‘You can’t stand on a corner and talk about it …’: Medicinal cannabis use,impression management and the analytical status of interviews

Introduction

In this article, I explore examples of impression management in interviews with participants, in which they discuss their use of cannabis as a medicinal substance. The 4 participants (out of a total of 33) discussed in this article have been selected because they articulated an especially pronounced concern for impression management (much more than the other 29 participants). However, the role of myself as the interviewer must also be considered as interviews are social encounters and, therefore, co-constructions. In recognising the presence of impression management and co-construction, issues are posed regarding how unproblematically we can treat the data produced in such interviews when trying to find out about events that have happened outside of the interview. I argue that broader insights into identities, communities, moralities and behaviour can still be gained.

I start by briefly addressing how cannabis is used by some people for medicinal purposes, yet many are uncomfortable with the possibility of being criminalised and the wider stigma associated with ‘drug’ use. I then discuss how such sensitivities can understandably give rise to articulations best understood as instances of impression management (Goffman, 1959), in a social encounter (the interview) in which the participants articulate events and experiences while simultaneously attending to the presentation of the self and making sense of the research process and the interviewer. I argue that these instances of impression management can also tell us much about the participants and their lives outside of this social event.

Background and context

Cannabis tincture was a popular medicinal substance for addressing pain between 1840 and 1900, but its popularity among medical practitioners soon declined after this period (Grinspoon, 1994). Over the span of the 20th century, cannabis increasingly became understood as a recreational ‘drug’, with various connotations of deviance, as opposed to a substance with medical potential. However, since the early 1970s in the United States (Dunn and Davis, 1974) and later in the United Kingdom (Ware et al., 2005) and other countries, a growing number of people have reported using cannabis for a range of medicinal benefits. Typically, they are individuals with a range of chronic illnesses (e.g. multiple sclerosis (MS) or rheumatoid arthritis) or impairments (e.g. spinal cord injury). Subject to the illness or impairment in question, they report a range of benefits from using cannabis. For more on this, see Coomber et al. (2003).

Despite what tends to be labelled as ‘anecdotal’ claims for cannabis’ medical efficacy, as well as clinical evidence (Dansak, 1997; Hollister, 2001; Kickman and King, 2014; Leung, 2011; Musty and Rossi, 2001; Zimmer and Morgan, 1995), cannabis remains a controlled class B substance under the 1971 Misuse of Drugs Act in the United Kingdom. Many medicinal users argue that cannabis is the only substance that provides them with symptomatic relief and a reasonable quality of life (Coomber et al., 2003), but in using it they live with the possibility of criminal prosecution. While some medicinal cannabis users will have used the substance recreationally prior to using it medicinally, many will not and often resent having to engage with the illegal market to obtain it and are uncomfortable with the stigma of using an illegal drug (Coomber et al., 2003). While it has been argued that cannabis has undergone a degree of ‘normalisation’ among a section of typically, though not exclusively, younger people in the United Kingdom (Parker, 2005; Parker et al., 1998), for others the use of any illegal substance, and the potential for social stigma and criminalisation, remains a very sensitive issue.

Talking about drug use: discourse, identity and stigma

The medicinal cannabis users whose interviews I address in this research had not used cannabis prior to acquiring an illness and started to use cannabis in ways that they understood as medicinal. While managing stigma and negotiating perceptions of the self are commonly found in interviews with drug users (Hellum, 2005; Mostaghim and Hathaway, 2013; Omel’Chenko, 2006; Pedersen, 2015; Peretti-Watel, 2003; Radcliffe and Stevens, 2008; Rødner, 2005; Sandberg, 2012), we might expect the accounts of people who understand their use of cannabis to be the outcome of the onset of an illness or impairment, as opposed to what they understand to be of recreational use, to be additionally sensitive.

Research on how drug users attend to symbolic boundaries regarding their identities and those of ‘others’ is mindful of Goffman’s (1968) advice that those who could be seen as ‘discreditable’ people can pass as ‘normal’, provided they can manage information about the source of that ‘shame’. How drug users present themselves in interviews is very much about managing that information and attempting to influence how they are perceived. In a discussion of how heroin users negotiate the ‘junkie’ identity in interviews, Radcliffe and Stevens (2008) note how their participants exclude themselves from this category, but also acknowledge its validity regarding ‘others’. Similarly, Rødner (2005) discussed how those she described as ‘socially integrated drug users’ (p. 333) contrast their drug use with those whom they construct as ‘drug abusers’. Peretti-Watel (2003) built on Sykes and Matza’s (1957) neutralisation theory, which discussed the verbal techniques that juvenile offenders used to justify or excuse their behaviour. Peretti-Watel (2003) found that French cannabis users he interviewed engaged in risk denial by contrasting their cannabis use with ‘hard’ drugs, comparing cannabis to alcohol and emphasising the risks of the latter, and by portraying themselves as able to control their own cannabis use. Mostaghim and Hathaway (2013) note how Canadian undergraduates exhibit a more nuanced and fluid understanding of being a ‘user’ or ‘non-user’ of cannabis, dependent on the context of the social situation. While the medicinal users that I discuss have arguably had cannabis use imposed on them by chronic illness or impairment (although of course they chose to use it), Mostaghim and Hathaway (2013) conclude by arguing for further theoretical engagement with Goffman’s situated self. I aim to address this. Sandberg (2012) identified ‘normalisation’, sub-cultural and risk denial discourses in his interviews with Norwegian cannabis users. He concluded that all of these can be seen as responses to stigmatisation, challenging ideas around the assumed normalisation of cannabis in Western societies minimising the amount of stigma that users might feel they have to manage.

Less research into the management of stigma around identities relating to medicinal cannabis use, than recreational drug use, exists. Pedersen (2015) interviewed Norwegian cannabis users who identified as medicinal users with self-diagnosed attention deficit hyperactivity disorder (ADHD). He discussed the need among his participants to engage in symbolic boundary work, rhetorically, as all of them had previously been recreational cannabis users. The symbolic boundary work attempted to establish and maintain ‘medical user’ identities in opposition to ‘recreational use’ of others and their own ‘recreational use’ in the past. However, while all of Pedersen’s participants had once been recreational users, none of the participants that I will be discussing had used cannabis before the onset of illness.

Hammersley et al. (2001) provide a framework that allows consideration of the role of the interviewer in the co-construction of discourse (something that interview-based research with drug users tends not to take account of). They present three aspects of the process by way of which cannabis users negotiate identities in social encounters. Signification involves considering the meanings of the drug in question, as understood by the social actors involved (here, the interviewer and the participant). I would suggest that this also involves how each actor imagines that the other understands the drug in question and, perhaps, the universe of meanings around ‘drugs’ more broadly (which tend to be negative). They discuss negotiation in relation to how the cannabis users negotiate their self-presentation, subject to the audience and the context of the social situation. For me, in relation to interviews, both participant and interviewer are negotiating identities, although the participant usually has much more work to do to negotiate stigma. Finally, the authors discussed categorisation. Categorisation is what is being negotiated, how one actor will be understood by another, in terms of extant categories, for example, Rødner’s (2005) ‘drug user’ or ‘drug abuser’. Subject to the social context, negotiation and categorisation could have serious consequences for the categorised, depending on the categoriser and his or her social position (e.g. quite serious if they are a Judge). In overall terms, the possible consequences of negative categorisation by an interviewer are less serious, but the participants I discuss here took their categorisation very seriously, treating them as moral charges even when I had not actually made such accusations.

Method and methodology

The research I report on here revisits data first reported in Coomber et al. (2003), obtained via interviews conducted by myself. Thirty-three chronically ill and/or impaired medicinal cannabis users were recruited, via adverts in the newsletters of London-based disabled peoples’ organisations, in Disability Now and then further via a snowball sample (Becker, 1963), using the initial participants to provide further participants for the research. Ethical approval was obtained from the University of Greenwich Research Ethics Committee. Semi-structured interviews were conducted with participants, in their homes (for reasons of participant mobility/comfort and to hopefully make them more relaxed). Participants were asked questions about how and why they came to use cannabis medicinally, their perceptions of the benefits and difficulties associated with this and the effects that this practice might have on negotiated health care. Interviews lasted between 90 and 180 minutes and were recorded, with informed consent having been acquired. After considering different approaches to transcription, a very simple approach was decided upon. This maximised readability while sacrificing interpretations of intonation, rising/falling pitch, timings of pauses and vocal emphasis. A critical discourse analysis approach to the data was employed. As has been noted already, I draw here on data from just four of the participants, due to their articulations being most relevant to exploring aspects of impression management.

In terms of demographic characteristics, the four participants (anonymous names given by myself) are introduced as follows:

Jennie, aged 27, had arthritis and was close to finishing an MA. Jennie described herself as White British and was identified as middle-middle class (higher in educational capital than the other three). She came from the East Midlands.

Mark, aged 46, had multiple sclerosis and had previously done office work. Mark described himself as White British and was identified as lower-middle class. He came from the Home Counties.

Vanessa, aged 58, had multiple sclerosis and had previously worked as a bank clerk. Vanessa described herself as White British and was identified as middle-middle class (higher in economic capital than Jenny and Mark). She came from the Home Counties and knew Deborah.

Deborah, aged 56, had multiple sclerosis and had previously worked as a bank clerk. Deborah described herself as White British and was identified as middle-middle class (higher in economic capital than Jenny and Mark). She came from the Home Counties and knew Vanessa.

As mentioned above, it is significant that none of the four participants had used cannabis prior to medicinal use (and tellingly, they all asserted this strongly in their interviews). In terms of the sample from which this article draws, this puts them in a minority. Of the 33 participants, only 12 had no experience of cannabis prior to medicinal use (i.e. use not starting until the onset of illness or having acquired an impairment). Three of the four participants discussed in this article were members of the Multiple Sclerosis Society (the fourth, Jennie, did not have MS), which had adopted a policy similar to that of the British Medical Association (BMA) (1997) in respect that it was in favour of research aimed at developing cannabis-based medicines but did not condone the use of cannabis itself among its members, although many members used it anyway (Coomber et al., 2003,).

In terms of methodological issues, my interviews can, in places, be understood as addressing sensitive issues. Drury et al. (2007) note that any interview can be sensitive because self-disclosure can lead to emotional response. However, some topics have the potential to be much more sensitive, in terms of distress elicited, than others (Dempsey et al., 2016). Lee (1993) discusses a number of criteria that may constitute sensitive topics and he includes research in which participants are asked to make revelations about themselves, or about others, which may have consequences or may make them feel uncomfortable. Making revelations to a stranger about illegal drug use, which are being recorded, would seem to fit these criteria. My participants were particularly uncomfortable, in places, with how the interview discussions could lead them to be perceived. Had I anticipated that questions about the ideas that participants had about cannabis, prior to their first use of it, or where they got the idea that cannabis could be used medicinally, would result in displays of impression management (Goffman, 1959), I may have taken Lee’s (1993) advice and asked them later in the interview. In the interviews with most participants, these questions were not seen as sensitive. As Corbin and Morse (2003) note, though, it is not always possible to predict areas of sensitivity at the outset of research. Nor, I would add, is it always possible to predict issues that will be sensitive to some participants, but not to others.

Research on how drug users construct their identity in interviews needs to consider the place of the interviewer, in these co-constructions. Drawing on the work of Hammersley et al. (2001), the interviewer’s signification of cannabis is relevant, at least in as much as the participant will try to discern it, in the process of negotiating the categorisation that they can be seen as aiming for and which the interviewer can be seen as trying to arrive at, as research tries to understand issues around their drug use. Understanding impression management involves considering how the participant might have made sense of the interviewer, the research project and the interview itself. As the interviewer, I presented myself as friendly and polite, grateful for the time and experiences of the participants. In terms of characteristics, I am a White, middle-class male (my social class having changed as an outcome of my trajectory into higher education). I spoke with less of a South London accent than I ‘naturally’ had since starting lecturing and research. Was this ‘being’ middle class, on my part, somehow related to the impression management of these four participants? Perhaps (Letherby, 2014). The research project itself was described, by myself, as to gather data for academic research on the subject, for my PhD work and, hopefully, to contribute to the broader issue. I described the interview as asking questions about them, about their cannabis use and trying to understand the benefits and problems associated with their use of cannabis, as they saw it. I emphasised confidentiality and anonymity and how they should say whatever they liked, but hopefully be full and frank in their answers. All of this, of course, is my interpretation of what I said and, I acknowledge, my impression management in writing it. Nevertheless, you now have some sense of me, as a co-constructing actor in the interviews, just as you have some sense of the participants.

Excerpts of impression management in interviews

I now turn to the data and draw on extracts from interviews with Jennie, Mark, Vanessa and Deborah. It is significant that the displays of impression management (Goffman, 1959) began early in the interviews, when they were asked what their thoughts were about cannabis prior to using it, how they came to use it and whether they required the sensation of a ‘high’ to experience symptomatic relief. For Goffman (1959), this is to be expected. The ‘initial definition of the situation projected by an individual tends to provide a plan for the cooperative activity that follows…’ (p. 24). From early on, these participants were working to establish a moral standing. With medicinal cannabis use being a contested and sensitive issue (Lee, 1993), one might expect to see aspects of impression management (Goffman, 1959) arising, as the participants attempted to negotiate what they saw as a favourable self-presentation. The articulations and narratives that are produced at this point of the interviews by these four participants can be understood as co-constructed, partially shaped by precisely how I phrased the questions, and negotiation subject to how they perceived my questions as potentially categorising them. Often, this is done with considerable rhetorical force, by the participants, which I take as indicative of the sensitivity with which they perceived potential stigmatisation by way of categories that they rejected as not being representative of them or of their cannabis use. All four did considerable discursive work to impose the categorisation of ‘respectable medicinal cannabis user’ (someone who uses cannabis for symptomatic purposes only). In some excerpts, this also articulated constructions of social class/community respectability. The ‘other’, that they partially articulated, differed slightly, participant-by-participant, was the ‘recreational cannabis user’ (someone who uses cannabis for pleasure). In some of the excerpts to follow, connotations of social class and use of other drugs are interwoven into that ‘other’. So not only are the participants’ identities constructed dynamically, so is the ‘other’ they contrasted themselves with. Interestingly, the participants could categorise me, the interviewer, should they have wanted to and could also conjure hypothetical ‘others’, as well as actual people, for rhetorical purposes. As Collins (1998) noted, when citing Simpsons’ discussion of Bakhtin, interviewees may be engaged in ‘… multiple dialogues with themselves, the interviewer and others imaginably present …’ (Simpsons, 1997: 4.3, cited in Collins, 1998). Those others can be actual people or amalgams, and I argue that what appear to be among the most interview-specific and particular articulations also tell us much about identities, communities, moralities and behaviour – the social world of the individual, outside of the interview:

In terms of Gee’s (1999) suggestion that discourse be considered in terms of ‘whos-doing-whats’, Jennie’s narrative was occasioned by my question, which notably made reference to ‘recreational use’. This was the point of sensitivity, and her response to this can be seen as a charge-rebuttal (Silverman, 1994), where Jennie sought to distance herself from possibly being perceived as a recreational user, a categorisation that she understands as being undesirable, even though I didn’t actually try to apply it to her. Such were the sensitivities in these interviews. Jennie articulated a self-presentation of someone who was law abiding, respectable and a member of a respectable community. This served rhetorical purposes of re-negotiating what she saw as a positive self-presentation, or categorisation in Hammersley et al.’s (2001) terms, but also gives us insights about identity, community, morality and behaviour that go beyond the immediate context of the interview.

Jennie constructed a narrative in which she presented herself as someone who would not normally have used cannabis, emphasised by her shock at the suggestion that she should. In effect, she was given permission to use cannabis by a person of authority, in this narrative. The unnamed authoritative figure that Jennie articulates in the interview is only able, in Jennie’s understanding, to give this permission because outside of the interview this person held authority within that community. This tells us that, in such communities (a rural one in this case), in Jennie’s understanding through her membership of it, this is a way in which authority, power and influence sometimes work. This person is a moral arbiter, understood as able to give permission to people to transgress established morality. This person, who I speculate was a general practitioner (GP), was constructed by Jennie as being a ‘very, very well spoken, very well respected person of the community’. Rhetorically, the unnamed individual’s claim to authority is emphasised by Jennie’s use of what Pomerantz (1986, cited in Wooffitt, 1993) calls extreme case formulation. They are ‘very, very well spoken …’, so respectability in a social-class-related way is alluded to. Furthermore, this person also functioned in Jennie’s narrative as a categoriser. In her narrative, they permit her ‘respectable medicinal use’ and, in doing so, categorise Jennie as such a user. The reader will note that, in a fairly short extract, Jennie ‘recounts’ what she said to the authoritative third party, brought momentarily into our interview, on no less than three occasions. In doing so, she was demonstrating her initially strong opposition to the idea of using cannabis, as she understands that a respectable person would be thus inclined. This narrative has the effect of negotiating the problem of accounting for why Jennie came to use cannabis when it was something she was initially very much opposed to while also maintaining a sense of still being opposed to the unnecessary use of cannabis. The narrative does what it needs to do, in a cohesive way, based on Jennie’s understandings of the issues, of the interview, of the respectable figure from her community and of her community’s moral expectations.

Indeed, the fact that all four participants exhibit impression management (Goffman, 1959) around this part of the interview is telling in itself because it hints at strongly habituated (Bourdieu, 1979) inclinations against cannabis and drugs, perceived as aspects of lifestyles that they did not approve of and identities that they did not want applied to them. To see justificatory articulations drawing on quite traditional notions of community and of position within community is insightful. Jennie had described growing up in a relatively small, rural town, very different from the metropolitan environments more commonly associated with the ‘normalisation’ of illicit drug use (Parker et al., 1998) and its attendant cultural values. For many people from such conservative social and cultural backgrounds, the thought of being associated with the negative connotations of ‘drugs’ would provoke a charge-rebuttal (Silverman, 1994), as was seen in Jennie’s narrative. The majority of participants’ interviews in the study as a whole did not feature strong impression management negotiation narratives at this point because for them there was nothing to negotiate (i.e. they did not have strong feelings about cannabis in the first place). In fact, most participants reported having a relatively open-minded attitude to cannabis prior to using it (if they had not already used it ‘recreationally’ prior to medicinal use anyway).

Jennie responded similarly, a little later in the interview, when I asked her whether it was necessary to experience a ‘high’ in order to obtain the benefits that she used cannabis for?

I don’t want that, I don’t need it, I don’t like it …

Jennie interpreted my question as an accusation of enjoying her use of cannabis, as opposed to it being something she, reluctantly, had to use, and as an attempt to categorise her as a ‘recreational user’. How she reacted indicates that she perceived this as an even more serious moral charge than in the previous excerpt. Rhetorically, what Jefferson (1991, cited in Wooffitt, 1993) described as three-partedness was used to strongly emphasise and resist any possible suggestion that a ‘high’ was an effect that she might have wanted. Many medicinal cannabis users, in my experience, are aware that while many among the general public are to some degree sympathetic to them, due to their illness/impairment, there does exist a degree of scepticism that some medicinal users may also ‘enjoy their medicine’. Parsons’ (1951) discussion of the sick role tells us that ill people are excused from certain social responsibilities, but are not supposed to enjoy this. However, the impression management exhibited by Jennie was relatively moderate compared to that of the other participants discussed here.

Mark exhibited similar discursive tendencies and values, early in his interview:

My occasioning of Mark’s narrative here does not contain an obvious moral charge, but early in the interviews (when all the excerpts I discuss in this article occurred) identities were far less established (mine and his), allowing more room for uncertainty and anxiety around negotiation and categorisation. In terms of thinking about what his narrative seeks to accomplish, Mark was trying to negotiate what he understood as a positive self-presentation, that of the ‘respectable medicinal user’. Again, as we saw with Jennie, we can see Mark employing extreme case formulations (Pomerantz, 1986, cited in Wooffitt, 1993): ‘no way I would have actually thought of using cannabis’ and ‘cannabis was only ever used as a recreational drug that never had any therapeutic values whatsoever’. The presence of an authoritative individual is this time accredited with being the source of the idea that cannabis could have medicinal uses and also functioned to justify Mark’s decision to use cannabis in this way and, as with Jennie, functioned to categorise his use and, therefore, Mark himself as a ‘respectable medicinal cannabis user’. Mark’s narrative is also interesting, as other moral issues are articulated. He constructed his previous thoughts about cannabis as being ‘recreational’ and then constructed this type of use as for people who wanted to get ‘happy and higher’ to ‘improve their sex lives’. The ‘recreational cannabis user’ ‘other’ is articulated by Mark as pleasure-seeking and not just regarding the drugs’ effects, but regarding sexual pleasure too. This contrasted significantly with the ‘respectable cannabis user’ who wouldn’t use it to seek chemical pleasures, let alone pleasures of other types.

Other than the rhetorical purpose of self-presentation within the interview to categorise himself using an identity that he understood as a positive one, Mark’s narrative also spoke of his views about ‘others’ in society, who use drugs like cannabis in very different ways. It provides insights about morality, identity, community and behaviour by way of his choices of symbolic resources when constructing an ‘upstanding’ in-group and a ‘transgressing’ out-group.

Deborah was one of two participants who most strongly exemplified a tendency towards self-presentational concerns in the interviews. Her narratives contained articulations that drew more on social class–related connotations, in negotiating her categorisation as ‘respectable cannabis user’. The excerpts that follow are taken from five pages of the interview transcript in which Deborah spoke about how she came to use cannabis. As the whole interview was only 35 pages long, this was quite a short interview with a relatively large part of her discourse featuring displays of impression management (Goffman, 1959). Indeed, the whole five pages could be seen as one long narrative that exhibited impression management, something that Deborah also did elsewhere in her interview:

This excerpt was occasioned by me trying to ascertain Deborah’s thoughts about cannabis before she came to use it. As I have said before, this was not a sensitive question for most participants. As with Jennie and Mark, we can see Deborah doing rhetorical work to negotiate what she saw as a positive self-presentation. Extreme case formulations (Pomerantz, 1986, cited in Wooffitt, 1993) are combined with three-partedness (Jefferson, 1991, cited in Wooffitt, 1993) in articulating her early thoughts that cannabis was ‘naughty, wrong and awful’. There is also some confusion in her account when she claims that her doctor had offered her cannabis. I can only speculate, but perhaps her doctor had, informally, suggested that she try it, as he couldn’t have actually prescribed cannabis. Either way, Deborah’s account of her response was emphatic, with six ‘no’s’. In my follow-up question, she clearly imposed the categorisation of ‘respectable medicinal user’ by distancing her use from using cannabis as a ‘leisure thing’. When I had one last try at this question, as all I was getting were negotiations, I was interrupted, shut down and thought it best to move on. For Deborah, the possibility that I might entertain the thought of her being sympathetic to recreational use was pushing my luck. I continued:

Deborah was still distancing herself, in this excerpt, despite already having recounted how her GP had suggested using cannabis, and now a friend, at a wedding. Interestingly, when she constructed the recounting of how she came to use cannabis here, the wedding conversation did lead to her use: ‘That’s how it started’. However, there is a clear articulation of her agency here, when she said, ‘maybe I would’. Unlike Jennie and Mark’s accounts, there is no construction of a permission giver here. The friend could have been constructed as a permission giver, Deborah being ‘respectable’ so the friend being that too, by association. The friend’s son was a medicinal cannabis user, but Deborah only articulated him when I asked. Deborah didn’t need a permission giver because she had already categorised herself as a ‘respectable medicinal cannabis user’, in the previously discussed excerpt, by interrupting me and closing the negotiation of identity down.

Finally, Deborah described a conversation with her doctor after she had smoked cannabis and felt light-headed and likely to fall over. She articulated her doctor as saying, ‘please don’t do that, it won’t look good in the local paper’. This is insightful regarding community life in respectable, well-to-do, middle-class circles of communities small enough for people to recognise you in the local media. It speaks of avoiding the visible transgression of established moral codes and connotations around ‘drugs’. It is insightful that the doctor’s remark is articulated as having more concern for being socially upstanding than the health risks of falling down. Maybe that is how he said and meant it, maybe it isn’t, but the local newspaper represents the risk of visibility within the local community and the potential for stigmatisation and shame.

The final participant, Vanessa, also exhibited similar discursive tendencies, again starting at the same point as the other participants. As with Deborah, there was an interruption, but it is even sooner. Also, similarly to Deborah, Vanessa was quite rhetorically forceful in relation to her categorisation. Interestingly, the two participants were acquaintances (one having referred me to the other), and it struck me, at the time, that they were both from middle-class fractions that were used to having more power in social situations (as indicated by their manner and also wealth). This is important in relation to understanding how interviews are co-constructions. Some participants will be more used to, more able and more willing to impose their will than others. I phrased this question in a particular way, and this occasioned a particular response:

Strength of feeling was evident from the start, with the interruption. The part about her son going to parties articulates her as a good mother, an additional aspect to the positive self-presentation. Interestingly, an ‘other’ is articulated here, but this ‘other’ is the ecstasy user and an inhabitant of ‘some of the classiest houses’. It is as if she was saying ‘surely such things ought not to be going on there’, which involved a momentary articulation about who would not and, thereby, who would be expected to consume such drugs. This is interesting, because an in-group/out-group contrasting rhetorical strategy, so commonly used in everyday discourse (McKinlay et al., 1993; Potter, 1997; Wooffitt, 1993), cannot as easily be articulated when the transgressor is part of the articulator’s in-group. Vanessa finished with ‘as for myself being part of it or access to it, no’. This was articulated as something that was simply not of her social milieu. Rhetorically, this allowed her to categorise herself as a ‘respectable medicinal cannabis user’, and it is the closing utterance. I followed up, trying to clarify that I wasn’t trying to classify Vanessa in that way and I was, again, closed down.

Vanessa later discussed the local MS Society, in which she was a member at the time of the interview. The remark at the end is perhaps the most revealing of all and is insightful with regard to why these four participants exhibited discourses that lacked a critical/politicised engagement with the issue (found in the discourses of eight other participants) of medicinal cannabis use, but were high in impression management:

My interjection, ‘No, it’s not a social thing’, seemed to neutralise what I had quickly realised Vanessa could perceive as a moral charge of being a ‘recreational user’. Having negotiated this, together, she continued with a narrative that first led me to become interested in how certain users seemed to spend much of their interview negotiating how they were perceived, with the image of numerous ‘respectable medicinal cannabis users’ being aware of each other’s use, but no one actually talking about it. I am glad that I interjected as I did, or I may have been deprived of a major insight. It is the last sentence that is most interesting – ‘You can’t stand on a corner and talk about it because it’s just not on, is it?’ This utterance suggests a middle-class habitus greatly concerned with social standing and highly averse to the notion of entering into public discourse on the matter (which some other participants among the sample of 33 actively pursued through the media). This is a habitual type of engagement with medicinal cannabis use – as something to do, but not to talk about, because ‘… it’s just not on, is it?’ It is fascinating to see how the internalisation of the ‘rules of respectability’ actually impaired the possibility of collective action, for such participants. Advantageously positioned middle-class individuals were more likely to have the opportunity to advance the issue of making cannabis legally available to medicinal users, as their voices were more likely to be heard in the social conversation about this issue. However, they censored themselves, denying themselves a place in that debate out of concern for respectability. Again, insights into identities, moralities, communities and behaviour are gained, this time regarding an interest in the political potential of medicinal cannabis users and their discourses.

Discussion

The excerpts from interviews with the four medicinal cannabis users examined in this article are examples of how impression management (Goffman, 1959), as a reaction to sensitive questions/topics in interviews often seemingly perceived as moral charges by the participant, featured articulations of narratives and strong rhetorical strategies. The immediate purpose that this served for the participants was to attempt to impose what they saw as a positive self-presentation, which I understand as that of ‘respectable medicinal cannabis user’. This categorisation is most meaningful in contrast to that of the ‘recreational cannabis user’, the construction of the former requiring articulations of the latter.

The discursive work that the participants were engaged in needs to be seen as operating on a number of levels. They negotiated a positive self-presentation, but these rhetorical strategies, narratives, negotiations and categorisations also tell us a lot about the participants’ lives outside of the interviews too, in which this negotiation of categorisation was an ongoing part of their lives. We have seen insights into the lives of people who had not used cannabis before having to use it, due to their lives taking difficult and unforeseen turns. To do this was to redirect the trajectory of their lives in ways that were uncomfortable to traverse and uncomfortable to discuss with me. In traversing this redirection in their lives and in negotiating categorisation in various situations, a dynamic process is at work, that of the fluid yet durable, that of the adaptive internalisation and remaking – that of the habitus (Bourdieu, 1979). We have seen glimpses of the habitus, of the habitual inclinations of four participants who were uncomfortable with using a substance that they know tends to be associated with a very different ‘type’ of substance user to how they see themselves. They sought to distance and differentiate themselves from this ‘other’ at every turn, to emphasise that such things as obtaining pleasure from ‘drug’ use or, in turn, other pleasures in Mark’s case were alien to their milieu. We have seen narratives recounted as to how they negotiated their identities in relation to cannabis in everyday life, when talking to doctors or when being at social gatherings. We have had insights into the idea that identities had to be negotiated regarding other roles, like being a mother, in the case of Deborah, or when engaging with other medicinal cannabis users, in the case of Vanessa. We saw narratives about people in positions of authority. Jennie cited one who, in her excerpt was constructed as a permission giver, was pivotal in the construction of her transitional experience, from non-user to user. I argued that, especially in small communities, such people can act as moral arbiters. I believe that not just in her narrative, but in her understanding of her own transition, this person allowed Jennie to come to use a substance that she believed helped her greatly, even if she remained uncomfortable with the potential for stigmatisation, because she clung to a morality that would allow her to feel stigmatisation. We also saw how some were more confident in their own categorisation, not feeling the need for a permission giver. Yet, ultimately, these four participants tended to keep their medicinal cannabis use to themselves (bar talking to me about it) and excluded themselves from the public debate on the issue, despite their middle-class social positioning, meaning that they would likely have been seen positively by many and be more likely to have their voices heard than some.

This negotiation of what you can be/ought to do and what you can’t be/should not do provides fascinating insights into identity, communities, morality and behaviour. What looks like the most morally charged and occasioned, therefore particular to the interview, type constructions often provides the most interesting insights beyond the interview. I now briefly revisit the debate about the analytical status of the interview and then present one argument as to why I think very morally charged and occasioned instances, like impression management, can still tell us much about the identities, communities, moralities and behaviour beyond the interview.

How should we regard the interview?

The idea that interviews need to be understood as social encounters is certainly not new (see Collins, 1998; Mishler, 1979). However, my discussion of the issue, and the suggestion I am going to make about impression management, may be of interest to others.

The interview, in one form or another, has been a central method in social research for a considerable period of time (Neuman, 2000) and is widely employed within sociological research (Bryman, 2004), as well as elsewhere. Mishler (1979) noted the paradox in which

As theorists and researchers, we tend to behave as if context were the enemy of understanding rather than the resource for understanding which it is in our everyday lives. (p. 2; my italics)

In the history of the interview, we have seen quantitative traditions that emphasised standardisation, neutrality objectivity and reduced the interview to the level of stimulus and response and that of a behavioural event. We have also seen the qualitative alternative that arose and followed the belief that certain types of interviews were more likely to provide subjectively authentic accounts than others (Seale, 1998). While there are strong ‘political’ arguments for adopting such approaches, which are importantly reflexive about the role of social sciences, the commitment to such realist epistemologies promotes what Seale (1998) calls a ‘somewhat romantic’ belief that certain approaches to interviewing can guarantee data that contain ‘how it really is’ (p. 209). Objective and distanced or subjective and involved, there exists the belief that an authentic account can be produced by following the right steps.

This position contrasts with one in which people’s talk is involved in presenting and preserving certain views of themselves, others and the social world in which they are involved. Goffman (1959) suggested that one feature of behaviour within interactions is a concern to protect the positive moral standing of the self and of others. This is one of the tasks that may be located within the performative conceptualisation of language, in which people use language as a form of social action. Within this view, interview responses are no longer simply true or false reports on reality, but are also, or only (depending on the position that a researcher adopts within this debate), articulations of the interviewee’s ‘reality’.

This view of language and what happens in interviews has a number of corresponding positions with regard to the analytic status of interview data. One position is that interviews are not only displays of subjective realities and ‘moral arenas’ but also contain references to a reality outside of them, a position in which realism has not been abandoned, at least not totally. As an example of this position, Glassner and Loughlin (n.d., cited in Silverman, 1994) took the view that interview responses could be treated as culturally defined narratives and also possibly as reports on reality. The authors discuss an example in which, when a participant says that she uses cannabis because her friends do, they take this to suggest two things: first that she has employed a culturally prevalent way of understanding and talking about this topic (narrative) and, second, that this is evidence that cannabis use is part of peer gatherings. In this example, the authors can be seen as having taken Garfinkel’s (1967) advice, that accounts are a part of the world that they describe. It is not necessary to merely accept what is being said, and the way this ‘revelation’ is being constructed, in an unquestioning and literal fashion. While the specificity of the meanings being produced and the aspects of ‘impression management’ (Goffman, 1959) are on one level contextual to the interview, it is essential to realise that broader social contexts also apply to what is being talked about within the interview. McKinlay et al. (1993) argue that a continuum runs from the local and immediate discursive context within which talk occurs, to the wider purposes that discourse may serve.

For me, the bridge between these concerns and insights into the world beyond the interview lies in using Goffman’s (1959) work on impression management in conjunction with Bourdieu’s (1979) work on habitus. When individuals are asked to talk about their involvement in an issue that they perceive as sensitive, we might expect to see instances of ‘impression management’. This impression management is one aspect of a co-construction between the participant and the researcher (or actor/audience and audience/actor in Goffman’s terms, depending on who is talking and who is listening). Goffman’s (1959) notion of impression management involves individuals attempting to portray themselves in idealised ways, and those portrayals will depend, to some extent, on how the participant understands the social situation, including the researcher. Goffman’s dramaturgical approach has at its heart the analogy of individuals as actors, giving performances, playing roles and appearing keen to protect the positive moral standing of the self. Goffman (1959) discussed various strategies used by individuals to effectively and convincingly perform their roles, and such strategies are significant to the research reported here, in relation to how individuals negotiated the perception of themselves.

In a discussion of Goffman and impression management, Ritzer (2003) argued that Goffman adopts quite a cynical view of the individual, that is to say that he conceptualised the individual as playing roles that may often consciously involve misleading those who are understood to be the audience. As Mostaghim and Hathaway (2013) point out, Goffman’s actor in the later edition of The Presentation of Self in Everyday Life is less cynical. However, this raises an interesting question about the degree to which such performances are consciously produced. In relation to this question, Goffman’s (1959) concept of impression management might benefit from insights derived from Bourdieu’s (1979, 1992) concept of habitus and later the linguistic habitus. If impression management involves the tendency among people to portray themselves in a favourable light, then this can be seen as being done in habitualised ways, the product of a socially constructed habitus negotiating its own portrayal. The way in which an individual may want to be seen is, itself, socially constituted, tied to the habitual ways in which they reflect upon themselves. In turn, the habitus is malleable, fluid, always in flux, shaping and shaped by the trajectory (Bourdieu, 1992) of the individual’s life and the fields they pass through, the amounts and types of capital they acquire, the modifying of tendencies and inclinations this produces and the complex interactions as the person remakes themselves and is remade. In his concluding remarks, Peretti-Watel (2003) argued that future research should address how what he conceptualises as ‘learnt cognitive skills’ (p. 39) are acquired. I would argue that a Bourdieusian framework using the concept of trajectory allows us to see such ‘skills’ as cultural capital and ask questions about how and where they were acquired.

Instances of impression management speak of the social world of the individual, as the participant presents themselves, subject to their understanding of the ‘interview’, of the interviewer and of how revelations might lead to them being perceived. Reconceptualised, I believe, they are an aid to insight. They can tell us much about the participant and their life outside of the interview, as narratives and rhetorical strategies are partly habitualised responses. They are situationally particular performances, but ultimately these improvisations are rooted in the social worlds they describe (Garfinkel, 1967) and the habitus that inhabits and acts within them.

Conclusion

Narratives around ‘drugs’ and ‘drug’ use are, above all, moralistic in character. Interviews about substance use need to be understood as discursive events in which, subject to the habitualised inclinations of the participants in question, impression management may be found. The literal accuracy of some of what is said in such interviews may be questionable; however, the argument here has been that instances of impression management in interviews should not be regarded as a problem but as an opportunity. Impression management articulated by way of narratives and powerful rhetorical devices are instances of a socially constructed habitus responding to sensitive questions and, thereby, providing insight into the social world of the participants. Their articulations are indicative of the meanings that the issues in question have for them. They provide great insight into aspects such as identities, communities, moralities and behaviour.

The excerpts examined offer great insight into the discomfort with which some medicinal cannabis users speak about coming to use the substance, a substance they had not used before and were not well disposed towards, mindful of the stigma that ‘drug’ use still holds in the minds of many. The excerpts told us about the social context of these concerns, of communities in which the social standing of those seen as being of position can permit one to try a substance that might have positive health benefits, but individuals from such communities remain wary of others finding out about this. In some excerpts, the participants were participating in that very same moralised ‘blame game’, by making judgements about the alleged drug use of others and distancing themselves from them. We also saw how, for medicinal cannabis users such as the four considered in this article, this overriding concern for ‘respectability’, to keep their cannabis use as private as possible, precludes them from engaging in a critical and politicised struggle to have the legal position of medicinal cannabis use changed – a change that would benefit them, freeing them from the fear of potential criminalisation and perhaps the very social stigma that makes them so uncomfortable. Their habitualised reluctance to speak, publicly, about the issue results in them acting against their own interests.

These accounts are full of the rhetorical devices of self-presentation – the extreme case formulations, three-partedness and charge-rebuttal sequences, and it has been argued that rather than this being a problem, insight has been gained that is applicable to the sociological business of understanding the place of medicinal cannabis use in the lives of these participants. The narratives and articulations involved in self-presentations are always habitual, as is the reality of our use of language itself. People articulate in ways they have learned and to which they are inclined (Bourdieu, 1992), and this learning to use language is always socially embedded. Thus, accounts ‘help us understand, the complex nature of values, identities, cultures, and communities’ (Sandberg, 2010: 447). Ultimately, as Garfinkel (1967) argued, accounts are part of the world they describe, and it is those social worlds that we wish to understand.