Extending the scope for resistance to gender-based devaluation

ABSTRACT

This review focuses on resistance to experiences of group-based devaluation, specifically in the context of gender. This literature has seen considerable development in recent years – we outline this development and review the empirical evidence that supports it. The first section of the review discusses definitional issues surrounding resistance. The second section describes how the literature on resistance has developed, with particular attention to the “broader” perspective on resistance that has gained prominence over the last decade or so. This perspective includes subtle and even implicit forms of resistance alongside its “traditional” forms. The third and fourth sections review empirical evidence for this perspective, derived primarily from the gender literature. The final section describes outstanding questions in research on resistance to group-based devaluation. Taken together, the evidence we review illustrates that members of devalued groups can employ a broad repertoire of strategies to resist group-based devaluation.

KEYWORDS:

• Resistance
• group-based devaluation
• gender
• stereotyping
• implicit strategies

In 2016, a technology company launched a chat-bot that was trained on Twitter interactions with human users to create and send out tweets of its own. Soon, the chat-bot had to be withdrawn because it had begun sending racist, sexist, and anti-Semitic tweets (Hunt, 2016). This modern-day example demonstrates how prejudice and discrimination based on group memberships continue to pervade our daily lives, even in quite subtle ways – reflected in the fact that artificial intelligences quickly extract these underlying patterns from large volumes of text. Prejudice and discrimination manifest in many ways, and their effects accumulate in disadvantaged groups’ everyday experience. In this review, we examine resistance to these experiences. We argue that disadvantaged groups have developed a varied spectrum of strategies to contest experiences of group-based devaluation. In particular, in line with the subtle dynamics of prejudice in the chat bot example, one of the key messages of this review is that strategies for resistance too can be subtle.

“Group-based devaluation” describes the experience of those social groups that occupy the lower ranks of social hierarchies. As a result of their low social status, these groups often face considerable prejudice and discrimination, as well as higher risk of poverty, reduced access to education and poor health outcomes (Pickett & Wilkinson, 2010). A lot of the work in this area focuses on the negative (psychological) effects that devaluation can have on the individual, for instance lowering their self-esteem (for a review see Barreto & Ellemers, 2015), but there is also a great deal of work examining how members of devalued groups cope with these experiences. Here, we are particularly focused on resistance to group-based devaluation. Research studying resistance to group-based devaluation has seen considerable investment and development over the last decade. We aim to give an overview of this development and review the empirical evidence that supports it. Much of the evidence supporting this new perspective on resistance comes from the gender literature. In line with this, the work reviewed below focuses primarily on resistance to gender-based devaluation – particularly women’s resistance to the idea that they are devalued relative to men. Synthesis of this evidence offers a perspective that views resistance not as an exception, but as an integral part of the psychological lives of devalued groups (Lawrence & Dodds, 2018).

The review is divided into five sections. In the first section we discuss definitional issues concerning resistance, and highlight differences from related concepts such as resilience. In the second section we chart the development of research on resistance in the field of Social Psychology. The third and fourth sections review empirical evidence that support recent perspectives on resistance, focusing specifically on evidence arising from studies on gender-based devaluation. In the fifth section we elaborate on some outstanding questions and issues in the field, before concluding.

Defining resistance

Resistance has an intuitive non-academic meaning as opposition to aversive forces and obstacles, or a refusal to accept a harmful status quo, and is often applied to political opposition movements. The most notable example of this is the term “the resistance”, to describe political movements and guerrilla warfare during WWII (Finkel, 2015). In the academic domain, the concept of resistance has received a great deal of attention, not only in Psychology, but also in fields such as Sociology (Hollander & Einwohner, 2004; Riessman, 2000), Political Science (Scott, 1985), and Women’s Studies (Weitz, 2001). Although it is beyond the scope of this review to discuss these other literatures exhaustively, we will briefly reflect on the Sociological literature. Beyond that, key areas of similarity and difference between the concept of resistance in different literatures are outlined in Table 1. Hollander and Einwohner (2004) provide a review of how resistance is viewed in the Sociological literature. They outline that there are two widely recognised features of resistance. Generally, resistance involves some form of “action”, that is, resistance is an active rather than passive response.¹ Second, resistance involves opposition; it “occurs in opposition to someone or something else” (Hollander & Einwohner, 2004, p. 539). Other scholars have made this same point by saying that the concept of resistance is most easily defined when “paired with something that warrants resisting or rejecting” (Liechty, 2019, p. 204). That is, to understand resistance, we must first define what it is that is being resisted. Here, we focus on resistance to experiences of group-based devaluation, particularly among women.

Table 1. Overview of the resistance literature from different (sub-)fields. We focus on key areas of overlap and difference with the conceptualisation of resistance that we develop in this review.

(Sub-)field	Resistance …	Aim of resistance	Example strategies	Further reading/References
Political science	… to political oppression	Challenge oppressive political systems	Protest, sabotage, wilful ignorance	Scott (1985); Lawrence and Dodds (2018)
Sociology	… to inequality	Challenge oppressive social structures	Protest, verbal confrontation, lifestyle choices that contravene prescriptions	Hollander and Einwohner (2004); Riessman (2000)
Gender studies	… to patriarchy	Liberate self from gendered restrictions and expectations	Choice of hairstyle that contravenes established notions of femininity	Weitz (2001)
Social Psychology – Social Influence	… to social influence	Act out against attempts to influence to maintain independence	Derogate source of influence attempt; Non-compliance	Knowles and Linn (2004); Sagarin & Miller Henningsen (2017)
Social Psychology – Reactance theory	… to restricted freedoms	Liberate self from restrictions on personal freedom	Engage in prohibited behaviour	Brehm (1966)
Social Psychology – Social representations theory	… to normative representations	Negotiation to maintain own preferred representation in face of normative pressures	Verbal negotiation in interactions	Moscovici (1980); Duveen & Lloyd (1990)
Psychology – Clinical/Health	… to stigma	Reject stigma; Avoid self-stigma	Advocacy; seek empowering experiences	Ritsher & Phelan, 2004; O’Connor et al, 2018
Social Psychology – Social Identity theory	… to group-based devaluation	Contest group-based devaluation	Collective Action; Inter-group bias	Van Zomeren et al. (2008); de Lemus et al. (2016)

Group-based devaluation is the experience of being devalued based on the social groups to which one belongs. Social hierarchies rank different social groups on valued characteristics, with some groups occupying the higher ranks and others the lower ranks of the social hierarchy. Members of groups that occupy the lower ranks of the social hierarchy experience disadvantages relatively to members of higher-ranking groups. Across contexts, group-based devaluation represents a threat to those who are members of devalued groups. This threat can take a number of different forms, for instance when the group is denied worth or value (value threats) or because the group is denied agency or control (control threats) or when the group experiences practical or material devaluation (resource threats;). Such threats cannot easily be ignored, and resistance represents one way of addressing them (Branscombe, Ellemers, et al., 1999; Ellemers et al., 2002).

In this review, we focus specifically on women’s experiences of devaluation relative to men. Much of the work we discuss below operationalises the devaluation of women relative to men by exposing female participants to gender stereotypes. Gender stereotypes are assumptions made about the traits, interests, characteristics, and social roles of individuals based on their gender group membership (Eagly & Steffen, 1984). For example, gender stereotypes suggest that women are more emotional than men, and that men are more aggressive than women. Gender stereotypes represent a form of gender-based devaluation to women, because it assigns women those roles and traits that are considered to be low-status, and men are assigned the traits and roles that are associated with high status. Gender stereotypes reinforce existing gender hierarchies because they convey that status hierarchies are based on real differences between groups. In this way, gender stereotypes offer legitimacy to gender hierarchies in which women are devalued compared to men (Ridgeway, 1991; Tajfel & Turner, 1979). Crucially, research has shown that – in fact – the causality operates in the other direction: status hierarchies establish which groups are valued and which are not, and they are then ascribed traits that “match” their supposed status position (Cuddy et al., 2015). A second issue that contributes to the power of stereotypes in shaping social relations is the fact that they reinforce the binary gender system – women and men are represented as “complementary” – whereby the strengths of one group are the weaknesses of the other (Glick & Fiske, 2001). As such, gender stereotypes serve to “separate out” men and women into two contrasting groups (Bastian & Haslam, 2006). In this review, we are interested in how women might resist the devaluation of their gender in-group relative to men. We define resistance to group-based devaluation as counteracting, challenging, or rejecting group-based devaluation. That is, resistance addresses group-based devaluation by tackling it directly – challenging, counteracting, rejecting the idea that the in-group is devalued. A key aspect of the definition we use here is captured by Knowles and Linn (2004, p. 5) who outline that within Psychology “resistance has acquired a dual definition”, reflecting both a certain outcome or response, and a motivational state. Thus, in this review we consider resistance to group-based devaluation as both a motivation and a category of responses. We define the motivation to resist as the motivation to challenge, counteract or contradict the devaluation of one’s social group. By the same token, resistance responses are the responses that counteract, challenge or reject the devaluation of one’s social group (for conceptual elaboration, see de Lemus et al., 2016). The experience of group-based devaluation triggers the motivation to resist, which in turn triggers the resistance response.

Resistance, reactance, resilience

At this stage it is important to distinguish resistance from related concepts, like resilience, or reactance. A key feature of resistance as a way of coping with devaluation, is that it aims to challenge, reject, or counteract the source of devaluation (e.g., unequal social relationships), and it is this feature that differentiates resistance from related concepts. The main difference between resistance and retaliation is that the primary aim of retaliation is to provide a sense of justice following negative treatment (“an eye for an eye”; Lee et al., 2012), whereas the primary aim of resistance (as defined here) is to challenge the social hierarchies and structures that give rise to such treatment. In turn, resilience, as defined by Shih (2004), focuses on compensating for, or overcoming stressful experiences. Similarly, Masten (2014) describes resilience as a system’s capacity to successfully adapt to threats or disruptions. That is, resilience can be seen as a capacity or potential, rather than a concrete response. Broadly speaking, then, resilience reflects the capacity to respond constructively to a stressful experience, whereas resistance focuses on addressing the source of that experience through concrete responses and actions To illustrate this difference with an example, resilience would be to recover physically and psychologically from sexual harassment with only limited disruption to day-to-day functioning, whereas resistance would be to lodge a complaint or engage in collective action to protest against the social conditions that facilitate sexual harassment. Finally, the literature on “reactance” (Brehm, 1966) argues that restrictions to one’s personal freedom motivate people to engage in behaviours designed to reclaim or reinstate those freedoms. This concept is closely related to the concept of resistance we focus on here – primarily due to the fact that reactance seeks not just to cope with, but to remove restrictions. Resistance, then, can be seen as a “broader” version of reactance (see de Lemus & Stroebe, 2015), that is not necessarily focused on restrictions to personal freedom, but can also be applied to address experiences of group-based devaluation.

Individual and collective concerns

As part of our definition of resistance (and its contrast with resilience) we argue that resistance focuses on tackling devaluation directly. In the case of group-based devaluation – by definition – the source of the problem is the devaluation of the group, and to resist it, a person must engage with the group element of that experience. Still this perspective does allow for the fact that the reason why people are concerned about the devaluation of the group is because it has implications for them personally, as a member of the group. That is, although resistance must engage with the group element of devaluation, it can incorporate individual concerns as well. However, responses that are purely focused on individual concerns, such as showing that “group devaluation does not apply to me personally” would not be considered resistance. Such responses leave group devaluation intact, and instead focus on minimising its consequences for the individual. Of course such responses can still serve to cope with threats – as observed for instance when individuals attempt to leave the devalued group (“social mobility”; Jackson et al., 1996, see below for elaboration), or self-group distancing responses (van Veelen et al., 2020), but do not address the root of the problem, namely the presumed low-status of the ingroup. Therefore, such responses would not count as resistance under our definition. To summarise, we argue that resistance to group-based devaluation must engage with the group-related element of the devaluation experience, because resistance focuses on tackling devaluation directly.

Perspectives on resistance

In the Social Psychological literature, early research on resistance focused on resistance to persuasion – whereby participants resist certain information based (for example) on the social source of the message (Moscovici, 1980). Early work on resistance to group-based devaluation is embedded within the field of inter-group relations research, and specifically Social Identity Theory (Tajfel & Turner, 1979). Seminal work in the social identity literature has described three categories of responses that can be used to address the threat arising from experiences of group-based devaluation (Jackson et al., 1996). The first of these, social mobility (sometimes called “exit”), reflects individual strategies of leaving the devalued group, either literally or symbolically, and joining a high-status group. The second category, social creativity, reflects strategies by which people reframe or reinterpret the meaning of their identity to deflect attention away from negative aspects and instead focus attention on positive aspects (see Ellemers (1993) on “identity management strategies”). The third category is social competition, and it is this category of responses that has been central to research on resistance. Social competition strategies resist group devaluation specifically by demonstrating disapproval and fighting for social change. This can happen at the individual level – for instance by verbally confronting a perpetrator (Barreto & Ellemers, 2015), but also as a collective, for instance through collective action and protest (Drury & Reicher, 2000; Van Zomeren et al., 2008). These strategies that rely on competition with and confrontation of (members of) the outgroup, constitute the bulk of classic research on resistance to group-based devaluation.

More recently, scholars have begun to consider resistance more broadly, and have argued that resistance can be expressed in many different strategies (see e.g., Becker & Barreto, 2019; C. W. Leach & Livingstone, 2015; de Lemus & Stroebe, 2015). For example, it has been argued that social creativity should be considered a form of resistance (C. W. Leach & Livingstone, 2015). As described above, social creativity refers to a group of identity management strategies by which members of devalued groups creatively re-define identity content to avoid negative implications (Tajfel & Turner, 1979). This can be viewed as resistance, in the sense that the group resist the pressure to define their identities in line with dominant (negative) representations, and instead creatively re-define identity content on their own terms. An example can be seen in the “black is beautiful” theme among Black Americans in the civil rights movement (Harlow & Benbrook, 2019; Tajfel & Turner, 1979). Recent research increasingly considers that such responses are also part of the “repertoire” of resistance strategies. After all, in the case of social creativity, the new construction of identity challenges, rejects and counteracts the normative one and reframes it in a more positive way.

One of the main reasons why researchers began to consider resistance more broadly than it is thought of in traditional approaches, is that resistance in the form of protest or confrontation is not always possible, practical or desirable, for a variety of reasons. First, experiences of group-based devaluation are continuously reinforced through everyday interactions, especially in cases where there is a great deal of contact with the out-group – such as in the case of gender (Hebl & King, 2013; Sue et al., 2007; Swim et al., 2001). As such, the need for resistance is ubiquitous, but it seems impractical for each of these experiences to trigger a very committed and effortful response from the individual or group who are targeted (such as protest). Instead, subtler low-key strategies might be required to maintain resistance to gender-based devaluation. Another complicating factor when it comes to opportunities for resistance more generally, are processes specifically designed to prevent responses that challenge devaluation – here we enter into the domain of hierarchy maintenance (e.g., Haley & Sidanius, 2005; Phelan & Rudman, 2010). Where resistance aims to challenge and counteract (devaluation arising from) social hierarchies, hierarchy maintenance processes are those that uphold and reinforce those hierarchies. For example, women who resist gender-based devaluation face backlash, in the form of negative judgements and behaviour towards women who are perceived as upsetting gender hierarchies (Rudman & Glick, 2001). That is, resistance to social hierarchies can be risky. Moreover, hierarchy maintenance processes can combine contradictory elements that serve to elicit role-congruent behaviour from members of the target group, even those who do not personally endorse those role divisions. For example, a recent study from our own lab focused on women’s willingness to express anger about gender-based devaluation. Anger is typically prescribed for women – that is, expressions of anger are incompatible with the social roles assigned to women. This notion is reflected in gender stereotypes, in two contradictory ways. First, anger disconfirms elements of gender stereotypes that present women as kind and caring. However, at the same time, expressions of anger can also be presented as confirming gender stereotypes of women as overly emotional. When expressions of anger are framed in this way – as confirming gender stereotypes – this creates a reluctance to express anger, particularly among women who identify as feminists. That is, the idea that anger might confirm a stereotype of women, leads feminists to avoid it. Ultimately, though, avoidance of anger expressions is in line with social roles at a different level. The finding that concern about confirming stereotypes can contribute to producing the stereotype-consistent behaviour is reminiscent of the stereotype threat literature, which also highlights how exposure to stereotypes can generate a “self-fulfilling prophecy” (Schmader et al., 2008). In sum, then, the hierarchy maintenance processes elicit role-congruent behaviour from the target group, even amongst those who do not endorse such role divisions to begin with (Sindic et al., 2018; van Breen & Barreto, 2022). This finding is relevant here because it leads us to a key observation – namely that hierarchy maintenance processes are flexible and routinely prevent social hierarchies from being easily challenged. The implications of this for resistance is that, to be effective, resistance too must be flexible and adapt to contextual constraints.

Opportunities for resistance can also be complicated by features of the devaluation experience itself, specifically the ambiguity that often characterises experiences of group-based devaluation. Again, we see this especially in the gender context, for instance in the literature on Benevolent Sexism, where the positive phrasing of a certain statement obscures the implied devaluation of women as a group (Glick & Fiske, 1996). When gender-based devaluation is ambiguous, women are less likely to confront the perpetrator directly (Barreto & Ellemers, 2005; Crocker & Major, 1989). This again highlights why a diverse “repertoire” of resistance strategies is beneficial – when one is unavailable others can be accessed.

Further, there is evidence that experiences of group-based devaluation are becoming more subtle under the influence of socio-cultural developments. In many Western societies, group-based devaluation is expressed in increasingly subtle ways (Pearson et al., 2009; Swim et al., 1995). The adoption of egalitarian norms means that prejudice, discrimination and other expressions of social devaluation are increasingly considered unacceptable, and sanctioned when expressed explicitly. Under the influence of egalitarian social norms, then, the expression of prejudicial attitudes is “driven underground”, and instead expressed more subtly. Indeed, research has demonstrated that processes that contribute to social devaluation can occur implicitly, outside of conscious awareness (Blair, 2002; Devine, 1989). It seems unlikely that such subtle experiences would trigger protest or confrontation. This begs the question of how members of target groups can cope with these subtle and ambiguous experiences. Does the more subtle form of devaluation experiences mean that resistance is not possible in these cases? Or perhaps there are other, more subtle response strategies that can be applied?

In sum, there are several reasons why a diverse repertoire of resistance strategies, including some more subtle strategies would be beneficial in resisting group-based devaluation. Note, further, that several of these considerations come across most clearly in the gender domain, which is the primary reason why research on this topic is often conducted in the context of gender. The gender domain provides a relevant test case for a “broadened” view on resistance.

The idea that resistance to gender-based devaluation can occur through diverse and subtle strategies raises the question of what these strategies might look like. Conceptually speaking, subtle resistance strategies are responses that arise from the motivation to resist, that is, the desire to counteract or challenge the devaluation of the group. However, when it comes to the strategies used to fulfil this motivation, the subtler resistance strategies include responses and behaviours that do not necessarily reflect resistance, but that can be used as such when the situation demands. In other words, these responses are contingent on context to determine whether a certain response is being employed to satisfy the motivation to resist. The inclusion of subtle resistance strategies under resistance is facilitated by the differentiation of the motivation to resist and the resistance response in the definition outlined above. The motivation to resist being de-coupled from the precise nature of the response allows the motivation to resist to be expressed in many different strategies. This also allows us to avoid categorising a response as “resistance” versus “not resistance” – a response can sometimes serve as resistance, and sometimes serve another function. In this way, we aim to avoid a-priori restrictions on what “counts” as a resistance response and what does not. Thus, we propose that subtle resistance includes responses and behaviours that do not necessarily reflect resistance, but that can be used as such when the situation demands. For a group that is stereotyped as lazy, resistance might take the form of task persistence. For a group that is stereotyped as aggressive, resistance might take the form of cooperative behaviour. In both cases, the behaviour counteracts and challenges the specific stereotype devaluing the group. In this view, then, the resistance response can vary, but these different responses are united by their underlying aims and motivations – to counteract or challenge group devaluation.

Recent perspectives: Empirical support

Recent perspectives on resistance emphasise the need for diverse, flexible and subtle strategies to successfully counteract group-based devaluation. In what follows, we turn to empirical evidence from the gender context demonstrating that – indeed – resistance to gender-based can occur through diverse strategies. Before we begin, it is worth noting here that not all the authors who have worked on this issue label the responses they observe as resistance. It is rather that, under our definition these responses would be considered resistance because they counteract group-based devaluation.

When the devaluation context relies on stereotypes, members might show counter-stereotypical behaviour as a way of disproving those stereotypes (e.g., Cihangir et al., 2013). In the context of gender-based devaluation, there is evidence that exposure to gender stereotypes can motivate women to perform better, for instance in the domain of negotiation (Kray et al., 2001) or engineering (Crisp et al., 2009). Crisp et al. (2009) examined how women’s maths performance is affected by gender stereotypes – specifically the idea that women are less skilled at maths than men. They compared women who studied Psychology, to women who studied Engineering – two fields that differed in the focus on maths, and in the representation of women in the field. Across two studies, they showed that women engineering students performed better after exposure to the stereotype than before. In other words, female Engineering majors showed a “stereotype boost effect” after stereotype exposure. This effect was not present among female Psychology students (in either study). This response from participants serves to counteract devaluation, through behaviour that disproves the stereotype, and it therefore falls under the definition of resistance we outline above.

In other cases, the positivity or value of the group might be threatened. Such devaluation can be counteracted by asserting the group’s value, through evaluative forms of in-group bias. In-group bias has been studied as a form of social competition (Tajfel & Turner, 1979), whereby biased allocation of resources serves to compete directly with the out-group. However, evaluative (rather than resource-based) forms of in-group bias can also be used to re-establish positive in-group identity, by reaffirming the value of the group (Oakes et al., 1994; Voci, 2006). Another strategy that comes to mind here is outlined in the rejection-identification model (Branscombe, Schmitt, et al., 1999; Ramos et al., 2012). Research in this area has demonstrated that experiences of inter-group rejection and devaluation can lead members of devalued groups to reaffirm their commitment to the group by reporting increased in-group identification. The central postulate of this model is that the underlying motivation for the rejection-identification response is a form of compensation. After exclusion by the dominant group, the individual must seek acceptance elsewhere, and the in-group provides a means of doing so. We suggest that this response could also arise from the motivation to resist. By claiming increased commitment to the low-status group, the individual signals (to the out-group and the in-group alike) that they reject this devaluation, thus challenging the narrative of devaluation and low group status (Barreto & Ellemers, 2003; C. Leach et al., 2010).

Interestingly, examples of resistance through diverse strategies can also be found outside the field of Social Psychology. In the field of Gender Studies, feminist theorists have argued that by adopting non-gender conforming hair or clothing styles women can signal rejection of dominant beauty standards (Ortner, 1995; Weitz, 2001). If we cast these findings in terms of gender-based devaluation, we might argue that such responses reflect a rejection of the traditional notions of femininity that arise from gendered hierarchies. As a side note – in the literature outside of Social Psychology, resistance is not normally directly operationalised, but rather used as an interpretative label. Nevertheless, for the purposes of the current review it is relevant to observe that the idea that gender-based devaluation can be resisted through a variety of different strategies, does not only come from the Social Psychological literature, but is supported by convergent findings from related literatures.

Implicit resistance strategies

When we consider resistance responses such as in-group bias and task persistence and performance, it is worth noting that these processes can also be captured on implicit measures, such as Implicit Associations tasks, evaluative decision tasks, lexical decision tasks, or Go-No Go associations tasks. These tasks typically rely on response time measures. Participants are presented with certain target words (e.g., positive and negative words) paired with certain primes (e.g., man/woman). Congruent pairings, where the participant perceives the prime and target as “matching”, typically generate faster response times, because they require less extensive processing than incongruent pairs. In this way, response times can inform us on participants’ implicit evaluation of the target stimulus. These responses are even more subtle than the subtle strategies outlined above. This suggests the possibility of resistance to gender-based devaluation on implicit measures. We define implicit responses as those responses (on e.g., attitudinal measures) that occur without the person being consciously aware of having made the response. Implicit resistance, then, is evident from implicit responses that signal rejection of group-based devaluation. In what follows, we highlight research that has used such reaction time measures to generate evidence for resistance against group devaluation, in implicit responses.

In a study using 85 female participants, we exposed half of our sample to gender stereotypes, through picture stimuli that showed women in a kitchen, and men in an office (de Lemus et al., 2013). The other half of the sample saw illustrations of counter-stereotypes, with women shown in the office and men in the kitchen. After exposure to 160 trials, participants then completed a lexical decision task. This task required the female participants to categorise positive and negative warmth and agency words (presented for 100 ms) which were preceded by gendered primes consisting of men’s and women’s faces, presented for 28 ms. Warmth/Communion traits are stereotypically associated with women whereas competence traits are stereotypically associated with men (Glick & Fiske, 1999). Accordingly, in baseline conditions when no stereotypes are salient, female primes facilitate the categorisation of warmth traits whereas male primes facilitate the categorisation of competence traits (de Lemus et al., 2008), thus we might expect that emphasising stereotypes during the exposure phase would strengthen the woman-communion association. In fact, the opposite effect was observed – women showed evidence for resistance: Previous exposure to gender stereotypes led them to activate counter-stereotypical associations. Specifically, for participants who had been exposed to the stereotype condition, positive words relating to competence were recognised faster when they were preceded by a female prime (RT = 473 ms) then a male prime (RT = 509 ms). Similarly, positive words relating to warmth were recognised faster when they were preceded by a male prime (RT = 466 ms) compared to a female prime (RT = 498 ms), though this effect was less pronounced. Moreover, these results were predicted by participants attitudes towards affirmative actions (Tougas et al., 1995). That is, the more participants held positive attitudes towards affirmative action policies to change women’s roles, the more they reversed the activation of gender stereotypes in the stereotype condition (β =.36, t(82) = 2.20, p = .03).

In later research, we examined how such exposure to stereotypes would impact on evaluative attitudes towards gendered social roles (de Lemus et al., 2018). Across two studies, we again exposed female participants (N Study 1 = 53; N Study 2 = 134) to women and men presented either stereotypically or counter-stereotypically, by superimposing female and male actors on backgrounds representing an office or a kitchen. Results showed – first – that women who were exposed to the stereotype condition experienced more negative emotion (M = 2.38; SD = 1.31) than women who were exposed to counter-stereotypes (M = 1.58; SD = .78). As a measure of implicit resistance, we asked participants to complete an evaluative decision task, in which they were asked to recognise positive or negative words (e.g., happiness, pain). As in the manipulation phase, these words were preceded by gender role primes (men and women in the office or kitchen) but – unlike the manipulation phase – these primes were now presented very briefly (28 ms),² to assess implicit responses. The structure of the task is presented graphically in Figure 1. The outcome of interest is the impact of the role prime on the ease with which the participants were able to recognise the positive and negative target words. Results of Study 1 showed that female participants who had been exposed to stereotypes were faster to recognise positive words when they were preceded by a picture of a woman (M = 430 ms; SD = 44 ms) than when they were preceded by a man (M = 441 ms; SD = 38 ms) in the kitchen context, but not in the office context. In sum, there was evidence for in-group bias on an implicit, associative measure: those women who had been exposed to stereotypes in the manipulation phase, were faster to recognise positive words when they were paired with in-group (“women”) than the outgroup. The kitchen context directly activates female-stereotypical roles, which are most threatening for women’s identity, this may explain why the resistance effect was evident in the kitchen context, more than the office context.

Figure 1. From de Lemus et al. (2018) – structure of the evaluative decision task.A figure showing the different screens participants saw during the evaluative decision task – starting with a fixation point, then the prime, then the target word, and finally the screen requesting the participant’s response.

Study 2 included some new features, relative to Study 1, that were designed to provide further insight into how implicit resistance comes about. First, we added an additional manipulated factor to the evaluative decision task: The primes in the evaluative decision task could be either face-primes (as in de Lemus et al., 2013) or “role primes” (as in Study 1). This offers insight into whether in-group bias arises in response to men and women per se or whether it arises in response to the social roles in which they are presented. Further, the target words were not just positive and negative words, we now added stereotype content (competence and warmth). That is, participants had to classify warmth and competence related words as positive or negative (as in de Lemus et al., 2013). This allowed us to examine whether the implicit in-group bias is a generalised evaluative response (reaffirming a positive view of women) or a more targeted representation of women as positive in some specific domain (stereotypical warmth or counter-stereotypical competence).

Results of Study 2 showed – first – that the results of Study 1 were replicated. As before, exposure to stereotypes again produced more negative emotion (M = 2.11; SD = 1.21) than counter-stereotype exposure (M = 1.58; SD = .68). Further, there was evidence for implicit in-group bias after exposure to stereotypes when participants were primed with roles (not just with male and female faces), indicating that resistance is triggered by the social roles in which women (and men) are presented, rather than men and women per se. As in the previous study, positive words were recognised more quickly when the prime showed a woman (M = 452 ms; SD = 51 ms) than when then prime showed a man (M = 459 ms; SD = 49 ms). Note that in this design the effect appears regardless of the context -kitchen or office- where the primes appeared in (this factor was nested in the design, as context was only present in half of the groups). Additionally, aside from the facilitation of positive targets by female primes, in this study we also observed that the recognition of negative targets was facilitated by primes showing men (M = 472 ms; SD = 46 ms) compared to primes involving women (M = 478 ms; SD = 47 ms). As such, the in-group bias effect seems to be more generalised in this study, compared to Study 1. Figure 2 summarises these findings. Finally, the positive words category was composed of positive words relating to competence and warmth. This dimension did not affect responses – as such these findings suggest that the implicit in-group bias response represents a general re-affirmation of the positivity of group membership.

Figure 2. From de Lemus et al. (2018) Study 2 - Reaction times (RTs) per exposure condition showing how gendered role primes impact reaction times to positive and negative target words, after stereotypic and counter-stereotypic training. A bar chart with reaction times (RTs) on the Y-axis and the different trial types on the X-axis. The bars show that – after stereotypic training – positive words were recognised faster when primed by a picture of a woman. Negative targets were recognised faster when primed by a picture of a man.

Finally, Study 2 included an additional measure of resistance – a measure of task persistence in a counter-stereotypical domain (spatial reasoning). As we have seen above, when group-based devaluation is communicated through stereotypes, counter-stereotypical behaviour can be used to resist that representation. In this case, resistance to stereotypes would be evident from the desire to demonstrate competence at spatial reasoning which is typically associated with men more than women (Hyde, 2014; Sanchis-Segura et al., 2018). Results showed exposure to stereotypes triggered persistence on the counter-stereotypical task. More specifically, female participants persisted more on the counter-stereotypical task when they had been exposed to stereotypic gender associations (M = 241 s) than when exposed to counter-stereotypical associations (M = 181 s).

Summarising this work, these studies showed evidence for resistance to gender-based devaluation in the form of implicit in-group bias, as well as task persistence in a counter-stereotypical domain. One interpretation of the implicit in-group bias effect would be to argue that the female participants approved of the stereotypical representation of women, which made them more attuned to positivity. However, a measure of self-reported emotion confirmed that the stereotype exposure (relative to the counter-stereotype condition) triggered negative emotions in both studies, suggesting that the automatic in-group bias here is related to negative affect (rather than positive). This is consistent with the idea that these female participants displayed a resistance effect – after exposure to a condition that triggers negative emotion, participants make more positive associations with their in-group. The in-group bias strategy is a group-focused strategy whereby women as a group are associated with positive targets. On the other hand, the measure of task persistence might be viewed as a relatively “individualistic” response, demonstrating that the stereotypical implications do not apply to the participant personally. However, in the study reviewed above, the in-group bias effect was a moderator of the persistence effect – the effect of stereotype versus counter-stereotype exposure on persistence was significant only for participants who also showed stronger implicit in-group bias, but not for participants low in implicit in-group bias. Thus, we believe the findings of this work support the group-oriented nature of the response.

In sum, the work reviewed in this section shows that resistance to gender-based devaluation can be found in a wide variety of strategies in addition to the classic forms of resistance. We have seen evidence for resistance to gender-based devaluation through subtle resistance strategies, such as task persistence, but also through implicit responses, such as evaluative associations. Such implicit forms of resistance might be conceived of as functioning like resistance in the physical immune system (vanDellen et al., 2011), fighting disease without the individual’s awareness or conscious intervention.

Resistance to subtle devaluation cues

The recognition of subtle resistance strategies, outlined above, ties in with another key issue in the literature on group-based devaluation. As noted above, group-based devaluation can be communicated very subtly, and sometimes even subliminally, that is, outside of conscious awareness (Blair, 2002; Devine, 1989). In other words, there is an increasing recognition that experiences of devaluation form a continuum – from experiences that are blatant, to experiences that are more subtle, and even occur outside of conscious awareness. This begs the question of how members of target groups can cope with these experiences. Here, the subtle and implicit resistance strategies considered above may provide a solution. Having established that resistance can occur through a variety of subtle and implicit strategies, we now turn to cases where these subtle resistance strategies are used to resist subtle forms of devaluation. In other words, we now examine whether the subtle and implicit resistance strategies identified in the research reviewed above, can help women to resist when they are confronted with subtle experiences of gender-based devaluation. Previous research suggests that subtle forms of gender-based devaluation undermine options for resistance (Barreto & Ellemers, 2005; Kray et al., 2001; Major et al., 2003). However, below we consider the possibility that very subtle cues of gender-based devaluation can nevertheless trigger resistance, particularly in the form of subtle and implicit resistance responses. We first discuss the conceptual aspects of this idea, before considering empirical evidence for such effects.

When experiences of group-based devaluation are subtle, or even occur outside of conscious awareness, the first question is whether this can still trigger the motivation to resist. Indeed, research confirms that motivational processes can be triggered without the individual being consciously aware of it (Bongers et al., 2009; Veltkamp et al., 2009). For example, work by Moskowitz et al. (2000) shows that the motivation to control stereotypes (amongst members of advantaged groups) can be triggered without the need for conscious awareness. This raises the possibility that the motivation to resist stereotypes may operate similarly amongst members of devalued groups.

If the motivation to resist is triggered outside of conscious awareness, then how might that translate to resistance responses? When an experience of group-based devaluation is subtle, and the motivation to resist it is triggered implicitly, it seems unlikely that this would translate to resistance responses such as collective action or interpersonal confrontation. Put simply – it seems unlike that a victim would directly confront a perpetrator if the victim is not aware of having been targeted. When we considered subtle and implicit resistance strategies above, we noted that these can occur alongside the more explicit resistance strategies. Here, however, there is something that makes subtle and implicit responses particularly effective – namely the fact that there is a degree of “structural fit” between the subtlety of the devaluation experience and the subtlety of the response. In terms of the how resistance responses may arise from an implicit motivation to resist, there is evidence that implicit motivations can impact attitudinal measures, such as stereotype activation (Moskowitz & Ignarri, 2009). Implicit processes can also affect behaviour more generally, for instance when more conscious opportunities to control the behaviour are absent (Friese et al., 2008). In our case, the opportunity to consciously control the influence of devaluation cues on behaviour is complicated by the fact that they occur outside conscious awareness. In such a case, then, we might expect that any subsequent behaviour is shaped by implicit processes (Friese et al., 2008).

In what follows, we consider empirical evidence for the idea that very subtle experiences of group-based devaluation can trigger resistance, primarily through subtle and implicit strategies. As before, many of the studies below operationalise gender-based devaluation through exposure to gender stereotypes specifically, but now these manipulations contain a subliminal element – to capture processes that occur outside of conscious awareness. Stereotypes are particularly suitable as the basis for subtle and subliminal manipulations, for several reasons. First, practically speaking, stereotypes can be conveyed in text, but also in images, and because they are simple in structure (e.g., Woman -> Emotional) can easily be used in reaction time measures such as Implicit Association Tasks, Go-No Go association tasks or priming procedures. Further, there is an inherent ambiguity in stereotypes that ensures that they are quite subtle in conveying devaluation. Specifically in the case of gender stereotypes, the devaluation implied by stereotypes can be subtle because gender stereotypes are often phrased positively (Jackman, 1994). This feature is beneficial to research on subtle processes, and can go some way to explaining why many of the studies reviewed below use manipulations based on gender stereotypes.

In 2016, Ramos et al. (2016) showed that women who were exposed to subtle reminders of sexism subsequently showed less gender stereotypical associations on an Implicit Association Task. The manipulation exposed female and male participants to either benevolent sexism, hostile sexism, or no sexism by having participants unscramble sentences and complete word stems that suggested sexist completions. This subtle manipulation of the two sexism conditions consisted of pictures and phrases (taken from the Ambivalent sexism inventory) representing sexist relations between men and women, whereas the non-sexism condition used neutral pictures and phrases. Participants then completed an Implicit Association Task (IAT; in Experiment 1) or Go/No Go Association Task (GNAT; in Experiment 2), both of which measured associations between men and women, and competence and warmth (controlling for the valence of the words). In Experiment 1, women who had been exposed to either of the sexism conditions made fewer gender stereotypic associations compared to women who were not exposed to sexism. Importantly, men’s stereotypical associations did not differ across conditions. These findings were replicated in Experiment 2, women who were not exposed to sexist beliefs made stereotypic responses – they responded more quickly when competence words shared a response key with male names than with female names. Women who were exposed to sexism made less stereotypical responses: they responded equally quickly when competence words shared a response key with male and female names. The alternative simple comparison also reached significance: when female names and competence words shared a response key, responses were faster amongst women who had been exposed to sexism than amongst women who were not exposed to sexism. Men were unaffected by exposure to sexism. Moreover, by applying the Quad model to the analysis of the GNAT responses (Conrey et al., 2005), the authors were able to establish that female participants’ accuracy-oriented detection (D) increased following exposure to sexism relative to the no sexism condition. In other words, exposure to sexism encourages women to place more emphasis on accuracy in their responses, which led to reductions in stereotypic responding.

A study from our own lab (van Breen et al., 2018) exposed women to subliminal cues of group devaluation (gender stereotypes). In the exposure phase, women were exposed to picture stimuli (e.g., an iron) paired with subliminal gender primes (presented for 42 ms) to create subliminal associations that were either stereotypical (woman – iron; woman – dishes) or counter-stereotypical (man – iron; man-dishes). Figure 3 illustrates this manipulation. After exposure to 120 trials, we examined the impact of this subliminal exposure on resistance tendencies. Specifically, building on the work discussed above, we examined resistance in the form of task persistence and in-group bias (de Lemus et al., 2018), as well as the tendency to associate women with counter-stereotypical attributes (de Lemus et al., 2013; Ramos et al., 2016). We further added a Moral Choice Dilemma task as an additional measure of in-group bias, this task presents participants with eight scenarios in which either an in-group member (woman 4×) or an out-group member (man 4×) must be sacrificed to save the lives of several others whose group membership was unspecified. In this measure, in-group bias is evident from a preference for sacrificing out-group members over in-group members. Across four studies (N1 = 121; N2 = 252; N3 = 103; N4 = 44) we observed evidence for resistance on three out of these four measures of resistance. Subliminal exposure to stereotypes (vs. counter-stereotypes) led to resistance in the form of implicit in-group bias, out-group derogation on the Moral Choice Dilemma task, and persistence in a counter-stereotypical performance domain (de Lemus et al., 2013, 2018). However, there was no evidence in this study that stereotype exposure led participants to associate women with counter-stereotypical attributes (de Lemus et al., 2013; Ramos et al., 2016). These resistance effects appeared specifically amongst those women with more politicised identities, that is, women who were strongly identified with feminists, but less identified with the broader group of women. This group of women (whom we have labelled “distinctive feminists”), has been shown by previous research to be particularly critical of gender stereotypes – as evidenced by self-reported disapproval and dislike of gender stereotypes (van Breen et al., 2017). We suggest that women who hold strong previous objections to stereotypes are those who are most likely to show resistance when exposure to stereotypes is subliminal (Moskowitz et al., 1999).

Figure 3. The manipulation used in van Breen et al. (2018), with examples of picture stimuli. The figure outlines how the stereotypical and counter-stereotypical exposure conditions were structured – the subliminal primes “woman” or “man” are paired with pictures of gendered activities to suggest either a stereotypical or counter-stereotypical association.

To elaborate briefly on the details of the results – after exposure to stereotypes, women who identified strongly with feminists, but not with women as a group, found it easier to identify positive words (e.g., “happiness”) when it was preceded by the prime “woman” (M = 606 ms), compared to the prime “man” (M = 625 ms). That is, this group showed evidence for evaluative in-group bias after subliminal exposure to stereotypes. The mean response times for the other trial types are shown in Table 2. The same group also showed evidence for in-group bias in the Moral Choice Dilemma task. After subliminal exposure to stereotypes, women who identified strongly with feminists but not with women were more willing to sacrifice men (M = 2.38 out of 4) than women (M = 1.77 out of 4). Finally, there was evidence for resistance in the form of persistence in a counter-stereotypical domain – a maths task. Women who identified with feminists, but not with women in general, persisted longer in trying to solve an unsolvable maths problem after subliminal exposure to stereotypes (M = 27s) compared to counter-stereotypes (M = 14s).

Table 2. Mean reaction times per trial type in the evaluative decision task reported in van Breen et al. (2018). After exposure to stereotypes, “distinctive feminists” associate positive words with women more than with men.

		Mean RT
Subliminal Exposure	Target valence	Prime: Woman	Prime: Man	Mean difference
Stereotypes	Positive target	606	625	-19	In-group bias
	Negative target	644	653	−9
Counter-stereotypes	Positive target	626	624	−2
	Negative target	644	650	−6

Note: The response times shown here represent the meta-analytic effect across the 4 studies.

In sum then, although exposure to stereotypes was very subtle, women who are strongly identified with feminists, but not with women in general, have various strategies at their disposal to resist stereotypes, either through their behaviour or through attitudinal responses. What these strategies have in common is that they serve to counteract, reject or challenge stereotypes. Persistence in a counter-stereotypical domain, such as a maths or spatial task, can serve to disprove gender stereotypes, and challenge their truthfulness. In-group bias can serve to reject the negative value and low status implications that stereotypes convey on the low-status group, by emphasising that the group is valued. Similarly, out-group derogation can serve to reject the supposed status advantage of the high-status group. An additional aspect that is relevant to note here is that the occurrence of resistance depended on gender identification: resistance was evident amongst women who identify strongly with feminists but not with women. The manipulation here was based on subliminal exposure to gender stereotypes, and previous work has shown that this group object to and disapprove of gender stereotypes (van Breen et al., 2017), more strongly than women with other identification “profiles”. As such, it is possible that women who are strongly identified with feminists, but not women, are particularly “sensitive” to gender stereotypes, more likely to pick up on them when they are presented subliminally, and more “practiced” at addressing them (see also Kaiser et al., 2006). In sum, we suggest that the strong role of gender identification in this study (as compared to for instance de Lemus et al., 2018), is related to the subliminal nature of the devaluation cues.

More recently, we have examined resistance to subtle cues of devaluation in the context of national identity (van Breen et al., 2021). Aside from replicating findings outside the gender context (an issue to which we return below), this study was designed to provide insight into how resistance to subliminal devaluation cues comes about. Previous research has shown that subliminal cues are most powerful when they do not “stand alone” but rather when they fit into a larger framework of experiences (Strahan et al., 2002). In other words, even very subtle cues can trigger relatively sophisticated responses (such as resistance) if they are embedded within a larger framework of experiences of group-based devaluation. Here, we expose Spanish participants to various conditions where subliminal cues reflect negatively on the Spanish in-group, and positively on a national outgroup (Germans). Crucially, although several conditions contain subliminal cues that reflect negatively on the in-group, only one of them taps into a salient experience of group-based devaluation in the lives of our Spanish participants. It is in this condition that we expected to observe resistance to subliminal devaluation cues.

The manipulation draws on the Spanish/German intergroup context, and consists of four conditions. The first condition makes reference to national stereotypes, pairing the subliminal prime “Spanish” with negative stereotypical traits such as “lazy”, and pairing “German” with “efficient”. The second condition presented associations that reflected negatively on the Spanish in-group by pairing the prime “Spanish” with “poor”, and the prime “German” with “rich”. A third condition mixed both these types of associations, pairing “Spanish” with “lazy” and “poor”, and German with “efficient” and “rich”. This condition in particular was expected to trigger resistance by our Spanish participants, because it directly reflects inter-group tensions associated with the 2012 financial crisis (Sierp & Karner, 2017). Specifically, the public discourse surrounding the financial crisis suggested that the detrimental impact of the financial crisis in Spain was due to the in-group’s stereotypical traits (e.g., “poor work ethic”, Friedman, 2011), thereby legitimising the effects of the crisis as just and deserved (Kressel & Uleman, 2015). In sum, this condition was designed to serve as a subliminal reminder of a relevant threat to Spanish identity. Finally, the fourth condition was a control condition, where the associations are reversed: Spanish is paired with efficient and rich, German is paired with lazy and poor.

Across three studies we observed that, as expected, it was the legitimising condition where resistance was observed. After exposure to subliminal cues that legitimised the in-group’s disadvantaged position, we observed evidence for in-group bias in an evaluative decision task: participants were faster to recognise positive words (e.g., “happiness”) that were preceded by in-group primes than out-group primes. The reaction times observed across the three studies are shown in Table 3. This work extends the work discussed above, not only because it extends this branch of research outside of the gender context, but also because it demonstrates how resistance to subliminal cues of devaluation can come about. In this study we observed that not all subliminal associations trigger resistance, even if they reflect negatively on the group. Only the condition that legitimised disadvantage triggered resistance. There was no evidence for resistance in the condition that referred only to stereotypes, or in the condition that referred only to in-group disadvantage. As such, these findings suggest that resistance is triggered not by negative representations of the in-group per se, but rather in response to subliminal cues that remind participants of a salient threat in their daily lives. Previous research has shown that subliminal cues can trigger relatively powerful responses if they do not stand alone, but instead serve as reminders of more pervasive issues (Strahan et al., 2002). Indeed, this is what we see here.

Table 3. Response times to positive and negative trials preceded by in-group and out-group primes, after exposure to subliminal cues that legitimise in-group disadvantage. Across the studies, reaction times to positive words are faster when preceded by an in-group prime compared to an out-group prime.

	Target Valence	Prime: In-group	Prime: Outgroup	Mdiff
Study 1	Negative	620.89	621.13	−0.23
	Positive	610.65	620.61	-9.96	In-group bias
Study 2	Negative	679.24	652.29	26.95
	Positive	634.13	652.50	-18.37	In-group bias
Study 3	Negative	620.28	623.10	−2.82
	Positive	606.70	615.04	-8.35	In-group bias

It is worth noting that, across these lines of research, we have measured explicit, subtle, and implicit resistance strategies. However, in response to subliminal cues, resistance seemed to occur primarily through implicit and subtle strategies. To some extent it seems intuitive that when the individual is not consciously aware of devaluation cues, direct resistance strategies (e.g., verbal confrontation) are not applicable, and resistance instead occurs through more subtle strategies. At the same time, this demonstrates how an exclusive focus on confrontation and protest might underestimate resistance amongst members of devalued groups.

One of the key implications of the evidence reviewed above is that resistance to gender-based devaluation can occur in circumstances where it had previously been considered impossible. For instance, our work on implicit forms of resistance demonstrates that these strategies can be used to resist even when cues of gender-based devaluation were presented subliminally. Previous research had shown that subtle forms of gender-based devaluation (e.g., benevolent sexism; ambiguous cues; implicit cues) undermine options for resistance through collective action or confrontation (e.g., Kray et al., 2001). Such work has often been interpreted to mean that, in such circumstances, resistance is not possible. However, our work shows that women can still counteract gender-based devaluation through other strategies, such as counter-stereotypical behaviours (de Lemus et al., 2018) or responses that reaffirm group value (van Breen et al., 2018). Likewise, our work demonstrates that resistance, even in its implicit form, is triggered specifically by devaluation experiences that tap into salient threats to the group (van Breen et al., 2021), demonstrating how resistance is tailored to the specific context in which it occurs. In sum, given evidence that “modern” forms of devaluation are increasingly subtle, this line of work demonstrates that this need not be an impediment to resistance. In sum, we believe that in this line of work we have taken the first steps in broadening the scope for resistance to group-based devaluation. While much work remains to be done (as the following section will discuss), we believe that the work reviewed above demonstrates the principle that resistance to gender-based devaluation is more flexible and more prevalent than previously thought.

Outstanding questions regarding resistance

There are several outstanding questions in research on resistance to group-based devaluation. In this last section, we will reflect on these questions and discuss interesting possible answers to these questions suggested by the work that has already been done in this line, which we have reviewed above.

Beyond the gender context

Much of the research we review above has been conducted in the context of gender – namely women being (implicitly) devalued relative to men. This begs the question of whether the diverse and subtle resistance strategies we outline above only occur in the gender context. We believe there are two points worth making on this issue. First, some of our recent work has shown evidence for resistance to subtle devaluation cues in the context of national identity (van Breen et al., 2021), specifically Spanish national identity. Second, in that study there was also evidence that resistance to subtle devaluation cues is aided by an “embeddedness” in the lived experience of those affected. This suggests that not *all* devaluation contexts would trigger resistance of this kind, but specifically those that have a relevance to participants’ daily lives. As such, we suggest that there are several contexts beyond gender that can trigger the kind of diverse and flexible forms of resistance we have outlined above.

A study by Lee et al. (2012) speaks to these issues in the context of race-based devaluation. The authors examined how women of Asian and Black ethnic groups in the United States respond to racism. In a first study, they exposed Black women (N = 36) and Asian women (N = 34) to either a racist comment made by an experiment confederate in an online discussion task, or a “generally rude” comment without racist implications. The participants were then given several options for how to respond. They could confront the confederate directly, but later in the procedure could also respond to the racism comment indirectly, by distributing good-tasting and bad-tasting jelly beans to the confederate, as part of an ostensibly separate task. Assigning bad-tasting jelly beans to the confederate was taken as an indirect response to the racism they had experienced earlier (i.e., retaliation). Results showed that participants preferred the indirect response over the direct confrontation. Although the sample was small, this study indicates that the use of indirect coping strategies is also observed in the context of race-based devaluation. Further examination of resistance to other forms of devaluation, such as race-based devaluation, or devaluation experienced by non-binary or transgender individuals (Doyle et al., 2021; Lewis et al., 2022) represents a key area for further research.

Mechanisms of implicit resistance

The possibility of resistance being triggered outside of conscious awareness is perhaps the most interesting issue raised in recent literature on resistance to group-based devaluation. One outstanding question in this area focuses on the mechanisms that allow such responses to arise. Where do implicit responses to devaluation come from? The literature reviewed above has provided indications for several possible mechanisms.

First, on the side of implicit motivation to resist – one motivational mechanism that contributes to implicit forms of resistance is evident in the paper by Ramos et al. (2016). The authors demonstrate that after exposure to sexism, female participants showed reduced implicit gender bias. The authors then apply the QUAD model (Conrey et al., 2005) to examine how this response comes about. Results showed that after exposure to the sexism condition (relative to the control condition) female participants’ accuracy-orientation was increased. That is, after exposure to sexism, women are more motivated to respond accurately in the Go-No Go association task where correct responses are stereotype-independent. Thus, it seems that the motivation to resist the impact of stereotypes (or other cues of devaluation) on behaviour produces an increasing emphasis on stereotype-independent behaviour. This finding is reminiscent of the findings discussed above regarding task persistence as a resistance strategy, as well as the stereotype-boost literature (Crisp et al., 2009) in which stereotype exposure produces better performance.

Learning and practice represent another mechanism that can facilitate resistance responses at the implicit level. A person might be motivated to resist devaluation generally, and with practice this tendency becomes increasingly ingrained or “chronic”. Subsequent resistance responses are then triggered “automatically” by the experience of group-based devaluation. This possibility is supported by evidence in the work reviewed above, for the role of politicised identities such as – in the context of gender – feminist identification. Feminist identifiers resist and object to gender-based devaluation in many different contexts (van Breen et al., 2017) – and as such can be said to have extensive practice in resistance. This tendency to resist can then be activated “automatically” and produce implicit resistance responses. Relatedly, those individuals with strong stigma consciousness (Mendoza Denton et al., 2006; Pinel, 1999) or high rejection sensitivity (London et al., 2012) maybe more likely to show implicit responses to group-based devaluation, as for them, too, group-based devaluation is a chronic concern.

A third possible mechanism arises from recent work on basic processes that may serve as the building blocks for resistance, or “triggers” for the motivation to resist. As we have noted previously, resistance is often thought of as a strategic, deliberative process. However, when we consider implicit forms of resistance, we must consider the contribution of more basic and routine processes. Research from our own lab has provided some indications that, indeed, resistance is evident as part of basic processes (van den Berg et al., 2022). In an EEG (Electroencephalogram) experiment, female participants were exposed to images of women and men represented neutrally (without any reference to social roles or devaluation), or engaged in either stereotypical or counter-stereotypical jobs, while EEG was recorded. There were four patterns of interest from the perspective of resistance. First, trials representing stereotypes and counter-stereotypes are clearly differentiated from neutral trials. That is, the social information contained in stereotypes (relative to neutral trials) is detected relatively early on. Second, our female participants were sensitive to the self-relevance of the images – stimuli containing female actors drew more attention than stimuli containing male actors. Third, participants responded to low status images (women in stereotypical roles and men in counter-stereotypical roles). Thus, in early processing, low-status representations of the in-group draw particular attention. For women, this corresponds to stereotypes. The task demands further required participants to press either “Yes” or “No” in response to each of the images, based on a feature of the stimulus that was unrelated to its stereotypical content. We observed that when a female actor was presented stereotypically (as opposed to counter-stereotypically) the requirement to respond yes triggered an response which in the literature has been described as Conflict-Related Negativity. That is, responding yes to a stereotypical representation of women triggered conflict (compared to saying yes to a counter-stereotypical representation). In sum, then, this work provides some indication that resistance not only relies on complex higher order cognitions, like awareness of socio-political issues or political orientation. Instead, the “building blocks” of resistance such as self-relevance, threat, and status, are evident in basic processes. Taking together the insights from this work, we argue that resistance to group-based devaluation need not be a deliberative, strategic response, but is an integral part of the psychological functioning of devalued group members, through its links to basic processes such as threat, stress, status, and self-relevance.

The question of the mechanisms underlying implicit forms of resistance remains an outstanding question that requires further study, particularly when it comes to the relative contributions and interactions of the mechanisms outlined above. Nevertheless, we believe that the work discussed above provides some interesting starting points for work in that direction.

Is intent required before a response can be seen as resistance?

In the broader literature on resistance, one central outstanding question is the issue of intent (Hollander & Einwohner, 2004). Does the individual need to intend for a certain act to be resistant before it can be interpreted as such? We believe that consideration of implicit forms of resistance provides an interesting perspective on this issue. Given that implicit responses are not subject to conscious deliberation, the person in question is not aware of having engaged in the response, and as such cannot be said to have intended a response to be resistant, at least not consciously. However, as discussed above, there is increasing evidence that processes like motivation, intention, and goal pursuit can operate outside of conscious awareness (Bongers et al., 2009; Glaser & Knowles, 2008; Moskowitz & Li, 2011). That is, goals, motivations and intentions, can be formed, triggered and pursued without the need for conscious intervention. In other words, previous work on implicit cognition suggests intent might operate outside of conscious awareness. Further, above we have considered a situation where the motivation to resist becomes “chronic”. In such a case, it might be possible for resistance responses to occur more automatically, without having to apply deliberate intent each time. As such, we would suggest that intent is crucial in the occurrence of resistance, but that intent itself need not be conscious or deliberated in each specific instance.

What is the function or aim of resistance?

Our approach to resistance also has relevant implications for what its aims are. In the social identity literature, the ultimate aim of social competition – and related strategies like resistance – is to produce social change, that is, to achieve a situation in which the in-group is no longer devalued. As such, it is relevant to ask here whether subtle resistance strategies can contribute to social change. We argue that it is certainly possible for more subtle forms of resistance to contribute to social change. Small acts on the part of an individual, such as counter-stereotypical behaviour – as outlined above – can disprove and undermine stereotypes and, over time, lead to their breakdown, which in turn contributes to social change. However, in our framework resistance, aside from producing social change, can also serve a number of other identity-related aims, such as maintaining collective self-esteem and well-being after value threats (Branscombe, Schmitt, et al., 1999; Rudman et al., 2007) or re-gaining a sense of control after control threats (Fritsche et al., 2017), which then in turn can contribute to improving the group’s social position. Importantly, as before, these different aims need not be mutually exclusive (Becker & Barreto, 2019).

Conclusions

This review synthesises evidence from recent literature to present a new perspective on resistance to experiences of gender-based devaluation. In the last decade or so, we have seen a renewed interest in resistance – and increasing attention to the broad “repertoire” of resistance strategies members of devalued groups have at their disposal. Here, we have reviewed research in the context of gender relations that supports that perspective, and have illustrated some of the implications of adopting such a broader view of resistance. We hope to convey that resistance to gender-based devaluation is not an exception, but rather an integral part of the coping repertoire of women. Women can counteract gender-based devaluation by means of diverse strategies, including some strategies that can take place outside of conscious awareness. This point has important theoretical implications, not least for existing theories of intergroup relations that tend to focus on ostensibly more conscious strategies of coping with group disadvantage (e.g., social identity theory) but also those that claim people consciously justify this disadvantage (e.g., system justification theory). Although contextual factors sometimes incline devalued groups to act in ways that can perpetuate the status quo, the motivation to counteract group-based devaluation (i.e., resistance) is a core aspect of devalued groups’ psychological responses. Extending the repertoire of resistance to the more implicit and subtle realms allows us to look at these theories in a new light, and suggest additional light at the end of the tunnel for disadvantaged groups.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The work was supported by the Ministerio de Ciencia Tecnología y Telecomunicaciones [PSI2016-79971-P].

Notes

¹ Note that “action” as used here is not synonymous with behaviour or physical action, but rather as the opposite to passive, as highlighted in the manuscript text.

² Primes that are presented very briefly would normally be labelled as “subliminal” but in this study the primes were not masked, so we prefer the description “very brief” in this case.