Fixed and fluid at the same time: how service providers make sense of relapse prevention in Swedish addiction treatment

ABSTRACT

This article explores how professionals within Swedish addiction treatment (n = 18) describe and make sense of relapse prevention (RP). RP is known as a self-control programme for maintaining behavioural change, helping people deal with high-risk situations. However, since self-control techniques have been incorporated widely in the addiction treatment field, the specificities of RP have become vague. To grasp what RP ‘is’, we draw on John Law’s and Annemarie Mol’s thoughts on how logics enact objects and realities. We thus follow critical scholarship in Science and Technology Studies and view treatment as a local knowledge-making practice that may depart from how it was originally designed. A key question is how RP is potentially transformed and made-to-matter when moved from the controlled settings of theorising and experimental studies to practice. The professionals used a logic of fixity to make RP stable, structured and evidence-based, easily distinguishable from other interventions. They also used a logic of fluidity to explain how and why they tinkered with it and adapted it to the preferences of both staff and attendees. The two logics enacted two different realities of addiction treatment: one in which RP is standardised, temporally demarcated and can solve most addiction problems, and another where interventions must be individualised, continuous and adapted to local settings and needs. It did not appear contradictory to ‘make up’ RP as both fixed and fluid; the two realities exist side by side, but with different material effects.

KEYWORDS:

• Relapse prevention
• professionals
• qualitative
• logic
• STS

Introduction

This study explores how professionals within Swedish addiction treatment describe and make sense of relapse prevention (RP). Besides a study by Theodoropoulou (2020) on relapse and recovery, surprisingly little attention has been given to RP in critical addiction research. What everyone ‘knows’ about it seems to come from inventors and proponents (e.g. Marlatt & George, 1984; Witkiewitz & Marlatt, 2004). The literature is dominated by efficacy studies that typically treat RP as a singular entity. Thus, detailed analyses of what RP can entail in different settings are needed (see Rhodes & Lancaster, 2019). Aligned with calls to explore such complexity in clinical practice (Byrne, 2013; Cartwright, 2013; Greenhalgh & Papoutsi, 2018), we take an in-depth look at how professionals make sense of RP. To scrutinise what happens with an intervention when moved from the controlled settings of theorising and experimental studies to local treatment, we interviewed Swedish treatment staff who claimed to work with RP. During data collection, we found that they described not one RP, but different ones, and here we analyse their discussions with theoretical concepts from Science and Technology Studies (STS) that can explain why certain objects appear slippery, multiple and ‘messy’ (Law & Singleton, 2005, p. 333).

The results centre on a key theme in debates on addiction treatment and public health: how can recovery best be achieved? We highlight the complex interrelationship between ideas and technologies on the one hand and what people do with these on the other (Greenhalgh & Papoutsi, 2018, p. 2). In an early study, Berg (1996) showed that the medical record is not just a container of patient information; it also shapes the very nature of medical practice. Similarly, research has shown that different treatment techniques typically individualise addiction problems and make strong assumptions about what is considered healthy and normal (Dennis, 2017, 2020; Fomiatti et al., 2017; Fraser & Ekendahl, 2018; Pienaar et al., 2017). We follow this analytical path and target what objects or entities ‘become’ through practice, which makes redundant any separation of the entity ‘itself’ and its discursive representations (e.g. Dennis, 2017; Law, 2000). The article scrutinises the multiplicity in practice of an understudied intervention, RP, and challenges the assumptions about recovery and agency that it rests on. As such, it adds to critical scholarship on local enactments and multiplicity in the public health field, found in this journal (e.g. Dennis, 2017; Rhodes, Lancaster et al., 2019; Will, 2017) and in others (Donetto et al., 2021).

Relapse prevention: popular but fuzzy

Emanating from a study by psychologist G. Alan Marlatt on men with alcohol problems during the 1970s, RP has gained a strong foothold in addiction treatment, and it now also includes behaviours such as smoking and gambling (Marlatt & Witkiewitz, 2005). Originally controversial, as it challenged the disease model of addiction (Dimeff & Marlatt, 1998, p. 514), RP is now a mainstream ‘talk therapy’ (Donovan & Witkiewitz, 2012; Roy & Miller, 2012).

RP is based on cognitive-behavioural principles that aim to help people avoid or cope with relapses into dysfunctional behaviour, where ‘high-risk situations’ are central (Marlatt & George, 1984; Witkiewitz & Marlatt, 2004). Many factors, including physiological states, cognitions and coping skills are said to determine how people respond to such situations (Witkiewitz & Marlatt, 2004, p. 229). RP concentrates on events, cognitions and emotions that occur outside of treatment, and tries through in-session identification, analysis and role-playing help people increase self-efficacy (Dimeff & Marlatt, 1998, pp. 521–522), and ‘continually assess their own relapse vulnerability’ (Witkiewitz & Marlatt, 2004, p. 231).

RP is today considered evidence-based. It is popular globally (Donovan & Witkiewitz, 2012) and is delivered by most addiction treatment agencies in Sweden (Ekendahl & Karlsson, 2021; Socialstyrelsen, 2019). However, distinguishing RP from other psychosocial interventions is challenging. As Hendershot et al. (2011, p. 2) note, ‘the diffuse application of RP approaches … tends to complicate efforts to define RP-based treatments and evaluate their overall efficacy’. While the Swedish national guidelines (Socialstyrelsen, 2019) recommends RP for several forms of substance use problems, it is listed together with the generic term ‘cognitive behavioural therapy’. Its core ingredients are present in most psychosocial treatments for substance use problems (McGovern et al., 2005) and RP is now not only employed to maintain change, but also to initiate it (Brandon et al., 2007).

Objects and realities in the addiction field

Critical scholars have showed how ‘addicts’ are made up as people with specific and often stigmatized problems in, for example, drug policy (Lancaster et al., 2015; Moore et al., 2015), bureaucratic systems (Moore & Fraser, 2013), the law (Seear, 2020) and in treatment (Ekendahl & Karlsson, 2021), but also through measuring scales (Dwyer & Fraser, 2016), diagnoses (Keane, 2012), and even personal accounts of substance use (Demant, 2009; Dilkes-Frayne & Duff, 2017). How ideas such as addiction (Fraser, 2016), volition (Karasaki et al., 2013), recovery (Fomiatti et al., 2017) and ‘triggers’ (Dennis, 2016) are understood and put to use have also been explored. A common theme in this work concerns tensions and contradictions. While neuroscientists claim that advances in the understanding of addiction have been made during the past decades, its critics view this knowledge as in ‘epistemological disorder’ (Fomiatti et al., 2017, p. 174), and as contributing to a situation where more and more behaviours are defined as addictions (Fraser et al., 2014).

Recent studies illustrate how treatment staff employ the brain disease model of addiction strategically in clinical practice (Barnett et al., 2018, 2020), and show that neuroscience is not implemented linearly, but is enmeshed in clinical negotiations, considerations and strategies. We can expect a similar reasoning among professionals working with RP in Sweden. The emphasis on self-efficacy in RP may align poorly with both the notion that addicted persons suffer from loss of control (Ekendahl & Karlsson, 2021), and with the Swedish tradition of a social-democratic welfare model that favours collective solutions to individual problems (Moore et al., 2015).

Theoretical position

Our study targets how RP is locally ‘made up’ in a complicated, dynamic and often unpredictable practice (Greenhalgh & Papoutsi, 2018; Lancaster & Rhodes, 2020). Rather than viewing RP as a universal intervention freed from historical or local contingencies, we assume, with the ‘evidence-making interventions’ (EMI) approach, recently proposed by Rhodes and Lancaster (2019), that it comes to existence through practice, and potentially does something with this practice: ‘An EMI approach focuses on what interventions become through their implementations; how they are worked-with into different things with multiple effects; and crucially, how they are made-to-matter locally’ (Rhodes & Lancaster, 2019, p. 2, original emphasis). The EMI approach has been employed, for example, to analyse scientific controversies surrounding hepatitis-C treatment (Rhodes, Lancaster et al., 2019) and how ‘multiple methadones’ are made up in prisons (Rhodes, Azbel et al., 2019). In this vein, RP is hardly detached from ‘pre-existing institutional, infrastructural, and material relations’ (Timmermans & Berg, 1997, p. 275), but is rather embedded in a network of other entities, including staff, clients, administrative procedures, policies, tools and technologies (Timmermans & Berg, 2003).

We draw upon the STS-literature on ‘objects’ (e.g. Law & Singleton, 2005) and scrutinize how technologies come into being as objects in realities (Fraser, 2020; Woolgar & Lezaun, 2013). According to Law and Singleton (2005, pp. 334–335, original emphasis), ‘realities, messy or otherwise, are enacted into being … such enactments take place in the practices of getting to know those realities’. When trying to understand how realities are enacted we need to rely on ‘different models for imagining objects’ (p. 335). Ontological performances where objects are imagined are ‘modes of ordering’ reality, that is, ‘arrangements that recursively perform themselves through different materials – speech, subjectivities, organizations, technical artefacts’ (Law, 2000, p. 23). Thus, different enactments are not reducible to different perspectives on the same phenomenon, but are rather indicative of different realities (Law & Urry, 2004, p. 397). Objects, including RP, can be multiple and enacted in practice through various modes of ordering, or logics (Law, 2002, p. 92; see also Law & Mol, 2001; Mol & Law, 1994).

We address two different logics at work when the service providers imagine objects and thus enact the realities of RP (Law & Singleton, 2005): a logic of fixity that makes RP easily distinguishable, demarcated and comparable (Mol & Law, 1994, p. 648), and a logic of fluidity that makes it unstable, gradient and open-ended (Mol, 1999, p. 83). Following the STS-literature, we assume that enactments of logics and enactments of objects are interrelated in and through practice.

Data and methods

The sample contained professional staff (13 females and 5 males) from two specialized clinics for addiction treatment, one social service agency, one outpatient treatment centre, and one local medical clinic in the Stockholm region. In Sweden, the responsibility for addiction treatment is divided by municipal social services and regional healthcare. Regional healthcare is in charge of medical treatment (e.g. detoxification, substitution treatment), and the social services have the main responsibility for assessing problems, providing psychosocial support and referring to treatment. Gatekeepers at these agencies informed staff about the research, and recruited the interviewees. They included nine social workers (who worked with needs assessments and different forms of psychosocial counselling), three nurses (who provided medicines and medical testing plus held group-based RP-sessions), four psychologists and two addiction specialists (who held individual and group-based RP-sessions). They were all at mid- to late career levels and catered to people with all kinds of substance use problems. Hour-long individual interviews (and one group interview) were conducted face-to-face by the researchers and targeted treatment approaches, concepts such as craving and relapse, and service user characteristics. The interviews were the main data used, but we were also given insights into the material arrangements surrounding treatment, including local facilities, counselling rooms, instruments measuring craving, figures on white-boards and RP manuals. While not subjected to formal analyses, such materialities did shape our understanding of the data. Even if RP is recommended for different types of substance use (Socialstyrelsen, 2019), the interviewees mainly referred to alcohol problems when discussing the intervention. The project followed standard ethical principles of social sciences and was approved by the Stockholm Board of vetting the ethics of research (no 2018/1064-31/5).

We started the analysis by identifying what the interviewees talked about, for example, ‘RP as treatment’, ‘conversation and interaction’, ‘respect and kindness’, ‘individualised interventions’ and ‘empowerment ambitions’. To increase understanding of how RP was made up as a relevant intervention, we turned to the STS-literature and its take on how objects can move through time and space. We thus continued with a more deductive coding, treating the data as illustrative of how logics enact objects. Extracts where RP was a stable and scientifically supported intervention were interpreted as drawing on a logic of fixity. Less conclusive and delimited discussions of the intervention that pointed towards flexible and adaptable characteristics were interpreted as drawing on a logic of fluidity. As will be seen, the interviewees oscillated between enacting RP as a fixed and a fluid object.

Relapse prevention as a fixed object to approach

According to a logic of fixity, RP rests on solid theory and evidence about recovery, and encompasses techniques that allocate responsibility to the patient. The intervention was a ‘class’, ‘course’ or ‘curriculum’, rather than formal treatment. The interviewees argued that genuine RP, through structured sessions and verbal interaction, let attendees grow as persons, learn how to stay abstinent and cope with problems. Mark (addiction specialist) summarises this idea and plays down his own role:

I usually tell the patients that ‘It’s called treatment, but don’t look at me as treatment staff, look at me as a coach, because I can’t fix your habits, you have to do that yourself’, I say. ‘But you can use me as a coach and I have some knowledge and I have some methods, but you have to do the main part yourself’. Then we know from these methods that there are counterproductive ways to work. To be too lecturing, to be too directive and so forth and not explore such things. (…) I mean, I think it’s really good, because offering people methods that are structured and evidence-based, they have effects. When it comes to these methods, you don’t have to be very, you must have attended a course on this subject and know how to work with it, but you don’t have to be an expert.

This quote fixes RP as a structured method whose efficacy is detached from context and staff. For Mark, the intervention can move from one setting to another without changing. By following key principles, ‘coaches’ can adopt it without specific expertise. This focus on the techniques in their own right and on getting the attendees onboard renounces the complexity of helping people recover. The patient takes a standardised class, learns some basic skills, changes behaviour, and goes on with life.

The interviewees also enacted RP as a fixed object through comparisons with other interventions. Such comparisons revealed that RP differed on key characteristics, being more or less manual-based, focused on group interaction, or ambitious in terms of outcomes. RP could also complement comprehensive interventions such as long-term psychotherapy or inpatient treatment. Mary (social worker), says that ‘one might need something else too’, and Andrew (social worker) claims that RP does not qualify as ‘regular treatment’, since that would ‘require so much more around and about [the client]’. While being a very important part of the supply of treatment, RP was anyhow considered insufficient for some people. Below, Dorothy (nurse) concludes that RP is more skills-based and centred on personal development compared to 12-step treatment.

The RP-course is about really looking into yourself and to see sort of how do I deal with these situations that we talk about? (…) How can you say no? So, it’s a lot about things that one may not work so much with within twelve-step treatment in the same way. And the difference from the AA-meetings is that you share more everyday life events I would say. ‘This happened to me last weekend. I had cravings. Then I chose to go for a walk’ or ‘I saw my daughter for dinner’ and the like. And then you may get some attention from the members on how to deal with it, or ‘Good work!’, you get a lot of cred. But most of all, the patients talk about community. So, I’d say the RP-course gives you skills in how to deal with your sober drug-free life but also how to go deep into yourself, sort of.

Here, RP not only highlights the importance of ‘saying no’, but also provides guidance in how this is achieved in everyday life. Dorothy mentions that craving experiences are relevant to AA-meetings (based on the 12 steps) too, but that they are given more serious attention in RP-classes, which fixes the latter as different and superior. Similarly, Karen (psychologist) says that the mindfulness-oriented RP that she also works with relies even more on individual responsibility and experimenting with emotions than ’traditional’ RP. She compares them as follows:

Because in traditional relapse prevention, you work somewhat with alternative activities sort of, maybe you should leave the situation, or divert the craving in some way. And here it’s almost the opposite perhaps. You should stop and notice what is happening. And that can really be more challenging. But I also believe it’s more helpful for those who are willing to give it a try.

This quote fixes RP by concluding that it, more than mindfulness training, rests on group dynamics, and on teaching people how to avoid difficulties. Its efforts to help attendees ‘divert craving’ is considered less challenging than interventions that build on introspection and staying in the situation. The interviewees emphasised, however, that the structure of RP was largely rooted in oral tradition. While Mark, above, maintains that RP is manual-based, the general notion was that there is a lack of research and theory on how to ‘do RP’. ‘Craving, thoughts, interaction, risk situations, problem solving’ (Dorothy), were agreed upon focal points, but information on how to operationalise them was outdated or irrelevant. The quote below from Barbara (nurse) illustrates how RP relies on but evades stabilisation.

Yes, there’s a lot of different research on relapse prevention and craving, or at least on relapse prevention primarily. And I have worked here for a very long time, read a lot. And there’s very little, ehm, how shall I put it, concrete practical, I mean working tools, and there’s not much, there are two books as far as I know. And I think the other treatment agencies use them, one of which is really old, surely twenty, thirty years old and is called ‘Väckarklockan’ [Alarm clock in English]. (…) Even if you read other theories, relapse prevention is concerned with looking at craving, looking at risk situations, looking at goals, looking at how one deals with thoughts about alcohol and may I do something different. These are usually the cornerstones and then they look somewhat different. We have these cornerstones too, but we also talk about confidence, self-esteem, you make an emergency plan and talk about problem solving.

Barbara draws attention to the weight of research, knowledge dissemination and manuals, but discusses a practice lacking guidelines and relevant textbooks. Here, RP boils down to some agreed upon ‘cornerstones’. This makes the intervention remain intact, but also vague enough to warrant local adaptations.

Relapse prevention as a fluid object to tinker with

Barbara later on in the interview clarifies that the material used as homework between sessions is ‘my own’. This challenges the depersonalised intervention described in manuals, and enacts RP as less standardised. According to this logic of fluidity, RP is a general approach to service users’ problems rather than a fixed technique. Numerous examples were given of how the intervention was tinkered with, and the interviewees were often non-discriminatory as to what counts as RP. Even interventions with a non-abstinence orientation (‘controlled drinking’) were made up as RP. The abovementioned ‘cornerstones’ were always useful, regardless if the goal was total abstinence, moderation or reduced harm. Andrew illustrates this when asserting that RP is flexible enough to embrace different individual goals within the same group:

I think you can adapt relapse prevention to encompass controlled drinking. If you deliver it in groups, and they all want controlled drinking, or if you adapt it individually to that person.

This ‘forming’ and ‘adapting’ of RP was made relevant in many ways. Number of sessions per ‘class’, temporal order of sessions, number of attendees and staff per session, choice of manual, fidelity to the manual, extent and character of homework, existence of follow up and booster sessions (a single extra session after the program) were variables that could be tinkered with. Below, Beatrice (psychologist) claims to authoritatively ‘pick’ from RP-cornerstones and the textbook ‘Väckarklockan’ [The Alarm Clock] (Österling, 2001) to make up RP.

We start out from the ‘Alarm Clock’, but we have reduced it. I usually do perhaps six RP sessions following, well, the ambition to be sober and risk situations and risk factors. Something about thoughts, cravings, apparently relevant decisions are also good, I think. And this with preventing emergency situations usually works fine too, but not everything. (…) Then you obviously perform an individual behaviour analysis, which may be about handling disturbing feelings or stress or conflicts. (…) Then I use the book ‘Back to Controlled Drinking’, with risk factors. I only do IDS-42 in groups, because I don’t think it is worthwhile to identify risk factors individually. So, when I think about it I actually pick from that one too.

While Beatrice mentions core ingredients of RP, such as risk situations and craving, she makes up an intervention attuned to the service users’ characteristics and behaviour. She also concludes that a screening-instrument of craving (IDS-42, Davis et al., 1987) is ‘worthwhile’ in groups but not individually, and that it is applied selectively. This logic of fluidity provides RP with rigor and versatility at the same time. While her intervention depends on personal decisions, she shows no doubts about whether it ‘is’ at all an example of RP. Her tinkering is rather a crucial part of enacting the intervention in line with local needs. The next quote from Mark illustrates the emphasis on balancing between following the manual and following the attendee.

Well, I am convinced that you should stick to the manual in a sober way. I mean, simply put, that you obviously should follow the manual, that’s your task. But if the individual needs to, something turns up, and you need to change the order, that’s what you do. One shouldn’t be rigid. (…) I’d put it like this: stick to the manual, but let the needs determine. But if we start to sort of mix and tweak too much, and improvise, I’m really afraid we’ll have difficulties saying: ‘Are we really doing this?’ (…) Because the manual is my map sort of, ‘where am I now’? And follow it. But when the map does not fit the terrain, well the terrain should determine.

Mark uses a metaphor to underline that RP needs to be fluid, that the terrain (the service user) can sometimes be more important than the map (the manual). Rigidity and standardisation must be attuned to individual wishes, which shows how Mark draws on both a logic of fixity and fluidity in the same account. While some interviewees described that they could disagree with attendees (e.g. regarding treatment goals), doubts about their prospects and abilities were downplayed: ‘we don’t communicate the negative parts to the client’ (Douglas & Betty, social workers). Thus, the intervention was flexible regarding both the weight placed on individual preferences and treatment goals. Everything was up to the service user. Aligned with this was the ambition to always ask for permission before proceeding with new topics during sessions. Corresponding with the idea that RP is a ‘class’, a start of a lifelong path to self-improvement, the interviewees emphasised that the clients themselves must decide what is relevant. Too much standardisation would limit the attendees’ self-control, which in turn would make them less able to learn. Faye (social worker) elaborates on this interrogative approach:

Same thing there, one asks for permission. Not ‘Now we do this!’, but rather ‘What do you think could help you, would it for example, be of help with a stop card [a tool to remind attendees about important decisions]?’ And then one explains what that is, and waits to see what the client thinks. (…) Nothing is compulsory. I think it is important not to think that we shall do this and it is obligatory, but rather that the client appreciates the advantages. Something for them to use preventively.

The fluidity of RP thus extends not only to the discretion of staff to pick and mix core ingredients, but also to client influence on treatment planning. In this way RP targets a neoliberal subject who can manage risks and live productively without interference from higher authorities, and who knows best what is advantageous for recovery (Andersen & Järvinen, 2007). As mentioned by Mark above, the individual is in charge of treatment-related decision making, and staff should merely ask, suggest and guide along the way. In the final quote, Barbara maintains that she adapts craving discussions to attendees who might find the issue irrelevant.

If a person drinks every day and sits on the park bench: ‘How will I handle craving?’. If you are homeless and really worn out, perhaps you don’t go about thinking like: ‘Oh, I feel restless and worried now, I think about alcohol much more!’ If you’d say that to a person, they would think you’re crazy. Those thoughts will perhaps have to wait until later. But rather, ‘What is craving for you?’ … ’How should I know?’ … ’All right, but as you enter the liquor store every day, how does that feel?’ … ’Well, it feels like: I must have this.’ … ’Ok, so that’s craving for you, great!’. One needs to really interrogate, be quite distinct, in short sequences.

The approach to craving is here adapted to the language and situation of the client. Implicit is the idea that understanding craving is more appropriate for those who are able to reflect and choose how to act. While the person in the quote is non-communicative and compulsive, it is nevertheless considered aligned with RP to encourage verbalisation of crude emotions. Barbara emphasises that craving manifests uniquely in different individuals, and relies on the therapeutic technique of ‘starting where the client is’ (Vakharia & Little, 2017) to inspire such talk. This provides an example of how RP is made up as similar to other ‘talk therapies’. It can flow in and out of treatment protocols and plans. While the logic of fixity stabilises RP as structured sessions targeting high-risk situations, the logic of fluidity makes it as multifaceted as the service users. RP maintains its emphasis on agency, rationality and self-awareness even when catering to experiences of compulsion and despair. It is made up as flexible in its very core.

Discussion

This study analysed how professionals in Swedish addiction treatment make up RP as a relevant intervention. Our prior work shows that addiction treatment staff refer quite creatively to the characteristics of substance users when establishing a reasonable match between the solutions they propose and the problems they identify (Ekendahl & Karlsson, 2021). Service users are many ‘things’, and are thus easy targets of an intervention such as RP. This study shifted focus to service delivery, but identified a striking similarity in how the professionals creatively talked about what ideas and technologies ‘do’ with people, and what people simultaneously do with these. Our analysis shows that the interviewees made up RP according to two logics that enacted different realities of addiction treatment. RP was on the one hand a standardised, temporally demarcated, evidence-based and universal solution to addiction problems; a fixed intervention. RP had close affinities with what it ‘is’ in treatment manuals and research papers. On the other hand, it was also individualised, continuous and fuzzy, tinkered with and adapted to local needs. Staff did not find it contradictory to make up RP as fixed and fluid at the same time, which indicates that different realities exist side by side.

Corresponding with previous studies of staff perspectives (Barnett et al., 2018; Barnett et al., 2020; Fraser, 2016, p. 12), this creativity may originate from the interviewees’ complex and multifaceted work. Both Fraser (2016) and Barnett et al. (2018, 2020) conclude that the brain disease model of addiction was considered somewhat simplistic by staff, but that they referred to it strategically in clinical practice to promote positive treatment outcomes. In our study, the two concurrent enactments of RP may serve to legitimise the Swedish treatment system’s reliance on one particular intervention as a solution to a multitude of problems among service users. The logic of fixity aligns with political demands on human service organisations to employ standardised and evidence-based interventions (e.g. Timmermans & Mauck, 2005), and the logic of fluidity with demands on individualisation and responsibilisation (Andersen & Järvinen, 2007; Järvinen & Mik-Meyer, 2012), and with practical feasibility and patient-centred care (Epstein et al., 2010). There is thus one ‘desktop product RP’ providing political legitimacy, and one ‘real life RP’ that provides professional legitimacy (for a similar reasoning on Motivational Interviewing, see Björk, 2016). According to Mol and Berg (1994), the coexistence of different logics, often referred to as the difference between principles and practice, is common in medicine and similar disciplines. Since different logics enact different objects, it is not necessarily incompatible to claim that RP is both standardised and individualised at the same time. Mol and Berg (1994, p. 250) would argue instead that such alleged contradictions ‘simply do not deal with the same’ object, that is, the same RP.

Then, which are the realities of addiction treatment that the interviewees enacted? Making up RP as a fixed object enacts a reality where substance use problems can be managed through accurate and timely interventions, and where involved actors take the right attitude towards treatment and recovery. For the service providers this means relying on treatment manuals, scientific evidence and service users’ ability to change lifestyle. For the service users, this means to submit to the authority of professionals and to the idea that cognitive and behavioural changes are possible and required. In this reality of ‘science-usage’, substance use problems are individual shortcomings that are best ironed out by voluntary and demarcated techniques.

Discussions about RP as a fluid object enacted instead a reality of ‘self-improvement’ which reflects neoliberal valorisations of non-authoritarianism and do-it-yourself philosophy (Berg et al., 2021). The professionals were coaches and the attendees were ordinary citizens on a quest for better versions of themselves. This justifies adapting RP to the situation of both service providers and users, since what counts is personal growth and not fidelity with a manual. Although this appears as a sympathetic and humanitarian approach, the reality that it enacts makes strong assumptions about ‘a healthy and balanced lifestyle’ (Hendershot et al., 2011, p. 13). Such assumptions can be traced back to the early writings of Dimeff and Marlatt (1998, p. 515), who argue that ‘dysfunctional indulgences’ ought to be substituted for ‘adaptive wants’ (e.g., recreational activities, massage, taking time to pursue a pottery class, reading the New York Times for hours on Sunday morning)’. Many people would agree to the positive health effects of exchanging substance use with massage, but the norms about productive lifestyles that surface through the reality of ‘self-improvement’ risk making RP irrelevant for substance users who perceive life differently (Lancaster et al., 2015). In other realities, the ability to constantly renegotiate identity and plan ahead may appear futile (Berg et al., 2021).

While otherwise different, we observe that both enactments of RP assume the existence of inherently rational and ‘good’ citizens who can reflect and think themselves out of precarious life situations. The emphasis on personal agency as necessary for successful recovery plays down the significance of whether interventions are fixed or fluid. As long as RP follows the rationale of cognitive behavioural therapy (see Gaudiano, 2008) it seems to matter less how it is made up in practice. The focus is on helping the individual deal with cognitive problems and structural conditions, rather than on ameliorating them. It is obvious that an intervention which responsibilises individuals has significant material effects (Fraser, 2020). While being delivered with the best of intentions, RP risks contributing to a streamlined supply of treatment focused on correcting the shortcomings of individuals. This article has hinted at how professionals make sense of RP, but more studies are needed of how its cornerstones are made-to-matter across locations.

Acknowledgements

Thanks to the professionals who engaged in this research, and to Eva Samuelsson and Josefin Månsson who conducted some of the interviews.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The work was funded by FORTE [Grant 2017-00290].