| Abstract | The conversion between phonetic message and acoustic signal, i.e., speech, is a grammatical code, similar in interesting ways to syntax and phonology. Being more accessible to experiment, speech should, therefore, be an inviting object of study for those interested in the psychology of grammar. Experiments on speech have already provided some information about the psychological processes associated with the use of grammatical codes. Each of these codes speeds communication by delivering the information in parallel. But the gain in speed is achieved at the cost of a considerable complication, since it is in the nature of the codes that they restructure the information; as a result, the levels they link do not correspond in the number or shape of their segments. It is argued that questions about the uniqueness of language are more likely to be answered satisfactorily at the level of speech than elsewhere. Studying language at the level of speech enables one to go beneath language behavior down to the mechanisms that underlie it. |