On the Observability of Inequality by T.J. Berard

From Concepts to Methods
On the Observability of Inequality
T. J. Berard
Kent State University

Journal of Contemporary Ethnography Volume 35 Number 3 June 2006 236-256 © 2006 Sage Publications 10.1177/0891241605285097 http://jce.sagepub.com hosted at http://online.sagepub.com

Inequality can be understood as a social fact, or as a concept used to claim the existence of a social fact. These two orientations are intimately related; inequal- ity is a social fact only insofar as people use the concept of “inequality” accord- ing to social conventions for its use, thereby contributing to the meaningful, socially structured, and meaningfully socially structured character of society. A social constructionist approach to inequality requires attending to the con- cept “inequality” as a cultural, sociolinguistic resource used to understand and describe social relations. Such use simultaneously recognizes and accom- plishes inequality as an intersubjective reality. By attending to inequality as a concept, the empirical study of inequality can be grounded conceptually in the natural language which inevitably informs analytic as well as lay methods of reasoning, and grounded empirically by reference to situated, practical lan- guage use as a constitutive feature of social identity, social relations, and social problems.

Keywords: inequality; social construction; ethnomethodology; reification; intersubjectivity

The literature on social inequalities in all their various forms is an impos- ing mountain comprising theory, research, social criticism, policy stud- ies, and many other genres of social-philosophical, social-scientific, politi- cal, legal, historical, and journalistic writings, informed by diverse theoretical and political orientations and representing a variety of disciplines and institutional interests. Although there are few blanket statements that can fairly be thrown over this mountain of literature, there is an overwhelming tendency to approach inequality in a practical and evaluative manner, as an ostensibly objective social problem.
Author’s Note: I acknowledge with thanks the comments from the anonymous referees and the editor of the special issue, Scott Harris.
Downloaded from jce.sagepub.com by Wayne Mellinger on October 20, 2011
Berard / Observability of Inequality 237

Conventional social-scientific analyses of social problems typically begin with unquestioned ontological and epistemological commitments (e.g., reifi- cation) and unquestioned evaluations or agendas (e.g., social criticism, social reform) and proceed on the foundation of largely unexplicated meth- ods of sense making, in effect precluding analytic attention to the fundamen- tal question of how it is that social problems such as inequality are observable in the first place. Thus, the study of inequality has rarely attended to how inequality is employed as a resource in making sense of social relations, with the exceptions being few and recent (e.g., Harris 2001, 2003).

While constructionist studies of cultural methods of practical reasoning and practical action may be less suited for addressing social problems con- ventionally—that is, discovering their etiology, measuring their magnitude and scope, documenting their harmful effects, or suggesting means for their eradication or amelioration—constructionist traditions of inquiry are very well suited for elucidating how it is that social problems such as inequality are made observable in social interaction and made available for study in the first place.

One important distinction within constructionist inquiries can be drawn along more or less moral and political lines. Due to their “debunking” effect or “consciousness-raising” potential, constructionist arguments are often pressed into service for social criticism and social reform, and construction- ism here takes on the flavor of ideology critique or institutional critique. In such cases, the application of constructionist insights is liable to be selective, and limited to the object of the critique (cf. Woolgar and Pawluch 1985). The group on whose behalf the critique is being offered, and the concepts used to mount the critique (e.g., inequality) and the counterfactual social relations held out as a political aspiration are often spared the dereifying, counter- essentialist logic of constructionist inquiry. As Coulter observes,

If you seek to change a social order, you must talk about it with evaluative con- structs because they alone enable you to see and describe the transformation- relevant features of the society. But you cannot claim then to be engaged in a distinctively technical discipline, because there are many critically relevant components of that vernacular conceptual apparatus which can only be employed in contestable ways. Notions such as “power”, “class”, “bureau- cracy”, “communism” and the like are not labels for unambiguously given objects in the world; they are normatively negotiable designators carrying implications and expressing commitments in every act of description or expla- nation in which they are used. . . . An attempt to argue for their “objectivity” or “scientificity” may then betray an unwillingness to acknowledge and face the consequences of engaging in a moral enterprise. (Coulter 1982, 41)

Downloaded from jce.sagepub.com by Wayne Mellinger on October 20, 2011
238 Journal of Contemporary Ethnography
In what follows, a constructionist approach to the social problem of inequal- ity is pursued with reference to ethnomethodological and conversation- analytic principles. This means not only that the study of inequality becomes the study of the pragmatic logic and methods of practical reasoning and natu- rally occurring talk about inequality but also that questions of how social relations should be described and how they should be changed, insofar as these are political questions, are bracketed as the practical concerns of mem- bers. The role of the analyst can include explicating such matters as empirical and moral phenomena but does not include ironically dismissing or profes- sionally credentialing political formulations of society and social relations. The meanings and uses of inequality therefore become true topics of inquiry rather than merely unexplicated resources for studies and critiques (cf. Zimmerman and Pollner 1971), and therefore inquiry proceeds by address- ing how inequality as an evaluative category enters into social relations rather than by evaluating social relations in terms of inequality (cf. Berard 2005a).

Ethnomethodology and the related field of conversation analysis have much to offer the social sciences, but in this article, I will single out some edi- fying contributions relevant to the study of inequality. These contributions don’t apply to all possible definitions and uses of the term “inequality” but are offered as relevant for understanding the concept of inequality as it often figures into the social sciences, with reference to social groups or members of social groups and comparisons of their respective attributes. These points will cumulatively address how the scholarly understanding of inequality can be both grounded and extended by treating inequality as a concept drawn from our natural language and drawn from commonsense knowledge of social structure and shared between members of society, including natively competent analysts.

The Reality of Inequality

To say that inequality is socially constructed is not to say that inequality is merely a subjective judgment or merely a moral or political evaluation. It is to say that inequality should not be reified1 as existing prior to and independent of social understandings and judgments about the relevant issues and ques- tions, including: how individuals can and should be identified; whether dif- ferences between individuals or relations between individuals should be understood in terms of differences between groups or relations between groups; which group memberships or group relations, if any, are relevant on a given occasion; and whether social differences should be understood as a so- cial problem such as “inequality.” All these questions can be and frequently

Downloaded from jce.sagepub.com by Wayne Mellinger on October 20, 2011
Berard / Observability of Inequality 239
are addressed in practice by members in the course of their interactions in society and their answers are often observable in their speech and social rela- tions. The concern to avoid reification in our theory and our research has arguably not been so widely shared as to effectively check the reifying ten- dencies of positivist, objectivist, and structuralist approaches, but reification is no less a fallacy for its continued practice in the social sciences. The entan- glement of contemporary social science with analytically untenable reifications has as a consequence that the avoidance of reification will have radical methodological implications for the conceptualization and research of many topics in the human sciences. But these implications don’t have to include the radical relativism that is so frequently conjured up to dismiss constructionist and qualitative approaches to social structure.

Given the controversial nature and enduring professional marginality of social constructionism, it is important to address just what implications social constructionist approaches have for understanding the ontology of inequal- ity. Discussions of constructionism often founder on the dualism of subjec- tivism and objectivism, despite Alfred Schutz’s definitive move beyond the terms of this dualism. Schutz’s introduction of the term “intersubjectivity” in his social phenomenology arguably rendered this dualism analytically obso- lete, displacing the phenomenologists’ traditional and philosophical concern with subjectivity and consciousness with the notion of intersubjective inter- pretive schemes that structure our life world. This move not only enabled the sociology of knowledge to move from ideology critique of the kind practiced by Marx or Mannheim in the direction of the sociology of commonsense knowledge, à la Berger and Luckmann’s (1966) seminal treatise on social construction, but was also pivotal for the contemporaneous ethnomethodo- logical project of elucidating cultural or ethno methods of practical reason- ing and sense-making practices, which addresses not individuals and their points of view but anonymously competent members of a culture and their conventional (intersubjective) methods of practical action and practical rea- soning.2 Some claims of inequality are liable to be dismissed as disingenuous or confused, and as Coulter (2001) observes, many macrostructural claims are certainly contestable on a variety of grounds. One of the ways such claims can be contested is to dismiss them as subjective, as might happen when a suspicion of unequal treatment emerges from a misunderstanding or incomplete understanding, but inequality is often much more than an indi- vidual’s subjective feeling or claim.

On the other hand, even when there is consensus among members about a condition or claim of inequality, inequality should not be understood as objective, if objectivity is understood in a reifying manner. Inequality is not a feature of a population like physical distribution—there are too many

Downloaded from jce.sagepub.com by Wayne Mellinger on October 20, 2011
240 Journal of Contemporary Ethnography
implicit judgments implied in inequality for it to refer to a natural kind or a natural fact. But in many cases, the existence of inequality cannot meaning- fully be questioned, even though inequality still involves implicit judgments about the relevance of group identities and group relations or implicit com- mitments to a shared moral language. In a comparative or historical analysis, such consensus would nevertheless have to be understood as relative rather than objective. Following Schutz, we can capture the nature of such consen- sus much better by describing it as intersubjective, or a shared and agreed upon aspect of our mutually experienced social world. Intersubjective agree- ment or ratification, rather than existence independent of human language and social practices, is what constitutes a social fact such as the existence of inequality, when members of society say that inequality does indeed exist. When inequality does exist in society, it exists as a social fact, which is inde- pendent of any one individual’s subjective understanding but not independ- ent of collective meanings and social conventions for language use. Social facts are neither universal nor timeless, but they are no less real for these rea- sons; the question is not whether inequality can be a real social condition or social relation but how inequality can be real.

The “Observability” of Inequality

The same characteristics that provide for the real-worldly nature of inequality provide as well for the observability of inequality, that is, the fact that inequality can be located by members and is visible to members, as a real-worldly feature of society (cf. Baccus 1986). Although phenomeno- logical and ethnomethodological scholarship often addresses how various social phenomena are available as intersubjective, accountable (observable, reportable) features of settings and society, the work of Jeff Coulter, drawing upon the traditions of Wittgensteinian conceptual analysis, ethnomethodo- logy, and conversation analysis, is especially relevant for an analytic under- standing of social structural phenomena, including inequality (Coulter 2001; Berard 2005c). Given that inequality can be a feature of intersubjective social reality, the question arises as to how (by what methods or logic) social struc- tures and social-structural relations, conceived as intersubjective phenom- ena, are observable features of society for members, and for those analysts who offer formulations of social structure in much the same way that members can be seen to do.

Coulter has suggested how macro social groups and institutions can be observable in the actions and speech of individuals in particular contexts. Coulter suggests that under certain circumstances, and in rule-governed ways, it is entirely sensible for members to speak of a number of individuals

Downloaded from jce.sagepub.com by Wayne Mellinger on October 20, 2011
Berard / Observability of Inequality 241
as constituting a collectivity (a crowd, a riot, a demonstration), or as carrying out a collective action (a platoon on the attack), or to speak of one individual as representing a collectivity (a customer service representative representing a bank), or to speak of one individual as formulating policy for a collectivity (the Pope announcing the position of the Catholic Church on abortion, the President declaring the foreign policy of the United States with respect to Haiti), or to speak of a group of individuals as representing an institution and acting on behalf of an institution (city officials representing the city of Boston and lowering the property taxes of the city) (Coulter 2001, 33-34).

If individuals and groups of individuals can constitute, instantiate, and speak for and act for collectivities in the way Coulter suggests, and clearly they can, then it is simply a matter of extending Coulter’s insights to the rela- tions between collectivities or to comparisons between collectivities, in order to illuminate inequality and a variety of other terms implicating macro social groups and macro social group relations. So by extension, group rela- tions including group inequalities are culturally observable when certain people who are observably members of certain groups are compared to or seen interacting with other people who are observably members of other groups. Thus, the relationship between a black employee and a white manager can be seen by members as an instance of race relations and racial inequality, and the relationship between a woman and her abusive husband can be seen as an instance of gender relations and gender inequality. Each individual can be seen as a member of a different group, and therefore in their relations with each other they can be seen as acting out racial or gender relations, in a man- ner Schutz would call typical or typified (cf. Schutz 1964, 237). Simply put, group relations become observable when people who are observably mem- bers of different groups can be observed relating to each other, and inequality becomes observable when people who are observably members of different groups can be observed to differ from each other in those ways that are tied to or suggested by conventional, typified group distinctions, such as income, wealth, power, status, social capital, cultural capital, etc.

This is not to say that every interaction between a white person and a black person is an instance of race relations, or every interaction between a man and a woman is a gender relation; when it is noted that such interactions can be seen as racial or gender relations, this is not to say that they will be or they should be, simply that the co-presence of individuals from different catego- ries satisfies one basic criterial feature of inequality and thus provides one of the grounds that can make inequality an observable feature of a setting or action, or an intelligible description or complaint.
With respect to controversial, evaluative terms such as “inequality” and “discrimination,” it may be helpful to conceive of observability as a matter of

Downloaded from jce.sagepub.com by Wayne Mellinger on October 20, 2011
242 Journal of Contemporary Ethnography
degree, or as a claim liable to provoke counterclaims, although Coulter prop- erly introduces his observations with less controversial examples of collec- tive identity and collective action. It may not be controversial now to say that in Montgomery, Alabama, in 1955, a white bus driver demanding a seated black passenger give up her seat for a boarding white passenger is observable as racial inequality. An important part of what makes this interaction observ- able as racial inequality, however, are the racial identities of the three individ- uals involved: the seated passenger described as black and the driver and boarding passenger described as white. Change the race of the driver and it becomes less clearly observable as racial inequality. Change the race of either the seated passenger or the boarding passenger, and it becomes quite difficult to see the interaction as racial inequality. Change the race of all three and it arguably may be seen as a quite different type of inequality—“reverse dis- crimination.” Formulate the identity of the boarding passenger as handi- capped or elderly instead of white, and it also becomes much more difficult, or at least much more controversial, to see the interaction in terms of race relations or racial inequality. The point is not to put empirical weight on an arm- chair discussion of hypotheticals but to note that there are logico-grammatical criteria and social conventions for the intelligible use of macrostructural terms such as “inequality.” In different contexts, the status of claims about inequality can vary from unquestionable to outspoken or incomprehensible.

Just because concepts such as inequality do not refer to objective features of our society does not mean that macrostructural entities or actions or rela- tions are difficult to observe. It can be quite easy for laypeople to observe inequality in society, and many such observations cannot be meaningfully questioned; but what makes inequality observable is not that it is an object in our field of vision but that individuals and small groups who might actually be in our field of vision can and do occasionally relate to each other in ways consistent with our typified knowledge of macrostructural group differences and group relations, in which case actions and interactions in one’s field of vision can be understood as local, observable instances of macrostructural social relations such as racial inequality or gender inequality or class inequality.

Inequality as a Member’s Formulation and as an Analytic Formulation

Descriptions of people as members of certain groups or representatives of certain institutions are often treated in the social sciences as matters of fact, without regard for whether particular group memberships or institutional positions are the operative, relevant identities in specific contexts or courses

Downloaded from jce.sagepub.com by Wayne Mellinger on October 20, 2011
Berard / Observability of Inequality 243
of action. But identity can also be understood as a local, practical orientation and achievement of the members of a culture in a particular context, consis- tent with methodological principles in ethnomethodology and conversation analysis (see, especially, Antaki and Widdicombe 1998; Widdicombe 1998), among many other traditions of qualitative social inquiry as well. Just as identity can be respecified in terms of how identity is formulated by mem- bers, so social relations can be respecified in terms of how they are formu- lated by members. Just as an identity such as poor, or woman, is not always relevant, social relations such as class relations or gender relations are not always relevant (cf. Berard 2005b).

A seminal insight in the field of membership categorization analysis was simply that individuals can correctly be described as members or incumbents of an indefinite number of categories, drawn from category devices such as class, race, gender, age cohort/generation, nationality, religion, profession, and an indefinite number of others that get less attention from sociologists but can be the most relevant identities in certain contexts—jazz enthusiasts, bloodhound owners, Sierra Club members, legal drivers, persons trained on the cash register, people with B negative blood type, friends with a VCR, etc. Given the indefinite number of identities that could correctly be ascribed to any individual (the “multiple category incumbency of persons”), the mere correctness of an identity does not begin to warrant an analyst to speak of one of these identities as the relevant identity for understanding what is happen- ing, how it is happening, or why it is happening; individuals have their own relevances, which should be neither legislated nor ignored by analysts (cf. Schutz 1964, 253-54). Unless the identity of persons is treated as a question to be answered empirically, in terms of members’ orientations and relevances, observable in their talk, their actions, and their interactions, then the identity of “subjects” or members will all too often be formulated by the analyst in a stipulative, theoretically driven manner that imposes a social structure on the data rather than explicating the social structure observable within the data. The possibility and heuristic value of treating social identi- ties and social-structural group relations empirically, as the occasional and practical concerns of members observable in their talk and interactions, will be suggested in some (necessarily brief) examples toward the Conclusion, below.

In one ethnomethodological study relevant to inequality, I have analyzed the Japanese Exclusion Order issued by American military authorities dur- ing World War II in terms of how identity categories are used by lawyers and Justices in the famous Korematsu case (Berard 2002). It would be very easy, and correct, to analytically identify internees either in terms of their racial/ ethnic heritage or in terms of their nationality, and it would be easy and cor-

Downloaded from jce.sagepub.com by Wayne Mellinger on October 20, 2011
244 Journal of Contemporary Ethnography
rect to analytically identify the relevant context either as an international mil- itary conflict or as domestic American race relations. But attention to how identity and context are accomplished by the participants to the legal case themselves reveals that categories and contexts are formulated differently by different parties, in an interesting and structured manner. Critiques of the Exclusion Order could be structured by reference to racial identity and race relations, emphasizing racial discrimination, and defenses could be struc- tured by reference to foreign or dual nationality and international war, emphasizing an ostensible danger of espionage and sabotage. This analysis therefore suggests that it is much more rewarding, analytically and empiri- cally, to consider relevant identity and relevant context as the work of partici- pants rather than background facts that can be described without reference to the participants’ own demonstrable orientations.

Just as identity and context can be respecified as the concerns and achievements of members, so too can other topics be respecified as the con- cerns and achievements of members. Indeed, most of the central topics in sociology originated with lay concepts and remain curiously close to these mundane origins; hence our publications, our classes, and the sections in our associations deal with such lay concerns and lay concepts as economics, pol- itics, race, religion, crime, urbanization, family, media, etc. “Stratification” is certainly an impressive specimen of professional jargon, but there is no doubting that this pillar of sociological inquiry is squarely rooted in a much less specialized, much less professional concern with social inequality. Members of our culture do not need to enroll in sociology courses to know what inequality is, to develop critiques or defenses of it, and to describe social conditions or social actions by formulating them as instances of inequality. Inequality is a lay concept that laypeople can certainly use to describe or to challenge social conditions and social relations; the question is when do they speak of inequality, and in any given situation, whether they speak of inequality or otherwise orient to inequality as a feature of the local context.

Just as people do not always understand their identity in terms of a certain socioeconomic class, a certain race, or a certain gender, people do not always understand their position in society and their relations with others in society in terms of class relations, race relations, or gender relations. Given that peo- ple have these identities and have these social positions and have these social relations, and given that people are capable of formulating their identities in these ways, and capable of formulating their social positions and social rela- tions in these ways, it is certainly worth considering that people don’t always formulate their identities in these ways, and don’t always understand their social positions and social relations in these ways. Given that inequality is a

Downloaded from jce.sagepub.com by Wayne Mellinger on October 20, 2011
Berard / Observability of Inequality 245
lay concept and that this concept could sensibly be used in countless circum- stances, it is certainly important to consider that people aren’t using this term to formulate their positions or their concerns in many circumstances where they clearly could do so.

Because most analysts of society share with their “subjects” a native com- petence with native concepts and native language, analysts often enjoy an invaluable ability to understand social conditions and social interactions by means of terms that they share with the “subjects” of their studies or cri- tiques. Analysts can formulate the identities of their subjects, their social positions, and their social relations, all with the same vocabulary and cultural competence as the subjects themselves. Given that these formulations can be made competently, correctly, and even “professionally,” it is not always required that these formulations be made relevantly, in terms of demonstra- ble relevance to participants. But a shared cultural competence and shared vocabulary does not guarantee agreement about which identities are relevant, and which terms are appropriate, to formulate persons’ identities and rela- tions in any specific context. A social constructionist approach to inequality would seem to require a principled attention to the question of what “counts” as inequality in society, and especially the questions of when and how and why people formulate social relations by using the term “inequality.” Such studies could take the shape of “micro,” qualitative studies of interactional data (cf. Harris 2004, 131) or more “macro” studies such as intellectual histo- ries of concepts of inequality in society (cf. Harris 2003, 225). But in either case, constructionist studies of inequality have to come to terms with the ramifications of inequality as a members’ category.

Inequality as a Gloss

Both the Wittgensteinian tradition and the ethnomethodological tradition recognize that language use is not determined by context but that the flexibil- ity of language allows for alternative or disjunctive descriptions of “the same” phenomena in the “same” context. I can (when wearing my member’s hat) call my neighbor’s pets “beagles” or “dogs,” and I can call my state’s system of public school financing “unequal” or “discriminatory.” This can be appreciated as a remarkably obvious point, but the methodological signifi- cance should not be underestimated. There are many (many many) times when inequality is and will be used in social science as a gloss (a catch-all category, an umbrella term, etc.) in reference to data in which members do not say or write “inequality.” There are legitimate theoretical and editorial reasons why such a gloss may be edifying, interesting, or necessary in a wide

Downloaded from jce.sagepub.com by Wayne Mellinger on October 20, 2011
246 Journal of Contemporary Ethnography
range of contexts. But such glosses may serve to identify a relevant theoreti- cal context, a relevant field of literature, or to convince readers of the signifi- cance of a study, much more than they assist in the actual analysis of the data. There seems to be a widespread neglect of the language that members use to display their understanding of themselves and their positions in society. And this problem is not necessarily obviated by the collection of relevant data and the analysis of transcripts, precisely because analysts can and often do offer their own formulations of the identities and relations ostensibly represented by their data, instead of explicating the formulations that are present in, and constitutive of, their data (cf. Harris 2003, 222).

When one is confronted with data, whether it be recordings of naturally occurring talk-in-interaction, as favored by conversation analysts; or that sta- ple of the social sciences, the research interview; or textual documents from social institutions; or productions of various media; or experiments; or any number of other types of data, one feature of data about inequality is simply this: much data that is about inequality is in a sense not about inequality. With recent advances in technology, it can be rather easy to find material of one kind or another in which someone or some institution is literally and explic- itly addressing inequality—it is frequently as easy as accessing a large data- base and using “inequality” as a search term—but if one asks a graduate assistant to code interviews or newspaper articles as to whether or not some- one or some text addresses inequality, one might begin to appreciate that inequality is indeed a gloss. Inequality is often, and rightly enough, under- stood as a massive, long-lived and multifaceted social fact that can be indi- cated or represented by (or “operationalized” in terms of) any number of other terms that might literally appear in transcripts or documentary sources, terms such as “discrimination,” “bias,” “racism,” “patriarchy,” “class structure,” “poverty,” “deprivation,” “underclass,” “imperialism,” “hegemony,” “privilege,” “segregation,” “exclusion,” “stigma,” “hierarchy,” “social ladder,” “caste,” “sta- tus,” “power,” or many other terms.

The term “inequality” is used as a gloss by members, and it is used as a gloss by virtually all analysts or scholars of inequality as well, arguably according to the very same logic. That is to say, people can and do claim that cases or issues are about inequality even when participants are literally talk- ing about racial discrimination in housing, or a glass ceiling for women in middle management, or the inability of homosexual couples to enjoy the legal protections afforded by the institution of marriage. Using inequality as an analytic gloss on empirical data is not necessarily problematic but clearly can have the effect of displacing attention away from the language actually used in one’s data. If our data do not include members formulating their com- plaints or contexts by means of the term “inequality,” then methodologically

Downloaded from jce.sagepub.com by Wayne Mellinger on October 20, 2011
Berard / Observability of Inequality 247
it becomes necessary for analysts to acknowledge the mundane members’ methods involved in the analyst’s own observations of the data as relevant to the question of “inequality.” This methodological problem has been ad- dressed with respect to the study of deviance, in a discussion of how partici- pants’ talk can be recognized as referring to deviance when it is not literally referring to “deviance” (Hester 1992), but it is still a remarkably neglected premise of social scientific methodology and reasoning. The study of inequality would be transformed and enriched by an understanding that research thus far has been built on a foundation of analytic glosses rather than on empirical observations of inequality as a members’ formulation.

Three Brief Examples

The conceptual and methodological points raised in the above discussion deserve to be and will be fleshed out with reference to existing and future studies, but such studies are largely beyond the scope of these immediate comments, which are in many ways necessarily preliminary and program- matic. I will offer three brief examples, however, to illustrate how the study of inequality might develop in a constructionist direction, consistent with the analytic contentions offered above, and emphasizing language use as a reflexively constitutive feature of social relations and social structure.

The first is taken from a textbook treatment of race and ethnic relations, one of the few to display ethnomethodological sensibilities about group rela- tions and inequality. Rod Watson concludes a brief treatment of ethnographic and interactional analyses of racial and ethnic relations with a discussion of an excerpt (Watson 1984, 63) originally published in Speier’s (1973) How to Observe Face to Face Communication, featuring a brief exchange between a black doctor (D) and a white police officer (P) in the Deep South many years ago:

P: “What’s your name, boy?”
D: “Dr. Poussaint. I’m a physician” P: “What’s your first name, boy?” D: “Alvin.”

Drawing upon Watson’s (1984) brief analysis, we can say there are two membership category devices operating here (at least): the membership cate- gory device of profession and that of race. Alvin Poussaint is an incumbent of a professional category that enjoys more status than does the category of police officers, and he is an incumbent of a racial category that enjoys less status than the category white. The white officer, by addressing Alvin Pous-

Downloaded from jce.sagepub.com by Wayne Mellinger on October 20, 2011
248 Journal of Contemporary Ethnography
saint as “boy,” employs what was at the time a conventional means of main- taining the color line, and Watson observes that this achieves a degradation or a downranking of Alvin Poussaint’s racial status by employing a term of address originating in adult-children interactions, in effect reproducing racial inequality with implicit reference to the model of generational inequal- ity. Each individual orients to a different set of identities as relevant for the purposes of establishing whether the etiquette of the encounter would be informed by racial etiquette or professional etiquette, as is suggested when Alvin Pouissant chooses to formulate his identity in terms of a professional title or status that implicitly challenges the appropriateness of being addressed as “boy.” Watson suggests that this type of interaction, which involves multiple individuals negotiating the contextual relevance of differ- ent group identities tied to disjunctive social hierarchies, constitutes what he calls “stratification practices” (Watson 1984). This is a term that the study of inequality might yet choose to rediscover. It correctly identifies that crucial feature of inequality: that it is constructed or achieved in and through prac- tices, including “mere” talk.

In an entirely different context, but also informed by ethnomethodolog- ical principles, Paul ten Have discusses “how asymmetries are produced in and through the details of physicians’ and patients’ situated interactions” (ten Have 1991, 138). The question of asymmetry is perhaps raised for the same reason that it is considered problematic in much of the literature; ten Have notes that “a ‘morality of equality’ is hard to avoid” (ten Have 1991, 139), but it is quite possible to attend to inequality or asymmetry as an accomplishment of members in local settings, rather than as a target for the analyst to critique or reform (ten Have 1991, 140). Ten Have summarizes some generalizations about doctor-patient encounters that are largely consis- tent with the conventional view that the doctor-patient relationship is asym- metrical. For example, after the patient’s initial appearance at an appoint- ment, physicians more often than patients take the initiative with questions, orders, and proposals; such initiative from patients seems to be “dispreferred” in medical encounters. These initiatives, disproportionately taken by doc- tors, prompt specific kinds of responses from patients and often prompt responses in the form of short factual answers, reinforcing the interactional control that can be established by the asymmetry in initiative between doc- tors and patients (ten Have 1991, 140).

While ten Have agrees with these generalizations qua generalizations, he observes that these asymmetries are locally produced by both doctors and patients in the contingencies of their talk-in-interaction, that “both of them have possibilities for less asymmetrical interaction” (ten Have 1991, 149), and that patients can be observed at times to circumvent the interactional

Downloaded from jce.sagepub.com by Wayne Mellinger on October 20, 2011
Berard / Observability of Inequality 249
control that can be effected by doctors’ questioning. The patient can in vari- ous ways assume the role of an informant rather than a respondent (ten Have 1991, 145), implying “a more active role for the patient, as a teller of his or her own story” (ten Have 1991, 143).

As an example, ten Have provides the following translated transcript of a medical encounter, between a doctor (“D”) and a patient (“P”):
17 D: That has been quite a history with you in uh 18 January huh
P: Yes ( )
20 (.)

21 D: And that has then that has been (.) going alright in the meantime? 22 P: Yes
23 ((telephone call interruption))
P: (it’s just that) for a couple of months I
25 had (.) a cold and then it was over and then it was gone 26 D: Yes
P: And Sunday I suddenly got uh (.) a real headache 28 D: Yes
29 P: (I had it all-) all over this side
D: Yes
P: And Sunday night I got uh pain in my chest here 32 D: With coughing or something or whatever-

As ten Have (1991) notes, in line 21 the doctor, as he is finishing reading the patient’s records, and presumably referring to the last entry, asks about what has happened since then, providing for the patient to tell the story. The patient responds with a story, and with a story that is made longer than might be necessary in a variety of ways, including by starting the story earlier in time than might have been expected (line 24), and by mentioning “relatively permanent states” (lines 24-25) that figure as background information that does not yet get to the matter at hand, and by seemingly constructing a list, comprising a cold, and then a headache, all before the story is evidently fin- ished, “which again heightens the expectation of more to come” (ten Have 1991, 145). The doctor, for his part, seems to patiently wait for the patient to finish the story, and a series of brief “yes’s” signal his sustained attention to the unfolding story.
After examining doctor-patient interactions with a respectful attention to when, how, why, and by whom asymmetries might be oriented to or accom- plished in particular contexts, ten Have (1991) suggests “consultations are sometimes almost like conversations. At other times they resemble interro- gation. But mostly they are somewhere in between, zigzagging between the

Downloaded from jce.sagepub.com by Wayne Mellinger on October 20, 2011
250 Journal of Contemporary Ethnography
two poles in a way that is negotiated on a turn-by-turn basis by the partici- pants themselves” (p. 162). His analysis illustrates his prefatory observation that “Institutional structures are not only external constraints on participants’ actions; they are also actively used as a resource for those actions” (ten Have 1991, 139).

Comparing these two illustrations, several points can be made. First, in neither of these cases do participants formulate their social relations in terms of equality or inequality. Even the analysts referenced here discuss the data in terms of stratification and asymmetry, respectively, rather than equality and inequality. But both interactions could be glossed in terms of equality and inequality, either in vernacular or analytic formulations, due simply to the possibility of identifying the individuals according to pairs of categories (white/black, policeman/doctor, doctor/patient), which can be hierarchically ordered in terms of unequal status, unequal power, unequal knowledge, etc. In noting the possibility of glossing these interactions in terms of equality and inequality, I am not offering glosses as empirical observations or empiri- cal analysis, which they are not, but drawing attention to a highly significant conceptual and methodological issue. Many data that do not satisfy empirical standards of relevance in terms of the demonstrable orientation of partici- pants to the concept of inequality do nevertheless satisfy the conventional (logico-grammatical) criteria for making accountably sensible, intelligible references to inequality. While it is crucial to distinguish empirical examples of talk about inequality from empirical examples of talk that can be glossed as illustrating inequality, this is not to say that the study of inequality could or should be limited to data in which participants explicitly use the terms “equality” or “inequality.” Such an exclusive methodological standard would drastically impoverish the study of inequality. So long as analytic glosses such as inequality are not read into the data as empirical observations and analysis, as is done in much critical inquiry, glosses can still play an invalu- able role in analysis in that they allow analysts to identify and discuss data that may satisfy cultural and conceptual criteria of relevance with respect to inequality even if data don’t satisfy the empirical criterion of observable, demonstrable relevance to participants, for example, through participants’ explicit usage of the term “inequality.”

Thus, although neither of the first two examples features inequality as a participant’s formulation endogenous to the data, the data nevertheless allow for observations that are quite topical to the discussion of equality and inequality. For example, the relevant, operative identity of each individual in these data is not an a priori demographic fact about these individuals, but a matter of contingent personal orientation and social interaction. Understand- ably, in the police-citizen encounter, the black doctor prefers his status as

Downloaded from jce.sagepub.com by Wayne Mellinger on October 20, 2011
Berard / Observability of Inequality 251
doctor be recognized as relevant, and in the doctor-patient encounter, the patient welcomes the opportunity to discuss as-yet-undocumented medical history and complaints in a rather unstructured and informal manner, taking the more active role of an informant rather than the more passive role of respondent (ten Have 1991, 143-45). It should be clear, though, that not all white officers will ignore the professional identities of black citizens, and not all black citizens would challenge, even implicitly, how a white police officer addresses them; similarly, patients may not typically enjoy informal, egali- tarian conversations in medical encounters—sometimes the doctor may not allow this, and other times the patient may not seek or respond to such oppor- tunities. Whether an encounter might be seen as illustrating the perpetuation of inequality, or as illustrating an avoidance or lack of inequality, the relevant identities and the relations between individuals can be seen as a feature of a local setting, accomplished in talk and in interaction, rather than settled in advance by the decontextualized demographic characteristics and macro- social group relations that are thought to be omnirelevant in a variety of structuralist and critical methods in the social sciences. It is not predeter- mined that an encounter will revolve around the relevance of any one set of identities or social relations, or that interactions will unfold according to a logic of repression, although structuralist and critical approaches can easily find instances that are amenable to such a hermeneutics of stratification.

A third example, which explicitly addresses inequality, can also be dis- cussed briefly. In this case, the researcher is interested explicitly in inequality and prompts the interviewee to discuss her marriage in terms of equality and inequality (Harris 2001). While interviews are not naturally occurring con- versation, the analyst has not schooled the interviewee in the meaning of inequality and doesn’t need to; as is clear in her responses, she displays her understanding of equality and inequality in terms that are both private in their details and yet recognizably conventional in their logic. The respondent, Deborah, says at one point,

. . . And I started asking him questions why we weren’t equal. And he gave a lot of reasons why we weren’t equal and it was pretty much well, I came into the relationship with nothing, with nothing. And he pretty much put a roof over my head. I was young. I was young, and he already had a job, and pretty much he had everything I didn’t have. And so therefore we weren’t equal . . . [That’s how he felt] even though I was working and I had my own money. He’s more of a old-fashioned type of guy . . . that he’s the man of the house and that’s how it goes. (Harris 2001)

Downloaded from jce.sagepub.com by Wayne Mellinger on October 20, 2011
252 Journal of Contemporary Ethnography
In this statement of Deborah’s relationship with her husband, a number of different identity categories are brought into play as displaying both the rea- sons for inequality in Deborah’s marriage, and more fundamentally display- ing Deborah’s understanding of inequality. She mentions differences in housing or property and differences in income and relates these to differences in age and differences in gender. Any of these could illustrate inequality in and of themselves, or any of them could be understood as more relevant than the others, or as structuring the others; Deborah and seemingly her husband seem to understand these myriad differences primarily with reference to gen- der categories; however, all these differences are summed up by Deborah when she says “he’s the man of the house.” This is entirely consistent with a cultural knowledge of what an “old-fashioned” marriage might consist of, including an age disparity between an older man and a younger woman, an income disparity that also favors the man, and a household division of labor structured largely in terms of dependence and control. Since it is Deborah’s own understanding of equality and inequality, and her husband’s, which informs and in a sense socially structures her marital relations, it seems only proper to follow Harris in attending to how inequality is used by Deborah as a sense-making device in her marriage, rather than as a technical concept to impose upon the interview in a stipulative fashion.


The above contentions and examples suggest that inequality is socially constructed, suggest how inequality is socially constructed, and suggest how inequality can be conceptualized and researched as socially constructed. The implications for the analysis and empirical study of inequality can be briefly summarized by way of conclusion.
The question of how inequality exists is not merely a question for the phi- losophy of social science—it is also relevant to the selection of appropriate research methods for the study of inequality. Inequality as a social fact can- not be understood analytically without appreciating that inequality is a con- cept, and without appreciating that members’ understandings of inequality as a meaningful type of social relation are constitutive elements of the social relations referred to as inequality. The concept of inequality is constitutive of social relations in a variety of ways, but among them we must consider how the word “inequality” is used for a variety of practical purposes, including criticizing differential conditions, differential treatment, or differential opportunities; attempting to reform such inequalities; or seeking compensa- tion or recognition for having suffered such inequalities. Although there are

Downloaded from jce.sagepub.com by Wayne Mellinger on October 20, 2011
Berard / Observability of Inequality 253
constraints on when inequality can be referred to sensibly, not to mention plausibly or authoritatively, the very claim of inequality is capable of chang- ing the social context in which it is used, by explicitly or implicitly claiming the contextual relevance of certain group identities and group relations, or suggesting the possibility that reform or redress is being sought. Just as a complaint can suffice to change the meaning of a conversation, a complaint of inequality can suffice to change the meaning and course of social relations in any number of contexts.

The essentially meaningful nature of “inequality,” and the fact that the term involves implicit judgments of relevant identities and relevant differ- ences, does not entail a dismissive or subjectivist relativism with respect to social problems such as poverty, racial discrimination, or violence against women. Such objections to constructionism need to be examined in their own turn. It is these critics of constructionism who suggest that a phenome- non cannot be taken seriously unless it is thought to exist independent of lan- guage, social practices, and human judgment, as if the human nature of social phenomena entailed an ontological denigration of some kind. Many con- structionists are either satisfied in distinguishing their politics from their sociology or satisfied that social conventions, traditions, and principles (democracy, human dignity, rule of law, equality, etc.) provide as much moral ground for judgments about social justice as might ever be needed, without agonizing over whether these principles are guaranteed by transcendentalist ethics, codified in natural law, or hard-wired into a universal human reason. When claims that social reality is socially constructed are rejected on the basis of ahistorical, universalist reasoning communicated from an implied Archimedean point, it is not the constructionists who should have to do the explaining.

Finally, and following from all that has preceded, inequality is no less observable for being a social construction or for being a relational property between multiple groups. Collective identity and collective action and rela- tions between collectivities have mystified and vexed the social sciences pri- marily because the social sciences have neglected to study how these phe- nomena are accomplished by laypeople using mundane cultural methods of practical reasoning. Social scientists and social theorists have been all too willing to employ these ethno methods as implicit resources that provide them with their phenomena but all too slow to treat these methods of sense making as phenomena in their own right.

Future studies of inequality would do well to remember that inequality is a formulation of social relations that members are competent to offer for and amongst themselves. Analytic formulations of inequality may be useful or

Downloaded from jce.sagepub.com by Wayne Mellinger on October 20, 2011
254 Journal of Contemporary Ethnography
interesting in a variety of contexts, but such formulations will have a prob- lematic analytic status if they are inconsistent with members’ own formula- tions, because it is members’ formulations of their concerns and their con- texts that are constitutive of these concerns and their contexts. Because members don’t talk about inequality in the vast majority of cases where ana- lysts would like to speak of inequality, scholars of inequality can easily be tempted to employ inequality as a gloss on their data. Such glosses have their uses but don’t carry the kind of methodological and theoretical weight fre- quently placed upon them. Whether inequality is a gloss that illuminates fea- tures endogenous to the data or whether inequality is a gloss that imposes the analyst’s relevances upon the participants’ words and actions—this is neces- sarily a judgment that has to be made on the basis of detailed familiarity with participants’ words and actions, by analysts who are sensitive to the differ- ence between finding order in one’s data and imposing order on one’s data.

The conceptual points made above rightfully led to examples of different types of empirical research that might inform the study of inequality. Unfor- tunately, there is a problematic lack of research in which participants in natu- rally occurring talk-in-interaction formulate their concerns as concerns about “inequality,” literally and explicitly. This type of data would clearly have advantages for understanding how inequality, specifically, figures into mundane social interaction and commonsense knowledge of social structure. Nevertheless, there is an enormous wealth of data that we can recognize (and fairly gloss) as being related to the study of inequality. The illustrations from the interracial police-citizen encounter and the doctor-patient consultation were both discussed as having some analytic relevance to the study of inequality, despite the absence of any formulation of inequality in the tran- scripts. Another source of data that is even more topical is the loosely struc- tured interview, in which the interview is formulated as an interview about a topic such as “inequality in marriage,” and then proceeds by soliciting the interviewee’s own understanding of the research topic (Harris 2001).

Social construction can be understood as an observable social process and social achievement, involving a reflexive interplay between social concepts, social practices, and social contexts. If inequality can be analytically respecified in this manner, the studies that follow will benefit from both greater conceptual clarity and greater empirical grounding. The challenge and the promise is to marry sociolinguistic and social-structural questions in our research, following the recognition that they are already married in our subject matter.

Downloaded from jce.sagepub.com by Wayne Mellinger on October 20, 2011
Berard / Observability of Inequality 255

1. On reification, see especially Maynard and Wilson (1980).
2. This brief take on some questions of intellectual history is very much indebted to unpub- lished comments from Jeff Coulter, but compare Heeren (1971).

Antaki, C., and S. Widdicombe. 1998. Identity as an achievement and as a tool. In Identities in talk, edited by C. Antaki and S. Widdicombe, 1-14. London: Sage.
Baccus, M. D. 1986. Sociological indication and the visibility criterion of real world social theo- rizing. In Ethnomethodological studies of work, edited by H. Garfinkel, 1-19. New York: Routledge and Kegan Paul.
Berard, T. J. 2002. “Japanese American” identity and the problem of multiple description: Dis- junctive versions of the Japanese Exclusion Order. In Language, interaction and national identity: Studies in the social organization of national identity in talk-in-interaction, edited by S. Hester and W. Housley, 144-68. Burlington, VT: Ashgate.
———. 2005a. Evaluative categories of action and identity in non-evaluative human studies research: Examples from ethnomethodology. Qualitative Sociology Review 1:5-29.
———. 2005b. On multiple identities and educational contexts: Remarks on the study of inequalities and discrimination. Journal of Language, Identity, and Education 4:67-76. ———. 2005c. Rethinking practices and structures. Journal for the Philosophy of the Social Sci-
ences 35:196-230.
Berger, P., and T. Luckmann. 1966.
The social construction of reality: A treatise in the sociology
of knowledge. Garden City, NY: Anchor.
Coulter, J. 1982. Remarks on the conceptualization of social structure.
Philosophy of the Social
Sciences 12:33-46.
———. 2001. Human practices and the observability of the “macro-social.” In
The practice turn
in contemporary theory, edited by T. Schatzki, K. Knorr Cetina, and E. Von Savigny, 29-41.
New York: Routledge.
Harris, S. 2001. What can interactionism contribute to the study of inequality? The case of mar-

riage and beyond. Symbolic Interaction 24:455-80.
———. 2003. Studying equality/inequality: Naturalist and constructionist approaches to equal-

ity in marriage. Journal of Contemporary Ethnography 32:200-32.
———. 2004. Challenging the conventional wisdom: Recent proposals for the interpretive study

of inequality. Human Studies 27:113-36.
Heeren, J. 1971. Schütz and the sociology of commonsense knowledge. In
Understanding every-
day life: Toward the reconstruction of sociological knowledge, edited by J. Douglas, 45-56.
London: Routledge and Kegan Paul.
Hester, S. 1992. Recognizing references to deviance in referral talk. In
Text in context: Contribu-
tions to ethnomethodology, edited by G. Watson and R. Seiler, 156-74. London: Sage. Maynard, D., and T. Wilson. 1980. On the reification of social structure. Current Perspectives in
Social Theory 1:287-322.
Schutz, A. 1964. Equality and the meaning structure of the social world. In
Collected papers II:
Studies in social theory, edited by A. Brodersen, 226-73. The Hague, the Netherlands: Martinus Nijhoff.
Downloaded from jce.sagepub.com by Wayne Mellinger on October 20, 2011
256 Journal of Contemporary Ethnography
Speier, M. 1973. How to observe face to face communication. Pacific Palisades, CA: Goodyear. ten Have, Paul. 1991. Talk and institution: A reconsideration of the “asymmetry” of doctor- patient interaction. In Talk and social structure: Studies in ethnomethodology and conversa- tion analysis, edited by D. Boden and D. Zimmerman, 138-63. Berkeley: University of
California Press.
Watson, R. 1984. Racial and ethnic relations. In
Applied sociological perspectives, edited by R.
Anderson and W. Sharrock, 43-65. Boston: Allen and Unwin.
Widdicombe, S. 1998. Identity as an analysts’ and a participants’ resource. In
Identities in talk,
edited by C. Antaki and S. Widdicombe, 191-206. London: Sage.
Woolgar, S., and D. Pawluch. 1985. Ontological gerrymandering.
Social Problems 32:214-27. Zimmerman, D., and M. Pollner. 1971. The everyday world as a phenomenon. In Understanding
everyday life: Toward a reconstruction of sociological knowledge, edited by J. Douglas, 80- 103. London: Routledge and Kegan Paul.

T. J. Berard is an assistant professor of justice studies at Kent State University in Ohio. His research addresses the relevance of qualitative methods for the study of macro-structural group relations and collective action, especially inequality and discrimination, and also addresses the centrality for the human sciences of phenomena including labeling, classifying, and categoriz- ing, considered both as research topics and as research methods. These issues are explored largely within the sociology of law and sociology of deviance.
Downloaded from jce.sagepub.com by Wayne Mellinger on October 20, 2011