Magyar Fonetikai Füzetek 24. - Tamás Szende: Phonological Representation and Lenition Processes (1992)
2. Systematic phonological representation - 2.2. Phonological representation in a functional perspective
ates in the stratum of concrete individual, and since its systematic analysis, indeed reference to it, is rather meagre in the literature (though cf. Kerek 1977 for certain types of elision, Vogel 1987, Vogel—Kenesei 1987 for interdependences between phrase structure and the blocking of certain accommodation rules, and Szende 1989 with respect to distortion phenomena in general), we have to discuss the relation between gestalt rules and phonological rules in the classical sense, obviously with constant reference to those phenomena in speech, distortion processes, in which their operation can be documented. The set of phenomena concerned can be characterized in general as follows. The repeated occurrence of distortion phenomena classifies these types of processes on the basis of associated phonological and other conditions. For instance, vowel devoicing invariably occurs at morpheme boundaries (and usually at a phrase boundary) with the latter, as it were, conditioning this case of reduction; the factor that gives rise to sequence size truncation is normally the semantic depreciation of that sequence; and so on. Direct observation thus raises the theoretical problem of the relationship between (the types of) distortion phenomena and (phonological) rules in a natural manner. The rules of phonology, especially those of a morphosyntactic character, are absolute. Accusative -t has a constant shape as [t], and conjugation paradigms have prelexically determined vowel-harmony properties that are likewise exceptionless, otherwise the opposition látnák 'they would see' vs. látnék 'I would see' would not be possible. On the contrary, in sequence reduction, e.g. in a pronunciation broadly transcribed as ötkör (with [ce] in the second syllable) for ötkor 'at five o'clock', the application of vowel harmony to regular -kor 'at' is an occasional phenomenon, and the categorization of the regularity that is responsible for this individual form is uncertain. In particular, we could assume that the surface form is due to centralization resulting in /o/ —* [<£] , or else to a morphophonemic alternation of the temporal suffix -kor 'at' under the analogical influence [Systemzwang] of vowel harmony alternations in other case suffixes like -ból /-bői 'from inside', -tól/-től 'from', -hoz/-hez/-höz 'to', etc. "Rules" resulting in distortion phenomena are, then, relative. Wanting to avoid trivial statements like the segments occurring in distortion are not arbitrary phonetic patterns ("it is not the case that anything can replace anything"),