rus. Naturally this distinction between the two sounds had this effect, that some words in which o was the original vowel-not merely a vulgar corruption-began to be spelt with au in the literary dialect. An example is ausculari (as in Plaut. Merc. 575, ed. Ritschl, and many other places). Here there can be no doubt that osculari is the true form: ostium and other words, derived from the same base os, are never spelt with au. But ausculari became the received form-perhaps on the false analogy of auscultare-to give a fashionable colour to so common a proceeding. Sometimes a false derivation may have helped to bring about the same result-or may itself have been only the result of the new spelling-as in aurichalcum, a word borrowed from the Greek opeixaλkos, and originally written with an o1. The diphthong passes regularly in classical Latin into o in composition. Thus we have suffoco (base fauc), explodo (base plaud). Sometimes it passes into u, as accuso (base causa), defrudo (base fraud). Indeed even frudavi (compare frus-tra) occurs, and this form together with cludo, the proper name Clusius, and others, seems to shew that the change was not confined to compounds3. Somewhat analogous to the change of sound from au to o in Latin is the pronunciation of au in French—and in some parts of the North of England "law" is pronounced like lo. The common pronunciation of au in English is a weakening of another kind. eu. CH. VII. The diphthong eu occurs very rarely in Latin; it was (v) Latin regularly weakened to long u. The few examples—mostly proper names where it occurs are in inscriptions, and have been mentioned in the account of Vowel-Intensification. There are a few occasions in which eu occurs in compounds, as neu from ne-ue, seu from se-ue, neuter from ne-uter, and some others. The two vowels should probably CH. VII. (vi) Latin ou. = be sounded as distinctly as possible; but when sounded The last diphthong has passed through much the same history as eu, except that it lingered later in use. It is often found in the old inscriptions: Loucana is on the tomb of Barbatus, plous and ioubeatis in the letter concerning the Bacchanalia, iouranto in the Bantine table. U begins to appear in the inscriptions of the age of the Gracchi'. Thus in the lex Thoria iubeo and ioubeo occur indifferently; iudex and ioudex; iuro and iouro. Sometimes the o drove out the u, which in such cases had probably become a glide: but after this success it always sank into u at a very early period: thus poplicus occurs frequently in inscriptions, beginning with the Ep. de Bacch., passing in the lex agraria Thoria into publicus. Similarly we find nountios, nontiatus, nuntius: and noundinum (contracted from nouendinum) in the Ep. de Bacch., nondinum in the Tab. Bant., and the common nundinum2. Sometimes the o weakened itself into u: so that the diphthong passed through the stages ou, uu, and then u as before. Thus souos, which occurs in the beautiful epitaph of Claudia, quoted by Mommsen, passed to suuos, and that to suos and suus. So also occur flouios, fluuios, and flu-ere. Roby* assigns to ou the sound of "Southern English” ō, a diphthong formed of o and u. He probably means the ō with a u-glide. But in this case I should have expected the Latin diphthong to have passed into o rather 1 Lucios on the tombs of Barbatus and his son, which Corssen gives as examples of the weakening at a still earlier period, is more probably from Leucios. 2 Corssen, 1. 670. 3 Rom. Hist. I. p. 60, Eng. trans. Souom mareitom corde deilexit souo. Grammar, p. 81. than into u: but it does pass into u most regularly. I therefore think that its sound was probably very near u, like the Greek ov1. The following table gives the results of our discussion of the probable sound of the diphthongs: the new sounds being those to which the Greek and Latin languages were respectively tending. The English equivalents are given, as before, in brackets, as nearly as possible. CH. VII. The full u of the Graeco-Italian was retained by the 4. U=uLatin peoples, but weakened by the Greeks perhaps to the psilon. sound of the German ü. This is shewn, first by the fact that when the Greeks transliterated the Roman u, they employed not v but ou, which, as we have already seen, approached very nearly, if not quite, to u: secondly, and more conclusively, by the variation of practice amongst the Romans in transliterating Greek words. In the early time of Roman intercourse with Greece, they were content to employ the best equivalents for Greek sounds which their language afforded; accordingly they employed u to denote v, as in Burrus, i.e. Pyrrhus. But in the last century of the Republic, when the respect of the Romans 1 See p. 230. CH. VII. for Greek literature had greatly increased, they were not content with this rough and inexact representation: and therefore they borrowed the symbol T as well as the sound. I do not know that any authority before Cicero mentions this borrowing: it dated from his lifetime. The sound of v is not known with exactness; but it is certainly a modified u, and cannot have differed greatly from the German ue or ü, that is, as we have seen, a sound between i and u, having a front position of the tongue, like i, but rounded like u. Only one Greek people, the Boeotians, retained the full sound in its original place, i. e. in those words whose corresponding forms in other dialects are spelt with v; but even they denote that sound by the symbol ov, like the other Greeks. Thus they wrote yλovкou for yλuxu, but the quantities are not different; so that the sound of ou, in Boeotia at least, cannot then have been double. In inscriptions we find ἀσουλία, τούχα (i.e. τύχη), σουν, and σουγγράφως, Διονούσιος'; these are all Theban; κάρουξ, Κουζικηνός, Μουρίνα are on a list of victors at the Xapiтeioia from Orchomenos2; but in the next inscription of the same class and from the same place, we have the usual forms; Boeckh dates it Ol. 145. As a rule, the full peculiarities are found only in Theban inscriptions. In fragments of Corinna we have οὐμές and οὐμίων (i. e. ὑμῶν) θουγάτειρ, οὐψιβίας, ὠνούμηνεν, τού which in common Boeotian is τούν) identical in sound as in meaning with Latin tu. According to Hesychius the full u was kept in Laconia also: he gives káρova, ovdpaívw, and others. But there is no appearance of it upon inscriptions; neither is it found in the fragments of Alkman. Probably, therefore, Hesychius confused Boeotian with Laconian forms: the two 3 In Oropian inscriptions the v is found constantly, no doubt because of the close connection of the place with Attica : e.g. άργυρος, συνήγορος &c. in 1566 and 1569 c. The same explanation however will hardly apply to those of Lebadea where we have Πυθόνικος (1571), Ολύμπιος (1575), or to those of Orchomenos (1579, 1580, &c.). dialects have much in common in detail, but not very CH. VII. much in principle. any This weaker u differs from the full one, not in alteration of the action of the lips, but from a different position of the tongue, which is allowed to come further forward in the mouth: the variation therefore is in the direction which all simply weakened articulation takes. The same vowel is much affected in many languages as well as in Greek: the French u is a parallel example; in une, according to Mr Bell, the vowel is a mid-front-wide-round: that is, the tongue is more nearly in the position for e, about half way between that for u (original) and that for ü in un he thinks that the vowel is not rounded at all, i.e. there is no motion of the lips, but the point of the tongue acts as well as the back1. In England also u has lost its true character in the great majority of words in which it occurs, e.g. in but, shut: there is no rounding of the lips at all: the vowel differs from the French in the simpler position of the tongue: the back of it only is called into use, not the point. This weakening indeed is principally confined to the south; Cumberland is still pronounced by natives with the full vowel sound which we denote by oo: the tendency however (as is always the case) is progressive: it is not long since Russia and Prussia were called in England Roossia and Proossia: and already we sometimes hear put pronounced like but. Parallel affections of u in other lan guages. 5. Further (sporadic) Vowel-substitution in Greek. We have seen that in Greek the original a is regularly broken up into a, e, and 0; and that u is weakened into v orü. The sound of the original i remained unaltered. Beyond this there was little variation in the main body of 1 Technically, it is a "mid-mixed-primary" vowel, like that of que (Fr.). See pages 81-83. 5. Spora dic change in Greek. |