規則再配列과 自由交替 = Rule Reordering and Free Alternation

다국어 초록 (Multilingual Abstract)

Since Kiparsky(1965), rule reodering has been instrumental in accounting for language change and dialectal differences. So far the concern of generative phonologists has centered around the directionality of rule reordering. Several notions have been introduced for the explanation,e.g., marked/unmarked order, the principle of maximal application, feeding/bleeding, transparency/opacity, paradigm regularity,etc.
In synchronic scene, rule reordering has been used in explaining the dialectal differences. However, there are interesting cases of free alternation where the only systematic and economical way of explanation seems to be rule reordering. In this case, the free alternation can be observed in one and the same informant as well as among different speakers.
Late Middle Korean also abounds in the examples of free alternation.For example, both [t??uyya] 'to become' and [t?uyy?] appear in the same work. These surface forms can be produced most economically by applying two rules, Vowel Harmony Rule and Glide Insertion Rule, in alternate order to the underlying form /t??uy+??/.
For these kinds of examples, Fought(1973) proposed symmetric alternate ordering as opposed to the asymmetric linear ordering proposed by Chomsky and Halle(1968).
Fought claims that the alternate ordering of two rules, a-b and b-a must be allowed in one grammatical system.
Fought's claim, however, lacks necessity and consistency. In accounting for dialectal differences of a language, we did not say that the language had tow rules applying in alternate order to produce the different dialects. Rather we took the dialects as totally different two grammatical systems and we accounted for the difference by different rule orders. Idiolectal differences could also be explained through reodering. Likewise, the free alternation can be taken as the manifestation of two different grammars. In other words, a speaker has several sets of grammar where same or different rules are ordered differntly. The speaker may choose one of the sets according to the situation, the results of which being the surface free alternation.