ERIC Number: ED198696
Record Type: RIE
Publication Date: 1980
Reference Count: 0
Prosodic Encoding in Silent Reading.
In silent reading, short-memory tasks, such as semantic and syntactic processing, require a stage of phonetic encoding between visual representation and the actual extraction of meaning, and this encoding includes prosodic as well as segmental features. To test for this suprasegmental coding, an experiment was conducted in which subjects were given word lists to read silently. The last word in each list was to be read aloud. This word in each list was phonetically ambiguous, its syllabic stress dependent upon whether its syntactic function was that of a noun or a verb. The other words in the list carried stress identical to that of the "less-preferred" pronunciation of the ambiguous words, as determined unknowingly by the subjects beforehand. The less-preferred pronunciation was elicited in a majority of cases, indicating that there is a prosodic phonetic representation in short-term memory that is biasing pronunciation of the ambiguous word. It is hypothesized that cognitive parsing of written language parallels parsing in reception of spoken language, and that prosody is also used to mentally represent the phonetic structure of a sentence once it has been computed. Internal structural representations of reading, then, are based in an auditory rather than a visual medium. (PJM)
Publication Type: Reports - Research
Education Level: N/A
Authoring Institution: City Univ. of New York, NY. Graduate School and Univ. Center. Program in Linguistics.
Note: In its CUNY Forum, Nos. 7 and 8, p230-236, Fall 1979/Spring 1980.