SECTION XXXIV. THE PHONETIC SYSTEM-ITS OCCASIONAL LIABILITY TO BE APPLIED PREMATURELY. ITS INFLUENCE UPON THE FIXATION OF A LANGUAGE. - HISTORICAL VIEW OF THE TWO SERIES K-TSH AND G-DZH. The question as to the extent of the applicability and the universality of the application of the Phonetic System has thus far been illustrated by two classes of facts; those that relate to the analysis of certain combinations, and those that relate to the effects of the Euphonic Principle: and it is clear that the two groups are very different. The limitations that they illustrate are what we may call Formal; inasmuch as it is with the particular forms which such or such an orthography may take that they are specially connected. In the division of the subject now coming under notice the question of Form will be wholly out of sight: for the question will be one of Time rather than aught else: time meaning that particular stage in the development-that date in the lifetime of a language -for which the application of the phonetic principle is, comparatively speaking, well or ill adapted. The introduction of an alphabet, and along with it, a reduction of the language to which it applies to writing is, of itself, a boon of such unparagoned value that there is no possible time or season at which it can, under any circumstances, be other than a good. It is one of those blessings which can never come either too early or too late. At the same time, the period during which the language, either wholly or partly, is, more or less, in a transitional state, is not the best time for its introduction. There are, probably, in all languages, at every period, certain words of which the pronunciation is not absolutely fixed; and that, not so much because certain speakers have not come to an agreement as to the way in which certain doubtful words are to be sounded, as because there are (as facts of language) certain combinations of an unstable character: unstable, be it noted, being a word which we have seen before, and one which we shall see again; a word of no slight importance in both this and the following section. The term was applied to the series 1. Kia, kya, ksha, tsha: and no series better illustrates it and no series better shows what would happen to an orthography which was introduced either just when the pronunciation of ka or ga, began to grow unsteady, or just when kya or gya was beginning to change into ksha or gzha; or these sounds, one or both, into tsha or dzha-where, for the first time we get a stable combination. The least that can be said of them is that they would run a chance of being very soon altered: and though this is an objection of very little weight it is sufficient to show that there are certain times in the lifetime of a language, when either the introduction of a new, or the change of an old, combination, is less likely to be permanent than at others: a remark which we may now conveniently illustrate by the actual history of these two series-not, however, for a reason which may have already suggested itself, so closely and so exactly as we wish. Nevertheless, by referring to one language for one, and to another for another, part of the process, we arrive at a reasonable conclusion; and invest a sequence of changes which has hitherto been based on a priori notions, with something like the reality of history. We begin with the two extremes. There was a time when the c in words like chest and Chester, was sounded as k and of this time, there is, in most cases, ample and superabundant evidence. Had the e to which we now give the power of s, always been so sounded, it would never have been introduced in any language where a sign for the genuine and original sexisted. What, indeed, would have been its use? If words like city, had always, or when they were first derived from words like civis, been sounded sity, how would they differ from words like save, slave and the like; and, if not differing, why would they require a special sign or letter? Again, if the original sound were not that of k why would the letter c be assigned to them? For c, to a great extent, is the same as k. It it needless, however, to limit ourselves to the consideration of the presumptions of the case when there is evidence enough to make them unnecessary. It is admitted that the ein such a word as city, was once sounded as k. Thus much, perhaps, would hardly have been denied by a respectable philologue in the days of our distant ancestors. At the present time it is insisted upon; in some instances to the extent of changing the existing pronunciation of English-Latin; or Latin as it is taught in England. Upon this, it is needless to enlarge. Now if the evidence of the Latin language be condemned as insufficient there is that of the Greek in the background. With words like κήτος and κιστη to give us the old sound of the o in cetaceous and chest, we may, if necessary, dispense with the evidence of the Latin language altogether. Do I hold, then, that in the Italian where the c before bef i is sounded just as it is the English chest, it has grown out of that of the k in κιστῃ, and that the connection between these two extreme forms is that of the sequence ke, kye, ksh, tsh, with an unstable, transitory, or even ephemeral ksh, as a missing link, may have had a function and played a part-a short but important one? Such is my belief: indeed, it is already to be found, implicitly, in what I have already written. If tsh, then, have been ksh, and ksh have been ky, and ky have been k, and the changes from one sound to the other have gone on more rapidly than the corresponding change in the spelling, it is possible that one or more than one of the intermediate forms may never have been recorded; never, indeed, indicated or even hinted at. It is more than possible. It is almost certain that such has been the case. That the e in Latin, at some comparatively early period, was sounded as k, is certain; certain, too, it is that at a comparatively late period (i.e. the present time) it is sounded both as tsh (in Italian) and as s (in French and English). The power, however, of the letter during the intermediate period is by no means certain; and the longer that period is the greater the chance of some intermediate stage falling within it. It does not, then, follow that because the Latin c=kat one era of the Latin language, it may not have been something else at another, namely, tsh or ksh as the case may be-tsh at a late date, ksh at an intermediate one. In our own language I hold that in some words we have the series of changes in full; from simple k to complex tsh. The simple kwe get in the Latin castrum or castra. The West-Saxon spelling, however, is "ceaster;" where the ethe German and Scandinavian i orj; which, in its turn, is equivalent to English y. Hence, ceaster = cyaster=kjaster in Norse and German. But the Norwegian reading of a word thus spelt is (approximately) kshaster, which, in Swedish, is tshaster. At one time the word under notice was sounded kyaster, at another kshaster, at another tshester-the change from a to e being no part of the present question. That Chester is now called Tshester is certain. That the poet Ceadmon called himself either Kyadmon or Kshadmon, and that his contemporaries said either kyaster or kshaster, I hold to be nearly certain. But, how some scores of words were sounded in this or that district, at this or that time, during the intervening period, I cannot say. I cannot say when an older form died out, nor yet when a newer one started into life. If g stands to dzh, as k to tsh, the history of the two series Ka, kya, ksha, tsha may be expected to run parallel to each other. But we have already said that this is the case only on the first view of the subject. In reality the lines diverge; the reason being, as has already been stated, the affinity, in an opposite direction, between gandy. Hence, while ka almost always runs through kya to ksha, ga often runs through gya to ya. Now the actual exhibition of these changes is not an easy matter: and we know the reason why. The chances are that a sound which changes quickly may pass away without ever being recorded; and without leaving any visible sign of its existence; more especially is this likely in languages where there is no sign for the sound of sh; as is often the case. The best illustration known to me is to ✓ be found in the triple history of the combination k+ y in the three countries of Denmark, Norway, and Sweden. The ordinary spelling is kj; j being, as in German, the sign for the semi-vowel y. In Denmark this is sounded sometimes as k simply, sometimes as ky. In Norway it is ky, and something very like ksh. In Sweden it is tsh. Whether, or when, the Danish will go farther in the direction of tsh, remains to be seen. The Swedish has attained that final goal. The Norwegian gives the transitional form. But there is no use in concealing the complications of the question. The transitional character of the Norwegian kj is itself transitional: and it is only dealing fairly with the reader to show the extent to which this intermediate sound is, or is not, found in Norway. The present writer may, also, add that, upon the nature of it, he is not only an exposition, but a witness, and not a mere second-rate authority. The connection of the Norwegian kj, with ksh cannot be seen by mere inspection. The Norwegian language, unlike the Swedish, has no specific grammar of its own; because, though the Swedish is treated as a different language from the Danish, the Norwegian, so far as the written, literary, classical, or standard language is concerned, is simply Danish. The explanation of this is found in the political history of the two countries; a point which needs only being indicated. There is, indeed, at the present moment, a movement in Norway to develop out of the provincial dialects an eclectic form of speech which shall, at one and the same time be non-Danish, non-Swedish, Sandinavian, Norse, Norwegian aad national; in short an artificial language: an attempt upon which we pass no opinion. Hence, as literary languages, the Norwegian and Danish are the same: and, though such may exist, I know of no grammar which makes the peculiar Norwegian sound of kj, the object of any special treatment. It is not, then, on the surface that this peculiarity is to be found. Yet everyone who has any practical knowledge of the Norwegian Danish knows that in words like kjöre = =drive, and Kjöbenhavn = Copenhagen, etc., the sound of the first two letters is different in the two languages, dialects, forms of speech or whatever we choose to call them. Sometimes, indeed, they are identical, but, in this, the influence of the written language plays a part. But, even when the difference is manifest, the pronunciation of the differentiated combination is not constant. Sometimes it is that of the true continuous surd of the k series; -a rare sound. Upon the whole, however, it is a sound which is to ksh as the ch in chest is to tsh. I speak upon the point as a witness, because I am not aware of this very important fact in phonetics having hitherto had more than a cursory notice-if it has had this. But, at the same time, I speak with the diffidence of one who is pronouncing upon a sound foreign to his own tongue; and as I make no secret of being suspicious of other persons in such cases, it is only right I should, to some extent, draw attention to an opinion of my own. I have, however, made a point of checking it by that of natives; and, although I cannot say that an Englishman who sounds kjöre exactly as kshöre will speak exactly like a Norwegian, I have no hesitation in saying that he will utter the sound into which ky ran, and out of which the Swedish tsh arose. This I believe, in the special pair of languages before us, it preceded, as a cause precedes an effect. I believe, also, that it has done so in many other languages; in other words that the order of sequence is a fact of very general occurrence. Upon its universality I see no need of delivering an opinion. Practically, however, I hold that the presence of tsh, connected with the k-series, implies a previous ky; ky; and that ky, wherever it occurs, is a tsh in posse. the evidence to the reality of this sequence of changes is so indefinite that, even in the best illustration I have been able to hit upon, the line of argument is circuitous and indirect. No wonder. The same fact in language which makes ky and ksh unstable combinations, exposes them to the risk of not being orthographically represented. But the Norwegian forms of the Scandinavian or Norse, tell us something more. Instead of gy running parallel with ky it diverges. Instead of running into gzh it runs into y; so that, while kjöre= kshöre, gjente (wench) yente: a difference which has already been indicated. There are two other sounds of the k series which still require notice: viz., the two which stand to those of k and g, respectively, as f and v stand to pand b; or as the th in thin and thine stand to t and d. There is much more to be said about them than is to be found in the ordinary grammars; in many of which there is much that requires correction. Neither is it very easy to write about them concisely; inasmuch as the terms which apply to them are less definite than they should be. In describing or classifying the mutes of either the English or any other language, it is barren work to talk of them as lenes and aspirates, or as hards and softs. But, as they actually form a system, and as that system must be exhibited, four terms at least are required. Explosive and continuous (sometimes, more conveniently, explodent and continuant); sonant and surd will be the terms used here: the last two from the grammars of the Sanskrit grammas, explosive and continuous from the older writers in general; especially those who treated the classification of articulate sounds as anatomists or physiologists. The b in ba, when pronounced separately from the vowel, is explosive, and sonant; explosive because the sound, as that of a consonant only, cannot be continued; and sonant because its sound is uttered at the ordinary pitch of the voice. Fon the other hand is continuous and surd; for the sound itself can be prolonged; but, whether prolonged or not, it is uttered as a whisper. Hence the following system: In the s series there are no explodents; so far, however, as sonancy and surdness go, it agrees with that of p, t, and k. If it were otherwise, and if we could add the surd and sonant continuants of the k series, or, changing the expression, the sounds which stand to k and gas |