ABSTRACT
This chapter examines the representation of handshapes in sign languages. The model proposed is motivated by three types of evidence, of which the first two are most plentiful: phonemic distinctions, markedness, and assimilatory rules. I have adopted two separate theories of phonological structure in this work: the feature geometry theory of Clements (1985), Sagey (1986), and others, and the dependency theory of Anderson and Ewen (1987), van der Hulst (1989), and others. Although these two theoretical approaches are incompatible in some respects, they are compatible in others. In particular, the notion of articulator-based, hierarchically structured feature classes is adopted, together with the notion that phonological primitives at lower levels are unary, and that complex units are built by combining these primitives according to certain structural relationships.
