Project Details
Projekt
Occurrences, properties, and perception of reductions and deletions in spoken language and their impact for models of speech perception.
Applicant
Professor Dr. Henning Reetz
Subject Area
General and Comparative Linguistics, Experimental Linguistics, Typology, Non-European Languages
Term
from 2006 to 2014
Project identifier
Deutsche Forschungsgemeinschaft (DFG) - Project number 25227379
Running speech, rather than careful pronunciation, has not been studied intensively in linguistic research although listeners are exhibited to it in everyday¿s conversation. Especially, acoustic manifestations of words and phrases show a deviation from a `perfect¿ or `canonical¿ pronunciation with individual sound segments being reduced, assimilated, or even deleted. Nevertheless, listeners are capable to identify words and understand phrases, seemingly with ease. The main questions that are investigated by this project are: How are the `reduced¿ forms in running speech processed or are they stored by a speaker? Does a speaker of a language have variants or exemplars stored in his or her mental lexicon? Or, if there are only one or few prototypes stored, how are the variant forms mapped onto them ¿ is there a normalization or `re-do¿ procedure, or is there a representation cum mapping that can `ignore¿ acoustic variation? To clarify these questions, the proposed project will investigate which type of reductions occur in what (phonological and morphological) contexts and whether listeners are actually aware of these reductions. The outcome of the research will not only describe reductions and provide a typology of them but investigate the impact for different phonological, psycholinguistic and phonetic models of speech perception. The project will investigate databases of running speech for German and carry out production studies to document reductions and their contexts, perform gating and lexical decision experiments to gain insight into the perception of these reduced forms, and will try to verify processing hypotheses with brain imaging techniques.
DFG Programme
Priority Programmes
