昨日の記事「#1154. 言語変化のゴミ箱としての analogy」 ([2012-06-24-1]) で言及した Postal は，「#1123. 言語変化の原因と歴史言語学」 ([2012-05-24-1]) で取り上げたように，言語変化の原因を fashion と考える論者である．同時に，言語変化とは "grammar change" （文法変化）にほかならなず，その文法変化は世代間で生じる規則の再解釈であるという生成的言語観の持ち主だ．そして，その文法変化の最初の段階を表わす "primary sound change which is independent of language contact" は，"'nonfunctional' stylistic change" であり，それ以外の原因説明はありえないとする．
Postal の議論は一見すると単純だが，実は但し書きがいろいろとついている．例えば，社会言語学的な差異化，言語変化の実現と拡散の区別，言語接触による言語変化，英語の Romance Stress Rule の歴史的獲得についても触れている．Postal の議論は，言語変化理論の大きな枠組みの1つとして注目すべきと考えるので，重要な1節 (283--85) の大部分を引用しよう．
SOME REMARKS ON THE 'CAUSES OF SOUND CHANGE'
It is perhaps not inappropriate to ask what implications the view that sound change consists in changes in grammars has for the much discussed and controversial question of the causes of sound change. Why should a language, independently of contact with other languages, add a new rule at a certain point? Of course there are some scholars who hold that all linguistic change is a result of language contact, but this position seems too radically improbable to demand serious consideration today. Assuming then that some if not all phonological changes are independent of contact, what is their basis? It seems clear to the present writer that there is no more reason for languages to change than there is for automobiles to add fins one year and remove them the next, for jackets to have three buttons one year and two the next, etc. That is, it seems evident within the framework of sound change as grammar change that the 'causes' of sound change without language contact lie in the general tendency of human cultural products to undergo 'nonfunctional' stylistic change. This is of course to be understood as a remark about what we might call 'primary change,' that is, change which interrupts an assumed stable and long-existing system. It is somewhat more plausible that such stylistic primary changes may yield a grammar which is in some sense not optimal, and that this may itself lead to 'functional' change to bring about an optimal state. Halle's suggestion that children may learn a grammar partially distinct from that underlying the utterances they hear and thus, from the point of view of the adult language, reformulate the grammar (while adults may only add rules), may be looked upon as a suggestion of this type. For as he noted, a grammar G2 which results from the addition of a rule to a grammar G1 may not be the optimal grammar of the set of sentences it generates. Hence one would expect that children learning the new language will internalize not G2 but rather the optimal grammar G3. It remains to be seen how many of the instances of so-called 'structural sound change,' discussed in the writings of scholars like A. Martinet, can be provided with a basis of this type.
It should be emphasized that the claim that contact independent sound change is due basically to the stylistic possibility for adults to add limited types of rules to their grammars does not preclude the fact that these changes may serve social functions, i.e. may be related to group differentiation, status differences, etc. That is, the claim that change is stylistic is not incompatible with the kinds of results reached by such investigators as Labov (1963). These latter matters concern more properly the social explanation for the spread of the change, a matter which seems more properly sociological than linguistic.
I have been careful above not to insist that all instances of what have been called sound change were necessarily independent of language contact. Although committed to the view that sound change can occur without contact, we can also accept the view that some changes result from borrowing. The view that sound change is grammar change in fact really eliminates much of the importance of this difference. For under the rule interpretation of change, the only issue is the forces which led to the grammar change. Studies of complex cases of phonological grammar change due to contact have actually been carried out recently within the framework of generative grammar. This work, to be reported in a forthcoming monograph (Keyser and Halle, to appear), shows quite clearly that English has borrowed a number of Romance phonological rules. In particular, English has incorporated essentially the Latin stress rule, that which softens Romance k to s in certain environments, etc. This work has important implications beyond the area of sound change and may affect radically concepts of genetic relationship, Mischsprache, etc., all of which will, I think, require reevaluation.
・ Postal, P. Aspects of Phonological Theory. New York: Harper & Row, 1968.
Powered by WinChalow1.0rc4 based on chalow