Explanations and expectations: drug narratives among young cannabis users in treatment.

This article analyses how young people enrolled in drug addiction treatment in Copenhagen, Denmark, explain their cannabis careers and how they view their possibilities for quitting drug use again. Inspired by Mead and narrative studies of health and illness, the article identifies four different drug use 'aetiologies' drawn upon by the interviewees. These cover childhood experiences, self-medication, the influence of friends and cannabis use as a specific lifestyle. A central argument of the article is that these explanations not only concern the past but also point towards the future by assigning the interviewee a more or less agential position in relation to drugs. Further, the drug narratives are viewed as interactional achievements, related to the social context in which they were produced, namely, the institutional setting of the treatment centres. The article is based on 30 qualitative interviews with young people in drug addiction treatment.

Introduction

This article uses qualitative interviews to analyse how young people enrolled in drug addiction treatment in Copenhagen, Denmark, explain their cannabis careers and how they view their possibilities for quitting drug use again. We develop a typology consisting of four different narratives, which the interviewees draw upon when depicting the development of their drug use. As we will show, the interviewees’ ‘affiliation’ with a specific type of narrative is to some extent contingent, meaning that for many of them, other ways of accounting for their drug use could also have fitted. A central aim of the article is to show that images of the past are related to images of the future: the way the interviewees narrate their cannabis career up until now is reflected in the way they anticipate their relationship to drugs in the years to come.

The article focuses on young people with cannabis problems, a group that has been the object of relatively few sociological studies (e.g. Hammersley et al. 2001; Miller and Plant 2002; Hammersley and Leon 2006; Hathaway et al. 2008). For decades, much research on drug addiction treatment has focused on opiate users, presumably motivated by this group’s dominant position among drug addicts in general. However, in the last ten years the characteristics of those seeking drug treatment have changed. Among 5,686 users enrolled in drug treatment in Denmark in 2011, 63 per cent reported cannabis to be their main drug, whereas 17 per cent reported opiates (National Board of Health 2012). In contrast, in 2000 only 30 per cent of 3,920 users reported that cannabis was their main drug, while 54 per cent reported opiates (National Board of Health 2005). Among young people (18–24 year-olds) enrolled in treatment in 2011, 80 per cent reported cannabis as their main drug; the corresponding number in 2003 was 46 per cent (National Board of Health 2012). A similar trend is seen in a number of other countries such as Germany, France, Hungary and Belgium (EMCDDA 2012). Thus, cannabis has become the main drug for a still larger share of the treatment population, particularly among young people.

When considering typologies in the existing literature on drug use, these have primarily centred either on the extent of the actual drug use or on the characteristics of the users. With respect to drug use, researchers have described this in terms of controlled use, problematic use, abuse, etc. (Hammersley and Leon 2006; Ramo et al. 2010; Järvinen and Ravn 2011). With respect to the characteristics of the users, studies tend to apply a psychological or psychiatric approach focusing on users’ behavioural patterns and/or personality traits (Yacoubian 1999; Miller and Plant 2002). Finally, some studies have taken a combined look at both users and drug use, for instance Patra and colleagues (2009), who showed how categories of drug users (e.g. opioid versus non-opioid) differed in terms of health status, housing situation, illegal income generation etc. (Wilkinson et al. 1987).

In contrast to these studies, we focus on the interviewees’ narratives about their drug use development, and not on the amount of drugs currently used or the consequences of drug use in terms of illness or material problems. One among few examples of such a line of research is Sandberg (2012), who analysed three different narratives, or ‘discursive repertoires’, which Norwegian cannabis users drew upon when describing their cannabis use: normalisation, celebration and risk denial. Other researchers as well have shown that drug narratives often serve the function of legitimising the interviewees’ current use by emphasising control or pleasure (e.g. Pennay and Moore 2010; Chatwin and Porteous 2013). However, in contrast to our article, these analyses do not relate the interviewees’ current situation to their relationship to the past and future. An approach that comes closer to ours is represented by Biernacki’s (1986) and McIntosh and McKeganey’s (2000) studies of recovering addicts’ narratives, showing how the interviewees struggle to construct a new non-addict identity through a reinterpretation of their former lifestyles. We will continue along these lines with an even greater focus on the temporal dimension – by relating the interviewees’ explanations of their drug use to their expectations regarding the future, and by addressing differences in the drug use aetiologies depicted in their accounts. In the following section we present the theoretical frame of the article.

Theoretical frame

To inform our analytical approach, we use Mead’s theories on time (1932/1959) and the self (1934/1964). When Mead is quoted in sociology, it is mostly for his theory of the self. Mead’s ‘remarkable theory of the past’, however, ‘has not been mined for its relevance to social concerns’, Strauss (1964: xiii) pointed out in his introduction to George Herbert Mead on social psychology. Fifty years later, this is still the case. Mead’s ‘perspective of the present’ has not received much attention, at least not within empirical sociological research.

In Mead’s (1932)/1959 theory of time, people’s narratives about the past are relative and dependent on their present life circumstances and orientations. In this sense, the ‘what it was’ is established through the ‘what it is’. Mead (1932/1959: 23) writes: ‘It [the present] marks out and in a sense selects what has made its peculiarity possible. It creates with its uniqueness a past and a future. As soon as we view it, it becomes a history and a prophecy’. From the perspective of the present, life stretches out in two directions: towards the past which we interpret in multiple and changing ways, and towards the future which we plan and try to anticipate (Järvinen 2004). Both our past and our future are, in Mead’s (1938/1972: 97) description, ‘hypothetical’. There is always more than one potential life story to be told in a person’s life, just like there is a range of imaginable futures. Narrating a life is a selective affair of making sense out of a myriad of experiences and possibilities. Hence, when the young drug users in our study ‘explain’ their cannabis problems in a specific way, the theory proposes that they could also have chosen other explanations, emphasising other life events, and that their choice of explanations is related to their present situation and to their orientation towards the future. In this perspective, people’s plans for the future may influence their interpretation of the past. In Mead’s (1932/1959: 3) words: ‘We realise that the world that will be cannot differ from the world that is without rewriting the past to which we now look back’. But the mechanism also works the other way around: our interpretation of the past has consequences for what we anticipate and plan for the future. In other words, the stories people tell about their past may change their life and give it a specific purpose and meaning, just like the purpose and meaning people focus on in their present life may change their understanding of the past.

Mead’s theory of time is closely interwoven with his theory of the self. With every reconstruction of the past, the self reconstructs itself, or in Mead’s (1938/1972: 72) words, ‘a new past is always a criterion of a new self’. Mead’s theory of time (and the self) is persistently social, emphasising the interactionist dimension of the present, in which the self’s interpretation of its past and future is anchored (Järvinen 2004). According to Mead (1934/1964: 244), an individual can narrate his/her life in a meaningful way ‘only on the basis of social relations and interactions, [and] only by means of his [or her] experiential transactions with other individuals in an organised social environment’. Yet, social selves are not socially determined selves. Although Mead stresses that social selves are always constructed with reference to the communities to which they belong, he also emphasises the active role of the individual in shaping his or her destiny (Järvinen 2004). The present, out of which people’s ‘organisation of a perspective’ (Morris 1972: iii) arises, is unconditionally a social present, but the individual human being is also creative and enterprising, and in a very profound sense unpredictable, in the ways he or she deciphers the past and projects the future.

Mead’s reasoning on time and the self has a lot in common with ‘the narrative turn’ in studies of health and illness – and in qualitative sociological research in general – beginning in the early 1980s (e.g. Bury 1982, 2001; Williams 1984; Riessman 1990; Hydén 1997). Despite this kinship, Mead’s theory of time has seldom been used in narrative studies of health and illness.

Three points from narrative health and illness research may be singled out as especially important for our Mead-inspired analysis. First, several scholars, with Bury (1982) and Williams (1984) being among the pioneers, have shown that illness forces people to reconstruct their identity and reorient themselves in order to create continuity in their life history. Illness narratives are a means for people to deal with their changed life situation, and with the identity problems that the transformed relationship between their body, self and surroundings has brought about (Bury 1982; Hydén 1997). Second, illness narratives may be seen as a medium for discussing possible explanations for the illness (Williams 1984; Hydén 1997; Bury 2001). Bury (2001) talks about ‘contingent narratives’, that is, narratives about which factors have caused or influenced the development of illness, and ‘moral narratives’ related to blame and self-worth. Such moral narratives may construct illness as caused by circumstances the individual cannot influence or, alternatively, as a self-afflicted condition resulting from ‘inappropriate’ behaviours (Bury 2001: 274–5). Third, health and illness researchers stress that illness narratives incorporate cultural ideas, forming the individual’s experiences and making them socially recognisable (Hydén 1997; Riessman 2003). Bury (2001: 278) uses the term ‘core narratives’ to capture the idea that narratives are always presented within ‘cultural settings which provide specific forms of language, clichés, motifs, references and other elements of linguistic and symbolic repertoires which allow and constrain what is said and how it is expressed’.

The combination of Mead (1932/1959) and the narrative tradition within health and illness studies leads us to focus on two dimensions in our interviews with drug users. One analytic dimension is the explanations of drug problems put forward by the participants. Narrating a life affected by drug problems means tracing the lines backwards, from the present towards individual events and circumstances identified as significant for the problems (Järvinen 2004). Descriptions of the past cannot, in the perspective of Mead, be separated from explanations, because narratives always extract configurations out of mere successions and because these configurations are typically causal (see also Ricoeur 1984: 152). Mead’s (1932/1959) theory of time puts emphasis on ‘emergence’. Discontinuity, or emergence of something novel and unexpected, demands that people reconsider their life (Mead 1932/1959: 31ff.). When an individual faces discontinuity, he/she starts the arduous task of reconstructing the past and future in terms of it. The abruptness of a new present is mitigated by the new perspective on the past and future, a perspective from which the emergent becomes understandable and the future tolerable and manageable. ‘The present structures time by finding in the past the conditions for the emerging event and predicting consequences for the future’ (Järvinen 2004: 50). According to Mead, explanations of the emergent are contingent. Narratives about the past are always simplifications of things that actually happened; they are stories that single out a limited number of factors, actions and incidents from a wide range of possibilities. In Tilly’s (2006: 70) words, stories ‘omit a large number of likely causes, necessary conditions and, especially, competing explanations of whatever happened’.

Another analytical dimension in the article is the relationship between the interviewees’ accounts of the past and their orientation towards the future – a relationship that, in our view, has received too little attention in narrative studies of health and illness. As mentioned above, we regard drug problems as a factor (an ‘emergent’ in Mead’s terminology) challenging the interviewees’ identity constructions and calling for a reconsideration of their life history. Inspired by Mead’s theory of time we privilege the temporal dimension of the narratives by including both the interviewees’ explanations of their drug problems (pointing back in time) and their imagined future relationship to drugs (pointing forward). In Stone’s (1989: 296) words, causal reasoning about the past is consequential for future actions because different choices of explanations ‘locate the responsibility and burden of reform differently’. Explanatory narratives are processes of image making, and images of causes are related to images of consequences, solutions and responsibility. Hence, a specific explanation of a problem (in our case drug problems) is not only retrospective but also prospective because it indicates ‘a locus of control, a plausible candidate to take responsibility for a problem [and] a point of leverage to fix a problem’ (Stone 1989: 289). Thus, in the analysis of our interviewees’ narratives we will show how some explanations of drug problems tend to be associated with relatively agentic relationships to problem solutions, while other explanations are more fatalist both when it comes to causes and future control of drug use. Before we turn to the analysis, we describe the data material that forms the basis of the article.

Methods and data

The article is based on 30 qualitative interviews1 with young people enrolled in drug treatment in Copenhagen, Denmark. The participants were enrolled at three small centres (Centre A: 14 interviewees, Centre B: 12 interviewees and Centre C: 4 interviewees), all offering outpatient treatment. Most interviewees (27) were recruited through the staff at the treatment centres who introduced the research project to potential participants. If interested, the young drug users would provide their mobile phone numbers and would then be contacted by the interviewers. The rest of the participants were contacted directly by one of the interviewers who was given the opportunity to observe a group therapy session where she introduced the study. After the session, those who were interested contacted her and made appointments for interviews (three people).

Most interviews took place in a meeting room at the treatment facilities; three were conducted at the researchers’ work place or in a nearby park. The interviews lasted 45–80 minutes, most of them around one hour. They were semi-structured and based on a schedule consisting of four overall themes, which covered drug use development and drug ‘career’, present drug use, expectations when seeking treatment, and experiences with being in treatment. Further, the interviewees were asked about their familial background, current living situation and plans for the future. The project followed prescribed ethical standards for social science qualitative research (cf. Flick 2007; Heath et al. 2009). Thus, participation was based on informed consent, and participants were allowed to withdraw at any time. All interviews were audio-recorded, fully transcribed and anonymised with respect to names, places and other identifiable markers. In Denmark there are no institutional boards for the approval of social science studies, but the study is in accordance with the Helsinki Declaration and with the ethical guidelines for the social sciences specified by the Danish Council for Independent Research.

The sample consisted of 19 men and 11 women. Their mean age was 21.7 years, ranging from 17 to 28, with the majority (21 people) being between 18 and 23 years of age. While 10 were in regular jobs (some half-time or temporary) or enrolled in education, the majority received welfare payments and took part in various activation or educational projects in relation to this.

The interviewees attended the treatment centre once or twice a week. Around half of them were enrolled in group therapy sessions, or group therapy in combination with individual therapy, while the other half only attended individual sessions. The type of treatment they received varied across the sample: some attended structured conversations with a social worker, others had therapeutic sessions with a psychologist, and some attended sessions with a psychiatrist. According to the centres’ homepages and as described in interviews with staff members, all three centres worked with combinations of systems therapy, cognitive therapy and narrative therapy. We will return to the centres’ characteristics later.

The narratives identified in our study should be viewed as interactional achievements that are co-constructed between interviewer and interviewee: ‘both are looking for a way to understand and articulate the illness and the illness events as a meaningful whole’, as Hydén (1997: 61) puts it when describing illness interviews (cf. Williams 1984; Holstein and Gubrium 1995; Järvinen 2008).

The interviews can be seen as an ‘elicited context’ (Hydén 1997: 62) in which the interviewee is expected to be able to provide a meaningful story about the development of his or her drug use. Thus, the narratives are social actions (Holstein and Gubrium 1995; Gubrium and Holstein 1998) which serve a specific purpose and underline a specific performance. Also, the narratives should be seen in relation to the institutional context in which they were produced: the fact that the interviewees had been enrolled in treatment for a shorter or longer period of time when we met with them may very well have played a part in the facilitation of specific narratives. The stories were not only presented in the context of the interview, but presumably also drew on interactions taking place during treatment sessions (Andersen 2015). Hence, among many other influences, variations in narratives may mirror differences in the centres’ treatment approaches to drug problems.

When analysing the data, both authors read the interviews several times in order to identify narratives about the aetiology of cannabis problems. Four distinct narratives were found (described below), each offering a specific set of reasons for and rationales in the interviewee’s drug use career. Contrary to some studies on lay attitudes to health and illness (cf. Blaxter 1997; Popay et al. 2003), the interviewees were not asked directly to discuss ‘causes’ or ‘responsibility’ for (in this case) drug use problems, neither in general nor in terms of their own use. Instead, the explanations and rationales that we analyse in this article were brought forward as part of the interviewees’ depictions of their drug use careers. In some instances, this took the form of an explicit ‘causal explanation’ provided by the interviewee (‘I started smoking cannabis, because …’), while in other cases this manifested itself in the framing of the interview as a whole, that is, that a specific explanation took up much time and reoccurred several times.

It is important to note that many interviews contained elements of more than one explanation but that most interviews tended to be dominated by one type of narrative. Accordingly, as a second step of the analysis, the participants were classified into four groups dependent on which narrative was most prominent in their interview. Both authors categorised the interviews separately after which we compared the categorisations. Twenty-five interviews were categorised in accordance with one dominant explanation, while in the five remaining interviews, two or more explanations seemed equally central. In the final step of the analysis, we looked at accounts describing the participants’ orientations towards future drug use, for example, their answers to the questions ‘What is your plan for treatment; do you want to stop or reduce your cannabis smoking?’ and ‘Do you find it difficult to stop or reduce your cannabis smoking?’ We then related the accounts about future prospects to the explanation(s) of cannabis use most prominent in each interview.

Four types of cannabis narratives

As mentioned, the four types of drug narratives each offer an explanation of why one (mis)uses cannabis. One explanation focused on childhood experiences as being decisive for the development of cannabis problems. Typically, such experiences would cover growing up with parents who had alcohol or drug problems, but other traumatic experiences in childhood (e.g. parents’ illness and death; violence) were depicted as well. Another explanation described cannabis use as self-medication, meaning that initiation and later escalation of cannabis use was seen as a consequence of emotional problems (ADHD, anxiety, depression). A third explanation focused on the influence of peers, the logic being that hanging out with the ‘wrong’ people (friends, lovers or acquaintances) was a decisive factor for the interviewee’s drug use development. The last explanation identified in our data viewed cannabis use as part of a deliberately chosen lifestyle and specific way of ‘thinking about the world’.

In the following subsections, we analyse the four narratives in turn. We describe individual cases at some length to illustrate the logics of the explanations activated in the interviews and how the four drug use aetiologies are combined with the interviewees’ orientations towards the future. The aim is to show that causal reasoning is a form of image making (Stone 1989) that may be consequential for the future. Hence, we will show that some types of narratives are relatively fatalist, depicting drug use as caused by forces the individual cannot control, and drug problems as difficult, if not impossible to solve. Other narratives are more agentic, describing drug problems as resulting from the individual’s own actions and often also as amenable to intervention and control.

Like parents, like children

Twelve of the 30 interviewees depicted ‘problematic’ childhoods or childhoods with traumatising experiences (parents’ alcohol and/or drug problems, parents’ illness and death, violence). While some of them merely mentioned such experiences and then moved on with their narrative, relating their drug use to other explanations, other interviewees (six people) described a direct causal relationship between their childhood problems and their drug use career – the present subsection addresses the accounts of these interviewees.

Sarah (22) explained her cannabis use in this way: ‘The reason why I took too much of everything was that my mum was an alcoholic and my dad smoked loads of pot, so my childhood wasn’t easy, right, and I just wanted to get away. I smoked pot in order to be completely gone’. According to Sarah, her cannabis smoking was not a choice but a necessity: ‘I didn’t choose to use drugs. How could an 11-year-old choose such a thing […] If there had been alternatives, I would certainly have acted differently. I just wanted to get away’.

Sarah’s depiction of cannabis smoking as ‘a necessity’ in her childhood was associated with a pessimistic view on her future relationship to the drug: ‘Once you become addicted to the crap, once you develop a taste for it, you can’t leave it’. Sarah repeatedly compared her own dependence to her parents’ addiction problems which they had ‘never been able to solve’. Her own wish was to stop smoking cannabis should she ever have children of her own but she expected cessation to be extremely hard: ‘It’s going to be a real tough, long way for me to go […] I guess I’ll always have this desire for it’. Hence, in Sarah’s narrative, her parents’ problems were not only the cause of her own problems but also an indication that she would probably never get rid of her cannabis addiction.

Other participants in this group described how they had been introduced to cannabis by their parents. Line (23) related:

I was 14 when I first smoked pot. It was with my mum because she was of the opinion that if I should try it, I should do it in her company because at least she knew what she was doing […] And I wanted to try it because I wanted to find out why my mum used it […] why did she sit there smoking this brownish matter and why did it make her so relaxed?

Line put a lot of emphasis on the fact that her parents were drug users, answering the very first question in the interview (‘Can you please tell me a bit about yourself, who are you and how old are you?’) like this: ‘I am Line, and I am 23, and I live in [part of Copenhagen]. I am the daughter of a pusher and a heroin addict, and in connection with that I have been let down, which has made me into the person I am, and this is also part of the reason why I smoke cannabis’.

Like Sarah above, Line was rather pessimistic about her possibilities to control her smoking. She said she had smoked 15–20 joints a day for many years: ‘That’s what my life was about, getting up in the morning in order to smoke, being so stoned that I couldn’t stand up […] I could only survive if I smoked all these joints’. At the time of the interview, Line had diminished her cannabis use to five joints per day and said she wanted to stop completely but that she found it difficult. She had always used cannabis as a way of coping with difficulties in life, she said, and even if she stopped, she would probably return to massive use at the ‘slightest adversity’ because this was how she had ‘learnt to handle problems’: ‘I am the way I am because I have been exposed to these things’.

When considering the interviews in which cannabis problems were explained by reference to parents’ alcohol and/or drug problems as a whole, one finds a certain degree of determinism in the narratives. Cannabis smoking was depicted as a ‘natural’ or ‘necessary’ reaction to childhood experiences and as somehow inevitable also in the future. Compared to other interviewees in the study, these participants tended to see cannabis problems as chronic and as something that depended, not so much on one’s own will and agency, as on external circumstances (in childhood as well as later in life) and on the behaviour of other people. Read as ‘moral narratives’ (Bury 2001) the accounts seemed to stretch the role of the child, comprehended as a helpless victim of his/her parents’ misbehaviours, into the present. Because the child ‘had no choice’ in the past, the young adult drug user has no choice either.

Using cannabis as self-medication

Another type of narrative that tended to place the cause of drug use – and the solution of the problems – outside of the individual’s sphere of influence was the one describing cannabis smoking as self-medication. Compared to the first narrative this explanation did not locate the cause of drug problems in other people’s behaviours but rather in an unwilled and uncontrollable force inside the individual. At least ten of the interviewees had been diagnosed with mental health problems, for example, ADHD, anxiety or mild depressions and some were still undergoing tests and awaiting a diagnosis. While some of these nonetheless emphasised other explanations such as a problematic childhood or ‘bad company’ (see below), we here focus on those interviewees (seven people) who saw their mental health condition as the primary explanation for their drug use. These were participants who offered direct causal explanations along these lines (‘I smoke cannabis because of X (diagnosis)’, ‘x is the reason behind my drug problem’) and/or whose narrative was filled with descriptions of using cannabis as ‘medicine’ for various mental health problems.

Tanja (25), who started drug treatment two months prior to the interview, explained why she sought treatment in this way:

Well, all the time I have known there was something emotional which I was just trying to soothe, something which I hadn’t taken care of. And so it just got worse and worse, and in the end the drugs did not work any longer, they just made me feel bad. So the turning point for me was when I was feeling so bad that I couldn’t take any more drugs nor could I bear to do without them.

Tanja’s story included references to a drug-using network of friends as well as childhood problems. She said she had ‘been exposed to certain things’ (not further specified) during her upbringing, which was ‘a bit difficult’. However, she did not dwell on any of these experiences when reasoning about her cannabis use but rather focused on and emphasised her emotional problems throughout the interview: ‘I have always had this urge [to dope myself], “well then there’s cannabis” which I could smoke, because then I could relax’. At the time of the interview Tanja had not used drugs for two weeks. In treatment, she saw a social worker as well as a psychiatrist, and while the first assisted her with ‘practical stuff’, her psychiatrist ‘wants to help me get rid of the reason I am taking drugs, which I guess is probably that anxiety thing’. Thus, because Tanja used cannabis to soothe her emotional problems, she viewed her future prospects for quitting drugs as closely related to whether her anxiety and other mental health problems would be ‘cured’.

Another interviewee who talked about her drug use as ‘self-medication’ was Lise (26). When Lise started treatment 1½ year ago, she took a whole range of drugs, but at the time of the interview she had stopped using drugs other than cannabis. She was diagnosed with ADHD and borderline personality disorder. Lise did not regard cannabis as a drug: ‘Well, I smoked cannabis [prior to trying other illegal drugs]. But in my view cannabis is not a drug. Cannabis, that’s my medication. So I have been smoking [cannabis] since I was 13. Every day. Drugs, that was not until I turned 17’.

Lise’s description of her drug use development was complex and included many potential explanations. She could have emphasised her upbringing and living situation (she was thrown out of home by her mother when she was 17 and had been alternating between short-term rentals and being homeless since then) or the influence of friends and boyfriends, but instead she focused on her own ‘addictive personality’ and ADHD:

If I don’t get [cannabis], I become aggressive, I become mad, I become sad, I cannot sleep at night […] There’s no reason to stop when it [smoking] is not a problem for me […] It’s just who I am […] I have tried being clean, I was clean for six months, from cannabis and everything, and it was the saddest time of my life. I have never felt so bad. I did not fit in anywhere. I am not a normal person […] I only have people with diagnoses as friends.

Unlike Tanja above who said she wanted to stop but was unable to do so, Lise neither could nor wanted to stop, because she felt a life without cannabis was intolerable. Further, her self-presentation as being different from ‘normal’ people also referred to the groups of friends she hung out with. Instead of describing them as ‘bad company’, that is people whose behaviour had influenced her own behaviour in negative ways (cf. the third type of narrative below), she regarded herself and her friends as persons who, each for their individual emotional reasons, needed drugs. Cessation of drug use therefore was not really a possibility for Lise and her friends.

Like the narrative tying drug use to problematic childhoods, the self-medication narrative contained static elements. Interviewees depicted their emotional condition as more or less chronic and as necessitating use of cannabis in the past as well as in the future, unless other effective drugs (such as ADHD medicine or tranquillisers) could relieve their suffering. Some interviewees were intensely preoccupied with their contacts with the treatment system, especially psychiatrists testing them for different conditions and prescribing them legal drugs. To the extent that they pictured agency in relation to their future drug use, it was more in terms of the system’s handling of them as clients, than agency as self-change and individual ‘becoming’ (Riessman 2003: 21).

Hanging out with the wrong people

In seven interviews friendship relationships were identified as the dominant causal narrative. Again it is important to note that some of the interviews contained elements from other types of narratives (negative childhood experiences and/or emotional problems) but that the participants quoted in this section did not ascribe these elements direct causal status. Instead they focused on relationships with friends (or sometimes, partners) when talking about why and how they developed a problematic cannabis use.

Michael (22) described how he started smoking cannabis at age 14 because this was common among his peers: ‘My childhood friends lived just down the block […] and some of their friends were into it as well, so everybody knew each other […] and everybody had their time with it [smoking]’. At the time of the interview, Michael had been in treatment for around 2½ years. His ambition was to gain control over his cannabis use and return to occasional use again, but this was a difficult process:

Often, in the morning, you think: ‘Today I won’t smoke’ but then your friends call and ask if you want to hang out. And of course you do! And ‘what should we do?’ ‘We’ll smoke cannabis as we always do!’ Then it just goes in circles from there, because everybody around you smokes.

However, four months prior to the interview, Michael got a flat in the city and moved away. Ten days later he quit smoking cannabis. He described the fact that he moved (20 km) as decisive for his ability to stop smoking because he could no longer see his friends every day. Further, Michael did not consider the company of his friends as interesting as he used to: ‘sometimes it’s okay [to be with them], and other times they are completely stoned and you can’t get into contact with them […] then I am happy that I have quit and that I am able to do other things than just sit there dangling’. Michael said he still hangs out with his friends now and then. For these occasions he had developed small strategies in terms of, for example, drinking a beer instead of smoking a joint.

In addition to groups of friends, social networks grounded in more institutional settings such as schools or work-places may also play a part in the process of both initiating and escalating one’s cannabis use. Like a number of other interviewees, Leo (21) described how he started in a new school and soon singled out the other cannabis users, with the result that his cannabis smoking increased considerably:

I almost immediately realised that more than half of them were smoking. I quickly located two other guys, and we ended up […] not getting a thing of what the teacher was explaining, and then we decided ‘well, let’s go to Christiania [where cannabis is easily available]’. And that’s when it became disastrous, because these two friends, with whom I spent most of the day, both smoked a lot.

Leo’s narrative emphasises how cannabis users may be seen as destined to hang out together; the fact that the student population at the school also had a large share of non-smoking students did not really matter. The interview with Leo also contains information about traumatic experiences in childhood – his father died of alcoholism when Leo was a young teenager. Unlike the interviewees in the first section (‘Like parents, like children’), however, Leo did not put forward this as an explanation of his drug use but merely mentioned it in passing. Instead he, all the way through his story about cannabis initiation and escalation, described his cannabis smoking as a consequence of ‘bad influence’ from friends. At the time of the interview, Leo had not used cannabis for two months and was optimistic about his possibilities for staying away from it. He had recently met a new girlfriend who did not smoke cannabis, and identified this as a main reason behind ‘the turn for the better’ in his life.

Michael and Leo are examples of interviewees emphasising the major influence of their social surroundings on the development of their cannabis use. In this, they presented themselves as being ‘victims of the circumstances’ as well as being ordinary, ‘curious adolescents’ fascinated by the cannabis experience. Starting to smoke cannabis and developing a problematic use was in this sense something that ‘just happened’ in consequence of their belonging to specific peer groups. Likewise, as demonstrated, they saw their cessation of cannabis use as equally influenced by social networks.

However, participants drawing on this narrative tended to describe their friends’ roles as more important for their past relationship to cannabis than for their present and future prospects. In this sense, they depicted themselves as more autonomous and enterprising today than they were when they, inspired by friends, started to smoke cannabis and gradually escalated their use. Thus, in terms of agency, the social networks narrative deviated from the ‘like parents, like children’ narrative above, although both stressed external influences on the interviewees’ drug use development. Whereas the latter mirrored well-known cultural rationales about problematic childhoods forming individuals’ whole lives, the former was less deterministic – and, read as a ‘moral narrative’ (Bury 2001), more ambivalent. On the one hand, social networks were offered as the explanation of drug use, structuring the ups and downs of the participant’s drug use development. On the other hand, many interviewees drawing on this narrative depicted themselves as agentic and accountable individuals who – in the present, if not also in the past – had the possibility to opt out of ‘negative’ networks. This again seems to reflect broader cultural understandings of ‘bad company’ being a contributor to the development of youth problems (here excessive drug use) but not a factor determining the individual’s identity and room for manoeuvre in the long run.

Cannabis as a ‘lifestyle thing’

The last way of explaining cannabis use (represented by five people) that we find in our interviews relates to a specific lifestyle and/or philosophy. This was the type of narrative that most clearly attributed the causes of drug problems to the individual drug user him/herself. It was also a narrative where interviewees were relatively optimistic when it came to drug use cessation or diminishment. One of the participants drawing on the lifestyle explanation was Christian (20), who had smoked cannabis on a daily basis since he was 16. He said the following about why he started to use cannabis:

I have talked about these things with my therapist […] and I can rather easily explain it. I had these feelings of meaninglessness related to the world and how the universe works. I have always been a thinker and when I was 16 I started to smoke for real […] All of a sudden I was smoking every day and it was because of this meaninglessness-trip […] And from then on it was a bit difficult to be with my peers because they were so ignorant.

Just like the ‘diagnosed’ interviewees described earlier, Christian felt that he was different from other people, but unlike them he saw this as a difference to his own advantage. He felt that others were unable to understand his philosophy of life: ‘People may very well be provoked: ‘But what do you do, and shouldn’t you get yourself an education?’ ‘Absolutely not, I just want to relax’. Christian called himself a ‘nihilist’, explaining that ‘there is nothing that is wrong or right, and morals are just bullshit […] I don’t like the thought of a perfect life, living in balance […] I like imperfection, I like it when you have some problems’. He also dissociated himself from other ‘hashers’: ‘Some hashers are just spaced out, boring and repulsive when they smoke large quantities, while people like us try to balance it [laughs] […] For us cannabis is a lifestyle thing’. Yet Christian also said that he had not been very successful in ‘balancing it’. He had had periods with a ‘massive use’ when he just ‘lied apathetically in [his] bed’, he had dropped out of school twice and lost two jobs ‘due to [his] smoking’ and felt bad about this: ‘I generally regard myself as a capable person, and people always tell me ‘You are resourceful’, but being kicked out [from school and work], these are real loser-things’.

Like the interviewees in the second subsection above, Christian also talked about cannabis smoking as a means with which to combat ‘restlessness’ and ‘swarming thoughts’: ‘Ever since I started to smoke cannabis, I have never been able to relax without it […] My head is rushing. I have never had an ADHD-diagnosis and I don’t necessarily think there is a diagnosis for this’. Thus, unlike the interviewees above, Christian did not identify himself with ADHD, or describe his cannabis use as medication. He also tended to regard his restlessness as something that had developed after he started to smoke cannabis regularly, rather than a reason for his smoking it in the first place.

Freja (20) was another interviewee who stressed ‘ideological’ reasons despite other potential explanations in her life story. For one thing, her mother died when Freja was a teenager and immediately after this Freja started to smoke cannabis. Freja also mentioned psychiatric problems, for which she had been hospitalised. However, she did not tie these things together in the interview but merely mentioned them. Furthermore, the interview with Freja was filled with accounts describing how she was introduced to drugs by her friends, and how her drug use fluctuated in accordance with her social surroundings. Yet she did not regard her cannabis use as dependent on her friends’ use; rather, she saw her choice of friends as a consequence of her ‘attitudes to life’ and of her interest in cannabis. Freja said the following about the background of her cannabis use:

I had this theory that I’d rather be stupid and happy than – you know […] Already at 16, I arrived at the conclusion that […] everything is artificial, all things are made up. There is just as much reason to be an alcoholic as to go out earning heaps of money. It’s absolutely unimportant what you do.

After several years of intense cannabis smoking, Freja started in treatment and she had now refrained from cannabis for six months. She hoped to develop a ‘normal relationship’ to cannabis in the future, meaning that she would be able to smoke it every now and then, without returning to daily use. She was of the opinion that drug problems are always solvable if only you are willing to do something about them yourself:

I think many of the others here [at the treatment centre] are a bit stupid because they don’t take responsibility for what they do […] Not that I want to sound self-righteous but we all have things in our life, we all have our reasons. But that doesn’t mean you can’t do something about your problems. I can’t stand it when people say they cannot quit.

More than other participants in the study, Freja, Christian (above) and other interviewees drawing on the lifestyle narrative focused on taking personal responsibility for the solution of their cannabis problems. They explicitly rejected other types of drug use aetiologies, describing them as attempts to avoid ‘personal blame’. As regards their own drug use, they saw it as controllable. Because cannabis smoking was a preferred lifestyle, and not the result of other people’s bad behaviours or one’s own emotional problems, cessation of cannabis use was a question of personal will and determination. If the interviewees had not yet reduced their smoking, it was because they had not yet decided to do so, they reasoned.

Discussion

In our complex world, Tilly (2006: 65) says, ‘causes and effects always join in complicated ways’. There are simultaneous explanations, incremental effects, unintended consequences and feedback, but stories ‘exclude these inconvenient complications’ by grossly simplifying the processes involved (Tilly 2006: 65). Yet causal stories are more than simplified claims about past happenings, ‘they are fights about the possibility of control and the assignment of responsibility’ (Stone 1989: 283) and may therefore have an influence on the future. In her reworking of Mead’s perspective on time, Adam (2004a; 2004b; see also Bell and Mau 1971) stresses two points. First, that future possibilities should be seen as conceived and actualised in the present. The future is never determined by the past; it is open for the ‘actually possible’ until it becomes the present. Compared to Mead (1932/1959) who gave privileged status to the present, Adam (2004a) stresses the future, pointing out that images of the future become ‘real’ in the present insofar as they orient human action. Second, the future is relative to the frame of reference employed to understand the past and the present. A frame of fatalism tends to close the future whereas a frame of choice tends to keep it open (Bell and Blau 1971; Adam 2004a). Adam (2004a: 306) distinguishes between the ‘push from the past’ and the ‘pull from the future’. The former focuses on explaining the present, and by extension the future, on the basis of the past; ‘from a known past we can project the future as trend and probability’ (Adam 2004b: 10). The latter is more preoccupied with human purpose and intentionally seeing the future, and not primarily the past, as the point of leverage for human action (Adam 2004b).

Table1 summarises the four causal narratives in our interviews along two dimensions, one dimension describing the past in terms of social level or individual level explanations, the other dimension describing the future in terms of the degree of fatalism vs. agency and voluntarism expressed in the interviews (or in Adam’s terms ‘a push from the past’ vs. ‘a pull from the future’).

Table 1

Summary of drug use narratives

		Future
		Fatalism +	Voluntarism +
Past	Social level of explanation	Like parents, like children	Hanging out with the wrong people
	Individual level of explanation	Cannabis use as self-medication	Cannabis use as lifestyle

Accounts drawing on the family background explanation contained most fatalist ‘push from the past’ elements, envisaging the interviewees’ present (and more important: future) drug problems as determined by their childhood experiences. Participants subscribing to the self-medication explanation also referred to unwilled forces (mental health problems) that made their cannabis smoking ‘necessary’. However, they were somewhat more optimistic when it comes to managing their drug problems than the interviewees in the family background group because they (or at least some of them) placed their trust in legal drugs (ADHD medication, tranquilisers) that could ‘take the place’ of cannabis. Overall, users drawing on the social network explanation or the lifestyle explanation were more agency-oriented with regard to stopping or reducing their cannabis use. Interviewees subscribing to the social network explanation were generally confident that they would be able to manage their cannabis use again, but they also stressed (to differing degrees) that this demanded changes in their relationships with peers. Participants drawing on the lifestyle explanation clearly regarded cannabis use as depending on their own (reasonable and sometimes unreasonable) choices and described themselves as able to cease their drug use when they were ready for it. The challenge for this group (seen from the point of view of treatment) was rather that many interviewees said they had not yet ‘decided’ to stop or reduce, although they experienced severe consequences of their cannabis smoking (related to health, social relations, school and/or work).

Of the four narratives the lifestyle explanation most clearly differed from the others (cf. Davison et al. 1992 on the distinction between lifestyle explanations and explanations that are perceived as being beyond the individual’s control) and to a certain extent even challenged them, because it positioned them as ‘excuses’ that free the drug users from responsibility. For instance, the lifestyle narrative turned the social network explanation on its head by stating that one does not use drugs because of the influence of friends; on the contrary, one chooses specific friends because one wants to use drugs. In a similar vein, the self-medication narrative was challenged by the lifestyle perspective. People do not use drugs because they have emotional problems; they choose to use drugs, and then the drugs (if used in inappropriate ways) may lead to emotional problems. And finally, the stories of unhappy childhoods were rejected by proponents of the lifestyle perspective. In their view, many people experience negative things when growing up – ‘we all have our reasons’ as Freja, who had lost her mother as a teenager, put it. Having experienced bad things in the past, however, does not mean that this should be used as an explanation for one’s drug problems. It is ‘your own choice’ if you start to smoke cannabis, and referring to external reasons is the same as denying responsibility, the proponents of the lifestyle perspective argued.

When people narrate their lives, they do so within socioculturally recognisable frameworks and with the help of well-known vocabularies (Gergen 1992; Hydén 1997; Riessman 2003). Much of the reasoning in the interviews reflects common sense conceptions of why and how young people develop drug problems, and these conceptions are also verified in research (Plumridge and Chetwynd 1999). Hence, epidemiological and psychological studies have emphasised the impact of drug users’ social background and upbringing (e.g. Merikangas et al. 1998; Biederman et al. 2000), personality traits (e.g. Wadsworth et al. 2004; Elkins et al. 2007) and peer influence (e.g. Terry et al. 2007; Fletcher and Bonell 2013) on the initiation and continuation of illicit drug use. Among the few studies actually looking at users’ own explanations for their cannabis use is a Norwegian study, which found that most cannabis users associated their use of the drug with cultural difference and opposition (Sandberg 2013), that is, explanations parallel to the lifestyle narrative in our analysis. The same empirical study also found evidence of the self-medication explanation among interviewees diagnosed with or experiencing ADHD symptoms (see also Pedersen and Sandberg 2013). Our research contributes to this sparse tradition of studying drug users’ own narratives about their cannabis careers, first by depicting more varying explanations than for example the Norwegian study; second, by showing that explanations of drug use are to a certain degree contingent (in most of our interviews, other explanations would have been possible); and finally, by tying the explanations used by the interviewees to their expectations for the future.

Participants’ expectations regarding the future were also related to their drug use patterns at the time of the interview, but in complex ways. For instance, some of the interviewees who said they could easily stop described an intense current use, sometimes with severe consequences for their quality of life, which made it difficult for them to explain why they continued. Also, some of the interviewees who said they were unable to stop using cannabis reported that they had not smoked it for months. There was a pattern though, tying explanations of drug use development to participants’ present smoking practice. By and large, current cannabis use was more common among interviewees drawing on the self-medication and lifestyle explanations than in the two other narrative categories. While for the former category this was because of a perceived medical ‘need’ for the drug, for the latter it was presented as a personal preference for continued smoking.

The interviewees’ explanations were not produced in a social vacuum; rather they should be seen in relation to specific contexts. First, the interview setting in itself may have influenced the drug use aetiologies depicted in the accounts. An interview is, as we stated earlier, a specific type of interaction (an ‘elicited context’, cf. Hydén 1997: 62) where the interviewee is supposed to produce a coherent narrative. The interactional dynamic of an interview may contribute to the development of ‘dominant’ explanations. Interviewers typically respond to the statements of interviewees by accepting them, showing empathy and understanding, asking the interviewees to elaborate on certain themes (and not others), etc. – and all this may be conductive to the creation of specific causal patterns in the interviewees’ accounts. In this sense, the interviews in themselves may have contributed to the shaping, and sharpening, of certain explanations at the expense of others, explanations that we later identified as ‘dominant’.

Second, the interviewees’ explanations may be related to the treatment contexts in which the interviewees were enrolled. As mentioned in the methods section, participants were recruited from three small treatment centres (Centre A: 14 interviewees; Centre B: 12 interviewees and Centre C: four interviewees). These three centres have a lot in common: they are out-patient units working with a combination of therapeutic approaches (systems therapy, cognitive therapy and narrative therapy) and they have young people under 25 as their primary target group. But there are also differences. Centre A is more oriented towards psychiatry and more focused on using tests (ADHD-tests, intelligence tests, addiction severity tests) than the two other centres. Centre B is more narratively oriented, and at least some of the staff members are sceptical towards tests. This centre is explicitly concerned with not defining their clients as drug abusers and with reducing the stigma often associated with illicit drug use. Cannabis smoking in itself is not necessarily a problem, according to this centre, and ‘controlled’ use is just as good a treatment goal as abstinence. Centre B is also preoccupied with ‘unravelling’ the clients’ family backgrounds. Drug use is often seen as an indication that something is wrong in the families, not with the young drug user.

Significantly, six out of seven participants who explained their cannabis use with reference to ADHD and other mental health problems were enrolled at Centre A, where clients are systematically tested for different psychiatric conditions. Participants using the problematic childhood explanation or the lifestyle explanation, on the other hand, were more often recruited from Centre B. Five out of six interviewees relating their drug problems to their family background and three out of five interviewees using the lifestyle explanation came from this centre. The social network explanation, finally, held equal shares of participants from all three centres. When pointing out that there is an association between explanations and institutional contexts, we do not claim that young people’s self-conceptions are determined by treatment centres, and that our interviewees would have told completely different stories had they been involved with another centre. We do claim, however, that their narratives are interactional achievements, reflecting complex processes of identity construction, where therapeutic relations influence drug users’ self-understandings, just like their self-understanding influence what kind of treatment they seek.

Related to this, one may ask if the association between explanations and future expectations is a consequence of differing degrees of problem severity among the interviewees. It could for instance be that participants from drug addict families and those with mental health problems had more serious drug problems than the other participants, and therefore were more pessimistic about their future prospects. There is no clear support for this interpretation in the interviews, however. We find participants with severe problems (intense current drug use, social isolation, marginalisation in relation to the school system and labour market) in all narrative categories. As previously mentioned, two thirds of our interviewees were living on social welfare, and most of those who worked or studied had a fragile relationship to their workplace/ school. This was also the case for participants using social network and lifestyle explanations. Many of them had experienced severe consequences of their cannabis use: interrupted education, job firings, psychiatric hospitalisations, severed personal relationships, etc. Nevertheless, they – as opposed to the participants referring to traumatic childhoods or emotional problems – felt that they were in charge of their own lives, and that they, if not at the time of the interview then at some time in the future, would be able to control their cannabis use.

Summing up, this article contributes to the narrative tradition in medical sociology by showing how drug use can be inscribed in different causal narratives; by illustrating how these explanations are to some extent contingent; and by highlighting the temporal dimensions – past, present and future – of the narratives, thereby linking the interviewees’ explanations to their expectations for the future. Explanations, here causal narratives about drug use, do not merely point back in time; they also point forwards. Hence, causal narratives are not innocent stories about the past – on the contrary, they attribute people varying levels of optimism and choice when it comes to the future.