Sociology, Theory, and the Feminist Sociological Canon

In this article, I revisit the concept of “doing gender” to query whether the framework as directly articulated by West and Zimmerman is meant to be a description of social life (a social theory) or a testable and potentially falsifiable explanation of the empirical world (a sociological theory). I document how much of the research that uses “doing gender” is a misapplication of the concept as a social, rather than sociological, theory. I conclude by making the case for the role of “doing gender” in, rather than as, sociological theory.

Article "Doing Gender" as an article and as a concept is widely influential in and outside of sociology. Although the original article itself had a difficult time coming to publication (see West & Zimmerman, 2009, for more details), most sociologists now expect graduate students (and likely undergraduates, too) to be familiar with the concept and/or to have read the original 1987 article. The symposium in Gender & Society on the 20th anniversary of the article's publication (Gender & Society,Vol. 23(1)) asked whether "doing gender" was part of the sociological canon, a call to research, or something else. Indeed, the authors in the symposium document the myriad of ways in which the concept of "doing gender" has informed research projects and theoretical explications of social behavior.
In this article, I revisit the concept of "doing gender" as West and Zimmerman (2009) implore in their response to the Gender & Society symposium. I ask whether "doing gender" as described in 1987 and rearticulated in 2009 is a social or a sociological theory. Building on Chafetz's (2004a) critique of much of feminist theory, I query whether the framework as directly articulated by the authors is meant to be a description of social life (a social theory) or a testable and potentially falsifiable explanation of the empirical world (a sociological theory). I then show how the critiques that West and Zimmerman (2009) bring to bear on much of the research that uses "doing gender" are due to researchers' use of the concept as a social, rather than sociological, theory. This misuse has yielded scores of publications providing valuable thick description on social life without direct tests of the mechanisms by which "doing gender" is purported to explain the reproduction of gender inequality. I conclude by making the case for the role of "doing gender" in, rather than as, sociological theory.

Defining Sociological Theory
I have been heavily influenced by Janet Chafetz's (2004a) article (a symposium presentation subsequently published as an article alongside the symposium discussants' critiques), where she critically assesses the state of feminist theory in social science and in sociology as a discipline. Chafetz (2004a) argues that the goal of social science is to develop explanations (theories)-that is, attempts to answer questions of why and how-of empirically documentable phenomena concerning human behavior and the structures and processes they create in the present and have created in other times and places. (p. 964) There are, she argues, two key components to the discipline of sociology: (a) an emphasis on using systematic approaches 697157S GOXXX10.1177/2158244017697157SAGE OpenDavis to learn about the world (methodologies) and (b) theory, or "the development of at least relatively abstract explanations of empirically testable and documentable phenomena" (Chafetz, 2004a, p. 965).
Sociology as a discipline has long fought with itself regarding our comfort with being called a science. Even Weber (1922) famously argued for sociology's usefulness as a science and in society. Much in the discipline's history has been written about the extent to which sociology is a science, but much of that discourse has been derived from a "positivist" version of science. That is, the truth, reality, can be discovered through objectivity, meaning removing the researcher's subjective bias. Feminists have critiqued this approach to doing science (e.g., DeVault, 1996;Sprague & Zimmerman, 1993;andeven Chafetz, 2004a, 2004b), but one key fact remains. If sociology is a science, we use systematic approaches to studying the social world, and through those approaches, we both derive and test relatively abstract explanations for empirical phenomena.
A sociological theory is not one that is philosophically oriented that discusses the nature of social life-that would be what Chafetz (2004a) calls social theory. Sociological theory would go beyond description and would be focused on explaining how and why the empirical world operates as it does. This includes articulation of mechanisms (Hedström & Swedberg, 1998;Reskin, 2003) such that the premise of the theory itself is testable and falsifiable. Theories are evaluated, therefore, not on their ability to completely explain one specific set of experiences but instead their ability to explain (and predict as Smith-Lovin, 2000, would argue) the "empirically knowable world" (Chafetz, 2004a, p. 965). Social theory, on the contrary, provides thick description of social experiences, often elucidating the mechanisms to be included subsequently in sociological theories as they document social life in particular, ultimately leading to the refinement of sociological theory. Both sociological and social theory are important in the doing of our work as social scientists, as can be seen, for example, in the case of Morris's (2012) examination of gender and education in a two low-income high schools. Highlighting the inadequacy of contemporary sociological theories about educational attainment in explaining the educational gender gap, Morris develops an argument about the importance of how gender is constructed in a place, how it is contextualized, as an additional mechanism that explains how young men and women are handicapped in the pursuit of education. His ethnographic work yields additional insight into mechanisms through which gender inequality is perpetuated. In her critique of and response to Chafetz (2004a), Baber (2004) argues that theory should guide scholarship but should simultaneously allow for the building of bridges between scholars and activists in the pursuit of reducing gender inequality. Therefore, social theory with its focus on how inequalities are experienced by individuals may be of more use and value to practitioners than is sociological theory with its focus on prediction and explanation.
My view of the discipline of sociology is one that is based on the distinction articulated by Chafetz (2004a). Our goal is to better understand the social world through the use of systematic data collection and analysis. The creation of better explanations (theories) is paramount. I also concur with Risman (1994), Sprague (2005), and others (Chafetz, 2004a, included) who note that a feminist empiricist approach (one that seeks to investigate power relations in a society through the best methods and analysis techniques for the question at hand) is not at odds with the notion of the purpose of sociology as a discipline nor the feminist epistemological framework that seeks to eliminate gender (and other forms of) inequality. Some feminist sociologists have argued for the use of "different" methodologies, pointing to the positivist straw man as a part of the problem of the continuation of gender inequality (e.g., DeVault, 1996;Luker, 2008). Like Sprague and her colleagues (Sprague, 2005;Sprague & Zimmerman, 1993), I disagree. If the goal of a feminist sociology is to (a) systematically study the social world and develop explanations for empirical reality which then lead to (b) the elimination of gender inequality, then the methods that we use to collect and analyze the empirical world must be as diverse and responsive as the world itself. Indeed, those feminist critiques of "science" and discomfort with "more objective" approaches to studying the social world miss the big picture. We cannot eradicate gender inequality by amassing many thick descriptions of women's (or men's) lives based on their standpoint, location in the matrix of domination, or documentation of how their social location reflects oppression (Chafetz, 2004a;Schrock & Schwalbe, 2009). These examples of social theory are but one piece of a larger puzzle. The creation of systematic explanations for aspects of the empirically knowable world (here that aspect of the social world is gender inequality) that can be tested and potentially altered and/or falsified given empirical scrutiny is the goal, and all approaches to understanding the social world are needed to complete this task. As Walker (2004) rightly argues in her response to Chafetz (2004a), "[c] ompared with theory developed by so-called objective observers, acknowledging our own perspectives and asking how the world looks to others eventually will lead to rich theory closely linked to social life" (p. 991).
That said, theories must imply how they would be empirically tested. We must have research methodologies in our science that facilitate our testing of our theories. Theories about the social world that do not readily lend themselves to empirical scrutiny and potential falsification, that is, have no methodological implications for how researchers would go about collecting and analyzing empirical materials, nor how the theory's premises, postulates, mechanisms, and processes can be evaluated and potentially falsified are not sociological theories. In this case, the content of the empirical material is irrelevant. If a theory that purports to describe gender inequality and its reproduction cannot be potentially falsified, it would not meet the definition of a sociological theory.

The Case of "Doing Gender"
"Doing Gender," the 1987 article, is the most cited article in the discipline by most metrics (West & Zimmerman, 2009). The argument, as restated by West and Zimmerman in 2009, is as follows. One's sex (female or male) is usually determined by the possession of female or male genitalia (see West & Zimmerman, 1987, for more on the fluid nature of this categorization). Sex categorization, building on Goffman (1956), occurs through "the display and recognition of socially regulated external insignia of sex-such as deportment, dress, and bearing" (West & Zimmerman, 2009, p. 113). Sex category and gender are related in that gender is about being recognized as someone inhabiting a sex category-and "being accountable to current cultural conceptions of conduct becoming to-or compatible with the 'essential natures' of-a woman or man" (West & Zimmerman, 2009, p. 113). Therefore, gender is conceptualized as an ongoing process, a doing rather than a being.
West and Zimmerman's (2009) article was the conclusion to a symposium on their original piece. In it, they lament that authors (including some in the symposium) have misused their concept, so much so that they felt the need to rearticulate its premises. One key critique was the use of "doing gender" as an explanation for the reproduction of gender inequality. This usage presumes that "gender" is a thing and that "doing it" reproduces the status quo. Indeed, they argue that because gender is done and that it requires individuals being held morally accountable to current cultural norms ascribed to a sex category, the performance of gender should and must change over time-but that has little to do (directly) with the reproduction of gender inequality.
However, many authors, including myself, have invoked "doing gender" as a theory for why gender inequality is being reproduced. We have argued that, among other things, the reason men do less housework than do women is because they believe they are being held morally accountable to the sex category of "male." Thus, not doing housework is "doing gender." But as we use the language of "doing gender," we are falling into the trap that West and Zimmerman argue is diametrically opposed to how they had conceived of the concept. Gender is not a thing that is done, that is, a noun. Instead, "doing gender" is a verb phrase, a process.
Other authors have noted that "doing gender" as a theory has been misused. Deutsch (2007) recounts many examples of research that, she argues, utilize "doing gender" as a theory of gender maintenance. Risman (2009) also notes that "doing gender" has been invoked to document multiple masculinities and femininities; she encourages researchers to investigate sites where women and men are "undoing gender" (Butler, 2004;Deutsch, 2007). However, as West and Zimmerman (2009) note, "an emphasis on 'undoing gender' deflects attention away from the situational character of gender accountability, and circumstantial modifications" (p. 118). While authors like Risman (2009) andDeutsch (2007) want us to talk about how we can "undo gender," that is, to highlight "social processes that underlie resistance against conventional gender relations and on how successful change in the power dynamics and inequities between men and women can be accomplished" (Deutsch, 2007, p. 107), there remains a presumption that gender is a thing that exists to maintain inequality between women and men. The focus is on the gerund-"doing" versus "undoing" rather than the phrase "doing gender" versus "undoing gender." West and Zimmerman (2009) argue that this semantic difference is one of the crucial points of their theory and the place where the usage of their theory has been problematic.
This process of "doing gender," as described by Zimmerman (1987, 2009), explains how and why people behave as they do. People believe they are being held morally accountable to a sex category and behave according to their understanding of contemporary cultural norms around that sex category. This explanation for human behavior fits the definition of a sociological theory on the face of it, in that it is potentially falsifiable. However, what are the testable hypotheses derived from this explanation? Is the underlying abstract process falsifiable? Is there ever a circumstance that humans will not be behaving as they do at least in part because they believe they will be held morally accountable for their behavior? For example, Hirschi's (1969) social control theory highlights the centrality of being held morally accountable to a particular peer group even among deviants. Thus, the use of "sociological theory" as a descriptor for "doing gender" may not be warranted. This distinction is more than semantic, as a large body of scholarship has been published that purports to find support for this theory. And much of this scholarship, including the articles in the 2009 Gender & Society symposium, frames "doing gender" as an explanation for the reproduction of gender inequality, thus solidifying its position in the feminist sociological canon.
Herein lies the irony. Based on the arguments presented here, "doing gender" does not meet the definition of a sociological theory that explains the reproduction of gender inequality. But West and Zimmerman never intended it to be. West and Fenstermaker (1995) extend West and Zimmerman's (1987) argument to other forms of difference (and categorical inequalities; Tilly, 1998) by noting that distinctions among groups are not essential but must be created and maintained. Interestingly, they argue that difference is a social doing, a mechanism that helps explain how categorical inequalities are reproduced. Following this logic, both "doing gender" and "doing difference" are themselves not theories, but are mechanisms through which inequalities are reproduced. Therefore, "doing gender" (and "doing difference") is useful for explaining the reproduction of inequality when incorporated as a mechanism into broader theoretical explanations. The notion of "doing gender" is a key explanatory mechanism invoked by Chafetz (1990) in her theory explaining the links between macro-and micro-level gender inequality as well as by Ridgeway (2011) in her specification of status expectations theory as applied to gender inequality. Clearly articulated mechanisms, such as those implied in the "doing gender" approach, provide insight into the "black box" of sociological theory. These mechanisms describe the how and why of a theory. As Gross (2009) argues, mechanisms must be centered around social action. In the case of Chafetz (1990), knowing one will be held accountable for one's behavior based on sex category leads women and men to behave in ways that reinforce patriarchal norms at the micro level. Furthermore, these norms of moral accountability are part of the cultural ideology that constructs appropriate behaviors for women and men in institutional settings. Ridgeway, in some ways building on this idea, notes that gender is like a ghost in all interactions; it is there even when it is not being directly invoked. Individuals are being held accountable to culturally constructed norms of behavior tied to sex categories even in situations when gender is irrelevant to the task at hand (Ridgeway, 2011). Thus, "doing gender" as a concept is one of the mechanisms invoked in Ridgeway's specification of status expectations theory, as it is one of the explanations for how gender inequality is reproduced through everyday interactions. Calls for research focusing on "undoing gender" (Deutsch, 2007), where scholars examine how behaviors reduce inequality, miss Zimmerman's (1987/2009) point that "doing gender" is not about the maintenance of gender inequality (although Deutsch (2007) is accurate in her description regarding how the theory is used). Instead, it is a set of explanations for the origins and maintenance of culturally constructed norms of behavior tied to sex categories. As a mechanism, "doing gender" explains how culturally constructed norms of behavior tied to sex categories are maintained as well as challenged and modified, as people are held accountable to the contemporary norms.As Deutsch (2007) demonstrates, norms are modified through interaction and structural shifts but the processes of holding people accountable to those culturally constructed norms of behavior tied to sex category remain the same. "Doing gender" is not intended to explain the maintenance of gender inequality; Zimmerman (1987/2009) intend it as a mechanism that can be used to explain the reproduction of, and possibly the disruption of, culturally constructed norms of behavior tied to sex category.

Resituating "Doing Gender" Into Sociological Theories
What does this mean, then, for the disposition of research that purports to test or utilize "doing gender" theory? I would argue that this voluminous body of scholarship is quite important to our understanding of the social world but as examples of "doing gender" as a mechanism for the reproduction of gender inequality. Understanding the lives of men and women as they are lived is important. We cannot ignore how social location impacts lives, nor should we minimize scholarship that has documented how oppression and domination, as well as privilege and entitlement, operate in people's lives. But as Schrock and Schwalbe (2009) argue, simply documenting "multiple masculinities" or "multiple femininities" based on interlocking axes of domination masks the underlying processes that reproduce gender inequality structurally. For sociologists, both those of us embracing the label of feminist and those who do not, our goal remains the same. If we want to change the world, we must first understand it. And true sociological understanding is derived from the principles that make sociology the discipline that it is-rigorous, systematic approaches for the collection and analysis of materials regarding empirically known phenomena and "relatively abstract explanations" for those phenomena that are empirically testable and falsifiable. "Doing gender" as a concept, and research invoking it, has a place in our science, albeit one more in line with what the concepts' originators had intended than how it has regularly been used.

Author's Note
This article originated in, and benefited from, conversation with students in my graduate Gender and Social Structure course in Fall Semester 2012. This article also benefited from discussion in the George Mason University Department of Sociology and Anthropology Colloquium Series.