Perceptual Chunking, Processing Time and Semantic Information

Abstract

One of the major Undings of current psycholinguistic research is that there is some sort of correspondence between the linguistic constituent structure and the perceptual process of segmentation of sentential material. Several data suggest that the incoming acoustic Information is chunked by the listener in terms of the constituent structure of the sentence which linguists would assign to it in formal linguistic analysis. The next crucial question is how the listener is able to assign constituent structure to acoustic Information in sentence processing.Attempts have been made to answer the question of how the incoming Information is chunked by the listener in speech processing. Fodor and Bever (1965), Garrett, Bever and Fodor (1966) and Bever, Lackner and Kirk (1969) claim that the process of perceptual segmentation of sentential Information is guided by the listener's own linguistic knowledge which has an innate basis. This claim is succinctly summarized by Miller (1967:34) äs follows: "The perceived boundary must have existed in the mind of the listener, not in the acoustic signal." In this claim the language user's ability to perceive syntactic relations is taken äs an empirical correlate of the abstract linguistic construct of competence. Levelt (1969:4) rightly complains that most of the theories of sentence decoding are "far more elaborate than can be corroborated from the data at hand." Fodor and associates' Interpretation of their click data to suggest that the underlying logical structure is perceived "immediately and directly" by the listener is an example of such theories. The ratio of inferences to data in Fodor and associates' work presented to account for the language user's ability to perceive syntactic relations in terms of linguistic competence is too high.The purpose of this paper is to show that the chunking of sentences cannot be explained in terms of linguistic competence. It cannot be the Brought to you by | provisional account Unauthenticated Download Date | 7/2/15 2:10 AM

Research Trends in Rate-Controlled Speech for Language Learning

“…A Canadian researcher (see Patel 1972) has challenged the psycholinguistic tenant that speech information coming to the listener is grouped (or chunked) according to his/her linguistic knowledge. Patel claims that as research on comprehension of compressed speech suggests that processing time is the important variable in sentence decoding, it follows we might well question evidence that syntactic processing must take place in some sort of 11 Semantic void."…”