Fine phonetic detail in the production and perception of reduced pronoun and auxiliary combinations in British English

This thesis investigates the production and perception of reduced function words, and the role of fine phonetic detail in spoken word recognition. In particular, it focusses on the phonetic features that convey linguistic information and maintain the contrast between pronoun and auxiliary combinations in British English. The thesis is divided into two parts. The first part reports on a qualitative and quantitative analysis of pronoun and cliticised auxiliary combinations, such as I’d, she’s, you’ll, in reduced speech. Data of high acoustic quality and high degree of reduction were elicited in a controlled phonological and prosodic environment. The auditory and acoustic analyses of the data collected revealed that in reduced speech, function words retain essential phonetic features that constitute the identity of the target words. The comparison between the phonetic features of contrasting paradigms that convey linguistic information, such as you’ll versus you’d, and she’s versus she was, revealed that the contrast is maintained by a combination of acoustic parameters, of which resonances and duration are the most prominent. The second part investigates the role of fine phonetic detail in the intelligibility of reduced function words, and the perceptual salience of selected acoustic parameters. A perception experiment indicated that listeners are sensitive to the fine phonetic detail that maintains the contrast in reduced speech, and that they can correctly identify highly reduced words even when they are presented in a minimal semantic context. A further investigation into the perceptual salience of duration and resonances in spoken word recognition indicated that the resonances play a primary role in the correct identification of reduced speech. Besides contributing to our knowledge of reduced speech, and confirming the role of fine phonetic detail in speech intelligibility, this thesis highlights the importance of carrying out a qualitative analysis and using a non-segmental approach in the analysis of reduced speech.