Studying Intimacies that Matter: Affective Assemblages in Research Interviews with Forced Migrants

ABSTRACT This article explores methodological challenges in studying intimacies through analysing affective assemblages in research interviews. The participants in this study are forced migrants trying to reunite with their transnational families. Building on scholarship where affect is considered precognitive, meaning that it happens prior to emotions, this article approaches affective knowledge through methodological exploration. The analysis explores two research encounters from a data set which include interviews with 55 forced migrants in Finland, some of which were done together with a research assistant working as a translator. In addition, data include the researcher’s notes on re-listening to six interviews. The research question is how affect influences research interview situations and what effects these affective assemblages have on all involved. Additionally, the interest is in what these assemblages reveal about empathy and difference. The results show that intimate relations may manifest through a shared, intense affective assemblage where the borders of the researcher and participant become blurred. Difficult experiences of losses and injustices in the realm of intimacies may also cause affective dissonance. In these research encounters, difference between the researcher(s) and the participant(s) is left unsolved. The unresolved difference may reveal power relations between the researcher and participant.

In my different research projects with refugees and forced migrants, it has seemed that emotions encourage action but could also almost paralyze the interviewees into catatonic states. This has especially been the case when interviewees and research participants have transnational family members in distress. I have also often found myself immersed in these intense atmospheres during interviews or other fieldwork. In this article, I explore methodological challenges in studying affective intimacies through analysing affective assemblages in research interviews. The research question, or a wondering of sorts, is how affect influences research interview situations and what effects these affective assemblages have on all involved. A secondary point of curiosity is what these assemblages reveal about empathy and difference. This discussion is an ethical exploration.
In the title I suggest that I am studying intimacies that matter. This refers to my understanding of the theoretical concept of affect, which I utilize in this article. In my understanding of affect, I rely on perspectives influenced by thinkers such as Spinoza, Deleuze and Massumi, who investigate the body's capacity to be and act. In this line of thought, affect is considered precognitive, meaning that it happens prior to emotions (Massumi's introduction in Deleuze & Guattari, 1988, p. xvi). Thinking with these scholars, affect itself is material. However, I am not suggesting that I follow these thinkers "purely". Instead, I approach this work in conversation with feminist thinkers, who focus on the study of emotions but also delve deeper into the "bodies capabilities to affect or to be affected by other bodies", as in Spinozian/Deleuzian thought bodies do not simply "have" affects but they affect and become affected by other bodies. (see Åhäll, 2018;Ahmed, 2000Ahmed, , 2014Liljeström & Paasonen, 2010). A strand of discussion in the analytical sections also concerns the feminist discussion (and critique) on empathy (see Hemmings, 2012Hemmings, , 2015Pedwell, 2012). I have not attempted to fit the analysis into a theoretical frame as much as I have tried to show emerging issues through data examples and consider them in terms of theory and methodology. This approach is influenced by Deleuzian methodologies, where an important part of the methodological exploration is allowing for emergent and non-predetermined themes (Coleman & Ringrose, 2013).
The premise of the analysis in this paper is that research encounters, and especially interviews, contain affective, embodied and sensory knowledge. This is often bypassed in qualitative research when the analysis concerns only textual data that is analysed though thematic arranging and categorizing (see also MacLure, 2013). Compared to the vast methodological discussion on embodiment in the field of ethnographic research, relatively few scholars have concentrated on the issues of methodology and embodied knowledge in research interviews (some notable exceptions include Burns, 2003;Del Busso, 2007;Jackson, 2013;Lahti, 2018;MacLure, 2013). These studies have highlighted the importance of the embodied and reflexive knowledge that emerge in the interview situations. This knowledge has been described as "little experiences" (e.g. Coleman & Ringrose, 2013). However, concentrating on embodied knowledge can help to tease out new dimensions of the phenomena studied, and affective knowledge can also be understood as "more" rather than less, or "little". For example, Stewart (2007, p. 3), who studied the affective dimensions of ordinary life through ethnographic exploration, finds that affects can be captured in small "scenes" of the ordinary. She describes her analysis as an "assemblage of disparate scenes" in which "thought is patchy and material" (Stewart, 2007, p. 5). This concentration on the short, fleeting moments is often in the centre of studying affect. It may be that because of this possibility to focus on the small, researchers often shy away from considering interview situations in terms of affect. After all, studies based on research interviews often aim at collecting more material rather than concentrating on less.
Although my focus is methodological, it is important to note that this study also contributes to a lively academic discussion on studying transnational family relationships and the affects involved in maintaining these relations across borders. Previous studies that also deal with emotions or affects focus on, for example, mothering, caregiving and care-chains (e.g. McKay, 2006;Parreñas, 2005). Melander, Schmulyar Gréen, and Höjer (2020, p. 104) speak about gendered care triangles and relational transnational flows within them. They point out that migrant women perform mothering even across borders, whereas migrating fathers have a more absent role in their children's lives. The scholarship on transnational family ties has also drawn attention to the emotions that underpin and maintain ties across borders (Baldassar & Merla, 2014). This area of scholarship reveals that especially intimate affects that circulate between family members are able to travel transnationally. This aspect of affect as transnational is very rarely studied (however, a notable exception is Wise & Velayutham, 2017, whose methodological insights I will discuss in the next section). My analysis will show that interview situations may involve affect jumping across spatial and temporal domains (see also Dragojlovic, 2015).
This research project and the interviews conducted in the project concerned the very affective issue of family separation of forced migrants. The data includes altogether 45 group and individual interviews. For the analysis here, I re-listened to six interview tapes (lengths from 60 to 150 minutes) to be able to (re)think about the data in terms of affect. I use two stories as examples in this article. In addition, I refer to fieldwork when discussing the affective geographies of the time period of conducting the interviews. I stayed in contact with many of the interview participants due to their being in need of all kinds of support (and often guidance to legal and/or social services). Because of this extensive correspondence beyond the interviews, this study also resembles ethnographic fieldwork.
I use the term forced migrant to refer to a group of people who arrived in Finland seeking international protection (asylum, subsidiary protection, compassionate grounds for protection, victims of human trafficking). However, it should be noted that the concept is highly fluid, as the circumstances of being forced to leave, and those with some level of decisiveness and agency, are far from being clear distinctions (see, for example, Erdal & Oeppen, 2018). Administrative categories of international protection versus work-related residency applications do not reveal the underlying conditions clearly either, as the permit type is often a coincidence of sorts. Nonetheless, all my research participants had permits granted in the asylum process. Consequently, I have stuck with the concept of the forced migrant, as I think it points to an important aspect of the vulnerability that the research participants experience with the precariousness of their own and their family members' residency statuses.

Notes on affect, methodology and ethics
It is a difficult or even impossible task to exhaustively describe what "affective methodology" is, as its very definition defies categorization and classification. There are, however, several signifiers that may help in grasping the approach in practice. First, as Knudsen and Stage (2015, p. 5) state, affective methodology has "a strong situational specificity", which is a necessity if one wishes to empirically investigate the workings of affect. Affect is often a joining or meeting of forces in an encounter between people or things, so it is important to go into detail in describing these moments. Knudsen and Stage (2015) further explain how capturing pre-discursive affect, which is an approach I more or less aim to adapt in this text, may mean focusing on "affect as somehow disturbing normal communication structures" (Knudsen & Stage, 2015, p. 8). My use of Hemmings's (2012) work on affective dissonance (and possible affective solidarity as an effect) is focused on disturbing moments wherein my and the interviewers' positions either intersect or, most often, not. The analysis presents two disturbing scenes that rupture my pre-imagined structure of the interview situation.
My analysis in this article follows the tradition of specificity by focusing on these two intensely affective scenes from the interview situations. However, it should be noted that the scenes can never be (fully) separate from other scenes and encounters that have influenced the research process. First of all, the selection of these particular scenes has involved all 45 different interview situations with 55 forced migrants and hundreds of scenes within them. After first selecting six recordings to relisten to, I have selected these two examples, as they reveal something more-than mere categorizations or descriptions of the whole.
A suitable reference for the selection process of those six particular interviews to listen to could be MacLure's (2013, p. 172) account of how, in developing new approaches of qualitative coding, "we could acknowledge those uncomfortable affects that swarm among our supposedly rational arguments-moments of nausea, complacency, disgust, embarrassment, guilt, fear and fascination, that threaten to undo our certainty and our self-certainty by, again, allowing bodily intensities to surge up into thought and decision making". I wanted to dive into the deep end of studying affective intimacies and explore the methodological challenges involved.
The interviews were part of a project entitled "Family Separation, Migration Status and Everyday Security", funded by the Academy of Finland (my part was conducted in 2018-2020). The initial aim was to conduct second and third interviews with the participants before and after family reunification, but as the administrative processes were extremely long at the time (often from two to four years), we only conducted second interviews with two families (one of which was also interviewed for a third time). This correspondence has been significant in my overall understanding of family separation.
The actual interviews in dealt with the refugee journey, experiences of gaining asylum, everyday life in Finland, the family reunification process, family separation and experiences of (in) security. During the fieldwork I had support from three different research assistants with Persian (Dari and Farsi) and Arabic language skills who worked on my sub-project for three to six months each.
Eighteen of the research participants were women and 37 were men. They had lived in Finland for periods ranging from less than a year to eight years. Sixteen interviewees had arrived in Finland as unaccompanied minors and were attempting to reunite with their parents and siblings. All but one, a 17-year-old boy, were adults at the time of the interview. Thirty-nine interviewees had come to Finland as adults and were attempting to reunite with their spouses and/or children. I had several ways of finding participants. My contacts from an earlier ethnographic study with Afghan migrants helped me in finding several interviewees. The research assistants also had important contacts with different migrant organizations. In addition, we published several calls in Facebook groups and various migrant NGOs' mailing lists to find research participants. At the end, there were many more people who wanted to take part in the study than we actually had time to interview.
The urgency of the situation created the main ethical issue of the fieldwork. We, of course, handled formal ethics, like the ethical approval from the ethics board at the University of Turku, Finland. In publications, I have only used pseudonyms for the interviewees, and I have carefully considered issues of privacy during the research process and considered data protection risks from early on. However, our biggest challenge was to make sure that the participants understood that we could not help in advancing their process and that research was a separate domain from their immigration processes. The main reason for turning down interested participants was a sense that the person did not fully understand our role. Our work on informed consent was, more importantly, to explain what research is rather than (only) giving out the complicated consent and information forms. Nonetheless, we did hand out translated forms, yet I suspect they were not of interest to most of the participants, some of whom did not know how to read or write in any language.
Although it was important to inform participants about not being able to help in their administrative process, this is not to say that we did nothing to help. Engaging in necessary help was an ethical issue as well. Recent scholarship on research with refugees stresses how, in addition to the important "do no harm" principle, it is essential to give back to the community and not just "steal stories" (see Krause, 2017;Pittaway, Linda, & Richard, 2010). As I will explain later in the analysis, often the forced migrants' experiences were affectively so intense that they in fact moved us to act. Together with the research assistants, we wrote newspaper articles and blogs, spoke at different practitioners' (social workers, teachers, policy makers) events and also engaged in quite a bit of supportive work after the interviews. Many participants lacked proper legal or social help, so we guided them to the relevant services and kept in contact with most for many months.
It would be the subject of a whole different article to think of all the layers of informed consent in this particular study. Perhaps I would crystallize the ethical legitimacy of this study by referring to recent critical scholarship on vulnerabilities, which stresses agency even in dire circumstances (see Virokangas, Liuski & Kuronen, 2020). The research assistants and I felt that it was important to respect the agency of the adults who engaged in our study. We did not feel it was our place to deny them participation even though many people were in distress-given that the limits for our possibilities for action were understood and we felt that the participants wanted to speak with us.
All of these issues, from how we were able to find participants to how we conceived of our roles in the research and how the interviewees understood the situations and positionalities within, were part of the specific assemblage that emerged in the research encounter. The concept of assemblage derives from Deleuzean theory and refers to a "wide variety of wholes constructed from heterogeneous parts. Entities ranging from atoms and molecules to biological organisms, species and ecosystems may be usefully treated as assemblages and therefore as entities that are products of historical processes" (see DeLanda, 2006, p. 8). There are two important distinctions I wish to make regarding the relation of assemblage to some constructivist concepts such as "discourse" or "frame", which also describe a sort of coming together of ideas. First, assemblage cannot be predetermined, as a frame or discourse would be. Assemblage is always emerging or becoming. Second, this becoming is a result of all human and non-human forces at play. Consequently, an assemblage is as much evolving in the moment as it is determined by past moments.
The categories formed in this analysis could also be called method assemblages. Law (2004, pp. 41-42) defines method assemblage as an idea with "no fixed formula or general rules for determining good and bad bundles, and that (what I will now call) 'method assemblage' grows out of but also creates its hinterlands which shift in shape as well as being largely tacit, unclear and impure" (Law, 2004, p. 42). In this article, the two assemblages described are emerging from the layers of former research encounters, practices of recruiting participants, events earlier on a particular day when the encounter takes place (whether one is rushed to the situation or well prepared, nervous or calm), and affects circulating in the situation influenced by all research participants as well as transnational linkages to people and things, which also emerge in the situation.
It is perhaps significant to note that when I started to listen to the six recordings again, I had no preconceived ideas about how my listening would shape the analysis. I only wanted to revisit the spaces that I remembered as being the most intense, in the hopes of gaining analytical insights. The process of writing this text and the listening coincided, and it is hard or even impossible to separate the resulting analysis into different types of data. Lahti (2018) also explored listening to old interviews as a methodological starting point. She found that having her research participants listen to interviews from years back provided a fruitful starting point for affective analysis. But here, I find that my methodological exploration was more of an attuning to affect (and re-remembering) than an actual method. I wanted to involve myself (again) with the material as an exercise in knowing differently. St. Pierre (1997, p. 183), for instance, notes that "knowing that is mapped beyond the mind/body trap produces lines of flight that remain uncoded". My visitation back to the situation was an attempt to cross some of the strict lines of how empirical analysis is or may be done.

It is not your fault
I began the listening "exercise" by listening to the interview I remembered as being the most difficult.
Here I concentrate on one aspect: a thought of the sentence "it is not your fault" (which is also a radical feminist notion about blaming oneself in connection to domestic violence). This thought was central in my mind during most of the interviews I conducted in this project and, as I will show through the analysis of one particularly affective interview situation, the thought materialized important aspects of one affective assemblage. This assemblage was, of course, not exactly the same in any of the interviews, but by concentrating on "it is not your fault", I am able to show how interviews with severely traumatized people are also felt and embodied by the researcher. This embodiment is discussed as a methodological and ethical issue. To describe the particular situation in question, or the scene, in the terminology of Coleman and Ringrose (2013), I will first describe the context of the situation, as the context closely relates to why the particular scene incited intense affect.
The scene involves two young women, aged 21 and 18 years, both of whom arrived in Finland as unaccompanied minors during 2015. In the interview they tell me about their lives and their separation from their mother. Their story is one with a continuum of extreme violence. To protect their identities, I will not even share their country of origin, as it is not essential for this analysis. I call the women Fatemeh (older sister) and Sara (younger sister).
Fatemeh and Sara's story begins with their mother being sold off to another man as a payment for their father's debts. Their father, a drug-addicted gambler, is extremely violent, the women tell me. The whole family lives as undocumented migrants in a country which is not their country of citizenship. They have no rights and no possibilities for getting help from any official sources. Fatemeh and Sara are sold for sex by their father when they are still young children, and they are raped, beaten and burned. Then they manage to escape. They turn to extended family members for help and travel to Europe with a smuggler. The women end up in Finland and eventually receive news of their mother's whereabouts after years of being apart.
When the women find their mother, her horrific life story is revealed. She has three small children and has escaped from the man to whom she was initially sold. In an effort to also get to Europe, she is exploited by sex traffickers and abused in captivity for-as I gathered from the quite fragmented interview-at least a year. She manages to get away from the traffickers, but again ends up in an abusive situation with the man who helped her escape. When she finds this man sexually abusing her young daughter, she again escapes. At this point in the story, the mother is already in contact with Fatemeh and Sara. They finally help her into a safe house for abused women in the country where she resides. This is where the mother is at the time of the interview.
The continuum of sexual violence does not end there. After coming to Finland, Fatemeh turns 18 and is transferred to an adult reception centre, while Sara is still at the youth group home, as she is a minor. Fatemeh tells us she was raped at the reception centre. Later, when she travels to meet their mother in the country where she is residing at a safe house, she escapes several attempted rapes but still experiences severe sexual abuse. Finally, when Fatemeh comes back to Finland, she is raped by a translator who had promised to help the sisters reunite with their mother. This crime is under investigation by the police at the time of the interview, as it is the only rape that Fatemeh has told her social worker about.
By the time of this interview, my research assistant, Zeinab, and I had already done many interviews in which we heard stories of violence, torture and rape. However, this particular story affected us more strongly than any other. After the women told their story and described the last rape in the long sequence of violence, I said in a faltering voice (J stands for Johanna, Z for Zeinab): J: I just want to. I know you know this, but this is not your fault in any way. He is a very bad man. Z (translates): This is not your fault. He was a very bad man. This is not your fault. J: I think this is something that you cannot say too many times, you just have to hear. (switching to Finnish) Se ei oo teidän syy. Eikä sinun syy. Eikä teidän äidin syy. Se mitä on tapahtunu.
In the quote above, some traces of the affect-fused atmosphere are visible in the transcription of the recording. When re-listening to the recording, I remembered how, as the story went on, I had a desperate need to hold and console the young women. It is not your fault-I wanted make that absolutely clear to the women, to hold and to rock them and to make them experience it. On the tape, I hear their very young voices talking about these horrific abuses and myself trying to mould my voice to sound as compassionate as I possibly can. When Fatemeh says, "I don't know why these things all happen only to us all the time", I struggle to find words. When re-listening, I again sense how this particular moment is defining. I have to answer in a way that does not deepen the women's sense of guilt or shame. It is not a long enough moment to have analytical thoughts; it is an instinctual affective space.
While I struggle with making "it is not your fault" an embodied message (rather an impossible task, but that is the experience and memory of the situation), my body reacts. I start to sweat. My throat clogs up. These bodily symptoms are intertwined with the affective assemblage that this situation reflects. The memory of the bodily discomfort is vivid when listening to the tape. Now I see it as a manifestation of the intense affect that was part of the assemblage.
This moment in the interview when I start to repeat "it's not your fault" is also a point when our rhythm changes in the interview. I lose my composure after Fatemeh's remark, "I don't know why these things all happen only to us all the time". I even switch to Finnish, as if my native language would further assure the women-who only understand very little Finnish-that they are not to blame. I hear the change from the tape and remember the moment.
The affective assemblage in the situation contains a complex web of relational ties. First, there are the young women, Sara and Fatemeh, who are desperately trying to find a way to reunite with their long-lost mother. They seem extremely anxious and even unable to act. However, with a closer look at the situation, Sara and Fetemeh did manage to incite a strong sense of urgency in myself and in the research assistant, Zeinab. As a result, we ended up helping a great deal in their actual reunification process. So, it could be interpreted that at this moment the women were also inspiring our action. They brought their mother, younger sisters and generational experiences of sexual abuse to the affective assemblage. Later, in our reflection about the interview situation with Zeinab, she described feeling utterly ashamed of representing a nationality of people who collectively oppressed the nationality the women represented. Zeinab said that she was shocked by the women's story, as it came so close to home. She knew people who had been in similarly precarious situations, but she had never understood the full precarity and vulnerability of their lives. Rather, perhaps (as it is my interpretation), she had never felt the precarity before this particular interview. We could say, then, that the people Zeinab knew also joined the interview. Third, there was me, the white middle-class researcher in a position of power in relation to all the other participants, but certainly with her own share of vulnerabilities, needs and wounds. It should be noted that I have engaged in feminist activism for decades and have also undergone training in supporting abuse victims. I came to the situation equipped with many types of knowledge. It is not your fault, I knew to say.
To unpack the affective assemblage, I turn to feminist discussions on empathy. In order to get to that discussion, let me begin by introducing how Elizabeth St. Pierre used Deleuze's concept of the fold to analyse the difficulty in separating herself from her research participants (St. Pierre, 1997). In her reading, the concept of the fold disrupts our notion of interiority because it defines "the inside as the operation of the outside" (Deleuze, 1986(Deleuze, /1988. For St. Pierre, her "subjectivity had no inside or outside" and was thus "like a fold" (St. Pierre, 1997, p. 178). This notion captures what I also experienced in the interviews, where "it's not your fault" was the main message. It was a deep dive into an empathetic state in which it was difficult (or perhaps impossible) to separate my own subjectivity from the other participants present. There were several interviews where the fold emerged at some point of the interview. This experience was a strong catalyst for myself and the three research assistants with whom I worked to start "working much harder to understand my participants, to respect their lives, to examine my relationship with them, and to question my interpretations. The examination of one's own frailty surely makes one more careful about the inscription of others" (St. Pierre, 1997, p. 181). These interpretations are close to many feminist discussions on empathy, which claim that empathy may have transformative potential (e.g. Pedwell, 2012, p. 166). Ahmed (2014) points out how a particular atmosphere (a "feeling in the room"), may be intense, but it is never felt identically by the different people present in the situation. Ahmed continues that "these feelings not only heighten tension, they are also in tension. Emotions in their very intensity involve miscommunication, such that even when we feel we have the same feeling, we don't necessarily have the same relationship to the feeling" (Ahmed, 2014, pp. 10-11). I think this is an important aspect to recognize when speaking about empathy or when analysing the experience of sharing an intense emotional space. Unlike the model of "emotional contagion" suggests, Ahmed goes on to say that "it is the objects of emotion that circulate, rather than emotion as such" (Ahmed, 2014, p. 11). In the particular affective assemblage of "it's not your fault", we all shared the space of intensity, but for very different reasons and, consequently, with a different content of emotion. In the next section, I will describe a very different affective assemblage where instead of slipping into a "fold", the assemblage emerged around difference.

How much money do you make?
In her paper "Differences Disturbing Identity", Elizabeth Grosz (2010) strongly states that instead of equalizing differences, "we need to address the question of whether a plurality of subject positions can be adequately accommodated by the ideals represented by the able-bodied, white, middle-class Eurocentric male heterosexual subject!" (Grosz, 2010, p. 108, exclamation mark in the original). She goes on to suggest that difference means that there cannot be one aim or goal for all. In fact, Grosz sees difference as the only thing that makes us "recognizable perhaps for a moment in our path of becoming and self-overcoming". I wish to attend to this notion (of total and never-fading difference) for a while, as many feminist discussions on empathy presuppose that empathetic experience may inform another subject about another's distress and oppression (e.g. Alexander & Mohanty, 1997). In this line of thought, empathy would, in a way, transcend difference. However, as Hemmings (2012) also points out, the definition of transformative empathy leaves out the encounters where antagonism arises from inequality and is not calmly resolved, but instead carries a possibility of "hostility as a genuine affective condition of some encounters" (Hemmings, 2012, p. 153). I will explore this issue in this section, where the affective assemblage is condensed into a question in the beginning of the interview: "how much money do you make?" I will call the interviewee in this example Ras. He was a man from Iraq, whose wife and three children were in a refugee camp waiting for family reunification. Ras's application had failed because of not having sufficient income in Finland. His status for receiving international protection was "subsidiary protection", a category from which the right to apply for family reunification without a high income was removed in 2016 (see Hiitola, 2019).
First of all, this interview started off as stressful. I had scheduled two interviews, one after another, not leaving enough time between them. My research assistant, Ahmed, and I ended up being terribly late for the interview with Ras. As we rushed to meet him, I could already sense from afar that he was very tense. Maybe it was the way in which he was smoking his cigarette, or his bodily posture, that made me feel uncomfortable. Later, I pondered Ahmed's (2000, p. 8) example: "when we face others, we seek to recognize who they are, by reading the signs of their body as a sign". Ahmed goes on to argue that such "acts of reading constitute the 'subject' in relation to 'the stranger', who is recognized as 'out of place' in a given place". Ahmed's important point in her book Strange Encounters is that a stranger is something recognizable, and we cannot identify a stranger unless we already know (from past encounters with humans and things) what a stranger would entail. I therefore wonder whether a cultural imaginary (of danger) was participating in the unfolding of the encounter by encouraging particular kinds of interpretations made by both parties and intensifying the affective atmosphere of tension.
As we walked towards the interview place together, Ras talked only with the research assistant, Ahmed, in Arabic, and I struggled to engage with them. I did not know how to settle into the moment, and it was certainly also due to the rush that the previous (also intense) interview had ended with. While walking, I tried to engage with Ras by smiling and nodding, but we had no common language. As we approached the place where the interview was to take place, Ahmed translated to me Ras's question (and first engagement with me): "How much money do you make doing this job?" I told him the amount, feeling even more unhinged. Ras replied, "See, even you couldn't get my family to Finland. How could I?" The first encounter framed the assemblage. I interpreted that the tension in the situation was stemming from Ras's struggle with different violent administrative barriers, but that did not help with my "gut reaction" of feeling uncomfortable. This affect surfaced again when re-listening to the interviews. I believe that Hemmings's (2012) concept of affective dissonance is useful here. She writes about an example of a woman who cleans the houses of others. To get to discussing or realizing the inherent inequality of the situation (of the woman not being able to tend to her own children, of the power dynamics present and the labour not appreciated), "someone has to look at those conditions and provide a counter narrative motivated by lack of acceptance" (Hemmings, 2012, p. 157). This counter narrative has effects: "anger, frustration or even rage", Hemmings suggests. Ras's bodily gestures signalled anger, and yes, I would say that I experienced "affective dissonance" of sorts. Here was Ras, not attuning to a position of a victim of family separation, but instead encountering me with a question that revealed all the troubling dynamics at play. I was made aware that I was already well-off in comparison and, furthermore, aiming to benefit my research career with Ras's story (my own addition to the assemblage). "How much money do you make" resulted in the awareness of my many privileges in terms of class, race and secure citizenship.
Once we got to the actual interview situation, I also made a mistake. I was assuming that Ras had paid for his family's reunification process. My question invited an affective response, where Ras strongly asserted that he had no possibility of paying the high fees. My reading of the situation is that he interpreted my question as criticism (J stands for Johanna; the research assistant's speech is not included).
J: Can you first briefly say when you came to Finland? And how long have you been here?
Ras: I arrived in Finland in 2015. It means it has been over three years since I arrived.
It's been a year and eight months since I have applied for family reunification. My family was interviewed, and they even paid the fees. Listening to the interview again brings me back to the constant feeling of failing to connect with Ras. I asked several questions, many of which were not put well, such as the example above. Despite my efforts in empathy, I interpreted him becoming more and more annoyed during the interview. I stayed in contact with Ras for many months after the interview. We called a couple of times, and later Ras would send me a message every morning saying "Huomenta [good morning]", a stark contrast to how I experienced his presence in the interview. Then Ras stopped sending me messages. I never enquired why. My silence was due to several pressures at work, I told myself. But perhaps the reasons also had something to do with the initial uneasy atmosphere, since I did stay in contact with many other interviewees and and would enquire how they were doing if I all of a sudden did not hear from them.
In late 2019, a year after our interview, I asked Ras if he would be interested in a second interview. We still had research funding left, and I wanted to do second interviews to see how the research participants' cases and life situations had developed. Ras did not answer my messages. I then asked my research assistant, Ahmed, to call Ras, as we did not have a shared language. I found out from Ahmed that Ras was very upset with me. First, I had stopped being in contact with him, as if I did not care. Then I asked him for a second interview, "not even saying 'hi' first", Ras had stated. I felt like I was put in my place, but also that I deserved this feedback. Now, when I reflect on my research process, it has in fact been the affective assemblages of tension (and perhaps antagonism) that have offered me the most valuable information as a feminist researcher. The people who are only sustaining their lives by a thread, but who are also very angry about the injustices they experience, are very seldomly heard. They are not deprived of agency, as is shown in Ras' example of inciting affective dissonance in me. However, often their anger is not the kind that would make a difference, unlike that of an activist who knows the language of oppression.
When thinking about Ras's agency, Ahmed's (2000, p. 55) analysis of the "skin" as a history of encounters may be relevant. She writes that "the skin is touched differently by different others" and says we need analysis of its exposure and touchability. While both of the affective assemblages I have presented in the analysis could be conceived of as piercing the border of the skin (my skin, their skin), I believe Ras's response to me was a stronger entering into my own skin. The difference between us could not be felt away through empathy.

Conclusions
In the analysis, I have identified two affective assemblages that emerged in my research encounters with forced migrants experiencing family separation. I point out how intimate relations may manifest through a "fold", which is a shared, intense affective assemblage where the borders of the researcher and participant become blurred. Difficult experiences of losses and injustices in the realm of intimacies may also cause affective dissonance. In these research encounters, difference between the researcher(s) and the participant(s) is left unsolved.
In the first assemblage, the "Other" is touched, but the interpretation of the touch is in the possession of the touched, not the one touching. There is an understanding of sorts in this emerging encounter, much like Ahmed (2000, p. 48) writes, "the giving between two bodies does not lead to two bodies which are positioned in the same way, but it does lead to a new form of inclusivity, in which what is shared is the very impossibility of having a body that belongs to me or you". In connection to this inclusivity, feminist scholarship on empathy as a transformative force is relevant (see Alexander & Mohanty, 1997;Pedwell, 2012). The described encounter in the first affective assemblage incited action in myself and the research assistant. We worked hard(er) to influence policies as well as the situation of the interviewed women.
The second assemblage was transformative in a different way. The encounter did not reveal the knowledge of what the innermost thoughts or feelings of the research participant were. There was a constant lingering feeling of difference present. Hemmings (2012) notes that the difficulty with the feminist discussion on empathy concerns what she considers "its embedded assumptions about reciprocity", and that "good empathy" (one which is not a lazy attempt to take the other's place) would always be appreciated. However, there are instances where the "other refuses the terms of empathetic recognition". What then may emerge from the encounter? Ras's example shows that difference that denies empathy compels uncomfortable (but necessary) power relations between the researcher and participants to become visible. Power imbalances are also at play in whose wounds are recognizable and whose are not. It was easier for me to relate to the young women's pain. For example, I have easily seen how to make it visible for others through academic writing. This again reveals another power dynamic, like Ahmed points out. Pain may fix the "other as the one who 'has' pain, and who can overcome that pain only when the Western subject feels moved enough to give" (Ahmed, 2014, p. 22).
The analysis in this article also revealed challenges embedded in research concerning forced migrants' intimate relationships. Since affect tends to travel temporally and spatially (see Dragojlovic, 2015), the research encounter often includes many human and non-human elements that contribute to the high intensity of the atmosphere. First, the transnational family members often "join" us in the interviews, as the affect was so intense and the descriptions vivid. Wise and Velayutham (2017) use the concept of transnational affect to describe this "travel" of intimate affects across borders. They see that "affects and emotions are embedded in and flow with remittances, objects and gifts, letters, phone calls, and text messages, return visits, and transnational symbolic fields". In addition, these affects may manifest through different embodied affects which can emerge through, for example, "slapping the newspaper or spitting on a floor" while reading news about upsetting events, such as amplifying conflicts in the country of origin (Wise & Valayutham, 2017, p. 177, 122).
Second, as the difficult encounter with Ras shows, it would be extremely important to conduct interviews with forced migrants without being rushed or otherwise stressed by external circumstances or tight schedules. However, I found during the research project that this task was often very difficult. Many interviewees were in stressful life situations and thus often very late for interviews. Materialities included set times booked for interview rooms, which did not allow for flexibility. When interviews were conducted in research participants' homes, I often suffered from very hot room temperatures during the wintertime, when I was warmly dressed. Finally, cultural imaginaries may influence what emerges in or from the research encounters and intensify the atmosphere of tension. It would be important to continue studying difficult research encounters to also develop ways of dismantling prejudice and possible "fear of the other", which influence our preconceptions. It is also imperative that we engage in critical reflection about empathy and avoid easy explanations of being able to transcend (all) differences.
Finally, I must also admit that an ethnographic data collection method might be more suitable and flexible for research that includes very high affective intensity. Consequently, this project also ended up resembling ethnographic research in the form of engagement and long-term relationships-even friendships-with the participants. The elements that form the particular method assemblage (see Law, 2004) in the research situation may be overly intense when research encounters concern traumatized research participants who are speaking about intimate affects, such as their family and close relationships.