Ideal of the best system of naming — the Name-Giver ought to be familiar with the Platonic Ideas or Essences, and apportion his names according to resemblances among them.
Herein, so far as I can understand, consists the amount of positive inference which Plato enables us to draw from the Kratylus. Sokrates began by saying that names having natural rectitude were the only materials out of which a language could be formed: he ends by affirming merely that this is the best and most perfect mode of formation: he admits that names may become significant, though loosely and imperfectly, by convention alone — yet the best scheme would be, that in which they are significant by inherent resemblance to the thing named. But this cannot be done until the Name-giver, instead of proceeding upon the false theory of Herakleitus, starts from the true theory recognising the reality of eternal, unchangeable, Ideas or Forms. He will distinguish, and embody in appropriate syllables, those Forms of Names which truly resemble, and have natural connection with, the Forms of Things.
Such is the ideal of perfect or philosophical Naming, as Plato conceives it — disengaged from those divinations of the origin and metamorphoses of existing names, which occupy so much of the dialogue.[89] He does not indeed attempt to construct a body of true names à priori, but he sets forth the real nameable permanent essences, to which these names might be assimilated: the principles upon which the construction ought to be founded, by the philosophic lawgiver following out a good theory:[90] and he contrasts this process with two rival processes, each defective in its own way. This same contrast, pervading Plato’s views on other subjects, deserves a few words of illustration.
[89] Deuschle (Die Platonische Sprachphilosophie, p. 57) tells us that in this dialogue “Plato intentionally presented many of his thoughts in a covert or contradictory and unintelligible manner”. (Vieles absichtlich verhüllt oder widersprechend und missverständlich dargestellt wird.)
I see no probability in such an hypothesis.
Respecting the origin and primordial signification of language, a great variety of different opinions have been started.
William von Humboldt (Werke, vi. 80) assumes that there must have been some primitive and natural bond between each sound and its meaning (i.e. that names were originally significant φύσει), though there are very few particular cases in which such connexion can be brought to evidence or even divined. (Here we see that the larger knowledge of etymology possessed at present deters the modern philologer from that which Plato undertakes in the Kratylus.) He distinguishes a threefold relation between the name and the thing signified. 1. Directly imitative. 2. Indirectly imitative or symbolical. 3. Imitative by one remove, or analogical: where a name becomes transferred from one object to another, by virtue of likeness between the two objects. (Ueber die Verschiedenheit des menschlichen Sprachbaues und ihren Einfluss auf die geistige Entwicklung des Menschengeschlechtes, p. 78, Berlin, 1880.)
Mr. Hensleigh Wedgwood, in his Etymology of the English Language (see Prelim. Disc. p. 10 seq.), recognises the same imitative origin, and tries to apply the principle to particular English words. Mr. F. W. Farrar, in his recent interesting work (Chapters on Language) has explained and enforced copiously the like thesis — onomatopœic origin for language generally. He has combated the objections of Professor Max Müller, who considers the principle to be of little applicability or avail. But M. Renan assigns to it not less importance than Mr. Wedgwood and Mr. Farrar. (See sixth chapter of his ingenious dissertation De l’Origine du Langage, pp. 135-146-148.)
“L’imitation, ou l’onomatopée, paraît avoir été le procédé ordinaire d’après lequel les premiers nomenclateurs formèrent les appellations.… D’ailleurs, comme le choix de l’appellation n’est point arbitraire, et que jamais l’homme ne se décide à assembler des sons au hasard pour en faire les signes de la pensée, on peut affirmer que de tous les mots actuellement usités, il n’en est pas un seul qui n’ait eu sa raison suffisante, et ne se rattache, à travers mille transformations, à une élection primitive. Or, le motif déterminant pour le choix des mots a dû être, dans la plupart des cas, le désir d’imiter l’objet qu’on voulait exprimer. L’instinct de certains animaux suffit pour les porter à ce genre d’imitation, qui, faute de principes rationnels, reste chez eux infécond.…
“En résumé, le caprice n’a eu aucune part dans la formation du langage. Sans doute, on ne peut admettre qu’il y ait une relation intrinsèque entre le nom et la chose. Le système que Platon a si subtilement développé dans le Cratyle — cette thèse qu’il y a des dénominations naturelles, et que la propriété des mots se reconnaîlt à l’imitation plus ou moins exacte de l’objet, — pourrait tout au plus s’appliquer aux noms formés par onomatopée, et pour ceux-ci mêmes, la loi dont nous parlous n’établit qu’une convenance. Les appellations n’ont pas uniquement leur cause dans l’objet appelé (sans quoi, elles seraient les mêmes dans toutes les langues), mais dans l’objet appelé, vu à travers les dispositions personnelles du sujet appelant.… La raison qui a déterminé le choix des premiers hommes peut nous échapper; mais elle a existé. La liaison du sens et du mot n’est jamais nécessaire, jamais arbitraire; toujours elle est motivée.”