DISCUSSION PAPER: COMMENTS ON GLEASON'S “GRAMMATICAL PREREQUISITES”

Gleason discusses expectancies of the hearer-reader, emphasizing the relation of nonlinguistic content to the interpretation of the meaning of a statement. Similarly, from my viewpoint, language can only be understood in relation to a theory of the structure of human behavior as a whole.’ In some sense, a sentence “means” expected impact-not the mere summation of bits of verbal features. Similarly, expectancies arise from memory of prior language experience. Thus, in coming into New York last night, I saw a sign reading, “The army wants to join you” (by paying you for training). It derives impact by contrast with the already known “Join the navy and see the world.” Expectancies from universe of discourse, in my view, are important also. Thus a context recognized as one of science fiction can markedly change permitted collocations -hence meanings-of words. When expectancies overlap or interlock, an expectancy set up by one feature may trigger a response which either does not need the redundant features for intelligibility-or one feature may drastically override contrasting clashing features. I remember my sharp surprise-and pained annoyance-when a student in Michigan, reading isolated sentences to me, led me to “hear” statements wrongly, i.e., to “hear” the wrong phoneme of a minimal pair, when lexical, grammatical, or cultural probabilities (nor impossibilities) teased me to false expectancies. And, finally, I am in agreement with Gleason that psycholinguistic reality must be treated along with-and controlling-the evaluation * and the choice of any formalism, if we wish to achieve practical results tying grammar more closely to lexicography.