| This dissertation explores some of the less well-studied aspects of prosodic focus in English Prosodic focus refers to the ways in which speakers use acoustic means such as fundamental frequency (F0) modulation to make one part of an utterance more prominent than other parts. A fundamental assumption is that the speaker aims to facilitate the hearer's understanding of the message at any particular time in a discourse, and uses intonation to guide the listener's interpretation of the utterance in relationship to the larger discourse context. We assume that the speaker is signaling the discourse purpose of the utterance presented by using differences in the intonational tune. One of the challenges for psycho linguistics, therefore, is to devise tasks that tap the listener's competence in interpreting the intended discourse purpose. That is, a challenge for psycho linguistics is to determine how exactly different prosodic manipulations contribute to the highlighting of a part of an utterance and how this highlighting contributes to the interpretation of the utterances.; While previous experimental work on the contribution of prosodic focus to the interpretation of an utterance has concentrated on matching question-answer pairs (Gussenhoven, 1983; Birch & Clifton, 1995), the experiments devised for this study required listeners to anticipate and make forward interpretations of the context and the moves set up by a speaker. The context in all cases is followed by a stimulus sentence with one of the following intonation contours: (1) H* H* L−L% (hat pattern); (2) H* L−L% (early nuclear accent); (3) H* L−H* L−L% (double nuclear accents); (4) H* L+H* L−L% (rising nuclear accent).; In one experiment, listeners were required to select the question that was best addressed by the stimulus sentence having one of these four intonation contours (Question Under Discussion), while in the second experiment, the listener was required to find the most appropriate continuation to the narrative, given the intonation contour.; Results from the QUD paradigm indicate that listeners reliably interpret the early peak contour as an early narrow focus, and that they can reliably differentiate between the phonetically rather similar contours 3 (more double focus responses) and 4 (more late focus responses). These phonetically similar contours differing primarily with regard to the alignment of the intervening low tone, were interpreted to be structurally different, making a phonological analysis of intonation contours imperative. This result points out deficiencies in studies using phonetic descriptions of tonal patterns, such as F0 only, to describe focal prominences.; Results from the continuation paradigm revealed that listeners needed to be pressed for time to perform the task, else they would only show a general preference for broad focus responses for all contours. We take this as an indication that listeners have to respond to a still fresh auditory memory to perform this task since listener's interpretation changed when the time interval allowed to reply to a stimulus was shortened. As expected, listeners then gave more double focus responses to contour 3 than 4, and more late ratings to contour 4 than 3, showing that they are sensitive to subtle phonetic differences when they are linguistically relevant. |
