Being Where? Navigating the Involvement Paradox in Qualitative Research Accounts

Choice 1: Visibility

Perhaps the most striking characteristic of ethnographic realism is the almost complete absence of the author from most segments of the finished text. Only what members of the studied culture say and do and, presumably, think are visible in the text. The fieldworker, having finished the job of collecting data, simply vanishes behind a steady descriptive narrative justified largely by the respectable image and ideology of ethnographic practice. (Van Maanen, 2011, p. 46)

This first choice is the most obvious and fundamental in authorial representation. This refers to the degree to which the researcher is integrated or woven into the manuscript as a character within the research account. From our perspective, visibility in a manuscript goes beyond the method section, where authors are most likely to discuss their roles and relationships to the field, and extends to the findings where one’s role in the study and in the creation of the data is made more or less transparent.

At one extreme of the visibility dimension, the researcher is virtually invisible in the empirical story. Van Maanen’s (2011) treatise on “Tales of the Field” quoted above refers to a positioning in which the author “disappears” (see Figure 1). Van Maanen describes this form as a “realist tale,” and notes that it has been the dominant approach to ethnographic writing. Similarly, in her discussion of different forms of authorial positioning, Hatch (1996) constructs a two-by-two typology based on Genette’s (1980) narrative theory where the two dimensions are labeled “perspective” (or “who sees,” equivalent to an insider or outsider view) and “voice” (or “who says,” referring to the presence or absence of the author as a character in the story). The outsider perspective in which the author is absent is seen as the most frequent, and similarly suggests invisibility. As Hatch (1996) puts it, “the story exists rather than being told.”

It is indeed easy to find examples of qualitative research and even of ethnography, where the author essentially disappears as a character in the findings section. For example, the empirical findings sections of Barley’s (1983) semiotic ethnographic study of funeral directors, Pentland’s (1992) ethnographic study of software hotline technicians, and Balogun and Johnson’s (2004) study of middle-manager sensemaking during strategic change report activities and meanings as they were or are. The researchers, though very clearly “there” to contribute to the creation of the data are “not there” in the empirical description, although they may be quite strongly present and quite active in the sections of the papers dealing with literature, method, and discussion. Our own published qualitative studies (e.g., Bucher & Langley, 2016; Denis, Dompierre, Langley, & Rouleau, 2011; Klag, Jansen, & Lee, 2015) have also tended to follow this pattern for the most part. The commonplace nature of this phenomenon suggests that it reflects implicit norms of the field. Researchers are expected to be visible as observers and analysts, but unobtrusive within the presentation of the data themselves.

It is the other extreme, however, that is more thought-provoking, where the representation of researcher presence holds a more prominent place throughout the manuscript. We see an example of this in Whiteman and Cooper’s (2011) ethnographic study in subarctic Canada in which, in addition to the authors’ detailed stories of how they collected and analyzed data, the first author “appears” (see Figure 1) as a major character in the findings. The reader is exposed to her own personal narratives and stories of her actions and interactions as a participant in the field as well as her reactions to events that unfolded during the study period. The reader clearly sees the authors as both researchers and deep participants in the study.

More striking still are the rare cases of qualitative research in which the author is represented not just as a character participating secondarily in events as they occur, but as a central focus of the research. A classic illustration is Gephart’s (1978) rarely imitated study of his own “status degradation” in an academic setting based on field notes collected by himself and others. Gephart (1978) is indeed, as he had to be, highly visible as a major character in the research account, identified by his initials “RG” throughout the paper. Of course, the study of the mechanisms of one’s own humiliation is a particularly sensitive and extraordinary initiative that raises a number of other issues concerning how one can conceivably study such a personally distressing event and yet do so in a credible way. Another example is Herepath’s (2014) research on structure and agency in the practice of strategy in which the author is described as the “principal strategist” and in which the process of strategy making was the key focus. As Gephart (1978, p. 579) himself points out, “Developing concepts for analyzing everyday life activities by examining one’s own experiences seems a useful and inexpensive way of enriching organizational theory.” And yet, we do not often see such initiatives in the mainstream literature (perhaps partly because of ethical concerns) particularly in recent years.

Between the two extremes are levels of more limited visibility. One example is Lingo and O’Mahony’s (2010) ethnographic study on the brokerage role of producers in the country music industry. In addition to a brief note about the first author’s role in interviewing participants, the first author also represents herself not as a major character, but as a data point in the analysis, having coproduced a song, and comanaged a recording session with other band members.

So how does the question of visibility relate to the involvement paradox we discussed above? When authors make themselves visible in their research accounts, this makes them more vulnerable to critique on the grounds of professional distance. The strategy of rendering the author invisible, even though they were actually involved and may even have exerted some influence, thus seems to be one way for authors to elide questions about their role, and to convey distance and rigor. Revealing too much might undermine, rather than help, their cause. For example, as Anteby (2013) notes in his discussion of the types of personal involvement related to prior insider status, personal similarity or emotional engagement (Cunliffe & Karunanayake, 2013):

If not for personal communications with authors, it would not be evident that the timing of a study of funeral home directors coincided with a proximity to death in the author’s family (Barley, 1983), that a study of hospital employees’ job crafting was partly shaped by one of the authors’ upbringing in a nurse’s household (Wrzesniewski et al., 2003), or that a study of airline pilots stemmed in part from the author’s having been raised in family of pilots (Ashcraft, 2007). Such omissions help eliminate any perceived taint associated with personal involvement. (Anteby, 2013, p. 1279)

However, it is important to recognize that part of the visibility issue is about transparency. The first author recalls editing a paper in which on the first round, characters were introduced into the empirical story without specifying that these characters were actually the authors. Knowing who the authors are and what they are doing is clearly important to understanding the research and the knowledge claims that emerge from it especially if it is possible that the researchers may have influenced events.

Our point is that transparency is the baseline for representation of involvement. If involvement of some kind is occurring, it needs to be declared, and if researchers were active in the situations described, it is clearly important to know what they were doing. We recognize that researchers may then need other ways to address the involvement paradox. We discuss some of these in relation to other choices below.

Choice 2: Voice

In the last few years…a number of scholars are deepening our understanding of the practice of fieldwork by doing something simple and straightforward: namely, adopting the first person narrative voice…. The use of the first person is more than a cute convention or a self-deprecating call for attention. It is more than just an assertion of fieldworker chutzpah or blind ignorance of a world-out-there. At its core, scholars are turning to the self in order to discover not only truths about their own experience but about the world out there. (Venkatesh, 2013, pp. 4-5)

Beyond the question of visibility as such, is the question of how authors represent their narrative voices in a text: in the passive voice, in the first person, in the third person, or alternatively as a multiplicity of distinct voices (which we express as from depersonalizing to personalizing in Figure 1). “Voice” is another key dimension of authorial positioning proposed by Hatch (1996) and based on Genette’s (1980) narrative theory. The dimension is related to the prior one of visibility, in that an invisible author naturally suppresses his or her own personal voice in describing the events and phenomena observed. However, the converse is not true. It is possible to present oneself as a visible character in the empirical story and yet narrate findings in several different ways, varying from a highly personalized first-person account to a more neutral third person style. When there are multiple authors, their voices may be presented as unified and singular or as distinct and individualized. We subdivide the choice relating to voice into two subcategories associated with personalization (whose voice?) and multiplicity (how many voices?).

Choice 3: Stance

In my research, I was always hoping that what I found could be of use to people less fortunately situated than I. But while in the Harvard Society of Fellows, I was conditioned to believe that if research was truly scientific, researchers’ values must be set aside…. As I gained experience in a wider range of research situations, I found myself gradually abandoning the idea that there must be a strict separation between scientific research and action projects. Through the rest of my career, I have been exploring how research can be integrated with action in ways that will advance science and enhance human progress at the same time. (Whyte, 1984, p. 20)

Clearly, when a researcher goes into a study setting to collect data, he or she may actually influence it in some way by mere virtue of his or her presence, or by the actions through which the data are constituted. In addition, researchers bring with them values that influence what they see and focus on. Yet despite recognition among scholars of qualitative inquiry that no research can be either entirely value-free (Lincoln & Guba, 1985) or entirely unobtrusive (Fine, 1993), it seems that the default mode of representation of qualitative research involves the assumption of neutrality and unobtrusiveness (i.e., noncontamination by the researcher’s personal concerns or active influence on the research site even when acting as a participant observer). Thus, the third critical authorial choice that we call “stance” is defined as the degree to which the researcher incorporates recognition of the role of personal values or influence on their research sites and subjects in their representations. Cunliffe and Karunanayake (2013, p. 381) refer to this same dimension when they describe what they call the “political activism-active neutrality” hyphen-space as “the most challenging for researchers in that it involves not only identity work around ‘who am I?’ and ‘what are my values and responsibilities to act?’ but also emotional, personal, professional, and political risk.”

Indeed, while visibility and voice both bring the involvement paradox to the fore to some degree, the dimension of stance does so with even greater acuity because a nonneutral stance that appropriates or even lauds the active engagement of the researcher with the researched (whether through some form of action research, critique or advocacy of change as expressed in the quotation from Whyte above) fundamentally challenges the notion of professional distance. This explains why many qualitative articles’ authors go to some lengths to argue that the research intervention itself did not influence what was going on in the site in any significant way. For example, in her study of investment bankers, Michel (2007, p. 516) notes,

Although my presence could have influenced my informants’ behavior, I believe that the bankers’ fast-paced and demanding work deflected attention from my research role. These high demands and my almost daily presence over the period of two years would have made it difficult for the bankers to alter their behavior systematically.

While a neutral stance appears, explicitly or implicitly, to be the norm in the more mainstream journals, some scholars have dared to position themselves more proactively, and these provide an instructive contrast. For example, Jay’s (2013) article on the paradoxes associated with hybrid organizations offers an example of how authors can creatively incorporate recognition of influence without undermining the value of the research. Playing the self-ascribed role of “engaged organizational historian,” Jay (2013) describes explicitly in his paper how a conversation between himself and organization members in a feedback session (transcribed verbatim in the text of the article) led these members to a realization of the paradoxical nature of performance, stimulating a change in orientation. Jay positions his overall perspective in his method section as follows:

As I gathered these data about the unfolding organizational history, the “engaged” aspect of the role meant periodically sharing my findings through conversations and written documents. Doing so ensured deep and sustained access to people and documents because organization members saw my presence as potentially contributing to organizational learning and effectiveness…. The interactive process of sharing findings and reflections also generated useful data about sensemaking processes as people responded to my interpretations with their own. At the same time, this research approach—combining elements of ethnography and action research—necessarily introduces biases in the research process. My interpretations of CEA’s context, actions, and outcomes clearly affected my informants’ interpretations. The key to rigor with this interactive methodology is to be reflexive and transparent about one’s own impact, to “triangulate” insights with multiple data sources, and to consider one’s own role when theorizing about organizational processes observed. (Jay, 2013, p. 142)

It is interesting to note that Jay (2013) describes his method as “combining elements of ethnography and action research.” The reference to action research no doubt helps legitimize his stance, because it signals an accepted framework within which researcher influence is expected and normalized—the research purpose is deliberately oriented toward intervention. In the case of other explicit action research studies we identified (Bate, Khan, & Pye, 2000; Bernstein, 2012; Lüscher & Lewis, 2008), influence is clearly a central element of the approach represented, and the legitimacy of this influence is not in question. Yet, in our experience, we found few such studies published in the premier organizational journals, suggesting that there may still be some timidity surrounding the proactive stance expressed in such studies.

Moreover, in these action research studies, while recognizing their influence on events in the site, researchers were often careful, as above, to mitigate concerns about professional distance even within the context of this process. For example, in their action research study of LEGO, Lüscher and Lewis (2008) state their goal of seeing managerial processes as objectively as possible, in order to be of most help to the organization:

In…collaborative approaches, a complementary researcher role entails instilling methodological rigor. This role involves three components…. First, to enable effective collaboration, researchers identify possible patterns and emerging categories to explore with the subjects. In our study, systematically reviewing intervention session notes, interview codings, and existing literature enabled this effort. Second, the researchers leverage outsider perspectives to critique the research process…. Third, researchers encourage subjects’ ongoing experimentation and reflection to assess the validity and value of findings. (Lüscher & Lewis, 2008, p. 227)

The notion of stance we use in this section refers, however, not only to acknowledgement of influence, but also to statements revealing overt ideological commitments or political advocacy. Such a stance is revealed for example in a paper by Whiteman and Cooper (2016) titled “Decoupling Rape” about the way in which a forestry company in Guyana managed to decouple itself from concerns about both environmental and gender issues relevant to its operations while gaining environmental certification. The paper includes explicit recognition of the first author’s advocacy agenda and of how this intervened in the research process. Van Maanen (2011, p. 171) referred to such studies as “advocacy tales,” that is, ethnographies with an agenda, aimed at righting perceived wrongs: “Advocacy tales put forth a strong, clear point of view in which no doubt is left in the reader as to what side the ethnographer is on. Such a moral stance is carried throughout the writing and not restricted to occasional asides.”

Whiteman and Cooper’s (2016) paper, published in the recently created journal Academy of Management Discoveries, is accompanied by recorded audio soundbites that include discussion of the authors’ motivations behind the research as well as the difficulties encountered in carrying it out, all revealing the researchers’ advocacy stance. In the paper, Gail Whiteman describes her role and influence in a rather nuanced manner as follows:

While the first author circulated research results broadly, asked multiple questions of many stakeholders involved in the certification in the interest of reducing discrepancies and recoupling CSR behavior, and used their responses for data, by no means was her primarily role one of activism or active NGO campaigner. Her role remained research focused in terms of the investigation of documented evidence and interview perceptions. Notably, however, her active engagement with the certification and post-certification process did influence institutional outcomes in minor ways. (Whiteman & Cooper, 2016, p. 124)

Of course, other examples of nonneutral papers manifest in a critical perspective, though not necessarily communicating a societal advocacy agenda per se. For example, Barros (2010) uses participatory action research via a critical management studies lens, which is rooted in a critique of and search for alternative approaches to work life. Somewhat like action research more generally, reference to critical management studies or methodologies associated with it provides legitimating language (if used appropriately) for authors to situate themselves in acceptable ways within particular research communities. Again however, the examples we find in the organization studies literature are quite rarely published in the field’s mainstream journals.

Nevertheless, we think that appropriating stance (i.e., acknowledging the value commitments that drive the researcher’s involvement and/or specific activities that might influence findings), is important; it enables readers to appreciate the lens through which knowledge is constructed. Arguably, it is quite impossible for a researcher to be completely neutral in a study setting no matter how many checks and balances are put in place. The question for an author may become one of the extent to which a nonneutral position might affect the study process and outcomes. If there might be an effect, then how important is it for the reader to know this? Making it explicit allows readers to better grasp the driving concerns behind the research questions being studied, placing them in greater perspective. This brings us to the fourth dimension, reflexivity.

Choice 4: Reflexivity

Radical-reflexivity turns the reflexive act upon ourselves to deconstruct our own constructions of realities, identities, and knowledge, and highlight the intersubjective and indexical nature of meaning (i.e., accounts are ongoing discursive social accomplishments taking place in shared, taken-for-granted interactions between people). Radically reflexive researchers explore how we as researchers and practitioners constitute meaning through our own taken-for-granted suppositions, actions, and linguistic practices. (Cunliffe, 2003, p. 986)

The final choice turns the spotlight directly on the “self,” that is, toward a written interrogation of one’s position in the field and on the question of how the author-researcher’s own personal and professional identities and relationships with research participants might influence either themselves or the nature of the findings and theorizations reported. Following Cunliffe (2003), we represent this dimension as a continuum from “unreflective” (ignoring questions of how one’s role might be implicated in findings—and essentially “normalizing it—see Figure 1) through “reflective” (taking an objectivist stance to analyzing one’s own role) to more radically “reflexive” (recognizing the co-construction of research accounts between researchers and researched as well as their fragility, labeled “interrogating” in Figure 1).

A significant proportion of qualitative studies that we reviewed fall into the unreflective camp. This is not surprising when researcher visibility is limited, voice is depersonalized and researcher stance is neutral because there is no obvious “self” to be placed in the spotlight. Clearly, many qualitative researchers do not see a reason to interrogate their own role. Given the norms of publication, they might also wonder whether such questioning might be counterproductive in terms of enhancing the credibility of their work: According to a French proverb, qui s’excuse s’accuse (he who excuses himself, accuses himself). Thus, this may be the first thing to go when there are space constraints.

A contrasting approach further along on this continuum is represented by articles in which authors describe and question in more detail their own role in the events investigated, usually in the method section of a paper or in the limitations. Such references are usually no longer than a paragraph or two, but they do involve addressing questions about reactivity and potential bias that readers might have about the research. Examples include some of the extracts we illustrated in other sections above (Jay, 2013; Michel, 2007; Pratt, 2000; Whiteman & Cooper, 2016) where authors declare the form of their involvement, but at the same time rhetorically reassure readers concerning the legitimacy of the knowledge products developed, mitigating concerns about their professional distance. Cunliffe (2003) would describe this level of self-analysis as “reflective” rather than fully “reflexive” because it still assumes that the researcher has fairly full insight into their own situation.

Such extracts nevertheless help take the reader “under the skin” of the researcher’s experience, providing at least a glimpse of what it was like to “be there,” and potentially revealing how the researcher dealt with the ambiguities of their role. Some authors may even choose to develop such reflective accounts into full-fledged articles or book chapters separate from an original theoretically oriented piece that takes a less self-reflective orientation (Barley, 1990; Gersick, 1992; Kisfalvi, 2006; A. D. Smith, 2002). Van Maanen (2011) used the term “confessional tale” to refer to ethnographies that focus more on how the process played out for the researcher than on theoretical issues emerging from the study itself. Such confessional tales can be a rich and instructive resource for learning about the realities of qualitative methods and the struggles associated with gaining insight from qualitative data (Klag & Langley, 2013). For example, Barley’s (1990) “confessional tale” of doing the ethnography that led to his well-cited study on the structuring role of technology (Barley, 1986) is a much loved reading in qualitative methods seminars. This article reveals in rich and unusual detail some of the emotional and political struggles associated with involvement, without necessarily suggesting that these were easily resolvable. For example, the author gives an account of an incident that placed him in a particularly problematic political minefield and that might have jeopardized the research:

Near the end of the study, I witnessed an instance of extreme hostility between a technologist and a radiologist which resulted in the radiologist calling the technologist a number of unflattering names. The technologist, who felt that she had quietly borne a long history of affront, was quick to recognize that my presence meant she could successfully lodge a formal grievance, if I were willing to testify on her behalf. The technologist lost no time in asking for my aid. After a moment of agony, I decided that I could not risk compromising my position as researcher. I therefore lamely told the technologist that I could not assist her because I could not attribute blame or malicious intent when I had no access to the radiologist’s thought processes. The technologist became justifiably indignant, and I feared that my refusal would prove costly. (Barley, 1990, p. 243)

There are indeed a few qualitative papers that do attempt to reach beyond simple descriptive reflections about doing research to question more deeply the researcher’s role and its implications. A challenging example of “radical reflexivity” is represented by Boje, Luhman, and Baack’s (1999, pp. 357-358) study of an organization called the “Choral Company” in which the authors offer an attempt to “tell a research story without being hegemonic” by fully recognizing the range of competing stories among respondents and others involved in the research. The authors include multiple references to the struggles this required:

It is a difficult challenge for any storyteller to not privilege his or her own point of view, to step back and be self-reflexive about the ways in which he or she selects quotes, position them as authors, and recontextualize their meaning in this write-up. We as writers did privilege our own voices in narrating this article, selecting segments of other writers: authors cited, editor, reviewer, doctoral student, CC member, and academy official transcripts. As an article with multiple authors, our focus was to juxtapose our own self-aware hegemonic science tales with a synchronic, co-constructed telling of the story (3B). We agree with Ricoeur (1984) that there is no “whole story,” only bits and pieces woven together in a narrative that invites readers to read and unfold a macro-tale. In our ironic manner, we did indeed want to make the hegemonic constructions of the way we write and otherwise construct “subjects” more visible (3B).

A second example of this kind of radical reflexive form of representation is Hardy, Phillips, and Clegg’s (2001) reflexive actor-network theory based analysis of how the authors contributed through their own research to constructing the subject of the “refugee” through their persuasive interactions with both respondents (i.e., the “interviewees”) and with the research community itself (composed of journals with specific norms and sets of reviewing practices). Here again, the authors’ purpose has shifted from representing the dynamics of the research site itself (achieved in Hardy, 1994; Hardy & Phillips, 1998) toward representing the process of representation in which they themselves along with the publishing system are imbricated. Both this and the Boje et al. (1999) study reveal the particular challenges of a radical form of reflexivity. Such forms may be a fertile terrain for experimentation in qualitative inquiry, but they tend to stretch to the limit the boundaries and potentialities of the qualitative research genre.

Indeed, with a radical reflexive approach, not only is the focus moved from the phenomenon studied to the researcher themselves, but in addition, the very legitimacy of qualitative research of the more usual kind appears placed in question. Unlike most of the studies discussed under previous headings, radically reflexive researchers do not attempt to mask, mitigate or explain away the ambiguous implications of involvement. Rather, they revel in them, drawing attention to the fragility of the research enterprise, while still attempting to avoid undermining it. This is done not by claiming professional distance, but by avowing vulnerability and by revealing a heightened awareness of the power relationships that underpin research, despite the best of intentions.

Whether the radical reflexive approach is ultimately a desirable solution to the involvement paradox is debatable. For one thing, it draws attention away from the phenomena ostensibly being studied to focus on the researchers’ own soul-searching. For another, as Boje et al.’s (1999) study suggests, it has a circular quality (see also Cunliffe, 2003) because even when being radically reflexive, authors are still exerting their power over the text and cannot avoid orienting what they say to what they think their readers are likely to appreciate. Such is the nature of the publication process. As Fine (1993, p. 283) comments:

Being candid becomes a situated choice that is forever linked with how the candor is likely to affect one’s reputation as a scholar. We have our careers to think of, and issues of honesty and ethics must be analyzed within this personal nexus (Barnes, 1979, 179)…. Recent experimental attempts to move oneself into the center of one’s ethnography can no more escape the dilemmas of exposing one’s candor than can attempts to pretend that one wasn’t there at all.

Reductive Approaches: Emphasizing Distance, Downplaying Involvement

Some reductive approaches to the involvement paradox work by minimizing references to or explicit recognition of involvement and its possible influence on the research. Of course, it is hard to detect when this is occurring. In fact, Anteby’s (2013) references to personal knowledge of elements in certain authors’ backgrounds that might be relevant to the studies they conducted suggest that it is not uncommon. Moreover, we think that it can sometimes be guessed at.

A case in point arises whenever the object of the research involves cognitively based themes such as sensemaking or identity combined with methods in which the manifestations of these phenomena are obtained through explicit solicitation such as through interviews, focus groups or diaries (Balogun & Johnson, 2004; Gioia et al., 2010). These methods are extremely effective in capturing respondents’ explicit understandings. However, if Weick’s (1995) conception of sensemaking is considered seriously, sensemaking is always social—sense is created in interaction with an audience. When the audience is the researcher, or people brought together and staged by the researcher, then the researcher’s role in enabling meanings to emerge can be significant. In other words, the sense that is made is not necessarily naturally occurring, and authors are inevitably coproducers of it. This starts to be important when reflexivity among respondents generated by such techniques (potentially desirable in itself, but that might not occur in the same way without the research) could play a role in the evolution of subsequent events that are themselves being studied. Yet, few of us (and the first author speaks for herself as well) ever even comment on how solicitation techniques might potentially influence the phenomena we are observing. This is a place where we think that researchers could be more transparent and reflective about their involvement and influence. The example of Jay’s (2013) research shows how this may be productively accomplished.

Other more reductive approaches work by objectifying involvement. The most striking examples we noticed are the studies by Gephart (1978) and Herepath (2014) where the researchers were key actors, but are viewed from an objectivist stance throughout, with no reference made to the personal implications of their role. This is unusual. Much more common is the phenomenon of splitting involvement—as we mentioned, an approach that to some extent succeeds in signaling desirable involvement (being there to obtain rich data) and desirable distance (not being too deeply implicated in the creation of these data). This is done by referring to the researcher in the first person and directly to describe data collection and analysis procedures as well as to introduce theory and discussion, but then presenting the data and findings without referring to the author at all—in other words objectifying observations in the findings. This approach seems so ingrained in our research practices that it is rarely questioned—unless perhaps the researcher has left a trail of breadcrumbs in other parts of the paper that suggest that involvement might be of more significance than implied (e.g., as reflected in the quotes that open the paper).

Are reductive approaches a problem? One of the points of our paper is that we as researchers should be able to confront the realities of the involvement paradox and that reductive accounts may do a disservice to the field by assuming away the very real tensions our research practice. This brings us to some examples of where authors seem to deliberately emphasize the other pole of this paradox.

Expansive Approaches: Emphasizing Proximity, Playing Up Involvement

As discussed above, because of institutionalized publishing norms and expectations, we believe that most authors tend to play down rather than play up their involvement in their research accounts. Nevertheless, there are some interesting exceptions.

An audacious example combining high visibility, personalization, appropriation of advocacy stance as well as reflexivity is Vickers’s (2015) creative writing study of disabled employees perceiving themselves as “dirty workers.” In this study, the author combines snippets of interviews from two respondents with multiple sclerosis with her own experiences of dealing with this affliction as a faculty member. The paper alternates stories from the respondents (recreated artistically from small pieces of information within their interviews and presented as if from their perspective) with accounts of how she understood and interpreted their situations, and how this related to her own painful experiences of coping with disability, stigmatization and cruelties—unintended or otherwise—from her coworkers. The paper draws together these sources to develop a moving portrait of the experience of disability in the workplace, an experience however that is clearly and transparently as much if not more that of the author than that of the people she studied. Vickers’s (2015) study is bold and original, and it is interesting to see how she describes and reflects on it:

Other scholars have confirmed my view of the rarely acknowledged proximity that more traditional written academic presentations have to more creative works. Academics do what creative writers do: They create, discover, and try to bring to life both the individual and the unique. And (this should never be admitted by any sensible person wanting a continuing academic career) we academics often use illusion and rhetoric to make our case. The really rebellious among us go even further, drawing on our own life experiences in our scholarly work while holding close the notion that it is, in fact, impossible to tell the whole truth (e.g., whose truth?) in any written research presentation. So, when working with others’ life stories, I knowingly and deliberately cut, compress, and combine people’s reported slices of life. Academics refer to this as “making interpretive choices.” (Vickers, 2015, p. 82)

As can be seen, expansive forms of representation can stretch rather radically the norms of qualitative research scholarship as accepted in the field’s most highly ranked journals. Fortunately, there are venues where novel representational forms can be experimented with, honed and assessed. Nevertheless, because such forms tend to center the reader’s attention more on the authors themselves than on the original object of study, the contribution to advancement of knowledge may be harder to see, notwithstanding a significant methodological contribution. Moreover, as the quote from Margaret Vickers’s paper explicitly indicates, authors who engage in such efforts need to accept vulnerability (the essence of radical reflexivity). Whiteman and Cooper’s (2016) study mentioned earlier reveals a similar concern.

Both-And Approaches: Transcending the Involvement Paradox

A paradox is defined by W. K. Smith and Lewis (2011) as contradictory yet interrelated elements that exist simultaneously and persist over time. In other words, in a paradox, the poles are irreducible because they are mutually dependent and therefore must coexist whether one pole or other is emphasized at a particular time. In the case of the involvement paradox, involvement is both desirable to guarantee the kind of proximity that will enable us to capture the richness of the phenomena, and potentially problematic because it suggests that our understandings might be biased by that very proximity or that our proximity may have altered the shape of what is there to be understood. We argue here that it is important to acknowledge this irreducible tension. In examining published articles, we noticed different ways through which authors may attempt to mobilize and transcend the involvement paradox, accepting the importance of both poles.

A first rhetorical strategy that occurs frequently in published qualitative studies is what we called “mitigating involvement” (bottom of Figure 1—see the sections on voice, stance, and reflexivity). In its most common form, this means recognizing that involvement occurred, but then declaring that this involvement actually had minimal effect on the data collected or the analysis conducted. This is a form of mitigation that tends to imply that involvement and influence are probably not desirable features of the study, leaning toward the distancing pole (Michel, 2007). A rather different kind of mitigation occurs when an author positively defends involvement as a desirable feature of the study arguing against distance and leaning toward the value of involvement, as in this extract from Herepath (2014):

In undertaking this [embedded] approach, appreciating the act of intellectual arbitrage in play, my then role as the principal strategist tasked to develop the focal strategy was pivotal. Yet for SAP [strategy as practice] academics, irrespective of how well-informed and intermittently immersed within a strategic episode, they routinely remain an “outsider”: part participant, part voyeur, forever lacking this essential element of “withness” (Shotter, 2006, p. 585)…. The need for SAP [strategy as practice research] to engage with strategists as research partners, not merely participants, has been recognized since the inception of the field. (Herepath, 2014, p. 875, most citations removed)

A second way in which the strengths of both involvement and distance can be combined is what we call “validating involvement” (bottom of Figure 1—see the section on voice). This is nicely illustrated by the findings section of Pratt’s (2000) above-noted study of Amway distributors in which personal experiences and more objectivized inferences are presented in separate paragraphs but are woven together within the text in such a way that the experiences of respondents serve at the same time to legitimate the more personalized experiences of the author, while the author’s personal experiences make the more abstract insights and observations come alive. Weaving these elements together in a credible way clearly needs to be done skillfully so that the reader can clearly distinguish when an author is taking distance and when he or she is bringing in his own personal perspective.

The mobilization of multiplicity or “distributing involvement” (bottom of Figure 1—see the section on voice) is another way of effectively signaling both involvement and distance. This is explicitly revealed, for example, in de Rond and Lok’s (2016) piece on medics operating in the war zone in Afghanistan. The authors note how de Rond’s personal experiences and conversations with Lok (the outsider who had a stronger capacity for distance) served as initial inspiration for the emerging categories, but it was then the systematic coding of both authors of the field notes and respondent’s words that led to the final analysis. Similarly, the explicit inclusion of the separate voice of Thomas (the dean involved) within Gioia et al.’s (2010) study of the formation of a new identity contributes to enriching the overall account.

Toward Expanding the Repertoire

Our central argument in this article is that there must be more recognition of the nature of the involvement paradox in qualitative research, and that it needs to be addressed more directly and creatively in our research outputs. Our article illustrated several ways this can be done and identified a series of choices researchers may make in representing their involvement. However, in order for authors to feel comfortable in avowing involvement, they need to know that their contributions will not be automatically shot down as soon as they offer any hint of participation that reaches beyond the fly on the wall. Our own experience as authors, editors and reviewers suggests that although there may be questions raised (as demonstrated in the opening quotations), there is also more openness to more intense approaches to involvement than many authors seem to assume. Reviewers do not like feeling that they do not have the whole story and want explanation and reflection. Yet all the papers referred to in the opening quotations were published. At least in the first two cases, though possibly not in the third based on Watson’s (2011) analysis, they were better and richer in their representations of researcher involvement than they would have been otherwise. The involvement paradox will always be there whether we recognize it or not. We argue that it is time to join together as authors and reviewers in accepting it and in finding creative ways to navigate its complexities.