ABSTRACT
Following Slobin et al. (2001), I prefer to use the term polycomponential verb to refer to this category of signs rather than alternatives such as classifier predicate or polymorphemic verb. This term is used because the claim that these forms include classifier morphemes is open to question, and the analysis of these signs as multimorphemic constructions is also problematic (Cogill, 1999). Although a subset of polycomponential constructions are clearly verbal in character, these constructions also include forms, which have also been described by other researchers as having an adjectival role (Schick, 1990). This appears to be the reason the term predicate, rather than verb, has been adopted by some researchers. I do not follow this practice, however, for two reasons. First, the term predicate has traditionally been used to refer to a constituent of sentence structure, rather than a lexical category (Crystal, 1991).4 Second, it is possible that the adjectival type of polycomponential verbs is best considered a kind of stative verb. Engberg-Pedersen (1993) has suggested that we need more evidence before we can claim that this subclass of complex constructions actually act as adjectives (Engberg-Pedersen, personal communication, April, 2000). This is in keeping with the debate in the literature as to whether or not there is any evidence for the lexical category o f adjective in signed languages (Bergman, 1983,1986; Johnston, 1989). There does appear to be some support for a class of nonpolycomponential adjectives in Auslan, but this claim does require more investigation. Some nonpolycomponential signs that express adjectival meanings have most of the grammatical properties of verbs (especially stative verbs), but some appear to have some distinctive properties not shared by verbs. Nonpolycomponential signs such as RED or BIG, for example, may be modified by combining with premodifiers such as VERY and MORE, and may also be used attributively (i.e., they may appear before a nominal). It is not clear if the polycomponential signs o f visual-geometric description that appear to have adjectival meanings, however, share these adjective-like properties. In any case, it is also not entirely evident that these properties are sufficient to suggest that such usages of nonpolycomponential signs qualify as adjectival. As a result, the inclusion of these forms in the category of verbs appears to be the most appropriate analysis, subject to later revision. I thus provisionally use verb as a superordinate term that includes both verbs of motion, location and handling, and adjectival-type constructions of visual-geometric description.
