
29–30 and Politis 2006: 90.
14 See Hoffmann 1997; O’Meara 2013: 188–93.

15 Dodds 1928.

16 As is customary, I will write ‘One’ (with a capital ‘O’) when referring to the
thus called principle.

17 Princ. 3, 21–25.

18 Proclus, In Parm. 1106, 31ff; Dillon 1996.

19 Princ. 4, 3–5.

20 Princ. 4, 13–15.

21 Princ. 4, 15–16.

22 As he later emphasises, even calling it transcendent is incorrect. For

even what is transcendent transcends something and is thus related to it.
See Princ. 21, 7–8.

23 See Princ. II,1,4–8.

24 Galpérine 1987: 167; O’Meara 2013.

25 Westerink and Combès 1986: LXXI and Combès 1989.

26 Perhaps the two interpretations can be reconciled. Damascius states in

the same context: ‘Why should we look for something beyond the One? But
perhaps Plato has made us ineffably ascend by means of the One to what is
our subject now, namely the Ineffable beyond the One … ’ (Princ. 9, 9–13).
Indeed, he suggests, with reference to the passage just cited, that Plato may
have arrived at the Ineffable by denying of the One that it is one.
Accordingly, when Plato states that the One is not one, he hints, Damascius
thinks, at the Ineffable.
27 Damascius refers to this passage reluctantly because he believes that it
deals, not with the Ineffable, but rather with non-being at a lower level that
he also finds described in the seventh hypothesis of the Parmenides.
However, at In Parm. IV 115, 12–116, 8 Damascius speaks about two ways
of non-being, one concerning the Ineffable, the other one being the subject
matter of the seventh hypothesis. Accordingly, he believes that the argument
in our passage from the Sophist also applies to the Ineffable. See Princ. 7,
24–25 and Princ. 8, 3–5.

28 This is clearly a problem for all defenders of the existence of the One.

29 Ar. SE 17, 175b7–14; Gell. NA 16.2.1–3; Bobzien 2002.

30 The title of this chapter in NA is: The nature of the rule of the logicians in
disputation and declamation, and the defect of that rule.

31 For the following see Bobzien 2002.

32 See e.g. S.E. M 1.68–70; Gal. Med.exp. chs. XVII and XX Frede and
Walzer.

33 Damascius seems to ascribe the expression ‘reversal’ to Plato even

though Plato nowhere uses it in a technical sense. However, Plato arguably
does use this sort of argument, for which see Castagnoli 2010: 205–47. I
argue in Caluori (forthcoming) that Damascius here specifically refers to
Soph. 238d–9b.

34 Rappe 1998.

35 The seminal paper for ancient self-refutation is Burnyeat 1976. See now
also the detailed study Castagnoli 2010.

36 See also Castagnoli 2010: 244–7.

37 E.g. D.L. 9.103; S.E. PH 1.193. Depending on how the Pyrrhonists use
the term ‘affection’, however, Damascius’ use may be different.
38 The term amēchanos that I translate here as ‘without means’ is also used
by the sceptics when they describe a state of being unable to either assent
or deny – a state that Sextus calls aporetic (S.E. PH I 7).

39 I do not mean to say that they are non-rationally accessible to us. I just
mean that by means of reason, we are unable to grasp the Ineffable and that
this fact reveals a limit of reason.

40 Again, this is a metaphysical claim. And it is not as outlandish as it may

appear at first sight. The reaction of many people when confronted with the
fact (if it is one) that, if we are brains in a vat, we cannot truly say that we
are brains in a vat, for example, is not that, therefore, we are not brains in a
vat. Rather, these people think: “Perhaps we are, even though we cannot
truly say or think that we are.’ This reaction may or may not be wrong but it
is in any case perfectly reasonable. As Nagel puts it: ‘ … Instead I must say
‘Perhaps I can’t even think the truth about what I am, because I lack the
necessary concepts and my circumstances make it impossible for me to
acquire them!” Nagel 1986: 73.

41 For pangs of labour, see also Princ. 8, 12–20. For more on pangs of
labour in Damascius, see Caluori (forthcoming).

42 Nagel 1974.
 Bibliography

Allen, J. (1994) ‘Academic Probabilism and Stoic Epistemology’, Classical

Quarterly 44: 85–113.

Allen, J. (1997) ‘Carneadean Argument in Cicero’s Academic Books’, in B.

Inwood and J. Mansfeld (eds.), Assent and Argument: Studies in Cicero’s
Academic Books, Leiden: Brill, 217–56.

Allen, R. E. (1997) Plato’s Parmenides, New Haven: Yale University Press.

Anderson, A. (1969) ‘Socratic Reasoning in the “Euthyphro”’, Review of

Metaphysics 22: 461–81.

Annas, J. (1992) ‘Plato the Sceptic’, Oxford Studies in Ancient Philosophy:

Methods of Interpreting Plato and His Dialogues, Supplementary Volume: 43–
72.

Annas, J. (2011) ‘Ancient Scepticism and Ancient Religion’, in B. Morrison and

K. Ierodiakonou (eds.), Episteme, etc. Essays in Honour of Jonathan Barnes,
Oxford: Oxford University Press, 75–90.

Annas, J. and Barnes, J. (1985) The Modes of Scepticism: Ancient Texts and
Modern Interpretations, Cambridge: Cambridge University Press.

Annas, J. and Barnes, J. (2000) Sextus Empiricus: Outlines of Scepticism, 2nd

ed., Cambridge: Cambridge University Press, 195–264.

Armstrong, H. (1967) ‘Plotinus’, in H. Armstrong (ed.), The Cambridge History

of Later Greek and Early Medieval Philosophy, Cambridge: Cambridge
University Press.

Aubenque, P. (1961a) ‘Aristoteles und das Problem der Metaphysik’, Zeitschrift

für philosophische Forschung 15: 321–33.
Aubenque, P. (1961b) ‘Sur la notion aristotélicienne d’aporie’, in S. Mansion
(ed.), Aristote et les problèmes de méthode, Louvain, Publications
Universitaires, 3–19.

Aubenque, P. (1966) Le problème de l’être chez Aristote. Essai sur la

problématique aristotélicienne, 2nd ed., Paris: Presses Universitaires de
France.

Babut, D. (1994) ‘Du scepticisme au dépassement de la raison: Philosophie et

foi réligieuse chez Plutarque’, in Parerga: choix d’articles de Daniel Babut
(1974–1994), Lyon: Maison de l’Orient Mediterranéen, 549–84.

Babut, D. (2007) ‘L’unité de l’Académie selon Plutarque. Notes en marge d’un

débat ancien et toujours actuel’, in M. Bonazzi, C. Lévy and C. Steel (eds.), A
Platonic Pythagoras: Platonism and Pythagoreanism in the Imperial Age,
Turnhout: Brepols, 63–98.

Ballériaux, O. (2001) ‘Platon, Sophiste’, in A. Motte and C. Rutten (eds.),

Aporia dans la Philosophie grecque. Des origines à Aristote, Louvain-la-
Neuve: Peeters, 81–91.

Barnes, J. (1981) ‘Proof and the Syllogism’, in E. Berti (ed.), Aristotle on

Science: The ‘Posterior Analytics’, Padua: Antenore, 17–59.

Barnes, J. (1982a) The Presocratic Philosophers, London: Routledge & Kegan

Paul.

Barnes, J. (1982b): ‘Medicine, Experience and Logic’, in J. Barnes, J.

Brunschwig, M. F. Burnyeat and M. Schofield (eds.), Science and Speculation,
Cambridge: Cambridge University Press, 24–68.

Barnes, J. (1983) ‘Aphorism and Argument’, in K. Robb (ed.), Language and

Thought in Early Greek Philosophy, La Salle, IL: Monist Library of Philosophy,
91–109.
Barnes, J. (1989) ‘Antiochus of Ascalon’, in M. Griffin and J. Barnes (eds.),
Philosophia Togata: Essays on Philosophy and Roman Society, vol. I, Oxford:
Clarendon Press, 51–96.

Barnes, J. (1990) The Toils of Scepticism, Cambridge: Cambridge University

Press.

Barnes, J. (1991/2011) ‘Philosophy and Dialectic’, (French version), in M. A.

Sinaceur (ed.), Penser avec Aristote, Toulouse: Édition Ères, 107–16.

Barnes, J. (1992) ‘Diogenes Laertius IX 61–116: The Philosophy of

Pyrrhonism’, in W. Hasse (ed.), Philosophie, Wissenschaften, Technik.
Philosophie, Berlin: De Gruyter, 4241–301.

Barnes, J. (2007) ‘Sextan Scepticism’, in D. Scott (ed.), Maieusis, Oxford:

Oxford University Press, 322–34.

Bell, I. H. (2004) Metaphysics as an Aristotelian Science, Sankt Augustin:

Academia Verlag.

Benson, H. H. (1992) ‘Why Is There a Discussion of False Belief in the

Theaetetus?’, Journal of the History of Philosophy 30: 171–99.

Benson, H. H. (2000) Socratic Wisdom: The Model of Knowledge in Plato’s

Early Dialogues, Oxford: Oxford University Press.

Benson, H. H. (2006) ‘Plato’s Method of Dialectic’, in H. Benson (ed.),

Blackwell Companion to Plato Oxford: Blackwell, 85–99.

Berti, E. (1997) ‘Philosophie, dialectique et sophistique dans Metaphysique 2’,

Revue Internationale de Philosophie 51: 379–96.

Berti, E. (2005) ‘Aristotelian Endoxa as a Basis for a Modern

Reasonableness’, in R. Dottori (ed.), Reason and Reasonableness, Münster:
LIT Verlag, 177–90.
Bett, R. (1997) Sextus Empiricus: Against the Ethicists, Oxford: Clarendon
Press.

Bett, R. (2000) Pyrrho, His Antecedents, and His Legacy, Oxford: Oxford
University Press.

Bett, R. (2009) ‘Sextus Empiricus’, in N. Trakakis and G. Oppy (eds.), History

of Western Philosophy and Religion, vol. I, Oxford: Oxford University Press,
173–85.

Bett, R. (2015) ‘God: M 9.13–194’, in K. Algra and K. Ieorodiakonou (eds.),

Sextus Empiricus and Ancient Physics, Cambridge: Cambridge University
Press, 33–73.

Beversluis, J. (2000) Cross-Examining Socrates: A Defense of the

Interlocutors in Plato’s Early Dialogues, Cambridge: Cambridge University
Press.

Binder, G. and Liesenborghs, L. (1976) ‘Eine Zuweisung der Sentenz οὐκ ἔστιν
ἀντιλέγειν an Prodikos von Keos’, in C. J. Classen (ed.), Sophistik, Darmstadt:
Wissenschaftliche Buchgesellschaft, 452–62.

Bobzien, S. (2002) ‘Chrysippus and the Epistemic Theory of Vagueness’,

Proceedings of the Aristotelian Society 102: 217–38.

Bolton, R. (1990) ‘The Epistemological Basis of Aristotelian Dialectic’, in D.

Devereux and P. Pellegrin (eds.), Biologie, logique et métaphysique chez
Aristote, Paris: Editions du CNRS, 185–236.

Bonazzi, M. (2003) Academici e Platonici. Il dibattito antico sullo scetticismo di

Platone, Milan: LED.

Bonelli, M. (2001) Alessandro di Afrodisia e la metafisica come scienza

dimostrativa, Naples: Bibliopolis.
Bonelli, M. (2006) ‘Continuité et rupture dans l’Académie et le platonisme’,
Etudes platoniciennes III, Paris: Société d’Études Platoniciennes, 231–44.

Bonelli, M. (2014) ‘Plutarch and the Sceptics’, in M. Beck (ed.), A Companion

to Plutarch, Chichester: Wiley Blackwell, 121–34.

Boylan, M. (1982) ‘The Digestive and ‘Circulatory’ Systems in Aristotle’s

Biology’, Journal of the History of Biology 15: 89–118.

Boys-Stones, G. (1997a) ‘Thyrsus-Bearer of the Academy or Enthusiast for

Plato? Plutarch’s De Stoicorum Repugnantiis’, in J. Mossman (ed.), Plutarch
and His Intellectual World, London: Classical Press of Wales, 41–58.

Boys-Stones, G. (1997b) ‘Plutarch on the Probable Principle of Cold:

Epistemology and the De Primo Frigido’, Classical Quarterly 47: 227–38.

Brague, R. (1978) Le Restand. Supplément aux commentaires du Ménon de

Platon, Paris: Vrin.

Brandwood, L. (1990) The Chronology of Plato’s Dialogues, Cambridge:

Cambridge University Press.

Bréhier, É. (1968) La philosophie de Plotin, Paris: Vrin.

Brickhouse, T. C. and Smith, N. D. (1994) Plato’s Socrates, Oxford: Oxford
University Press.

Brisson, L. (1999) Platon Parménide, Paris: Flammarion.

Brittain, C. (2001) Philo of Larissa: The Last of the Academic Sceptics, Oxford:
Oxford University Press.

Brittain, C. (2006) (trans.) Cicero: On Academic Scepticism, Indianapolis:

Hackett.
Brittain, C. (2007) ‘Middle Platonists on Academic Scepticism’, in R. W.
Sharples and R. Sorabji (eds.), Greek and Roman Philosophy 100 B.C–200
A.D., vol. II, London: Institute of Classical Studies, 297–315.

Brittain, C. and Palmer, J. (2001) ‘The New Academy’s Appeals to the

Presocratics’, Phronesis 48: 38–72.

Broadie, S. (2009) ‘Aporia 8’, in M. Crubellier and A. Laks (eds.), Aristotle’s

Metaphysics Beta, Oxford: Clarendon Press, 135–50.

Brown, L. (1998) ‘Innovation and Continuity: The Battle of Gods and Giants,
Sophist 245–249’, in J. Gentzler (ed.), Method in Ancient Philosophy, Oxford:
Oxford University Press, 181–208.

Brown, L. (2008) ‘The Sophist on Statements, Predication, and Falsehood’, in

G. Fine (ed.), The Oxford Handbook of Plato, Oxford: Oxford University Press,
437–62.

Brown, L. (2010) ‘Definition and Division in Plato’s Sophist’, in D. Charles

(ed.), Definition in Greek Philosophy, 151–71.

Brown, L. (2011) ‘Negation and Not-Being: Dark Matter in the Sophist’, in R.

Patterson, V. Karasmanis and A. Hermann (eds.), Presocratics and Plato: A
Festschrift for Charles Kahn, Las Vegas: Parmenides Publishing, 268–91.

Bruns, I. (1887 and 1892) Scripta Minora, vols. 1 and 2, Berlin: Reimer.

Brunschwig, J. (1967) Aristote, Topiques. Tome I, Livres I–IV, Paris: Les Belles
Lettres.

Brunschwig, J. (1984/85) ‘Aristotle on Arguments without Winners or Losers’,

in Wissenschaftskolleg Jahrbuch, Berlin, 31–40.

Brunschwig, J. (1994) ‘Sextus Empiricus on the Criterion’, in his Papers in

Hellenistic Philosophy, Cambridge: Cambridge University Press, 224–43.
Brunschwig, J. (2000) ‘Dialectique et philosophie chez Aristote à nouveau’, in
N. L. Cordero (ed.), Ontologie et dialogue. Hommage à Pierre Aubenque,
Paris: Vrin, 107–30.

Brunschwig, J. (2007) Aristote, Topiques. Tome II, Livres V–VIII, Paris: Les
Belle Lettres.

Buchheim, T. (1989) Gorgias von Leontini: Reden, Fragmente und

Testimonien. Hamburg: Felix Meiner.

Burnet, J. (1901) Platonis Opera II, Oxford: Oxford University Press.

Burnyeat, M. F. (1976) ‘Protagoras and Self-Refutation in Later Greek

Philosophy’, Philosophical Review 85: 44–69.

Burnyeat, M. F. (1977a) ‘Examples in Epistemology: Socrates, Theaetetus,

and G.E. Moore’, Philosophy 52: 381–98.

Burnyeat, M. F. (1977b) ‘Socratic Midwifery, Platonic Inspiration’, Bulletin of the

Institute of Classical Studies 24: 7–16.

Burnyeat, M. F. (1982) ‘Gods and Heaps’, in M. Schofield and M. Nussbaum

(eds.), Language and Logos: Studies in Ancient Greek Philosophy,
Cambridge: Cambridge University Press, 315–38.

Burnyeat, M. F. (1990) The Theaetetus of Plato, Indianapolis: Hackett.

Burnyeat, M. F. (1997) ‘Antipater and Self-Refutation: Elusive Arguments in

Cicero’s Academica’, in B. Inwood and J. Mansfeld (eds.), Assent and
Argument: Studies in Cicero’s Academic Books, Leiden: Brill, 277–310.

Calogero, G. (1932) Studi sull’Eleatismo, Rome: Tipografia del senato.

Caluori, D. (2015) Plotinus on the Soul, Cambridge: Cambridge University

Press.
Caluori, D. (forthcoming) ‘Socratic Methods in Damascius’, in C. Moore (ed.),
Brill’s Companion to the Reception of Socrates, Leiden: Brill.

Campbell, L. (1883) The Theaetetus of Plato, with a revised text and English
notes, 2nd Edition. Oxford: Oxford University Press.

Castelnérac, B. and Marion, M. (2009) ‘Arguing for Inconsistency: Dialectical

Games in the Academy’, in G. Primiero and S. Rahman (eds.), Acts of
Knowledge: History, Philosophy and Logic, London: College Publications, 43–
82.

Castagnoli, L. (2002) Review of R. Bett, Pyrrho, His Antecedents, and His

Legacy (New York: Oxford University Press, 2000), in Ancient Philosophy 22:
443–57.

Castagnoli, L. (2010) Ancient Self-Refutation: The Logic and History of the

Self-Refutation Argument from Democritus to Augustine, Cambridge:
Cambridge University Press.

Castagnoli, L. (2013) ‘Early Pyrrhonism: Pyrrho to Aenesidemus’, in F.

Sheffield and J. Warren (eds.), Routledge Companion to Ancient Philosophy,
London: Routledge, 496–510.

Castagnoli, L. (2018) ‘Anesidemus’, in D. Machuca and B. Reed (eds.),

Skepticism: From Antiquity to the Present, London: Bloomsbury, 67–80.

Castelnérac, B. and Marion, M. (2009) ‘Arguing for Inconsistency: Dialectical

Games in the Academy’, in G. Primiero and S. Rahman (eds.), Acts of
Knowledge: History, Philosophy and Logic, London: College Publications, 43–
82.

Chance, T. H. (1991) Plato’s Euthydemus. Analysis of What Is and What Is Not
Philosophy, Berkeley: University of California Press.

Charrue, P. (1978) Plotin lecteur de Platon, Paris: Les Belles Lettres.

Chiaradonna, R. (2013) ‘Alexander, Boethus, and the Other Peripatetics: The
Theory of Universals in the Aristotelian Commentators’, in R. Chiaradonna and
G. Galuzzo (eds.), Universals in Ancient Philosophy, Pisa: Edizioni della
Normale, 299–328.

Cleary, J. J. (1995) Aristotle and Mathematics: Aporetic Method in Cosmology

and Metaphysics, Leiden: Brill.

Combès, J. (1989) ‘Damascius Lecteur du Parménide’, in his Etudes

Néoplatoniciennes, Grénoble, 63–99.

Cooper, J. (2004) ‘Arcesilaus: Socratic and Sceptic’, in his Knowledge, Nature

and the Good, Princeton: Princeton University Press, 81–103.

Cooper, J. (2007) ‘Socrates and Philosophy as a Way of Life’, in D. Scott (ed.),

Maieusis: Essays on Ancient Philosophy in Honour of Myles Burnyeat, Oxford:
Oxford University Press, 20–43.

Cooper, J. (2012) ‘Conclusion—and Retrospect (Metaphysics A 10)’, in C.

Steel (ed.), Aristotle’s Metaphysics Alpha: Symposium Aristotelicum. Oxford:
Oxford University Press, 335–64.

Cornford, F. M. (1939) Plato and Parmenides, London: Routledge and Kegan
Paul.

Corrigan, K. (2005) Reading Plotinus: A Practical Introduction to

Neoplatonism, Indiana: Perdue University Press.

Cotton, A. K. (2014) Platonic Dialogue and the Education of the Reader,
Oxford: Oxford University Press.

Couissin, P. (1929) ‘Le Stoicisme de la Nouvelle Académie’, Revue d’histoire

de la Philosophie 3: 241–76. (English trans. in M. F. Burnyeat (ed.), (1983),
The Skeptical Tradition, Berkeley: University of California Press, 31–63.)
Couloubaritsis, L. (1990) ‘La problématique sceptique d’un impensé: ἡ σκέψις’,
in A. J. Voelke (ed.), Le scepticisme antique, Cahier de la Revue de Théologie
et de Philosophie 15: 9–28.

Crivelli, P. (1996) ‘The Argument from Knowing and Not Knowing in Plato’s
Theaetetus (187e5– 188c8)’, Proceedings of the Aristotelian Society 96: 177–
96.

Crivelli, P. (2012) Plato’s Account of Falsehood: A Study of the Sophist,

Cambridge: Cambridge University Press.

Crubellier, M. and Laks, A. (2009) ‘Introduction’, in M. Crubellier and M. Laks

(eds.), Aristotle: Metaphysics Beta: Symposium Aristotelicum, Oxford: Oxford
University Press, 1–23.

De Waal, T. (2009) Knowing Platonist Objects: The Access Problem and Its
Solution in Plato, PhD dissertation, King’s College London.

Decleva Caizzi, F. (1981) Pirrone: Testimonianze, Naples: Bibliopolis.

Decleva Caizzi, F. (1992) ‘Aenesidemus and the Academy’, Classical

Quarterly 42: 176–89.

Delcomminette, S. (2014) ‘Odysseus and the Home of the Stranger from Elea’,
Classical Quarterly 64: 533–41.

des Places, E. (1973) Numénius Fragments (ed. and trans.), Paris: Les Belles
Lettres.

Diès, A. (1956) Platon Parménide, Paris.

Dillon, J. M. (1996) ‘Damascius on the Ineffable’, Archiv für Geschichte der
Philosophie 78: 120–9.
Dillon, J. M. (1999) ‘Plutarch’s Debt to Xenocrates’, in Perez Jimenez A. et al.
(eds), Plutarco, Platon y Aristoteles: Actas del V Congreso internacional de la
I.P.S.,Madrid, 305–311. (repr. in Dillon, The Platonic Heritage, Farnham:
Ashgate, 2012).

Dillon, J. M. and Blumenthal, H. J. (2015) Ennead IV.3–4.29. Problems

concerning the Soul, Las Vegas: Parmenides Publishing.

Dodds, E. R. (1928) ‘The Parmenides of Plato and the Origins of the
Neoplatonic One’, Classical Quarterly 22: 129–42.

Donini, P. L. (1994) ‘Testi e commenti, manuali e insegnamento: la forma

sistematica e i metodi della filosofia in età postellenistica’, in ANRW II.36.7,
5027–5100.

Donini, P. L. (2002) ‘L’eredità academica e i fondamenti del platonismo in

Plutarco’, in M. Barbanti and F. Romano (eds.), Henōsis kai Philia: Unione e
amicizia. Omaggio a Francesco Romano, Catania: CUECM, 247–73.

Dooley, W. E. and Madigan, A. (1992) Alexander of Aphrodisias on Aristotle’s

Metaphysics 2 and 3, Ithaca: Cornell University Press.

Dorion, L.-A. (2013) ‘Aristotle and the Socratic elenchos’, in C. Rapp and P.-S.
Hasper (eds.), Fallacious Arguments in Ancient Philosophy, Logical Analysis
and the History of Philosophy, 15: 323–42.

Duncombe, M. (2013) ‘The Greatest Difficulty at Parmenides 133C-134E and

Plato’s Relative Terms’, Oxford Studies in Ancient Philosophy 45: 43–62.

Düring, I. (1968) ‘Aristotle’s Use of Examples in the Topics’, in G. E. L. Owen

(ed.), Aristotle on Dialectic, Oxford: Oxford University Press, 202–209.

Dye, G. and Vitrac, B. (2009) ‘Le Contre les géomètres de Sextus Empiricus:
sources, cible, structure’, Phronesis 54: 155–203.
Ellis, J. (1994) ‘Alexander’s Defense of Aristotle’s Categories’, Phronesis, 39:
69–89.

Emilsson, E. (2005) ‘Soul and μερισμός’, in R. Chiaradonna (ed.), Studi

sull’anima in Plotino, Naples: Bibliopolis, 79–93.

Erler, M. (1987) Der Sinn der Aporien in den Dialogen Platons, Berlin: de
Gruyter.

Fazzo, S. (2002) Aporia e sistema: La materia, la forma, il divino nelle

Quaestiones di Alessandro di Afrodisia, Pisa: ETS.

Fine, G. (1979) ‘False Belief in the Theaetetus’, Phronesis, 24: 70–80.

Fine, G. (1992) ‘Inquiry in the Meno’, in R. Kraut (ed.), Cambridge Companion

to Plato, Cambridge: Cambridge University Press, 200–26.

Fine, G. (2003), Plato on Knowledge and Forms: Selected Essays, Oxford:

Oxford University Press.

Fine, G. (2008) ‘Does Socrates Claim to Know that He Knows Nothing?’,

Oxford Studies in Ancient Philosophy, 35: 49–85.

Fine, G. (2010) ‘Sceptical Enquiry’, in D. Charles (ed.), Definition in Greek

Philosophy, Oxford: Oxford University Press, 493–525.

Fine, G. (2011) ‘Sceptics and Inquiry: Sextus and the Epicureans’, in B.

Morison and K. Ierodiakonou (eds.), Episteme, etc.: Essays in Honour of
Jonathan Barnes, Oxford: Oxford University Press, 90–114.

Fine, G. (2014) The Possibility of Inquiry. Meno’s Paradox from Socrates to

Sextus, Oxford: Oxford University Press.

Forster, M. N. (2006) ‘Socrates’ Demand for Definitions’, Oxford Studies in

Ancient Philosophy, 31: 1–47.
Frede, M. (1983) ‘Stoics and Sceptics on Clear and Distinct Impressions’, in M.
Burnyeat (ed.), The Sceptical Tradition, Berkeley: University of California
Press, 151–76.

Frede, M. (1992a) ‘Plato’s Arguments and the Dialogue Form’, in J. C. Klagge
and N. D. Smith (eds.), Oxford Studies in Ancient Philosophy, Supplementary
Volume: 201–220.

Frede, M. (1992b) ‘Introduction to the Protagoras’, in S. Lombardo, and K. Bell

(eds.), Plato: Protagoras, Indianapolis: Hackett, vii–xxxiv.

Frede, M. (1992c) ‘Plato’s Sophist on False Statements’, in R. Kraut (ed.), The

Cambridge Companion to Plato, Cambridge: Cambridge University Press,
397–424.

Frede, M. (1996) ‘The Literary Form of the Sophist’, in C. Gill and M. M.
McCabe (eds.), Form and Argument in Late Plato, Oxford: Oxford University
Press, 135–51.

Frede, M. (2004) ‘Aristotle’s Account of the Origins of Philosophy’, Rhizai 1: 9–

44.

Friedländer, P. (1964) Platon, Bd. 2: Die Platonischen Schriften. Erste Periode,

3rd ed., Berlin: de Gruyter.

Furley, D. (1989) ‘Truth as What Survives the Elenchos’, in his Cosmic

Problems: Essays on Greek and Roman Philosophy of Nature, Cambridge:
Cambridge University Press, 38–46.

Galpérine, M.-C. (1987) Damascius, Des Premiers Principes: Apories et

resolutions, Lagrasse: Editions Verdier.

Gatti, M. L. (1996) ‘Plotinus: The Platonic Tradition and the Foundation of
Neoplatonism’, in L. Gerson (ed.), The Cambridge Companion to Plotinus,
Cambridge: Cambridge University Press, 10–37.
Geach, P. T. (1966) ‘Plato’s Euthyphro’, The Monist 50: 369–82.

Gill, M. L. (1996) Plato Parmenides: Translated by Mary Louise Gill and Paul
Ryan. Introduction by Mary Louise Gill, Indianapolis: Hackett.

Goldschmidt, V. (1947) Les dialogues de Platon: Structure et méthode

dialectique, Paris: Presses Universitaires de France.

Gonzales, F. (2002) ‘The Socratic Elenchus as Constructive Protreptic’, in G.

A. Scott (ed.), Does Socrates Have a Method? Rethinking the Elenchus in
Plato’s Dialogues and Beyond, State College: Pennsylvania State University
Press, 161–182.

Granger, H. (2004) ‘Argumentation and Heraclitus’ Book’, Oxford Studies in

Ancient Philosophy 26: 1–17.

Grgić, F. (2006) ‘Sextus Empiricus on the Goal of Skepticism’, Ancient

Philosophy 26: 141–60.

Grgić, F. (2008) ‘Sextus on the Possibility of Inquiry’, Pacific Philosophical

Quarterly 89: 436–59.

Griffin, M. (1994) ‘The Composition of the Academica: Motives and Versions’,

in B. Inwood and J. Mansfeld (eds.), Assent and Argument: Studies in Cicero’s
Academic Books, Leiden: Brill, 1–35.

Gronewald, M. (1968) ‘Ein neues Protagoras-Fragment’, Zeitschrift für

Papyrologie und Epigraphik 2: 1–2.

Gronewald, M. and Gesché, A. (1969) (eds.) Didymus der Blinde:

Psalmenkommentar (Tura-Papyrus), Teil III. Papyrologische Texte und
Abhandlungen, Bd. 8, Bonn: R. Habelt.

Gurtler, G. (2015) Ennead IV.4.30–45 and IV.5, Las Vegas: Parmenides

Publishing.
Hagdopoulos, D. J. (1976) ‘Protasis and Problema in the Topics’, Phronesis
21: 266–76.

Halper, E. (1992) ‘The Origin of Aristotle’s Metaphysical Aporiai’, in A. Preus

and J. P. Anton (eds.), Essays in Ancient Greek Philosophy V: Aristotle’s
Ontology. Albany: SUNY Press, 151–75.

Halper, E. (2009) One and Many in Aristotle’s Metaphysics. Books Alpha–

Delta, Las Vegas: Parmenides Publishing.

Hankinson, R. (1995) The Sceptics, London: Routledge.

Hasper, P.-S. (2013) ‘Between Science and Dialectic: Aristotle’s Account of

Good and Bad Peirastic Arguments in the Sophistical Refutations’, in C. Rapp
and P.-S. Hasper (eds.), Fallacious Arguments in Ancient Philosophy, Logical
Analysis and the History of Philosophy 15: 286–322.

Heinemman, F. (1921), Plotin: Forschungen über die plotinische Frage, Plotins

Entwicklung und sein System, Leipzig: Felix Meiner.

Henry, D. (2005) ‘Embryological Models in Ancient Philosophy’, Phronesis 50:

1–42.

Hiley, D. (1987) ‘The Deep Challenge of Pyrrhonian Scepticism’, Journal of the

History of Philosophy 24: 185–213.

Hirzel, R. (1883) Untersuchungen zu Ciceros philosophischen Schriften,

Leipzig: Felix Meiner.

Hirzel, R. (1895) Der Dialog, Leipzig: Felix Meiner.

Hoffman, H. (2004) ‘Zeno’s tortoise’, Antike Kunst 47: 5–9.

Hoffmann, P. (1997) ‘L’expression de l’indicible dans le néoplatonisme grec, de

Plotin à Damascius’, in C. Lévy and L. Pernot (eds.), Dire l’Évidence, Paris:
Editions L’Harmattan, 335–90.

Ioppolo, A. M. (1993) ‘The Academic Position of Favorinus of Arelate’,

Phronesis 38: 183–213.

Ioppolo, A. M. (2009) La testimonianza di Sesto Empirico sull’accademia

scettica, Naples: Bibliopolis.

Irwin, T. H. (1977) Plato’s Moral Theory: The Early and Middle Dialogues,
Oxford: Oxford University Press.

Irwin, T. H. (1988) Aristotle’s First Principles, Oxford: Oxford University Press.

Irwin, T. H. (1995) Plato’s Ethics, Oxford: Oxford University Press.

Kahn, C. H. (1988) ‘On the Relative Date of the Gorgias and the Protagoras’,
Oxford Studies in Ancient Philosophy 6: 69–102.

Kahn, C. H. (1996) Plato and the Socratic Dialogue, Cambridge: Cambridge
University Press.

Kalligas, P. (2014) The Enneads of Plotinus. A Commentary, vol. 1, Princeton:

Princeton University Press.

Kanayama, Y. (2011) ‘Plato as a Wayfinder: To Know Meno, the Robbery Case

and the Road to Larissa’, Japan Studies in Classical Antiquity 1: 63–88.

Kapp, E. (1942) Greek Foundations of Traditional Logic, New York: Columbia

University Press.

Karamanolis, G. (2006) Plato and Aristotle in Agreement?, Oxford: Oxford

University Press.

Karbowski, J. (2014) ‘Empirical Eulogos Argumentation in GA III.10’, British

Journal for the History of Philosophy 22: 25–38.
Kennedy, G. A. (1991) Aristotle, On Rhetoric. A Theory of Civic Discourse,
Newly Translated, with Introduction, Notes and Appendices, Oxford: Oxford
University Press.

Kerferd, G. B. (1955) ‘Gorgias on Nature or That Which Is Not’, Phronesis 1:

3–25.

Kerferd, G. B. (1981) The Sophistic Movement, Cambridge: Cambridge

University Press.

King, R. A. H. (2009) Aristotle and Plotinus on Memory, Berlin: de Gruyter.

Kirk, G. S., Raven, J. E. and Schofield, M. (1983) The Presocratic

Philosophers, 2nd ed., Cambridge: Cambridge University Press.

Klein, J. (1965) A Commentary on Plato’s Meno, Chapel Hill: University of

North Carolina Press (reprinted by University of Chicago Press 1989).

Knuuttila, S. and Sihvola, J. (2000) ‘Ancient Scepticism and Philosophy of

Religion’, in J. Sihvola (ed.), Ancient Scepticism and the Sceptical Tradition,
Acta Philosophica Fennica 66, Helsinki, 125–44.

Laks, A. (2009) ‘Aporia Zero (Metaphysics B 1, 995a24–995b4)’, in M.

Crubellier and A. Laks (eds.), Aristotle: Metaphysics Beta. Symposium
Aristotelicum, Oxford: Oxford University Press, 25–46.

Lee, J.-H. (2010) Der philosophische Sinn der Aporie bei Aristoteles,
Hamburg: Verlag Dr. Kovac.

Leigh, F. (2010) ‘Being and Power in Plato’s Sophist’, Apeiron 43: 63–85.

Lennox, J. G. (2014) ‘Aristotle on the Emergence of Material Complexity:

Meteorology IV and Aristotle’s Biology’, HOPOS: The Journal of the
International Society for the History and Philosophy of Science 4: 272–305.
Leszl, W. (1975) Aristotle’s Conception of Ontology, Padua: Editrice Antenore.

Lloyd, A. (1990) The Anatomy of Neoplatonism, Oxford: Oxford University

Press.

Long, A. A. (1984) ‘Methods of Argument in Gorgias, Palamedes’, in K.

Boudouris (ed.), The Sophistic Movement, Athens, Kardamitsa, 233–41.

Long, A. A. (1998) ‘Plato’s Apologies and Socrates in the Theaetetus’, in J.

Gentzler (ed.), Method in Ancient Philosophy, Oxford: Oxford University Press,
113–36.

Machuca, D. (2006) ‘The Pyrrhonist’s Ataraxia and Philanthropia’, Ancient

Philosophy 26: 111–39.

Machuca, D. (2011) ‘Ancient Skepticism: Pyrrhonism’, Philosophy Compass 6:

246–58.

Machuca, D. (2013) ‘Pyrrhonism, Inquiry, and Rationality’, Elenchos 34: 201–

28.

Mackenzie, M. M. (1988a) ‘Heraclitus and the Art of Paradox’, Oxford Studies
in Ancient Philosophy 6: 1–37.

Mackenzie, M. M. (1988b) ‘Impasse and Explanation, from the Lysis to the
Phaedo’, Archiv für Geschichte der Philosophie 70: 15–45.

Madigan, A. (1992) Alexander of Aphrodisias On Aristotle Metaphysics 3,

Ithaca: Cornell University Press.

Madigan, A.(1987) ‘Alexander of Aphrodisias: the Book of Ethical Problems’,

Aufstieg und Niedergang der Römischen Welt II 36 2, 1260–79.

Madigan, A. (1999) Aristotle Metaphysics Book B and Book K 1–2, Oxford:

Oxford University Press.
Maier, H. (1896–1900) Die Syllogistik des Aristoteles, 3 vols., Tübingen:
Laupp.

Makin, S. (2005) ‘Melissus and His Opponents: The Argument of DK 30 B 8’,

Phronesis 50: 263–88.

Mansfeld, J. (1981) ‘Protagoras on Epistemological Obstacles and Persons’, in

G. B. Kerferd (ed.), The Sophists and Their Legacy: Proceedings of the Fourth
International Colloquium on Ancient Philosophy, Wiesbaden: F. Steiner. 38–
53.

Mansfeld, J. (1985) ‘Historical and Philosophical Aspects of Gorgias’ “On What

Is Not”’, in L. Montoneri and F. Romano (eds.), Gorgia e la Sofistica, Siculorum
Gymnasium 38: 243–71.

Mansfeld, J. (1995) ‘Aenesidemus and the Academics’, in L. Ayres (ed.), The

Passionate Intellect: Essays on the Transformation of the Classical Tradition
Presented to Prof. I. G. Kidd, New Brunswick, 235–48.

Mansion, S. (1955) ‘Les apories de la Métaphysique aristotélicienne’, in Autour

d’Aristote. Recueil d’études de philosophie ancienne et médiévale, offert à
Monseigneur A. Mansion, Louvain, Publications Universitaires de Louvain,
141–79.

Marchand, S. (2010) ‘Le sceptique cherche-t-il vraiment la vérité?’, Revue de

Métaphysique et de Morale, 65: 125–41.

Marzolo, C. (2006) Plotino, Che cos’è l’essere vivente e che cos’è l’uomo? I. 1.
[53]. Pisa.

Mates, B. (1996) The Skeptic Way; Sextus Empiricus’s Outlines of Pyrrhonism,

Oxford: Oxford University Press.

Matthews, G. (1999a) Socratic Perplexity and the Nature of Philosophy,

Oxford: Oxford University Press.
Matthews, G. (1999b) ‘The Normalization of Perplexity in Aristotle’, in M. Sim
(ed.), From Puzzles to Principles?: Essays on Aristotle’s Dialectic, Lanham:
Lexington Books, 125–36.

McCabe, M. M. (1994) Plato’s Individuals, Princeton: Princeton University

Press.

McCabe, M. M. (2000) Plato and His Predecessors: The Dramatisation of

Reason, Cambridge: Cambridge University Press.

McDowell, J. (1973) Plato, Theaetetus, translated with notes, Oxford:

Clarendon Press.

McPherran, M. (1996) The Religion of Socrates, State College: Pennsylvania

State University Press.

Mejer, J. (1972) ‘The Alleged New Fragment of Protagoras’, Hermes 100:

175–8.

Mette, H.-J. (1984) (ed.), ‘Zwei Akademiker heute: Krantor von Soloi und
Arkesilaos von Pitane’, Lustrum 26: 7–94.

Mette, H.-J. (1985) (ed.), ‘Weitere Akademiker heute: von Lakydes bis zu
Kleitomachos’, Lustrum 27: 39–148.

Mette, H.-J. (1986–87) (ed.) ‘Philon von Larisa und Antiochos von Askalon’,
Lustrum 28–9: 9–63.

Migliori, M. (1990) Dialettica e Verità: Commentario filosofico al ‘Parmenide’ di

Platone, Milan: Vita e Pensiero.

Miller, M. H. (1986) Plato’s Parmenides: The Conversion of the Soul,

Princeton: Princeton University Press.
Moller, D. (2004) ‘The Pyrrhonian Skeptic’s Telos’, Ancient Philosophy 24:
424–41.

Moreschini, C. (1966) Platonis Parmenides Phaedrus, Rome: Ed. dell’Ateneo.

Morsink, J. (1982) Aristotle on the Generation of Animals: A Philosophical

Study, Lanham: Rowman and Littlefield.

Motte, A. and Rutten, C. (2001) (eds.) Aporia dans la philosophie grecque des
origins à Aristote, Louivain-la-Neuve: Peeters.

Mueller, I. (1982) ‘Geometry and Scepticism’, in J. Barnes, J. Brunschwig,

M. F. Burnyeat and M. Schofield (eds.), Science and Speculation, Cambridge:
Cambridge University Press, 69–95.

Nagel, T. (1974) ‘What Is It Like to Be a Bat?’, The Philosophical Review 83:

435–50.

Nagel, T. (1986) The View from Nowhere, Oxford: Oxford University Press.

Nehamas, A. (1990) ‘Eristic, Antilogic, Sophistic, Dialectic: Plato’s

Demarcation of Philosophy from Sophistry’, History of Philosophy Quarterly 7:
3–16.

Nightingale, A. (2010) ‘Plato on Aporia and Self-Knowledge’, in A. Nightingale

and D. Sedley (eds.), Ancient Models of Mind: Studies in Human and Divine
Rationality, Cambridge: Cambridge University Press, 8–26.

Noble, C. (2013) ‘How Plotinus’ Soul Animates His Body: The Argument for
the Soul-Trace at Ennead 4.4.18.1–9’, Phronesis, 58: 249–79.

Notomi, N. (1999) The Unity of Plato’s Sophist: Between the Sophist and the
Philosopher, Cambridge: Cambridge University Press.
Olfert, C. M. (2015) ‘Skeptical Investigation and Its Perks: D.L. 9.66–70 and
79–89’, in K. M. Vogt (ed.), Pyrrhonian Skepticism in Diogenes Laertius,
Tübingen: Mohr Siebeck, 147–170.

O’Meara, D. (1993) Plotinus: An Introduction to the Enneads, Oxford: Oxford

University Press.

O’Meara, D. (2013) Sur les traces de l’Absolu, Fribourg: Academic Press

Fribourg.

Opsomer, J. (1998) In Search of the Truth: Academic Tendencies in Middle

Platonism, Brussels: KAWLSK.

Owen, G. E. L. (1961) ‘Tithenai ta phainomena’, in S. Mansion (ed.), Aristote et

les problèmes de méthode, Louvain-la-Neuve: Peeters, 83–103.

Owen, G. E. L. (1970) ‘Notes on Ryle’s Plato’, in O. P. Wood and G. Pitcher
(eds.), Ryle, London: Macmillan, 341–72. Repr. G. E. L. Owen (1986), 85–103;
and G. Fine (1999) Plato I: Metaphysics and Epistemology, Oxford: Oxford
University Press, 298–319.

Owen, G. E. L. (1986) Logic, Science and Dialectic: Collected Papers in Greek
Philosophy, M. Nussbaum (ed.), London: Duckworth.

Owens, J. (1978) The Doctrine of Being in the Aristotelian ‘Metaphysics’. A

Study in the Greek Background of Mediaeval Thought, 3rd ed., Toronto:
Pontifical Institute of Mediaeval Studies.

Palmer, J. (1999) Plato’s Reception of Parmenides, Oxford: Clarendon Press.

Palmer, J. (2000) ‘Skeptical Investigation’, Ancient Philosophy 30: 351–73.

Palmer, J. (2004) ‘Melissus and Parmenides’, Oxford Studies in Ancient

Philosophy 26: 19–54.
Palmer, J. (2009) Parmenides and Presocratic Philosophy, Oxford: Oxford
University Press.

Peck, A. L. (1942) (trans. and comm.) Aristotle: Generation of Animals,

Cambridge, MA: Harvard University Press.

Perin, C. (2006) ‘Pyrrhonian Scepticism and the Search for Truth’, Oxford
Studies in Ancient Philosophy 30: 337–60.

Perin, C. (2010) The Demands of Reason: An Essay on Pyrrhonian

Scepticism, Oxford: Oxford University Press.

Peterson, S. (1981) ‘The Greatest Difficulty for Plato’s Theory of Forms: The
Unknowability Argument of Parmenides 133c–134c,’ Archiv für Geschichte der
Philosophie 63: 1–16.

Pickard-Cambridge, W. A. (1984) (trans.) ‘Sophistical Refutations’, in J. Barnes

(ed.), The Complete Works of Aristotle, 2 vols., Princeton: Princeton University
Press.

Platt, A. (1984) (trans.) ‘Generation of Animals’, in J. Barnes (ed.), The

Complete Works of Aristotle: The Revised Oxford Translation, vol. 1,
Princeton: Princeton University Press, 1111–218.

Politis, Vasilis (2003) ‘Aristotle on Aporia and Searching in Metaphysics’,

Proceedings of the Boston Area Colloquium in Ancient Philosophy 18: 145–82.

Politis, Vasilis (2006) ‘Aporia and Searching in the Early Plato’, in L. Judson
and V. Karasmanis (eds.), Remembering Socrates: Philosophical Essays,
Oxford: Oxford University Press, 88–109.

Politis, Vasilis (2007) ‘Is Socrates Paralyzed by His State of Aporia? Meno
79e7-80d4’, in L. Brisson and M. Erler (eds.), Gorgias—Menon, Sankt
Augustin: Academia Verlag, 268–72.
Politis, Vasilis (2008) ‘The Place of Aporia in Plato’s Charmides’, Phronesis 53:
1–34.

Politis, Vasilis (2012a) ‘What Do the Arguments in the Protagoras Amount to?’,
Phronesis 57: 209–39.

Politis, Vasilis (2012b) ‘What Is Behind the ti esti Question?’, in J. L. Fink (ed.),
The Development of Dialectic from Plato to Aristotle, Cambridge: Cambridge
University Press, 199–223.

Politis, Vasilis (2015) The Structure of Enquiry in Plato’s Early Dialogues,

Cambridge: Cambridge University Press.

Politis, V. and Su, J. (2017) ‘The Concept of Ousia in Metaphysics Alpha, Beta,
and Gamma’, in R. Polansky and W. Wians (eds.), Reading Aristotle.
Exposition and Argument, Leiden: Brill, 257–276.

Polito, R. (2014) Aenesidemus of Cnossus: Testimonia, Cambridge:

Cambridge University Press.

Prantl, C. (1855) Geschichte der Logik im Abendlande, Leipzig: Verlag von S.

Hirzel.

Primavesi, O. (1996) Die Aristotelische Topik: ein Interpretationsmodell und

seine Erprobung am Beispiel von Topik B, München: C.H. Beck.

Primavesi, O. (2010) ‘Dialektik und Gespräch bei Aristoteles’, in K. W.

Hempfer and A. Traninger (eds.), Der Dialog im Diskursfeld seiner Zeit. Von
der Antike bis zur Aufklärung, Stuttgart: Franz Steiner Verlag, 47–73.

Prior, W. J. (1998) ‘Plato and the “Socratic Fallacy”’, Phronesis 43: 97–113.

Rapp, C. (2016) ‘Dialectic and Logic from a Rhetorical Point of View’, in J.-B.
Gourinat and J. Lemaire (eds.), Logique et dialectique dans l’Antiquité, Paris:
Vrin, 161–192.
Rapp, C. and Wagner, T. (2013) ‘On Some Aristotelian Sources of Modern
Argumentation Theory’, Argumentation 27: 7–30.

Rappe, S. (1998) ‘Scepticism in the Sixth Century? Damascius’ Doubts and

Solutions Concerning First Principles’, Journal of the History of Philosophy 36:
337–63.

Rappe, S. (2000) Reading Neoplatonism: Non-Discursive Thinking in the Texts

of Plotinus, Proclus, and Damascius, Cambridge: Cambridge University Press.

Rappe, S. (2010) Damascius’ Problems and Solutions Concerning First

Principles. Translated from the Greek. Oxford: Oxford University Press.

Rashed, M. (2007) Essentialisme: Alexandre d’Aphrodise entre logique,

physique et cosmologie, Berlin: de Gruyter.

Reale, G. (1980) The Concept of First Philosophy and the Unity of the
Metaphysics of Aristotle, Albany: SUNY Press.

Reeve, C. D. C. (1989) Socrates in the Apology, Indianapolis: Hackett.

Reid, J. S. (2007) (ed.) The Academica of Cicero, London: Evergreen.

Rickless, S. C. (2007) Plato’s Forms in Transition: A Reading of the

Parmenides, Cambridge: Cambridge University Press.

Rist, J. (1967) Plotinus: The Road to Reality, Cambridge: Cambridge

University Press.

Robinson, R. (1942) ‘Plato’s Parmenides II’, Classical Philology 37: 159–86.

Robinson, R. (1953) Plato’s Earlier Dialectic, Oxford: Oxford University Press.

Rodriguez, E. (2016) Exploring Both Sides: Plato’s New Method for First
Principles, doctoral dissertation, Yale University.
Ross, W. S. (1924) (ed.) Aristotle: Metaphysics: A Revised Text with
Introduction and Commentary, vols. 1–2, Oxford: Oxford University Press.

Ross, W. S (1958) Aristotelis Topica et Sophistici elenchi, rec. brevique

adnotatione critica instr., Oxford: Clarendon Press.

Santas, G. X. (1972) ‘The Socratic Fallacy’, Journal of the History of

Philosophy 10: 127–41.

Sayre, K. M. (1996) Parmenides’ Lesson: Translation and Explication of Plato’s

Parmenides, South Bend: Notre Dame Press.

Sayre, K. M. (2005) ‘The Method Revisited: Parmenides, 135E9-136C6’, in A.

Havlícek and F. Karfík (eds.), Plato’s Parmenides (Proceedings of the Fourth
Symposium Platonicum Pragense), Prague: OIKOYMENH, 125–40.

Schofield, M. (1999) ‘Academic Epistemology’, in K. Algra, J. Barnes, J.

Mansfeld and M. Schofield (eds.), The Cambridge History of Hellenistic
Philosophy, Cambridge: Cambridge University Press, 323–51.

Scolnicov, S. (2001) Plato’s Parmenides, Berkeley: University of California

Press.

Scott, D. (2006) Plato’s Meno, Cambridge: Cambridge University Press.

Sedley, D. (2004) The Midwife of Platonism: Text and Subtext in Plato’s

Theaetetus, Oxford: Oxford University Press.

Sedley, D. (2012) (ed.) The Philosophy of Antiochus, Cambridge: Cambridge

University Press.

Sharples, R. W. (1985) Plato. Meno, Oxford: Aris & Phillips.

Sharples, R. W. (1987) ‘Alexander of Aphrodisias: Tradition and Innovation’,

ANRW II.32.7.
Sharples, R. W. (1990) Alexander of Aphrodisias: Ethical Problems, London:
Bloomsbury Academic.

Sharples, R. W. (1992) Alexander of Aphrodisias: Quaestiones 1.1–2.15,

London: Bloomsbury Academic.

Sharples, R. W. (1994) Alexander of Aphrodisias: Quaestiones 2.16–3.15,

London: Bloomsbury Academic.

Sharples, R. W. (2004) Alexander of Aphrodisias: Supplement On the Soul,

London: Bloomsbury Academic.

Sharples, R. W. (2005) ‘Alexander of Aphrodisias on Universals: Two

Problematic Texts’, Phronesis, 50: 43–55.

Sharples, R. W. (2008) Alexander of Aphrodisias, Mantissa, Berlin: de Gruyter.

Smith, R. (1993) ‘Aristotle on the Uses of Dialectic’, Synthese, 96: 335–58.

Smith, R. (1997) Aristotle, Topics. Books I and VIII with Excerpts from Related
Texts, Oxford: Clarendon Press.

Smith, R. (1999) ‘Dialectic and Method in Aristotle’, in M. Sim (ed.), From

Puzzles to Principles?: Essays on Aristotle’s Dialectic, Lanham: Lexington
Books, 39–56.

Spinelli, E. (2000) ‘Sextus Empiricus, the Neighbouring Philosophies and the

Sceptical Tradition’, in J. Sihvola (ed.), Ancient Scepticism and the Sceptical
Tradition, Acta Philosophica Fennica, 66: 35–61.

Sprague, R. K. (1962) Plato’s Use of Fallacy. A Study of the Euthydemus and
some other Dialogues, London: Routledge.

Sprague, R. K. (1973) ‘Introduction’, in Plato, Laches and Charmides, Library

of Liberal Arts.
Stevens, A. (2001) ‘Aristote, Métaphysique’, in Motte, A. and Rutten, C. (eds.),
Aporia dans la Philosophie grecque. Des origines à Aristote, Louvain-la-
Neuve: Peeters, 257–68.

Stocks, J. L. (1984) (trans.) ‘de Caelo’, in J. Barnes (ed.), The Complete
Works of Aristotle: The Revised Oxford Translation, vol. 1, Princeton:
Princeton University Press, 447–511.

Striker, G. (1980) ‘Sceptical Strategies’, in M. Schofield, M. Burnyeat and J.

Barnes (eds.), Doubt and Dogmatism, Oxford: Oxford University Press, 54–83.

Striker, G. (1996) ‘On the Differences Between the Pyrrhonists and the
Academics’, in her own Essays in on Hellenistic Epistemology and Ethics,
Cambridge: Cambridge University Press 135–49.

Striker, G. (2001) ‘Scepticism as a Kind of Philosophy’, Archiv für Geschichte

der Philosophie 83: 113–29.

Szaif, J. (1996) Platons Begriff der Wahrheit, Freiburg/München: Alber.

Szaif, J. (2004) ‘Die Aletheia in Platons Tugendlehre’, in M. van Ackeren (ed.),

Platon Verstehen, Darmstadt: Wissenschaftliche Buchgesellschaft.

Tarrant, H. (1985) Scepticism or Platonism: The Philosophy of the Fourth

Academy, Cambridge: Cambridge University Press.

Taylor, C. C. W. (1982) ‘The End of the Euthyphro’, Phronesis, 27: 109–18.

Taylor, D. E. (2014) ‘Pyrrhonian Skepticism, Value Nihilism and the Good of
Knowledge’, Ancient Philosophy 34: 317–39.

Thorsrud, H. (2011) ‘Sextus Empiricus on Skeptical Piety’, in D. E. Machuca

(ed.), New Essays on Ancient Pyrrhonism, Philosophia Antiqua 126, Leiden:
Brill, 91–111.
Tsouna, V. (2015) ‘Plato’s Representations of the “Socratics”’, in U. Zilioli (ed.),
From the Socratics to the Socratic Schools, London: Routledge, 1–25.

Tuozzo, T. M. (2011) Plato’s Charmides. Positive Elenchus in a ‘Socratic’

Dialogue, Cambridge: Cambridge University Press.

van Ophuijsen, J. M. (2001) Alexander of Aphrodisias: On Aristotle Topics 1,

London: Bloomsbury Academic.

Vlastos, G. (1967) ‘Zeno of Elea’, in P. Edwards (ed.), The Encyclopedia of

Philosophy, vol. 8, New York: Macmillan, 369–79.

Vlastos, G. (1991) Socrates: Ironist and Moral Philosopher, Cambridge:

Cambridge University Press.

Vlastos, G. (1994a) ‘Socrates’ Disavowal of Knowledge’, in M. Burnyeat (ed.),

Socratic Studies, Cambridge: Cambridge University Press, 39–66.

Vlastos, G. (1994b) ‘Is the “Socratic Fallacy” Socratic?’, in M. Burnyeat (ed.),

Socratic Studies, Cambridge: Cambridge University Press, 67–86.

Vogt, K. M. (2010) ‘Scepticism and Action’, in R. Bett (ed.), The Cambridge
Companion to Ancient Scepticism, Cambridge: Cambridge University Press,
165–80.

Vogt, K. M. (2012) Belief and Truth: A Skeptic Reading of Plato, Oxford: Oxford
University Press.

Vogt, K. M. (2015) (ed.) Pyrrhonian Skepticism in Diogenes Laertius,

Tübingen: Mohr Siebeck.

Volkmann-Schluck, K. H. (1957) Plotin als Interpret der Ontologie Platons,

Frankfurt: Vittorio Klostermann.

von Kutschera, F. (1995) Platons ‘Parmenides’, Berlin: de Gruyter.

von Wilamowitz-Moellendorff, U. (1920) Platon, Berlin.

Wagner, T. and Rapp, C. (2004) Aristoteles, Topik. Neuübersetzung mit

Kommentar und Einleitung, Ditzingen, Ditzingen: Reclam Verlag.

Warren, J. (2002) ‘Socratic Scepticism in Plutarch’s Adversus Colotem’,

Elenchos 23: 333–56.

Weische, A. (1965) Cicero und die neue Akademie, Münster: Achendorf.

Westerink, L. G. and Combès, J. (1986) Damascius: Traité Des Premiers

Principes, Vol. I, Paris: Les Belles Lettres.

Włodarczyk, M. A. (2000) Pyrrhonian Inquiry, Cambridge: Cambridge

University Press.

Woodruff, P. (1985) ‘Didymus on Protagoras and the Protagoreans’, Journal of

the History of Philosophy 23: 483–97.

Woodruff, P. (1987) ‘Expert Knowledge in the Apology and Laches: What a

General Needs to Know’, Proceedings of the Boston Area Colloquium in
Ancient Philosophy 3: 79–115.

Woodruff, P. (1988) ‘Aporetic Pyrrhonism’, Oxford Studies in Ancient

Philosophy 6: 139–68.

Woodruff, P. (1990) ‘Plato’s Early Theory of Knowledge’, in S. Everson (ed.),

Epistemology, Cambridge: Cambridge University Press, 60–84.

Woodruff, P. (2010) ‘The Pyrrhonian Modes’, in R. Bett (ed.), The Cambridge

Companion to Ancient Scepticism, Cambridge: Cambridge University Press,
208–31.
 Index Locorum
Alcinous 
Didaskalikos 
4.6, 195
Alexander of Aphrodisias 
On the Metaphysics (in Meta.) 
173, 27–174, 4, 239
200, 18–21, 236
206, 12–13, 240
210, 20–1, 240
212, 25–7, 242
212, 27–35, 242
213, 11–13, 243
213, 19–23, 242
213, 26–214, 17, 243
213, 3–10, 243
214, 24–215, 18, 244
216, 8–11, 246
218, 17, 240
263, 26–29, 240
On the Topics (in Top.) 
1.1, 17, 3, 244
29, 23–30, 9, 233
29, 30–31, 233
3, 4–24, 232
30, 12–18, 234
30, 18–31, 4, 236
32, 12–34, 5, 239
32, 17–20, 238
32, 22–26, 238
458, 26–459, 3, 229
 Problems and Solutions (Quaest.) 
1.11, 245
1.17, 245
1.26, 245
1.3, 245
1.8, 245
2.27, 243
Suppliment to On the Soul (Mantissa) 
5, 245
Anonymous 
Theaetetus Commentary 
49–56, 193
Aristophanes 
Clouds, 38
Aristotle 
Generation of Animals 
716b12–13, 167
717a30–31, 168
717a5, 163
718a35–37, 167
718a36–37, 167
719b33–34, 168
720b21, 163
728b32–4, 159
732a32, 159
733b32, 160
733b32–734a4, 160
734a10, 161
734a11–13, 161
734a13–14, 161
734a14–16, 161
734a5–6, 161
734a7–8, 161
 734a9–10, 161
734b1–2, 161
734b18, 162
734b4–7, 161
734b7–19, 162
734b8–9, 161
735a29, 166
735a29–736a23, 158
735a5–7, 161
735b7–8, 158
736a24, 166
738b7, 163
740b2–5, 159
740b2–8, 166
740b5–8, 160
741a6, 166
743b32, 166
752a24, 158
754b20–21, 165
757a14, 166
757a2–13, 169
760a4–9, 156
760b31–2, 156
770b30, 166
776a8–9, 163
776a9–14, 163
778b23–25, 164
778b29–30, 165
778b32–35, 165
7a10–11, 161
Great Ethics 
1206b30–37, 261
History of Animals 
 556a11–2, 163
559b16, 159
Metaphysics, 5
1000a5, 142
1000a5–1001a3, 141
1001a4–5, 142, 148
1001b13–16, 19
1002b12–32, 142
1005b17–25, 12
1009b7–12, 23
1033a31–b10, 243
1034b7–19, 243
1042a17–18, 246
895a16–18, 145
895b19, 145
897a22–25, 145
981b27–29, 138
982a1–3, 138
982a14–17, 138
982a21–28, 151
982a4, 138
982a4–6, 138
982a8–14, 151
982b11–21, 230
982b12–13, 97
982b12–983a21, 205, 216
982b7–10, 138
983a11–21, 231
983a20–23, 138
983a24–26, 138
983b1–3, 151
983b4, 138
983b5–6, 139
984a18, 151
984b8–11, 151
985a13–16, 145
988a18–23, 139
988a20–23, 145
988a23, 139
988b14–16, 145
988b18, 139
988b20, 139
988b6–19, 139
993a11–17, 139
993a13–17, 145
993a24–27, 139
993a25–27, 139, 147
994a34, 117
995a24–25, 147
995a24–b4, 147, 205
995a25–26, 140
995a25–27, 136, 141, 176
995a27–28, 147
995a27–31, 253
995a28–30, 147
995a28–32, 176
995a28–33, 230
995a29, 117
995a31–33, 147
995a33–34, 147
995a34, 117
995a34–b1, 148
995a35, 117
995b1–2, 148
995b13–14, 142, 148
995b2, 117
 995b20–27, 142
995b23–24, 177
995b2–4, 148, 177
995b31–33, 143
995b31–36, 142
996a11, 142
996a1–2, 142
996a12–12, 143
996a15–17, 142
996a4–9, 142, 143
996b1–3, 139
996b33–997a15, 143
996b8–10, 138
997a16–25, 143
997a33, 142
997a34–998a19, 143
997b35–998a6, 236
998a20–21, 142
998a23–b13, 143
998b17–999a23, 143
999a24–26, 142, 148
999a24–b17, 143
999b4–28, 242
Meteorology 
381b23–5, 158
383b20, 158
Nicomachean Ethics 
1098b14–16, 262
1145b2–7, 205
1145b3, 117
I.6, 262
On Sleep 
454b25–29, 165
On the Heavens 
279b6–7, 152
292a15, 170
292a19–22, 160
On the Soul 
403a27–b19, 235
413a9, 255
Parts of Animals 
644b25–26, 170
644b28–29, 171
Physics 
185a5–12, 19
186a10–22, 19
203b6, 272
204a34–b22, 234
204a8–34, 235
204b5–7, 234
233a21–31, 19
239b14–18, 13
Posterior Analytics 
71b19–23, 231
71b26–29, 231
76b11–16, 231
Rhetoric 
1355a33–36, 177
1355a33–5, 112
1355a36–38, 177
1356b28–35, 125
1402a17–20, 20
1402a24–6, 20
1402a3–17, 161
1403a33–4, 127
Sophistical Refutations 
 116b37, 161
165b15–22, 175
165b3, 188
165b8, 188
169b11, 188
169b25, 188
169b27–9, 175
172a22, 188
183b7, 188
279b17–21, 19
Topics, 7
100a18–20, 112
100a18–21, 113, 232
100a29–30, 19
100b21–3, 123
101a25–36, 205
101a25–b5, 233
101a34–36, 150, 177
101a34–6, 116
101a35, 177
101a36–b4, 150
101a5–10, 238
101b28–36, 121
104a12–b5, 133
104a20–5, 133
104a3–8, 122
104a8–12, 123, 232
104b12–4, 132
104b1–5, 127, 130
104b19–24, 134
104b24–28, 134
104b29–34, 134
105a34–b1, 125
 105a3–5, 134
105a37, 124
108b19–22, 119
112a14–15, 119
112b27–113a19, 133
114b25–34, 133
121b29–30, 119
129b35–130a4, 119
129b4–6, 128
141b15–28, 235
145a37–b20, 119
145a38–b20, 212
145b1–2, 6, 120, 143
145b1–20, 229
145b16–20, 6, 120, 143, 229, 273
155b33–35, 177
155b7–16, 177
158b16–23, 119
159a39–b1, 188
159a4–6, 119
159b23–33, 126
159b26–27, 188
159b8, 126
163b9–16, 177

Cicero 
Letters to Atticus 
VII.XI.3, 174
VII.XII.4, 174
VIII.XIII.2, 174
XV.IV.2, 174
XVI.VIII.2, 174
Lucullus 
100, 185
102, 190
102–103, 182
103, 190
104, 177, 180, 190
105, 190
109, 181
110, 180, 181, 182, 190
113, 187
115, 180
12, 190
13–15, 184
132, 174
133, 180
138, 174
14, 184, 185
14–15, 183
142, 185
146, 190
15, 180, 184, 186, 187
16, 186
18, 190
2.93, 277
2.94, 277
27, 180
28–29, 181
29, 180
31, 182
32, 190
33, 182
34, 190
54, 182
60, 177, 178, 179, 181
 61, 184
61–62, 185
62, 182
66, 179, 187
68, 188
69, 174
7, 177, 178, 179, 184
70, 180
72–76, 184, 186
73, 186
76, 179
76–77, 186
77, 187
78, 188, 190
8, 178
95–97, 174
95–98, 180
98, 180
99, 180, 190
On Divination (Div.) 
II 150, 177, 178
On the Duties (Off.) 
II 8, 177
On the Ends of Good and Evil (Fin.) 
II 1–2, 177
II 2, 177
V 10, 177, 178
On the Nature of the Gods (DND) 
I 11, 177
II 168, 177
On the Orator (De oratore) 
I 84, 177
III 107, 177
 III 67, 177
III 67–16, 177
III 80, 177
III 84, 177
Tusculan Disputations (Tusc.) 
II 9, 177
IV 7, 178
Varro 
16, 177, 185
43, 183
44, 180, 185, 186
44–45, 182, 186
45, 184, 187
46, 177

Damascius 
Aporiai and Solutions Concerning First Principles (Princ.) 
1, 4–7, 270
11, 15–19, 278
12, 11–13, 279
12, 13–19, 279
12, 19–21, 279
12, 3–6, 278
12, 9–10, 278
2, 6–8, 271
2, 9–20, 271
21, 18–21, 281
21, 7–8, 274
3, 21–5, 274
4, 13–15, 274
4, 15–16, 274
4, 3–5, 274
6, 11–16, 282
 7, 24–5, 276
8, 12–20, 282
8, 3–5, 276
9, 9–13, 276
I.16.5–6, 280
II, 1, 4–8, 274
Commentary of Plato’s Parmenides (in Parm.) 
IV, 115, 12–116, 8, 276
Democritus 
B11, 23
B117, 23
Diogenes Laertius 
4.67, 173
7.175, 174
7.198, 174
7.44, 174
7.82, 174
9.103, 281
9.107, 222
9.51, 19, 20, 21
9.55, 20
9.6, 10
9.69–70, 208
9.79, 210

Euclid 
Elements 
I 1,2,5, 236

Gellius 
Attic Nights (NA) 
16.2.9–13, 277
Heraclitus 
B1, 11
B10, 11
B104, 11
B114, 194
B12, 11
B123, 11
B17, 11
B18, 54
B19, 11
B204, 11
B207, 11
B28a, 11
B34, 11
B40, 11
B41, 11
B50, 11
B51, 11
B57, 11
B60, 9, 11
B61, 11, 12
B88, 12

Isocrates 
Orations 
10.2–3, 19

Leibniz 
On the Ultimate Origination of Things, 272

Melissus 
B1, 17
B1–7, 17
 B2, 17
B7, 17
B8, 17
B8.2, 18
B8.4, 18
B8.5, 19

Numenius 
Fragments 
25, 198
26, 174
75–83, 198

Parmenides 
B8, 15
Photios 
Bibliotheca 
212.169b21–9, 214
212.169b32–34, 214
212.169b38–41, 213
212.170a26–33, 214
212.170b15–19, 214
212.170b3–8, 214
Plato 
Alcibiades I 
116e–118b, 45
Apology, 38
21a4–d8, 205
21b, 186
21b4–5, 213
21b7, 30
23d, 216
30e, 42
 38a, 47
Charmides, 29, 91
157a3–6, 34
158b5–c2, 34
158d8–e1, 34
159a1–8, 34
160de, 34
166c5, 39
167b, 31
167b–169a, 34
169c, 216
169cd, 31
169d1, 216
175e–176a, 34
176b1, 34
Cratylus 
415c2–9, 216
429d1–6, 100
429d4–6, 41
Euthydemus, 5, 29
272b, 107
275d5–6, 216
277d–278b, 40
279d–280b, 40
283e–284c, 40
283e7, 100
283e7–284a8, 41
285eb–286b, 40
286b8–c3, 21
289b–d, 40
293bc, 99
295b–296d, 40
300cd, 40
 300e–301c, 40
301ab, 107
302b, 39
303a, 41
Euthyphro, 29, 91
11b, 36
11b–e, 36
15b–c, 36
Gorgias 
527d–e, 29
449b9, 40
462b–463d, 40
489bc, 39
508e6–509b1, 63–65
521e–522b, 46
Hippias Major, 29, 66
287c–d, 264
288a, 264
289d, 264
Ion 
541e, 39
Laches, 29, 91
187e–88a, 40
190c, 31
190e, 31
192ab5–8, 34
192d–193e, 32
194a8, 39
194a8–b2, 32
194ab, 32
194b, 39
194b2, 33
194bc, 31, 33
 194c, 216
195ab, 38
195b–196a, 40
196ab, 38
196b2, 216
200e, 31, 41
Laws 
731e, 194
799c, 69
Lysis, 29
216c, 216
218ab, 45
218b1, 45
Meno 
70a–80b, 29
71b9–c2, 37
71e1, 37
72a, 175
75c, 31
76a, 235
79e7–8, 208
79e7–80d4, 206
79e–80a, 36, 216
79e–80d4, 216
80a1–2, 93
80a2, 36
80ab, 37
80b, 37
80b3, 37
80b4, 212
80b4–7, 37
80bc, 38
80c6, 37
 80cd, 31
80d, 37
80d–86c, 43
84a3–d2, 206
84a–c, 44, 45
84b, 93, 175
84b–c, 147
84cd, 176
86b, 38
86b–c, 92
86d, 38
86e–87c, 230
87bc, 176
Parmenides 
127d6–7, 81
127e1–4, 15, 81
127e2, 69
127e7, 82
128a2–3, 82
128a8–b1, 81
128b1–2, 81
128c6–d6, 16
128d1, 83
128d2, 83
128d5, 81, 83
128d5–6, 81
128d6, 82
128d7, 83
128d-e, 19
128e2, 83
128e5–130a2, 83
129b1, 70
129b2, 70
129d2–130a2, 70
129d5, 70
129e, 68
129e3, 70
129e5–130a2, 86
129e6, 68, 69, 80
130b, 68
130b1–5, 73, 79
130c, 68
130c1–d9, 71
130c3, 68, 71
130c7, 68, 71
130e5, 73
131b3–6, 87
131e9, 73
132a1, 73
132b3–6, 87
132c10, 73
133a10, 87
133a11–b2, 84
133a5–7, 75
133a8, 72, 73, 75
133a8–b2, 73
133a8–c1, 75
133a9, 79
133b1, 72
133b2, 73, 79
133b4, 73, 77
133b4–6, 79, 87
133b4–c1, 75, 79
133b5–6, 74
133b7–9, 76
133b7–c1, 75
 133b8, 75, 76, 79
133c1, 79
133c3–5, 74, 77
134e9–b2, 76, 78
135a, 68
135a2–3, 68
135a3, 68, 78
135a3–4, 79
135a3–5, 87
135a3–b2, 26
135a4–5, 79
135a5, 79
135a6, 79
135a6–7, 79
135a7–b1, 79
135b1–3, 80
135b2, 79
135b5–c2, 87
135b5–c3, 74, 80
135b7, 79
135d8, 89
135d8–e4, 86
135e1–3, 84
135e–136d, 230
137b3, 85
141e10–142a3, 275
141e9–142a1, 275
142a3–7, 275
166c5, 86
Phaedo 
100d, 264
65ab, 198
68c–69d, 47
 85c, 176
99c, 176
Phaedrus 
251e2, 45
261d6–8, 19
Protagoras, 29
324d, 69
324d2–e2, 48
324d–e, 6, 31
326e, 31
330ab, 34
334de, 40
335a, 39
348c–d, 92
348cd, 46
349bc, 34
359a, 34
360e3–5, 65
360e–361d, 29
361c2–3, 55
361c2–d2, 55
361c7, 56
361cd, 29
Republic, 35
336c, 39
350cd, 45
353e12, 65
429a, 195
485a–487a, 47
489e–495b, 42
509b, 273
515c4–d7, 206
515cd, 42
 518cd, 42
524a6–b5, 205
524e2–525a2, 205
598b, 104
Sophist, 27
216a–b, 19
216d, 101
229a6, 45
230a, 95
230b–c, 92
230bc, 42
230c1, 45
230c–231b, 195
230c–d, 94
230cd, 45
232d5–e1, 20
234d, 108
236–251, 102
236d2, 102
236d–241b, 276
236d9, 103
236e1, 109
237b7–8, 276
237ce, 41
237e1–2, 276
238c9–11, 276
238d–239b, 280
238d5–8, 103
239a3, 104
239a8–10, 104
239b1–5, 104
239c9–d4, 102
239d1–4, 104
239e1–240a6, 102
240a12, 104
240c3–5, 102
240d9, 104
240e3–4, 105
241a3–4, 104
241a3–b3, 102
241b, 105
241b1, 109
241d6–7, 102
242c–243a, 105
242c4–6, 105
243b, 69
243b2–7, 107
243b7–10, 105
243c, 105, 107
243d8–e7, 107
243e1–2, 106
245a8–9, 270
249d10, 107
249d10–11, 106, 108
250a, 106
250a11–b7, 107
250ab, 106
250b8, 106
250d2–4, 107
250d7–e6, 107
250e1–2, 105
250e6–7, 102
251a–b, 107
255e–256e, 108
256a10–b4, 108
256d8–9, 108
 257b–258e, 102
258c–d, 109
259a–b, 108
259c–d, 102
259d, 108
260c1–4, 109
260c9–d3, 109
261d–264b, 109
262d2–4, 110
263d–264b, 110
264b, 110
Symposium 
201c, 264
203b2–d8, 205
204a, 45
204e, 264
209ae, 44
210a–212a, 264
210a6, 44
210e–212a, 195
215e–216c, 42, 45
215e–217a, 46
216e4–5, 193
218ab, 42
218e, 46
221e–222a, 46
255d8, 45
Theaetetus 
145d5–e9, 227
145e8, 92
148e1, 92
148e1–5, 283
148e6–7, 92
149a–151d, 193
149a3–7, 92
149a8–9, 93
149c8–d3, 283
150c4, 92
150c7–8, 193
151a5–b1, 206, 216
151a6–7, 283
151a7–8, 93
151a8, 94
151b, 93
151c4–5, 93
151c5–d3, 193
152a6–8, 21
154c, 95, 96
154c8–10, 96
154e1–5, 97
155 c1, 97
155c6–7, 97
155c–d, 96
155d5–6, 96
160e5, 93
163–4, 97
164c8–d2, 97
165b–e, 97
167–168b, 39
167d5–b3, 97
167e–168b, 42
168c2, 97
170a–171d, 98
174c5, 95
174d1, 95
175d, 216
 175d5, 95
177c–179b, 98
187d1–6, 98
187e5–7, 98
187e5–8, 96
188a–d, 99
188c–189b, 41
188c2–3, 278
189a6–13, 100
190c–e, 99
191a, 216
191a3–6, 100
197e2–3, 93
200a11–12, 99
209b, 100
210c, 92, 95
Timaeus 
55d–56b, 203
90a, 194
Plotinus 
Enneads 
I.1, 252, 258, 259, 260
I.1.1, 250
I.1.1.1–4, 259
I.1.13, 252
I.1.7, 258
I.1–7, 249
I.3.4, 257
I.4, 261, 262
I.4.1.1–4, 261
I.5, 261
I.5.1.1–2, 262
I.6.1.20–50, 264
I.6.2.11–13, 263
I.6.9.42–44, 264
I.7, 261
I.7.1.1–4, 262
I.7.1.17–28, 262
II.5, 263
II.6, 260
II.6.1, 263
II.8, 260
II.9.17.20–21, 264
III.5, 260, 261
III.5.2, 261
III.7, 260
III.7.1.8, 260
III.7.6, 260
IV.2, 258, 261
IV.2.2, 258
IV.3, 251, 253, 254, 257
IV.3.1.14–37, 254
IV.3.1.15, 254
IV.3.2, 254
IV.3.2.29–32, 254
IV.3.2.57–8, 255
IV.3.20, 255
IV.3.20.36–51, 256
IV.3.21.10–21, 255
IV.3.21.7, 255
IV.3.21–22, 255
IV.3.22, 255
IV.3.22.7–11, 255
IV.3.25–32, 256
IV.3.3.1–3, 255
IV.3.4, 255
 IV.3.5, 255
IV.3.6–8, 255
IV.3.7, 256
IV.3.7.1–12, 254
IV.3.9, 255
IV.3–4, 256
IV.3–5, 251, 258, 259
IV.4, 251, 253, 254, 256, 257, 259
IV.4.18, 259
IV.4.22.6–12, 261
IV.4.28, 259
IV.5, 251, 253, 254, 256
IV.8, 261
IV.8.1, 261
IV.8.1.23–28, 261
IV.8.2, 258, 261
IV.8.5, 261
IV.9, 257
V.1.12.1–3, 267
V.3, 257
V.3.13.1–8, 267
V.5, 251
V.7, 251
V.8, 264
VI.1.2, 263
VI.1.3.3–4, 271
VI.7, 251, 259
VI.7.32.38, 264
VI.7.4.1–6, 251
VI.7.4.7–11, 259
VI.9.3, 266
VI.9.4, 266
Plutarch 
 Against Colotes (Adv. Colotem) 
1116E–1119C, 196
1118B, 197
1121E, 198
1121F–1122A, 183
1122A, 198
1122B, 199
1125D, 200
1125DE, 200
On the Contradictions of the Stoics (Stoic. repugnan) 
1036b–c, 185
1037c, 177
On the Daimonion of Socrates 
946A, 203
On the Principle of Cold 
946A, 201
946B–948A, 201
946BC, 201
948AB, 202
949F, 202
952CD, 202
952D, 203
955A, 203
Platonic Questions 
1000A, 194
1000BC, 195
1000DE, 195
999EF, 194
Porphyry 
Life of Plotinus (V.Plot.) 
13, 257
13.10–17, 252
18.10–19, 252
Proclus 
Elements of Theology (ET) 
Prop. 7, 271
On the Parmenides (in Parm.) 
1106, 31, 274
III 815–833, 201
Protagoras 
80B1, 21
Pseudo-Aristotle 
On Melissus, Xenophanes and Gorgias (MXG) 
975a23, 19
975b25, 19
975b37, 19
979a13–18, 25
979b21–34, 25
979b34–5, 25
980a20–b9, 28
980b12–18, 28
980b18–20, 27
980b9–12, 28

Seneca 
Letters 
44, 181
88, 181
Against the Mathematicians (M) 
1.108, 209
1.125, 209
1.131, 209
1.132, 209
1.15, 209
1.160, 209
1.163, 209
1.169, 209
1.170, 209
1.18, 209
1.205, 209
1.228, 209
1.231, 209
1.232, 209
1.29, 209
1.30, 209
1.33, 209
1.35, 209
1.57, 223
1.6, 217
1.68, 209, 210
1.68–70, 277
1.7, 209
1.70, 210
1.74, 209
1.84, 209, 210
10.103, 209
10.104, 210
10.105, 207
10.107, 209
10.112, 209
10.139, 209
10.142, 209
10.144, 209
10.153, 209
10.16, 209
10.169, 209
10.17, 209
10.181, 209
10.189, 209
10.190, 209
10.205, 209
10.211, 209
10.213, 209
10.215, 209
10.237, 209
10.245, 209
10.246, 207, 209
10.247, 209, 223
10.284, 209
10.291, 209
10.292, 209
10.298, 209
10.302, 212
10.319, 209
10.337, 209
10.340, 207
10.44, 209
10.45, 209
10.5, 209
10.53, 209
10.58, 208
10.61, 209
10.67, 207
10.68, 207
10.74, 209
10.86, 212
11.1, 209
11.110–61, 219
11.167, 209
11.21, 223
11.219, 209
11.232, 209
11.234, 209
11.235, 209
11.236, 209
11.239, 209
11.243, 210
11.246, 209
11.257, 209
11.89, 209
11.96, 209
2.100, 209
2.113, 209
2.69, 209
2.89, 209
2.90, 209
2.96, 209
2.99, 212
3.1, 209
3.102, 209
3.104, 209
3.115, 209
3.48, 209
3.57–8, 237
3.60, 209
3.77, 209
3.80, 209
3.82, 209
3.98, 209
4.15, 209
4.20, 209
4.21, 209
4.22, 209
4.31, 209
5.94, 209
5.65, 209
5.89, 210
5.91, 209
6.59, 209
7.262, 209
7.264, 212
7.283, 209
7.287, 209
7.30, 207
7.303, 209
7.304, 209
7.308, 209
7.314, 209
7.337–336a, 221
7.343, 209
7.364, 209
7.378, 210
7.384, 209
7.38–45, 223
7.388, 209
7.393, 223
7.410, 212
7.435, 209
7.446, 209
7.60, 21
7.77–78, 27
7.78–9, 27
7.83–6, 28
7.87, 209
8.118, 209
8.123, 209
8.124, 209
8.125, 209
8.130, 209
8.14, 209
8.140, 209
8.156, 223
8.160, 207
8.188, 209
8.198, 209
8.244, 209
8.278, 207
8.31, 209
8.32, 209
8.336, 209
8.35, 209
8.36, 209
8.379, 210
8.393, 209
8.394, 209
8.40, 210
8.402, 209
8.437, 209
8.445, 210
8.46, 209
8.470, 210
8.481, 209
8.52, 209
8.55, 209
8.65, 209
8.75, 207
8.77, 209
8.78, 207
8.80, 207
8.87, 209
8.99, 207
9.12, 209
9.13, 209
9.194, 209
9.2, 209
9.207, 207
9.218, 209
9.258, 209
9.267, 209
9.303, 207
9.31, 209
9.311, 209
9.326, 210
9.330, 209
9.348, 209
9.350, 209
9.351, 209
9.352, 209
9.356, 210
9.357, 210
9.358, 209
9.365, 209
9.368, 210
9.414, 209
9.42, 209
9.421, 209
9.430, 209
9.433, 209
9.436, 209
9.440, 209
9.47, 210
9.49, 221
I 235, 174
VIII 159–189, 197
Outlines of Pyrrhonism (PH) 
1.10, 212
1.12, 217
1.1–3, 207
1.169, 210
1.17, 220
1.178, 212
1.180, 209
1.184, 209
1.187–205, 212
1.192, 212
1.196, 212, 213
1.205, 225
1.210, 217
1.220–35, 213
1.221–3, 215
1.226, 224
1.232–3, 226
1.232–4, 215
1.23–4, 220
1.24, 221
1.25–30, 218
1.26, 217
1.27–8, 219
1.28–9, 217
1.7, 207
1.8, 213, 216, 222
2.1–11, 221, 225
2.115, 209
2.127, 209
2.183, 210
2.197, 210
2.199, 210
2.20, 210
2.253, 277
2.255, 209
2.61, 209
2.80–96, 223
2.88, 280
2.9, 210
2.91, 280
2.95, 209
3.102, 209
3.115, 209
3.13, 209
3.134, 209
3.139, 211, 212
3.142, 209
3.157, 209
3.16, 209
3.176, 209
3.2, 221
3.22, 210
3.235–8, 219
3.238, 209, 211
3.242, 210
3.258, 209
3.259, 209
3.259–65, 210
3.266, 209
3.270, 209
3.281, 217
3.54, 212
3.55, 209
3.73, 210
3.79, 209
 3.80, 210
8, 3–5, 281
8.156, 223
I 234, 198
Simplicius 
On the Physics (in Phys.) 
475, 11–19, 235
476, 23–29, 235

Xenophon 
Memorabilia 
1.1.14, 26
 Subject Index
Academy, 213, 214
New Academy, 4, 51, 53, 172–191, 192, 196, 197, 199–200, 203–204, 213
Sceptical Academy. See New Academy
Aenesidemus, 173, 210, 213–215, 217, 218
Agonistic, 38, 41, 42
Alcinous, 195
Alexander of Aphrodisias, 228–247, 255
Allen, J., 190
Allen, R., 67–68, 72, 76, 80, 88
Ambiguity, 35, 36, 40, 148, 242
Analogy, 11, 46, 65–66, 149, 162, 178, 216
Anaxagoras, 185
Anderson, A., 59
Annas, J., 53, 207, 210, 213, 215, 221
Anonymous Theaetetus Commentator, 53, 93, 193
Antilegein, 83
Antilogic, 19
Antinomy, 13–15
Antiochus of Ascalon, 173–174
Antisthenes, 134
Aporia, 226
and (lack of) empirical/observational evidence, 155, 157–160, 164, 167–
169, 230
and amazement/marvel/wonder, 70–71, 86, 97, 205, 216, 230
and dialectical reasoning, 129, 231, 234, 259
and dilemmatic argument/reasoning. See under Argument
and epistemology, 50–53, 58–64
and erōs, 45, 205
and exegesis, 228–247
and Geach’s charge, 50–51, 59, 61–63
and hupothesis, 80–90
and methodology, 155, 156, 228, 249, 268
and Scepticism, 4–5, 10, 30, 48–60, 172–173, 180–191, 192–204, 205–
227, 252–253, 256–257, 260, 262, 265, 267–268, 280, 281
and the demand for definitions, 49–50
aporetic philosopher/philosophy, 1–6, 9–10, 16–17, 28, 44, 45, 173, 189,
207–208, 213, 214–215, 222–223, 248–249, 268
aporia-based argument. See under Argument
aporos/aporon, 1, 33, 54, 102, 209, 210
as a dilemma, 9, 15, 24, 25, 53, 161, 164–166, 210, 226, 250–251, 271–
272
as a state of mind versus the cause/object of a state of mind, 6–8, 68, 70,
77, 78, 105, 120, 122, 135, 158, 205, 214–215, 230, 252–253,
265–266, 267–268, 272–273
as labour pains, 91–95, 195, 206, 216
as pollachōs legomenon, 169
as problem. See Problems
as problēma, 7, 112, 120, 122, 170, 193, 201
at the end of an enquiry versus at the beginning of an enquiry, 49, 205–206
dilemmatic structure of, 24, 26, 210, 226, 250–251
elenctic, 43, 49, 92, 252, 253–262
(vs. labour pain aporia), 93–95
ethical significance of, 11, 29–30, 33, 35, 41, 46–47, 216–217, 262
intellectual benefits of, 29–47, 91–95, 118–119, 150, 175, 177, 205–206
necessary or merely helpful?, 45, 55–56, 137–138, 140, 147–154, 217,
257
ontological, 263–265, 267–268
refutative. See elenctic
resolution (and resolvability), 5, 12, 16, 18, 24, 44, 48, 50, 54–59, 63–66,
83, 148, 156, 158–159, 161–162, 164, 167, 168, 169, 218, 219,
222, 252–253, 255–257, 258, 262, 264–268
the term aporia and its cognates, 1–2, 3, 6–7, 30–31, 33, 48–49, 67–80,
87, 88, 95, 102, 104, 105, 112, 119, 137, 155–156, 170, 172, 174,
 206–211, 216, 229–231, 269
zetetic (versus cathartic, and versus refutative), 4, 31, 45, 49, 94, 156,
157–164, 205, 214–215, 252, 259–262, 267
Appearances, 21, 22–24, 26, 33, 155, 197, 205, 217, 219
Apprehension. See Knowledge as apprehension
Arcesilaus, 51, 172–173, 177, 180, 182–184, 186–190, 192, 193, 196–197,
198–200, 204, 215, 226
Argument 
ad hominem, 49
aporetic, 10, 12, 68–69
aporetic, negative versus positive, 10, 16, 28, 37, 93, 143–144, 164, 205,
214–215
aporia-based, 4, 5, 7–8, 9, 48–66, 251
dialectical, 7–8, 82–83, 112–136, 176–177, 231–232, 237–238, 239–241,
243, 245–247, 277–278
dilemmatic, 26, 53, 210, 226, 250–251
disputative. See Disputation
elenctic argument/testing, 4, 8, 29, 30, 31–33, 36, 38, 39, 40, 42, 44, 143,
144, 214–215
eristic, 5, 19, 24, 29–30, 36, 39–41, 96–97, 99, 100–101, 103, 108, 111,
175
fallacious, 36, 40, 106–107, 277
refutative. See elenctic argument/testing
sophistic. See eristic
Aristophanes, 38
Aristotle, 3–8, 12–13, 19, 23, 25, 85–86, 97, 112–136, 137–154, 155–171,
172–173, 175, 176–177, 178, 205–206, 216–217
aporiai about (the science of) first principles, 137
aporiai and endoxa, 118
aporiai in natural science/biology, 155–171
definition of aporia, 112, 119–121, 122, 170, 229
dialectical aporiai, 53
dialectical method, 112–136, 238
 dialectical premises, 115, 122–127, 129–131, 133–134
dialectical problems (problemata), 112–113, 114–115, 121–123, 130–135
diaporēsai (working through an aporia), 113–119, 135, 137, 147, 150, 152–
153, 172, 177, 230
embryology, 158, 167
endoxa, 7–8, 114, 115, 118, 121, 123–129, 131, 132–133, 140–142, 146,
150, 176–177, 206, 235
Armstrong, H., 248
Ataraxia (imperturbability). See under Scepticism
Atomists, 13
Atticus, 174
Aubenque, P., 4, 146, 230, 231

Babut, D., 192

Ballériaux, O., 102
Barnes, J., 12, 14, 16, 187, 207, 208, 210, 211, 215, 218, 223, 231, 232
Being, 17, 25–28, 281–282
not being, 98, 100, 101–110, 276
Belief 
consistency among beliefs, 21, 36, 49, 93, 94–95, 212
credibility of, 7, 51–52, 60, 62
false. See false judgement under Judgement
justified, 48, 52, 54, 55, 58, 60, 62–63
reliable, 50, 52, 58–63, 118
Bell, I.H., 139, 232
Benson, H., 33, 36, 40, 81, 99
Berti, E., 124, 127, 128, 129
Bett, R., 214, 221
Beversluis, J., 39, 41
Binder, G., 21
Blumenthal, H.J., 253
Bobzien, S., 277
Bolton, R., 129
Bonazzi, M., 213, 215
Bonelli, M., 233
Boylan, M., 155
Boys-Stones, G., 203
Brague, R., 37
Bréhier, É., 248, 249
Brisson, L., 68, 75
Brittain, C., 174, 184, 192
Broadie, S., 241, 242, 243
Brown, L., 101, 106, 107, 108, 109
Brunschwig, J., 124, 129, 208, 221, 235
Buchheim, T., 24
Burnet, J., 68, 70, 75
Burnyeat, M.F., 93, 99, 181, 184, 186, 194, 211, 219, 280

Calogero, G., 24
Caluori, D., 255, 258
Campbell, L., 194
Carneades, 51, 173, 177, 180, 188, 190, 192, 193, 196, 197–198, 200, 202,
204
Castagnoli, L., 219, 280
Catulus, 174, 184
Chance, T., 40
Change and alteration, 11, 17–19, 26, 69, 97, 106–108, 160–161, 167, 169,
201, 241, 242
Charrue, P., 249
Chiaradonna, R., 245
Chrysippus, 178, 185, 200, 202–203, 277
Cicero, 50, 53, 173–175, 178–180, 182–191, 198, 277
Cleary, J., 140
Clitomachus, 173, 180, 190–191
Combès, J., 269, 270, 275
Common sense/commonsense, 10, 12, 13, 17, 18, 23, 26, 216
Consistency. See under Belief
Contradiction, 20–28, 32, 48–49, 70, 81, 84, 102–103, 104, 105, 108–109,
115, 127, 134, 145, 217, 280
as a source of aporia, 7, 9–19
compelling, 7
Cooper, J.M., 52, 139, 184
Corax of Syracuse, 20
Cornford, F.M., 75, 88, 93, 108, 194
Corrigan, K., 249
Cotton, A., 36
Couloubaritsis, L., 208
Coussin, P., 188
Crivelli, P., 97, 103, 104, 106, 107, 108, 109, 110
Crubellier, M., 143, 209, 230

Damascius, 267, 269–284

De Waal, T., 75
Decleva Caizzi, F., 208, 214
Decretum/decreta, 180, 181, 187, 190
Definition, 39, 54, 101, 107, 121–136, 229, 235
demand for definitions in Plato, 31, 34, 49–50, 57–66
Delcominette, S., 106
Democritus, 23, 182, 185, 186, 196
Dialectic, 19, 26, 36, 39, 97, 112–136, 175–176, 188–189, 228, 231–241,
277–278, 280
and aporia, 7–8, 82–83, 137, 170, 269
dialectical argument. See under Argument
dialectical method. See under Method
dialectical reasoning, 231–241, 259–260
dialegesthai, 87
Didymus the Blind, 21, 22, 23, 24
Diès, A., 75, 88
Dilemma. See under Aporia
Dillon, J., 253, 269
Diogenes Laertius, 20, 207, 214, 223
Disputation, 8, 19, 36, 53, 114, 118, 122, 124, 125, 129, 177, 178, 182, 189,
194
Dodds, E.R., 273
Dogma, 4, 178, 179, 180, 181, 185–186, 188–189, 192, 197, 198, 204, 206–
227, 248–249, 268, 282
Donini, P.L., 241
Duncombe, M., 75
Düring, I., 229
Dye, G., 236

Education, 33, 41, 42, 79–80, 94

Elenchos (‘refutation’, ‘testing’). See under Argument
Elenctic testing. See under Argument
Ellis, J., 245
Embarassment, 32, 38–39, 45
Emillson, E., 255
Empedocles, 185
Enquiry, 118, 180–191
and aporia, 1–6, 48–66, 216–227
and examples, 50–51, 61–63
and scepticism, 5, 175–182, 191, 205–227
aporia and the possibility of, 37–38
open-minded, 178–184, 191, 224
scientific versus philosophical, 114–115
Epistemology, 174, 189, 203
in Plato, 29, 50–53, 58–66
Equipollence (isosthenia), 53, 212–213, 222, 224, 226, 230
Eristic. See under Argument
Erler, M., 3
Eros, 45
Erotic longing, 45
Eubulides’ liar, 9
Examples, 32, 61–63, 120, 234
Excellence. See Virtue
Exhortation, 46, 54, 225
Experience, 5–6, 13, 17, 18, 22, 32–33, 34, 36–37, 40, 45, 74–75, 92, 97, 98,
145, 216, 217–221, 230, 257, 278, 282–284
everyday experience, 10, 15–16, 185, 237

Fallacy, 40, 106, 161

Fazzo, S., 228
Fine, G., 38, 59, 99, 221
First principles, 137, 149–154, 204, 231–239, 267, 269, 270, 272, 273
Forster, M., 49
Frede, M., 42, 49, 101, 103, 107, 145, 187

Galen, 173, 177, 178, 200

Galpérine, M.-C., 269, 274
Gatti, M.L., 248, 249
Geach, P., 50–51, 59, 61–63
Geach’s charge. See aporia and Geach’s charge, under Aporia
Gellius, 277
Generated versus ungenerated, 17, 25–26, 170, 241–244
Geometry/geometrical, 43, 175–176, 233, 235–238
Gesché, A., 22
Gill, M.L., 75, 87
Goldschmidt, V., 30
Gonzales, F., 47
Gorgias, 10, 19, 24–28, 36, 211
Granger, H., 12
Grgić, F., 217, 218, 221, 222, 224
Griffin, M., 174
Gronewald, M., 22
Gurtler, G., 256
Halper, E., 138, 140, 142
Hankinson, R.J., 213
Harmony, 11
Henry, D., 162
Heraclitus, 1, 3, 9–12, 16, 28, 54, 134, 194, 198
Hiley, D., 222
Hirzel, R., 184, 190
Hoffman, H., 13
Hoffmann, P., 273, 280
Hope, 1, 54, 175, 217
Husserl, E., 32
Hylomorphism/hylomorphic, 241, 243, 245–247

Ignorance. See knowledge versus ignorance, under Knowledge

Impasse. See Aporia
Ineffable, the, 274–284
Innate conceptions, 195–196
Intellect versus senses. See under Senses
Introspection, 34
Ioppolo, A.M., 206
Irwin, T., 34, 35, 59, 155, 230, 232, 234
Isocrates, 19

Judgement, 23, 30, 145, 149–150, 194–195, 258

false, 44–45, 95–96, 98–100, 104, 109–110
suspension of, 44–45, 178–179, 180–181, 183, 187–188, 199, 203, 207,
209–212, 217–227

Kalligas, P., 250

Karamanolis, G., 172
Karbowski, J., 156
Kerferd, G., 19, 24, 25
King, R., 256
Kirk, G., 11
Klein, J., 37
Knowledge, 31–35, 98–100, 118, 140–142, 149–154, 179
and aporia, 48–66
as apprehension, 186, 188–191
attainability of, 5–6, 51–52, 75, 79, 85
conceit of, 11–12, 44–45
criterion of, 5–6, 31
infallible, 6
knowing-how, 31–32
limitis of. See under Reason
nothing can be known, 180–191
versus belief, 48–66
versus ignorance, 36, 43, 278–280
Knuuttila, S., 221

Lacydes, 174
Laks, A., 137, 139, 140, 141, 143, 147, 209, 230
Language, 32, 104
and reality, 269–284
and thought, 27, 269–284
speaking correctly (orthologia), 96, 103–105, 108
Leibniz, 272
On the Ultimate Origination of Things, 272
Leigh, F., 106
Lennox, J., 158
Leszl, W., 141
Liesenborghs, L., 21
Limited-and-unlimited, 13–18, 25–26, 234–235
Lloyd, A., 263
Lloyd, G.E.R., 157
Logos, 11, 16, 20, 64, 102, 103, 105, 109–110, 161, 209, 275, 276, 280–282
Long, A.A., 95
Love. See Eros
Lucullus, 174, 178, 179, 180, 181–182, 184–185, 186–187, 189

Machuca, D., 207, 218, 219, 220

Mackenzie, M., 12, 31
Mackie, J.L., 280
Madigan, A., 4, 137, 139, 140, 143, 144, 150, 228, 229, 240
Makin, S., 19
Man-measure doctrine, 21–24, 26
Mansfeld, J., 22, 25, 173
Mansion, S., 139
Marchand, S., 222
Mates, B., 212, 215
Matter, 160–161, 163, 241–247, 255
prime matter, 243
Matthews, G., 4, 37, 43, 155, 158, 193, 273
McDowell, J., 92, 96, 100, 194
Meinwald, C., 68, 76, 85
Mejer, J., 22
Melissus of Samos, 10, 17–19, 25, 134
Meno’s paradox, 37–38, 40, 43, 223, 225
Method, 67, 86
dialectical, 112–136, 228, 230, 232, 238–240, 246
scientific, 231–239
Methodology. See and methodology, under Aporia
Metrodorus of Chios, 185
Migliori, M., 67, 88
Miller, M., 68
Moller, D., 218
Moreschini, C., 75
Morsink, J., 162
Motion, 12–16, 24, 28, 95, 201, 208, 241
Motte, A., 3, 205, 212
Mueller, I., 236, 237

Nagel, T., 282, 284

Natural order, 11
Natural science, 19, 168
Nature, 11, 19, 23, 24, 157–158, 168–169, 183, 193, 203, 243
Nehamas, A., 41
Nightingale, A., 216
Noble, C., 259
Notomi, N., 101
Numenius of Apamea, 174, 177, 192, 198, 200

O’Meara, D., 249, 264, 269, 273, 274

Olfert, C.M., 206, 223, 224
One-and-many, 13–19, 25, 69–70, 73, 81–84, 103–104, 107, 117, 143
Opposites, 6–7, 14, 27, 29, 128, 130–131, 143, 144, 164–165, 201–202, 229
compresent. See under Plato
Opsomer, J., 4, 192
Owen, G.E.L., 67, 232
Owens, J., 137, 139, 140, 141

Palmer, J., 17, 19, 24, 26, 102, 105, 184, 222, 223
Paradox, 9, 11, 12, 15–16, 19, 23–24, 37–38, 40–41, 43, 108, 123, 124, 129–
130, 134, 179–181, 196, 221, 223, 225
Parmenides of Elea, 13, 15, 16–18, 102, 105, 109, 185, 198
Particulars, 27, 61–63, 124, 146, 241–242
sensible, 83, 241
Perin, C., 219, 220
Perplexity. See Aporia
Peterson, S., 75
Philo of Larissa, 174, 184, 190–191, 197, 198
Philosophy/philosophia 
dogmatic, 4, 181, 185, 186, 188, 204, 205–227, 248–249, 268, 282
 sceptical. See Scepticism
speculative, 4, 5, 25
systematic, 4, 5, 248
versus sophistry, 29–30, 35, 39–41, 91, 101–105, 108, 110–111, 194
Placitum/placita, 179–181
Plato, 3–6, 9, 16, 19, 20–21, 23, 25, 26–28, 29–47, 48–66, 67–90, 91–111,
147, 172–173, 174–176, 185–186, 205–206, 216
and scepticism, 48–66, 215
aporetic (as opposed to doctrinal or sceptical) reading of aporetic
dialogues, 29–30, 35
aporetic dialogues, 9, 29–30, 35, 37, 39–47, 67, 80, 91, 205, 249
aporetic language, 67–80, 98, 99, 105, 213–214
aporetic structure, 67, 80–90, 91
articulation and elenctic testability requirement, 30–35
compresent opposites, 70–71, 72, 80–84, 86
correct speaking (orthologia), 96, 103–105, 108
early dialogues, 8, 29–66, 92, 205, 251, 252, 253, 260, 267
forms, 26–27, 67, 68, 69–80, 83–90, 195, 196, 199, 201, 215, 266, 270
Good, the, 45, 262, 264–265, 270, 273
hupothesis, 80–90
kinds (and communion of kinds), 101–102, 106, 107–108
marking off forms, 73, 74, 76, 77, 79–80, 84–85
One, the, 69, 86, 266–267, 270, 273–278
philosopher versus sophist, 29–30, 35, 39–41, 91, 101–105, 108
relation of participation, 70, 73, 75, 76, 83
scope of forms, 71–72, 78, 201
sensibles, 70–71, 73, 76, 270
separate forms, 69, 71–78, 79, 89
what-is-it questions, 57–58, 260
Plotinus, 4, 5, 248–268, 269, 270, 273
Plurality, 13, 15–17, 24, 69, 276
Plutarch, 177, 178, 179, 183, 185, 192–204, 213
Politis, V., 4, 8, 29, 31, 37, 48, 49, 50, 52, 54, 58, 64, 66, 94, 95, 138, 139,
146, 148–149, 205, 212, 273
Polito, R., 213
Porphyry, 13, 251–252, 253, 271, 274
Primavesi, O., 124
Principles. See First principles
transcendent principles, 269
Prior, W., 59
Problems, 6, 12, 16, 29, 30, 31, 43, 70, 91, 95–100, 112–115, 121–123, 130–
136, 139–140, 143–144, 155, 156–157, 172, 176, 193, 208, 230–
231, 242–246, 249, 265, 269, 278–280
about being, 27, 105–108
about not being, 101–105, 109–110
not-whether-but-how, 95–96, 97–98, 102–103
used both philosophically and eristically, 96–97, 103
Proclus, 201, 270, 273
Protagoras, 2, 10, 19–24, 26, 29, 39, 40, 42, 55, 63, 65, 95, 98, 236
Protreptic, 39–41
Puzzlement. See Aporia
Puzzles. See Problems
Pyrrho of Elis, 173, 207, 213–214, 216, 217
Pyrrhonism. See under Scepticism

Rappe, S., 269, 280

Rashed, M., 245
Raven, J., 11
Reale, G., 140
Reality, 108, 152, 153–154, 195, 202–203
as a whole, 269–272, 273–275 (see also, Totality of what there is)
Reason 
limits of, 267, 269, 280–282, 284
Reasoning. See argument
demonstrative, 67, 214, 231–247
Reasons 
conclusive versus inconclusive, 56–58, 63
conflict of, 48, 56, 65–66, 126–136, 163, 166–169, 217–227
inter-personal versus intra-personal conflict of, 65
of iron and adamant, 64–66
real versus apparent, 4, 6–8, 49
Recollection, 38, 43, 44, 93, 196
Reductio ad absurdum, 19, 27
as indirect proof versus as means of inducing aporia, 15
Reeve, C.D.C., 59
Relativism, 21
Rhetoric, 20, 26, 95, 124–125, 175, 195, 214, 238
Rickless, S., 68
Rist, J., 248
Robinson, R., 67, 89, 93
Rodriguez, E., 81, 85, 89
Rutten, C., 3, 205, 212

Santas, G., 50, 59

Sayre, K., 68, 85
Scepticism 
Academic/Platonic, 4, 53, 172–174, 177–191, 200, 204, 206, 213–214,
222, 223, 224
and aporia. See under Aporia
and ataraxia (imperturbability), 216–226
and enquiry. See under Enquiry
anti-scepticism in Plato, 49–50
Cartesian, 212
in Plato. See under Plato
Pyrrhonian, 4, 5, 173, 181, 183, 205–227, 236, 262, 268, 269, 280–281
Schofield, M., 11
Science, 115–117, 137–142, 144, 146, 147–154, 196, 228, 231–234, 236,
238
Scolnicov, S., 68, 81
Scott, D., 37
Sedley, D., 91, 93, 99, 100, 184
Self-image, 33, 36, 42, 45–46
Self-improvement, 35, 41
Self-knowledge, 44, 94, 216
Self-refutation, 219, 280, 283
Seneca, 20, 181
Senses, 182, 185, 197–198
versus intellect, 16–17, 19, 23
Sextus Empiricus, 21, 24, 27, 181, 197, 198, 206–227, 236–237, 277, 280,
281
Shame, 45–46, 94, 216
Sharples, R., 38, 228, 245
Sihvola, J., 221
Silence, 175, 201, 276–280, 281
Simplicius, 13–14, 162, 235
Socrates, 9, 25–26, 172–173, 174–176, 178, 185–186
and his daimonion, 192–193, 194, 195
as midwife, 91–95, 193, 196, 206, 282–284
pretence of ignorance, 11, 41, 50, 175, 182, 183, 184, 185, 186, 193, 196,
253
Socratic irony, 11, 184, 193
Socratic questioning/interrogation/examination. See elenctic
argument/testing, under Argument
Sophia. See Wisdom
Sophists, 10, 19, 39–41, 42, 96, 100, 101–105, 108, 109–110, 194, 196, 226
Soul, 34–35, 38, 45, 46, 47, 52, 87, 120–121, 158, 160–161, 195, 197, 199,
206, 216, 229–230, 243, 251–252, 253–262, 265–267, 274, 283
Spinelli, E., 215
Sprague, R., 32, 40
Steel, C., 4
Stevens, A., 137
Stoics, 6, 172–173, 174, 178–179, 185, 187–191, 192, 195, 196–197, 198–
200, 202, 213, 221, 236, 238, 257–258, 260–262, 264, 265, 277
Striker, G., 180, 188, 206, 212, 218, 219
Su, J., 4
Szaif, J., 38, 47

Tarrant, H., 213, 215

Taylor, D.E., 219
Therapy, 42, 214, 217, 219
Thorsrud, H., 221
Timon of Phlius, 10, 213, 217, 218
to be/is, 17–19
Totality of what there is, 17
Truth, 6, 9, 21–24, 38, 42, 43–47, 54, 127, 129, 130, 140–141, 143, 148–149,
152, 172, 176–177, 178–180, 181–183, 186–187, 189, 191, 192,
193, 207, 211, 214, 219, 220–221, 223–227, 230–231, 232–233,
238–240, 249, 281–282
Tsouna, V., 46
Turning-around, 41–42, 46, 47, 280–281
Tuzzo, T., 34

Understanding, 16–17, 23, 26–27, 79, 197–198

and aporia, 10–12, 145–147, 172, 175–176, 179

Vagueness, 35, 36, 40, 77, 148

van Ophuijsen, J.M., 233
Varro, 173–174, 182–187, 189
Virtue, 2, 29, 34, 175, 183, 214, 220, 250, 262
intellectualist understanding of, 31–35, 36–38, 47
whether or not virtue can be taught, 54, 62–63
Vitrac, B., 236
Vlastos, G., 14, 40, 49, 59
Vogt, K., 206, 208, 219, 220, 221, 224
Vollkmann-Schluck, K.H., 249
von Kutschera, F., 68
von Wilamowitz-Moellendorff, U., 75

Weische, A., 172

Wesensschau, 32
Westerink, L.G., 269, 270, 275
Wilde, O., 175
Wisdom, 2, 35, 40, 45, 47, 51–52, 92, 195, 213–214
conceit of, 11–12, 194
Wittgenstein, L., 32, 282
Włodarczyk, M.A., 223
Woodruff, P., 4, 22, 59, 214

Xenophanes, 185
Xenophon, 25

Zeller, E., 99
Zeno of Citium, 178, 182, 183, 186–187, 189, 199
Zeno of Elea, 3, 10, 12–17, 19, 23–24, 25–27, 28, 67, 69–71, 73, 77, 81–87,
95, 236