Morphologization of Phonological Alternations: a Theoretical Study Based on the Evidence from Germanic.

Abstract

This investigation seeks to achieve the following goals: (a) a deeper underst and ing of the interaction between morphology and phonology than is afforded by current linguistic models; (b) identification of the diachronic bases underlying the dephonologization of phonological processes; (c) development of a suitable linguistic theory, within which the description of phonological and morphological systems may be attempted; (d) a rigorously argued definition of the phenomenon of 'morphologization'. The data base for this study consists of selected problems in Germanic morphonology, though evidence from non-Germanic languages have been considered where appropriate. Chapter I examines a number of interactions between morphological and phonological parameters in order to test hypotheses of morphologization based on st and ard notions of "conditioning." The finding is a wide variety of alternations (Sanskrit vowel s and hi and the pragrhya duals, Thurneysen's Law in Gothic, Afrikaans nasal phonology) that involve 'reference' to morphosyntactic, lexical, and /or semantic parameters, but differ with respect to functional motivation or 'teleology'. While phonology and morphology converge and overlap at various points, notions of simultaneous morphological and phonological conditioning must be rejected as too imprecise. Furthermore, since phonologically motivated operations may also make reference to non-phonological information, changes among reference parameters do not provide a valid criterion for a hypothesis of morphologization. It is concluded, therefore, that the generative conception of rule structure (structural description/structural change) must be replaced with a trichotomy consisting of reference, telology, and alternancy (means). Chapter II discusses the major properties inherent to the phonological variable (quantal plausibility, contextual plausibility, productivity/exceptionality, semiotic capacity, systemic functionality) in order to isolate one that will provide a reliable basis from which to formulate a hypothesis of morphologization. Through a careful evaluation of the role of these properties in individual case histories (e.g., g-syncopation and final cluster reduction in Afrikaans, analogical umlaut in German, Sc and inavian u-umlaut, epenthesis in German and Dutch), it is shown that only systemic functionality emerges as a viable alternancy property from which to formulate our hypothesis. A fundamental conviction underlying this work is that the problem of morphologization cannot be meaningfully approached outside the framework of a comprehensive theory of morphology. Thus, Chapter III explicates the structure of the phonomorphological sign within the context of a teleological conception of phonology and morphology. The 'teleology' of an alternation denotes its intrinsic or assigned systemic value (functional motivation), or the systemically imposed exigency to which it is a resolution (e.g., phonotactics). Within the physical dimension of the sign, one can distinguish 'process' teleologies (i.e., causal relationships between phonetic parameters) and 'euphonic' teleologies (phonetically unmotivated and non-facultative). Within the representational dimension, one can distinguish four teleologies: exponence, relation (semantic opposition in derivation), concomitance (functionally null, 'external' allomorphy), and morphotactics (e.g., root structure and stem foundation). Chapter IV forges the conclusions arrived at in the foregoing chapters into a hypothesis of morphologization, which is said to take place if, and only if, the dephonologized alternation becomes a means of the language in question; that is, it must participate directly in the encoding of morphosyntactic/semantic features by means of phonological values (exponence and relation). Most dephonologized alternations, however, wind up as concomitant or morphotactic alternations, which are peripheral to the encoding of representations.