Discourses of behavioural addiction, normalisation and techniques of governmentality in inpatient substance abuse treatment

ABSTRACT The concept of addiction has expanded in recent decades to include diverse behaviours in addition to addiction to specific substances. Hence, the understanding of what constitutes normal behaviour and what constitutes addiction has been constantly changing. Substance-abusing clients are typically seen as having additional behavioural addictions, which manifest during their substance abuse treatment. In this article, we study the constructions of normality, deviance and the techniques of governmentality, produced by the discourses of behavioural addiction found in interviews with workers in an inpatient substance abuse treatment unit. Five identified discourses – psychological, disease, sociocultural, family and normalizing – differ from each other as regards to what is understood as addictive behaviour as opposed to normality and how it is explained; normality can be construed, for example, as the balance between internal emotions, health and adequate parenting, which may be beyond the reach of those addicted. What is considered a behavioural addiction is questioned in the normalizing discourse. Discourses also differ as regards to the techniques of governmentality and in the ways individual responsibility is understood.


Introduction
The concept of addiction is increasingly used loosely both in science and everyday life, frequently without further scrutiny. Central to its definition is dependence on a substance or activity (see e.g. Sussman and Sussman 2011). In recent decades, the concept of addiction has expanded to include a variety of excessive behaviours in addition to addiction to substances (Thombs and Osborn 2019). It would appear that ever more behaviours are seen as deviating from the norm and are deemed excessive, such as those relating to sex, sports or working, and are referred to as behavioural addictions. Thus, the conception of what constitutes normal behaviour and what constitutes addiction has been constantly changing (Thombs and Osborn 2019). It would further appear that dependence on any given activity or substance felt to be rewarding may be understood as compulsive (Alavi et al. 2012;Billieux et al. 2015;Orford 2001). Instead of a single diagnosis, addiction is rather an umbrella concept covering several interrelated clinical problems (Miller, Forcehimes, and Zweben 2019;Sinclair, Lochner, and Stein 2016;WHO, 2016). Different addictive behaviours are often defined as manifestations of the same phenomenon and, thus, are also intertwined (Chen 2016;Di Nicola et al. 2015;Sinclair, Lochner, and Stein 2016;Sussman and Sussman 2011;Thombs and Osborn 2019). For example, in the evaluation of dependence, Billieux et al. (2015) as well as Orford (2001) emphasize how disruptive and sustained the behavioural pattern is in terms of other spheres of life.
Behavioural addictions have been increasingly examined since the 2010s. At the same time, a discussion has ensued as to whether expanding the scope of addiction research to cover a wider range of human behaviours is pathologizing normal behaviour, rather than identifying problematic behaviour not previously identified (Kardefelt-Winther et al. 2017). It has been stated that intensive commitment to any specific activity can be deemed a behavioural addiction if diagnoses are continuously made (Billieux et al. 2015). This practice of expanding the concept of addiction reveals a need for society to identify and label any recurring, compulsive human behaviour (Hellman et al. 2016) that is difficult to distinguish between normal behaviour and dependence.
According to earlier research, clients entering substance abuse treatment often have concomitant behavioural addictions (see Di Nicola et al. 2015;Sinclair, Lochner, and Stein 2016) that manifest during the rehabilitation period. Behavioural patterns interpreted as addictions emerge in the institutional sphere, which become visible in the everyday processes of rehabilitation and in workers' talk and interpretations in relation to these problems and how they should be addressed. Perceptions of addiction have been studied among health and social service professionals as well as among lay persons and substance users themselves (e.g. Hellman et al. 2016;Samuelsson and Wallander 2015;Koski-Jännes, Pennonen, and Simmat-Durand 2016;Blomqvist et al. 2016). There are both consensus and variations regarding how dependence is understood (Thombs and Osborn 2019). For example, addiction can be interpreted as a disease or a learned behaviour. Hence, these various discourses exist in research about alcohol, drugs, tobacco, gambling and so forth as well as in everyday talk. In this study, we explore the kinds of discourses regarding behavioural addiction that are produced in interviews with workers in an inpatient substance abuse treatment unit. Our special focus is on examining what is significant within each discourse in terms of normality, deviance and the processes of normalization. In the field of addiction research, we wish to contribute to theoretical discussions on discipline and governmentality by illustrating how people with substance abuse disorders are disciplined and constructed as deviant, also in areas not related to their substance abuse.

Normalizing power and addiction
Normality is the opposite of addiction, which is defined as deviant. The production of normality is invariably linked to modifying and controlling deviant behaviour. In Foucault's terms, this can be described as a normalizing power that intertwines both objectifying and controlling disciplinary power and the power producing self-governing subjects with the aim to transform them towards normalcy (Foucault 1983). Normalizing power can be described as the process based on the difference between undesirable deviance and desirable normality, which operates through social institutions and practices (Schirato, Danaher, and Webb 2012). According to Foucault, normalizing and, thereby, normalizing power pervade present-day society ever more extensively, which has led to a situation in which the judges of normality are everywhere (Foucault 1995(Foucault [1975).
The line between normal and abnormal or deviant is fluid, yet always connected to cultural beliefs of 'good' and 'ordinary', i.e. how people are expected to live their lives (Raitakari, Juhila, and Räsänen 2018). Expanding the definition of abnormal behaviour defines culturally normal in a new way (Miller and Rose 1997; see also Waltraud 2006). Production of a new normal also creates new deviances, such as new addictions. Understanding addictive behaviour within a more extensive frame increases the use of normalizing power in governing people towards being more self-regulating, 'addiction free' subjects. (Kaisto and Pyykkönen 2010.) Foucault determines governmentality as the interface between techniques applied to the self (power to produce selfgovernance) and other techniques intended to control people (disciplinary power) (Foucault 1983;Hänninen and Karjalainen 1997). In reference to this, Jani Selin (2010) writes about therapeutic communities, which appear to be combinations of persuasive and controlling techniques with the aim to produce normality. A human being is understood simultaneously as an object of control and a controller of others and of one's self (Helén 2005(Helén , 2010Kaisto and Pyykkönen 2010). This normalizing power is present in individuals' behaviours and in the choices they make. It defines human action both consciously and unconsciously (Kaisto and Pyykkönen 2010).
At the core of rehabilitation in substance abuse treatment, professionals control and persuade clients, endeavouring to get them to work on their identities voluntarily and become more self-governing individuals. Governmentality thus appears as a process for ensuring welfare by guiding clients' behaviours and actions (see Rose 2000;Szott 2015;Helén 2010). The institutional task of substance abuse treatment is to produce behaviour defined as normal as opposed to dependence and the social problems often associated with addictive behaviour. In practice, this means envisioning a normal life, defining its requirements, and utilizing techniques aimed at progressing clients' self-government. (Kaisto and Pyykkönen 2010;Juhila, Raitakari, and Hall 2017.) Therapeutic communities are a fascinating target for research on forms of normalizing power (Hedlund, Landstad, and Tritter 2019;Selin 2010;Tourunen 2000;Weckroth 2006), because they carry both the history of authoritarian disciplinary power and the Western tradition of managing deviance by strengthening clients' self-governance. Hence, in today's community-based treatments, disciplinary techniques, such as restricted departures from the facility, expected active participation in treatment practices, abstinence and screening practices, merge with softer techniques that emphasize individuals' own choices and responsibilities, such as being meticulous, liable and attending the self-regulative practices in AA/NA. (Selin 2010.) Techniques of governmentality become visible in therapeutic communities, where clients encounter institutional norms and are expected to act accordingly.

The research unit
In Finland, institutional substance abuse treatment is almost without exception implemented as voluntary care provided by private care providers or municipalities. The institutions offer a relatively diverse field of services, albeit with much in common and also similar ideological foundations in their treatment practices. This study was conducted in an inpatient substance abuse treatment unit located in central Finland. The treatment facility in question provides nonmedical, therapeutic community-based treatment for both individuals and families. Treatment is based on cognitive behavioural therapy. In addition, the values in the treatment unit are influenced by 12step ideology. Treatment usually lasts one to two months, but the duration of therapy depends largely on referral from the municipality and the client's needs.
Community-based treatment is a therapy method, a way of organizing rehabilitation and an institutional position from which the workers speak. The discourses that the workers produce in the interviews contain rich descriptions of the everyday life and events of the treatment unit and their interpretations of clients' behaviours and their addictions. Thus, it is assumed that the discourses are closely linked to treatment practices, where the workers lead therapeutic groups and serve as parties to reciprocal therapeutic interactions (see e.g. Helén 2005). Discourses produce both parallel and competing interpretations of behavioural addiction (Thombs and Osborn 2019), but they all include normalizing power although with different emphases. The workers both use discourses in their everyday encounters with clients and are under the influence of these discourses in doing their work.
In studying discourses, we analyse practices built into the interview talk by substance abuse treatment unit workers, and at the same time, we examine the social and cultural meanings embedded in these practices, which in this case pertain to behavioural addictions.

Participants and data
The data was collected through individual semi-structured interviews ( Figure 1) with ten workers from the rehabilitation unit. Focus was set on behavioural addictions. Their work experience varied from a couple of years to 20 years. All participants were women. They were professionals in the health or social services, and most of them had training in therapy. More specific information about them was not collected in the interests of anonymity.
The participants did not know the exact theme of the interview beforehand, which was justifiable in light of the discourse analysis method to be used. In a close-knit work community, exchanging ideas on the research theme might have had a unifying effect on the discourses produced in the interviews. The duration of the interviews (27-54 mins) was determined depending on the workers' (W) and interviewer's (SE) interactions. The interviews lasted longer with more experienced workers.
Research permission was obtained in October 2017 from the background organization of the treatment unit. Participants' consent was requested after they had been informed about the study. They were free to withdraw from the study at any stage.

Discourse analysis as a method
The analysis started by reading carefully through the transcribed interview talk multiple times. This reading resulted in the observation that different interpretations of addictions were very common in the data. After that we started to code these interpretations systematically and ended up with different discourses defined in this study as relatively integrated systems of meanings that construct social reality (cf. Foucault 2002Foucault [1969; Parker 1992; Jokinen, Juhila, and Suoninen 2016). After differentiating the discourses, recurring themes, keywords and meanings present in each discourse were encoded.
In examining the discourses of behavioural addiction, our interest lies in the production of the normal, the deviant and the inherent techniques of governmentality. A connection exists between patterns of behaviours deemed socially acceptable and what are to be termed as behaviours subjected to normalizing power. Intrinsically, the concept of behavioural addiction is recurrently delineated.
The discourses reflect what is contemporarily perceived as normal and deviant behaviours in the context of substance abuse rehabilitation, as well as how accepted and deviant behaviours are produced. In the analysis involving parallel and competing discourses, we demonstrate how the discourses are construed in the interview data and their features. Discourses from the interview data were identified and examined by asking the following analytical questions: 1) How did the workers describe normality and deviances from it in relation to the behaviour of their clients? 2) What causes did the workers attribute to deviant behaviour? and 3) How were the treatment of deviance and the pursuit of normality, i.e. techniques of governmentality, construed in their talk?

Normality and deviance in workers' discourses of behavioural addiction
Each of the five discourses identified -psychological, disease, sociocultural, family and normalizing -constitute a unique and competing discourse of behavioural addiction. Table 1 below summarizes the main findings. These five discourses are analysed respectively; thus, every discourse is first described in terms of how behavioural dependency is explained. Next, we show what is significant within each discourse in terms of normality, deviance, and the process of normalization.
Excerpts are presented to illustrate these.

The psychological discourse
The psychological discourse is characterized by attributing behavioural dependencies to the individual's negative feelings, other psychological ill-being, or then attempting to explain the endeavour as a way to cut loose from substance addiction. The terms 'escaping' and 'coping' occur frequently in this discourse. The discourse can be distinguished by the prominent position of the individual's feelings and internal motivations when rendering behavioural addiction comprehensible. The reasons leading to an individual's excessive behaviours are attributed, for example, to traumatization, emotional deficiency, depression or feeling fearful, thereby evading confrontations in therapy with childhood emotional experiences and also genuine problems. The reason for the addictive behaviour and being rehabilitated from it are interpreted as reliant on the individuals' capacity to process their feelings. In her talk, the worker contextualizes the psychological discourse used in the institutional location, the rehabilitation unit. She refers especially to follow-up rehabilitation by stating that then the rehabilitation will no longer focus on problems such as withdrawal symptoms but will progress to processing entities. The worker directs the attention to the client's childhood and matters emerging from it. The style of the talk becomes psychological. Processing matters emerging from childhood may cause excessive behaviours. According to the worker's observations, such behaviour is considered to be an addictive behaviour. Next, the worker cites an example of behavioural dependence, baking deemed excessive. The style of talk brings out the difference between what is dependence and what is not. Baking figures as an example of an everyday action which, taken to extremes, becomes a limiting factor in rehabilitation. 'Normal' is defined as the opposite of excessive behaviour.
The interviewer asks if extending activities to the time after rehabilitation causes behaviour to be defined as dependency. The worker responds in the affirmative. She reverts to the baking theme by stating that this could be connected to seeking approval. The individual's internal motive is the decisive factor in the discourse, which causes the behaviour to be termed addiction. Finally, the worker moves the example to going to the gym, stating that there one can repair poor self-esteem. The explanation is again psychological. Compulsive behaviour originates from the individual's internal motivations, autobiographical experiences and the fears these give rise to, and the worker refers to this with the comment, 'so nobody can ever beat me up'.
In the psychological discourse, normality is the ability to face one's feelings and regulate them. To the client, this means an opportunity to achieve normality via an internal process of healing. Behavioural dependencies manifest as subordinate behaviours from which a person can achieve freedom by facing issues as yet unaddressed. From the perspective of regulation, the psychological discourse shifts responsibility for recovery onto the individual. Reasons for dependency are to be found in the individual's internal motivations and failures when facing up to feelings. The discourse turns the focus to the individual's internal world.

The disease discourse
In the disease discourse, dependency manifests as an incurable disease with which the client must learn to live. The discourse consists in opposites. Dependency isolates the individual permanently from the normal, healthy lives of the mainstream population; due to falling ill, normal life is no longer an option for the individual. What is at issue is the client's 'eternal' pursuit of the normal, which manifests as living with the disease, the ability to control the disease and a proneness to dependencies. This excerpt exemplifies the typical mode of interpreting behavioural dependencies as a phenomenon transferring from one object to another. W10: And so everything made a difference to it, hopes of a relationship and a wonderful life and children and a baby and all that. But working through it, she realised that it was the person herself or then those things, that if you want a child it might become obsessive. SE: So, if a dependency has developed for something, then that dependency manifests more easily in other things?
W10: Yes. Exercise is one where it easily comes out. Quite a lot of people have to consider if it's ok that I go to the gym five times a week or that I go there and don't go to the self-help group.
SE: A bit earlier you used the word diseases. Where do you think dependencies come from? W10: That's something I've thought about a lot. So, when you say that it goes from generation to generation. But what makes it a disease? Somehow, I think of it as there's nothing you can do about it yourself when you go and fall sick. So, those famous words, 'why don't you just stay off the bottle and stop drinking?' Well, I would stop if I could or if it was that easy. But somehow, it's that strong that a person can do nothing. A heart attack is a bad example, but if there's something wrong with your heart, well, that's a disease.
At the beginning of the excerpt, the interviewee describes matters pertaining to family life which stand for normality. Next, however, comes the word 'but', followed by an account of how the client realized by working through things that such matters may become obsessive. The comprehension of normality is, indeed, typically produced by first speaking of matters deemed normal and turning the talk immediately to dependency. The interviewer summarizes the interviewee's conception of the characteristics of dependency with a question as to whether the development of one dependency suggests a propensity to engage in dependencies in other matters, to which the interviewee accedes.
In the disease discourse, the normality of the healthy is forever beyond the reach of the individual suffering from dependency; their normality is habitually defined in terms of what is missing from life. In the excerpt, in which intensive use of the gym and failure to take part in self-help groups appear as symptoms of dependency, normality is defined as the opposite, i.e. attending self-help groups and spending less time in the gym. Vis-à-vis the mainstream population, the client is faced with her own normality. The client may come close to a normal life but not to a life totally free from dependency. It is expected that an individual attending treatment due to addiction will, by working through things, come to realize that in her behaviour anything may prove addictive. The pursuit of the normal manifests as self-control, which is attainable only by acknowledging the disease, thereby assuming the identity of a sick person. The disease discourse manifests as disciplinary power in which the client is assigned a permanent diseased identity, assuming the identity entails mental effort, admission of addiction and acknowledgement of being diseased. Having once fallen ill, normality in the client's life is narrowly defined; vis-á-vis the world of the healthy, she is permanently in a state of otherness. She must learn to live with the disease, which means constant selfgovernance and timely identification of possible new symptoms of dependency.

The sociocultural discourse
In the sociocultural discourse, the proliferating manifestations of dependency are interpreted as connected to contemporary society and culture. Here, dependencies are typically accounted for by perceiving them as socially learned. The predominating living habits, values and subcultures are thought to feed dependencies and give rise to new modes of addiction. When dependencies manifest in various social connections and at different times as different behaviours, the definitions of what is normal and what is proof of dependency vary across ages. The discourse is identified by the expressions used in which it is claimed that dependencies originate and take shape from the individual's life situations, cultural factors and social learning. SE: Have dependencies stayed the same over time? W3: They're not the same. Earlier it was purely doing work. Then there were no gyms and that, but then came this thing with the gym. Later then likely those computer games, that's one and, well, Facebook is one of them and the internet. It changes this way and that's the main thing in society. SE: But then came the gym? W3: We've had that, but it was only in the 1980s. Then there's been more in the younger age group, more drug users that do crimes. It's part of that culture that you should look good and have great muscles. Earlier, working was part of the old boozer culture. This gym concerns the younger drug users.
The interviewee claims that the phenomena in dependency have not remained unchanged. The discourse refers to a cultural change and its effects on behavioural addictions. In the traditional alcohol-oriented Finnish culture, normality is working hard when sober. In the world of drugtaking and crime, normality is appearing to be physically strong, which also subsumes the hierarchical nature of the drug and criminal world, the threat of violence and protecting oneself against that. The interviewee includes computer games and social media as new products of the culture predisposing one to dependency.
Dependencies are embroiled in subcultures and the values and ideals prevailing in these. The subject of the dependency will vary according to the culture or subculture with which the individual identifies. For example, the expression 'being part of that culture' suggests a normality that is redefined in each and every culture and subculture. Since dependency is perceived to be socially learned, recovery has a dimension of social relearning. As defined in the rehabilitation unit, entering a life free from dependencies appears as the way to normality.
In the sociocultural discourse, normality is produced as relatively and socially constructed. Defining normality is not freely available to the individual; it is tied to the social environment and the ways of comprehending the cultures therein. From the perspective of techniques of governmentality, the discourse invites the client to make a cultural change. Rehabilitation entails a transition from the subculture to the dominant culture. In the sociocultural discourse, dependency does not appear to brand the clients for the rest of their lives; rather, it is a learned behavioural model that they can change by means of new modes of thought and action.

The family discourse
The family discourse addresses clients as mothers, fathers, parents or families. The central subject is the child, the effect of the parent's behaviour and the effect of the interaction of the parents on the child. Behavioural addictions manifest notably as disruptions in behaviour detrimental to relations and actions within the family. Dependency taints everyday life and, consequently, in evaluating the effect on the child, the parent's excessive behaviour is paramount. Dependency is interpreted in terms of adequate parenting and is accounted for as an evasion of relations and obligations within the family. The reason for the dependency having developed is perceived to be learning from a model; dependency was acquired in the client's childhood home as a maladaptive model induced by negative conditions for growth. In the following excerpt, dependency is defined in a manner typical of the family discourse as detrimental to the everyday life of the family.
W8: [Addiction] does affect human relationships and the human psyche. You can be really angry, really freaking out, stressed and irritable with family if you can't do those things, that's the core of dependency. It's something you don't want to stop even if others think it's a bad thing. It can have such affects that a person stays awake more, sleeps less. Can't be bothered to take care of the family meals. Do the everyday routine, do laundry. But, in a way, the dependency always takes first place. Yes, for example [the one who did too much handicraft] did at least show that she had all sorts of other dependencies, and yes, it caused her quite a problem. Even her hands got sore and, physically, she had trouble. And then when the child was there, she got worked up really badly with the child. So that in her own opinion, she was there but still not there.
It is customary in the family discourse to define the disadvantages of dependency as detrimental to the everyday routines and interactions of the family. A situation is dubbed as dependency when the client does not desist from behaviour her family considers detrimental. Dependency manifests as irritability and anger when a person may not do the things on which she is dependent. The worker gives as examples the disruptions in the daily routine and dereliction of parental duties; the client does not prepare food or do the laundry.
Normality is defined as family life in which the interactions between parents and children are undisturbed; normality is consideration for one's immediate family and being present. The normality pursued is to be awake in the daytime, to sleep at night and to take care of the family in a way which might be challenging even in those families where the behaviour of the parents is not described as dependency. Any excessive behaviour deemed dependency may disrupt this. The discourse is decidedly normative. Frequently, the referring party is child protection services. Thus, techniques of governmentality extend to the work practices with families. The workers' obligation to sustain the discourse and monitor certain boundaries appears strong. Ensuring the child's everyday life becomes the core of treatment in the adult.

The normalizing discourse
The normalizing discourse is characterized by questioning workers' existing interpretations concerning the limits of behavioural addictions and what is normal. The discourse can be identified by its questioning talk in which normal, behavioural dependency and entitlement to control are interpreted anew. The roles of 'the ordinary person', the worker and the client are presented in reverse such that the client appears to be as normal or more normal than the ordinary person. It does not rely on a strict separation between the ingroup (us) and the outgroup (them). The talk tends to be critical of calling any excessive behaviour a behavioural addiction. The discourse features talk in favour of impassioned hobbies or seeking content in life in which any excessive behaviour at all is easily dubbed behavioural addiction. The normalizing discourse also challenges the notion that in the world outside substance abuse rehabilitation, many excessive behaviours are not deemed dependencies even though they may supersede everything else. In this discourse, the client's behaviour is compared with the worker's own behaviour, rendering the client's behaviour normal. SE: But then we're not talking about dependency. We're talking about some kind of survival strategy? And it might be like positive at the outset, because it helps to cope? W9: Yes. Certainly, over the change. But sometimes you can go a bit too far with over-identifying dependencies. Like everything is always seen as some obsessive action or dependency. So, you should look a bit further, so you don't go blaming that now you should stop this. Like exercise. I see from the perspective of the mental work what we go through here. So, there ought to be some sort of a counterweight.
The worker presents the client's behaviour as normal by comparing it to what she would do in a comparable situation. The worker goes on to state that behaviour easily interpreted as behavioural addiction may be an ad hoc survival strategy. The worker makes a direct appeal to the interviewer. Behavioural dependency comes to concern both the worker and the interviewer as a survival strategy to be interpreted as normality.
The worker attaches meaning to the client's situational survival strategies as behaviour intended to help her over the change and notes that things may be too easily interpreted as dependencies. She first makes use of a cautious expression, 'a bit', but she soon ventures some generalization. Finally, the client's exercising, which, interpreted as dependency, might assume an unwanted status, appears normative and an indispensable counterbalance to rehabilitation. The excerpt from the interview reveals the worker's mounting normalizing reflection. The excerpt illustrates a discourse of behavioural addiction as a matter of degree rather than kind. The discourse represents the worker's reflections on her work and the norms of the work organization; this also presents a redefinition of behavioural dependency and its boundaries.
When workers focus their talk on themselves, they challenge and question normality. The initial assumption of the rehabilitation facility is that clients are seeking to change their behaviour towards the normality of the mainstream culture and that the workers are there to help them. When a worker representing normality turns the talk to whether behavioural dependencies concern also herself, she questions her position of authority. In power relations, the focus is typically on the behaviour of the subordinate party in the hierarchy. In the normalizing discourse, this is turned upside down; the worker turns the scrutiny onto her own behaviour. From the perspective of techniques of governmentality, the discourse contains an element of similarity; dependency may equally concern the client as well as the worker. What is ultimately behavioural dependency and what is normal remains unresolved.

Discussion
The purpose of this research was to study the discourses of behavioural addiction produced in interviews with workers in an inpatient substance abuse treatment unit and, especially, to identify what conceptions of normality, deviance and techniques of governmentality are embedded in these discourses. We sought to learn about views of behavioural dependencies among substance abuse rehabilitation workers. Our study lends support to the view that a multitude of things are called addictions (see Billieux et al. 2015;Orford 2001;Thombs and Osborn 2019), but substance abuse workers also find it challenging at times to draw a line between behavioural addiction and normal behaviour. The workers assessed behavioural addiction from the perspective of the fluency of the rehabilitation routine (see Billieux et al. 2015).
As researchers, we were interested above all in how normality and deviance are produced, and the techniques of governmentality involved in this. The workers' propensity to classify a range of excessive behaviours as pathological was discovered. Even in the substance abuse treatment facility, the concept of addiction is not outlined solely regarding substances. Instead, it extends to a multitude of activities that are considered extensive. This has consequences for what is considered abnormal or deviant and how this kind of behaviour is managed in treatment practices. As much as individuals categorized as normal in present-day society enjoy the diversity of socially acceptable life, substance abuse treatment clients are not included in this. As earlier research has established, what is normal for a client in substance abuse treatment appears to differ from the normality of the mainstream culture and is more narrowly circumscribed (see Nettleton, Neale, and Pickering 2012;Selin 2010;Virokannas 2004), although normality was indeed questioned in the talk of the workers. Excessive involvement in any activity is easily considered an addiction (see Billieux et al. 2015). As a result, the boundaries of normal and abnormal are redefined and the concept of addiction is expanded (see Kardefelt-Winther et al. 2017).
We set out to shed light by identifying the five discourses of behavioural addictions perceptible in the interviews (see also Helén 2005). The discourses produce different social realities and influence the ways in which clients are encountered, their chances for autonomy, and workers' expectations regarding their clients' coping. In each discourse, the reasons for behavioural addiction as well as the border between normality and deviancy receive their own definitions and, therefore, also produce different ways of governing the clients.
Discourses come into being in certain institutional contexts, which may also revert from the explanatory models of dependency, such as sickness, conditioning or family system models, to cognitive, psychoanalytic and sociocultural models (see Thombs and Osborn 2019). The workers' talk becomes reflective of the workers' own attitudes and those of the work organization. We are aware of the strong presence of the psychological discourse in the data, as cognitive behavioural therapy is the main ideology and method in this facility (see Freeman and Freeman 2005). The disease discourse manifests in the rehabilitation unit because in its values and practices, a 12-step programme, medication-free treatment and cognitive behavioural therapy are central. Hence, it is also natural that moderation, balance and social learning are present in the interviews via the sociocultural discourse. Social learning mainly embodies the ideology of community rehabilitation in which peer support has an important mission. The family discourse is markedly context-bound and differs from the individual rehabilitation context, although both are located in the same rehabilitation unit. While people attend individual rehabilitation on a voluntary basis, they usually attend family rehabilitation on referral from child protection services. Parenthood becomes the object of assessment and is reflected in the workers' talk. Due to the fact that the working practices of the rehabilitation unit are based on elements from several different directions in therapy and several different ideologies, the normalizing discourse is inevitably present; the workers can be eclectic and vary their approaches and methods. Although the workers are in the last instance bound to the organizational culture, the professional skills in the field include critical reflection, which manifests in the normalizing discourse.
In this study, the substance abuse rehabilitation unit is a place where meanings are born, within which discourse, behavioural addiction is rendered comprehensible in substance abuse treatment; thus, the kinds of normality and deviance produced are not insignificant. There are consequences for work practices and also for the way in which the roles, responsibilities and normalizing power of the workers are understood in everyday interactions (see Selin 2010). Although the workers share the same institutional place and professional culture, within the institution, different interpretations of deviance, normality and the responsibility of the individual concurrently exist. The multitude of perspectives may give rise to tensions, but it may also be a matter of applying different points of view in work practices according to which ones are deemed efficient.
Regarding the normality they produce and the techniques of governmentality, the discourses represent differing and conflicting ways of comprehending. It is not a matter of the differing idiolects of different individuals but rather of varying discourses within the same individual interview. Likewise, similar modes of speech recur in different interviews, and these can be interpreted as conveying that the workers share the same institutional location and professional culture.
The conceptions of individual responsibility and ability to act differ widely between discourses. On the one hand, the alternation and concurrence of the techniques of governmentality in the workers' talk charges the client in relation to herself and her environment, and on the other hand, affords opportunities to build the identity of the rehabilitant without risk of stigmatization. This manifests the alternation of external guidance and control focused on the person herself (see Kaisto and Pyykkönen 2010;Raitakari, Juhila, and Räsänen 2018). The disease discourse differs from the others in that it liberates the individual from responsibility. It represents an orientation different from those of the other discourses, in which the individual is interpreted as capable of rehabilitation and achieving normality and, thus, to become a self-governing subject. In the disease discourse, techniques of governmentality build on external supports that clients need all their lives. In community rehabilitation, the disease discourse renders control of the client an indispensable part of treatment. As clients are incapable of limiting their behaviour, they need strict rules and surveillance. The role of the worker is to control but also to apply softer persuasive techniques to guide the client to acknowledge the diseased identity.
The normalizing discourse challenges the meanings of normality assigned by the other four discourses. Behavioural addiction is challenged and interpreted as a possible pathologization of the everyday (see Kardefelt-Winther et al. 2017). The result serves to express how drawing a line between normal and abnormal is fluid in relation to behavioural addictions (see Miller and Rose 1997). The existence of dependency as a pathological phenomenon is not denied, but behaviour previously characterized as dependency is normalized. From the perspective of treatment ethics, the manifestation of the normalizing discourse in the workers' talk is significant. Substance abuse treatment is conducted in a world of various techniques of governmentality and of producing identity (see Selin 2010;Szott 2015;Tourunen 2000;Weckroth 2006). At least in the case of the normalizing discourse, the use of normalizing power to keep clients and workers apart is lacking. Thus, the interpretive horizons converge on each other (see Weckroth 2006). It is encouraging to encounter a discourse that is lenient towards clients, although it may conflict with the conceptions of the mainstream culture. The existence of the normalizing discourse is significant for the therapeutic relationship between client and worker and also in regards to encountering clients in an ethical manner.
There are certain limitations to this study. One is the small number of interviews conducted in a single inpatient setting. Delineating the normal and the deviant might have turned out differently if we had interviewed workers working with clients whose problems were less severe. Also, an interview study does not reveal what the nature of the relationship between the discourses may be in the routines of rehabilitation. Yet, the data does indeed permit decisions on the existence of the discourses and their simultaneous occurrences.

Conclusion
Five identified discourses -psychological, disease, sociocultural, family and normalizing -differ from each other as regards to what is understood as addictive behaviour as opposed to normality and how it is explained; normality can be construed, for example, as a balance of internal emotions, health or adequate parenting that may be beyond the reach of those addicted. What is considered a behavioural addiction is expanding, but it is questioned in the normalizing discourse. Discourses differ as regards to the techniques of governmentality, and the ways in which individual responsibility is understood. In the future, it would be interesting to ascertain how the various discourses exert influence in daily rehabilitation practices. Even though the study was concerned with the micro level of talk, the findings can be connected to the foundation of treatment practices; the discourses of the workers are intertwined with wider societal discourses on the nature of culturally acceptable citizenship (see Foucault 2002) and conceptions of addiction (see Thombs and Osborn 2019). The work at hand continues the research tradition in behavioural addiction as it is characterized by polyphony (Thombs and Osborn 2019), but it inspects them further in terms of how normality and deviance are produced and the techniques of governmentality involved in inpatient substance abuse treatment. Substance abuse treatment professionals work amidst controlling, techniques of governmentality and producing identities. In this ethically loaded position, it is important to be able to see that roles can be converted and also to question one's position of authority.