[7] Humboldt’s Cosmos, ii. 107-109, ed. Sabine.
[8] Philology has been well defined as the cognitio cogniti, and Comparative Grammar, (the branch of Philology which occupies itself with the study of the birth, the development, and the decadence of various languages, together with their divergences and affinities), has deserved the title of Θριγκὸς μαθημάτων φιλολογικῶν, “the coping-stone of philological inquiries.” See Science Comparative des Langues, par Louis Benloew. Paris, 1858.
[9] Thus, though Zend and Sanskrit are the oldest languages of the Indo-European family, they are offsets of an older primitive one. “Among other evidences of this, may be mentioned the changes that words had already undergone in Zend and Sanscrit from the original form they had in the parent tongue; as in the number ‘twenty,’ which being in the Zend ‘visaiti,’ and in Sanscrit ‘vinsaiti,’ shews that they have thrown off the ‘d’ of the original ‘dva,’ two.”—Sir G. Wilkinson in Rawlinson’s Herod. i. p. 280.
[10]Charma, Essai sur le Langage, p. 60.
[11] “Ici comme ailleurs on a commencé par bâtir des systèmes, au lieu de se borner à l’observation de faits.”—Abel Rémusat.
[12] Bunsen, Phil. of Un. Hist. i. 40. The philosophers who held these views were called “Analogists,” while those who leaned to the conventional origin of language were styled “Anomalists.” But Plato and Aristotle admit the existence of both principles, and have written on the subject with a depth of philosophical insight, which, in spite of their defective knowledge, has never been surpassed. See Humboldt’s Cosmos, i. 41, ii. 261.
[13] Plato’s Cratylus, p. 423, et passim; and Schleiermacher’s Introduction. The great authority on the ancient views of philology is Lersch, Sprachphilosophie der Alten. (Bonn, 1838-1841.) The question which agitated the schools was, φύσει τὰ ὀνόματα ἢ θέσει; it was generally decided in favour of the “Analogists,” though often for frivolous reasons. See Aul. Gell. Noct. Att. x. 4. (Renan, p. 137.) Cf. Xen. Mem. iv. 6. 1. Arrian, Epict. i. 17, ii. 10. Marc. Aur. iii. 2; v. 8; x. 8. These views of the mimetic character of words (Arist. Rhet. iii. 1, 2), and their intrinsic connection with things, did not seem to be much disturbed by the fact of the multiplicity of languages, although this fact led Aristotle to place the conventional element first. The very word βάρβαρος implies a lofty contempt for all languages except Greek, and traces of a similar contempt may be found in the vocabulary of many nations. Cf. Timtim, Zamzummim, &c., Renan, p. 178. Pictet’s Origines Indo-Eur. p. 56, seqq. (1 Cor. xiv. 11.)
[14] ὃς ἂν τὰ ὀνόματα ἐπίστηται ἐπίστασθαι καὶ τὰ πράγματα. Plato, Crat. 435, c. In proof that Plato did recognise both elements of language—the absolute and the conventional, see Crat. 435, c., and Philol. Trans. iii. 137. For an able exposition of the Cratylus, see Dr. Donaldson’s New Crat. p. 93, seqq.
[15] Herodot. ii. 2.
[16] Raumer, Gesch. der Hohenstaufen, iii. 491, quoted by Baehr, Herod. l. c. For some other theories on the primitive language, see Cardinal Wiseman’s Lectures on Science, i. 19. Becanus supposed seriously that Low Dutch was spoken in Paradise. Hermathena, lib. ix. p. 204. “That children naturally speak Hebrew,” is one of the vulgar errors which had to be exploded even in the time of Sir T. Browne. Vulg. Err. v. ch. 26. When James IV. of Scotland repeated the experiment of Psammetichus, the infants were shut up with a dumb man, and spoke Hebrew spontaneously! Basque, Swedish, Russ, &c., have all had their advocates. Charma, Essai sur le Langage, p. 242, seqq. Leibnitz, Lettre à M. de Sparvenfeld, § 8.