The effect of semantic-pragmatic context on information status, as reflected in speakers' prosody | | Posted on:1996-01-05 | Degree:Ph.D | Type:Dissertation | | University:State University of New York at Stony Brook | Candidate:Chase, Polly Elizabeth | Full Text:PDF | | GTID:1465390014488362 | Subject:Experimental psychology | | Abstract/Summary: | PDF Full Text Request | | This paper describes the information status of words in a discourse context in terms of information theory. That is, the information represented by a word is relatively given when the word has a higher probability of occurring in its context, and is relatively new when the word has a lower probability of occurring in its context. Following information theory, I define the probability of a word's occurring in a particular discourse location in terms of its degree of contrast with other words that might have occurred at that location, but did not. This definition contrasts with other theories of given and new information, which tend to assume that information is either the surface forms of words or the referents of words.;Four experiments are described that provide supporting evidence. Subjects were given short texts to read aloud and their speech was recorded. Target words appeared in either of two conditions: in one condition they were relatively given and in the other they were relatively new, according to the specifications of the theory. The duration of the target words, relative to the clauses in which they occurred, was measured as an indication of information status.;Experiment 1 showed that a word may be a repetition of a recently mentioned word, and yet be treated as new information, when it forms a semantic-pragmatic contrast with a nearby word. Experiment 2 showed that a word may be newly introduced into a discourse and yet be treated as given information, when it has a high probability of occurring in its context. Experiment 3 showed that, under some conditions, a word may have a high probability of occurring in its context (as in Experiment 2) and yet be treated as new information, when it forms a semantic-pragmatic contrast with a nearby word (as in Experiment 1). Experiment 4 showed that the semantic-pragmatic contrast effect of Experiment 3 depends on the relative probability values of the contrasting words. | | Keywords/Search Tags: | Information, Word, Context, Semantic-pragmatic, Experiment, Probability | PDF Full Text Request | Related items |
| |
|