Analyzing agency and identity navigation in addiction stories by drawing on actor-network theory and narrative positioning analysis

Abstract In addiction stories in which dependence is experienced as an antagonistic force, agency manifests as enigmatic. As narrators in these stories usually describe how they lost their agency to a substance, we may ask who then acts. By drawing on the actor-network theory, I propose that addiction stories should be approached with an ontology that allows agency also to non-human actors and conceptualizes both human and non-human agencies as relational. Moreover, I argue that addiction stories perform complex identity navigation that can be captured by analysing them from the dimensions of ‘story,’ ‘interaction,’ and ‘identity claim’. As addiction stories describe what kinds of unique human and non-human elements and assemblages have contributed to the development of addiction, they provide expressive material to analyze how their narrators reassemble their addictive past (story), justify it to their audience (interaction) and articulate who they are (identity claim). By approaching addiction stories through these dimensions, we can produce knowledge on what kinds of identity alignments with particular human and non-human actors promote or hinder addiction as part of specific assemblages. This knowledge can help health practitioners focus their treatment interventions on the relational identities that act as barriers or facilitators of recovery.


Introduction
In life stories on addiction that describe addiction as a problem, agency manifests as enigmatic. As narrators in them typically describe the process of how they lost their agency to a substance, we may ask who then takes over the agency and is the actor. Can material things act and drive action toward undesirable goals? By drawing on actor-network theory (ANT), I propose that addiction stories should be approached with an ontology that allows agency also to non-human actors and conceptualizes both human and non-human agencies as relational. This ontology questions the idea that human beings are autonomous individuals who develop addiction because of problems with their independency. Instead, it emphasizes that addiction develops and circulates in relation to different kinds of attachments that are called 'assemblages' (Duff, 2014) or actor-networks (Latour, 2005). In this ontology, addiction does not have an essence or a stable identity. It is an unstable phenomenon, ontologically multiple, and constantly made and changed in practice (Pienaar & Dilkes-Frayne, 2017). It is a relationally evolving collective of actor-networks that can take multiple forms and serve many kinds of purposes. As the recent literature shows, addiction-related assemblages can develop to become antagonistic to health and well-being, but they can also operate as helpers that facilitate substance users to complete their day-to-day tasks, to deal with isolation, or to experience life more pleasurably (Dennis, 2019;Pienaar & Dilkes-Frayne, 2017).
Moreover, by drawing on narrative positioning analysis, I argue that addiction stories perform complex identity navigation that can be captured by analysing them from the dimensions of 'story,' 'interaction,' and 'identity claim' (Bamberg, 1997(Bamberg, , 2020T€ orr€ onen, 2021). Depending on the dimension from which we approach addiction stories, addictive trajectories become associated with different kinds of actor-networks, selves and bodies. When we approach them as 'stories,' addiction stories usually describe the confrontation between the trajectory of the self that is driven by addiction and the trajectory of the self that seeks mastery over one's life. This is the content, or the 'what is told' dimension of addiction stories. In the story dimension, narrators position actors and their temporal trajectories vis-a-vis one another (Bamberg, 2020) and elucidate through what kinds of actor-networks their addiction grew, became stabilized or ended.
When we approach addiction stories as 'interaction,' we notice that new kinds of actors are linked to them, and we witness the evolvement of actor-networks in which the relation between the narrator and the listener as well as the interaction context take the roles of mediators influencing how the life stories are reassembled. This is the 'how events are told' dimension of addiction stories in which narrators interact with their listeners and position them in a specific kind of relation to the story (Bamberg, 2020). Telling one's life story is an ongoing process of reassembling, an act that shapes its contours, relations, and meanings. Every time a life story is told, something changes in it (Frank, 2010).
Moreover, when we approach life stories as 'identity claims,' we follow how their narrators individualize and subjectify 'who they are' in relation to wider discourses and narratives that circulate around them. Then we follow what kinds of elements narrators incorporate from them into their life stories and how these elements act as mediators that distribute their action in relation to certain kinds of values and norms.
Life stories in which addiction develops into an antagonistic force resemble illness narratives. They both deal with forces that disturb stabilized identities and patterns of actornetworks and push them into a process where they seek new relational forms and continuities (Hyd en, 2008). The majority of the existing studies have approached addiction and illness narratives from the perspective of story by paying attention to their themes (Riessman, 2003), structures (Williams, 1984), trajectories (Robinson, 1990), or narrative resources (Frank, 1995). Some studies have also examined addiction and illness narratives from the perspective of interaction (e.g. Denzin, 1987;J€ arvinen, 2001. However, we lack studies in which addiction or illness narratives are examined as a practice of three intersecting perspectives. The study by Lora Arduser (2014) is an exception. She argues that in the research of illness narratives, the analysis needs to be sensitive to subtle positioning processes in relation to how the past events are constructed, how the interaction situation mediates their representation, and what kinds of discourses are drawn from the surrounding culture to clarify the development of a problem and its possible ending (Arduser, 2014).
I will first explain what the actor-network theory and the narrative positioning analysis by Bamberg can offer to the analysis of life stories on addiction. After this I will introduce two example stories on addiction that I then employ to demonstrate how we can analyze addiction stories as a practice in which three evolving actor-networks emerge and intersect.

Actor-network theory and narrative analysis
In actor-network theory (ANT), action such as addictive behavior or recovery from it can be studied 'as a conglomerate of many surprising sets of agencies that have to be slowly disentangled' (Latour, 2005, p. 44). While traditional social science approaches tend to assume that action is based on human actors, ANT also treats non-human actors as equally important participants of action. Human and nonhuman actors can participate in action either as routinized and ready-made intermediaries or as active and transformative mediators (Latour, 2005). Intermediaries transport meaning faithfully as unnoticed and invisible forces, whereas mediators are visible and transformative: they translate actors into a new kind of coexistence.
In ANT, actors' agency is relational: their ability to move action comes from their heterogeneous associations, networkness (Michael, 2017). When someone develops an addiction, this addiction is composed of surprising sets of agencies whose ability to build associations, form actor-networks, and circulate from an event to another is the focus of analysis.
Life stories on addiction provide expressive material for actor-network analysis. They describe what kinds of unique and concrete conditions, situations, life stages, events, objects, and relations have guided and modified the development of addiction or recovery (Bamberg, 2020;Latour, 2005). From them we can track down how their narrators reassemble the bits and pieces of their past life and weave them into meaningful trajectories that clarify what kinds of 'chains of translations' the attachments and assemblages of their addiction or recovery have undergone. Addiction stories exemplify how 'action is other-taken! How it 'is mysteriously carried out and at the same time distributed to others' (Latour, 2005, p. 45). In the analysis of the other-taken movement of addiction or recovery, we can pay attention to how they become circulated from one event to another by mediators that form a trajectory that becomes privileged out of an indefinite number of possibilities (Latour, 1992).
For example, when people lose their ability to regulate and moderate their online gambling, this activity may develop into a privileged trajectory that multiplies the attachments of online gambling (T€ orr€ onen et al., 2020). Then addictive associations grow and form an actor-network of othering (Law, 2011). They infiltrate into existing habits, dissociate them from earlier connections, and translate them to serve the associations of the addictive self. Actors' power to influence action comes from their ability to multiply their associations (Latour, 2005). Actors can move action further only if they manage to recruit other actors to serve their interests. The more attachments the actor is able to build, the more the actor exists, has transformative agency, is able to overcome resistance, and move the action further. Accordingly, when someone wants to get out of addiction, s/ he needs to build an actor-network of associations that assists this desire and makes it capable to act toward the desired goal. In addiction stories the dimension of 'story' describes how actors become powerful as they manage to multiply their associations or weaken when they no longer are able to build associations that help them overcome the difficulties they face in action.
Life stories on addiction are performative acts done to an audience, so situation of communication, the identity of the listener, and the way the communication proceedsas elements of the dimension of interactionaffect the way the narrators reassemble the main events of their lives: what they highlight in them, omit from them, and how they distribute agencies to their relational elements. As the life stories analyzed here have been told to a researcher who in the interaction situation represents the 'face' of established society, this invites certain kinds of 'over-hearers' (Goffman, 1981) or 'imagined audiences' (Boyd, 2014) to take part in it as actors around which new kinds of assemblages are built. Because addiction is related to a reduced capacity to work, financial problems, and difficulties to be a functional family member (Miczo, 2003), it is a sensitive and stigmatized issue. As addiction stories' narrators know that the surrounding society addresses their addiction as deviant behavior, this makes them imagine judging, 'normalizing,' and 'stigmatizing' over-hearers and orientates them to narrate their addiction story from a defensive perspective. A defensive frame of 'enunciational dynamics' (Mattozzi, 2020) encourages the narrators to perform an interactive self that persuades their listeners to identify with the narrators' adversities, to feel sympathy for their suffering, and to understand their personal choices as morally accountable (J€ arvinen, 2020). Erving Goffman (1967) calls this kind of interactive action positive face-work.
Besides representing how narrators reassemble their addictive self and justify it to the audience, addiction stories also embody a dimension of 'identity claim' by articulating how their narrators individualize and subjectify themselves (Latour, 2005;cf., Bamberg, 2020). This becomes traceable in how narrators appeal in their stories and justifications to dominant discourses (Fairclough, 2003) and master narratives (Hochman & Spector-Mersel, 2020): what kinds of elements they circulate from them in their life stories, how they translate them as building blocks and attachments to their action, and how they, as mediators, connect their action to particular values and norms.

Data examples
In what follows, I use two example life stories on addiction to show how we can analyze addiction stories as a practice in which storytellers reassemble the actor-networks of their 'addictive,' 'recovering,' 'interactive,' and 'individualized' selves. To demonstrate what kinds of analytical procedures are useful in this kind of analysis, I take influences both from ANT and Bamberg's narrative positioning analysis, and from Greimas's narrative semiotics (Greimas & Court es, 1982), symbolic interactionism (Goffman, 1967;J€ arvinen, 2020;Sykes & Matza, 1957), and critical discourse analysis (Fairclough, 2003).
The example life stories on addiction have been taken from the study the author did together with Eva Samuelsson and Malin Gunnarsson with a title 'Online gambling venues as relational actors in addiction: Applying the actor-network approach to life stories of online gamblers' (T€ orr€ onen et al., 2020). The life stories were produced through individual faceto-face interviews in response to open questions about how the interviewees' addiction started, developed, and changed over time.
In the two example life stories, addiction appears as a problem, and therefore they exemplify life stories in which narrators reassemble their addiction to antagonistic assemblages. Both narrators are currently recovering from their gambling addiction by participating in treatment. Example 1 is a life story of a man in his 30 s. By gambling addictively, he lost a lot of loaned money but with the help of his spouse who covered the costs, he did not drift outside society. He still has work and a family. Example 2 is a life story of a woman in her 50 s. Her addictive gambling developed into criminal activity and continued for so long that when it was revealed, she could not escape imprisonment. Moreover, in the process she lost her work and family. These two life stories demonstrate typical gender differences in identity navigation between men and women in the broader material of addiction stories (N ¼ 34; to know more about the data see T€ orr€ onen et al., 2020) and this is the reason why I chose them as examples here.
While in the article I develop narrative tools to analyze identity navigation in addiction stories in which addiction assemblages appear as antagonistic, these narrative tools are also applicable to examining life stories on addiction in which 'addictive' assemblages help actors complete their day-to-day tasks, deal with isolation, or experience life more pleasurably (Dennis, 2019;Pienaar & Dilkes-Frayne, 2017). In the life stories in which addiction appears as a positive force, the actors of 'normal society' may take the position of an 'enemy' and be reassembled to antagonistic assemblages that complicate narrators' possibilities to manage their lives (Dennis, 2019).

Story: Analyzing subject's and anti-subject's trajectories
Studies show that illness narratives and addiction stories typically describe an identity change toward a new identity through the disruption of the established identity (Arduser, 2014). Robinson (1990) has found that illness narratives may portray a stable, a progressive, or a regressive journey toward a new identity, while Frank's studies (1993Frank's studies ( , 1995 indicate that the identity transformation may follow the plots of 'restitution,' 'chaos,' and 'quest.' Similar kinds of typologies have also been proposed in terms of recovering from addiction problems. The study by H€ anninen and Koski-J€ annes (1999) demonstrates that the recovery narratives from addiction may represent the story types of 'AA story,' 'personal growth story,' 'co-dependence story,' 'love story,' and 'master story. ' Mellor et al. (2021), again, have documented that they may perform four distinctive storylines: 'emancipation,' 'discovery,' 'mastery,' and 'coping.' These typologies serve as heuristic examples of the possible directions that the identity navigation in them may take and what kinds of elements then drive the struggle against the addiction. However, they do not capture the whole variety of the trajectories in addiction stories, nor do they grasp how addiction in them may undergo multiple translations and does not then follow one overarching storyline.
Therefore, I suggest that ANT-as described above-provides more dynamic concepts to analyzing addiction stories' identity navigation between conflicting forces and storylines. As explained above, when addiction stories' narrators experience their dependence as antagonistic to their health and well-being and want to come out of it, addiction stories typically embody the struggle between a self whose action is driven by the addiction-related associations and a self who tries to tame them and replace them with actor-networks that enable, mediate, and stabilize a healthier lifestyle (see Denzin, 1987). Addiction stories are in these cases worth approaching as stories that deal with two storylines. I call the storyline in which an actor seeks mastery over his or her life the trajectory of the subject and the storyline in which an actor's action is driven by addictive relations the trajectory of the anti-subject. Moreover, depending on the kind of addiction with which one struggles, the confrontation between these two trajectories may take different forms and appear as an opposition between the addictive self and the recovering self, the uncontrollable self and the control-seeking self, or the harmful self and the health-seeking self, etc.
The opposition between the subject's trajectory to recover and anti-subject's trajectory to indulge in addictive relations reflects tensions between two powers: between the assemblages of normalizing environment and the assemblages of addictive relations. When the normalizing environment encourages relations that create expectations to the subject to become self-governing, productive, ordered and a socially functional actor for the needs of the community, employment and family life (Dean, 2010), addictive relations have a tendency to disrupt these disciplinary processes and turn them to serve other purposes.
To analyze how in addiction stories action faces these opposing forces, we can use Greimassian tools of actant model and canonical narrative schema. They help us follow how narrators in addiction stories deal with conflicting trajectories of the subject and the anti-subject, mediate them in association with heterogeneous actors, and assemble them as actor-networks of attachments that enable addiction or undermine it (Latour, 2005;T€ orr€ onen & Tigerstedt, 2020).
In an actantial model, actors are approached as 'actants' that accomplish or undergo an act (Czarniawska, 2004). The model identifies six possible 'actantial positions' that form relations and articulate relational agencies in a narrative trajectory: subject versus object, sender versus receiver, and helper versus opponent (Silverman, 2020). 'Subject' refers to the main actor, whereas 'object' is the goal of action (Bal, 1985). Both human beings and organizations, nations, animals, material things, concepts, emotions, technologies, genes, future expectations, virtual entities, and so on, may act as actant-subjects or actant-objects in a narrative trajectory. This holds true also for other actantial positions (Greimas & Court es, 1982). The relation between 'sender' and 'receiver' communicates what kind of power, values, or norms make the subject's action necessary by allowing, authorizing, or legitimating it. 'Sender' is something that rules over the whole enterprise. It commands the subject to act toward the object for the benefit of 'receiver' that can be the same actor as the subject or something else. A subject's action toward the object is further mediated by 'helper' and 'opponent' actants. They refer to concrete situational actors which either promote or hinder a subject's action (Bal, 1985). Similar kinds of actantial relations can be identified from the narrative trajectory of the anti-subject as well (Greimas & Court es, 1982).
The actantial positions the actors take in their course of action are temporal and changing. They develop throughout a trajectory (Czarniawska, 2004). Also, not all actantial positions are necessarily articulated in a trajectory. When the narrative trajectory is fully formed, it encompasses three phases-qualifying, realizing, and sanctifying. Greimas calls a fully formed narrative trajectory the canonical narrative schema. In it, the qualifying phase identifies the main actor (the subject or the anti-subject) and an object and builds motivation for the main actor to pursue the object; the realizing phase describes its action against opponents to attain the goal; and the final, sanctifying phase evaluates how successfully or unsuccessfully the main actor accomplished the task (Greimas & Court es, 1982).
Two examples will show how we can use Greimas's actantial model and canonical narrative schema in the analysis of addiction stories. These will help us follow how the narrators reassemble two conflicting trajectories of the subject and the anti-subject and link them to heterogeneous actor-networks.
Example 1: Swedish man recovering from gambling problem Example 1 comes from the life story of a Swedish man who has struggled with gambling problems. In it, the narrator describes to the interviewer how his gambling escalated in relation to online gambling.
1. When I started to play online casino, I first felt that I could quickly win with a little money 2. quite big sums of money. I started to gamble more, and the amounts I spent 3. on the gambling increased. In a short time, I lost 2000-3000 EUR. My partner helped 4. me with that sum that time, but she didn't see how big the problem really was. I just 5. said: 'I've lost this sum' and 'I'll put an end to this.' Then another year and a half 6. went by and we bought a house. At the same time, I resigned from my job and before 7. I started a new one, I was completely free at home all by myself for two months.
8. There somewhere, when I didn't have my social circles and I missed my colleagues, 9. I felt disappointed that things were not going in the direction I wanted. I started 10. gambling more instead, and this time it went really fast. After going back to work, 11. I even gambled with bigger amounts. I also took all these bank loans I could, SMS 12. loans and anything I managed to get. When I got home from work, I cooked and had 13. my mobile phone lying on the kitchen counter and gambled while cooking. Nobody 14. could see what I was doing. I really tried to salvage the house and prevent the kids 15. from ending up on the street, but it all came to an end when I was no longer able to 16. pay my share of the house and bills. I knew I had a problem, but I was not willing to 17. seek help. But now the consequences had become so terribly big that there was no 18. turning back. My partner said: 'now it's enough.' She found a place and said: 19. 'today you and I will go to a meeting.' I realized that I couldn't have done this 20. without help (T€ orr€ onen et al., 2020). We can approach this life story as a struggle between two trajectories that are related to the gambling self as an antisubject and to the recovering self as a subject. The story shows how online gambling circulates and infiltrates into the narrator's main everyday life relations and practices and transforms them-through multiple translations-to serve gambling. In this anti-subject's trajectory, many kinds of mediators as actant-helpers make him gamble more: online casino venues, understanding partner, buying a house, changing jobs, loneliness, lack of colleagues, disappointment, taking bank and SMS loans, fear of failing the family, mobile phone, invisibility of gambling, and the kitchen counter. They form an expanding assemblage in which online gambling's relational transformative agency grows as it is able to build new associations. In this anti-subject's trajectory, the actantsender is to win fast money and to provide for the family, the actant-receiver is the family, and the actant-opponents are the attachments that end gambling.
Even though gambling seems to be the main object of the man's action in almost the entire story, the temporal structure of the story tells us that lines 18 and 19 describe the main struggle of the story. In this realizing phase, his partner makes an intervention against gambling and forces him to go to treatment that ends his addictive gambling for now. The action described in lines from 1 to 18 prepares and moves him toward the main battle as a qualifying phase, and the last sentence in lines 19 and 20 performs the sanctifying phase.
When looking at this story from the perspective of the subject's trajectory, we may say that being no longer able to pay his bills and share of the house acts as an actant-helper that together with the partner as an actant-sender and treatment as an actant-helper translate his action toward recovering. In this subject's trajectory the actant-receiver is not the man alone but the family.
Comparison of the anti-subject's and subject's trajectories shows that the addictive self is active and the recovering self quite passive. The addictive self becomes related to multiple mediators, whereas the recovering self is only attached to those of few. The story has a happy ending for now because the partner and the treatment manage to unlink the man from his addictive attachments and replace them with healthier attachments which, although not so well-grounded, are currently able to keep him away from gambling. In contrast to Example 1, this life story embodies only a qualifying phase. The narrator does not summarize to the interviewer how her story on gambling ends, because her recovery process-the main battle and the realizing phase-is in progress. The narrator tells this to the interviewer in the later part of the interview, where she emphasizes still struggling with the consequences of her dependence and with her self-esteem: 'I need to see that I have the right to live.' As in the previous example, we can approach this life story as a struggle between two trajectories. The story demonstrates how online gambling takes the position of the anti-subject and develops through multiple translations a powerful relational agency that becomes associated with an expanding assemblage. This translates the woman's everyday life relations and practices to serve the purposes of gambling. In this anti-subject's trajectory, her childhood abuse as attached to post-traumatic stress, the good girl syndrome, and research knowledge act as actant-senders (lines 21-29) that authorize her gambling. The actant-helpers are the online gambling venues, worries, trance, stealing money, turning off emotions, secrecy, and shame. Together they form an actor-network of associations that fosters her gambling and circulates it further.
Because Example 2 performs only the qualifying phase, the subject's trajectory remains open. In it, mental exhaustion, sleeplessness, eating disorder, the revelation of gambling, the disclosure of stealing, suicidal thoughts, driving herself to the bottom, family, and the feeling of relief act as actant-helpers that end her gambling and mediate it toward an assemblage that may keep her away from gambling.
The later part of the interview gives us more information about the progress of the subject's trajectory of recovering. We get to know that the disclosure of her crime leads to a course of events in which she is convicted of the crime, goes to prison, gets treatment, and continues the treatment by participating in the meetings of Gamblers Anonymous. In the light of this information, we may say that the actant-sender for her recovering self is the disclosure of her stealing for gambling. It brings her dependence under public scrutiny, puts an end to her gambling, and mediates it under various kinds of attachments of surveillance and care, which in turn act as actant-helpers. Together, these form an actor-network endorsing her process of recovery and give her the ability to resist addictive attachments.

Interaction: Justifying the addictive self's action
Narrators' storytelling 'is tailored to fit the expected response of the listener(s)' (Frank, 2010, p. 127), whose behavior the narrators seek to anticipate (Latour, 1988). As our two life stories have been told in an interview situation to a researcher who represents the 'face' of established society, this interaction situation-as explained above-brings into it 'over-hearers' (Goffman, 1981) and 'imagined audiences' (Boyd, 2014) that are linked to judging, normalizing, and stigmatizing voices. This encourages the narrators to reassemble their life stories to their listeners in a frame of reference of justification (see Latour, 2013;Bamberg, 1997).
Thus, in the analysis of addiction stories as interaction we can first consider the way in which their narrators use 'legitimizing explanations' to neutralize their addictive trajectories to 'over-hearers' who perceive addiction as a deviant, unacceptable, and inappropriate behavior (Scott & Lyman, 1968;Sykes & Matza, 1957). In this, the narrators may use different techniques. They may blame others and reject their own responsibility by arguing that their action was othertaken by elements outside their control. They may stress the exceptional circumstances that hit them and argue that because of this their addiction is less deviant than that of others belonging to the same category. Or they may deny the negative consequences of their action by emphasizing, for example, that they were not aware of them and in this sense not responsible for them. Moreover, they may condemn the society and people who judge them as prejudiced, insincere, intolerant, and so on (J€ arvinen, 2020).
The narrator of Example 1 refers to the exceptional circumstances to neutralize the development of his addiction. He explains to the interviewer that the escalation of his gambling was instigated by the loneliness of being left out of social circles at home (lines 6-9) with a technology that enabled full-time secret gambling (lines 9-14). He argues that this empty uncontrolled space, solitude, and unlimited gambling possibilities as an assemblage pushed him into a downward spiral.
In Example 2, in turn, the narrator refers both to the exceptional circumstances and to her unawareness of the negative consequences of her action to neutralize her addictive action. She draws the interviewer's attention to how her gambling problem was initiated by post-traumatic stress of childhood abuse (lines 22-23), work pressures (line 21), and a good girl syndrome (lines 24-25). These mediators together made her gambling problem escalate. She further accentuates the image of being moved by uncontrollable othertaken forces by emphasizing that she hardly remembers anything about her addictive period (line 33). In this way she mitigates her own agency in the escalation of her addiction that led to criminal consequences.
Thus, both life stories have elements of a 'victim narrative,' and their narrators use neutralization techniques to undermine their responsibility in the development of addiction. The narrator of Example 2, in particular, places herself in the position of a 'victim' of circumstances, which earlier research has identified as a common strategy of doing positive facework in addiction stories (Sibley et al., 2020).
Second, in the analysis of how the narrators justify their addictive trajectory to the listener(s), we can examine how they modify its relational agencies and configurations or oppose them with modalities of obligation ('having-to-do), desire ('wanting-to-do,') ability ('being-able-to-do,') and competence ('knowing-how-to-do') (Greimas & Court es, 1982;Sulkunen & T€ orr€ onen, 1997). Here, we pay attention to how the narrators use these modalities to undermine their agency in relation to the antagonistic assemblages of their addictive self. The relation of sender-receiver actants expresses how a subject's or an anti-subject's action is authorized and related to obligation. The relation between subject/anti-subject and object clarifies what kind of desire, will, and subjective motivation urges the subject or the anti-subject to reach the object. And the relation between helpers and opponents expresses what kinds of physical, material, and cognitive means facilitate or hinder the subject's or the anti-subject's pursuit of the object as abilities and competences or inabilities and lack of skills (T€ orr€ onen, 2001).
In Example 1, the narrator legitimizes his addictive gambling as an obligation to provide for the family (lines 1-2, 14-16). He does not express any positive desires or emotions in relation to gambling. We get the impression that gambling for him was like work, guided by a rational capitalistic motive to make a profit. By attributing the position of actant-opponent to multiple relational agencies, such as online casino venues, invisibility of gambling, understanding partner, loneliness at home, lack of colleagues, ease of access to bank and SMS loans, and fear of failing the family, he further emphasizes the powerful abilities of the anti-subject over him, downplaying his own agency and responsibility in the development of addiction. Moreover, by acknowledging that he lacked motivation to recover and by pointing out how his wife ended his gambling and made him go to treatment (lines 18-19), he expresses having outsourced control and caretaking in his life to his wife (Miczo, 2003).
In Example 2, the narrator legitimizes the development of her addictive assemblage around gambling as an obligatory reaction against her extreme anxiety caused by childhood abuse, difficult work situation, post-traumatic stress, and a good girl syndrome (lines 21-26). She stresses that she did not have a choice: her gambling 'was simply a way to survive' (line 36). She also justifies her drifting into gambling by referring to scientific knowledge that recommended a technique to relieve work stress by taking short breaks and by playing solitaire (lines 27-29). Because solitaire felt boring to her, she turned to online lottery (lines 29-30). In contrast to Example 1, in which the narrator does not articulate his desire to gamble, the woman explicates this. She explains how the attachments of gambling distributed to her a feeling of trance, relief, and oblivion against the attachments of everyday life which made her worried, scared, feeling bad (line 22), and panicky (line 31). Her vague memories of the gambling period (line 33) further strengthen the image that she lacked abilities and competencies to resist gambling. Throughout her story she emphasizes the impression that her 'healthy' body was taken over by such kinds of imposing actant-opponents whose power she was unable to combat and did not have enough competences to resist until she hit rock bottom. This sinking to the bottom crushed the actornetwork of antagonistic associations, freed her for new kinds of associations, and mediated her action toward recovery.
Third, in the analysis of how narrators of addiction stories justify and neutralize their deviant behavior to the listener, we can examine in reference to what kinds of emotions they make it understandable and meaningful and how they use these emotions as forces that participate in circulating their addictive associations from one event to another and make it expand. In Example 1, the narrator explains the escalation of his gambling in reference to the emotion of loneliness (lines 6-9). He offers it to the interviewer as a mediator that clarifies why online gambling so quickly and forcefully sucked him into its world. In Example 2, the narrator refers to emotions along the entire passage to shed light on how they participated in every twist and turn in the escalation and collapse of her gambling. For her, gambling first turned off all her emotions (lines 31-32). It functioned as a medicine against the stress, worries, and panic but in the end became translated into an activity that contributed to mental exhaustion and suicidal thoughts, and when exposed, to feelings of relief and shame.

Identity claim: Individualizing actor-networks in relation to dominant discourses and master narratives
In addiction stories, narrators not only reassemble and neutralize their addictive trajectories but also perform an identity claim (Bamberg, 2020). When analyzing how narrators articulate how their 'inside' is being prepared (Latour, 2005) and who they are becoming in the process of storytelling we can pay attention to the ways in which narrative trajectories and interactive justifications incorporate dominant discourses (Fairclough, 2003) and master narratives or oppose them (Hochman & Spector-Mersel, 2020). This gives us clues to detect how various wider global, national, and regional assemblages become integrated and translated as a part of narrators' local individual histories, interaction, and futures. As dominant discourses and master narratives articulate ideological beliefs, justify specific institutional practices, and seek hegemonic acceptance for their moral orders (Fairclough, 2003;Harris et al., 2001), their localization as attachments in life stories on addiction inform us what kinds of values, norms, and subjectivities mediate narrators' action.
When analyzing how narrators individualize and subjectify themselves in relation to dominant discourses and master narratives, we can examine life stories' intertextuality: how discourses and narratives that are 'external' to it are brought into it (Fairclough, 2003). We can further ask which discursive and narrative elements are then highlighted, which undermined, and which excluded (Fairclough, 2003) and how they become attached to other elements.
In relation to what kinds of wider discourses and narratives does the man in Example 1 subjectify his action? Even though he relates his gambling problem to his feeling of loneliness and isolation, these feelings arose in relation to his being included in the 'normal' assemblages of work and family in the society he lives in. He is not a marginalized gambler. He legitimizes his addictive self's action with the motive to make money fast. In describing his gambling problem, he emphasizes that he is an active, hard-working, competitionoriented man who seeks economic success to be able to provide for his family (lines 1-2, 14-16). In this way, he incorporates into his life story elements from the master narrative of neoliberalism that highlights economic motives in action and treats them as core mediators for a successful life. Moreover, by underlining how he is a work-oriented man who focuses on making money and who leaves the emotional caretaking labor in the family to his wife, he aligns with the master narrative of 'traditional hegemonic masculinity' that distributes clear gendered division of labor between male and female family members (Connell & Messerschmidt, 2005).
It is noteworthy that this male life story does not explicitly draw on or oppose the dominant discourses and master narratives of addiction. However, between the lines he links his addiction to the self-medication against loneliness (West, 2005). The female narrator in Example 2, in contrast, individualizes and subjectivizes her gambling more explicitly in relation to the master narratives of addiction. When she links her addiction to post-traumatic stress and childhood abuse (lines 21-23), she makes sense of her addiction by incorporating elements from a master narrative that emphasizes family background as a core mediator for addiction. Also, she underlines having learnt from research that playing a game is a good technique to take a break from stress and follows this advice in practice. In this way her life story also encompasses elements from a master narrative of self-medication that emphasizes rational motives in the development of addiction (lines 26-30). Furthermore, by expressing 'I drove myself to the bottom' (line 41), she indicates that she has been in AA-based treatment, which she has experienced as rewarding and which then has provided to her a mediator (competence) to make sense of who she is and how she should act in the future. Lastly, by underscoring her good girl syndrome, vulnerability, emotionality, feelings of shame, selfcare, and adaptation to the acute stress situation, she incorporates into her life story features of 'emphasized femininity' (Connell & Messerschmidt, 2005) and claims for herself a gendered identity that aligns with the hegemonic understanding of being a woman. In line with this gendered identity, her gambling was not other-taken by actors that were related to making money as in the case of the man in Example 1. Instead, her addiction was other-taken by actors that were connected to taking care of her 'traumatic' self, even to the extent of using illegal methods.

Discussion
In the article I have demonstrated how we can analyze life stories on addiction through the dimensions of story, interaction, and identity claim. My demonstration shows that in the dimension of story, life stories on addiction reassemble and record the trajectories of the subject and the anti-subject and describe what kinds of chains of translations their actors, associations, and assemblages have undergone. In the dimension of interaction, addiction stories perform a justification and form an actor-network in which the narrator and the listener act as the main mediators whose relation modifies what becomes emphasized, undermined, and excluded and through whose 'enunciational dynamics' (Mattozzi, 2020) the act of reassembling becomes linked to diverse kinds of overhearers and imagined audiences. Lastly, from the dimension of identity claim, life stories document how their narrators individualize and subjectify themselves in relation to wider discourses and narratives. The analysis of life stories from this dimension shows us what kinds of value-and norm-related linkages move their action further.
These addictive, recovering, interactive, and individualized selves which narrators enact and embody in an interview situation could also be called bodies or bodily selves. It is important to note that they and their actor-networks are not reducible to each other, nor do they form hierarchical relations. They rather exemplify how multiple selves, bodies, and realities intersect in the act of telling one's life story.
We should not try to force the diverse selves, bodies, and actor-networks under the command of one coherent identity or one shared reality. Instead, we should approach them in their multiplicity and understand that their intertwining is complex (Mol, 2002). While they intertwine, they partly overlap but do not become harmonized; who would have such a powerful agency over one's attachments to life that one could have a full control over them? Rather, these multiple bodily selves are other-driven. They are moved by assemblages in which they can grow by increasing their attachments, become stabilized into predictable habits, or disappear by losing their attachments. Thus, from the perspective of ANT, addiction stories' narrators are productive to approach as 'intersections.' They are 'nodes' through which many surprising sets of agencies, actor-networks, and bodies traverse, struggle with each other for dominance, and become translated toward heterogeneous directions (Latour, 2005). Similar considerations are also important for addiction and recovery. If selves, bodies, and actor-networks in addiction stories should be approached as unstable, ontologically multiple, and continuously made in practice, so should addiction and recovery. They should be approached as relationally evolving assemblages where action is other-taken and where addiction and recovery can obtain multiple forms, serve many kinds of purposes, and develop into many kinds of unpredictable directions (Pienaar & Dilkes-Frayne, 2017).
By approaching addiction stories as enactments of unstable multiple realities, manifold selves, and diverse bodies, health professionals and practitioners have tools to map from different dimensions what kinds of attachments and assemblages may promote one's well-being or act as barriers to it. First, the mapping of the addictive self's attachments and establishing how they as assemblages make the subject act in a harmful way helps in identifying what kinds of attachments and actor-networks the intervention efforts need to address, heal, and replace. Here, it is important to pay attention to the actantsenders and actant-helpers that in the addictive assemblages move the harmful action further. Because they are able to make the subject lose her or his control over the substance and to resist the subject's efforts to recover, they should be the primary targets of therapeutic interventions. For example, in the case of the Swedish man (Example 1) the intervention should be targeted at the aspiration to make fast money and to the attachments and assemblages that facilitate his online gambling. In the case of the Swedish woman (Example 2) the intervention, in turn, should focus on the attachments and assemblages that are related to her post-traumatic stress as well as on the attachments and assemblages that empower her online gambling. In these interventions it is important to discover in what way these addictive relations are related to repetitive, sedimentary and stubborn habits, how long history they have, and in what way they could be unraveled to new kinds of associations. Correspondingly, the identification of the actant-senders and actant-helpers in the recovering self's trajectory and assemblages provides information about the attachments and actor-networks that have a potentiality to translate action toward healthier trajectories. In terms of this issue, our example addiction stories do not provide much information and one of the first tasks in intervention therefore would be to map what kinds of relations might have the capacity to provide for the persons of the examples positive motivation and helping instruments to continue the process of recovery.
Secondly, producing knowledge on how treatment-seeking bodies more or less consciously use legitimizing explanations to neutralize their addictive behavior and link their addiction to compulsion, powerlessness, ignorance, inability, and emotional landscape, helps detect in what way their recovering efforts are overshadowed by elements and relations that they experience as stigmatizing. To follow actornetworks through this dimension also provides health professionals and practitioners competence to recognize how recovering bodies' alignment to a specific kind of face-work may hinder them from taking responsibility for their past actions and from moving on. For example, as the man in Example 1 justifies and neutralizes his addictive gambling as based on a normal rational capitalistic motive to make a profit this should be addressed in therapeutic intervention so that he understands the irrational and addictive aspects linked to his orientation and takes responsibility for what he did. While the woman in Example 2 also undermines her responsibility in the development of addiction, she however is more aware of the problematic nature of her gambling. Thus, in her case the intervention can build on this awareness and feed it so that in her future life she is able to cultivate responsible relations.
Moreover, the knowledge on what kind of identity claims help-seeking bodies' current understanding of themselves is based on and with what kinds of cultural values and norms their action is individualized and translated further, helps distinguish what kinds of identifications and moral orders may impede or promote their efforts to recover. In Example 1, the man's identification with values of a breadwinner who leaves the emotional caretaking labor in the family to his wife serves as a good starting point for addressing moral orders that may stand as barriers for his recovery. In Example 2, again, woman's identification with the moral order of AA provides a fertile platform through which to promote auxiliary attachments for her recovery.

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

Funding
Funding for the study was provided by the project '"Addiction" as a changing pattern of relations: Comparing autobiographical narratives about different dependencies,' financed by the Swedish Research Council [VR; grant no 2018-01297].