We must remember that Sokrates in the Kratylus explicitly announces himself as having no formed opinion on the subject, and as competent only to the prosecution of the enquiry, jointly with the others. What he says must therefore be received as conjectures proposed for discussion. I see no ground for believing that he regarded any of them, even those which appear to us the strangest, as being absurd or extravagant — or that he proposed any of them in mockery and caricature, for the purpose of deriding other Etymologists. Because these etymologies, or many of them at least, appear to us obviously absurd, we are not warranted in believing that they must have appeared so to Plato. They did not appear so (as I have already observed) to Dionysius of Halikarnassus — nor to Diogenes, nor to the Platonists of antiquity nor to any critics earlier than the seventeenth century.[46] By many of these critics they were deemed not merely serious, but valuable. Nor are they more absurd than many of the etymologies proposed by Aristotle, by the Stoics, by the Alexandrine critics, by Varro, and by the grammatici or literary men of antiquity generally; moreover, even by Plato himself in other dialogues occasionally.[47] In determining what etymologies would appear to Plato reasonable or admissible, Dionysius, Plutarch, Proklus, and Alkinous, are more likely to judge rightly than we: partly because they had a larger knowledge of the etymologies proposed by Greek philosophers and grammatici than we possess — partly because they had no acquaintance with the enlarged views of modern etymologists — which, on the point here in question, are misleading rather than otherwise. Plato held the general theory that names, in so far as they were framed with perfect rectitude, held embodied in words and syllables a likeness or imitation of the essence of things. And if he tried to follow out such a theory into detail, without any knowledge of grammatical systems, without any large and well-chosen collection of analogies within his own language, or any comparison of different languages with each other — he could scarcely fail to lose himself in wonderful and violent transmutations of letters and syllables.[48]
[46] Dionys. Hal. De Comp. Verbor. c. 16, p. 96, Reiske; Plutarch, De Isid. et Osir. c. 60, p. 375.
Proklus advises that those who wish to become dialecticians should begin with the study of the Kratylus (Schol. ad Kratyl. p. 3, ed. Boiss.).
We read in the Phædrus of Plato (p. 244 B) in the second speech ascribed to Sokrates, two etymologies:— 1. μαντικὴ derived from μανικὴ by the insertion of τ, which Sokrates declares to be done in bad taste, οἱ δὲ νῦν ἀπειροκάλως τὸ ταῦ ἐπεμβάλλοντες μαντικὴν ἐκάλασαν. 2. οἰωνιστικὴ, quasi οἰονοϊστικὴ, from οἴησις, νοῦς, ἱστορία. Compare the etymology of ἔρως, p. 238 C. That these are real word-changes, which Plato believes to have taken place, is the natural and reasonable interpretation of the passage. Cicero (Divinat. i. 1) alludes to the first of the two as Plato’s real opinion; and Heindorf as well as Schleiermacher accept it in the same sense, while expressing their surprise at the want of etymological perspicacity in Plato. Ast and Stallbaum, on the contrary, declare that these two etymologies are mere irony and mockery, spoken by Plato, ex mente Sophistarum, and intended as a sneer at the perverse and silly Sophists. No reason is produced by Ast and Stallbaum to justify this hypothesis, except that you cannot imagine “Platonem tam cæcum fuisse,” &c. To me this reason is utterly insufficient; and I contend, moreover, that sneers at the Sophists would be quite out of place in a speech, such as the palinode of Sokrates about Eros.
[47] See what Aristotle says about Πάντη in the first chapter of the treatise De Cœlo; also about αὐτόματον from αὐτὸ μάτην, Physic. ii. 5, p. 197, b. 30.
Stallbaum, after having complimented Plato for his talent in caricaturing the etymologies of others, expresses his surprise to find Aristotle reproducing some of these very caricatures as serious, see Stallbaum’s note on Kratyl. p. 411 E.
Respecting the etymologies proposed by learned and able Romans in and before the Ciceronian and Augustan age, Ælius Stilo, Varro, Labeo, Nigidius, &c., see Aulus Gellius, xiii. 10; Quintilian, Inst. Or. i. 5; Varro, de Linguâ Latinâ.
Even to Quintilian, the etymologies of Varro appeared preposterous; and he observes, in reference to those proposed by Ælius Stilo and by others afterwards, “Cui non post Varronem sit venia?” (i. 6, 37). This critical remark, alike good tempered and reasonable, might be applied with still greater pertinence to the Kratylus of Plato. In regard to etymology, more might have been expected from Varro than from Plato; for in the days of Plato, etymological guesses were almost a novelty; while during the three centuries which elapsed between him and Varro, many such conjectures had been hazarded by various scholars, and more or less of improvement might be hoped from the conflict of opposite opinions and thinkers.
M. Gaston Boissier (in his interesting Étude sur la vie et les Ouvrages de M. Terentius Varron, p. 152, Paris, 1861) observes respecting Varro, what is still more applicable to Plato:— “Gardons nous bien d’ailleurs de demander à Varron ce qu’exige la science moderne: pour n’être pas trop sévères, remettons-le dans son époque et jugeons-le avec l’esprit de son temps. Il ne semble pas qu’alors on réclamât, de ceux qui recherchaient les étymologies, beaucoup d’exactitude et de sévérité. On se piquait moins d’arriver à l’origine réelle du mot, que de le décomposer d’une manière ingénieuse et qui en gravât le sens dans la mémoire. Les jurisconsultes eux-mêmes, malgré la gravité de leur profession et l’importance pratique de leurs recherches, ne suivaient pas une autre méthode. Trebatius trouvait dans sacellum les deux mots sacra cella: et Labéon faisait venir soror de seorsum, parce que la jeune fille se sépare de le maison paternelle pours suivre son époux: tout comme Nigidius trouvoit dans frater ferè alter — c’est à dire, un autre soi-même,” &c.
Lobeck has similar remarks in his Aglaophamus (pp. 867-869):— “Sané ita J. Capellus veteres juris consultos excusat, mutuum interpretantes quod ex meo tuum fiat, testamentum autem testationem mentis, non quod eam verborum originem esse putarent, sed ut significationem eorum altius in legentium animis defigerent. Similiterque ecclesiastici quidam auctores, quum nomen Pascha a græco verbo πάσχειν repetunt, non per ignorantiam lapsi, sed allusionis quandam gratiam aucupati videntur.”