Page images
PDF
EPUB

CH. I.

Apparent exceptions to the principle.

retained in the main the sounds, if not the symbols. From this it is obvious that the study of Comparative Philology can never supersede the necessity of thorough investigation of each particular language for itself. Greek and Latin etymology can only be known by historical investigation of the Greek and Latin languages themselves.

As a rule, the weaker sound is later than the stronger. But to this rule there are undoubtedly some exceptions. All such cases will however I believe be found on examination to be less apparent examples of the general lawnamely, that phonetic change arises from the strivingconscious or unconscious-for ease of articulation. A soft letter sometimes changes to a hard from the influence of neighbouring sounds: for example, the g of frag—whence fragor, &c.-is hardened to k in fractus. This of course takes place because it is much harder to articulate a soft consonant and then a hard one immediately afterwards than it is to pronounce two hards together. The principle of assimilation has come in and reversed the common rule of phonetic change; but assimilation itself is an instance of the wider principle. Similarly hiemps would seem to be a stronger form than hiems: and certainly the p is merely phonetic and belongs neither to the base nor to the case-suffix s. But because it is very difficult to sound s immediately after the labial nasal m, in an indistinct less energetic pronunciation of the word, a weak p was heard, to bridge over the difficulty: and this made its way at last into the written word. But the new form though heavier is still easier to sound than the old one. Ease of pronunciation was the reason why fragtus became fractus and hiem-s was increased to hiemps, just as much as it caused the weakening, e. g. of stlites into lites and esam into eram. Sometimes we find that the general endeavour for easier pronunciation takes the form of striving after greater distinctness of sound, and so has the effect of strengthening a weaker letter. Thus the Greeks unable to pronounce inμ clearly changed the first aspirate into

Ө

the stronger tenuis. But the change also is based on the fact that Tienu is an easier word to pronounce than Oionμi. Taken by itself 7 requires more effort to pronounce than : the check is more complete in pronouncing than in pronouncing (i.e. t'h, where the h is due. to a portion of the breath being allowed to escape before the t is fully sounded). But when occurs at the beginning of two consecutive syllables, a greater effort is required to place the organs of speech twice in the necessary position for producing it. All these and other apparent exceptions arising from assimilation and dissimilation of sounds, or from indistinct articulation, will be fully described in their proper places.

CH. I.

theories

upon this

A different cause has been assigned for certain varia- Different tions of sound by Prof. Max Müller in his valuable lectures on the Science of Language: he supposes an originally subject. indistinct sound, capable of passing into different forms in different languages or different dialects of the same language. In the fourth lecture of his second series, he gives several examples of "phonetic degeneracy:" and he says (p. 176) that the principal cause of this is "when people attempt to economize their breath and muscular energy." But beside this cause of variation, and distinct from it, he mentions another, which he calls "Dialectic Growth" (p. 180). By this he accounts for the phonetic diversity which is seen e.g. in the Sanskrit gharma, Greek depμó-s, Latin formus—all undoubtedly modifications of one IndoEuropean word meaning hot. These forms, he thinks, point to "a previous state of language, in which, as in the Polynesian dialects, the two or three principal points of consonantal contact were not yet felt as definitely separated from each other." Thus in the instance given above, the three forms were received by the three languages from some earlier stage, in which the articulation of the original word was so vague that it might take any one of the forms mentioned. This is possible, nor is the theory confuted by the a priori objection made to it by Prof. Cur

CH. I.

πεντε

tius', that such indistinctness of sound is inconsistent with
the strong articulation which peculiarly belongs to the oldest
languages. But there seems to me more weight in his
question, what the sound could have been which was capa-
ble of such strange variation. The numeral five is ex-
pressed by panchan in Sanskrit, Tévre in common Greek,
πéμπe in Aeolic, quinque in Latin, pomtis in Oscan, fimf
in Gothic, penki in Lithuanian. What can the two con-
sonantal sounds have originally been which could be
strengthened or weakened in so many ways? Prof. Müller
speaks of "phonetic idiosyncrasies" in particular lan-
guages: which seems to me only another title for weak-
nesses of articulation become hereditary by transmission
from one generation to another. But he allows 2 that
"these idiosyncrasies are quite inadequate to explain why
the Latin coquo should in Greek appear as TÉTTW." Pro-
fessor Curtius thinks that the change from original k to p
as in πеπτш, or from k to t as in Tɩs (Sk. kis, Lat. quis) is
to be explained by the involuntary springing up of para-
sitic sounds: thus that a u or v by relaxed articulation
was sounded after the k-as it actually did spring up in
the Latin, e.g. ting-u-o (Gr. Téyyw); and we may hear
similar cases of relaxed articulation in England, e.g. ne-a
for nay, and ge-ate for gate in Cumberland: and fi-ound
for found, &c. in Suffolk-then this labial v by degrees
corrupted the k to the labial p, and then vanished. Simi-
larly t might arise from k by the mediation of a parasitic
y—thus, k, ky, ty, t: the change from k to t being caused
by just the same indistinct articulation which in England
causes us often to hear tloth, and not cloth, and dlory not
glory: though Prof. Max Müller finds it hard to believe
it. These variations are of course not universal, only oc-
casional; it is only comparatively a small number of words
in which the Attic has weakened a k, which the Doric has
retained, to t or p: similarly the Doric has suffered change
in some roots as well as the Attic: √/Feπ (orig. VAK) is "to
1 Gr. Et. p. 380, note.

2

p. 184.

speak" in Doric as well as in Attic. I think that the theory given above is sufficient to explain most of the cases and thus they are all instances of a weakening tendency, gradually affecting different dialects and languages, and resisted by them in proportion to the firmness of their articulation; affecting for example the Doric least, the Attic considerably, the Aeolic (compare πéμπe and Tíoupes1 with the Attic Tévte and téoσapes) most of all; leaving the Latin untouched, but attacking the Oscan severely2.

1

CH. I.

Results of nation of phonetic laws of different

the combi

Many apparent and some real exceptions to this principle will be found in languages which have been largely affected by the introduction of foreign words: and still more where a whole people has adopted the language of another race. Such a people retains its own peculiarities peoples. of pronunciation; it finds in the new language some sounds which are strange to it, and which it cannot pronounce therefore it either drops them altogether, or more probably substitutes for them the nearest of its own,

1 It is not quite certain that πiσupes is Aeolic: it is old Ionic.

2 Dr Donaldson (New Cratylus, § 121) explained this difference of sound as having been produced by "the law of divergent articulations from "the union of the original guttural and labial kp." I am obliged to reject this terrible combination of sounds, because I see no reason to believe that our forefathers possessed much more flexible muscles than we do. He says (ib. § 110) that "the regular series of transitions, which such a combination of the guttural and labial would present, may easily be described: the guttural may be represented by k, q, g,j, s, h, the labial by p, b, v; and these sets of letters may be permuted with each other to any extent." Just before he says, Just before he says, "in those cases where a dental makes its appearance, it must be considered as having arisen by a fault of articulation from the sibilant:" so that t and d must be added to the permutations of the guttural. This is indeed etymology made very easy!

It may perhaps be thought that in these lectures too little reference is made to the works of one of the most active and independent of English philologists. As an old pupil, I should wish to do the fullest honour to the genius, learning, and untiring energy of Dr Donaldson: as such also I cannot but regret the failures in judgment (as they seem to me) which led him either to statements like those quoted above, which would degrade etymology to the mere juggling pastime that it is sometimes held to be, and render scientific treatment of it impossible-or to the wild and groundless ethnological theories which mar the Varronianus. If such theories were given as mere theories, no harm would be done; but they are put on the same footing with inductions as certain as those of any science can possibly be. It is this mixture of the proven and not-proven which must make Dr Donaldson's books unfit for students of comparative philology.

CH. I.

Erroneous spelling to suit supposed ctymologies.

especially if such sounds do not occur in the strange language. Thus old sounds are lost and new ones introduced; and it is quite uncertain whether the new sound will be an easier one than the old. We have a good illustration close at home, the way in which the Keltic ten-. dency to aspirate unaspirated sounds has affected the pronunciation of English in Ireland: e.g. car is sounded like k’har, which is nothing but a Kelto-English variation. The full sound of the a is also retained in Ireland, where we have weakened it in England. Many instances where admixture of race has operated on sound will be found in French, e.g. guêpe, guerre, &c. It is commonly supposed that guêpe is from vespa1, in which case the g must be due to the Frankish pronunciation of a Latin word, for there is no tendency in Latin to a parasitic g before a w. But it is certainly remarkable (as Max Müller has pointed out2) that all these French words beginning with gu can be traced to German words. It is better to suppose that this gu was the attempt of the descendants of the Romans to pronounce a German w, after they had let their own w sound become v, as it now is. But whichever explanation we take, we must recognise the change as resulting from the conflicting phonetic laws of two mixed peoples3.

It is well known that from this operation of a double set of phonetic laws the same word may exist in two forms in the same language: as in English "crab" and "crayfish," the latter being the French écrevisse from German krebiz, krebs. The etymological tendency exhibited in "cray-fish" is the last point which I wish to consider in this chapter. The French form being strange to the English ear, it was converted into something which would convey a meaning. We have here nothing but a highly irregular

[blocks in formation]

3 For the history of the effect of the Teutonic on the Roman languages see Diez, Introduction to the Grammar of the Romance Languages, trans. Cayley, p. 60.

4 Ib. p. 68.

« PreviousContinue »