A qualitative fallacy: Life trapped in interpretations and stories

This paper points out some problematic aspects of qualitative research based on interviews and uses examples from mental health. The narrative approach is explored while inquiring if the reality of life here is forced into the formula of a chronological story. The hermeneutic approach, in general, is also examined, and we ask if the reality of life in this scenario becomes caught up in a web of interpretations. Inspired by ideas from Bakhtin and phenomenology, we argue for interview-based research that stays with unresolvedness and constantly question the web of interpretations and narratives that determine our experiences. This also chimes with certain dialogical practices in mental health in which tolerance of uncertainty is the guiding principle. Concludingly, we suggest that interview-based research could be a practice of ‘un-resolving’ in which researchers, together with the participants, look for cracks, contradictions, and complexities to prevent the qualitative fallacies of well-organized meanings and well-composed stories.


Introduction
To set the scene of this article, we begin with two different points of departure. First, we look into how the anti-narrative character of a novel seemed to get the reader closer to something real. Second, we look into how a narrativizing research interview seemed to lead to abstractions and created a distance between the stories made and the reality of the life of the interviewed.
for ways of approaching and describing the world that refrains from adding interpretations or creating meanings and narratives.

Reading narrativizing interviews
In a previous study we conducted (Bøe et al., 2019), a man was interviewed about his recovery from mental health difficulties. When taking a closer look at the interview, we found that the interviewer and the interviewee seemed to be engaged in constructing a narrative-the story of the man's recovery. After some initial questions about when he became ill and when he got admitted to the hospital, the interviewer asked: "But what was it that made it turn around?" Before this, the man had not made any mention of a turning point. The idea of a turning point was introduced by the interviewer through her question. A turning point could be regarded as a key moment in a narrative (Llewellyn-Beardsley et al., 2019;McAdams and Bowman, 2001), and the question brings the interview on track in forming a narrative. The man answered the question by telling a story about his encounters with a male nurse at the hospital that had meant a great deal to him. "He saw something in me," he said. This seems to have been of great personal importance, as he goes on to repeat this statement in various ways. He then described different episodes from daily life at the ward. However, this does not seem to add any substantial content to the formation of the narrative of recovery. The interviewer asks, "What was it that he did that you thought was...?" and in this way, the interviewer continues the search for an ordered sequence of events in the man's recovery story. What struck us further was that even though they appeared interested in creating a story of his recovery, it seemed difficult for the man to find the right words to answer such questions. He frequently lost track of any storyline. The interviewee described various intense episodes from life at the ward, yet the ties between these episodes and his eventual recovery repeatedly seemed to slip away.
The inquiries of this article are motivated by a previous study with the title (Nothing matters: the significance of the unidentifiable, the superficial and nonsense) (Bøe et al., 2019). In this study, we did a reanalysis of interviews from three interview-based studies, searching for what we referred to as "small things" in the lives of people with mental health difficulties that seemed to matter for their recovery. Early in the process of reading the interviews, we made a preliminary discovery that the interviewees often spoke of something important, but this "something" of importance simultaneously seemed difficult to put into words for them. What we sought to examine from our initial research questions and what we thought of as identifiable small things seemed to turn into "nothing" when we wanted to identify these "somethings" and convert them into meaningful themes. In many of the interviews, we sensed that the participants could describe something with enthusiasm-perhaps hesitantly, stuttering, abrupt, but still with a tone and eagerness that indicated something of importance. However, when the interviewers pursued an elaboration, with questions coming from a hermeneutical and narrative gaze like: "What do you mean by…?"; "What did that lead to?"; "In what way did that help you?" the answers provided seemed to gradually miss some of the enthusiasm in the initial utterings.
In one sense, interview-based qualitative research has its own narrative structure. Beginning with verbalizations of past events in people's lives, utterances are fed into the qualitative research "machinery." Step by step, it digests the "data" through interviewing, writing up transcripts, conducting a formal analysis, producing scientific writing, and publishing results in a conventional scientific genre. Within this circuitry, there is the danger that the reality of life step by step is molded into the mechanical forms of order, meaning, structure, and stories. This seems to be done primarily on the premises of the research and the discourses of science, more than those of participants' lives. In this, we suggest, is a risk that the research results become more and more distant from the reality of life. Life becomes trapped in a web of interpretations and stories.
With the tension between the example of an anti-narrative novel and narrativizing interviews as a point of departure, we now discuss some dilemmas related to qualitative research based on interviews. In the pursuit of understanding and narratives, the researcher risks clearing away uncertainty, complexity, contradictions, doubt, and hesitations. In this way, the research procedures might work as shears to sever the links between the "results" and their roots in the "realness" of lived lives. We identify this risk as a qualitative fallacy that we as researchers may easily be blind to.
The McNamara fallacy, also known as the quantitative fallacy, points out a kind of blindness that follows a one-sided focus on what can be measured and quantified (Yankelovich, 1972). In our study presented above, we suggested that we may also speak of a qualitative fallacy. This qualitative fallacy points out a kind of blindness that also follows qualitative explorations if they neglect aspects of life that resist being represented in linguistic, hermeneutic, and narrative structures (Bøe et al., 2019). In the present article, we go on exploring what such a qualitative fallacy might be about. The questions we raise in the following will be related both to philosophical positions and analytical strategies. Our focus will be on qualitative research based on interviews, including the whole procedure from the interviews to publications. We will relate to research in our own field, mental health.
A narrative qualitative fallacy-life made into a story One common way of interpreting and giving meaning to human living is through narratives. However, Brinkmann highlights how Knausgaard succeeds in writing something that he felt as real because the text did not seem to construct such a narrative. The novel moves in the opposite direction, so to speak, into a realm of everyday life, events, situations, and experiences, in an attempt to describe them and carefully avoid providing them with the form of narrative construction. Bourdieu (1986) makes his critical point towards narratives this way: "To speak of a 'life history' implies the not insignificant presupposition that life is a history" (p. 210). 2 This way of looking at life implies a "tacit acceptance" of the notion of "life history." Life is regarded as a series of historical events in chronological order, one leading to the other. The notion that life is a history primes us to disregard any aspect or manner in which events of life may happen in discontinuity, unpredictability, and at random. Additionally, according to Bourdieu, seeing life as a story disregards the social spaces (topological order) in which the events of life are carried out and converts a multiplicity of interaction in space into an individual progression in time.
According to Bourdieu (1986), the notion of "life history" has become part of our common sense, yet it has also "been smuggled into the learned universe" (p. 210).
Bourdieu points to the interview situation in which the interviewer and the interviewee are both interested and motivated "by a concern to give meaning, to rationalize, to show the inherent logic, both for the past and for the future (…) like that of the cause (…) and effect between successive states" (p. 211). Bourdieu describes how research interviews reinforce the conversion of life and reality into stories in this way: This inclination toward making oneself [the interviewed], the ideologist of one's own life, through the selection of a few significant events with a view to elucidating an overall purpose, and through the creation of causal or final links between them which will make them coherent, is reinforced by the biographer [the interviewer], who is naturally inclined, especially through his formation as a professional interpreter, to accept this artificial creation of meaning (p. 211).
According to this, life histories are not there to be explored; they are produced, and it is in the interview (and in the ensuing work of analysis and writing) that such stories are produced. In a sense, Knausgaard, who told us to "be alert whenever events shape themselves into narratives, for narratives belong to literature and not to life," seems to echo the ideas of Bourdieu. Atkinson and Silverman (1997) also question the status of narratives in social science, taking Milan Kundera's novel "Immortality" as their point of departure. In this book, Kundera critically examines how the subject is constructed in a biography through structures available in the culture. The interview is the genre that elicits such subject-constructing biographies. Atkinson and Silverman suggest that rather than eliciting personal, authentic narratives, the interview evokes narratives according to standardized cultural patterns presented in the "guise of private confession" (p. 314). They suggest we live in the "interview society" because of the widespread notion that the interview is a device for revealing an authentic self and authentic experiences through the celebrated narratives produced in media and research. They oppose how the interview and its narrative products are privileged in contemporary sociological discourse. 3 In the interplay between the interviewer and the interviewee, the experiences of the interviewee are "narrativized" (Hydén, 1995) and fall victim to a "narrative reconstruction" (Bury, 2001) in order to fit the researcher's aims. When writing as if what is communicated is "his/her story," something is achieved, yet something is also lost. With its composition dependent on keeping something in but leaving much out, combined with the insertion of causational relations, the narrative casts the reality of life with its complexity, unresolvedness, contradictions, indefiniteness, and myriads of connections and interruptions into the dark.

A hermeneutic qualitative fallacy-life caught up in a web of interpretations
The qualitative fallacy we want to address is not only present in narrative approaches but also in hermeneutic approaches in general. What Gumbrecht (2004) refers to as "the hermeneutic world view" currently seems to permeate much of the human and social sciences. Within the hermeneutic paradigm, qualitative, interview-based research can be said to operate in terms of the production of interpretations and meanings. Gumbrecht suggests that the way human sciences have ended up in such a hermeneutic paradigm stands in the way of including the more direct ways in which we live-not through interpretations and production of meaning, but through what he refers to as production of presence. By this term, he refers to what is characteristic of the more direct, sensible, bodily, and material sides of human living. "Hermeneutic maximalists," (Gumbrecht, 2004, 55) says, "hold interpretation to be humankind's exclusive way of relating to the world." Gumbrecht (2004) argues that this attention towards understanding, interpretation, and meaning has led to "a loss of the world" (p. 49): that is, that the reality of lived life that "meaning cannot convey" (p. 65) is left out altogether by the interpretative machinery of the human (and social) sciences.
Brinkmann found in Knausgaard a way of writing that made a strong impression on him, and he relates this precisely to the anti-hermeneutic characteristics of the text. An attempt to "write life" not as something to be understood, but something to be described in ways perhaps never described before; a writing that strives for liberation from the imperative of understanding, liberated from the impulse toward interpretation that nearly unavoidably sneaks into any exposition of life.
Brinkmann (2017b) points out another layer to the way science turns living into a matter of interpretation through the notions of "double hermeneutics" from Anthony Giddens (1993) and "the looping effect" from Hacking (1995a) A science that is about interpreting a human reality that is already constituted by people's interpretations becomes what Giddens (1993) referred to as "double hermeneutics." A growing web of interpretations is reinforced by social scientists when they construct new concepts and theories to explain the lives of social actors who are already embedded in interpretations. When science's concepts and theories are fed back into the public discourse, they become part of the self-reflective everyday vocabulary that the (initially interpreted) social actors use to interpret themselves and each other. Hacking (1995a) called this the "looping effect." When researchers' interpretations are fed back into the lives of those who are interpreted, their lives become affected by these interpretations (Brinkmann, 2017b: 9). People interact with these interpretations and categories and "tend to conform or grow into the ways they are described" (Hacking, 1995b: 21, see also Brinkmann, 2005). This is in accordance with what is referred to as Thomas' Theorem: "if men define situations as real, they are real in their consequences" (Thomas and Thomas, 1928). This feedback not only affects how people understand themselves and their situations but also their subsequent actions.
This loop of interpretations in science is put into accelerating motion so that human lives and experiences become more and more caught up in a "machinery" of meaning production. Science seems to contribute to, and reinforce, this hermeneutical loop in which both people in their everyday life, as well as the scientists, become caught up in an ever thicker and tighter webs of interpretations.

The phenomenological ruin of representations-cracking the sediments
We now turn to ideas from phenomenology, which precisely offer a critical analysis of this web of interpretations. According to Lévinas et al. (1998), the phenomenology of Husserl contributed to what he calls "the ruin of representations." Levinas points to the way science before Husserl was "caught up" in representations. The science that came before Husserl's phenomenology did not exhibit any clear awareness of the representative aspects of knowledge. Instead, knowledge of reality was presumed to be equal to reality.
In "The Crisis of European Sciences and Transcendental Phenomenology," Husserl (1970) pointed out how science had lost its roots in what he identified as people's lifeworld. Husserl used the metaphor of sedimentation to explain how the representations that science took for knowledge about the world was a petrified "world" where layer upon layer of representations made by science created a growing distance between reality and the experiences of everyday life. Grelland (2005) explains the process of sedimentation this way: We get, because we can write, read, and continue to write, constantly deposited new "geological" layers of meaning through the development towards higher and higher abstraction, where the deeper layers of concrete experiences and evidence are gradually hidden by the higher levels with what Husserl identifies as an "associative" structure of meaning that becomes remote from experience (…) (p. 33, our translation from Norwegian). However, given the distinction between reality and the representation of reality, reality always escapes representation, and thus knowledge cannot, and should not, be seen as accumulative (Grelland, 2005). A science that depends on writing, reading, and constituting ideal and abstract objects comes with a price: a decay and loss of a more original relation to reality. Steinnes (2007), drawing on Bakhtin and Derrida, claims that man's history of giving meaning to the world through religion, philosophy, and science can be seen either as a history of enlightenment or a history of decay and shows how Derrida regarded modern science as decay. Derrida identifies an inflation in language; an accelerating, petrifying increase of linguistic structures grounded in the idea that knowledge can and should be accumulated (Steinnes, 2007).
As a response to this scientific chain of writing and reading that produces new abstractions in the process of sedimentation, phenomenology wants to offer a methodology that ruins accumulated petrified knowledge to "go back to the things 'themselves'" (Husserl, 2001: 168).

Keeping the magma from cooling-the two-faced Janus of experience
We have, up until now, questioned narratives and interpretations. What about experience? Can we there find a direct, unmediated relation to reality? Mikhail Bakhtin (1993) writes about experience as a "two-faced Janus." As we, in Bakhtin's words, participate in the "ongoing event of being," we both turn to accounts of the event, which are constructed employing the language of a culture, and simultaneously we turn to the "never-repeatable uniqueness of actually lived life" (p. 2). Holquist (1993), playing with phenomenology's metaphor of sedimentation, point out how Bakhtin repeatedly returned to... "...the heart of the heart, … the center of the dialogue between being and language, the world and the mind, 'the given' and 'the created' …[seeking to] get back to the naked immediacy of experience as it is felt from within the utmost particularity of a specific life (…) [in the] sheer quality of happening in life before the magma of such experiences cool, hardening into igneous theories, or accounts of what happened" (Holquist, 1993: x).
However, Bakhtin recognizes that this "naked immediate experience" is outside any reach and points out "the word" as our path to reality. This path is by no means straightforward; no, words highlight and dim, the path goes through a "complex play of light and shadow" (Bakhtin, 1981: 277). Or, as he puts it: "[N]o living word relates to its objects in a singular way: between the word and its object, between the word and the speaking subject, there is an elastic environment of other, alien words about the same object … Indeed, any concrete utterance finds the object at which it is directed already as it where overlain with qualifications, open to dispute, charged with value, already enveloped in an obscuring mist-or, on the contrary, by the "light" of alien words that have already been spoken about it." (Bakhtin, 1981: 276). According to Bakhtin, we are dependent on words and dialog to enter into a relation to the world; without words, the world outside us is out of reach. On the other side, the path to the world that words make possible is most complex and manifold. It goes through "a weave of thousands of dialogical threads" (ibid).
Similarly, explicitly relating to research, Denzin (1991) points out how the perceptions and experiences articulated in interviews are mediated through language. This "language, which is our window into (…) our world, plays tricks. It [language] displaces the very thing it is supposed to represent so that what is always given is a trace of other things, not the thing (…) itself" (p. 68, italics added. See also Silverman, 2017). The articulation of experience in language can be seen as a window to reality even if language "displaces and plays tricks" because traces of the world can be found in language. Silverman (2017) points out a crucial distinction between, on the one side, the experiences found in the interview data seen as a topic to be investigated in itself or, on the other side, seen as resources for investigating the world. If experiences are seen as a topic, the researcher's interest is "the window," to use Denzin's metaphor above, through which we view the world. If experiences are seen as a resource, it is the reality that we see through the window that is of the researcher's interest. One could perhaps say that the qualitative fallacy we propose is about how qualitative research risks mistaking the window for the world outside the window.

Staying with unresolvedness
Like Bakhtin's work, phenomenology is also engaged in questioning how reality shows itself in our consciousness. Phenomenological approaches in research present themselves as an alternative to the abstraction and idealization of science that produces knowledge that gradually becomes more and more remote to experience and disconnected from living. The methodological strategies of phenomenology could be seen as a "solution" to the qualitative fallacies we suggest in this article, an alternative to the interpretative loops and the narrativizing character of much qualitative research. However, we think that there will always remain an inevitable and unresolvable problem with science as it can never directly access the pure, unmediated "things in themselves" that Husserl believed to exist beyond interpretive frames. Science must not be tempted to think it can overcome this unresolvedness. On the contrary, science should stay with it and even become "a practice of unresolving." Bakhtin points out that our relation to the world through experience is mediated by "the mist and light of words." This is in phenomenology addressed through what Husserl called the natural attitude (Zahavi, 2019b, 903). Giorgi (1997) points out how phenomenology introduces description as the main task of research instead of explanation, construction, and interpretation (p. 241). However, descriptions are not to be considered as "pure" accounts of the world and events in life. Instead, when research inquires about people's descriptions, these descriptions are already "captured" within certain "associative structures of meaning" and the sedimentations of abstractions. This is in phenomenology referred to as the "natural attitude," and, as Giorgi writes, the purpose of phenomenology is to analyze and understand this natural attitude with all "the details, biases, errors, and prejudices that we carry with us in everyday life" (p. 243) where past interpretations predetermine present experience. Experience is, in fact, a kind of "'realist' prejudice that what appears to you is truly the state of the world" (Depraz et al., 2003: 25). To really pay attention to the way the world is presented before you, "you must break with 'the natural attitude'" (p. 25), and when you break with the natural attitude, "reality is not lost from sight, but for the first time made accessible for a proper investigation" (Zahavi, 2019b: 903).
Phenomenology's strategies offered by Husserl-such as epoché, reduction, suspension and bracketing-are all "tools" that help us to make "reality accessible for investigation" by questioning "the natural attitude" that we mistakenly take for an immediate, un-interpreted perception (Zahavi, 2019b). Zahavi (2019a) puts it this way: Epoché and reduction can (…) be seen as elements [with the] purpose to liberate us from our natural(istic) dogmatism and makes us aware of our own constitutive contribution, make us aware of the extent to which our own subjective accomplishments are at play when worldly objects appear in the way they do and with the validity and meaning that they have. (no pages) In a sense, phenomenology offers research strategies that repeatedly question (and "melt") the ("hardened") structures found in the hermeneutic webs of interpretations and the narratives of both everyday life and science.
This aligns with the way of writing that Brinkmann found in Knausgaard, and which, to repeat, Knausgaard himself characterizes as "to write is to write oneself through the prejudices, to a world on the other side, as it could be when we were children (…) that gaze, where something is seen as for the first time" (Knausgaard in Brinkmann, 2015: 166). Knausgaard's overwhelming, detailed, often unconventional, and surprising descriptions (may) affect the reader so that reality is revealed as for the first time. Instead of the satisfaction of understanding something or the excitement of the progress of a story, you feel the joy that something real is exhibited-or glimpsed-there right in front of you. One might even suggest that the reading of the text evokes a feeling of realness that surpasses the one you have when perceiving reality itself. Perhaps this is so because the text succeeds in making cracks in the sedimented associative structures that come with the natural attitude that determines our perception.

The status of experience-a phenomenological confusion?
As Brinkmann points out, researchers have learned from phenomenology "to place experience on center stage of qualitative studies" (Brinkmann, 2017a: 119), and interviews pursuing "experience" are regarded as the Gold standard of qualitative methods (Silverman, 2017: 145). However, the problem might be that the character and status of experience that is put center stage is often not in line with phenomenological ideas as introduced above. Some research presented under the label of "phenomenology" seems to take the participants' (articulations of) experience, let it pass unquestioned, and apply it as a source for knowledge through various analyses in which it is structured, condensed, and put forth as knowledge about the world.
However, as Giorgi pointed out, what we get when asking for descriptions of experience is the "natural attitude" with its errors, biases, and prejudices infused with previous interpretations. Phenomenological approaches are about bringing these errors, biases and prejudices into light. As Zahavi makes clear: " [We] cannot simply take our natural realist assumptions for granted but must instead engage in a reflective move that allows [us] to explore and assess the epistemic and metaphysical presuppositions of the [natural attitude]" (Zahavi, 2019b: 903).
What is asked for in interviews and refined in analytical strategies in published findings might end up reinforcing, in phenomenological terms, the sedimentation it was intended to overcome. Under the label of "phenomenological research," scientific writing might reinforce the production of meaning and narratives that in turn "loop" back into people's lives. This creates a fallacy based on a misunderstood status of experience, seen from a phenomenological perspective.

The fallacy of narrativization and a hermeneutic world view in mental health research
A growing body of research within mental health takes its point of departure in asking persons who have recovered from a condition labeled by a diagnosis about what it was that made it possible for them to recover (Leamy et al., 2011;Tew et al., 2012;Topor et al., 2018). This research does not primarily explore articulations of experience as a topic (the question of how recovery processes are experienced and articulated and the conditions for this articulation), but sets instead out to describe "the reality" of recovery processes, with articulations of experiences in interview data as a resource in such investigations. This means that descriptions given by the participants are seen as data describing the dynamics, structures, and causes involved in the recovery process (experience as a resource), and not only how they articulate their experiences and thoughts of such a recovery process (experience as topic). This research is conducted by placing "(lived) experience center stage," often with reference to phenomenology and phenomenological methods (e.g., Bøe et al., 2014;Borg et al., 2011;Borg and Davidson, 2007;Davidson, 1993;Eriksen et al., 2012;Lindvig et al., 2021;Sutton et al., 2012). It may seem that experiences collected in interviews are taken at face value and accumulated through qualitative research (both in the single studies and in reviews) and composed into well-organized truths about mental health and what helps.
We suggest that perhaps the researchers (the interviewers) give more validity to the articulated experience than the interviewees themselves. If we look carefully at the interviews, we may discover that the way the interviewed articulate their experiences is often characterized by hesitations, doubts, abruptness and an inherent questioning of their own articulations. Perhaps the participants already in the interviews are engaged in a questioning of, in phenomenological terms, the natural attitude.
One question to be asked here is whether there is a tendency that these recovery narratives, produced from interviews, via analysis, to publications, adhere to a master narrative in western cultures (i.e., a narrative with the person as a hero in his/her own life story and a "happy ending") (Rose and Kalathil, 2019). In mental health research: the person taking control and becoming an agent in his or her own life, defining needs and goals, finding means to fulfill them, and ultimately reaching a state of recovery. The research that "produces" recovery narratives may accumulate and give rise to a recovery master narrative. This master narrative may, in turn, loop back and affect the way people articulate their recovery as a normative structure for people's stories. If such a master narrative is not problematized, there is a risk that experiences and narratives which do not fit become suppressed and marginalized. This point was well expressed by a woman who thought that her story was not "the story you want, I'm sure" because it did not correspond to the "compulsory positivity" demanded (Llewellyn-Beardsley et al., 2019: 304). Frank (2013) identifies three different types of narratives used to structure people's experiences of illness, a story of restitution, a story of chaos, and the story of a quest. He points out that restitution is the story from the doctor's and health personnel's point of view. Since there is no role for the sufferer's own perspective in this, the story of a quest, in which the patient becomes the agent, is needed. Similarly, Bury (2001) identifies different genres of narratives that operate in health research: Medical narratives in which the expert becomes the hero and moral narratives in which the user/patient becomes the hero.
We could ask if narratives with the professional as the hero (medical narrative, restitution narrative), as told by the expert, become simplistic, structured according to the master narrative of "the doctor curing the patient." However, the same should also be asked about the narratives told by the user/patient (moral narrative, quest narrative), with the user/patient as the hero, also turn a complex situation into a simplistic narrative, now structuring the user/patient as the hero of the story.
Interview-based research as "un-resolving"-looking for cracks, contradictions, and complexities We have argued that the temptation to impose order, coherence, and meaning upon raw qualitative data is fallacious and misrepresentative, turning the reality of "life as lived" into researcher-driven stories. Inspired by Bakhtin and phenomenological ideas, we now propose that interviews could be seen as a practice of co-questioning (rather than co-creation) in which both interviewer and interviewee are engaged in questioning the interpretations and stories that the interviews with a gravity-like way seem to produce.
Perhaps, we can find ways of doing research that is more about looking for crackswithout smoothing them out; showing contradictions-rather than covering them up; increasing complexities-rather than unifying. The fallacies of the narrativizing and interpretative machinery of qualitative research call for strategies to examine what does not fit in and what does not make sense. This becomes a kind of research that stays with contradictions, stays with unresolvedness, and refrains from bringing into order. Could research interviews-instead of being regarded as a kind of co-creation-be regarded as a practice of co-questioning and as a practice of "un-resolving'?" Ideas from Bakhtin (1993Bakhtin ( , 1981 and the methodological considerations that Sullivan (2012) and Sullivan and McCarthy (2005) offers may be helpful here. Sullivan and McCarthy lend the distinction and concepts of centrifugal and centripetal forces from Bakhtin. They argue that research should be a dialogic enterprise of a centrifugal character where the aim is to create a disordered space through an "outward" movement toward an ever-increasing multitude of possible perspectives. Traditional research may tend toward the opposite, research as an "inward" movement, a centripetal endeavor, seeking order, regulation, and unity through arriving at fixed meanings. On the other hand, the centrifugal dialogical approach, inviting an ever-increasing multitude of descriptions in qualitative exploration, could be present both in the creation of data in the interview and in the further analyses and interpretations of the material as text.
Such a centrifugal process implies suspicion, doubt, and a willingness to question any and every description. Sullivan (2012) suggests that dialogical analyses of qualitative data offer a possible approach that encompasses an attitude of both trust and suspicion vis-a-vis the content of interviews. However, as he interestingly points out, the dilemmatic choice between the attitude of trust and suspicion resides not only in the researcher. Already in the utterance of the participants, suspicion and a felt-ness of ambiguity vis-a-vis their own experience is present. A hesitation in transforming experience into words seems to be inherent in all speech. The participants speak with self-suspicion and doubt. It is as if the participants themselves hesitate and hold back as if they wanted to lessen the weight of their words so that they do not make definite what is indefinite. The attitude of trust and suspicion that the researcher brings to the analysis is already there in the doubting nature of the speaker's utterances. As Sullivan points out, this entails a presence of multiple meanings and an ambiguity in the said. In the utterances constituting the qualitative data resides "uncertainty, ambiguity, the anticipation of another's judgment, dilemmas and search for resolution-even amidst claims of certainty" (Sullivan, 2012: 15). These aspects of ambiguity, doubt, and unresolvedness should not be seen as a weakness or limitation but rather should be regarded as prominent aspects that we should be particularly attentive to.
Hesitations, doubts, contradictions, uncertainties, multivoicedness, abrupt speech, ambiguity, and so on, all present in the words people find for articulating their world, could and should perhaps be of particular interest in studies based on interviews. Is it possible to imagine a kind of research that invites this, recognizes it and asks in ways that open up and hold on to this in the analytical process and the writing and publications? Such an approach would neither be inductive, data-driven, nor deductive, theory-driven, but in line with a kind of abduction that Brinkmann (2014) calls "breakdown-driven" which attends to instances where understanding seem to collapse, to situations of "breakdowns, surprise, bewilderment" (p. 722).
Staying with unresolvedness in mental health practice Frank (2013) argues that within the health field it is considered necessary to understand your life and to create a productive narrative to be able to heal and obtain a good life. Also recovery research within mental health signal that recovery depends on understanding your life and leading it according to a narrative. Considering the arguments of this article, perhaps there is a flipside to this? Suppose the "message" in therapy and mental health initiatives, backed up with accompanying research, is that you must understand your life (make the right interpretation) and make a story of your life in which you become the hero. In that case, many of us may fall short when confronted by such an imperative. Furthermore, such an imperative may miss out on the reality of living. A good life should not be equaled to our understanding of it nor the story we make of it.
However, within existential approaches (Spinelli, 2014, 12) has suggested a stance of "un-knowing" where what is "seemingly familiar, assumed to be understood and understandable" now is treated as "novel, unfixed in meanings and, hence accessible to previously unexamined lived possibilities" (p.12). Within family therapy and a dialogical framework, Seikkula and his colleagues have developed the Open dialog approach helping clients together with their family and network (Seikkula, 2002;Seikkula, 2011). The approach is inspired by ideas from Bakhtin already introduced above. In the dialogs facilitated to help the client and his network-in contrast to much traditional practice-it is not crucial that any specific problem is identified and understood in the right way. Neither is the approach about problem-solving. On the contrary, there is in a sense an imperative that you should not understand, and an imperative that you should not solve any specific problem. The most critical and perhaps radical guiding principle underlying the approach is that practitioners are urged to tolerate uncertainty (Arnkil and Seikkula, 2015). In this way, an open dialog may be facilitated and generated. A multiplicity of voices and expressions may be invited, without an agenda of reaching conclusions, agree, making decisions, taking specific actions, and so on. Bøe and Colleagues (2015) points out that expressive and ethical aspects outside the hermeneutic or narrative domain are perhaps the most important in this practice. The Open dialog approach is not (primarily) about the content of the told, the narratives, but the dialogical event of telling and the responsiveness that comes with it. The lives of the involved may change through the generation of a vitalizing multitude of expressivity that precisely remains indefinite and unresolved (Bøe et al., 2015).
The example of the open dialog approach could perhaps serve as an indication that it is possible to develop practices outside the fallacies and imperatives imposed by a hermeneutic and narrative world view and research.

No last words…
Etymologically, the word "research" means to look really hard or very meticulously (Online Etymological Dictionary, 2015). This thoroughness is often understood in terms of having a specific, straightforward research question as a point of departure. It involves systematically applying pre-established methodological procedures based on an epistemological framework to organize and structure data to turn it into a form of knowledge capable of answering the research question. Science is often understood as a procedure for procuring answers or solving problems. Could the thoroughness of research be thought of in other ways? Perhaps there exists a thoroughness that is about a kind of "dwelling," a way of staying with, and even the exhibition of cracks, contradictions, and complexities. Is it possible to hesitate, refrain from taking further steps toward order, well-ordered fixed meanings, or well-composed narratives? We call for research that does not answer but asks new questions. One that does not solve problems, but perhaps creates new ones? It is crucial to keep the vital doubt and uncertainty alive. Ultimately, this will allow it to be possible to, again and again, get a glimpse of the world as if for the first time.