Phonological representation of the sign: Linearity and nonlinearity in American sign language

Sign languages present us with an interesting puzzle. While all groups of hearing humans use a natural spoken language, and no group of hearing people has ever been known to use sign language as its primary means of communication, among every known community of deaf people left to their own devices a natural sign language has arisen. Deaf children whose parents are also deaf acquire sign language in the same way as hearing children acquire spoken language (Newport and Meier (1986)). That is, deaf children are presented with the same quality of input, they go through the same stages, and they overgeneralize grammatical rules, just as children acquiring spoken language do. In addition, deaf children with signing deaf parents learn sign language in about the same length of time as hearing children learn spoken language. Research has also demonstrated that sign language (or at least American Sign Language) is a complex language system with phonological, morphological, and syntactic levels and rules (see, for example Klima and Bellugi (1979), Siple (1978) Wilbur (1986b)). Rate and speed of acquisition of such a complex system are often pointed to as evidence that certain universal organizing principles of language are part of our genetic make-up (Chomsky (1959)). Thus, it appears that humans are genetically endowed with the capability to learn each of two language systems which are distinct from each other at the motoric and perceptual levels.