Version classiqueVersion mobile
OpenEdition Books

Greek and Latin. Expressions of Meaning

 | 
Andreas T. Zanker

Appendix III: Ancient Theories of Meaning

Texte intégral

  • 1 Mates (1973), Long (1974), Long & Sedley (1987), 1.97-101, 183-236; Manetti (1993), Sluiter (1990) (...)

1From the fifth century BC onwards we have a range of texts that engage closely with the topic of meaning and signification, but we can see the question of the meaning of words already being discussed in the fragments of Heraclitus and Parmenides. Problems of interpretation are also broached in the early verse of Homer and Hesiod. In this appendix I would like to provide something of an introduction to some of the main lines of classical philosophical debate, fully aware of the fact that I will not be able to do justice to the topic here. In spite of the brevity of the sketch, I hope that such an appendix may be useful for orientation. As mentioned in the Introduction, this book is not about ancient theories of signification, but about the vocabularies of meaning that were assembled in the ancient world. The following discussion will center on the conceptions found in Plato and Aristotle, as well as the Stoic and Epicurean theories of language (as far as we can reconstruct them).1

2As we have seen, no Greek noun that could be used in the sense of S-meaning is attested in that usage before the end of the sixth century. Verbs that would later be used of inanimate objects (including pieces of text) in the sense of “this means” are likewise absent from our early written sources. Despite this, there is abundant evidence that the Presocratic philosophers discussed language. In the fragments of Heraclitus (late sixth and early fifth century), we see equations made between near homonyms, for example the play on the similarity of the terms “βιός” (“bow”) and “βίος” (“life”): the name of the bow (“τόξον”) is “life” (“βίος”), but its work is death (22 B 48 DK). Parmenides mentions the way that the names that men have given to objects are inaccurate (28 B 8.38-41, 53-59, 9.1-4 DK), and Empedocles also discusses the relationship between word and thing (31 B 9 DK). From what we can tell, the sophists of the second half of the fifth century (Protagoras, Prodicus, and Hippias, for example) debated issues of language in earnest: Protagoras, for example, could criticize the use of the term “µῆνις” (“anger”) by Homer as a feminine noun, since from its meaning one would expect that it would be masculine (Aristotle, On Sophistical Refutations 173b17-22), and Prodicus seems to have been teaching on the topic of the correctness of names (Plato, Cratylus 384b).

  • 2 See Ademollo (2011); cf. Barney (2001); Sedley (2003); Ewegen (2014). There are a number of import (...)
  • 3 For the argumentative positions in the Cratylus, see Kretzmann (1971), 126, and Long (2005), 43.
  • 4 Kahn (1973), 168:“Thus the Cratylus urges us to study not linguistics but ontology”.

3It is with Plato, however, that the issue of language and meaning becomes central in our surviving texts. In his Cratylus, the philosopher engages with the question as to whether the meaning of words is fixed by nature (“φύσει”) or by imposition (“θέσει”) – that is to say, whether the forms of words reflects something about the nature of the things that they denote, or whether they are simply imposed conventionally.2 The former view (nature) had been implicit in many of the etymologies offered by Homer and Hesiod, while the latter (imposition) seems to have been maintained by Democritus in the late fifth century (68 B 26 DK). Both sides of the question are considered – the figure of Cratylus argues for the φύσει side, Hermogenes for that of θέσει – but both options are found to be lacking, albeit for different reasons.3 Hermogenes’ radical claim, that the relationship of words to things is a matter of convention and the desire of the speaker (according to him, Socrates can use the word “horse” for a man), appears to lead to relativism: it cannot be up to an individual speaker alone to determine the meaning of words. A name is a tool (“ὄργανον”) for teaching and separating reality (Cratylus 388a-c), and it cannot perform this function if speakers cannot distinguish between true and false propositions. Cratylus’ thesis, that words resemble the things that they represent naturally, also falls down: words change form over time, and terms such as “σκληρότης” (“hardness”, “harshness”) appear to contain letters and sounds (for example, the letter “λ”) that are otherwise associated with slipperiness and softness (Cratylus 434c-435d). The somewhat ambivalent conclusion is that both naturalism and conventionalism have their place, although what precisely this amounts to remains unresolved – Socrates notes that words do not adequately evoke things as they are and suggests that one should investigate the things themselves rather than trusting too much in words (Cratylus 438d-439d).4 In spite of this, the Cratylus provided a starting point for both Aristotelian and Stoic reflections on language – Aristotle ruling out the naturalist position that connected words with things, the Stoics taking it as the starting point for their own theories.

  • 5 On meaning in Aristotle, see Modrak (2001), especially 13-51. For investigations of this area of A (...)
  • 6 Kretzmann (1974), 3, describes this excerpt as “the most influential text in the history of semant (...)

4The bulk of Aristotle’s reflections of meaning is to be found in the logical treatises that make up the Organon (Categories, On Interpretation, Prior and Posterior Analytics, Topics, and On Sophistical Refutations) as well as in his works on rhetoric and poetry (at Poetics 1457a4-10, for example, on terms that are “ἄσηµος”). Additional engagement is to be found, however, throughout his other works (e.g. Metaphysics 1006a-b).5 At the beginning of his On Interpretation, Aristotle provides one of the seminal descriptions of the relationship between word and thing in the history of semantics (it is also the only major statement on the topic that he offers):6

1. ἔστι µὲν οὖν τὰ ἐν τῇ φωνῇ τῶν ἐν τῇ ψυχῇ παθηµάτων σύµβολα, καὶ τὰ γραφόµενα τῶν ἐν τῇ φωνῇ. καὶ ὥσπερ οὐδὲ γράµµατα πᾶσι τὰ αὐτά, οὐδὲ φωναὶ αἱ αὐταί· ὧν µέντοι ταῦτα σηµεῖα πρώτων, ταὐτὰ πᾶσι παθήµατα τῆς ψυχῆς, καὶ ὧν ταῦτα ὁµοιώµατα πράγµατα ἤδη ταὐτά.

“Things that are spoken are the symbols of affections in the soul, and the things that are written down are the symbols of the things that are spoken. Just as writing is not the same for all races, neither is spoken language the same for all. The primary things of which they are signs, i.e. the affections of the soul, are, however, the same for all races, as are the actual things of which these [i.e. the affections of the soul] are likenesses” (Aristotle, On Interpretation 16a3-8).

  • 7 At On Sophistical Refutations 165a7-8 we read that names are symbols in the place of things: “...τ (...)
  • 8 Aristotle of course adds a fourth entity – written words. See Manetti (1993), 55, who suggests tha (...)
  • 9 Cited by Lyons (1977), 96.
  • 10 This is, however, the only apparent reference to Plato in On Interpretation. See Kretzmann (1974), (...)

5Written words (“τὰ γραφόµενα”) are the symbols of spoken words (“τὰ ἐν τῇ φωνῇ”), while the spoken words are the symbols of affections (or impressions) in the soul (“τῶν ἐν τῇ ψυχῇ παθηµάτων σύµβολα”); these affections in the soul are caused by objects in the world.7 Aristotle thereupon states that he has dealt with the topic elsewhere, and summarily ends his discussion. This is, however, our first explicit articulation of the semiotic triangle,8 the relationship between thought, speech, and reality evoked by the scholastic maxim “uox significat [rem] mediantibus conceptibus” (“the word signifies [the thing] with concepts mediating between the two”).9 These affections in the soul are often taken to be mental states; these mental states mediated between words and objects. By means of this triangle, Aristotle seems to provide an answer to the dispute found in Plato’s Cratylus: written words, the words themselves, and speech are symbols of the affections in the soul “by convention”, whereas the relationship between the affections in the soul and the things themselves is a necessary one. Indeed, Aristotle seems to allude directly to Plato’s description of words as tools in the Cratylus (Cratylus 388a) in his statement that “every sentence is significant, not as a tool but, as we said, by convention” (On Interpretation 17a1-2).10

  • 11 See Ackrill (1963), 113-114; Long (1971), 121.
  • 12 Irwin (1982), 242, argues that the beginning of On Interpretation engages with signification rathe (...)
  • 13 On the development of the term “σύµβολον” (“symbol”), see Struck (2004).
  • 14 See Kretzmann (1974), 5.

6There are nevertheless problems with Aristotle’s account. His statement that the relationship between the affections in/of the soul and the things themselves is the same for all races of men is clearly incorrect, as it is hard to imagine objects exciting identical thoughts in all human beings.11 Different individuals form different thoughts when they observe the same object. Aristotle’s language is also notoriously obscure; it is unclear, for example, as to whether one should interpret the phrase “ἐν τῇ ψυχῇ παθηµάτων” as referring to thoughts at all: the “affections” mentioned here may simply refer to sense impressions.12 Besides this, some scholars argue for a distinction between the terms “σύµβολα” and “σηµεῖα”, the former designating conventional signs, the latter denoting natural ones. Historically, a symbol consisted of half of a token that was broken between two parties (guest-friends, for instance) in order to enable a service to be returned or to render it transferable, and the term’s usage here in conjunction with “σηµεῖα” may have weight to it.13 Boethius simply translated the two terms with the Latin noun “nota” (“sign”), but this may distort the nuances of the Greek.14 In any case, this short excerpt from the beginning of Aristotle’s On Interpretation continues to provoke considerable debate.

  • 15 For compact discussions of Stoic semantics, see Long & Sedley (1987), 1.199-202; Sluiter (1990), 2 (...)
  • 16 On the Stoics’ relationship to the naturalist position of Plato’s Cratylus, see Long (2005); Allen (...)
  • 17 See Barwick (1957); cf. Blank (1982), especially 21-22.
  • 18 Augustine also uses the terms “stirps” (“trunk”) and “sementum” (“seed”) of words at De Dialectica(...)
  • 19 Cf. “perspicis enim haec uerba ita sonare ut ipsae res quae his uerbis significantur” (“you see th (...)
  • 20 As Augustine states things, “uis” cannot be traced back to any more primitive word: “reddetur rati (...)

7The Stoic school moved in a different direction from that of Aristotle.15 The Stoics laid considerable stress on questions of semantics in order to guard themselves against sophistic arguments, such as those mobilized by the Platonic characters Euthydemus and Dionysodorus (see the Introduction). The issue of language looms large in the lists of book titles supplied by Diogenes Laertius, and Chrysippus set the study of the signifier and what is signified at the center of his conception of dialectic (Diogenes Laertius, Lives of the Philosophers 7.62). One of the key aims of Stoic philosophy was to reconcile phenomena with the overriding logos; this included the question of words and their origin.16 While the texts that disclose their work one tymology and historical linguistics are late and in large part fragmentary, scholars have reassembled a plausible theory. The narrative is one of decline:17 words originally had a specific, natural meaning but had over the years come to possess different significations. According to the Stoics, the original words, which Origen gave the name “πρῶται φωναί” (“first sounds” Origen, Against Celsus 1.24) and Augustine “cunabula uerborum” (“cradles of words” De Dialectica 6),18 were either onomatopoeic (such as tinnitus”, “hinnitus” or “balatus”, to use Augustine’s examples),19 or, when the sound of the object did not allow onomatopoeia, in some way reflected the things themselves (the smooth sound of the term for “honey”, “mel”, for example, cohered with the sweetness of the thing it described). In the course of time, language moved away from this tight iconic relationship to the world, in that terms came to be metaphorically transferred from one thing to another. It was perhaps on account of this that the Stoic philosopher Chrysippus could say that every word was ambiguous (Gellius, Noctes Atticae 11.12.1). At times, the forms of the word changed their appearance in the course of their transference; thus, Augustine could follow the term “uia” (“road”) back to “uis” asits πρῶτη φωνή (via “uietum”, “uitis”, and “uincere”).20

  • 21 See Manetti (1993), 92-110. For a neat comparison, see Long & Sedley (1987), 1.201.

8The Stoics developed a different semiotic triangle from that found in Aristotle;21 a late source (Sextus Empiricus, Against the Professors 8.11-13) records that the Stoics distinguished between the sign (“σηµαῖνον”), that which is signified (“σηµαινόµενον”, or “λεκτόν”), and the referent (“τυγχάνον”). Two of these are physical – signs in spoken language are disturbances of air and the referent was also conceived of as physical according to Stoic ontology – while the second, the thing said, was non-physical. As Sextus puts it:

2. τούτων δὲ δύο µὲν εἶναι σώµατα, καθάπερ τὴν φωνὴν καὶ τὸ τυγχάνον, ἓν δὲ ἀσώµατον, ὥσπερ τὸ σηµαινόµενον πρᾶγµα, καὶ λεκτόν, ὅπερ ἀληθές τε γίγνεται ἢ ψεῦδος.

  • 22 The quality of this excerpt as a guide to Stoic thought has been questioned; see Frede (1994).

“Of these things [sound, thing signified, and reference], two are bodies – that is to say, the sound and the reference – and one is incorporeal, namely the thing signified and what is said/the sayable, which is also true or false” (Sextus Empiricus, Against the Professors 8.12).22

  • 23 Frede (1994); cf. Graeser (1978), 88-90.
  • 24 Cf. Schenkeveld & Barnes (1999), 210.

9Besides the incompleteness of our texts, the difficulty of the term “λεκτόν” (“sayable”) issues from the fact that its usage may have changed over time; according to Frede, the Stoics employed it primarily in order to discuss propositions, but other types of illocution (questions, orders, and so on) later also came to be described by it.23 There are moreover further complications: according to Diocles in Diogenes Laertius (Lives of the Philosophers 7.63), λεκτά could be complete (full propositions, such as “Socrates writes”) or incomplete (predicates without specified subjects, such as “writes”), although why this should be so is unclear. Moreover, the precise status of an incorporeal entity of this type within a fundamentally materialist philosophical system such as that of the Stoics requires explanation.24 Despite its difficulty for modern interpreters, Stoic thought on language heavily influenced the grammatical tradition (for example, Dionysius Thrax, Varro, and Apollonius Dyscolus), whose usage of the vocabulary we have noted in the previous chapters, and was foundational for texts such as Augustine’s De Dialectica, De Magistro, and De Doctrina Christiana.

  • 25 A standard interpretation of Epicurus’ theory of the origin of language can be found at Sedley (19 (...)
  • 26 Cf. Allen (2005), 28.

10The Epicurean school provided an alternative view; according to it, there was no initial imposition of names in the sense of a rational selection made by an agent.25 The idea that words were imposed on things by gods or exceptional individuals is mocked by Epicurean authors such as Lucretius and Diogenes of Oenoanda (Lucretius, De Rerum Natura 5.1041-1090; Diogenes of Oenoanda, fr. 10, cols. 3-5). According to Epicurus (Diogenes Laertius, Lives of the Philosophers 10.75-76), language originated from human beings’ interaction with their environment: the first words were the result of instinctive acts of expression, human beings simply vocalizing in away similar to that of beasts. According to Proclus (In Cratylum 8.4-7 Pasquali), Epicurus compared the first forms of speech to involuntary actions such as coughing, sneezing, crying out, and so on.26 Gradually, over the course of time, full languages were developed and partly filled out by means of convention:

3. ὕστερον δὲ κοινῶς καθ᾽ ἕκαστα ἔθνη τὰ ἴδια τεθῆναι πρὸς τὸ τὰς δηλώσεις ἧττον ἀµφιβόλους γενέσθαι ἀλλήλοις καὶ συντοµωτέρως δηλουµένας.

“Later, particular forms were fixed by convention within the individual races, so as to make their significations less ambiguous and more concisely expressed” (Diogenes Laertius, Lives of the Philosophers 10.76).

  • 27 Cf. Sedley (1973), 18. For Plato’s own position on foreign languages, cf. Cratylus 389d-390a with (...)

11Epicurean views on language therefore represent something of a middle road between the naturalist and conventionalist positions: nature and convention both played a role, albeit at different moments in the history of language. This understanding of the genesis of language incidentally helped to account for the fact that different races spoke different languages: whereas Aristotle had posited a conventional link between words and things (via the human mind) in order to explain differences in tongue, Epicurus offers a different possibility, arguing that the type of vocalization depended on the environment of the region in which it was uttered.27

  • 28 On this interpretation, see Long (1971), 119-123; Manetti (1993), 119-121; for an opposing point o (...)

12Epicurus’ position on meaning itself is murky; our evidence is simply inadequate. Two ancient authors in fact criticized Epicurus for failing to set anything between words and things as had the Stoics (Sextus Empiricus, Against the Professors 8.13, 8.258; Plutarch, Against Colotes 1119a-1120f); if true, this would suggest that Epicurus had a rather direct account of the relationship between language and reality. In order to find a mediating entity set between words and their referents in the world, some scholars focus on the mental notion of πρόληψις (“anticipation”, “preconception”).28 They observe that in Epicurean epistemology πρόληψις plays a major role in the way in which human beings make inferences; in order for us to make a judgment about whether it is a cow that we see in front of us, we need to have experienced what a cow looks like (Diogenes Laertius, Lives of the Philosophers 10.33). We therefore need (a) to have seen an image of this type of animal, (b) have stored this in the mind, and (c) be able to compare this image with what we see before us. These προλήψεις, gathered over the course of one’s life, play a key role in memory and perception in general. It has been argued that προλήψεις in fact take up the same position as the Stoic λεκτά in a modified semiotic triangle. If προλήψεις are concepts, then it is possible that Epicurus conceived of them as playing a role in what humans meant in speaking and how they interpreted the speech of others. The following passage, from Epicurus’ Letter to Herodotus, would certainly seem to support the view that Epicurus did indeed conceive of something as mediating between words and things:

4. πρῶτον µὲν οὖν τὰ ὑποτεταγµένα τοῖς φθόγγοις, ὦ Ἡρόδοτε, δεῖ εἰληφέναι, ὅπως ἂν τὰ δοξαζόµενα ἢ ζητούµενα ἢ ἀπορούµενα ἔχωµεν εἰς ταῦτα ἀναγαγόντες ἐπικρίνειν, καὶ µὴ ἄκριτα πάντα ἡµῖν ᾖ εἰς ἄπειρον ἀποδεικνύουσιν ἢ κενοὺς φθόγγους ἔχωµεν. ἀνάγκη γὰρ τὸ πρῶτον ἐννόηµα καθ᾽ ἕκαστον φθόγγον βλέπεσθαι καὶ µηθὲν ἀποδείξεως προσδεῖσθαι, εἴπερ ἕξοµεν τὸ ζητούµενον ἢ ἀπορούµενον καὶ δοξαζόµενον ἐφ᾽ ὃ ἀνάξοµεν.

“First, Herodotus, you must understand what lies underneath the words, in order that by reference to this we may be in a position to test opinions, inquiries, or problems, so that our proofs may not run on untested ad infinitum, nor the terms we use be empty of meaning. For the primary meaning of each term must be clearly seen, and ought to need no proving; this being necessary, if we are to have something towards which we can refer the question or the problem or the opinion before us” (Diogenes Laertius, Lives of the Philosophers 10.37-38).

  • 29 Sedley (1973), 20, translates the first and second of these as “meaning”.

13Here, we see references to “the things placed under words”, “the primary concept”, and “empty words” – all of which would seem to suggest that there is a third, mediating entity between words and things.29

  • 30 On this, cf. Long (1971), 121; Glidden (1983), 191.
  • 31 Sedley (1973), 20.
  • 32 See Manetti (1993), 122-123.

14If this is in fact the case, there would seem to be similarities between the Epicurean theory and that of Aristotle.30 In On Interpretation, Aristotle takes the affections in the soul to be derived from objects in the world via the senses in much the same way as Epicurus takes προλήψεις to arise from external objects. Still, there are points of contrast between the two:31 (a) for Aristotle, the affections that arise in the soul from perceiving external things are the same for all races – it is only the relationship between the impressions and speech that is conventional – whereas for Epicurus differences in language are the result of differences in the environment. In addition, (b) Epicurus considers at least some of the ways that language interacts with the world to be determined by nature. While Aristotle had stated that the relationship between language and the world was conventional, by virtue of the fact that the link between language and the affections of the soul was conventional, Epicurus suggested that the first words originated through instinctive reactions to the environment. In this last point, Epicurus shows a loose affinity with the naturalist position presented in the Cratylus and the Stoic conception of language that descended from it, although for Epicurus the emphasis in determining correctness in speech is not on producing etymologies to get back to the πρῶται φωναί, since these are obscured by the metaphorical transferences that occurred in the period after the initial stage of naming, but rather to grasp the πρῶτον ἐννόηµα (“the primary meaning/concept”) associated with the name (Diogenes Laertius, Lives of the Philosophers 10.37-38), which may or may not be identical to a πρόληψις.32

15To sum up: the semiotic triangles assembled by Aristotle, the Stoics, and possibly the Epicureans set an entity between words and their referents: for Aristotle, this consisted in affections (“παθήµατα”) in/of the soul, for the Stoics it was “τὸ σηµαινόµενον πρᾶγµα, καὶ λεκτόν”, and for the Epicureans some scholars suggest that the role may have been taken up by προλήψεις. The interpretation of the theories in which these expressions are housed is difficult, due in large part to the fragmentary or otherwise incomplete nature of the texts, and the debate surrounding them continues.

Notes

1 Mates (1973), Long (1974), Long & Sedley (1987), 1.97-101, 183-236; Manetti (1993), Sluiter (1990), Sluiter (1997), Schenkeveld & Barnes (1999), Allan (2010), and de Jonge & van Ophuijsen (2010) provide introductions for those interested in various aspects of the topic. For a concise introduction to theories of meaning in the ancient world, see Bächli & Graeser (2000), 34-39.

2 See Ademollo (2011); cf. Barney (2001); Sedley (2003); Ewegen (2014). There are a number of important essays on the Cratylus contained in Schofield & Nussbaum (1982). For a concise account of the Cratylus, see Modrak (2001), 14-19. On the translation of “θέσις” as “imposition”, see e.g. Allen (2005), 18-19.

3 For the argumentative positions in the Cratylus, see Kretzmann (1971), 126, and Long (2005), 43.

4 Kahn (1973), 168:“Thus the Cratylus urges us to study not linguistics but ontology”.

5 On meaning in Aristotle, see Modrak (2001), especially 13-51. For investigations of this area of Aristotle’s thought, see Charles (2000) and Noriega-Olmos (2013).

6 Kretzmann (1974), 3, describes this excerpt as “the most influential text in the history of semantics”. For Aristotle’s “footprints in the linguist’s garden”, see Allan (2008).

7 At On Sophistical Refutations 165a7-8 we read that names are symbols in the place of things: “...τοῖς ὀνόµασιν ἀντὶ τῶν πραγµάτων χρώµεθα συµβόλοις (“we employ names as symbols in the place of the things [themselves]”).

8 Aristotle of course adds a fourth entity – written words. See Manetti (1993), 55, who suggests that Plato implied the triangle in his explication of the relationship between name and object in texts such as the Phaedrus and Cratylus: name-δύναµις-thing. Manetti (1996), 15, provides a chart of different “semiotic triangles” from antiquity onwards.

9 Cited by Lyons (1977), 96.

10 This is, however, the only apparent reference to Plato in On Interpretation. See Kretzmann (1974), 13; Ademollo (2011), 114.

11 See Ackrill (1963), 113-114; Long (1971), 121.

12 Irwin (1982), 242, argues that the beginning of On Interpretation engages with signification rather than meaning: “To have a concept of meaning and to have a word meaning ‘mean’ is not yet to have a theory of meaning”.

13 On the development of the term “σύµβολον” (“symbol”), see Struck (2004).

14 See Kretzmann (1974), 5.

15 For compact discussions of Stoic semantics, see Long & Sedley (1987), 1.199-202; Sluiter (1990), 22-36.

16 On the Stoics’ relationship to the naturalist position of Plato’s Cratylus, see Long (2005); Allen (2005).

17 See Barwick (1957); cf. Blank (1982), especially 21-22.

18 Augustine also uses the terms “stirps” (“trunk”) and “sementum” (“seed”) of words at De Dialectica 6.

19 Cf. “perspicis enim haec uerba ita sonare ut ipsae res quae his uerbis significantur” (“you see that these words sound like the very things that are signified by the words” Augustine, De Dialectica 6).

20 As Augustine states things, “uis” cannot be traced back to any more primitive word: “reddetur ratio, quod robusto et quasi ualido sono uerbum rei quam significat congruit. ultra quod requirat non habet” (“the reason is this: that the word accords with the thing that it signifies on account of its robust and strong, so to speak, noise. There is nothing further that can be stated” Augustine, De Dialectica 6).

21 See Manetti (1993), 92-110. For a neat comparison, see Long & Sedley (1987), 1.201.

22 The quality of this excerpt as a guide to Stoic thought has been questioned; see Frede (1994).

23 Frede (1994); cf. Graeser (1978), 88-90.

24 Cf. Schenkeveld & Barnes (1999), 210.

25 A standard interpretation of Epicurus’ theory of the origin of language can be found at Sedley (1973), 18.

26 Cf. Allen (2005), 28.

27 Cf. Sedley (1973), 18. For Plato’s own position on foreign languages, cf. Cratylus 389d-390a with Kretzmann (1971), 129-130: “That one and the same model correct name should appear as “hippos” among the Greeks and “cheval” among the barbarians is of no more consequence than that the shuttle for weaving thick woolen cloaks should be painted red in this weaver’s shop and yellow in that one”.

28 On this interpretation, see Long (1971), 119-123; Manetti (1993), 119-121; for an opposing point of view, see Glidden (1983).

29 Sedley (1973), 20, translates the first and second of these as “meaning”.

30 On this, cf. Long (1971), 121; Glidden (1983), 191.

31 Sedley (1973), 20.

32 See Manetti (1993), 122-123.

© C.H.Beck, 2016

Conditions d’utilisation : http://www.openedition.org/6540