Literatūra ISSN 0258-0802 eISSN 1648-1143

2025, vol. 67(3), pp. 8–28 DOI: https://doi.org/10.15388/Litera.2025.67.3.1

Straipsniai / Articles

Theory of Communication in Gorgias and Plato’s Cratylus

Mantas Adomėnas
Baltic Institute of Advanced Technology / Vilnius University, Faculty of Philology,
Institute of English, Romance, and Classical Studies
E-mail: mantas@adomenas.lt
ORCID ID https://orcid.org/0000-0003-1395-7648

Abstract. This essay attempts to establish a link between the third part of the Περὶ τοῦ μὴ ὄντος that contains Gorgias’ critique of communication and Plato’s theory of communication (as developed in the Cratylus). After analysing the text of, and attempting to reconstruct the original structure of Gorgias’ argument in, Part 3 of the Περὶ τοῦ μὴ ὄντος, the author contends that Plato, in spite of never addressing Gorgias’ work directly, formulates his theory of communication in the Cratylus within the conceptual schemes present in Gorgias’ treatise. Plato implicitly criticises Gorgias’ refutation of the possibility of communication, while retaining, nevertheless, certain features of Gorgias’ argument.
Keywords: Gorgias, Plato, Sophists, Cratylus, philosophy of language, communication.

Komunikacijos teorija: Gorgijas ir Platono Kratilas

Santrauka. Straipsnyje siekiama nustatyti ryšį tarp Gorgijo veikalo Περὶ τοῦ μὴ ὄντος trečiosios dalies, kurioje pateikiama Gorgijo komunikacijos kritika, ir Platono komunikacijos teorijos, išplėtotos dialoge Kratilas. Išanalizavus tekstą ir pamėginus atkurti Gorgijo argumento, pateikiamo Περὶ τοῦ μὴ ὄντος trečiojoje dalyje, originalią struktūrą, daroma išvada, kad Platonas, tiesiogiai neaptardamas Gorgijo veikalo, savo komunikacijos teoriją dialoge Kratilas formuoja remdamasis Gorgijo traktato konceptualinėmis schemomis. Platonas netiesiogiai kritikuoja Gorgijo komunikacijos galimybės atmetimą, tačiau išlaiko tam tikras Gorgijo argumento ypatybes.
Reikšminiai žodžiai: Gorgijas, Platonas, sofistai, Kratilas, kalbos filosofija, komunikacija.

_________

Received: 25/09/2025. Accepted: 24/10/2025.
Copyright © Mantas Adomėnas, 2025. Published by Vilnius University Press.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

I. Introduction

The aim of this paper is to reconstruct the features of Gorgias’ argument against communication in the third part of his lost treatise Περὶ τοῦ μὴ ὄντος which are picked up by, and seemingly exert some impact upon, the conceptualisation of communication in Plato’s Cratylus. The reconstruction is far from a trivial task, insofar as the treatise survives only in two incomplete, tendentious, and sometimes mutually inconsistent epitomes contained in Sextus’ treatise Against Mathematicians (Adv. Math. 7.65–87) and in the Pseudo-Aristotelian De Melisso Xenophane Gorgia (MXG 5.1–5 (979a11–33), 6.9–25 979b20–980b21). Thus the movement of interpretation is twofold: on the one hand, restoring the often-obscured original Gorgias’ argument structure from fragmentary retellings, and, on the other hand, to trace lines of convergence and potential influence and/or reaction with the relevant sections in Plato’s Cratylus.

Before we start reconstruction of Gorgias’ arguments, however, it is useful to recall the informative description of the argumentative pattern in the Περὶ τοῦ μὴ ὄντος provided by Sedley: “one aspect which remains constant in both [MXG and Sextus’] versions is the concessive structure of the overall argument: p, and even if not p, q, and even if not q, r. This device is a direct legacy of Gorgias” (Sedley, 1992, p. 25–26, n. 8). The other question that has to be provisionally settled before the discussion of the text is the interrelation between the Sextus’ and MXG versions. Prima facie MXG version is superior: it is acknowledged that MXG presents a more complete summary in general, and, as regards the third part of the Περὶ τοῦ μὴ ὄντος, MXG gives four distinct arguments where Sextus seemingly adduces only one. There is also the abundance of technical terminology of later schools in Sextus’ version. Nonetheless, the clear presence of the concessive structure of argument in Sextus (which is especially cogent since it is not clearly recognised as such in the text), and twice repeated indication of direct borrowing from the author (cf. φησίν in § 85 & § 86) requires a closer examination before any final pronouncement is made.

II. Reconstructing Gorgias’ Argument against the Possibility of Communication in the Περὶ τοῦ μὴ ὄντος

II.1. Logos, Sight, and Sound (MXG 980a19–b8, Sextus §§ 83–86)

The argument that is proposed in this section of MXG was well summarised by Kerferd (regardless of the fact that he was talking about Sextus’ argument): “externally existing objects are perceived by sight and hearing (and presumably by the other senses as well) in each case exclusively according to the sensible characteristics appropriate to the sense in question – visible characteristics by sight, and audible characteristics by the faculty of hearing. There is no possibility of interchange – we never see sounds nor do we hear sights. [...] It is impossible for us to communicate such perceptions to other people. This is because communication is exclusively by means of speech or words, and the externally existing objects are not words. There is no possibility of converting things into words, and as a result there is no possibility of communicating things through, or by means of, words” (Kerferd, 1984, pp. 217–218).

This is a fair summary of the argument in MXG. Rhetorical questions are posed in the opening of the section (let us note concessionary transition from the previous section): εἰ δὲ καὶ γνωστά, πῶς ἄν τις, φησί, δηλώσειεν ἄλλῳ; ὃ γὰρ εἶδε, πῶς ἄν τις, φησί, τοῦτο εἴποι λόγῳ; ἢ πῶς ἂν ἐκεῖνο δῆλον ἀκούσαντι γίγνοιτο, μὴ ἰδόντι; “But even if they are knowable, how could someone, he asks, indicate them to someone else? For what one sees, how, he asks, could one say this by a speech? Or how could that thing become clear to someone who hears, but does not see it?” (980a20–22)1 – are answered by the separation of the sensory domains of sight and sound, both mutually and in respect to λόγος: ὥσπερ γὰρ οὐδὲ ἡ ὄψις τοὺς φθόγγους γιγνώσκει, οὕτως οὐδὲ ἡ ἀκοὴ τὰ χρώματα ἀκούει, ἀλλὰ φθόγγους· καὶ λέγει ὁ λέγων, ἀλλ’ οὐ χρῶμα οὐδὲ πρᾶγμα, “For just as sight does not know sounds, so too hearing does not hear colours, but sounds: and someone who speaks utters a speech, but not a colour or a thing” (980b1–3).

So far, this separation might seem to allow for the communication within the same sensory domain: one might be able to communicate certain phonetic data through the medium of sound, and certain visual data through visual medium, e.g., by drawing the object of communication.2 The last words of the section, however, indicate that Gorgias was concerned with more fundamental issues than such borderline cases: πρᾶγμα can be the content of language no more than a colour is (taking both χρῶμα and πρᾶγμα to be internal objects of λέγει; cf. 980b6). Sensory perception here functions as a vehicle of an analogy3 in order to make a point about separateness of speech and πράγματα (cf. introductory ὥσπερ at 980b1).

What is the meaning of the πρᾶγμα in this context? Apparently, πρᾶγμα here picks up πράγματα from the previous section, where it summarises the impossibility of knowledge: ἡμῖν γ’ ἄγνωστ’ ἂν εἶναι πράγματα (980a18–19). Therefore, the meaning of πρᾶγμα / πράγματα ought not to be restricted to ‘things’ in the sense of ‘perceptible things’ (as the source of Sextus understands it, cf. § 85, 3), for, when Gorgias speaks about concrete senses in the Περὶ τοῦ μὴ ὄντος, he uses χρῶμα to describe the object of sight, and ψόφος / φθόγγος for the object of hearing. πράγματα thus seems to be the totality of ‘things’ outside the knowing subject (cf. 980b19–20, as well as 979a27–28), of which the senses give an imperfect rendering and whose existence and knowability are in doubt. On this reading, what this clause states is not only the mutual exclusion of language and sensory domains, but also the radical separation, on the basis of analogy from sensory domains, of language and non-linguistic reality.4 In fact, this statement is the only basis of the employment of such radical difference between λόγος and πρᾶγμα one line later, where it is adduced as something already proven and agreed to: πῶς αὐτὸ παρ’ ἄλλου λόγῳ ἢ σημείῳ τινὶ ἑτέρῳ τοῦ πράγματος ἐννοήσει: “how can he understand it from someone else by means of a speech or of some sign different from the thing?” (980b4–5).

The section that follows introduces a further dimension: if something is not among the contents of one’s mind (ὃ οὖν τις μὴ ἐννοεῖ 980b3), how is it possible to grasp it “through language or through some sign, different from the thing [itself]”? (πῶς αὐτὸ παρ’ ἄλλου λόγῳ ἢ σημείῳ τινὶ ἑτέρῳ τοῦ πράγματος ἐννοήσει; 980b4–5). The response to this question implies that, in Gorgias’ understanding, λόγος has nothing in common with the phonetic (or, in the case of the written communication, visual) aspect of language – cf. the differentiation of λόγος from ψόφος and χρῶμα at 980b5: ἀλλ’ ἢ ἐὰν μὲν χρῶμα, ἰδών, ἐὰν δὲ <ψόφος, ἀκού>σας: “if it is a colour, by seeing it, or, if it is <a sound, by hearing it>”. But then, λόγος has to be used in the sense of ‘intended meaning’: οὐ <ψόφον> λέγει <ὁ λέ>γων οὐδὲ χρῶμα, ἀλλὰ λόγον “<someone who speaks> does not say <a sound> or a colour, but a speech” (980b6). As such, it is identical to ‘thought’: [...] ἀλλὰ λόγον· ὥστ’ οὐδὲ διανοεῖσθαι χρῶμά ἐστιν, ἀλλ’ ὁρᾶν, οὐδὲ ψόφον, ἀλλ’ ἀκούειν: “but a speech; so that it is not possible to think a colour, but to see it, nor a sound, but to hear it” (980b6–8).5

Let us now return to the Sextus’ version of Part 3 of the Περὶ τοῦ μὴ ὄντος. It is usually maintained that Sextus gives a somewhat expanded paraphrasis of the argument that is essentially identical to the one discussed – viz., the argument from the analogy of separateness of sensory domains. Thus, in the words of Kerferd, “[MXG] makes the very same point that was found in the main argument at the commencement of Sextus’ third section. It is however expressed more briefly” (Kerferd, 1984, p. 219), whereas “[t]he argument in Sextus’ third section consists of one main argument [...] and two subordinate arguments which are in effect answers to possible objections to the main argument” (Kerferd, 1984, p. 217). And, while the ‘main’ argument (83–85, 1) clearly paraphrases the same text as the one analysed by us, at least certain authenticity of the parts of the texts which Kerferd (op. cit.) calls ‘subordinate arguments’ (85–86), and which, for Newiger, are clear indications of deficiency of the Sextus’ version (Newiger, 1973, pp. 163–167), is vouchsafed by the occurrence of φησίν (85, 2; 86, 4) which indicates Gorgias’ authorship of at least the core of the arguments in question.

An even clearer indication of the authenticity of this version is the vestiges of the concessionary structure of the argument. Let us turn our attention to the only survival of that structure (symptomatically, it also contains one of the two φησίν): εἰ γὰρ καὶ ὑπόκειται, φησίν, ὁ λόγος, ἀλλὰ διαφέρει τῶν λοιπῶν ὑποκειμένων: “For even if speech exists, he says, it nonetheless differs from all the other things that exist” (86, 4–5). This sentence implies that concession is made so that λόγος can be considered to be ‘existent’ (ὑποκεῖσθαι is used by Sextus interchangeably with εἶναι – cf. τὰ ὑποκείμενα καὶ ὄντα at 84, 1–2). Thus the previous position must have been that λόγος is ‘non-existent’, μὴ ὤν. And indeed, slightly above, we find the following expressions: λόγος δὲ οὐκ ἐστι τὰ ὑποκείμενα καὶ ὄντα: “but speech is not the things that exist and that are” (84, 1–2). Thus, the text contains clear evidence that, prior to the assumption that λόγος must be ‘existent’, it was not considered to be one of the existing things.6 The concession concedes (after having first denied) the existence to λόγος.

II.2. Characteristic Traits of Gorgias’ Argument throughout the Περὶ τοῦ μὴ ὄντος

Now I would like to draw attention to certain features of the previous two parts of the Περὶ τοῦ μὴ ὄντος, preserved in both Sextus’ and MXG versions, namely, (i) the dilemmatic opening of the major subdivisions of the text7 and (ii) the argument from the convertibility of a negative expression.

All the major sections of the text, both in Sextus’ and in the MXG versions, open with outlining a dilemma – by stating the only possible alternative answers to the question at hand:8

εἰ δέ ἐστιν, ἤτοι ἀγένητον ἢ γενόμενον εἶναι
“if [scil. something] is, it is either ungenerated or generated” (979b20–1);

ἐτι εἴπερ ἐστι <τι, ἢ> ἓν ἢ πλείω, φησίν, ἐστίν
“Again, if [scil. something] is, he says, it is either one or more” (979b35 & cf. 73);

εἰ γάρ ἐστι <τι>, ἤτοι τὸ ὄν ἐστιν ἢ τὸ μὴ ὄν9
“if [scil. something] is, then it is either what is or what is not” (66; the same disjunction is implied in οὐκ ἐστιν οὔτε εἶναι, οὔτε μὴ εἶναι (979a24));

εἰ γὰρ τὸ ὄν ἐστιν, ἤτοι ἀΐδιόν ἐστιν ἢ γενητόν
“For if what is is, it is either eternal, or generated” (68);

εἰ γὰρ γέγονεν, ἤτοι ἐξ ὄντος ἢ ἐκ μὴ ὄντος γέγονεν
“For if it has come to be, it has come to be generated either out of what is or out of what is not” (71);

εἰ γὰρ ἕν ἐστιν, ἤτοι ποσόν ἐστιν ἢ συνεχές ἐστιν ἢ μέγεθός ἐστιν ἢ σῶμά ἐστιν
“For if it is one, it is either a [scil. discrete] quantity, or continuous, or a magnitude, or a body” (73).

Besides, collation of the openings of §§ 77–79 yields one more disjunction:

<ἤτοι> τὰ φρονούμενα οὐκ ἐστιν ὄντα <ἢ> τὰ φρονούμενά ἐστιν ὄντα
“<Either> thoughts are not things that are <or> thoughts are things that are”.

The same disjunction is preserved and integrated into the argument at 980a12–15 in MXG:

καὶ γὰρ τὰ ὁρώμενα καὶ ἀκουόμενα διὰ τοῦτο ἐστιν, ὅτι φρονεῖται ἕκαστα αὐτῶν· εἰ δὲ μὴ διὰ τοῦτο, ἀλλ’ ὥσπερ οὐδὲν μᾶλλον ἃ ὁρῶμεν ἐστιν...
“For things seen and things heard are for the reason that each of them is an object of thought (phroneitai) – and even if it is not for this reason, but just as what we see is not more [scil. by the fact that it is seen]...”

Secondly, both Part 1 and Part 2 of the Περὶ τοῦ μὴ ὄντος seem to contain an argument from the convertibility of a negative expression, or ex communicatione negationis – “If non-A f, then A non-f” and vice versa. Its fullest statement is provided by Sextus’ paraphrase:

εἰ γὰρ τὰ φρονούμενα, φησὶν ὁ Γοργίας, οὐκ ἐστιν ὄντα, τὸ ὂν οὐ φρονεῖται. καὶ κατὰ λόγον· ὥσπερ γὰρ εἰ τοῖς φρονουμένοις συμβέβηκεν εἶναι λευκοῖς, κἂν συμβεβήκει τοῖς λευκοῖς φρονεῖσθαι, οὕτως εἰ τοῖς φρονουμένοις συμβέβηκεν μὴ εἶναι οὖσι, κατ’ ἀνάγκην συμβήσεται τοῖς οὖσι μὴ φρονεῖσθαι. διόπερ ὑγιὲς καὶ σῷζον τὴν ἀκολουθίαν ἐστὶ τὸ «εἰ τὰ φρονούμενα οὐκ ἐστιν ὄντα, τὸ ὂν οὐ φρονεῖται»
“For if thoughts, says Gorgias, are not things that are, then what is is not thought of. And this is in conformity with reason: for just as, if it happened to thoughts to be white, it would also happen to white things to be thoughts of, so too if it happens to thoughts to not be things that are, then of necessity it will happen to the things that are to not be thoughts of. [78] That is why [scil. the argument] ‘if thoughts are not things that are, then what is is not something thought of’ is sound and preserves logical consequence” (77–78).10

An argument of the same form is preserved in the section of MXG corresponding to Part 1 of the Περὶ τοῦ μὴ ὄντος:

εἰ δ’ ὅμως τὸ μὴ εἶναί ἐστι, τὸ εἶναι, φησί, οὐκ ἐστι, τὸ ἀντικείμενον. εἰ γὰρ τὸ μὴ εἶναί ἐστι τὸ εἶναι μὴ εἶναι προσήκει
“But if nevertheless nonbeing is, being, he says, is not – on the contrary. For if nonbeing is, it is fitting that being not be” (979a30–32).

Again, a more expanded and paraphrased version can be found in Sextus:11

καὶ ἄλλως, εἰ τὸ μὴ ὄν ἐστι, τὸ ὂν οὐκ ἔσται· ἐναντία γάρ ἐστι ταῦτα ἀλλήλοις, καὶ εἰ τῷ μὴ ὄντι συμβέβηκε τὸ εἶναι, τῷ ὄντι συμβήσεται τὸ μὴ εἶναι
“And in a different way: if what is not is, what is will not be. For these are opposed to each other, and if being comes about for what is not, nonbeing will come about for what is” (67).

The special attention paid by the paraphrasts to this argument12 leads to believe that the argument ἀπὸ ἐναντίων / ἀντικειμένων was something peculiar to this treatise, maybe of Gorgias’ own invention. Thus we see that Gorgias in his treatise was using certain argumentative procedures and stock-arguments that could be applied in all three parts, thereby giving his treatise the impression of exhaustiveness and of the logical necessity of its conclusions. One of these procedures was dilemmatic opening: ‘Either A or B’, and one of the stock-arguments that we see being employed in both Part 1 and Part 2 is the argument from convertibility of negative expression.

II.3. Reconstruction of Gorgias’ Argument against the Possibility of Communication Continued

In the view of this, it may be surmised that the same features would have applied to the third part. Then, the tentative reconstruction of the beginning of Part 3 would be the following:

εἰ δὲ καὶ γνωστά, <ἀλλ’ οὐ δηλωτὰ ἄλλοις. ᾧ γὰρ δηλοῦμεν ἤτοι τὸ ὄν ἐστι ἤ μὴ ὄν· τοῦτο δ’ ὁ λόγος. καὶ εἰ μὲν λόγος οὐκ ἐστι τὸ ὄν, οὐδὲ τὸ ὄν λέγεται>
“Even if [things] can be known, <they are nevertheless not capable being communicated to others. For what we communicate with is either existent or non existent – and this is speech. And if speech is not an existent, neither will the existent be spoken of” (=argument from convertibility of negative expression; author’s translation).

This is, in fact, what is contained in the first sentence of § 84: ᾧ γὰρ μηνύομεν, ἐστι λόγος, λόγος δὲ οὐκ ἐστι τὰ ὑποκείμενα καὶ ὄντα· οὐκ ἄρα τὰ ὄντα μηνύομεν: “For what we indicate by is speech, but the things that exist and that are are not speech. So it is not the things that are that we indicate to other people”. Then Gorgias proceeds to draw one more conclusion from the same argument: if what we signify is not τὰ ὄντα (and granted the premise that λόγος is not one of τὰ ὄντα), then what we signify is (only) λόγος, which is radically different from the πράγματα: οὐκ ἄρα τὰ ὄντα μηνύομεν τοῖς πέλας ἀλλὰ λόγον, ὃς ἕτερός ἐστι τῶν ὑποκειμένων: “So it is not the things that are that we indicate to other people, but rather speech, which is different from the things that exist” (§ 84). Analogy of sensory domains is adduced to illustrate this radical difference: καθάπερ οὖν τὸ ὁρατὸν οὐκ ἂν γένοιτο ἀκουστὸν καὶ ἀνάπαλιν, οὕτως ἐπεὶ ὑπόκειται τὸ ὂν ἐκτός, οὐκ ἂν γένοιτο λόγος ὁ ἡμέτερος “So just as what is visible could not become audible and vice versa, so too, since what is exists externally, it could not become our speech” (ibid.). That is to say, just as what is visible cannot become audible and vice versa, logos cannot become ‘things’, and ‘existing’ things cannot become logos (without, we may surmise, ceasing to be onta). Thus we can see that the radical difference between λόγος and πράγματα is based not solely upon the analogy of the sensory domains. To prove it, Gorgias also employs the argument from convertibility of negative expression – and then goes on to elaborate upon its conclusion.

Neither the evidence of the Gorgias’ method elsewhere, nor the state of the text (especially since we are dealing with a very heavy paraphrasis here) permits us to expect a clear and logical transition to the next argument. The first sentence of § 85 indicates that we are still within the first part of the (hypothetical) initial disjunction (λόγος as μὴ ὄν), but it also reveals that the accent is on the impossibility of communicating to the ‘other’: μὴ ὢν δὲ λόγος οὐκ ἂν δηλωθείη ἑτέρῳ: “and as speech is nonexistent, it could not be revealed to another person” (§ 85). It would be a mistake to look, in what follows, for an account of mechanism of perception (as Mazzara does; see Mazzara, 1983, pp. 130–132, and cf. Mazzara, 1982, ad loc.) or for behavioural theory of communication (as Mourelatos’ hint leads us to believe, cf. Mourelatos, 1985, p. 613–614). What this passage says is that λόγος is secondary to πράγματα, being the response to certain external experience in the case of sensations:

ὅ γε μὴν λόγος, φησίν, ἀπὸ τῶν ἐξωθεν προσπιπτόντων ἡμῖν πραγμάτων συνίσταται [...].13 ἐκ γὰρ τῆς τοῦ χυλοῦ ἐγκυρήσεως ἐγγίνεται ἡμῖν ὁ κατὰ ταύτης τῆς ποιότητος ἐκφερόμενος λόγος, καὶ ἐκ τῆς τοῦ χρώματος ὑποπτώσεως ὁ κατὰ τοῦ χρώματος
“Surely speech, he says, is constituted out of the external things that strike us [...]. For from the encounter with flavour there comes to us speech that expresses this quality, and from the occurrence of colour that which expresses colour” (§ 85).

Therefore, it is not the case that λόγος indicates or represents things, but things provide explanation / meaning to the λόγος: οὐχ ὁ λόγος τοῦ ἐκτὸς παραστατικός ἐστιν, ἀλλὰ τὸ ἐκτὸς τοῦ λόγου μηνυτικὸν γίνεται: “But if this is the case, it is not the speech that presents the external thing, but the external thing that indicates the speech” (ibid.). Certain λόγος can be understood only in reference to the object that caused it (this is the meaning τὸ ἐκτὸς τοῦ λόγου μηνυτικὸν γίνεται). What does it mean? If we assume that sensory perception is once again used to provide an analogy or a contrast, a possible implication would be that even if it is possible to verify a certain piece of communication about the external objects by referring to the objects themselves (or rather, to put it another way, exposure to the same πράγματα would cause identical impressions), there is no possibility of such identical impressions in the case of διάνοια, even though it may be the ‘cause’ of λόγος (just as in the case of sensory perceptions and πράγματα). In the case of thought processes, there is no objectively existing reality to which both participants of the act of communication could be exposed, and thus it is impossible for λόγος, a certain unit of meaning, to be communicated to another person.

Before proceeding to the next section, it has to be remarked that the notion of ‘existence’ qua ‘objectivity’ that is operating here confirms the opening statement of the paragraph – that λόγος is μὴ ὤν. The latter must mean something like ‘not objectively existent’, or ‘not existing in the sense that πράγματα exist’ (cf. carriages in the sea at § 79 – although there is διάνοια of them, they are not existent). λόγος is not existent insofar as it is something subjectively and not objectively existent. Gorgias is not arguing that acts of speech do not take place, but rather that λόγος – meaning – remains confined to its performer. What he advocates seems to be a certain solipsism – λόγος, the meaning that one tries to convey, is confined to the person performing the communication in the same fashion that thoughts remain confined in the individual. This, in its turn, fits very well with the picture we get in MXG, 980b3–8.

Words in the beginning of § 86 up until καὶ ὄντα μηνύεσθαι are a gloss and ought to be disregarded. What follows is the concession “Even if λόγος exists...” and its consequences: εἰ γὰρ καὶ ὑπόκειται, φησίν, ὁ λόγος, ἀλλὰ διαφέρει τῶν λοιπῶν ὑποκειμένων: “For even if speech exists, he says, it nonetheless differs from all the other things that exist”. I should not venture any further, since the remaining text is very strongly tinged by the inferences stemming from the weak reading of the difference between sensory domains (cf. n. 2 supra). Statement πλείστῳ διενήνοχε τὰ ὁρατὰ σώματα τῶν λόγων: “there is nothing that differs more than visible bodies and speeches” is asserted on the basis of difference in (sensory) organs: δι’ ἑτέρου γὰρ ὀργάνου ληπτόν ἐστι τὸ ὁρατὸν καὶ δι’ ἄλλου ὁ λόγος: “For what is visible is grasped by one organ, speech by a different one”. The most one can say about λόγος from MXG evidence is that it has no organ at all – otherwise Gorgias’ arguments against communication are futile. The concluding sentence: οὐκ ἄρα ἐνδείκνυται τὰ πολλὰ τῶν ὑποκειμένων ὁ λόγος, ὥσπερ οὐδὲ ἐκεῖνα τὴν ἀλλήλων διαδηλοῖ φύσιν: “So speech does not indicate the multitude of things that exist, just as these do not reveal their nature to each other” – betrays traces of the (possibly genuine) analogy of the sensory domains, but otherwise it follows the same weak reading.

However, the remaining bit of reliable evidence in the middle of the section makes it safe to assume that this is the point where MXG presentation of Part 3 actually starts. MXG omits the dilemmatic opening and the argument from the convertibility of negative expression, and preserves only the concessionary opening tag of Part 3: εἰ δὲ καὶ γνωστά, πῶς ἄν τις, φησί, δηλώσειεν ἄλλῳ; “But even if they are knowable, how could someone, he asks, indicate them to someone else?”.

What about the beginning of Sextus’ account, though (§ 83)? Where does that fit in? So far it seemed that Sextus’ account has preserved, although in a mangled, heavily paraphrased and interpolated form, the actual sequence of the beginning of Part 3 of the Περὶ τοῦ μὴ ὄντος. The first and the last phrases of the paragraph correspond to MXG opening – cf. Sextus’ § 83: καὶ εἰ καταλαμβάνοιτο δέ, ἀνέξοιστον ἑτέρῳ. [...] πῶς οὖν δύναται ταῦτα ἑτέρῳ μηνύεσθαι; “But even if it were apprehended, it would not be expressible for someone else. [...] [H]ow then can these be indicated to someone else?” and MXG 980a19–20: εἰ δὲ καὶ γνωστά, πῶς ἄν τις [...] δηλώσειεν ἄλλῳ; “But even if they are knowable, how could someone [...] indicate them to someone else?”. The assumption τὰ ὄντα ὁρατά ἐστι καὶ ἀκουστὰ καὶ κοινῶς αἰσθητά, ἅπερ ἐκτός ὑπόκειται: “the things that are are visible and audible and, in general, perceptible, those that exist externally” (§ 83) may be safely ascribed to Sextus. What remains is a statement of a mutual exclusivity of the sensory domains: τὰ μὲν ὁρατὰ ὁράσει καταληπτά ἐστι τὰ δὲ ἀκουστὰ ἀκοῇ καὶ οὐκ ἐναλλάξ: “among these, the visible ones are apprehended by sight and the audible ones by hearing, and not reciprocally” (ibid.). If we maintain that this statement corresponds to some feature of the original text, it is reasonable to assume that Gorgias, before proceeding with his argument, here introduces his understanding of the senses and sense perception that will function as an analogy throughout Part 3.

In this connection, in particular, testimony from Plato’s Meno springs to mind, where Socrates gives definition of colour à la Gorgias in order to suit Meno’s, Gorgias’ disciple’s, intellectual tastes: Βούλει οὖν σοι κατὰ Γοργίαν ἀποκρίνωμαι, ᾗ ἂν σὺ μάλιστα ἀκολουθήσαις; “Do you want me to answer after the manner of Gorgias, which you would most easily follow?” (76c4–5; for the Meno, G.M.A. Grube’s translation is quoted throughout). The definition itself – ἐστιν γὰρ χρόα ἀπορροὴ σχημάτων ὄψει σύμμετρος καὶ αἰσθητός: “colour is an effluvium from shapes which fits the sight and is perceived” (76d4–5) – presupposes Empedoclean theory of sensation, where sensations are constituted by the effluvies (ἀπορροαί) of things getting into appropriate ducts, or channels (πόροι):

Οὐκοῦν λέγετε ἀπορροάς τινας τῶν ὄντων κατὰ Ἐμπεδοκλέα; [...] Καὶ πόρους εἰς οὓς καὶ δι’ ὧν αἱ ἀπορροαὶ πορεύονται; [...] Καὶ τῶν ἀπορροῶν τὰς μὲν ἁρμόττειν ἐνίοις τῶν πόρων, τὰς δὲ ἐλάττους ἢ μείζους εἶναι;
“Do you both say there are effluvia of things, as Empedocles does? [...] And that there are channels through which the effluvia make their way? [...] And some effluvia fit some of the channels, while others are too small or too big?” (76c7–d1).

According to Socrates, it will be possible to build definitions of other sensations following the same model:

καὶ ἅμα οἶμαι ἐννοεῖς ὅτι ἔχοις ἂν ἐξ αὐτῆς εἰπεῖν καὶ φωνὴν ὅ ἐστι, καὶ ὀσμὴν καὶ ἄλλα πολλὰ τῶν τοιούτων
“At the same time I think that you can deduce from this answer what sound is, and smell, and many such things” (76d8–e1)

The ascription of this theory to Empedocles is unproblematic as we have several other testimonies that Empedocles held some such view (cf. Theophr. De sensu 7, Arist. De gen. et corr. 324b26 sqq., Ps.-Plut. De plac. philos. 4, 9, as well as fr. B 89 DK). If it can be maintained that this theory was also taken up by Gorgias (and, in spite of Bluck, 1961, pp. 251–253, there are no real grounds to entertain doubt on this matter),14 the most natural context for it to be would be the opening of Part 3 of the Περὶ τοῦ μὴ ὄντος, where through the reference to that theory Gorgias would establish separateness of the sensory domains which will serve as a vehicle of analogy throughout Part 3. If each sensory organ receives only the effluvies commensurable with it, this theory provides a ‘scientific’ argument for Gorgias’ position which, of course, can alternatively be argued from the self-identity of the senses. It would, however, be useful for Gorgias in ‘proving’ the incommunicability: having said that each sense is generated by commensurate effluvies, he could further ask, “And how can one receive the effluvies of λόγος (provided that there were such)?” It can be neither hearing nor sight, since οὐ <ψόφον> λέγει <ὁ λέ>γων οὐδὲ χρῶμα, ἀλλὰ λόγον κτλ. “<someone who speaks> does not say <a sound> or a colour, but a speech etc.” (MXG 980b6).15

Mourelatos’ objections to reading Plato’s testimony in the context of the Περὶ τοῦ μὴ ὄντος are based on his assumptions concerning the overall character of Gorgias’ treatise: “[M]odern scholars have deployed the strategy of importing doctrines of other pre-Socratics as premises or considerations for Gorgias’ argument. The strategy is in all cases ill-conceived. Since Gorgias has highly paradoxical conclusions to defend, he would have been foolish to draw on speculative theses that are in themselves controversial” (Mourelatos, 1985, p. 610), and further: “Gorgias [...] had better not borrow premises from Empedoclean physiology” (ibid., p. 622). Maybe he had better not, and maybe, “It would [...] be poor strategy” (ibid., n. 30 p. 622); but the question is, did he or did he not? The testimony of MXG indicates that he was drawing on the arguments of others – notably, Melissus (979a22, b22) and Zeno (979a23, b25, b37); that there was a considerable correspondence between some of his views and those of Leucippus (980a7–8); that he was arguing his position by contrasting the opinions of others (979a14–18). In the context of this eclecticism, why not Empedocles, whose disciple he is reported to be? As to the question whether Gorgias wanted to confer visibility of philosophical sophistication to his argument, or to argue from the commonly accepted premises, from what we know, the former option seems to be more plausible. Besides, we must not forget that we are dealing – at best – with a summary and epitome of Gorgias’ argument, and that the majority of supporting arguments and considerations must be presumed lost.

III. Plato’s Theory of Communication in the Cratylus

III.1. Plato’s Conceptualisation of the Problem

If the passage in the Meno goes back to the Περὶ τοῦ μὴ ὄντος, apparently, it is the most explicit allusion to Part 3 of the Περὶ τοῦ μὴ ὄντος in Plato.16 If Plato was familiar with Part 3 of the Περὶ τοῦ μὴ ὄντος, the reading of the ostensibly definitive refutation of the possibility of communication given by the Περὶ τοῦ μὴ ὄντος was bound to raise the following questions:17 How, on what model, can communication be possible? What must be the ontological status of the sign? How is solipsism posited by Gorgias to be overcome, and what philosophical presuppositions would this model require? In what follows, I shall attempt to delineate Plato’s answers to these questions, while highlighting the points of affinity between Gorgias’ critique of communication in the Περὶ τοῦ μὴ ὄντος, and Plato’s theory of communication that occupies the second half of the Cratylus (from 422d onwards).18

Gorgias’ critique, despite its elaborate character, hinges upon one fundamental insight: the separation of the sensory domains from one another, and of sense-perception from internal experience. While communicating, one does not transmit meaning, but only sound or a visual signal. Meaning (λόγος) belongs to a different order, non-overlapping with that of sensory experiences. For Gorgias, it is a sign of impossibility of communication. Plato accepts Gorgias’ premises, but posits the question differently: given that, how does it happen that we, having received a sensory signal, reconstruct its meaning?19 Plato’s answer is given in the context of the search for the ‘correctness of names’ – in fact, correct names are those that satisfy the criteria for successful communication (387b11–d8), and therefore it is justified to say that conditions of the correctness of names are identical to those for the possibility of communication. ‘Correctness of names’ is the same for all names: μία ὀρθότης παντὸς ὀνόματος: “there is only one kind of correctness in all names” (422c7–8; for the Cratylus, C.D.C. Reeve’s translation is quoted throughout, slightly adapted where necessary), namely, to reveal what things are: τῶν ὀνομάτων ὀρθότης τοιαύτη τις ἐβούλετο εἶναι, οἵα δηλοῦν οἷον ἕκαστόν ἐστι τῶν ὄντων: “the correctness of every name we analyzed was intended to consist in its expressing (dēloun) the nature of one of the things that are” (422d1–3). Composite and derivative ‘later’ words are able to achieve that (ἀπεργάζεσθαι) through ‘simple’ and primary ‘first names’, πρῶτα ὀνόματα (422d8–9). Provided that Gorgias’ critique is taken seriously, certain regular combinations of sounds coming from outside is all we can assume when talking about ‘names’. In order to ‘reveal’ or ‘disclose’ things, names must contain information about things. How can combinations of sounds (pure noise, we might say) contain information about things? One possible answer is: by convention. An answer of this kind had been discarded, on different grounds, earlier in the dialogue. Besides, by what means is one to make such a convention? In order to make it, one has to possess a functioning communication already, etc. So the question actually becomes that of the beginning, of a first successful act of communication whereby intersubjective agreement could be established. In the context of the Cratylus, this question is conceptualised as the question about the first names which do not have any antecedents: τὰ δὲ δὴ πρῶτα, οἷς οὔπω ἕτερα ὑπόκειται, τίνι τρόπῳ κατὰ τὸ δυνατὸν ὅτι μάλιστα φανερὰ ἡμῖν ποιήσει τὰ ὄντα, εἴπερ μέλλει ὀνόματα εἶναι; “And if the primary names are indeed names, they must make the things that are as clear as possible to us. But how can they do this when they aren’t based on other names?” (422d11–e1).

III.2. Platonic Solution: Mimēsis

Plato’s answer is, “By imitation”. Sensory signals through which we are trying to communicate must be made to resemble the intended object of communication (cf. 422e1–423a6, esp. μιμούμενοι αὐτὴν τὴν φύσιν τοῦ πράγματος: “in imitation of the very nature of the thing”). If, for example, we did not possess voice and were trying to convey something through gestures, gestures themselves would have to be made to resemble objects we would be trying to communicate (ὡς ὁμοιότατ’ ἂν τὰ ἡμέτερα αὐτῶν σώματα καὶ σχήματα ἐποιοῦμεν ἐκείνοις: “we’d make our bodies and our gestures as much like theirs as possible” 423a5–6, and cf. 423a8–b2). In what follows, Plato applies this principle to sounds, developing it futher until a definition of naming is reached: ὀνομαστικὸς […] αὐτὸ τοῦτο μιμεῖσθαι δύναιτο ἑκάστου, τὴν οὐσίαν, γράμμασί τε καὶ συλλαβαῖς “the namer [would be] someone who is able to imitate in letters and syllables this being or essence that each thing has” (423e7–8 & 424a6) (οὐσία need not mean here anything more than ‘What thing is’, as opposed to its properties). Thus, in order to be able to communicate, phonetic signals have to resemble by themselves the very things they are trying to communicate. Plato gives a display of how this happens in the analysis of certain primary words: ρ imitates movement, because tongue moves most when pronouncing this sound (426c1–2 & 426d3–e6); ι imitates small, subtle and penetrating things (426e6–427a2), sibilant sounds like φ, ψ, σ and ζ imitate blowing or hard breathing (427a2–8). The compression and stopping of the tongue in δ and τ imitates binding or shackling and rest (427a8–b2). λ imitates gliding and smooth things, because the tongue glides in pronouncing it (427b2–5), whereas the stopping of the λ-gliding tongue with γ imitates stickiness and glueyness (427b5–7).

This account has often been ridiculed for the crudeness of its model and the obvious limits it imposes on the mimetic function of names. In the view of Gorgias’ critique of communication, however, Plato may be more perceptive here than is usually maintained. In the cases that involve the above-mentioned sounds – the medium of communication – they not only imitate certain features or objects of the external world. They also carry with themselves their interpretation, since their imitation proceeds on the basis of the similarity of the anatomical processes involved in their pronunciation, and the imitated object. Thus the recipient has only to pronounce the sound in order to establish its fundamental meaning.

We need not dwell upon the details of elaboration of this stage of the theory. A couple of points, however, have to be made.

(1) Socrates seems to be constantly oscillating between the discussion of the atemporal, ‘synchronic’ conditions for successful communication, and the account of the historical genesis of language. The most obvious indication of this ambiguity is the way he checks himself at the end of the long passage describing how language is to be (re?)constructed upon mimetic principles (424c6–425b4):

καὶ πάλιν ἐκ τῶν ὀνομάτων καὶ ῥημάτων μέγα ἤδη τι καὶ καλὸν καὶ ὅλον συστήσομεν, ὥσπερ ἐκεῖ τὸ ζῷον τῇ γραφικῇ, ἐνταῦθα τὸν λόγον τῇ ὀνομαστικῇ ἢ ῥητορικῇ ἢ ἥτις ἐστὶν ἡ τέχνη. μᾶλλον δὲ οὐχ ἡμεῖς, ἀλλὰ λέγων ἐξηνέχθην. συνέθεσαν μὲν γὰρ οὕτως ᾗπερ σύγκειται οἱ παλαιοί· ἡμᾶς δὲ δεῖ, εἴπερ τεχνικῶς ἐπιστησόμεθα σκοπεῖσθαι αὐτὰ πάντα, οὕτω διελομένους, εἴτε κατὰ τρόπον τά τε πρῶτα ὀνόματα κεῖται καὶ τὰ ὕστερα εἴτε μή, οὕτω θεᾶσθαι
“From names and verbs, in turn, we shall finally construct something important, beautiful, and whole. And just as the painter painted an animal, so – by means of the craft of naming or rhetoric or whatever it is – we shall construct sentences. Of course, I don’t really mean we ourselves – I was carried away by the discussion. It was the ancients who combined things in this way. Our job – if indeed we are to examine all these things with scientific knowledge—is to divide where they put together, so as to see whether or not both the primary and derivative names are given in accord with nature” (425a1–b3).

From this passage, it would seem that the same criteria of ‘correctness’, of successful communication apply both to the present acts of communication, and to the historical constitution of language. In this context, one is inclined to read the passage about gesture-language as a tentative reconstruction of the very first act of communication that established the first intersubjective contact and made further communication possible:

“If we hadn’t a voice or a tongue, and wanted to express things to one another, wouldn’t we try to make signs by moving our hands, head, and the rest of our body, just as dumb people do at present (ὥσπερ νῦν οἱ ἐνεοί)?20 [...] So, if we wanted to express something light in weight or above us, I think we’d raise our hand towards the sky in imitation of the very nature of the thing. And if we wanted to express something heavy or below us, we’d move our hand towards the earth” (422e2–423a4).

Even regardless of the exegesis of gesture-passage, however, the historical dimension is firmly there through the postulating of νομοθέται, the primordial namegivers.

(2) Later in the dialogue, Socrates remarks that, in order to be able to imitate, the elements that constitute the medium of the imitation have to possess a natural similarity to the object of imitation, namely, elements of the first names must possess a natural similarity to the things named: εἴπερ ἔσται τὸ ὄνομα ὅμοιον τῷ πράγματι, ἀναγκαῖον πεφυκέναι τὰ στοιχεῖα ὅμοια τοῖς πράγμασιν, ἐξ ὧν τὰ πρῶτα ὀνόματά τις συνθήσει: “But if a name is indeed to be like a thing, mustn’t the letters or elements out of which primary names are composed be naturally like things?”, just as pigments are naturally (φύσει) similar to the objects depicted (434a3–b2). This remark summarises the belief that is implicit in the practice of the passage dealing with meanings of individual sounds: if the medium of imitation is a sensory phenomenon like sound, it can only imitate things that are essentially like it, i.e., only those things that can be objects of sense-perception. Of course, a certain moment of universalisation is allowed, e.g., a rather specific way of moving the tongue in pronunciation of ρ can mimetically signify ‘movement in general’ (τὸ ῥῶ ἔμοιγε φαίνεται ὥσπερ ὄργανον εἶναι πάσης τῆς κινήσεως “‘r’ seems to me to be a tool for copying every sort of motion” 426c1–2), but it would be absurd to assume natural mimetic similarity between any sound, or any combination of sounds, and a non-perceptible quality, e.g., ‘justice’. Apparently Socrates has in mind the restrictions that the mimetic theory of communication imposes upon language’s scope of representation, when, just before the passage where the meanings of sounds are discussed starts, he says: ἃ μὲν τοίνυν ἐγὼ ᾔσθημαι περὶ τῶν πρώτων ὀνομάτων πάνυ μοι δοκεῖ ὑβριστικὰ εἶναι καὶ γελοῖα: “my impressions about primary names seem to me to be entirely outrageous and absurd” (426b5–6); also, cf. previous remark: γελοῖα μὲν οἶμαι φανεῖσθαι [...] γράμμασι καὶ συλλαβαῖς τὰ πράγματα μεμιμημένα κατάδηλα γιγνόμενα· ὅμως δὲ ἀνάγκη: “Perhaps it will seem absurd [...] to think that things become clear by being imitated in letters and syllables, but it is absolutely unavoidable” (425d1–3). It is, however, ‘necessary’, insofar as all the communication has to use external means, and the only contact with the external world is through senses. The deficiency of the (purely) mimetic model, however, is already implicitly indicated at this stage.

III.3. Socrates Reverses Course: Convention, rather than Imitation

The mimetic project is essentially incomplete (and uncompletable). Immediately after the section analysing the sounds, a major reversal of the dialogue takes place, where the role of the interlocutor passes from Hermogenes to Cratylus, and where Socrates turns on his tracks in order to examine critically that which has been achieved (cf. 428d5–8: δεῖ δή, ὡς ἔοικε, θαμὰ μεταστρέφεσθαι ἐπὶ τὰ προειρημένα, καὶ πειρᾶσθαι, τὸ ἐκείνου τοῦ ποιητοῦ, βλέπειν «ἅμα πρόσσω καὶ ὀπίσσω»: “Therefore, I think we have to turn back frequently to what we’ve already said, in order to test it by looking at it ‘backwards and forwards simultaneously’, as the aforementioned poet puts it”).

The details of Socrates’ critique, again, do not need to concern us here. Socrates’ argumentation in the Cratylus-section is somewhat double-edged: on the one hand, he revises, and brings qualifications into, the previously proposed mimetic theory of language, yet, on the other hand, he combats Cratylus’ philosophical tenets: that all names, if they are such, represent their designates equally well and absolutely precisely, and that the only and the best way of discovering the truth about reality is ‘through names’, i.e., through the analysis of the picture of reality that language presents. Accordingly, the argument of the Cratylus-section breaks into two halves (428e5–435d1 and 435d1–439b8). It must be noted, however, that the subsequent critique does not refute the mimetic principle as a foundation of a communication theory, but rather qualifies and modifies the previous conclusions.

(1) Among the arguments against Cratylus’ ‘absolute representation’ in this section there are two points that interest us in particular. First of them argues by drawing further conclusions from the mimetic principle: a name cannot be an absolute representation of a thing, since it would contradict the imitation principle that was put as a foundation of the mimetic model of language – viz., that names imitate things. For if the names (or, for that matter, images in general) were capable of perfectly resembling their object in all respects, it would result in reduplication of reality. Therefore, in order to maintain the mimetic theory of names, one has to admit that names are of necessity imperfect imitations, varying in degree of resemblance and succeeding in representing their object insofar as they preserve its ‘pattern’ (431a9–433b5, cf. esp. 432e5–7: μηδὲν ἧττον ὀνομάζεσθαι τὸ πρᾶγμα καὶ λέγεσθαι, ἕως ἂν ὁ τύπος ἐνῇ τοῦ πράγματος περὶ οὗ ἂν ὁ λόγος ᾖ: “Things are still named and described when this happens, provided the phrases include the pattern of the things they’re about”). Thus the critique does not subvert the earlier positive construction, but rather completes it and draws its further consequences. Furthermore, it may be noted that imperfection inherent in imitation appears not only as a necessary evil – it transpires to be teleologically necessary in order that communication might be possible at all, insofar as only the mimetic model can enable communication. The difference between the imaged and the image becomes that which constitutes the condition of (the possibility of) language:

τοῦδε ποιοῦ τινος, καὶ συμπάσης εἰκόνος μὴ οὐχ αὕτη <ᾖ> ἡ ὀρθότης, ἀλλὰ τὸ ἐναντίον οὐδὲ τὸ παράπαν δέῃ πάντα ἀποδοῦναι οἷόν ἐστιν ὃ εἰκάζει, εἰ μέλλει εἰκὼν εἶναι
“But this isn’t the sort of correctness that belongs to things with sensory qualities, such as images in general. Indeed, the opposite is true of them – an image cannot remain an image if it presents all the details of what it represents” (432b1–4).

In the second argument that runs from 434a2 to 435d1, Socrates, by a single ingenious move, introduces ‘convention’ (συνθήκη; equivalent to ἔθος ‘habit’, cf. 434e5–8) as a necessary supplement of the mimetic principle in communication: if two dialect forms of the same word that possess different endings are understood, and if even the same word may contain a sound that, in the mimetic theory, contradicts its meaning, and still be understood, language must be able to signify ‘from the dissimilar’, ἀπὸ τοῦ ἀνομοίου (435a5), as well as from similar, and this cannot be achieved unless convention if introduced (if mimetic representation was the only model, such words could not be understood at all). Socrates gives the following definition of ‘convention’: ὅταν τοῦτο φθέγγωμαι, διανοοῦμαι ἐκεῖνο, σὺ δὲ γιγνώσκεις ὅτι ἐκεῖνο διανοοῦμαι:21 “when I utter this name and mean [something] by it, you know that this is what I mean?” (434e6–8).

It has been assumed that, at this point, συνθήκη-theory replaces the previously held mimetic theory of names (as in, e.g., Ewegen, 2014, pp. 170–171). Indeed, some of the expressions used by Socrates immediately after ‘convention’ and which have been introduced afterwards may be taken to mean that: οὐκ ἂν καλῶς ἐτι ἔχοι λέγειν τὴν ὁμοιότητα δήλωμα εἶναι, ἀλλὰ τὸ ἔθος· ἐκεῖνο γὰρ, ὡς ἔοικε, καὶ ὁμοίῳ καὶ ἀνομοίῳ δηλοῖ: “you must say that expressing something isn’t a matter of likeness but of usage, since usage, it seems, enables both like and unlike names to express things” (435b1–3), as well as γίγνεται ἡ ὀρθότης τοῦ ὀνόματος συνθήκη, ἐπειδή γε δηλοῖ καὶ τὰ ὅμοια καὶ τὰ ἀνόμοια γράμματα, ἔθους τε καὶ συνθήκης τυχόντα: “the correctness of names has become a matter of convention, for isn’t it the chancing of usage and convention that makes both like and unlike letters express things?” (435a8–9).

A more careful analysis, however, reveals that the statements quoted above are meant only to explain how communication ‘by dissimilar’ may be possible at all, and they do not extend beyond the cases where such communication takes place. The ending of the second statement (specifically, δηλοῖ καὶ τὰ ὅμοια καὶ τὰ ἀνόμοια γράμματα, ἔθους τε καὶ συνθήκης τυχόντα (435a9–10)), in fact, confirms such reading, and so does the rest of the Socrates’ passage (cf. ἀναγκαῖόν που καὶ συνθήκην τι καὶ ἔθος συμβάλλεσθαι πρὸς δήλωσιν ὧν διανοούμενοι λέγομεν: “both convention and usage must contribute something to expressing what we mean when we speak” (435b5–6), συνθήκην κῦρος ἔχειν: “convention to have some control” (435c1), ἀναγκαῖον δὲ ᾖ καὶ τῷ φορτικῷ τούτῳ προσχρῆσθαι, τῇ συνθήκῃ: “have to make use of this worthless thing, convention” (435c4–6), and cf. κατὰ τὸ δυνατόν: “as far as possible” (at 435a7). It transpires that ‘convention’ is the ‘second-best’ factor in communication, and it does not abolish – but rather supplements and perfects – the mimetic principle (at 435b6–c2 numbers are quoted as an example of entity, communication by imitation of which is practically impossible).22

(2) Immediately afterwards, Socrates addresses the second major assumption by Cratylus: that the function of the name is to ‘teach’, i.e., to reveal information about things, and whoever knows names knows things (435d4–6). Besides, investigation of names (exemplified by Socrates’ etymological activities earlier in the dialogue) is the only and the best way to acquire knowledge of reality (435e6–436a8). Socrates’ most powerful argument against Cratylus’ position is that νομοθέται, in order to establish names knowingly, had to know the reality before there were any names (Versio A 437e1–438b7 & Versio B 438a3–b3). This quasi-historical argument is rendered into the sychronic mode immediately afterwards: even provided (as Cratylus suggests in 438b8–c2) that a “force higher than human” (μείζω τινὰ δύναμιν ἢ ἀνθρωπείαν) posited first names, i.e., even if one does not need to account for its knowledge of reality before names were made, given the contradiction between the pictures of reality that the names give, an extra-linguistic point of reference, a certain entscheidende Instanz (cf. τίνι ἔτι διακρινοῦμεν, ἢ ἐπὶ τί ἐλθόντες;) is needed in order to decide which of the conflicting pictures is correct: ἄττα ζητητέα πλὴν ὀνομάτων, ἃ ἡμῖν ἐμφανιεῖ ἄνευ ὀνομάτων ὁπότερα τούτων ἐστὶ τἀληθῆ, δείξαντα δῆλον ὅτι τὴν ἀλήθειαν τῶν ὄντων: “we’ll have to look for something other than names, something that will make plain to us without using names which of these two kinds of names are the true ones – that is to say, the ones that express the truth about the things that are” (438c4–d8). Therefore, it must be possible to learn reality without names (438e1–2).

III.4. Image-Theory of Words and its Gorgianic Antecedents

Immediately below, Socrates recasts the latter conclusion in terms of the image-theory of words. It is better to learn reality from itself, and from reality – the image, whether it imitates reality properly, than vice versa – to learn from image the reality that it imitates and try to establish whether it imitates it well (439a6–b2). This attitude was neatly captured in Robert Wardy’s description: “One might [...] suppose that words deputise for their objects, and that, since no substitute matches the original, we should never be satisfied with the pale imitation of reality which is the best logos can ever achieve, but should rather hold out for the real thing: if the more accurate a copy is the better it is, then by most stringent standards things must ‘represent’ themselves“ (Wardy, 1996, p. 21). The fact, however, that the subject of Wardy’s remark is Gorgias’ Περὶ τοῦ μὴ ὄντος, is already sufficient to indicate the proximity between Plato’s position here and Gorgias’ critique.

This proximity is especially transparent in Socrates’ brief remark which postulates an ontological (and semantic) gap between things and words:

διὰ τίνος ἄλλου οὖν ἔτι προσδοκᾷς ἂν [τὰ ὄντα] μαθεῖν; ἆρα δι’ ἄλλου του ἢ οὗπερ εἰκός τε καὶ δικαιότατον, δι’ ἀλλήλων γε, εἴ πῃ συγγενῆ ἐστιν, καὶ αὐτὰ δι’ αὑτῶν; τὸ γάρ που ἕτερον ἐκείνων καὶ ἀλλοῖον ἕτερον ἄν τι καὶ ἀλλοῖον σημαίνοι ἀλλ’ οὐκ ἐκεῖνα
“How else would you expect to learn about them? How else than in the most legitimate and natural way, namely, learning them through one another, if they are somehow akin, and through themselves? For something different, something that was other than they, wouldn’t signify them, but something different, something other” (438e5–9).

One has only to compare the corresponding sections of the Περὶ τοῦ μὴ ὄντος from MXG and Sextus in order to realise the affinity:

ὃ οὖν τις μὴ ἐννοεῖ, πῶς αὐτὸ παρ ἄλλου λόγῳ ἢ σημείῳ τινὶ ἑτέρῳ τοῦ πράγματος ἐννοήσει
“So what someone does not think, how can he understand it from someone else by means of a speech, or by means of some sign different from the thing” (980b3–5),

as well as:

ᾧ γὰρ μηνύομεν, ἐστι λόγος, λόγος δὲ οὐκ ἐστι τὰ ὑποκείμενα καὶ ὄντα· οὐκ ἄρα τὰ ὄντα μηνύομεν τοῖς πέλας ἀλλὰ λόγον, ὃς ἕτερός ἐστι τῶν ὑποκειμένων
“For what we indicate by is speech, but the things that exist and that are are not speech. So it is not the things that are that we indicate to other people, but rather speech, which is different from the things that exist” (§ 84).

Thus Plato seems to retain the fundamental ontological difference between words and things; for him, this difference is expressed in the ontological status of words as ‘images’ (εἰκόνες) of things, and it does not preclude the communication (in fact, vice versa – it is mimetic theory that re-establishes the possibility of communication in the wake of Gorgias’ critique). The difference in the ontological status between words and things is used as an argument against Cratylus for the epistemological priority of the ‘reality itself’ over its verbal images. Contrary to Gorgias’ train of argument, the ontological difference between words and things not only does not preclude the possibility of communication, but in fact makes it possible.

As regards the characteristic feature of Gorgias’ theory of sense perception that underlies his critique of communication (and, insofar as we can see from what survives of Part 2 of the Περὶ τοῦ μὴ ὄντος, his critique of knowledge, as well – cf. 980a12-15 MXG & § 81 Sextus), viz., radical mutual separation of the senses, we see it addressed in the Theaetetus:

ΣΩ. [...] Ἀπόκρισις ποτέρα ὀρθοτέρα, ᾧ ὁρῶμεν τοῦτο εἶναι ὀφθαλμούς, δι’ οὗ ὁρῶμεν, καὶ ᾧ ἀκούομεν ὦτα, ἢ δι’ οὗ ἀκούομεν; – ΘΕΑΙ. Δι’ ὧν ἕκαστα αἰσθανόμεθα [...] μᾶλλον ἢ οἷς. – ΣΩ. Δεινὸν γάρ που, ὦ παῖ, εἰ πολλαί τινες ἐν ἡμῖν ὥσπερ ἐν δουρείοις ἵπποις αἰσθήσεις ἐγκάθηνται, ἀλλὰ μὴ εἰς μίαν τινὰ ἰδέαν, εἴτε ψυχὴν εἴτε ὅτι δεῖ καλεῖν, πάντα ταῦτα συντείνει, ᾗ διὰ τούτων οἷον ὀργάνων αἰσθανόμεθα ὅσα αἰσθητά
“Socrates. Is it more correct to say that the eyes are that with which we see, or that through which we see? Do we hear with the ears or through the ears? – Theaetetus: Well, I should think, Socrates, that it is ‘through which’ we perceive in each case, rather than ‘with which’. – Socrates: Yes, my son. It would be a very strange thing, I must say, if there were a number of perceptions sitting inside us as if we were Wooden Horses, and there were not some single form, soul or whatever one ought to call it, to which all these converge – something with which, through those things, as if they were instruments, we perceive all that is perceptible” (184c5–d5; here and infra, M. J. Levett’s translation, revised by Myles Burnyeat, is used for the Theaetetus),

as well as:

ἦ καὶ ἐθελήσεις ὁμολογεῖν ἃ δι’ ἑτέρας δυνάμεως αἰσθάνῃ, ἀδύνατον εἶναι δι’ ἄλλης ταῦτ’ αἰσθέσθαι, οἷον ἃ δι’ ἀκοῆς, δι’ ὄψεως, ἢ ἃ δι’ ὄψεως, δι’ ἀκοῆς;
“And are you also willing to admit that what you perceive through one power, you can’t perceive through another? For instance, what you perceive through hearing, you couldn’t perceive through sight, and similarly what you perceive through sight you couldn’t perceive through hearing?” (184e8–185a2 & ff).

Here, mutual exclusivity of the senses does not lead to the impossibility of communication and knowledge, but, rather characteristically, contributes towards the proof that sense perception is separate from, and subordinate to, the kind of knowledge that is performed by soul.

III.5. Some Implications of Plato’s Theory of Communication: A Corollary

Just to add one more remark in the way of corollary. After Socrates finishes criticising Cratylus’ reliance on names, he introduces an unmistakably new point (cf. ἔτι τοίνυν τόδε σκεψώμεθα 439b10), and sets out on his last attack – against the conception of reality as being in the state of universal flux. While he shifts from the problematic of names to the problems of the nature of reality, he makes, en passant, a probably most straightforward statement about the character of language that we get in the Cratylus:

ἔτι τοίνυν τόδε σκεψώμεθα, ὅπως μὴ ἡμᾶς τὰ πολλὰ ταῦτα ὀνόματα ἐς ταὐτὸν τείνοντα ἐξαπατᾷ, εἰ τῷ ὄντι μὲν οἱ θέμενοι αὐτὰ διανοηθέντες γε ἔθεντο ὡς ἰόντων ἁπάντων ἀεὶ καὶ ῥεόντων – φαίνονται γὰρ ἔμοιγε αὐτοὶ οὕτω διανοηθῆναι – τὸ δ’, εἰ ἔτυχεν, οὐχ οὕτως ἔχει
“Still, let’s investigate one further issue so as to avoid being deceived by the fact that so many of these names seem to lean in the same direction – as we will be if, as seems to me to be the case, the name-givers really did give them in the belief that everything is always moving and flowing, and as it happens things aren’t really that way at all” (439b10–c4).

Albeit cautiously, Socrates endorses the view that the picture of reality given by the language is one of the constant flux and change.23

In the view of the passage just quoted, an attractive extrapolation of Plato’s theory of communication presents itself. If it can be maintained that the mimetic principle functions both as an explanation of the possibility of communication ‘synchronically’, and as a theory of the historical genesis of language (v. supra.), and given the inevitably sensory character of vocal imitation (as Plato reminds us, the media of imitation must be naturally akin to its object, 434a3–b2), it may be conjectured that the initial stage of language – to which πρῶτα ὀνόματα belong – can only imitate the sensory phenomena. One could find some support for this conjecture in the extremely empirical character of the ‘primary words’ discussed by Socrates (cf. 426c1–427d1). Besides, semantic characteristics of the individual signs are grounded in the physiological characteristics of their articulation. Since articulation is a movement, movement is given semantic priority in the mimetic system of language – as στᾶσις is explained as ‘negation of movement’, ἀπόφασις τοῦ ἰέναι (426d1–3). The sensory character of the primitive stage of language cannot but lead to ‘fluxy’ representation of reality, and thus it corresponds to the adherence to the flux-theory by all the early thinkers.24

IV. Conclusion

Gorgias’ tenets are not an explicit, let alone the foremost object of Plato’s preoccupations in the Cratylus. Rather, Gorgias’ critique of communication constitutes an intellectual background for Plato’s speculative constructions. Nevertheless, verbal correspondences and our independent knowledge of Plato’s awareness of the problems raised by Gorgias’ Περὶ τοῦ μὴ ὄντος point towards the conclusion that Gorgias’ critique was, for Plato, a significant point of departure: he constructs his theory of names in such a way as to satisfy Gorgias’ criteria (proposed negatively, as a critique) for the possibility of communication. Simultaneously, he preserves (though refashioning it in terms of the conceptual structures of his own philosophy) probably the most important of Gorgias’ insights – a specific ontological status of language, the fundamental difference between language and extralinguistic reality, as well as its implications for epistemology and the theory of communication.

References

Bett, R., 2020. Gorgias’ Περὶ τοῦ μὴ ὄντος and Its Relation to Skepticism. International Journal for the Study of Skepticism, 10(3-4), pp. 187–208. https://doi.org/10.1163/22105700-bja10010

Bluck, R. S., 1961. Plato’s Meno. Ed. with introd. and comm. by R. S. Bluck. Cambridge: University Press.

Bonazzi, M., 2023. The Turn to Language. In: J. Billings and C. Moore, eds. The Cambridge Companion to the Sophists. Cambridge: University Press, pp. 179–199. https://doi.org/10.1017/9781108859639.007

Buchheim, T., 2012. Gorgias von Leontinoi. Reden, Fragmente und Testimonien. Übersetzt, mit einer Einleitung und Anmerkungen herausgegeben von T. Buchheim. 2., verbesserte Auflage. Hamburg: Felix Meiner Verlag. https://doi.org/10.28937/978-3-7873-3223-6

Di Iulio, E., 2021. What Is Gorgias’ ‘Not Being’? A brief journey through the Treatise, the Apology of Palamedes and the Encomium of Helen. Archai, (31), pp. 1–23. https://doi.org/10.14195/1984-249x_31_26

Di Iulio, E., 2023. Gorgias’s Thought. An Epistemological Reading. London & New York: Routledge. https://doi.org/10.4324/9781003277620

Ewegen, S. M., 2014. Plato’s Cratylus. The Comedy of Language. Bloomington and Indianapolis: Indiana University Press. https://doi.org/10.2979/6971.0

Kerferd, G. B., 1981. The interpretation of Gorgias’ treatise the Περὶ τοῦ μὴ ὄντος ἢ Περὶ φύσεως. Deucalion, 36, pp. 319–327.

Kerferd, G. B., 1984. Meaning and reference. Gorgias and the relation between language and reality. In: G. B. Kerferd. The Sophistic Movement. Athens: Athenian Library of Philosophy, pp. 215–222.

Kerferd, G. B., 1985. Gorgias and Empedocles. Siculorum Gymnasium, 38, pp. 595–605.

Kolby, K., 2023. Discovering Nothing to Create Anything: Gorgias’s On the Nonexistant and the World-Building Power of Logos. Scientia et Humanitas: A Journal of Student Research, 13, pp. 35–46.

Laks, A. and Most, G. W., 2016. Early Greek Philosophy. Vol. VIII. Sophists. Part 1. Cambridge (Mass.)-London: Harvard University Press.

Mazzara, G., 1982. Gorgia. Ontologo e metafisico. Palermo: ila palma.

Mazzara, G., 1983. Gorgia. Origine e struttura materiale della parola. L’Antiquité Classique, 52, pp. 130–140. https://doi.org/10.3406/antiq.1983.2087

Mourelatos, A. P. D., 1985. Gorgias on the function of language. Siculorum Gymnasium, 38, pp. 607–638.

Newiger, H.-J., 1973. Untersuchungen zu Gorgias’ Schrift Über das Nichtseiende. Berlin-New York: Walter de Gruyter. https://doi.org/10.1515/9783110834673

Palmer, J. A., 1999. Plato’s Reception of Parmenides. Oxford: Clarendon Press. https://doi.org/10.1093/oso/9780198238003.001.0001

Rodriguez, E., 2019. Untying the Gorgianic ‘Not’: Argumentative Structure in On Not-Being. Classical Quarterly, 69(1), pp. 87–106. https://doi.org/10.1017/s0009838819000648

Schiappa, E., 1997. Interpreting Gorgias’s ‘Being’ in “On Not-Being or On Nature”. Philosophy & Rhetoric, 30(1), pp. 13–30.

Sedley, D. N., 1992. Sextus Empiricus and the atomist criterion of truth. Elenchos, 13, pp. 21–56.

Spangenberg, P., 2021. El paradigma de la sensibilidad y la fragmentación en el PTMO de Gorgias. Archai, (31), pp. 1–25. https://doi.org/10.14195/1984-249X_31_29

Volf, M., 2021. Gorgias’ revising of ancient epistemology: on non-being by Gorgias and its paraphrases. Archai, (31), pp. 1–25. https://doi.org/10.14195/1984-249x_31_27

Wardy, R., 1996. The Birth of Rhetoric. Gorgias, Plato and their successors. London & New York: Routledge. https://doi.org/10.4324/9780203981962