We nevertheless lay it down in our grammars, that adjectives in f form their feminine by ‘changing’ f into ve. The correspondence of sounds which we thus register, though it is a consequence of phonetic development, does not, strictly speaking, express a law of sound-change; we might call it ‘a law of sound-correspondence’ or ‘sound-interchange.’ The ‘law of sound-interchange’ states in a convenient form the aggregate results of events which have occurred in accordance with some ‘law of sound-change.’ Our question, then, refers to the ‘laws of sound-change’ proper, and not to those of ‘sound-interchange;’ and if we say that a law of sound-change admits of no exceptions, we can only mean that, within the limits of some definite language or dialect, all cases which fulfil the same phonetic conditions have had the same fate: i.e. the same sound must there have changed into the same other sound throughout the language, or, where various sounds are seen to replace one and the same other sound of the older language, the cause for this difference must be sought in the difference of phonetic conditions, such as accent, contact with or proximity to other sounds, etc.
It must be clear, after all that has been said in this chapter, that laws of sound-change, in the correct meaning of this term, must be consistent and absolutely regular. As regards the case of the individual speaker, we have seen that the utterance of each sound depends on the memory-picture of motion and position, and that these pictures exert their influence without the speaker being conscious of it. It will then naturally follow that if these pictures alter gradually in the case of any one sound in any one word, they will do so for the same sound in all other cases where it occurs under like conditions.
It is indeed often stated that the sense of etymological connection of a particular word with others which retain a certain sound unaltered may prevent that sound from taking the same course in that word as it does in other words not so influenced; but the existence and efficacy of some counteracting influence does not disprove the existence of the force against which it operates, and which it overcomes or neutralizes. Nor, again, could the inter-communication between the individual speakers cause occasional suspension of the law of sound-change.
We have seen that the association which arises between memory-pictures of the sound, and of the motion of our vocal organs, etc., for its utterance, is—though but external—nevertheless very close, and that it soon becomes indissoluble. The slight and gradual changes in the utterance of the surrounding speakers alter the memory-pictures of the sound, and the corresponding memory-picture of motion and position follows in the same way. It is, then, only in case of mixture of dialect, i.e. when a considerable group of speakers of one dialect becomes mixed and scattered among speakers of another, that the following generation may adopt one sound from the one dialect and another from the second; thus apparently exhibiting the differentiation of the same sound, under the same phonetic circumstances, into two, of which the one appears as the rule, the other as the exception. But then, again, such a case—though when it has happened we may not always be aware of it, and consequently may not always be able to assign the phenomenon to its true cause—does not prove that the law of sound-change admitted of exception. We merely have the results of two such laws mixed and confused.
CHAPTER IV.
CHANGE IN WORD-SIGNIFICATION.
Sound-change is brought about by the repeated substitution of a sound or sounds almost imperceptibly differing from the original. The A.S. hláfmesse is now represented by the English Lammas: though the mm sound is clearly easier to pronounce than the combination represented by fm, generations passed away before the word as we have it in English became the recognised form. In the case of sound-change, we must notice that the rise of the new sound is simultaneous with the disappearance of the old one. In the case of change of signification, it is possible for the old meaning to be maintained by the side of the new one; as when we speak of ‘the House,’ meaning the House of Parliament, we do not exclude the original and proper meaning of the word, but we merely narrow and define its signification. Indeed, change in signification consists invariably in a widening or narrowing of the extent of the signification, corresponding to which we find an impoverishment or an enrichment of the contents. As we saw that the employment of ‘House’ to denote the House of Parliament implied a narrowing or specialising of the extent of the signification of the ordinary meaning of house, so we may take a word like moon, properly and originally applied only to the earth’s satellite, and apply it to a whole class, which we regard in some way as resembling it, as when we speak of Jupiter’s moons. In this case we widen the application of the word by narrowing its contents, but even when thus widened the meaning still includes its original denotation. Frequently such a widened application becomes once more narrowed, by the widening of the contents: an instance of this double process we have, e.g., in the word crane.[6] Originally only meaning the bird of that name, it was, by a metaphor, applied to a class of objects similar in some respects to the bird. A process of narrowing this application led to the use of the word as a specific name for a certain machine. The word, in this sense, no longer includes its original meaning, and is transferred. It is only by such a succession of widening and narrowing that a word can assume a signification absolutely different from its original meaning. This transference may be more or less occasional, or become usual. Thus in the case of green for unripe (cf. blackberries are red when they are green) the meaning is in a certain sense an ‘occasional’ one, the real and original meaning being still clearly felt. This original meaning is, however, quite lost sight of when we use grain in to dye in the grain, for ‘to dye of a fast colour’ by means of cochineal, etc., grain here being the name given to fibre of wood, etc.[7]
Change in signification, however, has this in common with sound-change, that it is effected by individual usage which departs from the common usage; and that this departure passes only gradually into common usage. Change in signification is a law of language; it is a necessity: and change is rendered possible by the fact that the signification attaching to a word each time it is employed need not be identical with that which usage attaches to it. As we shall have to consider this discrepancy, we shall employ the expressions ‘usual’ and ‘occasional’ signification: and by the ‘usual’ we shall understand the ordinary or general signification; by the ‘occasional’ we shall understand that which the individual attaches to it at the particular moment when he uses the word. The ‘usual’ signification means, as we employ it, the entire contents of any word as it presents itself to a member of any linguistic community: the ‘occasional’ signification means the contents of the conception which the speaker, as he utters the word, connects therewith, and expects the listener to connect with it likewise. The word shade, used by itself and without any interpretation from the context or the situation, would suggest to a hearer its USUAL signification of ‘interruption of light;’ but the individual who employs the word may have in mind, as he may easily disclose, the shade of a tree or a lamp-shade.
The ‘occasional’ signification is commonly richer than the ‘usual’ one in content and narrower in extent. For instance, the word in its occasional sense may denote something concrete: while, in its usual sense, it denotes something abstract only; i.e. some general conception under which different concrete conceptions may be ranged. By a ‘concrete’ conception is here meant something presupposed as actually existing, subject to definite limits of time and space; by an ‘abstract’ one is here meant a general conception, the contents of a mere idea and nothing more, freed from all trammels of time and place. The House of Commons is concrete: a house is abstract. This division has nothing to do with the ordinary division of substantives into abstract and concrete. The substantives which in ordinary grammar we call ‘concrete’ often denote a conception as general as the so-called abstract nouns; as in England’s battles: and, conversely, the latter are occasionally used as what we here call ‘concretes’ when they are used to express a single quality or activity defined by limits of space and time; as, The days of thy youth. In the phrase ‘My horse has run well to-day,’ horse is concrete in the sense which we attach to the term: but in the phrase ‘A horse has four legs,’ it is what we call ‘abstract;’ because the statement does not refer to any one definite concrete horse, but to horses generally, and the predicate therefore is associated with the abstract idea of horse.
The greater number of words can be employed in occasional use in either abstract or concrete significations. There are some words, indeed, essentially concrete, such as thou, thine, he, there, to-day, yesterday;—which, however, need individual application to render them immediately and definitely concrete. Words like I, here, there, serve to define some one’s position in the concrete world; but it requires the aid of other words, or of the circumstances in which they are uttered, to render them thus definite. Even our demonstrative pronouns, and the word the, may be employed to denote abstract conceptions; as, The whale is a mammal; it has warm blood. Pity the widow and the orphan. Even proper names, which we might be inclined at first to take as the type of concrete words, as denoting a single object or person, may be used either ‘usually’ as concrete, or ‘occasionally’ as abstract, since the same name may be borne by various people and various localities, as Newton, Brighton: and, indeed, may be applied to objects named after localities; as Stilton, Champagne, etc. Then there is a small class of words which express an object conceived of as existing once and once only, such as God, devil, world, universe, earth, sun. These nouns are concrete both in their ‘usual’ and in almost all their ‘occasional’ meanings; but even they may be regarded as abstract if regarded from a definite point of view. Indeed, a proper name is essentially concrete; if it becomes abstract, this can only be because it has become a generic name, i.e. because it has become a common noun, a common noun being such in virtue of its standing as the name of each individual of a class or group of things. On the other hand, there are some words which from their very nature are abstract; such are the pronouns ever, any; the Latin quisquam, ullus, unquam, uspiam; but the abstract character even of words like these suffers certain limitations in occasional usage; cf. Did he ever (i.e. on any particular occasion) act so, and Should he ever really do it. In these cases ever is in the first instance limited to the past, and in the second to the future.