Dilemmas of resistance: How concerns for cultural aspects of identity shape and constrain resistance among minority groups

ABSTRACT A major theme in social psychological models of collective action is that shared social identity is a critical foundation for resistance and collective action. In this review, we suggest that this foundational role of social identity can be double edged for many minority groups because material disadvantage is often coupled with the historical erosion of key aspects of ingroup culture and identity. Focusing on the role of ingroup language in Welsh national identity and history in relation to African identity, we present an integrative review of our research on five different dilemmas of resistance that can arise when perceived threats to socio-cultural aspects identity sit alongside threats to the material position (in terms of status and power) of the group. We conclude that the central role of social identity in collective action and resistance can itself present challenges for groups whose core sense of who they are has been eroded..

Research on the social psychological determinants of collective action has flourished in recent years, offering considerable insight into when and why members of disadvantaged groups take action to improve their group's position.A major theme in this research is that a sense of shared social identity is a critical foundation for collective action.In this review, we suggest that for many minority groups, this foundational role of social identity can be double-edged.This is because material disadvantage is also often coupled with the historical erosion of key aspects of ingroup culture and other group-defining attributes, constituting a threat to the very sense of who "we" are.This combination of minority status and threats to ingroup-defining cultural attributes presents a set of dilemmas of resistance that must be navigated by minority groups seeking to improve their ingroup's position.
In this paper we present an overview of our research into how members of minority groups deal with these dilemmas of resistance, focusing on the cases of Welsh national identity (in relation to England) and African identity because of the strong components of language, culture, and history in each case.We outline five such dilemmas, concentrating on the potential tensions between strategies that emphasise the distinctiveness and "purity" of ingroup identity and those that focus more on redressing outgroup advantage.As we elaborate later, these dilemmas are particular to cases in which perceived threats to socio-cultural aspects of identity sit alongside threats to the material position (in terms of status and power) of the group, and thus are located within a much wider array of dilemmas and non-obvious choices faced by disadvantaged groups.The dilemmas we address relate to how minority groups (1) deploy "distinctive" cultural attributes; (2) build support for potential political goals; (3) spread "distinctive" attributes within the ingroup; (4) orient towards legal and illegal forms of action; and (5) use narratives of rich ingroup history to mobilise collective action in the present.
Before presenting our research, we provide a brief overview of social identity perspectives on resistance and collective action in minority groups, highlighting the relative absence of concerns about threats to identity and culture in these models.We then review research that has focused more directly on how group members react to the experience of perceived threats to the distinctiveness and vitality of ingroup identity and culture, and also consider how concerns about ingroup culture and identity have historically been implicated in a wide range of social movements.Focusing on the role of ingroup language and history in shaping minority group strategies, we then present an integrative review of our research on five dilemmas of resistance faced by minority groups.We conclude that the central role of social identity in collective action and resistance can itself present challenges for groups whose core sense of who they are has been eroded.

An overview of perspectives on social identities, collective action, and resistance
Several prominent social psychological models of collective action by disadvantaged groups posit a crucial role for social identities: a shared sense of "us" that underpins action taken for the benefit of, and together with, others who share that social identity (e.g., Simon & Klandermans, 2001;Simon et al., 1998;Tausch et al., 2011;Thomas et al., 2009Thomas et al., , 2012;;van Zomeren et al., 2008;Wright et al., 1990).This research has drawn on many perspectives, but a frequently cited framework is social identity theory (SIT; Tajfel & Turner, 1979), and in particular its conception of social competition as a strategy for dealing with relatively low ingroup status.The "classic" SIT position is that social competition occurs whenever boundaries between an ingroup and higher-status outgroups are relatively impermeable (ruling out individual mobility), and when status relations between the groups are perceived as insecure (illegitimate and/or unstable).Accordingly, recent models of collective action by disadvantaged groups have focused in different ways on appraisals of illegitimacy and the associated emotion of anger, and on appraisals of social change as possible (e.g., when group efficacy is appraised as high).

Implications of appraising an ingroup identity itself as under threat
While not necessarily a formal assumption of these models, it is nevertheless tacit that the social identities underlying collective action are clear and obvious to those involved, both in terms of their meaning (e.g., the defining features of "our" culture) and also in terms of their integrity, in the sense that the distinctiveness and even existence of the group is not in question.Put another way, in pre-existing groups, collective action by group members arises to the extent that they identify with the group.
It is helpful to question how well this tacit assumption maps on to the struggles in which minority groups engage, and how resistance might be shaped by other sorts of concerns that arise when ingroup identities are themselves not so clearly distinct and are facing erosion.Indeed, these concerns for the distinctiveness and integrity of an ingroup identity are considered in Tajfel's earlier theorising in which he observed that "We have previously characterised 'social competition' as based on the minority's aims to achieve parity with the majority; but in other ways, the minority aims to remain different" (Tajfel, 1978, p. 16).For Tajfel, the twin concern with equality and collective distinctiveness underlay a strategy of accommodation, consisting of "the minority's attempts to retain their own identity and separateness while at the same time becoming more like the majority in their opportunities of achieving goals and marks of respect" (Tajfel, 1978, p. 16).
This strategy of accommodation clearly overlaps with the definition of social competition offered later by Tajfel and Turner (1979), but suggests that the social psychological challenges faced by minority groups are often multiple.Of particular concern here is that potential threats to the worth or value of one's social identity -characterised as status threat (Branscombe et al., 1999;Riek et al., 2006) -often covary with potential threats to the identity itself (e.g., distinctiveness threat: Jetten et al., 2004;Spears et al., 2002).
Separately from research on collective action, an extensive tradition of social identity research on ingroup members' concerns for the integrity and vitality of an ingroup identity shows that these are also important drivers of intergroup behaviour.For instance, research on basic principles derived from SIT has shown that ingroup bias can be driven by group members' motivation to establish a positively-distinct ingroup identity (Jetten et al., 2004;Spears et al., 2002).When the distinctiveness of an ingroup from a relevant comparison outgroup is unclear, group members use available dimensions of comparison to establish favourable differences between the ingroup and outgroup, resulting in well-documented patterns of ingroup bias and favouritism (Jetten et al., 2004;Scheepers et al., 2002;Spears et al., 2002Spears et al., , 2009)).Even at the most basic motivational level, group members are thus concerned not just with how positively their ingroup compares with an outgroup, but also with how meaningfully distinct their ingroup is.
A concern for the integrity and vitality of ingroup identity has been connected directly to collective action by other more recent research.Echoing the concept of "ethnolinguistic vitality" in early social identity work on language and identity (Bourhis et al., 2019;Giles & Johnson, 1987;Giles, Bourhis, et al., 1977), Wohl and colleagues (e.g., Wohl et al., 2020) have examined the experience and consequences of what they term "collective angst": a group-based emotion experienced when "the future vitality of the cherished ingroup is perceived to be under existential threat" (Wohl et al., 2020, p. 482).The experience of collective angst has been found to predict support for group-based action aimed at alleviating that threat, including action that strengthens ingroup identity (e.g., through encouraging the learning of ingroup language, history, and culture; Wohl et al., 2010Wohl et al., , 2011)), and action against outgroups that are perceived to be a source of that existential threat (Jetten & Wohl, 2012;Wohl et al., 2014).Of particular relevance here, these actions can include support for both peaceful and violent forms of protest, with the latter receiving more support among diasporic Somali and Tamil communities when they held a politicised collective identity, but also experienced collective angst (Wohl et al., 2014).Similarly, Kachanoff and colleagues found that support for collective action among Black Americans was greater when they experienced collective autonomy restriction, defined as the perception that outgroups were attempting to limit Black American's ability to determine and express their identity (Kachanoff et al., 2020).
The relationship between social change and the experience of existential threat to cultural identity is also emphasised in de la Sablonnière and colleagues' typology of social change (de la Sablonnière & Taylor, 2020;de la Sablonnière, 2017).A perceived threat to cultural identity is a key condition for what de la Sablonnière (2017) terms dramatic social change, defined as "a situation where a rapid event leads to a profound societal transformation and produces a rupture in the equilibrium of the social and normative structures and changes/threatens the cultural identity of group" (p.12).
Coping with dramatic social change thus depends in part on the ability to address that threat to cultural identity.Otherwise, a state of collective inertia (de la Sablonnière & Taylor, 2020; de la Sablonnière, 2017) can arise when a dysfunctional societal structure cannot be addressed because concurrent threats to cultural identity inhibit functional collective responses.
More broadly, Sweetman et al. (2013) conceptual overview of possible goals of social change highlights that different social movements have quite different goals, ranging from collective mobility within a subjectively-legitimate system, through to separatism or revolution.Amongst these, forms of separatism in particular highlight the central role of concerns about being meaningfully different in shaping how minority groups resist the influence of dominant outgroups.However, this range of possible social change goals is not well integrated into prominent models of collective action, for which social change goals are centred more narrowly on overturning injustices, and where the goal is in some sense to become more like an outgroup in terms of social, economic, and political position.This narrow focus of collective action research may also be contrasted with the more general concept of resistance, which encompasses not only direct and overt collective challenges to the status quo, but also a wider range of more subtle practices and strategies (often focusing on ingroup culture) through which minority groups push back against an outgroup's hegemony (Hollander & Einwohner, 2004;J. R. Vollhardt et al., 2020;Leach & Livingstone, 2015;Raby, 2005;Rosales & Langhout, 2020;Scott, 2000).

Identity and resistance in minority groups: lessons from social movements
It is also instructive to consider the sheer range of social movements for which the establishment of meaningful difference is their goal, in the form of some sort of official recognition of cultural, social, economic, and/or political separateness.In Europe alone this is evident in the number of separatist movements that are based at least in part on distinctive language and cultural traditions (Anderson, 2018;Kostov, 2020).In some cases these movements may be marginal and benign, but in others they are implicated in major social change, with action ranging from peaceful protest and political action through to terrorism (Anderson, 2018).
The case of nationalism in the British Isles provides a good illustration: a feature of the nationalist movement in Ireland which secured independence from the UK following violent conflict in the early 20 th century was the cultural revival and assertiveness -based on the Irish language, traditional sports, and Gaelic culture more generally -that developed alongside political and economic claims.The development of (peaceful) modern nationalisms in Scotland and Wales is likewise characterised by claims of cultural and, particularly in the case of Wales, linguistic distinctiveness (Christian et al., 1976;Davies, 2007;Giles, Taylor, et al., 1977;Lynch, 2011).Indeed, Douglas Hyde, who helped to found the Gaelic League in Ireland in 1893, couched the need for Irish national independence as being primarily about cultural distinctiveness, because only addressing material disadvantage would merely involve . . .making Ireland a land of wealth and factories, whilst they extinguished every thought and every idea that was Irish, and left us, at last, after a hundred years of good government, fat, wealthy, and populous, but with all our characteristics gone, with every external that at present differentiates us from the English lost or dropped (Hyde, 1892, p. 4) These European examples also echo phenomena in many other parts of the world, especially when considering how the current geopolitical map of the world emerged following the collapse of the major Europe-centred empires in the last couple of centuries.As historical analyses suggest, the anti-colonial resistance that literally reshaped the world during this period was characterised by the assertion of cultural distinctiveness of various forms, manifested in examples such as Négritude in the context of Black and pan-African liberation, Islamic and Arabic cultural influences in the context of Middle Eastern and North African resistance to British and French colonialism, and Hindu cultural and religious revivalism in anticolonial resistance in India (Gopal, 2019;Pieterse, 1990;Robinson et al., 2021;Said, 1994).As Said puts it, this resistance required that "rebellious 'natives' impress upon the metropolitan culture the independence and integrity of their own culture, free from colonial encroachment" (p.241).

Five dilemmas of resistance: an overview
The above review of both social psychological research and the historical record presents a powerful case for the importance of cultural and linguistic distinctiveness to understanding collective action, resistance, and social change.Balancing concerns for equality and the integrity of ingroup identity and culture is easier said than done, however.If one is acting from the position of having a clearly-defined and distinct ingroup identity, then responding to illegitimate low ingroup status is -notwithstanding the stability of intergroup status differences -arguably a relatively straightforward proposition, at least in terms of knowing what the point of acting actually is.In contrast, we suggest that the combination of threats to status and threats to identity is particularly challenging because it presents choices or dilemmas for minority group members when it comes to choosing paths of resistance.These dilemmas arise because of the way in which the "ideal" of achieving equality while remaining or becoming different is often viewed in the light of potentially more achievable outcomes such as equality at the expense of difference (e.g., through assimilation), or forgoing specific political/economic equalities (at least in the short term) in favour of protecting/promoting the distinctiveness of ingroup identity -especially where "distinctive" features of that identity (culture; language) have been eroded.
In the following sections we outline five such dilemmas.As noted above, these dilemmas relate to how minority groups (1) deploy "distinctive" cultural attributes; (2) build support for potential political goals; (3) spread "distinctive" attributes within the ingroup; (4) orient towards legal and illegal forms of action; and (5) use narratives of rich ingroup history to mobilise collective action in the present.

Dilemma 1: How to deploy 'distinctive' cultural attributes such as ingroup language in relation to political goals
Minority groups often draw on a range of cultural or stereotypical attributes in the face of a hegemonic majority outgroup (e.g., Gopal, 2019;Pieterse, 1990;Robinson et al., 2021;Said, 1994).Minority group members can therefore use the ingroup's defining characteristics and attributes as flexible identity management resources, directing them towards the ingroup's identity needs and instrumental goals.This produces its own challenges, though, and the first dilemma we address arises because there is often considerable variation in terms of (a) which attributes or characteristics are seen as ingroup-defining, and (b) the meaning and value ascribed to particular attributes (Haslam et al., 1997).Moreover, the meaning of these attributes and their value are often actively debated by ingroup members: who "we" are and what defines "us" are not just cognitions or intra-psychic calculations; they are also constructed, posited, and debated through rhetoric and discourse (Antaki & Widdicombe, 1998;Billig, 1985;Edwards, 1991).This is important because of the consequences of defining the ingroup in particular ways as a basis for legitimating particular forms of intergroup relations (e.g., Wetherell & Potter, 1992), and as a strategy of influence (e.g., S. Reicher & Hopkins, 2001;S. Reicher et al., 1997).
The aim of the first phase of our research (Livingstone, Spears, & Manstead, 2009) was therefore to examine how identity-defining attributes can be deployed in different ways by ingroup members, and how these are given impetus by characterisations of the ingroup's intergroup position.We conducted a thematic analysis of 17 conversational interviews, focusing on how the Welsh language is deployed as an identity-defining attribute by selfdefined Welsh people.We focused on how this deployment was located within characterisations of intergroup power relations between Wales and England, the legitimacy and stability of these relations, and of the nature of intergroup threat that flows from them.In this way, we were interested in how these characterisations worked as entrepreneurial products by examining the ways in which they can implicitly or explicitly create impetus towards alternative social relations (Billig, 1985;S. Reicher & Hopkins, 2001).The interviewees included Welsh speakers and non-Welsh speakers and represented a wide range of identity and political positions (e.g., regarding Wales' position in the UK).Ages ranged from 20s to late 60s.Interview questions focused on (1) participants' own sense of national identity, including important characteristics of "Welshness", (2) the role of the Welsh language, and (3) the relationship between Wales and England.
The findings showed that Welsh speakers characterised the importance and role of the Welsh language in very different ways, and that these characterisations were given impetus by their perceptions of the wider Wales-England relationship.Extracts from three interviews help to illustrate these differences.The typical view of the Wales-England relationship was that it was historically and/or currently unfair or illegitimate.Within such characterisations, there was in turn variation as to whether this relationship was stable/unchangeable, or unstable/changeable.
Focusing first on characterisations of the Wales-England relationship as illegitimate and stable, the issue of Welsh visibility or distinctiveness became most apparent, particularly in relation to England and Englishness.Extracts from two interviewees, I.4 and I.5, are presented in Table 1, and contain similar characterisations of the Wales-England relationship.On the one hand, I.4 defines Welshness in terms that do not explicitly centre the Welsh language as a defining attribute.After stressing the importance of political and social values, and the tendency to resist identity threats in "subtle", "pragmatic" ways, I.4 in turn characterises the Welsh language as wholly non-essential to claiming Welshness: instead, they assert that speaking Welsh is a matter of individual conscience, rather than a collective imperative.I.5 characterises Welsh identity quite differently.In addition to Wales' stable and illegitimate low status/power, they deploy the Welsh language as the only available means by which to protect or assert Welsh distinctiveness, due to the historical erosion of other possible bases for a distinctive culture.This characterisation is used to generate an imperative on anyone who claims Welshness to learn and use the Welsh language.
The key contrast between the positions offered by I.4 and I.5 is therefore not in terms of the material relationship between the ingroup and hegemonic outgroup -it is in terms of the concurrent vitality or erosion of ingroup (Welsh) culture, and what this means in terms for the role of the Welsh language.Characterisations of stable, illegitimate low status/power can therefore be used to deploy particular attributes to create different action impetuses, depending on concurrent characterisations of threat to ingroup distinctiveness.
Characterisations of illegitimate low power: possibilities for change.In contrast, illegitimate low status/power could also be characterised as unstable or changeable, as was the case for I9 (summarised in Table 2).In our analysis, Table 1.Extract from participants I4 and I5 (Livingstone, Spears, Manstead, & Bruder, 2009), illustrating contrasting positions regarding the relationship between Wales and England, and the importance of the Welsh language to Welsh national identity.
Participant I4 Historically it's antagonistic (. ..) primarily from England to, towards Wales because they had the best deal, like, and they had the most to gain from kind of eroding any kind of, cultural identity but also political identity.But, (. ..) we just accept that kind of, um, we have a lesser status, I think, um, not in the sense of we don't want to do anything about it, but in the sense that these things happened, and, you know, there's very little we can do about it.One of the first things I would say is that Wales would be, uh, primarily a kind of socialist country, from my perspective anyway, whereas England would be not necessarily (. ..), secondly I would say that, um, being Welsh has some, some connotations related to, uh, resistance of a, an imposed culture (. ..) there's more of a subtle resistance in being Welsh and also a more pragmatic approach to being Welsh.
I don't think it's a necessary thing, in any way, shape or form, it doesn't confer Welshness and neither does it kind of, um, stop you being Welsh if you don't speak Welsh.I think it's just again a matter of personal preference, (. ..) just a matter of personal taste.

Participant I5
We're just like the forgotten ones, we're, we're almost ignored or not taken seriously or whatever, um, and yeah I've, that frustrates me and I must confess to having a bit of a chip on my shoulder about it (. ..) we're really, we're really hidden behind England, I think that's the major, kind of, um, relationship we have to England, and from a Welshman's point of view it really is far from ideal.
If you cut out the language, the language makes a heck of a difference because Anglo-Welsh culture or whatever it's called really isn't that distinctive.When you look at Welsh culture, Welsh, Welsh language culture, you've got a whole lot of things (. ..) the problem is when you get rid of the religion, the industry and stuff like that, and the language, there's not that much that sort of defines us as a people.
People who feel very Welsh (. ..)I think they should, um, I think they should learn the language (. ..) it's such an important part of their identity (. ..) it's important that they speak Welsh, that they learn Welsh in order to keep their birthright alive.
we suggested that the characterisation offered by I9 emphasises change as an ongoing process involving not only the strengthening and consolidation of Welsh identity, but also the realisation of political equality and ultimately separation from the United Kingdom.The broader project of Welsh identity is therefore explicitly bound up with a specific vision of the current and future political status of Wales.It is within this framework that I9 deploys the Welsh language: as a resource for political mobilisation and consciousnessraising, directing ingroup members towards particular political goals, in addition to its role as a marker of cultural distinctiveness.This is also an overtly strategic characterisation, as is evident from I.9"s claims regarding non-Welsh speakers" position in and claim to membership of the national category.The cultural and political importance of the language is here balanced against the reality that the majority of Welsh people do not speak the language.While the deployment of the Welsh language as a means of collective mobilisation goes hand-in-hand with a particular characterisation of the Wales-England relationship, it is thus also constrained by other structural realities within Wales that may impede the goal of national mobilisation.
The dual concerns with an ingroup's status position and the integrity of ingroup identity can thus produce very different positions with respect to whether, why, and how a particular attribute is important to ingroup identity.This in turn has implications for intergroup strategies, on the one hand (e.g., cultural regeneration; political separatism), and for the management and policing of the ingroup category, on the other (e.g., who is truly "one of us").The essence of this first dilemma is the difficulty of coalescing around a truly shared definition of ingroup identity that can be geared towards effective resistance.This difficulty arises because the precise nature of that outgroup's influence is itself often contested.

Dilemma 2: How to build support for potential political goals given the importance of specific cultural attributes such as ingroup language
During the time we were conducting our research on Welsh identity, an influential radio current affairs programme in the UK (the Today programme on BBC Radio 4) ran a feature on the promotion of the Welsh language in Wales.The feature began with an interview with a Welsh speaker who expressed worry about the supposed zeal with which the Welsh language was being promoted.The most striking aspect of this interview was that the interviewee insisted on meeting the interviewer in a secret location and that his voice was disguised, such was his concern about the consequences of expressing such an opinion.(Livingstone, Spears, Manstead, & Bruder, 2009).

Participant I9
We're no fans of the English, you know (. ..)I'm not a big fan of the English, never have been, and probably never will be because of what they've done to Wales (. ..) but the way things are going now (. ..) we don't fear the English any more.
That's what's, what's brought about the big renaissance, say from 1950 onwards . . . the Welsh language, and people like Saunders Lewis kicked off, the students got hold of it, and it's, it became a political football at the time . . .you cannot divorce politics as well (. ..) This is what I call the silent revolution (. ..) you cannot, uh, divorce the language and compartment it, can't divorce the culture, you know, um, it's all part of the Welsh whole you see (. ..) they can exist on their own, but when they come in, together, that's when the Welsh identity is really good (. ..)So you put all the economies, the politics, the cultural and the language (. ..)Before you're very vulnerable to England walking over us.They don't walk over us any more.I don't mind it, because I think we need to take them on board, because you can't, you know, alienate, it's a bonus that you speak the language, I wouldn't think less of anybody because they don't speak the language.
This incident highlights a more general point: that while the definition of ingroup identity is clearly contested, certain definitions, attributes and values can become dominant.Processes of consensualisation (Haslam et al., 1997), and the efforts of political elites within the minority category (S.Reicher & Hopkins, 2001) can mean that the importance and meaning of particular attributes become difficult to contest.This can be seen as a necessary step towards consolidating or reasserting the distinctiveness and vitality of the minority identity.However, it may have other less positive consequences for the minority group's social movements and political goals.As already noted, cultural resurgence within a minority group is often married with political campaigns geared towards attaining recognition, and sometimes separation from a hegemonic majority.However, this can be difficult because the distinctive attributes that are held up as ingroup defining may not be accessible to many within the minority category.A possible consequence is that when an attribute such as ingroup language is positioned as essential, ingroup members who do not have that attribute are less likely to support shared group goals because they feel positioned outside of the ingroup category.This is the basis of the second dilemma of resistance.
When some -or even many -members of a minority category do not have specific attributes, what are the consequences of reifying an ingroup identity in terms of those attributes?In our research, we focused on the potential consequences of such reification for patterns of national identification, and for support for the specific potential political goal of national autonomy for Wales.We drew in particular on approaches that emphasise that, rather than being fixed or a given, an individual's social identification is sensitive to the ingroup's wider intergroup position (Doosje et al., 2002;Jetten et al., 2001;Tajfel, 1978), and one's own position within the ingroup (Jetten et al., 2003).The analyses we describe below therefore tested the indirect association between individuals' Welsh language ability (as a specific instance of having or not having a criterial ingroup attribute) and support for national autonomy for Wales, via national identification.

Indirect association between Welsh language ability and support for national autonomy
In a first study (Livingstone, Manstead, et al., 2011, Study 1) we conducted a basic test of these hypotheses.We sampled 116 sixth-form students (aged 16-18) at a school in which Welsh was the language of instruction, and took questionnaire measures of (1) their self-rated Welsh language ability; (2) Welsh identification; and (3) support for greater national autonomy for Wales.Consistent with our predictions, Welsh language ability positively predicted Welsh identification (β = .45,p < .001),which in turn positively predicted support for national autonomy (β = .42,p < .001).The indirect association between Welsh language ability and support for national autonomy via Welsh identification was also significant, b = 0.17, SE = 0.05, 95% CIs [0.07, 0.28].

Consequences of Welsh language ability on national identity: the role of linguistic context
We examined this pattern further in a second, larger study of 646 adults who defined themselves as Welsh.Participants were recruited from one of two types of regional context in Wales.For the purposes of the study, we termed these "Welsh-speaking regions" (190 participants largely from Northern and Western areas of Wales) and "non-Welsh-speaking regions" (456 participants from Eastern and Southern areas).
Two additional issues were of concern.First, social identification is typically a more complex and nuanced position than simply identifying with a specific category.As several approaches in the social identity tradition highlight, multiple social categorisations are available to us in most situations (Crisp & Hewstone, 2006;Hornsey & Hogg, 2000;Mummendey & Wenzel, 1999;Turner et al., 1987).This is very much the case in Wales, where "Welsh" and "British" (which usually represents something of a superordinate category that includes various combinations of people within the UK, but which is also often equated with "English"; see Langlands, 1999) are available.Also potentially available as a national identity is "English".This is usually the archetypical "outgroup" category against which "Welshness" is defined.However, the identity dynamics surrounding the Welsh language present intriguing possibilities for how even self-defined Welsh people orient towards "Englishness".The relatively consensual position of the Welsh language as defining of Welsh national identity, combined with the fact that the majority of people who claim Welsh national identity cannot speak Welsh, means that non-Welsh speakers who would otherwise see themselves as Welsh can also find themselves positioned as "English" by Welsh speakers -especially in predominantly Welsh-speaking communities (Bowie, 1993;Trosset, 1986).The extent to which not being able to speak Welsh impacts on English identification, as well as Welsh and British identification, was therefore of particular interest in this study.
The question of when Welsh language ability might impact on patterns of national identity (and English identification in particular) led us to the second novel feature of this study -the comparison of effects across linguistic/regional contexts.The Welsh language and perceptions of its importance are not spread uniformly across Wales.We reasoned that, as with many large-scale minority categories, the ways in which identity dynamics play out in Wales are therefore likely to depend on features of the local context.Specifically, we hypothesised that the implications of Welsh language ability for patterns of national identification would be more pronounced in regional contexts in which the Welsh language was more prominent and perceived as important to Welsh identity.

Welsh language ability and patterns of national identification
The first step in our analysis was to test the predictive effect of Welsh language ability on ingroup (Welsh), superordinate (British), and outgroup (English) identification, and how these associations varied as a function of linguistic context.The associations between Welsh language fluency and ingroup (Welsh) and superordinate (British) identification were straightforward, and not moderated by linguistic context: greater Welsh language ability predicted greater Welsh identification, and lower British identification.
The pattern for outgroup (English) identification was more striking: its association with Welsh language ability was moderated by linguistic context, with the association being virtually zero in non-Welsh-speaking regions.In contrast, Welsh language ability was strongly associated with English identification in Welsh-speaking regions, such that lower levels of Welsh language ability were associated with much higher levels of English identification.This was despite the fact that Welsh identification remained extremely high (>6 on a 7-point scale) across the board.

Predicting support for national autonomy
The next step was to test a multi-group path model, in which Welsh language ability was the predictor, Welsh, British, and English identification were parallel mediators, and support for national autonomy was the outcome.Linguistic context was the grouping factor.This model is summarised in Figure 1.
Echoing the findings of the first study, the indirect path from Welsh language ability to support for national autonomy via ingroup (Welsh) identification was positive and significant.The indirect path via British identification was also positive and significant (Welsh language ability predicting lower British identification, and lower British identification predicting greater support for national autonomy).
More intriguingly, in Welsh-speaking regions, but not in non-Welshspeaking regions (as per the interaction described above), lower Welsh language ability predicted greater English identification.When the effect of British identification was fixed to zero (due to the statistical covariance and cultural overlap between "English" and "British" identification), English identification in turn predicted lower support for national autonomy.
The findings therefore indicated that in contexts where the Welsh language was both more widely-spoken and positioned as essential to Welsh identity, not being able to speak Welsh was associated with much stronger outgroup (English) identification despite concurrent high levels of Welsh identification.Greater English identification was in turn associated with lower support for national autonomy.

Balancing ingroup mobilisation with ingroup cultural vitality
The findings of this second study crystalise Dilemma 2: positioning a cultural attribute as central to a minority ingroup identity might protect against cultural erosion, but in doing so it may also have unintended consequences for mobilising towards collective political goals if not everyone who claims that identity can also claim that cultural attribute.The process implicated is that not having a key cultural attribute can lead one to be positioned and come to identify as "other" or even "outgroup", no matter how strongly one claims that minority ingroup identity.

Dilemma 3: Whether to spread 'distinctive' cultural attributes within the ingroup
If the inaccessibility of an important attribute to category members sows the seeds of the previous dilemma, then it may seem that this dilemma at least has a straightforward remedy: Spread the criterial attribute within the category to be mobilised.Indeed, efforts to do exactly this are characteristic of many social movements among minority groups.These movements clearly vary in the extent to which the spread of the cultural attribute is explicitly aligned with political and social struggles, ranging from relatively apolitical efforts to get people to learn a language as a way of reconnecting with their culture at a personal level, through efforts to ensure that everyday interactions and transactions are conducted through the medium of a particular language, to explicitly politicised efforts at consciousness-raising through the regeneration of distinctive cultural practices.What unites them is a concern to enhance the vitality of the attributes that are held up as ingroup-defining.
We suggest that such efforts risk creating another dilemma, this time for those who already have a criterial attribute.The roots of this dilemma lie in the potential consequences of the spread of an attribute for their own position within the minority category.It may seem intuitive to suggest that an increase in the strength or prevalence of a defining attribute should be seen as positive, and that the larger this increase, the better for the ingroup.However, while the "distinctive" attribute becoming more widespread within the ingroup might strengthen the ingroup's position vis-à-vis an outgroup, it also holds the potential to undermine the position of those within the ingroup who already have that attribute, as their own distinctive position wanes (Brewer, 1991;Tajfel & Turner, 1979).
Rather than always being an accessible feature of most group members' everyday lives, cultural attributes may be "iconised" (Coupland et al., 2003); that is, they are given a value that is more symbolic.As Coupland et al. (2003, p. 171) put it with regard to the Welsh language, "In its representation . . .Welsh is focused and dignified.That is, the distance achieved by the metalinguistic framing of Welsh seems to be a vertical distance -Welsh being elevated and wondered at."Asserting the value of a cultural attribute in this way can reflect and reinforce the inaccessibility of that attribute to many group members.It also positions those who do have access to that attribute (e.g., Welsh speakers) in a relatively special, prototypical position: a "cultural elite", to use Jones ' (1996, p. 41) term.
We hypothesised that the dilemma posed by spreading a defining cultural attribute (strengthening cultural vitality, distinctiveness, and mobilisation potential vs. losing one's own relatively "special" position within the group) would under some circumstances lead to ambivalence regarding the prospect of large increases in the prevalence of cultural attributes.A factor that we expected would shape whether the spread of a cultural attribute is seen positively by those who already have that attribute is the comparative context.When the potential spread of a defining cultural attribute is appraised in the context of the wider relationship between the minority ingroup and hegemonic outgroup, we expected that such a spread would be seen positively.In contrast, we expected that the potential spread of a defining cultural attribute would be seen less positively when it is appraised in the narrower context of relations of different subgroups within the minority category.Here, the salient distinction is between those who do and do not have that cultural attribute already (Welsh speakers and non-Welsh speakers, for instance).This should bring concerns about the loss of a "special", prototypical position within the minority to the fore, with the result that widespread adoption of that cultural attribute within the ingroup is viewed less positively.
In one study we tested this hypothesis among Welsh speakers (N = 163), focusing on their responses to the possible future spread of the Welsh language amongst Welsh people (Livingstone, Spears, et al., 2011, Study 2).We manipulated the future prospects for how widely-spoken the Welsh language would be in 40 years' time, creating three conditions that respectively stated that 25% of the Welsh population would be able to speak Welsh (no increase from present numbers); that 50% of the Welsh population would be able to speak Welsh (a moderate increase); or that 75% of the Welsh population would be able to speak Welsh (a large increase, which would make Welsh speakers a clear majority among Welsh people).We also manipulated the comparative context, such that the study was framed as being about relations between Wales and England (inter-group context) or as being about relations between Welsh speakers and non-Welsh speakers in Wales (intra-group context).
The main outcome variable was an index of how positively or negatively the projected future extent of Welsh language ability was perceived to be.On this outcome, the interaction we predicted was large and highly significant (η 2 p = .109),indicating that the effect of the potential spread of the ingroup language did indeed depend on the comparative context in which it was appraised (see panel A of Figure 2).In the inter-category (intergroup) context, the prospects of a moderate or a large increase in Welsh language use were seen much more positively than no increase, and similarly to each other: compared to the current low levels, any type of increase was seen very positively.
In the intra-category (intragroup) context, the effect of the potential spread of the ingroup language was markedly different.Here the prospect of a large increase in Welsh language use was viewed much less positively than the prospect of a moderate increase.Further, the simple effect of comparative context was also highly significant in the large increase condition: such an increase was viewed much less positively in the intra-group condition than in the inter-group condition.
We also tested this hypothesis in another study (Livingstone, Spears, et al., 2011, Study 1) using a similar design but in a different minority group (women; N = 198) and with a different stereotypical, group-related characteristic (emotional intelligence).The idea in this study was to provide a conceptual analogue of the situation of the Welsh language in Wales -that is, a setting in which a defining ingroup attribute is accessible to only a minority within the minority group.As a first step in this study, our British women participants completed a questionnaire that supposedly assessed their emotional intelligence.Regardless of their actual answers, all participants were told that their scores placed them in a subgroup of women who score highly on emotional intelligence.They were also told that emotional intelligence is often seen as an important characteristic that differentiates women from men, but that in reality only around 25% of women were (like them) in the high emotional intelligence subgroup.This created an analogue of Welsh speakers' position in Wales: as a minority within the minority ingroup who have an ingroup-defining attribute.
The procedure thereafter was very similar to the study involving Welsh speakers.We manipulated the future prospects for how widespread high emotional intelligence would be among women as a whole in 20 years' time.The three conditions respectively stated that 25% of women would be high in emotional intelligence (no increase from present numbers); that 50% of women would be high in emotional intelligence (moderate increase); or that 75% of women would be high in emotional intelligence (large increase).We also orthogonally manipulated the comparative context, such that the study was framed as being about relations between women and men (intergroup context) or as being about relations between high emotional intelligence women and low emotional intelligence women (intra-group context).The main outcome variable was again how positively the future prospects were appraised.
The results were very similar to those in the Welsh speaker study.The interaction between the future prospects and comparative context manipulations was significant (although smaller, η 2 p = .039,perhaps not surprisingly given the more contrived nature of the context -see panel B of Figure 2).In the inter-group context, the prospects of a moderate or a large increase were seen more positively than no increase, and similarly to each other.In contrast, in the intra-group context the prospect of a large increase in emotional intelligence was viewed less positively in the intra-group condition than in the inter-group condition.

The mediating role of concerns about subgroup identity being undermined
In the study involving Welsh speakers, we also examined a potential process through which the comparative context influenced on the perceived positivity of a large increase in Welsh language use: concerns about the identity of Welsh speakers as an important subgroup being undermined.Participants in this study also completed a scale assessing these concerns (e.g., "Any special importance of Welsh speakers in Wales would be undermined").Echoing the main positive appraisal outcome, the interaction between future prospects and comparative context was significant on this measure.The simple main effect of comparative context was again significant only in the large increase condition: when faced with the prospect of a large increase in Welsh language use, Welsh speakers reported stronger concerns that their position as an important subgroup would be undermined when that increase was framed in intra-group terms than in inter-group terms.The indirect effect of comparative context on positive appraisal via concerns about subgroup identity being undermined was in turn significant in this large increase condition.In other words, framing a large increase in terms of relations within the ingroup heightened concerns about the subgroup's special position being undermined, which in turn predicted a less positive view of the prospect of a large increase in Welsh language use (see Figure 3).
In summary, a third dilemma of resistance emerges because spreading a defining ingroup attribute such as language within an ingroup may have a double-edged effect among those who already have that attribute and are thus a prototypical minority within the ingroup (Livingstone, Spears, et al., 2011).The roots of this dilemma lie in the potential consequences of the spread of an attribute for relations within the minority category, and whether it is seen as undermining the "special", distinctive position of people who already have that attribute -although whether this occurs only when the attribute is relatively rare (as opposed to already widespread) is unclear, and is a key question for future research.As it is, the findings of these studies suggest that rather than being a straightforward means of resisting an outgroup's hegemony, spreading "distinctive" attributes thus involves balancing distinctiveness and status concerns within as well as between minority groups.

Dilemma 4: Orienting towards conventional/legal or radical/ illegal forms of action
Summarising our argument so far, minority group members' orientations towards a majority outgroup are complicated by the influence of different forms of intergroup threat -particularly threats to status and threats to ingroup identity and culture itself -and the diversity of (sometimes competing) responses that these forms of threat can demand.It might be objected that in the research described so far, one or both of these forms of threat have been assumed, rather than directly examined.Is there direct evidence that appraisals of the illegitimacy of ingroup-outgroup relations (as an indicator of status threat) interact with appraisals that ingroup identity itself is under threat?And if so, how do these forms of threat combine and translate into specific intergroup orientations and preferences for action?The dilemma that these questions speak to is one that is common in a range of social movements: the dilemma of what form (and focus) of action in which to engage.Our research suggests that at least part of this dilemma lies in how perceptions of ingroup identity as vulnerable and threatened shape the consequences of perceiving the ingroup's relationship with a dominant outgroup as illegitimate.
Recent research on collective action has shed light on the conditions under which group members support and engage in different forms of collective action, and particularly when support may emerge for radical forms of action (often referred to as non-normative collective action), including violence and destruction of property.Relevant factors include appraisals of the relative efficacy of legal and illegal forms of action (Saab et al., 2016), group-based emotions such as anger (Sweetman et al., 2019) and contempt (Tausch et al., 2011), intergroup dynamics that redefine "violence" as legitimate self-defence against an aggressing outgroup (Drury & Reicher, 2000; S. D. Reicher, 1996;Stott et al., 2007), and perceptions of an ingroup identity as facing an existential threat (de la Sablonnière & Taylor, 2020; Wohl et al., 2014).Our own research links several of the these insights by focusing on intergroup appraisals and emotions, and anger in particular as the proximal emotional predictor of collective action intentions (e.g., Mackie et al., 2000;Thomas et al., 2009;van Zomeren et al., 2004).Anger in intergroup terms has been widely found to arise from appraisals of the illegitimacy of an outgroup's actions (Iyer & Leach, 2008;Lazarus, 1991;Mackie et al., 2000;Roseman, 1991;Smith & Mackie, 2015), and to in turn mediate the effect of illegitimacy on collective action intentions (van Zomeren et al., 2004).Our more novel hypothesis is that the emotional Figure 3. Indirect effect of comparative context (intergroup, rather than intragroup) on affective reaction to the prospect of a large increase in an identity-defining attribute, via the appraisal of one's subgroup identity as potentially being undermined (Livingstone, Spears, et al., 2011;Study 2).
impact of illegitimacy perceptions (in terms of anger specifically) depends upon concurrent levels of identity threat (defined as perceived threats to the cultural distinctiveness of ingroup identity, in contrast to other forms of threat such as to status).In other words, we predicted that the illegitimacy and identity threat perceptions would interact to predict intergroup anger, which would in turn predict support for different intergroup strategies.
An interaction between illegitimacy and identity threat reflects the way in which reactions to specific perceived social structural conditions are framed by the perceived vitality of one's identity.Reacting to high levels of illegitimacy may be considerably easier if the ingroup has a strong, vibrant identity base on which to draw and mobilise.When the ingroup's identity has been eroded or lacks vitality, however, the illegitimacy of the ingroupoutgroup relationship may serve to frame the ingroup's fragile identity in terms of malevolent outgroup influence, exacerbating a sense of anger.This reasoning echoes recent research on how the perceived vulnerability of an ingroup identity predicts support for collective action, including relatively radical, illegal forms (Chayinska et al., 2021;de la Sablonnière, 2017;Wohl et al., 2014).The contribution of our research is to examine how perceptions of threat to the vitality of an ingroup identity moderate responses to appraisals of relations with a hegemonic outgroup as illegitimate.
We tested the interactive effect of illegitimacy perceptions and identity threat in the same survey study of 646 Welsh-identified adults described in relation to Dilemma 2, but extending the analysis to a set of different collective action outcomes (Livingstone, Spears, & Manstead, 2009).As with Dilemma 2, we analysed support for national autonomy for Wales, which we conceptualised as action geared primarily towards change in the status of the ingroup (Wales) in relation to the hegemonic outgroup (England).Two other outcome variables, in contrast, focused on what we termed identity protective orientations, geared more towards shoring up the vitality of the ingroup's identity.One of these represented a constitutional, legal form of action: support for a petition to protect and promote the Welsh language.The other represented support for a more radical, illegal form of action: a well-known campaign of arson that took place against second homes (i.e., residences that were not the owner's main home) owned by English people in Wales.This campaign is widely recognised to have been motivated by a desire to protect Welsh-speaking communities from cultural and linguistic dilution through "Anglicisation" (Humphries, 2008).

Illegitimacy and identity threat interact to predict group-based anger
Our analysis focused on whether appraisals of illegitimacy and identity threat would interactively predict these different forms of resistance indirectly, through group-based anger.Identity threat did indeed moderate the association between illegitimacy and anger: illegitimacy was particularly predictive of anger when identity threat perceptions were also high, with the highest levels of anger associated with high illegitimacy and high identity threat.
When it came to the different action orientations, the results revealed differential associations with the appraisals and with anger, as indicated in Figure 4.The desire for status change (support for national autonomy) was predicted by illegitimacy appraisals, but not by identity threat or by anger.Conversely, constitutional/legal action to protect ingroup identity was predicted by identity threat (and anger), but not by illegitimacy appraisals.In turn, anger was most strongly predictive of support for unconstitutional/ illegal identity-protective action.
In path analytic terms, these associations meant that identity threat (1) intensified the association between illegitimacy and anger, and (2) thus strengthened the indirect path from illegitimacy to radical/illegal identityprotective action, in the form of support for the arson campaign -but, not for status change.It is important to stress that these different intergroup orientations are related to one another (they covaried significantly), and in practice it is perfectly possible to support none, some, or all of these orientations at once.Nonetheless, our findings suggest that orienting to one form of action (e.g., identity-protective, including illegal forms) over others (e.g., seeking political parity through separatism) depends at least in part on how appraisals of "our" relationship with "them" are modulated by Figure 4.The interaction between illegitimacy and identity threat perceptions explains additional variance in intergroup anger: the predictive role of illegitimacy was stronger when identity threat was also high (from Livingstone, Spears, & Manstead, 2009).appraisals of the vitality of "our" identity, (re)shaping what is proper and possible as a form of resistance.
A further nuance to these findings came from the regional breakdown of the sample according to whether it was predominantly Welsh-speaking, or non-Welsh speaking, as described above in relation to Dilemma 2. Following similar reasoning, we tested whether the interactive effect of illegitimacy and identity threat would in turn be contingent on whether ingroup identity was defined in terms of the ingroup language (i.e., in Welsh-speaking regions), compared to where it was not (i.e., in non-Welsh speaking regions).This was indeed the case: as illustrated in Figure 5, the interactive effect of illegitimacy and identity threat was significant only in Welsh-speaking regions, where ingroup identity hinged to a greater extent on the language, and identity threat in turn was likely attributable more to the cultural influence of the dominant outgroup.
The interplay between illegitimacy and identity threat appraisals in shaping group-based anger and action intentions has important implications, not least for approaches to intergroup emotion that have generally attempted to map specific emotional experiences such as guilt, anger, or shame onto a precisely-defined appraisal base (Iyer & Leach, 2008;Mackie et al., 2000).Our analysis of this dilemma of resistance suggests that the emotional (and in turn, social and political) consequences of a particular appraisal can depend on other concurrent appraisals -a point that is quite central to appraisal theories of emotion at an individual level (e.g., Lazarus, 1991;Roseman, 1991).
More directly relevant to the question of resistance, these results underscore the importance of distinguishing between different social change goals and the functions of intergroup action (Sweetman et al., 2013).This is especially so in terms of how it can be geared towards status change (arguably a more instrumental goal), or towards shoring up, protecting, or asserting facets of an ingroup identity (actions that serve more of an identity function; Scheepers et al., 2003;Turner-Zwinkels & van Zomeren, 2021).These different goals and functions can clearly be interwoven in resistance movements, and there is no inherent tension between them.Yet, the relative emphasis placed on seeking large-scale constitutional change rather than action (legal or illegal) to protect the culture of dwindling language communities (for example) is a very real dilemma when it comes to choosing paths of resistance (for instance, in anti-colonial movements; Gopal, 2019;Pieterse, 1990).The roots of this dilemma lie in the extent to which the erosion of minority group identity changes the action that is proper and possible in the face of historical injustice.

Dilemma 5: Whether and how to use narratives of rich ingroup history to mobilise collective action in the present
Colonialism is not satisfied merely with holding a people in its grip and emptying the native's brain of all form and content.By a kind of perverted logic, it turns to the past of the people, and distorts, disfigures, and destroys it.(Fanon, 2001, p. 169) In a frenzy I excavated black antiquity.What I discovered left me speechless . . .Segu, Djenne, cities with over 100,000 inhabitants; accounts of learned black men (doctors of theology who travelled to Mecca to discuss the Koran). . . it allowed me to regain a valuable historic identity.The white man was wrong, I was not a primitive or a subhuman.(Fanon, 2021, p. 109) The first of the above quotes from Frantz Fanon makes the point that a central strategy of domination characteristic of colonial relations has been the dilution or even erasure of a colonised group's culture and history (see Robinson et al., 2021;Said, 1994).The second quote, in turn, illustrates that resistance to such oppression has often involved efforts to increase awareness of and pride in an ingroup's collective history (Choudry & Vally, 2017) -a point evident in movements as diverse as the "Gaelic revival" in Ireland that preceded its national independence, and forms of pan-Africanism that informed resistance to European colonisation in Africa (Pieterse, 1990;Said, 1994).
Shifting focus from language to collective history as an important, but often threatened dimension of an ingroup's culture, our recent research (Makanju et al., 2020(Makanju et al., , 2022) ) has addressed the effects of using positive representations of collective history (i.e., representations of the ingroup as having a meaningful history that can be seen positively by ingroup members) as a strategy of resistance.Focusing on African identity, we tested whether exposure to positive representations of African history would (1) produce a more positive sense of African social identity (e.g., in terms of identification as African, and perceptions of positive norms), and (2) lead to stronger intentions to pursue goals that benefit the ingroup (e.g., collective action to address social problems, and to assert the value and importance of Africa relative to Western countries) among African participants.
As detailed below, our findings suggest that exposure to positive representations of collective history has little direct effect on how participants appraised their African identity, or their intentions to take collective action on behalf of Africa.Instead, using narratives of a rich collective history to promote collective action and civic engagement in post-colonial societies may create a dilemma because appraisals of an ingroup's history are multidimensional, and can thus shape how group members engage with the ingroup and seek to shape its future in different and contradictory ways.

History and social identities
Examining the role of representations of collective history in minority group resistance connects our research to other approaches in social psychology that recognise the importance of collective histories and the appraisal of "our" past in shaping the present and future of an ingroup (Liu & Hilton, 2005).In general terms, the perception of collective continuity -that an ingroup identity stretches back through time as an enduring entity -has been found to be associated with ingroup identification and collective selfesteem (Sani et al., 2007), and with greater social wellbeing (Sani et al., 2008).The valence of an ingroup's collective past also matters: collective continuity is positive to the extent that the ingroup's history itself is also positive; in contrast, continuity with a past in which the ingroup behaved negatively can pose a social identity threat (Roth et al., 2017).The negative effects of knowing that an ingroup has perpetrated injustices in the past are also central to the large literature on group-based guilt (Branscombe & Doosje, 2004;Doosje et al., 1998;Roccas et al., 2006;Zimmermann et al., 2011).
Turning to groups that have suffered injustice, the effect of appraising the ingroup's past as negative has been found to weaken intentions to take collective action in the future (Rabinovich & Morton, 2012).The impact of an ingroup's history of suffering is also the focus of research on transgenerational trauma (Danieli, 1998; see J. R. Vollhardt, 2012, for a review) and collective victimhood (Bilali & Vollhardt, 2019;J. Vollhardt, 2020), with evidence of the potential for a collective history of suffering to lead to poorer wellbeing in the present, and to negative cycles of intergroup behaviour in the future (Noor et al., 2012).
Other research has indicated how more specific historical narratives are used (typically in politicians' discourse) to mobilise group members towards different group goals (Jetten & Wohl, 2012;Klein & Licata, 2003;Obradović & Howarth, 2018;S. Reicher & Hopkins, 2001).As an example, S. Reicher and Hopkins (2001) analysis illustrates how Scottish politicians invoke aspects of Scotland's history to create impetus towards political projects such as Scottish national independence.Indeed, the very same historical touchstones can be invoked either to support or to reject Scottish independence.
For minority and disadvantaged groups, positive historical representations thus have the potential to positively affect both group members' own relationship with their group, and their inclinations to work towards shared group goals.However, there are blind spots in this research, including the relative absence of work in post-colonial contexts, especially in Africa (with some exceptions; e.g., Klein & Licata, 2003;Licata et al., 2018), and the absence of a direct test of whether exposure to positive representations of collective history in post-colonial settings has positive effects on group members' perceptions of that ingroup identity, and on their willingness to act collectively for the group.

Do positive representations of collective history positively affect ingroup engagement?
Our first study (Makanju et al., 2020, Study 1) addressed these gaps by randomly allocating 162 African participants (the majority of whom were Nigerian) to view either a video portraying pre-colonial African history in positive terms, or a video portraying pre-colonial African history in the more negative terms in which it has typically been cast in Eurocentric accounts of history.A second study extended and refined this design by randomly allocating 431 African participants (again, the majority were Nigerian) to one of three different conditions.In all conditions, participants first watched a video that represented predominant colonialist views; specifically, of Africa and Africans as backward, lacking in history and culture, and in line with common racial stereotypes of "uncivilised", "tribal" people.This traditional colonialist representation of African history was presented first in order to make salient the negative colonial representations of African history.In one condition, this video was followed by a second video that presented a much more positive and fact-based representation of African history.In a second condition, a second video presented a more negative, but fact-based representation of African history (e.g., focusing on complicity of some Africans in the trans-Atlantic slave trade).The final condition was a quasi-control, in which there was no second video following the colonial representation video.
In both studies, the positive historical representation video described several different aspects of Africa's pre-colonial past, including sophisticated ancient civilisations, technical achievements, scholarship, and trade with other civilisations.Our aim was to assess the effect of viewing this on two broad sets of outcomes: (1) measures of how participants perceive and relate to African identity (including African identification, collective self-esteem, and group entitativity), and (2) identity management strategies and collective engagement, including collective political action to support pan-African political goals, and asserting the equality of Africa and Africans with Western culture.
Did exposure to a positive representation of African history have the positive effects we predicted?The answer across both studies was a very consistent "no": The manipulations had no significant effects on any outcome, with very small effect sizes (η p 2 < .01 in nearly all cases) across the board.These findings are on the face of it inconsistent both with published social psychological research on the effects of historical representations, and with many accounts of actual anti-colonial struggle (e.g., Césaire, 2000;Fanon, 2001Fanon, , 2021;;Robinson et al., 2021;Said, 1994).As such, our ongoing research focuses on understanding variation in the relationship between different representations of in ingroup's history, and the sorts of psychological and behavioural outcomes that such representations might spur.

Dimensions of appraisal of ingroup history
One part of this process is to unpack some of the unacknowledged multidimensionality of appraisals of "our" collective history.A feature of much research that has quantitatively examined the impact of representations of history on group processes is that an ingroup's history is appraised first and foremost, or even solely, in terms of its valence: was "our" history positive or negative?It is conceivable, though, that collective history is also appraised along other dimensions: we may see history of given valence as more or less relevant (subjectively important) to ourselves, our group, and to present-day challenges.Likewise, we might appraise an ingroup's history in terms of its richness, independently of its valence, both in terms of the amount of detail and events, and in terms of its temporal depth.The influence of historical representations may also be shaped by the appraised clarity of that history: regardless of its valence, the extent to which "our" past is perceived as easyto-understand, consensual, and vivid (vs.hard-to-understand, contested, and vague) may shape whether and how it is invoked and used in relation to present-day challenges.
Our most recent work (Makanju et al., 2022) has qualitatively examined the complexity of (1) how an ingroup's history is appraised, and (2) how representations of collective history are (or are not) employed as a basis for ingroup engagement in the present and future.We asked 33 participants who identified as African to write short essay-style narratives about African history in terms of how much they felt they knew about it, how rich that history was, its valence and complexity, and how important participants saw African history as being.We then also asked them to write down their thoughts about what the contemporary problems facing Africa are, and what they think can be done to solve these problems.As with the research described above in relation to Dilemma 1, a key focus of the analysis was on the rhetorical aspects of participants' responses; namely, if and how their characterisations of African history are used to create impetus towards ingroup engagement in the present and future.
Echoing the quantitative findings described above, in most cases participants' representations of history had an unclear or ambiguous relation with ingroup engagement (perhaps reflecting the fact that the question asking about contemporary problems facing Africa intentionally did not explicitly mention history).Nevertheless, some participants did make explicit connections between collective history and ingroup engagement, drawing on their representations of African history as a justification for why and how other Africans should work to address contemporary challenges.
Importantly, however, these narratives took two very different forms, as illustrated in the extracts in Table 3.We describe these as a history-asinspiration narrative, and a history-as-contrast narrative, respectively.On the one hand, P31 (a 44-year-old male from Nigeria) articulates the history-as-inspiration narrative.Their representation of African history is not only positive, but also emphasises its richness and complexity.Moreover, the role of hegemonic outgroups such as European colonist countries is explicitly invoked when characterising this richness and complexity: they describe African history as being every bit as rich as that of Europe (for example), but also as having been rendered as contested, and misunderstood because of misleading, negative Western counter-narratives about African history.The claim here is that African history provides a basis for positively evaluating African identity relative to comparators such as Europe, and the reason that it is not more widely seen as such is because of the malign influence of colonisation on Africans' (and others') understanding of that history.In turn, P31 asserts the importance of African history not only to themselves, but to other Africans, the world more generally, and to a specific understanding of contemporary challenges in Africa.It is on the basis of this characterisation that P31 asserts that knowledge of African history is crucial to addressing challenges in the present and the future -something to draw upon as an inspirational resource, echoing the quote from Fanon at the start of this section.
In contrast, P8 (an 18-year-old female from Tanzania) articulates the history-as-contrast narrative.They also emphasise the richness and complexity of African history, but in somewhat different terms to those of P31.Its complexity, in particular, is presented as a challenge to comprehensibility, and as a basis for disagreement and lack of consensus.Moreover, P8 characterises the most well-known aspects of African history as negative (e.g., as marked by conflict), and as something against which the more positive aspects of (contemporary) African cultures can be contrasted.While acknowledging the importance of African history to world history more generally, P8 is ambivalent about the importance of that history to herself and to other Africans, especially relative to other bases for a positive African identity, such as contemporary culture.On the basis of this characterisation of African history as negative, rich, complex, and subjectively unimportant to their sense of identity, P8 explicitly invokes African history as being something from which Africans must collectively break away in order to build a successful future.
The complexity with which an ingroup's history might be appraised helps to create a fifth dilemma facing minority groups whose identity and culture has been eroded: how might representations of "our" history be used to  (Makanju et al., 2022), respectively illustrating history-as-inspiration and history-as-contrast narratives.

P31: History-asinspiration On the richness of African history:
African history is as rich as, if not to say richer than, the history of Europe, Asia and the Americas.I am aware of the input Africa and Africans have made to the high ancient cultures of Europe and the Americas. . .Then there are various other empires/civilisations across the rest of Africa.All these have some level of records that go back to as early as 4500 BC . . .Much of the "great civilisations" of the ancient world were actually taught or learned from African civilisation

On the complexity of African history:
There are political positions that mitigate against honest, truthful, and transparent discussion/analysis of African history.This is the political position of the West that had to propagate lies about Africans in order to facilitate its designs to carry out the mass enslavement and dehumanisation of Africans . . .Thus African history would remain more complex than it should, until Africans begin to take a keen interest in its own history and conduct its own research into this.As regards the ease of visualising African history I consider this in two ways 1. Being able to picture events in mind: this is largely difficult, non-existent as a lot of Africans have very little information about their history.2. Being able to understand the flow of actions and actors, the actions-reactions sequences and how they lead to where things are today . . .I think a lot of Africans find this a bit difficult as they can't see parallels between where things are today to what has caused by viewing the same action-reaction . . .sequence. (Continued)

On the valence of African history:
There are two sets of African history, that which is contemporary and is generally negative and that which is older and generally not identified to be African.As an example of the second set the Egyptian civilisation is hardly ever portrayed as African . . .Of the former set popular media generally focuses on the negative aspects of the African experience as that plays to the gallery and also helps their personal financial agenda (they are in business to make money P8: History-as-contrast On the richness of African history: I do believe that noteworthy African history exists . . . the riches of Africa, the beauty, culture, food and so much more . . .There is so much to learn about African history, if we just consider the simple fact that we are talking about a whole continent and not a country then this alone should highlight how much there is to learn . . .I do feel it goes back thousands of years but I do not feel it is well documented . . .I feel that the history dating back hundreds of years ago is better documented.
On the complexity of African history: While I do not know much about the history of Africa as a whole, I believe it can be complex in some cases and straightforward in others, I cannot really explain why I feel this way.It is very controversial, many people would disagree about the real history of Africa, many people would find it hard to believe that Africa is actually not as bad as the media portrays it to be.Is it hard to visualise?Yes and No, sometimes it is hard to visualise when I think of the hardships, for example; when I think of slavery, and even the good things are hard to visualise.However, some aspects are easier to visualise than other

On the valence of African history:
The African history that is most spoken about is in my opinion generally negative while it should be positive because there is so much beauty and goodness in Africa.Most of the history that is publicised is of suffering, you will hear about war, famine and so on -while there is so much happiness and joy and I feel like that is what may people don't realise about Africa -people are so happy, whilst there are problems do not get me wrong, but people are happy. . . . . .African history . . . is mostly unpleasant.I feel this is because we only get to see one side of the coin.But if we looked deeper . . .you realise it is actually not all bad -you realise Africa is so beautiful and pleasant.

On the importance of African history:
In terms of relevance, I believe a lot of the history explains/helps in understanding why things are the way they are.African history is very very important to human and world history, I mean Africa is the mother land. . .Is African history important to how I see myself.
Yes and no -Some aspects yes but with some I would say no . . .I would say some of the culture back home is more important to how I see myself than the history is.African history to other Africans is very important but I feel the importance has faded over the past years -especially in the case of people who have lived outside the continent. . .(

On contemporary problems facing Africa:
Focusing on the past.Not seeing the beauty in the beautiful things that we have (the minds, creativity and so much more).We need to be educated about our countries and how brilliant and beautiful they are.

On Africa's future:
The future is bright if we stop looking at the past and focus on all the brilliance and beauty of our countries and culture.
mobilise action towards shared goals?This dilemma lies in the potential for positive or negative representations of collective history to lead to ingroup engagement or disengagement.A closer examination of other research on the impact of historical representations also reveals under-appreciated complexity in the effect of the valence of history.For instance, while a focus on a negative past has been suggested to undermine collective action (Rabinovich & Morton, 2012), Licata et al. (2018) found in an African context that appraising a negative history of colonisation and exploitation led to greater willingness to take collective action to seek reparations.Overall, our findings suggest that a sense of collective history can indeed inspire action in the present, but that there may be no simple effect of specific historical narratives in that regard.Understanding when and why historical narratives have specific mobilising or demobilising effects in turn likely requires a multidimensional approach to the appraisal of history.

Limitations and future directions
Before concluding, we want to reflect on some of the limitations of the work described above, and on possibilities for future research that might further our understanding of dilemmas of resistance faced by minority groups.A key limitation of our research is that it focused on a limited set of contexts and identities (e.g., Welsh national identity in dilemmas 1-4; African identity in dilemma 5).A priority in taking this research forward would be to examine the interplay of concerns about social/political disadvantage and threatened socio-cultural aspects of identity in other contexts.However, this is not a straightforward call for direct replication attempts of specific quantitative findings (important as such attempts are).Rather, we urge future work to test theoretical principles as they may play out in a given context, rather than simply attempt to replicate the bald quantitative effects described above.The reason for this is our belief that the form these effects take in our research on Welsh national identity, for instance, are shaped by the particularities of that context (e.g., that the identity-defining characteristic of the Welsh language is combined with the fact that it is only spoken by a minority of the population).In other contexts where a minority identity is less clearly dependent on one specific (and not widely shared) characteristic, the dynamics of resistance may be different.As an example, the dilemma of how to build support for potential political goals given the importance of a specific cultural attribute such as ingroup language (Dilemma 2) may be less pronounced when the group can draw on a range of other cultural resources, but this in turn may intensify the dilemma of how to deploy "distinctive" cultural attributes in relation to political goals (Dilemma 1) because of the many different bases of ingroup identity.The same theoretical principlesthat resistance to outgroup hegemony will be shaped and constrained by perceptions of the vitality of socio-cultural aspects of ingroup identity -may therefore give rise to different patterns depending on the setting in question.
We are also mindful that the quantitative evidence we have discussed in relation to some of the dilemmas came from non-experimental designs, including cross-sectional surveys (e.g., Dilemmas 3 and 4).While our supposition is that the processes described in those sections do reflect causal relationships, we do not yet have data capable of verifying the causal nature of those relationships, and we have tried to be careful not to infer causation as a result.Given that these are also contexts and topics that are not especially amenable to controlled experimental designs, a more viable approach would be longitudinal designs, such as panel studies.
Finally, it is important to stress that the dilemmas we have outlined above are simply those that are evident in our own research, and certainly do not exhaust the dilemmas that minority and disadvantaged groups face when it comes to resisting outgroup hegemony.For instance, the choice at the heart of Dilemma 4 (whether to orient towards conventional or more radical forms of collective action) also produces a number of other dilemmas that are discernible in research on resistance and collective action.These might relate to the communicative and strategic value of specific courses of action, such as the costs and benefits of violent versus non-violent forms of protest when it comes to the success of a movement and how the movement is viewed by broader society (Feinberg et al., 2020;Teixeira et al., 2020).Other dilemmas might relate to the risks posed by resistance in oppressive contexts (Ayanian et al., 2021), or more generally the social costs of resistance in terms of the reactions of close others such as family (Cornejo et al., 2021;González et al., 2021;Klandermans, 1997).While the dilemmas we have reviewed in this paper each relate to the confluence of threats to status and to socio-cultural aspects of identity, they therefore sit within a much wider array of dilemmas and non-obvious choices faced by disadvantaged groups when it comes to claiming social value and integrity.

Conclusions: The nature of threat and the nature of resistance
For members of minority groups the twin concerns of achieving parity and maintaining the integrity of the ingroup's identity and culture are interlinked but not always complementary, and these linkages can present dilemmas when it comes to choosing paths of resistance.While many social movements do seek to align cultural and political aspects of activism within a minority ingroup, our research suggests that this can pose dilemmas because one priority (e.g., revitalising or consolidating a distinctive ingroup culture) can easily be pitted against another (e.g., addressing socio-economic inequalities).We should emphasise that these dilemmas may not be apparent at the level of the individual group member, and do not necessarily reflect intrapsychic states of uncertainty or conflict.On the contrary, individual group members may be quite certain about the direction that their group should take.Instead, these are truly group-level dilemmas that arise because of, and are apparent through, contestation over the nature and value of ingroup identity and culture, and the ingroup's intergroup position.The pragmatism and flexibility required to mobilise a broad constituency of ingroup members towards a specific political goal may not always sit easily with cultural movements that seek to expunge outgroup cultural influences (especially when many ingroup members are not averse to those influences).A major challenge for social psychological perspectives on minority group resistance is to address when and why minority group members prefer one approach over another.

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
Some of the research reported here was made possible by support from the ESRC (RES-148-25-0014) and the Leverhulme Trust (ECF/2007/0050).

Figure 2 .
Figure2.The effect of future prospects for the numerical strength of an identitydefining attribute on affective reactions to those prospects is moderated by the comparative context (inter-group vs. intra-group).Panel a illustrates data from a study of Welsh speakers' reactions to potential increases in Welsh language use, while Panel B illustrates data from a study of women's reactions to potential increases in emotional intelligence among women (fromLivingstone, Spears, et al., 2011).

Figure 5 .
Figure 5. Path estimates for non-Welsh-speaking regions (upper panel) and Welshspeaking regions (lower panel).Values are standardised (β) regression coefficients, with the exception of paths from the illegitimacy X identity threat interaction term, which instead report unstandardised (B) coefficients followed by the standard error in parentheses (from Livingstone, Spears, & Manstead, 2009).

Table 3 .
Extracts from participants P31 and P8

Table 3 .
).When you apply both of these together you get people you say Africa has added nothing to world history, civilisation and I wonder then what was the Egyptian civilisation and what it added to the Greek and Roman civilisations of the West and Mesopotamia of the EastOn the importance of African history:African history is the history of the world.If Africa is the cradle of human life and the cradle of human civilisation, then its history IS the history of the world. . . . the day the world begins to take actual truthful interest in African history that is the day a higher level of justice being to prevail world over. . .I see myself as myself first and foremost then as African.However I do desire people around to be better informed about the African experience in order to foster better relations . . .I think a lot of other Africans are more easily negatively impacted by the negative outlook on the African image.I think this is largely due to a lack of knowledge of real African history and the understanding of how this history has shaped where Africa is today and how it is still shaping it into the future On contemporary problems facing Africa: Image/Self-image I think this is probably the area requiring greatest attention.Africans need to take time to understand themselves -who they are, who other Africans are, who Africans in the diaspora are, what our individual and joint histories are, and re-imagine ourselves in a positive light.Then begin to (actually simultaneously) work to address the negative image, information that Africa has with the rest of the World. . .Political Africans need to understand that the system of politics it has inherited from the West needs to be re-evaluated, keep what is positive and practical for it as well look inward to apply good practices from its own political structures/frameworks that it had before the incursion of the West On Africa's future: Continued manipulation and enslavement by the rest of the world, except: 1.It begins to take interest in its own history.Researching such histories itself, teaching this to its next generations and ensuring they continue to take interest and pride in their rich history 2. It begins to debunk the lies that have been perpetrated and are still being perpetrated about it.3. It begins to learn from its history and past mistakes, errors, etc. (Continued).