Morphological classes and grammatical organization

Abstract

One moral which is drawn from the detailed studies of morphological classes presented here is that morphology is an independent component of the grammar. As such, morphology has its own internal logic, with its own rules, patterns, and system of defaults. Morphological categories often correspond in a general way to phonological, semantic, and syntactic categories, but the correspondence is not close enough to permit a reduction. Thus, the suffixes that are added to a noun in Latvian, Spanish or Arapesh (the noun’s inflection class) cannot be predicted completely on the basis of the sex of the referent of that noun (the noun’s semantics) or the pattern of agreement that it induces on words in construction with it (the noun’s gender, or syntax). Similarly, whether a verb takes one agreement affix or two in Georgian or Algonquian (the verb’s inflection class) cannot be predicted purely on the basis of how many participants are involved in the action referred to by the verb (the verb’s semantics), nor on the basis of how many noun phrases the verb is subcategorized for (the verb’s syntax). The authors conclude that morphology, while interrelated with other aspects of the grammar in various ways, has a life of its own.

The paper is a written version of comments made at the symposium on Morphological Classes during the 1990/1991 annual meeting of the Linguistics Society of America, Chicago, III. Thanks go to Mark Aronoff for organizing the symposium, and to all of the participants for making their work available in advance. I have tried to include some of the most relevant points that were brought up in general discussion at the symposium; these are indicated in the text as ‘personal communications.’