CHAPTER IX. CH. IX. Changes produced by want of clearness in pronun ciation. INDISTINCT ARTICULATION. I HAVE now described at some length the changes arising in Greek and Latin from a weak articulation. For example, we have seen how a stronger could be displaced by a weaker sound. This is the simplest instance of absolute weakening. Sometimes, again, we saw that a stronger took the place of a weaker sound, when that sound formed part of a compound which could be pronounced more easily after such change: here, therefore, also there was weakening; a violent contrast of sounds was done away with. In a word, the new sound or new compound was always an easier sound to pronounce under the circumstances. I wish now briefly to consider a different kind of change, caused by what I call indistinct articulation. It is possible to alter a language in another way than by merely substituting an easier for a more difficult sound; in which case the new sound, weaker though it be, is clearly heard. It is possible to pronounce a word, generally through laziness, without sufficient sharpness to give each letter its full and proper sound. In this case no other recognised letter is at first heard; but an indefinite amount of indistinct sound is produced after the letter thus slurred; which in time, if this relaxed pronunciation become common, often takes the form of the nearest sound in the existing alphabet. Thus two letters grow out of one; and a word is often actually increased; and so it may happen that the new form is not really easier to pronounce than the old one. The old saying is here justified, that lazy people give themselves most trouble. It is, I think, unquestionably the desire to save labourto avoid the exertion required to pronounce clearly and distinctly a difficult sound-which produced this change, just as much as it produced substitution and assimilation, as we have already seen. Both kinds of change are due to that one and the same principle which causes all phonetic change: but as the sacrifice of clearness is much greater in this second kind, I see no real economy in it, and believe that laziness was generally its immediate cause1. I have given a few examples of this change from our own language in the first chapter". I now proceed to give some of its more remarkable operations in Greek and Latin. It affects most (as we should naturally expect) the strongest sounds-as the gutturals—or combination of sound, as e.g. sum-sit, causing the insertion of a nonoriginal p; or, lastly, such sounds as were especially difficult to a particular people, as the spirants to the Greeks. I take first the passage of the gutturals in both Greek and Latin into the labials or the dentals. CH. IX. 1. Labialism. 3 K to π and p. This name has been given (first, so far as I know, by Change of Professor Curtius) to the change from K to T and p3. He believes the change to have been produced through the influence of a parasitic u or w (v): k is the hardest of all consonants, as he says, to pronounce, and requires the most distinct articulation to keep the sound pure from subsidiary breaths. If we pronounce it lazily without fully opening the mouth, the result is that together with it a slight w-sound is quite unconsciously pronounced, because the position of the tongue is almost exactly the 2 At p. 14. 3 See Gr. Et. p. 45, &c. 1 See p. 5. CH. IX. Possibility of this change. same for k and g as for w, and if the lips be nearly shut an imperfect labial is necessarily produced: the k or g is followed by a labial after-sound; a "halbvocalischer labialer Nachklang," Corssen calls it': though the sound is at genuine consonant by the definition given at page 57; the mouth-aperture is so nearly closed that no sound can escape through it without rustling or friction; it has nothing of the pure vowel-sound of u. Other imperfect placing of the organs leads to other similar sounds, as y, whence arises Dentalism, which we shall next consider. In order to make this clear, two points must be proved first, that v (that is, w) following after k could change it to p: next, that the v is really almost always adventitious, and not part of an original Graeco-Italian or Indo-European compound. The first must be proved by the exceptions to the second statement. The change from k to p through an intermediate kv, does occur in a few cases where the v is probably Indo-European: thus the possibility of it is proved. On the other hand, it will, I think, be made clear by several examples, taken from within the Latin itself, that this lv (or qu) was commonly later than k. As the first example of original ku we may take the often quoted akva, "a horse." Here the va is the termination: the noun is formed from AK, and the horse is conceived of as "the swift." The v is found in the Sanskrit açva, the Lithuanian aszva, the old Saxon ehu2. By the side of these and the Latin equ-os we cannot doubt that ππоs stands for k-Fo-s; especially as the assimilated form Kкos is preserved in the Etym. Mag.: the has sunk from the Graeco-Italian e. Here the original kv has passed into π in Greek. Rather oddly, the same original form must be assumed for the cognate words in many languages denoting water; Latin aqua, Gothic ahva, and Sanskrit ap or apas, the nominative plural, which alone ΠΟ occurs in classical literature: the word does not seem to occur in any simple Greek form; but Curtius conjectures, with great probability, that it occurs in the name Meσoάπ-o1, the dwellers between the two waters, on the analogy of Μεσοποταμία, Μεθύδριον, Inter-amna, &c. Even the interrogative pronoun must apparently be added to this list, as having, at least, a secondary form-kva as well as ka-before the separation: whence come qui, the Gothic hva, the Sanskrit ku-tas, "whence," &c., and, consequently, the Greek To in TO-OEV, TOLOS (To-yo-s), &c. : but that the simpler form ka still survived is shewn by its use in Sanskrit and Lithuanian, by the middle Ionic кóðeν and κoîos, and by the fact that it was corrupted in a different way to Greek ris and тe, which can come from ka but not from kva. Again, the Latin qui-es, Gothic hvi-la, would seem also to shew a second form kvi by the side of ki, whence reipai: the Greek does not help us, as it never took the secondary form. Sanskrit and Lithuanian agree (at least initially) with the Greek in the forms panchan, penki, and Téμπе: hence we should infer an Indo-European by-form krankan beside the original kankan: the modification of the second k seems to be almost confined to Graeco-Italian. Sufficient examples have been given, I think, to shew that ku when original could pass into a labial; and Grassmann assumes in every case such a compound for the origin of the change3. But these are nearly all the certain examples which can be given of the compound sound occurring in several IndoEuropean languages; and though useful as establishing the possibility of the transition, they are certainly by far too few to prove that the labial always results from an original Indo-European kv. 2 Next we have to shew that the v in other cases sprang merely from phonetic causes, and was not a suffix. This will be sufficiently clear from the cases where kv(qu) is CH. IX. 1 Caused by parasitic V. CH. IX. found in Latin as well as k(c), compared with the corre- 1 Gr. Et. No. 625. 2 Compare also Plaut. Miles G. 1220; cum ipso pol sum secuta: which is altered by Fleckeisen to locuta. 3 See Corssen, 1. 426. |