Comprehensive or Comprehensible Experience? A Case Study of Religion and Traumatic Bereavement

The first half of this article provides a brief overview of two respective projects concerning traumatic bereavement, in which religious faith appeared to feature amid a constellation of significant coping and sense-making mechanisms for survivors. After presenting some illustrative examples of the kind of data produced in the course of our research, the second half of the article develops a retrospectively critical appraisal of our data collection and corresponding analysis practices. In questioning the extent to which our accounts of our participants’ accounts can be considered adequate representations of social order, we critically explore the relative potential of ‘reflexivity’ for bridging the experiential gap between researchers and participants. Taken together, these reflections prompt a return to the salutary question: what counts as sociologically ‘see-able’?


Introduction
Researching violence, bereavement, and mourning inevitably leads us to recognisable, yet unfamiliar stories. We are forced to negotiate moments during our fieldwork which confront our own life worlds. Some stories we hear may even seem ubiquitous, yet, if anything, serve to highlight the ontological chasms that exist between us and our participants. Religion and faith, as qualitative features of social narratives, may stand as one exemplar (among many) of this much broader 'problem' of interpretation. Based on retrospective reflection on past fieldwork and interview data, this article grapples with our shared concern that experiential reality for 'believers' and their intrinsic ways of being in the world are too often abstracted into familiar categories of analysis.
Interactions between researchers and participants with either no 'shared faith' in a particular religious deity or spiritual doctrine or, conversely, no 'faith shared' at all by one side in any religious or spiritual worldview are common and deserve our attention. At best, under these circumstances, we might attempt to convey stories of religious experience as they were conveyed to us without necessarily understanding or reconciling them with our own beliefs. At worst, we might dilute them within our analyses or perhaps even abandon them altogether, truly in bad faith from an ethical standpoint, for fear of 'getting it wrong' or misinterpreting and misrepresenting experiences divergent from our own. Yet despite these apparent difficulties, stories told between members of one religion and another or, in our case, between participants with a religious or spiritual conviction and researchers without one, are rendered comprehensible (i.e. understandable).
This article aims to explore how we might navigate the unfamiliar and how we might engage with material that we do not necessarily identify with on a personal, spiritual, or ideological level. It contains five main sections, which are structured as follows.
The first gives a brief overview of two studies that we conducted, respectively, and from which the interview data we consider here and present later in the article derives. Our shared curiosity about the contingencies and dynamics of qualitative research, including the methodological questions it provokes, led to a series of conversations concerning rapport with participants, conversation topics covered during interviews, and data analysis processes. One problem, or interest, each of us returned to was how to deal with and situate unfamiliar (non)knowledge within our fieldwork practices and interview data after the event. By retrospective virtue of such questions, we hope our puzzlement can prove constructive, if not exactly resolvable, epistemically speaking.
The second section presents a series of extracts from our data to simply illustrate the kinds of experiences prompting these thoughts. Despite their fascinating content and conventional 'feel' as segments of narrative, our inability to fully identify with participants' stories only heightened a perceived tension between faithful data collection and subsequent representation.
The third section explores the fact that, despite this perceived tension, religious narratives were nonetheless comprehensible and decipherable products of our research interactions. Here, we critically ask whether the brief extracts and analyses given in section 'Background to our respective studies' represent adequate accounts of our participants' experiences.
The fourth section considers how practices of reflexivity, positionality, and self-critique have become one stock methodological answer to this question. Reflexivity, as Lynch (2000) argues, is often hailed as a 'methodological virtue and source of superior insight, perspicacity or awareness ' (p. 26). While its proponents have pointed to its emancipatory and critical potential, we argue that the practice of reflexivity is in fact ordinary and mundane; it does not, in this case, carry inherent methodological value alone for grappling with faith and religious worldviews, or for wholly reconstructing participants' partial or unchallenged accounts.
In light of these arguments, the final section considers alternative research design strategies we might employ to resist methodological tendencies that create detached, 'expert' knowledge only of epistemic value to social scientific communities. One way in which we encourage readers to question unfamiliar narratives and seek further clarity in their research is through ongoing critical engagement with the 'do no harm' principle typically followed to satisfy institutional ethical approval. The article concludes by reflecting on the gap between comprehension and comprehensibility in social research, returning to the salutary question: what counts as sociologically 'see-able'?

Background to our respective studies
Making sense of material that we do not necessarily identify with on a personal, spiritual, or ideological level is, of course, a task potentially faced by all researchers. Our focus on religion stems from similar experiences we shared during our respective doctoral research projects which were completed in 2018. William McGowan's research centred on bereaved families and survivors of political violence and terror attacks who are engaged in storytelling and dialogue workshops at The Tim Parry Johnathan Ball Foundation for Peace, a non-governmental organisation (NGO) based in Warrington, England. Elizabeth Cook focused on the experiences of anti-violence campaigners affiliated with an NGO called Mothers Against Violence, who have been affected primarily by violent crime in Manchester, among other cities in the UK. Both projects explored traumatic bereavement utilising participant observation and in-depth interviews as primary data collection methods. Each project was approved by Ethics Committees at our respective institutions and informed, written consent was provided by all participants; all respondents' names used here are pseudonyms.
Our reflections on interview topics, rapport with participants, and data analysis processes led to recurrent conversations about how closely we felt we had been able to understand and represent our participants' experiences in our finished PhDs. Several shared themes naturally emerged across our projects, but it was the way in which some participants framed their experiences of, and responses to, violence in religious or spiritual ways that intrigued us. Neither of us shared a religious conviction with our participants, nor did we want to ignore or overlook the significance of their stories. We recalled instances where participants would speak at-length about religious scripture, or transcendental experiences, to which we could only listen and nod, simultaneously feeling pangs of inadequacy and an all-too-familiar sense of 'imposter syndrome' (Breeze, 2018).
Our data were collected and eventually written-up relatively 'unproblematically', but we harboured frustrations at our lack of shared knowledge with interviewees. So why should it be then, sociologically speaking, that we negotiated these research processes with simultaneous ease and difficulty? In making sense of this question, we are not occupied with religion exclusively, but also with questions of ontology, epistemology, ethics, and reflexivity more broadly.
Religion, of course, continues to occupy a central space within sociology. While it is beyond our scope to offer a protracted discussion of this classic and contemporary history (though see, inter alia, Horii, 2018;Smith et al., 2013;Spickard, 2017), suffice it to say that debates about religion vis-a-vis knowledge, objectivity, science, tradition, and belief have not abated in the 21 century. According to some histories, science and technologies associated with modernity may have undermined traditional religious practice and belief, creating an erroneous intellectual dichotomy between science as rational, modern, and reasoned and religion as 'parochial and unscientific' (Spalek and Imtoual, 2008: 2; see also Thomas, 1971). Other analyses reveal the intriguing debt owed to presociological theologies by contemporary social sciences (Burdziej, 2014). Conventional approaches to the 'religious/secular' dichotomy within sociologies of religion, particularly within social constructionist methodologies, have often favoured bracketing questions of the sacred when dealing with religious experience empirically -this forming the basis of Peter Berger's widely cited 'methodological atheism' (Cantrell, 2016;Porpora, 2006). Despite not sharing a religious conviction with our participants, we found this familiar dichotomy to be unhelpful and reductive and, consequently, this style of bracketing inadequate and superficial.
Following Porpora (2006), who offers a methodological critique of Berger's work on religion, subsequent conversations and retrospective (re)analysis of our data have been partly shaped by the alterative notion of 'methodological agnosticism'. While Berger's work is not discredited wholesale by Porpora, it is scrutinised for setting aside supernatural, transcendental, or sacred experiences in favour of a 'value free' social (i.e. not psychic) science. A Christian Lutheran himself, Berger's writings on religion were both illuminating and changing, as he adapted his approach to empirical study and knowledge for both sociological and theological audiences throughout his career (Cantrell, 2016;Porpora, 2006). An alternative methodological agnosticism would dictate that experiences pointing to religious or spiritual beliefs and attitudes, seemingly invisible to the researcher, 'must at least be explored rather than set aside' (Porpora, 2006: 70). Examples of this principle are not confined to the quiet, modest, and arguably individualistic introspection alluded to later in this article, but are found in both Elisha (2011) and Luhrmann's (2012) studies of public and communal evangelism (for an excellent discussion of these works and their ontological treatment of God see also Bialecki, 2014). This sentiment seemed to capture our desire of hearing our participants (as opposed to simply listening to them), and our simultaneous lack of understanding and occasional reticence towards their religious narratives.
Consequently, we believe that there is value in revisiting and critically reflecting on our data and fieldwork experiences by flagging up their methodological deficiencies. One intention is to highlight those ostensibly direct empirical examples of 'non-knowledge' that we encounter during our fieldwork. As we will try to convey in doing so, the 'problem' of religion for social research or of not sharing religious experiences with participants is, in many respects, not at all a problem as such, at least not in the fundamental ways described above. What both authors experienced during their respective fieldwork encounters were narratives that presented themselves in simultaneously recognisable and yet inaccessible ways.

Narrative extracts about religious experience
Critically reflecting on the suitability of our data collection and analyses completed at the time, the following three subsections illustrate some of the kinds of experiences described to us. As we argue later in the article, these extracts raise a series of methodological questions, both inclusive and independent of religiosity. This level admittance of both the natural (i.e. observable) and supernatural, neither asserting nor precluding either (Porpora, 2006: 58), represents the first stage of the methodological agnosticism described above, opening up an ample space in which to critique our accounts of our participants' accounts later in the article.

June
June, an anti-violence campaigner, shared the story of the loss of her son who was shot dead during a wave of inner-city gun violence. She described how faith, belief, and religion had come to inform her reading of this tragic event. Speaking about the night that her son was shot, June described hearing what she believes now to be 'God's voice' urging her not to let her son leave the house that evening. Expressing her role in the aftermath of her son's death in terms such as 'calling, 'sacrifice', and 'fate', she reflected that 'by the time my son was killed I was ready for what God had called me to do'.
While religion featured centrally within June's narrative, there were periodic, implicit reminders of the gap that existed between our assumed worldviews, as the following passage suggests: June: He never came back, yeah? And every time I think about that, I think, you know something June, if you had obeyed, your son would have been still alive. But again, because the Bible tells me all things are working together for my good . . . So, whether I obey or not obey God is still . . . because what he's saying is I will work it for your good. It might look not so good, but he said I will work it for your good. He said all things work together for those who love the Lord and are called accordant to His purpose.
June spoke for nearly 4 hours about these experiences, habitually referring to religious passages, readings, and personal meditations that she had learned over her lifetime. She was, of course, not alone in orienting conversations about traumatic bereavement around her faith. Our attempts to make sense of the 'data' that we had amassed, however, were also influenced by feelings of confusion and uncertainty over how exactly to understand these unfamiliar meanings, passages, and stories from scripture.
A secular 'reading' of June's story might, for example, see this as a tale of coping and recovery, where the role of faith serves as a defence mechanism against the extreme distress of violence. Indeed, it might serve such functions, providing a narrative of comfort and solace at a time of traumatic upheaval and disruption or, for others, represent a moment of crisis which confronts faith and confirmation.
Nevertheless, reducing religion and faith to a set of more familiar (for the researchers) psychological and emotional phenomena seems speculative and abstract. We note as an aside that much criminological research around violence, for example, often displays a tendency to see the importance of religion only in how it manifests in preventive factors, desistance from offending, and as an instrumental coping mechanism in the aftermath of violence (see Adamczyk et al., 2017). Does this risk rationalising religion in aid of accepted (secular) standards of analytical procedure and knowledge practices? The language of 'calling', 'vessel', and 'sacrifice' employed by June and others to describe their experiences sometimes created confusion and ambiguity over how to faithfully and respectfully represent unfamiliar experiences, though we would soon come to ask a seemingly more basic methodological question: how far were we really able to represent any experiences described to us? This question seemed compounded, given the religious tenet of June's narrative and those that follow, our lack of shared faith, and the modern, secular roots of so many of our sociological canons.
Practically speaking, the extract above also contains an obvious interpretive ambiguity: is it life itself, insofar as this represents God's plan, that 'might look not so good' (to June, the believer, whose son has lost his life as part of this higher plan)? Or is June, in fact, performing quite a deliberate awareness-turn; namely that the idea of someone believing in such a seemingly unjust system of monotheistic reasoning 'might look not so good' (to the researcher, the assumed non-believer -this being one possible assumption June may have held of Researcher B, itself riddled with more complexity than such an assumption would likely reconcile in the moment)? Without seeking clarification at the time for fear of interrupting or disrupting June's interview 'flow', such subtle, ordinary language turns can be difficult to retrospectively reconstruct -an obvious shortcoming we explore further below.

Lynn
Perhaps similarly difficult to comprehend is Lynn's account. Lynn's husband Jim was killed in Northern Ireland during a bombing campaign in the early 1990s. For several years after his death Lynn would reflect on how merciless her husband's assailants had been, 'meticulously planning every minute, from occupation of this house to the minute the bomb exploded. I just couldn't understand how human beings could do that' (Lynn).
We reproduce part of her account here to show how she came to experience transcendental connections with her husband several years after his murder. While the visions and experiences she recounts may be hard to grasp from a secular, worldly viewpoint, her story was compelling, coherent, and moving -immanently comprehensible (understandable), if not entirely comprehensive (complete): Lynn: I firmly believe in a life hereafter. I was always a bit sceptical [laughs] about spirits and ghosts and all that sort of stuff but now I fully believe in a presence because Jim has come back to me a few times. Now, he hasn't come back to me now for the past two years. [. . .] I was having horrific nightmares about Jim, er, being all stuck together wrong . . . you know, foot sticking out of here and his head down there and . . . it was horrific, it really was horrific. One night I went to bed and I thought I can't go to sleep because I'm going to have these nightmares again. I propped myself on my pillows and I took out my book and my glasses and I looked up and Jim was standing by the bedroom door and I thought . . . 'aye right' [expressing disbelief], so put my head down in my book again and I looked up again and Jim was still there. It was him just standing there and he had on the grey cardigan that he had on him when they [his perpetrators] took him away. He just looked at me and he held out his arms and he says, he always called me 'girl', he says 'look at me girl, I'm ok now you go to sleep' so I thought to myself 'oh, right ok' and I have never had a nightmare again, never! I was at church that Sunday evening and they were doing a healing service and in the healing service the priest was standing up in the pulpit and he was preaching about on the last day we will all arise perfect and I thought there you go. That's Jim telling me again that he's alright. He's alright where he is and that I'm to continue on in my life and do the best I can and that, that was one of the good times when he appeared to me.
Lynn's references to Jim reappearing after his death punctuated our interview. She did not speak of these interactions in overtly religious terms, but in matter-of-fact, colloquial ways, indicating, it seemed, no required input or understanding from the interviewer. The stories that she shared exhibited quite ordinary, conventional characteristics, even though their content described out-of-the-ordinary experience (Sacks, 1992: 215-221). This is difficult to decipher, however, when we 'read' the data without sufficient context: Lynn: I was lying in bed reading and this wall opposite my bed lit up. I put the book down and the wall was like a meadow full of flowers and they were all pastel coloured, lilacs and pinks and pale yellows. There was a path coming down, winding down the meadow and this figure was moving down the path . . . as it got to the bottom of the path I realised it was Jim and he had his arms out like that [gestures]. I went to get up and he waved me to sit down, you know, as much as saying 'no, no you're alright' and he turned and he started walking back up the path again and he waved. I could see his back and I'm waving like that there and I thought I'm never gonna see Jim again.
Visions reportedly experienced by Lynn certainly exhibited worldly and practical consequences, including the very conversation we were having. The above extract is taken from a section of our interview in which Lynn described forging a relationship with her new partner. Taken as a sign that she was now safe and cared for in a loving relationship, the story about Jim's final appearance was used as one way of describing this newly formed relationship. Lynn implied that she felt conflicted about starting a new relationship, speaking at-length about how she and her new partner lived separately, found it important to have their own spaces, and that she would never see herself as 'Jim's widow' but always as 'Jim's wife'. She discussed this earlier in our interview, but when it came to her experiences of seeing Jim, she once again oriented her discussion to this new relationship: Do you know my interpretation of that [seeing Jim walking away for the final time]? I have now met someone who will not shit me around or do anything to hurt me, who will look after me, so he [Jim] doesn't need to look after me anymore.

Anne and Kevin
Anne and Kevin, whose daughter was killed during the 2005 '7/7' attacks in London, explicitly described circumstances which renewed and fortified their Christian faith: Their desire to combat post-7/7 Islamophobia was also linked by Anne to her experience of forgiving the perpetrators and accused, a process which she described as the last hurdle in moving on from the attack and its debilitating effect on her well-being. This occurred during a trial several years later in which three men accused of assisting the bombers were acquitted: Anne: One of the hardest things I've ever done . . . I actually was [long pause] led to pray for them . . . and I have never done anything quite so hard in my life, it left me feeling absolutely exhausted. Whether they were guilty or not was beside the point. I could pray for them because, as it came out so clearly in Northern Ireland, there but for the grace of God go I. If we had been in those same circumstances, those same people, we would have probably, almost certainly done exactly the same thing and so to be able to see these people, who may or may not have helped people kill my daughter, in a way as victims themselves, victims of what has been done over the years to them by various sources, was very releasing to me.
In finding space to pray for the accused, seeing them as victims of structural forces, we might say that for Anne the suicide attackers responsible for 7/7 became 'transformed from being a dangerous other (not to be pitied) to being the subjects of pity' (Walklate, 2011: 189). Anne's own words elsewhere about forgiveness virtually mirror those of Agamben's (1999) on the 'lacuna' or absences within testimony. He talks about the proxy role played by survivors on behalf of those who die. They cannot speak for the dead since they themselves did not experience death, and the dead themselves cannot speak about exactly what death is like. Anne's remarks about forgiveness suggest this lack of ability to speak for those who no longer can, which again is articulated through a primarily religious lens: Anne: I don't have to forgive him [perpetrator] for killing Lauren. Lauren has to do that. That's the dreadful thing about murder, because the person you've killed can't forgive you, in this life anyway, but what I have to forgive is my pain and my loss. Now if you keep it to what has been, the trespass against us, against me, the harm that's been done to me, then it's much easier to forgive and I think that is the problem with people who cannot forgive, is they're not actually analysing what has the harm been to them.
The fact that both perpetrator and victim are dead does not detract from the paralleled ways in which Anne rationalises her decision, as a Christian, not to forgive unconditionally. This way of rationalising bereavement via faith and forgiveness necessarily means that making sense of loss, coping, and simply 'going on' (Giddens, 1991: 39) is understood, for Anne, as something entirely contingent upon 'looking' and 'moving' forwards in time, not revisiting past factual events. Simultaneously forgiving the grief felt directly as Lauren's mother, Anne's refusal to speak for Lauren discharges her of some responsibility to forgive her perpetrators unconditionally. Forgiveness, at least in a narratological sense, constitutes one way in which negative attachments to the past are transcended.
Two features of this narrative confronted us as intriguing and yet difficult to grasp. At one level, the requirement for Anne to forgive the perpetrator of her daughter's murder may not be of automatic or primary concern. It is an easy enough concept to grasp, in principle, however, if not in practice. Once accepted as a requisite response according to religious principles, however, the issue of how one should forgive and who is to be forgiven rears its head. Rather than presenting itself as a religious principle set in stone, here we see the kind of interpretive work done among survivors who navigate their understandings of forgiveness alongside unimaginable grief. It is not, in short, unconditional, nor preordained.

Are these adequate accounts of experience?
Returning to Porpora's (2006) methodological agnosticism, the 'truth status' of the above religious experiences is set aside and bracketed, but precisely for the purpose of taking it, and the context around it, seriously. While religious experience (i.e. the actual supernatural quality of the experience) is 'forever debarred' (Porpora, 2006: 58) within Bergerian methodological atheism and similar approaches to 'social constructionism' discussed above, this momentary bracketing has no problem in accepting such experiences as 'real'. But what is meant by real? The question instead becomes: are we, as social scientists, providing adequate accounts of experience? Having bracketed the supernatural as something potentially true but impossible to know, ontologically speaking for us, what does the natural look like? What natural observations can we make of this phenomenon? As we have already intimated, these extraordinary stories are articulated within ordinary social conventions.
A corollary point thus requires justification: what is meant by 'adequate', sociologically speaking? Adequacy, as a methodological benchmark, is differentially located across the discipline. Subjecting our data collection and analysis to closer self-scrutiny, we found value in subsequently thinking through ethnomethodological approaches to problems of social order. In her oft-cited introduction to Harold Garfinkel's Ethnomethodology's Program, Anne Rawls (2002) explains the methodological ideal known as 'unique adequacy' (p. 6). This ideal holds that immersion in the situation being studied, to the extent that we 'learn to be competent practitioners of whatever social phenomena' that represents, means 'unique adequacy can be assumed for most persons' (Rawls, 2002: 6). This suggests a tacit familiarity with the situation, people, problems, actions, and communications in question. However, as Rawls also points out, unique adequacy can be difficult to achieve among so-called 'specialized populations' (such as scientists or police officers) who, in reality, could be encountered through any activities beyond the immediately competent participation of researchers. Furthermore, unique adequacy might be assumed under certain circumstances but not others. Hence, due to our lack of practical familiarity with our participants' religious worlds, we can only fall back on our comprehension of their stories.
Religion or faith was articulated as a major source of support for our participants, often providing a comforting or guiding influence when almost all other areas of their lives seemed chaotic, lonely, or stressful. Religious faith was said to have been forged by experience of losing a loved one. The support structures borne from it cut across both individual, personal characteristics, and communal dynamics -both serve to mould, and are moulded by, religious conviction. However, while the retrospective character of these accounts might be illustrative in a Millsian sense (see Rawls, 2002: 6), suggestive of fascinating and moving personal accounts, the data analysis and presentation conventions we adopt above (and which are so typical of much qualitative sociology) fail to demonstrably reproduce/represent the ordinary quality of our interactions with participants.
For Sacks (1992: 218), even extraordinary, exciting, bizarre, boring, memorable, humorous, or upsetting stories are recounted within parameters of ordinary conversational conventions and are thus recognisable as such. Thus, they were simply 'ordinary stories' (Sacks, 1992: 215-221) which we simultaneously understood and yet could not faithfully relate to as secular researchers -incredibly moving and extraordinary in their content, yet assuming a plainly ordinary storied form which we could recognise as such. There was no barrier to meaningful interaction or respectful contemplation at Lynn, June, Anne, or Kevin's stories. Indeed, they simply played out as we listened and stood on their own terms, an internal coherence all their own, no matter how hard to understand or believe they may be to some.
They did, however, reveal something of the ontological space, or lacunae, between witness and listener, raising a series of important questions. How close are we really to our participants' narratives at any time? How far can we truly comprehend the experiences they relay to us? If that comprehension does not always extend very far, how much of a 'problem' does this necessarily pose for our research? Interview data such as ours remains problematic for its 'thinness'; we cannot subsequently reconstruct some innate meaning to these stories by appealing to academic strategies of reflexive practice. Conversely, our interactions with participants and the ordinary, shared, sense-making practices enacted therein constitute the very reflexive stuff of everyday social life.
Unlike, for example, standpoint feminism and participant action derivatives which often evince tacit understanding between 'researcher' and 'researched' (see, inter alia, Downes et al., 2014;Harding, 2004;Reinharz, 1992), our fieldwork can only be seen as an attempt to hear, and perhaps comprehend, fragments of our participants' lives (of course, we remain relationally located somewhere, as do all researchers of religion, see Neitz, 2013). Indeed, when these stories were presented to us, they often raised more questions than answers. These questions highlight a series of absences, both literally in our participants' lives and stories, but also by holding up a mirror to absences in our own understanding of spiritual, religious, and transcendental phenomena.
The stories relayed here and our framing of them represent what is broadly comprehensible (i.e. understandable) to us. Relatively speaking, the plots are internally coherent even for someone who does not share essential beliefs with the storyteller and, as Riessman (2008) notes, hold a rhetorical persuasiveness despite disagreements or disbeliefs that we might hold personally. On the contrary, one might reasonably ask what is potentially lost from only pursuing the comprehensible in research? Is the comprehensible enough for sociological research to aspire to? Afterall, the possible existence of tangible non-knowledge, or nichtwissen in Ulrich Beck's terms (Mythen and Walklate, 2013), in fieldwork and knowledge production represent valuable, if oblique, sources of understanding and the occasional formation of novel, even paradigmatic insights (Kuhn, 1962).
However, there is a tension. At its best, sociology offers descriptions and social explanations of empirical phenomena. Conveying comprehensible experiences is a staple pursuit of sociological research: making meaning and sense of experiences other than our own for the purpose of identifying patterns, making contributions to policy-relevant knowledge and acknowledging harms that have previously gone unheard. Without speciously imitating the natural sciences (see Latour, 2000), we remain cautious about how best to describe and represent things we cannot see or comprehend for ourselves. The question, then, throughout our research becomes: what is 'see-able'? More accurately, what qualifies as sociologically 'see-able' (Garfinkel, 2016)?
Each of the accounts considered above are selections taken from the eventual products (i.e. interview transcripts) of our participants' self-conscious agreements to sit down and articulate various aspects of their lives with us. For Rawls (2002), this relationship between people's self-conscious and reflexive actions and our attempts to document and study them poses a fundamental sociological problem: 'How is this reflexive character to be approached by the analyst in a way that is both empirically adequate and reflexively sensitive? ' (p. 48). It is to these matters that we now turn.

Confession as the sociological solution? Reflexivity and the virtuous researcher
A common strategy employed by qualitative researchers at this point is to offer a more 'reflexive' account, hopefully filling the gap of comprehension between themselves and their participants. Having declared a particular topic important, work is quickly done to ensure that the aforementioned processes of data collection and subsequent analyses 'make sense' on their own terms despite differences in positionality, worldviews, or extant knowledge between researchers and participants. Often, where data collection and analysis experience stumbling blocks as a result, 'doing reflexivity' might take the form of a declaration of weakness or mitigation concerning analysis. While there are a range of ways we might define 'reflexivity' (see Lynch, 2000), this summary treatment of methodological provisos is one commonly found within the social sciences. Our concern here is that such summary understandings of reflexivity elide the fact that as actors making practical sense of each other's lives in mutually recognisable ways, reflexivity is an 'always-already' present feature of social life (Lynch, 2000;Slack, 2000). This is to remember that reflexivity is not a panacea for transparency, truth, or authenticity within the research process, nor as Lynch (2000) puts it, does it represent a virtuous source of epistemic knowledge.
Sociologically speaking, what even 'counts' as data and those aspects of it we deem most important revolve around established and maintained social orders. For us, this manifested itself in two ways. The first is when mutual understandings of topics were forged despite an ostensible lack of commonality between us and our participants. Put simply, our inability to faithfully identify with our participants' experiences did not prevent the practical negotiation of those stories with them in ways rendered intelligible for, and by, both parties. This does not mean that we can lay claim to knowing anything of our participants' ontological experiences of religion. The second is where there are breakdowns of order or a total misunderstanding between researchers and participants. We maintain that these are just as useful, for example, when seeking further clarity or explanation from actors in an observed field setting.
Learning from our lack of understanding in aspects of our participant's lives, while simultaneously recognising something of their storied character, has been a first step for us in rethinking 'reflexivity' in ethnomethodological terms. As Coulon (1995: 22-23) insightfully puts it: Reflexivity refers to the equivalence between describing and producing an action, between its comprehension and the expression of this comprehension. 'Doing' an interaction is telling it. Reflexivity presupposes 'that the activities whereby members produce and manage settings of organized everyday affairs are identical with members' procedures for making those settings "account-able"' (Garfinkel, 1984(Garfinkel, [1967: 1).
We are not implying that everything will necessarily be insightful or complex, but rather that our everyday adjudications of what constitute those things, the information we designate 'data', and the process of reflecting on our position towards that information are inseparable. To suggest otherwise is to cleave our own tacit understandings away from our research practice, reserving the term 'reflexivity' for a much more specific form of doctored and catalogued introspection which is subsequently hailed as evidence of rigour and transparency.
Other scholars have expressed similar concerns (Mauthner and Doucet, 2003;Pillow, 2003) and it is not our contention, in step with Pillow (2003: 177), that we should 'stop talking about our positions'. Rather, we hope that these collective critiques provide better understandings of research as an embodied experience, to acknowledge the anxieties that accompany this, and to reflect on what this might tell us about the conclusions we reach within our research. Of course, 'our positions' are ever-present and thus inform research practice; a more useful admonition might be to urge each other to 'stop assigning special status to our positions'.
For reflexivity to produce better research we must recognise in situ reflexivity, move away from ad hoc rationalisations and towards thick descriptions of these experiences that are grounded and built up around relationships, exchanges, and situational interactions in the foreground and the substantive matters and relational practices of community, heritage, and ethnicity in the background. To be clear, we are not suggesting that religion or ethnicity themselves are abstract. Instead, we are suggesting that it is through relational practice that they are already (and so become) manifest. Speaking of their existence is insufficient; how are they understood, misunderstood, or quietly assumed within our research practices? If we marginalise such questions in favour of conceptual and analytical harmony, we ignore the very fabric of such phenomena.

Narrowing the gap between comprehensible and comprehensive experience: methodological implications
This argument carries two practical methodological implications that we consciously strive to take seriously in our future research, including around faith and religion.
The first is a more active, inquisitive, and dialogical approach to conducting qualitative fieldwork, be that direct participant observations in physical spaces of worship, for example, or the more incidental exploration of religion through interviews and ethnographic activity such as that described here. Our experiences in the field often left us with a sense of dissatisfaction and even shame that we could not more fully identify with our participants' life worlds. Yet we still managed to conduct our respective fieldwork with relative ease -our incomprehension did not put barriers in the way of doing 'good research projects' in that regard.
Based on our preceding argument, the problems with this scenario are less to do with the practical accomplishment of the fieldwork itself and more to do with our comprehension of the resulting data. Certainly, from our perspectives and with the benefit of hindsight, projects relying solely on interview data invite such ambiguities by design and we would be reluctant to replicate them again. Even where this is unavoidable, we believe that more can be done in seeking demonstrable clarity from participants to locate the importance of religion in the practical accomplishments of everyday life. Garfinkel's (1991) principles of respecification serve as important, and unashamedly demanding, criteria. Namely, that 'the reported phenomena cannot be reduced by using the familiar reduction procedures in the social science movement without losing those phenomena' and that the reported phenomena are only inspectably the case. They are unavailable to the arts of designing and interpreting definitions, metaphors, models, constructions, types, or ideals. They cannot be recovered by attempts, no matter how thoughtful, to specify an examinable practice by detailing a generality. (p. 16) We should not shy away from sources of 'trouble' (Lynch, 2011: 931-932;Garfinkel, 1984Garfinkel, [1967) where fraught understandings between participants and researchers force issues more fully into the open. Surely, this is likely to yield more valuable data than if we quietly exit the field, fearful that we might be discredited in the eyes of our participants and only thankful to have momentarily 'passed' (Goffman, 1963). If we truly wish to share the voices and perspectives of researched actors and communities, we should question unfamiliar narratives and seek clarity on their significance before conveying them, rather than misunderstanding them but trying to convey them anyway.
The second, related implication of this approach would be to re-evaluate some of our typical institutional approaches to certain ethical and safeguarding norms, including a more nuanced approach to harm. From most institutional ethical review perspectives, successfully entering and exiting research fields hinges almost entirely on our ability to do no harm (McGowan, 2020: 8). This should, of course, remain an absolute ethical requisite of any project -to the extent that it should almost go without saying. However, we suspect that such an imperative has perhaps created a number of unintended consequences for researchers, whereby the 'do no harm' principle has caused us to sometimes lose sight of our responsibility to seek a deeper understanding of the research field than when we entered it. As a deeply personal and sometimes quite private sphere of people's lives, religion is often perceived to constitute a quintessentially 'sensitive research topic' (Renzetti and Lee, 1993), which should not be challenged, questioned, or interrogated.
In our view, however, understanding a research field and doing no harm should mutually overlap, not sit dichotomously opposed. Seeking clarity, explanation, or elaboration from participants can be done without causing distress, alarm, or annoyance, just as we might broach somewhat 'sensitive' discussion areas in other routine settings. It can be done without causing 'harm' in an ethical sense, yet failing to pursue discussion in this way may not be so ingenuous if we go away and misrepresent people's accounts, presenting them simply to 'speak for themselves'. One danger of an excessively risk-averse approach to 'doing ethical research', including exploring beliefs we are not committed to ourselves, is that the misreading of accounts of minority and often marginalised groups is potentially exacerbated. We should not simply be looking at the text our fieldwork produces, but the context in which the narrative was produced, the exchanges and interactions leading to that point, including points of confusion or incertitude.
As we have reiterated, to ignore the importance of moments of incertitude in favour of 'getting on' with the analysis at hand is to divorce notions of reflexivity from its everyday and practical nature. The 'epistemic virtue' (Lynch, 2000) this claims, if only implicitly, in fact risks a strange sort of 'objectivised ignorance' whereby researchers can distance themselves from the very source of their data, often to give the impression of systematicity or rigour (see Rawls, 2002: 27). However, the practices of co-construction found within in-depth interview techniques highlights that researchers and research data are intimately connected. Researchers experiencing difficulty in understanding the significance of their data by trying to establish yet more distance away from it may see this gap between comprehensible, and comprehensively documented, experience widen.

Conclusion
In this article, we have utilised religious experiences reported to us as interesting examples which illuminate broader issues of sociological and methodological significance. The data extracts discussed here were articulated by people, in part, for their distinctly 'storyable' possibilities (Sacks, 1992: 218). Yet these narratives represent second-order phenomena and are not themselves reproductions of the experiences being relayed. Unlike out-of-body experiences or visceral emotions -the subject of some of our participants' stories -the narratives themselves were articulated in plainly comprehensible (i.e. understandable) ways. While original experiences contained within these narratives may be out of any shared ontological grasp, our participants' stories were rendered according to shared conversational and interactional convention. To conflate these two elements (i.e. original experiences and stories to which they now belong) would be analytically and ethically erroneous.
There is an important difference between comprehension as knowing what someone else is describing to you based on ontologically tacit knowledge that you identify with based on past transcendental experiences, and recognising what someone else is describing to you based on a shared framework of ordinary language which renders the stories they tell about themselves externally intelligible or comprehensible despite inner doubts, questions, confusion, or scepticism you might privately harbour. To discuss participants' experiences as the former, when we might only recognise the significance of their stories as a result of the latter, is perhaps disingenuous and abstract. If our task is to document the voices of our participants, then we must do so demonstrably. While researchers often share tacit emotions with their participants, something we seldom know unless they offer more detail for us to see (see, for example, Wakeman, 2014), those that do not are still responsible for offering more transparent accounts.
Our accounts have been offered in recognition of this responsibility. However, we argue they are ultimately inadequate for withstanding thoroughgoing methodological scrutiny, exhibiting an insufficiently 'detailed documentation of the research site' (Rawls, 2002: 27) for rigorously meaningful reconstruction. That said, the sort of narrative (re) analysis made possible here by Sacks (1992) has usefully illuminated our data, going some way to reconciling the issues highlighted in our Introduction. As such, we have also problematised our past approaches to researching violence using interview data, vowing to pursue alternative approaches in the future where possible. If action, meaning, and understanding are inseparable elements of a demonstrable account of social order (Mair and Sharrock, 2019), then there is value in remaining openly inquisitive about worldviews other than our own and 'working up' these experiences in subsequent analysis. The notion of 'privileging the data' is an important one and consists of 'unmotivated looking' in data analysis and collection (Sacks, 1984).
This returns us to the salutary question: what counts as sociologically 'see-able'? An insistence on empirically observable phenomena within future analysis need not equate to disenchantment in our own personal psyches, nor should it be misinterpreted as ignorance of, or disinterest in, contemporary expressions of religion and spirituality beyond our immediate grasp. Where and how we look for sociological answers when on spiritual terrain represents far too interesting a problem for such convenient conclusions. The writing up of our sociological research practice has much to gain from describing that which is comprehensible, documented as comprehensively as possible, while remaining ever curious about what might lie beyond our comprehension.