Plato; drawing by David Levine

The great difficulty in writing about Plato is to combine the depth and strength of the Platonic vision with the Socratic subtlety of the arguments by which it is conveyed. Plato’s dialogues are a miraculous blend of philosophical imagination and logic. The interpreter must somehow respond to both, for if the imaginative vision is cut loose from the arguments it becomes grandiloquent posturing, and the arguments on their own are arid, the mere skeleton of a philosophy. So it is already a criticism to say of the books under review that Professor Findlay’s work is all vision, without argument, and that Profesor Irwin’s is all argument with no vision.

In these failings, however, we see two opposed styles of Platonic interpretation taken to extremes. One might describe Findlay’s book1 as a caricature of the old-style visionary approach to Plato, Irwin’s as a caricature of the contemporary analytic mode—were it not that the unusual intensity with which each cultivates his chosen style brings compensating virtues. One-sided both may be, but there have been few interpreters so gripped with the Platonic vision as Findlay, few so remorseless as Irwin in following the twists and turns of the logic. Findlay’s prose is a mildly pompous amble, Irwin’s a formidable obstacle course of reasoning through whose densities one may not pass without first mastering the special code of abbreviations under which all the key propositions are securely camouflaged, but from both one can learn. In the comparison between them one can weigh the rival merits of two powerful styles of interpretation. If, in the end, Irwin’s book is a significant and challenging contribution to Platonic studies, and Findlay’s is not, that is because in the interpretation of Plato, as of any great philosopher, it is the argument, not the vision, to which the controlling priority belongs.

The plain fact of the matter is that if one is going to discourse about a large philosophical vision, one has to get it right, and one will not get it right without a close study of the arguments by which it is conveyed. This is not because the content of the vision is wholly determined by the arguments. In Plato’s case, I believe that he quite often means more than he has arguments to prove, and it is a characteristic fault of Irwin’s methodology that he is inclined to limit Plato’s positive commitments to what his arguments allow a coolly analytic mind to conclude. But Findlay is downright contemptuous of the arguments, and in consequence he is deeply wrong about what the grand Platonic vision actually is.

The cornerstone of Findlay’s Platonism is what he calls the Great Inversion. The ordinary man’s view is that there exist in the world about us sticks and stones, men and horses, etc., and among these many particular things some are in various ways to be accounted good or beautiful, others bad or ugly, some things are straight, others crooked, some are equal to something else, others unequal, and so on for a host of properties of practical or theoretical interest. But when it comes to the properties themselves, goodness and beauty, straightness and equality, and their opposites, and the kinds, stick or man, of which the many particular sticks and men are instances, these abstract general things—if the common view will call them things at all—have at best a secondary, dependent existence.

Goodness in the abstract is a pale shadow beside the good men, good deeds, and good furniture we seek to have about us. Indeed, many philosophers would argue that “goodness” should be explained away altogether as a mere abstraction or generalization from the particular good things of ordinary empirical experience. But for Plato it is the other way round. Goodness, Beauty, Equality, and the like (the capitals are now inevitable) enjoy an eternal unchanging existence as transcendent Forms (Eidæ). The shadows are the transitory empirical instances—earthly goods and sensible beauties—which the ordinary man takes for real but which the philosopher sees as imperfect likenesses of the Forms.

Thus far most Plato scholars would (more or less) agree. The peculiarity of Findlay’s Plato is that he goes on to deny that the empirical world exists at all. The sticks and men of common experience are not things in themselves but merely modes in which the eternal Forms are expressed. Platonism is not the two-world theory which has inspired artists and poets through the ages. It is a single-world theory, but the world in question is a realm of abstract transcendent entities accessible only to those capable of something Findlay calls the “eidetic experience.”

Now Findlay knows very well that there are clear statements in the dialogues which affirm the existence of both Forms and particulars even while emphasizing the differences between them. “Shall we assume two sorts of things that are, the one visible, the other invisible? The invisible being always invariably the same, the visible never the same?” (Phaedo 79a). The interlocutor in the dialogue replies “Yes,” but Findlay says of a similar passage in the Timaeus that Plato was not such a dolt as to believe that reality comprises more than one category of thing, Forms and particulars. He should be understood to mean that Forms and particulars “are” or “exist” in radically distinct senses of the word. In the strict sense these sticks and men are not and can do nothing. It is the Forms alone which properly exist and act. You and I are mere “outflows and manifests” of the causality whereby Man and other Forms exemplify themselves. Readers familiar with the dialogues may object that the explanatory function of Forms has nothing to do with efficient causality. Plato is clear that unchanging Forms cannot be active in space and time, and that is why he provides in the Timaeus a divine Craftsman to make the world on the model of the Forms. Findlay replies by transmuting the divine Craftsman into another Form. The cosmic Craftsman is not a particular mind but the Form Thinkingness and Knowingness, the eidetic Intellection itself. So the thesis is saved: strictly there is nothing but Forms and the doings of Forms.


Findlay also knows very well that the analytic Platonist will think that the single-world theory which has come out of this tinkering with the text is a farrago of nonsense. How are we to conceive the abstract and general thing Equality-as-such activating itself in a pair of equal sticks? Or the abstract general Thinkingness seeing to it that this happens? Findlay does not explain. He writes as if Thinkingness undertakes thought to organize the exemplification of itself and other Forms, but ordinarily he denies (brushing aside rather considerable evidence to the contrary) that Plato was ever tempted to regard his Forms as “self-predicative.” (Examples of self-predication would be “Goodness is good,” “Beauty is beautiful,” “Equality is equal,” so here “Thinkingness thinks”: it is not unimportant that some examples are more patently nonsense than others.) Nor does Findlay deign to justify his overruling Plato’s explicit indication that “the maker and father of this universe” is not a Form but the best of causes of the best of the things that have come to be on the model of the Forms. Promised a splendid metaphysical vision, we are left with eccentric posturing, flagrantly at variance with the text of the dialogues.

Fortunately, Plato is much more interesting than that—if you read rather than rewrite what he has to say. I mean by this, what he has to say in the dialogues. Plato is one of a mere two or three ancient Greek philosophers whose written works have come down to us in their entirety. Whatever “publication” meant in those days, we have everything he saw fit to publish plus some unfinished remains. The doctrines and arguments he wished to be judged by are there for us to see. Findlay presumes to “correct” the dialogues by the so-called “unwritten doctrines” deriving from fragmentary reports in Aristotle and others of certain obscure mathematicizing speculations which did not get written up in the dialogues. Findlay offers the rather good suggestion that they were not written up because they never got beyond the stage of programmatic aspiration. They were a project for a mathematical analysis of the Forms which Plato did not manage to work out satisfactorily. How extraordinary, therefore, not to say presumptuous, to set up this merely programmatic aspiration as the esoteric core of Plato’s philosophy, reading the dialogues as partial, allusive glimpses of the real thing. Findlay is not alone in this attitude—the esoteric interpretation has been much in vogue in certain Continental universities—but the historical basis on which the reconstruction rests is shaky in the extreme.2 On Findlay’s own showing, moreover, the “unwritten doctrines” presuppose an understanding of the Theory of Forms. Hence any reading of Plato which does not respect the dialogues’ more argumentative treatment of the Forms will be historically unsound.

Here it should be noticed that it is a condition on any “single-world” interpretation of Plato’s metaphysics that there should be nothing in this world (this world as the ordinary man experiences and conceives it) which does not derive its being from a corresponding Form. Let there be just one item “down here” which is not an “outflow and manifest” of the Forms, which enjoys an existence other than as a mere mode of Form, and the whole edifice falls. One stick will do—to my knowledge no rigorous Plato scholar has ever championed Forms for sticks and stones. True, Findlay postulates (without textual warrant) Secondary and Tertiary Memberships of the Ideal Order to cope with assorted bad, negative, or merely relative things. This is one of the points on which he uses “unwritten” material to “correct” the dialogues.3 For the dialogues give pride of place to Forms for relative notions such as Equality and Similarity, and treat the negative of a pair of opposites (Injustice, Inequality) on the same terms as the positive.


But the real crux is that Forms for natural species like Man and Horse do not make their appearance until a comparatively late work, the Timaeus. In the Phaedo, Plato’s first and in some respects his fullest presentation of the Theory of Forms, sticks and stones, men and horses are contrasted, as typical sensible particulars, with Forms such as Equality and Beauty which they may from time to time exemplify. There is not a whisper of a Form of Man or Horse, let alone of Stick or Stone. But these would be essential if Plato meant his theory to be a comprehensive account of existence. In another dialogue, the Parmenides, Plato has Socrates hesitate about admitting a Form of Man. Such uncertainty shows it was a real question whether to posit a Form corresponding to a given term. How so, if, as Findlay believes, the burden of the Theory of Forms is that everything “down here” is nothing but an “outflow and manifest” of the Forms? This objection, it must be said, applies to many standard accounts of the Theory of Forms, not to Findlay’s alone. One of the virtues of Findlay’s enthusiasm for his own version of Platonism is that he follows it through with such forceful commitment that one can see more clearly than with other writers that it is his own, not Plato’s.

Anyone who attends to the arguments in the Phaedo can see at once why the original Theory of Forms makes no mention of a Form of Man or Horse, but only of Forms for such terms as Equality and Beauty. The arguments which yield transcendent Forms of the latter type simply would not work for the former. For the arguments have nothing to do, with modes of existence. They are concerned exclusively with the manner in which predicates like “equal” and “beautiful” attach to their subjects. A stick which is equal to one thing is unequal to another; a man who is beautiful at one time or in one context of comparison is ugly at another time or in another context of comparison. Such things are both equal and unequal, both beautiful and ugly—a coincidence of opposites which would be a contradiction unless the predications are taken as relative to times and contexts. The point is that in this world we never meet pure instances of equality and beauty. Earthly beauties are always relative beauties, beauties which in another context or aspect turn out to be ugly; any sensible example of equality is in another relation just as good an example of inequality.

It should be obvious that there is no straightforward way of making out that “man” and “horse” are affected by the same kind of relativity. No opposite predicate stands to “man” or “horse” as “ugly” stands to “beautiful” or “unequal” to “equal.” Bucephalus is not a horse at one time or in one context and something else at another time or in another context. So if the thesis of the coincidence of opposites (at the same or at different times) is a ground for positing Forms of Equality and Beauty, it is not also a ground for positing Forms of Man and Horse. That is why the original Theory of Forms makes no provision for the latter.

But what about Equality and Beauty? Presented with the above considerations about the coincidence of opposites, a coolly analytic mind might conclude simply that predicates like “equal” and “beautiful” are irredeemably relative or comparative. Equality is always a relation between two things, beauty just is transitory and comparative, and that’s that. But not Plato. He demands an Equal that is not relative, a Beauty that is not transitory and comparative, and since for the reasons given no such thing is to be met with in the empirical world he locates them in a higher realm of transcendent Forms. Here is the true beginning of the great metaphysical vision. But what is its justification? Why won’t Plato stop with the coolly analytic conclusion of the relativity of equality and beauty and learn to live with it?

To my mind this is the hardest and most pressing question about Platonism, which surprisingly few scholars try to answer. Terence Irwin is one of the few. Not the least of the signal achievements of his book is the manner in which he exhibits the Theory of Forms as the outcome of Plato’s long engagement with the thought of his mentor Socrates. It is of course a commonplace that to understand Platonism we have to go back to Socrates. But usually this is taken to mean that Plato carried on where Socrates left off, developed the Socratic doctrines in new and more metaphysical directions. Irwin’s account is dramatically different. Mature Platonism, for him, is root and branch opposed to the Socratic philosophy. The Theory of Forms was an act of parricide. This challenging new perspective is reached by an exceptionally thorough and acute analysis of the arguments in the early “Socratic” dialogues, on the one hand, and the more characteristically “Platonic” dialogues like the Phaedo and Republic on the other, the two groups4 being set against each other on point after point to weave a developmental story of extraordinary coherence.

Since, then, we are to come at the Theory of Forms by way of Socrates’ theorizing about the good life, let us review some of the themes of the early dialogues. First and foremost, virtue is knowledge—knowledge of what is good or beneficial for oneself. The ordinary man distinguishes courage from piety and each of these from justice, but he can only do so because he identifies courage, say, with confidence in situations of danger and justice with something different, such as certain kinds of regard for other people’s interests. Socrates has no difficulty in showing that in some situations of danger confidence as such is overconfidence or rashness, not courage, i.e., it is bad rather than good. The qualities of character according to which one might seek to differentiate the virtues one from another are not unconditionally good or beneficial, as a real virtue worthy of the name must be; hence the ordinary man’s differentiating definitions do not fix real, reliable virtues. The only unconditionally beneficial attribute is the knowledge which will make the right use of confidence, choose the right way to express a regard for other people’s interests, and so on. True courage, genuine justice, is the possession of that knowledge, which knowledge therefore is what you need for the good life.

But conversely, according to Socrates this moral knowledge is all you need for the good life. The ordinary man thinks that one may have the wisdom to know the right thing to do and still need courage to carry it out or temperance to resist the temptations that lead away from the path of virtue. But if, as Socrates has argued, courage and temperance are the wisdom that knows the right thing to do, there is nothing further to add. A man has only to know the good and he has all he needs to carry it out. Contrary to common belief, there is no such thing as acting against one’s better judgment; no such thing as moral weakness. Hence (to compound the paradox) if a man makes the wrong choice and acts badly, it must be because he did not know it was wrong. And if he did not know, the wrong he did was involuntary. No evil-doer, however monstrous, is a voluntary agent in his crimes.

It may well seem scarcely credible that anyone should arrive at conclusions of such violent paradox without stopping to ask whether something hasn’t gone wrong on the way. But Socrates does ask this, often. He does not claim to know that his conclusions represent certain truth. He does not himself have the moral knowledge he is looking for. When he is confident of his paradoxical conclusions, it is because they have emerged and been tested in the fire of many argumentative discussions, have held their own against all manner of rival views, have won the reluctant or willing assent of every type of interlocutor. This process cannot be described in a summary way. It must be read first-hand in the Laches, the Protagoras, the Gorgias, and those other early dialogues where Socrates’ living presence stamps his every utterance with a heroic integrity of mind and character, a total commitment to follow the argument wherever it may lead and to believe on the strength of it that violent paradox (as it seems to everyone else) is plain unvarnished truth.

There is something else we can do to understand Socrates, something that (as Irwin brilliantly shows) Plato must have done as he worked over the Socratic arguments in the early dialogues. We can look for the assumptions or principles which so guide Socrates’ discussions that the argument always tends toward the same paradoxical conclusions. This is the project Irwin has made his own.

We have already met one such principle, which bulks large in Irwin’s reconstruction: the idea that any virtue worthy of the name must be an unconditionally beneficial possession. And this might be challenged. (Are there not tragedies in which the very virtues of a protagonist prove to be his undoing?) But what is meant by “beneficial?” We must consider two alternatives which have important differences: that virtue, properly appreciated, is supremely valuable in itself, for its own sake; or that virtue contributes to some further end already valued by the agent. Irwin’s hypothesis, and a startlingly controversial hypothesis it is, is that Socrates means the second. We are not virtuous for virtue’s own sake but solely for the sake of the happiness which will come out of it. Grant this hypothesis and a number of the pieces fall into place.

First, if virtue is knowledge of what is good or beneficial to oneself in the instrumental sense, it becomes, in effect, a case of craft knowledge, like carpentry or cobbling. It is a principled skill or ability to engineer a certain product, namely happiness or the final good. That is why Socratic arguments so often draw on the analogy of the crafts, because Socrates thinks of virtue as the art of making oneself happy. But second, happiness is something that all of us desire, both the virtuous and the nonvirtuous. So if the difference between the virtuous and the nonvirtuous is a difference between the man who has and the man who lacks the art of securing happiness, it is a difference not in desire or ultimate aim but in belief about what will satisfy our desire, about the means to happiness. We can now see why Socrates puts so much emphasis on argument and discussion. The wicked man who exploits his fellows for selfish ends has made a mistake about what will make him happy: he is not gaining the final good he ultimately desires. Help him to correct his mistaken beliefs, put him on the road to acquiring the true art of happiness, and he will change his ways. He is bound to change his ways because his heart, as it were, is already in the right place; only the mind errs.

If this seems like explaining one set of incredible paradoxes by another, it is in part because I have left out all the elaboration and detail of Irwin’s reconstruction: the skillful marshalling of evidence from Socrates’ use of the craft analogy, from his arguments about virtue and happiness, from his analysis of the springs of human action, in short everything that makes the book a tour de force of philosophical scholarship. But when all that is taken on board, the unease remains. The key premise, that virtue has only instrumental value, is just as paradoxical as the paradoxes it helps to explain. It puts courage, temperance, and justice on a level with being a safe driver or following a sensible diet, and to my mind this rings untrue to the man who could accept the sentence of death, from the very citizens he had tried to educate, with the words “No evil can befall a good man either in life or in death” (Apology 40d).

Do those words mean that if I have the art of making myself happy (Socratic virtue as Irwin construes it), not only can I achieve my aims so far as this depends on my own actions, but I can prevent evil coming to me by way of other people or natural disasters frustrating my plans? It would be plain silly to believe in a craft with such superhuman powers as that. Socrates must rather mean that no outside disaster can take away the wisdom which is courage, temperance, and justice, and that for a good man only the loss of virtue itself would really count as evil. Anyone who prefers the second interpretation, as I do, is bound to conclude that for Socrates virtue and knowledge are of supreme value in and for themselves.5 In which case, when Socrates ascribes happiness to the life of the virtuous man, he must mean, as Aristotle did later, the happiness of living that life itself, not a separate product to which virtuous living is a mere means.

There is a difficulty in any case about specifying, on the instrumental interpretation, what happiness can amount to, if it is to be the final good which both the virtuous and the nonvirtuous can agree they want more than anything else in life. More, indeed, than life itself, if a commitment to justice can demand, as it did from Socrates, the surrender of one’s own life. Socrates never achieves a settled specification of happiness, and no wonder. What on earth could the virtuous and the nonvirtuous agree upon, beyond the name “happiness” or “the good”? Irwin acknowledges the difficulty, but not as a weakness in his interpretation of Socrates. Rather, it is a weakness in the Socratic theory itself, as he has interpreted it. And on that claim he mounts the second major hypothesis of the book: that one determining consideration which led Plato to the Theory of Forms was precisely the difficulty of specifying the good in neutral, undisputed terms on which both the virtuous and the non-virtuous could be supposed to agree.

The difficulty is that if you try to define the good by some neutral, empirically decidable feature of good things, you come up against the now familiar coincidence of opposites: any neutral feature will be good in some circumstances, bad in others. If, on the other hand, the good cannot be so defined, it is no longer reasonable to maintain that the evildoer desires the good and is mistaken only about the means to it. His desire is misdirected or corrupt, not only his belief—and once that is admitted, a host of further changes in moral theory must ensue. The craft analogy for virtue goes: there is no chance now of the wrongdoer mending his ways unless he can be brought to value virtue intrinsically, as something good in itself. And for this, logical argument of the kind Socrates practices will scarcely be enough. Desire itself must be trained, as in the elaborate social program devised for Plato’s ideal state in the Republic; or it must be taught to find through philosophical reflection new and more satisfying objects to aim for, as in the famous ascent of desire toward ideal Beauty in the Symposium—on these and other developments in the “Platonic” dialogues Irwin has interesting and helpful things to say from the innovating anti-Socratic perspective he wants us to adopt. But our question was: Does Irwin’s story of Plato’s dissatisfaction with Socrates give a reasonable explanation of the Theory of Forms?

Many philosophers have learned to live with the idea that moral predicates like “good” and “just” are not definable in neutral, empirically decidable terms. Why didn’t Plato? According to Irwin, because he did not formulate sharply for himself the difference between (1) justice or goodness being independent of (not definable in terms of) the empirical properties of sensible things, (2) justice or goodness being independent of sensible things. (1) is supported by the argument from the coincidence of opposites, and Irwin rightly denies that it is meant to yield a Form for every general term. But even for the cases to which the argument does apply, (2) alone yields Platonism, the thesis that justice and goodness exist on their own separately from the sensible things which are concrete instances of justice and goodness. Because Plato does not distinguish (1) and (2), he confusedly takes the coincidence of opposites as a ground for positing transcendent Forms of Justice and Goodness.

So Plato’s great metaphysical vision rests on nothing more than an unclarity or confusion? I would find that the hardest paradox of all. The truth is, there is an important dimension of Plato’s thinking about these issues which Irwin has left out. Curiously, it is a dimension to which Findlay is partially blind as well. I refer to Plato’s provocative doctrine that, in morals and mathematics and perhaps in other branches of knowledge as well, learning is recollection, a recovery from within oneself of knowledge enjoyed by the soul before it became incarnate in a human body.

In a celebrated passage of the Meno, a work intermediate between the “Socratic” and the “Platonic” dialogues, Socrates confronts an uneducated slave with a geometrical problem and shows how through skillful questioning he may be led, first to see what is wrong with his initial mistaken solutions, and then to arrive at the correct answer—at each step relying solely on internal resources of his own, not on information proferred by his questioner. The scene is a model demonstration of educational methods emphasizing “discovery.” But even the most favorable reading of it will not support the conclusion Plato draws, that the slave is recollecting knowledge which his soul enjoyed before birth. The most Plato shows is that the slave brings to the scene an implicit knowledge of certain mathematical notions which Socrates’ questions help to make explicit; it remains open to dispute where he got that knowledge from. Very likely someone did dispute that very point, for in the Phaedo Plato tries again, focusing this time on the key mathematical notion of equality. The argument about equal sticks and stones in the Phaedo is designed to show that, because any sensible example of equality is also an example of inequality, our knowledge of equality as such cannot be derived from the observation of equal things. But it is an explicit premise of the argument that we have such knowledge, as demonstrated in the Meno, even if it takes dialectical cross-examination to bring it to an adequate and fully conscious formulation. So Plato concludes that the knowledge of equality is independent of our empirical existence.

It would be possible, on the strength of this argument, to hold simply that mathematical knowledge is innate, something we bring to this world in the sense that we are born with it. Modern rationalist philosophers have indeed claimed that the scene in the Meno proves just that; they share Findlay’s inclination to play down pre-existence as “mythic,” or Irwin’s to ignore it altogether. But a demythologized innateness hypothesis would run counter to something Findlay himself rightly stresses, Plato’s deeply realist conviction that knowledge is of something which exists independently of the knowing mind. We come into the world carrying within us an implicit knowledge of equality, of goodness, of justice—for the argument is supposed to hold for morals as well as mathematics—but the equality we know is not born with us. I: exists to be known independently of our knowing it. We know it independently of our empirical existence, with a knowledge rooted in the very nature of our souls; so Plato infers that the thing known exists independently of all empirical reality. The epistemological premise is crucial to the positing of Forms: “If understanding (knowledge) is an altogether different kind of thing from mere true opinion, it is certain that these Forms exist in their own right” (Timaeus 51d).

If this is correct, it means that what Forms Plato posits will depend on what subject-matters he thinks are such that knowledge or understanding of them is rooted in the original, nonempirical nature of the soul. It is from that perspective, to mention only one of the many issues this suggestion raises, that I would investigate the “unwritten” program for a mathematical analysis of the Forms. For certainly not all subject-matters, and not all that creditably passes for knowledge, can be grounded in the very nature of the soul; and it may be that in the end Plato came to think that only highly abstract mathematical structures would fit the bill.

I do not, however, claim that the epistemological premise is sufficient justification for the Theory of Forms. Here, if anywhere, Plato’s vision outruns the resources of his argument. But Plato himself tells us that the Theory of Forms and the pre-existence of the soul stand or fall together (Phaedo 76e). My contention is that we shall not understand the Theory of Forms unless we see it as correlative to a cosmic vision of the soul’s existence in which our present embodiment is but one passing phase. In due course the vision is crowned by the appearance in the Timaeus of a divine Craftsman to construct the soul and the whole physical universe on the model of the Forms: it is God’s knowledge which requires Forms of natural species such as Man and Horse.

All this is heady stuff, hard for the analytic Platonist to enjoy. But it is no less a part of the essential Plato than the arguments, no less crucial to his metaphysics and his moral theory. In particular, Plato’s moral theory is less Aristotelian than Irwin makes it appear, because Irwin is uncomfortable with the transcendent, otherworldly dimension which Plato, unlike Socrates, has constantly in view. No doubt it is idle to think there could be such a thing as a definitive interpretation of a philosopher as rich in ideas and as many-sided as Plato, but if there is a permanent lesson to be learned from the endlessly fascinating attempt to comprehend his thought, it is that in a truly philosophical reading vision and argument must contrive to meet, as they met in the towering genius it is all about.

This Issue

September 27, 1979