Contextual expertise and the development of organization and management theory

Generating theory from research settings requires researchers to adeptly engage with the social intricacies of the field. They need to develop a contextual understanding by gaining in ‐ depth insights into the setting, while retaining a critical distance from it. Researchers must practice this along the entire research journey, from site selection, through data collection and analyses, to theory development and explanation. However, contextual expertise has been assumed rather than actively considered as a critical component in organization and management studies. We lay out steps for understanding the nature of contextual expertise, offer systematic advice for research projects and outline methodological practices across the research process to choose, capture, comprehend, convey, and confirm contextual specificities. The growing social complexity of research settings renders the development of contextual expertise increasingly important for the generation of enriched and more diverse theory. It also has implications for organization and management studies as a scientific community.


Introduction
Discussions of the role of those who conduct research on the outcome of their work has centered on researcher involvement and bias, long a staple of philosophical debates across the social sciences (e.g., Kuhn, 1962;Berger and Luckmann, 1967;Louis and Bartunek, 1992;Cunliffe and Karunanayake, 2013;Jones and Bartunek, 2019). What has not been subject to the same scrutiny, but is at least as important to understanding how theory is developed in empirical management research, is the capacity of investigators to engage with the social complexities of research settings. An indication of the importance of this is provided by Jarzabkowski et al.'s (2015) reflection on how their work on the global reinsurance industry was made more complicated in ways that they had not anticipated because of a need to collect data in different languages and accommodate insights generated in different geographical locations. In addition to language and geographical differences, research complexity may also be increased by the need to accommodate diverse cultures, institutions, organizations, professions, and sectors. Evidence of this has arisen from work spanning European healthcare systems (Kyratsis et al., 2016), soccer ball manufacturing in Pakistan (Khan et al., 2007) and Australian professions (Liu, 2017).
While contextual expertise has been ignored, taken for granted, or deemed too obvious for consideration, the cases above demonstrate the importance for researchers to be able to cope with the distinct characteristics of the empirical setting. Not only does it influence the quality of case selection, data collection and analysis, but also the ways in which researchers draw, present and explain theoretical inferences. Thus, the make-up of the research team is directly related to research outcomes. However, personal connections to one's research tend to remain taboo topics in organization and management studies (Anteby, 2012;Jones and Bartunek, 2019) and their influence on research outcomes is correspondingly under-developed.
Thus, our purpose with this article, in line with Lee's (2020) recent exhortations, is to extend the methodological research agenda by bringing to the fore the influence of contextual expertise and its implications for the execution of complex research projects. We define contextual expertise as the ability to have both competent as well as critical engagement with the full range of contextual specificities, some of which may only become apparent during data collection or analysis. We see this as a balancing act between gaining in-depth insight into the empirical setting while retaining a critical distance from it. This foregrounds the inevitably distinctive characteristics of researchers and the intricacies with which they must contend. While work has shown that researchersboth individually and as teamshave biases, views, and values that shape their research activity (Kuhn, 1962;Berger and Luckmann, 1967), we re-focus on the relationship between researcher and researched (Van Maanen, 2011). In so doing, we challenge the implicit assumption that researchers possess equivalent skills and that, if a robust process is followed, the findings and subsequent theory will be a credible reflection of the research undertaken. As interpretivist researchers, we are not claiming that having a certain combination of individuals in a research team would allow them to unveil the 'true' understanding of what is occurring in a research context. Rather, we need to develop our understanding of how different skills within the research team will be correlated with the nature of the theoretical inferences drawn. Further, it is apparent that, as the complexity of the research setting increases, so does the potential for missing, misreading, and/or misrepresenting data.
We contribute to methodology advancement, theory development and community building. First, we offer insights into how to improve contextual expertise by outlining methodological practices for researchers across qualitative data selection, collection, analysis, presentation, and explanation to more completely capture, comprehend, convey, and confirm contextual specificities. Second, we show that the consideration of contextual expertise leads to richer, more robust and heterogeneous theory development. Third, we highlight that embracing complexity as well as diversity requires coordinated efforts by the scientific community of organization and management studies scholars. We thus lay out implications for the recruitment and training of scholars, for collaboration and teamwork, and for reviewing and evaluating research.

Methodological rigor and the role of researchers
Both our understanding and practice of what constitutes methodological rigor have progressed extensively in organizational research (Shah and Corley, 2006;Gioia et al., 2013). While this has been well documented in quantitative studies with advances in the use of data collection and statistical analyses, the same is also true in qualitative research. For example, the so-called Gioia method has emerged as a popular framework for demonstrating the analytical processes involved in moving from raw data to theoretical extrapolation (Gioia et al., 2013). Other scholars have developed alternative ways and templates for achieving similarly compelling outcomes (see, e.g., Eisenhardt, 1989;Langley and Abdallah, 2011). However, in organization and management research we have neglected the competencies and characteristics of those who conduct research vis-à-vis diverse research settings. This is somewhat surprising given the voluminous literature on the role of the researcher in social science scholarship. For example, more than fifty years ago Berger and Luckmann (1967) famously asserted that reality and meaning are socially constructed while, more recently, Mantere and Ketokivi (2013) have noted the limits of the cognitive capacity of researchers: reasoning is neither exclusively computational nor researcherinvariant. Thus, it is perceived that researchers are more likely to operate within paradigms rather than across them (Kuhn, 1962;Pfeffer, 1993) and are heavily influenced by their own cultural, ethical, historical, political, social, and technical impregnation (Latour and Woolgar, 1979;Knorr-Cetina, 1981). Researchers commonly use intuition, emotion and guesswork rather than reasoning based on a strict application of rules (Lipton, 2005;Swedberg, 2016;Harley and Cornelissen, 2019). This raises questions for how we determine the quality of scholarship. Amis and Silk (2008, p. 456), for example, contend that the evaluation of research 'cannot be divorced from the political, axiological, ontological, and epistemological orientations of the scholarship'.
This body of work squarely locates the researcher within the research project, and further points to the importance of the capabilities of researchers in drawing theoretical inferences from their fieldwork. Yet, 'the general understanding of how scientists reason and formulate explanations is surprisingly limited' (Mantere and Ketokivi, 2013, p. 70). Harley and Cornelissen (2019, p. 8) note that '[r]easoning involves the conscious and deliberate use of relevant assumptions, explanatory principles, and observations'. This implies it is the role of the researcher to infer what they deem relevant in a conscious and deliberate way. As there are "multiple possible theoretical inferences that could be made" (Harley and Cornelissen, 2019, p. 8) from any data set, the competence and contextual understanding of the researcher is even more important. Not only do methods influence the types of questions scholars can ask and the theories they can develop, so also do the expertise and identity of the researchers.
Other work has more explicitly considered the role of the researcher in the technical execution of the research. Gold (1958), for example, presents the impact of varying degrees of participation versus observation during data collection. Alderfer and Smith (1982) show how researchers may themselves intervene in the setting to obtain relevant data and thus move between basic and action research. Others have outlined different levels of involvement by ethnographers (Orbe, 2000;Cunliffe, 2010;Langley and Klag, 2019) that have led to the development of research approaches such as autoethnography (Karra and Phillips, 2008;Boylorn and Orbe, 2014) and para-ethnography (Islam, 2015). As Pratt (2009, p. 859) noted, and as these articles convey, researchers should be very clear about their 'position in the field'. While useful, in each of these articles, the competency and qualification of those doing the research is assumed, ignored or conflated with their involvement.
Adjacent disciplines such as anthropology and ethnography have a more established tradition of exploring the role of researchers and research teams (Gottlieb, 1995;Kennedy, 1995;Erickson and Stull, 1997;Hafernik et al., 1997;Woods et al., 2000). This work conveys the importance of deep researcher involvement with empirics through methods such as thick description (Geertz, 1973), and also the significance of the increased complexity of the field, for instance due to multi-site research (Marcus, 1995;Hannerz, 2003). Some offer instrumental warnings and advice for team work (Erickson and Stull, 1997), while others are critical about actual team practices. For instance, Mauthner and Doucet (2008) criticize collaborative academic modes of production for dividing and distributing knowledge production across the team instead of practicing team interaction across the research process, although there are positive counterexamples (e.g., Creese et al., 2008). Despite this work, the critical examination of expertiserather than familiarityalong with its methodological and theoretical significance, requires further exploration, particularly in organization and management studies.
Linked to the role of the researcher in determining research outcomes, calls have emerged for greater reflexivity when it comes to determining the nature of knowledge making and knowledge dissemination (Calás and Smircich, 1999;Cunliffe, 2003;Burawoy, 2013;Jarzabkowski et al., 2015). These have been accompanied by increasingly sophisticated and fine-grained techniques to incorporate reflexivity into research practice (Reissner, 2018). Reflexivity is particularly important when scholars work within hyphen-spaces (Cunliffe and Karunanayake, 2013) at the intersection of the insider-outsider duality between researcher and organizational roles (Griffith, 1998;Brannick and Coghlan, 2007;Dwyer and Buckle, 2009). For instance, participant observers take both notes and action (Spradley, 1980;Pass, 2020). We know that such involvement can be beneficial (Spradley, 1980;Cunliffe, 2010). At the same time methodological work has addressed the concern that such scholarly involvement is a double-edged sword that facilitates deeper insights but also potentially gives rise to deepening blind-spots as scholars become too familiar with the settings (Langley and Klag, 2019).
Such work on reflexivity is important yet not sufficient for addressing the issues we describe here. While it can help to draw out potential biases and blind-spots as well as alert us to potential deficits in our approach or methods of data analyses, a lack of contextual understanding with respect to imperatives such as culture, field, history, or language cannot be resolved through reflexivity alone. While reflexivity is important for confronting our own ideological biases and assumptions, a lack of contextual understanding also negatively impacts our theorizing. Reflexivity is about being very clear about assumptions, experiences, and personal characteristics, while contextual expertise is about having more and deeper understanding about the setting. Of course, reflexivity may lead to an awareness of the need for more expertise.
Irrespective of whether we consider researchers to be more or less subjective, they are certainly more or less able to appreciate what is unfurling in specific research settings. If we wish to improve theoretical inferences, it is insufficient to merely use multiple methods or collect more data without a certain understanding of the context. In other words, even if scholars can control for some of their biases, if they are unable to satisfactorily comprehend the empirical setting, their theoretical inferences, judged from their own ontological and epistemological standpoints, will likely miss or misread important data. Jarzabkowski et al. (2015) explained how their research team went about dealing with the problems of assessing a complex empirical setting, noting how they had to put integrative structures in place to facilitate team interaction as they developed a greater understanding of the complexity of their undertaking in their ethnographic study. This is central to our work here. In terms of team composition, Zyphur (2009, p. 686) suggested that 'populating research teams with at least one methodologist could greatly benefit the team as a whole'. In terms of data engagement, Gioia et al. (2013, p. 19) described how they use a member of the research team as a 'devil's advocate', who brings an outsider perspective to the data. While steps such as these are undoubtedly useful, the impact of the research team on the veracity of site selection, data collection, data analysis, and theoretical extension remains largely opaque.
As with the examples provided from Zyphur (2009) and Gioia et al. (2013), studies of method have focused almost exclusively on analytical rigor, the relationship between data and theory, or the methodological approach and ideological bias of researchers. The characteristics of the research team and their impact on collecting, analyzing and developing data-driven theory are rarely considered. For instance, Creed received formal training in religious studies at a divinity school, but does not mention his prior expertise in an article on identity work in the Anglican Church (Creed et al., 2010). Dacin et al. (2010Dacin et al. ( , p. 1399) write about 'excellent access to the research site', but do not note how their personal affiliations and experiences with Cambridge University helped them to put college dining in context. Kornberger et al. (2018) had been living and working in Vienna for some time when they commenced their work on the refugee challenge, but do not mention how this helped them to understand how the City's leaders acted differently in this case compared to other challenges that they have faced. In each case, such details are presumably deemed less relevant then the overall rigor of the research process. Thus, while scholarly biases are sometimes acknowledged, interpretive expertise, despite its centrality to empirical interpretation and theory development, is implicitly assumed.

Social intricacies and complexities
A research setting may be theoretically 'misrepresented' for a number of reasonsbecause it is underexplored and ignored, because it is only examined with a partial lens, or because its complexity and interlinkages are not considered. To capture and theorize social reality, we need to consider social intricacies and complexitiesthroughout the entire research process. There have thus been calls for scholars to extend the scope and breadth of empirical work. Scott (2005, p. 478), for example, has lamented that an 'embarrassingly large proportion' of theoretical understandings are based on data from US organizations. Similarly, Üsdiken (2014) notes that management scholarship is dominated by US-style research. More generally, scholarship often ignores and misrepresents certain groups, concerns, and identities, including gender, race, disability, class, and sexual orientation, and their intersections (Crenshaw, 1989;Holvino, 2010;Bothello et al., 2019). Or, if they are not ignored, they are othered, so that the body of theory is not impacted (Nkomo, 1992). As Minnich (2010) pointed out, it is difficult to add novel knowledge if existing knowledge is already defined as the whole. We hence need to 're-vision the very way we "see" organizations' (Nkomo, 1992, p. 505; see also Bansal et al., 2018) that is, our 're-presentation' of social reality. This re-visioning requires a widening of perspectives, including diverse, complex, and interrelated identities, categories, processes, and systems (Dhamoon, 2011) as well as a broadening of methodologies by considering senses such as orality (Willox et al., 2013), visuality (Bell and Davison, 2013), and scent (Gümüsay et al., 2018). As work on intersectionality (Castro and Holvino, 2016) and multimodality (Boxenbaum et al., 2018;Höllerer et al., 2018) highlights, these concerns should not be understood as operating in isolation but rather as interwoven. Further, we need to recognize not only the diversity of interlinked social stratifications but also the multiplicities and connectedness of cultures, fields, institutions, organizations, and professions. We thus concur with Suddaby et al. (2011) that organization and management theories do not sufficiently capture the empirical complexities of organizations.
If we are to respond to such critiques, then we will need more multifaceted research designs that enable researchers to capture complex data that reflect what happens in organizations across different settings. This issue has been rendered more acute by calls for scholars to investigate societal concerns such as 'grand challenges' that are often viewed to be so complex as to be intractable (Ferraro et al., 2015;George et al., 2016;Gümüsay et al., 2020a). This further emphasizes the need for research teams to reflect these demands. While we assume that researchers are able to engage with the research setting independent of its complexity, this is problematic. If the fit between method and object of inquiry is important (Edmondson and Mcmanus, 2007), the fit between researcher and research setting is equally so.

Contextual expertise
Thought is bound by being, which has given rise to the renowned idiom that reality is socially constructed due to the 'existential determination (Seinsgebundenheit) of thought ' (Berger and Luckmann, 1967, p. 16, italics in original). Social construction is dependent on those who construct it: it relies on their reasoning in thought (Mantere and Ketokivi, 2013), methodologies (Law, 2004), and underlying substance or '"social location" of thought' (Berger and Luckmann, 1967, p. 19). For organization and management scholars, this signifies their contextuality when theorizing. For the field, the social construction of its theories is shaped by the underlying cognitive locationality of its collectivity of members.
We construe contextual expertise as the ability of an individual researcher, or more likely a research team, to be able to generate a depth and breadth of understanding of the empirical context across the entire research process (competence) while also maintaining the critical and respectful distance required to grasp the relationship between concepts (criticality). It complements work on involvement (Anteby, 2012; Langley and Klag, 2019) and recognizes that researchers are often engaged in research without having acquired the necessary depth of understanding of the research site, nor the skills and competencies to collect the requisite data. Contextual expertise is not about the degree of involvement or representation of presence, but about the ability to sufficiently: (1) choose to engage; (2) capture; (3) comprehend; (4) convey; and (5) confirm the specificities of the research setting. As shown in Table 1, the expertise of the research team should be systematically incorporated along these five dimensions.
Similar to researcher involvement, contextual expertise requires an almost paradoxical proximity to and distance from the setting. In contrast to this, however, it is not about the familiarity with the research setting but the ability to avoid overlooking key data points and their meaning. Independent of the involvement in the setting, the researcher or research team can acquire contextual expertise through critical engagement with contextual factors such as culture, gender, language, race or profession, without entering the specific field itself. For instance, in a discussion, a senior colleague told us that when he commences a new research project on professional service firms, he does not start at zero but takes with him his experience of hundreds of interviews, numerous days of observation, and thousands of hours of critical thinking and discursive exchange in and about similar organizations.
Scholarly work on class, gender, race, religion, and sexual orientation has long grappled with the intersection of the researcher and the researched highlighting how much the organization and management literature has neglected the perspectives of less privileged groups (e.g., Calás and Smircich, 2006). There have been correspondingly strong calls to not just focus on marginalized groups as objects of inquiry but also to build a research team that reflects more completely the diversity of social reality (Harding, 1987;Eichler, 1988;Merriam et al., 2001;Archer, 2002;Berger, 2015). For instance, Alderfer (1982) noted that prevailing societal race relations influence researchers' racial identities and experiences. As the recent #BlackLivesMatter protests highlight, even if white and black people inhabit the same physical space, they still experience a very different social space. Similarly, Sato (2004) spoke of her positionality in the field and its impact on her research subjects and Liu (2017) explained how her shared identity with her Chinese Australian interviewees helped her to engage in shared reflexivity and dialogic engagement. We argue that this rationale needs to be applied to organization and management studies but that on its own it is insufficient. We assert that contextual expertise is related to but is not the same as researchers' identity (for a related discussion on conflating individual behavior and expertise with collective identities, see Hark and Villa, 2020).
Thus, while competent researchers do not necessarily need to mirror the attributional characteristics of the object of inquiry, a deeply ingrained understanding of a setting due to one's own identity can certainly be beneficial. We are reminded of the tale of an ethnographer who spoke to an inmate about doing research on prisoners. The prisoner asked him: you can leave any time you want, right? After the ethnographer affirmed, the prisoner responded: then you can never understand what it means to be a prisoner. In complex social settings, an intimate understanding of identity characteristics such as age, class, disability, gender, national identity, race, religion, or sexual orientation may be hard or impossible to acquire. However, research teams require diverse expertise, which is potentially overlapping and sometimes connected to research team identities, to capture a more complete empirical and theoretical picture of their setting.

Selection, collection, and analysis
Site selection, data collection and analysis require a careful approach towards a research setting. By ignoring contextual imperatives, researchers will inevitably fail to theorize certain social realities. For example, Holvino (2010, p. 248) has argued that people whose identity crosses the boundaries of constructed groups have been ignored as undocumented and 'hidden stories at the intersection of race, ethnicity, gender, class, nation and secularity'. To address this, scholars should shift their 'focus on particular social groups at neglected points of intersection' (McCall, 2005(McCall, , p. 1774). This requires intentional and competent zooming in. It also involves disentangling intersectional social identities that entail contradictory and colliding elements (Nash, 2008). When Liu (2017) identifies mythtapping, mythkeeping and mythbusting as processes deployed by Chinese Australian professionals to engage with stereotypes, she effectively shows a double understanding of these practices. She identifies characteristics that are considered Chinese and then highlights that these myths are consciously enacted and sometimes strategically resisted. In a recent article, Gümüsay et al. (2020b) describe the methodological collaboration process they used when researching the founding of the first Islamic bank in Germany. They state that as a team, they were able to combine competence of and distance from the setting that would prove crucial for developing empirical and theoretical insights. In the article, they depict how the first author was able to understand the research setting due to his prior knowledge of language, religion, and the German socio-cultural environment. During data collection, bank members used English, German and Turkish regularly in their day-to-day activities and also referred to important concepts in Arabic. While the first author was able to pick up on these because of his command of each of the languages used, it was his co-authors' probing questions that revealed that certain phrases in particular languages were used strategically to convey particular messages. This enabled them to uncover conceptual connections that were theoretically meaningful. Further, the role of religion in the bank was only accessed because of the first author's prior religious studies. Symbols, expressions, principles, and certain practices had implicit religious connotations. The researchers would likely not have noticed these cues without a specific knowledge of Islam. Finally, they explain that the significance of certain bank practices used to adapt to the German socio-legal setting would not have been fully revealed without the cultural understanding garnered by the first author during his time living in Germany and the investigative questions by his coauthors. In this case, then, the research team combined both comprehensive expertise with critical distance to engage with the Islamic banking setting on topics related to language, religion, and culture that would have in other circumstances potentially been left unexplored. This begs the question of what a team with other characteristics would have found and, more generally, how differences in contextual expertise have the potential to substantively influence theory development.
A further example is provided by Smets et al.'s (2012) examination of the merger of a German and an English law firm. The first author was fluent in both English and German and had lived in Germany and England for extensive periods of time. He was thus able to draw on skills acquired before entering the field that were vital in allowing access to the linguistic and cultural nuances that were required to understand the ways in which the merger was interpreted within each firm. The other two members of the research team were not directly involved in data collection, but as experienced scholars of professional service firms they were able to offer insights into the practices that emerged during the project. Without this combination of competent but critical engagement, and in particular without the apprehension of the German and English contexts, this comparative piece of research would have been unable to deliver the level of empirical insight and theoretical extension that was provided. These examples also highlight that contextual expertise is not additive but deduced through team interaction and development.
Despite the importance of the type of contextual understanding described above, existing organization and management studies articles commonly take the expertise of the research team for granted. The underlying assumption is that provided a robust process is employed, the findings and subsequent theory will be a credible reflection of the research undertaken. This holds for all scholarly work, irrespective of it embracing a positivistic, post-positivistic, interpretivist or other paradigmatic perspective. It is, for example, apparent in interpretive approaches such as the standardized Gioia design (Gioia et al., 2013) and in positivistic or post-positivistic ones such as Eisenhardt's comparative case method (Eisenhardt, 1989). However, as we showed above with reference to Gümüsay et al. (2020b) and Smets et al. (2012), the contextual expertise of researchers, far from being neutral, has a significant impact on the theoretical inferences that are drawn from the data.
The importance of contextual expertise is also attested by Langley's (1999, p. 691) observation of the importance of making 'uncodifiable creative leaps' when moving from data analysis to theory development. Without an understanding of the research site, these creative leaps can become fatal jumps. Contextual expertise can help in two ways. First, an in-depth understanding of the empirical site should result in shorter leaps, increasing the likely correspondence between observation and theory. Second, the less-speculative ground from which researchers leap will be firmer, resulting in more compelling insights on which to base theoretical inferences.
Contextual expertise can be situated with an individual or result from the combined skills of a research team engaged in collaborative data gathering, analysis, and theorizing. For instance, in a multi-country study, one researcher may speak one language used in an empirical setting and a second another language used elsewhere. This is very much in line with the more general notion of the complementarity of team skills (Katzenbach and Smith, 1993). Jarzabkowski et al. (2015) were faced with this situation when they collected data from different national settings. Lacking the requisite language skills and being aware of their 'linguistic positionality' (Cormier, 2018, p. 328), the project leader 'parachuted' in research assistants fluent in French and German. This gave the team a depth of understanding that would have otherwise been lacking. Alternatively, multiple researchers may speak the same relevant languages and can therefore code independently of one another for data in multiple languages. In both cases, the composition of the research team improves the analytical engagement with the research setting in collecting and analyzing the data. However, in the latter case, the researchers can critically evaluate each others' interpretations, potentially increasing the depth and robustness of the analysis. Such differences in complementary skillsets is also emphasized by Kowal et al. (2017) who first used a service provider and later graduate students and postdoctoral fellows as cultural and linguistic brokers. These students and fellows, they argue, had the additional benefits of having formal research training, which fed back into how they assisted in language translation and interpretation.
If no members of the research team possess a deep understanding of context, the team is likely to miss and/ or misread contextual specificities and have less appreciation for empirical nuances. For instance, they may not comprehend the significance of a cultural symbol with multiple meanings. They may also, of course, face linguistic difficulties if some participants in the setting use languages that the team does not understand. Thus, the team may decide to only collect data from research sites in which their home language is used while ignoring those places in which different languages are spoken, thereby likely diminishing the depth and richness of any empirical and theoretical inferences. As organization and management theory has neglected certain groups, topics, and places, and their intersections (Holvino, 2010;Rodriguez et al., 2016), it is important for theorizing to become more reflective of social reality by accounting for its empirical heterogeneity (e.g., Van de Ven, 2011). With the corresponding research designs having to cope with the difficulties of engaging with transnational organizations, multi-site research, and global interconnectedness (e.g., Jarzabkowski et al., 2015), so the issues that we raise here will become increasingly germane.

Presentation and explanation
Reasoning is a social process among members of a scientific community (Toulmin, 2003). Research teams not only need to produce robust findings; they also need to outline, explain and justify their results to their scholarly community, particularly to the editors and reviewers who serve as gatekeepers to research journals (Easterby-Smith et al., 2008;Pratt, 2008;Jonsen et al., 2018). Golden-Biddle and Locke (1993) distinguish among three dimensions that are central to convincing reviewers of the viability of texts: authenticity, plausibility, and criticality. While their work is focused on ethnographic research, their insights are also relevant for other types of qualitative research. Authenticity is more than simply showing first-hand experience. Rather it means that 'the text conveys that the researchers grasped and understood the members' world as much as possible according to the members' constructions of it' (Golden-Biddle and Locke, 1993, p. 599). We concur with this assessment. Importantly, the term 'as much as possible' not only depends on showing a deep immersion in the setting that provides 'a genuine experience; (Golden-Biddle and Locke, 1993, p. 599), but also on presenting the researchers' contextual expertise. This relates to the research team's ability to open the 'black box' of research team interaction to convey their competence and critical distance. It is important not just from a data collection standpoint but also with respect to convincing editors and reviewers of the suitability of a work for publication. Both the individual and collective credibility of researchers increasingly needs to be established to pass scrutiny and thus to become part of our published body of work.
However, while research methodologies outline criteria for what constitutes a credible argument (Locke and Golden-Biddle, 1997), we have much less insight into what constitutes a credible arguer. Given the constraints on the length of journal articles, researchers are forced to make choices about what to explain, and what to leave unsaid. The vast majority of researchers focus attention on methods of data collection and analysis, particularly in North American journals in which extensive space is accorded to explanations of methodological decisions and processes. There is also often a short section describing the research context. However, there is almost never any description of the contextual expertise of the research team. As a consequence, decisions about researcher suitability are based on their skill in explaining the methods used, while their expertise is implied or ignored. Very few scholars have laid out why they have sufficient competence to unveil the complexities of a research setting. There are notable exceptions, such as Bartunek's (1984, p. 357) description of her work on a religious order: 'My interpretation is based in part on my experience as a member of the order since 1966 and as a consultant for the restructuring after the decision was made to do it'. However, these are rare and often tend to provide little more than a brief account of how one or more researchers gained technical understanding of a research context. For instance, Lok (2010Lok ( , p. 1310 described how he addressed his lack of understanding of shareholder value: "I first immersed myself in the shareholder value literature, such as value management textbooks, academic and media articles, and books discussing the logic of shareholder value and its implications, in order to develop a feel for the structure of shareholder value discourse and its historical development in the U.S. and the U.K. I then focused more specifically on the role of the media and the U.K.'s corporate governance reform reports in reproducing and legitimating a particular version of the logic of shareholder value in the U.K'. Tracey et al. (2011, p. 63) state that a member of their research team was a manager from the organization, Aspire, which formed the focus of their study. This is a case of academic-practitioner or insider/outsider collaboration (Louis and Bartunek, 1992;Bartunek, 2007). However, the benefits of his involvement for the research project are implied rather than explicitly laid out: 'The third author led an Aspire outlet from March 2001 until July 2004, which allowed him to witness firsthand the series of events that underpinned development and subsequent breakdown of the venture'.
Commendable as the examples above are, the brevity of such accounts is problematic. As complexity is growing due to the diversity of cultures, industries, languages, and so on, it is becoming increasingly important, and challenging, to convince reviewers that the research team has the contextual expertise to construct theoretical inferences. Research teams could address this by outlining how diverse expertise is iteratively developed and combined, presenting a genealogy of team interaction as the unfolding of research 'team work'. For instance, research teams need to be clear on who translated primary data such as interviews as well as when and how (Cormier, 2018). In parallel, we believe that reviewers and editors need to respond to the increasing complexity of research sites with greater scrutiny of the research team's capabilities. Research teams that are not able to sufficiently convey their analytical fit with the research site may be unable to convince editors, reviewers and readers that their findings have veracity. As we noted earlier, given that the challenge of establishing contextual expertise increases with social complexity, so a likely outcome of this will be that we observe less engagement with complex settings as scholars see themselves unfit to pursue and convey research in certain settings. Less engagement and less acceptance will result in theoretical inferences being based on an underrepresentation of the complex organizational settings that typify the field.
Research teams need to be selective in disclosing information not only on the expertise-setting fit, but also in accounting for existing biases. For instance, paradoxically and erroneously, reviewers may consider men to be more credible than women, ceteris paribus, when doing research about gender in management, because they may think that women have an ideological agenda (Jané et al., 2018). Thus, Holvino (2010, p. 266) notes that an '"outsider within"' status is not such an advantage, for our knowledge production becomes suspect when we are caught in between the power relations of our disciplines, research institutions and academic practices and the communities and women we seek to give voice to through our research'. Presenting contextual expertise thus would focus on competency without necessarily disclosing identities, and would in some cases consciously hide them. For instance, women may not invoke their own gender but rather their expertise on gender issues through direct experience and exposure to previous work on gender and related fields.
Given that a paper's scholarly impact formally starts with publication, post-acceptance formats such as commentary from editors, authors, scholars, or other forms of engagement such as acknowledgements, blog posts, interviews, and summaries are also increasingly appearing. Here, we notice a second layer of presenting and confirming contextual expertise that influences the dissemination. For instance, the Academy of Management Discoveries populates articles with author's voice commentary, a video summary and an editor's comment. In such a commentary on an article by Whiteman and Cooper (2016), the first author mentions the characteristics of the two partners doing the field research with her: one came from the specific regional area and the other was an eminent professor with contextual expertise.

Discussion
In previous sections we have described the potential challenges for scholars facing socially intricate and complex research settings. Here we move on to discuss the implications of contextual expertise for qualitative research methods, theory development, and community practices across the research process (for an overview, see Table 2).

Methodological implications
We suggest specific methodological practices that can be used throughout the entire research process. Researchers should engage in pre-data gathering periods of knowledge acquisition and development with regards to pertinent characteristics of the empirical setting. Those with the requisite expertise can be recruited, or existing team members can be trained. This latter strategy can involve spending a prolonged amount of time within different industries or in different countries, taking courses, developing technological literacy, or even learning a new language or dialect. For instance, Cole (2015) trained in martial arts for 15 years and was fluent in Japanese and English prior to his research on a Japanese martial arts organization. Of course, such competency acquisition is resource and time consuming. Even after an extended period, it may not be possible to properly learn and capture the intricacies of professional (or prison!) settings, let alone the different languages or cultural nuances of geographically diverse organizations or sub-units. A further example of the importance of this is provided by Brandl and Bullinger (2017) who, having both lived in Vienna for some time, were able to draw on their contextual expertise to understand the significance of how a CEO shifted from what they call 'standard language' to a local Viennese dialect to craft a critical yet lenient response about the performance of the human resources department, while the human resources manager was present. In addition to the capabilities of individuals, we have stressed the importance of research team composition, polyphony, and practices in the overall understanding and analysis of the research setting. In particular, complex problems often surpass the limited capability of individual researchers and thus require acquisition of knowledge by other researchers, practitioners or research assistants (Van de Ven, 2007Avenier and Cajaiba, 2012;Kowal et al., 2017;Schumacher, 2018). While research has shown how teams can share data gathering work and coordinate data analysis across jurisdictions, boundaries, and fields (Köhler et al., 2012;Jarzabkowski et al., 2015;Smets et al., 2015), we stress the significance of a fit between team capability and research setting complexity. This also requires a reflection about how research teams share and exchange knowledge (Mauthner and Doucet, 2008), for instance through a team-based 'interpretive zone' (Wasser and Bresler, 1996). Both individual and collective capabilities increase the likelihood that theoretical inferences will reflect the research setting. Complex settings almost inevitably require team-based research projects in which the compositional diversity of expertise of the team somewhat matches the empirical context. In fact, it has been pointed out that there is a shift underway towards more team-based research and a growth in team sizes (Wuchty et al., 2007). In practice, this should lead to further considerations of team compositionsincluding the role of research assistants and organization insidersand their expertise vis-à-vis a particular research setting.
Developing context-research team fit requires a process of identifying relevant expertise and complementarities to achieve a sufficient level of intimacy as well as considering team dynamics and commitment to make the project feasible. Such teams need to subsequently structure, reflect on, and present their team interactions. By taking into account complementary skills before, during and after the gathering of data, the research process becomes more robust. With increasingly complex research settings, research team composition becomes more relevant to bring in the complexity from the setting as well as capture and comprehend data for theory development.
At the same time, larger and more diverse teams may face a tradeoff. While they may be better equipped to examine the research setting, such teams may also find it more difficult to coordinate and reach agreement amongst themselves (Wuchty et al., 2007). Further work needs to attend to these intra-research team processes and their impact on data collection, analysis, and theorization.
Moreover, complex research settings require engagement that is not only competent but also credible. Extant work has looked at methods to collect, analyze and convey research results (Miles and Huberman, 1994) and to increase their credibility through, for example, member checking, prolonged field engagement and triangulation (Lincoln and Guba, 1985). We complement this body of work by detailing the significance of the credibility of researchers and the evaluation of their capability fit with the research setting. In other words, while scholars have previously assessed the importance of credible analysis (Gibbert and Ruigrok, 2010), we underline the need to similarly consider the credibility of expertise across the research team.
In practice, a description of relevant research team characteristics needs to feature more prominently in the methodology section. For instance, Danner-Schröder and Geiger (2016) outline in detail the involvement of the first author in the data setting, but do not mention the second author's expertise on catastrophe management. While  researchers may not describe authorial capabilities due to space constraints, taken-for-granted professional norms and the need for author anonymity to preserve the double-blind review process, this should not lead to neglecting the research team characteristics and their contextual fit to conduct the research project and evaluate the findings. On the contrary, research outputs should specify how the skills of the research team were particularly conducive for the collection and examination of the data, and how team members confirmed, complemented and questioned findings due to overlapping contextual expertise. In addition, post-publication commentary and the presentation of additional representative material and artifacts can further strengthen the perception of contextual expertise. This is likely to not only further legitimize the article but also increase discourse about and citations of the piece, thereby assisting in establishing the article within a scholarly community's body of work.

Theoretical implications
Theorizing is based on a 'recursive movement of zooming in and zooming out' (Nicolini, 2009, p. 120). In this respect, a lack of contextual expertise can lead to at least two undesirable related outcomes. First, scholars may avoid zooming in on certain empirical phenomena. As research complexity places increasing demands on research teams, it becomes more likely that scholars will try to minimize the problems of data collection and analyses, and study settings with which they are familiar. The outcome is likely to be more incremental theoretical advancement based on a limited range of research settings instead of enriched theory that captures complex and diverse settings. Second, scholars may not be able to zoom out to extract and educe findings from settings. An incomplete comprehension of the research setting due to cultural, linguistic, or other contextual challenges likely means that relevant relationships between concepts are not captured in theory development.
Malinowski, for example, one of the earliest cultural anthropologists, admitted to 'inadequacies in all his material, whether photographic or linguistic or descriptive' (Firth, 1957, p. 79). Burawoy (2013) has also reflected on how problems in his research designs over a career of forty years resulted in 'fallacies' that, at times, led him to incomplete contextual appreciation, analyses, and theoretical inferences. More recently, Bothello et al. (2019, p. 11) stressed how a lack of reflective capabilities and context-driven research have led to patronizing vocabulary and pejorative labelling of non-Western countries' governance practices. A lack of contextual expertise results in underexploring and ignoring a setting, misunderstanding its intricacies, and not capturing its complexity and interlinkages.
While it is well established that theory and method are interlinked (Van Maanen et al., 2007;Cornelissen, 2017), contextual expertise emphasizes in particular the role of the researcher and the setting and their connection to theory and method. Rather than a dyadic or oppositional relationship (Bourdieu and Wacquant, 1992), we note a quadratic relationship between theory, method, researchers, and setting. In line with our suppositions, Grodal et al. (2020, p. 37) have argued that '[t]he theoretical insights that are drawn from the data are thus not simply "given" in the data but actively constructed by researchers to address puzzles that they find interesting and important'. This is then not simply a question of adding researchers to a team, but quite a dynamic process to acquire certain expertise while retaining workability to engage theory with the setting in a methodologically thorough manner. In fact, as Grodal et al. (2020, p. 37) continue '[o]ther researchers might view the same data very differently: As we know, many valuable insights emerge from the interaction of different individuals in distinct fields'. The types of methods as well as the epistemological position of the research team members are thus also likely to impact what contextual expertise is sought, how it is valued and how it is used for theory development. Overall then, this engages with the problem of 'faulty generalization and noninclusive universalization' (Nkomo, 1992, p. 489). It leads to increased theoretical richness and more heterogeneous theory development, as more scholars zoom in and engage with as well as zoom out and educe from socially complex and diverse settings.

Implications for organization and management studies as a scientific community
Our line of reasoning does not only have implications for individual manuscripts and projects, but also for how the loosely coupled organization and management studies community practices research. First, as the neutrality of the social sciences is not to look from no standpoint, but to look from multiple standpoints, our implications feed back on how the community recruits, trains, and evaluates scholars. To cover the breadth of social settings, the community may be more active in recruiting individuals with certain contextual expertise of under-researched settings, or offer resources such as training in this regard. Thus, to theorize social reality the organization and management studies community needs not only a diversity of theories and methodologies, but also a range of contextual expertise and related scholarly identities. Scholars may not have the competence or simply not an interest in a certain perspective and as a result disregard it. This is an argument scholars have made with regard to the tendency to neglect and misrepresent women's experiences (Scott, 1986). Moreover, the skill of what

Contextual Expertise & Management Theory 11
Hurtado (1996, p. 384) calls 'shifting consciousness' depicted as 'the ability of many women of color to shift from one group's perception of social reality to another' is equally relevant for researchers and research teams with other identities working in other contexts. In addition, there may also be further structural modifications such as taking into account the contextual richness and complexity of an article that may require a long-term development of expertise as a tenure or promotion criterion. Second, the community should reflect on its actual practices of collaborative research (Katz and Martin, 1997;Thomas et al., 2009;Jonsen et al., 2013). Pressures to publish and to increase citations lead to team publishing but not necessarily team research with many names on the article, but only one person in the field. This is a division of labor between data collection and analysis on the one hand and crafting the paper on the other hand. As Mauthner and Doucet (2008) highlight, this type of research at least requires a bridging between contextual work 'in the field' and textual work 'in the office' through team relations instead of team divisions of labor. More practically, it questions the extent to which team members understand field notes and how these should be shared among the research team. For instance, a second researcher may spend a shorter period of time in the field to obtain additional 'first-hand' exposure or an elaborate and documented engagement process with primary data within the research team takes place. This is not only a methodological but also a community concern, as it shifts taken-for-granted practices of separating field and textual work, which is not only potentially questionable from a normative standpoint but also with regards to the potential negative implications of separating data collection, analysis and write-up.
Third, our community of scholars needs to become more reflective of its biases. On the one hand, it has to encourage and enable context-intricate as well as multisite, multi-disciplinary, multi-institution, multi-method, multi-cultural, multi-linguistic, and multi-perspective methodologies. On the other hand, our methodological implications hinge on the acceptance of colleagues, reviewers, and editors that authors are welcome to unfold more of their contextual expertise in manuscripts. In that sense, it requires a legitimizing process by our community of scholars. However, reviewers are not bias-free experts that evaluate the quality of a manuscript and the authors' ability to accomplish what they have outlined solely on merit. The research community thus needs to develop mechanisms to train reviewers to deal with information about the research team that (re)present their contextual expertise and may also reveal some of their personal attributes. The community thus needs to engage with its own biasesa concern that has been picked up, inter alia, by debates around decolonizing the university and the invisibility of black and ethnic minority scholarship (Bhambra et al., 2020). For instance, languages other than English, both literally and figuratively as speech and representation, are oftentimes treated as peripheral (Meriläinen et al., 2008) or subaltern (Spivak, 1988)yet for contextual insights they should be considered core or superaltern.

Conclusion
While the cognitive limitations and biases of researchers have received extensive scrutiny, the individual and collective skills that allow them to immerse themselves in and theorize from data have been relatively neglected. As the social complexity of research settings grows, it is becoming increasingly important to obtain a researcher-context fit. Focusing on the skills of researchers through contextual expertise with its combination of both competence and criticality is likely to play an increasingly defining role in the generation of theory. Contextual expertise shapes the propensity of the research team to choose, capture, comprehend, convey, and confirm field intricacies so that our theorization is more fully able to account for the complex and diverse social settings that will be increasingly characteristic of future organization and management research.