Listening above the Din: The Potential of Language in Organizational Research

Abstract

In this article, the authors critically examine the use of language in organizational research. A literature review of management articles since the 1950s reveals that qualitative research is increasingly using quantitative methods to bolster arguments and argue for increased reliability. The authors found relatively few articles that included any analysis of the language or context embedded in any empirical study. They offer suggestions for the creation of alternative research paradigms that incorporate language and provide a voice to previously muted methods and voices.

Introduction

Empirical research is language rich, both in the manner in which we ask questions and in the ways in which we decipher responses. Despite the important issues related to the use of language and the social construction of meaning, however, organizational studies rarely mention these realities (for important exceptions, see Czarniawska-Joerges, 1997; Martin, 1990). This argument is not new, especially among organizational theorists (e.g., Alvesson & Wilmot, 1992; Forester, 1992; Gergen, 1992), but language deconstruction has not translated into mainstream management research or education.

Although the increased focus on language has been described as “one of the most profound trends within the social sciences” (Alvesson & Karreman, 2000, p. 136), we argue that the field of management is falling behind this trend. Furthermore, other trends, such as the quantification of qualitative research, are also constricting the lens through which we analyze our research. We also contend that the dominant paradigm for management research is becoming increasingly prescribed and predictable. Stiles (1993), in discussing “quality control” in qualitative research, emphasized the importance of creating a clear set of guidelines for qualitative research. We argue that such prescriptions inherently reduce the richness of divergent views or multiple layers of meaning. Consistent with the views of Astley and Zammuto (1992), we believe that although researchers as a group espouse innovative thinking and approaches to research, these important goals are not what we have come to value in terms of publication. The dangers inherent in homogenizing research have been well documented, especially in feminist writings (e.g., Gilligan, 1982), where alternative discourses have been suggested.

In this article, we have used the results of a literature review of organizational research published since 1950 to argue that a free flow of thought and ideas with open deference and invitation for the field to comment has been displaced by studies employing sophisticated quantitative methods that rarely ask the difficult questions that underlie their causal outcomes. One of the biggest problems with the current situation facing the management field is the seeming lack of desire, or incentive, to examine the rich layers of meaning that can be at least partially uncovered through analysis of language.

First, we present the findings of a literature review to substantiate empirically our claims that the number of qualitative works expressly discussing multiple layers of meaning, language, interpretation, or the co-construction of narratives is extremely low in relation to the total number of qualitative studies. Second, the literature review uncovers a growing trend toward quantifying qualitative research. Using these results as a backdrop, we examine the traditional view of management research and language. Next, we offer an alternative conceptualization of language use in organizational analysis focusing on the multiple layers of meaning imbedded in any text, the manner in which people use language, and the tools available for linguistic analysis. We then examine alternative discourses and conclude with suggestions for the field and future research.

Literature review

The first portion of the literature review, we identified research that examined language or meaning in organizational settings. In the second portion, we analyzed the number of qualitative research studies that employed quantitative methods in their results. We identified all studies using the electronic database PsycINFO holdings from 1950 to August 2003. This database covers a wide range of management disciplines. The literature review included only empirical studies that appeared in peer-reviewed journals. The authors generated and evaluated keywords used in the search to reflect a broad scope of qualitative and quantitative methods. We chose a literature review methodology because it permits an in-depth analysis of published articles that allows for the identification of themes and trends.

The identification of research studies that examined language or meaning in organizations entailed a three-step process. First, the keywords qualitative, interview, ethnography, content analysis, grounded theory, action research, narrative, theme, case study, field study, and focus group were used to identify studies that used a qualitative approach. We realize that this list in not nearly exhaustive but contend that it covers the most common qualitative methods used in management research. Next, the search was limited to studies that included management, business, or organization to assess studies that employed a qualitative technique in an organizational setting. Finally, only studies with the keywords language, meaning, or interpretation were kept for consideration. With these keywords, 348 articles were uncovered. These studies were examined in detail to verify that language or meaning was analyzed in an empirical qualitative design, in an organizational setting. In addition, to be included in the review, a component of the research had to demonstrate an effort to gain a deeper understanding of language, meaning, or interpretation in an organizational context. For example, how language is used by various stakeholders can convey multiple layers of meaning when examining a research question. We make no claim that our review is exhaustive, only that we examined studies with keywords typically found in qualitative management studies. The review resulted in 36 studies that explicitly studied language in a qualitative design within an organizational setting. These studies, and an explanation on how language, meaning, or interpretation was analyzed, are presented in Appendix A.

The results of the review suggest that compared to the total number of qualitative studies in organizational settings since 1950 (4,959), very few discuss language, meaning, or interpretation. In fact, we found no studies before 1980 that met the search criteria. All of the studies identified in the review could be classified as employing either positivist or postpositivist paradigms; that is, management forays into qualitative research are still heavily based in the scientific method of uncovering some objective truth through the use of qualitative techniques. These studies are based on naïve critical realist positions regarding perceptions of reality (Denzin & Lincoln, 2005). Very few studies alluded to multiple realities or the notion that reality can never be fully apprehended, only approximated (Guba, 1990). The postpositivist studies typically employed multiple methods and indications of reliability as their traditional evaluation criteria (consistent with Denzin & Lincoln, 2005).

In the management field, language and meaning have been studied in a variety of ways. One of the major themes identified from the review is the meaning of work (e.g., Isaksen, 2000; McAuley, Duberley, & Cohen, 2000; Poarch, 1998). These studies investigated how meaning is constructed, experienced, and developed. Researchers used semistructured interviews and ethnographic accounts to examine the meaning that people derive from their work. Others examined the meaning attributed to various work judgments, such as job performance (e.g., Dewe, 1992; Eisenhart & Ruff, 1983; Singh & Vinnicombe, 2000). These researchers examined the reactions to feedback and the use of equity theory (comparing ourselves to others) in determining perceptions of justice surrounding performance appraisals. Another theme that emerged from the articles pertained to organizational culture. Some researchers explored the role of language and meaning to understand how it shapes and influences the culture of the organization (e.g., Bate, 1990; Covaleski, Dirsmith, Heian, & Samuel, 1998; Gabriel, 1997; Jermier, Slocum, Fry, & Gaines, 1991). Narratives of participants were analyzed, as well as language and symbols used in formal and informal communication within the organization. There seem to be some promising signs that qualitative methods can produce important accounts of culture, mirroring the feats of anthropologic research. Language and meaning were also examined in relation to change management (e.g., Beech, 2000; Gioia & Thomas, 1996; Haddadj, 2003; Isabella, 1990; Sillince, 1999). How language and meaning can help us to better understand the change process was revealed to be an important and evolving theme in organizational theory; that is, what people say can inspire either action or cynicism, and ethnographic accounts revealed that words played a powerful role in mobilizing and sustaining organizational change efforts.

The second stage of the literature review involved a database search of PsycINFO to determine the number of qualitative studies published between 1950 and August 2003 that used quantitative techniques to bolster arguments. The second search was limited to studies that included the following keywords: quantitative, interrater reliability, statistical validity, NUD*IST, QSR, interrater agreement, alpha, or kappa. These keywords were chosen to represent the range of quantitative methods or measures that can be employed with qualitative data. There were 384 articles that met the search criteria. Next, the articles were examined to determine whether the qualitative results were quantified; that is, we were interested in studies that used software or statistical techniques to bolster their qualitative findings. Further, studies that had both a qualitative and a quantitative component but did not quantify the qualitative aspects of the results were not included in the literature review. Ultimately, 40 studies fit the criteria, as shown in Appendix B, along with the method of quantification.

The results of this review illustrate a trend, largely beginning in the 1990s, to quantify qualitative studies rather than letting the qualitative results stand on their own. Many studies in the review used computer software to aid in analyzing the qualitative data, such as QSR's NUD*IST software (e.g., Bell, Taylor, & Thorpe, 2002; McTavish, Litkowski, & Schrader, 1997; Price & Arnould, 1999; Zimmerman, Haddock, Current, & Ziemba, 2003). This illustrates the postpositivist trend in qualitative research in the management sciences, whereby traditional evaluation criteria such as internal and external reliability and validity are stressed as a means of bolstering the perceived legitimacy of the scientific method (Denzin & Lincoln, 2005). Although computer software represents a helpful method to classify and analyze large amounts of qualitative data, we argue its use should be subordinate to acquiring deeper knowledge of the subject through the analysis of language and multiple meanings. In other studies, researchers used multiple raters and calculated the interrater agreement or the interrater reliability (e.g., Clark & Montgomery, 1999; Prehar, 2001; Ruderman, Ohlott, Panzer, & King, 2002). Additional reliability coefficients were also applied in the reviewed studies, including the kappa coefficient (e.g., Coffman, 1992; Crump, Earp, Kozma, & Hertz-Picciotto, 1996; Schneider, Wheeler, & Cox, 1992).

The literature review left little doubt that an increasing number of qualitative studies are employing quantitative techniques to support their arguments. There is nothing inherently wrong with using technology or statistical techniques to improve the quality of qualitative interpretations. However, our review revealed that authors are increasingly quantifying their qualitative results, as though this somehow increases the legitimacy of the findings. As we examined each article, we uncovered a disturbing trend: that many researchers are using technology, but the result provides very little added value in terms of understanding the narrative. To be specific, by and large, the quantitative techniques neither shed additional interpretations on a narrative nor placed the researcher within the scope of inquiry but, instead, were focused primarily on legitimizing the themes that the researchers had uncovered. Our argument is that when technology is used as a means of providing “objectivity” to the “subjective nature” of narrative interpretation, we are at risk of losing sight of the value inherent in the richness of meanings and interpretation associated with qualitative data.

The literature review uncovered two primary trends that characterize the current dilemma facing organizational research. First, there are precious few studies that have analyzed language or meaning in a qualitative design that go beyond a positivist or postpositivist approach. The result of this trend is that we might be losing the potentially rich insights and layers of meaning imbedded in any narrative. Second, there is an emerging trend to quantify qualitative results. Here, we question the motive and not the technology. These trends can be partly attributed to the traditional view of research in the management literature, wherein quantification and cause-and-effect relationships appear to be increasingly rewarded. We are acutely aware of the irony involved in building our arguments from an essentially quantitative perspective; however, we believe that such an approach underscores our primary argument that quantification leads to increased perceived legitimacy, and find ourselves in the same trap in trying to illustrate empirically the publication trends our colleagues have noted anecdotally.

We will now examine the roots of what we perceive to be disturbing trends by first discussing the traditional view of research and language. In other words, we will now shift to the questions of how we got to the present circumstances, the tools available for language analyses, and the alternative paradigms available to management researchers.

Traditional view of research and language

The presumption of a concrete reality and an objective world capable of empirical study is rooted in the pure sciences. Scientific inquiry has operated under the assumption of narrowing the range of explanations toward that which best approximates some objective truth. Although the desired result of these assumptions is some causa finalis, it has been argued that these assumptions offer a restricted view of the sciences (e.g., Nietzsche, 1968). Indeed, the scientific method has been the dominant force in shaping management's reliance on positivist paradigms toward qualitative inquiry that isolates causes and effects, operationalizes theoretical phenomena, and allows for the generalization of findings (Flick, 2002). In social scientific inquiry, the dangers associated with the search for some objective truth are potentially disastrous. Gergen and Thatchenkery (1996) stated,

Scientific research may lead to technical accomplishments but it does not improve our descriptions and explanations of reality; descriptions and explanations are, rather, like markers by which we index our accomplishments. As research operates to displace one scientific theory with another we are not moving ineluctably “forward” on the road to truth—we are as so many would say, simply replacing one way of putting things with another. (p. 6)

In the premiere issue of the Academy of Management Journal, Wolf (1958) set the stage for management research to be built on the assumptions of the pure sciences. Wolf, expounded, “We can describe an organization as a living thing, it has a concrete social environment, a formal structure, recognised goals, and a variety of needs” (p. 14). Describing the organization as a living organism remains in vogue today, but we argue that organizations, with their rich social fabric, require alternative paradigms for analysis. Of course, similar concerns were also noted early in the management literature along with the potential implications of the dominant paradigm. Simonds (1959) argued,

As [the field of management] develops … there will be more and more stress on stating rather precisely the cause and effect relationships and on securing empirical data to substantiate or disprove these statements. Then the results of one investigation may be integrated with another until very substantial evidence is accumulated in support of a set of scientific principles. (p. 36)

Simonds proved quite prophetic, as the management literature has become increasingly focused on cause-and-effect relationships in search of universal truths at the expense of alternative discourses, interpretations, and levels of meaning. As Kemmis and McTaggart (2005) have pointed out, such an orthodox view of social science, despite its claim of value neutrality, normally serves the ideological function of justifying the position and interests of the wealthy and powerful.

From the traditional view of research in the social sciences emerged the dominant view of language. Alvesson and Karreman (2000) asserted that the traditional view of language uses words to represent and correspond to objects, whether they are people's inner lives (cognitions, emotions, values) or external lives (social interactions, relations). This view interprets language as an objective variable with far-reaching implications for methodology and practice. In almost all empirical research, the research design and text are written as if language is strictly controlled by the researcher; a simple tool through which she or he mirrors the world. However, not only are various paradigms of qualitative work being foreshadowed in management research, the notion that “behind all these terms stands the personal biography of the researcher who speaks from a particular class, gender, racial, cultural, and ethnic community perspective” (Denzin & Lincoln, 2005, p. 21) is absent from management's deliberation of findings. Although in quantitative studies, the language of the questions forms its own narrative that should be taken into account, instead statistical data are presented as if these offer a window to the context behind the questionnaire-filling situation. Lost behind reams of data and structural equation models are real people interpreting surveys in potentially very different fashions. We wholeheartedly concur with Gergen and Thatchenkery (1996), who have called for methodology to lose its status as the chief arbitrator of truth. Although intricate research methods and technologically sophisticated data analyses are essential to the research process, both the production and the interpretation of data must inevitably rely on the use of language.

In qualitative work, the narrative is simply more obvious than in quantitative studies, but the context surrounding interviews, focus groups, case studies, and so on has also been somewhat marginalized. In almost all qualitative studies in management, the researchers present selected portions of a narrative to “prove their case.” We argue that both the collection and the interpretation of the context of the respondent and researcher are critical to all empirical work. As Wittgenstein (1963) proposed, language gains its meaning not from its mental or subjective underpinning but from its use in action, thus underscoring the importance of context. The context, texture, rapport with the researcher, hypotheses derived prior to the study, and conclusions drawn thereafter are all part of the data. The trend toward personal experience narratives and, more recently, autoethnography (Holman Jones, 2005), show much promise for future qualitative management research.

Language cannot work as a simple tool for measuring reality. Language does not reflect or mirror reality, but as Gergen (1992) has suggested, it is a “mere messenger from the kingdom of reality” (p. 218). Research texts are not objective or clinical accounts of the facts; rather, they act in the persuasive construction of the facts through the powerful and biased view of the researcher. Poststructuralists have argued that positivist approaches clinging to the scientific method are just one way of telling stories. Similarly, critical theorists (e.g., Huber, 1995) have argued that positivists and postpositivist paradigms silence too many voices, a theme we will return to in the alternative paradigms section of the article.

The countercultural view of language as an arbitrator of multiple meanings and innovative theory creation has been dealt a striking blow by the manner in which peer-reviewed journals (ourselves) increasingly call for empirical evidence of reliability. This trend is especially evident in qualitative studies. The notion of interrater reliability, for instance, calls for an examination of the extent to which researchers interpret (or code) a narrative in a similar fashion. We argue that such reliability alphas undermine any effort to foster divergent interpretation and the presentation of multiple meanings. The management sciences are locked in the postpositivist paradigm's trap of traditional evaluation criteria such as internal and external reliability and validity as a means of gaining perceived legitimacy. High interrater reliability can be interpreted to mean that the researchers have similar backgrounds (e.g., level of education, interest in topic) and worldviews. The point here is that alternative interpretations of the same text would lead to lower interrater reliability yet potentially more rich and interesting discussions. Such work, however, would face inevitable criticism from reviewers focusing on low alphas.

Generalizability theory (see Cronbach, Gleser, Nanda, & Rajaratnam, 1972) is a useful statistical approach, rarely used in qualitative research in management, that can be used to explain variance caused by a range of factors. This theory appears to be a promising alternative to the narrow focus on interrater reliability coefficients in business research, by substituting the emphasis on reliability with the broader notion of generalizability (see Marcoulides, 1998). Although this approach appears promising, the point remains that postpositivist statistics should not obstruct the more pertinent and meaningful issue of language analysis.

In most qualitative studies in management today, what we are left with is a prescribed, narrowly focused account of some meaning-laden phenomena. Some authors have expounded on the potential benefits of the status quo. “Given the observation that those organizational theories that are viewed as most interesting seldom have received much empirical support (Astley, 1985), this may be celebrated to the extent that it liberates theory development from the straightjacket of verification” (Alvesson & Karreman, 2000, p. 153). Although this line of argumentation is provocative, we argue that as a field there is an implied responsibility to question and build interesting, even daring (counterdominant paradigm) theory and present it to our peers as tentative and speculative. Such an approach is consistent with the views of poststructuralists and postmodernists, who argue that there is no such thing as an “objective” observation and no clear window into the inner lives of individuals (Alvesson & Karreman, 2000). As shown in the literature review, this type of theory development is quite rare, and the current methodological dogma clearly begs for a more receptive use of language.

The potential role of language

An alternative view of language

Understanding how language works is central to understanding how organizations work (Musson & Cohen, 1999). The nature of language is contextually dependent, metaphorical, active, built on repressed meanings and capable of shaping all sorts of organizational perceptions (Alvesson & Karreman, 2000). One obvious metaphor for organizations is to view them as evolving texts or multilayered narratives taking place over time. To examine an organization in a cross-sectional design is, at best, capturing and interpreting meaning from one particular context at one particular time.

Understanding organizations in relation to the context in which they exist is a well-established perspective in organizational theory (e.g., Clegg, 1990; Gergen, 1992; Musson & Cohen, 1999; Pettigrew, 1987). However, expressly discussing the role of language within a given context has not received adequate attention. Alvesson and Karreman (2000) have discussed the predicament facing the field: “It is not that our understanding is poor; rather it is that social realities are so extraordinarily rich” (p. 147). Organizations are made of individuals with complex lives who collectively (re)create a culture that is (re)constructed through everyday activities and practices. The culture of any organization is created within, and articulated through, organizational language and symbols. Mundane, everyday talk during telephone calls, formal and informal meetings, water cooler conversations, and so on are mechanisms in which people “inform, amuse, update, gossip, review, reassess, reason, instruct, revise, argue, debate, contest, and actually constitute the moments, myths, and through time the very structuring of organizations” (Boden, 1994, p. 8). Thus, consistent with participatory action research paradigms (e.g., Howell, 2004), the context and how it evolves should factor into interpretations of language and research should be reconceptualized in the management sciences as a social practice.

Influence of postmodernism

Contemporary debates over language in organizational analysis have been argued to be an artifact of postmodernism (Alvesson, 1995; Knights, 1997), with the belief that there is no single, enduring reality. Postmodernists have received their fair share of just criticism for outlining the problems with contemporary thought and analysis yet offering few pragmatic suggestions or solutions. Postmodernist language scholars (e.g., Cooper, 1989; Deetz, 1992) have argued that language is often built on the repression of hidden meanings (i.e., metaphorical) and have asserted that efforts to say something definite rely on shaky foundations that ought to be deconstructed. The Enlightenment view, traced back to Locke (1825), was based on the argument that language and words are signs of internal conceptions. As Alvesson and Karreman (2000) have discussed, rather than telling it like it is, the challenge for the postmodern researcher is to tell it as it might become. This does not exempt postmodernist scholars from theory construction. Quite to the contrary, existing theories represent a discourse that is potentially available for many purposes in a variety of contexts, for if the goal of organizational research is to ascertain what respondents are thinking with regard to a topic, perhaps more weight needs to be placed on their use of language as a source of meaning making. Postmodernists would argue that the traditional management positivist and postpositivist paradigms silence too many organizational stakeholders and stimulate important questions about why certain voices have been largely absent from ethnographic accounts in management. Organizational members are actually quite well suited to developing their own theories to suit their context and needs. Although such theories “may lack the elegance and sophistication of official theory, in terms of immediate needs they can be more powerful” (Gergen & Thatchenkery, 1996, p. 371).

Levels of meaning

The rigorous assessment of language allows for the construction of multiple layers of meaning. To consider the level of meaning implies an interest in what people mean by the expressions they use. Musson and Cohen (1999) have asserted that meaning making is both contextual and relational; that is, not only is the context ever changing but the narrative is also formed in relation to other persons. Meaning, therefore, changes over time and from person to person (Gabriel, 1995). Management must come to terms with the fundamental tensions between “self” and “other” in qualitative research. Because of the retrenching into postpositivist camps, this artificial distinction continues to gain strength.

Seeking meaning, in essence, is a desire to assess what might be happening in terms of assumptions and beliefs about the world behind the scenes (Beech & Cairns, 2001). Interpretations of meaning are often subconscious, yet they inform the way we act and interpret the world. What becomes apparent in an organizational context is the extent to which all organizational members must be seen as “meaning makers” in their own right. Similarly, the researcher is implicated in the narrative, and autoethnographic accounts of organizational culture would be a valuable addition to the literature. Organizational researchers and practitioners alike can benefit from becoming more aware of multiple meanings or divergent views of “reality” to describe or manage organizational situations more effectively. Such an approach is consistent with the view that management problems are complex and require paradoxical thinking (see Lewis, 2000).

Musson and Cohen (1999) have argued for diligence in ensuring that language is not used to homogenize meaning, but instead propose that we focus on the diversity of coexisting (and often colliding) worlds. The implications for research are far reaching, as offering the speculative (the layers of meaning behind the transcript) in favor of the obvious (the content of the transcript), subjects one to criticism regarding lack of causality (and/or lack of immediate practical utility). We argue that theory building might be better served by the use of language as the gateway to multiple meanings offered up to the field for discussion.

Use of language

People use language for a variety of purposes, and even if the presumed intention is to communicate, language is a rich medium for expression and interpretation. People primarily use language not to make accurate representations of real world phenomena but, rather, to accomplish things (Alvesson & Karreman, 2000). The variety of means employed to achieve these accomplishments, as well as the meaning behind the accomplishments themselves, have been argued to be vastly underrepresented in organizational research (Potter & Wetherell, 1987). For instance, statements in an interview might say something about how language is used in the context of the researcher and surroundings but might or might not be applicable to other situations. Despite these important distinctions, very little empirical research explicitly discusses the role of language and context in the meaning-making exercise. This line of argument does not preclude empirical approaches important in any constructionist study, but we argue that the empirical outcomes of a study should be subordinate to the cognitive input or textual output presented by the researcher as language use. Empirical research might be better served if we acknowledge that respondents might be using the research opportunity to accomplish something (e.g., to send a message to upper management). The goal of the researcher is to deconstruct the potential layers of meaning in the narrative while acknowledging his or her own research bias (or preconceptions) and their active engagement as part of the research, not separate from it.

Tools for language analysis

The study of language and discourse is rich with tools that might have utility for organizational research. Discourse analysis, largely attributed to Potter and Wetherell (1987), and conversational analysis (Silverman, 1993) are both methods of studying language use in social situations. Discourse analysts attempt to show that an emphasis on the representational capacities of language conceal and obfuscate the more productive question of its creative and functional capacities (Alvesson & Karreman, 2000). The study of discourse allows one to identify the subissues, including organizational taboos and cultural norms, that might otherwise go undetected. In any data constructionist research, including discourse analysis, how the research pragmatically and metaphorically structures the researcher's perceptions is crucial for the outcome of the study (Grant & Oswick, 1996). In fact, Fairclough (1992) expressed caution that “discourse analysts are increasingly at risk of becoming incorporated into bureaucratic and managerial agendas” (p. 239). Denzin and Lincoln (2005) have suggested that there no longer exists the option of ignoring agendas in our research as the social sciences, and most certainly the field of management, are already imbedded in issues of ideology, power, sexism, racism, domination, and control. In a sense, all qualitative management research is fundamentally mediated by power relations that are socially or historically constituted (Kincheloe & McLaren, 2005). Kemmis and McTaggart (2005) have asserted that despite the claims of orthodox social science of value neutrality, such inquiry normally serves the ideological function of justifying the position and interests of the wealthy and powerful. This concern speaks directly to the issue of language as a tool used to accomplish things. We argue that all data constructionist work is, by definition, at risk of being incorporated into alternative agendas, yet it is the role of the researcher to explicitly unpack these potential agendas (including the intricate role of the “self”) and expose the multiple meanings that lay behind any narrative.

Trends in ethnographic studies and dialogic research appear promising for their potential incorporation into organizational studies. In ethnographic studies, the emphasis has changed from a focus on fieldwork to work on narratives as an equally critical focus of attention (Geertz, 1988). More recently, Gergen and Thatchenkery (1996) discussed dialogic research that allows participants to formulate modes of understanding or action that incorporate multiple inputs. Individuals' view of themselves and the world might be even more strongly influenced by social and historical forces than previously believed (Kincheloe & McLaren, 2005). An acknowledgement of this shift is promising, as the issue of context is more fully accounted for using experience sampling or other innovative designs (e.g., autoethnography). In fact, it would be a valuable addition to the organizational literature to see experience sampling being used in conjunction with autoethnography to compare narratives at different points of time, and in potentially different settings, all through the openly discussed filters of the researcher. Despite some promising signs of change, the literature review revealed that precious little recent empirical organizational research pays attention to issues of language. This reality is disturbing, as it might be indicative of the very limited, circumscribed fashion that peer-reviewed journals (we) expect and demand of empirical study.

Alternative paradigms

The dominant organizational research paradigm mirrors back language as if it were an objective variable. As a generator and purveyor of meanings, the field inherently operates to the benefit of certain stakeholders and to the detriment of others. Clearly, research that examines multiple layers of meaning and deconstructs language has been relegated to the sidelines of organizational inquiry. However, with the proper emphasis placed on language, organizational science could no longer afford to extricate itself from alternative paradigms including sociopolitical, moral, feminist, third world, and so on. Voices that have been muted would be heard with increased attention on our own research bias, the deconstruction and construction of language and multiple meanings, and the inclusion of novel methods and samples. For example, autoethnography has much to offer in the study of organizational phenomena, not the least of which being organizational cultures. Autoethnography can be defined as “research, writing and the method that connect the autobiographical and personal to the cultural and social. This form usually features concrete action, emotion, embodiment, self-consciousness introspection [and] claims the conventions of literary writing” (Ellis, 2004, p. xix).

Autoethnography might be a particularly effective tool in all forms of organizational research. We each have our own needs and desires from an occupation, and as such form implicit theories about how work should be structured, executed, and managed. Autoethnography is a marriage of the “self” and “other,” such that the researcher's fears, beliefs, and assumptions play an important role in the social construction of any narrative. By drawing on collective experience, insight, and shortcomings, we might be able to learn much more about what it means to be a supportive supervisor, why employees burn out, and what makes individuals feel an emotional connection to a person or organization (to list but an obvious few).

Language is powerful and has been argued to hold the key to political influence and control in organizations (e.g., Bate, 1994; Conger, 1991; Musson & Cohen, 1999). Harlow, Hearn, and Parkin's (1995) seminal research into gender and language treats the concepts of silence and din both metaphorically and literally. The din is the public, male-dominated patriarchy that has largely shaped organizational research. Silence refers to the private, female-dominated sphere, which has, with notable exceptions (e.g., Gilligan, 1982), gone unheard in most organizational research. Harlow and colleagues' (1995) work can be directly translated into an organizational context in which the “din of male leadership, management, rules, powers, and decision making structures is contrasted to the silence of women who are frequently not heard and who find progress difficult within these male-dominated spaces” (p. 917). The authors go on to examine the dins and silences in organizational theory itself. In this context, the din encompasses that which is dominant (published in top tier journals) including not only theories but approaches and the theorists themselves, and the silence of that which is seen as irrelevant, unimportant or simply absent. To “listen above the din” entails using language as a means of interpreting narratives outside of the dominant, restrictive status quo and providing a voice to alternative organizational stories and theories.

The voices of women, minorities, persons occupying minimum wage jobs, and workers in foreign countries are too often completely forgotten in mainstream management discourse (Gilligan, 1982). Management is still shackled by the implicit and explicit bound of patriarchy, as we worship the bottom line. It is still common in many management journals to ask authors to explain how their findings could lend themselves to increased organizational effectiveness, whereas social justice, gender equality and moral obligations play de facto secondary roles (Harlow et al., 1995). By illuminating these hidden voices, management scholars could shed light on complexities that go beyond current conceptualizations of management theory.

More participatory forms of ethnography, including autoethnography, must not be viewed as a panacea for the organizational sciences, as communication scholars have outlined some of the many tensions ethnographers must grapple with in the 21st century. Howell (2004) has provided a moving account of the difficulties associated with becoming embedded in research and sharing in the confidences of participants. In her study of women in southern Mexico, Howell found that it might only be through the sharing of confidences that firsthand accounts of responses to violence can enter the ethnographic record. She asked the difficult question of “how does one present these unsolicited, sensitive data?” (p. 28). Parallels can be drawn to management research on social ostracism, stress leave, and other psychologically traumatizing events that researchers might understand more fully by walking this tightrope between friend and researcher. Indeed, once you have established yourself for some time in an organization, “the lines between participant and observer, friend and stranger … are no longer so easily drawn (Behar, 1997, p. 28). Behar also reminded us of an important guiding principle in participatory ethnography: ”Act as a participant, but don't forget to keep your eyes open“ (p. 5). For management scholars, keeping our eyes and ears open to the voices of those in nonmanagerial positions and marginalized organizational roles might provide a more holistic portrayal of the organizational culture or phenomena under study.

Suggestions for the field

Although it is easy to knock the dominant power structure underlying organizational research and then selectively use studies to build our arguments, we believe that we need not diminish the value of past or present research to improve future research. Perhaps more to the point, we feel an obligation to offer suggestions to the field that might serve to enhance future research by illuminating the richness that language provides. In reconsidering the manner in which language operates, it is hoped that new and novel research agendas will emerge.

Our first suggestion involves being explicit about the speculative elements involved in any research and a call for increased humility in data interpretation. There might be a temptation when employing, for example, longitudinal designs using structural equation modeling in large samples to tout causal relationships and strong reliability coefficients. However, all empirical study rests on the manner and context in which the language (of the questionnaire or qualitative technique) was designed and interpreted by both the respondent and the researcher. We argue that discussions providing multiple meanings and interpretations might yield higher value than overly conclusive statements of “fact.”

Second, we argue that there might also be value added in being less ambitious in research scope and being more concerned with the depth of analysis. Here, we argue for more ambitious rigor and thoughtfulness concerning the linguistic dimension of any empirical study. Novel research agendas examining specific phenomena, but also incorporating language analysis in an autoethnographic design, would be a particularly useful addition to the field.

Third, we suggest that respondents be given their narratives to review. We realize that this protocol is part of the research routine of many researchers, but we argue that the practice should be made more explicit in method sections, as it provides yet another opportunity for both the respondent and the researcher to examine the use of language. For example, examining a specific emotion such as guilt in the workplace using experience sampling to uncover when and why people felt guilty over a lengthy period might be more fruitful than developing a language-laden guilt scale in a cross-sectional design. Providing a respondent with the opportunity to examine why he or she selected certain words to explain her guilt, for example, might provide a deeper understanding of the narrative than the text alone.

Finally, and most strikingly, we believe that the increased calls for the quantification of qualitative research should be displaced by the more vital issues of incorporating language, developing multiple layers of meaning, promoting collective theory creation, and using qualitative analysis technologies, while not being ruled by them. We need to stimulate debate surrounding Lincoln and Denzin's (2005) ”great divide“ of the scientific method versus the pursuit of socially and culturally responsive studies. Qualitative data analysis has never been ”easier“ in terms of the technologies available; we must therefore collectively ensure that technology is not restricted to the role of quantifying qualitative research for the sake of perceived legitimacy.

Language has much to offer organizational research; it remains an open question as to whether alternative voices will be heard above the din.