Textbook Philosophy and Theory of Artificial Intelligence 2017 Vincent C Muller Ebook All Chapter PDF

Philosophy and Theory of Artificial
Intelligence 2017 Vincent C. Müller

Visit to download the full and correct content document:
https://textbookfull.com/product/philosophy-and-theory-of-artificial-intelligence-2017-vi
ncent-c-muller/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
Beginning Game AI with Unity: Programming Artificial

Intelligence with C# Sebastiano M. Cossu
https://textbookfull.com/product/beginning-game-ai-with-unity-
programming-artificial-intelligence-with-c-sebastiano-m-cossu/
AI IA 2017 Advances in Artificial Intelligence XVIth

International Conference of the Italian Association for
Artificial Intelligence Bari Italy November 14 17 2017
Proceedings 1st Edition Floriana Esposito
https://textbookfull.com/product/ai-ia-2017-advances-in-
artificial-intelligence-xvith-international-conference-of-the-
italian-association-for-artificial-intelligence-bari-italy-
november-14-17-2017-proceedings-1st-edition-floriana-esp/
Artificial Intelligence Systems Based on Hybrid Neural

Networks: Theory and Applications Michael Zgurovsky
https://textbookfull.com/product/artificial-intelligence-systems-
based-on-hybrid-neural-networks-theory-and-applications-michael-
zgurovsky/
Artificial Intelligence and Robotics Huimin Lu
https://textbookfull.com/product/artificial-intelligence-and-
robotics-huimin-lu/
Ethics Of Artificial Intelligence S. Matthew Liao
https://textbookfull.com/product/ethics-of-artificial-
intelligence-s-matthew-liao/
Artificial Intelligence By Example Develop machine

intelligence from scratch using real artificial
intelligence use cases Denis Rothman
https://textbookfull.com/product/artificial-intelligence-by-
example-develop-machine-intelligence-from-scratch-using-real-
artificial-intelligence-use-cases-denis-rothman/
Practical Artificial Intelligence with Swift From

Fundamental Theory to Development of AI Driven Apps
Mars Geldard
https://textbookfull.com/product/practical-artificial-
intelligence-with-swift-from-fundamental-theory-to-development-
of-ai-driven-apps-mars-geldard/
Philosophy 2nd Edition QuickStudy Academic M C O

Donnell Phd
https://textbookfull.com/product/philosophy-2nd-edition-
quickstudy-academic-m-c-o-donnell-phd/
Practical Artificial Intelligence: Machine Learning,

Bots, and Agent Solutions Using C# 1st Edition Arnaldo
Pérez Castaño
https://textbookfull.com/product/practical-artificial-
intelligence-machine-learning-bots-and-agent-solutions-
using-c-1st-edition-arnaldo-perez-castano/
Studies in Applied Philosophy,
Epistemology and Rational Ethics
Vincent C. Müller Editor
Philosophy
and Theory
of Artificial
Intelligence 2017
Studies in Applied Philosophy, Epistemology
and Rational Ethics
Volume 44
Series editor
Lorenzo Magnani, University of Pavia, Pavia, Italy
e-mail: lmagnani@unipv.it
Editorial Board
Atocha Aliseda
Universidad Nacional Autónoma de México (UNAM), Coyoacan, Mexico
Giuseppe Longo
Centre Cavaillès, CNRS—Ecole Normale Supérieure, Paris, France
Chris Sinha
School of Foreign Languages, Hunan University, Changsha, P.R. China
Paul Thagard
Waterloo University, Waterloo, ON, Canada
John Woods
University of British Columbia, Vancouver, BC, Canada
Studies in Applied Philosophy, Epistemology and Rational Ethics (SAPERE) publishes new developments
and advances in all the fields of philosophy, epistemology, and ethics, bringing them together with a cluster of
scientific disciplines and technological outcomes: ranging from computer science to life sciences, from
economics, law, and education to engineering, logic, and mathematics, from medicine to physics, human
sciences, and politics. The series aims at covering all the challenging philosophical and ethical themes of
contemporary society, making them appropriately applicable to contemporary theoretical and practical
problems, impasses, controversies, and conflicts. Our scientific and technological era has offered “new” topics
to all areas of philosophy and ethics – for instance concerning scientific rationality, creativity, human and
artificial intelligence, social and folk epistemology, ordinary reasoning, cognitive niches and cultural evolution,
ecological crisis, ecologically situated rationality, consciousness, freedom and responsibility, human identity
and uniqueness, cooperation, altruism, intersubjectivity and empathy, spirituality, violence. The impact of such
topics has been mainly undermined by contemporary cultural settings, whereas they should increase the demand
of interdisciplinary applied knowledge and fresh and original understanding. In turn, traditional philosophical
and ethical themes have been profoundly affected and transformed as well: they should be further examined as
embedded and applied within their scientific and technological environments so to update their received and
often old-fashioned disciplinary treatment and appeal. Applying philosophy individuates therefore a new
research commitment for the 21st century, focused on the main problems of recent methodological, logical,
epistemological, and cognitive aspects of modeling activities employed both in intellectual and scientific
discovery, and in technological innovation, including the computational tools intertwined with such practices, to
understand them in a wide and integrated perspective.
Advisory Board
W. Park, Daejeon, South Korea
A. Abe, Chiba, Japan A. Pereira, São Paulo, Brazil
H. Andersen, Copenhagen, Denmark L. M. Pereira, Caparica, Portugal
O. Bueno, Coral Gables, USA A.-V. Pietarinen, Helsinki, Finland
S. Chandrasekharan, Mumbai, India D. Portides, Nicosia, Cyprus
M. Dascal, Tel Aviv, Israel D. Provijn, Ghent, Belgium
G. D. Crnkovic, Göteborg, Sweden J. Queiroz, Juiz de Fora, Brazil
M. Ghins, Lovain-la-Neuve, Belgium A. Raftopoulos, Nicosia, Cyprus
M. Guarini, Windsor, Canada C. Sakama, Wakayama, Japan
R. Gudwin, Campinas, Brazil C. Schmidt, Le Mans, France
A. Heeffer, Ghent, Belgium G. Schurz, Dusseldorf, Germany
M. Hildebrandt, Rotterdam, The Netherlands N. Schwartz, Buenos Aires, Argentina
K. E. Himma, Seattle, USA C. Shelley, Waterloo, Canada
M. Hoffmann, Atlanta, USA F. Stjernfelt, Aarhus, Denmark
P. Li, Guangzhou, P.R. China M. Suarez, Madrid, Spain
G. Minnameier, Frankfurt, Germany J. van den Hoven, Delft, The Netherlands
M. Morrison, Toronto, Canada P.-P. Verbeek, Enschede, The Netherlands
Y. Ohsawa, Tokyo, Japan R. Viale, Milan, Italy
S. Paavola, Helsinki, Finland M. Vorms, Paris, France
More information about this series at http://www.springer.com/series/10087

Vincent C. Müller
Editor
Philosophy and Theory

of Artificial Intelligence 2017
123
Editor
Vincent C. Müller
Interdisciplinary Ethics Applied (IDEA)
Centre
University of Leeds
Leeds, West Yorkshire
UK
ISSN 2192-6255 ISSN 2192-6263 (electronic)

Studies in Applied Philosophy, Epistemology and Rational Ethics
ISBN 978-3-319-96447-8 ISBN 978-3-319-96448-5 (eBook)
https://doi.org/10.1007/978-3-319-96448-5
Library of Congress Control Number: 2018948694
© Springer Nature Switzerland AG 2018

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made. The publisher remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Editorial Note
The papers in this volume result from the 3rd conference on the “Philosophy and
Theory of Artificial Intelligence” (PT-AI) 4–5 November 2017 which I organised in
Leeds where I am a university fellow—for details on the conference, see http://
www.pt-ai.org/.
For this conference, we had 77 extended abstract submissions by the deadline,
which were reviewed double-blind by two to four referees. A total of 28 submis-
sions, i.e. 36%, were accepted for presentation. We also accepted 18 posters to be
presented. The invited speakers were Thomas Metzinger, Mark Sprevak, José
Hernández-Orallo, Yi Zeng, Susan Schneider, David C. Hogg and Peter Millican.
All papers and posters were submitted in January at full length (two pages for
posters) and reviewed another time by at least two referees among the authors. In
the end, we have 32 papers here that represent the current state of the art in the
philosophy of AI. We grouped the papers broadly into three categories: “Cognition–
Reasoning–Consciousness”, “Computation–Intelligence–Machine Learning” and
“Ethics–Law”. This year, we see a significant increase in ethics, more work on
machine learning, perhaps less on embodiment or computation—and a stronger
“feel” that our area of work has entered the mainstream. This is also evident from
more papers in “standard” journals, more book publications with mainstream phi-
losophy presses and more philosophers from neighbouring fields joining us, such as
philosophy of mind or philosophy of science. These are encouraging developments,
and we are looking forward to PT-AI 2019!
We gratefully acknowledge support from the journal Artificial Intelligence, and
the IDEA Centre at the University of Leeds.
May 2018 Vincent C. Müller
v
Contents
Cognition - Reasoning - Consciousness

Artificial Consciousness: From Impossibility to Multiplicity . . . . . . . . . . 3
Chuanfei Chin
Cognition as Embodied Morphological Computation . . . . . . . . . . . . . . . 19
Gordana Dodig-Crnkovic
“The Action of the Brain”: Machine Models and Adaptive
Functions in Turing and Ashby . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Hajo Greif
An Epistemological Approach to the Symbol Grounding Problem . . . . . 36
Jodi Guazzini
An Enactive Theory of Need Satisfaction . . . . . . . . . . . . . . . . . . . . . . . . 40
Soheil Human, Golnaz Bidabadi, Markus F. Peschl, and Vadim Savenkov
Agency, Qualia and Life: Connecting Mind and Body Biologically . . . . 43
David Longinotti
Dynamic Concept Spaces in Computational Creativity for Music . . . . . . 57
René Mogensen
Creative AI: Music Composition Programs as an Extension
of the Composer’s Mind . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Caterina Moruzzi
How Are Robots’ Reasons for Action Grounded? . . . . . . . . . . . . . . . . . 73
Bryony Pierce
Artificial Brains and Hybrid Minds . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Paul Schweizer
vii
viii Contents
Huge, but Unnoticed, Gaps Between Current AI

and Natural Intelligence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Aaron Sloman
Social Cognition and Artificial Agents . . . . . . . . . . . . . . . . . . . . . . . . . . 106
Anna Strasser
Computation - Intelligence - Machine Learning

Mapping Intelligence: Requirements and Possibilities . . . . . . . . . . . . . . 117
Sankalp Bhatnagar, Anna Alexandrova, Shahar Avin, Stephen Cave,
Lucy Cheke, Matthew Crosby, Jan Feyereisl, Marta Halina,
Bao Sheng Loe, Seán Ó hÉigeartaigh, Fernando Martínez-Plumed,
Huw Price, Henry Shevlin, Adrian Weller, Alan Winfield,
and José Hernández-Orallo
Do Machine-Learning Machines Learn? . . . . . . . . . . . . . . . . . . . . . . . . 136
Selmer Bringsjord, Naveen Sundar Govindarajulu, Shreya Banerjee,
and John Hummel
Where Intelligence Lies: Externalist and Sociolinguistic
Perspectives on the Turing Test and AI . . . . . . . . . . . . . . . . . . . . . . . . . 158
Shlomo Danziger
Modelling Machine Learning Models . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
Raül Fabra-Boluda, Cèsar Ferri, José Hernández-Orallo,
Fernando Martínez-Plumed, and M. José Ramírez-Quintana
Is Programming Done by Projection and Introspection? . . . . . . . . . . . . 187
Sam Freed
Supporting Pluralism by Artificial Intelligence: Conceptualizing
Epistemic Disagreements as Digital Artifacts . . . . . . . . . . . . . . . . . . . . . 190
Soheil Human, Golnaz Bidabadi, and Vadim Savenkov
The Frame Problem, Gödelian Incompleteness, and the Lucas-Penrose
Argument: A Structural Analysis of Arguments About Limits
of AI, and Its Physical and Metaphysical Consequences . . . . . . . . . . . . 194
Yoshihiro Maruyama
Quantum Pancomputationalism and Statistical Data Science: From
Symbolic to Statistical AI, and to Quantum AI . . . . . . . . . . . . . . . . . . . 207
Yoshihiro Maruyama
Getting Clarity by Defining Artificial Intelligence—A Survey . . . . . . . . 212
Dagmar Monett and Colin W. P. Lewis
Epistemic Computation and Artificial Intelligence . . . . . . . . . . . . . . . . . 215
Jiří Wiedermann and Jan van Leeuwen
Contents ix
Will Machine Learning Yield Machine Intelligence? . . . . . . . . . . . . . . . 225

Carlos Zednik
Ethics - Law
In Critique of RoboLaw: The Model of SmartLaw . . . . . . . . . . . . . . . . 231
Paulius Astromskis
AAAI: An Argument Against Artificial Intelligence . . . . . . . . . . . . . . . . 235
Sander Beckers
Institutional Facts and AMAs in Society . . . . . . . . . . . . . . . . . . . . . . . . 248
Arzu Gokmen
A Systematic Account of Machine Moral Agency . . . . . . . . . . . . . . . . . . 252
Mahi Hardalupas
A Framework for Exploring Intelligent Artificial Personhood . . . . . . . . 255
Thomas B. Kane
Against Leben’s Rawlsian Collision Algorithm
for Autonomous Vehicles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
Geoff Keeling
Moral Status of Digital Agents: Acting Under Uncertainty . . . . . . . . . . 273
Abhishek Mishra
Friendly Superintelligent AI: All You Need Is Love . . . . . . . . . . . . . . . . 288
Michael Prinzing
Autonomous Weapon Systems - An Alleged Responsibility Gap . . . . . . 302
Torben Swoboda
Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
Cognition - Reasoning - Consciousness
Artificial Consciousness: From Impossibility
to Multiplicity
Chuanfei Chin(&)
Department of Philosophy, National University of Singapore,

Singapore 117570, Singapore
phiccf@nus.edu.sg
Abstract. How has multiplicity superseded impossibility in philosophical

challenges to artificial consciousness? I assess a trajectory in recent debates on
artificial consciousness, in which metaphysical and explanatory challenges to
the possibility of building conscious machines lead to epistemological concerns
about the multiplicity underlying ‘what it is like’ to be a conscious creature or be
in a conscious state. First, I analyse earlier challenges which claim that phe-
nomenal consciousness cannot arise, or cannot be built, in machines. These are
based on Block’s Chinese Nation and Chalmers’ Hard Problem. To defuse such
challenges, theorists of artificial consciousness can appeal to empirical methods
and models of explanation. Second, I explain why this naturalistic approach
produces an epistemological puzzle on the role of biological properties in
phenomenal consciousness. Neither behavioural tests nor theoretical inferences
seem to settle whether our machines are conscious. Third, I evaluate whether the
new challenge can be managed through a more fine-grained taxonomy of
conscious states. This strategy is supported by the development of similar tax-
onomies for biological species and animal consciousness. Although it makes
sense of some current models of artificial consciousness, it raises questions
about their subjective and moral significance.
Keywords: Artificial consciousness Machine consciousness

Phenomenal consciousness Scientific taxonomy Subjectivity
1 Introduction
I want to trace a trajectory in recent philosophical debates on artificial consciousness. In

this trajectory, metaphysical and explanatory challenges to the possibility of building
conscious machines are supplanted by epistemological concerns about the multiplicity
underlying ‘what it is like’ to be a conscious creature or be in a conscious state. Here
artificial consciousness refers, primarily, to phenomenal consciousness in machines
built from non-organic materials. Like most of the philosophers and scientists whom I
discuss, I will follow Block (1995) in using the concept of phenomenal consciousness
to refer subjective experience. By Block’s definition, the sum of a state’s phenomenal
properties is what it is like to be in that conscious state, and the sum of a creature’s
phenomenal states is what it is like to be that conscious creature. The paradigms of such
conscious states include having sensations, feelings, and perceptions.

V. C. Müller (Ed.): PT-AI 2017, SAPERE 44, pp. 3–18, 2018.
https://doi.org/10.1007/978-3-319-96448-5_1
4 C. Chin
Many surveys on artificial consciousness stress that this sub-field in artificial

intelligence research has multiple interests (Gamez 2008; Holland and Gamez 2009;
Reggia 2013; Scheutz 2014). Its research programmes aim to build machines which
mimic behaviour associated with consciousness, machines with the cognitive structure
of consciousness, or machines with conscious states. Often a distinction is drawn
between strong artificial consciousness, which aims for conscious machines, and weak
artificial consciousness, which builds machines that simulate some significant corre-
lates of consciousness. Of course, a research programme may nurture interests in both
strong and weak artificial consciousness; and the same model may be used to inves-
tigate both strong and weak artificial consciousness.
I shall focus on philosophical challenges to strong artificial consciousness. First, in
the next section, I will analyse two earlier challenges which claim that phenomenal
consciousness cannot arise, or cannot be built, in machines. These are based on Block’s
Chinese Nation and Chalmers’ Hard Problem. To defuse such challenges, we can
appeal to empirical methods and models of explanation. Second, I will explain why this
naturalistic approach leads to an epistemological puzzle on the role of biological
properties in phenomenal consciousness. Neither behavioural tests nor theoretical
inferences seem to settle whether our machines are conscious. Third, I will evaluate
whether the new challenge can be handled by a more fine-grained taxonomy of con-
scious states. This strategy is supported by the development of more fine-grained
taxonomies for biological species and animal consciousness. Although it makes sense
of some current models of artificial consciousness, it raises questions about their
subjective meaning and moral status.
2 The Impossibility of Artificial Consciousness
The literature on artificial consciousness contains several philosophical challenges to

the possibility of building conscious machines (Bishop 2009; Gamez 2008; McDermott
2007; Prinz 2003; Reggia 2013; Scheutz 2014). Such challenges draw on philosophical
arguments about the nature of consciousness and our access to it. One set of challenges
is against the metaphysical possibility of artificial consciousness. These are based on
the provocative thought experiments in Block (1978), Searle (1980), Maudlin (1989),
which suggest that machines, however sophisticated in functional or computational
terms, cannot be conscious. Another set of challenges is directed at the practical
possibility of building conscious machines. They are based on philosophical claims,
made by McGinn (1991), Levine (1983), Chalmers (1995), about our ignorance of how
conscious states arise from physical states. According to these challenges, we can
hardly expect to produce consciousness in machines if we cannot explain it in human
brains.
Most theorists of artificial consciousness are not troubled by such challenges. In his
survey, Scheutz (2014) describes two attitudes that support this stance. Here is how I
understand them. First, some theorists hold a pragmatic attitude towards the concept of
consciousness. They define this concept in an operational way, in terms of the pro-
cesses and principles which psychologists take to underlie consciousness. Their aim is
to use these processes and principles to improve performance in machines. They do not
Artificial Consciousness: From Impossibility to Multiplicity 5
want to replicate consciousness, so they need not worry if consciousness can arise, or
be produced, in machines. This attitude particularly suits those whose research lies in
weak artificial consciousness. Second, other theorists hold a revisionary attitude. They
want to refine or replace the concept of consciousness through their empirical inves-
tigation of the underlying processes and principles identified by psychologists. In doing
so, they wish to contribute to both psychology and philosophy. For instance, their
models of the relevant processes and principles may enable new psychological
experiments and produce new theories of consciousness. These may, in turn, influence
philosophical intuitions and views about consciousness.
I take this last point to mean that empirical research into strong artificial con-
sciousness need not be halted by the intuitions and views current in philosophy. To
demonstrate this, I will show that theorists of artificial consciousness can appeal to
empirical methods and models of explanation to defuse some philosophical challenges.
In particular, I will look at how we can respond to two challenges to the possibility of
building conscious machines – one based on Block’s Chinese Nation thought experi-
ment, the other on Chalmers’ Hard Problem of consciousness.1 Even those theorists
who are less inclined to take philosophical challenges seriously can clarify their
methodological commitments by considering these responses. Moreover, in the next
section, I will show why the commitments underlying these responses lead to an
epistemological puzzle which should interest all theorists of artificial consciousness.
(a) The first challenge centres on the nature of consciousness. It suggests that
conscious machines cannot be built since machines cannot be conscious. More pre-
cisely, it suggests that the functional properties realisable by machines are not sufficient
for consciousness. In Block’s thought experiment, a billion people in China are
instructed to duplicate the functional organisation of mental states in a human mind.
Through radio connections and satellite displays, they control an artificial body just as
neurons control a human body. They respond to various sensory inputs into the body
with appropriate behavioural outputs. But, according to Block (1978), we are loath to
attribute consciousness to this system: ‘there is prima facie doubt whether it has any
mental states at all – especially whether it has what philosophers have variously called
“qualitative states,” “raw feels,” or “immediate phenomenological qualities”’ (73). If
our intuition about the Chinese Nation is sound, then consciousness requires more than
the functional properties discovered in psychology. If so, the machines that realise only
these functional properties cannot be conscious.
I do not think that we need to defer to this intuition about the Chinese Nation.
Rather we should use empirical methods to uncover more about the nature of con-
sciousness. Our best research – in psychology, neuroscience, and artificial con-
sciousness – may determine that functional properties at a coarse-grained psychological
level are sufficient for consciousness. Or it may determine that functional properties at a
more fine-grained neurological level are necessary too. Whether the relevant properties
are realisable in our machines is a further question, also to be determined by empirical
1
I learnt especially from the responses offered in Prinz (2003) and Gamez (2008). I have put aside
challenges based on Searle’s Chinese Room thought experiment: they are analysed exhaustively in
the literature on artificial consciousness, with what looks to be diminishing returns. One response to
these challenges can be modelled after my response in (a).
6 C. Chin
investigation. None of this research should be pre-empted a priori by what our intuition
says in a thought experiment and what that supposedly implies about the possibility of
conscious machines.
Even Block would agree on this methodological point. He notes that, intuitively,
the human brain also does not seem to be the right kind of system to have what he calls
‘qualia’, the subjective aspect of experience. So our intuition, on its own, cannot be
relied on to judge which system does or does not have qualia. According to Block, we
can overrule intuition if we have independent reason to believe that a system has qualia,
and if we can explain away the apparent absurdity of believing this. Here his qualm
about a system like the Chinese Nation rests mainly on our lack of a theoretical ground
to believe that it has qualia. No psychological theory that he considers seems to explain
qualia. That is why he insists of the system: ‘any doubt that it has qualia is a doubt that
qualia are in the domain of psychology’ (84). To assuage this qualm, we need to build
an empirical theory of consciousness which explains qualia and evaluates whether
Chinese Nations, machines, and other systems have them.
(b) The second challenge directly addresses our explanation of consciousness. It
suggests that we cannot build machines to be conscious even if machines can be
conscious. According to Chalmers (1995), the Hard Problem we face is to explain how
conscious experiences arise from physical processes and mechanisms in the brain. He
distinguishes this from easy problems which require us to explain various psycho-
logical functions and behaviours in terms of computational or neural mechanisms. We
have yet to solve the Hard Problem because we do not know how consciousness is
produced in the human brain. But, until we do so, we cannot produce consciousness in
a machine except by accident. Here is how Gamez (2008) sums up this line of rea-
soning based on our ignorance: “if we don’t understand how human consciousness is
produced, then it makes little sense to attempt to make a robot phenomenally con-
scious” (892).
I find two related reasons to reject this challenge. First, the production of con-
sciousness may not require its explanation. Through empirical investigation, we may be
able to produce consciousness without explaining it in terms of physical processes and
mechanisms in the brain. If so, it suffices for us to create in machines the conditions
which give rise to consciousness in humans; we need not understand, in philosophi-
cally satisfying terms, how the conditions do this. Our research to produce con-
sciousness in machines may then help our research to explain consciousness in
humans. This cross-fertilisation between research programmes would be in keeping
with the revisionary attitude that Scheutz highlights.
Second, even if we need some kind of explanation to enable production, the
explanation of consciousness in empirical terms may not require a solution to the Hard
Problem. Through their empirical theories, scientists do not aim to explain, in some
metaphysically intelligible way, how the properties of a phenomenon ‘arise from’ other
properties at lower levels. Instead, they aim to establish a theoretical identity for the
phenomenon in terms of its underlying properties (Block and Stalnaker 1999;
McLaughlin 2003; Prinz 2003; Shea and Bayne 2010). (I say more about how this
applies to consciousness science in the next section.) To build their theories, scientists
draw correlations between levels, tying together some higher-level and lower-level
properties. In the biological and psychological sciences, what requires this kind of
explanation between levels depends on context: it is often determined by which

properties, at higher or lower levels, appear anomalous (Wimsatt 1976; Craver 2009,
Chap. 6; Prinz 2012, 287–8). These practices suggest that an empirically successful
theory of consciousness need not fill in the gap between phenomenal and physical
properties – at least, not in the terms defined by Chalmers’ Hard Problem.
3 The Multiplicity in Phenomenal Consciousness
I have shown how empirical methods and models of explanation can defuse philo-
sophical challenges to the possibility of artificial consciousness. They allow us to
counter intuitions drawn from thought experiments on the nature of consciousness, and
to undercut arguments derived from our ignorance of how conscious states arise from
physical states. By appealing to these empirical methods and models, we adopt a
naturalistic approach to the study of artificial consciousness. We use empirical meth-
ods, as far as possible, to answer questions about the nature of consciousness and our
access to it. We thereby allow empirical discoveries about phenomenal consciousness
to inform our conceptual understanding of artificial consciousness. But that naturalistic
approach produces a different philosophical challenge, arising from what we discover
to be the multiplicity underlying consciousness. This new challenge to artificial con-
sciousness is epistemological: it suggests that, even if we can build conscious
machines, we cannot tell that the machines are conscious.
The challenge rests on our difficulty in determining the role of biological properties
in phenomenal consciousness. Unless we determine their role, we cannot discover
whether our machines, lacking at least some of these properties, are conscious. Several
philosophers analyse this difficulty (Block 2002; Papineau 2002, Chap. 7; Prinz 2003,
2005; Tye 2016, Chap. 10). Yet their arguments are largely ignored by theorists of
artificial consciousness. I will focus on Prinz’s arguments – since they arise naturally
from his work on an empirical theory of consciousness and are addressed directly to
theorists of artificial consciousness.
Prinz begins by analysing, at the psychological level, the contents of our conscious
states and the conditions under which they become conscious. Following Nagel, he
considers having a perspective to be fundamental to consciousness: ‘We cannot have a
conscious experience of a view from nowhere’ (2003, 118). In his analysis, humans
experience the world, through our senses, ‘from a particular vantage point’. So the
contents of our consciousness are both perceptual and perspectival. These contents
become conscious when we are paying attention. When these contents become avail-
able for our deliberation and deliberate control of action, they enable our flexible
responses to the world. Putting together these hypotheses, Prinz proposes that con-
sciousness arises in humans when we attend to phenomena such that our perspectival
perceptual states become available for deliberation and deliberate control of action.
Next, by drawing on empirical studies, Prinz maps these contents and conditions of
conscious states onto the computational and neural levels. In information processing, the
contents of consciousness seem to lie at the intermediate level. Our intermediate-level
representations are ‘vantage-point specific and coherent’ (2003, 119). They are distinct
from higher-level representations which are too abstract to preserve perspective,
8 C. Chin
and lower-level representations which are too local to be coherent. In computational

models of cognition, attention is a process that filters representations onto the next stage,
while deliberate control is handled by working memory, a short-term storage capacity
with executive abilities. In the human brain, these computational processes are imple-
mented by a neural circuit between perceptual centres in the temporal cortex, attentional
centres in the parietal cortex, and working memory centres in the frontal cortex (2003,
119; 2005, 388). Prinz (2012) cites several lines of evidence indicating that gamma
vectorwaves play the crucial role in these brain regions. So, according to his latest theory,
consciousness arises in us ‘when and only when vectorwaves that realize intermediate-
level representations fire in the gamma range, and thereby become available to working
memory’ (293). That is, in empirical terms, a good candidate for the neurofunctional basis
of consciousness in humans.
Despite this progress, Prinz (2003, 2005) highlights an epistemological limitation,
which is independent of whatever empirical theory of consciousness we settle on. He
argues that we cannot determine if our biological properties are constitutive of con-
sciousness. So we cannot discover if our machines, which will lack at least some of
these properties, are conscious. This is the basis of his pessimism about research in
strong artificial consciousness: ‘It simply isn’t the case that scientific investigations into
the nature of consciousness will make questions of machine consciousness disappear.
Even if scientific theories of consciousness succeed by their own standards, we must
remain agnostic about artificial experience’ (117).
Like others who share his pessimism, Prinz cites the in-principle failure of beha-
vioural tests to settle these questions (Prinz 2003, IV; Block 2002; Papineau 2002,
Chap. 7, 2003). How do we find constitutive properties of consciousness? The standard
method is to test for what Prinz calls ‘difference-makers’ (121). It involves changing
processes at a tested level while keeping constant processes at other levels. If this
change makes a difference to conscious behaviour in humans, then some properties at
this tested level are constitutive of consciousness. Suppose that it is technically possible
to substitute silicon chips for neurons in the human brain. And suppose that it is
nomologically possible to do so while keeping constant the relevant processes at the
psychological and computational levels.2 This surgically altered person will become a
‘functional duplicate of a normal person with a normal brain’ (123). By design, the
functional duplicate will behave exactly as conscious humans do – reporting pain,
showing signs of anger, apparently ‘seeing sunsets and smelling roses’. Yet our current
tests for consciousness centre on behaviour. So we do not have a genuine test for
consciousness in the duplicate. We cannot, by these tests, tell if our properties at the
biological level are constitutive of consciousness.
I agree with Prinz (2003) that this thought experiment highlights a ‘serious epis-
temological problem’ (130). Indeed, I believe that he and others understate the depth of
2
This is a common idealisation in the thought experiment. In reality, we will find more than one
psychological level and more than one computational one (Prinz 2003, 120–1). During chip
replacement, we are more likely to keep constant processes at less fine-grained psychological and
computational levels. The epistemological difficulty with testing remains, though it is made more
complicated. Elsewhere, in Chin (2016), I analyse more complicated versions of the multiple-kinds
problem in consciousness science; see also Irvine (2013), Chap. 6.
the problem. They focus on the failure of behavioural tests to discover if biological
properties make a difference to conscious states. Prinz claims that this ‘method of
difference-makers seems to be the only way to find out what levels matter’ (130). Yet,
like other philosophers, he also recommends that we use inference to the best expla-
nation to establish a theoretical identity for consciousness (116).3 He does not explain
why this theoretical inference cannot clarify the role of biological properties in con-
sciousness and, thereby, improve the current tests for consciousness.
Let me make these connections explicit through the multiple-kinds problem shown
in Fig. 1. As the thought experiment suggests, we will discover at least two functional
structures responsible for conscious behaviour in humans. One is a neurofunctional
structure, such as that identified in Prinz’s theory. Another is a functional structure that
abstracts away from some biological mechanisms in the neurofunctional structure.
Therefore, the kind defined by the neurofunctional structure (kind2) is nested within the
kind defined by the more abstract functional structure (kind1). Kind1 includes conscious
humans and our functional duplicates, while kind2 excludes the functional duplicates.
So which is the structure of consciousness? Which structure defines a kind formed by
all and only conscious beings?
Fig. 1. The multiple kinds in phenomenal consciousness
Prinz’s argument shows that current tests, based on behaviour, cannot solve this
multiple-kinds problem. I want to extend this argument, to show why inference to the
best explanation does not help. Both the neurofunctional structure and the more
abstract structure are correlated with consciousness in humans. Both are also
3
Other philosophers include Block and Stalnaker (1999), McLaughlin (2003), Shea and Bayne
(2010), Allen and Trestman (2016), Sect. 4.3.
10 C. Chin
systematically related to conscious behaviour in humans. By focusing on the systematic

relations between the neurofunctional structure, consciousness in humans, and their
conscious behaviour, we can support an identity between consciousness and the neu-
rofunctional structure. But this move is ad hoc, classifying our functional duplicates by
fiat as not conscious. On the other hand, by focusing on the equally systematic relations
between the more abstract functional structure, consciousness in humans, and their
conscious behaviour, we can support an identity between consciousness and that
structure. Yet this is equally ad hoc, re-classifying the duplicates by fiat as conscious.
Neither hypothesis offers a simpler explanation. Whether we identify consciousness
with the neurofunctional structure or the more abstract structure, we must invoke both
structures to account for the total explananda. If we identify consciousness with the
neurofunctional structure, then we must use the more abstract structure to explain why
the duplicates share the same behaviour as humans even though the duplicates do not
have human brains. If we identify consciousness with the more abstract structure, then
we must use the neurofunctional structure to explain how the more abstract structure is
implemented differently in conscious humans and their duplicates. The first hypothesis
interprets consciousness as only one implementation of the more abstract structure,
while the second interprets the neurofunctional structure as only one implementation of
consciousness. So the familiar norms of explanatory simplicity do not help to choose
between these hypotheses. That is why the multiple-kinds problem seems intractable. If
we cannot solve this problem, then we cannot tell whether the biological properties that
our machines lack are constitutive of consciousness. And, therefore, we cannot tell
whether our machines are conscious.
4 The Development of Scientific Taxonomies
I have shown why the naturalistic approach that defuses earlier philosophical chal-
lenges on artificial consciousness produces an epistemological puzzle on the role of
biological properties in consciousness. Through empirical investigation, we will dis-
cover multiple functional structures underlying consciousness in humans. Neither
behavioural tests nor theoretical inferences are able to pick out one structure from
among them, in order to define a kind formed by all and only conscious beings. Unless
we solve this multiple-kinds problem, we cannot determine whether the biological
properties that our machines lack are constitutive of consciousness. In this section, I
want to examine how other scientists develop more fine-grained taxonomies to manage
their multiple kinds. Then I will evaluate how theorists of artificial consciousness can
use this taxonomic strategy.
How does the multiple-kinds problem arise elsewhere? One prominent instance is
what biologists call the ‘species problem’.4 When biologists try to classify organisms
4
This problem is analysed by both biologists and philosophers: see the surveys in Coyne and Orr
(2004); Cracraft (2000); Ereshefsky (2010, 2017); Richards (2010). I also learnt from the analysis in
LaPorte (2004), though we come to different conclusions. Richards (2010) argues that the problem
goes back to pre-Darwinian times: Darwin himself was confronted by ‘a multiplicity of species
concepts’ (75).
into species, they discover multiple structures underlying biodiversity. These structures
centre on interbreeding, genetic or phenotypic similarity, ecological niche, evolutionary
tendency, or phylogeny. They lead to conflicting definitions of what a species is.
Different structures define overlapping kinds, consisting of different populations of
organisms. According to the biologists Coyne and Orr (2004), at least nine species
definitions remain ‘serious competitors’. Three of them are often mentioned in the
philosophical literature: the Biological Species Concept (BSC), the Phylogenetic
Species Concept (PSC), and the Ecological Species Concept (ESC).5 They focus,
respectively, on three primary processes involved in evolution: sexual reproduction,
descent from common ancestry, and environmental selection pressures. Of the three,
which defines the nature of species?
Proponents of the BSC, the PSC, and the ESC sometimes claim that their definition
of species is the ‘best’.6 But, in practice, biologists choose between these definitions
according to their empirical interests. As de Queiroz (1999) explains, ‘they differ with
regard to the properties of lineage segments that they consider most important, which is
reflected in their preferences concerning species criteria’ (65). Their choice of the BSC,
the PSC, or the ESC allows them to investigate the wider explanatory structures
associated, respectively, with sexual reproduction, descent from common ancestry, or
ecological niche. For instance, those who are interested in the history of life prefer the
PSC over the BSC because they believe that reproductive isolation is ‘largely irrelevant
to reconstructing history’ (Coyne and Orr 2004, 281). Those who are interested in the
explanation of biodiversity reject the PSC because they see phylogeny as ‘largely
irrelevant to understanding the discreteness of nature’. Instead they use the BSC to
study populations that sexually reproduce or use the ESC to study adaptive zones in
ecology.
The result is a more fine-grained taxonomy of species, which can be used to
manage the multiple kinds found within biodiversity. Biologists now distinguish
between species which arise from interbreeding, species which arise from phylogenetic
connection, and species which arise from environmental selection (Ereshefsky 2010).
As Fig. 2 shows, the BSC and the PSC tend to define overlapping kinds of populations.
When genealogically distinct populations can reproduce with each other, the popula-
tions of a phylogenetic species are nested within the populations of an interbreeding
species. Through their taxonomy, biologists can clarify the relations between these
kinds and demarcate the explanatory structures involving these kinds.
5
The BSC defines species as ‘groups of interbreeding natural populations that are reproductively
isolated from other such groups’ (Mayr 1969). The PSC defines them as the ‘smallest diagnosable
cluster of individual organisms within which there is a parental pattern of ancestry and descent’
(Cracraft 1983). The ESC defines them as ‘a lineage (or a closely related set of lineages) which
occupies an adaptive zone minimally different from that of any other lineage in its range and which
evolves separate from all lineages outside its range’ (Van Valen 1976).
6
As Cracraft (2000) warns, ‘the notion of “best” is always relative’ (10). He urges us to ‘look hard at
the context of what best might mean’, including how general in application a definition is meant to
be, and whether a more general definition is always more useful.
12 C. Chin
Fig. 2. Two overlapping kinds of biological species
With the more fine-grained taxonomy in place, what matters to biological expla-
nation is not whether the BSC or the PSC offers the ‘best’ definition of species. Rather
biologists have to ensure that those who are interested in interbreeding species not
confuse classifications with those who are interested in phylogenetic species. In a
context with shared interests, such confusion is unlikely to arise. For instance, most
biologists interested in sexual reproduction and its effects focus on interbreeding
species. Their interests already pick out these relevant kinds from the overlapping ones
associated with sexual reproduction, descent from ancestry, and environmental selec-
tion pressures. In a context with competing interests, biologists can avoid misunder-
standing by making explicit reference to either interbreeding species, phylogenetic
species, or ecological species. However, in some general contexts, biologists need not
specify the kinds to which they refer. They may be keen to make generalisations across
different branches of biology (Brigandt 2003). So their claims apply uniformly to
interbreeding species, phylogenetic species, and ecological species.
The multiple-kinds problem also afflicts debates on animal consciousness. Here it
lies closer to our epistemological puzzle on artificial consciousness. For animal con-
sciousness, the problem arises because we discover at least two cognitive structures
underlying consciousness in humans. Both structures are responsible, in different ways,
for conscious behaviour in humans. I will follow how Godfrey-Smith (2016a, b) dis-
tinguishes these structures. The first involves simple modes of information processing
associated with pain and other primitive bodily feelings, such as thirst and feeling short
of breath. This structure enables us to respond to actual and potential injury with
flexible non-reflexive behaviour. The second structure involves more sophisticated
modes of information processing which integrate information from different senses and
bodily feelings, through the use of memory, attention, and executive control.
According to some theories of cognition, this structure allows us to model the world
before responding to it.
Figure 3 shows that these two cognitive structures define two overlapping kinds of
animals. The kind of animals with cognitive integration is nested within the kind with
primitive bodily feelings, because cognitive integration requires more machinery, such
as memory, attention, and executive control. So which is the cognitive structure of
consciousness? Which structure defines a kind formed by all and only conscious
animals? If cognitive integration is necessary for consciousness, then only animals with
memory, attention, and executive control count as conscious. But if primitive bodily
feelings are sufficient for consciousness, many more animals count as conscious, so
long as they have the sensorimotor capacities associated with primitive bodily feelings.
Faced with this multiple-kinds problem, Godfrey-Smith (2016a, b) proposes a more
fine-grained taxonomy of subjective experiences in animals. There are at least two
Fig. 3. Two overlapping kinds of animals
kinds of subjective experiences. The basic kind, which evolved first, consists of
experiences of pain and other primitive bodily feelings; the complex kind, which
evolved later, consists of experiences which integrate information from different senses
and bodily feelings. Both kinds of subjective experiences are found in conscious
humans: ‘Much human experience does involve the integration of different senses,
integration of the senses with memory, and so on, but there is also an ongoing role for
what seems to be old forms of experience that appear as intrusions into more organized
kinds of processing’ (2016b, 500). Through his taxonomy, we can clarify the relations
between both kinds of experiences and demarcate the explanatory structures involving
both kinds.
With the more fine-grained taxonomy in place, we can see that what matters in the
explanation of animal behaviour is not whether the basic or complex kind of subjective
experiences counts as conscious. Rather theorists of animal consciousness can focus on
either kind of experiences according to their empirical interests, so long as their ter-
minology does not obscure the differences between both kinds. For instance, Godfrey-
Smith classifies only experiences with cognitive integration as conscious: ‘“Con-
sciousness” is something beyond mere subjective experience, something richer or more
sophisticated’ (2016a, 53). Animals which experience pain and other primitive bodily
feelings have qualia; it feels like something to be them. But, without cognitive inte-
gration, they do not count for him as conscious: ‘I wonder whether squid feel pain,
whether damage feels like anything to them, but I do not see this as wondering whether
squid are conscious’ (2016b, 484). As he acknowledges, other theorists with different
interests tend to equate qualia with phenomenal consciousness: ‘If there is something it
feels like to be a system, then the system is said to have a kind of consciousness’ (483–
4). In turn, these theorists have to distinguish phenomenal consciousness from other,
more sophisticated, kinds of consciousness that require cognitive integration.
How might this taxonomic strategy address the epistemological puzzle on artificial
consciousness? We can develop a more fine-grained taxonomy of conscious states, in
order to manage the multiplicity that troubles theorists of artificial consciousness. If
Prinz is right, then we need to distinguish at least two kinds of states. The first consists
of neurofunctional states, such as those specified in his theory of consciousness. Our
functional duplicates do not have this kind of states. The second consists of functional
states that abstract away from some biological mechanisms in the neurofunctional
states; both humans and the duplicates share this kind of states. With this taxonomy, we
can clarify the relations between the neurofunctional and functional states, then
demarcate the explanatory structures involving both kinds of states. What matters in
explaining humans and duplicates is not whether the neurofunctional or functional
states count as conscious. Rather theorists of consciousness can focus on either kind of
14 C. Chin
states according to their empirical interests, so long as their terminology does not
obscure the differences between both kinds of states. Those who classify only the
neurofunctional states as conscious still need to acknowledge the role of the functional
states, which explain why the duplicates behave in ways that indicate consciousness in
humans. Those who classify the functional states as conscious still need to acknowl-
edge the role of the neurofunctional states; they explain how the functional states are
realised in humans.
This analysis brings out an epistemological difference between the case of bio-
logical species and that of artificial consciousness. Biologists are now confident that
interbreeding species, phylogenetic species, and ecological species play significant
explanatory roles. They know that the kinds associated with the BSC, the PSC, and the
ESC are involved in different explanatory structures associated with sexual reproduc-
tion, ancestral descent, and ecological niche. In contrast, we do not yet know, in any
precise terms, the states that will play significant explanatory roles in research on
artificial consciousness. However, this difference does not invalidate our use of the
taxonomic strategy. We need only begin with a provisional taxonomy of conscious
states to explore the different explanatory structures that interest us. As we discover
more about these explanatory structures, we can refine the taxonomy so that it reflects,
in more precise terms, the computational and biological processes cited in our expla-
nations. That is similar to how biologists developed their taxonomy for species.
Indeed, this taxonomic strategy can already make sense of some current models of
artificial consciousness. Some theorists suggest that building the right computational
processes into machines is sufficient to make them conscious. For instance, Dehaene
et al. (2017) propose that machines are conscious if they can select information for
global broadcasting, making it flexibly available for computations, and if they can self-
monitor those computations. To support their proposal, they claim that a machine with
both computational processes will behave ‘as though it were conscious’ (492). They
also cite evidence suggesting that subjective experience in humans ‘appears to cohere
with’ global broadcasting and self-monitoring (492). Other theorists believe that
building the right biological processes into machines is necessary to make them con-
scious. Haladjian and Montemayor (2016) connect consciousness to biological pro-
cesses in humans that endow them with emotion and empathy. So, in their view,
machines designed purely to compute with artificial intelligence will not have sub-
jective experiences. According to Godfrey-Smith (2016b), machines can have sub-
jective experiences only if they have some functional properties associated with ‘living
activity’ (505). For him, these properties include the robustness and adaptability typical
of complex biological systems in humans.
From our perspective, these models of artificial consciousness need not come into
conflict. Rather we can see them as jointly clarifying the more fine-grained taxonomy
of conscious states needed in research on artificial consciousness. On one hand,
Dehaene et al. (2017) are investigating the kind of states which are defined purely in
computational terms without reference to biological mechanisms; in particular they are
interested in the explanatory structures associated with global broadcasting and self-
monitoring. On the other hand, Haladjian and Montemayor (2016), Godfrey-Smith
(2016b) are interested in another kind of states, defined partly in biological terms; they
raise different difficulties for realising such states in machines.
5 Conclusion
In this paper, I assessed a trajectory in which multiplicity superseded impossibility in

philosophical challenges to artificial consciousness. First, I tackled two earlier chal-
lenges which claim that phenomenal consciousness cannot arise, or cannot be built, in
machines. The first challenge, from the nature of consciousness, is based on Block’s
Chinese Nation thought experiment. The second challenge, from the explanation of
consciousness, is based on Chalmers’ Hard Problem. I showed how a naturalistic
approach, appealing to empirical methods and models of explanation, can defuse these
challenges. To discover if machines can be conscious, we should rely on theories of
consciousness developed through empirical methods, rather than the intuitions about
consciousness provoked by thought experiments. To explain consciousness in empir-
ical terms, we need not supply a philosophically satisfying account of how phenomenal
properties arise from physical ones.
Second, I explained why this naturalistic approach leads to an epistemological
puzzle on the role of biological properties in phenomenal consciousness. Through
empirical investigation, we will discover multiple functional structures underlying
consciousness in humans. As several philosophers argued, behavioural tests cannot
pick out one structure from among them, in order to define a kind formed by all and
only conscious beings. I argued that inference to the best explanation cannot help too.
If we cannot solve this multiple-kinds problem, then we cannot determine whether the
biological properties that our machines lack are constitutive of consciousness. We also
cannot determine whether these machines are conscious.
Third, I evaluated whether a taxonomic strategy used in other sciences can address
this new challenge. To manage the overlapping kinds which they cite in explanations,
theorists of biological species and animal consciousness develop more fine-grained
taxonomies. I argued that, similarly, theorists of artificial consciousness can develop a
more fine-grained taxonomy of conscious states, which distinguishes between the
neurofunctional states specified in an empirical theory of consciousness and the
functional states that abstract away from some biological mechanisms in the neuro-
functional states. Such a taxonomy enables us to clarify the relations between both
kinds of states and demarcate the explanatory structures involving both kinds. In
addition, I argued that this taxonomic strategy helps to make sense of current models of
artificial consciousness, including those which require only computational states and
those which require partly biological states. We can interpret them as models for
investigating different kinds of conscious states.
This strategy presents us with three related challenges, on the explanatory, sub-
jective, and moral significance of the kinds in any new taxonomy. First, we need to
establish that these kinds of states play significant explanatory roles in research on
artificial consciousness. This is primarily an empirical challenge, depending on theo-
rists of artificial consciousness to explore different explanatory structures that interest
us. Second, we need to examine the subjective significance of these kinds of states.
Thus far, we have construed a conscious state’s phenomenal properties as capturing
‘what it is like to be’ in that state. But this construal does not help to discriminate what
the multiple kinds mean in subjective terms. We may do so by investigating the
16 C. Chin
capacities and interactions made possible by the underlying structures that define these
kinds. For instance, some basic structures may support what it is like to be an artificial
patient, while others may support what it is like to be an artificial agent. Third, we need
to explore the moral significance of these kinds of states. In what ways do the artificial
patients count as moral patients whose suffering we must ameliorate? In what ways do
the artificial agents count as moral agents whose lives we must attend to?
Acknowledgements. I thank Arzu Gokmen, Michael Prinzing, and Kaine Yeo for their sug-
gestions. Abhishek Mishra, Susan Schneider, Paul Schweitzer, and Alexandra Serrenti com-
mented on the talk. This research was supported by an NUS Early Career Award.
References
Allen, C., Trestman, M.: Animal Consciousness. In: Zalta, E.N. (ed.) The Stanford Encyclopedia
of Philosophy, Winter 2016 (2016). https://plato.stanford.edu/archives/win2016/entries/
consciousness-animal/. Accessed 10 Jan 2018
Bishop, J.M.: Why computers can’t feel pain. Minds Mach. 19(4), 507–516 (2009)
Block, N.: Troubles with functionalism. In: Block, N. (ed.) Consciousness, Function, and
Representation: Collected Papers, Volume 1 (2007), pp. 63–101. MIT Press, Cambridge
(1978)
Block, N.: On a confusion about the function of consciousness. In: Block, N. (ed.)
Consciousness, Function, and Representation: Collected Papers, Volume 1 (2007),
pp. 159–213. MIT Press, Cambridge (1995)
Block, N.: The harder problem of consciousness. In: Block, N. (ed.) Consciousness, Function,
and Representation: Collected Papers, Volume 1 (2007), pp. 397–433. MIT Press, Cambridge
(2002)
Block, N., Stalnaker, R.: Conceptual analysis, dualism, and the explanatory gap. Philos. Rev. 108
(1), 1–46 (1999)
Brigandt, I.: Species pluralism does not imply species eliminativism. Philos. Sci. 70(5), 1305–
1316 (2003)
Chalmers, D.J.: Facing up to the problem of consciousness. J. Conscious. Stud. 2(3), 200–219
(1995)
Chin, C.: Borderline Consciousness, Phenomenal Consciousness, and Artificial Consciousness:
A Unified Approach, DPhil thesis. University of Oxford (2016)
Coyne, J.A., Orr, H.A.: Speciation: a catalogue and critique of species concepts. In: Rosenberg,
A., Arp, R. (eds.) Philosophy of Biology: An Anthology, pp. 272–292. Wiley-Blackwell,
Oxford (2004)
Cracraft, J.: Species concepts and speciation analysis. In: Johnston, R.F. (ed.) Current
Ornithology, pp. 159–187. Springer, New York (1983)
Cracraft, J.: Species concepts in theoretical and applied biology: a systematic debate with
consequences. In: Wheeler, Q.D., Meier, R. (eds.) Species Concepts and Phylogenetic
Theory: A Debate, pp. 3–14. Columbia University Press, New York (2000)
Craver, C.F.: Explaining the Brain. Oxford University Press, Oxford (2009)
Dehaene, S., Lau, H., Kouider, S.: What is consciousness, and could machines have it? Science
358(6362), 486–492 (2017)
Ereshefsky, M.: Species, taxonomy, and systematics. In: Rosenberg, A., Arp, R. (eds.)
Philosophy of Biology: An Anthology, pp. 255–271. Wiley-Blackwell, Oxford (2010)
Ereshefsky, M.: Species. In: Zalta, E.N. (ed.) The Stanford Encyclopedia of Philosophy, Fall
2017 (2017). https://plato.stanford.edu/archives/fall2017/entries/species/. Accessed 10 Jan
2018
Gamez, D.: Progress in machine consciousness. Conscious. Cogn. 17(3), 887–910 (2008)
Godfrey-Smith, P.: Animal evolution and the origins of experience. In: Smith, D.L. (ed.) How
Biology Shapes Philosophy: New Foundations for Naturalism, pp. 51–71. Cambridge
University Press, Cambridge (2016a)
Godfrey-Smith, P.: Mind, matter, and metabolism. J. Philos. 113(10), 481–506 (2016b)
Haladjian, H.H., Montemayor, C.: Artificial consciousness and the consciousness-attention
dissociation. Conscious. Cogn. 45, 210–225 (2016)
Holland, O., Gamez, D.: Artificial intelligence and consciousness. In: Banks, W.P. (ed.)
Encyclopedia of Consciousness, pp. 37–45. Academic Press, Oxford (2009)
Irvine, E.: Consciousness as a Scientific Concept: A Philosophy of Science Perspective. Springer,
Dordrecht (2013)
LaPorte, J.: Natural Kinds and Conceptual Change. Cambridge University Press, Cambridge
(2004)
Levine, J.: Materialism and qualia: the explanatory gap. Pac. Philos. Q. 64(October), 354–361
(1983)
Maudlin, T.: Computation and consciousness. J. Philos. 86(8), 407–432 (1989)
McDermott, D.: Artificial intelligence and consciousness. In: Zelazo, P.D., Moscovitch, M.,
Thompson, E. (eds.) The Cambridge Handbook of Consciousness, pp. 117–150. Cambridge
University Press, Cambridge (2007)
McGinn, C.: The Problem of Consciousness: Essays Towards a Resolution. Blackwell, Oxford
(1991)
McLaughlin, B.P.: A naturalist-phenomenal realist response to block’s harder problem. Philos.
Issues 13(1), 163–204 (2003)
Papineau, D.: Thinking about Consciousness. Clarendon Press, Oxford (2002)
Prinz, J.: Level-headed mysterianism and artificial experience. In: Holland, O. (ed.) Machine
Consciousness, pp. 111–132. Imprint Academic, Exeter (2003)
Prinz, J.: A neurofunctional theory of consciousness. In: Brook, A., Akins, K. (eds.) Cognition
and the Brain: The Philosophy and Neuroscience Movement, pp. 381–396. Cambridge
Prinz, J.: The Conscious Brain: How Attention Engenders Experience. Oxford University Press,
Oxford (2012)
de Queiroz, K.: The general lineage concept of species and the defining properties of the species
category. In: Wilson, R.A. (ed.) Species: New Interdisciplinary Essays, pp. 49–89. MIT Press,
Cambridge (1999)
Reggia, J.A.: The rise of machine consciousness: studying consciousness with computational
models. Neural Netw. 44, 112–131 (2013)
Richards, R.A.: The Species Problem: A Philosophical Analysis. Cambridge University Press,
Cambridge (2010)
Scheutz, M.: Artificial Emotions and Machine Consciousness. In: Frankish, K., Ramsey,
W.M. (eds.) The Cambridge Handbook of Artificial Intelligence, pp. 247–266. Cambridge
Searle, J.R.: Minds, brains and programs. Behav. Brain Sci. 3(3), 417–457 (1980)
Shea, N., Bayne, T.: The vegetative state and the science of consciousness. Br. J. Philos. Sci. 61
(3), 459–484 (2010)
18 C. Chin
Tye, M.: Tense Bees and Shell-Shocked Crabs: Are Animals Conscious?. Oxford University
Press, New York (2016)
Valen, L.V.: Ecological species, multispecies, and oaks. Taxon 25(2/3), 233–239 (1976)
Wimsatt, W.C.: Reductionism, levels of organization, and the mind-body problem. In: Globus,
G.G., Maxwell, G., Savodnik, I. (eds.) Consciousness and the Brain, pp. 205–267. Springer,
Dordrecht (1976)
Cognition as Embodied Morphological
Computation
Gordana Dodig-Crnkovic1,2 ✉
( )
1
Chalmers University of Technology, Gothenburg, Sweden
dodig@chalmers.se
2
Gothenburg University, Gothenburg, Sweden
Abstract. Cognitive science is considered to be the study of mind (consciousness

and thought) and intelligence in humans. Under such definition variety of
unsolved/unsolvable problems appear. This article argues for a broad under‐
standing of cognition based on empirical results from i.a. natural sciences, self-
organization, artificial intelligence and artificial life, network science and neuro‐
science, that apart from the high level mental activities in humans, includes sub-
symbolic and sub-conscious processes, such as emotions, recognizes cognition
in other living beings as well as extended and distributed/social cognition. The
new idea of cognition as complex multiscale phenomenon evolved in living
organisms based on bodily structures that process information, linking cogniti‐
vists and EEEE (embodied, embedded, enactive, extended) cognition approaches
with the idea of morphological computation (info-computational self-organisa‐
tion) in cognizing agents, emerging in evolution through interactions of a (living/
cognizing) agent with the environment.
1 Understanding Cognition
Cognitive science is currently defined as a study of processes of knowledge generation

through perception, thinking (reasoning), memory, learning, problem solving, and
similar. Thagard (2013) makes an extension of the idea of “thinking” to include
emotional experience. This move bridges some of the distance between cognition as
thinking and its (sub-)processes, but the fundamental problem of generative mechanisms
that can dynamically overarch the chasm between matter and mind remains. The defi‐
nition of cognitive science does not mention biology, chemistry, (quantum- nano-, etc.)
physics or chaos theory, self-organisation, and artificial life, artificial intelligence or data
science, extended mind, or distributed cognition as studied with help of network science,
sociology or ecology.
On the current view, cognition is about high-level processes remote from physical-
chemical-biological substrate. It is modeled either by classical sequential computation,
understood as symbol manipulation, or by neural networks. On the other hand, histori‐
cally, behaviorism offered an alternative view of cognition with the focus on the observ‐
able behavior of a subject. This divide is mirrored in the present day schism between
cognitivism/computationalism on one side and EEEE (embodied, embedded, enactive,
extended) cognition on the other. There have been numerous attempts to bridge this gap

V. C. Müller (Ed.): PT-AI 2017, SAPERE 44, pp. 19–23, 2018.
https://doi.org/10.1007/978-3-319-96448-5_2
20 G. Dodig-Crnkovic
(Clark 2013), (Scheutz 2002), (Pfeifer and Iida 2005) and others, offering connection
between lower level sub-symbolic signal processing and higher-level processes of (clas‐
sical, mental) cognition.
The most frequent view of cognition is still human-centric and not evolutionary,
generative model. Thagard (2014) lists open philosophical problems of this approach to
cognition. Majority of those problems can only be solved on the basis of empirical data,
experiments and adequate generative models and simulations.
The idea of morphological computing has been proposed by (Paul 2004) (Pfeifer and
Iida 2005), (Hauser et al. 2014) and (Müller and Hoffmann 2017) defining computation
in a more general way than the traditional symbol manipulation, or connectionist models.
It is taking into account physical embodiment of computational mechanisms, thus
presenting suitable tool for modeling of a broader range of cognitive phenomena. In a
related approach, (Dodig-Crnkovic 2014) takes cognition in a cognitive agent to be
morphological computation, defined as information processing performed by
morphology on several levels of organization. Cognition in this framework is capacity
possessed by all living organisms, as (Maturana and Varela 1980) and (Stewart 1996)
argued. Every single cell, while alive, constantly cognizes. It registers inputs from the
world and its own body, ensures continuous existence through morphological processes
run on metabolic production of energy. It is avoiding dangers that could cause disinte‐
gration or damage, adapting its morphology to the environmental constraints. Physico-
chemical-biological processes present morphological computation on different levels of
organization. They depend on the morphology of the organism: its material, form and
structure.
Morphological computation is modeled as a dynamics of a structure of nodes (agents)
that exchange (communicate) information. Single living cell presents such a structure.
Groups of unicellular organisms (such as bacteria) communicate and build swarms or
films through morphological computation that presents social/distributed cognition.
Groups of cells through morphological computation cluster into multicellular assemblies
with specific control mechanisms, forming the tissues, organs, organisms and groups of
organisms. This layered organization of networks within networks provides information
processing speed-up.
A new quality in morphological computing in living organisms emerges with the
development of nervous system. With it, multicellular organisms as cognizing agents
acquire ability of self-representation, which enables distinction between “me” and the
“other” and presents basic functionality that supports locomotion. Animals that possess
nervous systems with centralized control connected to sensors and actuators, are capable
of locomotion which increases probability of survival. Brains in animals consist of large
number of mutually communicating cells. A single neuron is a relatively simple infor‐
mation processor, while the whole brain possesses advanced information processing/
computational capacities. We see the similar mechanism as in bacteria swarms with
distributed cognition implemented as morphological computation.
Besides the ability to model cognition as embodied, embedded, enactive, and
extended through interactions with the environment, morphological computing provides
means of understanding how this capacity evolved and how it develops during the life
of an organism.
Cognition as Embodied Morphological Computation 21
2 Problems Solutions with a Broader View of Cognition
Revisiting the list of unsolved/unsolvable problems of cognitive science under the

current idea of cognition (Thagard 2014) we can see their natural solution under a more
general concept of cognition as morphological computation:
The Emotion Challenge: Morphological computing of embodied cognition has layered

computational architecture. Sub-symbolic electro-chemical processes present the basic
layer in the information processing related to emotion (von Haugwitz and Dodig-
Crnkovic 2015).
The Consciousness Challenge: Consciousness is proposed as information integration

that has central role in the control of behavior (Tononi 2004) (Freeman 2009).
The World Challenge: Distributed morphological computation processes representing

hierarchies of computation solves this problem (Abramsky and Coecke 2007) (Sloman
2011) (Piccinini and Shagrir 2014) (Dodig-Crnkovic 2016, 2017).
The Body Challenge: Explicit modeling of a body is a consequence of the inclusion of

morphological computational processes in the substrate as an integral part of cognition
(Matsushita et al. 2005) (Pfeifer and Bongard 2006) (MacLennan 2010).
The Dynamical Systems Challenge: Dynamical systems are a very important class of
computational systems, as argued in (van Leeuwen and Wiedermann 2017) (Burgin and
Dodig-Crnkovic 2015).
The Social Challenge: Adopting cognition that is not only individual but also distrib‐
uted/social, solves this problem (Epstein 2007) (Barabasi 2010).
The Mathematics Challenge (brain cannot be conventional computer): Morphological

computing in living beings (unconventional computing) starts at quantum level and
propagates to higher levels of organisation as different kinds of physical, chemical,
biological, cognitive and social computing. (Cooper 2012) (Zenil 2012).
This short account presents an outline of an argument for the adoption of a broader
view of cognition then the one that presents the current received view. For the future
work, it remains to study the exact mechanisms of morphological computation at variety
of levels of organisation of living organisms in terms of computation as information
self-structuring (Dodig-Crnkovic 2016 and 2017). At the same time, cognitive compu‐
tational models are being tested in artifactual cognitive systems with artificial intelli‐
gence and cognitive computing.
References
Abramsky, S., Coecke, B.: Physics from computer science. Int. J. Unconv. Comput. 3(3), 179–
197 (2007)
Barabasi, A.-L.: Bursts: The Hidden Pattern Behind Everything We Do. Dutton, London (2010)
Another random document with
no related content on Scribd:
Riffe, there may be some justice in the claim of Strasburg to be the
birthplace of the printing-press. The completed press, however, was
not produced until Gutenberg had returned to the old home city of
the family—Mayence.
After the close of the suit brought by Dritzehn against Gutenberg,
that is to say, after 1440, there are no further references to
Gutenberg’s undertakings in Strasburg. It is not even known whether
or not he continued business operations there, but it appears that he
was dwelling there as late as 1444. In 1448, he is recorded as again
a citizen of Mayence, and it was in Mayence that, in 1450, the
completed invention became known to the world.
Gutenberg’s name stands on no title-page and is connected with
no colophon. The fact, however, that the full responsibility for the
invention belongs to him is borne witness to by his contemporaries,
Peter Schöffer, Ulrich Zell, the Abbot Trithemius, Jacob Wimpheling,
and others. In a chronicle of the archbishop of Mayence, continued
to the year 1555 and compiled by Count Wilhelm von Zimmern, it is
recorded that the noble art of book-printing was discovered in
Mayence by a worthy citizen named Gutenberg, who devoted to the
invention all his time and resources until he had brought it to a
successful completion.
In 1470, a letter was written by the scholar, Wilhelm Fichet, of
Paris, to the historian, Robert Gaguin, which letter was later printed
on the last sheet of a volume published in Paris and in Basel,
entitled: Gasparini Pergamensis Orthographiæ Liber. This letter
contains an enthusiastic description of the new art of book-printing
discovered in Germany by Gutenberg. The writer says: “There has
been discovered in Germany a wonderful new method for the
production of books, and those who have mastered this method are
taking their invention from Mayence out into the world somewhat as
the old Grecian warriors took their weapons from the belly of the
Trojan horse. The light of this wonderful discovery will spread from
Germany to all parts of the earth. I have been told by three
foreigners—Kranz, Freiburger, and Gering—that Gutenberg has
succeeded in producing books by means of metal letters in place of
using the handiwork of the scribes.”
Fichet goes on to speak of Gutenberg as “bringing more blessings
upon the world than were given by the goddess Ceres, for Ceres
could bestow only material food, while through Gutenberg the
productions of the thinkers could be brought within the reach of all
people.” This letter was written only two years after the death of
Gutenberg, and as it came from Basel, one of the first cities to which
the new art had been carried from Mayence, it constitutes very good
contemporary evidence as to the immediate credit that was given to
Gutenberg for the invention.[428]
The historical date now given for the completion of the invention is
August 22, 1450. On this date Gutenberg entered into a contract with
Johann Fust, a wealthy citizen and goldsmith of Mayence, under
which contract Fust loaned to Gutenberg, with interest at 6 per cent.
(a low rate for that period), the sum of 800 gulden in gold. This sum
Gutenberg agreed to utilise in developing his invention, while the
material of the workshop to be instituted was pledged to Fust as
security for the repayment of the loan. The sum proved insufficient
for establishing the necessary plant, and two years later Fust added
a further sum of 800 gulden.
Gutenberg pledged himself, as afterwards stated in the lawsuit
which arose between Fust and himself, to use this money for the
printing of books,—“das werk der bücher.” At the time Gutenberg
secured this loan, it seemed evident that, in experimenting with and
in developing his invention, he had exhausted his own entire
resources.
Gutenberg could, of course, lay no claim to being in any literal
sense of the term the first printer. Printing in one form or another had
been carried on in Germany and elsewhere for a number of years,
and printing from movable blocks had, in fact, been done in China
400 years or more before the beginning of Gutenberg’s work. As
early as the twelfth century, says Kapp, there are numerous
references to cloth printers, stampers of letters, and printers of
maps. The oldest wood-cut known to have been produced in Europe,
is a representation of S. Christopher, and bears date 1423. At about
this time, and probably, in fact, some years earlier, was begun in
Holland, as previously stated, the work of printing from wooden
blocks, the designs being principally devoted to holy subjects. In
connection with such designs, there had been printing also from
letterings cut out of solid wooden blocks, and these letterings had
even in some cases been cut upon blocks sufficient to occupy an
entire page.
The practical contribution made by Gutenberg, which developed
from the easy processes of stamping designs and brief lines of
lettering, a method by means of which whole books could be
produced, was first, in the use of movable metal type, produced by
casting, and second, in an improvement made in the mechanism of
the hand presses by which larger sheets could be worked.
The first work produced with this movable metal type was a Latin
version of the Bible. The description of this volume is first given in a
chronicle of Cologne, dating from the year 1499, the statements in
which rest upon the authority of Ulrich Zell, who was the first printer
in Cologne.
Concerning the further operations of Gutenberg, we are mainly
dependent upon the references in the records of the suit brought by
Fust, in 1445, for the repayment of his loan, and upon a document of
1468 in which a certain Dr. Humery entered into an undertaking with
the Archbishop of Mayence that the printing-office plant left by the
deceased Johann Gutenberg shall not be permitted to be taken out
of the city of Mayence. This later reference had to do with a second
printing-press established by Gutenberg with the aid of the said
Humery.
In the suit brought by Fust, Gutenberg contended that the second
payment of 800 gulden agreed upon had never been given to him in
full. He stated further that Fust had agreed to advance 300 gulden
per year for use in the purchase of materials, paper, parchment,
type-metal, and ink. The matter of the later accountings between
Fust and Gutenberg is evidently a complicated one and need not be
considered here in detail. Gutenberg’s inability to repay the first and
more important loan for the payment of which his first printing-press
had been mortgaged, caused the ownership of this office to come
into the control of Fust.
Fortunately, by the time his first venture had thus been closed, as
far at least as he was concerned, he had been able to give sufficient
evidence of the importance and of the commercial value of the
undertaking to be in a position to interest others in his schemes.
His second printing-press was in like manner pledged to the
associate who provided the capital,—Dr. Humery,—and the business
of this office appears to have been continued without break until the
time of Gutenberg’s death in 1468. With these new resources at
hand, Gutenberg was able to cast some new fonts of type, and to
make various improvements in his working methods.
The first issues of the new press, the organisation of which
appears to have been completed about 1457, were volumes
containing the writings of Mätthaus de Cracovia and Thomas
Aquinas. The third book was the famous first edition of the
Catholicon, a grammatical compilation of the Dominican monk
Balbus from Genoa. The Catholicon was a folio containing no less
than 373 rather closely printed sheets. In the meantime, Fust had
associated with him Schöffer or Schoiffher, who had been an
assistant of Gutenberg, and the two were continuing work in the
original printing-office.
The sacking of Mayence, in 1462, by Adolph of Nassau, put an
end, for the time, to all business in the city, including the work of the
new printing-presses. Gutenberg betook himself to the neighbouring
town of Eltville, which, as early as 1420, had given shelter to his
parents, and there he carried on his printing for a time under the
protection of Archbishop Adolph.
Kapp points out that the printing art had its development, not in a
university centre, but in a commercial town, and was from the outset
carried on, not by scholars, but by workers of the people, and that
this fact doubtless had an important influence in bringing the whole
business of the production of books and the distribution of literature
into closer relations with the mass of the German people than was
the case in France.
In France, as will be noted later, the first printers were directly
associated with the university, succeeding immediately to the official
university scribes, and the production of books through the presses
continued to be under direct control of the university, as had been
the case from the beginning with the production of books in
manuscript. The fact that the control of the first French presses
rested with the university Faculty, undoubtedly exercised an
important influence on the choice of the books to be printed, and the
first issues of the French presses were, therefore, in the main
restricted to editions of the classics or to works of jurisprudence and
medicine belonging to the official lists of the university texts. The
earlier issues of the German press, on the other hand, were books
belonging in no way to the university curriculum, but were addressed
directly to the interests of the people at large.
While the modifications introduced by Gutenberg into the methods
of printing, under which the old engraved blocks were replaced by
movable leaden type, seem slight in themselves, they constituted
nevertheless a new art. The actual changes were but inconsiderable,
but the practical result was a revolution in the possibilities of the
press.
Gutenberg’s work as a printer was, from a commercial point of
view, never successful. During the eighteen years which elapsed
between the time of his invention and the date of his death, he
seems to have been always under the pressure of debt and money
difficulties. He had in fact no time to make money. He had given up,
in his devotion to his invention, previous business undertakings
which were remunerative, and he had absorbed in the development
of the printing-press all the resources that he could control. His
interest, however, was evidently that of perfecting an art rather than
of creating a business; and in spite of his various difficulties and his
several lawsuits with his associates, it is in evidence as part of the
testimony in these very suits, that he was recognised by all as a man
of knowledge and character, and as a born leader, whose integrity of
purpose and whose nobility of aim were acknowledged by all with
whom he had to do. With all his misfortunes, he seems never for a
moment to have lost confidence in the value to the world of his idea,
and to this idea, with no thought of personal gain or advantage, he
was willing to devote his means and his life.
The difference between the production each year of a few hundred
copies of religious or classical works by the laborious toil of the
monks or the university scribes, works which could at best benefit
only the limited circle of readers who were within reach either of the
monasteries or of the universities, and a world-wide distribution, as
well of the great books of the earlier times which belonged to the
world’s literature as of the current thoughts of the contemporary
generation, was a difference, not of degree, but of kind. It was a
revolution in the history of human thought and in the influence of
thought upon humanity.
If the invention of printing had not taken shape in the brain of
Gutenberg, it would doubtless have come to the world through some
other worker, and, in fact, with no very great delay, for other men
were already busying themselves with the same great need and
were on the track of the same means of supplying the need. As the
history stands, however, the credit for the revolution must be given to
the mirror-maker of Mayence. Other sailors would certainly have
found their way to the Western Continent if the opportunity or the
attempt of Columbus had failed, but it is to Columbus that history
gives the laurel crown.
Gutenberg, and the printers who followed him, naturally selected
as the first models for their newly founded type the script letters with
which they were familiar in the best manuscripts. The first font of
type manufactured by Gutenberg, which was used in his earliest
publication, The Folio Bible, was known as the “missal type,” having
been copied from the script adopted by the monks for the books of
worship. This style of type was followed for a long time for Bibles and
for religious works generally. One of the earlier objections against
printed books was that they were so much less beautiful in their
appearance than the work of the best scribes, and it was the finest
script that remained as the ideal to be attained by the type-founders
and the clear black impression of the best oak-gall writing ink that
was to be imitated by the impressions from the presses.
The scholarly lovers of fine books in Germany regarded the new
art at the outset with no little disapproval and criticism. The collectors
who had brought together, with much labour and expenditure, stores
of valuable manuscripts dreaded lest, through the multiplication of
comparatively inexpensive copies of their texts, the value of their
collections should be taken away. When the messengers of Cardinal
Bessarion were shown by the Greek Laskaris (later the author of the
first Greek grammar that came into print), a specimen of one of the
earlier printed books, they spoke sneeringly of this so-called
discovery which had been made by a barbarian from a German
city.[429] The great manuscript-dealer, Vespasiano, writing in 1482
concerning the magnificent ducal library in Urbino, the volumes in
which had been largely either collected or purchased by himself or
under his own direction, says: “In this library all the volumes are of
perfect beauty, all written, by skilled scribes, on parchment and many
of them adorned with exquisite miniatures. The collection contains
no single printed book. The Duke (Frederick) would be ashamed to
have a printed book in his library.”[430] By collectors like Frederick
and manuscript-dealers like Vespasiano, the new art was considered
to be merely a mechanical method of producing inartistic volumes,
with which none but uncultivated people could be satisfied.
For a number of years, therefore, after the work of the first
presses, there were still produced beautiful specimens of
manuscripts, more particularly of Italian and French books of
worship, and for this class of manuscripts the work of the hand
illuminators and miniature painters continued to be utilised. In
Germany there are various examples of books which had been
printed, being again produced in written copy, as for instance, the
Chronicon Urspergense, of Hroswitha.[431] It was also the case that
for the production of large choir-books the work of the scribes
continued to be useful.
Trithemius, Abbot of Sponheim, wrote to Gerlach, Abbot of Deutz,
a letter which was printed in 1494 in Mayence, under the title, De
Laude Scriptorum Manualium. In this he says:
“A work written on parchment could be preserved for a thousand
years, while it is probable that no volume printed on paper will last
for more than two centuries. Many important works have not been
printed, and the copies required of these must be prepared by
scribes. The scribe who ceases his work because of the invention of
the printing-press can be no true lover of books, in that, regarding
only the present, he gives no due thought to the intellectual
cultivation of his successors. The printer has no care for the beauty
and the artistic form of books, while with the scribe this is a labour of
love.”[432]
Notwithstanding such criticism on the part of a few scholarly
churchmen, the influence of Rome and of the Church generally,
during the earlier work of the printers, was very largely favourable
and had not a little to do with the support given to the work which
might easily otherwise have been given up for lack of adequate
business return. The Church of Rome felt itself at this time
sufficiently secure in its control of the minds of men to be prepared to
utilise to full advantage all methods for distributing its doctrinal
literature, and to have no dread as to these same means being used
for the scattering of heretical teachings. The popes of the time,
largely influenced by the spirit of the Renaissance, gave a cordial
welcome to the revival of scholarly interests and to the printing-press
as an important means for furthering the general education and the
intellectual development of the community. Their interest was by no
means limited to the distribution of doctrinal works, but in these
earlier years of publishing they welcomed, and to a considerable
extent co-operated in, the production of editions, for general
circulation, of the works of the pagan classics.
Hegel says, in his Philosophy of History, that the renewed interest
in the studying of the writings of the ancients found an important
support in the service of the printing-press. He goes on to point out
that the Church felt no anxiety concerning this renewed interest in
pagan literature, and evidently did not imagine that this literature was
introducing into the minds of men a new element of suggestion and
of inquiry.
It may be considered as one of the fortunate circumstances
attending the introduction of the art of printing that the popes of the
time were largely men of liberal education and of intellectual tastes,
while one or two, such as Nicholas V., Julius II., and Leo X., had a
very keen personal interest in literature and were collectors of books.
The fact that Leo X. was a luxury-loving, free-thinking prince rather
than a devoted Christian leader or teacher, may very probably have
been in the end a service for the enlightenment and development of
his own generation and of the generations that were to come. An
earnest and narrow-minded head of the Church could, during the
first years of the sixteenth century, have retarded not a little the
development of the work of producing books for the community at
large.
It was a number of years before the dread of the use of the
printing-press for the spread of heretical doctrines and of a
consequent undermining of the authority of the Church assumed
such proportions in the minds of the popes in Rome and with the
bishops elsewhere, as to cause the influence of the Church to be
placed against the interests of the world of literature. As a result of
this early acceptance by the Church of the printing-press as a useful
ally and servant, the first Italian presses were supported by bishops
and cardinals in the work of producing classics for scholarly readers,
while at the other extremity of the Church organisation, and at a
distance of a thousand miles or more from Rome, the Brothers of
Common Life were using the presses in their Brotherhood homes for
the distribution of cheap books among the people.
Berthold von Henneberg, Elector of Mayence, speaks of “The
Divine Art of Printing.”[433] The Carthusian monk, Werner Rolewinck,
writes, in his Outline History of the World (Fasciculus Temporum):
“The art of printing which has been discovered in Mayence is the art
of arts, the science of sciences, by means of which it will be possible
to place in the hands of all men treasures of literature and of
knowledge which have heretofore been out of their reach.”
Joh. Rauchler, the first Rector of the Tübingen High School (later
the University of Tübingen), rejoices that through the new art so
many authors can now be brought within the reach of students in
Latin, Greek, and Hebrew, authors who are witnesses for the
Christian faith, and the service of whose writings to the Church and
to the world is so great, that he can but consider “this art as a gift
directly from God himself.”[434] Felix Fabri, Prior of the Dominican
monastery in Ulm, says, in his Historia Suevorum, issued in the year
1459, that “no art that the world has known can be considered so
worthy, so useful, so much to be esteemed, indeed, so divine as that
which has now, through the Grace of God, been discovered in
Mayence.”
The first printing work done by the Brothers of Common Life dates
from 1468. They appear to have promptly utilised their scribes as
compositors and their illuminators as designers for the new form in
which their books were produced. Many of the Benedictine
monasteries which had for so many centuries led the way in the
preservation and the multiplication of literature at once associated
presses with their monasteries and had their monks trained in the art
of setting type and of printing sheets.
Among the monastery printing-presses were those of the
Carthusian monastery in Strasburg, the monastery of S. Ulrich and
Afra, in Augsburg, and the Benedictine monasteries in Nuremberg
and Rostock. As a rule, in places where the work of scribes had
been active, the printing-press found a ready acceptance. It was not
long, however, before so great a development in the methods of the
printing business was brought about that it became difficult for the
monasteries to carry on the work effectively, and by the middle of the
sixteenth century the production of books in monasteries had
practically ceased.
The favourable relations between the Church and the printers
were checked by the Humanistic movement, which, a generation or
more before the Reformation, began to bring into question the
authority of the Church and the infallibility of papacy. The influence of
the Humanistic teachers was so largely furthered by the co-operation
of the printers that the jealousy and dread of the ecclesiastical
authorities were promptly aroused, and they began to utter
fulminations against the wicked and ignorant men who were using
the art of printing for misleading the community and for the
circulation of error. The ecclesiastics, who had at first favoured the
widest possible circulation of the Scriptures, now contended that
much of the heretical teaching was due to the misunderstanding of
the Scriptures on the part of readers who were acting without the
guidance of their spiritual advisers.
The authorities of the Church now began to take the ground that
the reading of the Scriptures by individuals was not to be permitted,
and that the Bible was to be given to the community only through the
interpretation of the Church. At the same time, the authority of the
Church was exerted to repress, or at least to restrict, the operations
of the printing-press, and to bring printers and publishers under a
close ecclesiastical supervision and censorship. It was now,
however, too late to stand between the printing-press and the
people. Large portions of the community had become accustomed to
a wide circulation of books and to the selection without restriction of
such reading-matter as might be placed within their reach, and this
privilege they were no longer willing to forego.
It was nevertheless true that in certain countries, particularly in
Italy and in France, the censorship of the Church was strong enough
seriously to hamper and interfere with publishing undertakings and to
check the natural development of literary production. Even in Italy,
however, the critical spirit was found to be too strong to be entirely
crushed out, and from Venice, the most important of the Italian
publishing centres, it proved possible to secure for the productions of
the printing-press a circulation that was practically independent of
the censorship of Rome.
The Humanistic movement was, on other grounds, of immediate
service for the printers and publishers, in that it brought about an
active demand for the works of classical writers, a demand which it
required the fullest resources of the earlier printers to supply.
If the invention of Gutenberg had taken shape during the period
when there happened to be no such active intellectual literary
interests, the first printers might easily have found it difficult to
secure business for their presses and the development of the
business of book production would have been seriously hampered.
The long series of controversies which were brought into being by
the Reformation, and the large mass of controversial literature which
was the result of the Reformation, constituted, a generation later,
another favourable influence in securing an assured foundation for
the business of the printers. If it be the case that the work of the
leaders of the Reformation could hardly have been carried on
without the aid of the printing-press, it is also true that at a time when
the business of the early printers was in a very critical and
unremunerative condition, the impetus given to the production of
literature, and the increased eagerness on the part of the common
people for literature, formed an essential factor in making an assured
foundation for the business of the printers and the publishers.
In 1462, on the 28th of October, Archbishop Adolph of Nassau
captured the city of Mayence and gave it over to his soldiers for
plunder. The typesetters and printers, with all other artisans whose
work depended upon the commerce of the city, were driven to flight,
and it appeared for the moment as if the newly instituted printing
business had been crushed out. The result of the scattering of the
printers, however, was the introduction of the new art into a number
of other centres where the influences were favourable for its
development.
The typesetters of Mayence, driven from their printing-offices by
the heavy hand of the Church, journeyed throughout the world,
carrying their new knowledge and training and they were able to give
to many communities the means of education and enlightenment
through which the great revolt against the Church was finally
instituted. The work of the printers, checked for the time in Mayence,
took shape promptly in Strasburg, and from there was taken down
the Rhine to Cologne, and in a few years was also in active
operation in Basel, Augsburg, Ulm, and Nuremberg. In 1464, as
elsewhere described, German printers carried their invention into
Italy and erected the first Italian printing-press in Subiaco. And in
1470, also through Germans, the work of the printers began in Paris.
The shrewd and enterprising merchant Fust, by means of whose
capital Gutenberg had been able to begin his business operations,
would hardly have pressed his suit against his associate, if he had
not had confidence in the value of the invention. As soon as, through
the decision of 1455, he came into possession of the presses, he at
once put these again into operation. He found a practical
superintendent or co-worker in Peter Schöffer. Schöffer was a
German by birth, but had carried on work in Paris as a scribe or
writer of higher class manuscripts, as illuminator, and as a
manuscript-dealer. Returning to Mayence in 1454, he had entered
the employ of Gutenberg as type-setter and proof-reader. Later,
having married the daughter of Fust, he was taken into partnership
by his father-in-law, and was able to make a satisfactory organisation
and a wide development for the business of the printing-office. The
first publication issued by Fust & Schöffer was a psalter printed (in
Latin) on parchment, with the great missal type.
The second work, undertaken, not at the risk of the printers, but at
the cost of two of the Mayence monasteries, was an edition of a
great choir-book. This psalter, or rather psalterium, is the first printed
work in which the name of the printer is given and the date of the
publication. It apparently proved possible to secure for this book
even with the very inadequate distributing machinery that was
available, a remunerative sale, as it was printed again in 1490, in
1502, in 1515, and in 1516.
Among the earlier publications of Fust & Schöffer are the
Rationale Divinorum Officiorum of the Dominican monk Durandus,
which was issued in 1459, the Codex Constitutionum of Pope
Clement, issued in 1460, and the Bull of Emperor Frederic III.
against Diether von Isenburg, printed in 1461. The most beautiful
and most important production of their press was, however, the great
Latin Bible issued in 1462, in two folio volumes, and which is known
as the “48 line Bible.”
The work of the printing-office was, as previously stated, stopped
by the sacking of the city, and the two partners appear to have
migrated for the time to Frankfort. In 1464, they were again in
Mayence, and in that year they published the sixth book of the
Decretals of Pope Boniface VIII., and the De Officiis of Cicero. The
latter was the first of the German editions of the classics, and
remained a favourite book with the German printers, being
repeatedly reprinted.
In 1453, Fust made a journey to Paris in order to find sale there for
his big Bible. This was four years before the first Paris printing-press
began its work, and it was in connection with this big Bible that the
gossip arose of Fust being able, through compact with the Devil, to
produce an indefinite number of copies of a book. It could not be
understood how in any other way these copies could be offered so
cheaply. The University of Paris was at that date the most important
in Europe, and the influence of the University upon the cultivation of
the city and its close relations with the old book-trade in manuscripts,
had made Paris the most important European centre for literary
production and the place where scholars were in the habit of looking
for their material. It was in Paris, if anywhere, that it should prove
possible to find sale for the Latin Bible, and Fust’s efforts appear to
have met with a prompt success. The first Bible bearing a date was
completed in 1462, and is known as the Mayence Bible. At the time it
was in readiness (in October) nothing could be done in getting it into
the market, as Mayence was being besieged by Adolph of Nassau.
In 1466, Fust is again in Paris with copies of the second edition of
his De Officiis, and with other of his publications.
There is still preserved in the city library of Geneva a copy of this
edition of Cicero, which contains the record that it was bought by
Louis de la Vernada, in Paris, in July, 1466, from Fust.[435]
Fust & Schöffer may claim to have been the first printers who
acted also as publishers and booksellers. Notwithstanding the many
difficulties with which they had had to contend, they were able to
offer their books at prices which, to the old dealers in manuscripts,
seemed astounding and which gave some pretext for the charge of
magic. Madden says that a copy of the “48 line Bible” printed on
parchment, could be bought in Paris, in 1470, for 2000 francs, and
that the cost of the same text a few years earlier in manuscript form
would have been five times as great. Bishop John of Aleria, writing in
1467 to Pope Paul II., says that it is now possible to purchase in
Rome for 20 gulden, gold, works which a few years earlier would
have cost not less than 100 gulden, and that other books now selling
as low as 4 gulden would previously have cost not less than 20
gulden. The first results of the printing-press appear, therefore, to
have been a reduction of about four fifths in the price of work of a
scholarly character.
Fust is entitled to the description, not only of the second printer
and of the first publisher, but of the first pirate in printed books. In
1465, Mentel printed in Strasburg under the title of De Arte
Prædicatoria, the fourth book of S. Augustine’s De Doctrina
Christiana. The editor states that he had, for the purpose of this
edition, collected manuscript texts in the libraries of Heidelberg,
Speyer, Worms, and Strasburg, and that he had induced Joh.
Mentel, a “master of the art of printing,” to put the volume into a form
available for the general use of clerics.
Fust reprinted this volume in 1466, following the text with
precision, and simply replacing Mentel’s name with his own. This is
the first instance of literary appropriation of which there is any
record, after the beginning of printing.[436]
After the death of Fust, which occurred early in 1467, Schöffer
continued the business with Fust’s sons, and established branches
in Paris and in Angers. His name appears for the first time alone on
the title-page of the Thomas Aquinas, published in a folio of 516
pages in March, 1467. He prints it in full as Petrus Schoiffher de
Gernsheim. In a receipt for 15 gold crowns, paid by the College of
Autun for a copy of this, Schoiffher styles himself Impressor
Librorum. He appears to have made sale in Paris not only of his own
publications, but of the books issued by other German printers.
In a copy of the work of Johannus Scotus, printed by Koberger in
1474, now contained in the library of the Paris Arsenal, appears the
entry, “I, Peter Schöffer, printer from Mayence, acknowledge that I
have received from the worthy Magistrate, Johannus Henrici, of Pisa,
three scuta as the price of this book.”[437] Schöffer seems to have
acted in some measure also as purchasing agent for the University
of Paris, through an associate, Guimier, who was a licensed member
of the Paris guild. The Paris branch of the business was given up a
few years later, and Schöffer devoted his energies to extending his
trade in Germany. In 1479, his name appears in the list of the
citizens of Frankfort, and the removal to Frankfort of his publishing
headquarters constituted the first step towards the selection of that
city as the centre of the publishing and bookselling trade of
Germany, a position that it retained for more than a century. Schöffer
continued, however, to do the work of his printing in Mayence.
Some light is thrown upon the extent of the publishing
undertakings carried on at the time by Schöffer with his associate
Hancquis, by the record of a suit brought by the two partners in 1480
against a certain Bernhard Inkus, of Frankfort. They charged Inkus
with having begun the publication of a considerable series of books,
the property right in which (Eigentumsrecht) vested in themselves
and in Conrad Henki (who was a son of Fust). It does not appear
from the record of this trial on what grounds Schöffer and his
associates claimed the right to control these books, or whether the
unauthorised issues of which they complained had been printed by
the defendant Inkus or were simply being offered for sale by him on
behalf of other printers.
The case appears to have been referred or possibly appealed to a
court in Basel, and by this court was issued some preliminary
injunction against the continued sale of the books complained of.
The record giving the final decision of the case is, however, missing.
The lack of full details of the suit is the more to be regretted as it
appears to have been the first case after the invention of printing
involving, if not copyright ownership, at least a certain control by
contract.
In the same year we find the Magistracy of the City of Frankfort
applying to the Magistracy of the City of Lübeck for the protection of
Schöffer against some illegitimate infringement of Schöffer’s
business rights on the part of the Lübeck citizen Hans Bitz. Here also
is there no record as to the result of the application. The firm also
had dealings with Ulm, as appears from a claim made, in 1481, for
the collection of the moneys due from certain citizens in Ulm—
Harscher, Ruwinger, and Ofener, for books delivered. They sent to
Ulm, with a protection certificate given by Elector Diether of
Mayence, a representative who was empowered to collect the
money. There was at the outset some delay in connection with an
alleged informality in his authorisation, but the Magistracy of Ulm
sent back word that as soon as the requisite authorisation was
secured, the collection of the money would be enforced in due
course.
These cases are evidence of a certain organisation of machinery
for the distribution of books and for the management of a publishing
business, within a comparatively brief period after the beginning of
the work of the printing-presses, and they indicate also that the
second firm which entered into the business of printing had
succeeded in establishing such business on fairly assured
foundations, and in carrying on successfully large undertakings. It is
to be noted further that Fust & Schöffer, and other of the earlier
German printers, did their work without the assistance of any
patronage, and without even the advantage of a university
connection. The early printers of Italy would have found it
impracticable to carry on their operations without the assistance of
certain wealthy and enterprising noblemen who were prepared to
interest themselves in the new art either from curiosity or from
philanthropy, and as late even as 1495, that is to say nearly half a
century after the beginning of printing, the organisation of the
business of Aldus was dependent upon the favour and services of
certain of his noble friends. In Paris the first printers were helped,
and in part supported, by the money of patrons or of the Crown, and
by the co-operation and influence of the University. In England also
the influence of Oxford was of material importance in securing for the
first printers some assured foundation and support, while the work of
Caxton and his immediate successors in London was also largely
furthered by, if not actually dependent upon, the work or help of
noble and wealthy friends.
In Germany, however, the printing work began, as we have seen,
through the enterprise and ingenuity of a citizen manufacturer who
was supported by the middle class of the community, and who made
his first connections directly with the townspeople. The help of the
universities appears to have been of comparatively smaller
importance. It probably counted for more in Cologne than in any
other of the university cities in which the earlier printers did their
work.
In the course of the thirty-six years of his independent business
activity, that is from the death of Fust in 1466 to the time of his own
death, Schöffer printed in all fifty-nine works which bear date and
which have been identified as his. His firm took rank for this period
as by far the most important printing and publishing concern in
existence.
With hardly an exception, the books issued from his press were
folios. They were printed with fifty to sixty lines on the page, and
contained an average of about 300 pages or 150 sheets.
Among the works included in the list are the Constitutiones of
Clement V., the Institutions of Justinian, the Expositio Sententiarum
of Thomas Aquinas, the Epistolæ of S. Jerome, the sixth book of the
Decretals of Boniface VIII., the Decretals of Gregory IX., the Codex
of Justinian, and the Expositio Psalterii of Joh. Torquemada. The last
named volume had already been printed in Subiaco and again in
Rome under the direct supervision of the author, who was supplying
the funds for carrying on the first printing-office established in Italy.
After Schöffer’s death in 1502, his son printed an edition of the
Mercurius Trimegistus.
I do not find record of the arrangement entered into by Schöffer for
the editing of the texts of the works printed by him. The collection of
manuscripts for use as “copy” for printers, and the collection of
different manuscripts in order to secure the most complete and
accurate texts, must have called for a considerable measure of
scholarly and of general literary knowledge.
It does not appear that Schöffer had enjoyed opportunities for
making himself a scholarly authority, or that he ever made claim to
any special scholarly attainments. There is no record of editorial
work done by himself in the books issued from his press, as was the
case to so exceptional a degree a few years later with the books
printed by Aldus; nor has Schöffer preserved in connection with his
editions the names of the editors who supervised their publication,
as came to be the practice later with the issues of the Aldine press,
of Froben in Basel, and of Koberger in Nuremberg. As far as I can
ascertain, however, the Schöffer texts compared favourably for
accuracy and for authority with other of the earlier printed books, and
it is to be assumed, therefore, that he had been able to organise an
adequate critical staff or to secure from time to time, as required, the
services of competent scholars.
The business founded by Gutenberg, taken possession of under
mortgage by Fust, and carried on first by Fust & Schöffer, and later
by Schöffer and other associates, lasted nearly one hundred years.
The first publication was, as noted, the big Psalterium, printed in
1457, and the last, an edition of the German version of the books of
Livy, printed by Ivo Schöffer in 1557.
It seems evident that while the credit for the great invention fairly
belongs to Gutenberg, and the original planning and initiative of the
business were his, a large measure of business capacity must have
belonged to his partner Fust, who had also, to be sure, the
advantage of being a capitalist to begin with, a factor as important in
the earliest time of publishing as in the present day.
One of the most definite pieces of testimony in regard to the
connection of Gutenberg with the invention of printing, testimony
which possesses special value as coming from a person possessing
first-hand knowledge of the facts, is contained in an Epilogue written
in verse by John Schöffer (son of Peter), and printed at the end of
the Livy published by him in 1505. It is addressed to the Emperor
Maximilian, and reads as follows:
“May your Majesty deign to accept this book which was printed at
Mayence, the town in which the admirable art of typography was
invented, in the year 1450, by John Gutenberg, and afterwards
brought to perfection at the expense, and by the labour, of John Fust
and Peter Schöffer.”
It would not belong to the plan of this historical sketch to give in
detail a record of the successive concerns which carried on
throughout Germany, with increasing rapidity and with undertakings
of ever widening importance, the business of printing and publishing.
I propose merely to present the records of a few of the earlier
concerns, and to make such reference to typical firms of later
generations as may give an impression of the gradual development
of the book-trade, and as may serve also as examples from which to
judge of the development of the idea of the literary property in
Germany, and the varying positions taken under the enactments and
other governmental measures in regard to such property.
The books printed during the first half-century were, as we shall
note, almost exclusively reissues of ecclesiastical or pagan classics,
and apart from such original work as may have been put into
introductions or notes, did not call for the labour of contemporary
authors. Among the earlier of original German publications is to be
classed a German grammar entitled Die Leyenschul, printed by
Peter Jordan in Mayence in 1531. This grammar, which remained for
a considerable time an authority on its subject, does not bear the
name of the author or editor.
Another of the earlier original works for the sale of which the
author may have secured some compensation was the Astronomie
of Joh. Stöffler, which was printed in Ottenheim in 1513.
One of the more important of the earlier publishing concerns of
Mayence was that of Franz Behem, who printed in the ten years
succeeding 1539 an important series of theological works. With the
close of Behem’s business in 1552, Mayence appears to have lost
its relative importance in connection with the work of printing and
publishing.
In Strasburg, which had contested with Mayence the prestige of
being the actual birthplace of the printing-press, important publishing
undertakings were carried on from a very early date, and for a
number of years the city of Basel alone could compete with
Strasburg in the number and importance of the books issued from its
presses. The two publishing concerns whose individual enterprises
and whose rivalry with each other did so much to bring Strasburg
into importance as a factor in the German book-trade, were those of
Johann Mentel and of Heinrich Eggestein. Mentel’s first publications
were a Latin Bible in two folio volumes, which was the first Bible
printed in the smaller Gothic type; an edition of De Doctrina
Christiana of S. Augustine; an edition of the Summa of Thomas
Aquinas; an edition of the Bible in German, which appeared in 1466;

Textbook Philosophy and Theory of Artificial Intelligence 2017 Vincent C Muller Ebook All Chapter PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Textbook Philosophy and Theory of Artificial Intelligence 2017 Vincent C Muller Ebook All Chapter PDF

Uploaded by

Copyright:

Available Formats

Philosophy and Theory of Artificial

Intelligence 2017 Vincent C. Müller

Beginning Game AI with Unity: Programming Artificial

AI IA 2017 Advances in Artificial Intelligence XVIth

Artificial Intelligence Systems Based on Hybrid Neural

Artificial Intelligence and Robotics Huimin Lu

Artificial Intelligence By Example Develop machine

Practical Artificial Intelligence with Swift From

Philosophy 2nd Edition QuickStudy Academic M C O

Practical Artificial Intelligence: Machine Learning,

Vincent C. Müller Editor

More information about this series at http://www.springer.com/series/10087

Philosophy and Theory

ISSN 2192-6255 ISSN 2192-6263 (electronic)

Library of Congress Control Number: 2018948694

© Springer Nature Switzerland AG 2018

May 2018 Vincent C. Müller

Cognition - Reasoning - Consciousness

Huge, but Unnoticed, Gaps Between Current AI

Computation - Intelligence - Machine Learning

Will Machine Learning Yield Machine Intelligence? . . . . . . . . . . . . . . . 225

Department of Philosophy, National University of Singapore,

Abstract. How has multiplicity superseded impossibility in philosophical

Keywords: Artiﬁcial consciousness Machine consciousness

I want to trace a trajectory in recent philosophical debates on artiﬁcial consciousness. In

© Springer Nature Switzerland AG 2018

Many surveys on artificial consciousness stress that this sub-field in artificial

2 The Impossibility of Artiﬁcial Consciousness

The literature on artiﬁcial consciousness contains several philosophical challenges to

explanation between levels depends on context: it is often determined by which

3 The Multiplicity in Phenomenal Consciousness

and lower-level representations which are too local to be coherent. In computational

Fig. 1. The multiple kinds in phenomenal consciousness

systematically related to conscious behaviour in humans. By focusing on the systematic

4 The Development of Scientiﬁc Taxonomies

Fig. 2. Two overlapping kinds of biological species

Fig. 3. Two overlapping kinds of animals

In this paper, I assessed a trajectory in which multiplicity superseded impossibility in

Abstract. Cognitive science is considered to be the study of mind (consciousness

Cognitive science is currently deﬁned as a study of processes of knowledge generation

© Springer Nature Switzerland AG 2018

2 Problems Solutions with a Broader View of Cognition

Revisiting the list of unsolved/unsolvable problems of cognitive science under the

The Emotion Challenge: Morphological computing of embodied cognition has layered

The Consciousness Challenge: Consciousness is proposed as information integration

The World Challenge: Distributed morphological computation processes representing

The Body Challenge: Explicit modeling of a body is a consequence of the inclusion of

The Mathematics Challenge (brain cannot be conventional computer): Morphological

You might also like