Gender Uptake: Theorizing the Semiotics of (Un)Doing Gender

Abstract

“Doing Gender” is often read as a theory of how people signify femininity/masculinity via expressive “performances.” This prevailing interpretation falls short of what the theory calls for—an ethnomethodologically informed analysis of gender’s emergent “naturalness.” To further this agenda, I theorize the audience’s “gender uptake” as a central component of doing gender by elaborating ethnomethodology’s attention to interpretive acts. Integrating Dorothy Smith’s intellectual legacy for feminist interpretive sociology, ethnomethodology’s neglected insight on categorization, and cultural sociology’s recent rediscovery of Peircean semiotics, I argue the facticity of gender’s “naturalness” remains underdetermined until the audience makes a series of ideological moves that cannot be predetermined by the performer. With case studies of how feminist parents account for their children’s gender-stereotypical interests, I illustrate how this audience-centered approach helps us unpack the interactional production/naturalization of categorical differences processually as open-ended negotiation of sign relations, where meaning emerges from selective attention.

One of the most cited articles in sociology, West and Zimmerman’s (1987:126) “Doing Gender” offers an “ethnomethodologically informed, therefore distinctively sociological” theory of gender as an “emergent feature of social situations.” Since its publication, scholars have largely used it as a theory of “gender performance” to explore how people enact their masculinity/femininity in different contexts. Partly due to such simplification, the notion of “doing gender” quickly became a basic, commonsensical resource for sociologists, even though ethnomethodology is increasingly marginalized in our discipline (Crawley 2022). Still, two central issues remain underexplored in theorizing gender as an “emergent feature of social situations”: How does gender appear natural and inevitable? When does this sense of gender as “natural” and “inevitable” emerge?

The political stakes of naturalization are paramount, as is the sociological importance of theorizing the naturalization process. Indeed, these questions concern not just gender scholars but also those studying ageism, racism, nationalism, heterosexism, and other forms of oppression (Atkinson 1980; Lie 2001; Loveman 2014; Schilt and Westbrook 2009). Categorical inequalities are often justified by an appeal to “natural difference” between people lumped into different categories, even though such difference is a social product. Recentering West and Zimmerman’s concern with gender’s emergent sense of “naturalness” is thus a productive initial step to examine how various socially produced categorical distinctions are experienced as an “inevitable” reality.

But although West and Zimmerman (1987:138) offer us a theory for the “labor that helps to constitute the essential nature of women” and men, they do not specify an analytic approach to unpack the naturalization process. Researchers thus often evade the situated and emergent “naturalness” of gender by assuming people simply get used to repeated performances. Such simplification further cuts the theory from its ethnomethodological roots, leading gender scholars to conflate the “accountability” of social action with the internalized “responsibility” to perform gender properly (Hollander 2013). Importantly, Garfinkel (1967:1) originally hyphenated the term as “account-able” to stress the “observable-and-reportable” quality of action—the capacity that any action can be described in an account. ¹ But as gender scholars became preoccupied with the dramaturgical performer, they lost sight of ethnomethodology’s keen attention to the work the audience does in accounting for a performance as “only natural” for members of the performer’s sex category. Consequently, researchers often take the interactional emergence of gender’s “naturalness” for granted once they identify a “gender performance.”

To address these problems, this article first restores West and Zimmerman’s (1987) ethnomethodological sensibility by bringing the audience back in; I then elaborate their theory with recent advancements in pragmatist semiotics. As my theoretical exegesis and integration suggest, taking up West and Zimmerman’s concern with gender’s emergent “naturalness” calls for an audience-centered approach to the study of doing gender that foregrounds people’s meaning-making practices. By rearticulating the process of doing gender from the audience’s vantage point, I show how it is the audience’s ideological work during/after a dramaturgical performance that retrospectively transforms it into what Goffman (1976) called a “natural indexical sign” of gender and thereby accomplishes gender’s emergent “naturalness.” The facticity of gender’s “naturalness,” I argue, remains underdetermined until an audience makes a “gender uptake.” Specifically, I theorize gender uptake as interpretive acts that take up a performer’s conduct and sex-categorical “essence” as a relevant sign-object pair on an indexical or iconic ground. Such gender uptake is both ideologized and ideologizing: ideologized because it is motivated by existing ideas about gender difference and ideologizing because it, in turn, produces empirical instantiations of such difference, which make gender stereotypes more powerful.

In what follows, I first discuss how gender scholars’ conflation of doing with performing compromised the theory’s analytic scope and ethnomethodological sensibility. By elaborating West’s empirical work with Smith’s (1990) interpretive sociology, I recover the original theory’s attention to the audience and explicate how audience accounts—formulated in situ or post hoc—weave local performances into extralocal discourses that travel beyond the here and now of face-to-face interaction. Then, I unpack the categorical basis of audience accounts by tracing West and Zimmerman’s (1987) idea of sex category to Sacks’s (1972) ethnomethodology of categorization. Doing so attunes us to how the audience’s selective use of social category delimits the context for interpreting an ongoing/past interaction. Finally, I dissect the audience’s meaning-making process in account formulation by turning to cultural sociology’s rediscovery of Peirce’s (1955) pragmatist theory of signs. This focus on meaning production via signs in action (i.e., semiosis) helps us identify four ideological moves the audience makes to facilitate the emergence of gender’s “naturalness”: typification, anchoring, downshifting, and erasure.

To illustrate how these ideological moves contribute to the semiotic process of doing gender, I present a pair of case studies drawn from a larger project on feminist parents, featuring two of these parents as the primary audience of their sons’ conduct. By dissecting the accounts these parents use to render their children’s gender-stereotypical interests account-able and intelligible, both to themselves and to me as an interlocutor, I demonstrate how gender’s sense of “naturalness” does not always emerge when a boy pushes a toy car. Rather, the wide range of discourses about gender and other differences in our society enable us to project different—that is, gendered or ungendered—meanings onto our interactant’s action. What this article offers, then, is not only a semiotic lens onto how the audience’s categorical accounts accomplish gender’s emergent sense of “naturalness” but also an analytic scaffolding to explore how the audience can render social actions account-able in creative ways that legitimate different ideologies. As such, this article invites scholars across fields to expand the analytic scope of social action and consider the cultural processes via which categorical distinction and inequality are produced and naturalized—not simply by habitual expressive acts but, rather, by methodic interpretive acts.

The Limits Of The Performance Paradigm

As Crawley (2022) observed, gender scholars often read “Doing Gender” as dependent on a simplified version of Goffman’s dramaturgy. Such a reading largely assumes the performer’s standpoint. Conceptualizing doing gender as “the way we [italics added] embody and enact masculinities and femininities” (Robinson 2020:9), scholars focus on how we signify our masculine/feminine selves by modifying our bodies and coordinating our comportments (Hoang 2015; Yang 2022), and how we internalize these performances as our habitus (Bridges 2009; Martin 1998). In this view, gender becomes “natural” and “inevitable” as we get used to routine gender performances, which can result from the developmental process of gender socialization and the disciplinary power of gender policing. Simply put, gender is supposedly “done” when a girl asks for a Barbie doll, no matter whether she identifies with Barbie’s femininity or her earlier preference for trucks was sanctioned by other children at school.

Despite its many critical insights, this “performance paradigm” (Silvio 2019) leaves us with a much narrower view of social action than what the doing gender framework originally affords: By equating “doing” with “performing,” it reduces social action to expressive acts—how people carry themselves and what they do with/to their bodies. Accordingly, gender scholars tend to focus on how people construct social selves by externalizing gendered qualities, casting aside how we construct social others by projecting such qualities onto “performers” during and/or after a given interaction (Silvio 2019). But from the ethnomethodological perspective West and Zimmerman (1987) intended to introduce, “doing” is also about the audience’s interpretive acts, such as their documentary method of interpretation (Garfinkel 1967), their selective application of social categories (Sacks 1972), and their methodic accounting of performances (Heritage 1984).

Albeit rarely taken up by gender scholars, this keen attention to interpretive acts is evident in West’s own empirical work on doing gender. In her study of cosmetic surgery (Dull and West 1991), West did not simply ask patients to recount how they decided to pursue such surgery but, rather, examined how their audience—in this case, their surgeons—thought of these decisions in terms of gender. As she showed, the post hoc accounts surgeons produced during research interviews legitimated the pursuit for cosmetic surgery as “‘normal’ and ‘natural’ for a woman” but not for a man, thereby accomplishing gender difference as “objective, factual, and transsituational” (Dull and West 1991:57, 64). As the word “transsituational” suggests, the audience’s accounting involves a “scaling project” (Carr and Lempert 2016). The gender difference was talked into being in the here and now of an interview with the researchers, and the surgeons discursively extended it to the there and then of the narrated doctor-patient interaction—it was scaled up as equally “real” in both situations. In this sense, the audience’s account, whether produced during or after the original interaction, is not just a “response” to doing gender but also what gender’s doing consists of.

If reconsidering West’s own empirical work helps rescue ethnomethodology’s attention to the audience’s role in the interactional production of gender’s “naturalness,” elaborating her ideas with Smith’s (1990) feminist interpretive sociology enables us to rearticulate the process of doing gender from the audience’s vantage point and defend the theory against its trivialization as a “micro” theory that stops short at the local level. As Crawley (2022) notes, Smith shares not only West and Zimmerman’s (1987) ethnomethodological sensibility but also their rejection of the micro/macro split. Similarly concerned with “facticity and the social construction of ostensibly objective knowledge” (DeVault 2023:318), Smith emphasizes how people’s activities of “reading texts” link local actualities with extra-local relations of ruling and thereby stabilize these material relations (Crawley, Whitlock, and Earles 2021:128). In this view, the audience’s accounts are the necessary medium for an ephemeral performance to travel beyond a specific moment of interaction—they solidify performances into concrete texts that can be cited from time to time as the context for making sense of people, their conduct, and interactional events.

As Smith (1990:176) argued, people’s dress, hairstyles, and all the activities Goffman (1959) analyzed as the presentation of self, “are to be ‘read,’ ‘interpreted,’ not merely seen.” To explicate this, Smith turned to Garfinkel’s (1967) work on “the documentary method of interpretation.” Gender, she argued, is not simply a “seen but unnoticed” background of everyday affairs. Rather, there is more to “seeing” in the accomplishment of gender. To see something as feminine is to read it through certain interpretive schemata, and to read something through these schemata is to activate some texts—in her case the texts of femininity—in the event of interaction. Once invoked, these texts “provide paradigmatic interpretive circles, supplying images, icons, or descriptions of behaviors” that enable us to construe our interactant’s performances as “documents” pointing to her underlying femininity (Smith 1990:177). The documents we cumulate in our reading over time then “accomplish the accountable presence of what is and has been happening, what is there to be spoken of” (Smith 1990:177). In this sense, a performer’s femininity does not exist independently from the audience’s uptakes or the discourse the audience invokes. Instead, it is the audience’s account that renders her presentation of self as account-ably feminine and “builds it into the organization of social reality” (Smith 1990:171).

To wit, in doing the reading and describing the performance in an account that renders the performance account-able, we find ourselves at the intersection of ideology and practice. One such ideology is what Goffman (1976:74) called “the doctrine of natural expression,” which leads people to misread each other’s gender display as “natural indexical signs” that express their interactant’s innate essence. Despite West and Zimmerman’s (1987) critique of Goffman’s notion of “gender display,” West (1996) later noted this badly neglected insight into the doctrine of natural expression is where Goffman gets closest to understanding the naturalizing power of audience accounts, which is central to their ethnomethodological intervention in theorizing “doing gender.” An ethnomethodologically informed analysis of doing gender thus must give considerable attention to the naturalizing power of interpretive acts. Practically, this demands us to go beyond asking people how/why they present their gendered selves in the ways they do and toward dissecting audience accounts collected directly from observers of those performances. ²

Hence, instead of simplifying “doing gender” as expressive performances, I argue we can approach it more effectively by foregrounding the interpretive process by which the audience activates the text of femininity/masculinity in their accounts to render such performance account-able as “only natural” for those in a certain sex category (West and Fenstermaker 2002:548). Doing so helps us better attend to West and Zimmerman’s (2009:116) emphasis on account-ability to sex category as “the key to understanding gender’s doing”: “Nothing is assured until the audience acts . . . [they] are the sole arbiters of performative success” (Taylor 2022:74). Although the performer’s gender display often precedes the audience’s uptake, its gendered-ness remains underdetermined until the audience uses sex category “to analyze settings for features [that sex category] made observable” (Smith 1987:161). Only after examining how the audience acted and what categories they used to analyze the performer’s conduct can we sensibly talk about the accomplishment of gender. To do so, we need an ethnomethodological perspective on categorization that elucidates the contextualizing effect of the audience’s selective category use.

The Pragmatics Of Sex Categorization

Since the 2010s, some sociologists have rightly started calling attention to the audience’s categorization practices (Lagos 2019; Ridgeway 2011; Westbrook and Schilt 2014). But this literature often treats “sex categorization” as synonymous to what psychologists Kessler and McKenna (1978) experimentally studied as “gender attribution”—the cognitive process whereby we decide if someone is female or male, usually by using visual cues as proxies for often-hidden genitals. Building on this, Westbrook and Schilt (2014:36) coined “determining gender” as an umbrella term for “the different subprocesses of attributing or . . . officially deciding another person’s gender.” Gender determination, they argue, is context-dependent: Genitals are critical in sexual interactions, whereas chromosomes can override visual cues in legal settings. We may thus consider people as having “social gender,” “sexual gender,” “sports gender,” and so on (Westbrook and Schilt 2014:50).

However, from West and Zimmerman’s (1987) ethnomethodological perspective, sex categorization is not equivalent to gender attribution. Informed by Sacks’s (1972) work on membership categorization analysis, their conceptualization of sex categorization is primarily concerned with how viewers and hearers lift out sex category as more relevant than other social categories, such as age or religion, for describing what is going on (West and Fenstermaker 2002). ³ As Sacks (1972:333) showed, every society has at least two categorization devices—sex and age—that can classify any member of that society, ⁴ but “a single category from any membership categorization device can be referentially adequate.” These conditions create a “problem of relevance” that the audience must resolve: Because people can be correctly described in infinite ways, the “fact” that someone is male, Jewish, or middle-aged cannot warrant them being referred to as such (Schegloff 1991; West and Fenstermaker 2002). This problem thus fissures one’s known identity and interactionally relevant identity: One may be “known to be a woman” but treated as relevantly a student and “not relevantly a woman” (Land and Kitzinger 2011:54).

Thus, to grasp the audience’s active role in doing gender, we must go beyond the social psychology of gender attribution to examine the sociolinguistic processes and interactional effects of their selective use of categorization device. Instead of asking what criteria the audience uses to determine if a person is a woman or man, we must dissect how they make sex category relevant to a situation when there are always alternative ways to categorize people and how orientation to this category defines the context of interaction. After all, even if a parent knows their child identifies as a boy, they may not always refer to him as “my boy” but may instead invoke their knowledge of their child’s age, skin tone, or zodiac sign to categorize them as “my 2-year-old,” “my privileged White child,” or “my little Scorpio” and render the child’s conduct account-able as something “only natural” (West and Fenstermaker 2002:548) for members of this category.

It is in this sense that West and Fenstermaker (1995:30) argue that “[w]hile sex category, race category and class category are potentially omnirelevant to social life . . . these may be stressed or muted [italics added].” Achieving a shared understanding of reality requires rendering interactional performances account-able; but because no account can exhaust the state of affair it describes, people must choose what to describe and how to describe what they observed (Heritage 1984). Audiences thus constantly stress and mute categories—both in the original interaction with the performer and in future interactions when the audience takes the footing of the “animator” (Goffman 1981) and recounts the original interaction to someone else (see Dull and West 1991; Smith 1978). Via such muting and stressing, the audience delimits the context of the original interaction as gendered, classed, racialized, and so on, which, in turn, regiments how the performer’s conduct should be read/heard.

In other words, categorization is not simply context-dependent but context-renewing (Goodwin and Heritage 1990) in the ethnomethodological perspective that underpins the theorization of doing gender. Consider Thorne’s (1993:69) school ethnography: On the playground, children refer to members of the opposite team not by name but by sex category—“Help, a girl’s chasin’ me!” ⁵ Although the child could have said a third-grader or a White person was chasing him, calling the chaser a girl invokes gender rivalry—rather than generational or racial conflict—as the relevant context for the chasing. Hence, despite gender’s minimal relevance in some other interactions Thorne observed, it became the basis of separate collectives here, as an “emergent feature” of the social situation on the playground (West and Zimmerman 1987:126).

Recovering the ethnomethodological sensibility of the doing gender framework thus requires us to foreground not just the audience’s interpretive acts but also the inevitable selectivity of their categorical accounts. More specifically, for the audience, to do gender is to hold a performer’s conduct account-able to a sex category, which requires lifting out some aspects of the performer’s identity into prominence while discarding others (Heritage 1984). It is via such selective categorization processes that the audience activates the text of femininity/masculinity as a relevant context for interpreting behaviors and events (Smith 1990). But to unpack how a sense of “naturalness” emerges from an audience’s interpretation, we need a social theory of meaning: “Every [expressive] act involves the use of signs. We must . . . turn next to the problem of the nature of signs” (Schutz 1967:118).

The Semiotics Of (Un)Doing Gender

From an interactionist perspective, gender embodiment and performances are “signifying practices” (Schrock and Schwalbe 2009). But how do certain embodiments and performances become meaningful signs that signify and naturalize social difference while others do not? Peirce’s (1955) pragmatist semiotics—as an elaborate theory of meaning—is well suited for tackling this issue. First, sharing ethnomethodology’s attention to concrete experience, creative problem-solving, and language in use, it echoes West and Zimmerman’s (1987) emphasis on situation over identity, process over variable, and selectivity over exhaustion (Emirbayer and Maynard 2011). Second, focusing on “how meaning emerges from a flow of signs in action,” it helps explicate “the evolving chains of meaning making that human interpreters follow in the journey from doubt to certainty” (Timmermans 2017:142, 171). I leverage these strengths of Peircean semiotics to identify four ideological moves the audience makes in formulating sex-categorical accounts: typification, anchoring, downshifting, and erasure. These four moves, as I will show, work together to accomplish the facticity of gender difference as “natural” and “inevitable.”

According to Peirce (1955), a sign is something which stands for something else in some aspect to someone. Hence, selective attention underpins every step of semiosis, that is, the interpretive process by which we use signs to make meaning out of the world: Anything can be a sign of any aspect of one’s identity, but nothing is a sign until it has some effect (an uptake or interpretant) on the audience, such as an understanding or conjecture (Gal and Irvine 2019). Only with such an uptake can a sign become meaningful as “a part of an act of signification” (Timmermans 2017:142). Interpreting interactional events thus requires the audience to assess what is a sign and what is not, what can be its object and what cannot. To a given audience, a made-in-U.S. pink shirt can be a sign of a person’s femininity or her patriotism: In one case, the audience fix their selective attention to the shirt’s color, in the other, the place of manufacture.

Existing ideologies about categorical distinctions shape our ability to identify signs (Gal and Irvine 2019). With the cultural knowledge that pink is a type of “feminine color,” when we see an American girl, Amy, wearing a pink shirt, what we take from the observation may not be “Amy likes pink” or “Americans like pink” but “girls like pink.” From this, we may then conclude, “girls and boys are so different.” In such an uptake, we selectively typify Amy as a representative member of the sex category “girl” (Timmermans and Tavory 2020) and anchor her clothing choice into a gendered “axis of differentiation” (Gal and Irvine 2019), where two ideologically contrasted images are linked to two juxtaposed person types: “girls love pink” versus “boys love blue.” Thus, through typification and anchoring, a local actuality (what a specific person wears on a specific day) gains meaning as a sign of a general pattern, which then reanimates the gender ideology that made it readable as such (Smith 1990).

Typification and anchoring are critical for the accomplishment of gender, but they are only possible with erasure. When taking up something as a sign, we also turn our selective attention away from something else such that some activities, qualities, and person types are erased from our chain of semiosis—they are rendered invisible or explained away, thus losing meaning and significance (Gal and Irvine 2019). For example, when we take up a specific piece of pink clothing Amy wears on a specific day as an instantiation of “what girls like to wear in general,” we not only withhold the demographic variation of girls (whom we have seen or believe to be wearing pink shirts) but also the qualitative variations across their shirts (e.g., shades, fabrics). Only then can we construe their clothing as identical embodiments of the same quality that signifies a shared substrate among girls. Conversely, if we typify Amy as an American and anchor her pink outfit to a geopolitical axis of differentiation, the same incident will not ideologize gender difference but something else—perhaps the flamboyancy of American capitalism (compared to the modesty of Maoist socialism, for example). Therefore, even if Amy anticipated others will see her pink shirt as a sign of her femininity, such an imagined interpretant may not actually materialize during an interaction. When the audience semiotically erase observations that could reify girls’ and boys’ “essential natures,” gender cannot be done.

Peircean semiotics thus sharpens ethnomethodology’s focus on the selectivity of categorization and account-ability by pushing us to examine which general type an audience made an empirical token to instantiate and what other types the audience erased. But it also provides us the analytic language to examine the ground on which the audience conjecture a sign to represent its object—a task ethnomethodology stops short at but is central to unpacking the naturalization process. Peirce (1955) distinguished three kinds of signs by their relations with their objects: “Icons” resemble objects by similarity, “indexes” point to objects by contiguity, and “symbols” stand for objects by convention. Among the three, symbols are the highest ranking and least natural signs because they are only related to their objects by cultural representations (Parmentier 1994). For gender scholars, the gendering of colors, toys, and clothes is all about creating representations that relate something to femininity/masculinity. Pink, for example, has no inherent relation to girlhood and was actually considered a powerful color in the early twentieth century (Paoletti 2012). In Peircean terminology, pink is a symbol of femininity.

But this is rarely how we experience pink things as signs. In today’s United States, we commonly experience pink and girlhood as a “natural” sign-object pair precisely because the convention that makes them recognizable to us as a pair is erased in our meaning-making. When seeing pink confetti at a “gender reveal,” most people will not think about the history of how pink went from a “boy color” to a “girl color” but simply conjecture the message as “They’re having a girl!” or just, “It’s a girl!” One may even start imagining a baby in pink shirts. In these uptakes, pink is “apperceived by their interpretants as being lower-ranking signs” in terms of complexity—a semiotic process Peirce called “downshifting” (Parmentier 1994:18). Pink is by definition an arbitrary symbol (highest ranking, most complex), but the audience’s uptake transforms it into a causal index or even a spontaneous icon (lowest ranking, simplest) of femininity. ⁶ It is in this sense that Goffman (1976:74) described gender displays as “natural indexical signs” that “sometimes hav[e] ‘iconic’ features”—indexical because the audience takes up the performer’s conduct as contiguous to her “nature”; iconic because the audience construes the softness of her feminine performance as identical to the softness of her innate, feminine substrate.

Semiotically, downshifting is what allows a sense of “naturalness” and “inevitability” to emerge from meaning-making. As our ideologized uptake erases its own action in relating a symbol to an object (e.g., a masculine/feminine substrate) and in turning our selective attention to the sign, it gives the sign-object relation a “separate existence” on its own, thus ideologizing the relation’s presumed independence from our uptakes—it feels as if “the mediation of a sign enables an object to have an effect” on our mind when “it is actually the conjecture itself that has allowed this, by linking the sign and object” (Gal and Irvine 2019:90). Downshifting thus effectively objectifies the audience’s projected object (e.g., a substrate) as an objective feature of the material world in relation to which the sign they observe is simply inevitable or spontaneous. As Timmermans and Tavory (2020:304) argue, when bundled with typification, such a downshift produces a “compound effect” that makes the sign-object relation “more abstract and lawlike”—the “naturalness” of the difference thus produced is done at a categorical level.

Thus, Peircean semiotics helps us examine naturalization as an emergent, open-ended process by centering the ideological work that underlies an audience’s category-based accounts. It offers us a more precise analytic language to unpack how certain embodiments and performances became meaningful signs in the first place and transformed into what Goffman (1976:74) specified as “natural indexical signs” that sometimes have “iconic features.” As this processual view of meaning-in-action reveals, the accomplishment of gender is neither simple nor inevitable. A performance cannot become a “natural indexical sign” of one’s female/male essence unless the audience (1) typifies the performer as a sex category member, (2) anchors the performance into a gendered axis of differentiation, (3) downshifts the conventional ground on which the performance signifies a substrate shared by the category’s members, and (4) erases other possible sign relations that could be taken to signify/naturalize other differences.

These interpretive acts are motivated by an essentialist gender ideology, but they are also what generate empirical instantiations for it (Smith 1990). Yet because our society harbors multiple ideologies about gender and other differences (Bem 1983; Risman 2018), there is no guarantee an audience will make all these ideological moves in all situations. Competing ideologies may direct one’s attention to different aspects of what sociologists may assume is a gendered interaction (Yang 2023). If we follow West and Zimmerman’s (2009) definition of “undoing gender” as the abandonment of sex category in accounting, these alternative uptakes suggest the possibility of momentarily undoing gender by naturalizing other stereotypes instead, although these uptakes, like all uptakes, are always subject to renegotiation.

Methodologically, this means analysts cannot simply assume people just get used to repeated performances or presume the relevance of some pregiven categories to a predefined context. With its focus on “gender uptake,” an audience-centered approach to the study of doing gender calls for three analytic tasks. First, to understand what a dramaturgical performance signifies, we must pay attention to not just how the performer staged it but also how their audience accounted for it. Second, to unpack the ideological moves the audience made during such accounting, we have to examine what kind of conduct or person type was taken up as a sign, assess the ground for its signification, and identify what conduct, identity, or category was erased in this process. Third, to grasp the relevant context for this event, we need to be cautious about imposing our own judgment as analysts and look for how the audience’s uptake contextualized the performance and the accounting itself.

The Case Of Feminist Parents

To illustrate how these three tasks enable a more rigorous analysis of (un)doing gender that attends to the naturalizing power of audience uptake, I present a pair of case studies of how two parents, Grace and Alison, accounted for their sons’ gender-stereotypical interests during my research interviews. ⁷ These two cases come from a larger study about how feminist parents in the United States navigate gender-related issues in childrearing at a time when ideas about gender and sexuality are shifting and developing rapidly. This larger study was based on semistructured interviews with 72 parents who self-identified as feminists and had at least one child under age 14. ⁸ Early in the data collection phase, I noticed that for many of these parents, their politics of childrearing often centered around combating the aggressive gendering of children’s toys, clothes, and activities. Yet many respondents were surprised to find that even when they offered their daughters a “gender buffet” (Averett 2016) of differently gendered options or gave their sons dolls instead of Nerf guns, their children still developed gender-stereotypical interests that breached these parents’ constructionist views of gender. For example, one respondent told me that despite her effort to buy things from both the “boys’ section” and the “girls’ section,” her son still developed “a fascination with trucks.” She recalled lamenting to her sister, “Oh My God! He loves trucks. . . . Ugh! I can’t believe trucks gonna be his thing!”

I soon realized this recurring theme offered a unique window into the semiotic process of doing gender. As both pragmatism and ethnomethodology have shown, meaning-making is most palpable when what one observes breaches what one believes (Emirbayer and Maynard 2011; Garfinkel 1967). To reconcile such contradiction, people need to repair their existing knowledge to accommodate their empirical observations or abductively espouse another ideology as more convincing than their former belief (Gal and Irvine 2019). Hence, feminist parents’ discussions of their children’s gendered interests can provide rich material for tracing “the evolving chains of meaning making” via which people reach a sense of certainty in their knowledge about gender difference as “natural” (Timmermans 2017:171).

To explore these moments of perplexity more systematically, I developed a new set of questions that invited my respondents to talk about their curation of their children’s toy and clothing options and reflect on their children’s interests and preferences. Amid the growing literature on parents’ accounts of their children’s gender variance (Meadow 2018; Rahilly 2020), these questions allowed me to take up queer theory’s invitation to invert our exoticizing gaze (Schilt and Westbrook 2009) and orient my respondents to their children’s gender conformity as a perplexing puzzle that needs to be accounted for. Moreover, because doing gender is about “cast[ing] particular pursuits as expressions of masculine and feminine ‘natures’” (West and Zimmerman 1987:126), these questions turned the interview itself into a potential occasion for doing gender. When a respondent invokes biological-determinist accounts to resolve or foreclose the perplexity of their children’s gender typing, they are effectively “creating differences between girls and boys” (West and Zimmerman 1987:137) during their conversation with me.

Following the “flexible coding” approach (Deterding and Waters 2021), I first indexed major themes in each transcript and then dived into respondents’ discussion of their children’s gender-stereotypical interests. As with the accounts Dull and West (1991) analyzed in their interview study of cosmetic surgery, the parental accounts I obtained were “given after the fact” and varied in their level of detail. Still, these accounts showcase how a performance could be understood by the performer’s immediate audience and made intelligible to a much wider audience who were not present in the original interaction, such as a research interviewer. Hence, my subcoding focused on identifying the type of accounts respondents used to render their children’s gender-stereotypical interests account-able during the interview. In doing so, I looked for accounts that illustrate the “doctrine of natural expression” (Goffman 1976) and accounts that contest or reconfigure the facticity of “natural” gender difference while being open to empirical “surprises.” This combination of deduction and induction attuned me to “the range of practical possibilities” (Connell 1992:738) of category-based meaning-making and information-rich cases that can yield “greater analytical clarity” for theory-driven research (Chen 1999:591).

Grace and Alison stood out as such cases. During subcoding, I noticed that their accounts were particularly rich with “magnified moments” (Hochschild 1994:4) that offered “a window into the social construction of reality” (Messner 2000:766). Although they used different accounts to render their sons’ gender-stereotypical interests account-able in gendered versus ungendered ways, both of their accounts exemplify the semiotic process of categorical account-making with “especial lucidity” (Chen 1999:591). Because qualitative analysis is about words rather than numbers (Lareau and Rao 2016), I focus on the depth of my data and refrain from the misfitted practice of reporting percentages in the style of survey research (see Small and Calarco 2022:64). The rest of this article thus presents thick descriptions of these two theoretically informative cases to illustrate how attending to the ideological work behind—and the contextualizing effect of—the audience’s uptake can enable a more fine-grained analysis of the interactional production and naturalization of gender difference.

As with any interview data, Grace’s and Alison’s words cannot be taken at face value—they are accounts given in an interaction about other interactions, which are not equivalent to the original parent-child interactions Grace and Alison referenced. But as West and Zimmerman (2009:116) remind us, “any method that captures members of society’s ‘descriptive accounting of states of affairs to one another’ (Heritage 1984:136–37) can be deployed for the study of doing gender.” To wit, interview data are suitable for the study of doing gender not because the analyst can “recover” the original sequence of action and response from interviewees’ accounts but because interviews provide concrete instances of accounting—by the interviewee, to the interviewer. Hence, an interactionist analysis of these accounts demands as much attention to what Grace and Alison said as to how they said it to me.

Informed by the ethnomethodological sensibility of West, Zimmerman, and Smith, my analyses of Grace’s and Alison’s accounts focus on “the putting together of social reality in situations” (Crawley et al. 2021:136). Practically, this means examining the interactional processes whereby interviewees were “inviting us to view the world we share with them” (Packer 2010:119) and contextualizing their children’s interests into an intersubjectively accomplished reality where these interests are account-ably “natural.” To enhance my analytic grip on these processes, I rewatched my video recording numerous times for each quote I used, at different speeds and with a level of intensity more typical for conversation analysis than for traditional interview studies. I paid close attention to subtle interactional cues, such as gestures, facial expressions, prosody, and personal deictics. I also followed the three analytic tasks outlined in the previous section, which required me to look for moments when Grace and Alison selectively invoked a membership categorization device (Sacks 1972) to anchor the account-ability of their children’s interests, and then use the Peircean terminology of typification, anchoring, downshifting, and erasure (Gal and Irvine 2019; Timmermans and Tavory 2020) to identify the ideological moves they made during accounting.

In what follows, I first dissect Grace’s account to show how typification to sex category, anchoring to a gendered axis of differentiation, downshifting from symbol, and erasure of other potential sign relations contribute to doing gender. Enhancing my argument from a different angle, I then use Alison’s account to illustrate how gender’s “naturalness” cannot emerge if the audience’s uptake diverges from the four ideological moves: Alison typified her son as a “kid” (rather than a “boy”) and anchored his interest to a generational (rather than gendered) axis of differentiation, thus undoing gender while “doing childhood.” This is not to say parents will never use other accounts to interpret children’s gendered interests. Rather, these two examples highlight important heterogeneity (Small and Calarco 2022) in how people’s gender performance can be held account-able. Such empirical variation in plausible accounts shows the indeterminacy of expressive acts’ gendered meaning and the limits of interviewing people about their own gender performances.

Doing Gender As Usual

Existing literature suggests parents will take up their children’s gender typing as a “natural indexical sign” of their innate biological feminine/masculine substrates (Goffman 1976; Kane 2012). In my interviews, some parents also rendered their children’s interests account-able with such a biological-determinist account. Consider Grace, a White, middle-class woman with an associate degree and a husband who is half Native American. According to Grace, her 7-year-old son Craig was interested in “anything big motor, anything he can knock over.” After listing some of the toys he likes, such as Nerf guns and army guys, Grace quickly commented:

Yesterday, we had a friend over who . . . has a 7-year-old daughter. . . . And she was like, “You have to come over and just see what it’s like having a girl, a 7-year-old girl. It’s very qui-et and it’s very art-sy and there’s gli-tter everywhere.” And I was like, “Not in my house. There’s Legos everywhere.”

Grace drawled the three qualities her friend associated with girls, stressed Craig’s interest in Legos, and endorsed her friend’s proposal for anchoring their children’s divergent interests into a gendered axis of differentiation. While Legos could be considered gender-neutral, this uptake typified Legos as a big-motor toy and linked it with boys, as opposed to fine-motor toys that are anchored to the contrasting sex category, “girl.” The children’s divergent interests thus gained gendered meaning in Grace’s reported speech, as local instantiations of a general gender difference.

Having invoked this gendered contrast, Grace continued to talk about her struggle with Nerf guns. After telling me her concern about safety, Grace frowned, shrugged, and said as she turned her palm upward, “That’s just how boys play.” In this short comment, Grace typified an empirical observation of Craig (a token) into a general regularity among boys (a type), making Craig’s interest in Nerf guns an index of his boyhood. She shrugged and added:

Like we know that from—Ugh, and if I could cite the study, I would, but there’s a study . . . about this remote Southern Amazon village or something where they had no . . . exposure to weapons at all, and boys were running around with sticks and just like sword-fighting . . . so I just think it’s just a completely natural thing, and I had to figure out how to embrace it because . . . I think if I would have said, “No, we’re not doing any sort of guns or shooting,” he was gonna do it anyways.

Previously, Grace controlled age to construct gender difference; here, she created a different-system-similar-outcome comparison and introduced such competing variables as hemisphere, residential area, and even civilization in her reflexive consideration of factors that might have shaped Craig’s interest. Anticipating the counterargument that the link between boys and guns is culturally specific (an arbitrary symbol), she first picked these contextual variables as possible axes of differentiation that set Craig apart from boys in a “remote Southern Amazon village.” One could differentiate Nerf guns and sticks by their material qualities and anchor them to a contrast in civilizations, but Grace erased this potentially noticeable sign-object pair from her semiotic chain. Instead, she typified Nerf-gun fights and sword fights as similar cases of “how boys play” across time/space. In so doing, she both posited that Nerf-gun fights, as tokens of playfights (a type), signify nothing but gender and downshifted the ground for this signification from symbol to index: For Grace, boys’ playfights exist independently from cultural conventions. The relevant context for interpreting Craig’s interest, Grace suggested, is not his upbringing in a North American middle-class family with a White mother and a half Native American father but that he is a boy in a world where “real” gender difference exists “objectively” between genders.

This downshift not only muted Craig’s account-ability to other membership categories, such as race and nationality, but also construed a spatial reference, where externally observable behaviors are contiguous to an internal force that is assumed to be the biological root of gender. Having anchored Craig’s Nerf-gun play to his sex category membership, Grace posited it as “a completely natural thing.” Her only option then was to embrace Craig’s interest by teaching him how to express it in responsible, consensual ways, telling him, “You can shoot knees or below . . . and if somebody says, ‘don’t shoot me,’ you don’t shoot them.” When asked to elaborate, Grace raised her hand and said:

We are not forcing it on him ((raising shoulders)). If he didn’t want to play with them, he wouldn’t play with them, just like . . . he’s never asked for a Barbie doll ((raising her palm)), like ((slightly shrugging)) he’s had chances to when we go to the toy store ((raising her palm again)). . . . So, yeah, I think ((shrugging left shoulder)) that’s . . . in his testosterone.

After Grace ended her sentence with another shrug, I repeated, “In his testosterone?” Shrugging once again, she said, “Yeah, yeah, I just think he is very much a boy. Like whatever’s made him that way, like he’s embraced it.” Here, Grace construed boys’ interest in weapons as indexing an innate “essence” they share, and she posited testosterone as the source of this “essence.” For Grace, insofar as boys have more testosterone than girls, they will be more interested in weapons.

From an interactionist perspective, Grace’s use of “testosterone” not only provided an account for Craig’s interest but also functioned as a “contextualization cue” (Gumperz 1992) for how I should hear this account in our interaction. Like her earlier citation of the “study,” her use of this biomedical term invited me to position her in a certain way: Although Grace is a White, middle-class woman with an associate degree, she was not speaking as a person with these attributes but as a parent who is familiar with “science” (West and Fenstermaker 2002). In other words, she was implicitly guiding me to hear her account as “scientific” rather than raced, classed, or gendered. Furthermore, while the denotational content and pragmatic wording of Grace’s account construed the narrated event of Nerf-gun fights as a natural result of Craig’s hormone, her nonverbal cues in the storytelling event of our interview also invited me to see Craig’s interest as inevitable (Wortham 2001). Grace’s repeated shrugs and palm-ups conveyed her powerlessness in front of nature and consolidated the ongoing construction of gender difference.

The “naturalness” of Craig’s interest thus emerged from Grace’s accounting practices. In describing Craig’s interest as “that’s how boys play,” she typified him as a sex category member. In co-selecting “boys” and “fighting” as subject-predicate and contrasting them with “girls” and “artsy,” she anchored his interest to a gendered axis of differentiation. In conjecturing playfights as an index contiguous from boys’ hormones rather than a symbol specific to any culture, she downshifted the ground of signification as a “natural” occurrence. In disattending the difference between guns and sticks, Craig’s family background, and other demographic attributes, she erased the many other categorical differences her observation could be made “a case of.”

Thus, Craig’s Nerf-gun fights entered Grace’s meaning-making of the narrated event and the joint attention of our storytelling event (the interview) as an indexical sign pointing to a quality underlying a gendered person type and was treated as evidence for the “facticity” of gender difference (Smith 1990). Whatever intention Craig had, his action was taken up as an unintentional expression of his innate male essence rather than a choreographed display that conformed to others’ expectations (Goffman 1976). Importantly, Craig’s interest became a meaningful sign that instantiated “natural” gender difference only because Grace’s uptake and erasure rendered it account-able to his sex category membership (not his citizenship) and on a ground independent from cultural conventions (not an arbitrary symbol). Gender, in effect, was done—not simply by Craig’s Nerf-gun fights but by Grace’s accounting of it.

Undoing Gender With Life-Stage Categorization

In Grace’s account, sex category serves as a key semiotic resource for holding Craig’s interests account-able as “how boys play.” However, sex categories are not always the anchor for account-ability. Given the selectivity of meaning-making (Gal and Irvine 2019) and accounts (Heritage 1984), membership in sex category may also be erased, or in West and Fenstermaker’s (1995:30) words, “muted.” Hence, insofar as doing gender requires the audience to “lift out” the performer’s sex category membership as the anchor of intelligibility of the performer’s conduct and formulate a “category account” (Raymond 2019), the possibility of undoing gender can be found in moments when the audience abandons sex category in their accounting as something irrelevant to the intelligibility of the performer’s conduct. As West and Zimmerman (2009:117) remind us, “‘Undoing’ implies abandonment—that sex category . . . is no longer something to which we are accountable.” Yet because a great deal of our knowledge about social order is stored in membership categories (Sacks 1972, 1995; Schegloff 2007; Smith 1990; Whitehead 2009), abandoning sex category often requires us to invoke a different membership categorization device to reestablish the intelligibility of our observations and render our interactant’s performance account-able to a different set of ideology and discourse. Under the cultural hegemony of developmental psychology, life-stage categories like “child” and “kid” offer a persuasive alternative to gender account-ability. Just as typification to sex category sets the stage for naturalizing differences between girls and boys, typification to life-stage category helps accomplish differences between children and adults. Alison, a straight, upper-class, White woman, provides a telling case of how “doing childhood” can be a means for undoing gender.

Despite Alison’s counterbalancing, her 11-year-old son, Sidney, latched onto stereotypically boyish things, such as dinosaurs, sea creatures, construction vehicles, and trains. When asked if she had figured out why he liked those things, Alison said she had “no idea” but then recast her lack of explanation as an intelligible response by typifying Sidney as a “kid”:

I mean . . . as a baby, he loved to carry around spatulas too and I don’t know why . . . he was carrying it. He always had like a spatula or like . . . you know, who knows? I don’t know. Kids do weird things. My nephew carried around a bag of snake, like rubber snakes for a couple of years. . . . I don’t know why he love[s] construction vehicles.

In typifying her son and nephew as “kids,” Alison erased the many spatiotemporal differences to create a qualitative similarity among Sidney’s early fascination with spatulas, his later interest in construction vehicles, and his cousin’s obsession with rubber snakes. For Alison, these are all “weird things” that one can hardly explain. But still, these weird interests are account-able insofar as the person showing these interests is seeable as a “kid.” By inviting me to hear these inexplicably weird interests as appropriate and “only natural” for someone at Sidney’s life stage, regardless of sex category membership (cf. West and Fenstermaker 2002:548), Alison’s comparison sidelined the gendered connotation of liking construction vehicles and generalized it instead as a case of how “kids do weird things.”

Once Alison typified Sidney as a member of the life-stage category “kids,” she started seeing a nongendered order in Sidney’s gender-stereotypical interests:

I think he liked the scariness of them, but in a controlled way. So, I see that with like construction vehicles and dinosaurs that they’re incredibly scary, right? Like, they could run you over; they’re loud; dinosaurs are scary. But like, working through that and learning about them with your parents is a very safe way to take this . . . scary thing and . . . work through it or understand it maybe a little better. So, I don’t know what that was like, if he was drawn to . . . the exciting, dangerous? ((slightly tilting her head)) Kind of like danger in a 2-year-old’s mind . . . and then want to discuss those with me?

Alison uttered “dangerous” with a raising tone and slightly tilted her head, suggesting to me that others might not consider “dangerous” the best description. But she preempted this potential disagreement by characterizing such danger as an age-specific perception rather than a universal agreement. In so doing, she invoked a generational axis of differentiation: Although adults like Alison and myself might not find construction vehicles and dinosaurs dangerous, we could still understand Sidney and his agemates’ reaction if we take the standpoint of 2-year-olds. Indeed, when Alison framed Sidney’s interest as “learning about them with your parents” instead of “his parents,” the gendered “he” became interchangeable with a child-ed but ungendered “you.” As this shift in personal deictics indicates, Alison treated Sidney’s sex category membership as irrelevant for the intelligibility and account-ability of his interest—instead, he was relevantly a “kid” (Land and Kitzinger 2011:54).

Albeit still speculative, Alison found this account applicable to Sidney’s other interests. Counting with her fingers, she continued, “Like he loves sharks, and he loves stonefish that are poisonous.” To wit, these items are similar not because they are all stereotypically boyish but because they trigger similar emotional responses in children. This similarity then serves as a context for seeing Sidney’s interests as developmental rather than gendered. By describing them as what he needed to work through with his parents as a 2-year-old instead of what he identified with as a boy, Alison left the cultural association between masculinity and danger aside.

This pattern further motivated Alison to conjecture an “undercurrent” beneath Sidney’s progression from one interest to another, where each new interest gave Sidney a little more adrenaline rush that he needed to navigate. After explaining adrenaline to me, Alison said:

Remember I said that he liked birds for a while? Like that was his first real obsession. And the reason why he liked birds was because we had a bird outside our windowsill, and I brought him over to look at it. And the bird was just like quietly sitting there, and we were talking about it for like a minute or two. And then the bird flew off and . . . startled my son, right? So . . . all of a sudden this . . . cute little thing . . . scared him, right? So, he got an adrenaline rush, and then for the next six weeks . . . we talked about nothing but birds, cuz I feel like he was kind of . . . working through that fear.

Alison’s description of birds as a “cute little thing” is in stark contrast with her earlier description of construction vehicles and dinosaurs as big, loud, and dangerous. This juxtaposition links the scale of development with the scale of scariness: As Sidney grew up, the things he became obsessed with and the things that could cause fear in him also got bigger, louder, and scarier. Unlike dinosaurs and sea creatures, birds are generally not considered masculine. By incorporating birds into her account, Alison disrupted the coherence of Sidney’s interests along the gender line while demonstrating their developmental coherence around each corresponding life stage.

After recounting Sidney’s obsession with birds, Alison continued, “Okay, so birds were the first one, and now . . . .” Alison drew some air ladders as she spoke, as if she was gesticulating the progression of Sidney’s interests. Then she said, “Yeah, now I’m thinking through it, I’m thinking like, ‘Maybe this is why!’” By framing each observation as elaborating an underlying pattern documented by previous observations, Alison’s “documentary method of interpretation” (Garfinkel 1967) enabled the “naturalness” and intelligibility of Sidney’s interest progression to emerge—and the gendered-ness of his interests to dissipate—from our talk.

As Alison’s words and gestures reveal, developmentalism gives a more encompassing order to Sidney’s variegated interests, which had hitherto been typified as “stereotypically boy.” When I commented that these interests seemed related to child development, Alison agreed excitingly. Counting with her fingers, she said, “I guess they’re stereotypical kid fascinations, right? Like, a lo-t of kids love construction vehicles. A lo-t of kids love sea creatures. A lo-t of kids love . . . dinosaurs, right?” Sidney’s interests in all these items were thus taken up as similar instantiations of a general regularity among members of the life-stage category “kid.” As such, Alison held Sidney’s interests account-able without reifying them as essential male “nature” but instead projected an underlying quality among children that differentiated them from adults by the “nature” of human development. As Alison oriented me to this “natural lifetime” (Atkinson 1980) as the relevant context for interpreting Sidney’s interests and disoriented me from his sex category membership, “doing childhood” opened a momentary possibility for undoing gender.

Competing Categorizations And The Creativity Of Account-Ability

“The actual events are not facts,” Smith (1978:35) reminds us, “[i]t is the use of proper procedure for categorizing events which transforms them into facts.” Whereas previous studies have demonstrated that individual transgression cannot undo gender if one’s audience still holds one account-able to sex category (Meadow 2018; Schilt 2010), Grace’s and Alison’s cases illustrate how following a gender script—whether intentionally or not—is inadequate for doing gender because one’s audience may hold one account-able to other social categories and naturalize other differences instead. Simply put, the facticity of gender’s “naturalness” is underdetermined when a boy picks up a Nerf gun. More specifically, the “naturalness” of gender cannot emerge from an interaction until the audience makes four ideological moves that cannot be predetermined by the actor’s impression management: typification to sex category, anchoring to a gendered axis of differentiation, downshifting from symbol, and erasure of other potential sign relations.

Hence, against the prevailing simplification of “doing gender” into a theory of “gender performance,” I argue the audience plays an indispensable role in the “doing” of gender. To retheorize the process of doing gender from the audience’s vantage point, this article integrates neglected ethnomethodological insights on categorization and recent advancements in pragmatist semiotics. In so doing, I propose an audience-centered approach to study gender’s doing that not only helps us identify how different ideological moves work together to accomplish the “naturalness” of gender difference but also reveals how these interdependent moves leave room for creativity (Emirbayer and Maynard 2011). In particular, whereas the commonplace conflation of account-ability with responsibility has led scholars to reduce it to a social control mechanism manifest in gender-policing, recovering account-ability’s ethnomethodological connotation as the “observable-and-reportable” potential of action (Garfinkel 1967:1) helps us step back and explore how the performer’s conduct can be described and made intelligible in gendered or ungendered ways.

When accounting for her son’s interests during our interview, Grace followed the prototypical semiotic process of doing gender. Although she did talk about racial socialization at other moments of our interview, here, she selectively put Craig into a sex category and muted his membership in race and other social categories. Such temporary erasure enabled her to construe Craig’s Nerf-gun fights as typical for boys across cultures and construct gender difference as a “natural” and “inevitable” result of hormonal difference. Yet as Alison’s case suggests, when the audience’s uptake diverges from these ideological moves, they may momentarily undo gender for the signification and naturalization of other differences, such as life stages. In other words, the selective nature of account-ability permits alternative ways of meaning-making and problem-solving that allow the audience to uphold, repair, or contest the “naturalness” of gender. The audience’s uptakes are thus creative in both senses of the term—they are productive of a semiotic reality that reinforces their motivating ideology but not rigidly confined to a uniform semiotic chain where each expressive act is fixated with a single interpretation.

Highlighting the audience’s active and creative role in doing gender, therefore, can help bring our attention from individuated performers back to the interactional space where the performer and audience do gender together. As Crawley (2022:377) argues, “Doing gender was never just about what one person performs, but rather about joint action.” Insofar as an interaction involves both expressive acts and interpretive acts, a maximalist approach to the study of doing gender ought to consider inputs by both the performer and the audience. Because the performer’s impression management may anticipate certain interpretations by their audience, their existing beliefs about the “generalized other” can direct them to choose certain expressive acts (see Hoang 2015; Yang 2022), which then become the “raw material” for the audience’s meaning-making. But because the audience does not have direct access to the performer’s psyche, there is no guarantee that such intentional performance “will be understood by observers in the way the performer intends” (Lagos 2019:804; Schutz 1967). Therefore, what the performer’s expressive acts actually signify and whether they successfully challenge/uphold existing gender ideologies depend on the audience’s interpretive acts (Taylor 2022).

By proposing an audience-centered approach to study gender’s doing, I do not mean we should simply replace individuated performers with individuated observers. Rather, amid the dominance of the “performance paradigm” in gender studies (Silvio 2019), I foreground ethnomethodology’s attention to the audience’s interpretive acts to help us revision laypeople and academics alike as situated and relational knowledge producers who methodically analyze the empirical world around them to produce, revise, or transform their knowledge about gender difference. ⁹ Notably, people may continue to engage in interpretive acts when the original “performer” is absent, such as when my respondents tell me or their friends about their children’s fascination with trucks. They may also project gendered qualities onto things that are not capable of “performing” gender, such as when biologists construct a romance about egg and sperm based on stereotypical female versus male roles (Martin 1991). My call to foreground the audience’s interpretive acts is thus not a turn away from interaction but an invitation to expand our understanding of interaction. With an explicit focus on gender uptake, we can more effectively examine how people produce and naturalize gender difference not only in interactions between a performer and an audience but also in interactions with nonhumans and in interactions about past interactions. But in any case, they are embedded in social relations—audience also have their own audience.

In other words, although the selectivity of categorical accounts leaves room for creativity, meaning-making is not a free-will terrain where anything goes. The range of uptakes an audience can intelligibly make is limited by the range of ideologies around them (Gal and Irvine 2019). Widely circulating discourses such as biological determinism and child development offered my respondents and me mutually intelligible frames for interpreting a child’s interest, but it is hard to imagine one can convince someone else with accounts like “he likes Nerf guns because he is White” or “she likes pink dresses because she is a Scorpio.” As West and Zimmerman (1987:136) point out, “descriptions [i.e., accounts of activities] are themselves accountable.” To achieve a shared social reality, one must convey their uptake intelligibly and convincingly to their interactants—whether that is the performer whom they produced an account of or a further audience (e.g., a friend, relative, stranger, interviewer) whom they are accounting to. The audience’s uptakes are thus also open to contestation, both in the subsequent turn by the performer and in a future interaction with someone else.

As such, although I used Alison’s case to illustrate the semiotic process of undoing gender, it should not be read as a pronouncement of gender being finally and truly eradicated, or “undone” in the past-perfect tense. In the past two decades, the scholarly discourse around “undoing gender” has mistakenly assumed a realist ontology to posit that gender can be willfully “undone” when volitional actors “move beyond traditional gendered scripts” and move into a “post-gender” world (Risman 2009:82). But this misses a key lesson from West and Zimmerman’s (1987) ethnomethodological intervention—reality is always under active production (Crawley 2022). Indeed, to theorize gender as “an emergent feature of social situations” (West and Zimmerman 1987:126) is to acknowledge it must come from somewhere—where it was muted. As Thorne (1993:84) elegantly put, “Gender boundaries are episodic and ambiguous,” like “many short fences that are quickly built and quickly dismantled.”

To wit, the doing and undoing of gender must be viewed as a never-ending process of negotiation. At best, we can only talk about gender being “undone” in a particular moment. But that moment is always open to reinterpretations and recontextualizations that can turn “undoing” into “doing” and vice versa. In an interaction, the performer may signal a self-categorization (e.g., a boy) that the audience does not take up. But when the audience makes an alternative categorization of the performer (e.g., a child) to account for the performer’s conduct, the audience may also be challenged by the performer in a subsequent turn. And even if the performer and audience agree on a shared definition of the situation, each party may recount the interaction to another person afterward using different sets of social categories available in their “cultural toolkit” (Swidler 1986) to redefine what “really happened” in a gendered or ungendered account.

In short, reality production is “future-facing” (Crawley 2022) because social interactions are open-ended: Any interaction can be cited and (re)interpreted in a future interaction, just as any interpretive act can itself be the object of another interpretive act. As Peirce (1955) reminds us, meaning-making happens in an infinite chain of sign activities—there always will be a next uptake (Gal and Irvine 2019). Because giving accounts “involves creating new ways of articulating recognizable accounts” (Crawley 2022:378), every time we cite and reinterpret an interactional event/performance in a different categorical account, we add another layer of meaning to the original performance, which enables us to create a new reality with our interlocutors.

In my empirical case studies, I mainly focused on how Grace and Alison created different versions of reality with me. Given the norms around in-depth interviewing, I was less inclined to directly challenge my respondents’ accounts when I personally disagreed with their explanations. But such contestation is likely to happen in everyday interactions. Although this is beyond the scope of this article, tracing the negotiation between performer and audience can help us better understand what ensues for both parties when an intersubjective agreement is not established—when an audience interprets a performer’s gender practice differently from what the performer intended (e.g., as not gendered). Alternatively, by examining the negotiation between the audience and the audience’s audience, we can gain deeper insight into how an intended gender performance may take on a life of its own beyond the here and now of the original face-to-face interaction, evolving into different signs that entail different realities as different audiences take it up for their own ideological projects. In either case, an explicit focus on audience uptake will allow us to closely examine how people selectively place one another into some social categories (as opposed to other equally correct alternatives) and hold each other’s conduct (and accounts) account-able to these categories to achieve such interactional outcomes as the creation, naturalization, reconfiguration, or erasure of gender difference.

Implications Beyond Sociology Of Gender

Besides clarifying a key theoretical issue in feminist sociology, my semiotic elaboration of West and Zimmerman’s (1987) ethnomethodological sensibility also contributes to several other fields. First, categorization and classification have garnered much general sociological interest not just in gender and sexuality studies but also in the sociology of race and ethnicity (Brubaker and Cooper 2000; Lie 2001; Loveman 2014; Monk 2022) and the sociology of knowledge (Eyal 2013; Fourcade 2016; Hacking 2006). But this burgeoning literature often focuses on the institutionalization of categories at the macro level. By drawing attention to the semiotic process whereby people invoke categories in interaction, my theorization of gender uptake both complicates and complements this literature. As my case studies demonstrate, the institutional availability of categories does not warrant their use in actual social situations. Both sex category and life-stage category are institutionalized, but they only become meaningful in interaction when invoked by interactants as anchors for the account-ability of interactional events. Once invoked, they enable accounts of how certain activity is “only natural” for people in a certain category, which then legitimate the categorization device as capturing “real” and “factual” differences between those whom it classifies.

Yet because our attention is inevitably partial, our use of social category is always selective (Gal and Irvine 2019). As the proliferation of institutionalized categories and discourses around them expands the pool of interpretive frames we can choose from, the meaning of expressive acts can become more underdetermined. Whereas Grace typified Craig into the sex category “boy” and conjectured his interests as indexing an innate male “essence,” Alison resisted naturalizing gender difference by reifying developmental difference instead. When Alison typified Sidney as a “kid” and conjectured that his interest indexed where he was developmentally, she erased potentially signifiable differences along gender lines. As the accomplishment of gender is forestalled in the naturalization of developmental difference, “doing childhood” becomes a means of undoing gender.

In this sense, centering audience uptake also enables us to start developing a more dynamic theory of childhood. In the past 30 years, the sociology of childhood has moved away from the future-oriented process of becoming adults to focus on the lived experiences of being children (Thorne 1993). But this paradigm shift may also inadvertently objectify and essentialize “children” and “childhood” as individual attributes based on age (Hammersley 2017). Amid such intellectual stasis, examining the accounting practices that produce and naturalize difference between life stages can help us transcend the being-versus-becoming binary to understand how childhood is an active process of doing, a discursive formation ideologized via account-in-action, just like gender: “‘[C]hildren’ are cultural events that members make happen in concrete situations via methodic categorization practices” (Atkinson 1980:37).

Ultimately, Grace’s and Alison’s cases reveal that different categorization devices can compete with each other to structure the intelligibility of the world we observe and legitimate certain ideologies while discrediting others. As such, these case studies call for more productive engagement across intersectional, pragmatist, and ethnomethodological frameworks. An intersectional lens has helped us see how doing gender can be intertwined with doing class, racial, and other differences (Moore 2001; West and Fenstermaker 1995); pragmatism and ethnomethodology further reveal how such interactional processes are also coupled with “doing erasure.” Indeed, for a performance to be meaningful and account-able in gendered ways, more differences must be erased than “done” in any incident of doing gender. This perspective offers a “context-driven” alternative to the “how-many-variables-do-I-need approach” to intersectionality (McKinzie and Richards 2019). Instead of treating contexts as top-down “buckets” in which social life happens, my analysis draws attention to how context(ualization) emerges from people’s uptakes and orientations that define the relevant frame for interaction and interpretation.

Hence, researchers may benefit from approaching intersectionality not just as a “sociologists’ resource” but as a “members’ resource” (Zimmerman and Pollner 1970). Having used intersectional frameworks to justify interaction terms in regression models, we must explore how people draw on intersectional theories in their pragmatic use of social categories in everyday or interview conversations. Indeed, the incorporation of intersectionality into daily vocabulary may motivate people to make certain uptakes that allow the interlocking relation between some identities (but not others) to emerge intersubjectively as an objective reality. The empirical question for future research thus becomes what other differences are erased when people orient to multiple categories and invoke axes of differentiation anchored in compound person types. By tackling this question with the analytic framework developed in this article, scholars will gain deeper insight into how multiple social distinctions can amplify each other in some situations but “contradict, mute, and twist” one another in other situations (Thorne 2004:404).

Conclusions

In this article, I argued that West and Zimmerman’s (1987) concern with gender’s emergent “naturalness” can be most effectively addressed by shifting our analytic focus from the impression-managing tactics of performers to the ideological work behind their audience’s uptake of such performance. This audience-centered approach leverages Dorothy Smith’s theoretical legacy to recover doing gender’s roots in ethnomethodology and extends this intellectual lineage with cultural sociology’s rediscovery of Peircean semiotics during its pragmatist revival (Gross, Reed, and Winship 2022). By expending social action from expressive acts to interpretive acts, my theorization of “gender uptake” rectifies commonplace conflation of doing with performing. By situating the presumed “omnirelevance of gender” into the empirical “problem of relevance,” it problematizes the hasty equation of sex categorization with gender attribution. By shifting units of analysis from biographic individuals to moments in meaning-making, I highlight how account-ability is a creative process of reality production, not just a social control mechanism that reinforces our responsibility to perform gender in normative ways.

“Without theory we are blind” (Burawoy 2009:13). The strength of my theory of “gender uptake” lies not in identifying causal propositions between variables but in guiding careful attention to the minute details of interaction (Hoang 2022). Specifically, retheorizing the process of doing gender from the audience’s vantage point calls attention to underexplored terrains of social life, such as how an audience’s ideological work shapes the meaningfulness of a performer’s expressive acts and how an audience’s pragmatic use of social category shapes the context for interpretation. This explicit focus on the audience may appear to be an overcorrection for the prevailing performer-centered underreading of “doing gender,” but my goal is to offer an alternative way of seeing that makes visible what has been obscured by existing paradigms and analytic approaches. Amid sociology’s relational-processual turn, this framework is especially useful for unpacking the gradual unfolding of social life—where categorical differences are produced, naturalized, contested, and negotiated via accounts-in-action; where our research subjects are not structural dopes but methodic meaning-makers, or rather, pattern-seeking knowledge producers, just like us.