Recent theoretical positions in both liguistics and psychology have emphasized the interaction of various levels of analysis in the production and perception of speech. The proposed research will explore the interaction of various domains by looking at acoustic measurements of spoken homonyms which differ in frequency either directly (e.g., """"""""through""""""""/""""""""threw"""""""") or by a difference in the proper name/common word dimension (e.g., """"""""Art""""""""/""""""""art""""""""). Acoustic differences, especially duration, will be examined to see if thay are evenly distributed across phonemes, indicating that the lexical status of a word exerts an influence throughout its production, as with small changes in speaking rate. Alternatively, thedifferences could be concentrated in one class of phonemes, suggesting that the articulatory plans themselves differ. Small changes in speaking rate will be induced experimentally to help sort out the possibilities. Proposing separate representations will require some rethinking of the links between phonological representations, but would argue for a division between phonology and the lexicon. Postulating a continuing link between the sound production and the lexicon will be compatible with theories which have no divisions within the language system, but will require an exacting control mechanism to account for the small size of the effects of frequency. While the primary purpose of using proper names as stimuli is to increase the number of homonymous pairs of English words, it is also of interest to establish whether they belong in the lexicon proper. This will be tested directly by seeing if the auditory presentation of a homonymous first name (e.g., """"""""Art"""""""") primes the associates of its homonym (e.g., """"""""museum"""""""". from """"""""art""""""""), has been found for common word homonyms. If names behave juat as common homonyms, then the definition of the lexicon must be expanded to include them, changing the conception of lexical neighborhoods, phoneme frequencies, and letter frequencies. If names seem to exist in their own lexicon, it will be important to establish that lexicon's connection with the phonology (since proper names are phonologically regular) and with the tests of language capabilitythat occur in such deficits as aphasia. The results will give us further insight into the central role of the lexicon not just in determining the distributional characteristics of a language, but of its ongoing implementation as well.
