Title

ALLOMORPHY IN LINGUISTIC THEORY: STRONG VERBS AND DERIVED NOUNS IN GERMAN

Date of Completion

January 1981

Keywords

Language, Linguistics

Degree

Ph.D.

Abstract

This thesis investigates properties of German strong verbs and noun counterparts within a restrictive theory in which related words are derived from a uniform base representation in the lexicon. I compare this approach to a morpholexical approach (Lieber, 1980) in which all stem variants are in the lexicon. I will assume the general theory of word formation proposed by Allen (1978), in which word formation rules are level-ordered. I will also assume a theory of phonology (Michaels, 1980a) in which three components of rules are distinguished: (i)allomorphy rules, (ii)derivational rules, and (iii)phonetic rules. ^ In Chapter 2, preliminary discussions of issues such as level-ordering, zero-derivation, conversion, allomorphy and morpholexical rules are presented. ^ Chapter 3 is an analysis of German strong verbs and noun counterparts in which class designations are lexical primitives. Allomorphy rules which are sensitive to class membership derive the stem vowels of the principal parts of the verb from a base stem vowel. I argue that the noun counterparts in Class I and Class II are derived from the verbs based on the fact that a single nominalization rule can predict the properties of all derived nouns in a class. Noun counterparts in the remaining classes are not derived from verbs and have distinct lexical entries.^ Chapter 4 is an alternative analysis of German strong verbs and noun counterparts. Here, no primitive notion of strong verb or class membership is assumed. A strong verb is one which is marked for allomorphy rules; all verbs which are marked for the same allomorphy rules constitute a class. I present data from verbs with separable prefixes which offer support for this alternative analysis. ^ In Chapter 5, I consider data from Old English strong verbs and noun counterparts. My analysis here closely parallels the analysis of the German data in Chapter 4. The Old English data provide further support for my claim that the generalizations are best captured by the more restrictive theory. ^