Abstract

Available from UMI in association with The British Library. Requires signed TDF. ;This thesis is a metatheoretical discussion of approaches to the study of speech perception. It consists of three separate reviews, and an argument connecting them. The first review is of speech perception research as carried out in the information-processing approach, that of almost all work in this area over the last thirty years. The second is an overview of cognitivist philosophy, the conceptual foundation of the information-processing approach. The major tenets of this foundation are that cognition can be seen as computation, with information being represented symbolically, and the representations processed according to formal rules before being interpreted. The third review is of some relevant parts of the phenomenological philosophies of Husserl, Heidegger, and Merleau-Ponty. ;Detailed arguments are presented against the appropriacy of cognitivist philosophy as a foundation for speech perception research, using the work of Heidegger and Merleau-Ponty. The major factor in this inappropriacy is found to be cognitivism's view of the perceiving subject as a symbol-manipulator, for whom the task of speech perception is basically one of transformation of one representation of speech into another . Such a view neglects, among other things, the necessity of accounting for the possibility of there being any representation at all. Providing this account involves consideration of the nature of the human subject and its relation to the world. It is argued that, since descriptions are not given by the world in itself , but rather arise from a relation of meaningfulness between a person and the world, a human being must be more than a formal system. This discussion involves consideration of the researcher as a human subject, and the conclusion that descriptions of speech should give explicit consideration of the subject for whom the description is relevant, and especially should maintain a strict distinction between the perspective of the researcher and that of the subject being studied. ;The last part of the thesis shows the consequences of these arguments for speech perception theory, first by analysing the effects of the information-processing framework on speech perception research; and secondly by outlining an alternative framework for speech perception research based on more appropriate assumptions about the nature of human beings, arising from the preceding arguments. In particular, the task of speech perception is seen not as transformation but as constitution of descriptions. Finally some comparison is made of the interpretation of observations about speech perception in the two frameworks, and some suggested directions for future research are indicated.