Ebin - Pub - Platos Essentialism Reinterpreting The Theory of Forms 97811089862439781108833660

PLATO’S ESSENTIALISM
In this book, Vasilis Politis argues that Plato’s Forms are essences, not
merely things that have an essence. Politis shows that understanding
Plato’s theory of Forms as a theory of essence presents a serious
challenge to contemporary philosophers who regard essentialism
as little more than an optional item on the philosophical menu.
This approach, he suggests, also constitutes a sharp critique of those
who view Aristotelian essentialism as the only sensible position:
Plato’s essentialism, Politis demonstrates, is a well-argued, rigorous,
and coherent theory, and a viable competitor to that of Aristotle.
This book will appeal to students and scholars with an interest in the
intersection between philosophy and the history of philosophy.
  is Professor of Philosophy at Trinity College

Dublin. He is author of numerous books, including The Structure
of Enquiry in Plato’s Early Dialogues (Cambridge University Press,
) and The Aporetic Tradition in Ancient Philosophy (with George
Karamanolis, Cambridge University Press, ).
Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108986243
PLATO’S ESSENTIALISM
Reinterpreting the Theory of Forms
VASILIS POLITIS
Trinity College Dublin

University Printing House, Cambridge  , United Kingdom
One Liberty Plaza, th Floor, New York,  , USA
 Williamstown Road, Port Melbourne,  , Australia
–, rd Floor, Plot , Splendor Forum, Jasola District Centre, New Delhi – , India
 Anson Road, #–/, Singapore 
Cambridge University Press is part of the University of Cambridge.

It furthers the University’s mission by disseminating knowledge in the pursuit of
education, learning, and research at the highest international levels of excellence.
www.cambridge.org
Information on this title: www.cambridge.org/
: ./
© Vasilis Politis 
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 
A catalogue record for this publication is available from the British Library.
 ---- Hardback
Cambridge University Press has no responsibility for the persistence or accuracy
of URLs for external or third-party internet websites referred to in this publication
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.

For Lesley Brown, John Dillon, David Berman
mentors, friends, interlocutors

Contents
Acknowledgements page ix
Introduction 
 Why cannot the ti esti question be answered by example
and exemplar?
Hippias Major 
 Why cannot essences, or Forms, be perceived by the senses?
Hippias Major. Phaedo. Republic 
 Why are essences, or Forms, unitary, uniform and
non-composite? Why are they changeless? Eternal?
Are they logically independent of each other?
Phaedo and Republic 
 The relation between knowledge and enquiry
in the Phaedo 
 Why are essences, or Forms, distinct from sense-perceptible
things?
Phaedo  and Republic V. – 
 Why are essences, or Forms, the basis of all causation
and explanation?
Phaedo – 
 What is the role of essences, or Forms, in judgements
about sense-perceptible and physical things?
Republic VII. – 
vii

viii Contents
 Why does thinking of things require essences, or Forms?
Parmenides 
 Why are essences, or Forms, separate from physical things?
Also Timaeus and Philebus 
 What yokes together mind and world?
Phaedo – and Republic VI. – 
Conclusion: Forms simply are essences, not things that
have essences 
Bibliography 
General Index 
Index Locorum 

Acknowledgements
My gratitude extends, above all, to John Dillon, The Trinity Plato Centre
(which, as it happens, is located in Trinity College Dublin) and all those
who work there and have every day been providing me with a lived space
that we have been shaping jointly. I had the pleasure and privilege to
experience a similar environment during the winter of  in The
Durham Institute of Advanced Study, and to it and its excellent directors,
and Nicholas Saul not least, I am grateful. At the outset of this project
I had two wonderful stays, first at Wuhan University in the summer and
autumn of , made even better by my host there, Hao Chanchi, and
his hospitality, and then at Uppsala University in the spring of , made
just as enjoyable by my host and friend there, Pauliina Remes, and her
hospitality. In both places, I enjoyed many seminars with excellent grad-
uate students. I had the opportunity to present individual talks on one or
the other item in this book in several places in China, Europe and at São
Paulo University and Campinas University in Brazil. I thank my friend,
George Karamanolis, my host at Vienna University, and my friend, Klaus
Corcilius, my host at Tübingen University. I am also grateful to my friend
and host Richard King during my time at Bern University. Several uni-
versities in China provided more extended points of refuge during the past
four years and the writing of this book, most especially Wuhan University,
Renmin University and China University of Political Science and Law.
I am grateful to you all. Many people have contributed generously to this
book, and I fear I will forget some, but I will not forget Matthew Adams,
Keith Begley, Nicolò Benzi, David Berman, Lesley Brown, Friedemann
Buddensiek, Damian Caluori, Laura Candiotto, Luca Castagnoli, Nick
Clairmont, Niall Connolly, Klaus Corcilius, Giulio Di Basilio, Filomena
Di Paola, John Dillon, Nathalie Ek, Paolo Fait, Ge Tianqin, Zuzanna
Gnatek, Margaret Hampson, Vivil Valvik Haraldsen, Verity Harte, Blake
Hestir, Thomas Hodgson, David Horan, Martin Jacobson, Jiao Liming,
George Karamanolis, Kate Kiernan, Richard King, Hermann Körner, Inna
ix

x Acknowledgements
Kupreeva, Jens Kristian Larsen, Peter Larsen, Jim Levine, Ruairi Maguire,
Mary Margaret McCabe, David Meissner, James Miller, Nie Minli, John
Nugent, Kenneth Pearce, Sverre Raffnsøe, Nigel Rapport, Pauliina Remes,
Samuel Rickless, Evan Rodriguez, Henrik Rojahn, Pauline Sabrier, Kara
Schechtman, Colm Shanahan, Philipp Steinkrüger, Damien Storey, Su
Jun, Jan Szaif, Oda Tvedt, Daniel Vazquez, Tom Vogt, Wang Wei, Daniel
Watts, Wei Liu, Manfred Weltecke, Benjamin White, Xin Liu and Zhang
Jiayu. I thank you all.

Introduction
The topic of the present study is Plato’s theory of Forms, as it used to be

called. The thesis of the study is that Plato’s Forms simply are essences and
that Plato’s theory of Forms is a theory of essence – essences, in the sense
of what we are committed to by the supposition that the ti esti (‘What is
it?’) question can be posed and, all going well, answered. This thesis says
that the characteristics that, as is generally recognised, Plato attributes to
Forms, he attributes to them because he thinks that it can be shown that
essences, on the original and minimal sense of essence, must be so
characterised. The characteristics that Plato attributes to Forms include
the following: Forms are changeless, uniform, not perceptible by the
senses, knowable only by reasoning, the basis of causation and explanation,
distinct from sense-perceptible things, necessary for thought and speech,
separate from physical things. According to the thesis of this study, each
and all of these characteristics of Forms can be derived, Plato thinks, from
the supposition that we can ask and, all going well, answer the ti esti
question adequately and truly, and the supposition that, whatever else
Forms are and is characteristic of them, they are essences, essences in the
sense of that which is designated by an adequate and true answer to the ti
esti question.
For Plato, the question ‘What is . . . ?’ is not, originally and according to
its original meaning and use – the meaning and use shared by Socrates’
understanding of it and the understanding of it by his interlocutors – a
philosophical, much less technical question, the posing of which commits
one to the existence of essences in a disputable or controversial sense. In
several dialogues, Plato presents Socrates’ interlocutors, those who are
without philosophical background or training, as taking themselves to be
immediately capable of understanding this question, and indeed of answer-
ing it, and presents Socrates as having to do much work to persuade them,
and Plato the reader, that things are not so simple. That which the ti esti
question asks for, in its original meaning and use, is a standard


of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108986243.001
 Introduction
(paradeigma) for a thing’s being of a certain quality and a way of deter-
mining whether a thing is such as to be of a certain quality.
If there is one thing that Socrates, as Plato represents him, is convinced
of, it is that the ti esti question, especially when asked of certain things or
qualities, such as beauty, equality, unity, justice, is a most important and
profoundly difficult one, the answering of which is a major undertaking
and requires demanding enquiry. At the same time, there is a dialogue, the
Hippias Major, in which, as we shall see, Plato has a character, Hippias,
present Socrates and his distinctive convictions about the ti esti question
with a monumental challenge. For, Hippias insists, the question, ‘What is
beauty?’, is ‘trivial and worth practically nothing’ (e–). He does this
because he argues that the ti esti question can be answered, with ease and
without any enquiry to speak of, by pointing to an example of a particular
thing of exemplary beauty, such as a girl, or a horse or a lyre; a thing,
therefore, capable of serving as an adequate standard for a thing’s being
beautiful and for determining of a thing whether or not it is such as to
be beautiful.
At the same time, there can be no doubt that Plato thinks that the ti esti
question, when properly considered and especially when asked of certain
things or qualities, is a philosophical one, whose answer, to be adequate,
must conform to certain requirements that are substantive and potentially
subject to controversy and dispute. The most important of these require-
ments says that, when asked of certain things or qualities, the ti esti
question cannot be adequately answered by example and exemplar, that
is, in the way Hippias insists and argues it can. It does not follow from this
that the appeal to an example and exemplar of a thing that is F cannot
contribute to the search for an answer to the question ‘What is F?’; what it
means, rather, is that this appeal is not, by itself, adequate for answering
the question. The reason why this is the most important requirement that
Plato associates with the ti esti question, when asked of certain things or
qualities, is that, if the ti esti question can be answered by example and
exemplar, it is, as Hippias points out, so easy to answer as to be trivial and
worth practically nothing.
Plato associates further substantive requirements with the ti esti ques-
tion; in particular, the answer to the question must be unitary, and it must
be explanatory. I shall not spell out these requirements here; I have done so
elsewhere (Politis ; , ch. ), and we will have the opportunity to
consider them, and their consequences and relevance for the theory of
Forms, at the proper junctures of the study. What is important to observe
is that, by adding these requirements – generality (i.e., not by example and

Introduction 
exemplar), unity, explanatoriness – to the original and basic meaning and
use of the ti esti question, that is, the meaning and use that leaves open
what is, or provides, an adequate standard (paradeigma) for a thing or
quality, Plato is, in effect, introducing a philosophical, and potentially
controversial and disputable, notion of essence: the essence of a thing or
quality, F, is that which is designated by a true answer to the question
‘What is F?’, and this answer has to conform to certain
substantive requirements.
For brevity, I shall simply say that the essence of a thing or quality, F, is
that which is designated by an adequate and true answer to the question
‘What is F?’ If this should make a reader object that, on this account of
essence, even Hippias is committed to essences, I think the following
answer will do: On a strictly minimal notion of essence, so be it; on a
strictly maximal notion of essence, only a true answer to the ti esti question
that satisfies all the requirements that a philosopher may associate with the
question – generality, unity, explanatoriness, as well as any further require-
ments that, for particular reasons, she may associate with it – designates an
essence; and, obviously, there are notions of essence in between these
two extremes.
Let me, without further delay, return to the point of the present project.
I argue that, for Plato and in regard to certain things or qualities, the
answer to the ti esti question commits us to entities whose existence is
controversial and disputed by people in general, namely, what Plato calls
Forms (eidē). An important reason why the existence of such entities is
not evident, but disputable, is that, in such dialogues as Phaedo, Republic,
Parmenides and likewise Timaeus and Philebus, Forms are characterised as
having a number of unfamiliar and remarkable characteristics: Forms are

Plato sometimes uses eidos and idea interchangeably, when talking about Forms or about essences
(such as Republic V. a; and, perhaps, Euthyphro d–e); but not always. He sometimes uses idea
when he says that an eidos is, precisely, an idea (Greek term) that is always the same. (See, e.g.,
Parmenides a– and b–c. In passages such as Republic VI. b–, and perhaps even
V. a, it is not clear whether he uses idea for eidos or, rather, in this other way.) When he uses
idea in this way, it would not be right to translate ‘Form’ for idea; we may translate, rather,
‘character’ or ‘quality’. Plato’s point will then be that a Form is, precisely, a character or quality
that is always the same. Remarkably, this shows that Plato does not simply assume that a quality must
be always the same in every one of its instantiations or occurrences.

The inclusion of the Timaeus and Philebus, in addition to Phaedo-Republic-Parmenides, may raise
some eyebrows, since it is generally thought that they are much later. My reason for including them
is that, as I will argue, they are, in regard to the theory of Forms, continuous with the Republic, and
appear to be intended by Plato as so being. Of course, once we include the Timaeus and Philebus, we
ought, ideally, to consider, in regard to Forms, also dialogues generally thought to be later than the
Parmenides but earlier than them; such as, especially, Theaetetus, Sophist, Statesman, as well as
Phaedrus and Symposium wherever we place them in the relative chronology overall and indeed in

 Introduction
changeless, uniform, not perceptible by the senses, knowable only by
reasoning, the basis of causation and explanation, distinct from sense-
perceptible things, necessary for thought and speech, separate from phys-
ical things. I cannot seriously entertain the idea that Plato intends the
existence of such entities to be taken on trust. In particular, we must
wonder why he thinks that these characteristics are all true of the same
entities (that they are co-referring and co-extensive) and, most important,
why he thinks that they are true of, precisely, those entities to which we
come to be committed through the posing of the ti esti question.
Unless and until we take up and properly consider these critical ques-
tions – by what process of reasoning, in Plato, does the posing of the ti esti
question turn into the commitment to Forms? What justifies the suppo-
sition that the several characteristics of Forms are all true of the same
things, and of the very things we are committed to by the posing of the ti
esti question? – we are not in a position to suppose that Plato has, or
intends to have, a theory of essence, or a theory of Forms: a theory, in the
sense of a single account that he intends to be coherent and unitary, such
that accepting one element in it commits one to accepting all the elements
in it. These critical questions are at the core of the present study.
If I dedicate a book-length study to this task, it is not because I want to
take issue with one or another interpretation of Plato’s theory of Forms –
as it used to be called, before such systematic interpretations were practi-
cally displaced by a single-minded preoccupation, by critics setting the
tone, with the dialogical drama displayed by a Platonic dialogue, based on
one or another single dialogue, warning against dialogue-crossing, and
predisposed against the search for systematicity, or theory, in Plato. My
motives derive, in part, from a basic concern I have about the current state
of the art regarding Plato’s Forms and their relation to his essences, that is,
essences, in the sense of those entities, whatever they may be, and however
disputable or not they may be, to which one is committed by the
relation to each other. It is simply that doing this would take us too far . . . and we would need many
more words than Cambridge University Press admits.

Thus Annas (, ): ‘It is often said that Plato has a “Theory” of Forms and even that it
dominates his entire work. In fact Forms appear rarely and are always discussed non-technically; they
answer to a variety of needs which are never systematically brought together . . .. If we ask “What are
Forms?” we find a variety of answers.’ If the argument of the present study is on the right lines, this
statement is the opposite of the truth.

I recommend Christopher Rowe’s salutary opposition to this tendency: ‘Talk of “versatility” is in
danger of suggesting that we can retreat into interpreting each dialogue on its own (as some scholars
in the last two centuries have attempted to do), and there are too many connections between them,
too many constants, to make that a viable proposition’ (, ). Rowe goes on to spell out the need
to read the dialogues together.

Introduction 
supposition that the ti esti question can be posed and, all going well,
answered. My impression is that those critics who are still occupied with
Plato’s Forms, including the best and most perceptive of them such as
Verity Harte, consider it adequate to characterise Plato’s Forms as being
those entities that satisfy a set of characteristics such as those listed above,
to which is commonly added (though sometimes this seems to be forgot-
ten!) that Forms are essences, in the sense of that which is designated by a
true answer to a ti esti question. This is to characterise Plato’s Forms
through a list of characteristics; and this, I believe, is not at all to
understand what Plato’s Forms are or why one would believe in
such things.
I know of a single exception, of some time ago, to this tendency among
critics, by a critic who very much marks the alternative approach I want to
take up and defend. This is Alan Code, when he says: ‘Plato’s realm of
separable being is not the realm of existence, though of course its inhab-
itants are supposed to exist. Rather, it is the domain of definable entities –
the objects about which one asks the Socratic “What is F?” question’
(, ). However, Code did not defend this statement; or, if he
did, no one appears to have taken notice, perhaps because of the exceed-
ingly compressed way in which he did defend it. In his  classic,
Plato’s Theory of Ideas, David Ross, for one, asked whether Plato’s theory of
Forms contains ‘an essential core’ () – this being the question of the
present study.
By itself, the proposition that there are entities that satisfy this, or some
suitably similar, list of characteristics, must remain a philosophical curios-
ity, no doubt fascinating and worthy of a visionary mind, but otherwise
something it is hard to know what to do with except admire it and wonder
at how strange and incredible it is. If this is what Plato’s Forms are and
what the theory of Forms is, the recommendation would not be unrea-
sonable that said that, having duly noted Plato’s commitment to such

Harte () begins (–) by stating, as basic characteristics of Forms, the following four, in
which the notion of essence does not figure: . Forms are primary beings; . they have causal
responsibility; . they are privileged bearers of certain terms and . they are objects of a knowledge of
a privileged sort. Later Harte characterises Forms as being essences, when she says (in regard of
Phaedo c–d): ‘Socratic questions ask “What is F?” for some range of properties. The Form is
identified as “What is [F]” – that is, as the referent of the answer to this Socratic inquiry’ (). See
Rowe (, ) for a similar characterisation. Rowe is especially clear and explicit that the whole
idea of Forms goes back to the Socratic ti esti question. See Chapter .

Silverman’s The Dialectic of Essence () promises to bring Forms closer to essences, but in fact
Silverman argues against the view that a Form is identical with its essence, or that Forms are essences,
and he considers Forms to be, rather, the ‘bearers’ of essences. I come back to this issue in the
Conclusion of the present study – which the reader is welcome to read in advance.

 Introduction
entities, we had better limit ourselves to the examination of the dialogical,
dramatical and argumentative ins and outs of each of Plato’s dialogues and
leave it at that.
If I may be allowed some pathos and a little exaggeration, this is to miss
the point not only of Plato’s Forms but of his philosophy. For I submit
that, for Plato, Forms simply are essences. Essences, Plato thinks, simply
are that which we are committed to, in one way or another, directly or
indirectly, by the pursuit of ti esti questions and this pursuit’s logical
ramifications; and ti esti questions are thoroughly caught up in the dialog-
ical, dramatical and argumentative enquiries, and the aporiai that all this is
ultimately rooted in, that make up the dialogues.
Plato has the reader work hard to identify the basic elements in his
philosophy, whatever they are, especially by placing in particular dramatic
settings, and by practising philosophy as drama, the ways in which they are
worked out, and situating this drama in historically inspired contexts
engaging with intellectual figures alive and dead (e.g., lesser and greater
sophists, generals, poets, dramatists, Heraclitus, Parmenides, Homer . . .)
as well as with other forces and presences such as the war and Athens and
the fatal mix-up of the two. In looking for these elements, I try to be
sensitive to the dialogical character of Plato’s arguments; such as the
distinction between arguments addressed to those who have not yet
accepted this, that or the other feature of essences and Forms and argu-
ments addressed to those who have already accepted this. I shall,
I recognise, be less occupied with the historical dimension, both in and
out of the dialogues – much as I’d have liked to have integrated this more
into the study.
The defence of the thesis that Plato’s Forms simply are essences and that
Plato’s theory of Forms is a theory of essence must, of necessity, take a
certain general form and proceed by a certain series of logical steps. It must
start with the recognition, not only that the ti esti question is at the root of
numerous enquiries in the dialogues – so much is commonly recognised –
but that Plato is acutely aware of a debunking and, as we would say,
deflationary response to this question, which says that the question is
‘trivial and worth practically nothing’, because it can be answered, with
ease and without any enquiry to speak of, by pointing to an example and
exemplar, and therefore an adequate standard, of what it asks for. For, if
the ti esti question can be answered by example and exemplar, and if a
standard of a thing or quality can be provided in this way, there is no
reason to think that more is needed to answer it and afford the desired
standard; and so the great substance that Plato affords this question, and

Introduction 
the difficult and weighty enquiries he associates with it, is vain. Plato did
not have to wait for Wittgenstein or Geach to be presented with this
radical challenge, he has Hippias present it in the Hippias Major. This is a
dialogue the question of whose authenticity is rebutted once and for all by
the recognition of the function the dialogue serves, through the radical
dispute it contains regarding the ti esti question and whether it can be
answered by example and exemplar: Hippias argues (he does not simply
assume) that it can, Socrates that it cannot. This dialogue points to a basic
element in Plato’s approach to philosophy, philosophical questions and
philosophical enquiry, namely, the ti esti question associated with certain
substantial and demanding requirements for its answer; it acknowledges
that this element is disputable; and it provides a dialectical defence of it,
that is, an argument against the debunking and deflationary alternative.
This is the point at which I begin, in Chapter .
Having taken this first, basic, step, of determining why Plato thinks that
there is no easy or readily available way of answering the ti esti question and
of providing a standard for certain things or qualities, it is of the essence
that we proceed with particular caution and care, in order to determine
what is the next move that Plato makes, and makes on just this basis: the
supposition, itself properly defended (in the Hippias Major), that the ti esti
question, at least when asked of certain things or qualities, cannot be
answered by example and exemplar. It will not do to proceed by supposing
that, at a certain point in his development, such as when he wrote the
Phaedo and Republic, Plato came to think that that which is designated by a
true answer to a ti esti question is a Platonic Form, that is, an entity that
satisfies some or all of a set of the mentioned characteristics. This will not
do, not because of any general misgivings one may have about develop-
mentalism, the view that Plato’s philosophy develops through certain
relatively distinct stages, but because it is not at all evident, but, on the
contrary, perfectly obscure, why the alternative to the quick and easy way
(as recommended by Hippias) of providing a standard of what something
is, that is, by example and exemplar, implies the commitment to entities
that have a single, much less some or all, of these characteristics.

Chapter  is a revised version of Politis (a).

It is a consequence of the thesis defended in the present study that developmentalism, in regard to
Forms and the theory of Forms, is mistaken; and it is mistaken irrespective of whether by
developmentalism we mean the view that Plato’s claims about Forms in such dialogues as Phaedo
and Republic are inconsistent with his claims in dialogues we consider to be earlier, or, on the
contrary, we mean the view that Plato’s claims about Forms in such dialogues as Phaedo and
Republic, though consistent with his earlier claims, introduce new claims that are not continuous

 Introduction
Nor will it help to add, as critics commonly have done, and still do even
while having, on the whole, turned against developmentalism, that this
step in Plato’s development, when he came to think of essences as Platonic
Forms, marks the philosopher’s recognisable and archetypical turn to
metaphysics and the questions of what there is and what are the most
basic and primary entities there are. It is not only that this narrative does
not begin to indicate why Plato thinks that all the mentioned character-
istics are true of the same entities (are co-referring and co-extensive) and of
those entities that are designated by a true answer to the ti esti question –
these being absolutely critical questions without a sense of the answer to
which we are entirely in the dark about Plato’s Forms. The narrative begs a
monumental question, namely, that Plato’s theory of Forms is, basically, a
theory about what there is and what are the most basic and primary entities
there are: a metaphysical theory in this sense. To beg this question is no
mean sin, for it is to ignore the possibility that what Plato’s theory of
Forms basically is, is a theory of essence, in the sense of a theory of what we
are committed to in thinking that the ti esti question can be posed and, all
going well, answered: a metaphysical theory, if you like, but in this quite
different sense. It is to confuse things to suppose, from the start, that a
theory of essence is a theory of what there is and what are the most basic
and primary entities there are; a supposition that, to be worthy of consid-
eration, requires a major and ambitious argument, such as we find in
Aristotle’s Metaphysics, and is anything but straightforward.
What we must do, rather, is identify a characteristic that Plato thinks
essences, or Forms, must satisfy, and of which it can be shown that he
thinks that this characteristic follows from the supposition, itself properly
defended (in the Hippias Major), that the ti esti question, at least when
asked of certain qualities, such as beauty, equality, unity, justice, cannot be
answered by example and exemplar. I argue, in Chapter , that there is
such a characteristic: it is that essences, or Forms, cannot be perceived by
the senses, which Plato, famously, asserts in Phaedo and Republic. I argue
that the reason why Plato thinks that certain essences, or Forms, cannot be
perceived by the senses is, precisely, that he thinks that what certain things
or qualities are cannot be specified (determined, defined, known) by
example and exemplar. If this is correct, there is absolutely no need to
with, or based on, or justifiable on the basis of earlier claims. (For a recent defence of
developmentalism in regard to Forms, see Dancy (). For critical assessments of
developmentalism in general in regard to Plato, see Annas and Rowe (), also Kahn ().)

For the conception of metaphysics that is associated with questions about essences and what things
are like essentially, see the classic paper by Kit Fine ().

Introduction 
suppose that Plato’s commitment to Forms marks a turn to metaphysics,
or any turn at all; all it marks is another step in Plato’s process of
determining, by logic and the art of reasoning when in the hands of a
master craftsman, what we are committed to in thinking that the ti esti
question can be posed and, all going well, answered.
My intention is to proceed in this way from the beginning to the end,
practically, of the present study. By relying on Hippias Major, Phaedo,
Republic, Parmenides (also, more briefly, Timaeus and Philebus, and the
occasional mention of other dialogues, too), I want to demonstrate that
Plato defends each of the characteristics of his essences, or Forms, on the
basis of the supposition alone that the ti esti question can be posed and, all
going well, answered, and everything that follows from that. If this effort
can be successfully sustained, we will have shown that Plato defends a
comprehensive, coherent and well-argued theory of essence, based ulti-
mately on a single question, and one that is not philosophically contro-
versial or predisposed towards one substantial theory or prejudiced
against another.
So far, I have anticipated two elements in this, Plato’s theory of essence,
which, as I shall demonstrate, Plato defends on the basis of the supposition
that the ti esti question can be posed and, all going well, answered: that,
when asked of certain things or qualities, the ti esti question cannot be
adequately answered by example and exemplar (Chapter ) and that, when
asked of such things or qualities, that which is designated by an adequate
and true answer to the ti esti question cannot be perceived by the senses
(Chapter ). A further element in Plato’s theory of essence is his claim that
essences, or Forms, are changeless and that they are uniform and non-
composite; which, famously, he makes in the Phaedo. In Chapter , I argue
that the claim that Forms are uniform and non-composite is derived from
the claim that the ti esti question must be answered with an account that is
unitary; and I argue that the claim that Forms are changeless is derived
from the fact that the ti esti question is a request for a standard of what a
thing is and from the requirement that this standard must be unitary. It
follows, I conclude contra a prominent line of critics, that we have no
reason to think that Plato’s Forms, in the Phaedo or Republic, are supposed
to be logically independent of each other.
A yet further element in Plato’s theory of essence is his claim (in Phaedo)
that essences, or Forms, are distinct from and not identical with sense-
perceptible things. In Chapter , I argue that this claim is derived from the
claims that the ti esti question, when asked of certain things or qualities,
cannot be answered by example and exemplar, and that that which it

 Introduction
designates, if it is answered adequately and truly, cannot be perceived by
the senses.
Let me not anticipate further how I intend to show that Plato defends,
in this thoroughly essence-based way, each of the elements in his theory of
essence – which is indeed what the theory of Forms is, if the argument of
the present study is on the right lines. I do, though, need to call attention
to a particular feature of this, the thoroughly essence-based way in which,
as I argue, Plato defends the elements in his theory of Forms. So far, the
elements whose distinctively essence-based defence I have anticipated are
derived, by Plato, in a direct way from the supposition that the ti esti
question can be posed and, all going well, answered. However, not all of
Plato’s essence-based arguments for an element in his theory of Forms are
as direct. Let me explain with what is a particularly important, and
clear, case.
One of the elements in Plato’s theory of Forms is the claim that
essences, or Forms, are necessary for, and provide the basis of, all causation
and explanation; a claim that, famously, he makes and defends towards the
end of Phaedo (e ff.). I argue (in Chapter ) that this claim, too, is
defended by Plato in a thoroughly essence-based way. However, in this
case, he does not rely only on the suppositions that there are essences and
that essences cannot be specified by example and exemplar or perceived by
the senses. For he also relies on the supposition, itself defended in the
Phaedo, that causation and explanation is uniform: same cause and expla-
nans if, and only if, same effect and explanandum.
This, I argue, is all that Plato relies on in defence of the claim that
essences, or Forms, are necessary for, and provide the basis of, all causation
and explanation. In particular, he does not rely on a claim that says that the
cause transmits its quality to the effect and hence must be like the effect –
a claim from which it follows that Forms, at any rate in so far as they are
causes, are self-predicative: the Form of F is itself F. Critics have com-
monly attributed to Plato these suppositions – the transmission theory of
causation and self-predication – to make sense of his argument for the
claim that causation requires Forms and is based on Forms. Neither of
these suppositions, I argue, are needed to make sense of Plato’s argument,

Chapter  is a revised version of Politis ().

That Plato is not committed to this principle of causation (i.e., that causation works by causes
transmitting their character to their effect) was argued, in regard to dialogues before Phaedo, by
Malcolm (, –, –). However, the principle is now commonly (e.g., by Sedley [])
invoked in the account of Plato’s argument in Phaedo e ff.

Introduction 
once that argument is properly understood; and neither are present in the
text of the Phaedo. The reason why critics have thought that Plato’s
argument is deficient without these suppositions, I argue, is that they have
failed to recognise the crux of Plato’s argument: that Forms – the Forms
that are argued to be necessary for and the basis of causation and expla-
nation – are, precisely, essences.
I hope the interest and significance of this undertaking is evident: the
undertaking of showing that Plato defends a comprehensive, coherent, and
well-argued theory of essence, based ultimately on a single question, the ti
esti question, which is understood, originally, in a philosophically unpre-
judiced and uncompromising way. This, it is true, is not by itself to show
how important such a theory of essence is. To consider how important
such a theory of essence is, thoroughly based on the ti esti question, we
would have to consider how important it is to pose, and try to answer, ti
esti questions, both in general and for this, that or the other thing or
quality. For this task, there is no better way than to study Plato’s dialogues,
which abound in such questions, the pursuit of answers to them and
the philosophical problems and aporiai with which ti esti questions are
caught up from the start. I note that, in The Structure of Enquiry in
Plato’s Early Dialogues (Cambridge, ), I have offered a general
account of why the posing of ti esti questions is so important (see also
Politis ; b).
Let me point to some ways in which this undertaking is of particular
interest and significance; and this in addition to the point I have already
been making: that Plato’s Forms simply are essences, and that his theory of
Forms simply is a theory of essence. For it follows from this that, contra a
common understanding of Plato’s Forms, we have no reason at all to
suppose that Forms are substances that have essences; all we have reason to
suppose is that they simply are essences. I shall conclude the present study
with this result and add that this shows that Plato’s Forms are not self-
predicative, or self-predicative in the way they would be if they were
substances having essences and distinct from their essence.
I think it is fair to say that, today, the original idea of essence and a
theory of essence is associated, thoroughly and exclusively, with Aristotle.
When telling colleagues and friends, both experts in ancient essentialism
and other philosophers, including neo-Aristotelians and other current so-
called analytical metaphysicians, of my project and its central thesis, they
have commonly reacted by observing that, surely, a theory of essence is a
theory of substances belonging to natural kinds and having essences and
that it is Aristotle, not Plato, who introduced the idea of substances

 Introduction
belonging to natural kinds and having essences. It seems to me that there
is something clearly and significantly mistaken about the view that an
original theory of essence is a theory of substances belonging to natural
kinds and having essences – at any rate if such a theory is associated with
the name of Aristotle. For, it seems quite clear that Aristotle’s theory of
essence, which is indeed a theory of substances having essences, is not, and
is not intended by Aristotle as being, purely and simply, a theory of
essence. Rather, it is, and is intended by Aristotle as being, a combination
of two theories, addressed to two different questions: a theory of what are
the most primary things there are, which says that the most primary things
there are are substances in the sense of ultimate subjects of predication (to
eschaton hupokeimenon) and a theory of essence, associated with the notion
of what something is (ti esti) and what it is for a thing to be the thing it is
(to ti ēn einai). That these are distinct theories ought to be evident; for they
are addressed to different questions. It is far from evident that these two
theories must be combined, that is, that either of the two questions can be
properly considered only if both questions are considered, and considered
jointly and together. This is not to object to Aristotle’s theory of essence,
combined, as it is in Aristotle, with a theory of substance and primary
being. Aristotle does not assume that the two theories must be combined.
On the contrary, he expends great effort (in the Metaphysics, starting
especially in book Zeta) arguing that the two theories must be combined
(especially in Zeta –, when he begins to argue that only substances have
an essence strictly speaking). It is to object to the assumption, common
today and made without argument, that an original theory of essence is an
Aristotelian theory of substance: the view that there are substances, and
that they have properties and that some of these properties are essential to
them. To assume this is to assume that there can be no theory of essence
except in combination with a theory of primary being and the view that
the primary beings are substances: a monumental assumption. If philoso-
phers today make it without argument, I suspect it is because the
Aristotelian categories of being, with substances at the ontological base,
have come to appear to many philosophers, not as the commitments of a
highly complex and ambitious philosophical theory, but as plain
common sense.

The following statement from Oderberg (, ), however incredible, is characteristic: ‘There is
also Platonist essentialism, but I do not count this as real essentialism according to the sort of
position I defend.’

Introduction 
I have already pointed to another, if closely related, remarkable conse-
quence of the present undertaking, which is worth repeating. For if Plato
defends a theory of essence, and if this is what the theory of Forms
basically is and if this theory is based ultimately on a single question, the
ti esti question and what we are committed to in thinking that it can be
posed and, all going well, answered, then we may not suppose that Plato’s
theory of essence, or his theory of Forms, is a theory of what there is and
what are the basic and most primary things there are; at any rate this is not
the primary aim of Plato’s theory. This is not to rule out that, in Plato, the
theory of essence is, at some particular point, combined with a theory of
what there is and what are the most primary things there are. (I am
inclined to think this happens in the Sophist  ff.) It is to insist
that Plato’s theory of essence is not based in these questions – What is
there? What is primary being? – and that, if at some point the questions
come up in Plato, we must ask why they come up at that point and
whether the posing of them by Plato is, once again, motivated by
the pursuit of ti esti questions. (This task, which requires a careful
study of the Sophist, goes beyond the present study.) None of this is to
deny that, in the course of the logical unfolding of the implications of the
ti esti question, Plato commits himself to substantial claims about
what there is – for example, to entities that are changeless and
cannot be perceived by the senses, which he thinks some people, whom
he calls ‘lovers of sights’, philotheamenoi, refuse to admit (Republic, end of
book V), and to two distinct sorts of things (duo eidē tōn ontōn), objects of
the senses (aisthēta) and objects of the intellect (noēta, Phaedo a). It is,
rather, to insist that, in Plato, such claims are consequences, relating to
being and what there is, of Plato’s considering what we are committed to
when we think that the ti esti question can be posed and, all going
well, answered.
There is also the whole question of Plato’s epistemology, as it is
commonly called. It is a, no less remarkable, consequence of the present
undertaking that, just as Plato’s theory of essence and his theory of Forms
is not, or not primarily, a metaphysical theory – a theory of what there is
and what are the basic and most primary things there are – so, too, it is not
an epistemological theory – a theory of what knowledge is and what can be
known. This is not to deny that there are epistemological claims, as we
would call them, in Plato’s theory of essence and his theory of Forms. The
claim that essences, or Forms, cannot be perceived by the senses is
evidently such a claim. It is, rather, to insist that, in Plato, such claims
are consequences, relating to knowledge, of his considering what we are

 Introduction
committed to when we think that the ti esti question can be posed and, all
going well, answered.
In regard to those claims in Plato’s theory of essence that we may
characterise as epistemological, I want to anticipate, finally and in the
remainder of this introduction, three notable consequences of the
present project.
First (I argue in Chapters  and ), it is a mistake to suppose, as critics
commonly have, that Plato’s claim that Forms cannot be perceived by the
senses amounts to, or implies, the claim that Forms can be known only a
priori and in a way that is independent of sense-perception. That this is a
mistake becomes apparent once we recognise why Plato thinks that Forms
cannot be perceived by the senses; namely, because he thinks that what
certain things or qualities are cannot be specified by example and exem-
plar. The claim that what certain things or qualities are cannot be specified
by example and exemplar implies, I argue, that what these things or
qualities are cannot be perceived by the senses; but it does not imply that
what these things or qualities are can be known only a priori and in a way
that is independent of sense-perception. This is not, of course, to deny
that, for Plato, Forms are objects of the intellect (noēta); it is to deny that
Plato’s distinction and opposition between objects of the senses (aisthēta)
and objects of the intellect (noēta) is equivalent with the modern distinc-
tion between things known a posteriori and things known a priori.
Secondly (I argue in Chapter ), Plato’s view, familiar especially from
Phaedo and Republic, that essences and Forms must be known by reason-
ing, does not rule out that such reasoning involves, and even depends on,
sense-perception. In the Phaedo (a) Plato expressly says that such rea-
soning requires and depends on sense-perception. What it rules out is that
such reasoning can be bypassed in favour of answering the ti esti question
directly and by appeal to what is evident to sense-perception and anything
based on that. This is a short cut especially attractive to the mindset, both
philosophical and lay, that takes delight, and passion, in debunking and
deflating what appears to it to be an unnecessary attachment to pathos-
filled philosophical enquiry and philosophical depth. This deflationist
mindset, while he thinks it is a mark of one’s having lost one’s way,
Plato takes seriously and engages with through robust argument.
Thirdly, in Chapter  I argue that, while we may speak of Plato’s theory
of essence and his theory of Forms as containing epistemological claims, it
is a major, and questionable, assumption, common to many critics, to

Chapter  is a revised version of Politis ().

Introduction 
suppose that these claims are specifically about knowledge, as opposed to
being about enquiry. What we need to do, rather, I argue, is ask whether
the epistemological claims we find in Plato’s theory are only about knowl-
edge or also about enquiry; and, if we find that they are about both
knowledge and enquiry, we need to ask whether they are primarily about
knowledge and only as a consequence about enquiry or, on the contrary,
they are primarily about enquiry and only as a consequence about knowl-
edge, or things are even more complicated. In Chapter , I begin this task,
with reference to the Phaedo. I argue that a central epistemological claim in
the Phaedo (a), which says that sense-perception is necessary for think-
ing of, and is the one and only means by which one may arrive at the
thinking of, certain essences and Forms, is both about knowledge and
about enquiry and that this claim is primarily about enquiry and only as a
consequence about knowledge. I show that the most central epistemolog-
ical claim in the dialogue (esp. at a–b), which says that knowledge
depends on thinking and reasoning rather than on sensory perception, is
expressly formulated as being just as much about enquiry as about knowl-
edge. However, I do not pursue this task further in the present study; it
would require an independent study. Instead, in Chapter , I return to the
task at hand.
Addendum
I want to note that the thesis I am defending in this study – it says that
every item in Plato’s theory of Forms can, Plato thinks, be derived from
the posing of the ti esti question of certain things or qualities, and from
holding out that this question can be answered adequately and truly – is
intended to be compatible with the view that, for Plato, the knowledge of
(certain) Forms requires a sort of intellectual vision. The thesis I am
defending implies that the knowledge of Forms requires a verbal account
(a logos) of what something is. But it does not imply that such a verbal
account is sufficient for such knowledge. Notably, the thesis I am
defending in this study implies that, if the knowledge of (certain) Forms
requires a sort of intellectual vision, then this vision can be arrived at only
through engaging – discursively and dialectically – with verbal accounts;

Phaedo b says that one who knows is able to give an account, logos. If we suppose the logos
includes an account of what the thing known is, this says that knowing what a thing is is necessary
for knowledge; it does not say that knowing what a thing is sufficient for knowledge.

 Introduction
and so there is not a short cut to such a vision. But it would be no more
than a mistake to think that the thesis I am defending rules out the view
that, for Plato, the knowledge of (certain) Forms requires a sort of
intellectual vision.
I want to note this, for three reasons.
First, while I shall not, in this study, consider whether, for Plato, the
knowledge of (certain) Forms requires a sort of intellectual vision, I want
to take up this question in a follow-up project and (all going well)
resulting study.
Secondly, I think we scholars on Plato ought to go back to a debate that
was embarked on very well by Cross () and Bluck () almost
seventy years ago, but that has since (and with such notable exceptions as
Nightingale [] and McCabe []) been allowed to lapse. This is the
debate over whether, for Plato, the knowledge of Forms requires a sort of
intellectual vision; and, if so, whether this is Plato’s view in regard to only
some, very special, Forms, such as the Form of Beauty (this was Cross’
view), or, on the contrary, it is Plato’s view in regard to all Forms (this was
Bluck’s view).
Thirdly, some critics have recently wanted to exclude The Seventh Letter
from the Platonic corpus and have been confident and insistent in this.
I suspect that their insistence has something to do with the fact that, in this
letter, Plato, or ‘Plato’, has knowledge of Forms require intellectual vision.

For this reason, I am sceptical of Rowett’s () recent claim that, after early dialogues in which
the request for definitions is prominent, Plato throws out this request.

Myles Burnyeat and Michael Frede (a; b).

 
Why cannot the ti esti question be answered

by example and exemplar?
Hippias Major
There have been many responses to Geach’s charge that Plato dismissed
out of hand the possibility of definition by example and exemplar – that is,
the possibility of answering the question ‘What is F?’ by pointing to, or in
some other way calling to mind, a particular thing that is conspicuously
F and thus capable of serving as a standard for a thing’s being F. What
critics have missed, remarkably, is that Plato casts a particular character as
an able proponent of a cogent defence of definition by example and
exemplar. When, in the Hippias Major d, Socrates asks Hippias
‘What is beauty?’, he answers by invoking, as standards for a thing’s being
beautiful, and hence as proper answers to the ti esti question, first a
particular girl, then a particular horse and a particular lyre, of exemplary
beauty. He urges that, since such examples and exemplars of beautiful
things are ready to hand, no reasoning, argument or enquiry is necessary
for answering this seemingly momentous question, which is, on the
contrary, ‘something trivial and worth practically nothing’ (e–).
I confess that, since I was struck by this some time ago, I have found it
remarkable that critics, apparently without exception, have missed this

Geach (, –). The literature in response to Geach is large and seemingly never-ending.
Representative examples include: Anderson (); Santas (); Irwin (, –); Vlastos
(; ); Woodruff (; ); Reeve (); Fine (; ); Prior (); Wolfsdorf
().

Balaudé (,  ff.) is the only critic I know who recognises that Hippias puts forward the girl as
an example and exemplar of a beautiful thing: see , , . With the exception of Woodruff’s
() judicious and insightful commentary, the literature has not been kind to this dialogue, even
setting aside the issue of authenticity. For a long while critics were agreed that Hippias simply
misunderstands Socrates’ ti esti question and in doing so confuses particulars and universals. This
view is an old one, eloquently formulated by Apelt (, ) more than a century ago, who thinks
Hippias’ confusion is rather between things, or qualities, and essences: ‘Also was ist das Schöne?
Wohlverstanden das Schöne und nicht, was ist schön. Dieser Unterschied dünkt dem Sophisten so
fremd wie böhmische Dörfer . . . Meine [i.e., Socrates’] Frage aber ging auf das Wesen der Schönheit,
auf den Begriff derselben’. Nehamas (b) did much to dispel this approach but then himself
defended an incredible reading of how Hippias understands the ti esti question; and he refused to


 Plato’s Essentialism
moment of an original Geach in Plato’s own backyard, through whom
Plato raises some serious doubts about his own distinctive mode of
enquiry. For if such questions as ‘What is beauty?’ can be answered by
example and exemplar, then their answer is evident to us and we need not
engage in difficult and demanding enquiry in search of an answer. Perhaps
the suspicion that the Hippias Major is not by Plato, or at any rate that it is
not a dialogue with the potential of making a real difference to our
understanding of Plato, has contributed to this oversight. But if there is
an argument in the dialogue that has this potential, then this is good
reason for dismissing the suspicion of inauthenticity.
Hippias is originally characterised as a man of immense commercial
nous, but he is also of a philosophical bent of a deflationary variety. He is
not impressed by Socrates’ challenge to state on what grounds (pothen,
c) he is able to tell what things are such as to be beautiful; as,
apparently, he must be able to do since he is a master of the crafting of
beautiful things (speeches especially) and of teaching others beautiful
pursuits. On the contrary, he argues, in a subtle way worthy of any
latter-day deflationist, that the ti esti question as formulated by Socrates
(Q, ‘What is this very thing, beauty?’, ti esti to kalon;) can be answered by
answering the plain question, Q, ‘What things are beautiful?’ (ti esti
kalon; See d–e) in a certain way. For, the thinking goes, if a thing
is so conspicuously beautiful as to be capable of serving as a standard for a
thing’s being beautiful, then it is, indeed, a beautiful thing; and other
things will be beautiful on account of it, that is, on account of their
conforming to the standard that it embodies. Hippias puts this by saying
that ‘it makes no difference’ (ouden diapherei, d) whether one asks Q
or Q. What he means is not that there is no difference in meaning
between these two interrogative sentences (‘ti esti to kalon;’ versus ‘ti esti
kalon;’) but that one can answer Q by, precisely, answering Q, and so
any difference between them ‘makes no difference’.
admit the crucial point, which is that Hippias defends answering this question by example and
exemplar and that this is the target of Socrates’ argument.

Kahn () is perhaps the main living exponent of the denial of authenticity. For a comprehensive
summary of the history of the question of authenticity, see Ludlam (,  ff.). For a recent
defence of authenticity, see Petrucci (). For the question of the relative dating of the Hippias
Major, see Malcolm (); Lee (a, , n. ).

Note the di’ ho (‘because of which’) at a, which refers to the girl as an exemplar and a standard
of beauty; and note further that this di’ ho picks up on the explicative datives in the explanations of
the form ‘It is on account of the F that things that are F are F’ at b–d. I am taking it that these
explanations are intended to mean: ‘It is on account of the standard for a thing’s being F that things
that are F are F.’

The ti esti question 
Hippias’ philosophical acumen is demonstrated again when, on
Socrates’ prompting, he offers not one example and exemplar of a beautiful
thing, but several. Socrates did not ask ‘What is beauty, in regard to things
such as humans?’, he simply asked ‘What is beauty?’ – meaning, in effect,
‘What is beauty, in regard to each and every thing?’ It is, therefore,
appropriate that, once Hippias has proposed the girl, Socrates should ask
him about beautiful things of other kinds, such as non-human animals or
artefacts. In this way, Plato impresses on us the question whether this
example and exemplar, the beautiful girl, is suitable also as a standard for
the beauty of things of other kinds. Hippias recognises this question and
its force, and his response is to accept different examples and exemplars for
different kinds of cases.
I propose that we analyse Hippias’ defence of definition by example and
exemplar in terms of the argued commitment to three propositions:
P Definition by example and exemplar It is possible to give an account of
what a quality, F, is by appeal to a particular thing that is F, if this
thing is chosen for its suitability as an example and exemplar of a
thing that is F and, therefore, as a standard for a thing’s being F.
P The one-standard-for-many-cases requirement of definition A particular
thing that is F, in so far as it is used as a standard for a thing’s being F, is
suitable for determining of a plurality of things whether or not they are F.
P The rejection of the one-standard-for-all-cases requirement of definition
It is not the case that there is some one thing that is F and is suitable
for determining of all things whether or not they are F.
These propositions do not appear to be independent of each other. P
clearly follows from P. And it is arguable that P is intended to be based on
P: it is because he wants to give an account of what this quality, beauty, is,
by example and exemplar that Hippias thinks that one may not, and certainly
one need not, suppose that a single example and exemplar will be suitable for
determining all cases. This is important, because it suggests that what Plato
is doing through Hippias is, precisely, articulating an account of definition
by example and exemplar, in regard to certain qualities.
So far, it is Hippias who has been in the driver’s seat, providing an
apparently cogent account of definition by example and exemplar. But
things are about to turn. Hippias commits himself to a further proposition,
when, on Socrates’ prompting, he says the following:
This utensil also [the pot], it is true, is beautiful, supposing it has been
beautifully crafted; however, in general this [the pot] is not worthy (ouk

 Plato’s Essentialism
axion) of being considered as being something beautiful when compared
with the horse, the girl and all those other beautiful things. (e–, trans.
Woodruff, adapted)
In this, the ouk axion statement, Hippias accepts not only that the pot can
be beautiful but, by implication from, and by parity of reasoning based on,
his earlier use of the girl, the horse and the lyre, that it too can serve as a
standard of beauty for a yet further kind of thing, such as artefacts of
quotidian utility.
Soon after that ouk axion statement, Hippias gives up answering this ti
esti question by example and exemplar and proposes an answer that,
however absurd it may otherwise be, stands out for its generality and
unity: he proposes that it is on account of being golden that any thing
that is beautiful is beautiful. And the last lines before this U-turn have
Socrates ask:
Does it still really seem to you that the beautiful itself, i.e. that on account
of which all the other things are adorned and come to appear beautiful
when that character comes to be present, is a girl or a horse or a lyre?
(d–, trans. Woodruff, adapted)
Decisively for Plato’s argument against the account of definition by
example and exemplar, Hippias commits himself, in the ouk axion state-
ment, to comparisons in beauty not only between the girl, horse, lyre and
pot on the one hand, and any everyday object on the other, but also
between the girl, horse and lyre on the one hand, and the pot on the other.
He commits himself to comparisons in beauty between the examples and
exemplars of beauty themselves, however different they may be.
What is distinctive about this commitment is not the belief, which is
surely indisputably true, that it is possible to compare different things (e.g.,
two humans) in respect of beauty. Nor is it the belief, which on a
moment’s reflection is no less indisputably true, that it is possible to
compare different things in respect of beauty all of which can function
as examples and exemplars of beauty (e.g., the girl and, let us imagine, her
brother who is also of exemplary beauty). What is distinctive is the belief
that it is possible to compare different things in respect of beauty that can
function as examples and exemplars of beauty, however different these
examples and exemplars may be.
We may, now, formulate the final Hippian proposition:
P The possibility of any horizontal comparisons in F It is possible to make
comparisons in F between the examples and exemplars of
F themselves, however different they may be.

The ti esti question 
Now it will, I suspect, be evident that these four propositions, P–, make
up an inconsistent set. According to P, it is not the case that there is some
one thing that is F and is suitable for determining of all things whether or
not they are F. But P, in conjunction with P, that is, with the commitment
to definition by example and exemplar, requires that there be such a thing.
For, if there is not such a thing, then there is not a thing to serve as an
example and exemplar, and in this way as a standard, for making the
comparison in F between the examples and exemplars of F themselves, if
these are as different as can be and cumulatively serve to determine the
beauty or otherwise of all things.
If we suppose that propositions P and P are based on proposition P,
then we may conclude that the inconsistency is, at bottom, between, on
the one hand, the commitment to the possibility of giving an account of
what beauty is by example and exemplar, and, on the other hand, the
belief in the possibility of making comparisons in beauty between
the examples and exemplars of beauty themselves, however different
they may be.
We are left with one or two questions. Does Plato’s argument against
the possibility of definition by example and exemplar of a quality such as
beauty – in response to an incisive and cogent defence of defining this
quality by example and exemplar – generalise to all qualities? Plato’s
argument depends on two propositions, P and P, in regard to which it
does not seem true to say that they generalise to all qualities. In regard to
P, it does not seem true to say that there is not some one red thing that is
suitable for determining of each and every thing whether or not it is red. In
regard to P (and with the possible exception of things such as my worn
woollen jumper, once bright red but now quite faint), it does not seem
true to say that some red things are not worthy of being considered red
when compared to some others.
Does Plato recognise that his argument, in regard to this quality, beauty,
may not generalise to all qualities? There is one passage in the corpus that
suggests that he does. This is Parmenides d–, when Socrates says of
certain things, such as hair, mud and dirt, that ‘these things are, just as we
see them to be’ (tauta men ge haper horōmen, tauta kai einai). For it is hard
to see how, if a quality is characterised in this way, we can know what
quality this is, if not simply by pointing to an exemplary case of it. But
then we may expect that Plato does not, on pain of plain inconsistency,
intend the Hippias Major argument to generalise to such qualities: he

 Plato’s Essentialism
allows that, for certain qualities, definition by example and exemplar may
be possible.
Finally, why cannot Hippias, or one who wants to defend definition by
example and exemplar along such promising lines, avoid the inconsistency
to which Socrates drives him, simply by abandoning P? This would be a
way out, if the principal purpose of Plato’s argument were to determine
what an interlocutor has to do to ensure consistency among her beliefs.
But we should not suppose that this is Plato’s principal purpose. The
purpose includes inviting the reader to reflect on the content of those
beliefs, or propositions, and decide whether they appear true to her, the
reader. It is true that my Miele is a beautiful vacuum cleaner compared to
most, and certainly it is fit to serve as an example and exemplar of a fine
household appliance. But it is, I contend, hardly worthy of being consid-
ered a beautiful thing, compared to my daughter, my son, my dog Bobo,
or the Stradivarius and Shostakovich to which I am listening.

Commenting on this passage from the Parmenides, Harte (, ) says: ‘Forms are not needed in
those cases where things are “just as we see them to be”.’ Republic X. a– is sometimes read as
implying that there is a Form for every quality. For why this reading is not necessary, see Harte
(, ), when she says that Smith () ‘proposed that the passage should be construed,
rather, as making the claim that we commonly assume, “[as a rule of procedure,] that the Idea which
corresponds to a group of particulars, each to each, is always one, in which case we call the group of
particulars by the same name as the [Form].” On this construal, the passage does not carry any
implication about the scope of Forms’ (square brackets are Harte’s).

 
Why cannot essences, or Forms, be perceived

by the senses?
Hippias Major. Phaedo. Republic
. Problems with the standard answer, and in defence

of a very different answer
I suppose a standard answer, with reference to several dialogues and
especially Phaedo and Republic, is that a Platonic Form cannot be perceived
by the senses because such Forms are changeless, non-physical things, and
only physical things, things in space and time and subject to physical
change, can be perceived by the senses. To this it is commonly added
that Plato’s Forms can only be known a priori and on the basis of a
knowledge that is not dependent on the senses. This is because critics
commonly suppose that Plato’s distinction and opposition between objects
of the senses (aisthēta) and objects of the intellect and of reasoning (noēta),
which he first makes in the Phaedo (a), repeats in the Republic
(VI. a–b) and repeats again in the Timaeus (e–a), is equivalent
to the modern distinction between things known a posteriori and things
known a priori.
One need not be satisfied with this answer. One may worry that it is not
specifically about Forms; for the same could be said of our knowledge of
any thing that one, or a Platonist, takes to be non-physical, such as
numbers or God: we cannot perceive any such thing by the senses and
we can only know it a priori. Or one may worry that Plato only says of the
Forms of certain qualities, such as the Forms of beauty, equality, oneness,
that they cannot be perceived by the senses; and it is difficult to see how
the standard answer can account for the scope-sensitivity of Plato’s claim.
One may also worry that the standard answer has the order of explanation
and justification the wrong way round: it is not because they are non-
physical things that Forms cannot be perceived by the senses; for the claim
that they cannot be perceived by the senses is, rather, part of the reasoning,

Examples are aplenty, a recent exemplary one being Wedgwood ().


 Plato’s Essentialism
in such dialogues as Phaedo and Republic, for thinking that Forms are non-
physical things.
There is, I believe, a most particular and serious problem with the
standard answer: it ignores, or else gets the wrong way round, the relation
between Plato’s claim, in such dialogues as Phaedo, Republic and Timaeus,
that the Forms of certain qualities cannot be perceived by the senses, and
his claim, in such dialogues as Euthyphro and Hippias Major, that what
certain qualities are cannot be specified by example and exemplar. I want
to argue that it is in this relation, if anywhere, that we find Plato’s reason
for thinking that the Forms of certain qualities cannot be perceived by
the senses.
For suppose that what a certain quality is, such as, for instance, beauty,
can be specified by example and exemplar, that is, it is possible to specify
what this quality, beauty, is by pointing to a particular thing, such as, for
instance, a particular girl, or horse or lyre, that is evidently beautiful and,
for this reason, capable of functioning as an exemplar of a beautiful thing
and as a standard (paradeigma) for a thing’s being beautiful. It follows that
what this quality, beauty, is can be perceived by the senses. For if a thing’s
being beautiful is evident, and if this being evident involves pointing to
that thing, then we can directly perceive (see, or hear, etc.) that it is
beautiful: its beauty is conspicuous to the senses. And if this – that a
thing’s having a certain quality is evident and conspicuous to the senses –
allows that thing to function as an exemplar and as a standard for a thing’s
having this quality, then it is evident and conspicuous to the senses what
this quality is. What this quality is can, therefore, be perceived by
the senses.
And suppose, conversely, that what a certain quality is, such as, for
instance, beauty, can be perceived by the senses. It follows that what this
quality is can be specified by example and exemplar. What is perceived by
a sense, such as sight, here, we are supposing, is what beauty is. However,
sight cannot perceive what beauty is, except by perceiving a particular
beautiful thing; for what sight and the senses, in general, perceive are
particular things and their qualities. Indeed, it is hard to think of another
way in which one can perceive what a quality is by perceiving a particular
thing that has that quality, except by perceiving that particular thing’s
having that quality and using the thing as an exemplar and a standard for a
thing’s having that quality.
This reasoning (in the last two paragraphs) may be a little dense, but
I hope it is clear and compelling. What it demonstrates is the following
equivalences:

The sensory imperceptibility of Forms 
It is possible to perceive by the senses what a certain quality, F, is
if, and only if,
it is possible to specify what this quality, F, is by example and exemplar.
Which is to say that:
It is not possible to perceive by the senses what a certain quality, F, is
if, and only if,
it is not possible to specify what this quality, F, is by example and exemplar.
I propose that these equivalences demonstrate why Plato thinks that it is
not possible to perceive by the senses what a certain quality, F, is. For
Plato, it is because what a certain quality, F, is cannot be specified by
example and exemplar that the Form of this quality cannot be perceived by
the senses. This is what I shall argue in this chapter.
What more could we wish for, if we want to know why Plato thinks that
the Form of a certain quality cannot be perceived by the senses? The
account goes to the heart of the matter, for it is based on what a
Platonic Form basically is: a Platonic Form of a quality, F, is the essence
of that quality, F, in the sense of that which is designated by an
adequate and true answer to the question ‘What is F?’ And Plato thinks,
for particular reasons, that, for certain qualities, F, an answer to the
question ‘What is F?’ that is by example and exemplar is not an
adequate answer.
It is true, one could write a whole book about why this is the right way
of thinking of Plato’s Forms, namely, as being basically essences thus
understood. This would be no small task; and it would involve, among
other things, showing that it is wrong to think of this – that Forms are
essences – as simply another characteristic of Plato’s Forms, on a par
with their several other characteristics: Forms are changeless, uniform,
not perceptible by the senses, knowable only by reasoning, the basis of
causation and explanation, distinct from sense-perceptible things,
necessary for thought and speech, separate from physical things and
more. The task would involve showing that while the Forms have
many distinctive characteristics, just this is revealing of what a
Platonic Form basically is: it is that which is designated by an adequate
and true answer to the question ‘What is F?’ This is the task of the
present study.
Let me spell out more fully the reasoning for thinking that this is why
Plato thinks that Forms cannot be perceived by the senses: because he
thinks that the qualities of which they are the Forms cannot be defined by
example and exemplar. Consider the following three propositions:

 Plato’s Essentialism
PROPOSITION-I What a Platonic Form basically is is that which is
designated by an adequate and true answer to the
question ‘What is F?’
PROPOSITION-II Plato thinks, for particular reasons, that, for certain
qualities, F, an answer to the question ‘What is F?’
that is by example and exemplar is not an
adequate answer.
PROPOSITION-III The Forms of these qualities cannot be perceived
by the senses.
I submit that, if these propositions are true, then we have all we need, and
all we could wish for, if we want to know why Plato thinks that the Forms
of certain qualities cannot be perceived by the senses. PROPOSITION-I
makes sure we start where we should: in an understanding of what a
Platonic Form is. PROPOSITION-III, we have seen, is a readily
recognisable consequence of PROPOSITIONS-I and II. Moreover,
PROPOSITION-II is the minimal and least committal premise that we
can think of in addition to PROPOSITION-I. This is because, if the
question ‘What is F?’ can be answered by example and exemplar, then this
question (as Plato reminds us through the character of Hippias in the
Hippias Major e–) is trivial and worth practically nothing, because it
is so easy to answer: answers are ready to hand and available without the
least effort and without any enquiry to speak of, which means that any
further theorising about essences and Forms, including the question
whether essences and Forms are physical or non-physical things, is unnec-
essary and vain. QED.
What I am arguing is that what we need, in order to determine why
Plato thinks that the Forms of certain qualities cannot be perceived by
the senses, is, basically, to observe that he thinks that, for certain
qualities, what a quality is cannot be specified by example and exemplar.
And I am urging that it follows from this that it is wrong to think that
Plato thinks that Forms cannot be perceived by the senses because they
are changeless, non-physical things, and only physical things can be
perceived by the senses, and non-physical things can only be known
a priori.
This is not to deny that, once he has argued that there are Forms, and
that they cannot be perceived by the senses, and that they are distinct from
corporeal and physical things, and once he has supposed that something is
sense-perceptible if, and only if, it is corporeal and physical, at that point
Plato may go on to argue that Forms are not sense-perceptible because they

The sensory imperceptibility of Forms 
are not corporeal or physical. Such argument, however, will be acceptable
only to one who has, for different and more radical reasons, already
accepted that essences and Forms cannot be perceived by the senses; it
will not justify, or adequately explain, the basic claim that essences and
Forms cannot be perceived by the senses. Such argument, among such
advanced interlocutors and addressed to such an advanced reader, can,
I think, be found in the late dialogue Timaeus (and in Philebus). There the
investigation proper begins with the division, introduced abruptly and
without preparation, between changeless things known by the intellect
and reasoning – they will soon emerge as being, or at any rate including,
Forms and a world of Forms – and changing, generated things (gignomena)
known by the senses and judgement involving the senses (doxa met’
aisthēseōs). A little later in the Timaeus (at b; anticipated at b), it is
argued that anything that is changing and generated (ginomenon) is both
corporeal (sōmatoeides) and visible-tangible (horaton hapton), and that it is
visible-tangible by having been made corporeal in a certain way (namely,
out of fire and earth). It is hard to see how anyone could properly
understand the distinction at Timaeus e–a, much less why it is
acceptable, if she were not familiar with Phaedo or Republic. And it
appears that Plato indicates that this is the intended reader, when, in the
extended lead-up to e–a, he repeatedly marks the occasion of the
present discussion as taking place the day after a discussion which, in
content if not in regard to the particular occasion in which it took place, is
manifestly that of the Republic.
It is elsewhere, and going back to Plato’s original introduction of the
distinction between objects of the senses and objects of the intellect and
reasoning (it is introduced in Phaedo a), and to his argument in
preparation for this distinction (I shall argue that the Hippias Major is
the source here), that we must look for Plato’s argument for the conclusion
that essences and Forms cannot be perceived by the senses.
. The origin of this answer in the Hippias Major

To recognise that the standard answer is wrong, in response to the
question why Plato thinks Forms cannot be perceived by the senses, we

The reader may wonder how Plato can provide both an ‘epistemological’ and a ‘metaphysical’
argument for the same conclusion. I answer this in Chapter , ‘What yokes together mind
and world?’

This is one reason why I cannot agree with Broadie () when she argues that we can, and should,
study the Timaeus on its own and separately from other dialogues of Plato.

 Plato’s Essentialism
need to consider why, and for what particular reasons, Plato thinks that,
for certain qualities, what that quality is cannot be specified by example
and exemplar. For, unless we consider this, we cannot rule out that, if he
uses the crucial equivalence (It is not possible to perceive by the senses what a
certain quality, F, is if, and only if, it is not possible to specify what this
quality, F, is by example and exemplar) in either of two explanatory and
justificatory directions (i.e., from left to right, or, on the contrary, from
right to left), he uses it to argue that it is not possible to specify what
beauty is by example and exemplar because it is not possible to perceive by
the senses what beauty is. And if it is in this explanatory and justificatory
direction that he uses the equivalence, the standard answer may still be
invoked, to determine why, in turn, Plato thinks that it is not possible to
perceive by the senses what beauty is. I believe this is to get the explanatory
and justificatory direction behind the crucial equivalence the wrong way
round; but showing this requires showing that Plato has particular reasons
for thinking that, for certain qualities, what a quality is cannot be specified
by example and exemplar, and that these reasons do not include any of the
following claims: the Forms of these qualities are changeless, non-physical
things; only physical things can be perceived by the senses; non-physical
things can only be known a priori.
In the Hippias Major (–), Plato argues that what beauty is
cannot be specified by example and exemplar. On the analysis I have
defended in Chapter , the nub of the argument is that, if what beauty is is
specified by example and exemplar, then we cannot make unrestricted
comparisons in beauty, that is, comparisons in beauty between things,
however different they may be; but we may well want to make unrestricted
comparisons in beauty. More fully spelled out, and suitably generalised,
the argument is that, for certain qualities, the commitment to specifying
what a quality, F, is by example and exemplar implies a commitment to
the following three propositions:
P Definition by example and exemplar It is possible to give an account
of what a quality, F, is by appeal to a particular thing that is
F, if this thing is chosen for its suitability as an example and
exemplar of a thing that is F and, therefore, as a standard for a thing’s
being F.
P The one-standard-for-many-cases requirement of definition A particular
thing that is F, in so far as it is used as a standard for a thing’s being F,
is suitable for determining of a plurality of things whether or not
they are F.

The sensory imperceptibility of Forms 
P The rejection of the one-standard-for-all-cases requirement of definition
It is not the case that there is some one thing that is F and is suitable
for determining of all things whether or not they are F.
But these propositions are incompatible with the following proposition,
to which we may find ourselves committed in regard to a certain
quality, F:
P The possibility of any horizontal comparison in F It is possible to
make comparisons in F between the examples and exemplars of
F themselves, however different they may be.
The lesson of Plato’s argument is that, if we find that, in regard to a certain
quality, F, such as beauty, we want to make unrestricted comparisons in
F (i.e., P), then we must recognise that we cannot specify what this
quality is by example and exemplar.
It is apparent that Plato’s argument in the Hippias Major, for the claim
that what beauty is cannot be specified by example and exemplar, does not
rely on any of the claims distinctive of the standard answer to why Forms
cannot be perceived by the senses. We may conclude that, if indeed Plato
uses the crucial equivalence (It is not possible to perceive by the senses what a
certain quality, F, is if, and only if, it is not possible to specify what this
quality, F, is by example and exemplar) in either of the two explanatory and
justificatory directions, he uses it to argue that it is not possible to perceive
by the senses what beauty is, because it is not possible to specify what
beauty is by example and exemplar.
If only there were a passage in which Plato said, in so many words, that
this is why what a certain quality is cannot be perceived by the senses:
because what this quality is cannot be specified by example and exemplar.
It seems to me that, once we know what we are looking for, the argument
in the Hippias Major (–) for the claim that what beauty is cannot
be specified by example and exemplar, and certain notable passages fol-
lowing this argument, can be understood as amounting to such a state-
ment. Hippias intends his distinctive answer to Socrates’ question, ‘What
is beauty?’, the answer by reference to a particular girl (later also a horse
and a lyre), as showing that this question is ‘a trifle’ (smikron) and ‘worth
practically nothing’ (oudenos axion hōs epos eipein, e–). It makes good
sense that he should intend this, if we suppose that he thinks that the
exemplary beauty of these things is evident and evident to the senses;
because, then, no particular effort and no search to speak of is required to
answer the ti esti question, only acquaintance with beautiful things. We

 Plato’s Essentialism
may expect that Plato intends it to be obvious to the reader that he, Plato,
thinks that this question, ‘What is beauty?’, is anything but a trifle or easy
to answer; rather, answering it requires genuine and demanding search.
That Plato intends this is confirmed later in the dialogue (at b), when
Socrates says, with sticky irony, that, even though he is sure Hippias will
be able to discover (ekseurein) what beauty is easily enough on his own and
by himself, when in his, Socrates’, company, they will have to search
together (suzētein) and to be prepared just as much for failure as for success
in this search.
That, for Hippias, the girl, the horse and the lyre of exemplary beauty
can function as examples and exemplars of beautiful things, precisely
because their beauty is evident and evident to the senses, is confirmed by
the prominent use of the verb phainesthai, both in the argument at
– and, repeatedly, later in the dialogue, and the use of this verb
in a sense associated with what is evident and evident to the senses. In the
argument at –, it is notable that when Socrates quotes Heraclitus as
saying that ‘the most beautiful of monkeys is ugly when set next to
humankind’ (a–), he immediately goes on to formulate this point
(or rather, a more general version of it) in terms of how the beautiful things
of one kind will appear (phaneitai, b and b) when set next to the
beautiful things of another kind. That, yet again, the conclusion is stated
in terms of how the beautiful things of one kind are when set next to the
beautiful things of another kind (Socrates goes on to say that they ‘may
admit that the most beautiful girl is ugly in relation to the family of the
gods’, b–) suggests that these claims, about how certain things are,
are intended to be based on how the same things appear, in a sense of
phainesthai associated with what is evident and evident to the senses.
I mean that they are so intended by Hippias, for it is he who thinks that
what beauty is and what things are beautiful is evident and evident to the
senses; and Socrates’ argument and dialectical intercourse with Hippias is
premised on Hippias’ premises. We cannot infer that this is what Plato
thinks, or that Socrates shares these premises; indeed, it is clear that
Socrates is not presented as sharing these premises, but, if anything, as
rejecting them.

Contra Rickless, who appeals to this passage in the Hippias Major, a–b, to argue that: ‘Socrates
moves to the claim that sensible F things are con-F (that is, partake of both the F and the con-F)
from the claim that sensible F things appear to be con-F. Since the Greek word for “appear” in this
context is ambiguous as between a veridical and a non-veridical sense, Socrates’ move from the one
claim to the other strongly suggests that his use of “appear” here is veridical’ (, , n. ,
original emphasis). It is clear that Rickless intends this to be true not only of Socrates in this passage,

The sensory imperceptibility of Forms 
This theme and this issue, that is, of the distinction between, and at the
same time the relation between, a thing’s directly appearing beautiful and
its being beautiful, is not only taken up here (at b), it is kept in play as
a major theme in the extended argument that follows: starting with the
question whether being golden at all times and in all circumstances
contributes to a thing’s appearing beautiful, that is, appearing so to the
senses and especially sight (see b, phainesthai, and d, phainesthai;
also e–c). It remains in play right up to Hippias’ confident
statement that there is a single thing, ‘that which is fitting’ (to prepon),
whose presence makes a thing both appear (phainesthai) and be (einai)
beautiful, and does this in such a way that it is not possible (adunaton) for
these two things, that is, a thing’s appearing beautiful and its being
beautiful, to come apart (c–).
We may conclude that this is Hippias’ aim here (at c f.): to
demonstrate that, for certain qualities, such as beauty, it is possible to give
an account of what one of those qualities, F, is, by exhibiting for all to see
(or hear, or feel, etc.) a thing that is evidently F, that is, evident to the senses.
He thinks that it is because we can perceive by the senses what beauty is
that we can specify what beauty is by example and exemplar. Socrates is
presented, by the author, Plato, as successfully refuting Hippias’ claim that
what beauty is can be specified by example and exemplar. We may
conclude, finally and definitively, that, in the Hippias Major, Socrates
is presented as demonstrating all the following: first, both that what
beauty is cannot be specified by example and exemplar and that what
beauty is cannot be perceived by the senses; second, that the two claims –
that what beauty is can be specified by example and exemplar and that
what beauty is can be perceived by the senses – stand and fall together;
and, finally, that it is because what beauty is cannot be specified by
example and exemplar that what beauty is cannot be perceived by
the senses.
. This is what is behind the claim in the Phaedo that essences and
Forms cannot be perceived by the senses
It is familiar that it is in the Phaedo that Plato states, loud and clear, that
what certain qualities are, their essences and Forms, cannot be perceived by
but of Plato generally. For he is arguing that Plato’s (so-called) argument from co-presence of
opposites is not limited to how physical things appear, but is about how physical things are.
I consider this issue in Chapter .

 Plato’s Essentialism
the senses. In an extended passage early in the dialogue (e–c, see esp.
a–a; picked up again later, esp. at a), Socrates argues that the
philosopher, that is, he whose supreme aim is knowledge and wisdom,
will, as far as is possible, avoid relying on sense-perception, and will, rather,
rely on thinking, reasoning and what the soul does itself by itself and not in
communion with the senses. In several passages, both here (see esp.
d–e) and later in the dialogue, Plato indicates what is the knowledge
that the philosopher, so understood, is especially aiming at, namely, the
knowledge of essences and Forms:
I am referring to all those things, such as largeness and health and strength
and, in one word, about the essence of all those things: that which each of
them is (kai tōn allōn, heni logō[i], hapantōn tēs ousias ho tugchanei hekaston
on). (d–e)
Other notable and familiar passages include:
For our present argument is no more about the equal than it is about the
beautiful itself and the good itself and the just and the pious, I mean about
all the things to which we attach this stamp – ‘that which a thing is’ (‘ho
esti’) – both in our questions when we are posing questions and in our
answers when we are proposing answers. (c–d)
This essence [or, ‘being’], of whose being we give an account (hautē hē ousia
hēs logon didomen tou einai) through the asking of questions and the
proposing of answers. (d–)
It is evident, from such passages as these, that the knowledge that Plato has
especially in mind, when he has Socrates say that the philosopher will, as
far as possible, avoid relying on sense-perception, and will, rather, rely on
thinking, reasoning and what the soul does itself by itself and not in

I consider this more closely in Chapter .

My reason for translating ‘that which a thing is’, and not ‘that which is’, for ho esti here is due to the
reference to Socrates’ practice of asking questions and receiving answers, since this is a practice
consisting of, or centrally involving, the asking of ti esti questions.

Vlastos (, ) comments on these lines as follows: ‘For Plato the definition of a concept is “the
account of the essence” (logos tēs ousias) of its Form, Φ’. Therefore, according to Vlastos, that which is
defined is not the Form, but the essence of the Form. Vlastos defends his statement by appeal to this
passage, Phaedo d, in his translation of it: ‘That reality itself, of whose essence we give the account
[hēs logon didomen tou einai] when we ask and answer our questions . . .’. I beg to differ. The genitive
here (tou einai) can, and I think should, be read in such a way that it is precisely the ousia that we give
an account of when we give an account of what it is (tou einai). This is most important, because, on
Vlastos’ translation and commentary, this passage confirms the common view, which is indeed
Vlastos’ view, that Forms are substances that have essences, rather than simply being essences. This
passage in no way supports this view.

The sensory imperceptibility of Forms 
communion with the senses, is the knowledge the search for which he has
exhibited in other dialogues, and dialogues commonly supposed to be
before the Phaedo, and is exhibiting again here. It is the knowledge aimed
at by a joint enquiry among a small group of people, addressed to a ti esti
question (such as ‘What is death?’ and ‘What is it to die?’, see c–), or a
ti esti question and an associated whether-or-not question (such as whether
or not it is possible for the soul to exist and to think without the body). We
may note that most of the specific qualities whose essence and Form is
mentioned at c–d, and mentioned in association with the ti esti
question in regard to these qualities, correspond to ti esti questions that
have been occupying Plato in other dialogues: ‘What is beauty?’, in Hippias
Major; ‘What is virtue?’, in Protagoras and Meno; ‘What is piety?’, in
Euthyphro.
The remarkable thing is that, as it is introduced in the Phaedo, the claim
that the essence and Form of these qualities cannot be perceived by the
senses appears unmotivated and unprepared for – as if it came from
nowhere. Socrates, abruptly, asks Simmias whether he has ever seen any
such things (toioutōn, i.e., essences and Forms) with his eyes (tois ophthal-
mois, d), which Simmias immediately denies – and that is pretty much
the end of that. It is also remarkable that what Socrates asks is whether
Simmias has ever seen any essence or Form with his eyes, not whether
anyone has. I think this is significant, and it is picked up again at c–,
when Socrates asks, not whether the Form of the quality equal has appeared
to anyone to be both equal and unequal, but whether it has so appeared to
Simmias. The reason for this, I shall argue later, is that Simmias is special,
in that he has understood that, and why, essences and Forms cannot be
perceived by the senses. Had the interlocutor been Hippias, he would no
doubt have insisted that he has often and without difficulty seen what
these things or qualities are.

See Taylor (, –): ‘It is assumed [in Phaedo] that the technical language of the theory of
Forms is so familiar a thing that Socrates needs to warn the lads not to be misled by it: an odd
representation if the whole theory has been invented by Plato after Socrates’ death.’ See also Rowe
(, ): ‘[Plato] writes, in the Phaedo and the Republic, as if “forms” are already a familiar topic,
so in any case projecting them back into the pre-Republic dialogues’ (he explicates the ‘in any case’
with ‘That is, if they were not there already’, note ). And see esp. (, ), where Rowe says:
‘So far as the Phaedo is concerned, the forms are those things, whatever they are, that the
philosopher seeks to “grasp” through dialectic; the very things that Socrates is seeking to grasp in
the Euthyphro, the Laches, the Meno . . .. That is one reason why Plato can have him, in the Phaedo,
describe “beautiful and good and all such being (ousia)” (D–) as “the things we’re always
chattering about”, and then later refer to them just as “those much-chattered-about (poluthrēleta)
things” (B–)’ (ellipsis in original).

See Chapter .

 Plato’s Essentialism
We have, therefore, two options. We may look closer in the Phaedo for
the justification of this claim: that Forms cannot be perceived by the
senses. Or we may suppose that Plato intends the claim and its justification
to be familiar to his audience, in the first instance from other dialogues.
I doubt that we can find in the Phaedo a defence of the claim. We may set
aside, as far too thin and as mentioned only in passing, such general
misgivings about sense-perception as are alluded to at b–, when it is
said that the poets keep harping on the apparent fact that nothing of what
we see or hear is distinct or clear. The only other potential locus for this
justification, that I can think of, is the famous passage, a–c, where it
is argued that the Form of the quality, equal, is distinct from, and not
identical with, any (pair of ) things that appear equal to the senses. This
passage, however, if it is read as containing the claim that the Form of
equal cannot be perceived by the senses, does not defend the claim;
rather, it uses it as one of the premises in an argument for the conclusion
that this Form is distinct from and not identical with any (pair of ) things
that appear equal to the senses.
The hypothesis recommends itself, therefore, that, in Phaedo, Plato
presents as familiar to his audience the claim that the Forms of certain
qualities cannot be perceived by the senses. I contend that we may accept
this hypothesis, if we recall that we may indeed expect it to be familiar to
his audience that Plato thinks that what certain qualities are cannot be
specified by example and exemplar. For we may recall the Euthyphro, when
Socrates castigates Euthyphro for wanting to answer the ti esti question
about piety by pointing to his particular action of, as he thinks, exemplary
piety (d–e), and, especially, as I have been at great pains to argue, the
Hippias Major, when it is argued that the view that essences and Forms can
be perceived by the senses stands and falls with the view that what certain
things or qualities are can be known by example and exemplar: Hippias
argues, Yes; Socrates, No.
. From beautiful to equal and one

To examine this hypothesis, we must ask whether we may expect that
Plato thinks that the argument in the Hippias Major, for the claim that

In Republic VII. b–, Socrates expressly points out that this is not the worry he had in mind.

I think it does contain it; see Chapter , where I consider this passage and the argument in it.

See Chapter .

This is to suppose that the Hippias Major is written before the Phaedo, but in relative proximity to
it. This supposition is defended by Malcolm .

The sensory imperceptibility of Forms 
what beauty is cannot be specified by example and exemplar, extends to
those qualities, such as equal, of which it is said in Phaedo that what they
are, their essences and Forms, cannot be perceived by the senses and
extends to qualities, such as one, of which the same is said in Republic
(see VII. e–a). Indeed, if we want to demonstrate that the
proposition that says that, for certain things or qualities, the ti esti question
cannot be answered by example and exemplar provides Plato’s reason for
thinking that Forms cannot be perceived by the senses, we must show,
further, that this proposition provides Plato’s reason for thinking that the
Form of any thing or quality for which he thinks there is a Form,
including, apparently, the Forms of such things or qualities as the follow-
ing, cannot be perceived by the senses: human and bed; desk; that part of
the instrument for weaving, the shuttle, mentioned in the Cratylus; and
indeed that whole instrument for weaving, the loom.
On the proposed extension of the Hippias Major argument, in the first
instance to such qualities as equal and one, the problem with specifying
what equal is by example and exemplar is not that there are no exemplary
examples of a pair of equal things; there are plenty such (i.e., P and P, in
our analysis of the Hippias Major argument, are acceptable). The problem
is that there is not an exemplary example of a pair of equal things that can
be used to determine of any pair of things whether they are equal (i.e., P
is acceptable). And this is a problem, if we want to make comparisons in
equality, and comparisons based on the appeal to an exemplary example,
between the exemplary examples themselves of a pair of equal things
however different they may be (i.e., P is acceptable). For, the possibility
of making such comparisons is incompatible with P–.
In regard to the quality, equal, we would, therefore, have to expect Plato
to argue as follows. Suppose we cut two sticks or two stones to equal size,
as carefully as we can, and use them as exemplars and as standards of
equality in, say, housebuilding or land surveying; and suppose we use a
certain recurring celestial pattern, visible to the trained eye, as an exemplar
and a standard of equality when, say, measuring space or time. We may
find that, on reflection, the latter standard is more accurate than the
former: two equal distances, spatial or temporal, measured by this recur-
ring celestial pattern, are more equal than two equal walls, or two equal

I have argued, in Chapter , that the argument does not extend, and is not intended by Plato to
extend, to all qualities.

It is notable that, in Republic VII. d–a, the issue is the use of visible celestial patterns as
exemplars and standards, paradeigmata, of, precisely, equalities and the adequacy of such sense-
perceptible exemplars and standards of equal (see esp. the isōn at a).

 Plato’s Essentialism
pieces of land, measured by the exemplary stones or sticks however
carefully cut. We may even say that, compared to those distances, these
walls or pieces of land are, or may be for all that we can tell by measuring
them in this way, unequal. However, if we try to find a single sense-
perceptible example and exemplar on which to base this comparison, as we
must if we think that what equal is can be specified by example and
exemplar, we find that we cannot find any.
In regard to the quality, one, we would have to expect Plato to argue as
follows. Suppose we pick out a particular finger, or indeed the human
being of which it is a part, and use it as an exemplar and a standard of
oneness for, say, living things; and suppose we pick out a certain geomet-
rical pattern, say that of a particular crystal, and use it as an exemplar and a
standard of oneness for, say, non-living things. We may find that, on
reflection, the latter standard is more accurate than the former: compared
to the crystal pattern, even the most perfect finger, or human being, is
deficient in oneness. However, if we try to find a single, sense-perceptible
example and exemplar on which to base this comparison, as we must if we
think that what one is can be specified by example and exemplar, we find
that we cannot find any.
I think it will be recognised that this pattern of argument is present in
Plato, and sufficiently recurrent to be a pattern of argument intended by
him as such, also in Phaedo and Republic. Candidate passages in the two
dialogues for exhibiting this pattern are Phaedo a–c; Republic I. e–
b, V. e–a, VII. a–a, and e–c. Critics commonly
refer to it as the argument from the co-presence of opposites, understood
in a distinctive way, that is, as concerning types rather than particulars.
And the same critics generally, if not invariably, associate Plato’s claim of
the co-presence of opposites, so understood, with his claim that it is wrong
to suppose that a standard for a thing’s being F can be opposite-to-F.
We may conclude that, according to Plato, the Hippias Major argument
for thinking that what the quality, beauty, is cannot be specified by
example and exemplar extends to such qualities as equal and one and that
Plato’s view that the argument extends in this way is not unreasonable.

This understanding is summed up succinctly by Verity Harte (, ): ‘Is the claim (PC) that,
for any particular perceptible, having some relevant feature, F, necessarily, that particular
perceptible also has the opposite feature, un-F? Or is it the claim (UC) that, for any perceptible
type, a token of which is F, for some relevant feature, necessarily that type has un-F tokens also?’
I note that, as Harte formulates the claim of the co-presence of opposites, the notion of a standard
(paradeigma) is not part of that argument.

The sensory imperceptibility of Forms 
In my estimation, our understanding of the pattern of argument com-
monly referred to by critics as Plato’s argument from co-presence of
opposites is substantially improved if we trace it back to the argument in
the Hippias Major against specifying what certain qualities are by example
and exemplar. The Hippias Major argument, too, concludes that it is
wrong, or problematic, to suppose that a standard for a thing’s being
F can be both F and opposite-to-F. Socrates concludes by objecting,
against Hippias, that Hippias’ exemplar and standard of beauty, the girl,
may turn out to be ugly, when compared to, say, a goddess; just as the
infamous pot, even while being a perfectly good standard for the beauty of
quotidian utensils, has turned out to be ugly when compared to the girl,
the horse and the lyre.
Indeed, if we ignore Socrates’ argument in the Hippias Major (on the
analysis I have defended of it), it is not clear what, if anything, is wrong
with an exemplar and a standard for a thing’s being F being itself opposite-
to-F when compared to a different standard for a thing’s being F. It is
Socrates’ argument that shows what is wrong, or problematical, with this:
we now need a single standard to determine of any thing whether it is F or
not F (a one over all, not just a one over many); and definition by example
and exemplar does not provide us with such a standard; and we need this,
if we want to make unrestricted comparisons in F.
Likewise, it is not clear what, if anything, is wrong with supposing that
‘for any perceptible type, a token of which is F, for some relevant feature,
necessarily that type has un-F tokens also’ (Harte , ), which is a
common way in which critics state the claim of the co-presence of
opposites. Indeed, I do not see how we can maintain that there is anything
wrong with this, unless we understand the argument for the co-presence of
opposites as being, fundamentally and from the start, about standards
(paradeigmata) and about whether it is possible to provide a standard for
a thing’s having a certain quality, F, by example and exemplar, if we want
to make unrestricted comparisons in F.
There is yet another benefit of tracing this pattern of argument (we may
call it the problem of the apparent co-presence of opposites in the example and
exemplar) back to the argument against definition by example and exem-
plar in the Hippias Major. One remarkable feature of this pattern of
argument, in such passages as Phaedo a–c and Republic V. e–a,
is that Plato consistently, if not exclusively, formulates the problem as
being that sense-perceptible standards appear (phainontai) no more F than
opposite-to-F; and it is arguable that by ‘to appear’ (phainesthai) here, he
means ‘appear to the senses’. Critics have not always recognised this, or its

 Plato’s Essentialism
significance; they have commonly spoken of co-presence of opposites,
when, it seems, what Plato is concerned with is apparent co-presence of
opposites, that is, apparent to the senses. Taking our starting point in the
Hippias Major argument shows why Plato is concerned, especially, with
apparent co-presence of opposites: because the common appeal, by such
people as Hippias and (in Republic) the lovers of sounds and sights, to a
sense-perceptible standard for a thing’s being F is motivated by the view
that what F is is apparent and apparent to the senses, that is, we can
directly perceive what F is.
. From beautiful, equal and one to human and bed

If I am not mistaken, the crucial equivalence, which says that It is possible
to perceive by the senses what a certain thing or quality, F, is if, and only if, it
is possible to specify what this thing or quality, F, is by example and exemplar,
holds good just as much of human and bed as it does of beauty, equality and
oneness. This is because the reasoning behind this equivalence (which I set
out in paragraphs four and five of this chapter) holds good just as much of
human and bed as it does of beauty, equality and oneness. I say this because
Plato commits himself to the Form of beds (in Republic X. b) and to
the Form of humans. He commits himself to the Form of humans in,
among other places, Philebus a, where he adds the Form of oxen; or
rather, what he commits himself to there is that there are henads (‘unities’,
from hen) of such things as humans and oxen. On the other hand,
notoriously, in Parmenides c he expresses doubts about whether there
are Forms of things such as humans, fire and water; or perhaps rather, he
expresses doubts about whether we should speak and think of the Forms of
such things in the same way as we think of the Forms of things such as
oneness, likeness and beauty, or in a different way (cf. potera chrē phanai
hōsper peri ekeinōn ē allōs). We need, therefore, to consider whether our
argument, for the claim that says that Plato thinks that Forms cannot be
perceived by the senses because he thinks that the qualities or things of
which they are the Forms cannot be defined by example and exemplar,
extends to the Forms of such things as humans and beds.
There is a problem with extending our argument from the Forms of
qualities such as beauty, equality and oneness to all Forms and specifically to
Forms of such things as humans and beds. For, if the crucial equivalence
can be used to explain why the essences, or Forms, of certain things or

A critic who does recognise this, and makes a particular point of it, is Kirwan ().

The sensory imperceptibility of Forms 
qualities cannot be perceived by the senses, namely, because these things or
qualities cannot be defined by example and exemplar, it can just as much
be used to explain why certain things or qualities cannot be defined by
example and exemplar, namely, because their essences and Forms cannot
be perceived by the senses, which, in turn, it will be said, is because these
essences and Forms are not physical entities and because they can only be
known a priori. In regard to the Forms of such qualities as beauty, equality
and oneness, we were able to answer this problem, by arguing that Plato’s
argument, in the Hippias Major, for the claim that beauty cannot be
defined by example and exemplar, does not rely on the view that the
essence and Form of this quality is not a physical entity, or that it can be
known only a priori; what it relies on is considerations about how ade-
quately to define this quality, especially if we want to make unrestricted
comparisons regarding it and the things that have it. Furthermore, we saw
that a Hippias Major–style argument can, without too much difficulty, be
extended to qualities such as equal and one, qualities the search for the
essence of which Plato is concerned with, to some degree, in Phaedo and
Republic and, even more, in Parmenides, Philebus and Timaeus; and we saw
that there is reason to think that Plato extends the argument along
such lines.
The problem is that there seems to be no immediate way of extending
this Hippias Major–style argument to things such as humans and beds and
their essences and Forms. For it is not immediately plausible to say that
certain humans are not worth considering human when compared to
others – the way in which Hippias insists that a beautiful pot is not worth
considering beautiful when compared to a beautiful girl, lyre and horse.
Even if we are prepared to sincerely say this, it is not immediately clear
why we should despair of finding a particular exemplary human that can
serve as a standard of whether any thing is such as to be human, and indeed
as a standard for determining the relative humanity of different humans in
their quality as humans. I shall not repeat all this regarding beds, but it is
worth noting straightaway that a bed is not just any odd artefact, but an
artefact essentially concerned with enabling and promoting the way in
which humans pass a significant part of their time, that is, asleep or resting,
and concerned with the quality of this so important time of life. I suspect
that the Greeks of Plato’s time, or those that could afford doing so, would
occupy themselves just as seriously with what makes a good bed, mattress
included, as do we.
To make too much of all this is, it seems to me, to ignore the
significance of the conclusion of our argument so far. For, even as it

 Plato’s Essentialism
stands, the conclusion says that Plato’s reason for thinking that most of his
Forms cannot be perceived by the senses is that, as he thinks and argues,
what these qualities are, their essence and Form, cannot be defined, and in
general specified and determined, by example and exemplar. When I say
most of his Forms, I mean: those that he typically and routinely mentions
already in Phaedo and Republic; those that, in Parmenides b–c he is
most sure about and those that are so highly prominent also in such
dialogues as Philebus and Timaeus. For these include, precisely, the
Forms of such qualities as one, equal, like, beautiful, just and good. If this
conclusion, and Plato’s argument for it, does not give his full reasoning for
all Forms, and if work is still needed to determine why he is also commit-
ted to Forms of such things as humans and beds, this is to be welcomed,
not feared, provided that it is consistent and coherent with the conclusion
of our argument so far.
If I may straightaway speak in general terms, there is no doubt that Plato
does not treat of such things as humans as being separate and independent
from such qualities as one, beautiful and good. On the contrary, both what
it is to be a human and what it is to be a good human depends, he thinks,
on what it is for a thing to be unified and one, which in turn is bound up
with what it is to be good. So much is familiar especially from the
dialogues Philebus and Timaeus. There are, it is true, those critics who
argue that, before the Parmenides and Sophist, and so also in Phaedo and
Republic, Plato’s theory of Forms is distinguished by the view that Forms
are logically independent of each other. I consider this reading, and argue
that it is false and ill-founded, elsewhere in the study (Chapter ). It is
worth recalling, straightaway, how, in Republic VII. –, immediately
after having brought up the question ‘What is a finger?’ and wondering
whether one, be one layman or philosopher, needs to raise this question,
he argues that we absolutely do need to raise the question ‘What is the
one?’ (ti pot’ esti auto to hen, e).
This holds out prospect of extending our argument and its conclusion
to things such as humans, water and fire along the following lines. Suppose
one thinks that, whatever about beauty, equality and oneness, it is possible
to specify what it is to be a human by pointing to an exemplary example of
a particular human. It is then pointed out to one that humans have parts,
such as fingers, and that we do not begin to know what a human is unless
we know its principal parts and how they add up to a single, unitary thing
like a human. Still, one may dig in and insists that what the unity of a

I consider this in Chapter .

The sensory imperceptibility of Forms 
human is can be specified by pointing to an example of a particular human
of exemplary human unity. It will then be pointed out to one that humans
come in different degrees of unity; some have supreme, or golden, unity,
others intermediate, or silver, unity, yet others a lesser, or bronze, unity.
Of course, one may resist this view and insist that Plato never intended the
division of humans into gold, silver and bronze to be taken this way
(Republic III. –; see Rowett ). Still, it will be pointed out,
and argued, we cannot really understand the unity of humans unless we
grasp that they are so much more unified, and their unity is so much more
complex and different, than the unity of, say, molluscs or, in general,
creatures that have no other occupation and pastime than to enjoy and
please themselves (Philebus c; compare Socrates’ discussion with
Callicles in Gorgias). Now, I hope it is plain, we can directly revert to
the Hippias Major–style argument that we have been at pains to spell out.
For, to make this comparison between the unity of humans and that of
molluscs, one will need a standard of unity that ranges over the unity of
both these kinds of thing. And, as the Hippias Major–style argument
shows, it is questionable whether such a standard of unity can be provided
by example and exemplar.
As to the proverbial bed of Republic X, I do not want to insist that a
Hippias Major–style argument can be applied, immediately and directly, in
regard to whether or not it is possible to define bed by example and
exemplar. Nor can we assume, from Plato’s examples there of bed and
desk, that he is committed to Forms of any and all artefacts, even the quite
useless ones with which our world is getting more full by the day – no
doubt some of these existed even in Plato’s time, trinkets and gadgets of all
sorts. The point is that beds and desks are artefacts made for the good of
humans; and for this reason they cannot be made any way one likes and is
able to sell, rather, to make them properly and well, one needs to under-
stand a lot about what humans beings are and what is good for them.
Finally, we may recall that, in Parmenides b–c, those things that
Plato has Socrates insist lack Forms, and of which Socrates says that ‘they
are just as we see them to be’, are things that conspicuously lack unity: dirt,
mud and hair – I would add nail clippings and in general the waste that is
clogging up our world by the day. These ‘accidental compounds’, as an
Aristotelian would call them, lack a Form, precisely because they lack a
distinctive unity. This, apparently, is why ‘they are just as we see them to
be’, which, we may conclude, entails that what they are may, unlike other

Or perhaps it is jellyfish; see O’Reilly ().

 Plato’s Essentialism
things, be capable of being specified by example and exemplar: here is
some exemplary dirt (pointing to the contents of a full vacuum cleaner).
. What about numbers? Does Plato’s argument extend to them?

Let us retrace our steps in this chapter. The Hippias Major argument
demonstrates that what the quality, beauty, is cannot be specified by
example and exemplar; and the argument directly extends to such qualities
as one and equal, which means that what they are cannot be specified by
example and exemplar. This is to consider qualities about which, in
Parmenides , Socrates is confident that there are Forms – the Form
of the quality, F, we have been arguing, whatever else it also is, is that
which the quality F is. In regard to qualities such as human, about which
Plato has Socrates say in the Parmenides passage that he is not sure whether
he should take up the same position, but of which, elsewhere (e.g.,
Philebus a), he has him say that there are Forms (what Socrates says
there is that there are unities, henads, of them), I have argued that there is a
way in which the argument extends to them. This is because things such as
humans depend, for their goodness and indeed for being what they are, on
being unitary and one in certain distinctive ways. It follows that what
human is cannot be specified by example and exemplar, because: what one
is cannot be so specified, and being one in certain distinctive ways is part of
what it is to be human. When it comes to things such as beds, of which
Plato says that there are Forms late in the Republic (and likewise desks, and
the shuttle mentioned in the Cratylus and the like), I have argued that he
thinks this, purely and simply, because he thinks that there is a Form of
human, and things such as beds are made to serve humans, their needs and
in general their good. This is to say that, if there is something that human
is, and if a bed is something that is made to serve humans in certain
distinctive ways, then there is something that bed is. It follows that what
bed (or desk or shuttle) is cannot be specified by example and exemplar,
because what human is cannot be so specified, and a bed is something
made to serve humans in certain distinctive ways.
What we have argued, therefore, is that Plato has a distinctive argument,
put forward, originally and clearly, in the Hippias Major, which demon-
strates, either immediately or in certain mediate ways, that, for a whole
range of qualities, what a certain quality is cannot be specified by example
and exemplar. We have also argued, in its own right and in regard to Plato,

See Chapter , where I develop this line of argument more fully and properly.

The sensory imperceptibility of Forms 
that, first, if what a certain quality, F, is cannot be specified by example
and exemplar, then what this quality is, its essence and Form, cannot be
perceived by the senses; and, secondly, that this shows why the essence or
Form of a certain quality cannot be perceived by the senses: because what
this quality is, its essence and Form, cannot be specified by example and
exemplar. We may conclude that Plato has a distinctive argument, going
back to Hippias Major but taken further especially in Phaedo, which
demonstrates, either immediately or in certain mediate ways, that, and
why, for a whole range of qualities, what a certain quality is, its essence and
Form, cannot be perceived by the senses: because what such qualities are
cannot be specified by example and exemplar.
What about numbers? Does Plato’s argument extend to them? Can
what the number three is be specified by pointing to a clover and remark-
ing ‘This has three leaves’? Or what the number four is be specified by
pointing to a quartet of musicians and remarking ‘Four people are playing’?
If so, what these numbers are, their essence and Form, can be perceived by
the senses; if not, not. It is true, even if such small numbers can be defined
by example and exemplar, and hence their essence and Form can be
perceived by the senses, it is not plausible to think that the same will be
true of larger numbers – as if one were to point to a beehive and remark,
‘This is occupied by a million bees’, to define the number one million. We
may still need to appeal to calculation (logismos), and hence to a manner of
reasoning (likewise logismos) as opposed to sense-perception, to define a
number such as ; that is, numbers such that we are not able to take in
at a glance that a collection of sense-perceptible things of a certain number
is a collection of things of that number. Such reasoning, it may be thought,
will not only involve sense-perception in some way, it will be strictly based
in sense-perception.
Clearly, the Hippias Major argument, if it extends to numbers and to
specifying of a number what it is, does not do so directly. For there is no
plausibility in thinking that some collections of n things are not worthy of
being judged to be collections of n things when compared to certain other
collections of n things – the way in which some beautiful things are not
worthy of being judged beautiful when compared to certain other beautiful
things, or the way in which some unitary things are not worthy of being
judged unitary when compared to certain other unitary things.
The question, therefore, is whether the Hippias Major argument
extends, in an indirect way and comparable to the way in which it extends
from one to human, to numbers; indeed, from one to numbers. There is a
passage in Republic VII that allows us to consider this question with

 Plato’s Essentialism
particular benefit. The passage is d–a, and it raises some difficult
and important issues of interpretation. Fortunately, on any reading of it,
it says that appeal to things that are not both one and many is necessary for
defining numbers. It follows that it is not possible to define a number, any
number, by appeal to a thing (or body, sōma, d) that can be perceived
by the senses. For, as Socrates and Glaucon have just argued and agreed
(see d–a), any sense-perceptible thing that is one will also be
many. Indeed, the passage (d–a) expressly says that numbers are
objects of thought and not of sense-perception. On the basis of our
argument so far in this chapter, we may conclude that it is not possible
to define a number by example and exemplar and that this is the reason
why it is not possible to define a number by appeal to a thing that can be
perceived by the senses.
Is it perhaps precipitate to conclude this? For, it may be objected, the
argument in Republic VII, d–a, against defining numbers by appeal
to sense-perceptible things, may not be related, or as closely related as we
have supposed, to the argument going back to the Hippias Major against
defining such qualities as beauty, and, by extension, one, by example and
exemplar and by appeal to sense-perceptible things. The question is
whether the argument in d–a relies on the supposition that defin-
ing numbers requires defining one: if it does, our conclusion stands; if not,
it may not stand. There is a particular reason to think that the argument in
d–a relies on this supposition, and though the reason may be
circumstantial, it is, I think, strong. The fact is that Socrates and
Glaucon have just argued that any sense-perceptible thing that is one will
also be many; and they have argued this, precisely, for the sake of
considering whether the question ‘What is the one itself?’ (cf. anerōtan ti
pote estin auto to hen, a) can be answered by appeal to sense-
perceptible things – by example and exemplar, as we have been saying –

There are three such issues. First, is Plato here criticising the mathematicians for treating of things,
including sense-perceptible things, as unitary, without them asking, with any serious intent,
whether these things genuinely are unitary and partless? (This is Burnyeat’s reading (
[originally, ], –); I am inclined to agree with him.) Secondly, is Plato committing
himself to the view that things that are only one and not also many are necessary for defining
numbers or even are that which numbers are? This question is left open by the fact, which is evident
from the text, that he is committing himself to things that are not both one and many, for defining
numbers. Thirdly, if Plato is committing himself to things that are only one and not also many (and
not only to things that are not both one and many), are these things supposed to be Forms or are
they supposed to be something distinct from both Forms and sense-perceptible things, namely,
numbers so understood? (The latter is what Aristotle thinks Plato, or perhaps rather, certain other
Platonists, supposes numbers to be: neither Forms nor sense-perceptible things, but a third kind of
thing.)

The sensory imperceptibility of Forms 
and for concluding, in the strongest terms, that it cannot (see
d–a). The question, in d–a, which is the question what
is required to define numbers, and whether sense-perceptible things are
sufficient to do so, follows on directly from this argument. We have,
I think, reason to conclude that Plato thinks that defining numbers
requires defining one.
This circumstantial reason, for concluding this, is not conclusive. It will
amount to conclusive reason if supplemented by the supposition that, in
d–a, Plato commits himself to the view that things that are only
one and not also many are necessary for defining numbers, or even are that
which numbers are (whether these things are Forms or are things that are
neither Forms nor sense-perceptible things). However, I do not want to
commit to this supposition without examination and to examine it
would take us too far into Plato’s account of mathematics, which is not
where I want to go more than necessary for our present purpose. I shall,
therefore, leave it at that.
. That Forms cannot be perceived by the senses does not imply
that they can be known only a priori
I have argued that the reason why Plato thinks that the Forms of certain
qualities cannot be perceived by the senses is, purely and simply, that he
thinks that what these qualities are cannot be specified by example and
exemplar. It is, therefore, wrong to suppose, as critics commonly have, that
Plato’s reason for this claim, the sensory imperceptibility of Forms, is that
he thinks that Forms are non-physical things; and only physical things can
be perceived by the senses; and non-physical things can only be known
a priori.
I want to end with what seems a remarkable and significant conse-
quence of this conclusion. For, if this is why the Forms of certain qualities
cannot be perceived by the senses, namely, because what these qualities are
cannot be defined by example and exemplar, then we cannot infer that the
Forms of these qualities can only be known a priori. This is not, of course,
to question that, for Plato, Forms are objects of the intellect and reasoning
(noēta), it is to question that Plato’s distinction and opposition between

I consider this closely in Chapter .

It may seem that he commits himself to this view in a–. But since in these lines he is reporting
what certain people, apparently philosophically inclined mathematicians, would say, it is hard
to tell.

 Plato’s Essentialism
objects of the senses (aisthēta) and objects of the intellect (noēta) is
equivalent with the modern distinction between things known a posteriori
and things known a priori.
It is, if I am not mistaken, a mere fallacy to think that if an object, O,
cannot be perceived by the senses, then the knowledge of O is a priori.
This is no less a fallacy, if the object, O, is an essence and Form of a
quality, F, in the sense of that which is designated by an adequate and
true answer to the question ‘What is F?’ Things that we take to exist
but not to be perceptible, even in principle, by the senses, are what are
today commonly called ‘theoretical entities’; but it is wrong to suppose
that theoretical entities can only be known a priori – the natural
sciences, as we have them today, are full of theoretical entities, but no
scientist or philosopher of science would think that all these entities are
known a priori.
What is the fallacy? One way of diagnosing the fallacy is to point out
that the inference from ‘O is not perceptible by the senses’ to ‘O can be
known only a priori’ is valid only if it is true that: If O is not perceptible by
the senses, then O is not (either in its own right or in relation to us and our
epistemic and other interests and needs) bound up with things that are
perceptible by the senses. However, this is a manifest untruth – witness the
idea of saving the appearances, or the idea of argument to the best
explanation, or, indeed, Plato’s idea of a sense-based, aporia-based justifi-
cation for sensorily imperceptible essences and Forms.
I was, some time ago, reading a decently good book on Brexit, which,
for all its virtues, fails to ask what national sovereignty is and what is a good
mark of it. In wondering about this, I confess I am not inclined to think
that the answer can be known by pointing to an exemplary example of a
sovereign nation. Perhaps this is philosophical prejudice. The problem is
that it seems hard to find such a nation, without wondering whether and
to what extent it really is sovereign and how we are to tell. Does it follow
that the answer to the question ‘What is national sovereignty?’ can, or
must, be known a priori? The mere suggestion that this is how this is
known seems absurd. One needs to study a lot of history, including some
economic history, and even some law, and reflect carefully on what one is
studying, properly to address the ti esti question regarding national sover-
eignty. It seems no less absurd than to think that when, in such dialogues

Harte (, –) refers to Plato’s Forms as ‘theoretical entities’; and she says: ‘where there is
no theoretical work for Forms to do, there is no reason to posit them’.

For this last idea, see Chapters  and .

The sensory imperceptibility of Forms 
as Phaedrus and Symposium, Plato is occupied with what love (erōs) is, and
since he does not appear to think that this question can be answered by
pointing to an exemplary lover, he must think that the answer to this
question can, or must, be known a priori. This is because, among other
things, to consider what erōs is, one needs, Plato argues (in the Phaedrus
this is especially clear), to consider whether erōs is something irrational in
us and whether all irrationality in us is as such bad; and it is hard to
imagine how one could begin to think properly about such questions, the
latter not least, without some serious life experience – lived experience. It
seems to me the height of absurdity to think that when Plato argues, in the
Phaedo, that fire is essentially hot and snow essentially cold, these propo-
sitions are supposed, by Plato, to be knowable a priori.
When I put this result to a colleague and friend – that Plato’s claim that
Forms cannot be perceived by the senses does not imply that they can be
known only a priori – he advised that, even though he likes it, it all
depends on what is meant by the term ‘a priori’, a term that, as he pointed
out, is used by philosophers today in a variety of different ways, which,
therefore, I ought to disambiguate. I cannot help the sinking feeling
I experience at this kind of response in general – death by disambiguation,
if I may – as invariably it takes the life out of what one is trying to say. The
present point does not at all depend on the term, a priori, or the many
ways in which it is used today. It depends on the fact that commentators
on Plato have equated Plato’s distinction, in Phaedo and Republic, and in
regard to the knowledge of Forms, between sense-perception and reason-
ing with the distinction between empirical knowledge and a priori knowl-
edge, and, in doing so, have supposed that the reasoning by which Forms
are known cannot involve sense-perception. Bostock, for example, having
made, and defended, this equation in regard to the Phaedo, wonders how,
at the end of the dialogue, Plato can credibly think that the truths that fire
is essentially hot and snow essentially cold (Plato defends that these are
truths) are known a priori.
These same critics need never have used the term a priori; they could
simply have said: Since Plato thinks that Forms cannot be perceived by the
senses and can be known only by reasoning, we may understand him as

Bostock (, ) says: ‘The new method [i.e., the method in Phaedo  ff.] is supposed to be
one that does not rely on the senses: it is meant to proceed wholly a priori and without the benefit of
any lessons from experience.’ He responds to Plato: ‘I see no contradiction in the supposition that
there could be a fire that was not hot, or a lump of snow that was not cold. But I imagine that Plato
would not have agreed on this point, and he would have thought that we can tell a priori that fire is
hot and snow cold’ (; original emphasis).

 Plato’s Essentialism
thinking that this reasoning ‘proceed[s] . . . without the benefit of any
lessons from experience’ (Bostock , ). What I take myself to have
demonstrated is that this inference is mistaken.
To recognise that this inference is mistaken, we have had to consider
why Plato thinks that Forms cannot be perceived by the senses. And we
have had to consider this by considering what, for Plato, Forms basically
are, and not, as critics commonly have, by supposing that a Platonic Form
is simply anything that has a set of characteristics, including being change-
less and non-physical (from which it may indeed follow that Forms cannot
be perceived by the senses). Only if we consider this, and in this way, can
we properly determine in what sense Plato intends the claim that Forms
cannot be perceived by the senses. Once we consider, in this way, why
Plato thinks that Forms cannot be perceived by the senses, and once we
recognise, as I have argued we should, that this is because he thinks that
what the qualities are that these Forms are of cannot be specified by
example and exemplar, we see immediately that this inference is fallacious.
The same critics would have recognised immediately that the inference is
fallacious, except that they did not ask why, and in what sense, Plato
thinks that Forms cannot be perceived by the senses; or, if they did, they
did not do so in the way on which I have insisted, that is, by considering
what Forms basically are; and they did not associate Plato’s justification for
the claim with his view that what certain things or qualities are cannot be
defined by example and exemplar.

 
Why are essences, or Forms, unitary, uniform and

non-composite? Why are they changeless? Eternal?
Are they logically independent of each other?
Phaedo and Republic
. The unity, uniformity and non-compositeness of Forms

The statement that a Form is something uniform (monoeides, ‘uni-form’) and
non-composite (asuntheton) is afforded its clearest expression in Phaedo
a–b. This happens when Socrates responds to Kebes’ objection that,
even if, by his previous argument (the argument from Recollection,
e–a), Socrates has established that the soul pre-exists the body, he
has not established that it does not dissolve upon the dissolution of the body.
In response to this objection Socrates argues that the soul is not such as to be
capable of dissolution (dialuesthai, ‘to dissolve’; also diaskedannusthai, ‘to be
scattered’). This is because, first, the soul is akin to (suggenēs, d, b; also
e) the Forms, with which it is ideally occupied and is at home; secondly,
the Forms are uniform (monoeides) and non-composite (asuntheta) and,
thirdly, only that which is composite is capable of dissolution. The terms
monoeides and asuntheton are used interchangeably throughout, both for the
Forms and for the soul (see c, c, d, b, e).
In Republic and Parmenides, Forms are characterised as being one (hen),
apparently in the sense of being unitary; and it is plausible to think that
this is closely related to, and perhaps the same as, the characterisation of
them as uniform and non-composite. This is plausible especially in view of
the following passage – a most important passage, to which I shall come
back later in this chapter:
And the same account is true of the just and the unjust, the good and the
bad, and all the Forms. Each of them is itself one (hen), but because they
manifest themselves everywhere in communion with actions, bodies, and
one another, each of them appears to be many. (Republic V. a–; trans.
Grube-Reeve, adapted)

Cf. hēmeteran ousan [hē ousia], ‘being ours’, at e–; i.e., the Forms, or essences, referred to as hē
ousia, ‘essence’ or ‘being’, are characterised as ‘being ours’.


 Plato’s Essentialism
That these two claims – that a Form is one and unitary and that a Form is
uniform – are intended as equivalent is confirmed by the fact that in a
number of passages in Republic and Parmenides Plato explicates that by an
eidos, or Form, he means an idea (‘quality’) that is one. If we go back to
what appears to be the earliest occurrence of the term eidos when used for
the essence of a thing, which is Euthyphro d–e, we may observe that Plato
has Socrates insist that what he is looking for is one eidos to serve as the
standard for all cases, as opposed to several ideai that might serve as such a
standard when conjoined and combined; for he insists that all things that
are holy are holy on account of one quality (mia[i] idea[i]):
Bear in mind then that I did not bid you tell me one or two of the many
pious actions but that form (eidos) itself because of which (hō[i]) all pious
actions are pious, for you agreed that all impious actions are impious and all
pious actions pious because of one quality (mia[i] idea[i]), or don’t you
remember? (Euthyphro, d–e; trans. Grube, adapted)
In the immediately following lines (e–), Socrates characterises this eidos
as the ‘standard’ (paradeigma) by the use of which we can determine of
each and every thing whether or not it is such as to be pious. In the
Parmenides it is explicated that a Form (eidos) is a quality (idea) that is one
and unitary (a–, c–, b–c).
Why does Plato think that Forms are unitary, uniform and non-
composite? How is this claim to be understood? And what are its conse-
quences? Is it a consequence that Forms are logically independent of each
other? These are the questions I consider in this chapter.
.. A false start

There is no plausibility, it seems to me, in the suggestion that Plato thinks
that Forms are unitary, uniform and non-composite because he thinks that
they are non-physical things. Even if we suppose, on the basis of the
Phaedo passage in which he makes this claim (a–b), that being non-
physical is a characteristic of the Forms, and even if we suppose that being
non-physical is a necessary condition for being unitary, uniform and non-
composite in the way in which these terms are intended by Plato when
applied to the Forms, this, clearly, is not a sufficient condition for being
unitary, uniform and non-composite. The set of Socrates and the scram-
bled egg I had for breakfast is not a physical thing but an abstract object,
but one may seriously wonder about the non-compositeness of this thing;
and the same can be said about the number , if it is thought that it is
composed of a number of units.

The simplicity and changelessness of Forms 
Furthermore, it is not plausible to think that the supposition that Forms
are non-physical things provides a good reason for supposing that Forms
are unitary, uniform and non-composite. If anything, it is more question-
able, and in greater need of justification, to suppose that Forms are non-
physical things than to suppose that they are unitary, uniform and non-
composite; or at any rate there is no clear justificatory priority of the one
over the other.
.. Unitary accounts and unitary things

What is the difference between such apparently earlier passages as
Euthyphro d–e, Hippias Major – and Meno d ff., in which
Plato claims that essences, and eidē, are unitary and one, and such
apparently later passages as those in Phaedo, Republic and Parmenides, in
which he claims that essences, and Forms, are unitary, uniform and non-
composite? Or is there, perhaps, no difference? For, if we compare the
wording of the Euthyphro passage with the Republic passage, they are
remarkably similar.
There is, I think, a significant difference, which may not be apparent if
the passages are read in isolation but becomes apparent if they are read in
context and in regard to their dialectical function. It appears that the
function of the earlier passages is to explicate to an interlocutor, if he is
not familiar with Socrates’ demand for definitions and search for essences,
what it is that Socrates is asking for, when he is asking for an answer to a ti
esti question and an account (logos) of what a thing is. Explication is
needed, because the interlocutors tend to suppose that the question
‘What is F?’, which is understood by all involved as the request for a
standard (paradeigma) for a thing’s being such as to be F, can be answered
by any one of multiple answers; or, as we would say, by a conjunctive or
disjunctive account. Euthyphro thinks he can appeal to his own particular
action as a standard of pious action in general, thereby implying (as
Socrates makes clear) that one can equally appeal to any number of other
particular things, as standards of pious action. Meno proceeds on a similar
assumption with regard to virtue; he thinks one answer is appropriate if the
virtue of men is in question, another answer if the virtue of women is in
question, etc., and that this is all that needs to be said about what virtue is.

If the reader is sceptical about such chronological suppositions, they do not really matter to my
point; what matters is the different dialectical functions, and different audiences, of
different dialogues.

 Plato’s Essentialism
Hippias proceeds very much as do both Euthyphro and Meno, when he
maintains that the question ‘What is beauty?’ can be answered by invoking
a particular girl of exemplary beauty and when he accepts Socrates’
clarification that, in that case, invoking a particular horse of exemplary
beauty, or a particular lyre of exemplary beauty, will equally serve for an
answer, though perhaps appropriate for different kinds of things. In such
passages, which appear to be from dialogues before the Phaedo, Plato
argues that if an account of what a thing is is adequate, then it is a
unitary account.
On the other hand, in several passages in such dialogues as Phaedo,
Republic and Parmenides, Plato asserts that essences, and eidē, are unitary
(and uniform, and non-composite). In these passages, Socrates is addres-
sing interlocutors who are familiar with the distinctive character of Socratic
enquiry, its demand for definitions and the search for essences; and so,
apparently, Plato is addressing readers already familiar with his conception
of philosophical argument and enquiry. These interlocutors are not
inclined to answer the ti esti question (e.g., the question ‘What is justice?’
in the Republic, or the questions ‘What is dying?’ and ‘What is the soul?’ in
the Phaedo) with multiple answers that can at best be strung together in, as
we would say, a disjunctive or a conjunctive account; they know that this is
not an adequate way of proceeding. The function of these passages is not,
therefore, to explicate that an account of what a thing is must be unitary.
What is their function? It is, I want to suggest, to prompt such
interlocutors, and Plato’s readers, to consider the following question:
What must that which is designated by an account of what a thing is,
and hence an essence and an eidos, be like intrinsically, if the account is
adequate and true? And it is to affirm the following answer to this question
(The relation between unitary accounts and unitary things):
If an account of what a thing is is adequate (hence general, unitary and
explanatory) and true, then that which it designates is unitary, uniform and
non-composite.
I am supposing here that, in a number of dialogues that appear to be before
the Phaedo, and one of whose major functions is to introduce the reader to
Plato’s conception of philosophical argument and enquiry, Plato indicates,
in a variety of ways, that he thinks that an adequate answer to a ti esti

In the Phaedo, when Socrates is at the end of his life, this familiarity is underscored. In the Republic,
it is intended that Glaucon is well versed in their particular method of argument (see, e.g., e–).
In the Parmenides, even if Socrates is presented as a young man, the theory of Forms is presented as
familiar to all involved in the discussion.

The simplicity and changelessness of Forms 
question (or, to this question when asked of certain things or qualities)
must be: general (i.e., not by example and exemplar), unitary and
explanatory.
. Plato’s reasoning for the claim that Forms are unitary, uniform
and non-composite
It appears, therefore, that Plato’s reasoning for the claim that Forms are
unitary, uniform and non-composite is this:
P A Form is that which is designated by an adequate (hence general,
unitary and explanatory) and true account of what a thing is.
P If an account of what a thing is is adequate (hence general, unitary
and explanatory) and true, then that which it designates is unitary,
uniform and non-composite. (= The relation between unitary accounts
and unitary things)
Therefore,
 A Form is something unitary, uniform and non-composite.
That Plato holds Premise  we have already gone some way towards
showing, and to demonstrate this is the aim of this entire study. In regard
to Premise , we shall see that there is reason to suppose that Plato holds it.
We have reason, therefore, to think that this represents Plato’s reasoning
for the claim that Forms are uniform, unitary and non-composite.
It might be objected that the cogency of this argument is rendered
questionable by what might seem to be the following suppositions in it:
first, the word ‘unitary’ is used in the same sense when it is predicated of an
account as when it is predicated of that which is designated by an account;
secondly, the same unity is in question in both; thirdly, the nature of the
unity of that which is designated by an adequate and true account can be
directly derived from the nature of the unity of such an account. I want to
make plain that the argument does not depend on any of
these suppositions.
Of course, we would like more concrete textual basis for Premise .
I believe we have such a basis, in the following passage from Phaedo:

I have argued for this supposition in The Structure of Enquiry in Plato’s Early Dialogues, Part I (Politis
). It is shared by many critics.

In and through Chapters  and .

 Plato’s Essentialism
After this, he said, when I had wearied of investigating the things-that-are,
I thought that I must be careful to avoid the experience of those who watch
the sun in the course of an eclipse, for some of them ruin their eyes unless
they watch its reflection in water or some such material. A similar thought
crossed my mind, and I feared that my soul would be altogether blinded if
I looked at the things-that-are with my eyes and tried to grasp them with
each of my senses. So I thought I must take refuge in statements and
investigate in statements the truth of the things-that-are. However, perhaps
this analogy is inadequate, for I certainly do not admit that one who investigates
the things-that-are in statements is any more investigating them in images than
is one who investigates them directly. (d–a; trans. Grube, adapted)
Plato states here that we should investigate the things-that-are (ta onta) not
directly through the senses, lest we become blinded like those who inves-
tigate the sun directly and not by investigating its reflection in an image;
rather, we should investigate the things-that-are in logoi (‘statements’,
including, as it subsequently emerges, explanatory theories).
I want to consider what Plato does when, at the end of this statement,
he adds the clarification that we must not at all suppose that ‘one who
investigates the things-that-are in logoi is any more investigating them in
images than is one who investigates them directly (en ergois)’ (a–).
The meaning of this clarification, I take it, is not that both these ways of
investigating things is investigating them in images and indirectly; on the
contrary, what Plato clarifies is that, if anything, neither of these ways of
investigating things is investigating them in images or indirectly. It makes
sense to say that investigating a thing directly through the senses is not
investigating it in an image. What is less clear is how Plato can claim, with
any plausibility or justification, that investigating a thing in logoi is not
investigating it in (something like) an image, and that investigating a thing
in logoi is not investigating it in any less direct a way than investigating it
directly through the senses.
This ought, it seems to me, be a pressing critical question. Philosophers
today are commonly worried about, in particular, a Platonist philosophy of
mathematics, on the grounds of what is now commonly called a causal
account of knowledge. The worry is that unless we give a causal account of
our knowledge of some supposedly real thing, X, we cannot be sure that it
is possible for us to know this very thing, X, and not merely X as
represented and conceived by us in our statements and theories. If one
thing is clear about the above passage, it is that Plato rejects a causal
account of knowledge in favour of a theory-based account of knowledge.
At the same time, and through his concluding clarification, he rebuts any
suggestion that only a causal account of knowledge can afford us a

The simplicity and changelessness of Forms 
knowledge of the things themselves, or that a theory-based account of
knowledge can only afford us knowledge of things as represented and
conceived by us in our statements and theories.
Suppose, now, we ask of Plato the following, critical question:
Q How can one, with any plausibility and justification, claim that
investigating a thing in logoi (‘statements’, including theories) is
not investigating it in (something like) an image and that investigat-
ing a thing in logoi is not investigating it in any less direct a way than
investigating it directly through the senses?
I think it is Plato that invites us to ask this question of him. What he says
in this passage can naturally be understood as implying that any thing to
whose existence we are committed in and through our logoi is (what we
would call) a ‘theoretical entity’. And one notable critic has argued that
Plato’s Forms are, or are very much like, ‘theoretical entities’. However, if
we say this, then we are immediately obliged to clarify that, unlike us (or,
some of us), who are immediately inclined to question that a theoretical
entity can be simply real – a thing-that-is – and who are immediately
inclined to suppose an opposition between theoretical entities and that
which really exists, Plato does not see any necessary opposition between
thinking that X is a theoretical entity and thinking that X is a really
existing thing. And now we can hardly fail to notice that this is precisely
what he clarifies, in so many words, when (at a–) he adds the
clarification at the conclusion of the famous passage from the Phaedo.
It seems to me, therefore, that we ought to consider this critical
question, Q, as a straight question that calls for a straight answer from
Plato. We must ask, therefore:
Q What is Plato’s justification for thinking that (P): It is possible, in and
through our logoi (‘statements’, including theories), to grasp not merely
things as they are represented and conceived by us in our statements and
theories, but simply the things-that-are?
A good answer, it seems to me, is to suppose that Plato thinks that a good
way of justifying P is by arguing that, ultimately and in the final analysis,

Harte . I note that Forms are introduced shortly after d–a, as the contents and objects of
certain logoi.

This is an important qualification. Plato should not be understood to think that any logos should
immediately be grounded through a Euthyphro dilemma; he should rather be understood to think
that certain logoi need, in the first instance, to be based on a logos tēs ousias – a real definition – and
that such logoi tēs ousias need to be grounded through a Euthyphro dilemma.

 Plato’s Essentialism
our logoi, if they are to be adequate and true, must be grounded in the
nature of real things, the things-that-are, where this grounding is to be
understood in terms of the following Euthyphro dilemma:
Q Is it the case that (OPTION ): That which is designated by an
adequate and true account of what a thing is is unitary, uniform and
non-composite BECAUSE the account that designates it is unitary?
OR is it the case, on the contrary, that (OPTION ): An adequate
and true account of what a thing is is unitary BECAUSE that which
it designates is unitary, uniform and non-composite?
We may suppose that he thinks this is a good way of justifying P, because
he thinks that: i. If a grounding is based on a Euthyphro dilemma, then it
based in our logoi; and ii. If a grounding is based on the latter option in
response to a Euthyphro dilemma, then it provides an account of how
things really are, not merely how they are represented and conceived by us
in our statements and theories.
What the Phaedo passage indicates is that Plato is keenly aware of the
following questions:
Q What must be the character of that in which an account of what a
quality, F, is is grounded, if that account is to be, first, general,
unitary, and explanatory, and, secondly, true?
Q What is the relation between the unity of an adequate and true
account of what a thing is, and the unity, uniformity, and non-
compositeness of that which is designated by an adequate and true
account of what a thing is?
The Phaedo passage indicates, I want to suggest, that he considers these
questions in the terms with which we are familiar, in regard to a particular
question of a corresponding form, from the Euthyphro and the Euthyphro
dilemma there.
This, I propose, is why Plato holds The relation between unitary accounts
and unitary things (i.e., Premise ). He holds this claim, because he
properly considers the question: What grounds the unity of an adequate
(hence unitary) and true account of what a thing is? Or, as he would,
rather, formulate this question: What is that because of which (cf. the hoti
in the statement of the Euthyphro dilemma at Euthyphro a–) an
adequate (hence unitary) and true account of what a thing is, is unitary?
And he answers this question by affirming that the one and only thing that
can ground the unity of an adequate (hence unitary) and true account of

The simplicity and changelessness of Forms 
what a thing is, is the unity, uniformity and non-compositeness of that
which is designated by such an account.
. Are Forms logically independent of each other?

No one supposes that, in the dialogues before the Phaedo, Plato thinks that
essences are logically independent of each other and do not stand in logical
relations to one another, in particular, the relations of implication and of
incompatibility. No one supposes this, either in general or on the grounds
that Plato thinks that an essence is that which is designated by an adequate,
and hence unitary, and true account of what a thing is. It is familiar that, in
several places in the dialogues before the Phaedo, Plato indicates logical
relations among different essences and among the different accounts that
designate them. He does so, for example, when he considers the relations
between the particular virtues and virtue in general (see esp. Protagoras).
And no one supposes that his doing so is in tension with his view that the
essence of a thing must be capable of being stated in a unitary account. It is
evident that the unity requirement for the adequacy of an account of what
a thing is, is compatible with the view that the different accounts of what
different things are, and so too the different essences designated by such
accounts, stand in logical relations to one another.
It is also familiar that, in the Sophist most especially, and under the title
of ‘the interweaving of Forms’ (sumplokē eidōn, e–; Plato has repeat-
edly been using koinōnia and koinōnein, ‘communion’ and ‘communing’,
for what he here refers to as sumplokē, ‘interweaving’; see, e.g., a,
b, b), Plato takes up and considers, extensively and incisively,
this topic in general: the logical relations among Forms. On a natural
reading of the relevant section of the Sophist ( ff.), Plato there provides
an account of how it is possible for different Forms to be related to each
other: the account appeals to certain most important kinds, or Forms, such
as identity and difference, as being that which enables different Forms to be
related to each other. There is no immediate suggestion, much less
implication, that, before the Sophist and in such dialogues as Phaedo and
Republic, Plato thought this was not possible: what he provides in the
Sophist is, rather, an account of how it is possible.
At the same time, there are many critics who think that, between the
dialogues before the Phaedo, on the one hand, and the Sophist, on the
other, including in those dialogues in which he sets out the theory of
Forms, Plato thinks that Forms are logically independent of each other and

 Plato’s Essentialism
do not stand in logical relations to one another. What is this remarkable
view among critics based on? Plato nowhere, in such dialogues as Phaedo
and Republic, or anywhere when he sets out the theory of Forms, says that
Forms are logically independent of each other and do not stand in logical
relations to one another. We have reason to suppose that what he says in
the Sophist is intended as a correction of a view that he held earlier, only if
we have reason to think that this is a view that he held earlier. The only
evidence the logical independence view can appeal to is Plato’s claim, in
Phaedo and Republic, that Forms are unitary (hen), uniform (monoeides)
and non-composite (asuntheton). And this is, in fact, what defenders of the
logical independence view have appealed to in defence of this view.
Is it plausible to suppose that, when Plato claims that Forms are unitary,
uniform and non-composite, he intends this claim to imply that Forms
are logically independent of each other and do not stand in logical relations
to one another? It is not at all clear why we should suppose that
Plato intends this implication. What would a philosopher have to think,
to intend this implication? Apparently, she would have to think that
(Logical Atomism):
Logical relations are based in the composition of the elements that stand in
logical relations to one another.
The most familiar version of Logical Atomism is perhaps that of the early
Wittgenstein, who argued that logical relations are based in the truth-
functional composition of propositions; hence that propositions that are
not truth-functionally composed of other propositions (in this sense,
‘elementary propositions’) are logically independent of each other and do
not stand in logical relations to one another.
It would seem, therefore, that what we would have to suppose, if we
want to attribute to Plato the following inference: Forms are unitary,
uniform and non-composite; therefore, they are logically independent of each

Critics do this in a variety of different ways, and these ways would, ideally, have to be distinguished,
but, for present purposes, I shall mention a number of the relevant critics without regard to the
variations among them: Stenzel (, ); Cornford (, –, –); Ryle (;
originally ); Hamlyn (, , ); Prauss (), with a good response by Gadamer
(, –; , ). Bambrough (, ) does not seem to say much, but refers to
Crombie (): ‘Mr. I. M. Crombie has usefully remarked that in these later dialogues, and
especially in the Philebus, there is a more sophisticated awareness than in the earlier dialogues of the
nature and importance of the relations between universals and universals as opposed to relations
between two artificially divided worlds of universals and particulars (An Examination of Plato’s
Doctrines, Vol. II, pp.  ff ).’ Moravcsik (, ); and, most recently, M. L. Gill (,
introduction, ).

The simplicity and changelessness of Forms 
other, is that he holds some version or other of Logical Atomism. Looking
for Logical Atomism in Phaedo, the Republic or indeed the first part of the
Parmenides, does not promise to turn up anything – like the proverbial
goose chase.
Only at the end of the Theaetetus, in the passage commonly referred to
as ‘Socrates’ dream’ (d f.), does Plato engage with a view that may be
compared to Logical Atomism. What he sets out there is the view that
things that have an account (logos) are composed out of things that do not
have an account and that can only be named. Plato’s purpose there is to
argue against this view. It might be thought that what he intends to argue
against is a view that he used to hold. But this supposition is credible only
if we can find independent evidence, in such dialogues as Phaedo, Republic
and the first part of the Parmenides, that Plato held this view. I do not
know of such evidence, other than the supposition that Plato’s claim that
Forms are unitary, uniform and non-composite implies that they are
logically independent of each other. But the onus of proof is on the
defender of the claim that Plato’s Forms are logically independent of each
other to show that there is this implication.
If the account that we have proposed and defended of why Plato thinks
that Forms are unitary, uniform and non-composite, is correct, then it is
not the case that the claim that Forms are unitary, uniform and non-
composite implies that they are logically independent of each other. On
the account that we have proposed and defended, Plato’s claim that Forms
are unitary, uniform and non-composite is based on the claim that a Form
is that which is designated by an adequate, and hence unitary, and true
account of what a thing is. And it is evident that the unity requirement for
the adequacy of an account of what a thing is, is compatible with thinking
that the different accounts of what different things are, and so too the
different essences designated by such accounts, stand in logical relations to
one another.
This leaves the textual question: What, in such dialogues as Phaedo and
Republic, does Plato say on the topic of the logical relations among Forms?
This question deserves a proper investigation of its own; and this is not
part of our present remit. I shall readily admit that he does not say very
much at all, provided that this admission is not seen as evidence for the
view that Plato’s Forms are logically independent of each other – plainly it
is not such evidence. Suffice it to note that there is a single passage, in
Republic, in which he says that Forms ‘commune with one another’ (cf.
allēlōn koinōnia) and says, as expressly and clearly as one may wish for, that

 Plato’s Essentialism
their communing with one another is no reason at all for denying that they
are unitary:
Each of them [i.e., ‘all the Forms’] is itself one, but because they manifest
themselves (phantazomena) everywhere in communion with actions, bodies,
and one another, each of them appears to be (phainesthai) many. (Republic
V. a–; trans. Grube-Reeve, adapted)
This passage ought, if anything, to have given pause to those who argue
that the Forms of Phaedo and Republic are intended to be logically
independent of each other or who think they can detect some version
of Logical Atomism in these dialogues. It ought to have done this, in its own
right and in conjunction with Plato’s reference, at Republic (VI. c), to
the Forms as forming a well-ordered world, kosmos. I find a salutary
corrective in Nettleship’s words: ‘And forms are primarily spoken of as
elements of unity in a multiplicity of sensible things. But it is important
not to overlook the further application of the same principle which is

It may be said that the passage says that Forms appear (phainesthai) many and that this does not
imply that they are many. Correct. But the passage says that Forms manifest themselves
(phantazomena) in communion with one another, with actions, and with bodies, and that, as a
result of this, they appear many. It can hardly be denied that Plato intends this to imply that they do
commune with one another, with actions, and with bodies. But, if the passage says that Forms
commune with one another, it cannot, it seems to me, be read to imply that they are logically
independent of each other. Or, if we want to be extra cautious, it is incumbent on the critic who
thinks that, in Phaedo and Republic, Forms are logically independent of each other, to explain what
it means to say that they commune with one another; and to do so in a way that is consistent with
the fact that when, in the Sophist, he says that Forms commune with one another, he intends this to
mean, or directly imply, that they do stand in logical relations with each other (especially relations
of implication and incompatibility). I know of no such explanation. For this issue, and for the
translation of the passage, Gould (, ; with the Greek transliterated by me in this note) is
instructive: ‘Waterfield, taking the koinōnia as real, implies that the eide are immanent in sensible
particulars; Grube, treating the koinōnia as an appearance, suggests that the eide (while separate
from sensible particulars) are unrelated to one another. Either the koinōnia is real or it is apparent. If
real, then the Forms are related to particulars as they are to one another. If apparent, then they may
be as separate from one another as they are from sensible particulars. To adjudicate this would
require a thorough interpretation of Plato’s middle theory. Without going this far, one can yet argue
for the superiority of Grube’s translation of this passage. The crux of the matter thus is whether the
eide appear everywhere (a) because of their association with actions, bodies, and one another or (b)
as though in association with actions, bodies, and one another. How, then, are we to take the
complex subordinate clause? Let us note that it has a men/de construction. Grube’s translation
projects this more clearly, for it highlights the contrast between hekaston einai and phainesthai
hekaston. To opt for (a) is to overlook the parallel between the einai in the straightforward men
clause and the phainesthai in the more puzzling de, a parallel to which Plato draws our attention by
his use of the men/de structure and the contrasting pair of infinitives.’

This is how Nettleship (, ) understood the two passages: ‘in the Republic . . . each form is
itself related to other forms [N. is evidently referring to the a passage here], and ultimately all the
forms of things are connected together and make one system [N. is evidently referring to the c
passage here]’. Nettleship’s ‘system’ for ‘kosmos’ here is notable and, I find, entirely defensible.

The simplicity and changelessness of Forms 
implied in the Republic; each form is itself related to other forms, and
ultimately all the forms of things are connected together and make one
system’ (, ).
. The changelessness and the eternality of Forms

In numerous passages, in Phaedo and Republic and generally wherever he
characterises Forms, Plato characterises them as changeless. Only in a
few passages does he characterise Forms as eternal; and these are
typically passages, whether in Phaedo or in Republic, in which he considers
the soul and the Forms together and argues that the soul is akin to
the Forms.
Plato’s chosen term for the changelessness of Forms is aei hōsautōs echon,
‘being always the same way’, and variations thereof: very occasionally he
drops the aei, ‘always’ (see, e.g., Phaedo a); often he adds aei hōsautōs
kai kata tauta echon, ‘being always the same way and in the same respects’;
or he adds aei hōsautōs kai kata tauta kai peri to auto echon, ‘being always
the same way and in the same respects and in the same relations’. In later
dialogues, in particular the Sophist, he will use the term akinēton, ‘unmov-
ing’, for eidē, or also for eidē (see esp. Sophist b, where he is engaging
with ‘the friends of the Forms’, hoi tōn eidōn philoi). Indeed, he will use
this term in close association with the original term for the changelessness
of Forms (see Sophist b–c, to kata tauta kai hōsautōs kai peri to
auto). This term, akinēton, is not used for Forms in the Phaedo or in the
Republic. Related expressions are used, however, as when he says
(Phaedo c–d) that Forms do not admit of any change (metabolē) or
alteration (alloiōsis).
The fact that Plato chooses these terms for the changelessness of
Forms – aei hōsautōs echon, with the repeated additions kata tauta echon
and peri to auto echon – is, it seems to me, extremely important, especially
when read together with the characterisation of Forms as unitary, uniform
and non-composite. This indicates that the changelessness of Forms is
intended by Plato to be related to the principle of non-contradiction, as he
characterises it in the fourth book of the Republic (–). For these
characteristics and their opposites – being/not being unitary, being/not
being non-composite, being/not being always in the same states, or
respects or relations – are employed by Plato when he spells out the
principle of non-contradiction, and they are crucial in spelling out this
principle. I want to suggest that it is precisely because sense-perceptible
things have parts, and because they are not always in the same way (that is,

 Plato’s Essentialism
they are in different ways, or states, at different times), and because they do
not always stand in the same relations to other things of their kind, and
because they do not always exhibit the same respects in relation
to themselves, that they can be both F and contrary-to-F, and hence
not-F, without contravening the principle of non-contradiction. Forms,
on the other hand, precisely because they are always in the same way (or
state), and always exhibit the same respects, and always stand in the same
relations to things of their kind (i.e., other Forms), cannot be both F and
contrary-to-F, and hence not-F, without contravening the principle of non-
contradiction.
In both the Phaedo (e.g., d) and the Republic (X. e–), Plato
uses the term athanaton (‘immortal’) for the eternality of Forms and does
so in the context of arguing that the soul is akin to the Forms and is
immortal. That he is using the term athanaton for eternality is clear from
the fact that, in these two passages from Phaedo and Republic, he uses
athanaton together with aei on (‘always existing’). It is clear also from a
single passage in the Phaedo (d), when he says that if something is
athanaton, then it is aidion, ‘eternal’. The term aidion, ‘eternal’, is used
only once in the Phaedo, d, and only once in the Republic, b.
This is not to say that these two sets of terms, aei on and aidion, on the one
hand, and athanaton, on the other, are synonymous or in general inter-
changeable. To say of a thing that it is athanaton, is to imply, or at any rate
to suggest, that it is something alive; for it is to say that it is something that
never ceases to be alive. On the other hand, to say of a thing that it is aei on
or aidion has no such implication or suggestion, for it is simply to say that
the thing never ceases to exist. It appears, therefore, that Plato, in using the
term athanaton of the Forms, is allowing himself some liberty, since,
apparently, there is no suggestion that the Forms are alive: Forms are,
indeed, intelligible, noēta, but not intelligent, noun echonta; being intelli-
gent, noun echon is a characteristic reserved for the soul.
I think we may readily suppose that the distinction, such as it is,
between being eternal, aei on/aidion, and being immortal, athanaton,
may be set aside as sufficiently minor, when these terms are used of
Forms. It is natural to suppose that the basic reason why Plato chooses
the term athanaton, for Forms in Phaedo, and not simply the terms aei on
and aidion, is that the question of the dialogue is whether the soul, which
is argued to be akin to the Forms, is immortal and hence eternal
(cf. d); and the natural way of posing this question, in a way that

I return to this point in Chapter .

The simplicity and changelessness of Forms 
makes it immediately recognisable, is through the term athanaton. We may
also recall that Socrates’ last argument for the immortality of the soul in the
dialogue, the very long argument that starts at e and is not completed
before b, relies crucially on the claim that an essential quality of the
soul is to be, precisely, death-less, a-thanaton (see esp. c–e). In regard
to the Republic, we have noted that when Forms are characterised as
something athanaton, it is in the context of an argument for the immor-
tality of the soul, which is argued to be akin to (suggenēs, e) the
Forms. Only rarely does Plato refer to the eternality of Forms elsewhere in
the Republic, and there he uses the term aei on (see VI. b).
Our question is: Why does Plato think that Forms are something
changeless and something eternal? For the sake of properly considering
this question, we need to observe that it would be quite wrong to think
that these two sets of terms, aei hōsautōs echon and aei on/aidion/athanaton
are used interchangeably when used of Forms. This would be to overlook
that there is no implication, in Plato or in general, that if a thing is aei on/
aidion/athanaton (‘eternal’), then it is aei hōsautōs echon (‘changeless’). On
one reading of the Timaeus, Plato thinks that the heavenly bodies are
eternal; but they are evidently not changeless, since they move and
movement is argued to be a form of change (kinēsis). Likewise, Plato
argues (in Phaedo) that the soul is eternal; for he argues that it is immortal
(athanaton) and that, if a thing is immortal, then it is eternal (cf. d).
But in the Sophist () he argues that a certain kind of kinēsis is an
essential quality of the soul, that is, the kinēsis distinctive of the activity and
the process of thinking rationally (noein). And in the Phaedrus (e) self-
movement is said to be of the essence of the soul. We shall see that it is
most important, for considering our question (Why does Plato think that
Forms are something changeless and something eternal?), to bear in mind
that ‘X is eternal’ does not imply ‘X is changeless’.
. A false start, on the changelessness of Forms

One might think that if a thing is separate from physical things, then it is
changeless. And one might invoke this conditional to argue that Plato’s
defence of the claim that Forms are changeless is based on his claim that
Forms are separate from physical things. But the truth of this conditional
(i.e., ‘If something is separate from physical things, then it is changeless’) is
eminently questionable. A central claim in Phaedo is that the soul is
chōriston (‘separable’ and, ultimately and all going well, ‘separate’) from
the particular body whose soul it at some time is, and, ultimately and all

 Plato’s Essentialism
going well, also from all bodies and all body. This claim about the soul,
together with the claim that the soul is akin to (suggenēs; strictly, ‘kindred’,
‘of like kind’) the Forms, is one source of evidence that we have for
attributing to Plato the claim that Forms are separate from bodily and in
general physical things. But, clearly, the soul is not changeless. On the
contrary, since an essential feature of the soul is thinking and reasoning,
and since thinking and reasoning are activities and processes, there is an
important sense in which the soul is characterised by change, and a
distinctive form of change, as Plato indeed spells out in the Sophist
(). The proposition ‘X is separate from physical things’ is, therefore,
compatible with the proposition ‘X is changing’.
It might be said that, whether or not he does so cogently, in Phaedo
Plato as a matter of fact bases his argument for the claim that Forms are
changeless on the claim that they are separate from physical things. But it
is questionable whether this is correct. Certainly, it cannot mean that, in
Phaedo, Plato bases his argument for the claim that Forms are changeless
on the statement that Forms are something chōriston. Plato does not state,
in Phaedo, that Forms are something chōriston, rather, that Forms are
something chōriston is, at best, an implication of his statements that the
soul is something chōriston and that the soul is of like kind (suggenēs) as
the Forms.
If we want to consider what, in actual fact, Plato’s argument, in Phaedo,
for the claim that Forms are changeless, is based on, the following passage
is especially instructive:
Then let’s go back to those entities to which we turned in our earlier
argument. Is the being itself, whose being we give an account of in asking
and answering questions, unvarying and constant, or does it vary? Does the
equal itself, the beautiful itself, what each thing is itself, that which is, ever
admit of any change whatever? Or does what each of them is, being
uniform alone by itself, remain unvarying and constant, and never admit
of any kind of alteration in any way or respect whatever? (c–d; trans.
Gallop, adapted)
In this passage, changelessness is argued to be a feature of essences; and,
apparently, the changelessness of essences is based, in particular, in the
uniformity of essences (see monoeides on, ‘being uniform’ or ‘since it is
uniform’, at d). It is essences that are characterised as ‘being always the
same’ (hōsautōs aei echei), as ‘being always the same in the same respects’
(aei . . . hōsautōs kata tauta echei), and as ‘not ever admitting of any
alteration in any point or in any way’ (oudepote oudamē[i] oudamōs

The simplicity and changelessness of Forms 
alloiōsin oudemian endechetai; he also uses metabolē, ‘change’, parallel to
alloiōsis, ‘alteration’, in the same passage). Essences are referred to as ‘The
ousia itself, of whose being we give an account through the asking of
questions and the proposing of answers’; and as ‘what each thing is’ (auto
hekaston ho esti). The back-reference in the opening of the passage is to an
earlier passage that is likewise about essences:
For our present argument is no more about the equal than it is about the
beautiful itself and the good itself and the just and the holy and, this being
precisely what I mean, about all the things to which we apply the title, ‘that
which it is’ (‘ho esti’), both in our questions when we are asking these
questions and in our answers when we are proposing answers. (c–d;
see also d–e, ho tugchanei hekaston on)
We may conclude that we may not suppose that the changelessness of
Forms can be defended on the basis of their being separate from physical
things and that we have good reason, based on Plato’s text, to suppose that
the changelessness of Forms is based in their being essences and, in
particular, uniform essences.
. Plato’s reasoning for the claim that Forms are changeless
If we observe that changelessness is argued to be a feature of essences, and
that an essence is that which is designated by an adequate and true account
of what a thing is, and if we suppose that Forms are, basically, essences,
then we may propose that Plato’s defence of the claim that Forms are
changeless is this:
P A Form is that which is designated by an adequate and true account
of what a thing is.
P If an account of what a thing is is adequate and true, then that which
it designates is changeless.
Therefore,
 A Form is something changeless.
The question, therefore, is why and on what reasoning Plato thinks that
Premise  is true. There is a simple answer, and there is a complex answer
to this question. The simple answer may seem immediately compelling,
but on reflection there is, by Plato’s own lights, a particular objection to it.
The complex answer builds on the simple answer, but obviates
the objection.

 Plato’s Essentialism
On the simple answer, Plato’s reasoning for Premise  is this:
i. An adequate and true account of what a thing, F, is, provides an adequate
and true standard (paradeigma) for a thing’s being such as to be F.
ii. An adequate and true standard for a thing’s being such as to be F,
must be changeless, or else that of which it is a standard (i.e., F)
would not be the very thing it is.
iii. If an adequate and true standard is changeless, then that which it
designates is changeless.
Therefore,
iv. If an account of what a thing is is adequate and true, then that which
it designates is changeless. (= Premise )
Premise i is self-evident.
In regard to Premise iii, we may suppose, on the reasoning that we
spelled out earlier in this chapter, when we considered the unity, unifor-
mity and non-compositeness of Forms, that Plato holds it. That is, he
defends Premise iii by considering the question, What grounds the change-
lessness of a standard for a thing’s being such as to be F?, where this notion
of grounding is to be understood in terms of a certain distinctive version
the Euthyphro dilemma. Or, as he would, rather, formulate this question:
What is that because of which a standard for a thing’s being such as to be
F, is changeless?, where the because here is to be understood in terms of the
because in the Euthyphro dilemma. And he answers this question by
arguing that the one and only thing that could ground the changelessness
of a standard for a thing’s being such as to be F is the changelessness of that
which this standard designates.
Premise ii, however, is, I think, objectionable, by Plato’s own lights. We may
imagine a Hippias, like the one in the Hippias Major, objecting as follows.
Suppose, Socrates, that one set of people use, as a standard for a thing’s
being such as to be beautiful, either girls or horses of exemplary beauty,
while another set of people use, as a standard for at thing’s being such as to
be beautiful, either horses or lyres of exemplary beauty. Alternatively,
imagine a similar scenario, with people using the one standard of beauty
(i.e., girls or horses) at one time, the other standard (i.e., horses or lyres) at
another time. So the standard of beauty varies among these two sets of
people, or one set of people at two different times – it is, as one might say, a
changing standard. Is it evident, Socrates, that these two sets of people, or
one set of people at two different times, are not employing a standard of a
single thing, beauty? No, it is not evident, and nothing you have said about
the mere idea of a standard, or paradeigma, shows that it is evident.

The simplicity and changelessness of Forms 
I do not see how Plato would be able to respond to this objection, if he
relied only on the idea of a standard (paradeigma) for a thing’s being such
as to be F. On the other hand, he is able to respond to it, if he relies on the
view that an adequate standard must be unitary – that is, if he relies on the
unity requirement for the adequacy of an account of what a thing is. For
the objection of the imaginary Hippias was based, precisely, on the
supposition that a standard for a thing’s being F need not be unitary but
may, on the contrary, be disjunctive.
We may, therefore, suppose that Plato’s reasoning for Premise  is this:
I. An adequate and true account of what a thing, F, is provides a
unitary and true standard (paradeigma) for a thing’s being such as to
be F.
II. A unitary and true standard for a thing’s being such as to be F, must
be changeless, or else that of which it is a standard (i.e., F) would not
be the very thing it is.
III. If an adequate and true standard is changeless, then that which it
designates is changeless.
Therefore,
IV. If an account of what a thing is is adequate and true, then that which
it designates is changeless. (= Premise )
We may conclude that Plato’s claim, which says that Forms are changeless,
can be defended on the basis of the claim that a Form is a unitary standard
(a unitary paradeigma) for a thing’s being such as to have a certain quality.
And we saw that there is good reason, based in Plato’s text, to think that
this is how Plato defends it (see Phaedo c–d, as discussed above).
. Whether the eternality of Forms can be defended

in a similar way
The answer might seem a simple YES. We may suppose that, for Plato,
there is a standard for a thing’s being such as to be F, for as long as there is
a world with things in it of which the question can be posed: Is this thing,
O, such as to be F, or not? Supposing, therefore, that the world is eternal,
standards (paradeigmata) will be eternal, too. And since Forms are such
paradeigmata, Forms are eternal.
The matter is not, however, so simple. It may be questioned that Plato
thinks the world is eternal. This is a difficult issue, the answer to which

 Plato’s Essentialism
depends on how we read the Timaeus: on one reading of this dialogue (the
so-called literal reading), the world was created in time; on a different
reading (the so-called non-literal reading), the world is eternal looking
back in time. On any reading of the Timaeus, it is not clear that Plato
thinks the world is eternal looking forward in time – that is, that he thinks
the world is indestructible. On the contrary, there is reason to think that
Plato thinks the world is destructible, looking forward in time.
All we can safely conclude on this line of reasoning, therefore, is that
Forms are as long-lasting as there is a world that requires standards for
determining what it is like, or what things in it are like. This, it will be
said, falls short of demonstrating that Forms are eternal. Indeed, it may be
said that, even if we suppose that, for Plato, the world is eternal, and that it
is eternal both looking back and looking forward in time, and hence that
Forms are coetaneous with the world, still this falls short of demonstrating
that Forms are, in the relevant sense, eternal. For, it may be said, the claim
that Forms are eternal is understood by Plato to imply that there would
always have been be Forms, even if there had not, contrary to fact, been a
world that requires standards for determining what it is like.
It is a good question whether Plato anywhere speculates about the
implications for Forms of the supposition that, contrary to fact, the world
did not exist. I do not want to insist on this point. I think it may be
conceded that the claim that Forms are eternal, on a certain understanding
of this claim, cannot be defended on the basis of the claim that Forms are
unitary paradeigmata, or, in general, on the basis of Plato’s account of
essence that has emerged from the dialogues that are commonly supposed
to be before the Phaedo and, in particular, his view of what an account of
what a thing is must be like, to be an adequate and true account.
If, at this point, one wants to argue that the claim that Forms are
eternal, on a certain strong notion of eternality, can only be defended on
the basis of the claim that Forms are separate from physical things, then
I am happy to concede the point. This concession does not take away from
what we have already established: first, that the claim that Forms are
changeless cannot be defended on the basis of the claim that Forms
are separate from physical things; secondly, that the claim that Forms are
changeless can be defended on the basis of the claim that Forms are unitary
paradeigmata and, thirdly, that the same defence can be given of the claim
that Forms are, if not strictly eternal, then, at any rate, as long-lasting as
there is a world that requires standards for determining what it is like, or
what things in it are like.

 
The relation between knowledge and enquiry

in the Phaedo
What is the relation, in Plato, between the account of knowledge and the
account of enquiry? Is the account of knowledge independent of the
account of enquiry? Taking up this question is a large task, not least
because, while so much work has been done on Plato’s account of
knowledge, and quite a lot is being done on Plato’s account of
enquiry, I know of only the odd critic who has considered the relation
between the two. It is remarkable that critics have generally treated the
two topics – Plato’s account of knowledge and his account of enquiry – as
if they were separate topics, which suggests they have been tacitly sup-
posing that, for Plato, the account of knowledge is independent of the
account of enquiry.
The following is what is at issue, stated in general terms. To think that
Plato’s account of knowledge is independent of his account of enquiry is to
think that, according to Plato, it is possible to give an account of what
knowledge is that makes no reference to enquiry. To deny this is to
propose that, according to Plato, first, knowledge comes about through
enquiry, either as its end-state or in the course of the enquiry (or in both

In addition to Politis (), I am thinking, most especially, of Hintikka, who writes: ‘there does not
exist, and there cannot exist, a fully self-contained theory of justification independent of theories of
discovery’ (, ). Hintikka takes this view to be Platonic in origin and orientation: ‘Hence it is
seen that Plato had in one important respect the same focus as we: the quest for knowledge rather
than the justification of beliefs. The definition of knowledge was thought of by Plato as a means for
this quest’ (; and see esp. ch. ). One might have expected Fine () to have, at the very least,
raised this question, in a book on a problem about enquiry in the Meno, and in view of her expertise
on Plato’s epistemology. But she does not.

I am thinking of such critics as White (); Ferejohn (); Taylor (); Wedgwood ().
They typically focus on Plato’s account of knowledge and treat of his thinking about enquiry, if at
all, largely in relation to the so-called paradox of enquiry in the Meno. I note, however, that Taylor
argues that, for Plato, knowledge is the product of enquiry: ‘Those themes may be summed up as the
doctrine that the aim of inquiry is to achieve systematic understanding of the intelligible principles of
reality’ (, ); indeed, he says that ‘Given the metaphysical theory of Forms as the basic things
that there are, Forms are the primary objects of inquiry’ (, emphasis added).


 Plato’s Essentialism
ways); and, secondly, it is not possible to give an account of what the
knowledge is that comes about through enquiry without making reference
to the process of enquiry through which it comes about. On the indepen-
dence view, what knowledge is is independent of the process of enquiry, if
any, through which it comes about; on the dependence view, what
knowledge is is dependent on the process of enquiry through which it
comes about.
To make the project and task of this chapter more manageable and
focused, on this apparently first attempt at treating of it, I shall concentrate
on the Phaedo and the following two questions regarding the relation
between knowledge and enquiry in this dialogue:
Q When, in the Phaedo (a; prepared for at a–c), Plato says that
sense-perception is necessary for thinking of, and is the one and only
means by which we may arrive at the thinking of, certain
essences and Forms, is this claim only about the thinking involved
in searching for and enquiring into these essences and Forms, or is it
also about the thinking involved in knowing these essences
and Forms?
Q If the claim, that sense-perception is necessary for thinking of, and is
the one and only means by which we may arrive at the thinking of,
certain essences and Forms, is both about knowledge and about
enquiry, is it primarily about knowledge and only as a consequence
about enquiry, or is it primarily about enquiry and only as a
consequence about knowledge?
I think these questions are of particular interest in their own right and that
much depends on, and seriously different consequences regarding Plato’s
epistemology follow from, how we answer them. They provide a good way,
it seems to me, of articulating and getting an initial handle on the larger
task and project of determining how Plato’s account of knowledge is
related to his account of enquiry.
In response to Q, I argue that Plato’s claim, which says that sense-
perception is necessary for thinking of, and is the one and only means by
which one may arrive at the thinking of, certain essences and Forms, is
both about enquiry and about knowledge. In response to Q, I argue that
this claim is primarily about enquiry and only as a consequence is it
about knowledge.
These conclusions stand opposed to the view, common among critics,
which says that, in the Phaedo (esp. e–c), Plato argues that knowl-
edge, or the knowledge of certain Forms, is a priori and independent of

Knowledge and Enquiry in the Phaedo 
sense-perception. I argue that this view is incompatible with Plato’s claim
(at a) that sense-perception is necessary for thinking of, and is the one
and only means by which one may arrive at the thinking of, certain
essences and Forms. I consider two traditional and standard ways of
attempting to reconcile the a claim with the view that the knowledge
of Forms is a priori and argue that these attempts fail.
I propose a very different account of why, at e–c (and again at a),
Plato thinks that certain Forms cannot be perceived by the senses and can
only be known by reasoning. On this account, this view is a consequence
of, and it is justified by, the view that the Forms of such qualities cannot be
defined by example and exemplar, that is, by pointing to a sense-
perceptible thing that conspicuously exemplifies the quality. I argue that
this account of why certain Forms cannot be perceived by the senses
provides a good way of understanding Socrates’ famous claim, at Phaedo
, that we must not examine things directly by the senses, lest we become
blinded in our soul, but must examine them through logoi, that is, in
words and verbal accounts. I conclude that, on this account of why certain
Forms cannot be perceived by the senses, we need not suppose that the
knowledge of such Forms is a priori.
. Two seemingly conflicting epistemological claims

in the Phaedo
I want to take up the following problem in reading the Phaedo and what
we would refer to as its epistemology. At e–c (I shall refer to this as
‘the earlier epistemological claim’; it is repeated at a) Socrates says that
the philosopher will, as far as possible, avoid relying on sense-perception in
the search for, and in the acquisition and possession of, knowledge, and
will rely, for enquiring and knowing, only on thinking, reasoning and what
the soul does itself by itself and not in communion with the senses.
However, at a– (prepared for at a–c; I shall refer to this as ‘the
later epistemological claim’) he says that sense-perception is necessary for
thinking of, and is the one and only means by which one may arrive at the
thinking of, certain Forms (here the Form of the quality, equal). There
appears to be a tension, or contradiction, between the two claims.
Certainly, there appears to be a contradiction, if the earlier claim is
understood as meaning or implying that knowledge is a priori and inde-
pendent of sense-perception, which is how many critics have understood

Such critics include Annas (); Bostock (); Taylor (); Wedgwood ().

 Plato’s Essentialism
it. For, on the face of it, the later claim directly implies that knowledge is
dependent on sense-perception, since it says that thinking about an object
of knowledge, such as a Form, is dependent on sense-perception, and
since, for Plato, knowing an object is, precisely, the successful result of
thinking of it and reasoning about it.
Critics who think that the earlier epistemological claim means or implies
that knowledge is a priori and independent of sense-perception have
standardly sought to avoid the contradiction in either, or both, of the
following ways. One way is to argue that the later epistemological claim
need not be understood as saying that sense-perception is necessary for the
thinking of a Form that is involved in, specifically, the knowing of it; it
need only be understood as saying that sense-perception is necessary for
lesser ways of thinking of a Form. Another way is to argue that the later
epistemological claim need only be understood as stating that sense-
perception is causally necessary for thinking of a Form; it need not be
understood as stating that sense-perception is epistemically necessary for
thinking of a Form, either in sense-perception’s being part of the reasons
and the justification for thinking that there is such a Form or, in general, in
its being part of the account of what it is to know a Form.
I shall consider the latter way of reconciling Plato’s two seemingly
conflicting epistemological claims at the end of Section .. In regard to
the former way, I doubt that it will succeed. One problem is that the earlier
epistemological claim does not say only that the philosopher will, as far as
possible, avoid relying on sense-perception in the acquisition of and the
possession of knowledge; it says also that she will, as far as possible, avoid
relying on sense-perception in the search for knowledge. (I shall demon-
strate this presently.) Compared to the thinking involved in positively
knowing a Form, the thinking involved in searching for a Form is a lesser,
because less complete and perfect, way of thinking of a Form. It is not true,
therefore, that the earlier epistemological claim is specifically about know-
ing a Form, in contradistinction to other, lesser, ways of thinking of it.
Another problem with this standard way of attempting to reconcile Plato’s

My impression is that critics have tended to put more emphasis, by far, on the earlier epistemological
claim. A good example is Bostock (), who says a great deal about the earlier, and hardly
anything about the later, epistemological claim.

Bostock () is a good example of both ways. Notable also is how scholars writing on Aristotle on
science take this view of Plato’s epistemology for granted, which presents Aristotle as the innovator.
Thus Karbowski () rehearses it as being familiar that, for Plato, sense-perception is only a cause
of and a stimulus for the ascent to rational and a priori knowledge and is not part of the justification
that constitutes rational knowledge.

Knowledge and Enquiry in the Phaedo 
two epistemological claims is that, while it is true that at a– the later
epistemological claim is formulated in terms of thinking (ennenoēkenai and
ennoēsai, from ennoein), in the passage that immediately precedes it and
prepares for it the claim is formulated both in terms of thinking and in
terms of knowing (epistasthai; see the use of epistēmē at b, followed by
enenoēsamen, again from ennoein, at b; and see the use of epistēmē at
c). It is not true, therefore, that the later epistemological claim is
specifically about the thinking of a Form, in contradistinction to
knowing it.
It can be shown, by a careful reading of e–c, that the earlier
epistemological claim is both about enquiry and about knowledge.
Socrates starts (e–a) by stating that, as it seems to him, the person
who has dedicated his life to philosophy may be well-disposed and hopeful
in the face of death, that he will encounter good things there; and he
promises to demonstrate to Simmias and Kebes that this is so. His
reasoning in the demonstration is that it would be absurd (atopon,
a), for one who has spent his life trying not to be attached to the
body, as far as it is possible for a human being not to be so attached, then,
when faced with dying – dying being, precisely, the detachment of the soul
from the body and its coming to exist independently of the body – to fear
and to be ill-disposed to dying. The critical premise in the demonstration,
the three of them agree, is that the philosopher is one who lives in this way.
The one part of Socrates’ reasoning for this premise is that the philosopher
will care for the soul over the body, its pleasures and luxuries, and will care
for the body only to the extent that doing so is strictly necessary for living
(c–a).
The other part of Socrates’ reasoning (a ff.) is that the philosopher’s
supreme aim, which is knowledge and wisdom – and this crucially involves
the knowledge of essences and Forms (d–e) – depends on what the
soul does itself by itself, that is, reasoning (logizesthai) and thinking
(dianoein), and not on what it does in communion with the body, that
is, seeing, hearing and in general sensory perception. What is remarkable is
that, in articulating this claim, Socrates marks clearly that his claim is just
as much about enquiry as it is about knowledge. Having stated that
essences and Forms cannot be perceived by the senses and can be known

See Rowe (, ch. ), for a good defence of the view that the overall aim of the Phaedo is to argue
that the good person, and especially the person whose goodness is based on philosophy, has reason to
be cheerful and hopeful in the face of death.

koinōnein with dative; also ‘through’, dia with genitive, and ‘with’, meta with genitive, and ‘by’, using
simply the instrumental dative. All these locutions are used, apparently, without distinction.

 Plato’s Essentialism
only by thinking and reasoning (d–a), he immediately says that it
is in just this way, by the use of thinking (tē[i] dianoia[i] chrōmenos, a),
that one will ‘try to hunt after’ (epicheiroi thēreuein, a) the things that
are (ta onta; here, the essences and Forms) and shortly later again speaks of
this as ‘the hunt for that which is’ (tēn tou ontos thēran, c). Of course,
to hunt after a thing, indeed to try to hunt after it – apparently, even the
trying may be difficult – is aimed at the possession of that thing, through
the acquisition of it (see ktēsis at a, in the original question: ‘What
about the possession of knowledge?’, Tí de dē peri autēn tēn tēs phronēseōs
ktēsin;). What is remarkable is that Plato formulates the claim, that
knowledge depends on reasoning rather than on sense-perception, both
in terms of knowledge and in terms of enquiry. In case we might think of
the hunting metaphor as mere metaphor, reference to enquiry, expressly so
called – zētēsis – is woven into this original question:
What about the possession of knowledge? Is the body an obstacle [to the
possession of knowledge], or not, if one were to include it and the
communion with it in the enquiry? (en tē[i] zētēsei koinōnon
sumparalambanē[i], a–b)
This reference to enquiry, expressly so called, is repeated at d– (en tais
zētēsesin).
In sum, what Socrates argues here (at e–c, i.e., the earlier episte-
mological claim, repeated at a) is that the philosopher will, as far as
possible, avoid relying on sense-perception both in the search for and in the
acquisition and possession of knowledge and that the philosopher will rely,
both for enquiring and for knowing, on thinking, reasoning and what the
soul does itself by itself and not in communion with the body and
the senses.
If, therefore, we ask, as we must if we want to understand what all this
means, what is meant by the idea of the soul’s operating not together with
the body and the senses but ‘itself by itself’, we need to ask not only what is
meant by the soul’s acquiring and possessing knowledge, itself by itself and
not together with the body and the senses, but also what is meant by the
soul’s searching for knowledge, and doing so in the way in which a
philosopher, like Socrates, searches for knowledge, itself by itself and not
together with the body and the senses. We would expect the answers to
these two questions to be closely related and a single account to provide for
both answers. Since the epistemological claim is formulated as being at
once about knowledge and about enquiry, we would not want Plato to

Knowledge and Enquiry in the Phaedo 
mean one thing when he says that in the search for knowledge the soul
operates itself by itself and to mean something quite different when he says
that in the acquisition and possession of knowledge the soul operates itself
by itself. I say this because I fear that, while we, modern critics, may take it
to be familiar and sufficiently clear what it means to say that knowledge is a
priori and independent of sense-perception, we do not at all have a clear
idea what it means to say that enquiry, and in particular philosophical
enquiry of the kind referred to in the Phaedo, is a priori and independent
of sense-perception. I shall return to this point presently.
In several passages in the dialogue, Plato indicates the kind of philo-
sophical enquiry he has in mind. As is pointed out in the passage in which
Plato articulates the present epistemological claim, this is the enquiry into
and the search for what certain things are:
I am referring to all those things, such as largeness and health and strength
and, in one word, about the essence of all those things: that which each of
them is. (kai tōn allōn, heni logō[i], hapantōn tēs ousias ho tugchanei hekaston
on; d–e)
Other notable and familiar passages include c–d and d– (which
I quoted in Chapter ). It is clear from such passages that the philosophical
enquiry referred to by Plato, when he has Socrates say that philosophical
enquiry ought to be conducted, as far as possible, by the soul itself by itself
and not with the body and the senses, is the kind of enquiry that he has
exhibited in other dialogues and made typical of Socrates and his practice
of philosophy, and is again exhibiting here, in the Phaedo.
I confess I have no clear idea what it means to say that such enquiry is a
priori; not least since the term a priori, as used in contemporary episte-
mology, is typically applied not to enquiry and the search for knowledge
but to knowledge and its acquisition. If, following one or another current
account of a priori knowledge, we say that one knows a proposition a
priori, if either one’s justification or evidence for believing that it is true, or
the way in which one has acquired the belief through a reliable mechanism
of belief acquisition makes no reference to and does not involve sense-
perception, or sensory experience, then, it seems to me, it is not at all clear
what it means to say that the search for an answer to a question is a priori.
It is true that such a search will involve all kinds of beliefs, but the
distinctive attitude of mind in such a search is not that of belief, that is,
the kind of belief that something is the case that a person has when she
knows, and (on some accounts of knowledge) believes that she knows, that

 Plato’s Essentialism
it is the case. The distinctive attitude of mind, when we are searching for
an answer to a question, is that of asking a question and doing so for
particular reasons, proposing an answer, proposing a conflicting answer,
wondering which answer is the better, etc.
This is the kind of dialectic we find in the Phaedo as the dialogue
progresses, just as we do in many other dialogues, between Socrates and
his answers and Simmias and Kebes and their opposed answers, in
response to the question of the relation between the soul and the body:
Can the soul exist and think without the body? It is true, we may say
that Simmias (and Kebes?) are inclined to believe that the soul is an
attunement of the body, whereas Socrates does not believe this. But
neither Simmias, nor Kebes, nor Socrates knows this, or believes he
knows it. Rather, they are proposing different and conflicting answers
about the nature of the soul, while being aware that they do not know
these answers to be true; and they do not believe these answers to be true,
if by belief we mean an element in the mental state of knowing, or even
claiming to know.
This is not to say that we cannot imagine what Socrates may have in
mind when he says that the philosophical search for knowledge ought, as
far as possible, to be conducted by the soul itself by itself and not in
communion with the senses. He may mean, and I would like to propose
that he does mean, simply and in a way Plato expects to be readily familiar
to the reader who is familiar with the characteristic enquiries both in the
Phaedo and dialogues before it, that: in this joint enquiry, among a small
group of people, addressed to a ti esti question, or a ti esti question and an
associated whether-or-not question, the participants are conducting the
enquiry basically by talking to themselves and to each other. In particular,
they do not set out to answer the question ‘What is F?’ by pointing to a
thing that is conspicuously F and hence capable of serving as an example
and exemplar of a thing that is F and, therefore, as a standard (paradeigma)
of a thing’s being F. This, we may recall, was the way in which Hippias in
the Hippias Major proposed to answer the question ‘What is beauty?’: by
pointing to a girl (or a horse or a lyre) of conspicuous beauty and hence
capable of serving as an example and exemplar of a beautiful thing and,
therefore, as a standard for a thing’s being beautiful. I shall return to this
suggestion presently.

For the relation between ti esti questions and whether-or-not questions in Plato, see Politis .

See Chapter .

Knowledge and Enquiry in the Phaedo 
. Why sense-perception is necessary to think of a Form

I turn now to the later epistemological passage, where Plato says that sense-
perception is necessary for thinking of, and is the one and only means by
which one may arrive at the thinking of, certain Forms:
We also agree on this: we haven’t arrived at the thinking (ennenoēkenai) of it
[i.e., of the Form of equal] from anywhere (mē . . . allothen), nor is it
possible to think of it (mēde dunaton einai ennoēsai), except from (all’ ē
ek) seeing or touching or some other of the senses – I am counting all these
as the same. (a–; trans. Gallop; adapted)
A reader presented with this passage, if she did not already believe that
Plato believes that knowledge of a Form is a priori, would naturally
conclude that Plato thinks that sense-perception is necessary for the
knowledge of a Form and is part of the account of this knowledge. For
the passage says that thinking about an object of knowledge, such as a
Form, is dependent on sense-perception, and, for Plato, knowing an object
is, precisely, the successful result of thinking of it and reasoning about it.
But this reading is not available, if one believes that, for Plato, knowledge
of a Form is a priori – a view that, we saw, one might arrive at on the basis
of the earlier epistemological passage in the Phaedo.
Why does Plato think that sense-perception is necessary for thinking of,
and is the one and only means by which one may arrive at the thinking of,
certain Forms such as the Form of the quality equal? And why is it
appropriate – if it is appropriate – to suppose that this claim is, in the
first instance at any rate, about the thinking involved in the search for these
Forms? These are not easy questions; and while the Phaedo assists us in
considering them, it is difficult, I believe, to consider them fully or
properly without invoking a particular, and familiar, passage from the
Republic, VII. a–a, the so-called summoners passage. I shall,
presently, consider whether the same idea, of the senses summoning us,
in a certain particular way and for certain particular reasons, to think of
Forms, can be found in the Phaedo.
Difficult and controversial as this Republic passage is, its principal point
is that certain sensory perceptions, especially if they are in conflict with
each other, call upon or summon the soul to go beyond sense-perception
and to have recourse to reasoning (logismos, logizesthai) and reason and
thought (noēsis, nous, dianoia), indeed to distinguish things into objects of

I shall properly consider this passage later, in Chapter .

 Plato’s Essentialism
perception and objects of reasoning (horata versus noēta, c; cf.
Phaedo a). Such conflicting sensory perceptions have this power, to raise
the soul above sense-perception, because they present what appear to the
senses to be the same thing as having, at the same time, opposite qualities,
such as, for example, the same thing’s, a finger’s, being both hard and soft
at the same time.
How and why a sensory perception of co-present opposites calls upon
the soul to go beyond sense-perception is a difficult and disputed ques-
tion. I am inclined to think that this is, in crucial part, because, if the
soul were to rely simply on sense-perception for the corresponding judge-
ments, it would not be able to rule out that such judgements contravene
the principle of non-contradiction or non-opposites. This is the principle
that says that the same thing cannot be both F and opposite-to-F at the
same time, in the same respects, in the same relations, etc., which Plato has
formulated earlier in the Republic (IV. –). This, I think, explains
very well why it is said that the soul is put in a state of aporia as a result of
its realisation of, and reflection on, such conflicting appearances (see
a–) and that it is this state of aporia that forces it to search for
(zētein), in particular, the answer to a ti esti question, such as, above all, the
question ‘What is oneness or unity (to hen)?’ (see e–a, esp. ti pot’
esti auto to hen). The realisation that a cognitive state that one is in may
involve a contradiction, and that one does not immediately have the
resources to avoid the possibility that it does contain a contradiction, is a
source of aporia in one, if anything is.
Most important, in the ‘summoners’ passage there is no suggestion that
when the soul has recourse to reasoning and reason, it immediately knows
something, such as an essence and Form; rather, the suggestion is, clearly,
that what the soul does in having recourse to reasoning and reason is ask
certain ti esti questions, including what is, apparently, the most important
and most difficult ti esti question: ‘What is oneness or unity (to hen)?’ The
soul recognises that such questions cannot be answered by relying on
sense-perception, they can only be answered by relying on reason
and reasoning.
It is true that, if we suppose that in the Phaedo, as in the Republic, Plato
thinks of sense-perception, and especially the phenomenon of a sensory
perception attributing opposite qualities to what appears to the senses to be
one and the same thing, as a necessary part of what motivates and justifies
the supposition that there are essences and Forms, which cannot be

I consider it in Chapter .

Knowledge and Enquiry in the Phaedo 
perceived by the senses, then we ought, at the same time, to be mindful of
the fact that the context in the Phaedo is different from that in the Republic
passage. The context in the Phaedo, but not in the Republic, includes the
idea of recollection; and the idea of recollection is associated in the Phaedo
with the idea of the soul’s pre-existing the body. But it seems plausible to
think that, irrespective of its association with the pre-existence of the soul,
the idea of recollection is associated in the Phaedo with the idea of this
peculiar phenomenon of sense-based co-presence of opposites being a
necessary part of what motivates and justifies the supposition of non-
sense-perceptible essences and Forms. For it seems that one function of
the supposition of recollection is to explain how certain of our sensory
perceptions, and the soul’s reflection on these, can call to mind certain
thoughts that are about essences and Forms: such sensory perceptions can
call to mind such Form-directed thoughts, because we already have Form-
directed thoughts in us in a latent and non-conscious way, and our
reflection on these sensory perceptions causes us to become aware of these
latent thoughts.
Irrespective of how exactly we understand the epistemological claim in
the Phaedo (a–), which says that sense-perception is necessary for
thinking of, and is the one and only means by which one may arrive at the
thinking of, certain Forms, such as the Forms of the quality equal, it is
clear that it is supposed to be a consequence of the following familiar
passage, which immediately precedes it (I have italicised the lines that
prepare for the epistemological claim at a–):
Consider, he said, whether this is the case: we say that there is something
that is equal. I do not mean a stick equal to a stick or a stone to a stone, or
anything of that kind, but something else beyond all these, the Equal itself.
Shall we say that this exists or not? – Indeed we shall, by Zeus, said
Simmias, most definitely. – And do we know what this is? – Certainly. –
Whence have we acquired the knowledge of it? Is it not from the things we
mentioned just now, from seeing sticks or stones or some other things as equal
that we come to think of that other which is different from them? Or doesn’t it
appear to you to be different? Look at it also this way: do not equal stones
and sticks sometimes, while remaining the same, appear to one as equal and
to another as unequal? – Certainly they do. – But what of the equals
themselves? Have they ever appeared unequal to you, or Equality to be
Inequality? – Never, Socrates. – These equal things and the Equal itself are
therefore not the same? – I do not think they are the same at all, Socrates. –
But it is definitely from the equal things, though they are different from the
Equal, that you have derived and grasped the knowledge of equality? – Very
true, Socrates. (a–c)

 Plato’s Essentialism
While it is notoriously controversial how we should understand and
analyse the core argument in this passage – this is the argument for the
conclusion that the Form of the quality, equal, is distinct from, and not
identical with, any (pair of ) things that are conspicuously equal to the
senses – it seems clear that the passage contains a number of elements
similar to those in the ‘summoners’ passage of the Republic. The elements
include, the following: that the same things appear to our senses to be both
F (here, equal) and opposite-to-F (here, unequal); that our realisation of
and reflection on such conflicting appearances is what leads us to
suppose that there is such a thing as the Form of equal and that this
Form is distinct from and not-identical with the things we perceive by the
senses as equal and that we can know, and even that some of us do know,
this Form, even though (as was expressly said earlier, at d–e, and will be
repeated presently, at a) the Form is not something that can be per-
ceived by the senses, rather, it ‘can be grasped only by the reasoning power
of the mind’ (ouk estin hotō[i] pot’ an allō[i] epilaboio ē tō[i] tēs dianoias
logismō[i], a–).
It is true that the Phaedo passage (a–c) is about one quality, equal,
whereas the Republic passage, on which we have drawn to help understand
the Phaedo passage and its argumentative context, is about a different
quality, one. But it is notable that, in Republic VII. d–a,
the issue under consideration is the use of visible celestial patterns as
exemplars and standards, paradeigmata, of, precisely, equalities, and the
adequacy of such sense-perceptible exemplars and standards of equal (see
esp. the isōn at a).
This is, I think, sufficient (for present purposes) to indicate why Plato
thinks that sense-perception is necessary for thinking of, and is the one and
only means by which one may arrive at the thinking of, certain Forms such
as the Form of the quality equal. Plato thinks this because he thinks that it
is in response to certain questions, which articulate problems and aporiai,
and which we are motivated and justified in asking because of the peculiar
content of certain of our sensory perceptions and, in particular, those
sensory perceptions that attribute opposite qualities to what appears to
the senses to be the same thing at the same time, that we are motivated and
justified to suppose the following. What this motivates and justifies us to
suppose, is this: There are certain essences and Forms, and they cannot be
perceived by the senses but can be searched for and, all going well, known
only by reasoning. This also shows why it is appropriate to suppose that

I consider this argument in Chapter .

Knowledge and Enquiry in the Phaedo 
the claim, which says that sense-perception is necessary for thinking of,
and is the one and only means by which one may arrive at the thinking of,
certain Forms, is, in the first instance at any rate, about the thinking
involved in the search for and enquiry into these Forms. For, what these
problems and aporiai motivate and justify us to do is not, certainly not
immediately, to know these essences and Forms, but rather, search for
them and do so by an intellectual search, a search based on reasoning.
If sense-perception is, in this peculiar way, involved in our thinking that
there are certain questions, which articulate problems and aporiai and that
(i.e., the questions and the aporiai they articulate) motivate and justify us
to suppose that there are essences and Forms, and these cannot be
perceived by the senses, then we may indeed sum up this by saying that
sense-perception provides the stimulus for searching for such essences and
Forms. At the same time, however, we must recognise just how sense-
perception provides this stimulus. For it not only provides a causal
stimulus, it also provides part of the reason for thinking that there are
such questions, problems, aporiai, and what they are and how they ought
to be articulated. The stimulus it provides is, indeed, causal, but it is no
less epistemic. If sense-perception only provided a causal stimulus, one
might argue that when, in the later epistemological claim, Plato claims that
sense-perception is necessary for thinking of a Form, he need not be
understood as implying that sense-perception is epistemically involved in
this thinking, he need only be understood as implying that sense-
perception is a causally necessary condition for this thinking. But I don’t
see how one can argue this, once it is recognised in what peculiar way
sense-perception stimulates us to suppose that there are essences and
Forms that cannot be perceived by the senses: it does so because our
reflection on certain of our sense-perceptions motivates and justifies us to
ask certain questions and to articulate certain questions and aporiai in
response to which we suppose that there are essences and Forms, and
Forms that cannot be perceived by the senses.
. How to reconcile the two epistemological claims

in the Phaedo
We have seen that in the Phaedo Plato defends two epistemological claims,
which appear to be in tension with each other. According to the earlier
claim (e–c), knowledge and enquiry, and especially the knowledge of
and enquiry into certain essences and Forms, are the concern of the
soul and what it does itself by itself and not in communion with the

 Plato’s Essentialism
senses; but according to the later claim (a; prepared for at a–c), sense-
perception is necessary for, and is the one and only means by which one
may arrive at, the thinking of an object of knowledge such as these essences
and Forms.
It is a good question how serious the tension is between the two claims.
One might urge that the tension is merely apparent and that this is evident
once we realise, and take sufficiently seriously, that, in formulating the
earlier claim, Plato is careful to formulate it, repeatedly and consistently, as
saying that the knowledge of, and enquiry into, certain essences and Forms
is the concern of the soul and what it does, as far as possible, itself by itself
and not in communion with the senses. For one may urge that the
repeated qualification, ‘as far as is possible’, serves to allow for the view
that sense-perception contributes to this knowledge and enquiry, even that
it is necessary for, and the one and only means by which one may arrive at,
this knowledge and enquiry. On the other hand, one may think that this
qualification is not sufficient to reconcile the two claims. What is clear is
that, if, as many critics have, one thinks that in and through the earlier
claim Plato is arguing that knowledge, or the knowledge of certain Forms,
is a priori, then certainly there appears to be a tension, indeed a contra-
diction, between the two claims.
How can we reconcile Plato’s two epistemological claims in the Phaedo?
I have argued that the later epistemological claim is motivated and justified
by Plato’s view (call it The sense-based, aporia-based justification for Forms),
which says that:
It is in response to certain questions, which articulate problems and aporiai, and
which we are motivated and justified in asking because of the peculiar content of
certain of our sensory perceptions, in particular those sensory perceptions that
appear to attribute opposite qualities to the same thing at the same time, that we
are motivated and justified to suppose that there are certain essences and Forms,
which cannot be perceived by the senses but which can be searched for and, all
going well, known only by reasoning.
And I have argued that the earlier epistemological claim in the Phaedo is
motivated and justified by Plato’s view (call it The non-perceptibility of
Forms) that

kath’ hoson dunatai: e, c, c (kata to dunaton). Remarkably, and as Dillon (, ch. ,
–) perceptively remarks, this little, but important, qualification is there in the reference to
‘becoming like God in so far as that is possible’ (homoiōsis theō[i] kata to dunaton; Dillon translates
‘assimilation to God’) in Theaetetus b, and likewise in the similar reference in Republic X. a
(homoiousthai theō[i] eis hoson dunaton athrōpō[i], ‘to become like God as much as is humanly
possible’).

Knowledge and Enquiry in the Phaedo 
it is not possible to perceive by the senses what certain qualities, such as equal in
the Phaedo (a–a) and one in the Republic (e–a), are, that is, it is
not possible to perceive by the senses their essences and Forms.
It may look as if these two claims (i.e., The sense-based, aporia-based
justification for Forms and The non-perceptibility of Forms) are readily
compatible and that there is no reason to think that they are not. But
things are not so simple. Whether these two claims are compatible depends
on what account of knowledge, and of enquiry, and of the relation
between the two, are involved in and are behind them.
Suppose we think that Plato is committed to The non-perceptibility of
Forms because he thinks that, Forms being what they are – supernatural
entities – the knowledge of them is a priori. This would be
compatible with The sense-based, aporia-based justification for Forms, if
the supposition in this, which says that ‘There are certain essences and
Forms that cannot be perceived by the senses but that can be searched for
and, all going well, known only by reasoning’, were only about
knowledge and not also about other ways of thinking; but we have found
that this is not so.
One problem is that it is not at all clear what it means for enquiry to be
a priori. Certainly, none of the accounts I know of a priori knowledge can
be automatically or straightforwardly extended to a priori enquiry. The
accounts of a priori knowledge are, typically, about what it is to hold a
belief a priori; and, typically, they say that a belief is held a priori, if either
the justification or evidence because of which one holds it makes no
reference to and does not involve sense-perception or experience, or the
way one has acquired the belief, through a reliable mechanism of belief
acquisition, makes no reference to and does not involve sense-perception
or experience. It seems clear that the reason why such familiar accounts of
a priori knowledge are focused on how, for what reasons or causes (or
both), one holds a belief, is the background supposition that belief is a
necessary condition for knowledge. But enquiry is not primarily occupied
with beliefs, or beliefs that are part of knowledge. Enquiry is, if anything,
occupied with questions; and the philosophical enquiry with which Plato is
concerned here is occupied with questions that articulate problems and
aporiai. Until and unless we have a proposal about what enquiry is,
comparable to the proposal that belief is a necessary condition for knowl-
edge, we have, as far as I can see, no secure way of extending the account of
a priori knowledge to an account of a priori enquiry, and we deceive
ourselves if we think we know what it means for enquiry to be a priori.

 Plato’s Essentialism
One might think that the available accounts of a priori knowledge can
simply be extended to a priori enquiry; this can be done, it may be
thought, purely and simply by saying that a particular enquiry is a priori
if, and only if, the proposition corresponding to the question with which
the enquiry is occupied, and to which it is addressed, can only be known a
priori. But I doubt that this will work. One problem is that, to a question,
and even more so to a question that articulates an aporia, there corresponds
not one but very many propositions. This is evident, if we consider such
questions as whether or not virtue can be taught, or whether it is better to
enter into an intimate friendship with a person in love with one or, on the
contrary, with one not in love with one. It is manifest that, to address
questions such as these, one will have to determine the truth or falsity of
not one but very many propositions. Indeed, it is questionable whether it
can determined, in advanced of the enquiry, what propositions one will
have to determine the truth or falsity of to conduct the enquiry and bring
it to a successful end. This implies that it is possible that some of these
propositions will be knowable only a priori, some only a posteriori, and
(possibly) some in both ways – if, that is, we are at all prepared to think
that propositions can be neatly distinguished according to these epistemo-
logical categories and if we think we know what the propositions are that
are objects of the enquiry and of the question that sets off the enquiry.
Let me explain with an example from the Phaedrus. The original
question that sets off the enquiry in that dialogue is: Whether it is better
to enter into an intimate friendship with a person in love with one or, on the
contrary, with one not in love with one. What propositions do we need to
consider in order to answer this question? It is hard to know this, in
advance of taking up and getting involved in the enquiry into this
question. Indeed, I suspect there is no way of deriving these propositions
from this question, in the way in which one might derive the relevant
propositions from the question ‘What time is it?’ (i.e., the proposition that
is the disjunction: ‘It is midnight, or it is : , or . . . or it is :
’, such that one, and just one, disjunct is true). Rather, it takes such a
thing as philosophical creativity, of the kind displayed in the Phaedrus, to
determine what propositions we need to consider in order to answer this
question. I submit that Plato identifies three such crux questions, which it
is necessary to consider to answer the original question. They are, first,
whether all irrationality depends, for its goodness or badness, on how it is
guided by reason or, on the contrary, some irrationality has its own source of
goodness; secondly, whether love is a form of irrationality and, thirdly,
whether at least some love is a form of irrationality that has its own source of

Knowledge and Enquiry in the Phaedo 
goodness. There are lots of propositions to consider here, for the purpose of
answering the original question and these consequent questions. One
might think that some of these propositions are knowable only a priori,
such as those relating to the general distinction between reason and
irrationality. Perhaps. Even so, why think that all the propositions will
be knowable only a priori? It seems evident that there are lots of propo-
sitions to consider here of which it would be implausible, indeed absurd, to
think that they are knowable only a priori.
Finally, even if one extends the account of a priori knowledge to an
account of a priori enquiry, it seems that The sense-based, aporia-based
justification for Forms is incompatible with thinking that enquiry is a priori.
What The sense-based, aporia-based justification for Forms says is that the
supposition, which says that ‘There are certain essences and Forms, which
cannot be perceived by the senses but which can be searched for and, all
going well, known only by reasoning’, is motivated and justified as a
response to certain questions, which articulate problems and aporiai, and
which we are motivated and justified in asking because of the peculiar
content of certain of our sensory perceptions. But, whatever it means to say
that one asks a certain question a priori, it ought to be clear that, if we give
a meaning to this idea on the analogy of what it means to say that one
holds a certain belief a priori, we cannot say that one asks a question a
priori, if what motivates and justifies one in asking it is the peculiar
content of certain of one’s sensory perceptions.
I conclude that, if we suppose that Plato is committed to The non-
perceptibility of Forms because he thinks that the knowledge of Forms is a
priori, then there is no way we can reconcile his commitment to The non-
perceptibility of Forms with his commitment to The sense-based, aporia-based
justification for Forms, once we recognise that The sense-based, aporia-based
justification for Forms is just as much about enquiry as it is about knowledge.
Is there another way of understanding why Plato is committed to The
non-perceptibility of Forms, and a way that allows us to recognise that this
commitment is compatible with his commitment to The sense-based,
aporia-based justification for Forms? I want to propose that Plato is com-
mitted to The non-perceptibility of Forms purely and simply because he
thinks that (The non-definability of Forms by example and exemplar):
What certain qualities, F (such as equal or one), are, their essence and Form,
cannot be specified by example and exemplar; that is, by pointing to a thing that
is conspicuously F (i.e., conspicuously to the senses) and, therefore, can serve as
an exemplar of a thing that is F and as a standard (paradeigma) for a thing’s
being F.

 Plato’s Essentialism
We have already (in Chapter , which in turn builds on Chapter )
defended the proposal that Plato is committed to The non-perceptibility
of Forms precisely because he is committed to The non-definability of certain
qualities by example and exemplar.
Now, if this is the reason why Plato is committed to The non-
perceptibility of Forms, that is, because he is committed to The non-
definability of certain qualities by example and exemplar, then his commit-
ment to The non-perceptibility of Forms does not imply that he thinks
essences and Forms (or, certain essences and Forms) can be known only a
priori. We saw this in Chapter ..
I recognise that this conclusion – it says that we should not take Plato’s
claim that (certain) essences and Forms cannot be perceived by the senses
and can be known only by reasoning to imply that he thinks (such)
essences and Forms can be known only a priori – will seem momentous
to many critics, in particular, to those critics who think that the distinction
between a priori and a posteriori knowledge provides a good handle, and
even the only secure handle we have at our disposal, on Plato’s epistemol-
ogy. For the purpose of this chapter, I shall not enter into this controversy.
My point is that there is a way, and a way that deserves serious consider-
ation, of understanding why Plato is committed to The non-perceptibility of
Forms, such that this commitment is readily compatible with his commit-
ment to The sense-based, aporia-based justification for Forms.
In regard to the Phaedo, it may be objected that, whatever about us
humans in our present, embodied condition, it is undeniable that, for the
soul in its disembodied state, which, according to Socrates during these,
his last hours, is the state aspired to by the philosopher, knowledge can
only be a priori. But I think this can be denied. The objection is correct,
on the supposition that the acquisition of knowledge by the soul in its
disembodied state is independent of what it saw in its previous, embodied
state. But we cannot assume that Plato is committed to this supposition.
He may, rather, think that the soul in its disembodied state will need to
recall what it once saw, if it is to arrive at any knowledge then. And there is
reason to think that this is what he thinks, if we suppose that he thinks
that, even in its disembodied state, the soul must search for knowledge if it
is to arrive at any; that is, even for the disembodied soul, knowledge is
based on enquiry. For he implies (at a–), as we have seen, that enquiry
requires sense-perception; and this, we have argued, is because enquiry is
stimulated, motivated and justified by our reflection on certain sensory
perceptions. So, unless Plato has an epistemology for the knowledge of the
soul in its disembodied state that is quite different from the epistemology

Knowledge and Enquiry in the Phaedo 
he has for the knowledge of the soul in its embodied state, we may suppose
that even in its disembodied state, the soul will have to recall some of what
it once perceived, if it is to search for knowledge – even in that state, when
we shall be free, or as free as we can possibly expect to be, of the
disturbances and the distractions that come from being embodied and
attached to the senses.
It may be said, what about the soul that has never been embodied? Is it
not undeniable that its knowledge can only be a priori? I am not inclined
to speculate. My impression is that Plato is not especially occupied with
such a soul in the Phaedo and that this is because in this dialogue he is
occupied with the human soul, and because Plato begins with the human
soul as each of us knows it from his or her condition, which is our present
embodied condition. This is not a mere impression. We saw that the
claim, at a–b, that knowledge depends on thinking and reasoning
rather than on sensory perception, is formulated as being just as much
about enquiry as about knowledge and that special emphasis is placed on
such enquiry being a task and undertaking, no less than the knowledge at
which it is aimed. It is plausible to suppose that beings whose soul was
never embodied would not be in need of such enquiry in order to know;
they would already know. If they, too, relied on recollection for their
knowledge, recollecting for them would be as easy as it is for us to recall
our family names.
One may object that I have ignored a passage in the Phaedo that clearly
tells in favour of the view that, for Plato, both knowledge and the search
for knowledge are a priori: this is d–a, when he argues that we ought
not to try to grasp things by our eyes and each of our senses, lest in doing
so we get blinded in our soul, but ought, rather, to investigate the truth of

I thoroughly agree with Rowe (, –), when he says: ‘Nowhere in the Phaedo is there any
commitment, on Socrates’ or anyone else’s part, to the idea that the objects of knowledge (the
forms) will be available on demand, as it were, after death, even to the “purified” soul. The myth
makes no provision for it, and in fact doesn’t mention forms at all; and Socrates on his own account
associates access to, or approaching, true reality – the unseen, aïdes – specifically with this life, not
the next.’ And I agree no less, when he says: ‘My conclusion about forms in the Phaedo is that this
dialogue does nothing to alter the picture we seem to get from other parts of the corpus: that access
to forms, and to knowledge, for human beings, is by philosophy – dialectic – alone. There are no
short cuts, even after death. So far as the Phaedo is concerned, the forms are those things, whatever
they are, that the philosopher seeks to “grasp” through dialectic; the very things that Socrates is
seeking to grasp in the Euthyphro, the Laches, the Meno . . .. That is one reason why Plato can have
him, in the Phaedo, describe “beautiful and good and all such being (ousia)” (D–) as “the
things we’re always chattering about”, and then later refer to them just as “those much-chattered-
about (poluthrēleta) things” (B–)’ ().

 Plato’s Essentialism
things in logoi (‘things said’, ‘accounts’). I shall defer the examination of
this passage for later. My immediate answer will not come as a surprise.
The distinction that Plato is insisting on in this passage is that between
knowing a thing by directly perceiving it versus knowing a thing through
speaking of it and thinking of it; and this is very different from the
distinction between knowing a thing a posteriori and knowing it a priori.
To deny that we can know things by directly perceiving them, and to
require that we search for the knowledge of them in and through what we
say, is compatible with thinking that what we say in this search will be
epistemically informed by our sensory perceptions.
. The question (How can Plato’s two claims be

reconciled?) answered
We are now in a position to state why Plato’s two epistemological claims in
the Phaedo are compatible and why the tension between them is only
apparent. When, in the earlier epistemological claim (e–e), he argues
that the philosopher will, as far as possible, avoid relying on sense-perception
in the search for, and in the acquisition of, knowledge, and will rely, for
enquiring and knowing, only on thinking, reasoning and what the soul does
itself by itself and not in communion with the senses, he argues this because
he thinks that what such qualities as equal are, their essences and Forms,
cannot be specified by example and exemplar and cannot be perceived by the
senses; they can be specified only by engaging in an intellectual search
occupied with certain questions that articulate relevant problems and aporiai.
On the other hand, when, in the later epistemological claim (a; prepared for
in a–c), he argues that sense-perception is necessary for thinking of such
essences and Forms, he does so because he thinks that the supposition that
there are certain essences and Forms, and that they cannot be perceived by the
senses but can be searched for and, all going well, known only by reasoning, is
motivated and justified as a response to certain questions, which articulate
problems and aporiai and which we are motivated and justified in asking
because of the peculiar content of certain of our sensory perceptions.
. The priority of enquiry over knowledge in the Phaedo

I want to end by pointing to what is, I think, a remarkable and significant
consequence of this conclusion. On this conclusion, the reason why Plato
 
Bostock () reads the passage in this way. Chapter .

Knowledge and Enquiry in the Phaedo 
turns to reasoning, and away from sense-perception, is that he thinks that
what certain qualities are, their essences and Forms, cannot be specified by
example and exemplar and, consequently and because of this, cannot be
perceived by the senses. It is not at all due to a general rationalist epistemology,
which prioritises a priori knowledge over a posteriori knowledge or due to the
claim that Forms are supernatural entities that can be known only a priori. It
follows that what Plato’s invocation of reasoning, as opposed to sense-
perception, provides is, in the first instance at any rate, an account of what
is involved in searching for the essences and Forms of certain qualities, not an
account of what is involved in positively getting to know and knowing them.
Plato’s simple thought appears to be this: since we cannot know what these
important qualities are by example and exemplar, and since this entails that we
cannot perceive what they are directly by the senses, we need to start searching
for them in an intellectual way, a way set in motion by questions, and
questions whose answer is not immediately evident to us but, on the contrary,
involves and is based on reasoning. This is the kind of reasoning that the
enquiries that make up many Platonic dialogues exhibit and exemplify.
It is evident that, if something, X, can be known only by a reason-based
search, then not only the search for X, but also the knowledge of X, is based
on reason. This is immediately evident, and even more so once we recall that
what enquiry is is, precisely, a certain process aimed at knowledge. But what
is remarkable is that, on this account, Plato’s claim that knowledge is based
on reason, and not on sense-perception, is not primarily a claim about
knowledge, that is, knowledge as such and as independent of enquiry; on
the contrary, it is primarily a claim about enquiry, and only as a consequence
is it about knowledge, that is, the knowledge aimed at by the enquiry.
This is not, of course, to say that there may not be, in Plato or the Phaedo, an
account of knowledge that is specifically about knowledge and not the conse-
quence of a corresponding account of enquiry; whether there is such an account
of knowledge, in the Phaedo specifically or Plato in general, is a large question
that goes beyond the present study. It is to say that we should be particularly
cautious before we conclude, on the basis of this or that passage in the Phaedo,
or elsewhere, that Plato is providing part of an account of, specifically, knowl-
edge, when what he is providing may be just as much part of an account of
enquiry. And it is to say that, once we are properly aware of this issue, we need to
be particularly cautious to consider whether, if what Plato is providing is part of
an account of both knowledge and enquiry, the account is primarily about
knowledge and only as a consequence about enquiry, or is it primarily about
enquiry and only as a consequence about knowledge.

 
Why are essences, or Forms, distinct from

sense-perceptible things?
Phaedo  and Republic V. –
In the Phaedo (a–c), Plato argues that the Form of the quality equal is
distinct from any pair of sense-perceptible things that are equal. The
conclusion of this argument – which says that Forms are distinct from
sense-perceptible things – is manifestly of great significance. In Phaedo, in
which Socrates’ friends during his last hours are ready to grant him that
there are Forms, and that they cannot be perceived by the senses, it serves
to introduce a basic distinction between ‘two kinds of beings’ (duo eidē tōn
ontōn, a), sense-perceptible beings (aisthēta) and intelligible beings,
objects of reason (noēta); and in Republic (at the end of book V), those
who do not acknowledge that there are imperceptible Forms are set against
philosophers and characterised as ‘lovers of sights’, philotheamones.
The question I want to consider is how, in Phaedo, Plato arrives at the
conclusion that Forms are distinct from sense-perceptible things. I want to
contend that the argument, when properly understood, depends on the
supposition that Forms are essences, essences in the sense of that which is
designated by an adequate and true answer to a ti esti question. Crucial in
the argument is that what certain qualities are cannot be perceived by the
senses; and I shall argue that equally crucial for the argument and its
cogency is the justification of this view, which says that the reason for this
is that what these qualities are cannot be specified by example and
exemplar. At the end of the chapter, I want to compare the Phaedo
argument to the argument at the end of Republic V (e–d).
This is not how the argument is currently understood. For the past
decades, and going back especially to Owen’s  paper, a certain
orthodoxy about the argument in these passages has emerged among
critics, according to which Plato’s argument is about the character and
nature of sense-perceptible things, as opposed to the character and nature
of Forms: Sense-perceptible things are subject to co-presence of opposites
(e.g., the same sense-perceptible things that are equal are also unequal);
Forms are not so subject; therefore, Forms are distinct from sense-


Forms as distinct from sense-perceptible things 
perceptible things. As regards what co-presence of opposites consists in,
the current consensus contains a variety of different views. A notable
difference regards whether co-presence of opposites is thought to concern
tokens or it is thought to concern types of sense-perceptible things. This
difference is summed up succinctly by Verity Harte (, ): ‘Is the
claim (PC) that, for any particular perceptible, having some relevant
feature, F, necessarily, that particular perceptible also has the opposite
feature, un-F? Or is it the claim (UC) that, for any perceptible type, a
token of which is F, for some relevant feature, necessarily that type has un-
F tokens also?’
We may note that, for the cogency and validity of Plato’s argument on
the current, standard and orthodox understanding of it, it does not at all
matter what Forms essentially are; all that matters is that they do not
possess a property, co-presence of opposites, that is possessed by sense-
perceptible things. For, according to Leibniz’s Law, if there is a property
that is possessed by an object, O, but not by an object, O’, then these
objects, O and O’, are distinct and non-identical; and for the application
of this law, it does not, apparently, matter what these objects essentially
are. On my understanding of Plato’s argument, on the other hand, it is
crucial for it and its cogency that we recognise that Forms are essences,
essences in the sense of that which is designated by an adequate and true
answer to a ti esti question.
I shall argue that the form of Plato’s argument in these passages is quite
different from how it is standardly understood. On the understanding of
the argument’s form that I want to defend, its form is as follows: The
sense-perceptible things that appear equal to the senses also appear unequal
to the senses; the Form of the quality, equal, does not appear both equal
and unequal; therefore, this Form is distinct from those sense-perceptible
things. The argument relies on apparent co-presence of opposites; it does

Up to the late s, this debate has been carefully summarised by White (). Since then it has
continued apparently unabated: Fine ( [reprinted ], see esp. section IV); Annas (,
–); Bostock (, –); Fine ( [reprinted ], see esp. –); Fine (, e.g.,
). McCabe () says that ‘particulars are thought [by Plato] to suffer from compresent opposites
by their very nature’ (, emphasis added). See also Sedley (a); Rickless (); Hestir ().
Tuozzo () turns up the volume: ‘In favour of this interpretation [i.e., a certain interpretation of
the equals argument in Phaedo] is the fact that Plato incontestably does hold that sensibles suffer the
co-presence of opposites, and that the Forms do not’ ().

I am not claiming originality of this part of my interpretation. This, if I am not mistaken, is how the
argument was read by Ross (, ); Hackforth (, ); Bluck (, ), that is, as claiming
that the same sense-perceptible things appear to the senses to be both equal and unequal, whereas this
does not happen with Forms. It is a good question how, and why, this reading was displaced, in

 Plato’s Essentialism
not rely on real co-presence of opposites. This issue, we shall see, is crucial
for considering whether the argument depends on the view that Forms are
essences. For, if the argument is understood to rely only on apparent co-
presence of opposites, it can, and I shall argue that it should, be under-
stood to depend on this view; if, on the other hand, it is understood to rely
on real co-presence of opposites, the argument does not rely on Forms
being essences.
. Phaedo a–c

Here is the passage:
Consider, he said, whether this is the case: we say that there is something
that is equal. I do not mean a stick equal to a stick or a stone to a stone, or
anything of that kind, but something else beyond all these, the Equal itself.
Shall we say that this exists or not? – Indeed we shall, by Zeus, said
Simmias, most definitely. – And do we know what this is? – Certainly. –
Whence have we acquired the knowledge of it? Is it not from the things we
mentioned just now, from seeing sticks or stones or some other things as
equal that we come to think of that other which is different from them? Or
favour of the reading: the same sense-perceptible things are both equal and unequal, whereas the
Form equal is not both equal and unequal. Gallop (, –), is characteristically careful and
cautious, but in the end appears to think that Plato moves from the many equals appear both equal
and unequal to the many equals are both equal and unequal (see ). Nehamas contributed to
cementing the orthodoxy in the making, when he says (a, –): ‘My claim is that
Hackforth and those who agree with him have been misled by an accidental feature of the
particular example that Plato used in his argument. They thus took it that Plato claims that things
can appear both F and not F in the same context to different observers’ (). It seems to me that
Owen () did not assist in getting the debate off to a good start. He begins by observing that: ‘In
Republic VII (a–a) numbers are classed with such characteristics as light and heavy, large and
small, on the score that our senses can never discover any of them kath’ hauto, in isolation (d):
in perceptible things they are inseparable from their opposites’ (, my underlining). He then
drops, without explanation, this qualification and caveat (i.e., the clause I have underlined) for his
general account of Plato’s argument. I note that Owen’s reading was anticipated by Murphy (,
, n. ). Murphy’s reading was criticised by Hackforth (, ), which is some years before
Owen. Hackforth’s response to Murphy was later criticised by Nehamas (a), as noted above,
who followed Owen, as have so many others since.

I would like to acknowledge that a similar point has been made before: it was made by Kirwan in his
 paper, when he urged especial caution before we attribute to Plato, on foot of such passages,
not only apparent, but real co-presence of opposites. But Kirwan’s call for caution has not generally
been heeded, perhaps because he did not consider why, in certain passages (both in the Hippias
Major, as I have noted more than once, and, as we shall see, at the end of Republic V), Plato appears
to move from apparent to real co-presence of opposites. I argue that Plato makes this move in these
passages only on behalf of those he is arguing against, such as Hippias in the Hippias Major and ‘the
lovers of sights’ at the end of Republic V. For they do not believe in a principled distinction between
‘appearing’ and ‘being’, because they think that what a quality is can be defined by example and
exemplar and by how it appears to the senses.

Forms as distinct from sense-perceptible things 
doesn’t it appear to you to be different? Look at it also this way: do not equal
stones and sticks sometimes, while remaining the same, appear to one as
equal and to another as unequal? – Certainly they do. – But what of the
equals themselves? Have they ever appeared unequal to you, or Equality to be
Inequality? – Never, Socrates. – These equal things and the Equal itself are,
therefore, not the same? – I do not think they are the same at all, Socrates. –
But it is definitely from the equal things, though they are different from the
Equal, that you have derived and grasped the knowledge of equality? – Very
true, Socrates. (Phaedo a–c; trans. Gallop, adapted)
The challenge for understanding Plato’s argument in the Phaedo a–c, on
any reading of this passage, is to determine why Plato chooses to demon-
strate, in just this way and by reference to a quality such as just this, equal,
that Forms are not identical with sense-perceptible things. After all, there are
plenty of qualities possessed by Forms that are not possessed by sense-
perceptible things: Forms are changeless, uniform, not perceptible by the
senses, knowable only by reasoning, the basis of causation and explanation,
distinct from sense-perceptible things, necessary for thought and speech,
separate from physical things and more. If the validity of the argument is
based, purely and simply, on Leibniz’s Law, Plato could have chosen any of
these qualities to argue that Forms are distinct from and not identical with
sense-perceptible things. It will not at all do, I believe, to say that he could
indeed have arrived at the conclusion in any of these alternative ways but
chose to arrive at it the way he did. This leaves positively unexplained two
things. First, why is the argument conducted in regard to a certain Form, the
Form of the quality, equal, and not simply in regard to the Form of any
quality? Secondly, why does the argument speak of the things that we see as
equal (idontes isa, b) and that appear equal (isa phainetai, b)?
There is a further and deeper problem with the standard approach to
Plato’s argument. For if it is said, or implied by the way in which one
understands Plato’s actual argument, that Plato might as well have con-
ducted it in any of these alternative ways, and by reference to any of the
characteristics possessed by Forms but not by sense-perceptible things, it
will, in effect, be supposed that the argument relies on an already available
theory of Forms, or at any rate an already available commitment to entities
that satisfy all these characteristics, which are manifestly not displayed by
sense-perceptible things. But, as several critics have noted, it is not at all

For a thorough criticism of the view that the Phaedo comes equipped with an already available theory
of Forms, see Lee (). Lee likewise argues that ‘there are strong dramatic indications against a
particular metaphysical theory being a necessary presupposition of the argument [i.e., the “equals”
argument we are considering]’ (). This, however, is as far as I am prepared to follow Lee in his

 Plato’s Essentialism
clear where Plato puts forward this theory and commitment. And, if we
suppose, as some critics do, that this theory and commitment is sprung on
us wholesale in just this dialogue, the Phaedo, then, while we may be
impressed by the validity of Plato’s argument in Phaedo a–c, since it will
be simply an instance of Leibniz’s Law, we may doubt that the argument
amounts to a defence, or justification, as opposed to merely to a perspic-
uous presentation, of the commitment, already made on who knows what
grounds, to entities that satisfy all manner of characteristics that are
manifestly not displayed by sense-perceptible things.
I want to argue that this challenge can be met, and these issues resolved,
if we suppose, first, that Plato’s argument is of the form I am proposing:
The sense-perceptible things that appear equal to the senses also appear
unequal to the senses; the Form of the quality, equal, does not appear both
equal and unequal; therefore, this Form is distinct from those sense-
perceptible things. And, secondly, the argument relies on Forms being
essences. I shall demonstrate, by a close reading of the passage, that there
are clear textual indications that the argument is of this form and some
notable indication that it relies on Forms being essences.
To understand Plato’s argument, we need, first of all, to consider what is
meant by the first premise, which says that: The sense-perceptible things
that appear equal also appear unequal. The sentence that states this
premise says: ‘Do not equal stones and sticks sometimes, while remaining
the same, appear to one (or, on a different textual variant: ‘at one time’) as
equal and to another (or, ‘at another time’) as unequal?’ This allows for a
number of interpretations, especially because the tō[i] . . . tō[i] at b (‘to
one . . . to another’) can be taken as masculine, in which case it is naturally
taken to mean ‘to one person . . . to another person’, or it can be taken as
neuter, in which case it means ‘to one . . . to another’, where it is not
immediately clear what needs to be supplied (I shall return to this issue
presently). In addition, some manuscripts, instead of tō[i] . . . tō[i], have
reading of Plato’s argument. From the view that the argument does not depend on an already
available theory of Forms, Lee infers that we must make sense of the argument without referring to
anything beyond the Phaedo and the statement of the argument. This is a dubious inference, and, in
my judgement, what this approach yields, in Lee’s paper, is not a substantial interpretation but
innumerable trees, or bits of trees, without a contour of a wood in sight.

Thus Patterson (, ): ‘Plato’s “middle dialogues,” especially the Phaedo, Republic, Symposium
and Phaedrus, along with the late Timaeus, bring the reader’s attention to a previously unnoticed sort
of thing that he calls “forms” (eide). Forms are entirely imperceptible but grasped in thought; non-
spatial and non-temporal yet fully real; independent of and separate from worldly things yet
“participated in” by them and somehow responsible for their being what they are.’ I suspect that
the view that these dialogues ‘bring the reader’s attention to a previously unnoticed sort of thing’ is
not uncommon in the literature.

Forms as distinct from sense-perceptible things 
tote . . . tote, which means ‘at one time . . . at another time’; and this may
indicate either a distinction in time as such, where it is the difference in
time that matters to the argument, or a distinction of occasions at different
times, such that it is the difference of these occasions, irrespective of how it
and they are understood, that matters to the argument (I shall return to
this issue presently).
If, to understand Plato’s argument, we limit ourselves to this passage, it
appears that the one thing we have to rely on for choosing among these
several variant interpretations of the first premise is that the second
premise says that: The Form of the quality, equal, does not appear both
equal and unequal; and we would expect that the sense in which this
Form does not appear to be both equal and unequal is the same as the
sense in which the sense-perceptible things that appear equal also appear as
unequal. As many critics have noted, it is not very plausible to think that a
Form does not, or cannot, appear in different and opposite ways to
different people; what is plausible, on the contrary, is that this does happen
and is possible. For example, and as Socrates explains (Republic I. e–
a), that which appears to Thrasymachus to be virtue, namely injustice,
is just that which appears to Socrates to be vice; and, as Socrates argues
(Gorgias –), that which appears to Polus (and, later, Callicles) to be
what it is to be powerful and able to do what one wants, is just that which
appears to him, Socrates, to be what it is to lack power and the ability to do
what one wants. It is, therefore, surprising that some of the same critics
have persisted in reading tō[i] . . . tō[i] in the first premise and take the tō[i]
in the masculine as referring to different people. It seems to me preferable,
on these grounds, to rule out this interpretation of the tō[i] . . . tō[i] in the
first premise.

Sedley (a) argues that Plato’s argument works only if we read tote . . . tote; and he understands
the tote . . . tote as indicating a distinction in time as such, so that it is the difference in time that
matters to the argument: Forms are changeless, whereas sensibles are changing. I set aside the
possibility of reading tō[i] . . . tō[i] in the feminine, hence as referring to ‘relativity of equality to
different respects’ (Lee , ). Lee (, –) mentions this possibility and the one or two
rare proponents who appear to have noticed it.

I shall consider presently why what the second premise says is, strictly, that this Form does not
appear to be both equal and unequal according to Simmias. This is, we shall see, an important
question. For the view that the argument relies only on Leibniz’s Law, and that the reference to
Simmias is irrelevant, see Dancy (, ), when he says: ‘We are going to advert to Leibniz’s
Law: we need some defeating predicate true of any ordinary equal thing that is not true of the equal
itself. At first sight, it looks as if the defeating predicate is “ – shows itself to Simmias as unequal
under some conceivable circumstances”: this is true of any equal stick but not of the equal. But the
reference to Simmias is irrelevant: Simmias is standing in here for anyone at all. The defeating
predicate is, in effect, “ – shows itself to someone under some conceivable circumstances as unequal”:
any equal stick, stone, or whatever does this; the equal does not.’

 Plato’s Essentialism
This still leaves us with several possible interpretations of the first
premise; and it leaves us quite in the dark about what the sense is, in
which sense-perceptible things that appear equal also appear unequal,
whereas this does not happen, or at any rate it does not happen according
to such people as Simmias, to the Form of this quality, equal. I don’t see
how we can begin to address these questions, and so to understand Plato’s
argument, unless we go beyond this passage and unless we are open to
going beyond even this dialogue for the understanding of the passage.
I propose that we go back to the Hippias Major and observe that when,
there, Socrates argued against Hippias that what beauty is cannot be
specified by example and exemplar, he noted that a consequence of
Hippias’ view, which is that, on the contrary, what beauty is can be
specified by example and exemplar, is that the same things that appear
beautiful when compared to one example and exemplar, and so to one
standard, of beauty, will appear ugly when compared to another example
and exemplar, and so to another standard, of beauty; and this because the
one example, exemplar and standard (such as the menial pot) is not
beautiful, and may even be positively ugly, when compared to the other
example, exemplar and standard (such as the girl, horse or lyre). I recall
that we have already considered this (see Chapters  and ) and made a
number of observations some of which are worth recalling at this point.
First, there is nothing problematic about supposing, by itself, that the
same things that appear beautiful when compared to one example, exem-
plar and standard of beauty, will appear ugly when compared to another
example, exemplar and standard, of beauty or about supposing, by itself,
that one example, exemplar and standard is not beautiful, and may even be
positively ugly, when compared to another. What is problematic, indeed
involves a contradiction, is supposing this because of the view that what
beauty is can be specified by example and exemplar.
Secondly, when Socrates spells out this consequence of Hippias’ view,
he does so both in terms of how beauty appears (phainetai), and indeed
appears to the senses, and, apparently without distinction, in terms of how
beauty is (esti). But, if we note that Socrates is drawing out the conse-
quences of Hippias’ view, we recognise that we ought not to suppose that
this seeming laxity between ‘appears’ and ‘is’ is due to Socrates; we ought
to suppose, rather, that Socrates thinks Hippias’ view implies that we may
move from ‘appears’ to ‘is’.
Thirdly, the proposition that what beauty is can be specified by example
and exemplar implies the proposition that what beauty is is apparent to the
senses; and the proposition that what beauty is cannot be specified by

Forms as distinct from sense-perceptible things 
example and exemplar implies the proposition that what beauty is is not
apparent to the senses.
I propose that we understand in just this way the first premise in the
Phaedo argument; except that whereas the argument in the Hippias Major
is about beauty and what beauty is, its essence and Form, the argument in
the Phaedo is about equality and what equality is, its essence and Form.
We have already seen (in Chapter ) that the argument from the Hippias
Major, against specifying what beauty is by example and exemplar, and
hence against supposing that what beauty is is apparent to the senses,
extends to certain other qualities, including equal and one.
I can imagine a reader objecting, at this point, that I am making too
much of the Hippias Major argument, on my understanding of it, and of
the supposition that that argument is pivotal in Plato’s philosophy. My
response is that I am able to make the very same point by relying, not only
on the Hippias Major argument but also on the argument at the end of
Republic V – which we shall consider in a moment. To anticipate, the
lovers of sounds and sights, to which the Republic argument is addressed,
are characterised as thinking that beauty is apparent to the senses and that
this is all there is to an understanding of beauty (see d ff.). The similarity
to the argument in the Hippias Major is remarkable. For that argument is
about what beauty is; and Hippias is presented as defending (not simply
assuming) the view that what this quality is is apparent to the senses and
that what this quality is can be specified by example and exemplar and by
reference to what is apparent to the senses. Moreover, apparent co-
presence of opposites is centre-stage in both the Hippias Major argument
and the argument at the end of Republic V. For, in the Hippias Major
argument, that the beautiful things we perceive by the senses will also
appear – and, according to Hippias, be – ugly, was presented as a
consequence of Hippias’ view, that what beauty is can be specified by
example and exemplar and by reference to what is apparent to the senses.
Hippias is an original and archetypal lover of sounds and sights in Plato.
If we adopt this proposal (I shall consider in a moment whether we have
textual grounds for doing so), we obtain a number of notable, and I think
very attractive, results.
First, the the tō[i] in the first premise must be understood as neuter, not
as masculine (at any rate if the masculine is understood as referring to

Dancy () also connects these passages. He thinks that Socrates’ argument in the Phaedo would
have been ‘less odd’ had he used The Beautiful from the Hippias Major, rather than The Equal
(esp. , but also ).

 Plato’s Essentialism
persons); and the tō[i] . . . tō[i] refers to different standards of equality. On
this reading, the reason why the sense-perceptible things that appear equal
also appear unequal, is that different standards are involved in the two
cases or occasions, such that the one standard is not equal, and may be
positively unequal, when compared to the other.
Secondly, there is nothing to choose between the textual variant tō[i] . . .
tō[i] in the first premise, on this reading of it, and the textual variant
tote . . . tote in the same premise, provided that we take what matters to the
argument on the tote . . . tote reading to be, not a distinction in time as
such but a distinction of different occasions at different times, where this
difference consists in different standards being used in the
different occasions.
Thirdly (from the two points above), we have no longer four interpre-
tations of the first premise: the two readings of tō[i] . . . tō[i], that is, as
masculine or as neuter, plus the two readings of the tote . . . tote, that is, as a
distinction in time as such or as a distinction of different occasions at
different times. We have just one interpretation, and one that is
properly motivated, both textually – if we allow ourselves some dialogue-
crossing! – and philosophically – since it goes back to Plato’s view, which is
at the root of his whole philosophy, in so far as his philosophy revolves
around the ti esti question and the view that this question is most
important and most difficult to answer, that what certain qualities are
cannot be specified by example and exemplar, and therefore, cannot be
perceived by the senses.
Fourthly, we have secured that there is a single sense, and one well
worth recognising, in which the Form of this quality, equal, does not
appear to be both equal and unequal; or, at any rate, this Form does not
appear to be both equal and unequal to those who have understood that,
and why, the essence or Form of such a quality cannot be specified by
example and exemplar, or perceived by the senses. The reason why, and
the sense in which, the Form of this quality, equal, does not appear both
equal and unequal, is that it is the single standard for what it is for any
thing to be equal, such that it is not the case that there is another standard,
such that, compared to it, this Form is not equal and may be positively
unequal. And the reason why Plato thinks that such an essence and Form
is needed is, precisely, that he thinks that what such qualities as equal or
beautiful are cannot be specified by example and exemplar and that it is the
view that they can be so specified that implies that the same things appear
to have both these qualities and their opposites.

Forms as distinct from sense-perceptible things 
Fifthly, it is apparent now why the second premise is formulated as it is:
‘But what of the equals themselves? Have they ever appeared unequal to
you, or Equality to be Inequality?’ Suppose the interlocutor had been, not
Simmias, but Hippias. Hippias thinks that what a certain quality is, such
as beauty, sometimes appears ugly. For he thinks that what such qualities
are can be specified by example and exemplar and hence is apparent to the
senses; and from this it follows, as he may recognise without being overly
perturbed, that an example and exemplar, and hence standard, of what
such a quality is can display the opposite quality when compared to a
suitably different standard (e.g., when the menial pot is compared, for
beauty, to the girl, horse or lyre). If Simmias does not think this, it is, we
must suppose, because, being close to Socrates and his reasonings, he has
understood what is wrong with a view such as Hippias’. It is remarkable
that, at the opening of the argument, Simmias is said to know what this
Form, ‘the equal itself’ (auto to ison, a), is: ‘And do we know what this
is? – Certainly’ (b–; ē kai epistametha auto ho estin; panu ge). We may
also recall that, at d, what Socrates had asked Simmias is not whether
essences and Forms have ever been seen by the senses by anyone but
whether they have ever been seen by the senses by Simmias. Simmias
replied in the negative, but no further reason was offered for the claim
that essences and Forms cannot be perceived by the senses. As I argued
Chapter , it is hard to know what to make of this, unless we think Plato is
drawing on something he has already argued, presumably in another
dialogue, such as, as I have argued, the Hippias Major.
Sixthly, the Phaedo argument depends, through and through, on the
supposition that Forms are essences, and the argument cannot at all be
understood without this supposition. For it depends on the view that what
certain qualities are, their essence and Form, cannot be specified by
example and exemplar; and on the consequence of this view, which is that
what these qualities are is not evident directly to the senses, and so cannot
be perceived by the senses; and on another consequence of the view, which
is that the standard of what a quality is cannot display the opposite quality.
Finally, it is quite wrong to suppose that the Phaedo argument, for the
conclusion that the Form of equal is distinct from and not identical with
any pair of sense-perceptible things that are equal, makes use of a

On the term, ‘the equals themselves’, there is a traditional reading going back to at least to
Heindorf, and endorsed by Archer-Hind (, –), among others, which takes ‘the equals
themselves’ to be referring to the Form of equal, arguing that the plural here as forced by idea that
the Form of equal must be complex or multiple, since equality can only be determined through
comparison. For more on this reading, see Dancy (, , n. ).

 Plato’s Essentialism
supposition, already available, that there are entities, Forms, that possess a
whole range of characteristics not possessed by sense-perceptible things.
Rather, the argument serves to make another particular step in Plato’s
logical exposition and defence of just this supposition. This defence is
based on the principal supposition that these entities, the Forms, are
essences, in the sense of that which is designated by an adequate and true
answer to a ti esti question; and on certain consequences, more basic and
less in need of justification than the conclusion that Forms are distinct
from and not identical with sense-perceptible things, derived, elsewhere
and not in the Phaedo, from this principal supposition.
I hope this sounds like the climax it is intended to be. There are still
some issues to consider and especially the issue of the validity of Plato’s
argument. Critics have been impressed by the appearance, which cannot
be denied, that Plato’s argument displays an instance of Leibniz’s Law and
depends on this for its validity. Some of the same critics have reasoned
that, if this is what the validity of Plato’s argument depends on, the
argument is better read as relying on real co-presence of opposites than
on apparent co-presence of opposites. For we know (since Frege’s ‘On
Sense and Reference’, , and especially since Quine took it up) that ‘O
appears F to N’ introduces a so-called intensional context and that, for this
reason, the following argument is not valid: O appears F to N; O’ does not
appear F to N; therefore, O is distinct from and not identical with O’.
Therefore, as some of these critics have concluded, Plato’s argument, if
understood as relying on apparent co-presence of opposites (where F above
stands for both equal and unequal), is not valid.
The problem with this reasoning is not that Plato knew nothing of
intensional contexts; he did not have to know this to recognise that, for
whatever reason, an argument is fallacious if it is of the form: O appears
F to N; O’ does not appear F to N; therefore, O is distinct from and not
identical with O’. The problem is that the reasoning assumes that, if
Plato’s argument relies on apparent co-presence of opposites, and not on
real co-presence of opposites, the apparent co-presence of opposites must
be understood as person relative, that is, that the tō[i] at b must be
understood as masculine and the tō[i] . . . tō[i] as referring to different
people. I have already argued, on independent grounds, that there is good
reason to reject this reading. I conclude that, if there is a problem with the
validity of Plato’s argument, it is not due to the reading of it as relying on

Dancy (, –) makes quite a lot of this.

Forms as distinct from sense-perceptible things 
apparent, not real, co-presence of opposites, nor is it due to the possibility
that Plato was foxed by an intensional context.
Is there a problem with the validity of Plato’s argument? We today are
habituated to thinking of validity in purely formal terms: we assume that
to consider whether an argument is valid is to consider the form of the
argument, and this alone; the content of the premises, or the reasoning
behind and justification of the premises, is not relevant for considering the
validity of an argument, except perhaps in so far as this might be relevant
to determining its form. I doubt that this is a good way of assessing the
validity of an argument in a philosopher like Plato, who, not having an
explicit notion of the form of an argument, or, in general, of formal logic,
makes no explicit distinction between the form and the content of an
argument. This is not to deny that Plato recognised, in some implicit way
and without formulating it, Leibniz’s Law, which is an exemplary case of
picking out a certain form of valid argument or to deny that he chose to
present the Phaedo argument as he does, because of this recognition. It is to
deny that, if we want to assess the validity of Plato’s argument, we should
rely, purely and simply, on the apparent fact that the argument is pre-
sented as conforming to Leibniz’s Law.
It seems to me that to assess the validity, and, in general, the cogency, of
Plato’s argument requires an understanding not only of its form, but also
of the content of its premises, and even of the reasoning behind and
justification of them; it requires a proper interpretation of the argument –
such as the one I have defended. The question I shall consider, therefore, is
whether Plato’s argument is cogent, on this interpretation of it. This
amounts to asking the following question: If what a quality such as equal
(or beautiful, or one, etc.) is cannot be specified by example and exemplar,
and by reference to a standard provided in just this way, and if what this
quality is cannot be perceived by the senses, does it follow that an adequate
standard of this quality must designate an entity – the essence and Form of
this quality – that is distinct from and not identical with any sense-
perceptible thing or quality?
Skilled as philosophers are today in making distinctions, and habituated
as they are in using them not least for the purpose of demonstrating that
we are not logically compelled to accept this, that or the other substantial
philosophical claim, however well-reasoned it may appear to be, there are
readily available ways of answering this question in the negative and of
resisting Plato’s conclusion even while granting his premises. One such
way is to argue that, while Plato’s conclusion is concerned simply with
things, his premises need only be understood as being concerned with

 Plato’s Essentialism
concepts, since they are concerned with what we are committed to when
we think of the identity conditions of the objects of our thoughts and that
any direct move from concepts to things is, as Kant was the first to teach
us, questionable. Another such way is to introduce the notion of a
particular substance and argue that a single particular substance that,
according to the conclusion of Plato’s argument, is not perceptible by
the senses may, the very same one according to the strictest notion of
sameness and identity, also be perceptible by the senses. (This is the notion
of substance that caused Bishop Berkeley such offence.)
Do such moves show that Plato’s argument is invalid and lacks cogency?
It seems to me that what it shows, rather, is that while Plato’s argument
may be cogent and well-reasoned, in the sense that the premises provide
good reason for accepting the conclusion, there may be ways of resisting
the conclusion even while granting the premises, but ways that are them-
selves thoroughly substantial and disputable. No surprise, too, it seems to
me, since one will find that this is true of any philosophical argument,
and conclusion, worth considering and taking seriously – unless we are
committed, from the start, to philosophical quietism, deflationism,
or scepticism. This is a mindset that Plato identifies in the Phaedo and at
a juncture of the dialogue in which the interlocutors are inclined to
give up on the argument altogether (i.e., the argument for and against
the claim that the soul can exist independently of the body), because there
seems no prospect of conducting the argument definitively, conclusively
and without any remaining doubts. Plato calls the mindset misologia, the
aversion to philosophical argument and philosophical theorising (see
Phaedo d ff.).
The question, therefore, is whether the premises in Plato’s argument,
and the reasoning behind and justification of them, provide good reason,
even if it might be denied that it is conclusive reason, for the conclusion
that says that the essence and Form of a certain quality, equal, is distinct
from and not identical with any pair of sense-perceptible equal things. It
seems to me that the answer to this question is, manifestly, Yes: If what
certain qualities are cannot be specified by example and exemplar and by
reference to something that can be directly perceived by the senses, a
consequence of such a specification being that the things invoked in it
display apparent co-presence in regard to these qualities, then what these
qualities are must be specified in some other way, and by reference to
something else, something distinct from and not identical with the things
invoked in such a specification, and something that does not display
apparent co-presence in regard to these qualities. QED

Forms as distinct from sense-perceptible things 
So far my reasoning has relied on the proposal, which I have tried to
motivate in a number of ways, that we should understand Plato’s argument
in Phaedo a–c as building on the argument in the Hippias Major for the
claim that what certain qualities are cannot be specified by example and
exemplar and by reference to something that is evident to the senses and
can be directly perceived by the senses. It may be objected that this
proposal is not sufficiently motivated by the text of Phaedo a–c. It is
true that the proposal supposes that the argument in this passage cannot be
properly understood unless we go beyond the passage and indeed the
dialogue; this, I would like to think, is not by itself a reason against the
proposal. Is there anything in the passage, of a strictly textual nature, that
supports the proposal? One remarkable and puzzling feature of the argu-
ment, as formulated in this passage, is that the second premise does not say
that the Form of this quality, equal, does not, or cannot, appear both equal
and unequal; it says that it has never appeared both equal and unequal to
Simmias: ‘But what of the equals themselves? Have they ever appeared
unequal to you, or Equality to be Inequality? – Never, Socrates’ (c–).
If this is taken strictly and according to the letter, it is naturally taken to
imply that this Form may have appeared both equal and unequal to others.
But can any sense be made of this? Surely the argument cannot depend on
Simmias, or on what someone who entertains the idea of co-presence of
opposites in a Form, happens to think. I have offered an account, based on
the recommended proposal, that makes good sense of this. Unlike
Hippias, or another who has not had the benefit of witnessing and
following Plato’s argument against the view that what certain qualities
are, their essence and Form, can be specified by example and exemplar and
by reference to something that is evident to the senses, Simmias, being
close to Socrates and being characterised, at the opening of the argument,
as one who knows this Form, knows that, and why, what this quality is, its
essence or Form, is not subject to co-presence in regard to this quality.
I don’t know of any other plausible account of this peculiar, and strictly
textual, feature of Plato’s argument.
We are left with one issue: Why be confident, or as confident as I have
been, that the argument should be understood in terms of apparent co-
presence of opposites, not in terms of real co-presence of opposites? At this
late hour, and having gone through what we have gone through, I hope
this does not need belabouring. The reason is a wholly textual one. First,
the first premise of the argument is introduced by reference to sticks and

This has been spelled out by Kirwan ().

 Plato’s Essentialism
stones in so far as they are seen as equal (idontes isa, ‘seen as equal’, b).
Secondly, the idontes isa is picked up immediately by the phainetai (‘they
appear’) in the statement of the first premise: ‘Do not equal stones and
sticks sometimes, while remaining the same, appear to one as equal and to
another as unequal?’ (b–; I mean the phainetai in this statement, at
b). Thirdly, nowhere in the premises of the argument does Plato move
from phainesthai to einai, or indicate any readiness to do so. It is true that
the verb phainesthai need not be understood to mean ‘to appear’; it can be
understood to mean ‘to manifestly be’. It is also true that one grammatical
way of distinguishing between these two uses of phainesthai, namely,
whether it is followed by the infinitive or by the participle, is not available
to us here. But the idontes isa is surely sufficient reason to rule out that the
phainesthai in the statement of the first premise (i.e., at b), and so
likewise in the second premise (at c, ephanē; on the supposition that
the verb phainesthai does not change its meaning across the two premises),
can be understood to mean ‘to manifestly be’, and sufficient reason to
think that it should be understood to mean ‘to appear’.
This ought to be plain. Consider one who says ‘Things that are seen as
equal in one way are also seen as unequal in another way’ (I use ‘in one
way’ and ‘in another way’ to cover all the textual variants and grammatical
variations discussed above). Why would such a person ever use ‘are seen
as’, which he immediately picks up on by using ‘appear as’, in a veridical
sense? He would have to think that sight, or sensory perception in general,
is veridical. For his statement is about any chance things that are seen as
equal and any chance seeing of such things – ‘chance’, in the sense of ‘ein
beliebiger’, ‘un qualsiasi’. To use ‘are seen as’ and ‘appear as’ in a veridical
sense, such a person would, therefore, have to have a remarkable theory of
sensory perception, such as a theory of the kind spelled out at great length
in the Theaetetus for the very purpose of considering what theory of
sensory perception one is committed to if one thinks that perception never
errs. We cannot seriously entertain the idea that Plato intends the argu-
ment at Phaedo a–c to rely on such a theory of sensory perception or
that Plato is committed, in that argument, to such a theory.
. Republic V. e–d

Here is the passage:
‘Given all this, then, I’ll demand an answer from this splendid fellow who
thinks there’s no such thing as the beautiful taken by itself, any nature that
belongs to beauty, by itself, remaining forever exactly as it is, but at the

Forms as distinct from sense-perceptible things 
same time supposes that beautiful things abound – that fellow who loves to
be a spectator, and won’t put up with it for a minute if someone says the
beautiful is one thing only, and similarly with the just, and so on: “Tell me
this, fine fellow that you are,” we’ll say: “of all these many beautiful things
of yours, surely there’s not a single one of them that won’t appear as ugly?
Or of the many just things, that won’t appear as unjust? Or of the many
pious things, that won’t appear as impious?”’
‘No’, said Glaucon, ‘they must always appear as in a way both beautiful
and ugly, and similarly with your other examples’.
‘What about the many doubles? Do they appear any less as halves than as
doubles?’
‘No’.
[b] ‘And big things, small things, light things or heavy things – they
won’t be called by whatever names we give them, will they, any more than
by the opposite of these?’
‘No’, he said; ‘they’ll always keep both’.
[b] ‘So is it the case with any of these many things of yours that it is
whatever anyone claims it to be more than that it is not?’
‘It’s like with those double meanings people play with at parties’, he said,
‘like the children’s riddle about the eunuch hitting the bat, playing on what
he hit it with and what it was sitting on; they too seem to go both ways, so
that it becomes impossible to grasp any of them firmly in the mind as either
being or not being both or neither’.
‘So do you know what to do with them’, I asked, ‘or any better place to
locate them than in between being and not being? I don’t suppose they’ll
turn out to be darker than what is not by outdoing it in not being, or
brighter than what is by outdoing it in being’.
‘Very true’, he said.
‘Then it seems we’ve discovered that the many things ordinary people
think about beauty and the rest tumble around somewhere between what
purely and simply is not and what purely and simply is.’
‘We have.’ (Republic V. e–d; trans. Rowe, my underlinings)
I want to consider two questions: First, why, in this passage, does Plato
move from apparent to real co-presence of opposites? Secondly, may we, as
we did in the case of the Phaedo argument, suppose that Plato’s argument
here relies on the supposition that Forms are essences and on the view,
argued for in the Hippias Major, that what certain qualities are, their

The argument at the end of Republic V (i.e., from b to a) has attracted a huge amount of
attention and analysis during the past forty or fifty years especially. For a good summary of the state
of the art on it, see Lee (b). For a more recent, and different, account of the argument at the
end of Republic V, see Rowe (, ch. ). Rowe reads the argument as constrained by the fact that
it is addressed to the lovers of sounds and sights and constrained in such a way that we cannot
directly infer from it Plato’s own metaphysical or epistemological views.

 Plato’s Essentialism
essence or Form, cannot be specified by example and exemplar and all that
follows from that?
Let us first ask: Why, in this passage, does Plato move from apparent to
real co-presence of opposites?
Up to b, it is said that, for certain qualities, F (which include
beautiful, just, pious, double, large, heavy), things appear both F and oppo-
site-to-F. The verb phainesthai is used repeatedly. But at b–, and
especially b, the point is suddenly stated not in terms of appears but in
terms of is: ‘So is it the case with any of these many things of yours that it is
whatever anyone claims it to be more than that it is not?’ (poteron oun estin
mallon ē ouk estin hekaston tōn pollōn touto ho an tis phē[i] auto einai;).
What is concluded (from c on) is that the things to which the lovers
of sounds and sights are committed are things that can be opined (they are
doxasta) but cannot be known (they are not gnōsta) and that this is because
these are things that both are (F) and are not (F). Plato began preparing for
this conclusion already from a, and it was stated most clearly at
d–, when it is said that the object of doxa is ‘something that both
is and is not simultaneously’ (hama on te kai mē on), and at e–, when
it is said that the object of doxa is ‘the thing that shares in both being and
not being’ (to amphoterōn metechon, tou einai te kai mē einai).
The move from apparent to real co-presence of opposites is a surprising and
puzzling one; especially if, as it is here, it is made without comment or indication
of why it might be thought to be a sound move. A common view among many
critics today, with notable exceptions such as Rowe (), is that Plato really
intends this move and that this is what the Republic passage shows.
This seems to me to be a mistake. The argumentative and dialectical
context of the argument at the end of Republic V is significantly different
from the argumentative and dialectical context of the argument at Phaedo
. The Phaedo argument is addressed to people familiar with and well-
disposed towards the view that there are things that are not sense-
perceptible and that these things include essences and Forms. The
Republic argument, on the other hand, is addressed to people, referred to
as ‘the lovers of sounds and sights’, or simply ‘lovers of sights’, philothea-
mones, who think that all things are either directly perceptible by the senses
or are constructs of such things.

In Republic b– Plato characterises the view of certain people, or the mindset of many people –
people to whom he refers as ‘the lovers of sounds and sights’ – as being a commitment to thinking
that all there is is ‘colours and shapes and all the things constructed out of such things’ (chroas kai
schēmata kai panta ta ek tōn toioutōn dēmiourgoumena).

Forms as distinct from sense-perceptible things 
The aim of the two arguments is different, too. The aim of the Phaedo
argument is to defend the claim that Forms are distinct from, and not
identical with, sense-perceptible things. The aim of the Republic argument
is to defend the claim that the proposition that all there is is sense-
perceptible things is incompatible with the proposition that real knowl-
edge is possible. It seems clear that the Republic argument presupposes that
Forms are distinct from, and not identical with, sense-perceptible things,
the question in the Republic argument being whether, if one denies the
existence of such non-sense-perceptible things, namely, Forms, one can
believe in the possibility of real knowledge.
These differences in dialectical context and argumentative aim ought to
give us pause before we suppose that these two arguments – Phaedo a–c
and Republic V. e–d – are basically the same argument with
possible minor variations, a supposition that has prompted some critics
to use the Republic argument to interpret the Phaedo argument.
In the Republic passage, Plato has a clear motive for having the people to
which the argument is addressed, the lovers of sounds and sights, make this
move: the move from apparent to real co-presence of opposites. His aim at
the end of Republic V is to show that if, as do the lovers of sounds and
sights, one thinks that all things are either directly perceptible by the senses
or constructs of such things, then one cannot think that genuine knowl-
edge, epistēmē, is possible. But, according to Plato’s argument here, it is
true on any view, including that of the lovers of sounds and sights, that
genuine knowledge is of what is. If, therefore, Plato wants to use the claim
that, for certain qualities, F, a sense-perceptible thing, O, appears both
F and opposite-to-F, to demonstrate that the lovers of sounds and sights
cannot aspire to genuine knowledge – and it is clear that this is how he
proceeds – he will have to make, on their behalf, the move from apparent to
real co-presence of opposites.
I conclude that it is the lovers of sounds and sights, because of their
mindset and commitments about what there is, who see no problem in
moving from ‘O appears both F and opposite-to-F’ to ‘O is both F and
opposite-to-F’; we cannot suppose that Plato sees no problem in such
a move.

See esp. Nehamas (a, ); Bostock (, –). Bostock makes the point in a
characteristically candid way: ‘I conclude that the Republic passage demands interpretation (b)
[i.e., that sensibles are both F and opposite-to-F, for certain qualities, F], and will not tolerate
interpretation (a) [i.e., sensibles only appear both F and opposite-to-F, for certain qualities, F].
Since the Republic is reasonably close in date to the Phaedo this affords a strong presumption that
interpretation (b) is also right for the Phaedo’ ().

 Plato’s Essentialism
Let us now ask: Can we, as we did in the case of the Phaedo argument,
defend the proposal that Plato’s argument at the end of Republic V relies
on the supposition that Forms are essences, and on the view, argued for in
the Hippias Major, that what certain qualities are, their essence and Form,
cannot be specified by example and exemplar and all that follows from
that? By ‘all that follows from that’ I mean, in particular: that what certain
qualities are, their essence and Form, cannot be perceived by the senses
and that, if one thinks that what certain qualities are, their essence and
Form, can be specified by example and exemplar, and can be perceived by
the senses, then one will be committed to things displaying co-presence of
opposites in regard to these qualities.
At this late hour, I shall not belabour the answer: Yes, if we can defend
this proposal for the Phaedo argument, we can defend it for the Republic
argument, too. For, in regard to their reliance on the view that sense-
perceptible things are subject to apparent co-presence of opposites in
regard to certain qualities, there is nothing to tell the two arguments, the
Republic argument and the Phaedo argument, apart.
Furthermore, there are particular reasons, more readily apparent than in
the case of the Phaedo argument, for thinking that the Republic argument
relies on the content of our proposal.
First, the lovers of sounds and sights, to which the Republic argument is
addressed, are characterised, from the start (d ff.), as thinking that beauty
is apparent to the senses and that this is all there is to an understanding of
beauty. The similarity to the argument in the Hippias Major is remarkable.
For that argument is about what beauty is; and Hippias is presented as
defending (not simply assuming) the view that what this quality is is
apparent to the senses and that what this quality is can be specified by
example and exemplar and by reference to what is apparent to the senses.
Secondly, the Republic argument is conducted on the basis of what the
lovers of sounds and sights are committed to and on the basis of premises
that they are prepared to accept (see d–e); which, crucially, includes
their view that all there is is sense-perceptible things and constructs of such
things. It follows that all we may suppose is that: If it is thought that there
are only sense-perceptible things, then sense-perceptible things are subject
to co-presence of opposites in regard to these qualities. We may not
suppose that, for Plato, sense-perceptible things are subject to co-presence
of opposites in regard to these qualities. This is, I submit, hugely impor-
tant and contradicts a now standard and orthodox view in the literature,
according to which, for Plato, this world of the senses is subject to and
‘suffers from’ co-presence of opposites (see n. ).

Forms as distinct from sense-perceptible things 
Again, the similarity to the argument in the Hippias Major is remark-
able. For, in that argument, that the beautiful things we perceive by the
senses will also appear – and, according to Hippias, be – ugly, was
presented as a consequence of Hippias’ view, that what beauty is can be
specified by example and exemplar and by reference to what is apparent to
the senses. Hippias is an original and archetypal lover of sounds and sights
in Plato.
I recall, one last time, what such co-presence meant in the Hippias
Major argument, and what I have argued that it means in the Phaedo
argument, and what, we may conclude, it means in the Republic argument.
The meaning is this: The same things that appear beautiful when com-
pared to one example and exemplar, and so to one standard, of beauty, will
appear ugly when compared to another example and exemplar, and so to
another standard, of beauty, and this because the one example, exemplar
and standard (such as the menial pot) is not beautiful, and may even be
positively ugly, when compared to the other example, exemplar and
standard (such as the girl, horse or lyre). I submit that having this meaning
of co-presence of opposites in mind, in regard to such qualities as beauty, is
exceedingly helpful when reading the Republic argument, which extends
these qualities so as to include, in addition to beauty, also justice, piety,
double, large and heavy.

 
Why are essences, or Forms, the basis of all

causation and explanation?
Phaedo –
In the Phaedo, e–c, Plato argues that essences or Forms are the basis
of all causation and explanation. This, it seems to me, is the right way to
formulate the conclusion of this supremely important argument. The
conclusion is not simply that essences or Forms are causes; for it is also
that causes are, or imply, essences or Forms. Nor is the conclusion that
causes are essences or Forms; for Plato argues that causes, provided that
they are grounded in essences or Forms, may also involve other things.
Above all, as I shall argue, Plato’s argument for this conclusion depends
crucially on the supposition that Forms are, precisely, essences.
It is right to think that Plato’s concept of aitia here should be under-
stood as referring to both causation and explanation. As we shall
see, Plato’s principal concern in this argument is a general understanding
of ‘Why?’ questions and what is required for an answer to a ‘Why?’
question to be adequate: to be, genuinely, explanatory. His
concern, therefore, is with what it is to explain something and with
explanation. At the same time, his concern is with what an answer to a
‘Why?’ question designates, if the answer is adequate and true. His
concern, therefore, is with causes or, as I shall generally prefer to say,
explanantia, and, of course, with those things that they cause and explain,
that is, effects or explananda. Plato is just as much concerned with
causation as with explanation.
The thesis of the present chapter is that Plato’s argument, for the conclu-
sion that Forms are the basis of all causation and explanation, relies on just
two premises. First, the supposition that Forms are, precisely, essences.
And, secondly, the supposition that says that explanations must be

For this point, see Michael Frede (, esp. –, ). Frede offers a general account, both
historically and more generally, of the distinction between causation and explanation and of the
relation between the two. For assessment and criticism of Frede’s account, see Ledbetter ();
Wolfsdorf ().


Forms as the basis of causation and explanation 
uniform: same cause/explanans if, and only if, same effect/explanandum
(see Section . for a full formulation).
If this account of Plato’s argument is correct, then Plato’s has a very real
possibility of being a cogent and compelling argument and an argument
that deserves being taken altogether seriously. For we – most of us at any
rate, and unless we have been overly influenced by Humean-style argu-
ments to the effect that the principle of the uniformity of causation cannot
be rationally grounded – will, on reflection, agree to the uniformity
requirement of causation and explanation. And, if the efforts of this
chapter are not in vain, we will be able to follow, step by step, how, from
this requirement – the requirement of the uniformity of causation and
explanation – and the need to find a way of securing that causation and
explanation conforms to it, Plato arrives at the conclusion that essences
and Forms are the basis of all causation and explanation.
The philosophical significance of Plato’s argument is paramount. Plato
attempts to establish a necessary connection between ‘Why?’ questions and
‘What is it?’ questions; these being two important types of question of
which we do not, pre-philosophically, or as a matter of course, suppose
that they are connected. The connection between these two types of
question, which Plato’s argument attempts to establish, goes well beyond
the point, which he has made in dialogues before the Phaedo (e.g., in
Hippias Major a through the use of di’ ho) and which says that an
answer to a ti esti question, if this question is understood as a request for a
standard (paradeigma) for a thing’s being F (irrespective of what form the
answer takes and of whether or not it conforms to certain stringent
requirements) will be explanatory in a certain way: a thing will be
F because it conforms to the standard for a thing’s being F. All that is
implied by that familiar point is that, if we want to know what a thing is,
its essence, then we want to know a certain kind of explanation. Plato’s
argument in the Phaedo is much stronger, for its conclusion implies that if
we want to explain something and know why a thing has a certain quality,
then we must look for what that quality is, its essence. To put it simply:
Plato’s present argument derives the need for the notion of essence from
the notion of explanation, whereas what he had done earlier was observe
that the notion of essence implies the notion of a certain kind
of explanation.
On our account of Plato’s argument, its conclusion is not about a
certain kind of causation or explanation – sometimes referred to, in the
literature and apparently due to a tendency to Aristotelianise Plato, as
formal causes – for it is about all causation and explanation and about

 Plato’s Essentialism
causation and explanation as such. This is because, on our account, Plato’s
argument is based on a single premise, that causation and explanation
must conform to the uniformity requirement; and this requirement, as we
shall see, is intended by Plato to be about all causation and explanation and
about causation and explanation as such – it is about the very nature of
causation and explanation.
It is worth pointing out that, on our account of it, Plato’s argument
amounts to an argument for believing in essences: unless we believe in
essences, we cannot explain anything; at any rate, we cannot explain
anything in a uniform way; or, at any rate, we cannot explain anything
in a uniform way, if we think that the uniformity requirement of causation
and explanation must, to be at all valid, be grounded in something, as
opposed to simply being imposed on things by us for the benefit of
satisfying our scientific or practical needs and desires. No doubt Plato’s
argument for believing in essences will not satisfy certain Humeans, who
prefer to be sceptical about the possibility of grounding the uniformity
requirement of causation and explanation in anything, than to believe in
such things as essences. And it will not generally satisfy Kantians either.
But for those who are not in fear of essences, Plato’s argument may even
be compelling.
. What is new?

So much has been written on this argument from the Phaedo, especially in
the past fifty or so years, that the reader may naturally wonder how there
can be anything new or significant left to say. I believe the present account
of Plato’s argument contains two very significant claims that stand out
from the current scholarship. First, Plato’s claim is about Forms in the
sense of, precisely, essences – an essence being that which is designated by
an adequate and true account of what a thing or quality is. Plato’s claim,
therefore, is that: Essences are the basis of all causation and explanation.
The term eidos and the notion of a Form, if these are understood to mark a
substantive and not merely verbal distinction to the notion of an essence
and the terms used by Plato for this notion, have no place in, and make no
difference to, Plato’s argument for the conclusion that essences and Forms
are the basis of all causation and explanation. Secondly, contrary to
what many critics think (e.g., Annas ; Fine ; Meinwald ;
Sedley ), the supposition (sometimes referred to in the scholarship as
‘the transmission principle of causation’), which says that causation works
by the cause transmitting its character to the effect, and hence that the

Forms as the basis of causation and explanation 
cause must be like the effect (so that hot things are caused to be hot by
other things that are hot), has no place in and is not part of
Plato’s argument.
It seems to me that these distinctive features of our account of Plato’s
argument are of particular significance for understanding Plato’s account
of causation and explanation. Current scholarship generally understands
Plato’s argument as being about, specifically, Forms. This is not to say that
critics do not recognise, or that they deny, that, among the many things
that are characteristic of Plato’s Forms, Forms are also essences. What
I mean is that in the current scholarship it is not generally supposed that it
is, precisely, this feature of Plato’s Forms – that they are essences – that is
the crux of Plato’s argument and that, single-handedly, sustains the argu-
ment. If anything, critics think it is the transmission principle of causation,
in conjunction with the idea that a Platonic Form has the quality of which
it is the Form (e.g., the Form of heat is hot), that sustains the argument:
the now standard view is that Plato’s argument relies on the transmission
principle of causation plus the supposition that Forms are self-predicative.
This account, I argue, misrepresents Plato’s argument. All we need in
order to understand Plato’s argument, I argue, is, first, that Forms are
essences, and, secondly, that explanations must be uniform, that is, must
conform to the principle: same explanandum/effect if, and only if, same
explanans/cause. The principle of the uniformity of causation and expla-
nation does not at all imply the (so-called) transmission principle
of causation.
This feature of current scholarship on Plato’s argument – that critics do
not generally suppose that, for the purpose of this argument, Forms are
precisely essences – has the consequence, it seems to me, that it is not at all
clear what these things, the Forms, of which Plato argues that they are
causes, are supposed to be, other than that they are, for certain particular
reasons, things suitable for being causes. It seems to me that this conse-
quence amounts to a serious problem for the now standard view. Suppose
two philosophers agree on the transmission principle of causation but
disagree on what the causes are that satisfy this principle: the one thinks

I follow those critics who think that the transmission theory of causation implies that the cause has
the quality that it transmits to the effect. I am supposing, therefore, that, if Forms are causes in
accordance with the transmission theory of causation, then Forms are self-predicative. Meinwald is
especially clear on this: ‘[Socrates] is thinking that when something acquires a property, that property
is transferred to the participant by coming with a participant’s share of a form that itself has that
property’ (, ; original emphasis). Meinwald is here commenting on the Parmenides, but she
thinks this view goes back to Phaedo-Republic.

 Plato’s Essentialism
they are Platonic Forms, the other disagrees. It seems to me that we cannot
make sense of this dispute unless we have a firm grip on what Forms are.
For the purpose of such a grip, it is not adequate to specify that a Platonic
Form is not only an entity that is a cause, and a cause conforming to the
transmission principle of causation, but also an entity that is simple,
changeless, imperceptible by the senses, knowable only by reason, self-
predicative, separate from physical things, and even . . . an essence. This
mode of specification is not adequate, unless we can be confident, or we
think Plato has reason to be confident, that these several characteristics all
refer to one and the same thing, Forms.
What about the (so-called) transmission principle of causation? Or the
claim that Forms are self-predicative? None of this is present, I submit, in
the text of the Phaedo; and we shall see that none of it is necessary for
understanding Plato’s argument. What is present in the text is a single line
that says that it would be absurd to explain why something is large by
appeal to something small:
I think you would be afraid that some opposite argument would confront
you if you said that someone is bigger or smaller by a head, first (prōton),
because the bigger is bigger and the smaller smaller by the same, and further
(epeita) because the bigger is bigger by a head which is small, and this would be
strange, namely, that someone is made bigger by something small. Would you
not be afraid of this? I certainly would, said Kebes, laughing. (a–b;
trans. Grube, adapted)
Should we, on the strength of this line (i.e., the line I have italicised,
a–b), attribute to Plato the transmission principle of causation and
the view that Forms are self-predicative? If Plato’s argument depended on
our doing so, perhaps we might, though at a considerable stretch of what
Plato actually says. But we shall see that Plato’s argument is not at all based
on these claims. And let us note just how weak this evidence is. The line
makes a negative point: that we must not suppose that the cause of a
thing’s being F is something that is opposite to F. It does not follow that
we must suppose that the cause of a thing’s being F must be something
that is itself F. The claim that Plato is in the process of defending – and
which in fact he has already defended – when this line comes up, says that
we must suppose that the cause of a thing’s being F is based in the Form of
the quality F and that we must suppose this if we want causation and
explanation to conform to the requirement of the uniformity of causation
and explanation. This claim is compatible and consonant with the claim
that we must not suppose that the cause of a thing’s being F is something
that is opposite to F; and it does not at all imply the transmission principle.

Forms as the basis of causation and explanation 
It is worth noting that this line, a–b, comes up some time after
Plato has articulated the requirement of the uniformity of causation and
explanation (it was articulated from a–b); for it comes up when he
sums up this articulation, and relates it to the proposal, which he is
introducing at this point, that causes are, basically, essences and Forms
(i.e., at e–d). It is clearly indicated that what this line says is not
intended to be part of the articulation of the uniformity requirement; for,
the line that precedes it (‘first, because the bigger is bigger and the smaller
smaller by the same’) and that serves to refer back to, and to remind the
reader of, the requirement of uniformity, is set off against it by the ‘first’
(prōton), while the line in question is marked by the ‘further’ (epeita).
Nothing in the articulation of the uniformity requirement, when this
requirement was properly articulated earlier (i.e., at a–b), indicated
this addition or the need for it.
Why, it will be asked, does Plato have Socrates utter the line that he
utters at a–b, if not to reveal his commitment to the transmission
principle of causation and to the view that Forms are self-predicative? This
is a good question. But an answer is readily available. Socrates has been
articulating the principle of the uniformity of causation and explanation
for the benefit of his friends – and Plato for the benefit of his audience – by
spelling out that if we explain why a thing is F by appealing to another
thing, the cause, that is F, then we must not also explain why other things
are F by appealing to a cause that is opposite to F; we must not do this, or
we will contravene the principle of the uniformity of causation and
explanation: same cause/explanans if, and only if, same effect/explanan-
dum. But this does not at all imply that we must explain why a thing is
F by appealing to another thing, the cause, that is F.
Why be so opposed, as I am, to the idea that Plato’s argument relies on
the transmission principle of causation? One reason is that this (so-called)
principle has nothing in particular to do with Forms or with what Forms
basically are. Taking Plato to be relying on this principle is another
example of the tendency in the scholarship of thinking of Plato’s theory
of Forms as one of those remarkably protean Swiss Army knives, to which
one can add any and every additional attachment and still rightfully call it
by the same name, ‘The Perfect Tool For All Your Needs’.
Another reason for opposing the idea that Plato’s argument relies on the
transmission principle of causation is that, as is familiar, and as we shall
see, a large part of the argument is made up of Socrates’ complaint that the
explanations propounded by physicalist thinkers have left him utterly

 Plato’s Essentialism
puzzled – indeed ‘seriously blinded’ (cf. sphodra etuphlōthēn). We shall see
that the reason why they left him utterly puzzled is that he thinks that such
explanations do not provide a means of satisfying the requirement of the
uniformity of causation and explanation and that without such a means no
apparent explanation is genuinely explanatory. But it is hard to see how
Socrates could seriously intend this complaint – as he clearly does, no whiff
of irony here – if he were committed to the transmission principle of
causation. For that principle is, if anything, the intellectual property of
physicalist thinkers; it is they, if anyone, who think that the cause must be
like the effect. If Plato’s intention were to take over this principle, and to
adapt it in a way suitable for his theory of Forms, he could hardly complain
that physicalist explanations, by themselves, leave us entirely puzzled and
in the dark. He could at most complain that physicalist explanations,
though they get the logic and structure of causation basically right, need
to be suitably adapted – a complaint whose moderation stands in stark
contrast to the radical character of Plato’s actual complaint in the Phaedo.
. Plato’s aporia about explanation (Phaedo e–d)

Socrates opens the narration of his intellectual journey with the remark:
A thorough examination is, therefore, needed of the explanation of gener-
ation and destruction in general. (e–a)
This examination is needed because, in spite of the extended discussion
and argument (going back virtually to the beginning of the dialogue, at
b), no agreement has been reached on the question under discussion,
whether the soul is or is not subject to generation and destruction.
Socrates, therefore, proposes to make another attempt at answering this
question, by shifting the attention from the question whether, in partic-
ular, the soul is or is not subject to generation and destruction, to the more
general and radical question: Why is any thing that is (or is not) subject to
generation and destruction subject to (or not subject to) generation and

See c–, not to be confused with the more general blinding image at e.

I read the scope of the holōs to be peri geneseōs kai phthoras. The point of the holōs is to shift the
discussion from the question whether a particular thing, the soul, is or is not subject to generation
and destruction, to the more general and radical question: Why is any thing that is (or is not) subject
to generation and destruction subject to (or not subject to) generation and destruction? I read the gar
to mean ‘therefore’, rather than ‘for’. The point of the gar is that, since the extended discussion and
argument (going back to b) has failed to establish agreement (on the question whether or not the
soul is subject to generation and destruction), a more general and radical examination is needed (dei).

Forms as the basis of causation and explanation 
destruction? Eventually he will apply the results of this examination to the
soul (c f.).
To understand why the dialogue takes this more general and radical
turn, we must ask why Socrates turns to the question of why any thing that
is (or is not) subject to generation and destruction is (or is not) subject to
generation and destruction. It is here that the general appeal to aitiai and
the question ‘Why?’ is introduced for the first time in the dialogue. It
appears that this turn of the dialogue is motivated by the outcome of the
extended preceding discussion and argument about the soul and the
question of its immortality. This goes back virtually to the beginning of
the dialogue (b), when Socrates begins defending the claim that the soul
is immortal, and, in particular, it goes back to Kebes’ initial challenge (at
e–b) of Socrates’ initial defence.
Socrates’ narration of his intellectual journey is a response to Simmias’
and Kebes’ views about the relation between soul and body, and these
views have emerged in the course of, and as a result of, this extended
discussion and argument. On their views, the soul depends on the body:
either on the individual body (which is Simmias’ view, see e–d) or on
the body in general (which is Kebes’ view, e–b). From this, they infer
that, since the body is subject to generation and destruction, the same
must be true of the soul (this was the point of Kebes’ initial challenge, at
e–b). It appears, therefore, that they think that the relation between
the soul and the body not only implies that, it also explains why, the soul is
subject to generation and destruction: because it depends on something
that is subject to generation and destruction. It is, therefore, natural and
appropriate that Socrates, when he broadens the discussion, should take up
the question concerning the explanation of generation and destruction. It
is also notable that, as soon as he begins the narration, he focuses especially
on those explanations that appeal purely to physical things (b–c),
which is the kind of explanation favoured by Simmias and Kebes.
What, we ought to ask, is distinctive of the explanations favoured by
Simmias and Kebes – what I have been calling ‘physicalist’ explanations,
that is, explanations that appeal purely to ‘physical’ things, things in space
and time? Socrates refers to such explanations generically as the explana-
tions traditionally favoured by those who engage in ‘the enquiry into

That the question is, in the first instance at least, about the explanation of things that are subject to
generation and destruction is also indicated by the repeated use of the phrase: ‘why something is, or
comes to be, or ceases to be’ (a–, b–, c). The issue is, in the first instance at any rate,
about the enquiry into natural phenomena (cf. peri phuseōs historian, a).

 Plato’s Essentialism
nature’ (phusis, a–). He goes on (b–c) to clarify what is distinc-
tive of such explanations, by citing some examples of the things appealed
to in them as being aitiai: the hot and the cold; blood, air and fire; the
brain. Does Plato provide a general account of such things – ‘physical’
things? Here (a–) he describes these things as things that are subject
to generation and destruction; and in general he describes such things as
the things that are, first, capable of being perceived and, secondly, subject
to change (or, the kind of change that is capable of being perceived; see
e.g., e–a). We may, therefore, also introduce the notion of a ‘physical
constituent’ to mean a constituent of a thing that is capable of being
perceived and subject to change (or, the kind of change that is capable of
being perceived). For it is notable that blood, air, fire and the brain, as well
as (later) bones and sinews and (later again) fire and snow, would appear to
be physical constituents of things in just this sense.
We may note, incidentally, that the introduction, at this point of the
dialogue, of a general appeal to aitiai and the question ‘Why?’ accords
with, and hence would also appear to prepare for, the methodological
recommendation later in the narration, when Socrates says that we ought
to investigate the truth about any thing not directly but through state-
ments (logoi) and, in particular, statements that put forward explanatory
hypotheses (see d–a). For, it appears that what this methodolog-
ical recommendation says is that an adequate investigation of whether p is
the case (e.g., whether the soul is immortal) must involve an investigation
of why p is the case, if it is the case, and why p is not the case, if it is not
the case (e.g., why the soul is immortal, if it is, and why the soul is not
immortal, if it is not).
Socrates begins the narration of his intellectual journey with the ques-
tion: What is the explanation of generation and destruction? But when we
look closer at the narration, it emerges that the question he wants to
address includes the more radical one: What is an explanation, whether
of generation and destruction or of anything else? He appears to move
directly from the former to the latter question. For he begins by recounting
how he lost trust especially in those explanations of generation and
destruction that appeal purely to physical things (b–c) but immedi-
ately goes on to describe how this directly led him to lose trust in
explanations altogether, even such familiar and apparently unobjectionable
ones as that human beings grow because they eat and drink (c–d) or
that ten is greater than eight because it is the sum of eight and two
(e–). He must, apparently, think that the question about the expla-
nation of generation and destruction is directly associated with a more

Forms as the basis of causation and explanation 
radical aporia about explanation, the kind of aporia it would be natural to
associate with the Socratic-type question: What is an explanation?
This reading will be corroborated when we consider why Socrates lost
trust in physicalist explanations. He did so, we will see, because he came to
think that such explanations do not satisfy certain fundamental require-
ments of explanation. These requirements, we will see, are not specifically
about physicalist explanations, and they are not specifically about the
explanation of generation and destruction, rather, they are about any
explanations and about explanation itself. This shows why he lost trust
in explanations altogether, even familiar and apparently unobjectionable
ones. He did so because he came to doubt that any familiar explanation
satisfies these requirements of explanation.
It is the following two requirements of explanation that traditional
explanations fail to satisfy:
REQ If same explanandum, then same explanans;
that is, it is impossible for numerically different particular
things, in so far as they have the same quality, to have different
and incompatible explanantia;
that is, it is necessary that, if E is the explanans of why O is F,
and E is the explanans of why O’ is F (where O and O’ are
numerically different particulars), then E = E.
REQ If same explanans, then same explanandum;

Socrates does not formulate this aporia in so many words, but he describes his intellectual journey as
a path from youthful confidence in explanations of natural phenomena to a state, as he grew in
maturity, of loss of trust in such explanations and indeed in explanations altogether. It is, I think,
natural and appropriate to describe this as a path from his being concerned entirely with the question
‘What is the explanation of this or that?’ to his being motivated, because of a particular aporia, to
concentrate rather on the question ‘What is an explanation?’ We may note that, in a later dialogue,
the Sophist (b–), Plato uses the very same topos, the contrast between youthful confidence
and mature aporia, to describe a similar path: from being concerned entirely with the question ‘What
things are there?’ to being motivated, because of a particular aporia (expressly so called), to
concentrate also on the question ‘What is being?’.

See Fine (, –); Annas (, –; see also Annas, n.  for further critics of a similar
view. My account differs in a number of significant ways. First, I consider how these requirements
ought to be formulated. Secondly, I consider why it is supposed that physicalist explanations do not
satisfy them. Thirdly, I argue that the same charge is levelled at teleological explanations as
traditionally conceived. Fourthly, I consider what is the motivation behind these requirements and
why they are supposed to be plausible. I argue that they are well-motivated but not because of the
principle sometimes referred to as the transmission principle of causation. Finally, I consider how it
is that explanations in conformity with the account that says that explanantia are primarily essences/
Forms is supposed to satisfy them. Here I argue that Fine’s and Annas’ responses are not satisfactory,
unless, as Annas thinks we are forced to do, we add the transmission principle of causation.

See e–b, read together with b–c.

 Plato’s Essentialism
that is, it is impossible for numerically different particular
things, in so far as they have different and incompatible qualities,
to have the same explanantia;
that is, it is necessary that, if E is the explanans of why O is F,
and E is the explanans of why O’ is G (where O and O’ are
numerically different particulars), then F = G.
It is appropriate to characterise these requirements as embodying the
constraints of the uniqueness, generality and uniformity of explanation.
They are requirements of the uniqueness of explanation because they
imply that each explanandum has a single explanans, the one proper to
it, and conversely each explanans explains a single explanandum, the one
proper to it. They are requirements of the generality of explanation
because they imply that, if a particular thing, O, has an explanation, it
does so on account of having a general quality, F. This is a general quality
in the sense that numerically different particular things, O and O’, can
have the same quality, F. What the requirements assert is that explanations
must be uniform, indeed, to articulate the constraint that explanations
must be uniform is precisely to state these requirements.
It appears that these requirements are supposed to be requirements
precisely of explanation, that is, it is in so far as something is an explana-
tion that it must satisfy them. Otherwise their apparent breach would not
be grounds for Socrates’ loss of trust in explanations altogether, even in
such familiar and apparently unobjectionable ones as that human beings
grow because they eat and drink or that ten is greater than eight because it
is the sum of eight and two. If, on the other hand, they are indeed
supposed to be requirements of explanation in general and as such, that
is, requirements that contribute to the very essence of explanation and to
the question ‘What is an explanation?’, then Socrates’ being in a state of
radical aporia is readily intelligible. For, by coming to think that traditional
explanations, whether the sophisticated ones offered by physicalist philos-
ophers (e.g., b–c) or everyday ones (e.g., c f.) fail to satisfy these
requirements, he came to think that traditional explanations are not
explanations at all.
Remarkably, no general and abstract defence, indeed no general and
abstract formulation, is offered of these requirements. They are, rather,
articulated and made plausible entirely through the appeal to apparent
examples of their breach and, no less, through the appeal to Socrates’

See e–a (also d–e), read together with e–b..

Forms as the basis of causation and explanation 
reaction of radical aporia in the face of these examples – a reaction we are
invited to recognise as natural and reasonable. What we gather from
Socrates’ narration of his intellectual journey is that the apparent breach
of these requirements by traditional and, in particular, physicalist expla-
nations is what led him to lose trust in such explanations and to suspect
that they are not explanations at all. He even lost trust in explanations
altogether, presumably because, and for as long as, he thought that he did
not know how to satisfy these requirements.
Let us for a moment look at how, in particular, these requirements are
supposed to emerge through the appeal to apparent examples of their
breach. In the argument at e–b, Socrates considers whether we
may suppose that the explanation of why some particular thing, O, ‘comes
to be two’ is that some other thing is added to it and joined together with
it. And he objects (a–b) to this attempted explanation on the grounds
that, if it were an adequate explanation, we could expect to find ourselves
in the position, in regard to some other thing, O’, of having to say that the
explanation of why this thing ‘comes to be two’ is that it is divided and
something is taken away from it. The objection is that this latter explanans,
the one in terms of division (schisis, a), is ‘contrary’ (enantion, a, see
below for the meaning of enantia) to the original one, the one in terms of
addition (prosthesis, a), but that one and the same explanandum,
coming to be two, is involved. The objection, therefore, relies crucially
on the appeal to the requirement, if same explanandum (coming to be
two), then same explanans, which is articulated and made plausible
through the appeal to an apparent example of its breach.
In the argument at e–a, Socrates considers whether we may
suppose that the explanation of why he, Socrates, stays imprisoned in
Athens and does not run away to Megara or Boeotia, is his bones and
sinews and their condition. And he objects to this attempted explanation
on the grounds that, if it were an adequate explanation, it could equally
have explained why the same thing, he Socrates, had run away to Megara
or Boeotia, in the hypothetical situation in which he had judged that it is
better to run away than to stay imprisoned. The objection, therefore,
appears to rely crucially on the appeal to the requirement, if same expla-
nans (the particular constituents of a particular body), then same expla-
nandum, which is articulated and made plausible through the appeal to an
apparent example of its breach.
This famous argument demands particular care. As I read it, its main
purpose is to exhibit the requirement, if same explanans, then same
explanandum, by appealing to an apparent example of its breach. On a

 Plato’s Essentialism
common reading, however, its purpose is simply to reject physicalist
explanations, that is, explanations that appeal only to physical things, in
favour of teleological or good-based ones. Let us briefly compare the two
readings. In general, I agree that the purpose of the argument is to reject
physicalist explanations in favour of teleological or good-based ones –
though we must not forget that traditional good-based explanations will
disappoint Socrates no less – but I think this purpose is supposed to be
achieved precisely by arguing that physicalist explanations fail to satisfy
those requirements of explanation.
First, if the purpose of the argument were simply to reject physicalist
explanations, explanations that appeal only to physical things and their
constituents, it would plainly fail. For, suppose a physicalist philosopher
asserts that the explanation of why a particular thing, O, behaves in a
particular way, W, is that it has a particular physical constituent, M. And
suppose another philosopher, for no other purpose than to reject physi-
calist explanations, objects that, if this were an adequate explanation, it
could equally have explained why this thing, O, behaved in a different
way, W*. Evidently all the physicalist philosopher needs to do, to defend
his favoured type of explanation against this objection, is respond that, had
this thing, O, behaved in a different way, W*, the explanation would have
been not that it has this physical constituent, M, but rather that it has a
different physical constituent, M*.
Secondly, it is of course true that, in this argument and its context (i.e.,
b–c), Socrates objects to physicalist explanations. But his grounds
for objecting are precisely that physicalist explanations do not satisfy those
requirements of explanation, and in particular the requirement, if same
explanans, then same explanandum. We should note that the force of the
objection is not that a physicalist philosopher could not simply require of
us that we formulate our statements, or hypotheses, of physicalist expla-
nations in conformity with these requirements. Evidently, he could require
this, and we could try our best to comply – as does every good Humean,
Kantian and pragmatist. Of such a physicalist philosopher we might say,
echoing a Kantian slogan, that he would conceive of the requirements as
being only regulative principles of scientific enquiry, not constitutive
principles of things. The force of Socrates’ objection is that nothing in
what it is to be a physicalist explanation, that is, an explanation that

See, e.g., Taylor (, –). Likewise, Annas (, ); Fine (, ), who suppose that
the appeal to teleological explanations is independent of the objection against physicalist
explanations based on their not satisfying the above requirements.

Forms as the basis of causation and explanation 
appeals only to physical things, implies that physicalist explanations satisfy
these requirements and that, therefore, we (or the physicalist philosopher)
have nothing to appeal to for the purpose of justifying the claim (regarding
whether physicalist explanations satisfy the requirement, if same explanans,
then same explanandum):
Q. Necessarily: if a particular thing, O, is F because it has a particular
physical constituent, M (i.e., its having this physical constituent, M, is the
entire explanation of why it is F), then it is impossible for a numerically
different thing, O’, if it has a quality, G, that is different from and
incompatible with the quality F, to be G because it has that same physical
constituent, M.
(An analogous formulation may be offered of the corresponding claim
regarding whether physicalist explanations satisfy the converse require-
ment, if same explanandum, then same explanans.) But evidently this
claim, Q, needs justification. For it is not at all obvious, nor does it in
any way go without saying, that physicalist explanations, or indeed any of
the explanations with which we are readily familiar (such as that human
beings grow because they eat and drink or that ten is greater than eight
because it is the sum of eight and two), satisfy these requirements of
explanation. This, I submit, is the force of Socrates’ objection that phys-
icalist explanations, and indeed any of the apparently unobjectionable
explanations with which we are readily familiar, fail to satisfy the require-
ments of explanation.
This means, we should note, that Plato, at this point, leaves open the
possibility that, on a reformed conception of them, explanations that
appeal to physical things may, provided that this is not all that they appeal
to, satisfy these requirements – in which case they will be of good
standing. We will see that Plato defends such a reformed conception;
for he argues that if physical things are appropriately related to essences,
and if essences are conceived of as the primary elements in explanantia,
then physical things can be elements in explanantia and can, therefore, be
genuine parts of explanations.
Thirdly, it is, of course, true that, in this argument and its context (i.e.,
b–c), Socrates sets teleological, or good-based, explanations against
physical ones and objects against the latter but holds out hope for the
former. But it is equally true that his hopes, to the extent that they were
founded in good-based explanations traditionally available, were dashed

See Larsen (, ).

 Plato’s Essentialism
(b) and that, as a result of this, he turned to essence-based or Form-
based explanations. We must ask, therefore, why his hopes in good-based
explanations were dashed and why he does not fear that his turn to
essence-based or Form-based explanations may suffer a similar fate.
I shall address these questions at the end of the chapter, and I shall argue
that the reason why Socrates despairs of traditional good-based explana-
tions, like the reason why he despairs of traditional physicalist explana-
tions, is that he thinks that, as traditionally conceived, they do not satisfy
those requirements of explanation, the uniformity requirements.
How does Plato intend that these requirements should be formulated?
What Socrates claims, when he describes how traditional explanations
breach these requirements, is that the same explanandum cannot be
explained by contrary (enantia) explanantia, and that, conversely, the same
explanans cannot explain contrary explananda (see, e.g., a–b and
a–b). What does he mean by ‘contrary’ here? It appears that he
means ‘different and incompatible’. Explananda are contrary if they ascribe
different and incompatible qualities to one and the same particular thing.
For example, the explananda Socrates’ staying in Athens and Socrates’
running away to Megara (see e–a) are evidently contrary in this
sense. In general, two qualities are incompatible if, and only if, one and the
same particular thing cannot (at the same time, in the same respect, etc.)
have both qualities. Similarly, explanantia are contrary if, and only if, they
explain things in terms of different and incompatible kinds of things. In
general, two kinds of things are incompatible if, and only if, one and the
same particular thing cannot instantiate both kinds. For example, the
explanations in terms of the appeal to addition (a) and division
(a), which are expressly referred to as contrary (enantia, a), are
contrary in this sense. For the same particular thing, in this case it is a
particular process, cannot at once be the process of addition and process of
division. Socrates’ claim, therefore, is that the same explanandum cannot
be explained by different and incompatible explanantia and that, con-
versely, the same explanans cannot explain different and incompatible
explananda.
Is the lack of a general and abstract defence, indeed formulation, of
these two requirements of explanation (i.e., REQ and REQ) a defect in

For convenience, we will often say simply that the same explanandum cannot have different
explanantia (and the same explanans cannot explain different explananda) to mean that the same
explanandum cannot have different and incompatible explanantia (and the same explanans cannot
explain different and incompatible explananda).

Forms as the basis of causation and explanation 
Plato’s argument? That depends on what the aim and ambition of the
argument is supposed to be. If it were to demonstrate the requirements, or
in general to address the question of their ultimate plausibility and their
truth, then certainly this would be a defect. It appears, however, that the
aim is a different one, namely, first, to introduce the requirements in a way
that exhibits their immediate plausibility and secondly, to appeal to them
in order to defend a particular account of explanation, namely, the account
that says that explanantia are, primarily, essences. In that case, a general
and abstract defence of the requirements is not part of Plato’s aim. Indeed,
it appears that we ought to understand his overall conclusion as having the
form of a conditional: on the supposition of the truth of these require-
ments of explanation, it follows that explanantia are, primarily, essences. If
this is the force of his overall conclusion, then what is required in order to
avoid the impression that the whole thing is merely hypothetical is,
precisely, some appropriate indication of why it is plausible to believe in
these requirements, and this is just what the narration of Socrates’ intel-
lectual journey aims at providing. In that case, it may also be appropriate
that no general and abstract formulation is given of them. For it is
plausible to think that it is precisely through particular examples, examples
that are supposed to strike us by their exemplifying, or failing to exemplify,
this or that general principle, that general principles first become familiar
and plausible to us. It may also be appropriate that the requirements
should be articulated and made plausible primarily through the appeal to
apparent examples of their breach. For it is plausible to think that it is
precisely when it appears to us that some particular case fails to satisfy
them that these requirements first strike us. As long as things run
smoothly, we need not wonder how they work.
We must ask, finally, why Plato thinks that these requirements are
plausible, and why he thinks they belong to the essence of explanation.
Apparently, he thinks that they are plausible because he thinks that:
INTELL If things do not satisfy these requirements of explanation (i.e.,
REQ and REQ), then their behaviour is not intelligible (or, what is
equivalent, if the behaviour of things is intelligible, then they satisfy these
requirements).

By ‘intelligible’ I mean ‘subject to explanation’ (see below for a defence of this). It is worth pointing
out that Plato is not committed to the converse claim: If things satisfy these requirements of
explanation, then their behaviour is intelligible. For this claim would depend on the view that these
(i.e., REQ and REQ) are the only requirements of explanation, and we do not have reason to
think that Plato holds this view.

 Plato’s Essentialism
The reasoning behind this central claim would appear to be along the
following lines:
P If numerically different things, though they have the same qualities,
have different explanantia, then we want to ask why this is, that is,
what is the relevant difference in the things that explains why they
have different explanantia; and likewise, if numerically different
things, though they have different qualities, have the same explanans,
then we want to ask why this is, that is, what is the relevant similarity
in the things that explains why they have the same explanans.
P Unless we suppose that there is an answer to this question, the
behaviour of the things will positively appear not to be
intelligible. But
P we can be satisfied that there is an answer to this question if, and only
if, we suppose that these requirements (i.e., REQ and REQ), are
true, and that they are true of the things in question.
It follows that,
INTELL If things do not satisfy these requirements, then their behaviour
is not intelligible (or, what is equivalent, if the behaviour of things is
intelligible, then they satisfy these requirements).
Moreover, if we suppose that there is an essential relation between a thing’s
being subject to explanation and the thing’s being intelligible (I will
consider this supposition in a moment), it follows that these requirements
belong to the essence of explanation. This line of reasoning, I suggest, is
what is supposed to motivate and indicate the immediate plausibility of
these requirements and of the view that they belong to the essence
of explanation.
Let me clarify this line of reasoning and indicate how it is supported by
the text. The central claim, INTELL, says that if things do not satisfy these
requirements of explanation, then their behaviour is not intelligible (or,
what is equivalent, if the behaviour of things is intelligible, then they
satisfy these requirements). First, we must avert a possible misunderstand-
ing. This claim does not say, and it does not imply, that things are
intelligible or that they satisfy these requirements of explanation. What it
says is, rather, that if things are intelligible, then they satisfy these require-
ments of explanation. It would, therefore, be misplaced to think that
Plato’s argument rests on the supposition that nature (phusis) – the totality
of things subject to generation and destruction – is governed by reason
(nous) – reason as the source of intelligibility – and a mistake, therefore, to
think that Plato’s argument is of no philosophical or scientific interest to

Forms as the basis of causation and explanation 
those of us who have, since Francis Bacon and Descartes lost faith in the
idea that nature is governed by reason. If Plato’s argument rests on a
supposition about the relation between nature and reason, it is that, if
nature is governed by reason, then it is subject to these requirements of
explanation. The central claim, we may say, is about the essence of
explanation and of intelligibility, not about what things, if any, are subject
to explanation and are intelligible.
Secondly, what, in the central claim, is supposed to be the relation
between a thing’s being subject to explanation and the thing’s being
intelligible? Suppose that being subject to explanation and being intelligi-
ble are, precisely, one and the same thing. In that case, the claim says that
if things do not satisfy these requirements of explanation, their behaviour
is not subject to explanation. Is this uninformative or tautologous? I think
not. What is perhaps uninformative and tautologous is the claim that if
things do not satisfy any requirements of explanation (i.e., do not satisfy
whatever it may be that explanation requires), then their behaviour is not
subject to explanation. But the claim is genuinely informative, and far
from tautologous, which says that if things do not satisfy these require-
ments of explanation, viz. REQ and REQ, then their behaviour is not
subject to explanation. I do not, therefore, see any immediate objection
against supposing that a thing’s being subject to explanation and the
thing’s being intelligible are, precisely, one and the same thing. It also
seems plausible to suppose this, if we bear in mind the relevant meaning of
‘a thing is subject to explanation’, namely, ‘there is something, whatever it
is, which is that because of which (the di’ hoti, see, e.g., Phaedo b–) the
thing, or some quality of it, is as it is’. For it seems plausible to think that
there being something which is that because of which a thing, or some
quality of it, is as it is, and the thing’s, or some quality of it, being
intelligible, are, precisely, one and the same thing.
Thirdly, we must suppose that the relation between a thing’s being
subject to explanation and the thing’s being intelligible, even if it is not
that of identity, is a particularly close one. Otherwise the main claim,
INTELL, would not directly imply – as evidently it is supposed to imply –
that there is something wrong with supposing that a thing satisfies the
requirements of explanation (i.e., satisfies the requirements of explanation,
whatever they may be) but that, nevertheless, it is not the case that its
behaviour is intelligible. It seems plausible to think, moreover, that what is
wrong with supposing this is, precisely, that it is a necessary truth, and one
apparently rooted in the essence of explanation and intelligibility, that if a
thing satisfies the requirements of explanation (i.e., satisfies the

 Plato’s Essentialism
requirements of explanation, whatever they may be), then its behaviour is
intelligible. In general, this necessary truth indicates what, minimally, we
must suppose to be the relation between a thing’s being subject to
explanation and the thing’s being intelligible.
Let me comment briefly on the premises, P–, of the argument – its
validity being evident. In regard to Premise , I suppose it is evident that we
do want to ask this. My daughter once did, when she was six years old. She
asked me why I was wearing rubber shoes in the pool. I said, because of the
need to avoid foot infections. She objected, in her insistent fashion (still with
her today), that in the sea I would swim barefoot. I said something to the
effect that there is a difference, but that I would tell her what it is some other
time. It appears that what my daughter said was an expression of some grasp
of the requirement, if same explanans (the need to avoid foot infections), then
same explanandum (the wearing of rubber shoes). Of course, her grasp did not
involve, or imply, her being able to formulate this requirement in abstract
and general terms, it involved rather, if anything, her being sensitive to a case
of its apparent breach (and no doubt its real breach, as my behaviour is not
always intelligible). What is striking is that Socrates, when he describes the
radical aporia that befell him in his youth about explanations, describes his
reaction, in the face of the would-be explanations offered him by his elders,
philosophers and non-philosophers alike, as being similarly childlike – and
no less insistent. This is of course a deliberate literary device on Plato’s part,
and one that appears to be modelled on something like a topos, the topos of
the innocence and apparent simple-mindedness needed for philosophy (cf.
euēthōs echō par’ emautō[i], d).
In regard to Premise , the point is not, of course, that, unless we
suppose that we know the answer to the question ‘What is the relevant
difference (or similarity)?’, or that we are able to find this out, the
behaviour of the things will appear not to be intelligible. We may have
no clue what the answer to this question is, or how to discover it. The
point is that, unless we suppose that there is an answer to this question, the
behaviour of the things will appear not to be intelligible. That is, it will
appear that there is no that because of which (the di’ hoti) the things, or
some qualities of them, are as they are. Of course, we may indeed suppose
that the behaviour of the things is not intelligible. There need not be
anything puzzling about supposing that. What is puzzling – whence
Socrates’ radical aporia – is to suppose that the behaviour of the things is
not intelligible, while, at the same time, supposing that the things are
subject to explanation. We have seen that the reason why this is altogether
puzzling is that, apparently, it involves a contradiction.

Forms as the basis of causation and explanation 
Premise  ought to be evident. The supposition that precisely these
requirements (i.e., REQ and REQ) are true, and, moreover, that they
are true of the things in question, is precisely what allows us to justify the
supposition that there is an answer to the question ‘What is the relevant
difference (or similarity)?’
It is sometimes thought (see, e.g., Ruben ) that Plato’s argument,
and in particular the requirements of the uniformity of explanation, which
are at the basis of it, presuppose that explanations and causes are deter-
ministic and rule out that they can be probabilistic. This is not the place
properly to consider whether this is correct. I am inclined to think that it is
not. Here is what an eminent physicist, Greene, says: ‘Even so, as long as
we can determine mathematically the precise form of probability waves,
their probabilistic predictions can be tested by repeating a given experi-
ment numerous times.’ This, if I am not mistaken, implies that the
requirements of the uniformity of explanation can, and arguably must, be
applied even in probabilistic explanations, provided that what they are
applied to in that case is not individual cause-effect relations but rather
probabilistic distributions of individual cause-effect relations. (I shall not
attempt to reformulate the requirements accordingly.)
This issue is of interest not only for the, non-historical, question of whether
Plato’s argument, rooted in the requirements of the uniformity of explana-
tion, for the conclusion that all explanation and causation is based in essences,
is relevant for any account of explanation, even explanations in current
physics. It is of interest also for the question of whether Plato’s method of
enquiry into nature and the physical world, in the Timaeus, based as it is on
the idea that the conclusions of such enquiry can be no more than likely (they
issue only in an eikos logos, a ‘likely account’, cf. c–d), is compatible with the
supposition that Plato thinks there is room for genuine explanations in this
enquiry: that he thinks a science of nature is genuinely possible.
. Plato’s resolution of the aporia about explanation:

explanantia are, primarily, essences (Phaedo d–b)
Socrates’ intellectual journey culminates in the introduction of the account
that says that explanantia are, primarily, essences (b f.). This is a
journey from complete loss of trust to perfectly restored trust in

Greene (, ).

I am not, of course, supposing here that what Plato intends by eikos logos in the Timaeus passage is
‘probabilistic explanation’.

 Plato’s Essentialism
explanations, and it is the account that says that explanantia are, primarily,
essences that restores this trust. Trust is restored, we will see, because Plato
thinks that explanations in conformity with this account, and such expla-
nations alone, satisfy the two fundamental requirements of explanation.
It is notable that the two requirements of explanation are invoked both
during Socrates’ description of how he lost trust in traditional explanations
(up to d) and when his trust in explanations has been restored by the
introduction of the account that says that explanantia are, primarily,
essences (beginning at d). This provides further indication that, just
as the apparent breach of these requirements by traditional explanations is
the source of his loss of trust in explanations, so the restoration of trust in
explanations is due to the satisfaction of these requirements by the account
that says that explanantia are, primarily, essences.
As it is formulated when it is first introduced (b–c), this account
says that:
The explanation of why O is F is that O is appropriately related to – it
‘partakes in’, or ‘communes with’, or ‘has present in it’ – the essence of the
quality F, Ess(F).
Here O is a changing thing, and F is a quality of that thing. It is pointed
out more than once that O is a changing thing, and apparently it is
changing in respect of the quality F (see a–, b–, c). It is,
therefore, natural to think that F is, as we would say, an accidental and
contingent quality of O.
The essence of the quality F, Ess(F), is the entity properly referred to as
the F itself by itself. What, we ought to ask, is the relation here between
the essence, or Form, of the quality F, and the physical thing that has that
quality, and whose having the quality is the explanandum? Apparently, the
two appear to be conceived here as distinct and non-identical. Plato has
argued, earlier in the dialogue (a–c), that Forms are distinct from and
not identical with sense-perceptible things. It is a good question whether
the two are, further, conceived as separate (see Chapter ).

For REQ, compare a–b with b–c. For REQ, compare esp. d–e with
e–b.

Remarkably, David Sedley thinks the schema ‘“It is because of the F that F things are F”’ is ‘treated
as [an example of] self-evident truths’ (, –). I confess I cannot see how it could be so
treated by Plato. Plato intends ‘the F’ here to signify an essence/Form, and he could hardly think it
is self-evident that causes are essences/Forms.

See, e.g., b–: [to kalon, agathon, megan] auto kath’ hauto. Earlier in the dialogue simply F itself
(auto).

Forms as the basis of causation and explanation 
Shortly later this entity, the essence of the quality, F, is referred to as,
precisely, the ousia of the quality F, or of things in so far as they are
F (c). We may single out this passage as central, for it says that
explanations (aitiai) are essences (ousiai):
And you would cry out loud that you know of no other way in which any
thing comes to be [including, that is, comes to be F] than by partaking in
the proper essence of any thing, whichever it may be that it partakes in; and
also in this case you know of no other aitia of coming to be two, save the
participation in two. (c–)
The term eidos is used a little later, with the same reference as ousia (see
b; also e and c). The relation between the changing thing
that is F and the essence of the quality F is variously called ‘participation’
(metaschesis, methexis), ‘communion’ (koinōnia) and (for the converse
relation) ‘presence’ (parousia, enousia), but its exact nature appears delib-
erately to be left open (see d–).
This is evidently an account about explanantia, which says that explanantia
are (primarily) essences. But it is just as much an account about explanation,
that is, about explananda, explanantia and their relation, A because B. Indeed,
the account is introduced as an attempt to satisfy the requirements, REQ
and REQ, which embody the constraints of the uniqueness, generality and
uniformity of explanation. Evidently these requirements are specifically about
how explananda and explanantia must be related to each other, that is, in such
a way that explanations are unique, general and uniform. So, the account,
though certainly about explanantia, is, above all, about explanation, that is,
about the whole triad: explananda, explanantia and their relation.
The account appeals to explananda, but how is an explanandum under-
stood here? Is it understood simply as an unanalysed whole, something’s
being such and such, or is it already analysed into a particular, O, and a
quality, F? Initially at least, it appears that it is understood as an unanalysed
whole. For this is the way in which natural phenomena present themselves
to us in experience and when we begin to ask simple and immediate
‘Why?’ questions, such as the question ‘Why do human beings grow?’
And in his narration Socrates appears deliberately to present such ‘Why?’
questions as arising naturally and directly from our experience of natural

For why ousia here may be understood to mean ‘essence’, see Politis , section , where I also
consider, at length, the relation, in the Phaedo, between the use of the term eidos and the use of the
term ousia and its cognates.

I put ‘primarily’ in brackets because, in this passage (b–c), the account says simply that
explanantia are essences. It is only later that it will emerge that the account is not that explanantia
are simply essences, but that explanantia, whatever else they may involve, are primarily essences.

 Plato’s Essentialism
phenomena. Unless, therefore, we have particular reason to think that
Plato understands explananda as already analysed into particulars and
qualities, we ought to leave it at that. We do, however, have such reason,
and this is that, precisely because of the requirements of explanation, Plato
needs to analyse explananda into particulars and qualities. For, in order to
state that explanations must be uniform, it is necessary to introduce the
idea of the sameness or difference of a quality and to suppose that the
sameness or difference of a quality is independent of the sameness or
difference of the things that possess it (I mean: one and the same quality
can be had by numerically different particulars; and the same particular can
have different qualities).
. The sufficiency claim

Why, then, does Plato think that any explanation in conformity with this
account satisfies the two fundamental requirements of explanation?
Why, that is, does he think that the account that says that explanantia
are (primarily) essences is sufficient to satisfy these requirements, REQ
and REQ? The account says that:
. O is F because O is appropriately related to the essence of the quality
F, Ess(F).

That this is indeed what Plato thinks is pointed out by, e.g., Annas (, –); Fine (,
). However, I do not think their accounts are satisfactory to explain why Plato thinks this. Fine
says: ‘Plato’s SA [i.e., the simple schema] disallows this [i.e., explanations by opposites]; something
is, or comes to be, F just in case it is, or comes to be, suitably related to the Form of F, and no Form
(Plato assumes) consists of opposites.’ Annas says: ‘The Form F explains an F instance by its
guaranteed freedom from being the opposite of F.’ But this, surely, is not sufficient to ensure
uniformity. We can see this if we name the Form of F ‘N’ and ask why it should not be possible that
N sometimes causes F in things, sometimes G (opposite to F) in things; or, conversely, why it
should not be possible that things that are F should sometimes be caused to be F by Form N,
sometimes by a different Form, M. As far as I can see, what Fine and Annas say does not address this
crucial question. Nor, in my view, would it be advisable to address it by invoking the transmission
principle of causation, for this would be to beg it. Annas appears to suppose that the question can
only be addressed by invoking this principle, and this is why she concludes that ‘Plato’s demand on
explanation is one which we no longer find compelling; it applies convincingly only to a few simple
cases where the explanandum is a quality that can be transmitted from one thing to another’ ().
I agree with her inference to her conclusion, but not with the supposition behind it, hence not with
her conclusion. What Fine and Annas fail to point out is that the Form F is the essence of the
quality F. This, and this alone, is why, if the Form is the cause, it can only cause F.

Sedley invokes the transmission principle of causation to explain why Plato thinks it is because of
the F that things are F (, –); and he argues that Plato thinks this principle is self-
evidently true ( f.). As if by the invocation of self-evidence, for a supposed principle, we could
dispense with considering why, and with what justification, the philosopher thinks it is indeed
a principle.

Forms as the basis of causation and explanation 
But if we reflect for a moment on the notion of essence, it is evident that:
. The essence of a quality, F, is that which determines and constitutes
the very identity of this quality: which quality this is.
If, however, we put these claims together (i.e.,  and ), we obtain the
following, central claim:
CC. That which explains why a particular thing has a certain quality, F, and
that which determines which quality this is and constitutes the identity of
this quality, are one and the same thing.
This is a striking claim, which indicates the precise way in which Plato
thinks that these fundamental concepts, explanation (aitia) and essence
(ousia), are related.
But this claim, CC, has an immediate consequence, and this is the truth
of the requirements REQ and REQ. On the supposition of this claim
(i.e., CC), the demonstration of REQ and REQ is straightforward. Here
is the argument:
It is impossible for different particulars, in so far as they have the same
quality, to have different explanantia; for, according to claim CC, if they
have different explanantia, then they have different qualities. So REQ is
true. Likewise, it is impossible for different particulars, in so far as they have
different qualities, to have the same explanans; for, according to claim CC,
if they have the same explanans, then they have the same quality. So REQ
is likewise true. On the supposition, therefore, of the account that says that
explanantia are (primarily) essences, the requirements of explanation,
REQ and REQ, are true. This account is, therefore, sufficient to satisfy
the requirements of explanation.
A brief comment on claim  – having already commented on claim . This
claim is supposed to articulate no more than Plato’s core concept of essence,
which says that the essence of any thing, X, is what it is to be this very thing,
X. (We should also note the expression einai hoper ēn, ‘to be the very thing it
is’, which is used twice later in the dialogue, at d and d.) Or, to say
the same thing in a different way, the essence of any thing, X, is what is
signified by the answer, if it is adequate and true, to the Socratic-type
question, ‘What is this very thing, X?’ We are of course familiar with this
Socratic-type question from dialogues we consider to be earlier than the
Phaedo. Earlier in the Phaedo itself, indeed, Plato reminds us of this type of
question, and he associates it directly with the theory of Forms.
It is worth observing that this defence of the requirements of explana-
tion appears to conform to the procedure of hypothesis, as this method is

 Plato’s Essentialism
introduced here (a– and d–). That is to say, on the hypothesis
of the account that says that explanantia are (primarily) essences, it follows
that these requirements are true. So, the requirements of explanation,
which perhaps had the status of hypotheses when initially introduced,
are now logically derived ‘from something higher’ (see anōthen, d),
that is, from the account that says that explanantia are (primarily) essences.
Indeed, it appears that this account is itself referred to as a hypothesis.
Plato, therefore, thinks that the account that says that explanantia are
(primarily) essences is sufficient to satisfy the requirements, REQ and
REQ, that embody the constraints of the uniqueness, generality and
uniformity of explanation. Indeed, he is right. Does he also think that
this account is necessary to satisfy these requirements? This does appear to
be his view. When he comments on the procedure of hypothesis here (in
d–), he says that if one is requested to defend (didonai logon) a
particular hypothesis, one must put forward a further hypothesis, namely,
‘the one that would appear best of those above (hētis tōn anōthen beltistē
phainoito), until one arrives at something adequate (hikanon)’. But if the
requirements of explanation were initially hypotheses, and if the task now
is to defend them by relying on a ‘higher’ hypothesis, that is, the account
that says that explanantia are (primarily) essences, then this comment on
the procedure of hypothesis shows that Plato thinks that there may be
other, and competing, ‘higher’ hypotheses for defending the requirements
and that his own account must be shown not only to provide a good
defence, but the best. To show this is to show that this account is not only
sufficient but also necessary to satisfy the requirements of explanation.
. The necessity claim

To consider why Plato thinks that the account that says that explanantia
are (primarily) essences is necessary to satisfy the requirements of explana-
tion, we need to distinguish the various elements in this account and to ask
why he thinks that each of them is necessary to satisfy the requirements of
explanation. We need to distinguish the following elements in the account:
i. the general schema of explanation, O is F because of E;
ii. the analysis of the explanandum into a particular and a quality had
by it;

See d–, supposing that the hupothesis mentioned here includes the general statement of the
account that explanantia are, primarily, essences, in c–, and not only its particular application
to the present case, in c–, i.e., the case of one thing coming to be two.

Forms as the basis of causation and explanation 
iii.the supposition that the quality F has an essence, Ess(F);
iv. the supposition, CC, which says that the entity, E, that explains why
a particular thing, O, has a particular quality, F, and the essence of
this quality, Ess(F), are one and the same thing (i.e., E = Ess(F)); and
v. the relation between the explanandum and the explanans, which
Plato variously refers to as ‘participation’, ‘communion’ and (for the
converse relation) ‘presence’, but whose exact nature he appears,
deliberately, to leave open (see d–).
The question, therefore, is why Plato thinks that each of these elements in
the account of explanation is necessary to satisfy the requirements
of explanation.
In regard to the first element (i), it ought to be evident that the general
schema of explanation, O is F because of E, is necessary to satisfy the
requirements of explanation, and one could hardly think otherwise. This is
evident if we consider that all this schema does is set out the form of any
answer to the question ‘Why is this thus and so?’: ‘This is thus and so
because of such and such.’ The variables in the original formulation, O,
F and E, are simply placeholders and do not add anything to the blueprint.
In regard to the second element (ii), we need to ask why it is justified to
think that explananda should be analysed into particulars and qualities had
by them. We saw that this is justified precisely by the constraint that says
that explanations must be uniform, that is, by the requirements REQ and
REQ. For, in order to state that explanations must be uniform, it is
necessary to introduce the idea of the sameness or difference of a quality
and to suppose that the sameness or difference of a quality is independent
of the sameness or difference of the things that have it (I mean: one and the
same quality can be had by numerically different particulars; and the same
particular can have different qualities). This, in turn, provides us with the
appropriate and minimal notion of a particular here: a particular is a this
and something that can be counted, it is a bearer of qualities and it is such
that one and the same quality can be had by numerically different
particulars. This shows that the analysis of explananda into particulars
and qualities had by them is necessary to formulate, and therefore to
satisfy, the requirements of explanation.
One might object that it would be no less justified to think that
explananda should be analysed into objects (e.g., Socrates) and states of
affairs (e.g., Socrates’ being in a state of growing). We ought to reply that,
if this analysis is to allow for the formulation of the requirements of
explanation, it will need a means of indicating that the following two

 Plato’s Essentialism
states of affairs have an element (what we have called, the quality) in
common: Socrates’ being in a state of growing and Plato’s being in state of
growing. Once such a means is provided, it will be evident that this analysis
is equivalent to ours: what is indicated by this means corresponds to what
we have called the quality, and the object appealed to in the analysis
corresponds to what we have called the particular. We can, therefore, avail
ourselves of both analyses without needing to choose between them.
In regard to the third element (iii), it is striking that the view that qualities
have essences is a presupposition of the requirements of explanation. This
is because, in order to state that explanations must be uniform, it is
necessary to introduce the idea of the sameness or difference of a quality
and to suppose that the sameness or difference of a quality is independent
of the sameness or difference of the particulars that have it. We can see this
also if we ask how it is possible to dispute the requirements of explanation.
For it appears that one way in which this is possible is, precisely, by
arguing that they cannot be true of anything and that this is so because
either qualities do not have determinate identity conditions or essences at
all, or, if they have identity conditions of a sort, these depend, in each case,
on the particulars that have them. The truth of the view that qualities have
essences, and that the essence of a quality is independent of the particular
that has it, is, therefore, a necessary condition for the truth of the
requirements of explanation, or rather for the possibility of there being
anything that satisfies them.
In regard to the fourth element (iv), we can, I think, readily recognise why
Plato may think that supposition CC is necessary to satisfy the require-
ments of explanation – though we will see that this is not quite his view.
Suppose that CC is not true, that is, the essence of the quality F, Ess(F), is
not identical with the explanans, E, that explains why a particular thing,
O, is F. In that case, it appears, the explanandum, O’s being F, will be
independent, for its essence, that is, for its being the very one it is, of the
explanans, E. (As we might say, echoing Hume, there will be no necessary
connection between the effect and the cause.) But then, it appears, there
will be nothing to appeal to for the purpose of showing that the putative
explanation, ‘O is F because of E’, satisfies the requirements
of explanation.
To recognise the force of this reasoning, we ought to recall why, if CC is
true, then the explanandum, O’s being F, will indeed be dependent, for its
being the very one it is, on the explanans, E. Evidently, the essence of the
quality F, that is, what this quality is, is part of the essence of this
explanandum, O’s being F, that is, part of what it is for this explanandum

Forms as the basis of causation and explanation 
to be the very one it is. But, according to CC, the essence of the quality
F is identical with the explanans, E. Therefore, if CC is true, the expla-
nandum, O’s being F, will be dependent, for its being the very one it is, on
the explanans, E. It was precisely this dependence and necessary connec-
tion that we made use of earlier for the purpose of showing that, if CC is
true, then the requirements of explanation, REQ and REQ, are true.
However, we have reason to object to this reasoning, as follows: even if
claim CC is false, that is, even if Ess(F) and E are not identical, there may
still be a necessary connection between them. In that case the explanan-
dum, O’s being F, may still be dependent, for its being the very one it is, on
the explanans, E – though the dependence may be less immediate, and less
immediately evident, than if it is based on the identity of Ess(F) and E. It
may, therefore, be possible, even if CC is false, to show that a putative
explanation, ‘O is F because of E’, satisfies the requirements of explana-
tion. Therefore, it is not the case that CC is necessary to satisfy the
requirements of explanation.
We shall see that Plato can acknowledge and accommodate this objection.
For we will see that he will argue (at c–c) that the explanans, E, of
O’s being F may not be the simple one, Ess(F), but may be rather a complex,
conjunctive one, [S + Ess(F)] – where S is, for example, a physical constit-
uent of O. For example, the explanans, E, of this body’s being hot may be not
simply the essence of heat but may be rather the conjunction of fire
(supposing the body has fire in it) and the essence of heat. Plato defends
not only the simple schema of explanation, but also a complex one, and he
thinks the complex schema is dependent on the simple one. It is evident,
however, that if the explanandum, O’s being F, is dependent, for its being the
very one it is, on the simple explanans, Ess(F), then it is also dependent, for
its being the very one it is, on the complex explanans, [S + Ess(F )]. For the
complex explanans is a conjunction of the simple one and a further explan-
atory element. On the complex schema of explanation, therefore, it will still
be true that the explanandum, O’s being F, is dependent, for its being the
very one it is, on the explanans. In this way, Plato can acknowledge and
accommodate the objection. Of course, if he acknowledges and accommo-
dates it in this way, this means that he does not, after all, think CC is
necessary to satisfy the requirements of explanation. What he thinks is
necessary to satisfy the requirements of explanation is rather:
CC-qualified. Either (E = Ess(F)) or (E = [S + Ess(F)]),
where this disjunction ought to be understood as being inclusive rather
than exclusive.

 Plato’s Essentialism
In regard to the fifth element (v), it is notable that Plato here appears
deliberately to leave open how exactly the relation between explananda and
explanantia is to be understood (see d–). This is indeed appropriate,
if he thinks that each element of the present account of explanation is
necessary to satisfy the requirements of explanation. For then the relation
between a particular explanandum and a particular explanans must be
specified simply as that relation, whatever it is, which, if it obtains, implies
that the particular explanandum, a particular thing’s being F, is explained
by the particular explanans, E. Suppose, on the other hand, that Plato had
a particular conception of the relation between explananda and explanantia
in mind here, such as, for instance, the view that it must be a relation
between separate and distinct things, where separation implies more than
the view that the essence of a quality is independent of the particular that
has it. In that case it would not be at all plausible for him to think that this
relation is necessary to satisfy the requirements of explanation. I doubt that
we have grounds for thinking that Plato has a particular conception of this
relation in mind here, and he appears to indicate the contrary (see
d–).
We may conclude that Plato thinks, for good reason, that the account of
explanation that says that explanantia are (primarily) essences is both
sufficient and necessary to satisfy the fundamental requirements of expla-
nations, REQ and REQ, which embody the constraints of the unique-
ness, generality and uniformity of explanation. This account, therefore, is
both necessary and sufficient for the purpose of resolving the original,
radical aporia about explanation.
. The simple and the complex schema of explanation

(Phaedo b–c)
Let us refer to the original schema of explanation (set out at b–b
and recalled at b–c) as the simple schema. This schema is of the
form, O is F because of its relation to the essence of the quality F, Ess(F). For
example (taken from c–), this body is ill because of its relation to the
essence of illness. And let us refer to the later schema of explanation (set
out at b ff., esp. c–c, and contrasted with the original
schema at b–c) as the complex schema. This schema is of the form,
O is F because of its relation to S and S’s relation to the essence of the quality F,
Ess(F). For example (from the same passage), this body is ill because of its
relation to fever and fever’s relation to the essence of illness. It is crucially
important that Plato defends both these schemata of explanation. For this

Forms as the basis of causation and explanation 
shows that his conclusion is not that explanantia are simply essences but
rather that explanantia, whatever else they may involve, are
primarily essences.
How are the two schemata of explanation related? The simple schema
was originally described as ‘safe’ (asphales, e, and d; indeed
‘perfectly safe’, asphalestaton, d), and this characterisation is recalled
when it is contrasted with the complex schema (b–c). What does
this safeness consist in? Not, certainly, in the simple schema’s being empty
of content or information – not in its being trivial, much less tautologi-
cal. If ‘being safe’ meant ‘being contentless’ or ‘being uninformative’,
then the complex schema of explanation could not itself likewise be
characterised as being safe; for it, at any rate, is evidently contentful and
informative. But this is how the complex schema is characterised (ek tōn
nun legomenōn [i.e., the description of the complex schema] allēn horōn
asphaleian, b–). In respect of safeness, the simple and the complex
schemata are on an equal footing.
I would like to suggest that the safeness of the simple schema of
explanation consists in the fact that, at any rate, explanations in conformity
with this schema meet the basic constraint for being explanations, namely,
the constraint of the uniqueness, generality and uniformity of explana-
tions. That explanations in conformity with this schema satisfy this con-
straint is not, however, something evident, for it is the upshot of the aporia
about what constitutes a genuine explanation and the solution to this
aporia provided by the simple schema. Hence the simple schema is
anything but contentless or uninformative. The simple schema states that
explanantia, whatever else they may involve, are primarily essences. It is
only natural that Socrates should refer to such explanations as safe,
trustworthy and reliable, since his concern is to find a method of explana-
tion that he can rely on as being genuinely explanatory. This interpretation
of the safeness of the simple schema also makes good sense of the fact that
the complex schema, too, is described as safe (b–). For Plato thinks
that the complex schema is dependent on the simple one (see the end of
this section for this dependence), and, therefore, if explanations in con-
formity with the simple schema are genuinely explanations, so too are
explanations in conformity with the complex one.

Taylor (, ) writes: ‘But the safety of the answer seems to lie in its total lack of information’.

Fine, who raises the question of the safety of the complex schema, answers as follows: ‘Whereas the
SA [i.e., the simple schema] specifies necessary and sufficient conditions, the CA [i.e., the complex
schema] need specify no more than sufficient conditions. Further, the CA allows some explanations
involving opposites. Fever, e.g. is adduced to explain illness; but, as Gallop notes, the opposite of

 Plato’s Essentialism
We may note that, on this interpretation, we must understand as playful
and ironical Socrates’ description of his attachment to the simple schema
as ‘simple-minded’ (cf. euēthōs echō par’ emautō[i], d) and ‘ignorant’
(amathē, c) and of the complex schema as ‘more sophisticated’
(kompsoteran, c). What is the point of this irony? Apparently, it is
yet again directed at those who look for complex and sophisticated
explanations without first pausing to ask what an explanation really is. It
is to them that the simple schema, especially when compared to more
complex explanations, will appear ‘simple-minded’, ‘ignorant’ and ‘plain’,
when in reality it is the basis of any more complex explanations. This
continues the irony that goes back to Socrates’ initial description of himself
as ‘unfit’ for the traditional method of scientific investigation (aphuēs,
c) and of his own alternative method as ‘random confusion’ (alla tin’
allon tropon [tēs methodou] autos eikē[i] phurō, b).
Let us look closer at the two schemata and their relation. The schemata
can be understood as follows.
Simple Schema
O is, accidentally, F
because
O is appropriately related (by relation R) to the essence of F, Ess(F).
For example (based on b–c), this body is hot because it is appropriately
related to the essence of heat; or, this thing has an odd number of parts
because it is appropriately related to the essence of odd.
Complex Schema
O is, accidentally, F
because
. O is appropriately related (by relation R) to S, and

. i. S is distinct from the essence of F, Ess(F), and
fever – hypothermia – can also explain illness’ (, ). However, it seems to me that, far from
showing that the complex schema is safe too, what Fine says here suggests that, compared with the
simple schema, the complex schema is not safe at all. With regard to the apparent objection that the
appeal to fever does not satisfy the one requirement of uniformity, we may respond that it is not
fever by itself that is adduced to explain illness but fever appropriately related to the essence of
illness (the same would be true of hypothermia). It is clearly fallacious to argue: fever and
hypothermia (if they are explanantia at all) are opposite explanantia; therefore, fever, if
appropriately related to illness (e.g., if understood as an effect of one kind of illness) and
hypothermia, if appropriately related to illness (e.g., if understood as an effect of a different kind
of illness) are opposite explanantia.

‘Accidentally’ because, as we saw, O is subject to change, and subject to change in particular with
regard to its being F.

Forms as the basis of causation and explanation 
ii. S is essentially F, and
iii. S is essentially F because it is appropriately related (by relation R) to
the essence of F, Ess(F).
For example (based on the same passage), this body is hot because . it is
appropriately related to fire, and . fire, which is distinct from the essence
of heat, is essentially hot, and is essentially hot because it is appropriately
related to the essence of heat. Or again: this thing has an odd number of
parts because . it is appropriately related to the unit, and . the unit,
which is distinct from the essence of odd, is essentially odd, and is
essentially odd because it is appropriately related to the essence of odd.
We should note that the complex schema must be understood as
representing a single, unitary explanation, not a conjunction of two
independent explanations. It is of the form, a because (b and c), and not
of the form, (a because b) and (b because c). If the complex schema
represented a conjunction of two independent explanations, then O is,
accidentally, F because O is appropriately related to S (i.e., the first conjunct
of the schema) would by itself be an explanation. But in that case there
would, after all, be genuine explanations whose explanantia do not involve
essences, for example, this body is hot because it is appropriately related to
fire – but this is what Plato denies. It follows, we ought to note, that Plato
rejects the following principle, we may call it the principle of the distrib-
utivity of explanation: if a because (b and c), then (a because b) and (b
because c).
Let me make a few comments in clarification and defence of this reading
of the two schemata and their relation. First, Plato distinguishes between a
thing’s being accidentally F and a thing’s being essentially F (see esp.
b–a). I am thinking of the following passage: ‘Because it isn’t,
surely, by nature that Simmias overtops him [Socrates], by virtue, that is,
of his being Simmias, but by virtue of the largeness that he happens to
have. Nor again does he overtop Socrates because Socrates is Socrates, but
because of smallness that Socrates has in relation to his largeness?’
(c–).
Secondly, he distinguishes between, on the one hand, things that are
essentially F, while being distinct from the essence of F (e.g., fire, which is
essentially hot but distinct from the essence of heat), and, on the other
hand, the essence of F itself (e.g., the essence of heat, see esp.

This passage does not, of course, say or imply that Simmias or Socrates have an individual essence,
rather it illustrates the distinction between a thing’s being accidentally F and a thing’s being
essentially F, by offering a positive instance of the former and a negative instance of the latter.

 Plato’s Essentialism
c–c). That the former side of this distinction concerns being
essentially F, rather than merely being necessarily F, is indicated by the fact
that the expression einai hoper ēn (‘to be the very thing it is’) is used of
snow and fire in relation to cold and heat (see d and d).
Thirdly, are fire, snow, the triad, etc., themselves essences? There is
good reason to think that they are not, and I have supposed this in the
interpretation of the complex schema of explanation. First, Plato expressly
sets fire, snow, the triad, etc., against auto to eidos (‘the Form itself’,
e–) and ta eidē (‘the Forms’, c–). Secondly, he allows for the
possibility that fire, snow, the triad, etc., are subject to destruction and
movement (d–, b–c). But essences and Forms are not
subject to any such thing. If we ask how something that is essentially
F and always F, such as fire, snow, the triad, etc., can nevertheless be
subject to destruction, the answer is that such things are always F, for as
long as they exist (see hotanper ē[i], e). Essences, on the other hand, are
eternal, everlasting or timeless. Thirdly, that fire is hot, that snow is cold,
that the triad is odd, etc., is supposed to stand in need of explanation by
appeal to something that is emphasised as being distinct from them, that
is, by appeal to the essence of hot, cold, odd, etc. And yet it is indicated
that hot, cold, odd, etc., are essential qualities of fire, snow, the triad, etc.
But Plato is not likely to think that an essence’s having a particular quality
needs to be explained by appeal to something distinct from this essence; on
the contrary, it is because of themselves that essences are what they are. It
follows that fire, snow, the triad, etc., though they have essential qualities,
are not themselves essences.
Fourthly, the complex schema of explanation indicates (see esp.
c–c) that Plato distinguishes between explananda that consist
in a thing’s being accidentally F (e.g., this body’s being hot) and expla-
nanda that consist in a thing’s being essentially F while being distinct from
the essence of F (e.g., fire’s being hot). And he assigns to these two kinds of
explananda different places in the unified account of explanation. For,
while accidental explananda (e.g., this body’s being hot) can figure only as
explananda and not as explanantia, essential explananda, which are distinct
from essences (e.g., fire’s being hot), though being themselves subject to
explanation in terms of essences (e.g., the essence of heat), can be elements
in the explanation of accidental explananda (e.g., this body’s being hot).
Essences, on the other hand, have just one place in explanation, that is, as
primary explanantia.
Plato thinks that, first, the simple schema of explanation is primary,
and, secondly, the complex schema is dependent on it. The simple schema

Forms as the basis of causation and explanation 
is primary because it gives expression to the fundamental claim that
explanantia, whatever else they may involve, are primarily essences; and
this claim is fundamental because it is what allows for the satisfaction of
the requirements of explanation (REQ and REQ), which embody the
constraints of the uniqueness, generality and uniformity of explanation.
Why does Plato think the complex schema of explanation is dependent
on the simple one? What this view means is that the complex schema is
true because the simple schema is true (but not conversely). But why does
Plato think that? To see this, let us recall that the complex schema is of the
form:
O is F because [(O-R-S) and (S-R-Ess(F))],
meaning that O is F because: O is appropriately related to S, and S is
appropriately related to the essence of F. And the simple schema is of the
form:
O is F because O-R-Ess(F),
meaning that O is F because: O is appropriately related to the essence of
F. Now, it is true that Plato does not provide an account of these relations,
R, R and R. But it is just as important to observe that he thinks there is a
crucial difference between relation R (in the simple schema), on the one
hand, and relations R and R (in the complex schema), on the other. For
he thinks the obtaining of R between O and Ess(F) is sufficient for an
explanation, E, of why O is F, whereas it is not the case that the obtaining
of R between O and S, or the obtaining of R between S and Ess(F), are
individually sufficient for an explanation of why O is F. (We recall that the
complex schema is of the form a because (b and c), but not of the form (a
because b) and (b because c).) Rather, he thinks that: i. the obtaining of R
between O and S is one conjunct in a conjunctive explanation, E*, of why
O is F, and the obtaining of R between S and Ess(F) is a further conjunct
in this same conjunctive explanation, E*; and ii. the conjunction of these
two conjuncts, which is E*, is sufficient (but not necessary) for the
explanation E (the same as in the simple schema), of why O is F.
. Addendum: The interlude about good-based (teleological)

explanations (Phaedo b–d)
Why, in the course of recounting his disillusionment with traditional
explanations, does Socrates turn to good-based (teleological) explanations,
only to become disappointed with them, too (b–d)? And why, in

 Plato’s Essentialism
response to this disappointment, does he turn to essence-based explana-
tions? The question is: First, why does Plato introduce good-based expla-
nations at this point of the dialogue? Secondly, why does he nevertheless
go on to defend not them, but essence-based ones? And thirdly, what does
he think is the relation between good-based (teleological) and essence-
based (formal) explanations?
One answer to the first question is clearly mistaken: Plato introduces
good-based explanations because his complaint against traditional expla-
nations is just that they are not good-based. This interpretation is
mistaken because Socrates’ complaint against traditional explanations is
independent of and prior to his becoming hopeful about good-based ones.
The complaint is lodged before the mention of good-based explanations,
and they are introduced because of the expectation that they, at any rate,
are not subject to the complaint. Indeed, if the complaint were simply that
traditional explanations are not good-based, we would have been given no
reason for thinking that explanations ought to be good-based.
In a central passage of the interlude about good-based explanations
(e–a), it is clearly indicated what the complaint is that good-based
explanations are expected to escape. As we saw, the complaint is that
traditional explanations, and in particular physicalist explanations (i.e.,
explanations that appeal only to physical things) do not satisfy the follow-
ing requirement: it is impossible that the same explanans (e.g., Socrates’
bones and sinews) should explain opposite explananda (e.g., his running
away to Megara or Boeotia as opposed to his staying imprisoned in
Athens). It is clearly indicated that it is just this complaint that good-
based explanations are expected to escape, because what is immediately
inferred from the objection, that physicalist explanations do not satisfy this
requirement, is that the real explanation of Socrates’ behaviour is his
judgement of what is best (hē doxa tou beltistou, a). A corresponding
complaint, however, was made before good-based explanations were men-
tioned (see esp. a–b, the complaint was that traditional explanations
do not satisfy the requirement which says that it is impossible that opposite
explanantia should explain the same explanandum). This shows that
Socrates’ complaint against traditional explanations is independent of,
and prior to, his hopeful appeal to good-based ones.
We may suppose, therefore, that good-based explanations are intro-
duced because of the expectation that they, at any rate, will satisfy the
requirements of explanation, REQ and REQ (same explanans if, and

This answer was defended by Taylor ().

Forms as the basis of causation and explanation 
only if, same explanandum), but that, as it turns out, this expectation is
satisfied not by them, at least not as they are traditionally conceived, but by
essence-based explanations. This also accords with a natural reading of the
characterisation of the upcoming defence of essence-based explanations as
a deuteros plous (‘second sailing’, d). Apparently, the phrase deuteros
plous originally referred to ‘those who use oars when the wind fails’ (LSJ,
ploos ). It is, therefore, natural to understand the present occurrence of
this phrase as meaning a slower and more laborious means to one’s
original destination.
The first sailing was the attempt, which has turned out unsuccessful, to
defend good-based explanations, that is, to show that they at any rate
satisfy the requirements of explanation. The second sailing is the upcom-
ing attempt to defend essence-based explanations in the same way, and this
will in fact succeed. The two sailings have one and the same destination: to
satisfy the requirements of explanation and thus answer the original aporia
about explanation. The first sailing set up high hopes of being quick and
painless, but failed; the second will require more time and effort, but
will succeed.
Why is the attempt to satisfy the requirements of explanation by appeal
to good-based explanations represented as unsuccessful, while the attempt
to satisfy the same requirements by appeal to essence-based explanations is
represented as successful? The following answer suggests itself. Plato thinks
that good-based explanations must, somehow, be capable of satisfying
the requirements of explanation, otherwise he would not have been
hopeful about them – and, indeed, associated them so intimately with
reason (nous). But he thinks that good-based explanations can satisfy the
requirements of explanation if, and only if, it is supposed that each
thing has its own proper (idion) good and that the good that is proper to
each thing depends on what the thing is, its essence. Without this
supposition, there is nothing to ensure that there being a good that is
proper to a particular thing (e.g., the earth) stands in a necessary connec-
tion with the thing’s having one particular quality (e.g., being round) as
opposed to its having a different and incompatible quality (e.g., being flat).
Hence, without this supposition, it is impossible to satisfy the require-
ment, if same explanans (e.g., the earth’s good), then same explanandum
(e.g., the earth’s being round). Whereas, with this supposition in place, the
satisfaction of this requirement is ensured if it is supposed that this
quality (e.g., being round) does indeed stand in a necessary connection
with (i.e., is either part of, or a consequence of ) the thing’s being the very
thing it is, its essence.

 Plato’s Essentialism
What this shows, we ought to note, is that good-based explanations
depend on essence-based ones. It also shows that, even if it is thought that
this dependence is mutual, still essence-based explanations must, for the
purpose of satisfying the requirements of explanation, REQ and REQ,
be defended first. If essence-based explanations are not defended first, the
attempt to satisfy the requirements of explanation by appeal to good-based
explanations is premature and bound to fail – as it did, Plato thinks, for
Anaxagoras.
We have found the answer to the question: What, in this part of the
Phaedo, is supposed to be the relation between good-based (teleological)
and essence-based (formal) explanations? Good-based explanations must, if
they are to satisfy the constraints of the uniqueness, generality and unifor-
mity of explanation, that is, the constraints embodied in the requirements
of explanation, REQ and REQ, depend on essence-based explanations.
Plato, we ought to note, does not think that essence-based and good-
based explanations are two different kinds of explanation. For he thinks all
explanations are essence-based. Rather, he thinks good-based explanations
are a kind of essence-based explanation. We are perhaps accustomed to
thinking of the distinction between teleological and essence based, or
formal, explanations as a distinction between two kinds of explanation.
Perhaps we are accustomed to think this because it is from Aristotle, or
from a certain interpretation of Aristotle, rather than from Plato that we
have become familiar with this distinction. But this is not the only way of
conceiving of the distinction, and Plato’s is an alternative.
If we think that Plato, in order to be able to draw such a distinction at
all, must conceive of the distinction between teleological and formal
explanations as a distinction between two kinds of explanation, then we
cannot make sense of the argument of the Phaedo for the claim that
explanantia are, primarily, essences. For the argument defends a single
account of all explanation, not only of formal or essence-based explana-
tions. This is because the argument, as we have seen, is motivated by, and
conceived as an answer to, the question ‘What is an explanation?’, and the
answer it defends, we have seen, is a unified one. The unifying feature is
indicated by the claim that any explanans, whatever other elements it may
contain (such as, we have seen, physical things), must, first and foremost,
contain an essence, the essence of the quality figuring in the explanandum.

 
What is the role of essences, or Forms, in judgements

about sense-perceptible and physical things?
Republic VII. –
I want to consider, in its own right and with a hopefully fresh pair of eyes,
this much-discussed passage from the Republic (VII. a–a). I shall
argue that, in it, Plato sets out a single, extended argument for what I call
Plato’s Principal Hypothesis, in the Republic, for essences and Forms. This
Hypothesis is that:
It is possible for us to make judgements and statements that are about sense-
perceptible and physical things, and that conform to the principle of non-
contradiction, ONLY IF we are capable of individuating certain of those
sense-perceptible things in accord with non-perceptible essences and Forms and
in particular the essence and Form of oneness or unity (to hen).
I argue that the principal question to which this passage is addressed is this:
How is it possible for us to speak of and make judgements about physical
things, if our statements and judgements are to be subject to the principle
of non-contradiction (as this principle is formulated in book IV of
Republic, see esp. b–a; sometimes referred to by critics as the
principle of opposites)? In response to this question, Plato argues that this
is possible, only if there is a kind of thought of which it is true to say that:
first, it is not based in and it is independent of sense-perception; secondly,
its primary objects are sensorily imperceptible essences and Forms and,
thirdly, one of its functions is to individuate physical things according to
essences and Forms and especially the essence and Form of unity. We shall
see that, in the passage, Forms are expressly understood as,
precisely, essences.
I want to note straightaway that this reading of the Republic passage is
compatible with thinking that, in this passage, Plato allows for a mode of
thought and judgement that is based in sense-perception. Our account,
therefore, is mindful of the claim, defended by a number of critics recently,

In this chapter, I use ‘physical things’ and ‘sense-perceptible things’ interchangeably.


 Plato’s Essentialism
which says that Plato’s distinction, in this passage, between sense-
perception and thought, is not as sharp as we might have expected. For
what, on our account, Plato argues in this passage, is not, simply, that
judgement requires a kind of thought that is independent of sense-
perception and whose immediate objects are sensorily imperceptible
essences and Forms; rather, what he argues is that judgement, if it is to
be subject to the principle of non-contradiction, requires this. To argue this
point, Plato does not need to suppose a sharp distinction between sense-
perception and judgement.
On our account of the argument in this passage, Plato supposes, at any
rate for the purpose of this argument, that sense-perception has proposi-
tional and conceptual content. (Whether Plato’s challenges this supposi-
tion in a particular passage of a later dialogue, namely, Theaetetus
–, is notoriously a vexed question, which goes back especially to
Cooper’s classic  paper.) A philosopher may, consistently and without
contradiction, argue that i. to make statements of the form ‘It sensorily
appears to me now that X is Y’ (where X and Y are conceptualised items,
and that X is Y is a proposition), sense-perception is sufficient and also
argue that ii. to make statements and judgements of the form ‘X is Y’, if it
is supposed that such statements and judgements are subject to the
principle of non-contradiction, sense-perception is not sufficient. It is
relatively easy to see that i and ii are consistent, especially if we observe
that prefixing a statement or judgement of the form ‘X is Y’ with the
operator ‘It sensorily appears to me now that . . .’ cancels out the possibility
of the resulting statement contradicting any other statement or judgement
to which the same prefix has been added.
Before we begin, allow me to anticipate a number of clarificatory
requests prompted by the formulation, above, of Plato’s Principal
Hypothesis, in the Republic, for essences and Forms. First, when I say that
judgements and statements ‘conform to the principle of non-contradic-
tion’, I mean that it is possible for the judgements, or statements, to
contradict each other, and it is possible for us take whatever precautions
we deem necessary to avoid that they actually contradict each other.
Secondly, when I say that it is possible to make judgements and statements
that are about sense-perceptible things and that conform to the principle of
non-contradiction, ‘ONLY IF we are capable of individuating certain of

This reading goes back to Burnyeat’s  paper. It has since been adopted by Lorenz (); Moss
(); Storey ().

I am grateful to Henrik Rojahn for impressing on me the need to make these clarifications.

Judgements about sense-perceptible things 
those sense-perceptible things in accord with non-perceptible essences and
Forms and in particular the essence and Form of oneness or unity (to hen)’,
I do not intend this to imply that to every sense-perceptible thing that such
a judgement and statement is about, there corresponds a non-perceptible
essence and Form. I do not intend this implication, both for general
reasons – as I do not think essences are simply universal properties – and
on pains of contradicting Parmenides , when Socrates says that there are
things, such as hair, mud and dirt, of which there are not ‘separate’ Forms.
For I take it that, whatever else Plato may also mean by ‘separate’ (chōris) in
the Parmenides passage, he means, or implies: not perceptible by the
senses; and, further, he does not doubt that we can make judgements
and statements, which conform to the principle of non-contradiction,
about such things (as hair, mud, dirt). Thirdly, what I mean by ‘ONLY
IF we are capable of individuating certain of those sense-perceptible things
in accord with non-perceptible essences and Forms and in particular the
essence and Form of oneness or unity’ is this: If we were not capable of
individuating certain sense-perceptible things in accord with non-
perceptible essences and Forms, and in particular the essence and Form
of oneness or unity, it would not be possible for us to make judgements
and statements, which conform to the principle of non-contradiction,
about any sense-perceptible things. Fourthly, this is compatible with what
Plato says in Parmenides  about hair, mud and dirt, if we make the
following supposition: Hair, mud and dirt, and in general those things
about which he says, in the Parmenides passage, that there are not separate
Forms, are causally dependent on other things, in a way in which those
other things are not causally dependent on them. We may call such things,
I propose, ‘waste’. Finally, this supposition provides a plausible reading of
the Parmenides passage (), if we take ‘hair’ there to refer to hair
clippings and hair clumps, just as ‘dirt’ there may plausibly be understood
to refer to a certain form of waste, and so too ‘mud’, provided we do not
confuse mud, as intended there, with such well-defined compounds as
potter’s clay.
. To whom is the argument in Republic VII. a–a

addressed?
How, on our account of it, does the argument in Republic VII. a–a
fit into the overall argument of the Republic? In particular, how does it fit
into the extended section that starts at the end of book V (c–a), is
picked up at the opening of the Sun Analogy (see VI. a, with its back

 Plato’s Essentialism
reference apparently to the end of book V), and from then on proceeds in
a continuous line up to our passage and some pages beyond it (a)? For
it appears that this extended section of the Republic is principally con-
cerned with a single topic, which is the question of what is distinctive of
philosophy and of how the thinking and character distinctive of the
philosopher can be generated, fostered and developed.
I shall argue that, in this argument, Plato is arguing from the ground up,
that is, in a way that is intended to engage with, and potentially persuade,
even those who do not believe in Forms; such as, in particular, ‘the lovers
of sounds and sights’, who, at the end of Republic V, Plato casts as his
dialectical opposition. As ‘the lovers of sounds and sights’ are characterised
at the end of book V, they are not so much thinkers who self-consciously
articulate and defend a certain view, as people of a certain mindset that
they may be more or less conscious of holding and more or less capable of
articulating and defending. At the same time, the view they hold, as this
view is characterised by Plato there, is, and is treated by Plato as being, a
view that can be held as a self-consciously articulated and defended view.
For it is the view that says that: All there is is sense-perceptible and physical
things. (We might refer to it as phenomenalism-cum-physicalism.)
When, at Republic V. c–a, Plato argues that if all there is is sense-
perceptible and physical things, then there can be no such thing as epistēmē
about anything, he is, precisely, treating the view of ‘the lovers of sounds
and sights’, irrespective of the question of how people may hold it, as a
certain articulate view addressed to the question ‘What is there?’
Why is it right to think, as I propose we do, that Plato returns to the
view of the lovers of sounds and sights at this late stage, towards the end of

a–: kai anamnēsas humas ta t’en tois emprosthen hrethenta kai allote ēdē pollakis eirēmena (‘and
reminding you of what was said earlier as well as spoken of on many previous occasions’).

In the Divided Line (VI. b), philosophy is, apparently for the first time in the Republic,
associated with the power of dialectic (hē tou dialegesthai dunamis); an association that Plato picks
up and stays with in book VII. It is worth noting that this association was foreshadowed at V. a, a
passage that begins to prepare for the distinction between the philosopher and the lover of sights and
sounds at the end of book V.

See Republic V. b–, when the lovers of sights and sounds are characterised as espousing only
‘colours and shapes and all the things constructed out of such things’ (chroas kai schēmata kai panta ta
ek tōn toioutōn dēmiourgoumena).

I choose this cumbersome label because, it seems to me, neither ‘phenomenalism’ nor ‘physicalism’
will do on their own. This is because it appears that the lovers of sights and sounds do not properly
distinguish between sense-perceptible things and physical things; on the contrary, it appears to be
distinctive of their view that they move without pause between ‘X sensorily appears Y’ and ‘X is Y’.
See Chapter .

For a good account of the relation of book VII to the book V argument in response to the lovers of
sounds and sights, see Rowe ,  f.

Judgements about sense-perceptible things 
book VII? Has he not argued against this view at the end of Republic V?
What he argued at the end of Republic V (c–a) is that the view that
says that all there is is sense-perceptible and physical things is incompatible
with the existence of, and hence also incompatible with the commitment
to, knowledge proper, epistēmē. So much is true on any reading of this
much studied and much-contested argument. For sure, this argument has
dialectical bite, against one who is committed to epistēmē. But what if, on
reflection and in response to Plato’s argument, the phenomenalist-cum-
physicalist is prepared to give up on this commitment? And, in particular,
if she is prepared to give up on this commitment by urging that even if
Plato’s argument shows that a phenomenalist-cum-physicalist cannot
believe in, or aspire to, epistēmē, there is nothing to prevent a phenome-
nalist-cum-physicalist from speaking and thinking about the world – the
sense-perceptible and physical world, which (as she thinks) is all there is –
just as much and just as well as anyone else? Unless and until such a
radicalised phenomenalist-cum-physicalist has been acknowledged and
addressed, Plato’s argument against ‘the lovers of sights and sounds’, from
the end of Republic V, will remain critically incomplete.
It is notable in this regard that when Plato, in book V, introduced the
argument against the lovers of sounds and sights, he was careful not to
imply that they would be offended if told that they cannot believe in or
aspire to knowledge proper; all he said was that they might be offended.
He was, therefore, careful to make room for the possibility that a lover of
sounds and sights would not be offended but would, on the contrary, hold
on to her position while recognising that it only allows one to have belief
and opinion (doxa), it does not allow one to have knowledge proper
(epistēmē), about the things that are.
I shall argue that it is to such a radicalised lover of sights and sounds that
our argument towards the end of Republic VII (a–a) is addressed.
For, on our account of this argument, the position of the lovers of sounds
and sights is incompatible not only with the existence of and commitment
to knowledge, epistēmē, but also with making about physical things state-
ments (logoi) and judgements (kriseis) that are intended to conform to the
principle of non-contradiction. This is a high price for anyone to be
prepared to pay. A phenomenalist-cum-physicalist who is prepared to give

For a good summary of the different recent accounts of the argument at the end of Republic V, see
Lee b.

See d–: ti oun ean hemin chalepainē[i] houtos, on phamen doxazein all’ ou gignōskein, kai
amphisbētē[i] hōs ouk alēthē legomen; (‘But what if this person should be offended with us, when we
say that he opines but does not know, and should object that we are not speaking truly?’)

 Plato’s Essentialism
up on epistēmē may not be prepared to give up on the possibility of making
about physical things statements and judgements that conform to the
principle of non-contradiction.
This does, I recognise, prompt the question of the possibility of an even
more extremely radicalised phenomenalist-cum-physicalist, who is pre-
pared to give up even on making about physical things statements and
judgements that conform to the principle of non-contradiction. From
what I can tell, Plato does not, in the Republic, attempt to respond to this
extreme variety of phenomenalism-cum-physicalism. However, he does so
in the Theaetetus; when, in the first, long argument in that dialogue, he
argues against the view that says that the only way we are cognitively
related to things is by being able to state, not how they are, but how they
appear to us in sense-perception. For it is part of that view, as characterised
in the Theaetetus argument, that the statements and judgements that we
make about things, all of which are of the form ‘It sensorily appears to me
now that X is Y’, cannot contradict each other. They cannot contradict
each other either inter-personally (i.e., in regard to different judgers, at the
same time or at different times) or intra-personally (i.e., in regard to the
same judger at different times or even at the same time in regard to
judgements based in different sense modalities). Having said this much,
I shall not further consider the question of how Plato’s account of thought
and judgement in the Republic, and their relation to sense-perception,
compares to his account of thought and judgement, and their relation to
sense-perception, in the Theaetetus.
. The general structure, and aim, of Plato’s argument

in Republic a–a
I shall proceed by going over the passage twice and in two rounds. In this,
the first round, I want to set out the general structure and aim of Plato’s
argument in a clear and easily surveyable way. In the second round I shall
then fill in the details and attend to the many issues that a reading of this
passage throws up and that need to be considered for an interpretation to
be properly defended.
.. Is the passage basically about the conversion-inducing power

of arithmetic?
It is sometimes said that the aim of this passage is to exhibit the power of
mathematics, and in particular arithmetic, to convert the soul from the

Judgements about sense-perceptible things 
attachment to physical things to the attention to non-physical things
including Forms. While this general characterisation of the passage may
not be wrong, it creates a most misleading expectation of the character of
the passage and misses its exceedingly argumentative character.
Plato begins by saying that the kind of arithmetic that, as he will argue,
has this conversion-inducing power has never before been used to this end:
‘It seems that it [i.e., the study of arithmetic] is one of those studies which
we are seeking that naturally lead to thought; but no one makes right use of
it, that is, as being altogether capable of drawing one up towards being’
(chrēsthai d’oudeis auto[i] orthōs, helktikō[i] onti pantapasi pros ousian;
a–). This suggests that it is not arithmetic as we commonly know
it that is intended to have this power. Perhaps it is obvious that a basic
arithmetic, as taught to children, is not suitable to, and so is never used for,
this end. For at an early stage of the teaching of arithmetic, it may be left
open whether or not numerals signify physical things. But Plato has
pointed out, earlier in book VII, that mathematicians recognise perfectly
well that numerals do not signify physical things and that numbers are not
physical things. It is hard, therefore, to see how he can say that the
arithmetic that has this conversion-inducing power has never before been
used to this end; unless, that is, he intends to imply that, as his argument
will reveal, it is not arithmetic as hitherto understood, not even by
professional mathematicians, that is supposed to have this conversion-
inducing power.
We shall see that the whole passage is principally occupied with the
question, Q: Is this thing that I am seeing (or touching, or hearing, etc.) a
single unitary thing (hen) or, on the contrary, a whole of separate parts? In
asking this question, which ultimately it belongs to the dialectician and
philosopher to investigate, the dialectician and philosopher is borrowing
from the mathematician and arithmetician the concept of oneness (to hen).
However, she is using this concept for a very different purpose than that
for which the arithmetician uses it, since evidently it is not part of the task
of the arithmetician to ask this question, Q. Nor is it part of the task of the
arithmetician to consider the following question, Q’: whether or not the
concept of oneness that is necessary for answering this question, Q, can be
specified by reference to things that are, to sense-perception, conspicuously
unitary. We shall see that the final and substantial part of Plato’s argument
in the Republic passage is dedicated to just this question, Q’ (see
d–a). As Plato will underscore immediately following the
Republic passage, the arithmetician simply takes for granted the concept
of oneness as being no more than the concept of something indivisible and

 Plato’s Essentialism
non-composite, without pausing to ask what it is that is indivisible and
non-composite; or how we can know of such a thing or how we can use
this concept, or the knowledge of it, for understanding what we see, hear,
etc. (see d–e).
I shall argue that the reason why this question – Q. Is this thing that
I am seeing (or touching, or hearing, etc.) a single unitary thing (hen) or,
on the contrary, a whole of separate parts? – occupies a central place in the
argument is that Plato thinks that only if we can pose and properly attend
to this question can we individuate the things that our judgements and
statements are about in such a way as to be entitled to suppose that these
judgements and statements conform to the principle of non-contradiction.
He has argued for this point in Republic IV (–), when he argued
that to suppose that a thing, O, cannot be both F and contrary-to-F (or,
both F and not-F), we must suppose that this thing, O, is a single, unitary
thing and not simply a whole of separate parts.
This point, from Republic IV, is not so hard to grasp. Consider my shirt,
which is an expensive Italian one (though I bought it greatly reduced)
woven together of countless white and countless black threads (or, alter-
natively, grey threads; it is so carefully and intricately woven that it is hard
to tell). Is it true to say, in regard to it, that it cannot be both white and
black? Not immediately, certainly. For, in an obvious way, it is both white
and black: some threads are black, others white. But what about the
individual threads? Surely, they cannot be both black and white. That
may be true, but only on the supposition that the number of the threads
that make up this shirt is not, literally, countless: the supposition that the
shirt can be analysed into a finite, or at any rate a definite, number of
threads. For otherwise, the point made about the whole shirt – that it is

It is sometimes argued that what Plato formulates in Republic IV is not a principle of non-
contradiction, because it is formulated entirely in terms of contraries (enantia), not
contradictories. I think the view that says that Plato’s principle is not one of contradiction can be
questioned. It can be argued (Lachance  has attempted this argument) that Plato’s formulation
of the principle in Republic IV is associated with his use (in Republic and in dialogues both before
and after) of a variety of phrases of the form: ‘saying things opposite to oneself/to themselves’,
enantia legein heautō[i]/heautois. Here the ‘opposite’ statements referred to by these phrases include
both contrary statements and contradictory statements. That Plato should do this is only to be
expected, since the statement ‘O is both F and con-F (at the same time, in the same respects, etc.)’
implies the statement ‘O is both F and not-F (at the same time, in the same respects, etc.)’. It is a
good question whether Plato also thinks that, conversely, the statement ‘O is both F and not-F (at
the same time, etc.)’ implies the statement ‘O is both F and con-F (at the same time, etc.)’. To
suppose that he does, we would have to suppose that he thinks that the truth of a negative
predication of a property of a subject, of the form, ‘O is not-F’, is grounded in the truth of a
positive predication of a contrary property of the same subject, ‘O is G’ (where G is contrary to F).
I cannot attend to this task here.

Judgements about sense-perceptible things 
both black and white – can be repeated about each and every thread. It
appears to follow that, to apply the principle of non-contradiction to this
thing, the shirt, we must suppose that either it or at any rate its ultimate
parts (supposing that it has ultimate parts), make up a single, unitary thing
and not simply a whole of separate parts.
.. A summary view of Plato’s argument and its structure

If we allow ourselves to read these two Stephanus pages, a–a, in a
single breath and as a single, extended argument, rather than attending
myopically to the finger passage and the issue of compresence of opposites,
the structure of Plato’s argument will, I believe, stand out clearly and be
readily recognisable.
STEP  (Lines a–a). It can be recognised that, with regard to
certain things or qualities, F, which generally come in opposites,
such as, for example, thick and thin, large and small, black or
white (or dark and light), hard and soft, the senses
SOMETIMES report to the soul that the same thing is at the
same time both F and contrary-to-F.
STEP  (Lines a–d). As a result, the soul that has these sensory
perceptions is, in such cases, puzzled about something appar-
ently inadequate about that which is reported by these sensory
perceptions and what these sensory perceptions say. And so the
soul calls upon thought to investigate the content and the object
of these sensory perceptions – what they are of or about – and
to consider whether it is one thing or more than one thing that
appears to the senses to be at the same time both F and
contrary-to-F.
STEP  (Lines d–a). But we (i.e., we dialecticians and philos-
ophers, here exemplified by Glaucon and Socrates) must ask
whether or not it is possible for the thought that asks this
question (i.e., the question whether it is one thing or more than
one thing that appear to the senses to be at the same time both
F and contrary-to-F) to answer it on the basis of sense-
perception.

I put ‘sometimes’ in CAPITALS to emphasise that there is no implication that always and in all cases
when we feel something, e.g., as hard, we also feel it as soft. I elaborate on this later. For the relevant
contrast, see n. .

 Plato’s Essentialism
STEP  (Lines a–). It is evident, on reflection, that it is not
possible for thought to answer this question on the basis of
sense-perception. This is because it is evident that ANY
THING that we see (or touch, or hear, etc.), we see at the
same time as being one and as being indefinitely many.
The obvious conclusion, apparently left to the reader to draw, is that, if
thought is to answer this question (i.e., whether it is one thing or more
than one thing that appears to the senses to be at the same time both F and
contrary-to-F), it must do so in a way that is, in part, independent of sense-
perception. We may suppose that this way will involve a kind of thought
that is independent of sense-perception and whose objects are sensorily
imperceptible essences and Forms.
Plato does the reader the service of expressly drawing this conclusion
when he says that it is on the basis of this whole reasoning that we arrive at
the distinction between sense-perceptible things (aisthēta) and intelligible
things (noēta): ‘And in this way we called the one “the intelligible” (to
noēton) and the other “the visible” (to horaton)’ (at c). This basic
division was, of course, introduced earlier in this extended section of the
Republic, namely, at the opening of the Sun Analogy in book VI; and it
was spelled out further in the Divided Line passage, which followed the
Sun Analogy. But here, in our passage (a–a), Plato corroborates
this basic division with a rigorous argument.
Even on this minimally interpretative and, as it seems to me, practically
purely paraphrastic account of Plato’s argument, it is clear what the gist of
the argument, and its overall conclusion, is intended to be: i. The senses,
or the soul on the basis of the senses, make certain reports or statements,
such as This same thing (that I feel) is at the same time both hard and soft; ii.
but there is something inadequate, puzzling and problematic about these
reports and statements, or their corresponding judgements, and this prob-
lem requires careful thought for its resolution; iii. the problem cannot be
resolved by a kind of thought that is based on sense-perception; therefore,
iv. the making, on the basis of sense-perception, of such judgements as
This thing (that I feel) is at the same time both hard and soft, requires, if the
judgements are to be adequate, a kind of thought that is independent of
sense-perception and whose objects are sensorily imperceptible essences
and Forms, and especially the Form of oneness or unity.

I put ‘any thing’ in CAPITALS to emphasise that here there is indeed an implication that this is true
in all cases.

Judgements about sense-perceptible things 
.. The crux (both substantial and interpretative) of Plato’s argument

The critical question is: What, in Plato’s view, is inadequate about such
reports by sense-perception, and the statements they contain, as This same
thing (that I feel) is at the same time both hard and soft? The answer, it seems
to me, is evident: such reported judgements are, as they stand and unless
they are further investigated and clarified, at immediate risk of contraven-
ing the principle of non-contradiction, which Plato has articulated earlier
in the Republic (book IV. b–a).
It would not be right to object, against this reading, that it cannot be
supposed that such a judgement contravenes the principle of non-
contradiction, because, as Plato has argued in book IV, it is possible to
construe the judgement in such a way that it conforms to the principle
(e.g., as [Even though I may not be able to tell apart these two parts of the
finger as I am feeling it], it is one part of the finger (i.e., the bone) that is hard,
another that is soft). For this is precisely Plato’s point here: that, as it stands
and as is reported simply on the basis of sense-perception, we cannot tell
whether or not this judgement contravenes or it conforms to the principle
of non-contradiction. To tell this, he argues, the judgement needs to be
investigated further – this is the investigation that the soul calls upon
thought to undertake – and to be revised and construed accordingly. What
he is arguing in the Republic VII passage is that, to operate with the
distinctions that were argued in book IV to be necessary for making about
sense-perceptible and physical things judgements that conform to the
principle of non-contradiction (and especially the distinction between a
single, unitary thing and a whole of parts), not only sense-perception, but
also a kind of thought that is independent of sense-perception are needed.
It would, likewise, not be right to object that it cannot be supposed that
Plato immediately intends the thought that is called upon to investigate
such judgements, reported simply on the basis of sense-perception, to be
independent of sense-perception. For what is remarkable is that Plato
immediately (i.e., at d–a; this was STEP  above) anticipates
this point, or objection. Having urged that such reports and statements are
inadequate and puzzling as they stand and unless they are further investi-
gated by thought, he goes on to consider how the thought that is suitable
for this investigation is related to sense-perception and, in particular:
whether or not thought can carry out this investigation on the basis of
sense-perception, and any thought based on sense-perception, alone. Plato
does not assume that the answer to this latter question is negative, he
expressly asks whether it should be answered in the affirmative or in the

 Plato’s Essentialism
negative. He answers it in the negative for a particular reason and argu-
ment: that that which we see (or touch, or hear, etc.), we see at the same
time as being one and as being indefinitely many (apeira to plēthos,
a–; STEP  above).
Why is it evident that the reason why such judgements, reported by, or
on the basis of, sense-perception, are puzzling and inadequate is that, as
they stand and unless they are properly revised by thought, they are at risk
of contravening the principle of non-contradiction?
First, it is hard to see what else could be thought to be inadequate
about them.
Secondly, Plato’s choice of two particular terms, when formulating these
judgements, indicates that he has the principle of non-contradiction in
mind: he repeatedly says that such judgements report a thing as being ‘at
the same time’ (hama) both F and contrary-to-F; and he says that such
judgements report ‘the same thing’ (tauton) as being at the same time both
F and contrary-to-F. He will go on to clarify that when such sameness, or
such unity, is attributed by sense-perception to an object of sense-
perception (such as the finger, or part of the finger, that I am currently
feeling), it is a confused (sugkechumenon), as opposed to a distinct
(kechōrismenon) and determinate (dihōrismenon) sameness, or unity, that
is attributed. We shall consider these clarifications presently.
Thirdly, he says that the question that these judgements immediately
prompt, and that thought is called upon to investigate, is whether it is one
thing or, on the contrary, more than one thing that is at the same time
both F and contrary-to-F. This is precisely the procedure Plato set out
when he first introduced the principle of non-contradiction in book IV
(b–a). There he recommends this way of proceeding in general,
not only in regard to the analysis of soul and the question whether the soul
is a single, unitary thing or, on the contrary, a whole of parts. In book IV,
he first states the procedure in general terms:
It is obvious that the same thing will not be willing to do or undergo
opposites in the same part of itself, in relation to the same thing, at the same
time. So, if ever we find this happening to the soul, we’ll know that we
aren’t dealing with one thing, but many. (b–c; trans. Grube-Reeve)
He then goes on carefully to explicate this procedure, in regard to the

application of the principle of non-contradiction to physical things: a
human being not moving from the spot but moving her hands and her
head, a spinning top spinning round its axis on the same spot on the
ground. He finally formulates the principle in general terms:

Judgements about sense-perceptible things 
No such statement will disturb us, then, or make us believe that the same
thing can have, do, or undergo opposites, at the same time, in the same
respect, and in relation to the same thing. (e–a; trans. Grube-
Reeve, slightly adapted)
Only then does he apply this procedure to the soul, in the way with which
we are all familiar (i.e., up to c). The explication of how the principle
applies to the corporeal human being and the spinning top is, precisely,
that it is not a single unitary thing that is both F and contrary-to-F (e.g.,
moving and standing still), but rather one thing, or the one part of the
compound thing, is F and the other thing, or the other part of the
compound thing, is contrary-to-F.
.. What judgements, as reported by sense-perception, need to be

investigated further by thought, before they can be allowed to stand?
It should be clear that Plato does not mean to say that anything that we
sense as F (e.g., as soft) we at the same time also sense as con-F (e.g., as
hard). This would be so obviously false as not to deserve consideration.
What he does is identify and spell out certain cases in which what is sensed
as F is at the same time also sensed as con-F; for example, the case in which
one is feeling a particular part of one’s finger. At the same time, it is clear
that, when he says that such reports by sense-perception need to be
investigated further by thought, before they and the judgements they
contain can be allowed to stand, he intends to generalise from these cases.
The question is: How is this generalisation to be understood? And how
extensive is it intended to be?
When I touch a certain part of my finger in a certain way, it may be
natural to report what I feel by saying that The same thing at the same time
feels, and apparently is, both hard and soft. Certainly, it need not be
apparent to this sensory perception that it is one part of the finger (i.e.,
the bone) that feels hard, another part of it (i.e., the skin and tissue) that
feels soft. But, Plato urges, the judgement, This same thing at the same time
is both hard and soft cannot be allowed to stand, before it is further
investigated by thought and revised and construed accordingly; because,
unless it is settled, above all, whether or not it is strictly the same thing – a
single, unitary thing – that is both hard and soft, the judgement is at risk of
contravening the principle of non-contradiction. Let us say, then, that, on

Or ‘be’, kai eiē. The kai eiē is most important, for it implies that the principle is not confined to
doings and sufferings.

 Plato’s Essentialism
reflection, I revise the judgement and now formulate it as: One part of this
thing is hard, another soft; and let us suppose, following Plato, that this
revision is conducted not simply by sense-perception, but by thought –
but bearing in mind that, at this point of Plato’s argument, it is left open
how sense-perception and thought are related.
How is the intended generalisation, based on a case such as this, to be
understood? For, evidently, Plato does not intend the lesson to be that
such revision by thought is relevant only for these cases. This ought to be
evident. Suppose I am touching an iron rod. It feels, purely and simply,
hard. We may hardly suppose that Plato thinks that, in this case, only
sense-perception is needed to make the adequate judgement: This thing is
hard. Why not? Why is it evident that Plato intends to generalise? Plato’s
general point (which he made already in book IV when he introduced the
principle of non-contradiction) is that, to judge of a thing, O, that it is F,
and not at the same time con-F (and hence not-F), we need to consider
whether the thing, O, is a single, unitary thing or a compound of
distinct parts; because this consideration is necessary in order to ensure
that this judgement conforms to the principle of non-contradiction. We
may put this point by saying that, for Plato, to apply the principle of
non-contradiction to a thing, for the purpose of making a judgement
about it that conforms to this principle, we need to consider whether this
thing is a single, unitary thing or a compound of distinct parts. In book IV,
as we have just seen, he made this point in general terms, before concen-
trating on its application to the soul and the phenomenon of the soul’s
being subject, consistently with the principle of non-contradiction, to
radically conflicting desires; at any rate the soul is subject to such conflict
occasionally. In this passage in book VII, he is considering this point in
regard to judgements made by, and on the basis of, sense-perception and
hence about the objects of the senses, such as a finger and its hardness and
softness. In regard to the cases in which that which appears to a single sense
to be a single thing, and appears both F and con-F, Plato argues that the
statement This same thing at the same time is both F and con-F, as reported
to the soul by the senses, cannot be allowed to stand, before it is further
investigated by thought for the purpose of ensuring that it conforms to the
principle of non-contradiction. Apparently, however, there is this need, to
ensure that judgements based in sense-perception conform to the principle
of non-contradiction, just as much in regard to other judgements based in
sense-perception, such as This thing (i.e., the iron rod) is hard. For what this
judgement is about, being a physical thing, is likewise a compound of
parts. (That physical things are compounds of parts was clearly stated in

Judgements about sense-perceptible things 
the Phaedo.) And so it needs to be considered in regard to what thing this
judgement is made: a particular part of the thing or the whole compound
thing? Unless this is considered, the thing will not be sufficiently individ-
uated, for the purpose of making judgements about it of which we are
entitled to suppose that they conform to the principle of non-
contradiction.
I turn now to the question: How extensive is this generalisation
intended to be? It seems to me clear that we may not suppose that this
generalisation is, purely and simply, unrestricted, that is, that it is intended
to apply to any and all judgements and statements, that is, to anything that
can rightfully be called a judgement (here krisis, see krinomena at
VII. b; or statement, logos, see esp. legein at a). To suppose this,
we would have to suppose that Plato thinks that being subject to the
principle of non-contradiction is (as we would say) part of the concept of a
judgement or statement, as this concept is encountered in everyday dis-
course. But we have no evidence, certainly not in this passage from the
Republic, to suppose this; and the supposition has the air of bad anachro-
nism. Much more reasonable is to suppose that, if the generalisation is
intended to apply to all judgements (we shall consider this if in a moment),
this needs to be understood to mean that it is intended to apply to all
judgements that conform to the principle of non-contradiction. On this
understanding, Plato leaves it open whether or not all judgements need to
conform to the principle of non-contradiction, indeed, whether any judge-
ments need to conform to this principle. From what I can tell, he does not
address these questions in Republic. But he does address them in the
Theaetetus, especially when, in the first, long argument of that dialogue,
he argues that the supposition that no judgements need to conform to the
principle of non-contradiction has seriously unattractive consequences. For
the purposes of the present argument in Republic VII, Plato is only
concerned with judgements that do conform to the principle of non-
contradiction and with judgements precisely in so far as they conform to
this principle.
The question, therefore, is whether, in the Republic argument
(VII. a–a), Plato thinks that (OPTION ): All judgements,
that are based in sense-perception and that conform to the principle of
non-contradiction, are the product of sense-perception and thought;
or, on the contrary, he thinks that (OPTION ): Some but not all
judgements, that are based in sense-perception and that conform to the
principle of non-contradiction, are the product of sense-perception
and thought.

 Plato’s Essentialism
It seems to me that there is a conclusive case in favour of OPTION
 and against OPTION . The conclusion of Plato’s argument is that, to
make a judgement about a physical and sense-perceptible thing (e.g., a
finger) that conforms to the principle of non-contradiction, we need to
know what oneness or unity is, for the purpose of adequately individuating
this thing, the finger, and that we cannot know this by means of sense-
perception, or a kind of thought that is based on sense-perception:
because, as he concludes at a–, any thing that we perceive by the
senses to be one, we at the same time perceive by the senses to be
indefinitely many. It seems clear that this conclusion is intended to hold
good of any judgement that is based in sense-perception and that conforms
to the principle of non-contradiction. Certainly, there is no indication in
Plato’s argument that it is restricted to some judgements that are based in
sense-perception and that conform to the principle of non-contradiction.
It would, it seems to me, be questionable – I think tendentious – to say
that there is a single, clear indication in favour of OPTION , the
indication being that Plato says that for some judgements based in sense-
perception, sense-perception is adequate (hikanon; see the hikanōs at
b), by itself and without the need for thought to investigate its
content and its reported judgements. As I understand Burnyeat’s bold
statement in his  paper, this is what he says, when he says (,
emphasis added): ‘There is no talk here [i.e., ‘in Book vii (a–a)’]
of different parts of the soul, but the senses have a considerable autonomy,
not only in that they do the perceiving, but also as constituting an indepen-
dent source of judgements.’ Burnyeat must, from what I can tell, be under-
stood to mean that, in some cases at any rate, the senses constitute a source
of judgements that is independent of thought and that, consequently,
these judgements are independent of thought. If I am not mistaken, some
recent critics have been inclined to follow Burnyeat in this.
It is, certainly, true that Plato expressly distinguishes between, on the
one hand, judgements (such as This [e.g., this finger] is at the same time both
hard and soft) based in sense-perception that immediately and directly
prompt the soul to call upon thought to investigate them, and, on the
other hand, judgements based in sense-perception (such as This is a finger),
that do not immediately or directly prompt the soul to call upon thought
to investigate them (see esp. a–b). But, as Harte (, ) has
been at pains to show, he also says, in regard to the latter, that it is the soul
of the lay person (of ‘the many’, hoi polloi) that is not immediately or

Lorenz (); Moss (); Storey ().

Judgements about sense-perceptible things 
directly prompted to investigate them, in this way implying that the soul of
the dialectician and philosopher may be so prompted. We have seen that
the soul of the dialectician will indeed be so prompted, for the purpose of
making out how we can make judgements about physical and sense-
perceptible things, and in general things that are compounds of distinct
parts, of which we are entitled to suppose that they conform to the
principle of contradiction. It would be very strange to suppose that fingers
are exempted.
It might be said that we owe the reader an explanation, textually based,
of how Plato intends the distinction, which he expressly makes, between
sense-based judgements that require thought and those that do not, if not
in the way indicated by Burnyeat. The answer is that this is not i. a
distinction between sense-based judgements that require thought and
sense-based judgements that do not; it is, rather, ii. a distinction between
sense-based judgements of which it is immediately and directly apparent,
and hence apparent even to the soul of ‘the many’ (hoi polloi), that thought
is required to make them good, and judgements of which this is not
immediately or directly apparent, though it may become, and be, apparent
to the dialectician and philosopher.
It will still be said that we owe the reader an explanation of how our
reading is compatible with the following sentence, and especially the two
underlined phrases in it: ta men en tois aisthēsesin ou parakalunta tēn noēsin
eis episkepsin, hōs hikanōs hupo tēs aisthēseōs krinomena, ta de <en tois
aisthēsesin> pantapasi diakeleuomena ekeinēn episkepsasthai, hōs tēs
aisthēseōs ouden hugies poiousēs (‘On the one hand, those contents of
sensory perceptions that do not call upon thought to investigate, on the
supposition that they are adequately judged by sense-perception; on the
other hand, those <contents of sensory perceptions> that altogether move
it [i.e., the soul] to investigate, on the supposition that sense-perception
cannot arrive at anything sound’, a–b; my translation). The answer
is, it seems to me, clear. The two phrases that I have underlined are clearly
parallel and form an important part of the way in which the distinction is
marked. But the hōs constructions in these two phrases (hōs hikanōs hupo
tēs aisthēseōs krinomena; hōs tēs aisthēseōs ouden hugies poiousēs) need not be
understood to state a fact, or a supposition simpliciter; they need only be
understood to state a supposition made from a particular point of view.
Evidently, the point of view here is that of any person, including, therefore,
the lay person and ‘the many’. For no one wants to give the impression of
contradicting herself in what she says, or even at being at apparent risk of
contradicting herself in what she says. Hence everyone is puzzled by, and

 Plato’s Essentialism
hence inclined to investigate further, sense-based judgements of the form
This same thing is at the same time F and con-F. But it is not the case that
everyone is puzzled by, and hence inclined to investigate further, sense-
based judgements that are not of this form, such as This is a finger.
We may conclude that the conclusion of Plato’s argument, in regard to
its content, its logical form and its scope, is this: Any judgement, if it is
based in sense-perception, and if it conforms to the principle of non-
contradiction, is the product of, on the one hand, sense-perception, and,
on the other hand, a kind of thought that is independent of sense-
perception and whose primary object is the essence and Form of oneness
or unity – that which oneness or unity is.
. An analysis of Plato’s argument, and his verbal means

of indicating it, in greater detail
.. The ‘finger’ passage and its place in the overall argument
(a–b)
The ‘finger’ passage is the most familiar part of Plato’s argument, when he
explicates the distinction between two uses of sense-perception and does so
through the example of one’s perceiving the three middle fingers of one’s
hand. If the question to which sense-perception is addressed is ‘Is this a
finger?’, then perception does not, at any rate not immediately and
directly, issue in opposite answers, depending on whether it is the finger
in the middle or on the outside, a black or a white finger, a thick or a thin
finger, that one perceives. In such cases, ‘it is not the case that the soul of
the lay person (tōn pollōn) is compelled to enquire of thought “What is a
finger?”’ (d–). On the other hand, if the question to which sense-
perception is addressed is ‘Is this (pointing to the finger) something big or
something small, something thick or something thin, something hard or
something soft?’, then perception does immediately issue in opposite
answers: ‘the sensory perception will report to the soul that it is perceiving
the same thing as being [e.g.] both hard and soft’ (paraggelei [hē aisthēsis] tē[i]
psuchē[i] hōs tauton sklēron te kai malakon aisthanomenē, e–a).
In such cases, the way in which sensory perceptions report their content

For this point, about the force of the hōs hikanōs here, see Mann (, –). Mann
(–) also mentions a number of critics, other than those I have mentioned, who hold the
view that I have found to go back to Burnyeat () and that both Mann and I are arguing against:
Nehamas (; a) and Annas (, , –). In favour of the view that Mann and
I are defending, he refers also to Patterson (, –).

Judgements about sense-perceptible things 
will not be adequate (cf. endeōs ta toiauta dēlousin, ‘they [the sensory
perceptions] declare such things in an inadequate way’, e); on the
contrary, these reports will be ‘absurd to the soul and in need of investiga-
tion’ (hautai ge atopai tē[i] psuchē[i] hai hermēneiai kai episkepseōs deomenai,
b–, b–). In these cases, it is necessary that ‘the soul should, in
its turn, puzzle over the question, “What does this sensory perception
indicate as the hard thing, since it says of the same thing that it is also
soft?”’ (tí pote sēmainei hautē hē aisthēsis to sklēron, eiper to auto kai malakon
legei, a–).
It is often thought that Plato’s argument is substantially complete by the
end of the ‘finger’ passage (i.e., by b), and that what follows (i.e., up
to a) serves basically to extend the range of characteristics that issue
in compresence of opposites, from such qualities as thin-thick, large-small,
hard-soft, to numerical characteristics: one, two, three, etc. We have seen,
and shall further see, that there is much more to what follows the ‘finger’
passage and that what follows is crucial to the structure and logic of Plato’s
argument. For we have not yet been told what the investigation is that the
soul needs to conduct by means of thought in regard to the reports by
certain sensory perceptions of their content. Or what is to be accomplished
by this investigation. Or why this investigation directs thought to intelli-
gible things. Or even why this investigation is necessary. Or how, and by
what ability of the soul, this investigation is to be conducted. We will be
told all this but only after the finger passage.
.. The locus of a problem with sense-perception, and the

general character of this problem
Plato locates a problem with sense-perception in its content: that which it
is of. The content of sense-perception is referred to as ‘the things that are
present in sensory perceptions’ (ta en tois aisthēsesin, a) and as ‘the
things that appear’ (ta phainomena, b; also b and c). It is this
content, so characterised, of which it is said that in some cases it does
not, or at any rate not immediately, direct the thought of the common
person to investigate it whereas in other cases it does.
The phrase ta en tois aisthēsesin (‘the things present in sensory percep-
tions’, a) is particularly important here (in a–b, quoted
above). In this sentence, the term hē aisthēsis is used in two different ways:
in the plural, to refer to a particular sensory perception (e.g., the one I am
having now when looking at my finger), and, in the singular, to refer to the
ability for sense-perception or to a sense-modality. For it is clear that the

 Plato’s Essentialism
term ta en tois aisthēsesin cannot mean ‘the things present in the ability for
sense-perception’ or ‘the things present in the modalities of sense-percep-
tion’; it must mean ‘the things present in sensory perceptions’. What this
shows is that the question that Plato is addressing here is whether or not
the content of a sensory perception is adequate for a judgement about a
sense-perceptible thing.
Plato says that the thought-provoking use of sense-perception is thought-
provoking ‘because sense-perception does not arrive at anything sound’
(hōs tēs aisthēseōs ouden hugies poiousēs, b–). Plato clarifies what kind
of deficiency of, or problem with, sense-perception he has in mind, when
he says that the deficiency or problem does not concern either the lack of
precision of the perceptions (e.g., due to the distance of the things that
appear, ta porrōthen phainomena) or their lack of clarity (as in the case of
the perception of shadow-paintings, b–; also c). This is an
important clarification, especially in view of the familiar complaint that
Plato makes, on behalf of others, against sense-perception in Phaedo
(a–b), when he says that, according to certain views, it is a hindrance
to knowledge because it is neither precise (akribēs) nor clear (saphēs). It
appears that it is not this epistemological problem with sense-perception
that he intends here.
A crucial question for understanding Plato’s argument, therefore, is
what kind of inadequacy and absurdity he has in mind in regard to the
second case. It is apparent that this is, as we would say, a logical absurdity.
The problem will be spelled out as being that, unless these reports by
sense-perceptions of their content are investigated by thought, the judge-
ments based on them will have the form of a potential contradiction: O is
F and opposite-of-F at the same time. If so, then these reports must say of
one and the same thing that it is both F and opposite-of-F at the same time.
That opposite predications can amount to a contradiction only if they are
predicated of one and the same thing is a point that Plato made with
particular attention when, in book IV, he formulated and defended the
principle of non-contradiction (see b–c and e–a). In our
passage it is pointed out that such reports are indeed of this form: ‘the
sensory perception will report to the soul that it is perceiving the same thing
(tauton) as being both hard and soft’ (e–a); and, ‘Is it not
necessary that in such cases the soul should, in its turn, puzzle over the

I follow Dancy (, ) here and disagree with Moss (, ), when she says that ‘The
famous finger passage of Republic  ( A ff.) makes the same claim [as Phaedo b–c]’.

Judgements about sense-perceptible things 
question, “What does this sensory perception indicate as the hard thing,
since it says of the same thing (to auto) that it is also soft?”?’ (a–).
.. In response to an objection

On a number of occasions when I have presented this account of Plato’s
argument, I have met with the following objection: This account of Plato’s
argument may be plausible in regard to his example of the same finger
feeling, to the sense of touch, at the same time both hard and soft (i.e., the
example at a–); but it is not plausible in regard to his example of the
same finger looking, to the sense of sight, at the same time both large and
small (i.e., the example at e–). For (so the objection goes), there is
no risk of a contradiction in the statement “This same thing is at the same
time both large and small”. And indeed, even if there is a risk of contra-
diction, it is evident that the senses, or any thought based on the senses, are
sufficient to construe the statement in such a way as to ensure that it is not
a contradiction.
The first thing to say, in response to this objection, is that, just as hard
and soft are contrary qualities, so too large and small are contrary qualities.
It is hard to see, therefore, why, if the statement ‘This same thing is at the
same time both hard and soft’ is at risk of expressing a contradiction, the
same is not also true of the statement ‘This same thing is at the same time
both large and small’. Of course, if Plato had been talking of the relational
qualities, larger than and smaller than, then there would have been no risk
of contradiction, since the relevant statement would have been: ‘This same
thing is at the same time larger than (one thing) and smaller than (another
thing).’ But Plato uses ‘large’ and ‘small’, not ‘larger than’ and ‘smaller
than’; and we cannot simply assume that, when he says ‘large’ and ‘small’,
he means ‘larger than’ and ‘smaller than’.
The next thing to do, in response to the objection, is consider whether it
is evident that the senses, or any thought based on the senses, are sufficient
to construe the statement ‘This same finger is at the same time both large
and small’ in such a way as to ensure that it is not a contradiction. The
impression that this is evident is due, I suppose, to the supposition that it is
evident that the senses, or any thought based on the senses, are sufficient
for making certain comparative judgements, such as ‘This one thing looks,
and is, larger than that other thing.’ Let it be granted that the senses, or any
thought based on the senses, are sufficient for making such comparative
judgements. It is not clear that it is such a judgement that Plato has in
mind in e–, when he says: ‘But what about the largeness and the

 Plato’s Essentialism
smallness of the finger? Does sight see these adequately, and does it make
no difference to it whether one of the fingers is located in the middle [of
the hand] or at one end?’ (trans. Rowe, slightly amended). On the
contrary, the fact that Plato thinks it matters to the force, and the
logic, of his example whether the finger in question is located in
the middle or at the end of the hand tells against the view that the example
simply concerns comparative judgements such as ‘This finger looks, and is,
larger than that finger.’
What, then, is the form of the judgement here, if it is not a simple
comparative judgement such as ‘This finger looks, and is, larger than that
finger’? And why is it not evident that the senses, or any thought based on
the senses, are sufficient for making such a judgement? These are not easy
questions; and we may wonder whether Plato’s example at e– is
sufficiently spelled out to answer them properly. However, it seems to me
that what Plato intends to indicate, by pointing up the question whether
the finger that is being observed is located in the middle or at the end of
the hand that one is looking at, is that the relevant judgement involves the
notion of a complex structure – indeed something like a geometrical and
topological structure – which is the structure exhibited by the arrangement
of the five fingers into a hand. Plato’s point seems to be, therefore, that,
whether a finger is large or small depends on where it is positioned in such
a complex structure.
Now, it would be presumptuous to think that it is evident that the
senses, or any thought based on the senses, are adequate for making
judgements involving the grasp of such structures. Certainly, this is not
evident to a philosopher who argues, as Plato will duly do (see
d–a, which I shall consider in detail in a moment), that the
notion of strict and determinate unity cannot be specified by pointing to a
thing that looks (feels, etc.) conspicuously unitary. For it is arguable, and
perhaps even immediately plausible to suppose, that the notion of deter-
minate unity is necessary for the grasp of such structures.
.. Why unity and number are brought into the argument
(b–c)
At this point unity is brought into the argument. Notably, this is not done
for the purpose of considering whether, as it appears in sense-perception,
unity issues in compresence of opposites: this question is posed only later
(d). Unity is brought in because it is through the use and application
of precisely this concept that thought, when called upon to investigate the

Judgements about sense-perceptible things 
reports of the content of sensory perceptions, carries out this investigation.
The passage is worth quoting in full:
[GLAUCON] These reports (hermēneiai) will indeed be absurd to the soul
and in need of investigation. [SOCRATES] It is appropriate, therefore, that
in these cases the soul should first of all (prōton) attempt, by calling upon
reasoning and thought, to consider whether each of the things reported
(hekasta tōn eisaggellomenōn) are one or two (eite hen eite duo). [GL.]
Certainly it is. [SOC.] Is it not the case, then, that if it [i.e., the thing
reported] should turn out to be two, then each of the two will emerge as
being one? [GL.] Indeed. [SOC.] If, therefore, the soul thinks of each of
these things as one and thinks of both as two, it will think of the two things
as distinct (or, ‘as separate’, kechōrismena); for if it did not think of them as
distinct (or, ‘if it thought of them as inseparate’, achōrista), it would not
think of them as two, but as one. [GL.] Correct. (b–c)
[SOC.] On the other hand, we maintain, don’t we, that sight saw some-
thing (ti) as being both large and small, and not something distinct
(kechōrismenon), but rather, confused (sungechumenon). [GL.] We do.
[SOC.] It is for the sake of clarity on this point (dia de tēn toutou
saphēneian) that thought was compelled to inspect a thing large and also
small, and to inspect them not as confused things (sungechumena) but as
determinate things (dihōrismena), as opposed to what sense-perception did.
[GL.] True. [SOC.] Is it not the case, then, that this is the basis of our
thought’s arriving at the question, ‘What then is the large thing and also the
small thing?’? [GL.] Most certainly. (c–)
[SOC.] And so we called one <kind> that which is intelligible and the
other <kind> that which is visible. [GL.] Perfectly correct. (c–;
my translation)
Why, we must ask, is thought’s first task, in response to the apparent
absurdity of the reports of the content of sensory perceptions, ‘to consider
whether each of the things reported (hekasta tōn eisaggellomenōn) are one or
two (eite hen eite duo)’? Apparently, what the soul considers in considering
this is how the object of perception is to be individuated. But we must ask
why the soul has to consider this, and to what end. The evident answer, in
view of Plato’s argument up to this point, is that the object of perception
must be individuated in a way that is adequate for making a judgement
about it (e.g., ‘This thing is hard’), and avoids the logical absurdity of the

tí oun pot’ esti to mega au kai to smikron (c–). See below for the meaning of this indirect
interrogative phrase.

 Plato’s Essentialism
original reports by the sensory perceptions, which, unless further investi-
gated, have the form of a contradiction.
What this shows is that the principal task of thought here is to consider
how the object of perception must be individuated for it to be possible to
make about this object statements and judgements that conform to the
principle of non-contradiction. And the first step in this task is to consider
whether the subject of predication (here referred to as ‘the thing reported’,
to eisaggellomenon) is a single, unitary thing or, on the contrary, a com-
pound of (at least) two distinct things. If, for example, the soul were,
through this investigation based on reasoning and thought, to arrive at the
conclusion that the thing reported, which is reported by the sensory
perception as a single thing (cf. tauton at a) that is both hard and
soft, is in fact a compound of (at least) two distinct things, then it may
revise the report so that the judgement based on it says that the one thing,
or the one part of the compound thing, is hard and the other thing, or the
other part of the compound thing, is soft.
Does the revision by thought of the original report by a sensory
perception amount to a rejection of the report or only to a clarification
of it? The original report said that one and the same thing (tauton) is both
hard and soft (see a–). It may seem to follow that, if thought were,
through its investigation, to arrive at the conclusion that the thing reported
is really a compound of two distinct things, it would be rejecting the
original report by the senses. This follows, however, only if we suppose
that, if a sensory perception reports that one and the same thing is both hard
and soft, this means that a single distinct and determine thing is both hard
and soft. However, this is just what Plato denies, when he clarifies that,
unlike the attribution of unity by thought, the attribution of sameness by
sense-perception can only amount to the attribution of a sameness that is
indistinct (achōriston), indeterminate (the denial of dihōrismenon; the term
aoriston is not used) and confused (sunkechumenon). We may conclude
that thought’s revision of the reports by sensory perceptions of their
content serves not to accept or reject these reports but rather to clarify
and interpret them in such a way that the subject of predication, which is
here the object of perception, is adequately individuated for it to be
possible to make statements and judgements about it that are subject to
the principle of non-contradiction.
What would be the implication, if the soul, while recognising that
certain reports by sensory perceptions of their content have the form a
contradiction unless further investigated, did not call upon thought to
investigate and clarify these reports? The wrong answer, obviously, is that

Judgements about sense-perceptible things 
there would be true contradictions: Plato nowhere contemplates the
possibility of true contradictions. The right answer, it seems to me, is
that the principle of non-contradiction would not, without serious restric-
tion, apply to sense-perceptible things and to phainomena. What this
means it that it would not be possible to make about sense-perceptible
and physical things statements and judgements that conform to the
principle of non-contradiction. (When I say that judgements and state-
ments ‘conform to the principle of non-contradiction’, I mean that it is
possible for the judgements, or statements, to contradict each other, and it
is possible for us take whatever precautions we deem necessary to avoid
that they actually contradict each other.)
It is a consequence of this reading of this part of Plato’s argument that the
adjective hen, here and for the remainder of the argument (see b, b,
b, c; and later, e and a), is used to mean ‘one’, in the sense of ‘a
single, unitary thing’, and, especially, a single, unitary whole of parts, and
that the noun phrase to hen in what follows (see esp. e, a, a, a) is
used to designate the quality itself, unity, in the sense of: that on account of
which a single, unitary thing is single and unitary (see esp. ‘. . . “What is this
thing itself, unity?” . . .’, tí pote estin auto to hen, at a).
Is this reading of hen and to hen consistent with the fact that Plato wants
to defend the claim that arithmetic is a study that is capable of directing
thought to intelligible things? And with the fact that, at the end of the
argument, Plato takes this claim, about arithmetic, to have been estab-
lished (a–)? Admittedly, it is not consistent, if we suppose that
what he wants to defend is that arithmetic is a study that is capable, by
itself, of directing thought to intelligible things. For, obviously, the ability
for arithmetic does not have to include the ability for the kind of intellec-
tual activity that, on the present reading, the soul has to engage in in order
to investigate the reports by sensory perceptions of their content. But we
saw earlier (with reference to a–; see the opening of Section .) that
what Plato wants to defend is, rather, that arithmetic is capable of this, if
used in a peculiar and new way. We may note that when, at the end of the
argument, Plato returns to arithmetic (a–), he does so as a
deliberate move from the claim that the study of unity is capable of

Unlike Priest (, ): ‘Another example: I walk out of the room; for an instant, I am
symmetrically poised, one foot in, one foot out, my center of gravity lying on the vertical plane
containing the center of gravity of the door. Am I in or not in the room?’ He concludes: ‘I am both
in and not in.’ This passage is a perfect example, it seems to me, of how one may find that a physical
thing can contravene the principle of non-contradiction, if one fails to consider how a thing needs
to be individuated for the principle to apply to it.

 Plato’s Essentialism
directing thought to intelligible things (see a–) to the conclusion
that the study of unity, and, as a consequence, the study of arithmetic, is
capable of this (see a). Notable here is that Plato, for the purpose of
this argument, thinks of the study of arithmetic not alone and by itself but
as part of the study of unity.
.. The meaning of the phrase tí oun pot’ esti to mega au kai to smikron: a
question of how to individuate the object of perception (c–)
This indirect interrogative phrase is usually understood to refer to the
question, Q, ‘What is largeness and smallness?’, and is translated accord-
ingly. However, the phrase can also be understood to refer to the question,
Q, ‘What is the large thing and also the small thing?’ Socrates says that
what they have just been considering prompts (cf. enteuthen pothen, c)
the question: tí oun pot’ esti to mega au kai to smikron. What they have just
been considering is how, without contradiction, one and the same thing can
at the same time be both large and small, thick and thin, hard and soft. This
immediately prompts the question ‘What then is the large thing and also the
small thing?’; in particular, ‘Is it a single thing or two things?’ (see b–).
Suppose the question is Q. And suppose this question is answered, by giving
an account of largeness and smallness, irrespective of whether the account is or
is not in terms of sense-perceptible things or qualities. It is not at all clear how
this assists with the issue Plato is addressing. For, the question he is addressing
is: How, without contradiction, can one and the same sense-perceptible thing
at the same time be both large and small, thick and thin, hard and soft?
It appears that one reason why the interrogative phrase has generally been
understood as referring to Q is that, a little later, Plato will use a seemingly
similar phrase, tí pote estin auto to hen, to refer to the question ‘What is that
thing itself, unity?’ (a) But this overlooks that, in the phrase tí oun pot’ esti
to mega au kai to smikron, the important word, auto (‘itself’), is absent. The
function of this word, auto, in such phrases in Plato is, typically, to refer to the
quality itself, F-ness, as opposed to a particular thing that is F. In the later
passage this word is not only present, it is underscored by the fact that it is said
that what is at issue is ‘unity itself by itself’ (auto kath’ hauto to hen, d–e).
.. The distinction between sense-perceptible things and intelligible

things (c–)
Socrates concludes their argument up to this point with the statement:
‘And so we called one <kind> that which is intelligible and the other

Judgements about sense-perceptible things 
<kind> that which is visible’ (c). We saw that this line refers back
to the fundamental division of all things into these two kinds, and this goes
back to the Sun Analogy (a f.) and indeed to the end of book V (to
which a– in turn refers back). What is the relation between the
present argument and that division? The conclusion of the present argu-
ment is that there is a thing, namely, the Form of unity, which is
independent of sense-perceptible, physical things and which is the object
of a kind of thought that is independent of sense-perception. Evidently, it
is necessary that there should be at least one thing that is not sense-
perceptible or physical, if all things divide into sense-perceptible and
intelligible things. We may suppose, therefore, that Plato’s argument in
Republic VII. a–a is a crucial particular step in the general argument
for the division of all things into two kinds.
The Divided Line passage is especially important for understanding this
one line in Plato’s present argument: c. The Divided Line represents all
things, and the basic division in it is that between these two kinds of things: the
lower section represents sense-perceptible things, the upper section represents
intelligible things and, especially if not exclusively, Forms. The relation
between the lower section and the upper section is that of asymmetric
ontological dependence: sense-perceptible things are dependent, for their
being and being what they are, on intelligible things and ultimately Forms,
whereas intelligible things are not dependent, for their being and being what
they are, on sense-perceptible things (I return to this in Chapter ).
.. Socrates’ last question: How can we specify a concept of unity?

(d–a)
Socrates’ last question has the form of a dilemma:
For if, on the one hand, [A] unity itself by itself (auto kath’ hauto to hen) is
adequately seen (horatai), or grasped (lambanetai) by some other sense, then
it will not be something that draws one towards being. But if, on the other
hand, [B] it is always the case (aei) that a thing (ti) is at the same time
(hama) seen with an opposition to this [i.e., to unity], so that it does not
appear to be one thing any more than the opposite (hōste mēden mallon hen
ē kai tounantion phainesthai), then something will already be required to
judge it, and in this case the soul will be compelled to puzzle over and to
enquire, by turning over within itself this concept [ennoia, i.e., the concept
of unity], and to pose the question, ‘What is this thing itself, unity?’ (tí pote
estin auto to hen); and thus the study of unity will be among the things that
lead one to, and turn one around towards, the contemplation of being.
(d–a; my translation)

 Plato’s Essentialism
Why does Socrates imply that what is at issue is ‘unity itself by itself’ (auto
kath’ hauto to hen, d–e); ‘this thing itself, unity’ (auto to hen, a);
the account of what this thing itself, unity, is (tí pote estin auto to hen,
a) and the very concept of, or thought of (ennoia, e), unity? And
why – which might appear quite strange – does he imply that this is what is
at issue, even if it is supposed, contrary to what they will in fact suppose,
that unity as it appears in sense-perception is not subject to compresence of
opposites (i.e., the first lemma, in which the phrase ‘unity itself by itself’,
auto kath’ hauto to hen, is used)?
The reason, it seems to me, is that it is intended that everyone,
irrespective of whether he or she believes in intelligible entities that are
independent of sense-perceptible things, or in Forms thus understood,
needs an account of what unity is, and needs a concept of unity based on
such an account. Everyone needs this, because everyone thinks that it is
possible to think of and make judgements about sense-perceptible things,
and Plato has just argued that: we speaking and thinking humans generally
suppose that the subject of predication in a judgement is subject to the
principle of non-contradiction; and the concept of unity is necessary for
adequately individuating the subject of predication so that the principle of
non-contradiction is applicable to it.
We may suppose, therefore, that what is at issue between the two sides
of Socrates’ dilemma is whether or not it is possible to use a sense-
perceptible thing of conspicuous unity – that is, conspicuous to sense-
perception, such as in a perfectly shaped rose or indeed a perfectly
shaped and manicured finger – as a paradeigma and standard of unity;
that is, as something by reference to which one can give an account of
what this thing itself, unity, is, and in this way specify the requisite
concept of unity. If, on the one hand (i.e., A), unity as it appears in
sense-perception is not subject to compresence of opposites, then,
Socrates intends, this may indeed be possible. In that case, evidently, the
thought of an intelligible unity that is independent of sense-perceptible
things will not be required for the purpose of thinking of and
making judgements about sense-perceptible things that are subject to the
principle of non-contradiction. If, on the other hand (i.e., B), unity as it
appears in sense-perception is subject to compresence of opposites, and
if it is so in the radical way in which Socrates’ formulation of the
second lemma implies (‘so that it [i.e., the thing perceived] does not
appear to be one thing any more than the opposite’), then,
Socrates intends, this is not possible. It is not possible, because, if we see
(or hear, or feel, etc.) a thing as no more one thing than indefinitely many

Judgements about sense-perceptible things 
things, then no sense-perceptible thing will be of conspicuous unity, and
hence no sense-perceptible thing can be used as a paradeigma and standard
of unity. In that case, therefore, it is arguable that the thought of an
intelligible unity that is independent of sense-perceptible things will be
required to think of and make judgements about sense-perceptible things,
at any rate if such judgements are to be subject to the principle of non-
contradiction.
The issue whether the question ‘What is F?’, understood as a request for
a paradeigma and standard of a thing’s being F, can be answered by appeal
to a particular thing that is conspicuously F, has been engaging Plato in
earlier dialogues. Thus in the Euthyphro, when Socrates dismisses
Euthyphro’s view that his, Euthyphro’s, particular action, the one he is
about to perform for all to see, can serve as a paradeigma of pious action in
general, and hence serve to answer the question ‘What is piety?’ (d–e);
or in the Hippias Major, when Socrates argues against Hippias’ view that
invoking a particular girl of conspicuous beauty can serve to answer the
question ‘What is beauty?’ (–). However, the way in which this
issue is taken up in the present argument is distinctive and pronounced in
two significant respects.
First, it is made quite clear (as, perhaps, it is not in the Euthyphro) that
the issue is whether the question ‘What is F?’ (here, ‘What is unity?’, tí pote
estin auto to hen; a) can be answered by appeal to a particular thing
that, as it appears in sense-perception, is conspicuously F (here, unitary).
Secondly, if, as Plato argues here, the answer is NO, then this answer
will, it seems, be of momentous consequence, for the particular reason
that, as it has already been argued (by c and before Socrates’ last
question), the application of unity to any thing, O, is a necessary condition
for the principle of non-contradiction to be applicable to O, and hence a
necessary condition for making a judgement about O that is subject to the
principle of non-contradiction – even such a simple judgement as ‘O is
hard’ (as opposed to the mere report that it feels hard). The consequence is
momentous, because it means that, even if it is supposed that such
questions as ‘What is hardness?’, ‘What is softness?’, ‘What is redness?’,
etc., can be answered by appealing to a thing that, to touch or sight, is
conspicuously hard, or soft, or red, these answers, and hence the

apeira to phēthos, as Glaucon will say when he speaks up for the second lemma; see a–.

See Chapters  and  of this study; see also Politis , ch. .

It should not be immediately assumed that this supposition is un-Platonic: in Parmenides d–,
Socrates says of some things that ‘they are, just as we see them to be’. Commenting on this passage,
Harte (, ) says: ‘Forms are not needed in those cases where things are “just as we see them

 Plato’s Essentialism
concepts of hard, soft and red specified in this way, even if they are
sufficient for reporting how things look (or feel, etc.) to one, will not be
sufficient for making any judgement involving the predication of these
concepts to a sense-perceptible thing, if this judgement is to be subject to
the principle of non-contradiction.
We can now see why, at the beginning of the argument (d–),
Plato implied that even in regard to the perception of finger, the soul of a
probing and inquisitive person, as opposed to that of the common person
(tōn pollōn), is compelled to enquire ‘What is a finger?’ The probing and
inquisitive person will, at any rate if her inquisitiveness is sufficiently
probing in (as we might say) a logico-metaphysical direction, recognise
that, to make any predication of an object, including such a predication as
‘This is a finger’, it is necessary adequately to individuate the subject of
predication, by applying to it the concept of unity, for the purpose of
ensuring that the principle of non-contradiction is applicable to it, and
that it is questionable whether this concept, viz. the concept of unity, can
be specified by pointing to a thing, such as a finger, that appears conspic-
uously unitary to the senses.
What does this imply, in regard to such a person’s particular judgement,
‘This is a finger?’, whenever she sees a finger? It would not be at all
plausible to suppose that, on every occasion in which she makes such a
judgement, she needs to pause to consider how adequately to individuate
the subject of predication by applying to it the concept of unity. What this
person will need to recognise is that, if her sensory perception reports ‘This
is a finger’, then she is entitled to judge ‘This is a finger’ and to suppose
that this judgement is subject to the principle of non-contradiction, only
if: either she actually pauses; or she is capable of pausing and prepared to
pause or someone, though not necessarily she, is capable of pausing and
prepared to pause, to consider how adequately to individuate the subject of
predication by applying to it the concept of unity.
.. Glaucon’s answer (d–a)

Glaucon answers Socrates’ question with the following bold statement:
The sight of one and the same thing is certainly more than anything subject
to this [i.e., to compresence of opposites in regard to unity]; for, we see one
to be”. What cases these are may be for us to discover.’ If these cases do not include those of
immediately perceptible qualities of individual sense modalities, such as the quality red, it is hard to
imagine what they could include.

Judgements about sense-perceptible things 
and the same thing at the same time both as one and as indefinite in number
(touto g’ echei ouch hēkista hē peri to auto opsis; hama gar tauton hōs hen te
horōmen kai hōs apeira to phēthos. (a–)
How does Glaucon arrive at this statement? The suggestion will not do,
that all he needs to do is extend the range of characteristics that are subject
to compresence of opposites, from those previously mentioned to num-
ber. If we suppose that Glaucon arrives at his statement in this way, then
we shall not be able to account for the fact that his statement is not that
‘We see one and the same thing at the same time both as one and as more
than one (e.g., two, or three or some determinate number)’ but is, rather:
‘We see one and the same thing at the same time both as one and as
indefinite in number’. The former statement might serve as a response to
the question ‘How many?’, for example, ‘How many things do you see
when you see a finger?’; that is, one might reply: ‘I see one thing, the whole
finger, and I also see three finger-parts (i.e., those divided by the two
joints)’. But it is hard to see how Glaucon’s statement could be an answer
to this question.
Glaucon’s statement is an emphatic assertion of the second lemma of
Socrates’ question. The part of his statement that says that ‘we see one and
the same thing at the same time both as one and as indefinite in number’
(kai hōs apeira to phēthos, a–) is Plato’s clear way of indicating in just
how radical a sense Socrates’ formulation of this lemma is intended, when
he said that what is seen ‘does not appear to be one thing any more than
the opposite’ (hōste mēden mallon hen ē kai tounantion phainesthai, e).
.. Unity (to hen) once again

We have seen that the concept of unity, or oneness (to hen), occupies a
central place in Plato’s argument in Republic VII. a–a. It is a good
question what exactly Plato means by to hen here. We have seen that the
function of this concept is to enable us to individuate objects, and in the
first instance the objects of our sensory perceptions, in such a way that it is
possible to make about them statements and judgements that conform to
the principle of non-contradiction. We have also seen that by to hen here,
Plato does not mean the number one, as this number is understood in

Contra Pappas (, ): ‘Why does mathematics suddenly enter the present argument? Because
numbers form a special case of opposable properties. They appear in particular things in the same
confusing way that other relative terms do: a may mean, for instance, that my hand is
simultaneously one (hand) and five (fingers).’ Pappas states clearly a view held by many critics.

 Plato’s Essentialism
arithmetic, whether everyday arithmetic or the arithmetic of the most
advanced mathematicians of his time. This is clear not least from the fact
that he begins the argument (see a–) by pointing out that the way in
which he will use arithmetic (hence, concepts from arithmetic) is entirely
new. It is also clear from how he characterises the function of this concept,
to hen, when (at b f.) he explicates that its function here is to
determine, in a sufficiently distinct and determinate way, whether the
object of perception, which is here the logical subject of a statement and
judgement, is one or two (hen ē duo).
As we know, this is not Plato’s last word on the concept of unity and
oneness: it is arguable that the entire second part of the Parmenides is an
analysis of this concept and the different ways it can, and perhaps must, be
understood. This is not the place to consider how the concept of unity in
the Republic argument is related to the concepts of unity distinguished in
the Parmenides. Suffice it to say that the concept of unity in the Republic
argument is not the concept of extreme absolute unity distinguished in the
First Hypothesis in the second part of the Parmenides.

 
Why does thinking of things require essences,

or Forms?
Parmenides
One thing about which critics are generally agreed in regard to the second
part of the Parmenides, notwithstanding many and profound disagree-
ments, is that a good reading of it should attempt to establish a credible
connection between it and the critical examination of, and dispute about,
Forms, from which it follows on and to which it is in some way a response.
This desideratum, I argue, is met only if we suppose that the second part is
as much about thinking as it is about being. I shall, therefore, examine
whether the second part of the dialogue considers not only being, but also
thinking and its relation to being. I argue that it does, in ways that are
continuous with certain crucially important elements in the first part that
are specifically about thinking and its relation to being. The elements
I have in mind are, first, the claim at b–c, which says that Forms,
and the defining of Forms, are necessary for thought; and, secondly, the
claim at b–c, which says that thought is of something, something
that is, and something that is one. I argue that the earlier claim signifi-
cantly prepares for the later claim and that both claims cry out for further
articulation and defence. I argue that Plato provides this in the second part
of the dialogue.
The critical examination of the theory of Forms, which makes up much
of the first part of the dialogue, does not conclude with the objections

The claim at b–c, that Forms are necessary for thought, has attracted much attention, as has
the claim, at b–c, that thinking is of something that is and is one. Especially since Burnyeat’s
 paper, this latter claim has been caught up in the debate over whether Plato’s Forms are mind-
independent – which is not an issue I want to consider here and I shall consider in Chapter .
However, if I am not mistaken, critics have not made the connection between the b–c claim and
the b–c claim; and they have not considered whether, in the second part of the dialogue, Plato
defends the claim, as made in the  passage and directly implied in the  passage, that Forms are
necessary for thought. M. L. Gill (, ) briefly considers this, or something like it, but she runs
together the question of Plato’s defence of the claim that Forms are necessary for thought with the
question of his defence of the claim that they are necessary for ‘the things themselves’.


 Plato’s Essentialism
against the theory; it concludes with the claim that, however difficult we
may expect it to be to answer one who, as a result of these objections, wants
to give up on Forms and their existence and knowability by us humans, we
have no choice but to answer him. This is because, it is claimed (b–c),
Forms and the defining of Forms are necessary for thought. Giving up on
thought is, apparently, not an option. If the exercise (gumnasia) of the
second part of the dialogue is only about being and not also about thinking,
it can contribute to answering one who wants to give up on Forms as a result
of the objections against them, only by ignoring the claim that Forms are
necessary for thought. To contribute to answering him in a way that does
not ignore this, the exercise has to be also about thinking. Due to this
particular juncture and crisis point of the argument in the dialogue, which,
we may expect, significantly contributes to motivating the introduction of
the extended exercise that follows, we may expect that, in the order of the
argument and enquiry, this exercise is, crucially, also about thinking.
It may be said that even if Forms are necessary for thought, and
therefore we have no choice but to defend their existence and knowability,
the claim that they are necessary for thought, or any reference to the
relation of Forms to thought, or in general to the relation of being to
thinking, need not be part of that defence. This may be a good point in
general, but it ignores the dialectical and argumentative juncture in the
dialogue at which it is claimed that Forms are necessary for thought. The
claim that Forms are necessary for thought cannot simply be accepted,
either in its own right, since it is utterly disputable, or by one who has
given up on Forms due to the objections raised against them. This sceptic
about Forms will want to see a defence of this claim, and will want the
claim and its defence to provide an important reason, even if it is not the
only reason, for believing in Forms. I do not see that we, the readers, are
expected to be significantly differently disposed than this sceptic.
. The relation between thinking and being, and its relevance
for the dispute about Forms, in the first part of the dialogue
Following a series of objections against the supposition that there are
Forms, and against the theory of Forms that Socrates introduced at the

Contra Allen (), who, remarkably, has nothing to say about the claim, at b–c, that Forms
are necessary for thought, or about the role of this claim in the overall argument and dialectic in the
dialogue so far.

I shall not consider here whether there are further arguments for Forms in the second part, not based
on Forms being necessary for thought and in general on the relation of thinking and being.

Forms as requirements of thought 
start of the dialogue in response to Zeno’s claim that the supposition that
things are many leads to absurdities, Parmenides spells out, twice, that
these objections amount to credible reasons for doubting that there are
Forms and that they can be known by us humans. He insists that they
provide so credible reasons for doubting this, that to persuade one that
they can be answered, both the respondent and the sceptic must be
outstandingly gifted and be ready to conduct and to follow an extended
investigation that goes well beyond the immediate objective at hand. We
may wonder whether the reference (at b–) to such an investigation
may not be to the kind of investigation conducted in the second part of the
dialogue, or even to that very investigation. If we ask why he must be
answered who, as a result of the objections raised against them, doubts that
there are Forms and that they can be known by us humans, the reason is
immediately pointed out, when Parmenides goes on to say that Forms and
the defining of Forms are necessary for thought (b–c).
The question of the relation of thinking and being has already been
posed, earlier in the dialogue, when Parmenides anticipated, and closed
off, the danger that thinking should become isolated from being (b–
c). This danger is considered, and closed off, when, in quick succession,
Parmenides states that it is not possible that a thought should be of
nothing (oudenos); it is necessary that a thought should be of something
(tinos), and of something that is ([tinos] ontos), and of something that is
one (henos tinos) and involves some one quality (mian tina idean). A Form
(eidos), he says, is that which is thought of as being one thing, always the
same in all cases: touto to nooumenon hen einai, aei on to auto epi pasin.
What Parmenides says here is a response to Socrates’ proposal, which is to
safeguard the unity of each Form by supposing that each Form is a
unitary thought of a plurality of things. Parmenides’ response is to say
that this is not a possible way of safeguarding the unity of each of
those beings, Forms, because, on the contrary, the unity of a thought
depends (either exclusively or in important part; the passage appears to
leave this open) on the unity of the beings of which it is the thought – the
objects of thought.

b–c and e–b. The later passage refers back to the earlier passage, and the two
passages need to be read together.

I intend to steer clear, at this point, of the issue whether Plato’s argument here is intended to show
that Forms are mind-independent. Burnyeat () notoriously argues that it is so intended. Allen
(, ) strikes a cautionary note: ‘Nothing in the mere observation that thought is of something
which is and is one over many requires the conclusion that what is thought exists independently of
the thinking of it.’ I take up this issue in the final chapter.

 Plato’s Essentialism
It is a good question why, for Plato, Forms and the defining of Forms
are necessary for thought. I argue that we need to consider two very
different answers. On one answer, and one that is familiar from the
literature, it is because each thought is the thought of a Form, such that
to each different thought there corresponds a different Form. On a very
different answer, which is not so easy to find in the literature, it is because,
since each thought is the thought of one thing, it involves the thought of,
in the first instance, the Form of oneness or unity. I shall defer this issue till
the end of the chapter, when I argue for the latter answer.
It seems to me that Parmenides’ statement, at b–c, that a thought
is of something that is and is one, and that Forms provide for the unity of
what a thought is of, is picked up on and made use of in the all-important
claim, at b–c. This says that admitting that there are Forms (see easei
eidē tōn ontōn einai) and engaging in defining some Form for each one
thing (see ti horieitai eidos henos hekastou), is necessary for being able to
think of things (see oude hopoi trepsei tēn dianoia eksei). As in the 
passage, a Form is characterised here as a quality (idea) that is always one
and the same.
It is important to note that the claim here, in the  passage, is not
simply that Forms are necessary for thought, but that Forms, and the
defining of Forms, are necessary for thought. This indicates in what way
Forms are necessary for thought, namely, by the thinker, in thinking of
something, being engaged in defining, or attempting to define, a Form or
some particular Form or Forms. I shall come back to this all-
important point.
The passage at b–c moves from saying that Forms and the
defining of Forms are necessary for dianoia to saying that they are neces-
sary for tēn tou dialegesthai dunamis and for philosophia. This may suggest

The view that says that, according to Plato, certain words or concepts signify Forms and would be
meaningless if they did not signify Forms has been traditional for the past hundred plus years, both
in its own right and as an account of why, in the Parmenides passage (b–c), Plato claims that
Forms are necessary for thought and speech. It is true that this view is not as common as it used to
be. Recently Crivelli (, –) rejected it out of hand, and Malcolm (, ch. ) attacked it
thirty years ago, as did Gosling (, ) half a century ago. This does not mean it is about to go
away. For a recent statement of the view, see Rickless (, ), who, commenting on Parmenides
b–c, says this: ‘names signify forms (or their natures), and thus would have no meaning if forms
did not exist’. Like other defenders of this view, Rickless appeals to the Cratylus in its defence, as do
Kutschera (, ) and Hestir (, ch. ). Going back at least to Cornford (, –),
critics that defend the semantic view include Ryle ( [originally ], ); Keyt (, );
Teloh (, , ); Bostock (, ); Bestor (, ). Several of these critics defend the
view both in its own right and as an account of why, in the Parmenides passage, Plato claims that
Forms are necessary for thought and speech.

Forms as requirements of thought 
that Plato is concerned here not with thought in general, but only with a
highly advanced kind of thought, namely, dialectic and philosophy. But
the suggestion should be resisted. The passage serves to motivate the idea
that we have no choice but to answer one who wants to give up on Forms
as a result of the objections raised against them. The passage can perform
this function, only if it says that Forms and the defining Forms are
necessary for thought in general. For, if one is prepared to give up on
Forms as a result of these objections, one is no less prepared to give up on
any thought that requires Forms; unless, that is, this is a kind of thought
that one is not willing to give up or is not capable of giving up on,
irrespective of one’s views (if one has any) on the existence and knowability
of Forms. The phrase ‘you would not have anywhere to turn your
thought’ (oude hopoi trepsei tēn dianoia eksei) confirms that thought in
general is in question. So does the term tēn tou dialegesthai dunamis, if we
suppose that dunamis here means ‘ability’, and that, while dialegesthai can
develop into a highly advanced kind of thought and, in particular, dialectic
and philosophy, the ability for dialegesthai is rooted in thought in general
and its practice begins with thought in general. On this reading, the
passage is, as we would expect it to be, both about thought in general
and about the most advanced forms of thought such as dialectic
and philosophy.
It is remarkable that Plato does not offer any justification for the claim
that Forms and the defining of Forms are necessary for thought, in making
it here (b–c); he simply invokes it to indicate that we have no choice
but to answer one who wants to give up on Forms as a result of the
objections raised against them. Parmenides’ statement in the earlier passage
(b–c) indicates an important justification for the  claim; it is,
from what I can see, the only justification that Plato provides so far and
before the second part of the dialogue. For suppose we could persuade one
who wants to give up on Forms, as a result of the objections raised against
them, of the claim, asserted in the  passage, that a thought is of some
one existing thing, and that Forms, or some particular Form or Forms, are
necessary to provide for the unity of what a thought is of. Evidently, we
would have gone a very considerable way towards demonstrating that there
is no choice but to defend some notion of Forms.
At the same time, we may note that all Parmenides has done in the
earlier passage, b–c, is assert this claim: that a thought is of some one
existing thing, and that Forms, or some particular Form or Forms, are
necessary to provide for the unity of what a thought is of. To answer one
who gives up on Forms as a result of the objections raised against them,

 Plato’s Essentialism
this will have to be argued, not simply asserted. We may expect the second
part of the dialogue to contribute to this – we will find that it does.
. Forms here, in the first part of the dialogue, are

basically essences
What, in the first part of the dialogue, are Forms supposed to be? This is a
most important question for understanding the Parmenides. It is not
adequate, it seems to me, to suppose that a Form is, purely and simply,
an entity characterised in the way in which, first, Socrates characterises
such entities when, in response to Zeno’s claim that the supposition that
things are many leads to absurdities, he introduces the supposition of
Forms, and, next, Parmenides further characterises these entities in and
through the series of objections that he goes on to level against Socrates’
supposition. The problem with such a characterisation of what Forms are
supposed to be here is that it ignores that Parmenides both subjects the
theory of Forms to serious objections and concludes that, this notwith-
standing, we cannot give up on Forms, because they are necessary for
thought. We need to determine, therefore, what it is that we cannot give
up on when Parmenides concludes that we cannot give up on Forms. But,
if we characterise Forms as simply those entities that have all the charac-
teristics that Socrates attributes to them in introducing them and
Parmenides further attributes to them in subjecting them to the objec-
tions, we have no way of telling which of these characteristics it is that we
cannot give up on when Parmenides concludes that we cannot give up on
Forms because they are necessary for thought. We cannot simply assume
that he intends all the characteristics to survive the objections. This would
be to assume that no revision of the theory of Forms might be envisaged
through the objections. These are questionable assumptions. Far better, it
seems to me, is to ask whether there is a core notion of Forms in the first
part of the dialogue, such that, when Parmenides concludes that Forms are
indispensable because they are necessary for thought, this core notion is
what is supposed to survive the objections, in the first instance and
whatever about other characteristics of Forms that may seem to go beyond
this core notion.
When Socrates first introduces the Form of likeness and unlikeness, he
glosses to eidos tēs homoiotētos/anhomoiotētos in terms of ho estin homoion/
anhomoion, ‘that which like/unlike is’ (e–a); and, likewise, when
he introduces the Form of oneness and plurality, he glosses this Form in
terms of ho estin hen/polla, ‘that which one/many is’ (b–). This

Forms as requirements of thought 
suggests that, however he and Parmenides are going to further characterise
Forms, what they begin by supposing is that Forms are essences: essences
in the sense of that which a thing or quality is.
This, remarkably, is also what Parmenides ends by supposing, when he
concludes that Forms are necessary for thought (b–c). The claim is that
Forms, and the defining (horizein, horizesthai) of Forms, is necessary for
thought. Quite generally, what is defined is what a thing or quality is, its
essence. This is true irrespective of whether we take horizein/horizesthai in
a strict sense of providing a definition, in the sense of an account that has
to satisfy stringent requirements of adequacy or, instead, we take it in a less
strict sense, which we might translate as ‘to mark off’ or ‘to delimit’.
Whichever of the two ways we go, the object of this human intellectual
activity, the attempt to horizein something, is that which a thing or quality
is, whether in a strict or in a less strict sense of this term, hence whether in
a strict or in looser sense of the term ‘essence’.
There is, I think, one more passage in the first part of the dialogue in
which, through Socrates, Plato indicates what Forms must, basically, be
supposed to be. It is at b–d, when Parmenides interrogates Socrates
about what things or qualities there are Forms of. In his answer, there is
much that Socrates is hesitant and diffident about, but he is adamant that
two things are beyond dispute if one believes in Forms: that there is a
Form of oneness and a Form of likeness (b–) and likewise a Form of
justice, beauty and goodness (b–); and, most important, that there
are no Forms for things that are just as they appear to our senses to be
(d–). This last claim implies that Forms, as understood here, cannot
be perceived by the senses.
We shall have to see whether the notion of a Form survives into the
second part of the dialogue, and, if it does, whether it is this core notion of
a Form, used to signify that which a thing or quality is, and such that this
essence cannot be perceived by the senses, that survives.
. The transition to the second part of the dialogue

When Parmenides first proposes that Socrates needs further intellectual
exercise if he is not to miss out on the truth (he says three times that the
exercise is necessary for arriving at the truth: d, c, ee), he
motivates this proposal by pointing to what he thinks is Socrates’ prema-
ture attempt to define each of the Forms (horizesthai epicheireis . . . hen
hekastōn tōn eidōn, c–d). This picks up on the immediately preced-
ing passage, which says that Forms, and the defining of Forms, are necessary

 Plato’s Essentialism
for thought. Parmenides says (d–a) that he was impressed and
delighted that Socrates, in the original discussion with Zeno, distinguished
between objects of the senses and ‘those things that one might especially
grasp by reasoning and might consider to be Forms’ (ekeina ha malista tis
an logō[i] laboi kai eidē an hēgēsaito einai, e–). The reference is to
Socrates’ original introduction, early in the dialogue, of this distinction,
which he made for the purpose of arguing that the objects of the senses can
without absurdity be both like and unlike, and in general be subject to
contrary qualities, provided that, first, there are objects of thought, and
Forms in particular, and they cannot without absurdity be subject to
contrary qualities in the same way; and, secondly, the objects of the
senses have the qualities they have as a result of partaking of Forms. In
Socrates’ original argument (e–a), there is no particular indica-
tion that the argument is concerned also with thinking and not only with
being. But the argument in the dialogue has moved on since Socrates’
original contribution, and we have seen that, through Parmenides and on
two occasions (at b–c and at b–c), the question has been
posed of the relation between being and thinking.
Parmenides goes on to say that, if one wants to become more properly
trained in this exercise (ei boulei mallon gumnasthēnai), it is necessary to
investigate not only the consequences (ta sumbainonta) of the supposition
that a thing, whatever it may be, is (ei estin hekaston hupotithemenon), but

Rickless () argues that the claim here is that the Form of oneness cannot in any way be many;
indeed, that quite generally Plato’s Forms, before the second part of the Parmenides, cannot in any
way be many. Along similar lines, see M. L. Gill: ‘I shall argue that Part II [of the Parmenides] is an
indirect argument demonstrating that to save the theory of forms and philosophy Socrates must
abandon his thesis about the one and admit that it is both one and many (in different ways)’ (,
–, original emphasis). This, it seems to me, cannot be right. When Socrates introduces ‘his
thesis’ early in the dialogue, what he says is that that which one is (ho estin hen) cannot be many
(b–), not simply that the one cannot be many; and, in the second part of the dialogue, he says
the very same about the one that is understood in this way, that is, as‘ the one by itself’ (a) and
‘the one itself’ (b) and ‘the truly one’ (c), and is set against the one in combination with
being (also referred to as ‘the one being’, to hen on, see, e.g., a). What he says is that, while the
one in combination with being is also many, the one by itself (or ‘itself’, or ‘the truly one’, or ‘that
which one is’) is not also many. What this shows is two things. First, the claim is that the Form of
F cannot be both F and contrary-to-F, hence the Form of one cannot be both one and many; it is
not, contra Rickless, that Forms cannot in any way be many (for a similar criticism of Rickless, see
Sayre ). Secondly, and contra Rickless and Gill, there is not an inconsistency or contradiction
between the first and the second part of the dialogue on the question whether the one, itself by itself,
can be both one and many: the first part of the dialogue denies this and so does the second part.
What the second part does is explain, or begin to explain, how the one can, in relation to other
things, hence not itself by itself, be both one and many, an explanation that is taken further in the
Sophist. Remarkably, the claim that Forms are one by themselves but many in relation to physical
bodies, and indeed to other Forms, is made in so many words in Republic V. a–.

Forms as requirements of thought 
also the consequences of the supposition that the same thing is not (ei mē
esti to auto touto; e–a). This anticipates the dilemmatic structure
of the argument in the second part of the dialogue. It also picks up on
Socrates’ original supposition, to which Parmenides has just referred, that
there are Forms, and Socrates’ original investigation of some important
consequences of this supposition. Why should Parmenides, at this point of
the argument in the dialogue, require that one also investigate the conse-
quences of the supposition that there are not Forms? Evidently, because it
has become apparent, in the meanwhile and as a result of the objections
raised against them, that we cannot simply assume that there are Forms; on
the contrary, there are serious reasons against this supposition. At the same
time, Parmenides has concluded by asserting a most unattractive conse-
quence of the supposition that there are not Forms: thinking will not
be possible.
A dilemmatic mode of investigating the consequences of the supposition
that a certain thing is, and likewise of the supposition that it is not, has
been in play from the beginning of the dialogue. Especially worth noting
is that each step in the dilemmatic investigation links up, logically and
dialectically, with the previous one: Zeno investigated the consequences of
the supposition that only the many (ta polla) are, in order to demonstrate
that they are not less absurd than are commonly thought to be the
consequences of Parmenides’ supposition that only the one (to hen) is;
Socrates investigated the supposition that Forms are, for the purpose of
demonstrating that the supposition that the many are does not lead to
absurd consequences; Parmenides investigated the supposition that Forms
are, to demonstrate that it leads to serious difficulties; and he concluded by
adding that the supposition that there are no Forms leads to even more
serious difficulties. It is, therefore, a good question: Does the rigorously
dilemmatic procedure, which will emerge as distinctive of the second part
of the dialogue, likewise link up, logically and dialectically, with what
directly precedes it?
Parmenides goes on to explicate, in response to Socrates’ request for
clarification, what is involved in investigating the consequences both of the
supposition that a thing is and the supposition that it is not (a–c).
He mentions a number of examples of suppositions in regard to which this
procedure can be conducted: that the many are and that the one is, that
likeness is and that unlikeness is, that change is and that rest is, that

See the recent work of Evan Rodriguez (; ). He is working further on this, with a number
of publications forthcoming.

 Plato’s Essentialism
coming into being and passing away are and that being and not being are.
He says that for each such thing, which is supposed to be or not to be, we
must consider the consequences both for the thing in relation to itself (pros
hauto) and for the thing in relation to the things other than it (pros ta alla).
This distinction is not easy to understand; we may doubt it can be
properly understood in advance of a careful study of the second part of the
dialogue. But it appears to pick up on the last objection against the theory
of Forms, which pointed to the risk of Forms becoming wholly dissociated
from other things, and in particular the things around us (ta par’ hēmin, see
c–d), and likewise those other things becoming wholly dissociated
from them, which in turn led to the conclusion that Forms are unknow-
able to us. This radical dissociation was expressed in the language of the
Forms being what they are pros hautas and not pros ta par’ hēmin, and,
conversely, too, the things other than Forms being pros hauta and not pros
ta eidē (c–d; repeated at d–).
Save for Parmenides’ choice to conduct this procedure in regard to his
own supposition regarding the one itself (tēs emautou hupotheseōs, peri tou
henos autou hupothemenos, b), this concludes the substance of this
transitional passage, in preparation of the undertaking of the long and
ambitious exercise to follow. This dilemmatic procedure, of investigating
the consequences of the supposition that a thing is, or that it is not, can be
conducted in regard to all kinds of things. However, we may suppose that,
at the present juncture and crisis point of the argument in the dialogue, the
procedure is related, most especially, to suppositions about Forms and the
relevance of Forms for thinking.
I think Parmenides’ choice to conduct this procedure in regard to his
own supposition regarding ‘the one itself’ supports this. The question is
why, at this juncture of the dialogue, Parmenides makes this choice. It is
not an adequate answer, it seems to me, to say that Plato has in mind the
historical Parmenides, whose monist views he has already mentioned at the
opening of the dialogue. This answer ignores the argument in the dia-
logue and the juncture and crisis point it has arrived at. We may propose,
rather, that Parmenides’ choice, to conduct the exercise in regard to the

Contra Brisson (). In view of what I have been arguing, it will not be a surprise that I find no
more credible Brisson’s view that the one of the second part of the dialogue signifies, exclusively or
primarily, the unitary cosmos and the all. While the cosmos, and the all, may be one theme of the
second part, it is surely not the only or the primary one. If, as Brisson does, we want to trace
connections between Plato’s dialogue and the poem of Parmenides, a reference that Brisson
remarkably overlooks is to fragment , in which Parmenides marks an unbreakable unity
between thought and being.

Forms as requirements of thought 
supposition that the one is, is, in significant part, motivated by the present
juncture of the argument in the dialogue. This means that the reference to
‘the one itself’ (b), in which the exercise will take its starting point,
may be expected to include a reference to the Form, or essence, of oneness
or unity, and to this Form conceived, in the first instance, as an object of
thought and as providing a necessary condition for thought.
. Being, thinking, and the Form of oneness in the second

part of the dialogue
I turn to the second part of the dialogue, and the question whether it is
linked to that which precedes it and, in particular, to the claim that Forms
and the defining of Forms, including, most especially the Form of oneness
or unity, are necessary for thought. I think we have reason to surmise that
this question should be answered in the affirmative, if only we take a
summary look at the First Supposition and the Second Supposition
(hupothesis; I use Supposition for the hupotheseis in the second part), which
are directly linked with each other and which set off the second part of the
dialogue. The First Supposition concludes with the claim that, if the one
(to hen, understood as the primal source of all unity and unitary things) is
understood in a certain way, then it is not a possible object of naming,
speaking, judging, perceiving, giving an account (a–); in short, it is
totally outside thought. The Second Supposition concludes with the claim
that, if the one is understood in a certain different way, then it is manifestly
an object of knowledge, judgement, perception, naming and having an
account (d–e) – the very things the ending of the first investigation
denied of the one.
Even setting aside the question whether the one here is the Form of the
one, and even setting aside the question what the relation is between the
one and thinking, we could hardly have wished for a more emphatic

It will be evident, to the reader who knows Paul Natorp‘s  classic, Platos Ideenlehre (translated
with introduction in Natorp ), that my account of Plato’s account of the relation between
thought and being in the second part of Parmenides is much indebted to Natorp’s chapter on the
Parmenides in that monumental work – and this even setting entirely to one side Natorp’s overall
view (which I do not share; see Chapter ) that, for Plato, thinking is prior to being. Years after my
having worked on Natorp, David Horan, in a paper that he published some time later (),
demonstrated to me that, on this question at any rate – the question of how the thought of the one
is necessary for, and is involved in, all thought – Plotinus is an astute and penetrating interpreter
of Plato.

I do not mean to imply that, if the one is understood in different ways in the two Suppositions, two
ones are in question.

 Plato’s Essentialism
indication that Plato intends the second part of the dialogue to pick up on
the claim, from the first part and b–c in particular, that we cannot
consider a noēma without immediately considering its proper nooumenon,
indeed, without considering the question whether and if so how its proper
nooumenon is a unitary object, a hen.
The First Supposition (c) says of the one that it is one: [to hen] hen
estin. The examples Parmenides gave earlier (a–c) of such supposi-
tions said of a thing that it is. It is surprising, therefore, that he should here
consider not whether the one is, but whether it is one. That the statement
hen estin is best understood to mean ‘[the one] is one’ is confirmed by what
is immediately inferred from it, namely, that the one is not many; and it is
suggested by the fact that this is how Parmenides has just stated his own
supposition (at b–). We may suggest that the supposition here is also
that the one is; but, because this first investigation investigates the conse-
quences of the one if the one is supposed to be only in relation to itself
(pros hauto, see a f.), the supposition that the one is is taken to be
tantamount to the supposition that the one is one.
The first consequence infers from the supposition that the one is one,
that all it is is one and that it does not have any kind of multiplicity (see ei
hen esti, allo ti ouk an eiē polla to hen, c–). From this crucial, initial
consequence a host of other consequences are inferred. They are typically
of the form ‘the one is neither F nor contrary-to-F’, where F is a quality
that, in one way or another, presupposes that that which has it is or implies
a multiplicity; such as, most especially, if F is the quality of being a whole
and contrary-to-F is the quality of having parts.
Practically all the consequences of this first investigation make no
reference to thinking, but the last few do, when it is inferred that it is
not possible for the one to be named, or spoken of, or judged, or known,
or perceived or given an account (a). This conclusion, it appears, serves
to identify and defend a necessary condition for thought, namely, that
(The First General Condition for Thought):
Thought is of something that is.
For it is from the consequence that the one is not something that is, and
hence that it is not a being that is one (e–), that it is inferred that it
is not possible for something to belong to the one (a–); from which,

I shall not consider why, or because of what (hidden) supposition, this might be thought to be a
plausible inference. Perhaps this is because of the (hidden) supposition that what a thing is, it is,
purely and simply, in relation to itself.

Forms as requirements of thought 
in turn, it is inferred that it, the one, is not a possible object of naming,
speaking, judging, perceiving, giving an account (a–).
It is notable that not only modes of thought are mentioned (judging,
knowing, giving an account), and likewise of speech (naming, speaking
of ), but also perception, aisthēsis. We may suggest that Plato here considers
aisthēsis as a thought-involving kind of sensory perception, and this is why
aisthēsis is grouped together with modes of thought, or thought and
speech. When I speak of thought in what follows, I include aisthēsis,
understood as a thought-involving kind of sensory perception. We shall
see that, towards the end of the dialogue, Plato uses not aisthanesthai, but
phainesthai for a sensory perception that is not thought-involving.
This first consequence, from which all the other consequences are
inferred, is based on the first investigation of the consequences for the
one, on the supposition that it is only in relation to itself and not in
relation to other things. It is apparent, therefore, that the conclusion that it
is not possible for the one to be named, or spoken of, or judged, or known,
or given an account or perceived, serves to identify and defend a further
necessary condition for thought, namely, that (The Second General
Condition for Thought):
Thought requires complexity in the object of thought;
and such complexity requires at least two things standing in relation to each
other, namely, the relation ‘X belongs to Y’.
It is because the one, if supposed to be only by itself and only in relation to
itself, does not meet this condition that it is not even an object of thought.
The character of this complexity is indicated at a–, when it is said
that it is not possible for something (ti) to belong to what is not; and the
unthinkability of the one is based on this statement. This shows that
the complexity in the object of thought is here characterised as being of
the form: X belongs to Y (I use ‘X belongs to Y’ for Plato’s ‘X esti tō[i] Y’;
he has also been using, and will go on to use consistently, ‘Y partakes of
X’ for ‘X belongs to Y’).
It is remarkable that this passage (a–) picks up on what
Parmenides had already said in the first part of the dialogue, when, at
b–c, he said that it is not possible that a thought should be of
nothing (oudenos), rather, a thought is necessarily of something (tinos) and

The claim that the one is not even a being is inferred from the claim that it does not partake of time
in any way (e–). Some critics think this inference is clearly fallacious, and intended as
fallacious. I don’t think it is clearly fallacious, and there is no indication that it is intended
as fallacious.

 Plato’s Essentialism
of something that is ([tinos] ontos). However, the present passage signifi-
cantly spells out what was already said there. In particular, it spells out why
we can validly move from ‘Thinking is of something’ to ‘Thinking is of
something that is’. This is a legitimate move, because thinking of some-
thing, X, implies thinking of two things, X and Y, and thinking that
Y belongs to X; and because it is possible for Y to belong to X, only if X is
(see ho de mē esti, toutō[i] tō[i] mē onti eiē an ti auto[i] ē autou;, ‘Would
something belong to, or be of, that which is not, the very thing that is
not?’, a–). I note that whereas originally (at b–c) Parmenides
had added that thought is necessarily of some one thing that is, this point is
not yet touched on in the second part of the dialogue. It will, we shall see,
be taken up later.
It seems to me that this preoccupation with thinking, however brief, is
crucial for understanding this argumentative juncture of the dialogue, in
particular, for understanding what the difference is between the First
Supposition (c) and the Second Supposition (b), and why
Plato moves from the one to the other. In the two lines in which the
transition is made, Parmenides proposes that they take up the First
Supposition again from the beginning (epi tēn hupothesin palin ex archēs
epanelthōmen), to see whether in going over it again (epaniousin) they arrive
at a different result (b–). Why do they need to start again from the
beginning? And how can they arrive at a different result, if the starting
point is the same?
The reason why they need to start again from the beginning is, it
appears, that there is no point of supposing that a thing is, if, as has just
emerged, it can be demonstrated that that which is supposed to be is not a
possible object of thought. That this is the reason why they need to start
again from the beginning is apparent not only in view of the character of
this particular juncture of the argument, where it has emerged that the one
is not a possible object of thought, but also in view of the ending of the
second investigation, which says that this investigation has demonstrated
that the one is manifestly an object of knowledge, judgement, perception,
naming and having an account (d–e) – the very things the ending of
the first investigation denied of the one.

It is familiar, especially since Dodds’ classic  paper, that Plotinus read the ending of the
investigation of the First Supposition as positively introducing the idea of an absolute One that is
beyond thought. It is hard to assess this reading, since demonstrating of something that it is not a
possible object of thought falls short of demonstrating a strict absurdity, or contradiction, of
supposing that there is, or could be, such a thing (see Castagnoli ).

Forms as requirements of thought 
How can the same supposition give rise to such contrary consequences,
including that the one is not an object of thought and that it is an object of
thought? We saw that, in the first investigation of it (c–a), the one
was supposed to be only in relation to itself and not in relation to other
things; it was this that led to its not being an object of thought. In the
second investigation, on the other hand (b–e), the one is not
supposed to be only in relation to itself, rather, it is supposed to be also
in relation to other things and, first of all, to being. Shortly into the second
investigation, it is pointed out that ‘if, in thinking, we take this [i.e., the
one] only by itself and without that which we say it partakes of, it will
emerge as being only one and not also many?’ (ean auto tē[i] dianoia[i]
monon kath’ hauto labōmen aneu toutou hou phamen metechein, ara ge hen
monon phanēsetai ē kai polla to auto touto;, a–). This serves to identify
what happened to the one in the first investigation, and why it happened;
except that, whereas in the first investigation it was supposed that the one
is only by itself and is not in relation to other things, here we are only
considering in thought the one by itself and not in relation to other things.
It seems a reasonable conclusion to draw, that what motivates the move
from the first investigation, in which the one is supposed to be only in
relation to itself, to the second investigation, in which it is supposed to be
in relation to other things, and first of all to being, is the fact that, on the
first way of taking the one, it turned out not to be a possible object of
thought (a), whereas on the second way of taking it, it has turned out
manifestly to be an object of thought (d–e).
Where in the second part of the dialogue, if anywhere, is it argued that
the one, and, in particular, the one conceived as a Form, is necessary for
thought? So far, thought has been mentioned at three places: the end of the
first investigation (a); the mention, at a–, of the one taken only
by itself in thought (tē[i] dianoia[i]) and the end of the second investigation
(d–e). The next time thought is mentioned is during the fourth
investigation, which is addressed to the question of what happens to the
things other than the one, if it is supposed that the one is (b–b).
The reference to thinking is at c–d, when it is argued that, if we were
to take away in thought (tē[i] dianoia[i], c) the supposition that a
whole and any of its parts, or the parts of those parts, partake in the one,
we will be left not with determinate or limited (peras echon) wholes and
parts, but simply with indeterminate multiplicity (apeiria, apeiria to
plēthos). This points to the opposite of what happened earlier, when it
was supposed that the one is only by itself and not in relation to other
things: here it is the things other than the one that are supposed to be only

 Plato’s Essentialism
by themselves and not in relation to the one, whereas there it was the one
that was supposed to be only by itself and not in relation to other things.
The question is whether Plato thinks the argument is parallel in regard
to the possibility of thought being of something: Does he think that, just as
the one, if supposed to be only by itself, is not a possible object of thought,
so too the things other than the one, if supposed to be only by themselves,
are not a possible object of thought? It is the question whether Plato
anywhere in the second part of the dialogue argues that that which is
simply indeterminate multiplicity is not a possible object of thought.
Originally and in the first part of the dialogue (b–c), Parmenides
had claimed that thought must be not only of something and of something
that is, but also of something that is one. It follows that that which is
simply indeterminate multiplicity, since it does not partake of the one and
is not one in any way, is not a possible object of thought. The question is
whether here, in the second part of the dialogue, this claim is defended, or
spelled out further, which originally was only asserted.
The last two investigations in the Parmenides consider the consequences
of the supposition that the one is not: the penultimate investigation
considers the consequences for the many in relation to themselves
(b–e); the ultimate investigation considers the consequences
for the many in relation to the one (e–c). Even on a first
acquaintance with these two final investigations, one cannot fail to be
struck by the impression that they are just as much about thinking as
about being, and they are concerned with the consequences for thinking of
the supposition that the one is not. In the absence of the one, the beings in
these investigations are characterised as, purely and simply, relational
multiplicities making up indeterminate magnitudes. The lesson in regard
to thinking, spelled out especially in the penultimate investigation, is that,
if the one is not, then thinking will be ‘like a dream during sleep’ (hōsper
onar en hupnō[i], d); the kind of dream – more like a nightmare – in
which any time one tries to focus one’s attention on one such multiplicity,
or some one aspect or relation of it, the object of thought will appear to
change beyond recognition and break up in front of one’s eyes. That
Plato’s concern is with the consequences for thinking is clear from the
reference, twice, to what happens in this scenario if one tries to grasp ‘in
thought’ (tē[i] dianoia[i]) something as being some one thing. It is likewise
spelled out what will happen to thought if, rather than trying to focus on
some one thing in this scenario, one simply attends to the radical indeter-
minacy that is characteristic, in this scenario, of being: the object of
thought will amount to a thoroughgoing illusion (phantasma, a; also

Forms as requirements of thought 
a), that is, the kind of thing that, any time it appears or seems to have
a certain quality, it merely appears and seems so, without really being so.
The penultimate investigation, in which are spelled out these conse-
quences, for thinking, of the supposition that the one is not, comes close to
saying that, if the one is not, thinking will not be possible. However, Plato
stops short of making this statement: he does not say, here or in the
ultimate investigation, that if the one is not, thinking will not be possible.
What he does, in the ultimate investigation, is spell out, yet again and even
more pointedly, just how exceedingly limited is the thinking (if this is the
right word for it) that, according to the penultimate investigation, is
reduced to momentary appearing and seeming:
And indeed, they [i.e., the objects of thought in this scenario] are neither the
same nor different, neither in contact nor separate, nor anything else that they
appeared to be in the argument we went through before [i.e., the penultimate
argument they have just been through]. The others neither are nor appear to be
any of those things, if one is not. (b–; trans. Gill & Ryan)
It is hard to think of a more suitable statement for a philosopher to make,
in trying to make manifest that thinking is not possible under a certain
supposition, even if it appears that the philosopher does not want to
conclude with the statement ‘Under this supposition and in this scenario,
thinking is not possible’. I am inclined to think that, if Plato does not
conclude with this statement, it is not because he thinks that a last refuge
might have been afforded one who thinks that it is possible to think even
without the one. Rather, it is because he is sensitive to the risk that such a
statement, if intended as the conclusion of a strict demonstration and as
going further than summing up just how etiolated and reduced thinking (if
this is the right name for it) in this scenario has been shown to be, will
appear to be the conclusion of an incoherent task: to demonstrate that one
cannot think without the one, by asking one to do just that: to think
without the one.
We are near the end of our task, of demonstrating that a major aim of
the second part of the Parmenides is to make good the claim, asserted but
not defended in the first part, that Forms and the defining of Forms, and,
most especially the Form of oneness or unity, are necessary for thought. It
remains to consider whether, when Plato argues here that the one is
necessary for thought, and that the one that is necessary for thought is
the one that is supposed to stand in a necessary connection to being and, as
a consequence, to a host of other things, it is the Form of the one he has
in mind.

 Plato’s Essentialism
In considering this, we need to recall how Forms are understood in the
first part of the dialogue. For we saw that the basic notion of a Form, in the
first part of the dialogue, is the notion of that which a thing or quality is,
and such that this essence cannot be perceived by the senses. We must ask
whether the one that, in the second part of the dialogue, is argued to be
and to be necessary for thought, is understood in this way.
We need to be clear that the one we are concerned with here is not the
one as understood in the First Supposition, which is understood to be
isolated from everything else and not to enter into any relations with
anything else. That understanding of the one has already been set aside,
since it precludes the one from being a possible object of thought.
Likewise, we are not concerned with the one that has parts and is a single
unitary whole of parts. For this one, which is introduced in the Second
Supposition, is the result of a more original one in combination with
being, and it is indeed referred to as ‘the one being’, to hen on (see, e.g.,
a). What we are concerned with is this more original one.
As far as I can see, Plato does not have a single consistent way here, in
the second part of the dialogue, of referring to this original one, which is at
the same time related to being; indeed, its being related to being and
partaking of being is necessary for its being an object of thought. If this one
is the essence of the quality one, we would expect him to refer to it as to
hen kath’ hauto (‘the one by itself’) and to hen auto (‘the one itself’) – and so
he does, on two occasions (a and b). On one occasion he even
calls it ‘the truly one’ (to hōs alēthōs hen, c), when he says that it does
not have parts. We may recall that when, early in the dialogue (e–
b), Socrates glosses eidos tou X as ho estin X (i.e., glosses the Form of
X as the essence of X), he characterises such an eidos as being ‘itself by
itself’ (auto kath’ hauto, e–a). All these, moreover, are familiar
variations of phrases that Plato standardly uses to refer to essences
and Forms.
In the Second Supposition (b ff.), he argues that this original one is
distinct from and not identical with being (b f.) and that the fact that
the two, one and being, are different from each other is not due to each of
them and what each of them is, but due to difference (to heteron; b).
Later in the same Supposition, this original one is said to be changeless and
at rest in relation to itself and subject to change in relation to others
(e–a); and again, that in relation to itself (pros heauto) it is
neither a whole nor has parts; and again, that it is ‘already identical itself
with itself’ (ēdē tauton einai auto heautō[i], c–). I would like to
propose that the ēdē (‘already’) here is to be understood to mean: in virtue

Forms as requirements of thought 
of being what it is, not in virtue of its relation to other things, and that the
self-identity this one ‘already’ has is, precisely, the identity a thing has in
virtue of being the very thing it is.
We may conclude that there is good reason to think that the one in the
second part of the Parmenides, that is, the one that enters into a relation
with being but is not, in virtue of itself, the one that is, is: that which the
quality, one, is, its essence. It is, therefore, a Form, on the basic notion of
Forms employed in the first part of the dialogue.
Is it ruled out, anywhere in the second part of the Parmenides, that the
one itself can be perceived by the senses? Or, what amounts to the same, is
it ruled out that that which oneness or unity is, its essence, can be
determined (specified, defined) by pointing to something that, to the
senses, is conspicuously one and unitary? Perhaps Plato could have made
things easier for us, if he had considered more explicitly, in the second part
of the dialogue, whether or not the one itself can be perceived by the
senses. It is of considerable assistance that, in this investigation he should
preface the second part with the explication that the thing supposed to be,
or not to be, is something one might especially grasp by thinking and
might consider to be a Form (e–). At the same time, one who is
sceptical about the existence and knowability of Forms, such as the sceptic
mentioned by Parmenides at b–c and e–b, will want to
know why we may not suppose that what oneness or unity is can be
perceived by the senses. I want to invoke here a notable passage from the
Republic, in which it is claimed (by Glaucon) that what oneness or unity is
cannot be perceived by the senses, because anything we perceive as one, we
no less perceive as indefinitely many. At the same time, this passage is
only of limited assistance to us here, because Glaucon’s claim cries out to
be spelled out and defended, especially if it is to answer one who is
prepared to believe that essences exist and can be known to us, provided
that he is satisfied that they can be perceived by the senses. In the Republic
passage, nothing further is said in defence of this crucial claim, and our
question is if the second part of the Parmenides does better.
There is a passage in the second part of the Parmenides that indicates
that, and why, that which oneness or unity is cannot be perceived by the
senses. The penultimate investigation (b–e) considers what

See Meinwald ().

For the view that these two things amount to the same, see Chapters  and .

Republic VII. a. At e this one is introduced in and through the question ‘What is this one
itself?’ (tí pote estin auto to hen).

 Plato’s Essentialism
happens to the many in relation to themselves, and to the thought of them,
if the one is not; that is, if the one, conceived as an object of thought, is not
(see e–). Perception, aisthēsis, is not mentioned; perhaps this is
because, as we saw, aisthēsis has been used for a thought-involving kind
of perception. But appearing, phainesthai, is mentioned repeatedly; and, as
we have seen, it is spelled out just how exceedingly limited this appearing
will be, and how extremely disjointed its object will be. We may suppose
that appearing here is, or includes, sensory appearing.
It is a good question whether the investigation that Plato provides here
is adequate for arguing against one who insists that what oneness or unity
is can be perceived by the senses, that is, insists that it can be determined
(specified, defined) what oneness or unity is by pointing to something
conspicuously unitary to the senses. This person is asked, by Plato in this
investigation, to imagine what the objects of sensory appearing, and
sensory appearing itself, would be like, if there were no such thing, either
for being or for thinking, as the one itself and the one as an object of
thought. If he agrees that this is the only or the best way of investigating
whether what oneness or unity is can be perceived by the senses, this may
well be adequate for persuading him.
Why does Plato think that the only or the best way of investigating
whether what oneness or unity is can be perceived by the senses is to
imagine what the objects of sensory appearing and sensory appearing itself
would be like, if there was no such thing, either for being or for thinking,
as the one itself and the one as an object of thought? Is this not to stack the
cards against the view that what oneness or unity is can be perceived by the
senses? For, it seems, the proponent of this view may insist that a different,
and fairer, way of testing the view would be to imagine a scenario in which
one is perceiving by the senses, such as sight, something conspicuously
unitary, such as a snow-crystal on a window pane (as in Hans Christian
Andersen’s tale) or a schematic crystal on a computer screen, and to
consider whether this conspicuously unitary object of the senses can serve
as an example and exemplar, and hence as a standard (paradeigma), of what
oneness or unity is.
It seems to me that Plato has a point in not conducting the test, of the
view that what oneness or unity is can be perceived by the senses, in this
way, and in proposing, rather, the very different way that he proposes. The
problem with this other, and seemingly fairer, way of conducting the test is
that it makes it hard to tell whether, when we perceive (e.g.) the snow-
crystal as perfectly unitary, or sufficiently unitary to be capable of serving as
an example and exemplar, and as a standard, of oneness or unity, we are,

Forms as requirements of thought 
strictly, using sense-perception alone, and this is how we are able to
recognise the snow-crystal as unitary, or, on the contrary, it is only because
sense-perception here is combined with thought, and, potentially, a kind
of thought that cannot be accounted for simply on the basis of sense-
perception, that we are able to recognise this. Plato’s alternative way of
conducting the test is, I suggest, due to this problem, and it serves to
obviate it.
We may conclude that there is good reason to think that the one in the
second part of the Parmenides, that is, the one that enters into a relation
with being but is not, in virtue of itself, the one that is, is: that which the
quality, one, is, its essence, and such that it cannot be perceived by
the senses.
We may conclude that, late in the second part of the Parmenides, Plato
identifies and defends a further necessary condition for thought, in addi-
tion to the two conditions he identified and defended earlier in the second
part (The Third General Condition for Thought):
Thought is of something that is one;
and thought can be of something that is one, only if there is something that
oneness or unity is, its essence, and this essence cannot be perceived by the senses
(it is, in this sense, a Form).
Let us recall the two conditions of thought that he identified and defended
earlier in the second part: (The First General Condition for Thought)
Thought is of something that is; and (The Second General Condition for
Thought) Thought requires complexity in the object of thought, and such
complexity requires at least two things standing in relation to each other,
namely, the relation ‘X belongs to Y’. It is remarkable that these three general
conditions for thought are anticipated in the first part of the dialogue – at
b–c (also at b–c) – but while there it was simply asserted that
thought is subject to these conditions, in the second part the conditions are
spelled out and defended.
. Conclusion
One remarkable consequence of the Third General Condition of Thought,
as this is articulated and defended in the second part of Plato’s Parmenides,
is this: It is not specific Forms (the Form of justice, of human beings, of
fire, etc.) that provide for the oneness of the object of thought, but, in the
first instance at any rate, the Form of oneness that does this. The same
condition was put forward, without articulation and defence, in the first

 Plato’s Essentialism
part of dialogue, at b–c. It is proper, therefore, to use this condition
of thought, as articulated and defended in the second part, to determine
how we are to understand the same condition as it was put forward in the
first part. Once we do this, we recognise, contra a distinguished tradition
of critics (see n.  above), that we should not understand the claim at
b–c to be that each thought is the thought of a Form, and such that
to each different thought there corresponds a different Form; rather, we
should understand it to be that, since each thought is the thought of
one thing, it involves the thought of, in the first instance, the Form of
oneness or unity.
To better appreciate all this, we would have to consider how the Form
of oneness enters into thought and what its function is in thought. Crucial
for considering this, it seems to me, is to recall that the claim at b–c
was not simply that Forms are necessary for thought, but that Forms, and
the defining of Forms, are necessary for thought. (We may also recall that
Parmenides took up this idea, of defining Forms, in what immediately
follows b–c and begins to prepare for the second part.) This suggest
that what the Form of oneness does, for us and in our thought, is, in the
first instance at any rate, enable us to define, indeed, to try to define,
anything. It does not seem at all plausible to think that undertaking this
activity, of trying to define something, requires that we are, in some way
and unbeknownst to us, already and in advance of having successfully
enquired, referring to the successful result of the activity – referring to the
specific Form that we would know if we were to have successfully defined
it. But it is not implausible to think, as Plato apparently does, that this
activity, of trying to define any thing, requires having a suitable notion of
oneness – suitable especially as the oneness and unity distinctive of
definition and essence, the oneness that we are after when we seek to
define something – and requires knowing what things would have to be
like to be one in this sense.
On the account we have defended of why Plato thinks that Forms, and
most especially the Form of oneness, are necessary for thought, Plato
thinks that the attempt to define things, hence anything that this attempt
requires and involves, is necessary for thought, and for thought in general,

It will be evident that the account I have defended of the way in which, for Plato, Forms are
necessary for thought is far removed from the view, defended by, e.g., Bestor (), that Plato’s
account of how Forms are involved in thought is a variation of a Putnam-style externalist account of
the object-directedness of thought. On a Putnam-style externalist account, a thought is of an object,
O, if it is suitably caused by O; and a thinker can be thinking of O without being aware that this is
the object of which she is thinking.

Forms as requirements of thought 
not just for a most advanced mode of thought. This, indeed, is directly
implied by what he says early in the Cratylus, when he says that a central
function of naming is to correctly divide things. It is a very good, and a very
big, question why Plato thinks that all thought involves, in one way or another,
the attempt to distinguish, divide and, ultimately, define things. This comes
down to the question of what, for Plato, is the place and function of the ti esti
question in thought in general and why we are motivated and justified in raising
the ti esti question, not just as specialist philosophers, or specialist scientists, but
simply as thinkers. I have addressed an important specific dimension of this
question in Chapter ; I have also addressed it more generally in The Structure of
Enquiry in Plato’s Early Dialogues (see also Politis ; b).

Cratylus b–c, picked up at the end of the dialogue at d–e (I am supposing that, for Plato,
naming is not only a linguistic act but also a mental and intellectual act). Contra several critics (e.g.,
Kutschera ; Rickless ), in the Cratylus Plato does not argue that names signify Forms and
would be meaningless if they did not signify Forms. It is a difficult question what Plato’s account is,
in the Cratylus, of how names name (onomazein), or signify (sēmainein), things, or whether it is even
part of Plato’s aim in this dialogue to provide a general account of how words are related, for their
having a meaning, to the things they signify. However, it ought to be evident that he allows that a
name can signify something, without signifying a Form. Towards the end of the Cratylus, he argues,
carefully and persistently, that a name can be assigned to a thing simply demonstratively and by
pointing (–c). He links up this argument with the account of the function of names and of
naming that he had set out at the beginning of the dialogue (see d–e, where he refers back to
b–c). According to that original account, a name is a tool (organon) whose function is to
distinguish how things are (b–c; a tool by the use of which ‘we distinguish how things are’,
ta pragmata diakrinomen hē[i] echei, b–; ‘a tool for distinguishing the way things are’, organon
diakritikon tēs ousias, b–c). And, according to the argument towards the end of the dialogue, this
function can be accomplished simply demonstratively and by pointing. It is true that, if it is
accomplished in this way, it may not – indeed will not – issue in a true and correct way of
distinguishing how things are, but it will have succeeded in distinguishing how things are in a way
that is sufficient for assigning a name to a thing. For a similar point about the Cratylus, and one
made long before this dialogue became again a focus of serious interest, I recommend Gosling
(, ): ‘But this position [i.e., that ‘all meanings are Forms’] is not open to Socrates. For at
[Cratylus] d– he has first argued that what a noise means is a matter of convention, showing
this by the fact that we readily understand words although they contravene Cratylus’ similarity rule
and so should not be intelligible. Second, he argues that words might well have been ill-devised. But
the danger of an ill-devised word is not that it is senseless, but that it is misleading . . .. Yet if they
[i.e., words] always in virtue of being meaningful isolated Forms which constituted their meaning,
they could not be misleading.’ It may be said that this account of the Cratylus ignores its last two
pages (d–e), where it is argued that without Forms it is not possible ‘to correctly address a
thing’ (proseipein auto orthōs, d). Does it follow that Forms are necessary for naming things,
hence for names to have a meaning? The term proseipein (‘to address’) is closely related to the term
onomazein (‘to name’). We may suppose, therefore, that when Plato says that Forms are necessary
for correctly addressing things, he intends to imply that they are necessary for correctly naming
things. But the ‘correctly’ (orthōs) makes all the difference. Socrates has, in the ten Stephanus pages
coming up to this juncture, been at pains to argue that it is possible to name a thing – and,
therefore, it is possible for a name to have a meaning – without naming it correctly; and the ending
of the dialogue carefully maintains consistency on this point.

 
Why are essences, or Forms, separate from

physical things?
Also Timaeus and Philebus
According to the account I shall now defend, the relevant notion of

separation, when we consider Plato’s view that Forms are separate from
physical things, is not, contra a long tradition of interpretation of Platonic
separation, the modal notion: It is possible for X to exist without Y, but it is
not possible for Y to exist without X. Rather, it is the essence-based notion:
The account of what X is need not make reference to Y, but the account of
Y must make reference to X. This account is altogether consonant with the
view that what Plato’s Forms basically are, is essences. The one thing we
can say with certainly, I shall argue, on the question of whether Plato
thinks that Forms, or the Forms of some qualities, and qualities of very
special and paramount importance, are separate, and in what sense they are
separate, is that he thinks that the account of what certain qualities are,
their essence, need not make reference to physical things. The qualities he
has in mind include such qualities as: one, like, equal, good, beautiful and
just. It is of such qualities that, in Parmenides , Socrates says that he is
confident that there are separate (chōris) Forms; and it is such qualities that
he repeatedly mentions, in Phaedo and Republic, when he mentions
qualities of which there are essences and Forms.
On the account I shall defend, the relevant difference between Plato’s
theory of essences and Forms and Aristotle’s theory of essences and forms,
is this: For Plato, the things or qualities that principally have an essence
and Form are such qualities as: one, like, equal, good, beautiful and just; and
the account of what such a quality is need not make reference to physical
things. For Aristotle, on the other hand, the things or qualities that
principally have an essence and form are such things as: human, fire and
water – things that he characterises as substances (ousiai) and we com-
monly characterise as natural kinds or things belonging to natural kinds.
The account of what such a thing is must make reference to physical
things – after all, human, fire and water are physical things, things
belonging to kinds in nature (phusis). This is, I believe, also the difference


The separation of Forms from physical things 
that Aristotle has in mind. For him, just as for Plato, this, the essence-
based, not the modal, is the crucial notion of separation, and the one
suited for distinguishing his, Aristotle’s, essentialism from Plato’s.
Finally, on the account I shall defend, Plato’s claim that Forms are
separate from physical things is a well-motivated, well-integrated and
undetachable part of his essentialism. For, if it is supposed that the things
or qualities that principally have an essence and Form are such qualities as
one, like, equal, good, beautiful and just, then it is indeed plausible to think
that Forms are separate from physical things. This is because there is no
particular reason to think that the account of what such a quality is, its
essence, must make reference to physical things. If, therefore, an essential-
ist wants to deny separation, she will need to argue that the things or
qualities that principally have an essence are not such things or qualities,
but, on the contrary, things or qualities that belong to natural kinds, such
as human, fire and water. I believe that Aristotle recognised that this is a
task that he has to undertake in order to differentiate his essentialism from
Plato’s and in order to resist separation. This is unlike the typical modern
neo-Aristotelian essentialist, who begins by assuming that, if there are
things that have an essence, they are things that belong to natural kinds –
no wonder the modern essentialist is able to conclude that essentialism is
compatible with the naturalism that is the order of the modern day today.
The account I shall defend does not deny that Plato thinks that it is
possible for the Form of a quality, F, to exist without there being a
physical thing that is F. Rather, what I want to assert is that, if Plato
thinks that it is possible for the Form of a quality, F, to exist without there
being a physical thing that is F – and there is reason to think that he does –
he does so because, and for no other reason than that, he thinks that the

See, e.g., the following passages – which should be read seriatim – from Aristotle’s Metaphysics (I am
grateful to Ge Tianqin for having impressed them on me): ‘Being and unity are the same – i.e. one
nature – in the sense that they are followed by one another, just like principle and cause, not as being
indicated by the same formula (though it makes no difference even if we believe them like that –
indeed it helps).’ (Metaphysics Γ. b–, trans. Kirwan, modified) ‘On the other hand, that
the one in some sense signifies the same as being, is clear in virtue of the fact that it follows the
categories in the same number of ways and that it is in none of them (e.g. it is neither in the category
of the what-it-is nor in the category of of-what-quality, but it behaves in the same way as being), and
in virtue of the fact that “one human being” does not add anything else in predication to “human
being” (as being, too, is nothing over and above being a certain something or of a certain quality or of
a certain quantity), and in virtue of the fact that being one is being for each thing.’ (Metaphysics I.
a–, trans. Castelli) See also Metaphysics Λ. a–. I note that the passage from
Metaphysics Γ. (b–) has been thoroughly analysed by Politis and Steinkrüger (). The
passage from Metaphysics Iota is, of course, thoroughly striking for our present purpose.

See Gerson ().

 Plato’s Essentialism
account of what this quality, F, is need not make reference to a physical
thing that is F.
At the same time, there is some indication, I shall argue, that Plato does
not think that the inference from The account of what X is need not make
reference to Y to It is possible for X to exist without Y, is straightforwardly
valid. Rather, what he thinks is that, if we have reason to think that It is
possible for X to exist without Y, then we do so because we have reason to
think that The account of what X is need not make reference to Y; and,
conversely, if we have reason to think that It is not possible for X to exist
without Y, then we do so because we have reason to think that The account
of what X is must make reference to Y. One consequence of Plato’s view is
that, if someone – such as a latter day naturalist essentialist – claims that it
is not possible for essences to exist without physical things existing that
have those essences, but she does not consider whether the account of what
those things are must, or whether it need not, make reference to physical
things, then she has lent herself to a claim dogmatically and without
honest work.
. Immortality and separation in the Phaedo

Imagine a person who knows she only has a few days, possibly only a few
hours, to live, and who, for reasons of her own, finds herself asking
whether she may hope for something, and something good, for herself
after her death. She talks to some friends of hers, who, finding that she
seems hopeful in this regard, wonder why she is so concerned lest such
hopes be entirely vain. She tells them that it is common and received
opinion that we are dependent on our bodies for our existence, and,
notwithstanding having heard of some who believe in the resurrection of
the body, she personally has no illusions about the prospect of her body, or
something remotely like it, surviving her death. Now imagine that one of
her friends sets out to demonstrate to her that that which each of us really
is, and which distinguishes our individual life at its core, does not have to
figure anything bodily. Being a lover of beauty, and no mean musician, she
is not at first convinced, but she pricks up her ears as her friends remind
her that even Beethoven was not so much worse off deaf and that even
without a body she will be able to enjoy her musical scores and keep
working on them.
However, she finds that, whatever else, there is one particular sticking
point in her friend’s well-intended attempt: even if that which one is and
which distinguishes one’s life at its core does not have to figure anything

The separation of Forms from physical things 
bodily, one’s being alive may depend on one’s having a body; and, if it
does so depend, then any hopes one may have of surviving one’s bodily
death will necessarily be vain – which means that the point about one’s life
and its core will be worth keeping in mind, if at all, only for as long as one
is alive body and all. If her hopes are not to be vain, she insists with
obvious good sense, she must be capable of existing without her body; it is
not enough to be assured that what she is does not have to figure anything
bodily, or that the account of what she is need not make reference to
anything bodily.
Socrates’ situation as described early in the Phaedo, and his reaction to it
when it comes to the question whether to face death with hope or, on the
contrary, dejection and/or fear, is comparable to our imaginary situation;
except that it is Socrates who spells out that that which is distinctive of our
life at its core, especially if the life is fuelled by the love of wisdom, does
not have to figure anything bodily. It is his friend, Kebes, who objects that,
even if that is so, the common opinion may also be true that says that one’s
soul, hence one’s life in whatever form, cannot exist without a body (see
e–b) – whatever else it also is, one’s soul is, of course, that on
account of which one is alive.
This is good indication that Plato is aware of the move from The account
of what X is does not have to make reference to Y to It is possible for X to exist
without Y and is aware that there may be an issue with this move. We shall
find further confirmation in the Phaedo of Plato’s awareness of this. The
issue is just, that it is not clear whether this is a valid move; and it is not
clear whether, even if it is not valid (i.e., the truth of the antecedent claim
is compatible with the falsity of the consequent claim), the antecedent
claim provides good reason, and perhaps even the best reason we can have,
for the consequent claim. In the first instance, the move in the Phaedo is
from considerations about what one’s soul, oneself, and one’s life is, its
essence, to the conclusion that one’s soul, oneself, and one’s life can exist
without a body. But there is reason to think that Plato is just as much
aware of the move, and of issues it raises: from considerations about what a
quality, F, is, its essence, to the conclusion that it is possible for the essence
of this quality, that is, the Form of F, to exist without being present in a
physical or bodily thing.

For Plato’s close association of the notion of one’s soul with the notion of oneself, see esp. Protagoras
–, where he moves freely back and forth between existential questions about ‘one’s soul’ to
the same questions about ‘oneself’.

Plato often uses ‘to be present’ (eneinai) for the converse relation of ‘to partake of’ (metechein).

 Plato’s Essentialism
A closely related move is made by Simmias later in the dialogue
(e–d), when he insists that Socrates’ argument so far, for the claim
that the soul can exist without the body, has been inconclusive, and this
because it has not been based on an account of what the soul is (cf. ei oun
tugchanei hē psuchē ousa harmonia tis, dēlon hoti . . .; ‘If, therefore, what
the soul is is a certain attunement, it will be evident that . . .’, c–).
For, Simmias argues, what Socrates has argued so far, about the soul and
its superiority over the body whose soul it is, is compatible with
supposing that what the soul is is a certain attunement of bodily elements,
in which case the soul will not be capable of existing without the body
whose soul it is.
There can be no doubt, it seems to me, that this juncture (and following
Kebes’ additional argument against Socrates, which follows Simmias’
attunement argument as advanced and engaged with so far) marks a
deliberate crisis point in the dialogue. Those around Socrates are described
as becoming now thoroughly uncertain about which of the two sides to
believe: Socrates’ arguments for the claim that the soul can exist without
the body, or, on the contrary, Simmias’ and Kebes’ arguments for the
contrary claim. Plato at this juncture even steps back into the frame of the
dialogue and engages, through the characters in the frame, in extended
reflection about methodological issues raised by such radical and existential
uncertainty. This uncertainty is not relieved, or properly addressed, until
considerably later in the dialogue (e ff.), when, at long last, Socrates
embarks on what will prove to be the final argument for the immortality of
the soul. This is, of course, an exceedingly ambitious and complex argu-
ment; but if one thing is clear about it, it is that, ultimately, it bases the
conclusion that the soul is immortal, or death-less (a-thanaton), on an
account of what the soul is, its essence: the soul is the ultimate cause of life,
life being the opposite of death. It is only here, at the very end of the
dialogue and through its final argument, that Socrates provides what
Simmias requested of him through the attunement argument: a defence
of the claim that the soul can exist without the body, based on an account,
itself properly defended and not a mere hypothesis or supposition, of what
the soul is, its essence.
What this shows is that Plato thinks that a claim of the form X can/
cannot exist without Y must, for its justification, be based on an account of
what X is, its essence. It shows this, in the first instance, in regard to the
claim that the soul can/cannot exist without the body, but we may suppose

We have considered a central part of it, in Chapter .

The separation of Forms from physical things 
that Plato thinks the same in regard to the claim that the Form of a quality,
F, can exist without being present in a physical or bodily thing: he thinks
this claim can be justified only by our relying on an account of what this
quality, F, is, its essence. It is familiar that, in the Phaedo (and indeed in
the Republic), the soul is said to be akin to the Forms (suggenēs; ‘kindred’,
‘of like kind’). And it is clear that this kinship concerns, crucially if not
exclusively, the relation of the soul, and the Forms, to physical or bodily
things. Our soul, it is argued in the dialogue, can exist without a body;
and, it is likewise argued, we may suppose that it has existed, before its
present embodiment, and will exist, after its present embodiment, without
a body. So, too, the Form of a quality, F, can exist without being present in
a physical or bodily thing, and, even if actually it is present in a physical or
bodily thing, it does not depend, for its existence, on being so present.
We may conclude that, even if we suppose that Plato thinks that the
Form of a quality, F, can exist without being present in physical things, he
thinks that the only way this claim can be justified is by determining what
this quality, F, is, and by determining whether the account of what it is
allows, or does not allow, that F can exist without being present in a
physical thing. It follows that the question we need to consider, if we
want to consider whether Plato thinks that Forms are separate from
physical things, and what he means by this claim, and why he makes it,
is the question whether Plato thinks that the account of what a quality, F,
is allows, or does not allow, that what F is can exist without being present
in a physical thing. We need to consider what, in his view, determines
whether it does or does not allow this and why, in his view, for certain
qualities, the account of what a quality, F, is does allow this.
We may wonder: Is Plato more interested in the question whether the
account of what a quality, F, is must make reference to physical things or in the
question whether the essence or Form of that quality can exist without being
present in physical things? It is hard to tell, but it seems to me that he is
aware that there is a substantial move, and one that raises issues, from a
negative answer to the former question to a positive answer to the latter –
no blithe commitment for Plato to separation and the argument behind it.
He indicates awareness of such a move in regard to the question whether
the soul can, and will, exist without being present in bodies. For, having
argued, with great commitment in his last argument (in the Phaedo), that
what the soul is is such that it cannot admit of death – it is athanaton,
deathless – Socrates appears to raise a question mark about whether we can

I am using ‘is’ here, in ‘is present in’, timelessly.

 Plato’s Essentialism
directly infer that the soul is indestructible – anōlethron. But, if there is an
issue with this move, it is hard to see how there will not be a similar issue
with the move from the supposition that the account of what a quality, F,
is need not make reference to anything physical to the conclusion that the
essence and Form of this quality can exist without being present in
anything physical.
What are we to conclude from Plato’s apparent awareness of this issue?
We cannot, I think, conclude that, in the final analysis and at the end of
the day, Plato simply wants to resist and throw out this move, whether in
regard to the soul or in regard to essences and Forms. That would
mean that Socrates’ whole final argument for the immortality and inde-
structibility of the soul – his practically final swan-song (‘practically’
because there is a weighty myth to follow before he takes his leave) – is
intended as vain; whereas, it seems to me, as it does to many critics, it is
intended as a powerful argument and one very much worth the effort Plato
has put into it.
I would like to propose the following way of reconciling the store Plato
puts in this argument with his apparent awareness that the basic move on
which it relies can be questioned. On the one hand, Plato thinks that the
best reason we can have for thinking that X can, or that it cannot, exist
without being present in, or in general mixed up with, Y – in general, that
X can, or that it cannot, exist separately, itself by itself, pure and unmixed –
is to think, with good reason, that the account of what X is need not, or
that it must, make reference to Y. On the other hand, he recognises that if
someone wants to assert that the account of what X is need not make
reference to Y and at the same time deny that X can exist without being
present in, or in general mixed up with, Y, this person has not contradicted
himself and his position cannot be accused of logical inconsistency. There
is more to good philosophy than asserting things whose denial entails
a contradiction.
Perhaps this is why Socrates’ final argument is not quite his final swan-
song, which is, rather, the myth he goes on to tell before being put to
death. Socrates’ final argument is intended, by Plato, as a very good
argument for the immortality and indestructibility of the soul, and one
that cannot be matched by an equally good argument for the opposite
conclusion – the way in which Plato had Socrates’ earlier arguments in the
dialogue be matched by counter-arguments put forward by Simmias and
Kebes. On the other hand, if this argument is consistent with the assump-
tion that our physical death is our final demise, then there is a way in
which this assumption is tenable. And this means that there is a way in

The separation of Forms from physical things 
which any view based on this assumption, such as the view that our
physical demise is an object of dejection and fear, is tenable. There is,
therefore, still work to do in order to engage with this assumption and with
any view based on it. I suggest it is to such work that the myth, Socrates’
last word, is addressed.
. Unitary (etc.) versus human (etc.) in Parmenides 

Here is Parmenides a–e:
In fact, what Parmenides said when Socrates had finished confirmed this
impression. ‘Socrates’, he said, ‘you are much to be admired for your
keenness for argument! Tell me. Have you yourself distinguished as sepa-
rate, in the way you mention, certain forms themselves, and also as separate
the things that partake of them? Do you think that likeness itself is
something, separate from the likeness we have? And one and many and
all the things you heard Zeno read about a while ago?’ ‘I do indeed’,
Socrates answered. ‘And what about these?’ asked Parmenides. ‘Is there a
form, itself by itself, of just, and beautiful, and good, and everything of that
sort?’ ‘Yes’, he said. ‘What about a form of human being, separate from us
and all those like us? Is there a form itself of human being, or fire, or water?’
Socrates said, ‘Parmenides, I’ve often found myself in doubt whether
I should talk about those in the same way as the others or in a different
way.’ ‘And what about these, Socrates? Things that might seem absurd, like
hair and mud and dirt, or anything else totally undignified and worthless?
Are you doubtful whether or not you should say that a form is separate for
each of these, too, which in turn is other than anything we touch with our
hands?’ ‘Not at all’, Socrates answered. ‘On the contrary, these things are in
fact just what we see them to be. Surely it’s too outlandish to think there is a
form for them. Not that the thought that the same thing might hold in all
cases hasn’t troubled me from time to time. Then, when I get bogged down
in that, I hurry away, afraid that I may fall into some pit of nonsense and
come to harm; but when I arrive back in the vicinity of the things we agreed
a moment ago have forms, I linger there and occupy myself with them.’
‘That’s because you are still young, Socrates’, said Parmenides, ‘and phi-
losophy has not yet gripped you as, in my opinion, it will in the future,
once you begin to consider none of the cases beneath your notice. Now,
though, you still care about what people think, because of your youth’.
(trans. Gill & Ryan, slightly changed)
The question that Parmenides puts to Socrates in this passage is whether
there are separate Forms, that is, apparently, whether there are Forms that
are separate from physical things (or, from bodies, or from sense-
perceptible things: these are different ways – indeed, significantly different

 Plato’s Essentialism
ways – of referring to the same things, and it is clear from the end of the
passage that Socrates intends separation from sense-perceptible things). On
a plain reading of this passage, Socrates’ answer is that it depends on the
things or qualities we are considering. If we are considering such qualities
as one (i.e., unitary) or like, or such qualities as good, beautiful and just, he is
confident that there are separate Forms of such qualities. At the other end,
if we are considering such things as mud, hair or dirt, he is confident that
there are no separate Forms of them; for, he says, these things are just as we
see them to be. However, if we are considering such things as a human
being, fire or water, he is unsure about what to say, and, in particular,
unsure about whether to answer in the same way as in his first and
confident affirmative answer.
One reason why this passage is so interesting, as is recognised by critics
and commentators, is that, as perhaps the only passage in Plato, it is
expressly addressed to the question of the scope of Forms: of what things
or qualities there are Forms, or separate Forms. I want to propose that the
passage is also of very particular interest and significance, if we want to
understand what Plato means by separation and the idea that Forms, or
some Forms, are separate from physical things, and, if we want to under-
stand why he thinks that Forms, or the Forms of certain qualities, and
qualities that Plato considers of special and paramount importance, are
separate from physical things. This is a plausible proposal, it seems to me,
because, first, those things in regard to which Socrates is unsure whether
there are separate Forms are things that it is natural to think of as,
precisely, natural kinds, or things belonging to natural kinds, or simply,
physical things, things in nature, phusis; indeed, as exemplary examples of
such things: humans, fire and water. Secondly, because those things in
regard to which Socrates is confident that there are separate Forms are not
things that it would be at all plausible to think of as natural kinds, or things
belonging to such kinds. This is because, whatever each of these things
(one, like, good, beautiful and just) turns out to be, on an adequate and true
account of it, it is not plausible – at any rate not immediately and pending
further particular, substantive, and disputable reasons – to think that the
account of what it is must make reference to things in nature,
physical things.
Why is it not plausible, or not immediately plausible, to think that the
account of what a quality such as one or unitary is must make reference to
physical things? Consider the contrast case: the account of what water is, or
what a human is or what fire, is. Whatever each of these things turns out to
be, on an adequate and true account of this, it is hard to imagine how the

The separation of Forms from physical things 
account could avoid specifying that water, fire, humans are things in space
and time, and subject to the kinds of changes distinctive of things in space
and time, and subject to generation and destruction, and the like. Now, set
against this the case of one or unitary. Suppose it is said that the account of
what this quality is must, likewise, make reference to things in space and
time, such as humans or water, that are one and unitary. The immediate
problem with this answer is that, as it stands, it implies that only things in
space and time can be one and unitary. But the supposition that only
things in space and time can be one and unitary is not immediately
plausible; if anything, it is not at all plausible, at least not immediately
and pending further particular, substantive and disputable reasons. In the
absence of such reasons, the supposition is too implausible. For, if there are
non-physical things, such as numbers, or geometrical objects or rational
souls, they, too, can be one and unitary. For similar reasons, if there are
non-physical things such as these, they, too, can be beautiful (think of a
geometrical figure or a mathematical proof ), and they, too, or at any rate
those of them that are alive, such as rational souls, can be good and just.
To be sure, we can imagine a philosopher, and one who has a stake in
naturalism, insisting that such objections can be circumvented and such
problems answered. She could do so by insisting that the quality one or
unitary is a pros hen legomenon: i. it has a primary sense and a derivative
sense; and ii. the account of what it is in its primary sense must make
reference to physical things that are unitary, but the account of what it is in
its derivative sense, even if it does not need to make reference to physical
things immediately, it will need to make such reference eventually, because
non-physical things depend, for being what they are and for being any-
thing at all, on physical things. We would have to consider such strategies,
if we were to consider, fully and properly, Aristotle’s answer to Plato’s
account of oneness and unity. However, our present question is not
whether Plato’s claim, which says that essences or Forms, or the essences
and Forms of some things or qualities, are separate from physical things,
can be resisted, and indeed can be resisted even by an essentialist – we

I want to be clear that, when I speak of such things as being ‘non-physical’ things, I do not mean, or
I do not have to be taken to mean, ‘things separate from physical things’, on any account of
separation. All I need be taken to mean is what, a generation or two ago, philosophers – including
analytic philosophers such as Peter Strawson and Donald Davidson – used to mean when they would
say such things as: a human being is a bearer of at once physical and mental (hence non-physical)
properties. I hope such ways of speaking are still intelligible, and respectable enough, even today
when naturalism, both metaphysical and methodological, is the order of the day, and that they are
intelligible and respectable not only in regard to such jejune entities as qualia.

 Plato’s Essentialism
should not be surprised if it can. Our question is what Plato means by this
claim, why he makes it, and whether the reasons why he makes it are
good reasons.
Let us look again at the Parmenides passage, and ask: What are the
salient differences between a thing or quality such as one or unitary and a
thing such as human? And what difference do these differences make,
regarding the issue of separation in Plato?
Three particular differences, it seems to me, stand out. First (The First
Difference), the quality one or unitary is true of all things or, at any rate, of
all things that have an essence and Form. On the other hand, not all things
are humans or water. Secondly (The Second Difference), the quality one or
unitary is true not only of individual things, such as this human or that
(mass of, or collection of ) water, but also of wholes of such things,
including the maximal whole, that is, the universe or unqualified totality
of things (to pan). On the other hand, only individual things are human
or (masses of, or collections of ) water. Thirdly (The Third Difference),
unity is involved in the account of what such things as human and water
are. It is a good question, and important to keep in mind: whether such
things as humans and water are involved in the account of what unity is. It
seems to me that, clearly and on the particular textual grounds provided by
the Parmenides passage, Plato thinks, No; it is Aristotle, in subtle but
significant and indeed decisive difference to Plato, who argues, Yes.
What difference do these three differences make in regard to the issue of
separation in Plato? This is a large and forbidding question, but what
I have in mind is the more manageable question: What difference do these
three differences make, in regard to the issue of separation in Plato, if we
take account of just the Parmenides passage and the fact that, in it, Socrates
is sure that there are separate Forms of such things or qualities as one or
unitary but unsure whether he should say the same of such things as
humans, water and fire. Consider the first and the second differences. By
themselves, they are compatible with thinking (as, apparently, Aristotle
does) that there is not a separate essence or Form of unity; for, they are
compatible with thinking that unity is a pros hen legomenon (in the sense
indicated above). Or, to put it more simply, they are compatible with
thinking that unity itself is dependent, for being what it is and for being

It is striking how, when, in Sophist, Plato formulates the various theories about what there is, he
formulates them as being not only about ‘that which is’ (to on) and ‘the things that are’ (ta onta), but
also about ‘the totality of things’ (to pan) and ‘all things’ (ta panta): e, d, e, b,
b, b, c, d, d, a.

The separation of Forms from physical things 
anything at all, on the distinctive unity of such things as humans and
water, but the distinctive unity of such things as humans and water is not
dependent on unity itself. Or, to put it in a different but, arguably,
equivalent way, they are compatible with thinking that unity is primarily
true of individual things, such as this human or that (mass of, or collection
of ) water, and that it is true of wholes of such things, including the
maximal whole, the universe, only as a consequence of being true of those
individual things.
The fact is that Plato has Socrates say, in no uncertain terms, that there
is a separate Form of unity, and it is of things such as humans that he is
uncertain whether to say the same. It follows that, for Plato, unity is not a
pros hen legomenon, in the sense indicated; and that unity itself is not
dependent, for being what it is and for being anything at all, on the
distinctive unity of such things as humans and water; and that it is not
the case that unity is primarily true of individual things and only as a
consequence is it true of wholes of such things, including the maximal
whole, the universe. We may even suggest that a good way of taking
Socrates’ surety in the one case, combined with his unsurety in the other
case, is to suppose that the view Plato is articulating here, by having
Socrates take up this combination of stances, is that the distinctive unity
of such things as humans, water and fire depends, for what it is and for
being anything at all, on unity itself, but unity itself does not depend, for
being what it is and for being anything at all, on the distinctive unity of
these unitary things.
We arrive at the same result, if we consider The Third Difference,
which says that unity is involved in the account of what such things as
humans and water are. On its own, this difference is compatible with
thinking that such things as humans and water are involved in the account
of what unity is; for, it is compatible with thinking that the distinctive
unity of such things as humans and water is involved in the account of
what unity itself is. In that case, however, Plato would have needed to have
Socrates say either that there are separate Forms both of unity and of
humans and water, or that there are not separate Forms of either unity or
humans and water. The fact is, he has Socrates say that there is a separate
Form of unity, but has him have doubts about whether there is a separate
Form of humans and water. It follows that Plato does not think that such
things as humans and water, or the distinctive unity of humans and water,
are involved in the account of what unity itself is.
We might wonder: Why, then, does not Plato have Socrates simply
deny that there is a separate Form of water, humans, fire? Why does he,

 Plato’s Essentialism
rather, have Socrates say that he is unsure whether to answer in the same
way as in the case of unity, etc.? The following answer suggests itself:
Simply to deny that there is a separate Form of humans, fire, water would
be misleading, for it would invite the impression that the case of humans,
fire, water can be considered on its own and apart from the case of unity,
etc. But if, as Socrates is sure, there is a separate Form of unity, and if this
unity is involved in the account of what humans, fire, water are, then the
case of humans, water, fire must be considered together with the case of
unity. What this means is that there is a sense in which, yes, there is not a
separate Form of humans, fire, water, because these are natural kinds, or
things belonging to natural kinds, and the account of what they are
must make reference to physical things; and yet, there is also a sense in
which, no, it is not the case that there is not a separate Form of them,
because they depend on something, unity itself, of which there is a separate
Form. No wonder Socrates is unsure whether to answer as in the first case,
the case of unity, etc.
. Is the Parmenides () passage consonant with, and

confirmed by, dialogues that went before?
The question is whether, before the Parmenides (which, we may suppose, is
written after such dialogues as the Republic, Phaedo and Hippias Major, but
before such dialogues as Theaetetus, Sophist, Statesman, Timaeus and
Philebus), when Plato commits himself to Forms of certain qualities, it is
to Forms of such qualities as one, like, good, beautiful and just that he
commits himself; and whether, if he commits himself to Forms of natural
kinds, such as fire, water or human beings, he indicates that the Forms of
such things are dependent on those, apparently primary Forms. If we were
considering these questions from scratch, they would require much and
difficult work. On foot of the findings of this study so far, they invite
immediate and unambiguous affirmation.
Our investigation took its starting point in Plato’s claim, in the Hippias
Major, that certain qualities cannot be defined by example and exemplar
and in his remarkable argument for this claim. The quality in question in
that dialogue was, beautiful. So we are immediately within the ambit and
remit of the qualities of which alone, in the Parmenides passage, Plato has
Socrates be confident that there are separate Forms.
We went on to argue that Plato’s claim, most familiar from the Phaedo,
that the Form of certain qualities cannot be perceived by the senses, is
simply a consequence of the claim, as argued for in the Hippias Major, that

The separation of Forms from physical things 
what these qualities are, their essence, cannot be defined by example and
exemplar. The qualities that the Hippias Major argument directly extends
to, we saw, are such qualities as equal (as in Phaedo ) and one (as in
Republic VII. –). The Hippias Major argument, we have now seen,
extends also to such things as water (or the proverbial humming bee
mentioned in the Meno as an object of the ti esti question), if at all, only
indirectly, that is, if we suppose that, for Plato, the account of what such
natural kinds are must make reference to such qualities such as one. Again,
we are immediately within the ambit and remit of the qualities of which
alone, in the Parmenides passage, Plato has Socrates be confident that there
are separate Forms: this is true of the quality, one; and there is not a big
difference between the quality, like, in the Parmenides passage, and the
quality, equal, in the Phaedo, because we may suppose that for a pair of
things to be equal is for the two things to be like in a certain way or
respect, that is, in regard to magnitude or quantity. And we have already
reason to think that the Forms of such things as humans are dependent on
those, apparently primary Forms.
Our investigation has amply made manifest just how prominent a
place – and in what multiple ways – Plato affords the quality, one or
unitary. To begin with, Forms are expressly characterised, in Phaedo and
Republic, as one and unitary and uniform; and we argued that this goes
back, in the order of justification, to the requirement, which we encounter
in several dialogues before the Phaedo and the Hippias Major, and which
says that the account of what a thing is must be unitary, if it is to be an
adequate and true account. Furthermore, the argument, at the end of the
Phaedo, for the claim that explanation and causation is based in essences
and Forms – and this is undoubtedly a central claim in Plato – relies,
crucially, on the idea that explanations must be uniform, if they are to be
genuine explanations. Yet again, we have seen that, according to the
Republic (VII. –), the quality, one or unitary, and the question
what this quality is, its essence and Form, especially as this question is
prompted by certain sensory perceptions, is involved in all judgements
about sense-perceptible things, if such judgements want to lay claim to
being subject to the principle of non-contradiction. Finally, we have seen
that the claim, in the Parmenides (), which says that Forms and the
defining of Forms is necessary for thought, relies, crucially, on just this
quality, one or unitary, and on the question what this quality is, its essence
and Form, and that it is on the basis of this quality that this claim is
articulated, both earlier in the dialogue () and later, in the second part
of the dialogue.

 Plato’s Essentialism
It may be said that what all this shows is, merely and relatively unexcit-
edly, that Plato’s commitment to essences and Forms of such qualities as
one, like, equal, beautiful, good and just is more prominent than his
commitment to essences and Forms of qualities of natural kinds, such as
humans, fire, water; it does not show that, for Plato, the account of what a
certain natural kind is, its essence and Form, must make reference to such
qualities as one, like, etc., whereas the account of what such qualities as one,
like, etc., are need not make reference to a natural kind. I respond: to show
this much, is already to have shown something substantial and significant.
For the question this raises is why Plato should afford the former Forms
greater prominence than the latter. I have defended an answer that is
prompted by the Parmenides () passage, is consonant with passages
from several dialogues before the Parmenides and is suggested and rendered
reasonable by such passages. Furthermore, we shall see presently that this
answer is confirmed in dialogues that come after the Parmenides.
Still, let us ask: Is there, before the Parmenides, strictly textual evidence
for thinking that, for Plato, the account of what a certain natural kind is,
its essence and Form, must make reference to such qualities as one, like,
etc., whereas the account of what such qualities as one, like, etc., are need
not make reference to any natural kind? Or, more simply, is there strictly
textual evidence for the claim that a natural kind depends, for what it is
and for being anything at all, on such qualities as one, like, etc., whereas
such qualities do not depend, for what they are and for being anything at
all, on natural kinds? I can think of two significant pieces of evidence. (By
strictly textual evidence, I mean evidence that requires no interpretation,
or little interpretation, to be recognised as evidence.)
First, it is expressly said, in Phaedo and Republic, that Forms – all
Forms – are uniform, unitary, one; and in Republic V. a it is said
that, even though each Form is one, it manifests itself as many as a result of
its communion (koinōnia) with other things, including other Forms. If we
put together these two statements, it is only a small step to conclude that,
for Plato and already in the Republic, each Form depends, for being a
Form, on the Form of one: by being the Form it is, it is one, and by
depending on another Form, that is, the Form of one, it is many, or, at
any rate, two. Admittedly, it is not before the Sophist, and using the
same term, ‘communion’ (koinōnia), that Plato says this, or something
that can be understood like this, or, arguably, already in the second part of
the Parmenides.
Secondly, and it seems to me most important, there is Plato’s introduc-
tion of things such as fire and snow, in the course of his account of the

The separation of Forms from physical things 
relation between essence and explanation/causation late in the Phaedo. We
may note that fire is expressly mentioned in the Parmenides () passage,
and so is snow, since water is expressly mentioned and it is obvious that
snow is congealed water. Does Plato think there are Forms of these things –
things that evidently belong to natural kinds? The Phaedo passage
tells both ways, and both ways consistently, hence we should not want
to choose between them but should want to have it both ways – just as
does the Parmenides passage. On the one hand, these things have an
essence, and an essence that contributes to why the things that are made
up of them are as they are and behave as they do – an essence that
contributes to the causation and explanation of the compounds they make
up. On the other hand, they only have such an essence because they are
suitably related to certain other things, which have an essence simply and
unqualifiedly. In other words, things such as fire and water (snow) have
an essence only in relation to, and dependent on, certain other things that
have an essence.
. How is the Parmenides () passage taken further, and

worked out, in dialogues that come after?
I shall, in the context of this study and at this late hour, treat of this
question briefly and very selectively, and not in the fullness or depth it
deserves. Save for the obvious, but practically universally overlooked,
relation between the idea of communion of Forms in the Sophist and
the statement that Forms commune with each other in Republic V. a,
on which I have already remarked at some length both here and earlier in
this study, I shall set aside the Sophist; and I shall limit myself to Timaeus
and Philebus.
In considering the Parmenides () passage, we wondered how a
quality such as one differs from a thing such as human or water; and we
noted three remarkable differences. First, the quality one or unitary is true
of all things or, at any rate, of all things that have an essence and Form.
Secondly, the quality one or unitary is true not only of individual things,
such as this human or that (mass of, or collection of ) water, but also of
wholes of such things, including the maximal whole, that is, the universe
or unqualified totality of things (to pan). Thirdly, unity is involved in the
account of what such things as human and water are, but, for Plato and in
this passage, such things are not involved in the account of unity.

This was our reading, in Chapter .

 Plato’s Essentialism
The first two differences, and especially the second, call to mind,
immediately and with some force, the Timaeus and its distinctively cosmic
perspective on things. The Timaeus, for the sake of giving an account both
of the universe and of us humans (see a–b, for a statement that this is the
dialogue’s dual aim), starts with the universe, or unqualified totality of
things (to pan), and provides an account of individual things, including us
humans, by conceiving of them as parts of the universe and appointing
them a place in it.
Plato’s account of the universe in Timaeus is exceedingly complex; not
least because he appears to distinguish two sorts of universe, or ordered,
maximal wholes and kosmoi: the kosmos made up of Forms and the
kosmos, made up of physical and bodily things. Through the idea of the
cosmic craftsman, dēmiourgos, Plato introduces the idea of a distinctive
relation between these two kosmoi. This relation, notoriously, and since
Xenokrates and Aristotle and up to Sarah Broadie, David Sedley and John
Dillon, has been understood in very different ways. However, there is, it
seems to me, two things that ought to be beyond question when we
consider the dialogue’s distinctively cosmic perspective; and they are both
directly relevant for our present question. First, the Timaeus conceives of
the notion of the (maximal) totality of things (to pan) not simply as the
collection that includes all things, but as an ordered whole, a kosmos. And
this is true both of the totality of physical things and of the totality of
Forms (and indeed of the totality comprising both together). Secondly, the
dialogue’s account of the kosmos relies, crucially, on the Form of oneness
(the quality one, or unitary) and the Forms of beauty and goodness –
these being practically understood as being ordering principles, whether in
regard to the totality of Forms or the totality of physical things.
What this shows, it seems to me, is two things. First, whatever else is
served by having Socrates, in the Parmenides () passage, commit

Some critics (such as Broadie ) argue that the fact that the Timaeus adopts a cosmic perspective
shows that it is concerned with cosmology and not with metaphysics. This is, surely, a false contrast
as far as Plato is concerned. We would hardly question that Sophist  ff. is concerned with
metaphysical questions, questions about being, what there is and what being is; and we would
hardly suggest what Plato takes himself to be concerned with there are cosmological in
contradistinction to metaphysical questions. And yet, Plato there moves back and forth between
speaking of being and beings (to on, ta onta) and speaking of all things and the totality of things (ta
panta, to pan); see: e, d, e, b, b, b, c, d, d, a.

See esp. b, where the totality of Forms is characterised as a perfect, or perfectly complete
(pantelēs), living thing.

See Dillon (); Sedley (b); Broadie (). See also Sorabji (); Johansen ();
Carone ().

See, e.g., a–b.

The separation of Forms from physical things 
himself to separate Forms only of such qualities as one, beautiful and good,
this serves to introduce, or develop further, the idea that the unity (and
beauty and goodness) that is involved in understanding what individual
things and qualities are, their essence, is also true of wholes made up of
such things and qualities, including the maximal whole, the universe,
conceived as an ordered whole, a cosmos; and it is the very same quality,
oneness or unity, in both cases. Secondly, and contra Aristotle’s position, it
is not the case that unity is primarily true of individual things, such as this
human or that (mass of, or collection of ) water, and it is true of wholes of
such things, including the maximal whole, the universe, only as a conse-
quence of being true of those individual things. If anything, unity (and
beauty and goodness) is primarily true of ordered wholes of things and,
ultimately, of the maximal, ordered whole, the cosmos; and this unity is
true just as much of essences and Forms, and of ordered wholes of essences
and Forms, as it is of physical things; and it is the very same quality,
oneness or unity, in both cases.
From this it follows, we have seen, that the essence and Form of unity
(and beauty and goodness) is separate from physical things, and this is
separation in the sense that it is not the case that the account of what unity
is must make reference to physical things. Since this Form is involved in
the account of every Form, and so every Form depends on this Form for
being what it is and being anything at all (including the Forms of physical
things such as water, fire and humans), it is wrong, Plato thinks, to say of
any Form that it is, purely and simply, non-separate or immanent, and, of
the principal and primary Forms, it is right to say with surety that they
are separate.
I think we arrive at the same result from out of the Philebus. In the
course of Socrates, Philebus and Protarchus debating the question which of
the two, pleasure or intelligence, is more conducive to a good and happy
life, Socrates has, for the purpose of advancing this debate, introduced the
distinction between two fundamental, and fundamentally different, kinds
of thing: things that have limit (peras echonta), as such and in virtue of
being what they are, and things that lack limit (apeira), as such and in
virtue of being what they are. These two kinds of thing, he has argued, are
capable of combining and, when they combine, they do so as a result of a
distinctive cause, that is, intelligence or reason. These combinations make
up physical things, to the extent that physical things belong to a universe

I want to leave open whether such a cosmic perspective is already present in Republic, noting its
reference to a kosmos of Forms (at VI. c).

 Plato’s Essentialism
governed by intelligence and reason as opposed to a universe distinguished
by randomness and chance.
Socrates makes a point of stating that it is not only individual things that
belong to these kinds, or to their combination: such as pleasure. (About
pleasure it is agreed by all, including the apparently unreformable hedo-
nist, Philebus, that, as such and in virtue of being what it is, it belongs to
the unlimited; but Socrates argues that pleasure is capable of combining
with limit while still being a kind of pleasure.) Rather, it is also wholes of
such things, including the maximal whole, the universe (to pan), that
divide into these two kinds, and into their combination.
The important thing, for the purpose of our present question, is that, from
the very start, and when, early in the dialogue, the idea of limit is introduced,
limit is associated with oneness, unity and determinate structure; and this
oneness, unity and determinate structure is appealed to not only for under-
standing physical things, whether individually or collectively, but also non-
physical things, and Forms in particular. Thus, when Forms are introduced at
a – they are referred to as non-physical things, and the Form of humans and
oxen are mentioned – they are characterised as henads (‘unities’). At the same
time, it is argued (at b) that a single henad is a determinate structure, or
order, of monads (‘unities’, or ‘unities on their own’).
We may, without hesitation, suppose that it is the same quality, oneness or
unity, that is involved in the account of individual Forms and is involved in
the combinations of limit and unlimited that make up physical things to the
extent that they belong to a universe governed by intelligence and reason. We
may conclude, just as we did in regard to the Timaeus, that the essence and
Form of unity is separate from physical things, in the sense that it is not the
case that the account of what unity is must make reference to physical things.
We may also conclude that, since this Form is involved in the account of every
Form, and so every Form depends on this Form for being what it is and being
anything at all (including the Forms of physical things such as water, fire,
humans and oxen), it is wrong, Plato thinks, to say of any Form that it is,
purely and simply, non-separate or immanent; and, of the principal and
primary Forms, it is right to say with surety that they are separate.
. Aristotle’s testimony

It is familiar that Aristotle is commonly inclined to cast Plato’s Forms and
essences as being just like his own forms and essences, except that, for

See Muniz and Rudebusch ().

The separation of Forms from physical things 
Plato, they are separate from physical things – and so Aristotle is able, to
considerable comic effect, to cast Plato’s Forms of humans and horses as
eternal and immobile humans and horses as it were frozen stiff in the
permafrost of the Platonic heaven. What this attack demonstrates is that, if
the things that have essences, or have essences principally and fundamen-
tally, are the same for Plato as for Aristotle, namely, natural kinds or things
belonging to natural kinds, then the view that the essences and Forms of
these things are separate from physical things is, nothing short of, confused
and incoherent – comical and farcical.
If we ask why this view is confused and incoherent, on the other hand,
I don’t see that we are able to articulate the sense that it is indeed confused
and incoherent, if we think of separation in modal terms and as the claim
that It is possible for X to exist without Y. The view that what a horse is can
exist without horses existing may be strange, and it is hard to think of what
use it would be to the science of biology since all that this science needs,
apparently and as Aristotle is wont of reminding us, is the view that, with
odd possible exceptions, horses give birth to horses and do so in virtue of
something in them that causes this uniform reproduction. However, if we
want to know whether this thing in individual horses, that which causes
them to reproduce uniformly, or the formula behind it, can exist without
being in individual horses, appealing to the strangeness of an affirmative
answer (its ‘queerness’), or to the foray it involves into a metaphysics that
leaves natural science behind, is hardly going to stop one or amount to
more than handwaving and slogan-slinging.
What will, at a single and simple stroke, demonstrate that it is confused
and incoherent to think that the essence and Form of a natural kind is
separate from the things belonging to this kind, is the account of separa-
tion that says that for X to be separate from Y is for the account of what
X is, its essence and Form, not having to make reference to Y. Or, this
account of separation combined with the view that the only genuine – as
opposed to dogmatic and question-begging – justification for the claim
that X cannot, or that it can, exist without Y, is the claim, provided that it
is itself properly justified, that the account of what X is must, or that it
need not, make reference to Y.
I believe this result is confirmed by Aristotle’s testimony of Plato’s
theory of Forms and the claim that Forms are separate from physical
things. What I have in mind is not, of course, his ridiculing of this claim
when, in his characteristically polemical moments, he casts Plato’s essences
and Forms as being just like his essences and forms except that they are
separate from physical things. What I have in mind first and foremost, is

 Plato’s Essentialism
his claim, made early (Gamma ), later (Iota ), and late (Lambda, Mu and
Nu) in the Metaphysics, that, just like, as he argues, the notion of being,
so too the notion of oneness or unity is a pros hen legomenon, in the sense
that it is, principally and strictly, the notion of the oneness of substances
belonging to natural kinds, such as human beings and horses, and it is,
derivatively, the oneness of properties that depend, for being what they are
and for being anything at all, on such substances. I also have in mind,
parallel and no less relevant for our present purposes, his claim, in the
Nicomachean Ethics, that there is not a single notion of goodness that is
predicable of all good things. We would expect him to think the same
about the notion of beauty (to kalon, kallos), namely, that there is not a
single notion of beauty that is predicable of all things and to think that,
when we say of mathematical entities that they are beautiful, we are not
predicating the same property of them as when we say of human actions
and characters that they are beautiful or, as translators commonly translate,
‘fine’ or ‘noble’.
There can be no doubt that the opposite view, which says that oneness
and unity is a single quality – as so too goodness and beauty – the same in
all cases and irrespective of whether these cases make reference to things in
nature, is a cause of some vexation and exasperation for Aristotle.
And there can be no doubt that he associates this view with Platonists
and Plato. But why does this view exercise him so? The answer is not
plausible, it seems to me, which says that he is committed to rejecting it by
his own theory of categories. First, what the theory of categories implies, if
anything, is that being is a pros hen legomenon in the sense stated; it is a
further question whether the same is true of unity. Secondly, it is
debatable to what extent Aristotle simply takes over, in the Metaphysics,
the theory of categories; for one thing, he wants to engage, without
begging the question against them, all those thinkers who are not com-
mitted to this theory, and have never heard of it, be they Platonists or
natural philosophers.
I want to propose what seems to me an eminently sensible answer.
Aristotle knows that, unless he opposes from the start the view that says
that oneness and unity is a single quality – as so too goodness and beauty –
the same in all cases, he will not be able to resist the view that the account
of what unity is (or what goodness is; or what beauty is) need not make
reference to physical things that are unitary (or good; or beautiful). This is

See Politis and Steinkrüger ().

The separation of Forms from physical things 
because he, like Plato, thinks that non-physical things, too, can correctly
be said to be unitary (and good; and beautiful). Aristotle knows that,
unless he is able to resist this view, his opposition will be baseless, against
the view that essences and Forms, or the essences and Forms of those
things that have essences and Forms principally and strictly, are separate
from physical things.

 
What yokes together mind and world?

Phaedo – and Republic VI. –
In Phaedo d–a, Plato argues, apparently quite generally and in regard

to any thing, that we ought to investigate things, not directly through the
senses, but through logoi, ‘statements’. At the same time, he makes a point
of insisting that to investigate things through logoi is not to investigate
them indirectly, that is, to investigate them only in and through our
representations and theories of them, any more than to investigate a thing
by looking at it is to investigate it only in and through an image of it.
I want, by way of moving towards a conclusion to this study, to return to
this methodological remark, which, it seems to me, occupies a special place
among all of Plato’s methodological remarks. The outstanding question is:
Why is to investigate things through our logoi of them to investigate the
things themselves and not just things as represented in our statements,
accounts and theories of them? Does Plato anywhere indicate an answer to
this question?
This question, I think it will be admitted, is of very great interest and
significance. It is also very relevant for indicating what seems to me to be
the right response to a worry that I have heard expressed in reaction to the
account of the theory of Forms that I am defending in this study. The
worry I have heard says that, to have Plato’s theory be rooted, to such an
extent as I have, in the ti esti question is to play down, quite unacceptably,
the fact that Plato’s theory is a metaphysical theory: an account of the
things themselves, not of our questions and answers about them. This
worry can be dramatised in a variety of ways. It may be cast by saying that
I have failed to take account of Plato’s metaphysical realism, established
beyond doubt by Myles Burnyeat in . Indeed, it may be said that,

See Chapter , where I also quoted this passage.

I would like to note, and acknowledge in gratitude – he was a mentor and friend – David Evans’
excellent  paper, in which he argues, incisively, that those critics who urge that Plato is a realist
fail to take account of the many and important arguments in Plato in which he derives a conclusion


Forms, mind and reality 
not only have I failed to take account of Plato’s metaphysical realism,
I have been moving in the direction of the dark days of Paul Natorp, who
argued, in the fashion typical of a transcendental idealist, that Plato’s
theory is primarily about our thought and knowledge of things and only
as a consequence is it a theory about the things themselves. Or to put it
simply, it may be said, I have been turning Plato into an early Michael
Dummett, who thinks metaphysics is based in logic.
My response will not come as a surprise. Most certainly, the theory of
Forms is about the things themselves, not only or primarily about things as
they show up in our questions and answers about them. But, in Plato, and
quite deliberately, consciously and explicitly, there is not an opposition, or
even mere tension, between investigating the things themselves and inves-
tigating things in and through our questions and answers about them. This
is what the Phaedo passage says. Perhaps this is Plato’s deepest indebted-
ness to Parmenides, as we know Parmenides from fragment , when he
claims that being and thinking are co-ordinate – neither is superordinate,
or subordinate, to either – and unbreakably linked.
I want to mobilise the Sun Analogy in the interest of arriving at the same
result through it. The crucial lines in the Sun Analogy, because they say just
what the function of the Good is, are e–: ‘So that what gives truth to
the things known and the power (dunamis) to know to the knower is the
idea of the good. And though it is the cause (aitia) of knowledge and
truth . . .’ (trans. Grube). If we observe also what immediately follows, and
especially when it is said that what the Good provides for the things is not
only their being capable of being known, but their very being and essence (to
einai te kai tēn ousian, b–), it is apparent that the Good is charac-
terised here as just this: the single cause (aitia) of, on the one hand, the soul’s
ability (dunamis) to think of and to know things, and, on the other hand,
the things’ ability (dunamis) to be thought of and known, and indeed their
being (cf. to einai) and essence (if we read hē ousia here as ‘the essence’).
What this shows is that a central aim in the Sun Analogy is to link up, as
we would say, mind and world, and to do so by appealing to a single
about beings or being from a premise about thinking and knowing. For, as Evans reminds us, such
argument ought not to be available to a realist, unless, that is, he is confused.

Unfortunately, the Slings edition, whose text of these lines is the same as the Burnet edition, has
these lines as d–e. I shall follow the Burnet edition.

The reading of ‘essence’ for ousia here is not crucial to my argument. But we may note that the ‘ho
estin’ at b is naturally understood as ‘that which it is’ (and not: ‘that which is’), and hence as a
way of referring to the essence of a thing or quality.

See also d, outō toinun kai to tēs psuchēs hōde noei, ‘Thus then also in regard of that in virtue of
which the soul thinks’.

 Plato’s Essentialism
ultimate explanation and cause, and positing that this is the explanation
and cause of, at once and on the one hand, our ability to think of and to
know things, and, on the other hand, the things themselves, including
their being, their essence and their being such as to be capable of
being known.
Just how important this aim of the Sun Analogy is, is confirmed,
I think, by some remarkable features of the way in which Plato formulates
this analogy and its structure.
In a sentence of hardly three lines (in which are used the words ‘a yoke’
[ho zugos], ‘to yoke’ [zeugnumi] and ‘a yoking together’ [hē suzeugsis]), light
(phōs) – light, in the analogy, stands for the causal power of the Good – is
afforded the task of yoking together the sense of sight (hē tou horan aisthēsis)
and the ability of things to be seen (hē tou horasthai dunamis; e–a).
This, in the analogy, stands for, on the one hand, the ability of the soul to
think of and to know things, and, on the other hand, the ability of things to
be thought of and to be known. We really have to think of this as a
figurative image, with two animals straddled by a third and very different
thing, a yoke, which yokes them firmly and unbreakably together.
Secondly, and, I think, most remarkably, light, which, in the analogy,
stands for the causal power the Good, is referred to as a triton genos, a ‘third
kind’, in addition, that is, to the two kinds of thing: the soul and reality –
mind and world. First (d) it is said that light is a ‘third thing’ (triton)
that is necessary, in addition to a soul with the sense of sight and to things
capable of being seen, if there is to be actual seeing and being seen. But,
shortly after (e), this same light, and hence the causal power of the
Good, is referred to as a genos triton.
A plain consequence of this is that (with the possible exception of the
Good) Plato in no way affords priority to things – reality, the world – over
our ability to think of and to know things – the mind, the soul. Nor, indeed,
does he afford priority to our ability to think of and to know things over
those things. This follows, plainly, because he subjects both, and both
equally, to a third thing, indeed a third kind of thing, whose function it is
to link these things perfectly together. Compared to, and in relation to, this
third kind, the Good and its causal power – the power that, in the analogy,
is represented by the light of the sun – these two other kinds, that is, mind
and world, are related by a relation that presents no asymmetrical aspect.

It is interesting to note that Natorp, who is famous for arguing that, for Plato, thinking is prior to
being, and who dedicates a -page work (which he published in ) to arguing this, did, some
twenty years later, for the second edition, and not long before he died, recant, and in suitably

Forms, mind and reality 
There are, no doubt, those critics who will cry out that this relation,
between mind and world in Plato’s Sun Analogy, is plainly asymmetrical,
since, after all, the Form of the Good belongs to the world of Forms, in
which it occupies a special and supreme position. The explanatory and
causal relation, which explains our ability to think of and to know things,
they conclude, goes from reality to mind – as we would expect of a true
metaphysical realist (they say).
I wonder, how can a yoke that serves to yoke together two animals really
be the head of the one animal? Whatever that strange tethering device
might be called (using the attached head of one animal to yoke together
another animal), it cannot rightfully be called a yoke (zugos). Perhaps it will
be said that this is to trade overly in images and words. But what about the
term genos triton (e)? How, without plain inconsistency and confu-
sion, can one count a part, even the head, of the one of two animals, which
are referred to as the first kind and the second kind, as ‘a third kind’? Or, to
step out of the analogy, how, without plain inconsistency and confusion,
can one begin by counting the soul and its ability to think and to know, as
the first kind (or, as the second kind), and the world of Forms as the second
kind (or, as the first kind), and then refer to one of those Forms, the
topmost one, as ‘a third kind’? One might as well count as follows:
inanimate things is one kind, animate things is a second kind and The
Lion is a third kind. Such a count would issue in indefinitely many kinds,
whereas the result was ‘three’. Or are we to suppose that, when Plato uses
the term genos triton at e, he is speaking loosely and carelessly and
does not mean genos at all? Or that he does not mean just those three
things, and kinds of things: soul, world, the Good? That looks to me very
like special pleading. Such pleading might, just about, be tolerable, if the
only alternative was to have Plato be a metaphysical anti-realist or tran-
scendental idealist. But, if the alternative is, simply and plainly, to recog-
nise that (with the possible exception of the Good) Plato does not privilege
either reality over mind or mind over reality, and this because he subjects
both, and both equally, to a third kind of thing, which serves to link the
two perfectly together, what can possibly motivate or justify such pleading?
This, it is true, leaves us with the question of the status – metaphysical,
epistemological, cosmological, teleological, ethical, political, aesthetic – of
the Good in the total scheme of things. How can the Good be a third kind
elevated, declamatory and near hierophantic style: he argues, in the Metakritischer Anhang, which is
the appendix to the second edition of Platos Ideenlehre, for something like the account, of Plato’s
view of the relation between being and thinking, that I am defending here.

 Plato’s Essentialism
of thing, in addition to rational souls and the things they think of and
know, the real things? And how can it be a third kind of thing, if it is itself
characterised as a Form?
The former question is not for this occasion, for it is far too large and
difficult for this late hour. Plato makes a point of saying, in preparation for
the Sun Analogy, that he cannot address this question properly, but only
through an analogy; and I have no desire to get any further – any further
than what I have already noted, namely, that the Good is referred to as a
triton genos – into the vexed question of what he means by saying that the
Good is beyond being (or, beyond essence, epekeina tēs ousias) in status and
power (presbeia[i] kai dunamei; b–).
As for the latter question, the fact is that he does not refer to the Good as
a Form (eidos). What he does is use the expression hē tou agathou idea
(e–), which is also the expression he used when he first introduced
the matter of the greatest object of learning (to megiston mathēma, a).
It seems to me a genuine question whether, when he first uses this
expression, at a, he means ‘the Form of the Good’ or, instead, he
simply means ‘that very quality, the Good’; and, if there is this question of
the expression as used at a, there is room for this question of the
expression as used at e–. However, even conceding that the Good is
referred to as an idea in the sense of a Form (eidos), why does this show that
the Good belongs to the kind (or class, or set) that includes all Forms? It
seems to me to be perfectly compatible with reading the plain text,
according to which the Good (or the idea of the Good or the Form of
the Good) does not belong to the kind that includes all Forms, any more
than it belongs to the kind that includes rational souls. For it is expressly
said to be: ‘a third kind of thing’ (triton genos, e).
The Sun Analogy, I believe, is Plato’s answer to the outstanding and
most pressing question: Why is to investigate things through our logoi of
them to investigate the things themselves and not just things as represented
in our statements, accounts and theories of them? The answer, on the Sun
Analogy, is ready to hand: Because there is a perfect fit between our
statements, accounts and theories about things, at their best, and the
things themselves, which is what our statements, accounts and theories
are about. And the reason why there is this perfect fit is that there is a third
thing, and kind of thing, in addition to and distinct from, on the one
hand, the soul, which is the place of statements, accounts and theories,
and, on the other hand, the things that these acts of the soul are about.
This third thing is the explanation and cause of, at once and on the one
hand, the soul’s ability to think about and to know things through

Forms, mind and reality 
statements, accounts theories, and, on the other hand, those very things’
being and being what they are and being such that they are capable of
being thought of and known.
. Addendum
I would like to make, loud and clear, the following clarificatory points,
which serve to make clear what my account in this chapter of Plato’s Sun
Analogy does not imply and what I am not committed to in and through
this account.
First, I am not, in and through this account of Plato’s Sun Analogy,
arguing that Plato’s account of the relation between mind and world is
anti-realist, in the sense of asserting the priority of mind over world.
Secondly, I am not, in and through this account of Plato’s Sun Analogy,
arguing that Plato thinks that the Good is beyond knowledge.
Thirdly, I am not, in and through this account of Plato’s Sun Analogy,
arguing that Plato thinks that the Good is beyond being or
beyond essence.
Fourthly, yes, I admit, the account I have defended of Plato’s Sun
Analogy would benefit from being complemented by our taking up and
investigating the following questions (which I have not posed in this
chapter): Does Plato, in the Sun Analogy, distinguish between the sun
and the light of the sun, and, correspondingly, between the Good and the
power of the Good? If he does make this distinction, what role, if any, does
it play in his argument in the Sun Analogy?
Taking up these questions would require a full interpretation of the Sun
Analogy, which I have not attempted. However, from what I can see, the
crux claim in my account in this chapter of Plato’s Sun Analogy does not
depend on taking a particular, and potentially disputable, stand on
these questions.

I am grateful to Daniel Vazquez for advising me to do something like this.

Conclusion
Forms simply are essences, not things that have essences
This is the conclusion of the present study: that Plato’s Forms simply are
essences, not things that have essences. For this is what is established, if it is
established that Plato’s Forms are, basically, that which is designated by an
adequate and true answer to a ti esti question, and if it is established that
the other principal characteristics of Forms – Forms are changeless, uni-
form, not perceptible by the senses, knowable only by reasoning, the basis
of causation and explanation, distinct from sense-perceptible things, nec-
essary for thought and speech, separate from physical things and more –
are based, for their justification, in one way or another, on this basic
identity of Forms.
This contradicts a tradition of interpretation, dominant among many
critics for some time now, according to which, up until the second part of
the Parmenides, when Plato came to think better of it, and before he drove
the point home in a single line in the Sophist (c–), he thought of
Forms as things that have properties, and have properties in just the way in
which any thing has properties. Have we provided an alternative to this
view? Do we not still have to refute the basis for it, which is the view that,
up until the second part of the Parmenides and driven home in the Sophist

Perhaps especially since Vlastos’  paper. As Malcolm (, ), notes, the term ‘self-predication’
was introduced by Vlastos in that paper.

‘But I think you agree that of the things that are, some are spoken of in and by themselves (auta kath’
hauta), while others are always spoken of in relation to others (pros alla)’ (Sophist c–; trans.
Rowe). Remarkably, M. Frede (, , emphasis added) says that the distinction made in this
statement in the Sophist, between that which is kath’ hauto and that which is pros allo, ‘in no way’
occurs here for the first time in Plato: ‘Diese beiden Formen [i.e., to kath’ hauto and to pros allo,
referred to at Sophist d] . . . tauchen übrigens im Sophistes keineswegs zum erstenmal bei Platon
auf. Im Charmides nämlich heißt es (a–) . . ..’ If only Frede had followed up on this
remarkable statement! – it would have saved us so much trouble! This directly implies that
Michael Frede cannot be saying what M. L. Gill thinks he is saying. She thinks, and, as indeed
does this whole tradition, she attributes to Michael Frede the view that the Sophist claim goes back
only to the Parmenides and it contradicts and corrects what Plato thought before the Parmenides
(I am supposing the Charmides is before the Parmenides).


Conclusion 
line, Plato thought that the Form of the quality F is itself F, and is F in the
same way as other things are F, that is, refute the spectre of self-
predication?
I submit that we have already done this, in two ways, and thoroughly
enough. First, and as some critics committed to this traditional view have
recognised, a crucial piece of evidence for thinking that, in such dialogues
as Phaedo and Republic, Plato is committed to self-predication (in the sense
just stated), is the view that he is committed to the transmission theory of
causation. But I have argued that, in the one passage in which these critics
have found this theory of causation, namely, the account of causation in
the Phaedo (e ff.), it is not there and it is not required for Plato’s
argument.
Secondly, and all-important, by my having argued that, from the start,
Plato’s Forms are, basically, that which is designated by an adequate and
true answer to a ti esti question and that the other principal characteristics
of Forms are based on and derived from this basic identity of Forms, we
have shown that it is not credible to think that Plato was ever committed
to the view that the Form of the quality F is itself F and is F in the same way
as other things are F.
Suppose a philosopher thinks that things divide into two kinds: things
that simply are essences of certain qualities and things that are what they
are only (or, ‘always’ as used in Sophist c–) in relation to those
essences. This, I have argued, is what Plato thinks from the moment he
introduces Forms, or essences – and, if I may, it is what the Sophist line
sums up. It is evident that, in making this division, the philosopher is
already thinking that a thing that is not the essence of the quality, F, if it is
F, is F in relation to another thing (pros allo), namely, the essence of
F. Perhaps it never occurs to this philosopher that the essence of the
quality, F, is itself F. But, if she thinks the essence of the quality, F, is
itself F, how can she, without immediately, evidently and unmistakably
contradicting herself, think that it is F in the same way as other things are
F (the other things are F pros allo)? For her view is, precisely, that it is
things other than the essence of the quality, F, that are F pros allo. She will,
therefore, immediately want to mark that, if the essence of the quality F is
itself F, it is not F pros allo. If Greek is her language, she may mark this by

See M. L. Gill: ‘In my view the self-predication assumption goes hand-in-hand with a view of
causation, sometimes called “the transmission theory of causation”’ (, ).

See Chapter .

 Plato’s Essentialism
saying that the essence of the quality, F, is F kath’ hauto or auto kath’ hauto
(the Latin per se, English ‘by itself’ or ‘itself by itself’).
Do we not still have to go through the passages in which Plato appears
to be committed to the view that the Form of the quality F is itself F, and is
F in the same way as other things are F? The problem, it seems to me, is to
find such passages, and to do so without relying on the (putative) evidence
of the transmission theory of causation. The third-large regress passage in
the Parmenides (a–b) may be thought to be such a passage, but only if
there is reason to think that, when Plato spells out this regress, which
appears to rely on self-predication in the sense stated, he is implying that
he was himself committed to such self-predication before he came to think
better of it – I doubt that such reason can be found.
If we set aside the (putative) evidence of the third-large regress passage
in the Parmenides, and of the transmission theory of causation in the
Phaedo, what passages are there to appeal to in support of the view that,
before the second part of the Parmenides, Plato was committed to self-
predication in the sense stated? To be sure, we can find many passages,
apparently before the second part of the Parmenides, in which Plato says
that the Form F cannot be contrary-to-F. But whoever could have thought
that a statement of the form ‘X is F’ can be validly inferred from a
statement of the form ‘X cannot be both F and contrary-to-F’? And even
if we can find one, or possibly two, passages in which Plato says, or implies,
that the Form F is F, whoever could have thought that it follows that the
Form F is F in the same way as other things are F?
I come back to the all-important point, which is that, by having argued
that, from the very start, Plato’s Forms simply are essences, we have shown
that it is not credible to think that Plato was ever committed to the view

Gill (, ) appeals to Symposium e–a as a clear case of self-predication; but what that
passage says is that the Form of beauty is not both beautiful and not beautiful, or ugly, in the variety
of ways in which sense-perceptible things are so.

One such passage is Phaedo c–: ‘Consider, then, he said, whether you share my opinion as to
what follows, for I think that, if there is anything beautiful, excepting (plēn, often translated here
with “besides”) the Beautiful itself (auto to kalon), it is beautiful for no other reason than that it
shares in that Beautiful (ekeino to kalon), and I say so with everything. Do you agree to this sort of
cause?’ Harte (, ) comments: ‘Since this passage assumes that the Beautiful itself is
beautiful . . . we have here a pretty clear statement of self-predication in what looks to be a sample
case: the Form, the Beautiful itself, is beautiful.’ I am not sure this passage is such a clear case, if we
take plēn in the sense of ‘excepting’, not in the sense of ‘besides’ or ‘in addition to’. But I shall not
press this point. Harte (, ), when she considers the question of self-predication, cites but this
passage, Phaedo c; and she is careful not to commit herself to the presence in it of self-
predication, for she sums up her examination with: ‘If there are good grounds for supposing that
Forms self-predicate . . .’ (, emphasis added).

Conclusion 
that the Form of the quality F is itself F and is F in the same way as other
things are F.
This point, I want to make loud and clear, is not to be confused with,
and it does not rely on, a certain view, which goes back to Meinwald
() and which has since been defended again by Silverman () and
M. L. Gill (). (That view does not, contra the intellectual autobiog-
raphy of these critics, go back to Michael Frede’s classic book in German,
Prädikation und Existenzaussage.) The Meinwald-Silverman-Gill view says
that when, in the second part of the Parmenides, and driven home in the
Sophist line, Plato distinguishes things that are kath’ hauta and things that
are pros alla, this is the distinction between two kinds of predication, that is,
two ways in which X can be Y, or two uses of ‘. . . is . . .’: X is essentially Y;
and X is non-essentially Y. As Meinwald herself recognises and under-
scores, the following is clearly false: If it is established that Plato distin-
guishes these two kinds of predication, it follows that the things that he
thinks are essentially F simply are essences. This follows no more than it
follows from the supposition that my dog, Bobo, is essentially a dog that
he, Bobo, simply is an essence. Those critics (Meinwald, Silverman, Gill
and others – I suspect all this goes back to Gwil Owen) who argue that the
distinction in the Sophist line and in the Parmenides is, purely and simply,
that between two kinds of predication, have been at pains to point out that
there is no implication that a Form is identical with its essence. Indeed,
they have combined the view that Plato distinguishes these two kinds of
predication in Parmenides and Sophist with the denial that Plato ever
identified a Form with its essence.
Let us not end with what may give the impression of being no more
than another fine scholarly dispute among critics. Something much larger
is at stake. The question is: Why have critics been so ready to suppose that

It is commonly thought that this view goes back to Michael Frede (). But this is far from clear.
In his  paper (–), in which he takes himself to be re-stating his  view of Plato’s
account of being in the Sophist (see esp. , –), he appears to dissociate himself from this
understanding of his view. Gill (, ) recognises this; but, incredibly, she refuses to take Frede
at his word, when he dissociates himself from this understanding of his own view, and she says that,
in so dissociating himself, ‘He misrepresents what he has just said.’

Meinwald  is careful not to commit herself to any particular account of the relation between a
Form and its nature (‘nature’ is her term for my ‘essence’). All she is prepared to say is that ‘there
must be a connection between forms and natures so close that a nature cannot remain when the
associated form is taken away’ (; also ). Notably, she says that ‘A straightforward way of
ensuring this would be to take natures to be in some way components of forms’ (). This is
notable not least because, if essences/natures are components of Forms, then it is not the case that
Forms are essences. Silverman (), too, denies that a Form is identical with its essence and that
Forms are essences; he speaks of Forms as being ‘bearers’ of essence.

 Plato’s Essentialism
Plato’s Forms are things that have essences, and why have they not
considered it, from the start, an open question whether Plato’s Forms
may not, rather, simply be essences? The answer, apparently, is that they
have, from the start, been thinking of the whole issue of essentialism in
Aristotelian terms: there are things (which, manifestly, need not be
essences), and they have some qualities essentially and other qualities
non-essentially. If the argument of this study has been on the right lines,
Plato’s essentialism is altogether different; it is the view that there are two
kinds of things: essences and things that are suitably related to essences.
Shall we, then, blame Aristotle for misrepresenting Plato? Aristotle may be
responsible for the fact that we today have difficulty of thinking of essentialism
except in Aristotelian terms. Is he to blame for thinking that Plato’s Forms are
just like his essences, except that they are separate from physical things? Is this
not, according to the argument of the present study, a bad misrepresentation
of Plato? It is, but I wonder whether it is Aristotle’s considered representation
of Plato’s Forms. True, Aristotle is wont to challenging Platonists to counter
this representation, which he presents to them with a touch of good-
humoured caricature. However, in his less polemical and more probing
moments, Aristotle gives an altogether different account of Plato’s Forms.
In chapter six of book Zeta of the Metaphysics, Aristotle says, as clearly as
one could wish for, that, for Plato or a Platonist, those things that have an
essence in the strict sense – which, for Plato and Platonists, are the Forms – are
identical with their essence. He appears to conclude that he, too, Aristotle, no
less than Plato, is committed to the view that, if a thing has an essence in the
strict sense, then it is identical with its essence; except that, he, Aristotle, does
not think that the things that have an essence in the strict sense are Platonic
Forms, and he wants to argue that physical things, too, have an essence in the
strict sense. Most remarkable, Aristotle candidly indicates that, in his view, a
Platonist, who denies that physical things have an essence, has an easier time
satisfying the thesis that a thing that in the strict sense has an essence is
identical with its essence, than does he, Aristotle.
I am at a loss to decide what is responsible for the fact that critics have
been so ready to suppose that Plato’s Forms are things that have essences,
and why they have not considered it, from the start, an open question
whether Plato’s Forms may not, rather, simply be essences. Be that as it
may, I hope I have done something to redress this sorry state of affairs.

Most clearly at b–.

Bibliography
Allen, R. E. () Plato’s Parmenides, London: Yale University Press.

Anderson, A. () ‘Socratic Reasoning in the Euthyphro’, The Review of
Metaphysics, (), –.
Annas, J. () An Introduction to Plato’s Republic, Oxford: Clarendon Press.
() ‘Aristotle on Inefficient Causes’, The Philosophical Quarterly, (),
–.
() ‘Classical Greek Philosophy’, in J. Boardman, J. Griffin and O. Murray,
eds., The Oxford History of the Classical World, Oxford: Oxford University
Press, –.
Annas, J., and Rowe, C. J. () New Perspectives on Plato, Modern and Ancient,
Centre for Hellenic Studies Colloquia , Cambridge, MA: Harvard
University Press.
Apelt, O. () Platonische Aufsätze, Berlin: B.G. Teubner Verlag.
Archer-Hind, R. D. () The Phaedo of Plato, nd ed., London: Macmillan.
Balaudé, J.-F. () Hippias Mineur, Hippias Majeur, Paris: Livre de Poche.
Bambrough, R. () ‘The Disunity of Plato’s Thought’, Philosophy, (),
–.
Bestor, T. W. () ‘Plato’s Semantics and Plato’s Cave’, Oxford Studies in
Ancient Philosophy, , –.
Bluck, R. S. () Plato’s Phaedo, London: Routledge & Kegan Paul.
() ‘Logos and Forms in Plato: A Reply to Professor Cross’, Mind, (),
–.
Bostock, D. () Plato’s Phaedo, Oxford: Clarendon Press.
Brisson, L. () ‘“Is the World One?” A New Interpretation of Plato’s
Parmenides’, Oxford Studies in Ancient Philosophy, , –.
Broadie, S. () Nature and Divinity in Plato’s Timaeus, Cambridge:
Cambridge University Press.
Burnyeat, M. F. () ‘Plato on the Grammar of Perceiving’, The Classical
Quarterly, (), –.
() ‘Idealism and Greek Philosophy: What Descartes Saw and Berkeley
Missed’, The Philosophical Review, (), –.
([] ) ‘Platonism and Mathematics: A Prelude to Discussion’, in A.
Graeser, ed., Mathematics and Metaphysics in Aristotle, Proceedings of the


 Bibliography
Tenth Symposium Aristotelicum, Bern: P. Haupt, –. Reprinted in
M. F. Burnyeat, Explorations in Ancient and Modern Philosophy, vol. ,
Cambridge: Cambridge University Press, –.
Burnyeat, M. F., and Frede, M. (a) The Pseudo-Platonic Seventh Letter, D.
Scott, ed., Oxford: Oxford University Press.
(b) The Seventh Platonic Letter: A Seminar, D. Scott, ed., Oxford: Oxford
University Press.
Carone, G. R. () Plato’s Cosmology and Its Ethical Dimensions, New York:
Castagnoli, L. () Ancient Self-Refutation: The Logic and History of the Self-
Refutation Argument from Democritus to Augustine, Cambridge: Cambridge
University Press.
Code, A. () ‘Aristotle: Essence and Accident’, in R. Warner and R. E.
Grandy, eds., Philosophical Grounds of Rationality, Oxford: Oxford
University Press, –.
Cooper, J. M. () ‘Plato on Sense-Perception and Knowledge (Theaetetus
–)’, Phronesis, (), –.
Cornford, F. M. () Plato’s Theory of Knowledge: The Theaetetus and Sophist of
Plato, London: Routledge & Kegan Paul.
() Plato and Parmenides, London: Kegan Paul, Trench, Trubner & Co.
Crivelli, P. () ‘Plato’s Philosophy of Language’, in G. Fine, ed., The Oxford
Handbook of Plato, Oxford: Oxford University Press, –.
Crombie, I. M. ([] ) An Examination of Plato’s Doctrines: II. Plato on
Knowledge and Reality, London: Routledge.
Cross, R. C. () ‘Logos and Forms in Plato’, Mind, (), –.
Dancy, R. M. () Plato’s Introduction of Forms, Cambridge: Cambridge
University Press.
Dillon, J. M. () ‘The Riddle of the Timaeus: Is Plato Sowing Clues?’, in M.
Joyal, ed., Studies in Plato and the Platonic Tradition: Essays Presented to John
Whittaker, Abingdon: Routledge, –.
() The Roots of Platonism: The Origins and Chief Features of a Philosophical
Tradition, Cambridge: Cambridge University Press.
Dodds, E. R. () ‘The Parmenides of Plato and the Origin of the Neoplatonic
“One”’, The Classical Quarterly, , –.
Evans, D. () ‘Platonic Arguments: I’, Proceedings of the Aristotelian Society,
Supplementary Volumes, , –.
Ferejohn, M. () ‘Knowledge and the Forms in Plato’, in H. H. Benson, ed.,
A Companion to Plato, Malden: Blackwell, –.
Fine, G. ([] ) ‘Knowledge and Belief in Republic V’, Archiv für
Geschichte der Philosophie, (), –. Reprinted in G. Fine, Plato on
Knowledge and Forms: Selected Essays, Oxford: Clarendon Press, –.
() ‘Forms As Causes: Plato and Aristotle’, in A. Graeser, ed., Mathematics
and Metaphysics in Aristotle, Bern: P. Haupt, –.
([] ) ‘Knowledge and Belief in Republic V–VII’, in S. Everson, ed.,
Epistemology, Companions to Ancient Thought, vol. , Cambridge:

Bibliography 
Cambridge University Press. Reprinted in G. Fine, Plato on Knowledge and
Forms: Selected Essays, Oxford: Clarendon Press, –.
() ‘Inquiry in the Meno’, in R. Kraut, ed., The Cambridge Companion to
Plato, Cambridge: Cambridge University Press, –.
() On Ideas: Aristotle’s Criticism of Plato’s Theory of Forms, Oxford:
Clarendon Press.
() ‘Does Socrates Claim to Know That He Knows Nothing?’, Oxford
Studies in Ancient Philosophy, , –.
() The Possibility of Inquiry: Meno’s Paradox from Socrates to Sextus, Oxford:
Oxford University Press.
() ‘Plato on the Grades of Perception: Theaetetus – and the
Phaedo’, Oxford Studies in Ancient Philosophy, , –.
Fine, K. () ‘Essence and Modality’, Philosophical Perspectives, , –.
Frede, M. () Prädikation und Existenzaussage: Platons Gebrauch von “. . .
ist . . .” und “. . . ist Nicht . . .” im Sophistes, Hypomnemata , Göttingen:
Vandenhoeck und Ruprecht.
() ‘The Original Notion of Cause’, in M. Schofield, M. Burnyeat and J.
Barnes, eds., Doubt and Dogmatism: Studies in Hellenistic Epistemology,
Oxford: Clarendon Press, –.
() ‘Plato’s Sophist on False Statements’, in R. Kraut, ed., The Cambridge
Companion to Plato, Cambridge: Cambridge University Press, –.
Frege, G. () ‘Über Sinn und Bedeutung’, Zeitschrift für Philosophie und
Philosophische Kritik, N.F. , –.
Gadamer, H.-G. () Gesammelte Werke, Band : Griechische Philosophie I,
Tübingen: J. C. B. Mohr.
() Gesammelte Werke, Band : Griechische Philosophie III: Plato Im Dialog,
Tübingen: J. C. B. Mohr.
Gallop, D. () Plato Phaedo, Oxford: Clarendon Press.
Geach, P. T. () ‘Plato’s Euthyphro: An Analysis and Commentary’, Monist,
, –.
Gerson, L. P. () Platonism and Naturalism, Ithaca: Cornell University Press.
Gill, M. L. () Philosophos: Plato’s Missing Dialogue, Oxford: Oxford
University Press.
Gosling, J. C. B. () Plato. London: Routledge & Kegan Paul.
Gould, C. S. () ‘Grube’s Translation of Republic a: A Response to
Waterfield’, Ancient Philosophy, (), –.
Greene, B. () The Elegant Universe: Superstrings, Hidden Dimensions, and the
Quest for the Ultimate Theory, New York: W. W. Norton.
Hackforth, R. () ‘Plato’s Republic’ [review of Murphy ], The Classical
Review, N.S. (/), –.
() Plato’s Phaedo, Cambridge: Cambridge University Press.
Harte, V. () ‘Plato’s Metaphysics’, in G. Fine, ed., The Oxford Handbook of
Plato, Oxford: Oxford University Press, –.
Hamlyn, D. W. () ‘The Communion of Forms and the Development of
Plato’s Logic’, The Philosophical Quarterly, (), –.

 Bibliography
Hestir, B. E. () Plato on the Metaphysical Foundation of Meaning and Truth,
Cambridge: Cambridge University Press.
Hintikka, J. () Socratic Epistemology, Cambridge: Cambridge University
Press.
Horan, D. () ‘Plotinus’ Use of the Third Hypothesis of Plato’s Parmenides in
Enneads: VI..’, Dionysius, N.S. , –.
Irwin, T. H. () Plato’s Moral Theory: The Early and Middle Dialogues,
Oxford: Oxford University Press.
Johansen, T. K. () Plato’s Natural Philosophy, Cambridge: Cambridge
University Press.
Kahn, C. H. () ‘The Beautiful and the Genuine’, Oxford Studies in Ancient
Philosophy, , –.
() ‘Vlastos’s Socrates’, Phronesis, (), –.
Karamanolis, G., and Politis, V. () The Aporetic Tradition in Ancient
Philosophy, Cambridge: Cambridge University Press.
Karbowski, J. () ‘Justification “by Argument” in Aristotle’s Natural Science’,
Oxford Studies in Ancient Philosophy, , –.
Keyt, D. () ‘Plato’s Paradox That the Immutable Is Unknowable’, The
Philosophical Quarterly, (), –.
Kirwan, C. () ‘Plato and Relativity’, Phronesis, (), –.
Kutschera, F. von () Platons Parmenides, Berlin: Walter de Gruyter.
Larsen, P. D. () ‘Are There Forms of Sensible Qualities in Plato?’, Journal of
the American Philosophical Association, (), –.
Lachance, G. () ‘Platonic Contrariety (enantia): Ancestor of the Aristotelian
Notion of Contradiction (antiphasis)?’, Logica Universalis, , –.
Ledbetter, G. M. () ‘Reasons and Causes in Plato’, Ancient Philosophy, (),
–.
Lee, D. C. (a) ‘Dialectic and Disagreement in the Hippias Major’, Oxford
Studies in Ancient Philosophy, , –.
(b) ‘Interpreting Plato’s Republic: Knowledge and Belief’, Philosophy
Compass, (), –.
() ‘Drama, Dogmatism, and the “Equals” Argument in Plato’s Phaedo’,
Oxford Studies in Ancient Philosophy, , –.
Lorenz, H. () The Brute Within: Appetitive Desire in Plato and Aristotle,
Oxford: Clarendon Press.
Ludlam, I. () Hippias Major: An Interpretation, Stuttgart: F. Steiner.
Malcolm, J. () ‘On the Place of the Hippias Major in the Development of
Plato’s Thought’, Archiv für Geschichte der Philosophie, (), –.
() Plato on the Self-Predication of Forms: Early and Middle Dialogues,
Oxford: Clarendon Press.
Mann, W.-R. () The Discovery of Things: Aristotle’s Categories and Their
Context, Princeton: Princeton University Press.
McCabe, M. M. () Plato’s Individuals, Princeton: Princeton University Press.
‘The Unity of Virtue: Plato’s Models of Philosophy’, Aristotelian Society,
supplementary volume XC, –.

Bibliography 
Meinwald, C. C. () Plato’s Parmenides, Oxford: Oxford University Press.
Moravcsik, J. M. E. () ‘The Anatomy of Plato’s Divisions’, in E. N. Lee, A.
P. D. Mourelatos and R. M. Rorty, eds., Exegesis and Argument: Studies in
Greek Philosophy Presented to Gregory Vlastos, Assen: Van Gorcum, –.
Moss, J. () ‘Appearances and Calculations: Plato’s Division of the Soul’,
Oxford Studies in Ancient Philosophy, , –.
Muniz, F., and Rudebusch, G. () ‘Plato, Philebus B: A Problem Solved’,
The Classical Quarterly, N.S. (), –.
Murphy, N. R. () The Interpretation of Plato’s Republic, Oxford: Clarendon
Press.
Natorp, P. () Platos Ideenlehre: Eine Einführung in den Idealismus, Leipzig:
Meiner.
() Plato’s Theory of Ideas: An Introduction to Idealism, International Plato
Studies , V. Politis, ed., Sankt Augustin: Academia Verlag.
Nehamas, A. () ‘Predication and Forms of Opposites in the Phaedo’, The
Review of Metaphysics, (), –.
(a) ‘Plato on the Imperfection of the Sensible World’, American
Philosophical Quarterly, (), –.
(b) ‘Confusing Universals and Particulars in Plato’s Early Dialogues’, The
Review of Metaphysics, (), –.
Nettleship, R. L. () Lectures on the Republic of Plato, nd ed., London:
Macmillan.
Nightingale, A. () Spectacles of Truth in Classical Greek Philosophy: Theoria in
Its Cultural Context, Cambridge: Cambridge University Press.
Oderberg, D. S. () Real Essentialism, London: Routledge.
O’Reilly, K. R. () ‘The Jellyfish’s Pleasures: Philebus b–d’, Phronesis, 
(), –.
Owen, G. E. L. ([] ) ‘A Proof in the ΠΕΡΙ ΙΔΕΩΝ’, The Journal of
Hellenic Studies, (), –. Reprinted in M. C. Nussbaum, ed., Logic,
Science and Dialectic: Collected Papers in Greek Philosophy, New York:
Cornell University Press, –.
Pappas, N. () Plato and the Republic, London: Routledge.
Patterson, R. () Image and Reality in Plato’s Metaphysics, Indianapolis:
Hackett.
() ‘Plato: Arguments for Forms’, in R. Le Poidevin, P. Simons, A.
McGonigal and R. P. Cameron, eds., The Routledge Companion to
Metaphysics, London: Routledge, –.
Petrucci, F. M. () ‘L’autenticità dell’ Ippia Maggiore’, in M. Tulli, ed., Testo e
Forme del Testo: Ricerche di Filologia Filosofica, Ricerche di Filologia Classica
VII, Pisa: Fabrizio Serra, –.
Politis, V. () ‘Explanation and Essence in Plato’s Phaedo’, in D. Charles, ed.,
Definition in Greek Philosophy, Oxford: Oxford University Press, –.
() ‘What Is Behind the ti esti Question?’, in J. L. Fink, ed., The
Development of Dialectic from Plato to Aristotle, Cambridge: Cambridge
University Press, –.

 Bibliography
() The Structure of Enquiry in Plato’s Early Dialogues, Cambridge:
(a) ‘Plato’s Geach Talks to Socrates: Definition by Example and Exemplar
in the Hippias Major’, Phronesis, (), –.
(b) ‘Aporia and Sceptical Argument in Plato’s Early Dialogues’, in G.
Karamanolis and V. Politis, eds., The Aporetic Tradition in Ancient
Philosophy, Cambridge: Cambridge University Press.
() ‘The Relation between Knowledge and Enquiry in the Phaedo’, History
of Philosophy & Logical Analysis, (), –.
(forthcoming) ‘Note on Plato’s Statesman D–E’, The Classical Quarterly.
Politis, V., and Jun, S. () ‘The Concept of Ousia in Metaphysics Alpha, Beta,
and Gamma’, in W. Wians and R. Polansky, eds., Reading Aristotle:
Argument and Exposition, Leiden: Brill, –.
Politis, V., and Steinkrüger, P. () ‘Aristotle’s Second Problem about a Science
of Being qua Being: A Reconsideration of Metaphysics iv ’, Ancient
Philosophy, (), –.
Prauss, G. () Platon und der Logische Eleatismus, Berlin: Walter de Gruyter.
Priest, G. () ‘What Is So Bad about Contradictions?’, The Journal of
Philosophy, (), –.
Prior, W. J. () ‘Plato and the “Socratic Fallacy”’, Phronesis, (), –.
Reeve, C. D. C. () Socrates in the Apology, Indianapolis: Hackett.
Rickless, S. C. () Plato’s Forms in Transition: A Reading of the Parmenides,
Cambridge: Cambridge University Press.
Robinson, R. () ‘Forms and Error in Plato’s Theaetetus’, The Philosophical
Review, (), –.
Rodriguez, E. () ‘More Than a Reductio: Plato’s Method in the Parmenides
and Lysis’, Études platoniciennes, . doi.org/./etudesplatoniciennes
..
() ‘Aristotle’s Platonic Response to the Problem of First Principles’,
Journal of the History of Philosophy, , –.
Ross, D. () Plato’s Theory of Ideas, Oxford: Clarendon Press.
Rowe, C. J. () Plato and the Art of Philosophical Writing, Cambridge:
Rowett, C. () ‘Why the Philosopher Kings Will Believe the Noble Lie’,
Oxford Studies in Ancient Philosophy, , –.
() Knowledge and Truth in Plato: Stepping Past the Shadow of Socrates,
Oxford: Oxford University Press.
Ruben, D.-H. () Explaining Explanation, London: Routledge.
Ryle, G. ([] ) ‘Plato’s Parmenides’, Mind, (/), –,
–. Reprinted in R. E. Allen, ed., Studies in Plato’s Metaphysics,
London: Routledge & Kegan Paul, –.
() ‘Logical Atomism in Plato’s Theaetetus’, Phronesis, (), –.
Santas, G. X. () ‘The Socratic Fallacy’, Journal of the History of Philosophy, 
(), –.

Bibliography 
Sayre, K. M. () ‘[Rickless ] (review)’, Journal of the History of Philosophy,
(), –.
Sedley, D. () ‘Platonic Causes’, Phronesis, (), –.
(a) ‘Equal Sticks and Stones’, in D. Scott, ed., Maieusis: Essays in Ancient
Philosophy in Honour of Myles Burnyeat, Oxford: Oxford University Press,
–.
(b) Creationism and Its Critics in Antiquity, Berkeley: University of
California Press.
Sedley, D., and Long, A. () Plato: Meno and Phaedo, Cambridge: Cambridge
University Press.
Silverman, A. () The Dialectic of Essence, Princeton: Princeton University
Press.
Smith, J. A. () ‘General Relative Clauses in Greek’, The Classical Review, 
(/), –.
Sorabji, R. () Matter, Space, and Motion, London: Duckworth.
Stenzel, J. () Studien zur Entwicklung der platonischen Dialektik von Sokrates
zu Aristoteles, nd ed., Leipzig: Teubner.
Storey, D. () ‘Appearance, Perception, and Non-Rational Belief: Republic
c–a’, Oxford Studies in Ancient Philosophy, , –.
Taylor, A. E. () Plato: The Man and His Work, London: Methuen.
Taylor, C. C. W. () ‘Forms As Causes in the Phaedo’, Mind, (), –.
() ‘Plato’s Epistemology’, in G. Fine, ed., The Oxford Handbook of Plato,
Oxford: Oxford University Press, –.
Teloh, H. () The Development of Plato’s Metaphysics, University Park:
Pennsylvania State University Press.
Tuozzo, T. M. () ‘“Appearing Equal” at Phaedo B–C: An Epistemic
Interpretation’, Oxford Studies in Ancient Philosophy, , –.
Vlastos, G. () ‘The Third Man Argument in the Parmenides’, The
Philosophical Review, (), –.
() ‘Reasons and Causes in the Phaedo’, The Philosophical Review, (),
–.
([] ) ‘Socrates’ Disavowal of Knowledge’, The Philosophical Quarterly,
(), –. Reprinted in M. Burnyeat, ed., Socratic Studies, Cambridge:
Cambridge University Press, –.
([] ) ‘Is the “Socratic Fallacy” Socratic?’, Ancient Philosophy, (),
–. Reprinted in M. Burnyeat, ed., Socratic Studies, Cambridge:
Cambridge University Press, –.
Wedgwood, R. () ‘Plato’s Theory of Knowledge’, in D. O. Brink, S. S.
Meyer and C. Shields, eds., Virtue, Happiness, Knowledge: Themes from the
Work of Gail Fine and Terence Irwin, Oxford: Oxford University Press,
–.
White, F. C. () Plato’s Theory of Particulars, New York: Arno Press.
White, N. P. () ‘Plato’s Metaphysical Epistemology’, in R. Kraut, ed., The
Cambridge Companion to Plato, Cambridge: Cambridge University Press,
–.

 Bibliography
Wolfsdorf, D. () ‘The Socratic Fallacy and the Epistemological Priority of
Definitional Knowledge’, Apeiron, (), –.
() ‘Αἴτιον and Αἰτία in Plato’, Ancient Philosophy, (), –.
Woodruff, P. () Hippias Major, Oxford: Blackwell.
() ‘Expert Knowledge in the Apology and Laches: What a General Needs
to Know’, Proceedings of the Boston Area Colloquium in Ancient Philosophy, 
(), –.
() ‘Plato’s Early Theory of Knowledge’, in S. Everson, ed., Epistemology,
Cambridge, –.

General Index
Allen, R. E., – Dancy, R. M., , , , 

Anaxagoras,  Davidson, D., 
Andersen, H. C.,  debunking. See philosophy
Annas, J., , , , , ,  deflation. See philosophy
Apelt, O.,  Descartes, R., 
Archer-Hind, R. D.,  Dillon, J. M., , 
Aristotle, , –, , , , –, Dodds, E. R., 
–, , , ,  Dummett, M., 
Athens, 
epistemology. See knowledge
Bacon, F.,  essence
Balaudé, J.-F.,  and causation/explanation, , –, ,
Bambrough, R.,  , Ch. 
Beethoven, L. V.,  and definition. See essence and the ti esti
being question
in Plato's Sophist, , ,  and judgement, , , Ch. 
primary being, , – and sense-perception, –, , –,
the question ‘What is there?’ versus the –, Ch. , Ch. , Ch. , Ch. 
question ‘What is being?’, ,  and substance, , , 
Berkeley, G.,  and thinking, –, –, , ,
Bestor, T. W., ,  , Ch. , Ch. 
Bluck, R. S., ,  and the ti esti question (‘What is . . .?’
Bostock, D., , ,  question), , –, , , –, ,
Brisson, L.,  , , , , , , , –,
Broadie, S., ,  –, , –, Introduction,
Burnet, J.,  Ch. 
Burnyeat, M. F., , , , –, , essences and examples/exemplars, , –, ,
,  , , , , , –, –,
–, , , Ch. , Ch. 
Carone, G. R.,  relation among essences, , –, –,
Castagnoli, L.,  –, 
Castelli, L. M.,  the view that Forms are essences, not things
causation that have essences, , , Conclusion
and essence. See essence: essence and essentialism
causation/explanation Aristotelian essentialism, , , , ,
transmission theory of causation, , 
–, , , – Aristotle’s essentialism versus Plato’s
Code, A.,  essentialism, , , –, –,
Cornford, F. M.,  –, 
Crivelli, P.,  modern essentialism, , , –
Cross, R. C.,  Evans, D., 


 General Index
explanation. See also causation Malcolm, J., , , 
the principle of the uniformity of explanation, Mann, W.-R., 
, , –, , , , , McCabe, M. M., 
–, , ,  Meinwald, C. C., –, 
metaphysics
Ferejohn, M.,  as essence-based ontology, , 
Fine, G., , , , , ,  in the sense of what there is and what is
Fine, K.,  primary, , 
forms. See essence Moss, J., , , 
Frede, M., , , ,  Murphy, N. R., 
Frege, G., 
Natorp, P., , –
Gallop, D., , , –,  Nehamas, A., , , , 
Geach, P., ,  Nettleship, R. L., 
Gerson, L. P.,  Nightingale, A., 
Gill, M. L., , , , , ,
– Oderberg, D. S., 
Gosling, J. C. B., ,  Owen, G. E. L., , , 
Gould, C. S., 
Greene, B.,  Pappas, N., 
Grube, G. M. A., –, , , ,  Parmenides, , , 
Patterson, R., , 
Hackforth, R.,  philosophy
Harte, V., , , –, , , , , , debunking attitude to philosophy/deflationary
 attitude to philosophy, , , , 
Heindorf, L. F.,  Plato and naturalism, , 
Heraclitus, ,  Plotinus, , 
Hestir, B. E.,  Priest, G., 
Hintikka, J.,  Putnam, H., 
Homer, 
Horan, D.,  Quine, W. V. O., 
Hume, D., –, , 
Reeve, C. D. C., , , 
Johansen, T. K.,  requirements for answering the ti esti question,
–, 
Kahn, C. H.,  requirements of definition. See requirements for
Kant, I., , ,  answering the ti esti question
Karbowski, J.,  Rickless, S. C., , , , 
Keyt, D.,  Ross, D., , 
Kirwan, C., , ,  Rowe, C. J., , , , , –, ,
knowledge 
a priori knowledge versus a posteriori Rowett, C., 
knowledge, , , , , , –, Ruben, D.-H., 
Ch.  Ryan, P., , 
does knowledge of Forms require intellectual Ryle, G., 
vision?, –
epistemology, , –, , , ,  Sedley, D., , , , , , 
the relation between (the theory of ) self-predication
knowledge and (the theory of ) enquiry, are Forms self-predicative?, –, –,
 Conclusion
Kutschera, F. von, ,  Shostakovich, D., 
Silverman, A., , 
Lachance, G.,  Slings, S. R., 
Lee, D. C., ,  Smith, J. A., 
Lorenz, H., ,  Sorabji, R., 

General Index 
Steinkrüger, P.,  Vazquez, D., 
Storey, D., ,  Vlastos, G., , 
Strawson, P., 
substances and essence. See essence Wedgwood, R., 
White, F. C., 
Taylor, A. E.,  White, N. P., 
Taylor, C. C. W., , , , Wittgenstein, L., , 
 Woodruff, P., , 
Teloh, H., 
Tuozzo, T. M.,  Xenokrates, 

Index Locorum
Aristotle a–, 
Metaphysics, ,  Gorgias, 
Gamma ,  –, 
b–,  Hippias Major, –, , , , , , –,
Iota ,  , , , –, –, ch.
a–,  –, –, , 
Lambda,  c f., 
Lambda  c, 
a–,  d, 
Mu,  e–, , , , 
Nu,  b–d, 
Zeta –,  d–e, 
Zeta ,  d, 
b–,  a, , 
Nicomachen Ethics,  e–, 
a–b, 
Heraclitus a–, 
B,  b, 
b, 
Parmenides b, 
B, ,  b–, 
Plato d–, 
Charmides b, 
a–,  d, 
Cratylus, , , ,  e–c, 
b–c,  c–, 
b–,  b, 
b–c,  Laches, , 
b–c,  Meno, , , , 
–c,  d ff., 
d–e,  Parmenides, , , –, , –, , ,
d–,  , , , , 
d–e,  e–b, 
d,  e–a, 
Euthyphro, , , ,  e–a, 
d–e,  e–a, 
d–e,  b–, 
d–e, , – b–, 
d–e,  , , , 
e–,  a–e, , , –


Index Locorum 
b–c, – d–e, 
b–,  d–e, , 
b–d,  b–b, 
b–,  c, 
c,  c–d, 
d–, , ,  c, 
,  b–e, , 
a–b,  d, 
a–, ,  a, 
b–c, , –, , –, e–c, 
, – a, 
c–,  b–, 
b–c, ,  Phaedo, , –, –, , , –,
b–,  –, , , , , , ,
c–d,  , , , , , ,
c–d,  –
d–,  b, –
e–b, ,  e–c, , –, –, 
,  e–e, 
b–c, , ,  e–a, 
b–c, –, – a, 
b–c,  c–, 
b–c, ,  c–a, 
c–d,  e, 
d,  a–b, 
d–a,  a–a, 
e–, , – a, 
e–a,  a ff., 
a f.,  a–b, , 
a–c, ,  a–b, 
c,  b–, 
ee,  b–c, 
b–,  c, 
b, – d–e, 
c–a,  d–e, 
c, ,  d–e, 
c–,  d, , 
e–,  d–a, 
e–,  d–e, , , 
a, ,  a, 
a–, – a, 
a–,  c, 
a–, ,  d–, 
b ff.,  c, 
b–e,  e–b, 
b–,  e–b, 
b,  e–a, 
b f.,  , , 
a, ,  a–c, –, –, , , , , ,
a, ,  , 
a–,  a–a, 
b,  a–c, , , –
e–a,  a, 
b, ,  b–, 
c–,  b, 

 Index Locorum
Plato (cont.) e–b, , 
b, ,  a, , 
b–,  a–b, 
b, ,  a, , 
c,  a, , 
c–,  a–b, 
c–,  a–b, , 
c,  b–, –
a, –, –, ,  b–, , 
a–, , , , ,  b, 
b, ,  b–c, 
c–d,  b–d, 
c–d, –, ,  c, , 
d–, ,  b, 
e–,  b–c, 
a–b, – e–a, –, , 
c,  , 
c,   ff., 
c–d, , ,  a, 
d,  d–a, , , –
d–, ,  d, 
d, ,  d, 
e–a,  d–a, , 
a, , , , , , , ,  d–a, 
a,  e, 
a–,  a–, –
a,  a–, 
d,  b f., 
d,  b–c, –
e,  b–b, 
b,  b–, , 
e,  b–, 
b,  c, 
e–d,  c–, 
e–d,  d, 
c–,  d–, , , 
e–b,  d, 
d ff.,  e, 
e ff., , , – e–d, 
e–c,  e–b, , 
e–b,  a–b, 
e–a,  a–b, 
a–b,  a–b, –
a–,  b–c, , 
a,  c–, 
a–,  c–, 
a–, ,  c, 
b–c, –,  c–, 
c,  d–, 
c f.,  d, 
c–,  d–, 
c–d,  d–, 
d–e, ,  d, 
e–,  b, 

Index Locorum 
b ff.,  a, –
b–a,  a–, 
c–,  a–, , 
c–c,  b–, 
c–c,  b–, 
c–c,  c–a, –
d–,  d–e, 
d, ,  d–, 
d, ,  a, 
e–,  d–, 
e,  e–, 
e,  e–d, 
b–c,  e–a, –
c,  e–d, –
c–,  a, 
c–c,  b, 
b–c,  b–, 
b–c,  b, 
b–c,  c, 
b–,  VI
c,  b, 
c–e,  c, , 
c,  a, 
c–,  a, 
c f.,  a f., 
d, – a–b, 
Phaedrus, , , ,  a–, , 
e,  b–, 
Philebus, , , , –, , ,  b, 
a, , ,  d, 
b,  e, –
c,  e–a, 
Protagoras, ,  d, 
–,  e–, 
Republic, , –, , , , , –, , e–, 
, , –, –, , , , b–, 
, , , , –, , b–, 
 b, 
I VII
e–b,  –, , 
e–a,  a ff., 
III a–, , , 
–,  a–a, 
IV a–b, 
–, , ,  a–a, , , , , , ,
b–a, , – , , –, , 
b–c, ,  a–a, 
e–a, ,  a, 
c,  a, 
V, , , ,  a–b, –, 
a,  b, 
b–a,  b, 
d ff., ,  b–b, 
e–,  b, 

 Index Locorum
Plato (cont.) a–, 
b–, ,  a–, , 
b,  a, 
c,  a–, , 
c,  a, 
d–, ,  a–, 
e–,  a, 
e–a, ,  a, 
e,  d–a, 
–,  d–a, 
a–,  d–e, 
a–,  d, 
a,  d, 
a–,  a–, 
a–, ,  e–c, 
a–,  d–a, ,
a–d,  
a,  a, , 
a–,  a, 
b–,  X, 
b–c,  a–, 
b–c,  b, 
b,  b, 
b f.,  e, 
b–,  e–, 
b,  a, 
b,  Sophist, , , , , , , –,
b,  , 
c,   ff., 
c–,   ff., 
c,  e, , 
c–, ,  b, , 
c, , ,  b–, 
c–, ,  d, , 
c,  e, , 
d–a,  b, , 
d–a,  b, , 
d,  c, , 
d–a, ,  , –
d–a,  b, 
d–a,  b–c, 
d–a, ,  d, , 
d–e, ,  d, , 
e, ,   ff., 
e,  a, , 
e–a,  a, 
e–a,  b, 
e–a,  c–, –
e, ,  d, 
e,  b, 
a, ,  e–, 
a, , –, – Statesman, , 
a–,  Symposium, , , 
a,  e–a, 

Index Locorum 
The Seventh Letter,  a–b, 
Theaetetus, , , , ,  e–a, , 
b,  b, 
–,  c–d, 
d f.,  a–b, 
Timaeus, , , , –, , , , b, 
–,  b, 


Ebin - Pub - Platos Essentialism Reinterpreting The Theory of Forms 97811089862439781108833660

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

Ebin - Pub - Platos Essentialism Reinterpreting The Theory of Forms 97811089862439781108833660

Uploaded by

Copyright:

Available Formats

PLATO’S ESSENTIALISM

  is Professor of Philosophy at Trinity College

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Cambridge University Press is part of the University of Cambridge.

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

The topic of the present study is Plato’s theory of Forms, as it used to be

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Why cannot the ti esti question be answered

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Why cannot essences, or Forms, be perceived

. Problems with the standard answer, and in defence

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

. The origin of this answer in the Hippias Major

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

. From beautiful to equal and one

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

. From beautiful, equal and one to human and bed

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

. What about numbers? Does Plato’s argument extend to them?

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms

Downloaded from https://www.cambridge.org/core. , on , subject to the Cambridge Core terms