publisher colophon



The Rationalism of Plato

THOUGH WE ARE CONCERNED in this book with individual philosophers only in so far as they modified the rationalistic tradition, there are certain individuals whose influence has been so great that they must be treated in separate chapters. One reason for this is that their contributions to philosophy are numerous, and if they were dealt with along with their neighbors in time, chapters would become unduly extended. Another is that their extant works are very copious even when a few exist only in truncated form. A third is possibly a corollary of the second, for when an author leaves a great many works to posterity and must, because of the nature of things, write them one after another, and therefore during a long lifetime, he exhibits certain peculiarities of style and certain inconsistencies which are the concomitants of a long life: a man of eighty does not usually express himself in the same way, hold to the same interests, believe in the same things, as a man of twenty. And if the philosopher in question for one reason or another is read by many people, even when most of them are disciples, there will arise a variety of interpretations of his thought. What was the philosophy of Plato, Aristotle, Plotinus, Saint Augustine, assuming that each one of them had one philosophy, is difficult to answer, since, unless one is unusually presumptuous, one has to admit that one’s interpretation may be erroneous and is usually at best only plausible.

In Plato’s case we meet with a philosopher who has suffered from idolatry as well as from downright dislike. Both the ancients and the early Christians looked to him as the source of their doctrines, as if their consistency with his dialogues would give added proof to their theories. This is the more curious among the early Fathers with their theory of a praeparatio evangelica, a notion revived in our own time by Father F. C. Copleston,1 for one might think that the Source of all light would produce His revelations without the help of heathen thinkers. The revelation to Abraham seems to have been preceded by no teaching from the Sumerian or Egyptian sages, and there is nothing, as far as I know, in the Prophets which any Biblical scholar asserts to have been prepared by Philistine, Persian, or Indian literature. It is true that in Saint Paul there are two references to Greek writers, but Paul was far from using them as authority for beliefs which he had been taught through revelation. We are fortunately not burdened with the task of untangling the pagan sources of Christian doctrine; our mention of the praeparatio evangelica is made only because one version of Platonism has been colored by those who hold to it. One of the dialogues, Timaeus, has been so Christianized by its readers that it is next to impossible to read it as something written three hundred and fifty years or so before the birth of Christ. In fact there is nothing in Plato’s works or in his life which has not been a matter of debate. Fact and legend have been confused; the chronological order of his works has been so jumbled that there is only one book whose position in the series is generally agreed upon, the Laws, and even that has been held to be spurious. We need not go into the question of his biography unless we think that the dialogues are inexplicable otherwise.2 And since the inconsistencies in his works are of minor importance, I shall omit any discussion of them. Also, since we are not concerned with the development of his thought, I need not go into the chronological order of his works. Finally, since there is enough material to meet our needs in the works mentioned by his most eminent pupil, Aristotle, and accepted by the great majority of scholars as genuine, we shall consider them as authentic and base our argument upon them.


The doctrine of the two worlds appears in Plato in a new form. Reality is no longer that which constitutes the substance of the physical world, a kind of matter which is diversified according to the conditions under which it appears, nor yet a law which governs the flux of things. It is not even the persistence of the flux. It is not the Pythagorean numbers, whatever those numbers may have been, though in Timaeus the world is best understood more geometrico. It seems to be rather that which classes of things have in common, even when it is not found incorporated in things themselves. These common properties appear to be the start of his conception of the Ideas. And it is also probable that Plato derived the conception of the Ideas from his master, Socrates, whom he so frequently represents as trying to induce his pupils to give him a definition of a class of things, or, if one prefer, of an abstract noun. This appears with special clarity in one of his less entertaining dialogues, Hippias Major. Here Plato, as if replying to Xenophon’s Symposium (ii and v) where Socrates identifies the beautiful with the fitting,3 gives us the Sophist Hippias struggling to see the need for general ideas, as he gives only examples of beauty to Socrates. Gold is beautiful (289e); riches, health, and a fine funeral are beautiful, says Hippias (291d), but he never reaches the point of clarifying the idea of beauty itself. But the same technique of trying to get a person to attain to a vision of common characteristics is found in many of the dialogues which have been called dialogues of search. In Charmides the goal is the idea of temperance or good behavior, in Lysis of friendship, in Laches of courage, in Theaetetus of knowledge, in the Symposium of love. In all of these dialogues the pupils, as one might call them, seem able to produce examples of the idea whose definition is being sought, but unable to frame the definition itself. Definitions can be given only of classes, not of individuals, and Plato seems to have been the first philosopher to recognize the ineffable nature of particulars. Particulars can be seen, felt, observed, but they are logical surds. For every time we know something, as distinguished from observing it, every time we put that knowledge into words, we are forced to use common nouns which of course are general terms. If then gold and fine funerals are beautiful and the word “beautiful” means the same thing wherever it is used, then all beautiful things must have something in common which we can define when we know it.

Yet nowhere that I know of does Plato say that the common characteristics are observable. Nowhere does he have Socrates make comparisons between things and ask what common perceptual properties they have. What he is looking for is some reason why two things should be given the same name and that reason is not a generalization from sensory experience at all. Socrates’ pupils, when they are not so obtuse as Hippias, offer definitions of the idea under consideration, and Socrates’ search is in the direction of seeing whether the definition can be made self-consistent. If, for instance, Plato was trying to define circularity, he would not be satisfied with examples of circles such as wheels, shields, coins, but would insist on a verbal statement which would be approximately true of the examples but which would not be abstracted from them. In the jargon of today’s philosophies, for Plato essence is prior to existence, and its priority is in the field of cognition. The distinction between knowledge and observation is ultimate.

The common properties are not differentiated from their particulars merely by the generality. They are also differentiated from them by their timelessness. Whereas all beautiful things may be wiped out in the course of time, beauty itself, he believes, is something which remains stable and immutable. There seems to be, moreover, no special reason why any idea should ever be exemplified, though in Timaeus he says that eventually all possibilities must be realized. The history of mathematics shows how certain ideas can be deduced from what used to be called self-evident premises and are now called postulates, ideas which are not only not exemplified, but probably never could be. We are all brought up to believe that no perfect circle will ever be drawn on paper, nor really parallel lines. We may refine our drawings as we will, we are told, but there will always be a margin between the perfection of the mathematical ideal and the crudity of our exemplifications. Since the rise of Christianity, the thesis has been extended to moral matters. No man will ever be perfectly good this side of Paradise, and in fact it was only through the Incarnation of God in man that goodness of such perfection was ever seen on earth. And if no premises are self-evident, then the truth of no system of theorems can ever be any greater than the truth of the postulates, and consistency has to take the place of truth-to-fact. Ideas then are not dated. They have no situation in history. They are eternal. In fact, in the Republic where a definition of justice is sought and found, Plato makes it clear that no such idea has ever been in existence and that it probably never will be. Nor does it make much difference. If in naming the Ideas he used a noun which is etymologically related to the verb to know, it is because they represent what can be known, thus sharply setting them apart from what can be experienced in the world of history.

But in Greek the verb to know and the verb to see are closely related, as if to know something was to see it. We preserve this metaphor in English when we say that we see what is meant when we understand something. What is “seen” is the meaning of the idea in question. Presumably what Plato is looking for is something which could be seen in this way and which would preserve its self-identity wherever it might be found. The analogy to the beings of mathematics is again clear. For whether circularity is found in a wheel or in the orbits of the planets, it is always a circle as described by geometry, and the mathematical definition gives us a test by means of which we can tell whether we have a circle before us or something else. The self-identity of the Ideas depends on their immutability and the only things which could be immutable are those to which time is irrelevant.

An idea in this sense of the word is similar to an ideal. Since particular things in the world of history, of time, of change, of growth and decay, are never what they ought to be, never what their definitions say they really are, there must be some standard by means of which we determine both what things really are and how far they realize their natures. If we know what a frog, a dog, or a man really is, and we see that a given frog, dog, or man falls short of what the definition states, then we also see that particulars are never quite what they should be. And we soon understand that there must be certain ideals which we use to determine the nature of things and the degree of perfection which the things in question exhibit. That is, a given man may approach closer to the ideal of humanity than another. But as soon as we say that, we have assumed that we know what ideal humanity is, and when we use that ideal to judge the perfection of given human beings, men of the past as well as of the present, men of other civilizations as well as of our own, we have admitted that the ideal is not changed by local or chronological circumstances. Here a paradox enters our philosophy and one may well raise the question of how the ideal both can be the common properties of a group of things and yet not be manifested in any of them.

Now Plato, as is well known, usually presents his theories through the character of Socrates, and Socrates is said by Aristotle to have introduced into philosophy “definition by induction.” This is supposed to mean that he first assembled a collection of similar things and then drew out of them by inspection that which they had in common. The paradox arises from the obvious fact that he must have known which things were similar when he began to assemble them and that consequently his search for their similarity was unnecessary. But the paradox is attenuated, if not resolved, by our common procedure of classifying things in accordance with some superficial similarity which can be easily perceived but which turns out to be only superficial. In the history of science we find many examples of this. The theory of the four elements is a case in point. We do not know why all material substances were grouped into earth, water, air, and fire, but we do know that when the classification was made, each of the elements had a single character which determined the behavior of the compounds into which it entered. Earth, for instance, always moved to the center of the world below water, and if anything which contained earth was moved upward by force, it would drop back to its normal position. To have imagined that diverse things could nevertheless have common properties also involved believing that such properties need not be sensory qualities, colors, shapes, sounds, and the like. It also involved assuming that if we could predict how anything would behave under definable circumstances, it was because of the presence of such properties, either overt or concealed. If we say that a good man will always be courageous or temperate, then we can abstract courage and temperance from the men who possess it and examine those traits independently of the conditions under which they appear. The scientist may say that a falling body ought to fall with an acceleration of about 32 feet per second, and he means that “other things being equal” it will so fall. Furthermore, he knows what other things have to be equal. But the use of “ought” suggests that if the thing in question does not behave as it ought to, then something is wrong. What is wrong may not be perversity or sinfulness; it may be simply something or other which prevents the thing from behaving the way the scientist says it ought to behave. As every freshman knows, a body falling through the air or water does not behave as the Law of Falling Bodies says it should behave. And he can make allowances for what he will call perhaps the resistance of the medium.

This would seem to presuppose that there is an order of things according to which classifications may be made such that all things of the same class will behave in the same way or be characterized by the same traits. That order is not the order of things observed outside of laboratories. It also presupposes that when that order is established, there are things in it which do what they ought to do and are what they ought to be. Now when we call them “things,” it looks as if they ought to be like the particular things in the world of space and time. And indeed some followers of Plato did talk about them as if they were such beings. But their very perfection, their timelessness, their lack of particularity, their capacity of being used as standards, differentiate them from particulars. Our vocabulary, and to a greater extent that of Plato, makes it impossible to speak of them other than in terms derived from “experience.” Metaphors and similes, myths, have to be elaborated in order to handle them intellectually, and the simple literal-minded person is bound either to misunderstand their nature or to conclude that Plato was as simple and literal-minded as he himself is. Thus when Plato speaks of everything seeking the good, it looks as if there were one good thing which all things sought and, since the adjective “good” is usually found in the context of morals, it begins to look as if Plato’s universe was a universe of moral goodness. But when it is seen that the good is the perfection of each kind of thing, that is, its nature as purified of all the accidents of space and time, the problem of the good is transformed into the metaphysical problem of generic natures.

It cannot be denied, however unfortunate it may seem to enthusiastic Platonists, that the good is always the character of the class, not of the individual. If one can discover the nature of humanity, it will be that which all human beings share in common when the accidents of history are eliminated, but it will also be an eternal standard which all human beings should try to exemplify. This is not an ethics of “to your own self be true.” It is on the contrary an ethics of universal ends. Plato in the Republic recognizes that individuals differ profoundly in their ability to attain those ends and maintains that two types of man, those who are dominated by their appetites and those who are dominated by irascibility, never will be really human. They are incomplete human beings. But he is not so pessimistic as to believe that no human beings are endowed with the capacity to be really human and merely insists that circumstances must be so ordered that they be permitted to be so. But similarly with all generic traits: members of the various classes fall short to different degrees of attaining their real natures. A dog with three legs, a calf with two heads, a mathematical proof which is either inelegant or fallacious, are also monstrous deviations from what they ought to be. The dog and the calf will be defined in part as quadrupeds which are monocephalous, and their monstrous character is recognized as monstrous only because we know that they ought to have four legs and one head. We can know what caninity and bovinity are; we can state their definitions clearly and precisely. If we could not do this, we should not be able to distinguish one kind of thing from another, though we could distinguish their existential plurality. But the very fact that we can frame and accept generic definitions, as we are used to doing in mathematics, permits us also to detect imperfections. One of the troubles with this technique is that the history of thought has shown us how the principles of classification have changed and how in consequence of this, the ideas—or ideals—of Plato have been abandoned or seriously modified. But this is a small trouble, for we still use classifications made in a manner similar to that of Plato, based on the assumption that the world is orderly and that the things in it fall into classes the traits of which we use as standards.

At a later time—especially in Philo Judaeus—a term was coined for the perfect order of the world, for those beings which are the standards of all things. This term is the “Intelligible World.” It is contrasted with the “Sensible World.” It would be anachronistic to use these terms in expounding Plato’s own philosophy, for there is no evidence in the dialogues that he attributed to the ideas any character which would help organize them into a world at all. Each dialogue which is relevant to the problem attempts to discover the nature of some one idea; no dialogue establishes any interrelation between the various ideas. This is something which one would do well to remember, for in later thinkers it became customary to assert that the ideas formed a sort of hierarchy with the Idea of the Good at the apex and all the other ideas stemming from it, sometimes in order of generality. Plato does say that the idea of the good is preeminent, but that is simply because each idea is the good for the members of the class which it controls. But that is in no way saying, as some geometers might in their ignorance say, that since all geometric figures are in space, they can be deduced from the idea of space, or that because one can arrange such figures in order of the number of their sides, the triangle, the quadrangle, the pentagon, the hexagon, and so on, they come into existence in that order in the world of spatial beings. One does not generate—or define—a quadrangle by adding a side to a triangle and, strictly speaking, one can define a quadrangle without any reference to triangles and indeed in ignorance of their existence. Logically one might imagine that if there are three-, four-, five-sided figures, there ought also to be one- and two-sided figures from which the polygons “flow.” There is then in Plato no evidence of a logical relationship between his various ideas; they all have certain common characters: eternality, unity, difference from one another, immutability, and of course the possibility of being incorporated. In fact, in Timaeus (29d, e), which is the source of most of the fantasies drawn from Plato, he maintains that all possibilities must be realized, a statement which, as Love joy has shown,4 gave rise to “the Principle of Plenitude.” This principle was not developed by Plato himself into its consequences, one of which was the scala naturae. On the contrary, in general Plato seemed to think of his ideas as related to one another in a network, not in a hierarchy.

This seems to be pretty well illustrated in such a dialogue as the Lysis. One might say that this dialogue attempts to define “friendship.” No satisfactory definition is found, though Lysis himself is both good and handsome and a friend. A series of tentative definitions is presented by the boy to Socrates. They run about as follows.

(1) Friendship is that which is shown in a mother’s relation to her children.

(2) It is based on one’s recognition of the usefulness of others. But since one man may find another useful, and thus of two men, A may be a friend to B, while B is not a friend to A, in the case of two friends one may not be a friend.

(3) Friendship is the attraction of similars. But then two bad men may be friends, and apparently Socrates and his pupil think that friendship ought to obtain only between good men.

(4) Hence perhaps good men are friends. But the good is the self-sufficient, a mark of goodness which we have already noted and which will recur in ancient philosophies, and the self-sufficient desire nothing, love nothing, and hence have no friends.

(5) Can it be then that friendship exists only between contraries and that we love that which is different from us? But this is absurd, for though a bad man might be friendly to a good man, a good man could not possibly be friendly to a bad one. It remains then that there might be some things which are neutral, neither good nor bad.

(6) In that case, friendship would be either the relation between the good man and the indifferent or between the good man and the good. The indifferent—or the mean—might love the good in order to replace the evil from which it suffers, so that we come to another tentative conclusion:

(7) Friendship is the love of the mean, or indifferent, or for the good caused by the presence of evil. This arouses, however, new doubts. In all friendship there must be an object. One cannot just be a friend in a vacuum. What is there in the objects of friendship which is common to them all? If one cannot love someone for the sake of the good, since the good has no need of love, and since the good would not exist if there were no evil, only the mean and the desire for the good would remain. Evil cannot be the source of friendship, for nothing good can come from evil. But it seems strange that only men in a middle position between good and evil could be friends. For we find men of different degrees of perfection being friends of each other. The concluding possibility is:

(8) That you love a person who is somehow like you “in nature” (Lysis 221), a phrase which has been translated “congenial.” If “congenial” means being of the same genus, sharing the same nature, then it will do. But it is unsatisfactory since all men share the same nature and yet all men are not friends. The outcome seems absurd to Socrates and the dialogue terminates.

It is obvious that the definitions grow out of difficulties, logical or existential, with their predecessors. They are not made in order of generality, nor are they discovered by observing friends and drawing off from them their observably common characters. The speakers are made to understand what is required of a satisfactory definition of “friendship” and simply to test each one proposed by its logical consequences.

The philia or friendship which is discussed in this dialogue is love and in the Symposium Plato was to examine the matter more closely. Love is not pederasty, for which Plato has nothing but contempt. It is a warm and very affectionate friendship which makes one man want the company of another, a desire which may have its roots in sexuality, but if so, the roots are well concealed. Lysis himself is an example of a friend and yet is incapable of defining his own nature. It is this incapacity which is at the heart of the drama. It is one which appears in regard to other traits in Charmides, in Laches, and in Theaetetus, and it illustrates a point which Plato frequently makes, that one may be the incarnation of an idea without knowing what it is. This is very important for an understanding of his theory of knowledge and throws some light on a remark of Socrates in the Apology to the effect that an unexamined life is not worth living. An examined life is one in which all our motives are brought into the light of day, clarified by rational analysis, and made to resist criticism. But for our immediate purposes it is enough to point out that though the Intelligible World is the archetype of the Sensible World, it is not observed through our sensations. This of course is just common sense. We observe thousands of things daily without knowing what they are.

There is another peculiarity of this point of view which should not be passed over. That is Plato’s belief that the nature of things is whatever it is independently of our knowledge of it. He is far from being a subjectivist in his metaphysics. We discover natures; we do not produce them either by our powers of observation or by our methods of inquiry. When he speaks of the love between beings which are alike in nature, it is the words “in nature” which are important. Regardless of the maddening ambiguities of that term, in Plato it names something which is certainly objective. He would be willing to say that what we call the nature of something might be determined by our special interests, as the nature of water for a cook or a sailor or a fireman would be different from what it would be for a chemist. But he would insist that, aside from all such considerations, water has a nature because of which the cook, sailor, fireman, or chemist can use the substance to satisfy his special interests. It is this nature which is to be identified with the idea and which Plato’s pupil, Aristotle, was to call its essence.


Along with the theory of the two realms runs a dichotomy in the realm of knowledge, one part of which is directed toward the particulars or sensible things, the other toward the ideas. As in Heraclitus, the Eleatics, and probably the Pythagoreans, one type of cognition leads us into reality, the other into the world of appearance. The former alone is called knowledge by Plato, the latter being called opinion. This terminology indicates that he does not entirely reject sensory contact with the world, but that while it may accidentally give us the truth, it cannot be relied upon to do so. It gives us fleeting impressions of things as they change. These fleeting impressions may be called true, if one wishes, but they are momentarily true and true only of whatever passes by at the moment of perception. If beyond the flux there is a stable order, then clearly, though we might get a glimpse of it by chance, we could not by the very nature of things get more than a glimpse. How then are we to reach the truth?

Whatever else the truth is, it must be self-identical through time. Its object therefore must be also self-identical and the only self-identical objects are the ideas. For since the ideas cannot change, knowledge about them when once acquired will be unchanging too. The instrument which we possess for acquiring such knowledge is called the reason. If now we select a group of things, all of which belong to a single class, and wish to know them, not simply to look at them, we must purge them of every characteristic which is peculiar to them as individuals, for such characteristics vary from observer to observer and are impermanent. Such traits as size, sensory qualities, pleasantness and unpleasantness, will have to go. Two wheels may differ in size; one may be white and the other black; one may seem ugly to one man and beautiful to another. Yet they are both an approximation to something, circularity, whose real nature is not apprehended by the eye but discovered after all particularity is removed from them. For the visual circularity of the two wheels is not the circularity which the geometer is talking about. If it were, then circles too small to be seen and the circular orbits of the planets, which cannot be seen at all, would not be circular. The visual circle may suggest the geometric notion of a real circle, but the real circle is defined in terms which cannot be seen. If we define a circle as a class of points on a plane all of which are equidistant from another point called the center, we have not made use of a single concept which applies to observation. For no one can see a point or a plane. One cannot attain to a knowledge of circularity by abstraction alone, for how can one abstract from something a trait which the thing does not possess? One could never, for instance, get the idea of man as a rational animal by abstracting the common properties of a group of imbeciles or the idea of an apple by abstracting the common properties of bananas.

Moreover, there is another objection to what has been called the empirical method. It is granted that one cannot imagine anything which one has not experienced either as a whole or in part. But when one collects a large group of things, one finds that they have several traits in common which no one thinks of adding to their real nature. Dogs, for instance, are of all sizes and shapes and of a great variety of colors. But no one would define caninity on the basis of the size, shape, or color of the animal. The differences in size which run from that of a Chihuahua to that of a Briard or Saint Bernard, the differences in shape between a spaniel and a greyhound, the difference in color between a German shepherd and a Labrador retriever, do not prevent us from lumping them all together as dogs. A German shepherd resembles a wolf much more than he resembles a cocker spaniel, but in spite of their phylogenetic relationship, we do not classify the wolf with the German shepherd and we do classify the German shepherd with the cocker. How do we know what traits are essential and what are not? Or, to put the question in another way, how do we know what specimens to collect when we are trying to assemble a group of objects with a view to defining their real nature?

This question is taken up in Meno. The concept of the idea in this dialogue is perhaps clearer than it is in any other.5 It is identified with the real nature of a thing (72b, oὐσία), with the pattern (72c, e, ). Ostensibly this dialogue is trying to answer the question of whether goodness can be taught, but it soon becomes evident that this question is involved with the larger one of how anything can be learned. We know that we do learn things and that the things which we learn may be eternal truths. Though the dialogue is too well known to require exposition here, it should be pointed out that first of all Socrates insists that one cannot say whether virtue can be taught or not until one knows what virtue is (70a–71d), that when it is pointed out that there are several kinds of virtue, Socrates insists again that all kinds must have an essential unity (71e–73c), and that the example of a good definition which Socrates gives (75b–76c) is one which overlooks the particular characteristics of the thing to be defined, for the general. We can have knowledge of the general. We do have it. How do we get it? The answer is that we have it “in our minds” before we begin our investigation and that all the investigation does is to clarify what we already know in an obscure fashion.

This is made clear in the famous myth of recollection. We come into the world, according to the myth, with a stock of ideas, to be made famous in the seventeenth century as innate ideas. These ideas are not consciously held by babies, but on certain occasions they are brought into the light of consciousness and then we know them clearly. The story of the slave boy who is made to solve a geometric theorem by the adroit questioning of Socrates illustrates this process of learning. And apparently what it proves is that, given a teacher who already knows the answers, anyone can be made conscious of the knowledge which he already possesses in obscure form. I see no reason to believe that Plato was saying that any ignoramus would be a successful teacher, though he does believe in self-education. All the talk about the Socratic method, about Socrates’ being only a midwife who brings to birth embryonic ideas, comes down to the simple fact that he possesses the correct method for clarifying obscure ideas. The theory of innate ideas has been ridiculed and there is plenty in it which is indeed ridiculous. But the fact remains that the appropriate rubrics under which the objects of knowledge are to be classified, the appropriate questions which we are to answer when we begin our investigations, are logically prior to the investigations themselves. Returning to our example of the dog, should someone ask, “What is a dog?” it is clear that one could answer by pointing to a dog. And that is indeed the technique utilized by the young men whom Socrates is presented to us as instructing. But the underlying question is, “What is this animal an example of?” And how is one to know that one has reached a satisfactory conclusion unless one already has accepted certain criteria of satisfaction?

The criteria which Plato uses are all involved in logical consistency. Now consistency can obtain only between assertions, sentences, not between sensory qualities or things or events. That is, there is no inconsistency between red and green, good and evil, war and peace. Inconsistency, as everyone knows and should remember, arises when the same predicate is both asserted and denied of the same subject in the same sentence. If then there can be consistency and inconsistency in the realm of Plato’s ideas, the ideas must be declarative in nature; they must assert something. Hence the attribution of the fallacy of “the third man” is unjust. This fallacy, elaborated by Aristotle,6 is the argument that between the individual human being and the idea of humanity there must be a third something which is common to the two of them. Once this is asserted, it is clear that one is caught in an infinite regress, since between the third man and both the individual and the idea of humanity there must be a fourth which they all share. Plato is partly responsible for this, since he spoke of the particulars as “participating” in the ideas. But if an idea is simply something which is true of all the members of a class, and if it is something which may be true or false, the idea must say something; it cannot be like a sensory quality pervasive of a number of individuals, like the color red. The Pythagorean theorem is true of all right-angled triangles, but there need be nothing common to the theorem and the triangles which resides between them.

The examination which Socrates gives his pupils then is an investigation into the consistency of their beliefs. It is concerned exclusively with that, not with the origin of their beliefs, nor with their feelings about them, their utility socially or economically, nor with any other feature which might ally them to history. But since consistency is not and cannot be found in a collection of individual things and events, there is no need to posit a community of qualities between the subject matter of a proposition and the proposition itself. The proposition that a given apple is red is not itself red, and the idea of redness is not red either. The fallacy of the third man would be serious if the idea of humanity, however it may be symbolized, had to possess the same properties as the men who “participate” in it. But to insist on that would be like insisting that the stars in the United States flag which stand for the states should have the same shape as the states for which they stand or that the states be star-shaped.

The ideas have been subjected to four interpretations. They have been thought of as Eleatic norms, as general principles, as scientific laws, and as common properties. They are indeed used by Plato in all four senses. The Idea of Justice in the Republic is a norm, a standard of what the just man or the just state should be. But it is to be observed that Plato gives us here a concrete illustration of his idea of justice, as in Lysis he gives us a concrete idea of friendship, or in Charmides a concrete picture of “temperance.” Almost every dialogue in fact is an example of something, even when, as in Parmenides or the Sophist, he is illustrating a method of thinking. This would seem to give one some reason for believing that Plato is not merely interested in abstractions, but wishes also to show his readers the abstraction embedded in concrete form or, using words technically, the idea incorporated. But the ideas are also general principles, since what they say is applicable to a group of things, not merely to one thing. And the dialectical method used to reach them is applicable only to groups. Plato undoubtedly thinks that knowledge of universal applicability is more worth while than the apprehension of individuals. The ideas in the third place may be interpreted as scientific laws, which was Natorp’s interpretation, insofar as such laws are regulative, general, and eternal. Inductive science, in the sense of drawing out of a group of supposedly similar things that which they have in common, rests upon certain presuppositions which are methodological, such as Mill’s canons or, to take another pair of examples, the principle of the uniformity of nature or the indispensability of mechanical models. Had Plato been criticizing such methods of investigation, he would have directed his criticisms toward the presuppositions, for in many of the dialogues he is more interested in the bases of belief than in its assertions.

Of one thing we can be sure: that Plato assumed the ideas, whatever else they might be, to be internally consistent. If someone holds to an idea which is self-contradictory or which contradicts another idea known to be true, then that idea has to be abandoned. But how does one discover an idea to be self-contradictory? We can waive the matter of round squares or senile adolescents which are contradictions in terms. It is more profitable to confine ourselves to a concrete example from Plato himself. In the Sophist we find the role of the teacher taken by a stranger from Elea. He is a disciple of Parmenides and presumably the method of his interrogation is one which he has learned from that philosopher. After a bit of satirical play, the Stranger begins to discuss the meaning of Being and Nonbeing. The Sophists’ habit of confusing two of the meanings of “to be”—“to exist” and “to have the attribute of”—leads to logical tangles from which the beginner in philosophy finds it almost impossible to extricate himself, so that he will assert both that Being is not and that Non-being is. The Stranger leads his young pupil to assert that truth is the assertion that what is exists and that falsehood is the assertion that what is not exists. “But Parmenides the Great, my boy, while we were still children and right up to the end of his life protested against this, saying on every occasion both in prose and in verse, ‘Never whatsoever maintain,’ he said, ‘that Nonbeing is, but do thou direct thy thoughts away from such a path’” (Sophist 237a). But even to say that Nonbeing is not is to say something about it and, if we are going to say something about anything, we ought to know what it is that we are talking about. But obviously we cannot know anything about that which is the polar antithesis of every subject matter. For if we say, “Nonbeing is unthinkable” (241a), we seem to be saying that it is, “seem” in the eyes of the Sophist, for we are playing on the ambiguity of the word “is.” And if we say it is anything else, we commit the same apparent error. But then it turns out that if we say anything about Being, we shall get caught in a similar trap, for the word “Being” can mean nothing unless it is contrasted with “Nonbeing.”

It becomes evident, as one reads through such a dialogue, that Plato’s technique consists first in distinguishing between the meanings of single words by assuming that for each word there must correspond a single thing if the word is the subject of a sentence. He also assumes quite properly that predicates are of such a nature that some are what the Scholastics were to call compossible and some noncompossible. One can therefore fall into self-contradiction by attributing two or more noncompossible predicates to a single subject. This is the case of round squares, four-sided triangles, liquid solids, hereditary sterility, and the like. When one says that a given square is round, one does not employ a negative, and yet one is involved in self-contradiction. How does one know what attributes are compossible and what not? Plato, as far as I know, seemed to believe that the human mind would know such things if left to itself in childhood. He does not say this in so many words, though in Meno this is suggested, and the praise of Socrates as an intellectual midwife would seem to mean that the human mind contains as its innate possession the power of distinguishing between viable and unviable ideas, a power which can be brought to fruition by proper teaching. Again, his criticism of the pre-Socratics (Sophist 242d–43b) seems to appeal to common sense. Yet he does not emphasize this belief in the Sophist; it operates as a guide to the argument. For the Stranger constantly asks his pupil whether he sees the difference between two ideas, or their identity, and never asks him to defend his vision. Thus, when the teacher identifies Nonbeing with Difference, for it turns out that to say that something is not red or round is simply to say that it is other than red or round, Theaetetus, the pupil, sees this; and when the question of compossibles comes up, he also sees that some ideas can be attributed to the same subject and others not. He assents to certain propositions not because he has been shown that they are logically consistent with other propositions to which he has already assented, but because they appear to be self-evident.

It is of course inevitable that certain propositions will have to be accepted as the case, that they will have to act as premises to any argument, and that such premises must be acquired or brought to the light of consciousness through what can only be called by some name such as insight, intuition, common sense, or faith. Aristotle overtly stated that the fundamental notions employed in thinking were grasped by the intellect spontaneously; Plato feels the situation to be about the same. “About the same,” since the myth of recollection makes a difference. But what is it after all that we know through intuition? It is the categories: sameness and difference, motion (change) and rest, and being. But it is also agreed that these are universally applicable and that a given subject may be self-identical and different from other things, change in some respects and not in others, and in all respects be. This is what Plato (Sophist 251d) calls the mingling and participation of attributes; not that all attributes mingle and participate in one another, but that some do and some do not. It is this which destroys any attempt at a monistic metaphysics in his opinion, whether it be a metaphysics of the flux (252a), of those who say that all is one and immutable, or finally of those who insist that all things fall into separate and unrelated classes. It then becomes the task of philosophy, accomplished through dialectic, to discover which classes intermingle and which do not. Dialectic makes distinctions according to kinds, and the dialectician perceives clearly one idea permeating many, each one lying off by itself, and many others gathered together into one, and finally many which are entirely separated from all others (253e). This mingling of ideas is illustrated by the difference between Being and Motion and Rest: Being mingling with all ideas, Motion and Rest only with some.

Plato’s method is clarified in Theaetetus, for that dialogue has as its object the definition of “knowledge.” The young Theaetetus himself, who resembles Socrates physically (143e) and has an aptitude for philosophy (144a, b), is made to propose several definitions: that knowledge is perception (151e), that it is true opinion (187c), that it is true opinion corrected by, or perhaps accompanied by, reason (201d). The first of these definitions is attributed to Protagoras by Plato and is demolished by showing that every perceiver would be right in everything he might say, since no percept can be shown to be common to two people. Hence each perceiver would live in a world of his own from which he could never escape, and if knowledge is interpersonal, then perception cannot be knowledge. At the risk of being tedious, I should now like to examine for a moment the Socratic method in more detail.

First, Socrates proposes a clarification of the doctrine (152a): “Whatever each thing appears to be to me, such it is to me, and what it appears to be to you, such it is to you. And you and I are man.” It should be observed that the text says “man” and not “men,” so that Protagoras, the epistemological individualist, is made to ascribe a common quality or universal to the various perceivers. And Theaetetus grants that this is a correct interpretation of the slogan, Man is the measure of all things. If this is deliberate on the part of Plato, and we have no reason to believe that it is not deliberate, then the founder of the theory has already committed himself to an idea which by his own theory could not be justified. For if each individual has his own body of knowledge which cannot be shared by others, then our common nouns or universals would have no ground for existing. In the second place Theaetetus is induced to grant that, whatever else knowledge means, it means something which is true. But this leads to the conclusion that if an object seems cold to one man and warm to another, it is truly both cold and warm. Here we run into the difficulty of relative statements, a difficulty of which the English Platonists of the seventeenth century were to make much capital. There is of course no inconsistency in saying that something is both hot and cold so long as the two statements are completed by adding the perceiver or the respects to whom and in which they appear both hot and cold. But Plato wants a situation in which something can be asserted which will be true for all perceivers. He then proceeds to make Socrates expound the metaphysics which lies behind this epistemology, the metaphysics of the flux, whose proponents are Heraclitus and Empedocles as well as Protagoras. According to his presentation of this doctrine, the things which we perceive come and go, as in Hume, and not only do no two men have the same perceptions, but no single man ever has the same perception twice, since he is never the same from moment to moment (154a). Here Plato is supporting his dialetical argument with historical considerations, by which I mean simply that it would not follow that because all things are as they seem, each man is different from every other man, that no man remains the same from moment to moment, or that whatever world may exist as a world of possible objects of knowledge is always in a state of change. The argument therefore turns out to be no simple testing of an epistemological hypothesis, but an attack on a definite school of metaphysicians. In other words, Plato is trying to demolish the doctrine of the historical Protagoras, and just as in such a dialogue as Charmides he was dealing with the incorporation of an idea as well as with the idea itself, so here he is dealing with an actual case of a doctrine and not merely with a hypothetical case. Having reached this point, the next step is in the direction of purifying the idea of its historical dress and setting it forth in logical nakedness.

The first problem is that of escaping from the relativism of an individualistic epistemology into a world of absolutes. And to show how impossible this escape has become, Socrates first (157e) points to dreams and diseases and insanity which cause sensory illusions. Are we to admit that these illusions, which of course do not seem to be illusory to those who suffer from them, are knowledge? Are we to admit (158c) that there is no legitimate distinction to be made between the dreamworld and the world of waking life? Theaetetus naturally will not admit this, though, if he had been a real Protagorean he might have asked on what grounds we make the distinction. But he is not a real Protagorean and agrees that it is desirable to find a set of absolutes. He must then first answer the question of why human perception is any more reliable than animal perception (161c). Second, why should anyone study with a teacher if each man has possession of the truth (161d)? Third, does not the examination of other people’s beliefs as well as of one’s own become absurd? It is apparently granted by the Sophist that animal perceptions are false, not because we are human and must base our knowledge pragmatically on human perceptions, but because the dog-faced baboon or the tadpole simply cannot have true knowledge. It is also assumed that teaching and argument do produce desirable results. These are assumptions which would be made by almost anyone, though probably not for Plato’s reasons, but again a convinced relativist might not grant them. Nor is Socrates represented as being satisfied with the acquiescence of his young pupil, for though it is probable, he says, that all men are not equal in wisdom or wiser than the beasts, that has not been proved. And what is wanted is proof.

Hence the analysis has to be begun all over again. Do we actually say that we know what we perceive? Do we say that we know a foreign language when we hear its sounds or see its letters? We know the sounds and the forms, says Theaetetus (163c), but we do not know their meaning. This is accepted by Socrates. But then a new difficulty presents itself. How about memory (163e)? Memory is not perception, and yet if we know a color when we see it and we are not seeing it when we remember it, then we do not know it when we remember it. But this again is absurd. If it is not absurd, then what we know, we know only at the moment when we are knowing it. The central interest of this conclusion, as far as Plato’s method is involved, is the light it throws on philosophic method in general. Is Plato simply arguing about how we should use words or is there an object of study, in this case knowledge, regardless of its name, which we can study as a zoologist would study a frog or a chemist an organic compound? If we are simply talking about the meaning of words, we are engaged in a kind of lexicography. And the correct meaning becomes the most usual meaning and at most we can discuss historical semantics. In the second case we are tangled up again in the problem of Meno and at best can say only that we have a dim idea of what we are talking about and that the purpose of philosophy is to dispel the clouds. We know in advance what criteria we shall use to determine the satisfactoriness of a philosophic doctrine: we shall want it to be self-consistent. But we are also faced with the difficulty of choosing our premises, and here Plato, for all his contempt for the Many, seems to go back to something like common sense once more. He is well aware of the fact that if knowledge is to be true it must be true for anyone who understands it. He is also aware of the fact that we must accept remembered knowledge as real knowledge and does not question the value of teaching and self-examination. But Socrates’ insistence on his midwifery must imply the belief that each of us, or, if we bear the Republic in mind, each of us who is rational, has within him the seeds of knowledge which the dialectical process can bring to germination. However metaphorical the statement of this, however the facts may be explained, the philosopher may legitimately be asked to say what he is taking for granted and what he will consider to be the proper test of philosophic truth. The distinction made by Theaetetus, for instance, between the meaning of a language and its perceived sounds, would probably be made by everyone, but the question remains of why, on what grounds, we accept it. Again, when it is flatly admitted that memory is not perception, are we to find some way, as Hobbes did, to turn percepts into memory images, or are we to deny that perception is knowledge?

Theaetetus ends with a confession that the theories of knowledge which have been proposed in it are “wind eggs and not worth rearing.” Is a dialogue of this sort to be thought of simply as a satirical drama of ideas? Plato himself gives us the answer. “If after this,” he has Socrates say to Theaetetus (210b, c), “you should undertake to give birth to other doctrines, and if you should give birth to them, you will turn out to be pregnant with better thoughts because of our present investigation. And should you be sterile, still you will be less rough with your associates and kinder, wisely not thinking that you know what you do not know. For such alone is my art able to do and nothing more, nor do I know any of the things which the others know, great and wonderful men who are now living and have lived in the past.” In short, leading a pupil to examine a doctrine which he has himself proposed, inducing him to realize its inconsistencies, may not show him what doctrines are true, but it may at least show him which are false. The progress of knowledge has been accomplished, according to K. R. Popper,7 one of Plato’s severest critics, by the refutation of regnant theories. This is precisely what the Platonic method consists in.

Moreover the presentation of philosophy in dialogue form is in itself of philosophic interest. When an author such as Berkeley or Hume writes dialogues to expound his own ideas, the dialogue form is unnecessary. Plato’s Laws would be a good example of such a dialogue. But where a man is discussing the conflict of ideas, none of which need be his own,8 then it is possible to illuminate the sources of the conflict, the assumptions upon which the two or more sides of the argument are based, and to move from history into logic. The points of view then become logical possibilities, hypotheses, so to speak, which one is testing, and it must be admitted that as far as Plato is concerned he sometimes succeeds, as in Protagoras, in doing this with sympathy. He seems to believe that ideas have a life of their own and we must all have had the impression, on our first reading of Plato, that ideas which turn out to be purely tentative and destined for refutation are those with which he is himself in agreement. There is no one formula of which I am aware which gives a satisfactory compendious account of all the dialogues, and I can see no reason to doubt the wisdom of the old editors who classified them as didactic, practical, expository, obstetric, controversial, and so on. Such divisions overlap and the labels are far from clear. Yet the recognition that Plato attempted to do various things in his writings and not just one thing is sound. It is not too difficult to realize that in a dialogue like Timaeus the author is not arguing but expounding, whereas in Charmides he leaves it up to the reader to draw his own conclusions. But two things may be said about most of the dialogues. (1) Where a definition is sought, it is a definition of an abstract idea, an essence, and it is not to be found in a recital of examples. (2) Where no conclusion is reached, a sharp distinction is made between our ability to recognize concrete examples of the idea whose definition we are seeking, and our ability to formulate the idea in words. This would seem to be essential, and one of the main problems of the Platonic epistemology would seem to be how we can have “true opinion,” how we can see, how we can recognize, for example, a moral quality and yet not be able to “know” it. From this point of view Meno would be central to his thought.

Now in Meno Plato has recourse to a myth to explain his theory. But a myth, it will be said, is not an explanation. To speak of the soul as a charioteer trying to guide two horses of opposing character, as in Phaedrus, to speak of the cosmic order as a great geometric system deduced by a divine mind, as in Timaeus, to speak of the objects of human cognition as shadows cast on the wall of a cave from eternal archetypes, as in the Republic, is obviously to set forth an as-if. It is analogous to the scientific practice of citing examples in the form of experiments; none of the experiments proposed in Newton’s Principia need be performed, for the theorems ought to suffice. We need no experiments in mathematics to prove or explain our conclusions. It would be absurd after adding two and two, then to take two books, lay them on a table, add two more books to them, and then add the total. Yet we demand precisely that sort of thing in arguments which demonstrate unfamiliar conclusions. It was known that the world was round and not flat in Alexandrine times—in fact, it was known to Aristotle—and all that Magellan’s journey could possibly do was to corroborate the theory. Logically, adding books, apples, pebbles, and other objects together would never prove an arithmetical theorem; at most it would apply only to those experiments from which it was induced. Why then do we need illustrations, even if the only available illustrations are fictions?

The situation, if we are not misreading Plato, is the congruence of experience and reason. To know something is to see it in its proper place in an orderly system. But experience comes to us in a random order and knowledge always begins with classification, though the classification may amount to nothing more than assigning common nouns to the things we encounter. But classifications may be bad, in the sense that they may be superficial though true. For instance, the classification of material things in gases, liquids, and solids is true but it does not suffice for chemistry. Hence once the traditional classifications are made, we proceed to divide the classes into smaller subclasses. But on what principle of division?9 What good would it do a chemist to discover that gases can be divided into those which are colorless and those which are colored? Only the good of not identifying a colored gas with a colorless gas. Sooner or later we come to the point where we realize that classifications are made on the basis of what we want to do with the things classified. To identify hydrogen with helium because both are colorless and then to use them indiscriminately might lead to disaster. I do not mean that we classify for practical ends, such as filling balloons indiscriminately with hydrogen or helium and then finding that the former burns, unless we include intellectual satisfaction under the rubric “practical.” We seem to have a hunger for some design or form or order or pattern—the name is immaterial—in accordance with which we can organize our experience. If the earliest recorded orders are cast in the form of creation myths, that may be because making and producing are the most familiar activities of men living in a nonindustrial society. And if teleological explanations are the usual type, that may be because men usually think that they act to attain ends. We obviously cannot do anything about the cosmic order in the restricted sense of practical action. But if we can see in it a duplicate of the orders of production and purposiveness, then we are at home in it intellectually. It is simply an enlarged picture of human life. A God to whom we may pray, who can grant our prayers, can reward and punish us, can be a father, is a God with whom we can sympathize since He acts as a perfected human being.10

The myth then will be introduced when one reaches a point of such ultimacy that nothing in experience will illuminate it. If we ask what is the beginning of all things or the end, there is clearly no experience on which we can found our answer. I do not say that such questions are inevitable—indeed to Aristotle the world was without beginning or end—but if they are raised, then one is forced into mythography. If again we make a distinction between body and soul, and assume that only material things can perish, since perishing is disintegration and not annihilation, then we can logically infer that the soul is immortal. But when we go on to ask what happens to the soul when released from the body, we once more must resort to myth, since obviously we have no empirical evidence of an afterlife. It may be retorted that such questions need not be raised and in fact some people have not raised them. But the fact remains that others have asked them and Plato was one of that number. His myths cannot be proved by empirical evidence, but they can be more or less plausible, depending on how closely they reflect rational deductions and the common events of human history. To translate them into literal language would be folly, for what they say cannot be either refuted or demonstrated. They are not true or they would not be myths. Yet they can serve as a standard for the truth of lesser ideas. They give us a concrete picture of what is neither concrete nor a picture. Our ability to do without them is measured by our ability to restrain our curiosity.

There is of course something paradoxical in Plato’s satisfying our desire for illustrations. We live in a world of concrete particular experiences, of things and events colored by emotion and sensory quality. Of this world we can have only opinion, as we have said above, since knowledge is bound to demand universals. We therefore strip them of their emotional and sensory vestments, classify them, and establish relations between them and the classes to which they belong. Having done this, we turn about and try to make them concrete again. And we do this in mythography. We are thus able to bring them back, as it were, to the level of experience without restoring them to the disorder and obscurity of experience. The knowledge which we acquire through them is in this way analogous to empirical knowledge in its concreteness and yet is not simply opinion. I have found no passage in Plato which says that myths are absolutely true or that they have a higher kind of truth than rational knowledge. He may have believed that they were a substitute for the kind of knowledge which we would have if our minds were capable of apprehending it. We know the ideas by contemplation; we see their truth. So we can see the truth of a myth because of its vividness and its concreteness. But myth in Plato is only a substitute for reason. I mention this simply because his later disciples were to make much of the allegorical wisdom supposedly contained not only in their master’s myths but also in those related, for instance, by Homer and Hesiod, for whom Plato himself had little admiration as philosophers.


Any philosopher who publishes his works does so presumably because he believes his ideas to be of some importance for the education of his fellows. In Plato’s case this belief is dominant. Living in a time which he felt to be degenerate, having seen his beloved master put to death for the crime of teaching self-knowledge, surrounded by Sophists, he expressed his contempt not only for the society in which he lived but also for the “unexamined life” with vigor and persistence. If Menexenus is genuine, it is a bitter parody of the kind of false patriotic oration which did not die out in the fourth century B.C. Put into the mouth of Aspasia, the mistress of Pericles, it extols the Athenians for all the virtues and condemns the other Greeks for pretty nearly all the vices. The first speech in Phaedrus is a parody of Lysias whose rhetorical elegance seems to have been sickening to Plato. His attacks on the Sophists are too well known to need any illustration here. But his attack on those who patronize the Sophists is even more savage. “Are you,” says Socrates to Adimantus in the Republic (vi. 492a, b), “like the crowd, going to maintain that youth is corrupted by the Sophists and that there are some Sophists who do their corrupting, if it is worth speaking of, in private, but are not they who say such things themselves the greatest Sophists and do they not carry out their educational program to perfection and form whomever they wish, both young and old and men and women? … When the people are all seated together in assemblies or courtrooms or theaters or encampments or some other common meeting place in gangs, they denounce with loud noises some of the things which are said and done and they shout the praises of others, exaggerating in both cases, and they bellow like bulls and clap their hands, upon which the rocks and the place in which they are gathered re-echo twofold the noise, both the noise of denunciation and that of praise. Well, then, in such a case, how do you think that the young man’s heart, as they say, is stirred?” No private education will withstand such an attack and youth will believe that what the public applauds and denounces is worthy of applause and denunciation. If anything turns out right in such a state of society, the man who attributes it to the will of God will be speaking the truth (vi. 493a). The private Sophists simply teach what the crowd holds dear and “they call this wisdom” (493a). But such wisdom is simply knowledge of the sounds made by a great strong beast on such and such occasions and what must be said to pacify or enrage it (493b). A wise man of this type would call those things fair and foul or good and evil or just and unjust which pleased or displeased, respectively, the great beast and would identify the necessary with the good without understanding the differences in their natures. The program of such people is to give the public what it wants, nor are they able to furnish more than a ridiculous proof that the public wants the really good (493d). This being so, it is folly to think that the crowd can learn philosophy. For that would rest on the belief that they could also believe that something is a universal, rather than a host of particulars (493d). They can see instances of beauty, of goodness, of justice, but can entertain no idea of beauty in itself, of goodness in itself, of justice in itself.

Plato’s contempt for the crowd goes hand in hand with his dispraise of life itself as it was lived in Athens. Earlier in the Republic (vi. 486a) he had asked whether a man who had been busied with great things and the contemplation of all time and all existence would think human life worth much. And Socrates himself at his trial is portrayed as thinking that death is a small price to pay if teaching the truth is to be denied him (Apology 29). There is, moreover, in every man a many-headed beast (Republic ix. 588c) which is concealed beneath his skin. Life is a battle between the two natures and the task of the man who would attain a life of virtue is to tame the beast within. But this is not easy, for the beast is swayed by “images and phantasms both by day and by night” (Timaeus 71a) and can be controlled only by force “since it has no share in reason or intelligence” (71d).11 That the soul has both rational and irrational faculties which are at war with each other is an essential part of Plato’s teachings and he sees little, if any, chance of ensuring a victory for the rational part. For not only is the realm of reason separated from that of perception, and is neither in space nor in time, but it is clear from the Republic (435e) that some men are born bestial and that only a few have the potency of reason within them. The former must be curbed by the latter and they would be so curbed in a perfect state. But Plato is under no illusions about the possibility of bringing the perfect state into existence down here on earth. Perfection is found in mathematics and logic, not in space and time, though it is necessary to know what it would be like if we are even to approach its realization down here. For life is pictured by him as either a continuous struggle to understand what ought to be, as contrasted with what is, or as a renunciation of the struggle and submission to our appetites. The sensual man as well as the irascible man are characters established at birth, and since they have no instinct to be other than what they are, life is a depressing spectacle to the philosopher.

It should be pointed out in passing that Plato’s descriptions of the various types of human nature started a tradition in psychology which is usually thought of as beginning with the Characters of Theophrastus. In discussing the various forms of state in the Republic, he says (viii. 544d) that there are as many species of man as there are of states. To the best kind of state there corresponds “the truly good and just man,” and to the inferior forms there correspond the contentious man and the lover of fame and honors, to say nothing of the tyrannical man. To take but one example, the timocratic man (viii. 549a) is self-willed and “rather uncultivated,” fond of music and of listening to talk, though not himself rhetorical, savage to his slaves, “not looking down on them as a properly educated man would do” (549a), humble before the free, always ready to listen to the rulers, loving high position and honors. He does not claim a right to rule because of his power of speaking or anything of that sort, but because of his military skill, his athletic prowess, and his ability in the hunting field. As he grows older, he becomes fonder of money, and does not see the folly of his life since he lacks reason, “united to true culture.”12 In such a manner Plato describes the types of character which correspond to the kinds of state. These men are not supposed to exhaust the kinds of human beings who may exist but are simply monarchy, aristocracy, timocracy, oligarchy, democracy, and tyranny reduced to human proportions.13 But as the tradition developed through Theophrastus, it came to be believed that every man had a permanent character, imprinted on his soul at birth, and in such an influential writer as Horace we discover that permanence of character must be observed by the poet—or in modern terms, the novelist—as a rule.14 Permanence of character in Plato, however, is not a literary rule but gives rise to political problems. For if some characters are inherently bad, some way must be devised to control them, and that way is set forth in detail in both the Republic and the Laws. For once it is assumed that the state exists for the sake of the good, all citizens must be so organized that they will co-operate in its realization. Since evil men will not do so willingly, they must be forced to.

All this, we must repeat once more, exists in the sphere of eternity. Plato nowhere says that such clear-cut divisions are to be found empirically. One must imagine what would be the case if the totality of things were arranged in logically distinct classes. Otherwise it is impossible to reason at all. Logical classes to his way of thinking are homogeneous, whereas groups which are found in experience are mixed. Logical classes, however, are timeless and whatever may be truly said of them is eternally true. The basic model of reason for Plato was geometry and each class of thing in the empirical world is thought of as if it were a geometric figure about which certain statements can be made by the reasoning process unaided by observation. He realizes fully (Republic vi. 510d–511b) that we cannot reach a clear intuition of first principles and that we have to utilize hypotheses as if they were first principles, but he also maintains that, once having laid them down, we should proceed from inference to inference “never utilizing sensory percepts” (511c). This is analogous to the use made of intellectual models in contemporary science: bodies moving in free space, the economic man, man in a state of nature, perfectly elastic bodies and ideal gases. He was quite aware of what he was doing. As he says in the Republic (x. 611e), before beginning the concluding myth, we must consider the soul as if it were “raised out of the depths of this sea in which it is now sunk, and were cleansed and scraped free of the rocks and barnacles which, because it now feeds on earth, cling to it in wild profusion of earthy and stony accretion by reason of these feastings that are accounted happy. And then one might see whether in its real nature it is manifold or single in its simplicity, or what is the truth about it and how.”15 Scraping away the rocks and barnacles leaves the nature of the soul, or of anything else which may be under examination, intact and open to the scrutiny of reason. When one has grasped his method of reasoning, one also understands why it is beside the point to criticize him for failing to reach results which could be reached only by another method.

Since he is interested in discovering the nature of the ideal, it would be useless for him to describe the world of existence, for that world he believes to fall short of the ideal. So no geometer would reason from the drawings made by compass and ruler. Hence when he wishes to point out the defects of men, he turns to history; when he wishes to indicate the road to improvement, he turns to eternity. For, as he said toward the end of his life in the Laws (803b), “Though human affairs are not worthy of great seriousness, yet we must treat them seriously.” And since he can find no worthy exemplar of the good life, he introduces a myth and gives in more or less fanciful terms a description of the earliest ancestors of the Athenians, men who lived nine thousand years previously.16 The society of that time, we are told, was divided into classes according to the functions performed by each, priests, artisans, shepherds, hunters, and farmers, and finally the soldiers. Emphasis is placed upon the separateness of each class; none mixed with the others. Thus they reproduced the distinctness of the ideas which also are cut off from one another. Thanks to their patroness, Athena, these men surpassed all others in Virtue (Timaeus 24d). This story, repeated a bit more sketchily in Critias (109b ff.), adds the detail that the military class, which lived somewhat as the Guardians lived in the Republic, had no private property, no gold or silver, and lived abstemiously, between luxury and want.

It is a society similar to this which Socrates describes in the second book of the Republic (369b ff.). The origin of all states lies in the individual’s inability to survive in isolation from his fellows. Food, housing, and clothing are the first of our needs and these demand specialization of labor. Since, moreover, most societies are far from being self-sufficient, imports will be needed (370e). And within the city itself there will have to be men to sell the producers’ goods to the consumers (371c). The men in such a city will lead a relatively simple life, but one which does not rise above that of the animals (372d). It is a healthy state but not a state which could serve as a model for one seeking the nature of justice and injustice (372e). For this distinction arises only in a fevered society in which there are luxury, private property, and the need to defend oneself from the attacks of the envious. Step by step Plato approaches the condition of what one supposes to be his conception of Athens, in which the untrained are attempting to perform tasks for which they are not suited, and in which unnecessary desires are creating unnecessary evils. To remedy this situation a class of men should be instituted to govern (374e). These men are the Guardians. They are clearly Plato’s ideal of human nature.

The Guardians are perceptive, brave, and high-spirited (375a, b) but also gentle. In short, they must be lovers of wisdom (philosophers). It requires no great knowledge of the dialogues to see how this fits in with the myth of the charioteer and his two horses in Phaedrus (246b ff.). The charioteer is the rational man; the good horse is modest, temperate, and “a lover of true doctrine” (253d); the evil horse is ugly, insolent, and proud. The good horse is obedient to rational command; the evil horse obeys only the cracking of the whip. Man thus is presented as torn between two impulses, the impulse to listen to reason, the impulse to listen only to appetite. Plato’s appraisal of human life depends upon the role which reason plays in it. And since reason is the contemplation of the ideas, all bodily instincts which might distract one from the realm of ideas to that of perception are to be discouraged. This is not so much asceticism for its own sake as self-discipline for the sake of true knowledge. And since the city is constructed on the model of the man, there must be in the city something to correspond to the reason in the man.17 And as Plato develops his model city, he introduces these analogous classes of men, the artisans who cater to the appetites and the warriors who nobly defend the city from its enemies. The former are to be chosen from men with inborn appetitiveness and the latter from those who are irascible—or spirited—by nature. Fundamentally Plato’s criticism of his own state is that men perform functions for which they are not naturally fitted and that no instrument is provided for remedying the situation. Indeed that criticism runs through many of the dialogues. In the Apology (20a) Socrates points out that whereas men hire adequately trained horsemen to school their horses, the best they can do is to employ Sophists to train their sons. In Laches (194d) Nicias is made to say to Socrates, “Frequently have I heard you say that each of us is good in those things in which he is wise, but wherein he is ignorant, therein is he evil.” Consequently the courageous man is “he who knows what is to be feared or encountered both in war and in all other things” (Laches 195a). Almost in the opening of Protagoras (311b) we find Socrates insisting that only a physician can instruct one in medicine, only a sculptor can teach sculpture, and only a Sophist sophistry, a sentiment echoed in Meno (90b). The one-to-one correspondence between ideas and knowledge is an essential axiom of Platonism.

But since knowledge is the apprehension of the ideas, it follows that education must be a technique of purging the soul of whatever obstructs, such apprehension. Dialectic is the technique in question. If everyone were able to employ dialectic profitably, then society and the life we live in it would be in no need of improvement. But Plato’s insistence on the special skill which the lover of wisdom possesses, suffices in itself to show that he did not believe in the ubiquity of rationality. And when this is added to his explicit arguments concerning the ineradicable differences in human beings, the conclusion is inevitable. Plato’s criticism of his society is to be attacked, if at all, at this point. There is little sense in attacking him for having fascistic leanings, totalitarian doctrines, communistic programs, all of which can be found in him. The vulnerable point is his theory of human nature.

That moral education—and in some respects all education is moral—consists in learning how to control our desires and our manner of satisfying them, is usually admitted. Differences of opinion arise when the methods of control are being discussed. The classical tradition maintained that one’s will could be made submissive to one’s reason and that normally one did a thing because one believed it to be good. So in art one was supposed to have an idea of what one wanted to achieve and then went ahead and tried to achieve it. The notion that one used the reason to justify one’s desires is relatively modern. One finds the most outspoken expression of this thesis in Spinoza. “We neither strive for, wish, seek, nor desire anything because we judge that it is good,” he says (Ethics iii. 9, schol.), “but on the contrary, we judge something to be good because we strive for it, wish it, seek it, and desire it.”18 But as early as the thirteenth century Duns Scotus maintained that God did not make the world because it was good, but that it was good because God made it. Similarly Machiavelli laid it down as an axiom that the Prince created good and bad by fiat, following apparently the definition of the law in the Justinian Code, lus est id quod principi placet,19 In any event, if we recognize the difference between reason and will, and feel the conflict which arises between them in our own lives when we are impelled to do something which we “know” to be wrong, then we can understand a theory of human nature which makes the distinction fundamental and also assigns primacy to one or the other. In Plato, as in Aristotle, only the rational could be taught; the irrational could be disciplined only through practice. Thus one can teach mathematics, but not liking or disliking. The latter are drilled into one by example, by imitation, by rewards and punishments, and by other similar exercises. When a geometric theorem is demonstrated and one has understood its premises, one cannot but accept its conclusion. But when, for instance, a literary critic writes an essay in praise of a poem, one may reach the end and say to oneself, “But I still don’t like it.” Similarly one may listen to a sermon on faith or hope or charity or all three combined and agree that the arguments invoked are sound and nevertheless not find one’s faith strengthened, one’s hope increased, or one’s charity broadened. If this were not so, the race after twenty centuries of such preaching would be well-nigh perfect.

Herein are the reasons, I think, why Plato could maintain that no man could do wrong willingly. For whatever a man does is done for the sake of what he believes at the time he does it to be good. However we may define the good, if it is to be binding on groups of people, it must be something about which we may be wrong. If one cannot be mistaken about its nature, then there is no sense whatsoever in urging others to be good. For any sentence about the good would be as true as any other. But if the sentence, “This is good,” can be true or false, and if it does not simply mean, “This is what I want to do on this particular occasion, regardless of what I may have done in the past on similar occasions and of what I may do tomorrow and of what others may have done or may do in the future,” then the good must be the name of something whose nature is not determined by the likes and dislikes of any person. What it names must be similar in status to what have been called objective facts. Moreover, if it is objectively true that the earth is an oblate sphere, then two men may be right about the weight of a given bit of matter and give two different weights, for what the thing weighs will vary in accordance with the latitude and altitude at which it is weighed. But in this case we know the system of relations in which the weight is determined. If the good, again, has factual existence, then the number of people who accept certain sentences about it is no indication of the truth or falsity of those sentences. If it was wrong to execute Socrates, it was wrong regardless of how many Athenians thought it was right. In short, I am trying to say that in Plato’s opinion values were facts. Presumably he also believed that if the reason were properly trained by the dialectical method, those men who had the good fortune to be rational would understand the nature of the good. But the understanding in question would be like a kind of revelation, the revelation of innate knowledge. Plato’s low opinion of his fellow citizens rests on his observation of their reluctance to clarify their minds. If virtue is to be taught, he says in Meno (89d), there must be teachers of it, and even the most virtuous of the Athenians, Themistocles, Aristides, Pericles, and their like had sons who never attained the moral stature of their sires. Yet this cannot be attributable to their fathers’ indifference to virtue.20

It will be remembered that in the Republic (372e) the distinction between justice and injustice was said to arise only in fevered societies. So in Meno (99e) virtue is said to be neither innate nor acquired, but something “given us by the will of the gods without our knowing it.”21 We may have true opinion about virtue, but none of us can claim to know its nature in the sense of having looked upon it face to face. This is not far from Saint Augustine’s idea of grace and election.


Contempt for the Many became an integral part of classical rationalism. The Many were incapable of wisdom and hence it was folly to attempt to impart wisdom to them. It was not long before philosophers withdrew from all social contacts and obligations and the legend developed that both Plato and Aristotle had secret doctrines which they would not reveal in public. The evidence on which this legend rests is so shaky that there is no reason to discuss it.22 But clearly, if people are divided into those who are capable of learning and those who are incapable, then there is no point in trying to teach the unteachable. If virtue cannot be taught, then schools of virtue are absurd and men who pretend to teach it are charlatans.

That virtue cannot be taught is the thesis of Meno, but the argument there rests upon the empirical observation that so far no teachers of virtue have appeared. It is not that virtue is not one of the ideas and that its nature cannot be apprehended. The Greek for virtue means also excellence and the virtue of a human being is the excellence of his humanity. But an excellent human being is one who has knowledge and, when that is grasped, the identity of virtue and knowledge is perhaps better understood. It is safe to say that for Plato the end of man is the realization of his essential nature, necessitated by the accidents of our terrestrial life.

One is always on slippery ground when one is interpreting one of the Platonic myths, but I see no reason to doubt that Plato believed in the ultimate dualism of soul and body. Though some of his myths are playful, I find it hard to believe that either that of Phaedrus or that of Phaedo is anything but serious. In the latter we find that the lover of wisdom cares nothing for the body, either for its adornment or its pleasures, but turns himself toward the soul (Phaedo 64e). For the body is a hindrance to knowledge and not even the senses of sight and hearing, on which we most rely, have any truth in them (65b). The soul, however, when it liberates itself from the sense organs and bodily pleasures, may attain real knowledge. It will then be capable of knowing absolute or ideal justice, beauty, and goodness (65d). None of these things can be seen with the eyes. And so with all essences (ousiai). Since this sort of knowledge cannot be acquired while we are associated with the body, it can be gained only when we are free of it, that is, after death. During life we may approach knowledge by avoiding any commitments to the body except those which are absolutely necessary. The asceticism preached here is purification undergone for the sake of apprehending the truth. This wisdom is true virtue, the excellence of a soul realizing its own nature.

From this it follows that the moral life is the life of self-criticism. And that is why the reason must assume authority over both the appetites and the irascible faculty. But unless the rules of criticism and the data upon which those rules operate are fixed, the reason will be as fluctuating as our sensory perceptions. By the simple procedure of arguing that if an aggregate has a common name, it must have a common property, Plato was able to conceive of a set of stable beings, obviously his Ideas, which could serve as the objects of rational insight, and the rules do not seem to have troubled him. This may have been because they seemed self-evident. He uses them in his arguments, as we all must, but their origin does not appear to have been discussed by him.23 The reason then is not simply the capacity of a man for drawing inferences from premises; it is also his power of seeing somehow or other the ideas. Just how he sees them, Plato does not tell us. It was a problem which he bequeathed to his successors.

To use one of his examples (Phaedo 74a), we have an idea of “equality.” When we say that two perceptible things are equal, where do we see the equality? If we have two pieces of wood and say that they are equal, we know that in some respects they are not equal and that in others they are. They may be equal in weight and not in length. Therefore it cannot be the mere perception of the two pieces of wood which gives us the idea of equality, since we perceive the length as well as the weight.24 We must consequently bring to the perceptual experience an idea by means of which we test the objects before us. In other words, we first have an idea of equality and for some reason or other we wish to discover in what respect the two pieces of wood are equal. We do not simply glare at them and wait for an idea of equality—or some other idea—to pop out of them and enter our minds. Since such an idea cannot have come from the perceptual world, it must have been projected into the perceptual world by us. And furthermore, we must have had this and other ideas in our minds when we were born (Phaedo 75c). But since the soul is immortal, incapable of being either created or destroyed, these “absolute ideas” are its everlasting possession. Such knowledge grows dim at birth and is recollected on certain perceptual occasions. The pre-existence of the soul is necessitated by the possibility of true knowledge. But how about its postexistence?

Plato introduces at this point (78b) an axiom which continued to be accepted by later thinkers, the axiom that only compounds could be destroyed. A simple thing, having by definition no parts, could not be broken up and dispersed. But first (78d), equality in itself, beauty in itself, being in itself, can suffer no change, since they are all of a piece and have no parts. Material beings, men or horses or clothes, which are characterized as equal and beautiful and existent, do change and are never the same from moment to moment. This sort of thing is apprehended by the senses, the former by the reason. This returns us to the two worlds, the world of the perceptible and that of the imperceptible. The body obviously belongs to the former world, the soul to the latter. The soul makes use of the body in its investigations and staggers about like a drunkard in the realm of change; when, however, it “investigates things by itself,” it withdraws into the realm of the pure and the eternal and the immortal and remains there in communion with the ideas. “And this condition of the soul is called wisdom (phronesis)” (79d). The soul then in its own nature is most like “the divine and immortal and intelligible and uniform and indissoluble and never changing” (80b), whereas the body has just the opposite traits. The body therefore can be disintegrated at death; but the soul returns to the realm of ideas. Unfortunately, as Plato develops this thought, he admits that some souls have been contaminated by their bodily associations, though how a being described as always the same could be contaminated or otherwise modified is left obscure. But when Plato indulges in mythography, he forgets the niceties of logic, and here he makes Socrates expound a doctrine of reincarnation in accordance with which the future history of souls depends on their terrestrial behavior, each being reincarnated in bodies suitable to their morals, the bodies of asses, wolves, hawks, and the like. In order to escape such a fate, the wise man will follow the teachings of philosophy (82d ff.) and thus will free his soul from the chains of bodily enslavement. The strongest of such chains are pleasure and pain (83c). They can be shaken off by leading an orderly and courageous life, withstanding—as Socrates did—the temptations both to indulge in pleasure and to flee from pain. The proper adornments of the soul are temperance and justice and courage and freedom and truth (115a). And each of these upon examination turns out to be a form of liberation from the body.

Yet again we must repeat that Plato did not preach extreme asceticism. He represents Socrates as temperate but not abstemious. As Eryximachus says of him in the Symposium (176c), Socrates can drink or not drink and be unchanged. And indeed that dialogue, in praise of love, is a demonstration of how a pleasure which in one of its forms is debasing can become ennobling. In its primary stages, as shown in the speech of Phaedrus, the love of man for man, or of wife for husband, it incites one to deeds of bravery and self-denial, exemplified by the action of Achilles in avenging the death of Patroclus or by that of Alcestis in offering to die for her husband. The presence of the beloved deters one from doing shameful deeds, for one wants to appear noble in his presence. This is then taken up by Pausanias, who distinguishes between the Heavenly and the Vulgar Aphrodites. All the gods must be praised of course, but the Vulgar Aphrodite is indiscriminate in her workings, leading men to love for the enjoyment of the body rather than for that of the soul and caring nothing for the good or evil of what is accomplished. The Heavenly Aphrodite inspires the love of boys only, not that of women, for boys are “more robust by nature and have more intelligence” (181c). Nor is such love awakened until a boy begins to show signs of intelligence. Its purpose is virtuous. It is, moreover, says Eryximachus (186), a universal force, expressed not only among human beings but in all things that exist. This speaker makes love the force which combines things into greater wholes and reduces antagonisms. It appears in music and medicine, producing harmony and health. But unfortunately it is often replaced by the vulgar goddess. The two Aphrodites seem to turn into the Love and Strife of Empedocles, the orderly one bringing fertility and health into being, the disorderly destruction and evil.

But the love in which Plato is more deeply interested is the attraction which the ideas exercise on the soul. The various speakers in the Symposium, as is well known, deliver themselves of a set of customary praises of love, but when Socrates takes the floor, it is to speak of that mysterious power which is exercised over the human mind by the desire for knowledge. Socrates is careful to point out that Love is neither a god nor a mortal, but rather one of those intermediate beings, a daimon. It will be recalled that Socrates himself at the opening of the Symposium is being stopped short in meditation on his way to the banquet (174d), as he was often prevented from taking action when the action might be evil. This force too was daimonic. In fact, one might compare the daimones to the guardian angels of Christian mythology. In any event Love is a daimon. He is described as one might describe Socrates himself—poor, barefooted, sleeping on the ground, “always a companion of want” (203d), and at the same time laying schemes to entrap the beautiful and the good, brave, thirsty for truth, philosophizing, standing halfway between wisdom and ignorance. What Love desires is what he lacks, the beautiful. There are, to be sure, other forms of love (205b), and they have been described in the speeches which preceded that of Socrates, but in general Love is love of the good. And also it is love of immortality, since one wants to preserve the good forever. Lovers may attempt to achieve immortality through their children (208e) or they may do the same through producing prudence and virtue in general and producing them in communion with young and beautiful souls fit to conceive them. “Everyone would prefer to beget such children rather than human children” (209c).

There follows the famous program by which a man of philosophic mind may perfect himself. First, he will associate with people beautiful in body, but will soon learn that bodily beauty wherever found is one and the same and will become a lover of all bodily beauty and dissociate himself from the love of one only (210b). He will next rise to seeing that beauty of soul is more to be honored than beauty of body and through his love will strive to make those who possess it better. This will lead him on to the beauty of the ideal, as shown in laws and wisdom. And “he who has been educated up to this point in love, as he sees step by step the things that are truly beautiful, when he approaches the end of his erotic adventure will see something wonderfully beautiful in its nature, that very thing … for the sake of which all previous toil was undertaken, something which first is everlasting, neither coming into being nor perishing, neither increasing nor decreasing; second, it is not beautiful in part and ugly in part, nor beautiful now and ugly then, nor beautiful in one respect and ugly in another, nor here beautiful and there ugly, so as to be beautiful to this man and ugly to that” (210e–211a). It is not visible or corporeal; it is not a particular, but a universal and “exists by itself, always being unique, and all beautiful things partake of it in such a way that, though they are all coming into being and passing away, it in no respect is either great or less, nor is it ever acted upon by anything” (211b). The contemplation of “absolute beauty” is purely intellectual and presumably imageless. And one who has achieved this vision has become immortal (212a). One can easily see from the speech of Alcibiades which follows that Plato had Socrates in mind as the exemplar of the true lover.

The emphasis upon temperance and courage is perhaps derivative from traditional Greek ethics, but at the same time it is clear that both keep the body in subjection to the soul. One is urged to be temperate in one’s courage and courageous in one’s temperance: Socrates can drink as deeply as the next man and never get drunk (214a); he can be tempted sexually and not yield (219c); in war he has withstood all hardships (220a); he has no love of money (219e). And when Alcibiades comes close to the end of his speech, he gives us a picture of Socrates standing in a brown study in the cold all night long, like the man who is enjoying the vision of absolute beauty. In short, the summum bonum in Plato’s opinion is knowledge of the truth and the restraints upon action are simply instruments for its attainment.

In the Republic (iv. 427e ff.) is a list of the virtues which make a state just. Heading them is wisdom, possessed only by a few (428e), the rulers, who, it will be recalled, are specially marked by the gift of reason. Then comes bravery, the outstanding virtue of the soldiers, which is the preservation of true and lawful belief about what is to be feared and what not (430b). The third is temperance which, unlike wisdom and courage, does not belong pre-eminently to any single class, but is found in the association between governors and governed, for temperance is self-control, the control of the lower self by the higher, and the artisans obviously are the lower self of the state (432a). When there is temperance in the state and the governors are wise and the military brave, then each class will be doing that for which it is best fitted by nature, and justice, the fourth virtue, will prevail. This scheme of virtues in the state is paralleled in the individual (435b). Corresponding to the three classes of the state are reason, irascibility, and appetite in the individual (436a). But since the appetites are not self-disciplining (439b), there must be something in the soul to control them and this something is obviously the reason. The irascible faculty too, though closer to the reason, needs control. The just individual, like the just state, is governed by the reason. He is the man in whom each faculty performs its proper function and never interferes with its fellow faculties. And the virtue or excellence of each faculty will be named as it was named in the state—wisdom, bravery, temperance.

Plato’s emphasis on the rightness of each man’s doing what he is naturally fitted to do is in harmony with his theory of ideas. The idea as archetype and the idea as class character unite in the idea as a standard. That the class character is a standard which the members of the class ought to exemplify is an assumption which became an integral part of the rationalistic tradition. The reasonableness of the assumption may be questioned, as may that of all assumptions, but nevertheless its influence has been such that one is forced to ask why it has been so persuasive. It is clear that once a concept has been defined, any candidate for subsumption under it must conform to its strictures. If we have decided to use the word “triangle” to name plane figures bounded by three straight lines, then no figure with more than three sides can claim to be a triangle. Similarly if we define justice as the harmony of the three psychic faculties, then clearly a disharmonious soul cannot be just. But so far nothing has been said to make one think that it is better to be a triangle than to fail to be one, or to be just than to be unjust. The claim to belong to a given class is made by the classifier, not by that which is being classified. But Plato, and indeed most philosophers, takes it for granted that every individual being does make a claim to belong to some class or other, and presumably, if that is so, then it ought to be as perfect an example of its class as possible. It ought to indeed if it wishes to make things easier for a classifying mind, but there is no reason to believe that things are constituted to that end. The history of scientific classification illustrates the conventionality of our genera. It is the human being who classifies and not the Goddess Nature. But since we inherit our common nouns and other classifiers with our mother tongue, it is normal to believe that they name groups of things which are not merely similar but identical. When we spot deviations from the statistical mode, we give them a bad name, such as “abnormal” or even “unnatural.” Thus we fall in line behind Plato, whether we know it or not.

This technique has been the necessary, if not the sufficient, condition of science, for no science can move forward at all until it has discovered the similarities which are concealed by differences. The ancients grouped all material substances under the headings of the four elements and thereby were able to erect an intellectual model of a relatively simple universe to take the place of the helter-skelter diversities of sensory perception. The nineteenth century replaced this classification, which by Aristotle’s time was seen to be based on the spatial position of the elements in the cosmos when all was according to Nature, by one based on atomic weights. But in scientific cases the interests of the investigators are paramount: the scientist need not think of the happiness or unhappiness of his subject matter. In ethical—as in aesthetic—situations, we may classify human beings and the types of human behavior as we will for purposes of understanding them, but the moment we try to make them conform to our class concepts, we move out of the sphere of understanding into that of control. If one of Plato’s artisans wants to be a soldier, it is the duty of the rulers to see that he remains an artisan, and if one of his soldiers should show less irascibility than normal, then he should be either reduced to the rank of artisan or encouraged to be more high-spirited. But since Plato has assumed that types of personality are fixed, such a problem would never arise to plague him. He, like the Baconian scientist, has made a preliminary submission to Nature and he is controlling her by obeying her. If Nature has decreed that we fall into certain classes, then we must see to it that her dictates are followed. The fundamental question then becomes that of how different two things must be to become essentially different. For the point must come somewhere at which we correct our classifications and our language as well.

There was, however, in all probability something more than pure intellectual interest which induced men to identify descriptive and normative concepts and to make the class concepts standards of goodness and beauty as well as of truth. The Athenian lived in a small city-state in which descent from a common ancestor gave all citizens a mythological family tie. As one reads classical Greek literature, one sees how the distinction between Greek and Barbarian, Athenian and Spartan and Theban, in short how purity of blood was a dominating idea. To be a member of the clan was to share kinship with the other members of the clan, in the sense that there was an actual identity of substance throughout the clan. This is very different, as far as its emotional overtones go, from the feeling of a subject for his king. One is not lower in the social hierarchy than anyone else in the group and neither poverty nor riches can influence one’s status. But to belong to the clan is to live up to certain ideals, in the case of Socrates to the Laws, ideals which by their very nature cannot enforce themselves but which exact obedience through consent. Plato, as we have seen, knew that sometimes noble fathers had ignoble sons, but nevertheless he also seems to have seen that it was the duty of children to live up to the nobility of their fathers. As a matter of common sense this is understandable, but as a matter of logic it is absurd. Since Plato was a bitter critic of his fellow countrymen, he tried to put the whole argument on a different foundation, for once he could induce people to admit that certain acts were just, brave, temperate, in other words right, it was no longer a question of holding up exemplars before them as anything more than illustrations of good conduct. If the clan of Jones is essentially a military clan, and if I am a Jones, then I cannot escape my destiny by trying to be a philosopher. The military spirit is inherent in me from birth; it is partly that which makes me a Jones. With such an attitude put into one by long tradition, it is easy to see why one should believe in the inherence of common traits throughout a class of beings of any sort, and also why one should believe that these traits were also ideals which the members of the class should make every effort to attain. The whole argument is metaphorical and rests on mythical bases. But that does not make it the less persuasive. It is interesting incidentally to see how in the Indian epics, for instance, in the Ramayana, the rank of the characters, their divine descent, plays a pre-eminent role, whereas in the Iliad, though Odysseus may rail against Thersites and the Rule of the Many, no hero claims more privileges than any other because of his lineage. It is true that all the main figures are kings and princes, so that there is social equality among them. But at the same time little is made of this.

If one bases one’s idea of Plato’s ethics on the Republic and the Laws, one will see little pleasure in the life which he thinks is the best. Yet it would be false to turn him into a long-faced Puritan. For the whole pursuit of knowledge is shown to be one of the most pleasurable of activities. Plato, it is true, does not use the hedonistic test of its goodness, but he does insist that to indulge in argument, to converse with intelligent men, to discover the truth, are the most agreeable ways of spending one’s life. Indeed he seems to take the point of view of one who would maintain that any type of activity may be either pleasant or unpleasant, that pleasure and displeasure cannot be found in separation from the acts to which they pertain. The pleasures of the body are bad, but they are not bad because they are pleasant; their evil comes from their attachment to the body. The pleasures of the soul are good, but again, not because they are pleasures but because they are psychic.25 This appears very clearly in that puzzling dialogue, Protagoras (351b ff.), when Socrates points out that when people object to overindulgence in food and drink and love-making, it is because these things bring in their train diseases and want, and not because they are painful. Similarly some painful things, such as gymnastic training, military service, medical treatment, are good. In the end, it is true, the momentarily painful may lead to future pleasure and the momentarily pleasurable may lead to future pain, but when a man has to act, he is not impelled by either pleasure or pain but by his knowledge of the consequences of his act. It is knowledge which is involved here and not emotion, and that knowledge may be that a pleasant good or a pleasant evil will ensue if one makes a given choice, or that an unpleasant good or unpleasant evil will ensue from it. But how one is to interpret the speech of Socrates is far from clear, since Plato may have had in mind merely an exhibition of his master’s skill at convincing a Sophist while he was still a young man.26

The insistence upon the goodness of the rational life is seen in Plato’s use of nature as a norm. For it is the nature of men to be rational; that is what distinguishes them from the other animals. And the fact that some men are irrational simply implies that some men are unnatural. Though Plato does not use these words, yet he argues as if he were assuming that the end of life is the realization to the full of one’s essential nature, and one’s essential nature is the idea of the natural class to which one belongs. We shall see this argument reappearing more overtly in Aristotle. That some things which exist are not natural but contrary to nature is difficult for us to accept, for we are more inclined than he was to think of our classes as statistical aggregates and not as perfectly homogeneous. Homogeneous classes ought not to contain any members which deviate from the norm, for if they did so deviate, how then would we know to what class they belonged? The problem of classification, and hence of identification, is all-important in the Platonic philosophy and we shall discuss it below. For the time being, we can simply say that to Plato classes were established by Nature and not by custom or convention, and that they were spatiotemporal expressions of the eternal ideas or archetypes. That in itself conferred goodness upon them, for it was inconceivable to Plato that Nature could be bad. The problem was to determine what was natural and what not.


That problem was tackled in his meditations on logic and epistemology. We find him saying in Phaedrus that there are two principles which we must learn if we are to think clearly. The first (265d) is “that of seeing and of bringing together into one idea things which are scattered about here and there, so that by defining each thing one may clarify that which one wishes to explain.” The second (265e) is “that of cutting things up into kinds at the joints in a natural manner, and not undertaking to break up any part like a bad carver.” These natural divisions are determined by the ideas, each of which is separate from all others.27 When then (270d) we are studying the nature of anything, we must first discover whether it is simple or complex; second, if it is simple, we must see what power it has of acting or of being acted upon; third, if it can be acted upon, we must ask by what; fourth, if it has many forms, i.e., belongs to many classes, we must number them and in each case put the same questions which we put in the case of simple things. We have then the following categories in accordance with which we may ask questions: the categories of simplicity and complexity, of action and passion, of cause or agent and perhaps of effect. To these must be added Plato’s conception of explanation, for without it logic in his opinion would be sterile.

That explanation must always be teleological appears as early as Phaedo (97d), if Phaedo really is early, in the passage where Socrates gives us something of his intellectual history. Having pointed out that he had been delighted by hearing that Anaxagoras had explained the order and the cause of all things as the action of mind, since that would mean that all things were arranged in the way which would be the best possible for them, he then expresses his disappointment at discovering that all Anaxagoras had done was to assign as causes “airs and aethers and waters and many other such absurdities” (98c). Such materialistic explanations are not explanations at all, since they would give as causes of Socrates’ sitting in jail his bones and joints and muscles, whereas the real reason is that the Athenians have condemned him and he has decided that it is right for him to sit still and await the execution of his sentence. What he wanted was some reason to believe that if, for instance, the earth was in the center of the universe or flat or round, it was better for it to be so than otherwise. The question now becomes that of what is meant by better in this context.

In Timaeus we have a myth, not of creation to be sure, but of the cosmic order, and there if anywhere one might expect to find this concept clarified. At the very opening of the myth (28a), we find that an eternal model of a work of art, if copied, will always be beautiful, whereas a temporal model will give rise only to something which is not beautiful. The universe must have been modeled after an eternal archetype, for it was made by God and (30a) God “wished that all things should be good and nothing evil as far as possible” and took all that was disordered and brought it into order. But the orderly is the rational, and “for this reason he established intelligence in the soul, and the soul in the body … so that the work he was undertaking might be most beautiful and naturally best” (30b). One of the criteria of perfection is unity (31a). But this unity is the unity of an organic being, a living creature, and hence it must have parts. But since the most perfect form of a physical object is spherical, the universe must be in the shape of a sphere. (33b).

“Best” then is synonymous with “the rational” and the rational is (1) eternal, (2) logically prior to its exemplifications, (3) autonomous, in the sense of depending on nothing else for its existence. Hence when we have hit upon the ideal nature of anything whatsoever, we have come back to logical form or idea or class concept, which will be unified and eternal and, for that reason alone, good. But at the same time such ideas, since they each have their own identity, are different from all other ideas, and to the categories of unity and eternality we must add that of difference. In Plato’s own language, the most general categories are the Same, the Other, and Being.28 And presumably all ideas are therefore self-identical (the Same), different from all others (the Other), and really exist (Being). Hence one of the main purposes of logic is to discover through the application of the categories what is eternally the same and really existent. For the things which we perceive in this life are never the same from moment to moment, differing from what they themselves have been, and, since they are in this state of flux, have no “real being.”

The inability, or perhaps the unwillingness, to conceive of anything transient as real is at the heart of the rationalistic tradition and, as we shall see, forced subsequent philosophers into further and further retreats from perceptual experience. The acceptance of time as a reality which cannot be derived from anything prior to it is modern. The archai of the pre-Socratics, the gods of the poets, the Ideas of Plato, the Laws of Socrates, were all believed to be eternal and immutable, and it was their immutability alone which sufficed to prove their ultimate reality and goodness. Time for Plato was “the moving image of Eternity” (Timaeus 37d), though he never explained how motion could be generated out of the immovable, and though one might have true opinion about temporal matters, one could not have knowledge of them. Logic then was concerned exclusively with the everlasting pattern of change, not with change itself, and the technique of logical thinking had to conform to the demands of eternal being. These demands were implicated in the three categories which we have mentioned.

For instance, as is well known, one of Plato’s devices for analyzing thoughts was that of dichotomy. Since the Same and the Other exhausted all possibilities, a thought could always be contrasted with its negation. This is the French technique of de deux choses l’une, of the reductio ad absurdum in geometry, foreshadowed in the fragments of Parmenides. We have already mentioned some of the difficulties involved in this. But this does not exhaust Plato’s logical technique. We find in his three categories the so-called Laws of Thought which were to be made explicit by Aristotle. The Law of Contradiction is entailed in the distinction between the Same and the Other, for if an idea has an enduring character, then it cannot be other-than itself. No reference is made to respects and times, since an idea is all of a piece and does not exist in time. The introduction of the specification of respect and time by Aristotle freezes the objective fact as of a certain date and thus nullifies its temporality. This, as I say, is unnecessary in the world of ideas. The Law of Identity is entailed in the permanence of the ideas: once one has grasped their nature, they are always self-identical. The Law of Excluded Middle is entailed in the only element of diversity admitted among the ideas: each being whatever it is, it cannot be anything else. There is no middle point between identity and difference. Once these three laws are granted, they can be used in the world of time to direct our thinking. For everything in the world of time is an image of items in the world of ideas and one’s logical task is to return to the ideas themselves from their incorporation. When dealing with their incorporation, one has to purify it of all chance impurities and then hope to uncover the idea itself. For we can recognize an instance of an idea without knowing the idea itself.

It should not be forgotten by those who wish to understand and not merely find fault with Plato that he was looking for real and not nominal definitions. Thrasymachus in the first book of the Republic is not only willful but downright wrong in defining justice as “the advantage of the stronger” (338c). For Plato cannot bring himself to believe that positive law is identical with what we have come to call divine law. In this he would have been on the side of Antigone and against her sister Ismene. Yet Antigone identified as divine law what after all may only have been tradition or custom. In the long run it may well be that we all follow in her footsteps when we obey our consciences but, if we were Platonists, we would still try to do what followed from given definitions. Thrasymachus could ask Socrates what difference it would make whether he was sincere in his beliefs or not (Republic i. 349a), and Socrates was able to answer that it made no difference since he was merely following the argument itself, wherever it might lead. Yet he comes to the point where he has to grant that it leads him to seeing no signs of justice whatsoever. And what can he mean by that except that most people would agree with him? So we too should probably say that it is unjust to have one law for the powerful and another for the weak, regardless of what happens from time to time. Plato would then ask whether our feelings arise from our verbal definition of justice or from our knowledge of the nature of justice. His own answer would be obvious.

There is no system of logical procedure given in the dialogues, no treatise on terms, laws of thought, definition, division, inference, and fallacies. Plato’s logic must be formulated from his practice. But the need for textbooks in logic was growing as Athens became more of a commercial and trading center, with a mixture of people with various traditions of civic rights. The activities of the Sophists alone would have been enough to make men more and more conscious of forensic tricks, and it was natural for a philosopher to believe that clarity and consistency of thought would expose them. It was perhaps naïve also to believe that when exposed they would lose their charm, for the mistakes of the fathers seem often like wisdom to their children. We sometimes wonder how the Athenians could have turned so deaf an ear to Demosthenes and forget how lightly we ourselves took the threats of Hitler. It was ironical enough that, once Athens was conquered, the tutor of the conqueror’s son produced the first disquisitions on logic.29

1 History of Philosophy (London: Oates and Washbourne, 1947), Vol. I.

2 I have already done so in “Fact and Legend in the Biography of Plato,” Philosophical Review, LVII (1948), 439 ff., without convincing anyone who did not agree with that article’s conclusions before reading it. Plato himself is made to say that none of the works attributed to him are genuine in Epistle 2. 314c, which makes the search for Platonism easy for those who believe in the authenticity of the Letters. Nor shall we discuss the problem of secret or esoteric doctrines or Plato’s lectures. For the former, see G. Boas, “Ancient Testimony to Secret Doctrines,” Philosophical Review, LXII (1953), 79 ff.; for the latter, see Harold Cherniss, The Riddle of the Early Academy (Berkeley and Los Angeles: University of California Press, 1945).

3 Cf. Memorabilia iii. 8 and iv. 6.

4 See his The Great Chain of Being (Cambridge: Harvard University Press, 1936).

5 Cf. A. E. Taylor, Plato: The Man and His Work (New York: Dial Press, 1917).

6 See Metaphysics 990b 17 and 1059b.

7 See his very interesting article, “Three Views concerning Human Knowledge,” in Contemporary British Philosophy (London: Allen and Unwin, 1956).

8 E.g., J. Loewenberg in his Dialogues from Delphi (Berkeley: University of California Press, 1949).

9 See Statesman 262.

10 Cf. Theaetetus 176c. On the use of myths in Plato, see L. Edelstein, “The Function of the Myth in Plato’s Philosophy,” Journal of the History of Ideas, X (1949), 463. For a bibliography of articles recently published on this subject, see Cherniss, Lustrum, IV (1959), 240.

11 Cf. Charmides 155 d, e, and Phaedrus 230a and 246a.

12 (Republic viii. 549b). In other words, he is not a bloodless rationalizer, but also shares in what the Muses can bestow (art, history, and astronomy), in short, a well-rounded human being, balanced, harmonious, never given to extremes. But no translation is adequate.

13 Cf. Phaedrus 271d.

14 Cf. the stock characters in the commedia dell’ arte, in the comedies of Molière, and in the repertory companies of the nineteenth century.

15 Shorey’s translation in the Loeb Classical Library.

16 “Earliest,” that is, within the period under discussion. Plato maintains (Timaeus 22c) that “there have been many and various kinds of destructions of men and there will be more, the greatest by fire and water, and lesser ones by a thousand other agencies.” This was to be repeated in the Laws (iii. 677a), “There have been many destructions of men by cataclysms and diseases and many other agencies, after which a small remnant of the race of men was left.”

17 In Phaedrus (248d, e) is a list of characters in order of decreasing goodness: (1) the philosopher or lover of beauty, (2) a legitimate king or warrior ruler, (3) the politician or businessman or financier, (4) the gymnast or someone who cares for the upkeep of the body, (5) the prophet or mystagogue, (6) the poet or other “imitative artist,” (7) the craftsman or farmer, (8) the Sophist or demagogue, (9) the tyrant. This ranking of souls clearly introduces as its principle of appraisal the amount of contemplative reason required for each kind of life, the amount of attention which is given to the body and its demands, and the legitimacy of the calling. It is in no sense of the word a purely logical ranking, but rather an expression of Plato’s admiration and contempt for actual occupations.

18 Constat itaque ex his omnibus [i.e., from what had been previously said] nihil nos conari, velle, appetere, neque cupere, quia id bonum esse judicamus, sed contra nos propterea, aliquid bonum esse, judicare, quia id conamur, volumus, appetimus, at que cupimus. Cf. iii. 39, schol.

19 The historical affiliation between this sentence and Machiavelli is obscure, but the thought is identical. Shocking though it has appeared to many political philosophers, the sentiment is not very different from that contained in the holier slogan, Vox populi, vox Dei, which Machiavelli, like Alcuin (Epistle 127) thought absurd. Both the princeps and the populus are sovereign.

20 Note that at this point in the dialogue Anytus, the accuser of Socrates, is brought in to defend the thesis that all one needs to attain virtue is the will to follow the customs of our elders. See esp. 93a.


22 But see, if it is desired, G. Boas, “Ancient Testimony to Secret Doctrines,” loc. cit.

23 This matter will be taken up below (Section V).

24 The late John Burnet, in his Platonism (Berkeley: University of California Press, 1928), pp. 41 ff., says that the word “idea” primarily means “form” or “figure,” that the theory of ideas was “a Socratic development of Pythagoreanism” which was later rejected by Plato, and that the word ought to be dropped in favor of “forms.” But “form” is no less visual in meaning than “idea” and if the word denotes something which can be true or false, the visual meaning is lost anyway.

25 On the aetiology of pleasure and pain, see Timaeus 64d and Gorgias 497.

26 The problem of interpreting this passage is discussed, as so many other exegetical problems are, by Paul Shorey in What Plato Said (Chicago: University of Chicago Press, 1933), pp. 129 ff.

27 This is true even of the most general ideas, Being, Identity, and Difference. See H. Cherniss, The Riddle of the Early Academy, p. 55. This should be recalled if one is tempted to read a hierarchy into Plato.

28 Cf. the discussion of Being and Unity in the Sophist 244–45.

29 We are far from maintaining that we have listed all of Plato’s contributions to philosophy and repeat the warning given in our preface: This is not a history of philosophy. The Timaeus alone in one sentence (29e) stated what A. O. Lovejoy has called “the Principle of Plenitude,” the consequences of which fill the pages of The Great Chain of Being and were elaborated throughout the centuries in treatise after treatise. The myth of the Soul as Charioteer in Phaedrus, the problem of Meno, the myth of recollection and of innate ideas, the doctrine of Love, the political works, all have had their echoes in philosophy, science, theology, and art. But in our attempt to develop the fortunes of rationalism, we have had to neglect everything which did not bear upon them. This is not a book about individual philosophers but about the history of selected ideas.

Previous Chapter

II. Rationalism in Athens

Next Chapter

IV. Aristotle

Additional Information

Related ISBN
MARC Record
Launched on MUSE
Open Access
Mellon/NEH / Hopkins Open Publishing: Encore Editions
Creative Commons
Creative Commons
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.