The thesis has as its consistent theme the questioning of the nature of empirically based sociological inquiry. Deriving its initial impetus from the Social Studies of Scientific Knowledge, Part I pursues the methodological implications of the problematic of reflexivity through an empirical research project - what counts for lay people, non-scientists, as knowledge - via the analysis of interviews with people for whom the irruption of Multiple Sclerosis has breached their belief in the nature and efficacy of medical science. The failed 'solution' of Part I's culminating chapter however - to embody reflexivity in the form of analytic writing - begins the turn of the thesis against its original grounds. In a pivotal central chapter, it is recognised that the transformative acts of recording and transcription means the transcripts (upon which the analysis has been based) are irrecoverably different from the original speech. The object of analysis is not talk, but texts. Attention shifts from the question of analysis to the data as writing, texts whose value for research depends on their being different from either fiction or pure speculation (philosophy). Drawing then on contemporary work in philosophy and literary theory, Part II pursues this as a threefold question of the datatexts' relation to their referents, re-reading them in turn through their bearing on the philsophical Question of the Self, against the quasi-literary form of autobiography and as dialogue, an event in (written) speech in which the knowledge was produced. As the work proceeds, it is recognised that their epistemological status derives, not from their ostensive referents, but in the work of reading against their form, producing particular iterpretable meanings. Finally, engendered by re-reading the datatexts against Plato's Phaedrus qua dialogic inquiry, the thesis shifts from the question of referents towards the chiasmic relation betwen the form of the written and the figurality, the rhetoric of reading, which raises fundamental questions about the knowledge produced in/by sociological inquiry on the basis of the interpretability of data. As itself a written text, the final form of the thesis - which has moved, not towards a forseen conclusion but by putting into question the grounds of each position as it is achieved yet consistently focused on 'the same' data - is of substantive theoretical import: essential to its questioning of what such knowledge 'is'.

Revised thesis (Ph.D.)--Cambridge University, 2002. / Includes bibliographical references (p. 153-164) and index.