Plato's Essentialism: Reinterpreting the Theory of Forms 97811089862439781108833660

Citation preview

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

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

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

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.

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

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

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

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

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

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

Contents

viii 

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

  

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

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 graduate 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 Tu¨bingen University. I am also grateful to my friend and host Richard King during my time at Bern University. Several universities 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

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

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 Steinkru¨ger, 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.

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

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 answering 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 

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/10.1017/9781108986243.001



Introduction

(paradeigma) for a thing’s being of a certain quality and a way of determining 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 requirements 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 question; 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

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/10.1017/9781108986243.001

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 requirements 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

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/10.1017/9781108986243.001



Introduction

changeless, uniform, not perceptible by the senses, knowable only by reasoning, the basis of causation and explanation, distinct from senseperceptible things, necessary for thought and speech, separate from physical 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 questions – by what process of reasoning, in Plato, does the posing of the ti esti question turn into the commitment to Forms? What justifies the supposition 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 practically 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.

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/10.1017/9781108986243.001

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 forgotten!) 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 inhabitants 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 exceedingly 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 curiosity, 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 unreasonable 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.

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/10.1017/9781108986243.001



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 dialogical, 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 arguments 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

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/10.1017/9781108986243.001

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 developmentalism, 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

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/10.1017/9781108986243.001



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 characteristics 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 consideration, 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 ().

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/10.1017/9781108986243.001

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 ultimately on a single question, and one that is not philosophically controversial 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 noncomposite; 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 senseperceptible 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

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/10.1017/9781108986243.001



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 explanans 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 commonly 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.

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/10.1017/9781108986243.001

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 explanation – 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 unprejudiced 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 selfpredicative, 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 socalled 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

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/10.1017/9781108986243.001



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 philosophers 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.’

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/10.1017/9781108986243.001

Introduction



I have already pointed to another, if closely related, remarkable consequence 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

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/10.1017/9781108986243.001



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 exemplar. 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 distinction 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 reasoning, does not rule out that such reasoning involves, and even depends on, sense-perception. In the Phaedo (a) Plato expressly says that such reasoning 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 pathosfilled 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 ().

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/10.1017/9781108986243.001

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 knowledge 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 knowledge, 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 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 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 epistemological 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 knowledge. 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.

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/10.1017/9781108986243.001



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).

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/10.1017/9781108986243.001

 

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 du¨nkt 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



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/10.1017/9781108986243.002



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.’

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/10.1017/9781108986243.002

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

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/10.1017/9781108986243.002



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 statement, 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.

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/10.1017/9781108986243.002

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

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/10.1017/9781108986243.002



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 considered 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).

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/10.1017/9781108986243.002

 

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 nonphysical 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 ().



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/10.1017/9781108986243.003



Plato’s Essentialism

in such dialogues as Phaedo and Republic, for thinking that Forms are nonphysical 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:

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/10.1017/9781108986243.003

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:

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/10.1017/9781108986243.003



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 unnecessary 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

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/10.1017/9781108986243.003

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.

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/10.1017/9781108986243.003



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.

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/10.1017/9781108986243.003

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 following this argument, can be understood as amounting to such a statement. 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

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/10.1017/9781108986243.003



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,

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/10.1017/9781108986243.003

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 .

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/10.1017/9781108986243.003



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.

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/10.1017/9781108986243.003

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 ophthalmois, 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 .

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/10.1017/9781108986243.003



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 .

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/10.1017/9781108986243.003

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 following, 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 recurring 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 senseperceptible exemplars and standards of equal (see esp. the isōn at a).

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/10.1017/9781108986243.003



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 senseperceptible 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 geometrical 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.

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/10.1017/9781108986243.003

The sensory imperceptibility of Forms



In my estimation, our understanding of the pattern of argument commonly 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 oppositeto-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 exemplar 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

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/10.1017/9781108986243.003



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 ().

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/10.1017/9781108986243.003

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 adequately 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

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/10.1017/9781108986243.003



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 committed 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 .

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/10.1017/9781108986243.003

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 understand 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 ().

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/10.1017/9781108986243.003



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 demonstrates, 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.

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/10.1017/9781108986243.003

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 remarking ‘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

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/10.1017/9781108986243.003



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 defining 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 senseperceptible 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.)

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/10.1017/9781108986243.003

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 consequence 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.

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/10.1017/9781108986243.003



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 justification 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 sovereignty. 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 .

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/10.1017/9781108986243.003

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 propositions 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 reasoning with the distinction between empirical knowledge and a priori knowledge, 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).

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/10.1017/9781108986243.003



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 changeless 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.

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/10.1017/9781108986243.003

 

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’.



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/10.1017/9781108986243.004



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 noncomposite? How is this claim to be understood? And what are its consequences? 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 nonphysical is a characteristic of the Forms, and even if we suppose that being non-physical is a necessary condition for being unitary, uniform and noncomposite 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 scrambled 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.

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/10.1017/9781108986243.004

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 questionable, and in greater need of justification, to suppose that Forms are nonphysical things than to suppose that they are unitary, uniform and noncomposite; 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 noncomposite? 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.

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/10.1017/9781108986243.004



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 addressing 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.

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/10.1017/9781108986243.004

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 .

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/10.1017/9781108986243.004



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 investigate 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

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/10.1017/9781108986243.004

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 investigating 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.

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/10.1017/9781108986243.004



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 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? What is the relation between the unity of an adequate and true account of what a thing is, and the unity, uniformity, and noncompositeness 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

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/10.1017/9781108986243.004

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 repeatedly 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

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/10.1017/9781108986243.004



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 truthfunctional 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, ).

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/10.1017/9781108986243.004

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 noncomposite 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

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/10.1017/9781108986243.004



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.

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/10.1017/9781108986243.004

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, ‘unmoving’, 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,

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/10.1017/9781108986243.004



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 noncontradiction. 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 interchangeable. 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 intelligent, 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 .

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/10.1017/9781108986243.004

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 immortality 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) selfmovement 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

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/10.1017/9781108986243.004



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

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/10.1017/9781108986243.004

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.

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/10.1017/9781108986243.004



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, uniformity and non-compositeness of Forms, that Plato holds it. That is, he defends Premise iii by considering the question, What grounds the changelessness 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.

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/10.1017/9781108986243.004

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

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/10.1017/9781108986243.004



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.

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/10.1017/9781108986243.004

 

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 supposing 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).



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/10.1017/9781108986243.005



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 independence 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

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? 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 senseperception 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 knowledge, or the knowledge of certain Forms, is a priori and independent of

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/10.1017/9781108986243.005

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 senseperceptible 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 independent of sense-perception, which is how many critics have understood 

Such critics include Annas (); Bostock (); Taylor (); Wedgwood ().

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/10.1017/9781108986243.005



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 senseperception 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 demonstrate 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 knowing 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.

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/10.1017/9781108986243.005

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.

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/10.1017/9781108986243.005



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 epistemological 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

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/10.1017/9781108986243.005

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 philosophical 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 epistemology, 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 senseperception, 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

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/10.1017/9781108986243.005



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 .

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/10.1017/9781108986243.005

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 senseperception 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 .

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/10.1017/9781108986243.005



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 question. 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 judgements, 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 .

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/10.1017/9781108986243.005

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 nonsense-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 Formdirected 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)

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/10.1017/9781108986243.005



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 perceived 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 .

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/10.1017/9781108986243.005

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 senseperception 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 senseperception 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

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/10.1017/9781108986243.005



Plato’s Essentialism

senses; but according to the later claim (a; prepared for at a–c), senseperception 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 contradiction, 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’).

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/10.1017/9781108986243.005

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 knowledge, 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.

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/10.1017/9781108986243.005



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 epistemological 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

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/10.1017/9781108986243.005

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 propositions 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 nonperceptibility 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 nonperceptibility 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 committed 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.

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/10.1017/9781108986243.005



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 nonperceptibility of Forms, that is, because he is committed to The nondefinability of certain qualities by example and exemplar, then his commitment 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 epistemology. 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 consideration, of understanding why Plato is committed to The non-perceptibility of Forms, such that this commitment is readily compatible with his commitment 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

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/10.1017/9781108986243.005

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-chatteredabout (poluthrēleta) things” (B–)’ ().

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/10.1017/9781108986243.005



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 .

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/10.1017/9781108986243.005

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 senseperception, 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 consequence 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, knowledge, 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.

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/10.1017/9781108986243.005

 

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

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/10.1017/9781108986243.006

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 unF 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 senseperceptible 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

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/10.1017/9781108986243.006



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 copresence of opposites, it can, and I shall argue that it should, be understood 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.

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/10.1017/9781108986243.006

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 demonstrate, 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 senseperceptible 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 conducted 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

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/10.1017/9781108986243.006



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 perspicuous 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 senseperceptible 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; nonspatial 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.

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/10.1017/9781108986243.006

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.’

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/10.1017/9781108986243.006



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, exemplar 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 consequences 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

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/10.1017/9781108986243.006

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 copresence 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 ).

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/10.1017/9781108986243.006



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 interpretations 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 dialoguecrossing! – 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.

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/10.1017/9781108986243.006

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. ).

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/10.1017/9781108986243.006



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.

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/10.1017/9781108986243.006

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 presented 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 senseperceptible 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

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/10.1017/9781108986243.006



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 themselves 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

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/10.1017/9781108986243.006

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 argument, 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 copresence 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 ().

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/10.1017/9781108986243.006



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 argument 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

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/10.1017/9781108986243.006

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.

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/10.1017/9781108986243.006



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 opposite-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 welldisposed towards the view that there are things that are not senseperceptible 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’, philotheamones, 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).

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/10.1017/9781108986243.006

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 senseperceptible things is incompatible with the proposition that real knowledge 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 knowledge, 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’ ().

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/10.1017/9781108986243.006



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 senseperceptible 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 important 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. ).

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/10.1017/9781108986243.006

Forms as distinct from sense-perceptible things



Again, the similarity to the argument in the Hippias Major is remarkable. 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 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 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.

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/10.1017/9781108986243.006

 

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 understood 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 conclusion 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 ().



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/10.1017/9781108986243.007

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 arguments 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

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/10.1017/9781108986243.007



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

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/10.1017/9781108986243.007

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 argument. 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 explanation 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 consequence 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.

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/10.1017/9781108986243.007



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, selfpredicative, 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.

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/10.1017/9781108986243.007

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/explanandum. 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

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/10.1017/9781108986243.007



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 generation 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 particular, 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).

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/10.1017/9781108986243.007

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 explanations 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).

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/10.1017/9781108986243.007



Plato’s Essentialism

nature’ (phusis, a–). He goes on (b–c) to clarify what is distinctive 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 statements (logoi) and, in particular, statements that put forward explanatory hypotheses (see d–a). For, it appears that what this methodological 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 question: 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 immediately 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 explanation of generation and destruction is directly associated with a more

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/10.1017/9781108986243.007

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 requirements 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

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. 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.

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/10.1017/9781108986243.007



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 explanation 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 philosophers (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..

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/10.1017/9781108986243.007

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 explanations 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 explanans (the particular constituents of a particular body), then same explanandum, 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

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/10.1017/9781108986243.007



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 physicalist 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 explanations 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.

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/10.1017/9781108986243.007

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 requirement, 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 physicalist explanations, and indeed any of the apparently unobjectionable explanations with which we are readily familiar, fail to satisfy the requirements 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 (, ).

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/10.1017/9781108986243.007



Plato’s Essentialism

(b) and that, as a result of this, he turned to essence-based or Formbased 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 explanations, like the reason why he despairs of traditional physicalist explanations, 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, conversely, 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).

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/10.1017/9781108986243.007

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 requirements 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’ intellectual 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.

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/10.1017/9781108986243.007



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 misunderstanding. 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 requirements 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

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/10.1017/9781108986243.007

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 intelligible 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 requirements 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

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/10.1017/9781108986243.007



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.

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/10.1017/9781108986243.007

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 deterministic 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 experiment 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 explanation, 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’.

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/10.1017/9781108986243.007



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 explanations 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).

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/10.1017/9781108986243.007

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 deliberately 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 understood 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.

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/10.1017/9781108986243.007



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 selfevidently 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.

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/10.1017/9781108986243.007

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 explanation appears to conform to the procedure of hypothesis, as this method is

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/10.1017/9781108986243.007



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 explanation, 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.

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/10.1017/9781108986243.007

Forms as the basis of causation and explanation



iii. iv.

the supposition that the quality F has an essence, Ess(F); 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

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/10.1017/9781108986243.007



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 requirements 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

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/10.1017/9781108986243.007

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 explanandum, 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 connection 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 explanandum, 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 explanation. 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 constituent 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 explanatory 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 accommodates 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.

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/10.1017/9781108986243.007



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 explanations, REQ and REQ, which embody the constraints of the uniqueness, 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

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/10.1017/9781108986243.007

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 tautological. 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 explanations. That explanations in conformity with this schema satisfy this constraint 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 explanation 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 conformity 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

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/10.1017/9781108986243.007



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.

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/10.1017/9781108986243.007

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 distributivity 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.

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/10.1017/9781108986243.007



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 explananda 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

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/10.1017/9781108986243.007

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

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/10.1017/9781108986243.007



Plato’s Essentialism

response to this disappointment, does he turn to essence-based explanations? The question is: First, why does Plato introduce good-based explanations 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 essencebased (formal) explanations? One answer to the first question is clearly mistaken: Plato introduces good-based explanations because his complaint against traditional explanations 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 following 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 goodbased 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 mentioned (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 introduced 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 ().

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/10.1017/9781108986243.007

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 upcoming 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 connection 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 requirement, 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.

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/10.1017/9781108986243.007



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 uniformity 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 goodbased 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 explanations. 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.

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/10.1017/9781108986243.007

 

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 senseperceptible and physical things, and that conform to the principle of noncontradiction, 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.



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/10.1017/9781108986243.008



Plato’s Essentialism

which says that Plato’s distinction, in this passage, between senseperception 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 senseperception 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 senseperception 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 propositional and conceptual content. (Whether Plato’s challenges this supposition 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-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. 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.

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/10.1017/9781108986243.008

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 nonperceptible 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

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/10.1017/9781108986243.008



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 concerned 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 senseperceptible 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.

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/10.1017/9781108986243.008

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-cumphysicalist 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 phenomenalist-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 statements (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?’)

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/10.1017/9781108986243.008



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 prepared 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

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/10.1017/9781108986243.008

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 conversioninducing 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

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/10.1017/9781108986243.008



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, alternatively, 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 noncontradiction, 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.

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/10.1017/9781108986243.008

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 apparently 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 philosophers, 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 senseperception. 

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. .

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/10.1017/9781108986243.008



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 senseperception. 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 problem 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.

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/10.1017/9781108986243.008

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 contravening 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 noncontradiction, 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 investigated 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

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/10.1017/9781108986243.008



Plato’s Essentialism

negative. He answers it in the negative for a particular reason and argument: 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 senseperception (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:

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/10.1017/9781108986243.008

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. GrubeReeve, 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.

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/10.1017/9781108986243.008



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 concentrating 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

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/10.1017/9781108986243.008

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 individuated, for the purpose of making judgements about it of which we are entitled to suppose that they conform to the principle of noncontradiction. 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 discourse. But we have no evidence, certainly not in this passage from the Republic, to suppose this; and the supposition has the air of bad anachronism. 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 judgements 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 noncontradiction 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.

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/10.1017/9781108986243.008



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 senseperception, 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 senseperception, 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 independent source of judgements.’ Burnyeat must, from what I can tell, be understood 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 ().

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/10.1017/9781108986243.008

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 senseperceptible 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 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 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

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/10.1017/9781108986243.008



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, sensebased 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 noncontradiction, is the product of, on the one hand, sense-perception, and, on the other hand, a kind of thought that is independent of senseperception 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 senseperception 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 (, –).

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/10.1017/9781108986243.008

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 investigation’ (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 intelligible 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 perceptions’, 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

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/10.1017/9781108986243.008



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-perception’; 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 thoughtprovoking ‘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 judgements 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]’.

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/10.1017/9781108986243.008

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 contradiction, 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

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/10.1017/9781108986243.008



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 determinate 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

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/10.1017/9781108986243.008

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 something (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 that which is intelligible and the other 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.

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/10.1017/9781108986243.008



Plato’s Essentialism

original reports by the sensory perceptions, which, unless further investigated, 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 compound 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

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/10.1017/9781108986243.008

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 restriction, 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 statements ‘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 established (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 intellectual 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.

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/10.1017/9781108986243.008



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 accordingly. 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 that which is intelligible and the other

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/10.1017/9781108986243.008

Judgements about sense-perceptible things



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 argument 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 senseperceptible 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)

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/10.1017/9781108986243.008



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 senseperceptible thing of conspicuous unity – that is, conspicuous to senseperception, 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

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/10.1017/9781108986243.008

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 noncontradiction. 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

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/10.1017/9781108986243.008



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 conspicuously 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.

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/10.1017/9781108986243.008

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 number. 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.

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/10.1017/9781108986243.008



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.

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/10.1017/9781108986243.008

 

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 disagreements, 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 significantly 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 mindindependent – 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’.



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/10.1017/9781108986243.009



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.

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/10.1017/9781108986243.009

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.

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/10.1017/9781108986243.009



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 allimportant 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 necessary 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.

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/10.1017/9781108986243.009

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,

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/10.1017/9781108986243.009



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 notwithstanding, 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 characteristics that Socrates attributes to them in introducing them and Parmenides further attributes to them in subjecting them to the objections, 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

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/10.1017/9781108986243.009

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’ premature attempt to define each of the Forms (horizesthai epicheireis . . . hen hekastōn tōn eidōn, c–d). This picks up on the immediately preceding passage, which says that Forms, and the defining of Forms, are necessary

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/10.1017/9781108986243.009



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 indication 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–.

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/10.1017/9781108986243.009

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 consequences 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 consequence 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.

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/10.1017/9781108986243.009



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 unknowable 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 dialogue 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.

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/10.1017/9781108986243.009

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.

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/10.1017/9781108986243.009



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 suppositions 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 consequences 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.

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/10.1017/9781108986243.009

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.

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/10.1017/9781108986243.009



Plato’s Essentialism

of something that is ([tinos] ontos). However, the present passage significantly 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 something, 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 ).

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/10.1017/9781108986243.009

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

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/10.1017/9781108986243.009



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 indeterminacy that is characteristic, in this scenario, of being: the object of thought will amount to a thoroughgoing illusion (phantasma, a; also

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/10.1017/9781108986243.009

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 consequences, 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.

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/10.1017/9781108986243.009



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

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/10.1017/9781108986243.009

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).

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/10.1017/9781108986243.009



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 snowcrystal 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,

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/10.1017/9781108986243.009

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 senseperception, 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 addition 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

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/10.1017/9781108986243.009



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.

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/10.1017/9781108986243.009

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.

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/10.1017/9781108986243.009

 

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 commonly 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 

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/10.1017/9781108986243.010

The separation of Forms from physical things



that Aristotle has in mind. For him, just as for Plato, this, the essencebased, 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 essentialist 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 Steinkru¨ger (). The passage from Metaphysics Iota is, of course, thoroughly striking for our present purpose. See Gerson ().

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/10.1017/9781108986243.010



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

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/10.1017/9781108986243.010

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).

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/10.1017/9781108986243.010



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 argument; 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 .

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/10.1017/9781108986243.010

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.

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/10.1017/9781108986243.010



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 indestructibility 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 swansong, 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 assumption 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

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/10.1017/9781108986243.010

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 separate, 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 philosophy 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 senseperceptible things: these are different ways – indeed, significantly different

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/10.1017/9781108986243.010



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 understand 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

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/10.1017/9781108986243.010

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 anything 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.

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/10.1017/9781108986243.010



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.

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/10.1017/9781108986243.010

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,

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/10.1017/9781108986243.010



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

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/10.1017/9781108986243.010

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.

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/10.1017/9781108986243.010



Plato’s Essentialism

It may be said that what all this shows is, merely and relatively unexcitedly, 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 introduction of things such as fire and snow, in the course of his account of the

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/10.1017/9781108986243.010

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 .

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/10.1017/9781108986243.010



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.

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/10.1017/9781108986243.010

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 consequence 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).

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/10.1017/9781108986243.010



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 hedonist, 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 understanding physical things, whether individually or collectively, but also nonphysical 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 ().

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/10.1017/9781108986243.010

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 fundamentally, 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 separation 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

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/10.1017/9781108986243.010



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 committed 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 Steinkru¨ger ().

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/10.1017/9781108986243.010

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.

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/10.1017/9781108986243.010

 

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



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/10.1017/9781108986243.011

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 investigating things in and through our questions and answers about them. This is what the Phaedo passage says. Perhaps this is Plato’s deepest indebtedness 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 characterised 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’.

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/10.1017/9781108986243.011



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

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/10.1017/9781108986243.011

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 confusion, 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 transcendental idealist. But, if the alternative is, simply and plainly, to recognise 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.

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/10.1017/9781108986243.011



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

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/10.1017/9781108986243.011

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.

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/10.1017/9781108986243.011

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, 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 – 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 u¨brigens 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).



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/10.1017/9781108986243.012

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 selfpredication? 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 .

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/10.1017/9781108986243.012



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 selfpredication 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 selfpredication, for she sums up her examination with: ‘If there are good grounds for supposing that Forms self-predicate . . .’ (, emphasis added).

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/10.1017/9781108986243.012

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 autobiography 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 underscores, the following is clearly false: If it is established that Plato distinguishes 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.

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/10.1017/9781108986243.012



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 goodhumoured 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–.

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/10.1017/9781108986243.012

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 

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/10.1017/9781108986243.013



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: Cambridge University Press. Castagnoli, L. () Ancient Self-Refutation: The Logic and History of the SelfRefutation 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 fu¨r 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:

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/10.1017/9781108986243.013

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 fu¨r Philosophie und Philosophische Kritik, N.F. , –. Gadamer, H.-G. () Gesammelte Werke, Band : Griechische Philosophie I, Tu¨bingen: J. C. B. Mohr. () Gesammelte Werke, Band : Griechische Philosophie III: Plato Im Dialog, Tu¨bingen: 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, (), –.

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/10.1017/9781108986243.013



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 fu¨r 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, –.

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/10.1017/9781108986243.013

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 Einfu¨hrung 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, –.

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/10.1017/9781108986243.013



Bibliography

() The Structure of Enquiry in Plato’s Early Dialogues, Cambridge: Cambridge University Press. (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 Steinkru¨ger, 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: Cambridge University Press. 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,  (), –.

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/10.1017/9781108986243.013

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, –.

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/10.1017/9781108986243.013



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, –.

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/10.1017/9781108986243.013

General Index

Allen, R. E., – Anaxagoras,  Andersen, H. C.,  Annas, J., , , , , ,  Apelt, O.,  Archer-Hind, R. D.,  Aristotle, , –, , , , –, –, , , ,  Athens,  Bacon, F.,  Balaudé, J.-F.,  Bambrough, R.,  Beethoven, L. V.,  being in Plato's Sophist, , ,  primary being, , – the question ‘What is there?’ versus the question ‘What is being?’, ,  Berkeley, G.,  Bestor, T. W., ,  Bluck, R. S., ,  Bostock, D., , ,  Brisson, L.,  Broadie, S., ,  Burnet, J.,  Burnyeat, M. F., , , , –, , ,  Carone, G. R.,  Castagnoli, L.,  Castelli, L. M.,  causation and essence. See essence: essence and causation/explanation transmission theory of causation, , –, , , – Code, A.,  Cornford, F. M.,  Crivelli, P.,  Cross, R. C., 

Dancy, R. M., , , ,  Davidson, D.,  debunking. See philosophy deflation. See philosophy Descartes, R.,  Dillon, J. M., ,  Dodds, E. R.,  Dummett, M.,  epistemology. See knowledge essence and causation/explanation, , –, , , Ch.  and definition. See essence and the ti esti question and judgement, , , Ch.  and sense-perception, –, , –, –, Ch. , Ch. , Ch. , Ch.  and substance, , ,  and thinking, –, –, , , , Ch. , Ch.  and the ti esti question (‘What is . . .?’ question), , –, , , –, , , , , , , , , –, –, , –, Introduction, Ch.  essences and examples/exemplars, , –, , , , , , , –, –, –, , , Ch. , Ch.  relation among essences, , –, –, –,  the view that Forms are essences, not things that have essences, , , Conclusion essentialism Aristotelian essentialism, , , , ,  Aristotle’s essentialism versus Plato’s essentialism, , , –, –, –,  modern essentialism, , , – Evans, D., 



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/10.1017/9781108986243.014



General Index

explanation. See also causation the principle of the uniformity of explanation, , , –, , , , , –, , ,  Ferejohn, M.,  Fine, G., , , , , ,  Fine, K.,  forms. See essence Frede, M., , , ,  Frege, G.,  Gallop, D., , , –,  Geach, P., ,  Gerson, L. P.,  Gill, M. L., , , , , , – Gosling, J. C. B., ,  Gould, C. S.,  Greene, B.,  Grube, G. M. A., –, , , ,  Hackforth, R.,  Harte, V., , , –, , , , , ,  Heindorf, L. F.,  Heraclitus, ,  Hestir, B. E.,  Hintikka, J.,  Homer,  Horan, D.,  Hume, D., –, ,  Johansen, T. K.,  Kahn, C. H.,  Kant, I., , ,  Karbowski, J.,  Keyt, D.,  Kirwan, C., , ,  knowledge a priori knowledge versus a posteriori knowledge, , , , , , –, Ch.  does knowledge of Forms require intellectual vision?, – epistemology, , –, , , ,  the relation between (the theory of ) knowledge and (the theory of ) enquiry,  Kutschera, F. von, ,  Lachance, G.,  Lee, D. C., ,  Lorenz, H., , 

Malcolm, J., , ,  Mann, W.-R.,  McCabe, M. M.,  Meinwald, C. C., –,  metaphysics as essence-based ontology, ,  in the sense of what there is and what is primary, ,  Moss, J., , ,  Murphy, N. R.,  Natorp, P., , – Nehamas, A., , , ,  Nettleship, R. L.,  Nightingale, A.,  Oderberg, D. S.,  Owen, G. E. L., , ,  Pappas, N.,  Parmenides, , ,  Patterson, R., ,  philosophy debunking attitude to philosophy/deflationary attitude to philosophy, , , ,  Plato and naturalism, ,  Plotinus, ,  Priest, G.,  Putnam, H.,  Quine, W. V. O.,  Reeve, C. D. C., , ,  requirements for answering the ti esti question, –,  requirements of definition. See requirements for answering the ti esti question Rickless, S. C., , , ,  Ross, D., ,  Rowe, C. J., , , , , –, ,  Rowett, C.,  Ruben, D.-H.,  Ryan, P., ,  Ryle, G.,  Sedley, D., , , , , ,  self-predication are Forms self-predicative?, –, –, Conclusion Shostakovich, D.,  Silverman, A., ,  Slings, S. R.,  Smith, J. A.,  Sorabji, R., 

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/10.1017/9781108986243.014

General Index Steinkru¨ger, P.,  Storey, D., ,  Strawson, P.,  substances and essence. See essence Taylor, A. E.,  Taylor, C. C. W., , , ,  Teloh, H.,  Tuozzo, T. M., 

Vazquez, D.,  Vlastos, G., ,  Wedgwood, R.,  White, F. C.,  White, N. P.,  Wittgenstein, L., ,  Woodruff, P., ,  Xenokrates, 

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/10.1017/9781108986243.014



Index Locorum

a–,  Gorgias,  –,  Hippias Major, –, , , , , , –, , , , –, –, ch. –, –, ,  c f.,  c,  d,  e–, , , ,  b–d,  d–e,  d,  a, ,  e–,  a–b,  a–,  b,  b,  b,  b–,  d–,  b,  d,  e–c,  c–,  b,  Laches, ,  Meno, , , ,  d ff.,  Parmenides, , , –, , –, , , , , , ,  e–b,  e–a,  e–a,  e–a,  b–,  b–,  , , ,  a–e, , , –

Aristotle Metaphysics, ,  Gamma ,  b–,  Iota ,  a–,  Lambda,  Lambda  a–,  Mu,  Nu,  Zeta –,  Zeta ,  b–,  Nicomachen Ethics,  Heraclitus B,  Parmenides B, ,  Plato Charmides a–,  Cratylus, , , ,  b–c,  b–,  b–c,  b–c,  –c,  d–e,  d–,  d–e,  d,  Euthyphro, , , ,  d–e,  d–e,  d–e, , – d–e,  e–, 



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/10.1017/9781108986243.015

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



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

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/10.1017/9781108986243.015

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

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

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/10.1017/9781108986243.015

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



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

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/10.1017/9781108986243.015

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

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

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/10.1017/9781108986243.015

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

a–b,  e–a, ,  b,  c–d,  a–b,  b,  b, 

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/10.1017/9781108986243.015



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/10.1017/9781108986243.015