true and false names in the "cratylus"

True and False Names in the "Cratylus"Author(s): Mary RichardsonSource: Phronesis, Vol. 21, No. 2 (1976), pp. 135-145Published by: BRILLStable URL: http://www.jstor.org/stable/4181985 .

Accessed: 24/08/2013 01:58

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

BRILL is collaborating with JSTOR to digitize, preserve and extend access to Phronesis.

http://www.jstor.org

This content downloaded from 131.211.208.19 on Sat, 24 Aug 2013 01:58:53 AMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=bap

http://www.jstor.org/stable/4181985?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


True and False Names in the "Cratylus" MARY RICHARDSON

Recent commentators on the Cratylus have typically passed one of the following (equally condemnatory) judgments on the passage 385 B-C, in which Plato asserts that words can be true

or false: Plato has given us a patently bad argument, Plato is confused, or Plato does not mean what he says.' I think that an examination of the text will vindicate Plato of all of these charges. In this paper I shall attempt to show that the charges are unfounded by arguing that only if one supposes that Plato means by 'logos' in the passage in question something like what contemporary philosophers mean by 'statement' can orne advance the arguments used by the aforementioned commentators in support of their verdicts, and, further, that it is most unlikely that the text can support such an interpretation. The major burden of this paper will be to try to give a reasonable analysis of the passage in question without imposing on it the contemporary philo- sophical use of the word 'statement'.

Let us begin by examining the context in which the passage appears. Cratylus and Hermogenes have been arguing about the correctness of names.2 Both hold that there is such a thing as the correctness of names, but Cratylus has been defending the thesis that things have correct names by nature, while Hermogenes holds that the correctness of names is a matter of convention. One gets the impression, from Hermogenes' complaint that he has been asking Cratylus to explain his position and can extract from him no explanation or even discussion of the matter, that Cratylus and Hermogenes have been merely exchanging the slogans, 'Correct naming is a matter of convention', and 'Correct naming is a matter of nature', without attempting to spell out the positions or trying to construct arguments for or against

I cf. R. Robinson, 'A Criticism of Plato's Cratylus,' Phil. Rev., 65 (1956), pp. 324-341, R. Weingartner, 'Making Sense of the Cratylus,' Phronesis, 15 (1970), pp. 5-25, C. Kahn, 'Language and Ontology in the Cratylus,' Exegesis and Argument, Studies in Greek Philosophy presented to Gregory Vlastos (Phronesis, Supp. Vol. I) Assen 1973, pp. 152-176. 2 I shall translate 'onoma' as 'name', although Plato uses that term to cover what we call nouns, as well as proper names.

135



them. As the dialogue opens, Socrates is being brought into the discussion. Hermogenes briefly states Cratylus' and his own positions as follows: Cratylus' position is that everything has a correct name, which comes by nature and which is the same for all men; that is not a name which some people agree to call a thing. (383 A) Hermogenes thinks that no name belongs to any particular thing by nature, but only by law and custom of its users, so any name which someone im- poses on a thing is the coreect one, and that names can be changed at will. (384 C-D) Socrates agrees to help sort out the matter. It is com- monly thought that Socrates immediately launches an attack on the convention thesis held by Hermogenes, and the passage ve are considering is often called 'the first argument against the conventionalist position'.3 But what exactly is the conventionalist position which Hermogenes is holding? I think Socrates undertakes to find that out before he launches an attack against it. One of the great accomplishments of the dialogue is that it shows the nest of problems and distinctions which are overlooked by those who depend upon slogans rather than upon serious inquiry. Given Hermogenes' statement of his position, it is not clear which individuals or groups he thinks can determine word usage, and it is possible but not necessary for him to hold that there is no such thing as falsehood, and that things have no determinate nature of their own. Socrates first asks Hermogenes to be more specific about who may determine word usage, to which Hermogenes replies that a particular individual or state may do that. If this claim were true, it would make discussion difficult, if not impossible (and for that reason alone Socrates would be at pains to refute it). But this claim would not make it impossible to distinguish between true and false logoi,4 if a distinction were recognized between, on the one hand, coining a word and stipulating how that word is to be used (establishing a convention) and, on the other hand, using that word once the convention has been established. If such a distinction were recognlized, it could be held that to use a word in- correctly is not to establish a new convention, but it may involve the

3 Robinson, 'A Criticism of Plato's Cratylus,' p. 328, N. Kretzmann, 'Plato on the Correctness of Names,' A.P.Q. 8 (1971) 127, K. Lorenz and M. Mittel- strass, 'On Rational Philosophy of Language: The Programme of Plato's Cratylus,' Mind, 76 (1967), 5-6. ' I shall translate 'logoi' as 'sentences' in what follows, but it should be made clear that this term is meant to be as neutral as possible with respect to philo- sophical associations.

136



speaker in saying something false. This is why I do not think that the passage at 385 B-C is intended as the beginning of a reductio ad ab- surdum argument against Hermogenes' position. Socrates simply wants to find out what Hermogenes thinks about the matter. Her- mogenes agrees without hesitation that sentences can be true or false, and accedes just as readily to Socrates' argument that names also can be true or false. The argument is as follows:

(1) Sentences can be true or false. (385 B 5) (2) The smallest parts of true sentences are all true. (385 C 5) (3) The smallest parts of sentences are names. (385 C 8) (4) Names are spoken as parts of sentences. (385 C 10) (5) Therefore the names spoken as parts of true sentences are

true. (385 C 12) (6) A part of a false sentence is false. (385 C 14) (7) Therefore names can be false. (385 C 16)

Why does Hermogenes accede so readily to this argument and still hold that the correctness of names consists in someone's giving a thing a name? Is Hermogenes simply a sloppy thinker? Should Socrates have claimed a victory here? I don't think so, for as I shall later argue, Socrates and Hermogenes agree that there is a distinction between true and false words on the one hand, and correct and incorrect words on the other.5 If Hermogenes is aware of this distinction at this point, he could readily agree that one need merely use words in a way not in accord with the convention one has established (which use does not create a new convention), in order to be speaking falsely, yet hold that the correctness of names is a matter of convention.

Before considering Socrates' next move, I want to point out something about the discussion so far. Whereas we are reasonably sure what Hermogenes thinks constitutes correctness of a name, we are simply not told at this point what either he or Socrates thinks constitutes truth or falsity of a name. Taylor, however, thinks that what this latter is is clear:

5 The Jowett translation would make one think otherwise, for it translates 385 D 5-6 as, 'And will there be so many names of each thing as everybody says that there are? and will they be true names at the time of uttering them?' The text reads, 'H xacl 6n6aa av -

-tq kxmatcp v6wmroa elVat, 'roaout-, lot xmcl ?6'r 6T6Tmv t; I think the Loeb translation is more faithful to the text. It reads, 'And whatever the number of names anyone says a thing has, it will really have that number at the time when he says it?'

137



His (Socrates') point is the sound one, that language is a social activity; it is primarily an instrument of communication. A "name"l given by me privately to something which everybody else calls differently does not discharge its function; it misleads, is a bad instrument for its purpose. This is what Socrates means by calling it a "false" name. It is a spurious substitute for the genuine article which would do the work required.6

The obvious objection to this interpretation is that if Taylor is right, Plato is using 'true' and 'false' ambiguously, for he tells us that by 'false sentence' he means 'that which says things as they are not', and not, 'that which does not succeed in communicating what is intended to the hearer', or something of the sort. I do not think there is any reason to suppose that Plato is using the words 'true' and 'false' ambiguously. In fact, there is evidence in the passage 430 E 9-431 B 9, which I shall discuss shortly, that he did not. Nor, as I shall show, is (2) meant to follow from (1). This is why I conclude that we simply aren't told at this point what assumptions (2) and (6) are based upon.

What is common to many critics of the passage 385 B-C is that they think that Plato assigns properties to words on the model of the properties of sentences. On the contrary, as is made clear from several passages, Plato assigns properties to sentences on the model of the properties of words (425 A 2-5, 431 B 3-9). In fact, as we shall see shortly, the argument that sentences can be true or false depends upon the premiss that words can be true or false. Thus we can say, contra Robinson,7 that Plato has not here committed a fallacy of division. Weingartner also seems to think that premiss (2) is meant to follow from (1), when he says,

The procedure at this point is simplistic and mechanical: if there are true and false statements, then their parts must be respectively true and false as well.8

Most commentators, whether or not they think that the argument contains a fallacy of division, think that the argument is a bad one, because they think that Plato means by 'logos' something like statement. Robinson states the criticism as follows:

'Socrates' approximates names to statements by asserting that if statements are true or false, names are true or false also. (385 C) In fact, statements are true or false because they describe

S A. E. Taylor, Plato (7th ed, London: Methuen, 1960), p. 79. 7 Robinson, 'A Criticism of Plato's Cratylus,' p. 335. 8 Weingartner, 'Making Sense of the Cratylus,' p. 14.

138



and assert; and names are neither true nor false because they do not assert or describe, but name or refer.9

We shall see, from what Plato means by saying that names can be either true or false, that it is most unlikely that Plato means by 'sentence' anything very close to what contemporary philosophers mean by 'statement'.

We return now to the text. Since Socrates is satisfied with Her- mogenes' agreement with his conclusion, it is not discussed again until Cratylus says that he thinks otherwise, and Socrates must defend his position. But in order to understand Socrates' arguments at 430 D- 431 B we must first follow Socrates' discussion with Hermogenes to its conclusion.

Socrates' next move is to focus on Hermogenes' opinions about the nature of things named. Hermogenes is readily persuaded, on the basis of the common-sense distinction between wise and foolish men, that things have a determinate nature which is not dependent upon men's opinions, and that 'not all things belong equally to all men at the same time and perpetually'.10 (386 D 4-8) Still, without further argument, this admission does not constitute a refutation of Her- mogenes' position, but it is a prerequisite for Socrates' position.

I think it is from this point that Socrates begins to develop his own views on the correctness of names. He first argues that activities which have specific goals cannot be performed in just any way, but the way they may be performed depends upon the nature of the task involved. (For example, I can cut down a tree with a saw, but I, at least, cannot fell a tree with a single blow of the side of my hand.) Socrates proposes to treat naming, both speaking and creating names, on an analogy with such crafts as weaving, cutting and burning. Given a particular goal, say, weaving a heavy cloak, the task may be performed well or badly, for the reasons that the shuttle may be well or poorly made, the yarn used may be suitable or not, the cloak may be woven evenly or not, it may or may not keep out the wind, it may have broken threads, and so on. Similarly, since speaking is an activity with a purpose, when we speak, we succeed only when the proper instrument is used, and is used properly. (387 B 12-C 4) The instrument

9 Robinson, 'A Criticism of Plato's Cratylus,' p. 335. 10 Unless otherwise indicated, the passages cited are in the Loeb Classical Library translation by H. N. Fowler, in Plato, Cratylus, Parmenides, Gyeater Hippias, Lesser Hippias (Cambridge, Mass: Harvard University Press, 1963), pp. 6-191.

139



is the name and the activity is 'teaching one another something and separating things according to their natures'. (388 B 10-12) Suppose we agree that a shuttle cannot be made in just any shape. Should we agree that not just any name will do for speaking of things, or that some names are better than others, or that the wordsmith must look to the form of a word in order to create the proper one? Certainly not at this point, for we aren't told what 'separating things according to their natures' comes to. (It might be reasonable to suppose that the name maker must look to the form of the name as the shuttle maker must look to the form of the shuttle if separating things according to their natures were something like putting different kinds of things into different piles. This activity would call for something like a crane or fork lift.)

After a barrage of examples from Socrates, Hermogenes says at 422 C 7-11 that he is prepared to concede that the correctness of naming, both constructing names and using them in speaking, is not arbitrary, but dependent upon the nature of the tasks involved. (Cratylus is probably gloating at this point, but we shall never know, since he hasn't said anything since the beginning of the dialogue.) Still, Hermogenes wants to know, and rightly so, wherein the correctness of names consists. Socrates answers Hermogenes' question about the correctness of names at 424 C 6-425 A 5:

Must not we, too, separate first the vowels, then in their several classes the consonants or mutes, as they are called by those who specialize in phonetics, and also the letters which are neither vowels nor mutes, as well as the various classes that exist among the vowels themselves? And when we have made all these divisions properly, we must in turn give names to all the things which ought to have them, if there are any names11 to which they can all, like

11 There is an important ambiguity in the sentence in which this word appears. The Greek text reads. XOCL &TCL8&V 'MU 8LEX6.LC?OO C5 7&VTC at4 ol8 8EL 6v6paTa' 17LMClVaL, et :LtV CLg & &VOCp;pE'TaL 7XV'r 4aCMep 'r&

ATOLXelX, g (LV IC 'rLV LaCV OUCT& t- xoXL et tv o(u'-roL4

IVea-rLV et xmnx -6v mu'Ctv Tp67rov e?oanp &V 'r0l4

aCOoLXeEot0 (424 D 1-5). What is the antecedent of & in 424 D 2? The Loeb translation takes it to be 'names' and Jowett takes it to be 'classes'. I think the Jowett translation is more likely to be correct, for the following reason. Plato assumes that we will be proceeding by the method of division (424 B 7-8) by which letters are put into groups according to whether they are vowels, consonants, etc. Simi-

140



the letters, be referred, from which it is possible to see what their (essential - 423 E 1-425 A 6) nature is and whether there are any classes among them, as there are among the letters. When we have properly examined all these points, we must know how to employ each letter with reference to its fitness, whether one letter is to be applied to one thing or many are to be combined; just as painters, when they wish to produce an imitation, sometimes use only red, sometimes some other color, and sometimes mix many colors, as when they are making a picture of a man or something of that sort, employing each color, I suppose, as they think the particular picture demands it. In just this way, we, too, shall apply letters to things, using one letter for one thing when that seems to be required, or many letters together, forming syllables, as they are called, and in turn combining syllables, and by their combination forming nouns and verbs. And from nouns and verbs again we shall finally construct something great and fair and complete. Just as in our comparison we made the picture by the art of painting, so now will we make language by the art of naming, or of rhetoric, or whatever it be.

We might wonder about two things at this point. One is that although the original controversy between Hermogenes and Cratylus seemed to be based on a conflict about the principles of correctness of proper names, we are presented here with a principle of correctness for all categorematic words, including common nouns and nominal and in- finitival verb forms. I think this is because in this dialogue Plato is interested in explicating what he takes to be the imitative function of language, and he notes early in the dialogue that proper names are not likely to imitate their bearers, for one reason or another.

We may also wonder whether Plato thinks names are names of forms of things, and if so, how we can ever talk about anything other than forms. This question presupposes, among other things, that to utter a word is to refer to the object of which it is the name. I do not thinik Plato thought this is the case, and will try to show that in what follows.

larly, he says, we should see into what groups things which are to get names fall before we name them. I suggest that Plato thinks we know the essential nature of each thing when we know the (natural) class to which it belongs, so that a thing's nature will be imitated by an imitation of the nature of the class to which it belongs.

141



It may seem at this point as though Socrates were placing himself in agreement with Cratylus, for now he seems to agree with Cratylus' position as stated at the beginning of the dialogue. Certainly Cratylus thinks so at 428 B-C. But Cratylus has not been attentive to Socrates' development of his position in terms of the activity of speaking, and so is not prepared for the deluge to follow. For Socrates has already introduced by means of the craft analogy an important distinction between names and the activity of using names in speaking. This distinction is not part of Hermogenes' original statement of Cratylus' position. Just as in the activity of weaving, the shuttle must be made correctly by the carpenter and used skilfully by the weaver, in the activity of instruction, the word must be made correctly by the law- maker and used skilfully by the teacher. Already Plato has made an enormous contribution to the nature-convention controversy by considering names in the context of the activity of speaking. No matter what we think of the details of the craft analogy, Socrates' move is vitally important for the subsequent refutation of Cratylus' position and the development of his own. Cratylus' statement of his own position is incomplete once we introduce the distinction between names and their use, in the same way that Hermogenes' was when Socrates began to discuss conventionalism with him. Does Cratylus really hold that no names are incorrect? Yes, Cratylus assures him. Does Cratylus really hold that there is no falsehood? Yes, Cratylus again assures him, using for his argument the slogan, 'How could anyone who says what he says, say that which is not?' (429 D 3-5) Socrates naively replies that this is all too deep and proceeds to offer an account of falsehood which is perfectly consistent with both Cratylus' original position and with his own. It seems to me that one or both of the following things may have occasioned Socrates' show of naivet6 at this point. One is that Plato may not yet have formulated the problem of falsehood as it is stated in the Sophist, the other, that Plato may have thought that the account of language he is developing in the Cratylus would solve the problem of falsehood. Socrates here chooses to leave aside for the moment the question of whether names can be incorrect, and focuses exclusively on what happens when names are used. According to the analogy developed at 424 C 6-425 A 5, words are imitations in the way that paintings are. (I presume that Plato has representational art in mind here.) Just as one can associate a given painting with the wrong bit of landscape, or a portrait with the wrong person, and so on, so can one associate a given word with the wrong

142



thing. How exactly is this possible? Socrates presents us with an interesting example:

Well, then, can I not step up to the same man again and say, "This is your name"? A name is an imitation, just as a picture is. Very well; can I not say to him, "This is your name," and then bring before his sense of hearing perhaps the imitation of himself, saying that it is a man, or perhaps the imitation of the female of the human species, saying that it is a woman? Do you not believe that this is possible and sometimes happens? (430 E 9-431 A 7)

Note that the details of this illustration are important. I would say that if someone came up to a man and said, 'Woman,' there would be no grounds for saying that what he had said were true or false, or that he had said anything (about anything) at all. But Socrates seems to understand 'some such assignment' so broadly that just uttering a word counts as assigning that word to a certain thing, for he says,

If, then, some such assignment of names takes place, we will call one kind speaking truth, and the other speaking falsehood. But if this is accepted, and if it is possible to assign names in- correctly and to give to objects not the names that befit them, but sometimes those that are unfitting, it would be possible to treat verbs in the same way. And if verbs and nouns can be assigned in this way, the same must be true of sentences, for sentences are, I conceive, a combination of verbs and nouns. (431 B 1-9)

I think that what is unsatisfactory about this account is that Plato seems to think that it is possible for a speaker, in uttering a single word, not in the course of uttering other words, both to make clear what it is he is talking about, and to say what that thing is. However unsatisfactory this account is because of its assumption that uttering a word counts as assigning (or matching) that word to a certain object, Cratylus accedes to it. Socrates uses the painting analogy to draw together his accounts of creating and using names in the following way: using words is like matching (representational) paintings with the things they represent, and creating words is like drawing (representational) pictures. Matching words or sentences with objects is an almost totally mysterious process at this point, and Plato does not elaborate enough on it for us to be sure what his theory of truth and falsehood is (and especially that of sentences, because we are not told exactly how words combine to form sentences). However, we can say something about the theory. To say that Plato treats words as

143



'little statements'12 on the basis of the fact that he says that words can be true or false is to turn Plato's theory on its head, for when his Socrates is put on the spot by Cratylus and forced to defend his claim that sentences are true or false, he argues that sentences can be true or false because they are made up of words which are true or false, rather than vice-versa. Words are not taken to have subjects and predicates, so they can hardly be considered as 'little statements'. It is not made clear how words are combined to form sentences, so it is by no means clear that even they are taken to have subjects and predicates. Perhaps the proponents of the 'little statement' interpretation could agree that Plato thinks that words have the force of statements when uttered. This interpretation has more plausibility than the former, for as we have seen, Plato thinks that when the word 'man' is uttered in the presence of a man, it is being said that he is a man. But we must still be careful not to give an analysis of the truth and falsity of words in terms of the truth and falsity of logoi (under- stood as statements) which are formed by their combination. Kahn gives the following reconstruction of Plato's reasoning concerning the truth of words, thereby making this mistake:

A simple, singular proposition of the form, "A is B" is true only if there is some x which is A and which is B, that is to say, only if there is some object of which "A" and "B" are both true. On the other hand, "A is B" is false only if there is some object of which "A" is true and "B" is false... Hence in a true sentence of this form both onomata must be true (of the extra-linguistic object), whereas in a false sentence at least one of the onomata must be false (in the same sense).13

I do not think that this is what Plato meant by saying that sentences are true or false, for he consistently held that names, both nouns and verbs, imitate the natures of things. (424 E 5-425 A 1) Now suppose that the sentence in question were, "A man walks", and suppose that the man in question were Theaetetus. According to Plato, the name 'man' would be true, since Theaetetus is essentially a man. However, it is not part of the nature of Theaetetus, or part of the nature of any man, to be walking, so the name 'walks' is not an appropriate name of Theaetetus. However, 'walks' is the name of the activity of waLking. (cf. 424 A 7-10) This interpretation is supported 12 Robinson, 'A Criticism of Plato's Cratylus,' p. 335, Lorenz and Mittelstrass, 'On Rational Philosophy of Language,' p. 6. 18 Kahn, 'Language and Ontology in the Cratylus', p. 160.

144



by the fact that precisely this analysis of verbs as names of actions gives Plato trouble in later dialogues in trying to explain how it is possible to say anything, that is, to do more than merely utter a string of names.

One last question remains to be answered. Does Plato hold the view that words and sentences, when uttered, can be more or less true, as he accepts the view that words coined are more or less correct? I think the answer is that he does not. Socrates argues that if the principle of correctness in names is imitation, no name could be perfect, for if some name were a perfect imitation of a thing, there would be no way to distinguish that word from the thing of which it is the name. Even so, we speak to each other and understand each other with the aid of some element of conventionality. About the use of imperfect words, Socrates says:

Grant that an inappropriate letter may be employed. But if a letter, then grant also that a noun in a clause, and if a noun, then also a clause in a sentence may be employed which is not appropriate to the things in question, and the thing may none the less be named and described, so long as the intrinsic quality of the thing named is retained. (432 E 3-433 A 1)

The point is that as in the construction of a word, a thing's nature can be imitated well or poorly, and still be named and described, so in the construction of a sentence can the elements of a situation be named well or badly and the resultant whole still be true - just how inac- curately the constituent elements can be named and the sentence still be true is not discussed.

University of Alberta

145



true and false names in the "cratylus"

Documents