In the course of time we see languages changing, though the rate of change varies a good deal. But why do languages change? The ultimate cause must be the same as that of the mutability of all human customs and institutions, namely human nature. In the case of languages we are able to specify some causes of change, chief among which are the nature of the speech-material itself (sounds and the ideas to be expressed) and the way in which language is transmitted from generation to generation or rather from older contemporaries to each new individual; a fur ther cause of change is the influence of persons speaking or writ ing other languages.
Some deviations from the norm are of such a momentary and fleeting character that they have no influence whatever ; but others show by the frequency of their occurrence in the mouths of different speakers that they are deeply motived by human nature or by the structure of the particular language in which they oc cur; these will therefore tend to persist and finally become so universal that they can no longer be considered lapses, but form part and parcel of the language in its new stage. These changes then are the chief objects of study for the linguistic historian.
Human laziness plays its part in the phonetic changes of lan guages, which very often follow the line of least resistance ; but there will always be a curbing influence in the mere fact that one speaks to be understood by others who will ask one to repeat one's words if they are not clearly recognized. Here the whole structure of a language comes into consideration, for some in distinct pronunciations are more apt to cause confusion and misunderstanding than others ; in English a great many words are kept apart by distinction of the final voiced and unvoiced stop (b, p; d, t; g, k) ; consequently these sounds are pronounced carefully, while in German, where the same sounds serve only in a few cases to distinguish words. there has not been the same check on the natural tendency to unvoice final consonants, the result being that -b, -d, and -g are pronounced as p, t, k (or ch ) in Grab, Geld, Tag, etc. Whenever a sound is significative in a language it resists change much more effectively than correspond ing sounds which are not used to distinguish words. A correlated
principle is seen in "stump-words," those clippings of words which abound especially in school and college slang, e.g., gym for gym nastics, lab(oratory), undergrad(uate), but which in some cases have penetrated into normal language, e.g., brig for brigantine, cab for cabriolet, photo (cf. also numerous Christian names like Fred). In all such cases the beginning of the word has sufficed to convey the sense, and speakers have therefore accustomed themselves to drop the rest.
In these last mentioned cases the phonetic change affects only one word at a time; but linguistic history abounds in cases in which a sound is changed, not in one word only, but in all the words, or a majority of the words, in which it occurs, or at any rate in which it occurs in the same position or under similar con ditions. Here we speak of "sound-laws" or "phonetic laws," and some of these are of the utmost importance in linguistic science in determining the relationship between various languages and the etymology of particular words.
In dealing with these sound-laws we should never lose sight of the fact that a speech-sound always exists as a part of the whole sound-system or sound-pattern of that language; all the important sound-changes therefore affect more than one sound ; and not infrequently we see that a sound-change is conditioned by the whole structure of a language; similar changes will therefore take place independently in languages of similar structure. This is the case with mutation ("umlaut"), of which there is no trace in Gothic, but which took place in subsequent periods in all the other Germanic languages, though not exactly in the same way every where.
When initial k was dropped in the beginning of knight, all words with the same group of sounds were affected in the same way; know, knowledge, knock, knit, etc., but k was retained after a vowel, acknowledge. Very often two or more sounds possess ing some articulatory elements in common are affected at the same time in the same way; thus g in gn was dropped simultane ously with k in kn; gnaw, gnat. The English vowel-shift which probably began about 1400, modified all long vowels (which had formerly had their "continental" values), e.g., in bite, beet, beat, abate, foul, fool, foal; the vowels were diphthongised or raised, but generally speaking they kept the same distance, so that no confusion was produced and no clashings occurred, until finally the vowel in the beat-group was raised and so fused with that of the beet-group. In the great prehistoric Germanic consonant shift, which affected all the stopped consonants of primitive Aryan, the various sub-classes were similarly kept distinct, though the distinctions were made in a different way from that of the old Aryan system.