Abstract
In the first stage of the study. five-hundred and forty-eight minimally structured 15 minute conversations were conducted with drug users in South Ayrshire, Glasgow. Lothian and Newcastle upon Tyne. Conversations were cued by a single question (“So what are you using at the moment?”) and took their own course thereafter. Each conversation was tape recorded and each tape recording was transcribed by a typist. A conceptual framework was developed based on the dimensions purposiveness, hedonism. generalisability. time, reductionism. attribution of addiction and contradictoriness. A typology of conversations was then identified on the basis of judges rating profiles on these six dimensions. Pilot studies of the reliability of the coding system showed high levels of replicability for this system between three judges. (r : range. 77 to. 90). The second stage of the study then examined the ways in which each of these types of drug conversation emerged in particular settings or with particular client groups. Note that no assumptions are made about the truth or falsity of the conversations obtained. The study showed that discourses of the types identified distinguished between groups who were in. or not in, treatment and could also distinguish between a variety of different patterns of drug use amongst clients both in and out of the treatment context. The implications of the study are that particular types of functional discourse. regardless of their semantic truth, emerge reliably as drug users move through the different stages of a drug use career.