In the discussion referred to above Sweet was seconded by Lecky, who said that “assimilations vastly multiplied the number of elementary sounds in a language, and therefore could not be described as facilitating pronunciation.” This is a great exaggeration, for in the vast majority of instances assimilation introduces no new sounds at all (see, for instance, the lists in my LPh ch. xi.). Lecky was probably thinking of such instances as when [k, g] before front vowels become [tʃ, dʒ] or similar combinations, or when mutation caused by changes [u, o] into [y, ø], which sounds were not previously found in the language. Here we might perhaps say that those individuals who for the sake of their own ease introduced new sounds made things more difficult for coming generations (though even that is not quite certain), and the case would then be analogous to that of a man who has learnt a foreign expression for a new idea and then introduces it into his own language, thus burdening his countrymen with a new word instead of thinking how the same idea might have been rendered by means of native speech-material—in both cases a momentary alleviation is obtained at the cost of a permanent disadvantage, but neither case can be alleged against the view that the prevalent tendency among human beings is to prefer the easiest and shortest cut.
XIV.—§ 8. Extreme Weakenings.
When this lazy tendency is indulged to the full, the result is an indistinct protracted vocal murmur, with here and there possibly one or other sound (most often an s) rising to the surface: think, for instance, of the way in which we often hear grace said, prayers mumbled and other similar formulas muttered inarticulately, with half-closed lips and the least possible movement of the rest of the vocal organs. This is tolerated more or less in cases in which the utterance is hardly meant as a communication to any human being; otherwise it will generally be met with a request to repeat what has been said, the social curb being thus applied to the easygoing tendencies of the individual. Now, as a matter of fact, there are in every language a certain number of word-forms that can only be explained by this very laziness in pronouncing, which in extreme cases leads to complete unintelligibility.
Russian sudar’ (gosudar’), ‘sir,’ is colloquially shortened into a mere s, which may in subservient speech be added to almost any word as a meaningless enclitic. And curiously enough the same sound is used in exactly the same way in conversational Spanish, as buenos for bueno ‘good,’ only here it is a weakening of señor (Hanssen, Span. gramm. 60): thus two entirely different words, from identical psychological motives, yield the same result in two distant countries. Fr. monsieur, instead of [mɔ̃sjœ·r], as might be expected, sounds [mɔsjø] and extremely frequently [msjø] and even [psjø], with a transition not otherwise found in French. Madame before a name is very often shortened into [mam]; in English the same word becomes a single sound in yes’m. The weakening of mistress into miss and the old-fashioned mas for master also belong here, as do It. forms for signore, signora: gnor si, gnor no, gnora si, sor Luigi, la sora sposa, and Sp. usted ‘you’ for vuestra merced. Formulas of greeting and of politeness are liable to similar truncations, e.g. E. how d(e) do, Dan. [gda’] or even [da’] for goddag, G. [gmɔ̃in, gmɔ̃] for guten morgen, [na·mt] for guten abend; Fr. s’il vous plaît often becomes [siuplɛ, splɛ], and the synonymous Dan. vær så god is shortened into værsgo, of which often only [sgo’] remains. In Russian popular speech some small words are frequently inserted as a vague indication that the utterance or idea belongs to some one else: griu, grit, grim, gril, various mutilated forms of the verb govorit’ ‘say,’ mol from molvit’ ‘speak,’ de from dejati (Boyer et Speranski, Manuel 293 ff.); cp. the obsolete E. co, quo, for quoth. In all the Balkan languages a particle vre is extensively used, which Hatzidakis has explained from the vocative of OGr. mōrós. Modern Gr. thà is now a particle of futurity, but originates in thená, from thélei, ‘he will’ + nà from hína, ‘that.’ These examples must suffice to show that we have here to do with a universal tendency in all languages.
XIV.—§ 9. The Principle of Value.
To explain such deviations from normal phonetic development some scholars have assumed that a word or form in frequent use is liable to suffer exceptional treatment. Thus Vilhelm Thomsen, in his brilliant paper (1879) on the Romanic verb andare, andar, anar, aller, which he explains convincingly from Lat. ambulare, says that this verb “belongs to a group of words which in all languages stand as it were without the pale of the laws, that is, words which from their frequent employment are exposed to far more violent changes than other words, and therefore to some extent follow paths of their own.”[61] Schuchardt (Ueber die lautgesetze, 1885) turned upon the ‘young grammarians,’ Paul among the rest, who did not recognize this principle, and said that one word (or one sound) may need 10,000 repetitions in order to be changed into another one, and that consequently another word, which in the same time is used only 8,000 times, must be behindhand in its phonetic development. Quite apart from the fact that this number is evidently too small (for a moderately loquacious woman will easily pronounce such a word as he half a dozen times as often as these figures every year), it is obvious that the reasoning must be wrong, for were frequency the only decisive factor, G. morgen would have been treated in every other connexion exactly as it is in guten morgen, and that is just what has not happened. Frequency of repetition would in itself tend to render the habitude firmly rooted, thus really capable of resisting change, rather than the opposite; and instead of the purely mechanical explanation from the number of times a word is repeated, we must look for a more psychological explanation. This naturally must be found in the ease with which a word is understood in the given connexion or situation, and especially in its worthlessness for the purpose of communication. Worthlessness, however, is not the moving power, but merely the reason why less restraint than usual is imposed on the ever-present inclination of speakers to minimize effort. A parallel from another, though cognate, sphere of human activity may perhaps bring out my point of view more clearly. The taking off of one’s hat, combined with a low bow, served from the first to mark a more or less servile submissiveness to a prince or conqueror; then the gesture was gradually weakened, and a slight raising of the hat came to be a polite greeting even between equals; this is reduced to a mere touching of the hat or cap, and among friends the slightest movement of the hand in the direction of the hat is thought a sufficient greeting. When, however, it is important to indicate deference, the full ceremonial gesture is still used (though not to the same extent by all nations); otherwise no value is attached to it, and the inclination to spare oneself all unnecessary exertion has caused it to dwindle down to the slightest muscular action possible.
The above instances of the truncation of everyday formulas, etc., illustrate the length to which the ease principle can be carried when a word has little significatory value and the intention of the speaker can therefore be vaguely, but sufficiently, understood if the proper sound is merely suggested or hinted at. But in most words, and even in the words mentioned above, when they are to bear their full meaning, the pronunciation cannot be slurred to the same extent, if the speaker is to make himself understood. It is consequently his interest to pronounce more carefully, and this means greater conservatism and slower phonetic development on the whole.
There are naturally many degrees of relative value or worthlessness, and words may vary accordingly. An illustration may be taken from my own mother-tongue: the two words rigtig nok, literally ‘correct enough,’ are pronounced ['recti 'nɔk] or ['regdi 'nɔk] when keeping their full signification, but when they are reduced to an adverb with the same import as the weakened English certainly or (it is) true (that), there are various shortened pronunciations in frequent use: ['rectnɔg, 'regdnɔg, 'regnɔg, 'renɔg, 'renəg]. The worthlessness may affect a whole phrase, a word, or merely one syllable or sound.
XIV.—§ 10. Application to Case System, etc.
Our principle is important in many domains of linguistic history. If it is asked why the elaborate Old English system of cases and genders has gradually disappeared, an answer that will meet with the approval of most linguists of the ordinary school is (in the words of J. A. H. Murray): “The total loss of grammatical gender in English, and the almost complete disappearance of cases, are purely phonetic phenomena”—supplemented, of course, by the recognition of the action of analogy, to which is due, for instance, the levelling of the nom. and dative plural OE. stanas and stanum under the single form stones. The main explanation thus is the following: a phonetic law, operating without regard to the signification, caused the OE. unstressed vowels -a, -e, -u to become merged in an obscure -e in Middle English; as these endings were very often distinctive of cases, the Old English cases were consequently lost. Another phonetic law was operating similarly by causing the loss of final -n, which also played an important rôle in the old case system. And in this way phonetic laws and analogy have between them made a clean sweep of it, and we need look nowhere else for an explanation of the decay of the old declensions.