Richard Robinson, Plato's Earlier Dialectic, 2nd edition (Clarendon Press, Oxford, 1953). Reprinted in Gregory Vlastos (ed.), The Philosophy of Socrates (Anchor, 1971). Edited in hypertext by Andrew Chrucky, June 2, 2005.

Chapter 2



The outstanding method in Plato's earlier dialogues is the Socratic elenchus. 'Elenchus' in the wider sense means examining a person with regard to a statement he has made, by putting to him questions calling for further statements, in the hope that they will determine the meaning and the truth-value of his first statement. Most otten the truth-value expected is falsehood; and so 'elenchus' in the narrower sense is a form of cross-examination or refutation. In this sense it is the most striking aspect of the behaviour of Socrates in Plato's early dialogues. He is always putting to somebody some general question, usually in the field of ethics. Having received an answer (let us call it the primary answer), he asks many more questions. These secondary questions differ from the primary one in that, whereas that was a matter of real doubt and difficulty, the answers to all these seem obvious and inescapable. Socrates usually phrases them so that the natural answer is yes; and if you say anything else you are likely to seem irrational or at least queer. In other words, they are not so much requests for information as demands for an assent that cannot very well be withheld. They often seem at first irrelevant to the primary question, and sometimes they seem to fall into two disconnected groups among themselves. But at last Socrates says: 'Come now, let us add our admissions together' (Prt. 332D); and the result of doing so turns out to be the contradictory of the primary answer. Propositions to which the answerer feels he must agree have entailed the falsehood of his original assertion.

Such is the Socratic elenchus, often referred to also as exetasis or scrutiny and as basanismus or assay. It is so common in the early dialogues that we may almost say that Socrates never talks to anyone without refuting him. An exception is his conversation with Cephalus in the first book of the Republic (the first book of the Republic may be regarded as an early dialogue); but there the subject is personal experience and not abstract ethics.

The sureness of the refutation gives the impression that Socrates possesses knowledge about the subject on which he refutes others. This, however, he invariably denies. 'You treat me', he says in the Charmides (165B), 'as if I professed to know the matters I ask about, and as if I might agree with you if I wished to. But that is not so. On the contrary, I inquire into the proposition along with you because I do not know. I will tell you whether I agree or not when I have examined it.' (Cf. Ap. 23A.) That is always his attitude; and in harmony therewith he always puts the primary question as a request for information and not as if he were examining a candidate. Throughout the early dialogues, whether engaged in elenchus or not, he usually declares himself ignorant of the answers to all the general ethical questions that he raises. There are some extremely confident statements in the Apology and the Crito, and in the Euthydemus (293B) he admits knowing many small matters; but Meno 98B seems to be the only place where he actually professes to know something important.

And we must now observe some other curious disclaimers. Not merely does Socrates sometimes deny by implication that it is the answerer who is refuted ('It is the logos that I chiefly examine', he says, Prt. 333C); at other times he even denies that it is Socrates who is doing the refuting. He speaks as if the logos were what was doing the refuting, and as if the logos were a person over whom he had no control, refuting not merely the answerer and himself but even the whole company with equal impartiality and inexorability. He denies that he resembles Daedalus, who made statues move; for the logoi run away without his agency, and he would rather they remained (Euthyph. 11D). His language implies that he himself did not foresee the course the argument has taken, but was led along by it blindfolded; and that for all he knew the argument might have turned out a proof instead of a disproof of the original thesis. He even implies at times that there is no refutation at all, of anybody or by anybody or anything. There is only a company of persons engaged in determining the truth-value of a proposition, engaged in an impersonal elenchus in the wider sense.

This denial that he is conducting an elenchus is insincere, and constitutes what is known as the Socratic slyness or irony. The arguments could not be so workmanlike and purposeful, the results could not be so invariably negative, by divine inspiration or by mathematical probability. When we examine one of the arguments in detail, and see just what its logical structure is, we become convinced that from the very first of the secondary questions Socrates saw and intended the refutation of the primary answer. There is an elenchus in the narrower sense; and it is Socrates' own work. When he says of an answer 'Well, that is good enough' (Grg. 498A), he gives away the fact that, though the answerer has not admitted as much as he expected, he has admitted enough for his downfall. In reality Socrates is always doing what he does openly in Republic I 348-49, looking for a way to persuade the answerer that his thesis is false (348A4); and if the answerer refuses to grant him a premiss (348E) he keeps the conversation going somehow (348E-349B) until he has thought of another starting-point which the answerer will admit and which will serve to refute him. The statements that he is 'seeing whether the answer is true' are insincere. So are the earnest requests for instruction by which he obtains the primary answer. So are his occasional invitations to reciprocity in elenchus (e.g., Grg. 462A); he makes them only to persuade the other man to submit to questioning; and, when he is taken at his word and made the answerer, his answers soon become speeches. Insincere also is the pose of suffering from bad memory. In the Meno (71C) it is a way to entrap Meno into pontificating, so that he can be refuted. In the Protagoras (334C-D) it is a way of forcing Protagoras to answer questions; and Plato makes an imprudent admirer of Socrates point out the inaccuracy (336D). Socrates seems prepared to employ any kind of deception in order to get people into this elenchus.

Plato depicts or asserts various effects as following immediately from this ironical elenchus. One of these is the bewilderment of the answerer. Though the word 'elenchus' is scarcely used in the Meno (apparently only 75D), the thing itself is very much there; and Plato puts the following description of its effect into the mouth of the victim:

Socrates, I heard before I met you that you never do anything but puzzle yourself and others too; and now it seems to me that you are bewitching and drugging and completely spellbinding me, so that I have become saturated with puzzlement. In fact, if I may make a little joke, you are absolutely like the broad electric ray of the sea, both in appearance and otherwise. That fish benumbs anyone who comes near and touches it, and that is what you seem to have done to me now; for really I am numb in mind and mouth, and I do not know how to answer you. Yet I have discoursed on virtue thousands of times and to many people; and done it very well too, as I thought at the time. But now I cannot even say what it is. I think it is a wise decision of yours never to leave Athens; for, if you did such things in another city where you were a stranger, you might be arrested as a wizard. (Mem. 8oA-B.)

Plato notes also that this ironical elenchus often had the effect of making its victims angry with Socrates and ill disposed towards him. He makes Thrasymachus complain of what he calls 'Socrates' usual slyness' (Rp. I 337A). Socrates refuses to make any contribution himself; but when any other person makes one he pulls it to pieces. He ought to realize that questionings is easier than answering (336C). Thrasymachus believes that Socrates deliberately tries to make trouble in arguments (341A). However that may be, there is no doubt that the actual result is sometimes the conversion of a pleasant discussion into a quarrel. Even in the Laches, where the elenchus is unusually benign in tone, its first effect is to make two old friends quarrel. In the Apology Plato makes Socrates attribute his unpopularity to the elenchus (e.g.,21C-D-E, 23A).

Plato also tells us that this elenchus is very amusing to the bystanders (Sph. 230C), especially to the young and rich (Ap. 23C, cf. 33C), and that young men treat it as a game and imitate it in and out of season (Rp. VII 539B). This effect would naturally increase the anger of the victim against Socrates.


The picture which we have so far obtained of the Socratic elenchus is by no means a favourable one. This elenchus involved persistent hypocrisy; it showed negative and destructive spirit; it caused pain to its victims; it thereby made them enemies of Socrates; it thereby brought him to trial, according to his own admission in Plato's Apology; and so it brought him to his death.

The question thus arises what Plato conceived to be the justification of the elenchus. For what end was it worth while to be so destructive and insincere, and to incur so much enmity?

Plato certainly thought that it could be justified. He did not regard it as a deplorable defect in Socrates' character, to be explained by medical or psychological doctrines but not to be justified. He held that it had a sufficient reason and was a valuable procedure, to be retained in spite of some undesirable consequences. Furthermore, he certainly held that its justification was not merely the amusement provided for the bystanders.

There are three passages in the dialogues that offer something like a general discussion of the purpose of the elenchus; and we shall examine each in turn. Let us take first a passage from Socrates' conversation with Meno's servant (Men. 84). The question is how to construct a square double the area of a given square. The solution is yet to come; but the elenctic part, in which Socrates disproves the servant's false suggestions, is over, and Socrates breaks off to discuss the elenchus with Meno.

Do you notice, Meno, how far he has advanced already in his recollecting? At first he falsely thought he knew which is the line belonging to the eight-foot square, and answered confidently as if he knew, and did not feel at a loss; whereas now, though he knows no more than he did before, he does at least feel at a loss, and no longer thinks he knows.
-- You are right.
-- So now he is better off about the thing he did not know?
-- I think that, too.
-- Then did we do him any harm in puzzling him and numbing him like the electric ray?
-- I think not.
-- At least it seems that we have made him more likely to find out the truth. For now he will be glad to search for it because he knows he does not know it, whereas formerly he might easily have supposed on many occasions that he was talking sense about the double square if he said that it must have a side of double length.
-- It seems so.
-- And do you think he would ever have tried to discover the truth, or to learn what he thought he knew though he did not, if he had not fallen into puzzlement and come to believe that he did not know and desired to know?
-- I do not think so, Socrates.
-- Then he was benefited by being numbed?
-- I think so.

Of two ignorant persons, this passage implies, the one who knows that he is ignorant is better off than the one who supposes that he knows; and that is because the one has, and the other has not, a drive within him that may in time lead him to real knowledge. The elenchus changes ignorant men from the state of falsely supposing that they know to the state of recognizing that they do not know; and this is an important step along the road to knowledge, because the recognition that we do not know at once arouses the desire to know, and thus supplies the motive that was lacking before. Philosophy begins in wonder, and the assertion here made is that elenchus supplies the wonder. Though the passage contains no such word as 'curiosity', we can say without fear of 'misinterpretation by abstraction' that Plato in writing the Meno believed that curiosity is essential to the acquisition of knowledge, and that elenchus is the way to arouse curiosity. Elenchus is thus a method of teaching, of instilling intellectual knowledge in other persons. It does not, however, actuaIly increase knowledge, but only prepares the ground for it.

Another discussion of the purpose of elenchus occurs in the late Sophist (229E-230E), put into the mouth of another than Socrates, and showing how Plato regarded the elenchus at that time of his life.

Of education one way seems to be rougher, while the other part of it is smoother.
-- What are these two parts?
-- One is the time-honoured and traditional method which men used to adopt with their sons when they did something wrong, and still do adopt very often. It consists partly in anger and partly in a gentler sort of exhortation, and the best name for it as a whole is admonition.
-- Yes.
-- But some men appear to have reached the conclusion that all ignorance is involuntary, and that no one will ever learn anything if he thinks he is already a wise man in that respect, and that the admonitory form of education involves great labour and achieves little result.
-- They are right.
-- So they aim at the removal of this opinion by another means.
-- What is that?
-- They question a man on those matters where he thinks he is saying something although he is really saying nothing. And as he is confused they easily convict his opinions, by bringing them together and putting them side by side, and thus showing that they are contrary to each other at the same time in the same respect about the same things. When the man sees this he becomes angry with himself and gentle towards others. Thus he is relieved of great and overbearing opinions about himself, and this relief is the pleasantest of all to hear and the surest for the patient. For just as the physicians of the body believe that the body cannot benefit from the nourishment it receives until the internal hindrances are removed, so do those who perform this purification believe about the soul. She cannot profit from the knowledge offered to her, until the elenchus is applied and the man is refuted and brought to shame, thus purifying him from opinions that hinder learning and causing him to think he knows only what he does know and no more.
-- That is the best and most temperate state to be in.
-- For all these reasons, Theaetetus, we must say that elenchus is the greatest and most sovereign of the purifications; and the man who has not been subjected to it, even if he be the great king himself, must be regarded by us as suffering from the greatest impurities, and as uneducated and base in the respects in which the truly happy man ought to be purest and noblest.

Here the elenchus is explicitly subsumed under the general notion of education, and explicitly preferred to another form of education in words of the highest praise. Its nature is illustrated by a comparison with medical purging, which brings out the doctrine that elenchus is not itself the instilling of knowledge, but an essential preliminary thereto, consisting in the removal of an all but complete bar to knowledge naturally present in man. This, bar is the conceit. that we already know.

The third passage that contains a discussion of the purpose of the elenchus is the Apology, one of Plato's early works. According to this work, Socrates at his trial regarded his habit of elenchus as one of the main counts against him, and set out to justify or at least to explain it. He declared that it arose from the Delphic god's response to Chaerephon, which was that no one was wiser than Socrates. He felt that he had no wisdom; but he also felt that the god could not lie. After a long time of perplexity, it occurred to him to approach a man with a reputation for wisdom and study him at first hand. He found that the man thought he was wise but was not. Going then to many other men of repute, he always had the same experience. In one class, the men of skilful hands, there was some real wisdom; but this led to so much conceit of other, non-existent wisdom as more than outweighed it. Socrates concluded that he was really wiser than the wise because, whereas they knew nothing, he knew the single fact that he knew nothing.

But why did Socrates continue the elenchus after he had ascertained these facts to his satisfaction? Because, he tells us (23A-B), he felt that the god had imposed upon him the duty of demonstrating to all men that no man is wise. Later he says that the god has told him to philosophize and to scrutinize himself and others (28E); that the purpose of his elenchus and the command of the god is to shame people into putting first things first, and that the first thing is the virtue of the soul (29D-E); that he is to the Athenian people as a gadfly to a noble but sluggish horse (30E). In his speech after the determination of the penalty he calls his elenchus an examination of men's lives, for that seems to be the meaning of ελεγχον του βιου (39C); and describes his purpose as to put men to shame for living wrongly.

The Apology, like the Meno and the Sophist, regards elenchus as a way of convincing men that they are ignorant of things they thought they knew; but it places this procedure in a strongly moral and religious setting of which the other two works show little trace. It tells us that the elenchus arose out of a divine oracle, and that Socrates continued it because he felt divinely commanded to do so. It represents the ultimate aim of the elenchus not as intellectual education but as moral improvement. Its purpose is, as it is expressed at the end of the Apology, to make rnen better men, to give them more of the highest virtue of a man: and in practising it Socrates is a moral reformer.

To many persons the Socratic elenchus would seem a most unsuitable instrument for moral education. They would argue that such logic-chopping cannot be followed by most persons, does not command respect, and at best improves only the agility of the mind while leaving the character untouched. Socrates was certainly a unique reformer if he hoped to make men virtuous by logic.

Yet it is clear that Plato consciously intends to depict Socrates as consciously aiming at the moral improvement of his fellows by means of his elenchus. Looking back on the picture from the late Sophist, he contrasts the elenchus with another method of altering men which he calls 'admonition'. 'Admonition' includes the more ordinary methods of moral education, such as rebuke and persuasion and harangue and advice. And Plato says that the practitioners of the elenchus deliberately prefer it to 'admonition'.

This is an aspect of the paradoxical intellectualism of the practical philosophy of Plato, and apparently also of the historical Socrates. It hangs together with the proposition that virtue is knowledge. The method of the Platonic Socrates differs from those of all other moral reformers because of his unusually intellectual conception of what virtue is. He believes that you cannot really be virtuous unless you have a philosophical understanding of the definition of virtue. The practice of virtue is identical with the theory of it. The way to become courageous is to find out what courage is. Contrariwise, he who does not know the definition of virtue will not behave in a virtuous manner. When Socrates says early in the Apology that he demonstrated to men that they knew nothing, he means that they knew nothing about wisdom and other forms of virtue; for this is the only matter that interests him. And "because you cannot be virtuous without knowing what virtue is, there is to him nothing strange or puzzling in representing as vice in his third speech what he represented as ignorance in his first. In order to make men virtuous, you must make them know what virtue is. And in order to make them know what virtue is, you must remove their false opinion that they already know. And in order to remove this false opinion, you must subject them to elenchus. That is the way in which, according to the Platonic Socrates, the elenchus comes to be the appropriate instrument for moral education.


The Socratic elenchus is a very personal affair, in spite of Socrates' ironical declarations that it is an impersonal search for the truth. If the ulterior end of the elenchus is to be attained, it is essential that the answerer himself be convinced, and quite indifferent whether anyone else is. In the first place, he must believe his own primary statement; otherwise the refutation of that statement will not convict him of thinking he knew when he did not. In the second place, the answerer must be quite convinced of the logical validity of the argument; if he thinks that the contrary of his thesis does not really follow from the premisses adduced, he will again not be convicted of ignorance. Lastly, he must genuinely accept the premisses; that is the implication of Gorgias 471D and many other passages. The art of elenchus is to find premisses believed by the answerer and yet entailing the contrary of his thesis. Polus fails to refute Socrates because he cannot find premisses that Socrates accepts. What the ordinary man believes would entail the contrary of Socrates' thesis; but Socrates does not believe what the ordinary man believes. When the refutation is a reduction to absurdity, the conclusion must seem absurd to the answerer himself. Here again Polus fails, for he reduces Socrates' thesis to results that seem absurd to Polus and to most men, but not to Socrates. Socrates and Caesippus fail in the same way to refute Euthydemus and Dionysodorus (Euthd. 294, 298).

Plato brings out the personal nature of elenchus in the Gorgias. That dialogue, which contains the root ελεγχ- over fifty times in its eighty pages, represents Socrates as contrasting his own procedure with that of the law-courts. Whereas in law-courts you have to convince a third party, namely the judges, in the Socratic elenchus you have to convince your opponent himself. Hence the witnesses who are so effective at trials are useless here. The only true witness and authority is the answerer himself; and if he does not admit the fact it is irrelevant how many others do. The result depends not on a majority of votes, but on the single vote of the answerer (471E-472C, 474A, 475E).

Possibly this aspect of the elenchus explains why Socrates sometimes seems to start the argument with premisses that immediately decide the point, and then to hammer out the inference at most unnecessary length (e.g., Grg. 474B-479E). The whole essence of the elenchus lies in making visible to the answerer the link between certain of his actual beliefs and the contradictory of his present thesis. This link must be visible to the questioner before the process begins; and so may well be visible to the onlookers too, including ourselves.

In conscious opposition to the ideal of an argument addressed to this man personally, and really convincing him by starting from premisses that he really believes, which receives its clearest statement in the Gorgias, Aristotle set up the ideal of the listener who has the sense to recognize the expert, and accept on faith what the expert tells him are the principles of the subject. 'The learner ought to believe', he says (S.E. 2,165B3).

By addressing itself always to this person here and now, elenchus takes on particularity and accidentalness, which are defects. In this respect it is inferior to the impersonal and universal and rational march of a science axiomatized according to Aristotle's prescription. Plato might urge, however, that elenchus is the means by which the irrational and accidental individual is brought to the appreciation of universal science, brought out of his individual arbitrariness into the common world of reason.

If the actual Socrates practised elenchus, how did he come to it? Did he first decide to make men virtuous, then cast about for a means of doing so, and then hit upon elenchus? That seems unlikely. More probably he practised it at first simply because it was his nature to inquire into things more deeply than other men, and to be puzzled by difficulties that had not occurred even to the experts. Only when he had been asking questions for some time would he perceive that he knew better than other men in that he knew his ignorance and they did not know theirs. And only when he had thoroughly realized this would he give up asking questions for his old reason (for he would see it to be futile), and ask them now for the new end of reforming the answerer by showing him his ignorance. And very likely the new end would never have occurred to him but for the experience that his questions did actually have that result.


The following objection may be made to the method of elenchus: it only tells you that you are wrong, and does not also tell you why. Real conversion makes you no longer even want to hold your former thesis, because it shows you the reason why you held it and the inefficiency of that reason. But Socrates rarely does this; there are few parallels to that part of the Gorgias where, having disproved the view that Pericles and the rest were good statesmen, he goes on to show us why we thought they were. And this is why the elenchus so often misses its avowed aim, the actual convincing of the answerer (Grg. 513C), and why what seems to Socrates a conviction may be described by others thus: 'he was bound and gagged by you in the discussion' (Grg. 482E).

Plato does not talce account of this objection anywhere in his writings. Yet we can indicate with confidence the sort of answer that his writings suggest, the sort of answer that he would have given if the thought had been brought to his notice. The aim of the elenchus is not to switch a man from an opinion that happens to be false to an opinion that happens to be true. It is not satisfied by any exchange of one set of opinions for another, even if the new set is true and consistent whereas the old set was false and inconsistent. Thie aim of the elenchus is to wake men out of their dogmatic slumbers into genuine intellectual curiosity. The conviction of one's own ignorance involves and includes some dim realization of the difference between knowledge and all opinions whether false or true. In other words, the notion of the elenchus contains a germ of the Platonic conception of knowledge as absolutely distinct from opinion. The elenchus does not directly give a man any positive knowledge; but it gives him for the first time the idea of real knowledge, without which he can never have any positive knowledge even if he has all the propositions that express it. It is important to separate the realization that you are wrong from the realization why you are wrong or what the truth is, in order that the mind may dwell on the question what constitutes being wrong or right.

It may be urged that the elenchus would be more successful without the irony. The insincerity of pretending not to be conducting an elenchus must surely lessen the moral effect. It is not possible to make men good by a kind of behaviour that is not itself good. Furthermore, the irony seems to be a main cause of the anger which, as Socrates declares (Ap. 21E, &c.), often results from the elenchus; and if elenchus really makes people hate you, surely it is bad teaching and a bad form of intercourse in general. We can hardly suppose that after the victims' anger has cooled they admit their ignorance and start to reform their lives, for the Apology implies that most of them have remained angry and unconvinced to the end of their days. The beneficial shame that Alcibiades felt in the presence of Socrates (Smp. 216), the pleasantness and utility that Nicias found in being refuted (La. 188), must have come from a straightforward and unconcealed elenchus; for Socrates could not refute his intimates many times and still prevent their knowing when he was about to do so.

This objection, like the former one, is not noticed or met in the dialogues. We may conjecture, however, that Plato would have dealt with it differently at different periods of his life. While he was writing the earlier dialogues he would probably have defended the ironical form of the elenchus on the ground that it supplied a necessary shock. For it may be argued that he who announces beforehand that he is going to prove you ignorant thereby destroys his chance of doing so, because you will instantly close your mind against him. Especially is this so on matters of right and wrong and good and bad. What is required, therefore, is a drastic shock, a practical demonstration of ignorance accompanied by shame. For this purpose the victim must be drawn into a parade of knowledge, and then there must be a violent reversal of the situation, which can only be accomplished by some such mummery as Socrates practised.

At a later period of his life, however, Plato would probably have dealt with the objection by admitting it and abandoning the irony. The passage in the Sophist (translated above, pp. 84-85) makes no mention of irony, and asserts that elenchus makes the 'patient' angry only with himself, but gentle towards others. The seventh Letter requires that elenchus shall be conducted in a friendly manner (εν ευμενεσιν ελεγχοις 344B). The elenchus which Plato came to approve was a contest in which both parties openly admitted that the questioner was trying to refute and the answerer was trying not to be refuted. It was the formal and open exercise for which Aristotle wrote rules and hints in his Topics.

Three things happen to the elenchus in the middle and later dialogues. First, as we have just seen, it loses its irony. Second, it is incorporated into the larger whole of dialectic, which somewhat changes its character. Though still negative and destructive in essence, it is harnessed to the car of construction. Though still moral in its purpose, the ultimate moral end recedes a great deal, and a large scientific programme occupies the middle view. Third, while often referred to and recommended, it gradually ceases to be actually depicted in the dialogues. Refutations take less of the total space. Those that do occur are less obvious in form; it is not so easy to point to the separate premisses, to the manner in which each is obtained, and to the place where Socrates puts them together and draws the conclusion. They are less purely negative; there is often positive doctrine that is unnecessary to the proof. In the pure form of elenchus, moreover, there tends to be only one refutation to each thesis. For the refutation professes to be final and absolute, or 'iron and adamantine' as Socrates puts it (Grg. 509A); but if you add a second you seem to confess that the first was not so. In the middle dialogues, however, we do find more than one argument for the same conclusion; and this is a distinct change in character. Plato now offers a series of considerations making towards a conclusion, and it is possible to admit of any one that it is not conclusive by itself. He has given up the claim to be incontrovertible, and become in truth more persuasive. Thus elenchus changes into dialectic, the negative into the positive, pedagogy into discovery, morality into science.