Semantic Phonology

Sign phonology can be as complicated as anyone wants to make it; in this respect it differs not at all from phonology generally. As evidence for this I cite a review in the international newsletter Signpost of a book that gets to the bottom, says its reviewer, “of autosegmental, metrical, and also lexical phonology.”¹ Once highly regarded (by philosophers at least) as a safeguard against unnecessary over-elaboration, Ockham’s Razor and even the computer programmers’ vernacular KISS rule (Keep It Simple, Stupid) seem to have been forgotten in recent treatments of phonology—treatments that are almost, I am tempted to say, independent of language, certainly of language as laymen use it.

What I propose is not complicated at all; it is dead simple to begin with. I call it semantic phonology. It invites one to look at a sign—i.e. a word of a primary sign language—as simply a marriage of a noun and a verb. In semantic terminology, appropriate here, the sign is an agent-verb construction. The agent is so called because it is what acts (in signing as in generative semantics), and the verb is what the agent does. What could be simpler?

Semantic-phonological, or s-p verbs are, like common verbs, transitive or intransitive. For example: when a signer of American Sign Language signifies “yes,” the sign agent (i.e. the signer’s active arm including the hand) flexes at the wrist; it is intransitive, it has no object, it acts on no patient. But if the signer signifies “stupid,” the agent action continues until it strikes the forehead; the s-p verb in this sign is thus transitive: it has an object (the grammatical term); or it takes a patient (the semantic term).

This semantic phonology is not only simple; it is also nondiscriminatory. Most phonologies of sign language now in the literature accept hands of all kinds as members but draw the line there. Like private clubs they exclude certain other kinds of agents. Semantic phonology admits all kinds and colors of sign activity, welcoming nonmanual as well as manual agents and all their actions. For instance, an eyebrow [n] raises [v]; or the tongue [n] protrudes [v]. Facial expression of the language user is just as free to join in as are the hands and fingers. The only restriction is a natural phenomenon: the s-p nouns of nonmanual signs are pretty much constrained to combine only with intransitive s-p verbs; unless of course a signer is able to touch his nose with his tongue.

Some complication is unavoidable, but only as much as is needed. It is not a good deal to trade in Ockham’s Razor, as most sign phonologists seem to have done, for something more modish but less effective. In semantic phonology the s-p noun, the agent, can be described simply (and accurately and completely) as the net effect of several unitary motor acts. Of course, nonmanual agents do not need forming; they are there to be used, to act. Motor actions supply verbs for nonmanual as well as manual agents.

The upper limbs, not the hands only, are formed into s-p nouns by motor acts—this is so clear and simple that one wonders why signs have not been described this way before. In the previous examples, “yes” and “stupid,” the agent or s-p noun is formed by fully flexing the fingers with aligned thumb to make—what else?—a fist.

But it is right at this point that previous phonologies have gone down any of several garden paths, all leading to the East it seems in search of that mysterious something they call “orientation.” Complexity and confusion begin here. Does the palm of the moving hand in these two signs face in or out, or up or down? Or has the palm disappeared altogether—as the lap disappears when a sitter stands up? Some sign linguists would reckon orientation as the direction in which a finger, or a...