Sütterlin thinks it enough to mention some sound changes in which the new sound is more difficult than the old; these being admitted, he concludes (and others have said the same thing) that those other instances in which the new sound is evidently easier than the old one cannot be explained by the principle of ease. But it seems clear that this conclusion is not valid: the correct inference can only be that the tendency towards ease may be at work in some cases, though not in all, because there are other forces which may at times neutralize it or prove stronger than it. We shall meet a similar all-or-nothing fallacy in the chapter on Sound Symbolism.
Now, it is sometimes said that natives do not feel any difficulty in the sounds of their own language, however difficult these may be to foreigners. This is quite true if we speak of a conscious perception of this or that sound being difficult to produce; but it is no less true that the act of speaking always requires some exertion, muscular as well as psychical, on the part of the speaker, and that he is therefore apt on many occasions to speak with as little effort as possible, often with the result that his voice is not loud enough, or that his words become indistinct if he does not move his tongue, lips, etc., with the required precision or force. You may as well say that when once one has learnt the art of writing, it is no longer any effort to form one’s letters properly; and yet how many written communications do we not receive in which many of the letters are formed so badly that we can do little but guess from the context what each form is meant for! There can be no doubt that the main direction of change in the development of our written alphabet has been towards forms requiring less and less exertion—and similar causes have led to analogous results in the development of spoken sounds.
It is not always easy to decide which of two articulations is the easier one, and opinions may in some instances differ—we may also find in two neighbouring nations opposite phonetic developments, each of which may perhaps be asserted by speakers of the language to be in the direction of greater ease. “To judge of the difficulty of muscular activity, the muscular quantity at play cannot serve as an absolute measure. Is [d] absolutely more awkward to produce than [ð]? When a man is running full tilt, it is under certain circumstances easier for him to rush against the wall than to stop suddenly at some distance from it: when the tongue is in motion, it may be easier for it to thrust itself against the roof of the mouth or the teeth, i.e. to form a stop (a plosive), than to halt at a millimetre’s distance, i.e. to form a fricative” (Verner 78). In the same sense I wrote in 1904: “Many an articulation which obviously requires greater muscular movements is yet easier of execution than another in which the movement is less, but has to be carried out with greater precision: it requires less effort to chip wood than to operate for cataract” (PhG 181).
In other cases, however, no such doubt is possible: , [f] or [x] require more muscular exertion than [h], and a replacement of one of them by [h] therefore necessarily means a lessening of effort. Now, I am firmly convinced that whenever a phonologist finds one of these oral fricatives standing regularly in one language against [h] in another, he will at once take the former sound to be the original and [h] to be the derived sound: an indisputable indication that the instinctive feeling of all linguists is still in favour of the view that a movement towards the easier sound is the rule, and not the exception.
In thus taking up the cudgels for the ease theory I am not afraid of hearing the objection that I ascribe too great power to human laziness, indolence, inertia, shirking, easygoingness, sloth, sluggishness, lack of energy, or whatever other beautiful synonyms have been invented for ‘economy of effort’ or ‘following the line of least resistance.’ The fact remains that there is such a ‘tendency’ in all human beings, and by taking it into account in explaining changes of sound we are doing nothing else than applying here the same principle that attributes many simplifications of form to ‘analogy’: we see the same psychological force at work in the two different domains of phonetics and morphology.
It is, of course, no serious objection to this view that if this had been always the direction of change, speaking must have been uncommonly troublesome to our earliest ancestors[57]—who says it wasn’t?—or that “if certain combinations were really irksome in themselves, why should they have been attempted at all; why should they often have been maintained so long?” (Oertel 204)—as if people at a remote age had been able to compare consciously two articulations and to choose the easier one! Neither in language nor in any other activity has mankind at once hit upon the best or easiest expedients.
XIV.—§ 7. Sounds in Connected Speech.
In the great majority of linguistic changes we have to consider the ease or difficulty, not of the isolated sound, but of the sound in that particular conjunction with other sounds in which it occurs in words.[58] Thus in the numerous phenomena comprised under the name of assimilation. There is an interesting account in the Proceedings of the Philological Society (December 17, 1886) of a discussion of these problems, in which Sweet, while maintaining that “cases of saving of effort were very rare or non-existent” and that “all the ordinary sounds of language were about on a par as to difficulty of production,” said that assimilation “sprang from the desire to save space in articulation and secure ease of transition. Thus pn became pm, or else mn.” But in both these changes there is saving of effort, for in the former the movement of the tip of the tongue required for [n], and in the latter the movement of the soft palate required for [p], is done away with[59]: the term “saving of space” can have no other meaning than economy of muscular energy. And the same is true of what Sweet terms “saving of time,” which he finds effected by dropping superfluous sounds, especially at the end of words, e.g. [g] after [ŋ] in E. sing. Here, of course, one articulation (of the velum) is saved and this need not even be accompanied by the saving of any time, for in such cases the remaining sound is often lengthened so as to make up for the loss.[60]
If, then, all assimilations are to be counted as instances of saving of effort, it is worth noting that a great many phonetic changes which are not always given under the heading of assimilation should really be looked upon as such. If Lat. saponem yields Fr. savon, this is the result of a whole series of assimilations: first [p] becomes , because the vocal vibrations continue from the vowel before to the vowel after the consonant, the opening of the glottis being thus saved; then the transition of to [v] between vowels may be considered a partial assimilation to the open lip position of the vowels; the vowel [o] is nasalized in consequence of an assimilation to the nasal [n] (anticipation of the low position of the velum), and the subsequent dropping of the consonant [n] is a clear case of a different kind of assimilation (saving of a tip movement); at an early stage the two final sounds of saponem had disappeared, first [m] and later the indistinct vowel resulting from e: whether we reckon these disappearances as assimilations or not, at any rate they constitute a saving of effort. All droppings of sounds, whether consonants (as t in E. castle, postman, etc.) or vowels (as in E. p’rhaps, bus’ness, etc.), are to be viewed in the same light, and thus by their enormous number in the history of all languages form a strong argument in favour of the ease theory.
There is one more thing to be considered which is generally overlooked. In such assimilations as It. otto, sette, from octo, septem, a greater ease is effected not only by the assimilation as such, by which one of the consonants is dropped—for that would have been obtained just as well if the result had been occo, seppe—but also by the fact that it is the tip action which has been retained in both cases, for the tip of the tongue is much more flexible and more easily moved than either the lips or the back of the tongue. On the whole, many sound changes show how the tip is favoured at the cost of other organs, thus in the frequent transition of final -m to -n, found, for instance, in old Gothonic, in Middle English, in ancient Greek, in Balto-Slavic, in Finnish and in Chinese.