Kevin Whitehead's Categorizing the Categorizer: The Management of Racial Common Sense in Interaction"

"Categorizing the Categorizer":
The Management of Racial Common Sense in Interaction*

Kevin A. Whitehead
University of California, Santa Barbara

* Address correspondence to Department of Sociology, University of California, 3005 Social Sciences & Media Studies, Santa Barbara, CA 93106-9430; This research was supported in part by a South African National Research Foundation Prestige Scholarship for Doctoral Study Abroad. I am deeply indebted to Geoff Raymond and Gene Lerner for their detailed and incisive comments on earlier versions of this paper. Helpful feedback on earlier drafts was also provided by Don Zimmerman, Emanuel Schegloff, and Howard Winant. All shortcomings and errors remain, of course, my own. I would also like to express my gratitude to Linda Croyle for allowing me access to the videotapes that formed the data set for this study. It would not have been possible for me to undertake this research without her kind assistance. Portions of this report were presented in March 2007 at the Annual Meeting of the Pacific Sociological Association in Oakland, California, May 2007 at the Annual Conference of the International Communication Association in San Francisco, California, and August 2007 at the Annual Meeting of the American Sociological Association in New York City.

In this paper, I consider one mechanism by which racial categories, racial "common sense," and thus the social organization of race itself, are reproduced in interaction. I approach these issues by using an ethnomethodological, conversation analytic approach to analyze a range of practices employed by participants of a "race-training" workshop. These practices manage the normative accountability involved in referring to the racial categories of others when describing their actions, and thus in using racial common sense in talk-in-interaction. This accountability arises in part because a speaker's use of a racial category to explain someone else's actions may serve as an invitation for recipients to treat the speaker's own racial category as relevant for understanding and assessing the speaker's actions. I describe three main ways in which speakers can manage this accountability, namely generalizing race, localizing race, and alluding to race. My analysis shows that, even in attempting to resist racial common sense in accounting for their own actions and those of others, speakers orient to race as a normative framework according to which individuals will produce their own actions and interpret those of others, and thus reproduce it as relevant for understanding social action. This research contributes to advancing knowledge in the fields of ethnomethodology, conversation analysis, racial studies, and categorical inequality.

In this paper, I show that understanding the ways in which racial categories, and common sense knowledge about them, are deployed and oriented to in interaction can provide insights into the mechanisms through which they are maintained and reproduced. These findings contribute to three areas of sociology, namely the study of racial discourse, "racial formation" (Omi and Winant 1994), and "durable inequality" (Tilly 1998). I briefly discuss the relationship between my study and these areas in the paragraphs that follow.
The recognized importance of studying racial discourse has resulted in the accumulation of a substantial and growing body of literature in this area in the past two decades (Condor, Figgou, Abell, Gibson, and Stevenson 2006; van den Berg, Wetherell, and Houtkoop-Steenstra 2004).This research has contributed to examining race talk in terms of its implications for understanding expressions of racial "attitudes" or prejudice, and for the ways in which such expressions relate to broader systems of racial inequality. While the research I report in this paper shares with much of the race talk literature an emphasis on fine-grained attention to empirical materials, the substantive focus and analytical aims are somewhat more basic: rather than examining race talk for its relation to prejudice or racism, I focus on the reproduction of the category system on which such social problems rest (cf. Whitehead and Lerner 2009).
Omi and Winant'sinfluential theory of racial formation describes the "sociohistorical process through which racial categories are created, inhabited, transformed, and destroyed" (1994:55). Omi and Winant argue that racial formation occurs through "racial projects," which occur both at the "macro-level" of social structural forces, and at the "micro-level" of everyday experience and interaction. While the reproduction of widespread notions of race through everyday interactions is an important matter in terms of Omi and Winant's theory, the authors do not elaborate on the mechanisms through which such reproduction of the social organization of race takes place, focusing instead largely on racial projects at the level of the state and in social movements. In this paper I take up this question of the reproduction of race in interaction, identifying and explicating a set of practices that implicate a mechanism through which it occurs. Tilly's (1998) study of durable inequality is concerned with accounting for the enduring nature of many forms of categorical inequality, i.e., inequality based on social categories such as gender, ethnicity, and race. Tilly argues that changes in prevailing attitudes towards members of "out-group" categories will have relatively little effect on durable inequality, while "the introduction of new organizational forms – for example, installing different categories or changing the relation between categories and rewards – will have great impact." (1998:15). As a consequence, describing the mechanisms through which category systems are reproduced is a central concern for Tilly. While the mechanisms Tilly identifies involve instances in which individuals and organizations, including oppressed groups, become invested in or "buy into" the category systems in question, the mechanism I explicate accounts for how one such category system, race, can be reproduced even in the absence of any investment or stake in it on the part of those who use it. I turn now to a discussion of the basis of this mechanism in the organization of social categories, common sense knowledge, and social action.

Categories, Common Sense Knowledge, and Social Action
In his pioneering work on social categories, Sacks (1972a; 1995) showed the way in which membership categories can serve as repositories for, and organize, bodies of common sense cultural knowledge. This common sense knowledge, in Garfinkel's terms, consists of "socially sanctioned grounds of inference and action that people use in everyday life, and which they assume that other members of the group use in the same way" (Garfinkel 1956:185).
A consequence of this feature of membership categories is that a reference to a category can be treated by speakers and recipients as an account for (or explanation of) social action, as a result of the way it mobilizes common sense knowledge about that category. For example, Kitzinger (2005a; 2005b) has shown how callers to institutional lines (a suicide-prevention center and an after-hours doctor's office) used references to kinship categories in order to provide unremarkable, non-accountable reasons for their actions (cf. Sacks 1972b; 1995; Whitehead and Lerner 2009). Thus, callers could account for seeking help on behalf of another simply by displaying their relationship to that person through a reference to a kinship category (Kitzinger 2005a).
In many cases, such use of categories as accounts for social action is treated by participants and analysts alike as an utterly mundane and unproblematic feature of social organization, and even as a resource that enables people to talk about and make sense of their everyday activities in apparently smooth and seamless ways (Kitzinger 2005a; Sacks 1995). However, the use of some categories to tacitly account for social action is potentially treatable as an objectionable attribution of "stereotypical" characteristics to members of those categories (see, for example, Stokoe and Edwards 2007). This is especially so for categories such as race, gender, and sexuality, as a result of additional normative constraints regarding their use that have emerged from relatively recent political mobilizations.
            These constraints result in part from the operation of what I call, following Sacks, "categorizing the categorizer." Sacks (1995(Volume I):45) notes that it may be necessary to "categorize the categorizer" in order to understand the "perspective" they are taking in categorizing another, and thus what type of person the categorized one actually is.(i) Sacks' observations about this reflexivity involved in categorization were restricted to categories that could be ordered on a scale relative to one another, including "categories like age and social class, in contrast to those like race and sex" (Sacks 1995 (Volume I):45). I demonstrate in what follows, however, that a distinct but parallel form of reflexivity appears to operate with respect to mentions of racial categories (and hence possibly a range of other social categories). As a result, a speaker's mobilization of common sense racial knowledge through the use of a racial category to tacitly explain another's actions (as described above) can serve as an invitation for recipients to use the speaker's own racial membership category in explaining the speaker's actions – thus reflexively mobilizing common sense racial knowledge about the speaker (also cf. Stokoe and Edwards' discussion of the "speaker-indexical" nature of racial insults, and complaints about them – Stokoe and Edwards 2007:354), as well as moral judgments of the speaker's conduct (cf. Jayyusi's 1984 description of the links between categorization and the moral order). For example, a white speaker who simply identifies the race of a person of color in describing some action that person performed may be labeled "racist" specifically by virtue of being seen (and heard) as a white person attributing a person of color's actions to that person's racial identity (and thus as using common sense racial knowledge to explain those actions).
In light of the systematic potential for difficulties resulting from referring to race, one might ask whether such difficulties could be avoided by simply not mentioning race at all. Indeed, this "color-blind" approach has become widely popular in recent decades, with many seeing it as the best way of dealing with matters of race (see, for example, Brown 2003; Williams 1998 for discussions of "color-blindness"). However, a problem with adopting such an approach is that ceasing to attend to race, or to act on the basis of race, does not guarantee that others will do the same.
As a result, individuals who act without considering the possible racial consequences of their actions still face the possibility that others might interpret their actions racially, even if they did not have any racial intent when they acted. In short, sometimes race is relevant, and speakers who fail to appropriately recognize its relevance as they interact with others may face difficulties.
The foregoing discussion implicates dual and competing normative constraints on formulations of racial categories in talk: On the one hand, mentioning race may result in difficulties if speakers are heard as endorsing the racial common sense associated with the categories they are mentioning. On the other hand, failing to acknowledge the role of race when it is relevant to do so may also result in difficulties. One way in which the operation of these constraints on everyday actions can be observed is by examining practices speakers use for producing racial categorizations in their talk. How can speakers introduce race in such a way as to demonstrate that they do not endorse the common sense knowledge associated with it? How can they manage the implications of their own racial category for what they are doing when introducing race into their talk? How can they display, and allow for, recognition of the relevance of race through their talk while showing that it was others, rather than themselves, who chose to make it relevant by using it as a basis for action? The practices I describe below constitute ways in which speakers both orient to and manage this normative accountability associated with producing racial categorizations.
Data and Method
Since the normative constraints I have described above are particularly relevant in interactional contexts in which antiracist norms are prevalent, such contexts are obvious places to examine in order to investigate how speakers manage the resulting accountability of referring to race. A prime example of such a context is what can broadly be described as "diversity training," "racial sensitivity training" or simply "race training." Such training sessions are designed to engage individuals in discussions about the problems of race and racism, with the ultimate aim of bringing about individual, institutional, or societal change with respect to these problems. As such, they are contexts that are characterized by strong antiracist norms, with many such sessions actually being referred to by those who run them as "antiracist training."Although race could be heard as topically relevant for virtually everything that participants say in such settings, speakers are still accountable for how and when they introduce race into their talk. Participants must thus balance the task of speaking openly about their experiences and views with respect to race against the risk of being negatively evaluated or sanctioned should they be heard as inappropriately using race to account for action. The race-training context thus represents a natural "home" for practices speakers may use in managing their use of references to race in anti-racist environments.In light of this, I use video data of a race-training workshop in order to investigate these matters.
The data excerpts presented below were drawn from a full-day workshop videotaped in 2001 with the intention of producing a documentary film. However, the film has as yet not been produced, and the organizers of the workshop gave approval for the tapes to be used for research purposes. A total of three different cameras were used in order to record the two groups that convened in the morning, and the larger group that met in the afternoon. During the morning session, the participants, all of whom were invited to take part in the workshop and did so voluntarily, were divided into a "White Group" (consisting of two facilitators and six participants), and a "People of Color Group" (consisting of two facilitators and five participants), which engaged in separate two-hour sessions. In the afternoon session the participants and facilitators from both groups formed one larger group that engaged in a two-hour "group dialogue": participants from both of the previous sessions took part in an open-ended discussion of the issues that each group had dealt with separately in the earlier sessions. Throughout all the workshop sessions, the facilitators generally adhered to the aim, which they stated explicitly at the beginning of each session, of guiding a discussion on the issues of race and racism, rather than directing the participants towards prespecified outcomes.
Analysis was conducted using an ethnomethodological, conversation analytic approach. This approach is centrally concerned with explicating actors' practices (i.e., what they do and how they do things), rather than with their motivations, or with whether their actions are conscious, intentional or calculated (i.e., why they do things). Moreover, this approach focuses on analyzing utterances as public actions (rather than, for example, treating them as indicators of underlying psychological processes), and thus treats talk as a form of public social action, analyzing it primarily for its social and interactional import, rather than for what it reveals about any particular individual (Clayman and Gill 2004).       Each excerpt presented in the sections that follow[ii] is preceded by information regarding the session in which it took place (WG = "White Group;" "PCG = "People of Color Group;" GD = "Group Dialogue"), the camera the interaction was recorded on and transcribed from (A, B, or C) and the time segment from the camera's counter corresponding to the excerpt. The facilitators are identified with the letter "F" followed by a number (1-4), and the participants' names have been replaced by pseudonyms.
Generalizing Race
A first way in which speakers can manage the normative accountability resulting from "categorizing the categorizer" is by generalizing race, which involves formulating race in general so as to show that no specific racial category has special explanatory properties. One practice through which speakers can generalize race is list construction (Jefferson 1990). Formulations of this sort consist of a list of racial categories, followed by what Jefferson (1990) calls a "generalized list completer," a word or phrase that shows that there are other relevant list items that need not be explicitly specified. Hence, a list of racial categories followed by a generalized list completer is a method for referring to all possible racial categories. In this way, speakers can use race in general as an account for action while not singling out any particular racial category or categories for criticism.
This practice is exemplified by Excerpt 1, in which Sammy disagrees with the description of (white) "privilege" that Kim has produced in the previous stretch of talk. Just as Kim's experience as a white person constitutes the basis for her claims about privilege, Sammy's denial of white privilege implicates his experience as a white person. As a consequence, his disagreement is potentially vulnerable to being treated as self-serving. The design of Sammy's disagreeing turn addresses this potential problem through the use of a list of racial categories in constructing the counter-claim that racial privilege is situational and variable.
Responding to Kim's description of the ways in which she is able to use her (white) "privilege," and hence her sadness "at what others lose" (lines 1-2), Sammy (who has previously mentioned that he lives in San Diego) begins to propose that "where you live" (line 5) can "cause you to look at things quite a bit differently" (lines 7-8). As evidence for this claim, he contrasts the location in which the workshop is taking place (a small city), with a "much more urban larger city area" (line 9). In characterizing the types of neighborhoods that might be found in such an area, Sammy lists three racial categories ("Hispanic," "African American" and "Asian", lines 11-12), followed by a generalized list completer ("whatever the case may be," lines 12-13), thereby producing an inclusive list of all possible racial categories of which he is not a member.

Excerpt 1: [WG, A, 11:14:19–11:15:08]
1     KIM:  I don't feel guilty for being white but I feel very sad
2           (0.7) um (0.3) at what others lose. .hh
3           (3.5)
4     SAM:  (Well) (.) I think oftentimes people are depending upon
5           (0.5) y'know, where you live, (.) (I been) driving
6           around here for you know, .hh for the last day and a
7           half or so, .hh um: can (.) °eh eh eh° can cause you to
8           look at things quite a bit differently too. <I mean when
9           you live in a- in a- in a much more urban .hh larger
10          city area where you have .hh you know, pockets of:
11 ®       either you know Hispanic neighborhoods or: .hh
12 ®       predominantly African American or Asian or whatever the
13 ®       case may be, there are places whe:re (1.2) where I-
14          where I don't have the privilege(h)s .h[h (.) that I
15    (?):                                         [Mm hm.
16    SAM:  would traditionally.       

By using a generalized list in this way, Sammy claims that the racial organization of neighborhoods, regardless of the specific racial category involved, creates places in which racial "outgroups" are unwelcome, and therefore lack privileges. Sammy thus treats racial privilege as a situational characteristic of whichever "ingroup" constitutes the majority in any given neighborhood, rather than being tied to whiteness in the way that Kim had proposed. This logic serves as a basis for Sammy to complain about how he would likely be treated in certain neighborhoods, but by formulating his complaint in this way he conveys that his claims are based not on his status as a white person per se, but on his status as a racial "outgroup" member in such neighborhoods. Moreover, he shows that his complaint arises from the racial organization of neighborhoods in general, rather than from behavior associated with any specific racial category. In this way, he 1) acknowledges the relevance of race in general for the allocation of privilege, 2) denies that he, as a white person, is particularly privileged, and 3) avoids blaming members of any particular racial category for denying him privileges.
            In all cases of the list construction practice collected in the data, the speakers employing this practice in formulations of race did so in the course of making broad generalizations about the social organization of race. This demonstrates the way in which the generalizing list is best suited for use in making generalized claims about race as a form of social organization, rather than describing specific racial categories or incidents. In the following section, I describe a set of practices through which speakers can formulate specific racial categories, while limiting their use of those formulations to specific circumstances or incidents.
Localizing Race
In the following sections I examine two practices for producing "localized" racial formulations, namely qualifying racial references, and producing race as an "afterthought." By using these practices, speakers can display a racial formulation to be contingently included as a result of its relevance for the local interactional context, rather than being produced gratuitously. In this way, they provide ways of using race as an account for action, while limiting the account to the particular case or occasion being described.
                               Qualifying Race                               
The use of a qualifier prior to a reference to race is shown in Excerpt 2. In this case, Megan suspends her utterance to parenthetically insert a qualifier as a preface to a racial formulation (cf. Mazeland 2007). She does this in the course of a response to another participant's story about his father's experience of believing he had been discriminated against because he was white.(iii) In her response, Megan (who identifies as white) describes the effects that being told by her family about her father's negative experiences with people of color had on her own racial attitudes as she was growing up.

Excerpt 2: [WG, A, 11:21:18–11:24:00]
1     MEG:  I would say that I- I actually had a similar experience
2           in terms of growing up that my- my father had (0.6)
3 ®        a: a difficult situation or experience with (.) ­in this
4 ®        case particularly an African American person. .hh And I
5 ®        will say that one of the things that- that it impacted
6           me on is that I too I think was raised with sort of
7           this (.) in one breath, (0.3) everyone's treated equal,
8           .hh yet (.) at the s- by the same token, (0.2) the
9           examples that I was ever given about (0.2) when my white
10          father was impacted by someone it was by a person of
11          of color and I think that that very much played .hh for
12          me growing up about what my (.) ideas are a- ideas are
13          and have been about .hh u:m (0.4) people of color: (.)
14          sort of victimizing my family.

            By following her characterization of a particular "difficult situation or experience" (line 3), with the word "with," Megan initially projects that she is about to formulate the actor(s) involved in the situation. However, she then inserts the qualifier "in this case particularly" (lines 3-4), before going on to produce the initially projected formulation "an African American person" (line 4). In this way, the qualifier is arrived at as the result of an in-the-moment adaptation to the specifics of the circumstance she is reporting.
By prefacing the racial formulation in this way, Megan treats the specific racial category concerned as incidental to this particular case, and shows that she is not attempting to generalize beyond this case. Megan's orientation in this regard is further displayed as she speculates about the effect that her father's systematic mentions of the racial categories of people who mistreated him had on her views as she grew up: by suggesting that the mere disclosure of the racial categories of her father's "victimizers" may have influenced her views of people of those categories, she explicitly displays her understanding that a reference to a category may be hearable as proposing that category to be an account for action. Thus, by using a qualifier in this way before referring to a racial category, Megan avoids producing a racial formulation similar to those she is reporting her father (problematically) produced. In this way, she re-tells her father's story while resisting the same racialized inferences, including inferences about her own racial motivations as a white person referring to people of color, that were available to her as a recipient when her father told the story. This enables her to admit to her own racial prejudices, while simultaneously accounting (and thereby mitigating responsibility) for their development.

Racial "Afterthoughts"

Localized racial formulations can also be produced by redoing an already adequate reference to a person to include a more specific (racial) formulation, thus treating the inclusion of race as an "afterthought"(iv)to the formulation. This is illustrated by Excerpt 3, in which Darlene (who has self-identified as black, and has disclosed that she was born in Jamaica and lived there until her teens) describes an incident that occurred while she was working at a customer service desk in a department store.

Excerpt 3: [PCG, B, 11:26:12-11:26:51]
1     DAR:  And so (1.0) I remember one day I was there and dis
2 ®        woman came, ┬║this:┬║ African American woman an .hhh she
3           was upset about somethi:ng, something that went wrong at
4           the register and
5     (?):  (Mm.)
6     DAR:  you know they sent her to the service desk and she
7           ca:me, .hhh and so in my (0.4) nice little accent, which
8           I see you notice (>hasn't gone away as yet<) but back
9           then (.) I had an accent. .hhh And so: in my nice little
10          accent tryna explain to he:r, you know, the policies and
11          everything and she said (0.5) .hh ya know she said "You:
12          damn (w-) West Indians," you know "why don't you go back
13          to whe:re the hell you come from and learn to speak and
14          everything," .hhh an:d (.) of- at that point it hit me
15          it's like it h:it me all of a sudden that I was .hhh I
16          was not even in my comfort zone you know ((continues))

In referring to the customer in this incident, Darlene first refers to her as "this woman" (line 1), before repairing (see Schegloff, Jefferson, and Sacks 1977) to insert her race ("this: African American woman," line 2). By inserting a reference to race in this fashion, Darlene treats the racial formulation as a "second choice," thereby displaying that its inclusion was contingent on its particular relevance for her story. In other words, she shows that, rather than gratuitously mentioning the woman's race, she has done so in this case through an in-the-moment choice, or "afterthought," as a result of its particular relevance for her description of this incident. Darlene's production of race as an "afterthought" in this way enables her to display that the woman's racial (and citizenship) category was a relevant feature of this incident and its impact on her,(v) while resisting potential inferences that she routinely (and gratuitously) racially identifies people, and therefore routinely treats race as relevant for understanding behavior.
By using the practices I have described in this section, speakers can display that their use of a racial formulation was locally occasioned and arrived at recently. This serves to circumscribe the extent to which they could be heard (and responded to) as treating race as an account, by showing it to be either completely incidental, or limiting it to the particular case they are describing. Perhaps as a by-product, these practices suggest that speakers who use them do not routinely or gratuitously use race when it might be deemed inappropriate.
Having examined a range of practices through which speakers can refer explicitly to racial categories while managing the normative accountability of doing so, I now turn to an explication of the use of tacit racial formulations, in the form of allusions to race, in order to manage this accountability.
 Alluding to Race
Allusions to race provide a way by which speakers can imply or "plant" a racial meaning in their talk to convey it indirectly and inexplicitly, without saying it "in so many words" (cf. Schegloff 1996). Talking allusively, rather than saying something directly, requires recipients to track not just what has been said, but also what has been conveyed allusively, in order to make sense of an utterance. The resulting distribution of authorship (Goffman 1981) may provide a means for managing the formulations of others' racial categories. In the discussion that follows, I describe one practice that can be employed to allude to race, namely posing puzzles about actions that require explanations.
            By describing actions in such a way that a puzzle is posed regarding why an actor behaved in this manner, speakers can make relevant an explanation or account for the behavior. If these puzzles are posed after race has already been referred to, while discounting other possible solutions to the puzzle, race becomes available to recipients as the "obvious" solution for understanding the conduct in question. Thus, recipients can make sense of the actions, and solve the puzzle, only by using common sense racial knowledge to infer the racial category, and hence motivations, of the actor. This practice is exemplified by Excerpt 4, in which a facilitator tells a story about his experience of trying to open a bank account. Just prior to telling this story, this facilitator has reported that he is currently watching his daughter come to terms with the implications of being "a black woman," and has noted how difficult it is for him to watch, even though he himself grew up "a black man." By telling a story shortly after self-categorizing in this way, the facilitator invites his recipients to hear the story as exemplifying the implications of his status as "a black man," thereby treating his racial category as relevant in accounting for what happened in the story.

Excerpt 4: [PCG, B, 12:02:18-12:02:47]
1     F4:   U::m I- I used to tell a story abou:t ((sniffs)) goin
2           to the- (.) to the ba:nk, (0.4) u::h a::nd tryin tuh get
3           a:: corporate account for a ­company >for the dialogue
4           consultants< gettin a corporate account, .hh and havin
5 ®        problems with thee manager because he thought I was
6 ®        tryin tuh open a fraudulent account. .hhh U:m (0.8) hh
7           a:nd (0.8) thinkin >you know< I have no idea why this
8           guy thinks that. But no:w, even (.) with that story I
9           mean (you all will find out) the older ya get the more
10          gray ya get ya think ya know peop[le gon' lea::ve you
11    LAU:                                   [Mm hmm.
12    F4:   alone.

In his story, the facilitator recounts that the bank manager accused him of trying to open a fraudulent account (lines 5-6). Since such accusations are only called for when based on some kind of warrant or evidence of wrongdoing, this report establishes a puzzle about what cause the bank manager may have had for making the accusation. Moreover, the facilitator explicitly treats the bank manager's actions as puzzling (lines 7-8), and claims that he has no explanation for them. In addition, he has discounted any possible notion that the bank manager's accusation was in any way valid, by showing that he was at the bank for thoroughly legitimate reasons, to open an account for a company he works for (lines 2-4). Recipients can solve this puzzle, however, by applying common sense racial knowledge. Since the facilitator has already made his status as "a black man" relevant, recipients could make sense of the bank manager's actions by inferring that he was a member of a different racial category than the facilitator, and that the accusation was racially motivated.
By alluding to the relevance of race for understanding the bank manager's actions, rather than claiming it to be so more explicitly, the facilitator is able to make an understanding of racial discrimination available to his recipients without explicitly identifying the race of the bank manager. In this way, the facilitator can make inferences about the bank manager's racial motivations available, while attenuating potential inferences about his own racial motivations, which would become available to his recipients if he explicitly identified the bank manager's racial category or ascribed it as the cause of his actions. In addition, he displays that he did not jump to any conclusions about the motivations (including possible racial motivations) of the bank manager, but in fact was unable to find an account for his actions (see lines 7-8). In this way, he displays to his recipients that he is not paranoid or oversensitive when it comes to matters of race (cf. Jefferson 2004; Sacks 1995), but instead is cautious about making accusations of racism, allowing his recipients to reach that conclusion themselves rather than directly claiming it to be the case. Alluding to race in this way enables the facilitator to show that it was not him, but rather the bank manager, who was attending to race and acting on the basis of it in this interaction.
The posing of such puzzles provides a practice for co-implicating recipients who employ the common sense knowledge about racial categories necessary to make the inferences a racial hearing requires. By not explicitly mentioning the race of the actors in their utterances, while producing them in such a way that recipients will be left in no doubt about it, speakers can force their recipients to supply the common sense knowledge required to hear race as (obviously and apparently) relevant for understanding what happened, thus making them complicit in that racial common sense. Speakers can thus show that actors were racially motivated in their actions, while minimizing possible inferences that they (the speakers) themselves were racially motivated in describing them (cf. Stokoe and Edwards 2007).


The emergence (and, in many cases, familiarity) of the practices I have described, and their recurrent use in the production of racial references to others, suggests that common sense reasoning and knowledge about race is sustained (i.e., reproduced and managed) at least in part through a specific mechanism. That is, in managing the accountability associated with "categorizing the categorizer," speakers display an orientation to the relevance of their own racial identities, and the common sense knowledge associated with them, for how they report on and intervene in matters of race. That is, they orient to and manage the possibility that they may be held accountable as members of a particular racial category for avoiding inappropriate references to race, while at the same time recognizing itspotential relevance when it is appropriate to do so. In this way, even as they work to manage the accountability of their conduct in terms of racial categories and common sense knowledge, speakers treat those categories and common sense knowledge as consequential, and thereby reproduce them as such. This demonstrates one way in which "racial projects"(Omi and Winant 1994) carried out at the level of everyday interactions contribute to the reproduction of the category system underpinning the durable (racial) inequality Tilly (1998) describes.
            While there is no doubt that race can be reproduced as a result of strong commitments to maintaining its importance as a form of social organization (cf. Tilly 1998), the mechanism for the reproduction of race that I have demonstrated does not require such commitments in order to operate in the way that it does. That is, the operation of "categorizing the categorizer" (i.e., its status as a factor observably shaping people's conduct) may be the result of the systematic potential for being held accountable for failing to attend to and use race appropriately, independently of whether one personally believes in the importance of race. Thus, individuals may design their actions according to a racial interpretive framework solely as a consequence of the expectation that others may be using such a framework to interpret their actions, and that others may hold them accountable for those actions on the basis of that framework (cf. West and Fenstermaker 1995). In this way, the status of race as a social structure that shapes the experiences of individuals (cf. Omi and Winant 1994), and which individuals treat as a constraint on their actions, is reproduced independently of whether any particular individual is invested in or has a stake in it (cf. Wieder's classic study of the operation of "the convict code" in a "halfway house" for drug offenders - Wieder 1974).
It is further apparent, particularly from the examination of allusions to race, that the foregoing points concerning the reproduction of race apply not only to speakers (or those producing actions), but also to recipients (or those observing or hearing actions). Regardless of their personal beliefs about the importance of race, if recipients don't apply racial common sense in interpreting the racial motivations of actors made relevant by speakers' allusions to race, they will not be able to make adequate sense of the actions the speakers are describing. This provides a strong warrant for observers attempting to make sense of others' actions to take account of their possible racial implications, even if those observers have no stake or investment in the relevance of race (cf. Wieder 1974).
It is also important to emphasize that the actions for which speakers employ these practices result in further reproduction of race, by reflecting their orientation to their position within a racial social structure and the contingencies associated with that position. In my data, white speakers used the practices in doing such things as denying racial privilege (Excerpt 1), and admitting to and accounting for racial prejudice (Excerpt 2). In using these practices, speakers in these cases displayed their orientation to managing the implications of their whiteness for matters of privilege and prejudice. By contrast, people of color used the practices in the service of describing the impacts of being subjected to prejudice and discrimination (Excerpts 3 and 4), and displaying caution in making accusations of discrimination (Excerpt 4). The use of these practices by speakers of color in the data thus demonstrates their management of the experiences associated with being in a stigmatized or subordinate position in a racial system. This constitutes another way in which durable inequalities of the sort that Tilly (1998) describes become consequential, and are reproduced, through "racial projects" (Omi and Winant 1994) undertaken in everyday interactions.
            It should be emphasized that the generality of these findings may be limited by the particular interactional context in which they were generated, and that the range of actions for which these practices can be employed is most likely more extensive than those I have identified above. However, the mechanisms for the reproduction of race that they point to could be observed in any context in which practices such as the ones I have examined are employed. Further investigation in this regard could describe the use of these and other similar practices in other interactional environments, including those that are not as specialized or as normatively "antiracist" as the one I have examined.
Finally, it is important to recognize that the practices I have described, and the forms of social organization they rely on, are underwritten by basic and ubiquitous features of talk-in-interaction. Practices for list construction, repair, and storytelling in conversation, and actions such as disagreeing, agreeing, admitting, denying, accusing, and complaining are not particular to talk concerning race, but were mobilized by speakers in these interactions to perform particular projects in terms of race. This is to be expected since, regardless of the topical particulars of what speakers were doing, they were doing it through talk-in-interaction, and thus needed to draw on available practices for doing talking-in-interaction (cf. Raymond and Heritage 2006). This makes clear the importance of understanding interaction at the level of detail at which participants attend to it in order to understand what people are doing in interactions in which important issues, such as race, are being dealt with.

[i] For example, if a ten-year-old child categorizes someone as being "old," then the categorized person may be considerably younger than a person who is categorized as "old" by an adult (Sacks 1995(Volume I):45).
[ii] A list of the conversation analytic conventions used in producing transcripts of the data can be found at
[iii] See Whitehead and Lerner (2009) for an analysis of Megan's references to her father, particularly her reference in lines 9-10 to "my white father."
[iv] It is important to emphasize that the term "afterthought" is used here to convey the way in which a speakerdisplays an utterance to have been changed in mid-course, rather than as a claim about the speaker's psychological processes or motivations for changing the utterance in any given case.
[v] Prior to this excerpt, Darlene has discussed her own racial identity and affiliation with "black Americans" in recounting her experiences of working in the same department store. Her identification of the woman's race in this case may thus serve to convey the added impact that Darlene experienced as a result of being discriminated against by someone she would ordinarily have considered to share her racial identity.


Brown, Michael K. 2003. Whitewashing Race: The Myth of a Color-blind Society. Berkeley: University of California Press.
Clayman, Steven and Virginia Gill. 2004. "Conversation Analysis." Pp. 589-606 in Handbook of Data Analysis, edited by Alan Bryman and Melissa A. Hardy. Beverly Hills: Sage.
Condor, Susan, Lia Figgou, Jackie Abell, Stephen Gibson, and Clifford Stevenson. 2006. "'They're not racist.' Prejudice, Denial, Mitigation and Suppression in Dialogue." British Journal of Social Psychology 45:441-462.
Garfinkel, Harold. 1956. "Some Sociological Concepts and Methods for Psychiatrists." Psychiatric Research Reports6:181-195.
Goffman, Erving. 1981. Forms of Talk. Philadelphia, PA: University of Pennsylvania Press.
Jayyusi, Lena. 1984. Categorization and the Moral Order. Boston: Routledge & Kegan Paul.
Jefferson, Gail. 1990. "List Construction as a Task and Resource." Pp. 63-92 in Interaction Competence, edited by George Psathas. Lanham, MD: University Press of America.
—. 2004. "'At First I thought': A Normalizing Device for Extraordinary Events." Pp. 131-167 in Conversation Analysis: Studies from the First Generation, edited by Gene H. Lerner. Philadelphia: John Benjamins.
Kitzinger, Celia. 2005a. "'Speaking as a Heterosexual': (How) Does Sexuality Matter for Talk-in-Interaction?" Research on Language and Social Interaction 38:221-265.
—. 2005b. "Heteronormativity in Action: Reproducing the Heterosexual Nuclear Family in After Hours Medical Calls."Social Problems 52:477-498.
Mazeland, Harrie. 2007. "Parenthetical Sequences." Journal of Pragmatics 39:1816-1869.
Omi, Michael and Howard Winant. 1994. Racial Formation in the United States: From the 1960s to the 1990s. New York: Routledge.
Raymond, Geoffrey and John Heritage. 2006. "The Epistemics of Social Relations: Owning Grandchildren." Language in Society 35:677-705.
Sacks, Harvey. 1972a. "On the Analyzability of Stories by Children." Pp. 325-345 in Directions in Sociolinguistics: The Ethnography of Communication, edited by John J. Gumperz and Dell Hymes. New York: Holt, Rinehart and Winston, Inc.
—. 1972b. "An Initial Investigation of the Usability of Conversational Materials for Doing Sociology." Pp. 31-74 in Studies in Social Interaction, edited by David N. Sudnow. New York: Free Press.
—. 1995. Lectures on Conversation, vol. I & II, Edited by Gail Jefferson. Oxford: Blackwell.
Schegloff, Emanuel A. 1996. "Confirming Allusions: Toward an Empirical Account of Action." American Journal of Sociology 102:161-216.
Schegloff, Emanuel A., Gail Jefferson, and Harvey Sacks. 1977. "The Preference for Self-Correction in the Organization of Repair in Conversation." Language 53:361-382.
Stokoe, Elizabeth and Derek Edwards. 2007. "'Black this, Black that': Racial Insults and Reported Speech in Neighbour Complaints and Police Interrogations." Discourse & Society 18:355-390.
Tilly, Charles. 1998. Durable Inequality. Berkeley, CA: University of California Press.
van den Berg, Harry, Margaret Wetherell, and Hanneke Houtkoop-Steenstra. 2004. "Analyzing Race Talk: Multidisciplinary Perspectives on the Research Interview." Cambridge: Cambridge University Press.
West, Candace and Sarah Fenstermaker. 1995. "Doing Difference." Gender & Society 9:8-37.
Whitehead, Kevin A. and Gene H. Lerner. 2009. "When Are Persons 'White'? On Some Practical Asymmetries of Racial Reference in Talk-in-Interaction." Discourse and Society 20:613-641.
Wieder, D. Lawrence. 1974. "Telling the Code." Pp. 144-172 in Ethnomethodology: Selected Readings, edited by Roy Turner. Markham, Ontario: Penguin Books Canada, Ltd.
Williams, Patricia J. 1998. Seeing a Color-Blind Future: The Paradox of Race. New York: Noonday Press.