Sophist, 251a8-c6: ὁ ἀγαθὸς ἄνθρωπος (the good man) and 白馬 (white horse)

Marion, Florian [UCL]

Before coming to the problem of the communication of forms and to the more general and Platonic aporia of the one and the many (in the Sph.: the unity of the subject through the multiplicity of its predications or 'appellations'), the Eleatic Stranger devotes a few words to ‘the late-learners and novices’ who refuse the non-identical predication ‘good man (ἀγαθὸς ἄνθρωπος)’ or the fact that the same item can have a multiplicity of names, i.e. ‘man’ and ‘good’ (Sph. 251a8-c6). Beyond the question of identifying the thinkers here targeted and mocked by Plato (Antisthenes? Euthydemus and Dionysodorus? Megaric dialecticians?), it is clear that the commentators devote little energy to interpreting this passage (exceptions: Crivelli (2012), p. 103-109 and Brown (2019), p. 312-324): either they are content to note that it suffices to discriminate (either by a strong separation à la Frege between these two meanings, or by a mere subordination à la Leśniewki) identity (≈ to be qua identity, symbol =) and predication (≈ to be qua copula, symbol ε) to escape the sophism of the ‘late-learners’, or take advantage of the opportunity to digress on Antisthenes’ philosophy (Heidegger (1924/25), p. 500-511 and De Rijk (1986), p. 113-122), on the relationship between the Euthydemus and the Sophist, or about the Megarics (insofar as Plutarch, Adv. Col. 22.1119cd, 23.1120ab attributes an argument very similar to Sph. 251a8-c6 to the Megaric Stilpo, cf. Opsomer (2013)). Curiously, however, an argument very close to that of the ‘late-learners and novices’ had great posterity in another philosophical ecosystem. Around 300BC, the dialectician of the school of Names Gōngsūn Lón has proposed a reasoning of that kind against the non-identical predication (or ‘nomination’): ‘the white horse (白馬) is not a horse’ is true, because ‘white (白)’ is the name of a color, ‘horse (馬)’ that of a shape, and what names a color is not what names a form (these predicates do not belong to the same kinds). The White Horse Dialogue (白馬論, Báimǎ Lùn), unlike Sph. 251a8-c6, was hotly debated in Chinese philosophical circles: it is mentioned in various Classics (Liezi, Zhuangzi (chap. 2), Mengzi), in the Han Feizi and in the Mozi. The interpretation of the White Horse Paradox is not easy. A popular way among commentators to untie the Gongsunian semantic knot is to consider that Gōngsūn Lón and his (fictional?) opponent play with the ideas of identity and predication (or class-membership): white horse is not identical to horse, although white horse is a member of the class horse (or: horse is predicated of white horse): while Gōngsūn Lón refuses to discriminate the true statement ‘white horse is not identical to horse’ from the false statement ‘horse is not predicated of white horse’, his interlocutor retorts that ‘horse is predicated of white horse’ is true. Another interpretation of more Stilpo-Antisthenian pedigree (see the famous anti-Platonic anecdote reported in Simpl. in Cat. 208.28-3: ‘Plato, I can see the horse, but not horseness’, a similar anecdote is attributed to Stilpo in DL 2.119) has also been defended: Gōngsūn Lón would oppose the idea of horseness to its particular instances (that white horse, that black horse, that brown horse, etc.), i.e. the type to its tokens, the abstract and universal concept to the concrete particulars. According to this interpretation, the sentence ‘the white horse is not a horse’ should be paraphrased as ‘the concrete white horse is not the same as the idea of horseness’. A more plausible variant of this interpretation attributes to Gōngsūn Lón a reflection on the concepts (qua classes, kinds, or sets) of horses and white horses, which are neither co-extensional (i.e. {x: x is a white horse}  {x: x is a horse}) nor synonyms (being a horse does not mean the same as being a white horse). A final interpretation introduces a semantic taxonomy of predicates and classes according to the sorts to which they belong, white horse would be neither horse nor white because these names would all denotate classes belonging to sorts whose domains are disjoint (respectively: color + shape, shape, and color). These interpretations are not necessarily exclusive, reading the entire dialogue invites us to consider that Gōngsūn Lón and his interlocutor have in mind both the difference between identity and predication, and the non-synonymy of two concepts of which the extension of one is a fortiori included in the other. Interpretive charity exhorts us to classify predicates and classes according to the sorts to which they belong. I shall try to show that these interpretations are also admissible for the analysis of the position of the ‘late-learners’: the good man is a non-identical predication rejected because man and good are not synonymous, and they are not synonymous because they are not co-extensional and belong to different sorts. Reading Sph. 251a8-c6 in the light of the White Horse Dialogue is not intended to fall into orientalist exoticism in philosophy, but to provide an exegesis of the Sph. sensitive to the fact that the rejection of non-identical predication (or nomination) constitutes a foil to which several types of answers can be provided: Plato and the Mediterranean tradition find in such a seemingly sophistic riddle an opportunity to distinguish identical or reflexive predication (AεA) from non-identical and non-reflexive predication (AεB), while for the Mohists (Xún Zǐ, part. 3) it is the opportunity to wonder about the way in which the reference of generic composed names is fixed. To do that, they compare the reference of the generic and composed name ‘white horse (白馬)’ (which denotes the class corresponding to the intersection of the horse and white classes, i.e. {x: x is a white horse} = {x: x is a horse} ∩ {x: x is white})) to the reference of the name ‘ox-and-horse (牛馬)’ (which denotes the class corresponding to the union of the ox and horse classes, i.e. {x: x is an ox-and-horse} = {x: x is an ox} ∪ {x: x is a horse}). In many ways, the Mohist reflections are not unattractive for an examination of the reference or denotation of forms resulting from the combination of other forms in Sph. 251d ff.