標記については「#1577. 言語変化の形式的説明と機能的説明」 ([2013-08-21-1]) で議論した．言語変化が諸変異形からの選択だとすると，その選択は形式的（機械的）に駆動されているのだろうか，あるいは機能的に駆動されているのだろうか．この根源的な問題について，Samuels (28--29) が次のように考察している．やや長いが，問題の議論の箇所を引用する．
Does it [=the process of selection] involve merely fitting a ready-made stretch of the inherited spoken chain, recalled from memory, to a given situation? If so, the mechanical factors in change would appear more important. Or is there fresh selection of each functional element in the utterance? In that case, the functional factor would outweigh the mechanical. These are questions for which the psycholinguist has not yet provided an answer, and perhaps no answer is possible. From a subjective viewpoint, it might be surmised that, since many tactless or otherwise inapposite utterances turn out, in retrospect, to have been uses of previously learnt stretches of the spoken chain that simply did not fit the situation, our overall use of the 'form-directed' spoken chain is greater than we might think, and in any event greater than the conventionalities of cliché and phatic communion to which it is often restricted. Information theory has shown that all common collocations, being predictable, are more easily understood than rare collocations; and since the same presumably applies to the speaker's selection of them from the memory-store, their use is but another demonstration of the principle of least effort. On the other hand, a brain that is alerted to the danger of inapposite utterance will presumably 'check' the functional value of the units it selects. It depends, therefore, on the degree of alertness, whether an utterance is to be called as 'form-directed' or 'function-directed'; naturally there are borderline cases, but it is probably wrong to assume that utterance normally results from a conflict, within the brain, between the need for intelligibility and the tendency to rely on stretches of actual or embryo cliché. In the genesis of a single utterance, one or other factor will have the upper hand (which one depends on the speakers and the situation); but there is no way of gauging from individual utterances the relative importance of each factor for the processes of change. But although the two factors cannot be separated at the level of idiolect, it is reasonable to suppose that the total utterances of a community will fall mainly into one or the other of the two types --- 'form-directed' and 'function-directed'. Both types belong to the same language, and neither (except in extreme cases of cliché or innovation) is overtly distinguishable; they are interdependent and interact, and neither would exist or be understood without the other. It follows that, for the study of change, both types must in the first instance be accorded equal consideration . . . .
Samuels は，形式と機能による駆動の両方を常に視野に入れつつ言語変化を考察し始めることが肝要だと説く．個々の言語変化では，結果として，いずれが優勢であるかが分かるのかもしれないが，最初からいずれか偏重のバイアスを持ってはいけない，ということだ．ときに Samuels （の議論）は functionalist として批判を招いてきたが，Samuels 自身はこの引用の最後からも明らかなように，形式と機能の両観点のあいだでバランスを取ろうとしている．
・ Samuels, M. L. Linguistic Evolution with Special Reference to English. London: CUP, 1972.
Powered by WinChalow1.0rc4 based on chalow