Language: Its Nature, Development and Origin

XV.—§ 1. Emotional Exaggerations.

In the preceding chapter we have dwelt at great length on those changes which tend to render articulations easier and more convenient. But, important as they are, these are not the only changes that speech sounds undergo: there are other moods than that of ordinary listless everyday conversation, and they may lead to modifications of pronunciation which are different from and may even be in direct opposition to those mentioned or hinted at above. Thus, anger or other violent emotions may cause emphatic utterance, in which, e.g., stops may be much more strongly aspirated than they are in usual quiet parlance; even French, which has normally unaspirated (‘sharp’) [t] and [k], under such circumstances may aspirate them strongly—‘Mais taisez-vous donc!’ Military commands are characterized by peculiar emphasizings, even in some cases distortions of sounds and words. Pomposity and consequential airs are manifested in the treatment of speech sounds as well as in other gestures. Irony, scoffing, banter, amiable chaffing—each different mood or temper leaves its traces on enunciation. Actors and orators will often use stronger articulations than are strictly necessary to avoid those misunderstandings or that unintelligibility which may ensue from slipshod or indistinct pronunciation.[65] In short, anyone who will take careful note of the way in which people do really talk will find in the most everyday conversation as well as on more solemn occasions the greatest variety of such modifications and deviations from what might be termed ‘normal’ pronunciation; these, however, pass unnoticed under ordinary circumstances, when the attention is directed exclusively to the contents and general purport of the spoken words. A vowel or a consonant will be made a trifle shorter or longer than usual, the lips will open a little too much, an [e] will approach [æ] or [i], the off-glide after a final [t] will sound nearly as [s], the closure of a [d] will be made so loosely that a little air will escape and the sound therefore will be approximately a [ð] or a weak fricative point [r], etc. Most of these modifications are so small that they cannot be represented by letters, even by those of a very exact phonetic alphabet, but they exist all the same, and are by no means insignificant to those who want to understand the real essence of speech and of linguistic change, for life is built up of such minutiæ. The great majority of such alterations are of course made quite unconsciously, but by the side of these we must recognize that there are some individuals who more or less consciously affect a certain mode of enunciation, either from artistic motives, because they think it beautiful, or simply to ‘show off’—and sometimes such pronunciations may set the fashion and be widely imitated (cf. below, p. 292).

Tender emotions may lead to certain lengthenings of sounds. The intensifying effect of lengthening was noticed by A. Gill, Milton’s teacher, in 1621, see Jiriczek’s reprint, p. 48: “Atque vt Hebræi, ad ampliorem vocis alicuius significationem, syllabas adaugent [cf. here below, Ch. XX § 9]; sic nos syllabarum tempora: vt, grët [the diæresis denotes vowel-length] magnus, grëet ingens; monstrus prodigiosum, mönstrus valde prodigiosum, möönstrus prodigiosum adeo vt hominem stupidet.” Cf. also the lengthening in the exclamation God!, by novelists sometimes written Gawd or Gord. But it is curious that the same emotional lengthening will sometimes affect a consonant (or first part of a diphthong) in a position in which otherwise we always have a short quantity; thus, Danish clergymen, when speaking with unction, will lengthen the [l] of glæde ‘joy,’ which is ridiculed by comic writers through the unphonetic spelling ge-læde; and in the same way I find in Kipling (Stalky 119): “We’ll make it a be-autiful house,” and in O. Henry (Roads of Destiny 133): “A regular Paradise Lost for elegance of scenery and be-yooty of geography.” I suppose that the spellings ber-luddy and bee-luddy, which I find in recent novels, are meant to indicate the pronunciation [bl·-ʌdi], thus the exact counterpart of the Danish example. An unstressed vowel before the stressed syllable is similarly lengthened in “Dee-lightful couple!” (Shaw, Doctor’s Dilemma 41); American girl students will often say ['di·liʃ] for delicious.

XV.—§ 2. Euphony.

It was not uncommon in the seventeenth and eighteenth centuries to ascribe phonetic changes to a desire for euphony, a view which is represented in Bopp’s earliest works. But as early as 1821 Bredsdorff says that “people will always find that euphonious which they are accustomed to hear: considerations of euphony consequently will not cause changes in a language, but rather make for keeping it unchanged. Those changes which are generally supposed to be based on euphony are due chiefly to convenience, in some instances to care of distinctness.” This is quite true, but scarcely the whole truth. Euphony depends not only on custom, but even more on ease of articulation and on ease of perception: what requires intricate or difficult movements of the organs of speech will always be felt as cacophonous, and so will anything that is indistinct or blurred. But nations, as well as individuals, have an artistic feeling for these things in different degrees, and that may influence the phonetic character of a language, though perhaps chiefly in its broad features, while it may be difficult to point out any particular details in phonological history which have been thus worked upon. There can be no doubt that the artistic feeling is much more developed in the French than in the English nation, and we find in French fewer obscure vowels and more clearly articulated consonants than in English (cf. also my remarks on French accent, GS § 28).

XV.—§ 3. Organic Influences.

Some modifications of speech sounds are due to the fact that the organs of speech are used for other purposes than that of speaking. We all know the effect of someone trying to speak with his mouth full of food, or with a cigar or a pipe hanging between his lips and to some extent impeding their action. Various emotions are expressed by facial movements which may interfere with the production of ordinary speech sounds. A child that is crying speaks differently from one that is smiling or laughing. A smile requires a retraction of the corners of the mouth and a partial opening of the lips, and thus impedes the formation of that lip-closure which is an essential part of the ordinary [m]; hence most people when smiling will substitute the labiodental m, which to the ear greatly resembles the bilabial [m]. A smile will also often modify the front-round vowel [y] so as to make it approach [i]. Sweet may be right in supposing that “the habit of speaking with a constant smile or grin” is the reason for the Cockney unrounding of the vowel in [nau] for no. Schuchardt (Zs. f. rom. Phil. 5. 314) says that in Andalusian quia! instead of ca! the lips, under the influence of a certain emotion, are drawn scoffingly aside. Inversely, the rounding in Josu! instead of Jesu! is due to wonder (ib.); and exactly in the same way we have the surprised or pitying exclamation jøses! from Jesus in Danish. Compare also the rounding in Dan. and G. [nø·] for [ne·, nɛ·] (nej, nein). Lundell mentions that in Swedish a caressing lilla vän often becomes lylla vön, and I have often observed the same rounding in Dan. min lille ven. Schuchardt also mentions an Italian [ʃ] instead of [s] under the influence of pain or anger (mi duole la teʃta; ti do uno ʃchiaffo); a Danish parallel is the frequent [ʃluð’ər] for sludder ‘nonsense.’ We are here verging on the subject of the symbolic value of speech sounds, which will occupy us in a later chapter (XX).

Observe, too, how people will pronounce under the influence of alcohol: the tongue is not under control and is incapable of accurately forming the closure necessary for [t], which therefore becomes [r], and the thin rill necessary for [s], which therefore comes to resemble [ʃ]; there is also a general tendency to run sounds and syllables together.[66]

XV.—§ 4. Lapses and Blendings.

All these deviations are due to influences from what is outside the sphere of language as such. But we now come to something of the greatest importance in the life of language, the fact, namely, that deviations from the usual or normal pronunciation are very often due to causes inside the language itself, either by lingering reminiscences of what has just been spoken or by anticipation of something that the speaker is just on the point of pronouncing. The process of speech is a very complicated one, and while one thing is being said, the mind is continually active in preparing what has to be said next, arranging the ideas and fashioning the linguistic expression in all its details. Each word is a succession of sounds, and for each of these a complicated set of orders has to be issued from the brain to the various speech organs. Sometimes these get mixed up, and a command is sent down to one organ a moment too early or too late. The inclination to make mistakes naturally increases with the number of identical or similar sounds in close proximity. This is well known from those ‘jaw-breaking’ tongue-tests with which people amuse themselves in all countries and of which I need give only one typical specimen:

If the mind is occupied with one sound while another is being pronounced, and thus either runs in advance of or lags behind what should be its immediate business, the linguistic result may be of various kinds. The simplest case of influencing is assimilation of two contiguous sounds, which we have already considered from a different point of view. Next we have assimilative influence on a sound at a distance, as when we lapse into she shells instead of sea shells or she sells; such is Fr. chercher for older sercher (whence E. search) from Lat. circare, Dan. and G. vulgar ʃerʃant for sergeant; a curious mixed case is the pronunciation of transition as [træn'siʒən]: the normal development is [træn'ziʃən], but the voice-articulation of the two hissing sounds is reversed (possibly under accessory influence from the numerous words in which we have [træns] with [s], and from words ending in [iʒən], such as vision, division). Further examples of such assimilation at a distance or consonant-harmonization (malmsey from malvesie, etc.) may be found in my LPh 11. 7, where there are also examples of the corresponding harmonizings of vowels: Fr. camarade, It. uguale, Braganza, from camerade, eguale, Brigantia, etc. In Ugro-Finnic and Turkish this harmony of vowels has been raised to a principle pervading the whole structure of the language, as seen, e.g., most clearly in the varying plural endings in Yakut agalar, äsälär, ogolor, dörölör, ‘fathers, bears, children, muzzles.’

What escapes at the wrong place and causes confusion may be a part of the same word or of a following word; as examples of the latter case may be given a few of the lapses recorded in Meringer and Mayer’s Versprechen und Verlesen (Stuttgart, 1895): instead of saying Lateinisches lehnwort Meringer said Latenisches ... and then corrected himself; paster noster instead of pater noster; wenn das wesser ... wetter wieder besser ist. This phenomenon is termed in Danish at bakke snagvendt (for snakke bagvendt) and in English Spoonerism, from an Oxford don, W. A. Spooner, about whom many comic lapses are related (“Don’t you ever feel a half-warmed fish” instead of “half-formed wish”).

The simplest and most frequently occurring cases in which the order for a sound is issued too early or too late are those transpositions of two sounds which the linguists term ‘metatheses.’ They occur most frequently with s in connexion with a stop (wasp, waps; ask, ax) and with r (chiefly, perhaps exclusively, the trilled form of the sound) and a vowel (third, OE. þridda). A more complicated instance is seen in Fr. trésor for tésor, thesaurum. If the mind does not realize how far the vocal organs have got, the result may be the skipping of some sound or sounds; this is particularly likely to happen when the same sound has to be repeated at some little distance, and we then have the phenomenon termed ‘haplology,’ as in eighteen, OE. eahtatiene, and in the frequent pronunciation probly for probably, Fr. contrôle, idolatrie for contrerôle, idololatrie, Lat. stipendium for stipipendium, and numerous similar instances in every language (LPh 11. 9). Sometimes a sound may be skipped because the mind is confused through the fact that the same sound has to be pronounced a little later; thus the old Gothonic word for ‘bird’ (G. vogel, OE. fugol; E. fowl with a modified meaning) is derived from the verb fly, OE. fleogan, and originally had some form like *fluglo (OE. had an adj. flugol); in recent times flugelman (G. flügelmann) has become fugleman. It. has Federigo for Frederigo—thus the exactly opposite result of what has been brought about in trésor from the same kind of mental confusion.

When words are often repeated in succession, sounds from one of them will often creep into another, as is seen very often in numerals: the nasal which was found in the old forms for 7, 9 and 10 and is still seen in E. seven, nine, ten, has no place in the word for 8, and accordingly we have in the ordinal ON. sjaundi, átti, níundi, tíundi, but already in ON. we find áttandi by the side of átti, and in Dan. the present-day forms are syvende, ottende, niende, tiende; in the same way OFr. had sedme, uidme, noefme, disme (which have all now disappeared with the exception of dîme as a substantive). In the names of the months we had the same formation of a series in OFr.: septembre, octembre, novembre, decembre, but learned influence has reinstated octobre. G. elf for older eilf owes its vowel to the following zwelf; and as now the latter has given way to zwölf (the vowel being rounded in consequence of the w) many dialects count zehn, ölf, zwölf. Similarly, it seems to be due to their frequent occurrence in close contact with the verbal forms in -no that the Italian plural pronouns egli, elle are extended with that ending: eglino amano, elleno dicono. Diez compares the curious Bavarian wo-st bist, dem-st gehörst, etc., in which the personal ending of the verb is transferred to some other word with which it has nothing to do (on this phenomenon see Herzog, Streitfragen d. roman. phil. 48, Buergel Goodwin, Umgangsspr. in Südbayern 99).

In speaking, the mind is occupied not only with the words one is already pronouncing or knows that one is going to pronounce, but also with the ideas which one has to express but for which one has not yet chosen the linguistic form. In many cases two synonyms will rise to the consciousness at the same time, and the hesitation between them will often result in a compromise which contains the head of one and the tail of another word. It is evident that this process of blending is intimately related to those we have just been considering; see the detailed treatment in Ch. XVI § 6.

Syntactical blends are very frequent. Hesitation between different from and other than will result in different than or another from, and similarly we occasionally find another to, different to, contrary than, contrary from, opposite from, anywhere than. After a clause introduced by hardly or scarcely the normal conjunction is when, but sometimes we find than, because that is regular after the synonymous no sooner.

XV.—§ 5. Latitude of Correctness.

It is a natural consequence of the essence of human speech and the way in which it is transmitted from generation to generation that we have everywhere to recognize a certain latitude of correctness, alike in the significations in which the words may be used, in syntax and in pronunciation. The nearer a speaker keeps to the centre of what is established or usual, the easier will it be to understand him. If he is ‘eccentric’ on one point or another, the result may not always be that he conveys no idea at all, or that he is misunderstood, but often merely that he is understood with some little difficulty, or that his hearers have a momentary feeling of something odd in his choice of words, or expressions or pronunciation. In many cases, when someone has overstepped the boundaries of what is established, his hearers do not at once catch his meaning and have to gather it from the whole context of what follows: not unfrequently the meaning of something you have heard as an incomprehensible string of syllables will suddenly flash upon you without your knowing how it has happened. Misunderstandings are, of course, most liable to occur if words of different meaning, which in themselves would give sense in the same collocation, are similar in sound: in that case a trifling alteration of one sound, which in other words would create no difficulty at all, may prove pernicious. Now, what is the bearing of these considerations on the question of sound changes?

The latitude of correctness is very far from being the same in different languages. Some sounds in each language move within narrow boundaries, while others have a much larger field assigned to them; each language is punctilious in some, but not in all points. Deviations which in one language would be considered trifling, in another would be intolerable perversions. In German, for instance, a wide margin is allowed for the (local and individual) pronunciation of the diphthong written eu or äu (in eule, träume): it may begin with [ɔ] or [œ] or even [æ, a], and it may end in [i], or the corresponding rounded vowel [y], or one of the mid front vowels, rounded or not, it does not matter much; the diphthong is recognized or acknowledged in many shapes, while the similar diphthong in English, as in toy, voice, allows a far less range of variation (for other examples see LPh 16. 22).

Now, it is very important to keep in mind that there is an intimate connexion between phonetic latitude and the significations of words. If there are in a language a great many pairs of words which are identical in sound except for, say, the difference between [e·] and [i·] (or between long and short [i], or between voiced [b] and voiceless [p], or between a high and a low tone, etc.), then the speakers of that language necessarily will make that distinction with great precision, as otherwise too many misunderstandings would result. If, on the other hand, no mistakes worth speaking of would ensue, there is not the same inducement to be careful. In English, and to a somewhat lesser degree in French, it is easy to make up long lists of pairs of words where the sole difference is between voice and voicelessness in the final consonant (cab cap, bad bat, frog frock, etc.); hence final [b] and [p], [d] and [t], [g] and [k] are kept apart conscientiously, while German possesses very few such pairs of words; in German, consequently, the natural tendency to make final consonants voiceless has not been checked, and all final stopped consonants have now become voiceless. In initial and medial position, too, there are very few examples in German of the same distinction (see the lists, LPh 6. 78), and this circumstance makes us understand why Germans are so apt to efface the difference between [b, d, g] and [p, t, k]. On the other hand, the distinction between a long and a short vowel is kept much more effectively in German than in French, because in German ten or twenty times as many words would be liable to confusion through pronouncing a long instead of a short vowel or vice versa. In French no two words are kept apart by means of stress, as in English or German; so the rule laid down in grammars that the stress falls on the final syllable of the word is very frequently broken through for rhythmic and other reasons. Other similar instances might easily be advanced.

XV.—§ 6. Equidistant and Convergent Changes.

Phonetic shifts are of two kinds: the shifted sound may be identical with one already found in the language, or it may be a new sound. In the former, but not in the latter kind, fresh possibilities of confusions and misunderstandings may arise. Now, in some cases one sound (or series of sounds) marches into a position which has just been abandoned by another sound (or series of sounds), which has in its turn shifted into some other place. A notable instance is the old Gothonic consonant shift: Aryan b, d, g cannot have become Gothonic p, t, k till after primitive p, t, k had already become fricatives [f, þ, x (h)], for had the shift taken place before, intolerable confusion would have reigned in all parts of the vocabulary. Another instructive example is seen in the history of English long vowels. Not till OE. long a had been rounded into something like [ɔ·] (OE. stan, ME. stoon, stone) could a new long a develop, chiefly through lengthening of an old short a in certain positions. Somewhat later we witness the great vowel-raising through which the phonetic value of the long vowels (written all the time in essentially the same way) has been constantly on the move and yet the distance between them has been kept, so that no confusions worth speaking of have ever occurred. If we here leave out of account the rounded back vowels and speak only of front vowels, the shift may be thus represented through typical examples (the first and the last columns show the spelling, the others the sounds):

When the sound of (2) was raised into [i·], the sound of (1) had already left that position and had been diphthongized, and when the sound of (3) was raised from an open into a close e, (2) had already become [i·]; (4) could not become [æ·] or [ɛ·] till (3) had become a comparatively close e sound. The four vowels, as it were, climbed the ladder without ever reaching each other—a climbing which took centuries and in each case implied intermediate steps not indicated in our survey. No clashings could occur so long as each category kept its distance from the sounds above and below, and thus we find that the Elizabethans as scrupulously as Chaucer kept the four classes of words apart in their rimes. But in the seventeenth century class (3) was raised, and as no corresponding change had taken place with (2), the two classes have now fallen together with the single sound [i·]. This entails a certain number of homophones such as had not been created through the preceding equidistant changes.

XV.—§ 7. Homophones.

It is, moreover, interesting to note how many of the words which were made identical with others through this change were already rare at the time or have at any rate become obsolete since: this is true of breech, lief, meed, mete (adj.), quean, weal, wheal, ween and perhaps a few others. Now, obsolescence of some words is always found in connexion with such convergent sound changes. In some cases the word had already become rare before the change in sound took place, and then it is obvious that it cannot have offered serious resistance to the change that was setting in. In other cases the dying out of a word must be looked upon as a consequence of the sound change which had actually taken place. Many scholars are now inclined to see in phonetic coalescence one of the chief reasons why words fall into disuse, see, e.g., Liebisch (PBB XXIII, 228, many German examples in O. Weise, Unsere Mutterspr., 3d ed., 206) and Gilliéron, La faillite de l’étymologie phonétique (Neuveville, 1919—a book whose sensational title is hardly justified by its contents).

The drawbacks of homophones[67] are counteracted in various ways. Very often a synonym steps forward, as when lad or boy is used in nearly all English dialects to supplant son, which has become identical in sound with sun (cf. above p. 120, a childish instance). Very often it becomes usual to avoid misunderstandings through some addition, as when we say the sole of her foot, because her sole might be taken to mean her soul, or when the French say un dé à coudre or un dé à jouer (cf. E. minister of religion and cabinet minister, the right-hand corner, the subject-matter, where the same expedient is used to obviate ambiguities arisen from other causes). Chinese, of course, is the classical example of a language abounding in homophones caused by convergent sound changes, and it is highly interesting to study the various ways in which that language has remedied the resulting drawbacks, see, e.g., B. Karlgren, Ordet och pennan i Mittens rike (Stockholm, 1918), p. 49 ff. But on the whole we must say that the ways in which these phonetic inconveniences are counteracted are the same as those in which speakers react against misunderstandings arising from semantic or syntactic causes: as soon as they perceive that their meaning is not apprehended they turn their phrases in a different way, choosing some other expression for their thought, and by this means language is gradually freed from ambiguity.

XV.—§ 8. Significative Sounds preserved.

There do not seem to be many quite unimpeachable examples of words which have received exceptional phonetic treatment to obviate misunderstandings arising from homophony; other explanations (analogy from other forms of the same word, etc.) can generally be alleged more or less plausibly. But this does seem to be the easiest explanation of the fact that the E. preposition on has always the full vowel [ɔ], though in nine cases out of ten it is weakly stressed and though all the other analogous prepositions (to, for, of, at) in the corresponding weak positions in sentences are generally pronounced with the ‘neutral’ vowel [ə]. But if on were similarly pronounced, ambiguity would very often result from its phonetic identity with the weak forms of the extremely frequent little words an (the indefinite article) and and (possibly also in), not to mention the great number of [ən]s in words like drunken, shaken, deepen, etc., where the forms without -en also exist. With the preposition upon the same considerations do not hold good, hence the frequency of the pronunciation [əpən] in weak position. Considerations of clearness have also led to the disuse of the formerly frequent form o (o’) which was the ‘natural’ development of each of the two prepositions on and of. The form written a survives only in some fossilized combinations like ashore; in several others it has now disappeared (set the clock going, formerly a-going, etc.).

Sometimes, when all ordinary words are affected by a certain sound change, some words prove refractory because in their case the old sound is found to be more expressive than the new one. When the long E. [i·] was diphthongized into [ai], the words pipe and whine ceased to be good echoisms, but some dialects have peep ‘complain,’ which keeps the old sound of the former, and the Irish say wheen (Joyce, English as we speak it in Ireland, 103). In squeeze the [i·] sound has been retained as more expressive—the earlier form was squize; and the same is the case with some words meaning ‘to look narrowly’: peer, peek, keek, earlier pire, pike, kike (cf. Dan. pippe, kikke, kige, G. kieken).[68] In the same way, when the old [a·] was changed into [ɛ·, ei], the word gape ceased to be expressive (as it is still in Dan. gabe), but in popular speech the tendency to raise the vowel was resisted, and the old sound [ga·p] persisted, spelt garp as a London form in 1817 (Ellis, EEP v. 228) and still common in many dialects (see gaup, garp in EDD); Professor Hempl told me that [ga·p] was also a common pronunciation in America. In the chapter on Sound Symbolism (XX) we shall see some other instances of exceptional phonetic treatment of symbolic words (especially tiny, teeny, little, cuckoo).

XV.—§ 9. Divergent Changes and Analogy.

Besides equidistant and convergent sound changes we have divergent changes, through which sounds at one time identical have separated themselves later. This is a mere consequence of the fact that it is rare for a sound to be changed equally in all positions in which it occurs. On the contrary, one must admit that the vast majority of sound changes are conditioned by some such circumstance as influence of neighbouring sounds, position as initial, medial or final (often with subdivisions, as position between vowels, etc.), place in a strongly or weakly stressed syllable, and so forth. One may take as examples some familiar instances from French: Latin c (pronounced [k]), is variously treated before o (corpus > corps), a (canem > chien), and e (centum > cent); in amicum > ami it has totally disappeared. Lat. a becomes e in a stressed open syllable (natum > né), except before a nasal (amat > aime); but after c we have a different treatment (canem > chien), and in a close syllable it is kept (arborem > arbre); in weak syllables it is kept initially (amorem > amour), but becomes [ə] (spelt e) finally (bona > bonne). This enumeration of the chief rules will serve to show the far-reaching differentiation which in this way may take place among words closely related as parts of the same paradigm or family of words; thus, for Lat. amo, amas, amat, amamus, amatis, amant we get OFr. aim, aimes, aime, amons, amez, aiment, until the discrepancy is removed through analogy, and we get the regular modern forms aime, aimes, aime, aimons, aimez, aiment. The levelling tendency, however, is not strong enough to affect the initial a in amour and amant, which are felt as less closely connected with the verbal forms. What were at first only small differences may in course of time become greater through subsequent changes, as when the difference between feel and felt, keep and kept, etc., which was originally one of length only, became one of vowel quality as well, through the raising of long [e·] to [i·], while short [e] was not raised. And thus in many other cases. Different nations differ greatly in the degree in which they permit differentiation of cognate words; most nations resent any differentiation in initial sounds, while the Kelts have no objection to ‘the same word’ having as many as four different beginnings (for instance t-, d-, n-, nh-) according to circumstances. In Icelandic the word for ‘other, second’ has for centuries in different cases assumed such different forms as annarr, önnur, öðrum, aðrir, forms which in the other Scandinavian languages have been levelled down.

It is a natural consequence of the manner in which phonology is usually investigated and represented in manuals of historical grammar—which start with some old stage and follow the various changes of each sound in later stages—that these divergent changes have attracted nearly the sole attention of scholars; this has led to the prevalent idea that sound laws and analogy are the two opposed principles in the life of languages, the former tending always to destroy regularity and harmony, and the latter reconstructing what would without it be chaos and confusion.[69]

This view, however, is too rigorous and does not take into account the manysidedness of linguistic life. It is not every irregularity that is due to the operation of phonetic laws, as we have in all languages many survivals of the confused manner in which ideas were arranged and expressed in the mind of primitive man. On the other hand, there are many phonetic changes which do not increase the number of existing irregularities, but make for regularity and a simpler system through abolishing phonetic distinctions which had no semantic or functional value; such are, for instance, those convergent changes of unstressed vowels which have simplified the English flexional system (Ch. XIV § 10 above). And if we were in the habit of looking at linguistic change from the other end, tracing present sounds back to former sounds instead of beginning with antiquity, we should see that convergent changes are just as frequent as divergent ones. Indeed, many changes may be counted under both heads; an a, which is dissociated from other a’s through becoming e, is identified with and from henceforth shares the destiny of other e’s, etc.

XV.—§ 10. Extension of Sound Laws.

If a phonetic change has given to some words two forms without any difference in signification, the same alternation may be extended to other cases in which the sound in question has a different origin (‘phonetic analogy’). An undoubted instance is the unhistoric r in recent English. When the consonantal [r] was dropped finally and before a consonant while it was retained before a vowel, and words like better, here thus came to have two forms [betə, hiə] and [betər (ɔf), hiər (ən ðɛ·ə)] better off, here and there, the same alternation was transferred to words like idea, drama [ai'diə, dra·mə], so that the sound [r] is now very frequently inserted before a word beginning with a vowel: I’d no idea-r-of this, a drama-r-of Ibsen (many references MEG i. 13. 42). In French final t and s have become mute, but are retained before a vowel: il est [ɛ] venu, il est [ɛt] arrivé; les [le] femmes, les [lez] hommes; and now vulgar speakers will insert [t] or [z] in the wrong place between vowels: pa-t assez, j’allai-t écrire, avant-z-hier, moi-z-aussi; this is called ‘cuir’ or ‘velours.’

In course of time a ‘phonetic law’ may undergo a kind of metamorphosis, being extended to a greater and greater number of combinations. As regards recent times we are sometimes able to trace such a gradual development. A case in point is the dropping of [j] in [ju·] after certain consonants in English (see MEG i. 13, 7). It began with r as in true, rude; next came l when preceded by a consonant, as in blue, clue; in these cases [j] is never heard. But after l not preceded by another consonant there is a good deal of vacillation, thus in Lucy, absolute; after [s, z] as in Susan, resume there is a strong tendency to suppress [j], though this pronunciation has not yet prevailed,[70] and after [t, d, n], as in tune, due, new, the suppression is in Britain only found in vulgar speakers, while in some parts of the United States it is heard from educated speakers as well. In the speech of these the sound law may be said to attack any [ju·] after any point consonant, while it will have to be formulated in various less comprehensive terms for British speakers belonging to older or younger generations. It is extremely difficult, not to say impossible, to reconcile such occurrences with the orthodox ‘young grammarian’ theory of sound changes being due to a shifting of the organic feeling or motor sensation (verschiebung des bewegungsgefühls) which is supposed to have necessarily taken place wherever the same sound was under the same phonetic conditions. For what are here the same phonetic conditions? The position after r, after l combinations, after l even when standing alone, after all point consonants? Each generation of English speakers will give a different answer to this question. Now, it is highly probable that many of the comprehensive prehistoric sound changes, of which we see only the final result, while possible intermediate stages evade our inquiry, have begun in the same modest way as the transition from [ju·] to [u·] in English: with regard to them we are in exactly the same position as a man who had heard only such speakers as say consistently [tru·, ru·d, blu·, lu·si, su·zn, ri'zu·m, tu·n, du·, nu·] and who would then naturally suppose that [j] in the combination [ju·] had been dropped all at once after any point consonant.

XV.—§ 11. Spreading of Sound Change.

Sound laws (to retain provisionally that firmly established term) have by some linguists, who rightly reject the comparison with natural laws (e.g. Meringer), been compared rather with the ‘laws’ of fashion in dress. But I think it is important to make a distinction here: the comparison with fashions throws no light whatever on the question how sound changes originate—it can tell us nothing about the first impulse to drop [j] in certain positions before [u·]; but the comparison is valid when we come to consider the question how such a change when first begun in one individual spreads to other individuals. While the former question has been dealt with at some length in the preceding investigation, it now remains for us to say something about the latter. The spreading of phonetic change, as of any other linguistic change, is due to imitation, conscious and unconscious, of the speech habits of other people. We have already met with imitation in the chapters dealing with the child and with the influence exerted by foreign languages. But man is apt to imitate throughout the whole of his life, and this statement applies to his language as much as to his other habits. What he imitates, in this as in other fields, is not always the best; a real valuation of what would be linguistically good or preferable does not of course enter the head of the ‘man in the street.’ But he may imitate what he thinks pretty, or funny, and especially what he thinks characteristic of those people whom for some reason or other he looks up to. Imitation is essentially a social phenomenon, and if people do not always imitate the best (the best thing, the best pronunciation), they will generally imitate ‘their betters,’ i.e. those that are superior to them—in rank, in social position, in wealth, in everything that is thought enviable. What constitutes this superiority cannot be stated once for all; it varies according to surroundings, age, etc. A schoolboy may feel tempted to imitate a rough, swaggering boy a year or two older than himself rather than his teachers or parents, and in later life he may find other people worthy of imitation, according to his occupation or profession or individual taste. But when he does imitate he is apt to imitate everything, even sometimes things that are not worth imitating. In this way Percy, in Henry IV, Second Part, II. 3. 24—