6_Sun

Abstract. This article reinterprets the theory of mimesis in the Cratylus, exploring how language functions as an image that imitates its objects. Contrary to prevailing studies contending that Socrates fails to reconcile naturalism and conventionalism, this article argues that Socrates proposes a form of naturalism that acknowledges the role of convention. This naturalism reveals that human language has a dual nature by demonstrating the relation between images and originals. Through instrumentalism and sound-symbolism, Socrates envisions language as an ideal instrument for imitating the Forms of things, whereas the difference between images and originals leads to inevitable falsehood in the establishment and use of ordinary language. The real purpose of the theory of mimesis is to defend the possibility of knowledge and language by opposing the sophists’ doctrine of flux.
Keywords: Plato, Cratylus, the theory of mimesis, correctness of speech, Conventionalism

Platono mimezės teorija dialoge Kratilas: nuo idealios kalbos prie kasdienės kalbos

Santrauka. Šis straipsnis naujai interpretuoja mimezės teoriją, pateiktą Platono dialoge Kratilas, tirdamas, kaip kalba veikia kaip atvaizdas, mėgdžiojantis savo objektus. Skirtingai nei didžioji dauguma tyrimų, teigiančių, kad Sokratui nepavyksta suderinti natūralizmo ir konvencionalizmo, šis straipsnis teigia, kad Sokrato pasiūlyta natūralizmo forma pripažįsta susitarimo vaidmenį. Natūralizmas, parodydamas santykį tarp atvaizdų ir originalų, atskleidžia, kad žmonių kalba yra dvilypio pobūdžio. Per instrumentalizmą ir garso simbolizmą Sokratas įsivaizduoja kalbą kaip tobulą instrumentą daiktų formoms imituoti, o skirtumas tarp atvaizdų ir originalų neišvengiamai veda prie klaidų formuojant ir vartojant kasdienę kalbą. Tikrasis mimezės teorijos tikslas yra apginti pažinimo ir kalbos galimybę, oponuojant sofistų tėkmės doktrinai.
Pagrindiniai žodžiai: Platonas, Kratilas, mimezė kalbos taisyklingumas, konvencionalizmas

Acknowledgement. This article was supported by the Major Program of the National Social Science Fund of China, “Research on Mediterranean Civilization and the Origins of Ancient Greek Philosophy” (23&ZD239).

Received: 02/06/2024. Accepted: 09/09/2024
Copyright © Liangxin Sun, 2024. Published by Vilnius University Press.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (CC BY), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Introduction

Plato’s Cratylus, a dialogue characterized by its intricacy, delves into the correctness of language. The interlocutors engage in debates concerning naturalism versus conventionalism, and the theory of Forms versus the doctrine of flux, without reaching a consensus. The ambiguity surrounding Plato’s view on language has led to three main interpretations among modern scholars: firstly, Plato asserts that correct names and language possess descriptive content or properties appropriate to the nature of their objects, with his attack on conventionalism dominating the dialogue (Demos 1964; Weingartner 1970); secondly, Plato rejects naturalism as impractical, and reluctantly concedes that convention plays a pivotal role in determining correctness (Robinson 1956; Schofield 1982); thirdly, Plato endeavors to reconcile naturalism and conventionalism, suggesting their validity within distinct domains (Kretzmann 1971).

Socrates’ theory of mimesis, detailed in 421d–426a and 430b–433b, constitutes a crucial aspect of the Cratylus. Both Hermogenes and Cratylus are eventually convinced that it offers a better way to explain the correctness of names (426b; 430a; 433e). Nonetheless, this theory posits a mimetic relation between names and objects based on similarity, conflicting with modern linguistic theories that view the relation between the signifiant and the signifié as arbitrary and conventional (Saussure 1996: 67–69). The disparities between ancient and modern theories have led many scholars to dismiss the theory of mimesis. Smith (2008: 147) underscores that the mimetic relation constitutes neither a necessary nor a sufficient condition for reference. Williams (1982: 91–92) argues that mimesis cannot explain how language functions, while attributing a quasi-magical power to language.

This article argues that, contrary to the interpretations of many contemporary scholars who highlight the supposed contradiction between naturalism and conventionalism, Socrates’ naturalism does not conflict with his recognition of the role of convention in ordinary language. The passages (427d–430a, 434c–435d), often seen as Socrates’ criticism of naturalism, are actually an integral step in his argument for the natural correctness of language. My interpretation assumes that the Cratylus, as a dialectic dialogue, constitutes a cohesive whole in which Socrates’ arguments are not contradictory but dialectic. This perspective becomes clearer when distinguishing the genuine discourse and agonistic display (Barney 1998: 75) of the rivals’ views. Socrates’ naturalism, centered on the theory of mimesis, addresses both the correctness and falsehood of language through the similarity and difference between the original and the image. To explore the different roles of nature and convention in the functioning of language, Socrates presents language in two dimensions: ideal and ordinary. In this view, the theory of mimesis aims to provide a normative theory for how language should be established and used, reinforcing the authority of philosophers over the relativism and sophistry of the sophists.

The theory of language as mimesis

At the beginning of the Cratylus, Hermogenes recounts a debate between naturalism and conventionalism to Socrates. Cratylus advocates for naturalism, arguing that (383a–b):

(C1) For each thing, there is a natural correctness of its name, a kind of correctness that is the same for all, both Greeks and barbarians.

(C2) Correct names are not simply agreed upon by people who assign sounds to things.

(C3) It is impossible to impose a name incorrectly; a false name is nonsense and merely a string of noises (430a).

In contrast, Hermogenes supports conventionalism, contending that there is no correctness of names other than convention and agreement. Any name imposed on a thing by people is correct if it aligns with laws and customs (384d–e). Socrates adds that the name-maker may reside with an individual or the state (385a). Therefore, conventionalism takes two potential versions:

(H1) An individual can arbitrarily impose any name on a thing, even contrary to communal consensus, such as calling what others recognize as ‘human’ a ‘horse’.

(H2) A community can collectively impose any name on a thing, following its laws and customs.

Both naturalism and conventionalism seek to find a correct way to describe the relation between language and the world. Naturalism, as stated in (C1), posits that there is a natural suitability or relation between names and objects. Conversely, conventionalism views this relation as contingent, with no objective standard governing naming acts except for laws and customs. (H1) is typically ascribed to Hermogenes, linking his conventionalism to the challenge of autonomous idiolects (Weingartner 1970: 7). In this view, all naming acts and decisions are equally correct, even if they lead to private names within an idiolect. (H2) extends (H1) by considering public conventions or social consensus as the normative basis for naming acts. However, as Barney (1997: 155) notes, Hermogenes’ concept of convention is limited to the position that it contains merely a decision giving rise to a custom. Since public conventions arise from individual decisions, they are either made by an authority within the community, or function as an aggregate composed of individual conventions. Consequently, both (H1) and (H2) presuppose that language is a human construct imposed on reality, with the relation between names and objects entirely created by humans rather than discovered. Hermogenes’ position equates correctness with factualness, making falsehood simply a matter of contradicting common opinions or ordinary sense. The correctness of names adopts a ‘redundancy conception’ (Ademollo 2011: 3–4). Once something has been successfully named ‘N’ in ordinary language, ‘N’ is the correct name for it.

Cratylus’ oracular view, as Hermogenes complains, also fails to provide an objective standard for assessing the correctness of names. Cratylus supposes that the correctness of names is independent of users’ opinions, but he does not explicitly explain how a natural relation between names and objects could be established. His commitment to Heraclitean doctrine of flux exacerbates the difficulty of this task (440e). According to (C3), what we say is either correct or mere nonsense, making it impossible to speak falsehoods meaningfully. Confronted with this lack of normativity, Socrates initiates a dialogue with Hermogenes, presenting a preliminary argument on the true and false statement in 385b–d,1 outlined as follows:

(L1) There are true and false speeches. True statements speak of the things as they are, while false statements speak of things as they are not. Both are intelligible and meaningful.

(L3) Names are the smallest elements of a statement; if a statement about things is true, then the names used within it are also true.

(L4) Names can be true or false. True names speak of things as they are, while false names speak of things as they are not. Both are intelligible and meaningful.

Socrates does not draw a strict distinction between correctness and truth; he assumes that a true name must be a correct name. (L1) implies a correspondence theory of truth, proposing that language should have an ‘external correctness’ – a correctness assessed based on the relation between names and objects rather than among names. The phrase ‘things as they are (τὰ ὄντα ὡς ἔστιν)’ is later substituted with ‘the being of things (οὐσία τῶν ὄντων)’ in 385e. In Greek, the participle ‘οὐσία’ is the nominalization of the copula ‘ἔστιν’, capable of representing both the properties predicated in the form ‘– is F’ and the essence of things (Ademollo 2011: 77). (L2) and (L3) elucidate the relation between the whole of a statement and its elements. A statement being true necessitates that all its elements are true. (L4) counters Hermogenes’ and Cratylus’ claims that false names are impossible. This argument suggests that a coherent account of the correctness of names must demonstrate the possibility of false names. Understanding the relation between correctness and falsehood is a key objective that Socrates’ naturalism aims to achieve.

Since correct names speak of things as they are, they inherently serve a descriptive function that expresses the being of things they name.2 At the end of the etymological section (421d–422c), Socrates introduces the theory of mimesis, further elaborating this argument by asserting that a name is fundamentally an imitation or image representing the object it names. This theory was initially proposed to resolve the paradox of infinite regress in etymological analysis. If one endlessly inquires through which words names are spoken, they would eventually have to give up. Therefore, certain primary names must be identified as the endpoint of etymological analysis. These names are not derived from other names and indicate things to us in a way different from secondary names. Socrates explains that the name-maker is able to imitate the essence of each thing with letters and syllables of primary names, thus he can reveal what each thing is (423e). The relation between primary names and objects is analogous to that between paintings and originals. Just as a painter uses pigments to depict an original, the name-maker or lawgiver creates primary names by imposing the appropriate letters and syllables.

The theory of mimesis develops Cratylus’ naturalism by framing the natural relation between names and objects as a stable, mimetic relation, and effectively resolves the lack of normativity by taking the being (οὐσία) of things as the external standard for assessing the correctness of names. In Socrates’ view, Hermogenes’ conventionalism can only be reasonably justified based on the sophists’ ontology that the being of things is determined by subjective cognition. He interprets Protagoras’ doctrine that man is the measure of all things as ontological relativism: things are to someone as they appear to someone (385e–386d). Not only do the properties of things depend on the sense-perception, but also their essence is subjectively determined. As individuals perceive the same thing differently, things lack a definitive essence, permitting people to name them arbitrarily by virtue of conventions and customs. Socrates emphasizes to Hermogenes that the being of a thing, which pertains to its nature, is distinct from its appearances (φάντασια) and sensible properties (386e). All beings, and the practices that deal with them, have their own nature.

In summary, Socrates’ naturalism contains two crucial points that set it apart from the positions of Cratylus and Hermogenes: (S1) Names imitate the nature (or essential properties) of things in some way, and (S2) Names can be correct or false. (S1) is a necessary but not sufficient condition for (S2). The concepts of imitation or mimesis have deep metaphysical roots across Plato’s dialogues, where the ‘original-image’ paradigm is a recurring theme. Plato establishes a dichotomy between the original (πράδειγμα, εἴδος) and the image, giving a negative connotation to the world of images (Esposito 2023: 95). In the Republic, Plato employs the metaphor of divided lines to delineate the hierarchy of images (510a–b). The relation between the original and the image is inherently relative; some images assume the role of original for others, with diverse images descending from the noblest Forms, vacillating between pure being and non-being (479d). The separation of the original and the image occurs through mimetic activity (Sophist 265b). In the Timaeus, Plato envisions the cosmos as a product of the gods’ mimetic art (29a–b). Language, the most expansive image of all, occupies dual positions: speech disclosing eternal entities is irrefutable and invincible (29b), whereas speech imitating changeable things is susceptible to rebuttal, as their perpetual alterations breed perplexity (Letters VII 343c).

When investigating the gods’ names (408c–d), Socrates notes that Pan (Πᾶν), the goat-like deity, is Hermes’ double-natured son. As an imitation of all things (πᾶν), human language also has a dual nature: its correct part, dwelling among the gods above, is smooth and divine, while the false part is vulgar and goatish, residing below, among human beings. These two parts are intimately connected, highlighting the relation between the correctness and falsehood of language. To clarify this dual nature, Socrates introduces instrumentalism (387a–391b) and sound-symbolism (426b–427d) in his discourse with Hermogenes. Both theories counter conventionalism, collectively forming the theory of mimesis for ideal language. In the dialogue with Cratylus (427e–440e), Socrates elaborates on a theory of mimesis concerning ordinary language. The first theory explains how names serve a descriptive function, considering the ways in which a name ‘N’ correctly expresses certain descriptive information about the being of its object X. The second theory focuses on the falsehood of names, explaining how a name ‘N’ that lacks descriptive function can still refer to an object X. For example, despite ‘Hermogenes’ failing to correctly represent the being of Hermogenes,3 it still remains meaningful to use ‘Hermogenes’ to refer to him.

Plato’s ideal language

Socrates treats names as ideal instruments within instrumentalism. The argument unfolds as follows:

(I1) Every practice must be carried out based on its own nature and the nature of things, while using the correct instruments; otherwise, errors occur.

(I2) Names are instruments for the naming and speaking practice; the correct naming practice must be conducted based on its own nature and the nature of objects.

(I3) Not everyone knows how to impose names correctly; only lawgivers skilled in the naming art know how to impose names according to the nature of things.

(I4) The function of names is to differentiate and teach beings (τὸ ὄν). Philosophers who are skilled in dialectic know how to use names well. Thus, only philosophers know whether names are finely imposed.

In steps (I1) and (I2), the ‘nature’ of things pertains to their Forms. An instrument is correct if it embodies (ἀποδίδωμι) the Form of a thing through its material (389c). Socrates distinguishes between the Form and the material of names; a name is not the same as its letters and syllables. In step (I3), the lawgiver’s naming art has two facets (Ademollo 2011: 129–132): on the one hand, he looks into that which is the name itself, the universal Form shared by all names; on the other hand, he discerns the proper Form of each thing and puts it into letters and syllables (389d–389e). The relation of the universal Form to the proper Form is akin to that of a genus to a species (Calvert 1970: 33). In other words, at the genus level, the universal Form provides a general definition of a name; it necessitates the embodiment of the proper Form at the species level, making this name belong to a certain thing. Kahn (1973: 162) points out that Socrates thus illustrates the dialectical relation between the general and the specific functions of names, similar to how shuttles with the same appearance may differ in size to suit different fabrics (389b). In steps (I4) and (I5), the descriptive function of names is specified as differentiation and instruction. Philosophers, a select group with the knowledge of the proper Form (390b), are qualified to guide lawgivers in imposing names by drawing from the universal Form. The correctness of a name is intrinsically tied to the good it achieves; a name is correct mainly because of its excellence, namely, being finely imposed and functioning well (390d).

Instrumentalism assimilates all names to proper names, by implying that each name uniquely corresponds to particular things in the non-linguistic world, with the proper Form of its object dictating the descriptive content of a proper name. However, the world is not merely a conglomerate of homogeneous beings. The scope of naming extends beyond particular things to encompass various composite beings and abstract properties. Lawgivers encounter the formidable task of imposing a finite set of names to teach and differentiate an infinite array of beings. This prompts Socrates to continue searching for whatever in the world the correctness of names consists of (391b). Unlike proper names, a common name represents a group of particular things or abstract properties, making it impossible to establish a direct correspondence with its object based solely on the proper Form. Socrates tackles this issue by reducing the correspondence between names and objects to the similarity between letters and fundamental beings, envisioning the process of a naming practice according to etymology. Through an extensive etymological analysis, he illustrates that most names are secondary names, which can be deconstructed into primary names. Ultimately, all names trace back to primary names composed of phonemes that imitate certain properties of objects (421e–422c).

Socrates proposes that names related to constantly changing things, such as ἰόν (motion), ῥέον (flow), and δοῦν (binding), are primary names. These names serve as the elements of language, as all names investigated in etymology trace back to them (422b–c). To explore how the primary names convey descriptive information, Socrates envisions the origins of language (422e–423e). Before names existed, people relied on gestures and other bodily movements to imitate what they wanted to express. With the introduction of names, repetitive bodily gestures were replaced with more flexible imitations using sounds and mouths. Socrates replaces the term ‘embody’ with ‘imitate (μιμέομαι)’, distinguishing the naming art from music and painting. Unlike these arts, names do not replicate the sensible properties of things such as sound and color. For instance, those who imitate animals through onomatopoeia are not engaged in naming.

The lawgiver’s naming practice is a kind of mimetic activity akin to painting, delineated into three primary phases (424c–425a). The initial phase entails classification, which involves differentiating various letters and all beings in the world to identify fundamental beings that serve as simple elements like letters. Subsequently, the assignment (ἐπιφέρω) phase involves applying one or more letters to these fundamental beings or combining multiple letters for composite beings, akin to how a painter applies pigments based on similarities. Lastly, the combination phase involves combining letters into syllables, then into nouns and verbs, ultimately creating the grand and beautiful Logos. To illustrate the similarity between letters and fundamental beings, Socrates introduces the theory of sound-symbolism. Here, different types of movements are regarded as fundamental beings, with vocal organs and their movements constituting part of language. During pronunciation, letters imitate the movements and positions of things through the tongue and mouth. For instance (based on 416c–427c), the tongue is least at rest in pronouncing the letter ‘ρ’, so ‘ρ’ seems to be an instrument for imitating every sort of motion in names like ῥεῖν (to flow) and ῥοή (stream); while ‘ν’, sounded inwardly, is used in words like ἔνδον (inside) and ἐντός (within). These letters are the roots of all names, and the similarities between letters and beings are continually transmitted across various names. Human language evolves from these roots, imitating the structure and order of the cosmos as it progresses from letters to syllables, primary names, secondary names, and speech. Reflecting on (L4), the theory of mimesis regarding the ideal language can be encapsulated as follows:

(T1) A name ‘N’ is correct for an object X if and only if ‘N’ speaks of the being of X.

(T2) ‘N’ speaks of the being of X if and only if ‘N’ participates in the universal Form of names while embodying the proper Form of X.

(T3) ‘N’ embodies the proper Form of X if and only if: (a) if ‘N’ is a primary name and X is a fundamental being, then ‘N’ imitates the being of X; (b) if ‘N’ is a secondary name and X is a composite being, then the primary names (N1, N2, …, Nn) obtained from ‘N’ through etymological analysis, and the fundamental beings (X1, X2, …, Xn) that compose X, both satisfy condition (a).

(T4) A primary name ‘N’ imitates the being of a fundamental being X if and only if ‘N’ adheres to the principle of similarity: (a) the letters and syllables of ‘N’ possess enough essential properties of X, distinguishing ‘N’ from the names of other beings; (b) ‘N’ excludes all essential properties of other beings.

Instrumentalism and sound-symbolism are based on different ontologies. Socrates cautions Hermogenes that etymological analysis aligned with the doctrine of flux might be absurd and ridiculous (426b). In the etymology section (391c–421c), Socrates imitates sophists like Euthyphro, Anaxagoras, and Heraclitus to explain the Greek etymology. The sophists mistake their state of perplexity for possessing knowledge, believing that names are created for things in a state of motion and flux (411c). Toward the end of the dialogue (437a–c), Socrates attempts to reconstruct etymology entirely. He suggests that if the being of things is static and unchangeable, then new primary names and fundamental beings must be identified, altering the analysis of names like ‘knowledge (ἐπιστήμην)’. Socrates’ reevaluation of etymology implies the possibility of imitating static things like Forms. He seeks to present a normative theory concerning the naming practice, scientifically conceiving how a lawgiver should create an ideal language that accurately describes Forms. In this language, all names and statements trace back to a set of primary names imitating fundamental beings. The name-maker’s ontological perspective is embedded in these primary names. Indirectly, the ideal language imitates Forms through the similarities between letters and beings, faithfully differentiating and teaching things by describing their essence. This mimetic relation assures that language has a natural correctness independent of historical events and subjective opinions, thereby achieving harmony between nature and convention (Sedley 2003: 68).

The falsehood of ordinary language

When Cratylus breaks his silence and joins the dialogue, Socrates decides to re-examine his own arguments (428b). In instrumentalism, the lawgiver is depicted as an exceptional craftsman among the masses (389a) denoted by the flawless naming art. However, reality reflects varying degrees of art among craftsmen, thus the names and laws in ordinary language created by lawgivers also vary in quality (428d–429b). To illustrate the possibility of false names, Socrates provides a case of misusing names: if someone met Cratylus abroad, grasped his hand, and said, “Hello, Athenian stranger, son of Smicrion, Hermogenes!” they would effectively address Cratylus with the false name ‘Hermogenes’ (429e). This example shows that the language practice encompasses not only vocal organ movements and accompanying sounds but also physical acts, tone, and grammatical variants that convey the speaker’s intention. Non-discursive acts like ‘grasping Cratylus’ hand’ indicate that the speaker intends to refer to Cratylus. Additionally, the imperatives like ‘χαῖρε (hello)’ and the vocatives like ‘ὦ ξένε Ἀθηναῖε (Athenian stranger)’ and ‘Ἑρμόγενες (Hermogenes)’ signal to the audience that the speaker is addressing the person present (Smith 2008, 143). These speech acts and grammatical factors collectively form the context of communication. For Cratylus, they serve as ‘assertability conditions’ (Kripke 1982: 74) rather than truth-value conditions. When these conditions occur, Cratylus can judge the speaker’s intention, focusing solely on the referential relation between the name and the referent. Based on this, Smith (2008: 147) argues that Socrates overrides the principle of similarity. The descriptive function or mimetic relation is not a necessary condition of reference. Even if a name bears no similarity to its object, it can still successfully refer to that object within the linguistic community.

Nevertheless, the case of Socrates should not be interpreted, as Smith (2008: 150) does, as a conventionalist rebuttal of naturalism. Instead, it demonstrates that if a name, like ‘Hermogenes’, does not describe the being of its object, it can still be used arbitrarily to refer to any object, even in violation of convention and custom. While the mimetic relation or the descriptive function is not necessary for reference, it remains a necessary condition for correct reference (Smith 2008: 150). Socrates utilizes this case to elucidate his own naturalism, which contrasts with that of Cratylus by allowing for the possibility of misreference. He proceeds to explain this case through the theory of mimesis. When we approach someone and refer to him with the name ‘N’, this speech act is akin to distributing (διανέμω) a picture to its original. Just as a painter might incorrectly distribute a painting of a man to a woman, a speaker might distribute a name to refer to the wrong object (430c). The audience can bypass the established convention and custom to understand the speaker’s intention based on the speech act. Socrates distinguishes between the name establishment and the subsequent use, by indicating that names might be correctly made but later misused. It is necessary to establish a mimetic relation between names and objects independently of the user’s intentions to avoid such misreference.

However, the lawgiver may make the same mistakes as users, leading to names getting falsely established. Socrates extends the concept of falsehood to the lawgiver’s naming practice: the lawgiver might distribute inappropriate elements during the second and third phases of naming, thereby creating false linguistic images (431b–e, 432e). This occurs when letters and syllables are either omitted or added while creating primary names, or when inappropriate primary names are used to combine secondary names and speeches. Even with these imperfections, names and speeches can still describe the being of their objects as long as most of their elements or main letters are appropriate (431c, cf. 393e). Socrates thus revises (L2), by arguing that ‘the whole speech being true’ is not always a sufficient condition for ‘all elements being true’. The falsehood of language does not entirely contradict correctness; rather, it is a negation of correctness, signifying ‘incorrectness’ and ‘untruth’. False names are flawed images that lie between the ‘perfect imitation’ and the ‘complete dissimilarity’, bearing some degree of similarity to objects, though with less accuracy than correct names.

The ‘two Cratyluses’ argument (432b–d) further demonstrates that whether perfect or imperfect, an image only resembles its original but is never identical to it. Socrates argues that if a god could replicate not just Cratylus’ color and shape, but also all internal properties, it would result in two indistinguishable Cratyluses. This highlights that an image must lack some properties of the original, and recognizing this difference is essential to identifying it as an image. The difference between the image and the original determines that ordinary language can never be a perfect imitation of beings. Socrates thus persuades Cratylus to acknowledge that as long as the ‘pattern (τύπος)’ of the object is present in names and speeches, the object is effectively named and spoken of (432e–433a). The term τύπος more aptly denotes ‘Form’ or ‘general property’ (Ademollo 2011: 371). In the Theaetetus, Socrates argues that the whole (ὅλος) is distinct from all (πᾶς) its elements, constituting an indivisible unity devoid of parts (203a, 204b). Here, he might be suggesting that names and speeches not only imitate the objects through their elements but through their whole. Just as a painting’s whole transcends the combination of its pigments to imitate the unity of the original, linguistic images by virtue of its whole imitate the Forms of objects. In summary, by combining (L4), the falsehood of ordinary language can be outlined as follows:

(F1) A name ‘N’ is a false name for object X if and only if ‘N’ falsely speaks of the essence of X; that is, either (a) according to (T1) through (T4), ‘N’ is the correct name for object Y but is effectively used to refer to X; or (b) according to (T3) and (T4), most of the letters or main syllables of ‘N’ possess essential properties of X, and yet ‘N’ contains some incorrect letters.

In his examination of the convention’s role (434c–e), Socrates discovers that the Greek name ‘σκληρότης (hardness)’ can imitate both the essential property and its opposite of the object being named. The mimetic force of the letter ρ that imitates hardness is countered by the force of the letter λ which express softness. Despite this, ‘σκληρότης’ remains intelligible and meaningful in communication. People understand this name not through similarity but by convention (ἔθος). Socrates proposes that ἔθος involves forming a convention not only with others but also with oneself (αὐτὸς σαυτῷ συνέθου). Consequently, the correctness of names becomes a matter of convention, as Socrates reluctantly concedes, “the power of similarity is actually poor; thus, it is necessary to make use also of this vulgar means, convention, for the correctness of names (435a–c).” Many scholars take this concession as proof of Socrates’ endorsement of conventionalism (Barney 1997: 145). Schofield (1982: 66) notes that it diminishes any remaining appeal of naturalism.

This interpretation actually confuses Socrates’ explanation of ordinary linguistic phenomena and his normative naturalism. Socrates is acutely aware of the role of convention in ordinary language practice. Once created, ordinary language is dynamic, with users typically employing names without delving into the essence of things. The descriptive function of names thus becomes secondary, supplanted by a referential function governed by the speaker’s intention. Regardless of their initial correctness, names tend to transition from images to pure signs akin to numbers (435b). Consequently, the mimetic relation fixed by similarity gives way to a referential relation. In Socrates’ argument, σκληρότης constitutes an extreme case that defines the limits of the theory of mimesis without overturning it, illustrating that there are certain names function as pure signs like numbers in ordinary language (435b). These signs rely entirely on convention to refer to their objects. The referential relation between signs and the referent is contingent and changeable. Socrates underscores to Cratylus that the correctness of names is intrinsically tied to its excellence and function (435d–436a). The primary function of names is teaching, requiring that one who knows ‘N’ also knows the object X it describes. Similarity is the only and best way for names to teach things. Socrates’ argument indicates that, whether within the realm of the ideal language or the ordinary language, the descriptive function of names – imitating the being of things – is a necessary condition for the correct reference and the achievement of knowledge. Conventions play a supplementary role in the operation of ordinary language, especially ensuring that those pure signs, which do not conform to (F1), can be successfully used.

What is the correctness of language?

Socrates’ theory of mimesis illuminates the connection between the dual nature of human language. The difference between images and originals determines that human language inevitably contains some falsehoods. However, since differences always presuppose a certain degree of similarity, these falsehoods necessitate a mimetic relation between language and beings, thereby endowing human language with its correctness. Socrates aims to offer a normative theory for naming and talking about beings, though his standard for correctness does not fully align with the ordinary language practice. So why does Socrates emphasize the mimetic power and descriptive function of language in explaining its correctness? The concept of mimesis not only poses a metaphysical dichotomy of ‘original-image’ but intertwines with the ‘philosopher vs. sophist’ debate. Reames (2018: 149) observes that, in confronting sophists’ relativism and sophistry, Socrates struggles to overcome the flux and contradictions in their doctrines by distinguishing between the truth and the image. In the Theaetetus, an army of poets and sophists led by Homer perceives everything as in constant flux (152e). Socrates integrates Heraclitus’ doctrine of flux, Protagoras’ relativism and Theaetetus’ perspective, leading to the impossibility of language (Burnyeat 1990, 9). Philosophers, by contrast, are depicted as those who transcend particular things, while delving into the essence of things (175c). In the Cratylus, Socrates proposes that one can only know the Form by directly ‘looking into (βλέπων πρός)’ the thing itself (389a). βλέπω entails abstaining from cognitive interference and immersing oneself in beings, allowing things to reveal themselves as they are. Conversely, sophists assert that the being of things resides in their sensible properties and appearances, with flux and motion constituting the essential properties of things (401d, 411c, 421b). They thus reject ‘original-image’ paradigm, by aligning their doctrines with the common sense and the opinions of masses.

The pre-Socratic sophists’ fascination with wordplay and semantic analysis is widely recognized (Barney 1998: 68). Sophists traverse the polis, while hawking their doctrines for lucrative fees through persuasive oratory. In fact, the etymological section vividly unveils the intellectual landscape of the times, with Plato orchestrating Socrates to imitate poets and sophists, elucidating the etymology of the ordinary Greek through dramatized imitation, devoid of his own beliefs. Socrates emphasizes that an initial mistake in imposing primary names can lead to a chain of errors, similar to how a small initial miscalculation in mathematics can cause subsequent steps to follow in agreement with each other (436b–d). As Sedley (2003: 32–33) observes, etymology employs the skill of decoding to clean away the phonetic accretions and distortions that names had acquired over the centuries, thereby recovering their underlying form and providing insight into the mindset of the ancients. Socrates discovers that the ancients were not divine and wise, despite pretending to be wise (411b–c). More significantly, Barney (1998: 84–85) argues that etymology is at the bottom of a form of agonistic display, offering an extremely clever imitation of the method and practice of Socrates’ rivals. Socrates’ success in performing this etymological display allows him to surpass the sophists’ views and ultimately dismiss them.

Etymology begins with the poets Homer and Hesiod as authorities, after which, Socrates is inspired by the Euthyphro (396c–397a), and, ultimately, he follows the doctrines of Anaxagoras and Heraclitus. The etymological analysis progresses systematically from natural things to human affairs, investigating the name of various beings: (1) heroes and humans; (2) natural beings; (3) Greek gods; (4) celestial bodies; (5) moral and intellectual virtues and vices; and (6) the greatest and finest things. Socrates critiques certain names assigned by the poets to heroes and humans, while noting that some were devised for supplications (397b). It was not trifling people, but rather ‘sky-watchers and chatterers’ who imposed the names of the gods (400e–401b). Drawing on the insights of natural philosophers and Heraclitus’ doctrine of flux, Socrates elucidates the material and semantic connections among different names. For example (400b, 401d), ‘soul (ψυχὴν)’ is interpreted as the ‘holder of nature (φύσιν ἔχει)’, while ‘being (ὐσίαν)’ is rendered ‘ὠσίαν’ in the vernacular, reflecting Heraclitus’ view of beings as propelled by ‘the thrust (ὠθοῦν)’. All the names of virtues presuppose movement, flux, and coming-to-be (411c–413d): ‘wisdom (φρόνησις)’ is ‘intellection of movement and flux (φορᾶς καὶ ῥοῦ ῦνόησις)’, ‘justice (δίκαιον)’ denotes ‘through (διαϊόν)’ for everything, and ‘courage (ἀνδρεία)’ is ‘the flow contrary (ἀνρεία)’, not to any flow, but to that opposing the justice. However, a single name may yield multiple explanations, lacking a consistent standard for adjudicating the correctness of etymological analysis (413b–d). Unexplained names are classified as exonyms (409d–410a, 416a, 421c–d).

Hermogenes confronts Socrates’ etymology of virtue names, by insinuating that Socrates has merely heard these explanations from others. Socrates admits to deceiving Hermogenes (413d), and further acknowledges that the entire etymology is self-deception (428d), imitating the sophists’ practice rather than an authentic exhibition of his own soul. The sophists, complaining about their fluctuating speech akin to the tides of the Euripus (Phaedo 90c), exploit the distance between the image and the original to misconstrue the meanings of names, particularly those like ‘justice’ and ‘good’ (Phaedrus 263a). Their speech caters to the opinions and perplexity of the masses, falsely representing the original as if it has been genuinely revealed (Sophist 234c). Plato expressed concern about the sophists’ sophistry and wordplay, enabling them to espouse contradictory propositions simultaneously (Euthydemus 275d–276c), manipulating the public opinion and undermining ethical values. Instead of teaching the being of things, sophists’ etymology occupies a position prior to the original, transforming knowledge into relatable opinions accessible to audiences, ultimately entrenching them in subjective explanations and linguistic constructions of the reality.

Concluding the dialogue (439d–440a), Socrates underscores that if things are constantly changing and never stay the same, people cannot correctly describe or refer to them, not even with the demonstratives like ‘this (ἐκεῖνό)’ and ‘that (τοιοῦτον)’. By the time we speak of things, they instantly become something else. Naming acts presuppose that the lawgiver knows the being of things, but they cannot define things in flux or form definite knowledge about them. The sophists’ doctrine of flux undermines not only the possibility of naming but also the very foundations of language and knowledge, which ultimately threatens the legislative and moral code of the polis. If language merely imitates the world of flux, it certainly carries an innate tendency to mislead and misinform. Socrates, therefore, asserts that no one with understanding will trust names or name-makers (440c). His negative attitude toward the ordinary language implies that a philosopher’s role is not to engage in wordplay and etymological analysis, but to guide the audience toward knowledge of the Forms beyond linguistic images. The correctness of language should be measured by its functionality and suitability in the pursuit and exchange of knowledge. By employing the dialectical art of questioning and answering, Socrates systematically uses the one part of language to manifest another. While Hermogenes seeks the precision of language to restore his own name, Socrates’ etymology reveals the inherent falsehood and ambiguity in the ordinary language, guiding him to recognize the distinction between correctness and factualness. Cratylus, who adheres to both Heraclitus’ doctrine of flux and naturalism (440e), is guided by Socrates to understand that it is impossible to reasonably define the correctness of names based on the doctrine of flux. Through ongoing dialogues and debates, interlocutors undergo introspection and self-discovery, shedding the initial opinions and confusions. Socrates’ dialectical art serves as midwifery, nurturing pregnant souls and discarding unreal images upon discovery (Theaetetus 151c), relying on the dual nature of language illustrated by the theory of mimesis.

Conclusion

The Cratylus primarily explores the debate between Cratylus’ naturalism and Hermogenes’ conventionalism, neither of which fully accounts for a standard of correctness or the possibility of false names. Socrates’ theory of mimesis, structured around the ‘original-image’ paradigm, effectively resolves these issues by reconfiguring the contradiction between naturalism and conventionalism into a tension between the ideal language and the ordinary language. While ideal linguistic images aim for the excellence of their descriptive function and faithfully imitate their originals through similarity, the inevitable difference between the image and the original introduces the potential for falsehood in naming practices. This study suggests that Socrates develops a refined form of naturalism, where language’s correctness is evaluated based on its function to facilitate the pursuit and communication of knowledge. Since ordinary language is prone to be misused and misunderstood once established, Socrates highlights the supportive role of convention in establishing a fixed reference while maintaining that the mimetic relation is a necessary condition for the correct reference. Sophists exploit the falsehoods of language to manipulate words and build sophistry on the doctrine of flux. To counter this, Socrates emphasizes the descriptive function and mimetic power of language, offering fresh insights into the balance between the normative correctness and the practical functionality in speech. Philosophers like Socrates choose to guide interlocutors toward the knowledge of Forms through dialectical art.

Reference

Burnyeat, M., 1990. The Theaetetus of Plato. Indianapolis and Cambridge: Hackett Publishing Company.

Barney, R., 1997. Plato on Conventionalism. Phronesis, 42(2): 143–162. https://doi.org/10.1163/
156852897762700070

Demos, R., 1964. Plato’s Philosophy of Language. Journal of Philosophy, 61(20): 595–610. https://doi.org/10.2307/2023441

Kahn, C. H., 1973. Language and Ontology in the Cratylus. In: Lee & A. P. D. Mourelatos & R. M. Rorty (eds.), Exegesis and Argument, 152–176. Assen: Van Gorcum & Comp. B. V.

Kripke, S. A., 1982. Wittgenstein on Rules and Private Language. Cambridge and Massachusetts: Harvard University Press.

Reames, R., 2018. Seeming and Being in Plato’s Rhetorical Theory. Chicago and London: UCP.

Robinson, R., 1956. A Criticism of Plato’s Cratylus. The Philosophical Review, 65(3): 324-341. https://doi.org/10.2307/2182143

Saussure, F., 1959. Course in General Linguistics. In: C. Bally & A. Sechehaye (eds.), W. Baskin (trans.). Philosophical Library.

Schofield, M., 1982. The dénouement of the Cratylus. In: M. Schofield & M. C. Nussbaum (eds.), Language and Logos: Studies in Ancient Greek Philosophy, 61–82. Cambridge: CUP.

Smith, I., 2008. False Names, Demonstratives and the Refutation of Linguistic Naturalism in Plato’s Cratylus 427d1–431c3. Phronesis 53(2): 125–151. https://doi.org/10.1163/156852808X278703

Williams, B., 1982. Cratylus’ theory of names and its refutation. In: M. Schofield & M. C. Nussbaum (eds.), Language and Logos: Studies in Ancient Greek Philosophy, 83–94. Cambridge: CUP.

1 There is debate regarding the placement of 385b2–d1 in the text. Schofield (1972) suggests moving this passage to after 387c5, while Mackenzie (1986) proposes the opposite view.
2 It is widely recognized that the term ‘name (ὄνομα)’ is generally understood to bear descriptive content in the Cratylus (Barney 1997: 143). Plato’s example of ‘ὄνομα’ includes not only proper names and common nouns, but also adjectives (412c, 433e) and infinitives (414a–e), and participles (421c).
3 In Cratylus’ view, ‘Hermogenes’ is not the name of Hermogenes (383b). Hermogenes is misnamed because ‘Hermogenes’ means the son of Hermes who is the god of wealth. Socrates points out that Cratylus might think Hermogenes has always tried to make money but has never succeeded in business (384c). Socrates also offers another interpretation of ‘Hermes’ (408a–e).

Plato’s Theory of Mimesis in the Cratylus: from Ideal Language to Ordinary Language

Platono mimezės teorija dialoge Kratilas: nuo idealios kalbos prie kasdienės kalbos

Introduction

The theory of language as mimesis

Plato’s ideal language

The falsehood of ordinary language

What is the correctness of language?

Conclusion

Reference