This is an open access article distributed under the terms of the Creative Commons
Attribution License (”Though you prove patriots a thousand times… some fine morning you find yourselves crossing the border and you are reminded by the mob that you are, after all, nothing but vagrants and parasites, outside the protection of the law.”
(Leo Pinsker, Auto-Emancipation – an early Zionist pamphlet, published in 1882, cited in Morris, 2001, pp. 16-17)
Such observations on the sorry fate of Russia’s Jews illustrate the point that however much individuals may regard themselves as belonging they are not necessarily regarded as such by those they regard as their fellows. Moreover, this example illustrates the potentially dramatic consequences of this asymmetry: To find one-self physically expelled speaks volumes about one’s place in the community.
Yet, dramatic as such examples are, there is a danger that an exclusive focus on hostile acts of rejection and discrimination neglects the more everyday occasions in which individuals may find their self-conception as members of the community questioned. Moreover, such a questioning of one’s inclusion is not only psychologically distressing but can be practically consequential for one’s ability to access the everyday benefits of group membership. These benefits are diverse. They can include the benefits of feeling valued and respected (Khan et al., in press-a, in press-b; Renger & Simon, 2011) and knowing that one can rely on others for support (Pandey, Stevenson, Shankar, Hopkins, & Reicher, 2014; Reicher & Haslam, 2009), etc.
Some of these benefits may appear rather distant from the rights we associate with ‘citizenship’. However, recent work has broadened our understanding of what citizenship entails and this requires that we attend to the everyday significance of being recognized as someone who ‘belongs’. Indeed, in the current paper we wish to explore how social psychological theorizing on the benefits of being recognized as a bone-fide group member can help illuminate the processes involved in the claiming of everyday citizenship rights. Specifically, we focus on the claiming of the right to be heard as a member of the national community and report experimental studies addressing the degree to which such claims count in the eyes of others and result in the individual being heard and treated as a co-national.
Before detailing our work, we consider (i) the nature of everyday citizenship, (ii) the claiming of rights on the basis of being a fellow national, and (iii) the reception of such claims.
Identity and Everyday Citizenship
Historically, citizenship has been conceptualized in terms of the formal rights associated with membership of the national community. However, contemporary citizenship research has increasingly broadened its focus beyond the formal and legal aspects of national citizenship (e.g., voting rights) to include hitherto un-theorized topics such as sexual rights and ‘sexual citizenship’ (Richardson, 2000), the recognition of minority identities (Soysal, 2000), etc.
This tendency to decouple citizenship from the formal rights associated with membership of the national community reflects a growing awareness that there are various social and institutional bases for claiming rights (e.g., relating to universal human rights; Soysal, 2000). It has also encouraged an interest in the variety of contexts (e.g., ‘the global city’; Sassen, 2002) in which people’s rights are claimed, negotiated and compromised. However, although the right to speak and be heard can be claimed on a variety of grounds not all of which equate to national identity (Barnes, Auburn, & Lea, 2004), the nation remains the most widely accepted political category that permeates everyday life (Billig, 1995). Certainly, many debates are framed in terms of debates about ‘the national interest’ (Elcheroth & Reicher, 2014; Reicher & Hopkins, 2001), and analyses of political rhetoric highlight the ways in which those who would achieve influence routinely present themselves as members of the national community (Burns & Stevenson, 2013; Reicher & Hopkins, 2001).
For all these reasons, the nation remains an important category and it should come as little surprise that those seeking to claim all manner of rights should work with this identity rather than against it (on this latter, see Haslam, 2014). This is well illustrated if we consider how gay and lesbian activists in Ireland sought to build popular support for legal reform though a campaign under the slogan ‘Proud to be Irish, Proud to be Gay’. This campaign characterised the Irish as ‘naturally’ tolerant and depicted homophobia as a legacy of British colonialism (Dunphy, 1997). In other words, although new rights – including sexual rights may be claimed on a variety of non-national bases (Richardson, 2000), all manner of rights may be claimed through identity performances that assert one’s full membership of the national community.
National Identities: Identification and Identity Claims-Making
Social psychological theorizing on group membership has long emphasized the importance of individuals’ subjective self-definition. Thus, self-categorization theory (Turner, Hogg, Oakes, Reicher, & Wetherell, 1987) maintains that defining oneself in terms of a group is the basis for group behavior (involving the conformity to group norms, etc.). Yet, one’s subjective identifications are not always recognized by others. For example, American-ness is implicitly associated with being white (Devos & Banaji, 2005) and although Asian Americans do not report being any less American than White Americans, they frequently experience interactions in which their American identification is questioned or denied (Cheryan & Monin, 2005). Moreover, such experiences are distressing (Huynh & Devos, 2011).
Such findings hint at the complexity to group membership. Certainly, an individual’s subjective identification with a group is important. However, as Jenkins (1996) puts it, “social identity is never unilateral” (p. 21, original emphasis) but requires validation by others. Drawing on this logic it follows that we can conceptualize social identity in terms of an identity claims-making process in which people actively lay claim to group membership and invite others to recognize such claims. This fits with research which shows that minority group members, aware that they may be judged ‘aliens’, display and perform their heartfelt identification with the national community (Cheryan & Monin, 2005). Such social identity performances can be complex, involving (among other things) decisions about their clothing (Hopkins & Greenwood, 2013) or food preferences (Guendelman, Cheryan, & Monin, 2011). However, such performances are not guaranteed acceptance. Indeed, the way in which minority group members perform their national identities may differ from how majority group members perform these identities, and these latter may well disparage the identity performances of the former (Joyce, Stevenson, & Muldoon, 2013).
When it comes to national identity, the claims-making process is particularly complex because the criteria for inclusion (e.g., ancestry, language, accent, subjective identification, cultural knowledge, etc.) are varied and contested. Some criteria (e.g., ethnic ancestry) are more deterministic than others (e.g., an individual’s subjective identification with the nation’s civic institutions), and although it is commonplace to differentiate between nations that prioritize either ‘ethnic’ or ‘civic’ criteria (the classic comparison being between Germany and France) the reality is that most countries use a mix of these criteria (Poole, 1999). This is well-illustrated in Scotland – the location for our research. Scottish surveys (McCrone & Bechhofer, 2008) show people endorse a range of more deterministic (ethnic) criteria (e.g., ancestry) and more open (civic) criteria in which, as Manzo puts it, membership is “handed out as a reward for loyalty and not on the basis of un-chosen criteria such as race” (Manzo, 1996, p. 19). Accordingly, national membership can be claimed on the basis of diverse criteria, and research shows that some people living in Scotland but born elsewhere and to non-Scottish parents (e.g., in England to English parents) claim a Scottish identity and do so through asserting their identification with Scotland, and its values and institutions (Kiely, Bechhofer, & McCrone, 2005; Kiely, Bechhofer, Stewart, & McCrone, 2001).
Yet, although such studies illustrate the point that claims to membership of the national community can be made on the basis of diverse criteria, there is little social psychological research concerning the reception of such claims and how they are socially consequential. Accordingly, our research addresses the degree to which majority group members differentiate between claims based on different criteria, and how this is consequential for the identity claimant’s ability to be heard as a fellow national.
Identity-Claims Reception
In some respects it is easy to use questionnaire surveys to research the criteria people believe are important in judging group membership and social identity. When this is done in Scotland, Scottish-born respondents tend to endorse ethnic over civic criteria when judging national ingroup membership. Thus, survey data show that whilst 43% of English-born people living in Scotland felt that to be ‘truly Scottish’ it was essential to be born in Scotland, the corresponding figure for those born and bred in Scotland was 64% (Hussain & Miller, 2006).
However, gathering such information does not allow us to conclude that Scots would necessarily reject a claim to Scottishness by someone born in England. Nor can we assume that any such rejection would be consequential for that individual’s treatment. Since LaPiere’s (1934) pioneering research, we know that what people say when the question is posed in the abstract is not always what they practice when faced with a concrete instance. Moreover, in this context there are good reasons why this may be so.
First, surveys of opinion about the criteria defining national identity necessarily broach the issue in an abstract and impersonal manner. Yet, claims to group membership are personal and can entail a heartfelt plea that one’s subjective identification be recognized. Accordingly, it is unclear whether abstract questions about the criteria for judging national inclusion necessarily predict behavioral responses to what can be impassioned, personalized identity claims.
Second, such identity claims are typically made in particular contexts (e.g., where one seeks to speak on a topic that one cares about and is of relevance to group members). Again, the reception of such context-bound claims cannot necessarily be predicted from de-contextualized weightings of various membership criteria. Moreover, questionnaire research which asks people to weigh the diverse criteria for group membership (e.g., ethnic heritage vs. civic commitment) is explicitly comparative which means we cannot know whether, in the absence of such an explicit comparative frame of reference, participants would actually differentiate between claims.
Third, when faced with a concrete instance of an identity claimant, people may be concerned about appearing exclusionary and narrow-minded (and thus apparently accept certain claims that they would actually rather not). Indeed, interview research with indigenous Scots suggests that in face-to-face interactions with incomers they find it hard to query ‘civic’ claims to Scottish national identification that are couched in terms of heartfelt identification with the nation and its culture (McCrone & Bechhofer, 2008). Such concerns may be particularly prominent if the reception of the claim carries any implication of differential treatment for different claims. Of course, it could be possible to present such claims in a context where the consequences of claims-rejection are zero. This could then allow participants to judge different claims without having to worry about there being any moral judgment of their response. However, this strategy immediately limits our ability to address a key issue motivating interest in claims-reception in the first place: whether the reception of such claims actually matters for the claimant’s treatment.
With these methodological considerations in mind we investigated the reception of identity claims in an experimental design which allowed us to present personalized claims by a particular individual in different conditions and in the context of warranting the claimant’s right to speak on a national issue.
The Present Research
We report three experiments conducted in Scotland, a country of a little over 5 million people which receives some 100,000 migrants per year, over half of whom are from England. English-born people are therefore easily the largest migrant group, constituting some 10% of the overall population. The fact the English share UK citizenship with the Scots (and therefore have full and equal rights in Scotland) separates out the question of formal citizenship from everyday citizenship and thus makes Scotland a particularly favorable site for investigating informal everyday citizenship.
As noted above, survey and interview research shows it is commonplace for ‘incomers’ (notably English migrants) to claim Scottishness on the basis of their present and future commitment to the country: whatever their history, the important thing is that their destiny is Scottish (Kiely, Bechhofer, & McCrone, 2005; McCrone & Bechhofer, 2008). Such people are sometimes referred to as ‘new Scots’ (Maan, 1992) and such civic claims are often encouraged in political debate. For example, the leadership of the Scottish National Party proclaimed its ambition “to see the cause of Scotland argued with English, French, Irish, Indian, Pakistani, Chinese and every other accent in the rich tapestry of what we should be proud to call (…) ‘the mongrel nation of Scotland’” (cited in Reicher & Hopkins, 2001, pp. 163-164).
Survey research suggests indigenous Scots endorse the significance of both more inclusive criteria (e.g., a subjective identification with Scotland, its values and its institutions) and more exclusive criteria (e.g., having a Scottish ancestry), and that the latter have greater weightings (McCrone & Bechhofer, 2008). However, for all the reasons discussed earlier, the predictive utility of such weightings cannot be assumed. Accordingly, our research involved presenting Scottish participants with an individual brought up and educated in England who self-defined as Scottish. In one condition their claim to Scottishness was based on their Scottish ancestry (an ethnic claim), and in another, on their identification with Scotland and Scottish culture (a civic claim).
Study 1 was designed to provide experimental confirmation of the survey and interview findings that ethnic claims to nationhood are accepted more than civic claims. Studies 2 and 3 built upon Study 1 in three important ways. First, whereas the claims to identity in Study 1 were conveyed in a written description, those in Studies 2 and 3 were brought to life in an individual’s identity performance (presented via video). Second, the claims to identity were woven into a context. Specifically, the video showed the claimant speaking on a topic concerning Scotland and its future, and the identity-claim was worked into their speech as they explained their concern for the topic. This means that the claim was introduced in a context where it was meaningful and socially relevant. Third, the fact that the claim was introduced in the context of a speech allowed investigation of the degree to which different claims impacted upon the acceptance of the claimant’s ideas (indicating something of the social consequentiality of the reception of different identity claims).
Needless to say, as our work is rooted in the social identity tradition it follows that the reception of identity claims is likely to be affected by observers’ own level of identification with the group (on the grounds that group processes are more relevant for those with higher identifications). Indeed, it is possible that any differential evaluation of such claims would only be exhibited by those highly invested in the group identity and our analyses sought to address this issue.
Study 1
This study sought to confirm previous survey findings concerning the priority of ethnic over civic claims to nationhood in Scotland. Participants were given a written vignette in which an individual provided a self-description. Embedded in this was a claim to Scottish identity based either (according to condition) on parentage and place of birth (ethnic condition) or commitment (civic condition). We predicted that, when asked to rate how Scottish the claimant was, indigenous Scottish respondents would provide higher ratings in the ethnic than the civic condition.
Method
Participants
Fifty-five Scottish-born undergraduate psychology students (male n = 7, female n = 48, age M = 21.8, SD = 5.72; civic = 27, ethnic = 28) acted as participants. Data were gathered in class time and participants received no compensation.
Materials
Participants read a passage reporting how someone working in Scotland described themselves. In both conditions they were described as having an English accent. In one (ethnic claim) he described how he was born in Scotland to Scottish parents who had then moved to England where he had been brought up and educated. In turn, he explained that he had now returned to Scotland. In the other (civic claim), he described how he was born in England to English parents but had chosen to move to live in Scotland and commit himself to Scottish interests. Both of these types of claim are commonplace in Scotland and their wording was based upon existing survey and interview data.
The text in the ethnic condition was as follows:
I was born in Scotland of Scottish parents. When I was young, my parents moved south of the border, but although I have lived most of life outside Scotland, I could not help but absorb the connection of my ancestors with Scotland’s natural environment. Scotland and its environment are in my blood. So, when I was offered the opportunity to play a part in shaping Scotland’s environmental future, I grabbed at the opportunity. It allowed me to give some tangible expression to my Scottish ancestry.
In contrast, the text in the civic condition was:
I was born in England of English parents. But from my earliest youth I felt a fascination and an affinity with Scotland and its natural environment. That is why I chose to study and to specialize in environmental issues. Scotland and its environment have always been on my mind. So, when I was offered the opportunity to play a part in shaping Scotland’s environmental future, I grabbed at the opportunity. It allowed me to give some tangible expression to my commitment to Scotland.
After reading either paragraph, participants responded to a number of statements. In this and subsequent studies they used seven-point scales (anchored 1 = strongly disagree and 7 = strongly agree). Four statements (see Appendix A) concerned the claimant’s Scottishness (Cronbach’s α = .90) with a higher score indicating greater acceptance. Participants also reported their nationality and completed four identification items (see Appendix A) which formed a scale (α = .90) with a higher score indicating greater Scottish identification.
Results
The level of identification scores was high (M = 5.86, SD = 1.12) and non-normally distributed (mode = 6.50, Mdn = 6.00, skewness = -1.69, kurtosis = 3.39). This precludes using participants’ level of identification as a moderator in our analyses of the effect of condition on our dependent measures. Accordingly, as those identifying as Scottish are most relevant to our hypotheses we report two sets of analyses. The first is for the full sample (regardless of identification level). The second excludes those exhibiting a lower identification and employed a criterion score of greater than 5.5 on a 7 point scale (this is the point at which, more often than not, people are expressing strong agreement with statements about their identification; see Reicher, Hopkins, & Harrison, 2006). This resulted in a sample size of 39 (civic n = 16, ethnic n = 23) with identification high in both conditions (civic M = 6.44, mode = 6.50, Mdn = 6.50; ethnic M = 6.41, mode = 7.00, Mdn 6.50).
Using the full sample we found ratings of the claimant’s Scottishness were higher in the ethnic condition (ethnic M = 5.12, SD = 1.33 vs. civic M = 3.75, SD = 1.70), t(53) = 3.33, p = .002, Cohen’s d = .89. When the sample was restricted to the higher identification sample the effect of condition was repeated: ratings of the claimant’s Scottishness were higher in the ethnic condition (ethnic M = 5.12, SD = 1.32 vs. civic M = 3.70, SD = 1.76), t(37) = 2.88, p = .007, Cohen’s d = .91.
Discussion
In this first study we presented participants with an individual who had an English accent and had been raised outside Scotland. We found that such an individual was judged as having a stronger claim to membership of the Scottish ingroup if they could invoke the Scottishness of their parents and place of birth than someone whose claim rested on their subjective identification with Scotland. In our next studies we investigated if this difference was consequential for these two individuals’ treatment.
Study 2
Much research shows influence to be moderated by group membership: ingroup members are more influential than outgroup members (Abrams, Wetherell, Cochrane, Hogg, & Turner, 1990; Turner et al., 1987). The moderating role of group membership is well-illustrated in research on the intergroup sensitivity effect which shows that the same criticism is more accepted when coming from ingroup members than outgroup members (Hornsey, Oppes, & Svensson, 2002). Furthermore, compared to criticism from an out-group member, criticism from a fellow ingroup member is viewed as more constructive, elicits less negative affect, and results in greater acceptance of the need for behavioral change (Esposo, Hornsey, & Spoor, 2013; Hornsey et al., 2002; Hornsey & Imani, 2004; Hornsey, Grice, Jetten, Paulsen, & Callan, 2007).
In Study 2 we presented the claims used in Study 1 in two films. In both, the speaker was the same (with an English accent) and the speech identical – except with regards to the identity claim. After viewing the film, participants then read criticisms of the Scottish ingroup ostensibly made by the speaker in a post-speech interview (with the cover story stating we were interested in how people form impressions of people). Participants then completed ratings of the claimant’s criticisms. We predicted the nature of the speaker’s identity claim would impact upon the reception of his criticisms (i.e., better reception in the ethnic than civic condition).
Method
Participants
Thirty-four Scottish born psychology undergraduate students participated (males = 11, females = 23; age M = 18.3 years, SD = 1.21; civic n = 17, ethnic n = 17). They took part in a practical class and received no compensation.
Experimental Materials
Participants watched one of two speeches presented in a video. Although filmed in a university seminar room with an experienced speaker (with an English accent) reading a specially prepared script, the cover story presented it as real. The camera focused on the speaker at a lectern but showed the heads and shoulders of an audience and so appeared entirely realistic. The speech concerned the Scottish environment and land access. In Phase 1 of the speech the speaker referred to the importance of the environment for Scottish identity (see Appendix B). In Phase 2 he made one of the identity claims used in Study 1.
After viewing the video, participants received a booklet containing an extract from an interview apparently conducted with the speaker they had just seen (in reality it was created by the experimenters). According to the cover story, the interview was conducted by one of the organizers of the conference where the speech was delivered. The interview was identical in both conditions and contained criticisms of Scottish people. Thus, in response to the question ‘how have Scottish people received your arguments so far, how have they reacted to your proposals?’ the speaker was reported as replying:
well there have been some positive reactions, but frankly mostly they were negative, which I .. um .. it .. reflects what you always find.. You know, to be honest .. we Scots can be quite conventional in the way we approach these things, you know, very inward looking. Most of the time not very flexible either, because we are .. um .. too bound up by Scottish traditions .. um .. very cautious. I believe Scots are too backward looking, you know, heritage can also be baggage. It can get in the way of building a modern Scotland, it can stop us thinking … I guess what I am trying to say is that it would be nice if we Scots could be a bit less hung up on our past and a bit more imaginative and a lot more ambitious, you know.
Dependent Measures
After reading the interview, participants completed the dependent measures (seven-point scales anchored 1 = strongly disagree and 7 = strongly agree).
Agreement with speaker’s criticisms
This was measured by seven questions (Appendix A) which formed a scale (α = .89) with a higher score indicating greater agreement.
Negative feelings
Participants’ feelings about the speaker’s comments in the interview were measured with nine items (Appendix A) which formed a scale (α = .95) with a higher score indicating greater negativity.
Constructiveness of the criticism
This was measured by eight items (Appendix A) which formed a scale (α = .83) with a higher score indicating the criticisms were judged constructive.
Behavioral reform
Support for behavioral reform amongst Scots was measured by three questions (α = .82) concerning the degree to which participants thought Scottish people should follow the advice offered in the critical comments (Appendix A).
The speaker’s Scottishness
As in Study 1, four items concerned the degree to which the speaker was seen as Scottish (α = .74).
Participants’ own Scottish identification
Again, this was measured by the same four items as in Study 1 (α = .76).
Results
As participants’ levels of Scottish identification were heavily skewed towards the upper end of the scale (mode = 7; Mdn = 6.75; M = 6.40, SD = 0.70; skewness = -1.88; kurtosis = 4.27) we again report two sets of analyses. The first is for the full sample (regardless of identification level). The second included only highly identified Scots (using the same criterion as in Study 1) for analysis. This reduced sample consisted of 29 Scots (ethnic n = 15, civic n = 14) with identification equally high in both conditions (civic M = 6.73, mode = 7.00, Mdn = 6.87; ethnic M = 6.60, mode = 7.00, Mdn = 6.75).
Evaluation of Speaker’s Comments
Using the full sample we found that (as predicted) agreement with the speaker’s criticisms was higher in the ethnic (M = 4.91, SD = 0.74) than civic condition (M = 3.76, SD = 1.11), t(27.7) = 3.55, p = .001, Cohen’s d = 1.22 (equal variances not assumed). Similarly, and as expected, negative feelings in relation to the criticisms were less prominent in the ethnic (M = 3.03, SD = 1.24) than civic condition (M = 4.05, SD = 1.50), t(32) = -2.15, p = .039, Cohen’s d = .74. Furthermore, reform was judged more necessary in the ethnic (M = 4.67, SD = 1.13) than civic condition (M = 3.53, SD = 1.34), t(32) = 2.72, p = .01, Cohen’s d = .94. However, there was no effect on the perceived constructiveness of the criticism (ethnic M = 5.21, SD = .89; civic M = 4.85, SD = .84), t(32) = 1.19, p = .24.
When the analysis was restricted to the higher identification sample, these findings were repeated. Again, agreement with the speaker’s criticisms was higher in the ethnic (M = 5.01, SD = 0.71) than civic condition (M = 3.64, SD = 1.15), t(21.46) = 3.81, p = .001, Cohen’s d = 1.43 (equal variances not assumed); negative feelings were less prominent in the ethnic (M = 2.92, SD = 1.25) than civic condition (M = 4.17, SD = 1.45), t(27) = -2.51, p = .018, Cohen’s d = .55; and reform was judged more necessary in the ethnic (M = 4.71, SD = 1.20) than civic condition (M = 3.45, SD = 1.42), t(27) = 2.58, p = .016, Cohen’s d = .96. As with the full sample, there was no effect on the perceived constructiveness of the criticism (ethnic M = 5.15, SD = .93; civic M = 4.91, SD = .86), t(27) = .72, p = .48.
Overall then (and regardless of the sample employed), on three of the four measures, criticism was received better when coming from the ethnic claimant. That is, the treatment of the claimant invoking ethnic criteria of belonging was better than the treatment of the claimant invoking civic criteria.
The Speaker’s Scottishness
Using the full sample we found no effect of condition on ratings of the claimant’s Scottishness (ethnic M = 3.69, SD = 1.20 vs. civic M = 3.15, SD = 1.08), t(32) = 1.39, p = .175. This was also the case when we focused on the higher identification sample: ethnic M = 3.58, SD = 1.23; civic M = 3.00, SD = 1.13, t(27) = 1.33, p = .19.
Discussion
As predicted, a speaker criticizing the Scottish ingroup received more positive reactions when they claimed national identity on the basis of place of birth and family heritage than on the basis of choice and commitment. The ethnic claim elicited more agreement, better emotional reactions, and greater endorsement of the need for behavioral change. The one dimension where our hypotheses were not confirmed concerns the perceived constructiveness of the criticisms. However, given the design to our study, the absence of a condition effect is understandable. The critical comments occur after participants had seen a film in which the interviewee had spoken enthusiastically about the need for imaginative measures to protect the Scottish environment. It is therefore quite understandable that both speakers would be judged as attempting to be constructive (the cell means on the constructiveness scale were above the midpoint, revealing the speakers to be judged as well-intentioned).
It is also noteworthy that although participants differentiated between the two claimants on these three measures, they did not do so when asked explicitly about the claimants’ Scottishness. This might reflect participants’ concerns about explicitly questioning another’s heartfelt declarations of identification. If so, this underlines the importance of having less obvious indices of inclusion (in this case, indices of agreement with the claimant’s criticisms). However, it might also reflect the fact that the questions came at the end of the study.
Although the findings on the three criticism-related measures (i.e., agreement, affect, and endorsement of the need for behavioral change) are as predicted, it is possible to question whether they really reflect the relative successes of these claims to community membership. Perhaps the negative reactions to criticism in the civic condition were driven more by dislike (associated with features of the speaker’s personal narrative) than by judgments of group membership per se. Fortunately, another established effect can be used to confirm our interpretation. This effect is the so-called black sheep effect (e.g., Abrams, Marques, Bown, & Henson, 2000) and it is useful here because it allows us to untangle whether the differential treatment of the claimants in Study 2 is really bound up with judgments of group membership or a more general dislike of the speaker in the civic condition. That is, it allows us to untangle ingroup treatment and positive treatment.
Study 3
Research shows that when it comes to pointing out shortcomings among group members, ingroup members are accorded more authority to speak than outgroup members (Hornsey et al., 2002). However, research also shows that when there is an agreed-upon ingroup norm, attitudes and behaviors that contradict that norm are judged more harshly if the deviant is a member of the ingroup rather than the outgroup (Abrams, Marques, et al., 2000). Indeed, it seems an ingroup member’s counter-normative position attracts more censure because it subverts the integrity of the group’s identity.
In the current context, this logic implies that a proposal discrepant with ingroup norms would elicit a harsher reaction when articulated by someone with a stronger claim to group membership. Thus, with the same identity claims used in Studies 1 and 2, it follows we could predict more negative reactions to a counter-normative position advocated by the person claiming national identity on ethnic as opposed to civic criteria. If confirmed, this would counter the idea that Study 2’s results simply show that certain claims are more attractive and result in better treatment for reasons unconnected with the judgments of group membership.
Study 3 used a film and script that started in exactly the same way as Study 2 (see Appendix B). As in Study 2, Phase 1 contained the speaker’s explanation of the importance of the environment to Scotland and Phase 2 contained the identity claim (ethnic/civic). In addition two further phases of the film were shown (see Appendix C). In Phase 3, the speaker argued for the need to protect the environment through responsible land access – a position which, based on existing evidence (e.g., McCrone, Morris, & Kiely, 1995) we expected to be highly normative for Scots. In Phase 4, the speaker argued for giving landowners the right to control access (which runs counter to Scottish traditions – see the Scottish Government’s Land Reform (Scotland) Act, 2003).i
Whilst watching the speech, participants provided real-time evaluations via a hand-held dial linked to a computer. This allowed fine-grained analyses that not only shed light on whether identity claims have an effect on the reception of the speaker’s message, but also, when they have an effect.
We expected no effect of condition in Phases 1 or 2 (as there is little to agree or disagree with). Nor did we expect an effect in Phase 3 where the text is so normative that we expected high levels of agreement regardless of the speaker’s identity. Our key predictions, then, concern what happens in Phase 4 and in the relationship between Phases 3 and 4. Assuming ethnic claims to Scottishness are stronger than civic claims, the logic of the black sheep effect predicts harsher Phase 4 judgments in the ethnic compared to the civic condition. In turn, it predicts the decline in evaluation from Phase 3 to Phase 4 to be greater in the ethnic than civic condition.
Method
Participants
Fifty-seven Scottish psychology undergraduates participated (males = 28, females = 29; age M = 26.10 years, SD = 9.3; civic n = 30 ethnic n = 27). Participants were recruited through invitation and received no compensation. Data were gathered in a laboratory.
Experimental Materials
The speech lasted just over seven minutes and had four Phases. Phases 1, 3 and 4 were constant across conditions. Phase 1 was identical to that in Study 2 (Appendix B). Phase 2 contained the identity claims (identical to those in Studies 1 and 2). In Phase 3 the speaker argued for responsible public access to the land in Scotland and in Phase 4 he argued for a policy option giving private landowners powers to limit access to the land (see Appendix C).
Throughout the course of the speech participants used a handset dial to record their evaluation. Turning the dial to the left indicated a negative evaluation and to the right a positive evaluation. The two extremes and the midpoint were marked on the dial. Piloting the use of the handset procedure showed people used the dial in two different ways. Some kept the dial at the midpoint and only responded when they agreed or disagreed with a particular aspect of the speech. When doing so they typically turned the dial to an extreme before quickly returning it to the midpoint. When graphed over the course of the speech these traces showed that for much of the time the dial remained at the midpoint and that points of agreement or disagreement were marked with sharp upward and downward ‘spikes’. Others used the dial to show their ongoing evaluation of the speech. When graphed these traces showed more gradual changes in evaluation. This latter usage was that envisaged in our design and we therefore guided our participants to use the dial in this manner. Following presentation of the speech, participants completed the questionnaire measures.
Dependent Measures
Participants’ real-time evaluations were recorded every .25 second in the form of numerical values ranging from zero (maximum disapproval) to 4000 (maximum approval) (2000 was the midpoint). Data were stored electronically. After the speech finished, participants completed a brief questionnaire including items concerning the speaker’s level of Scottishness (The speaker’s claim to be Scottish was very unconvincing; The speaker was very Scottish; r = .55) and their own level of Scottish identification (I see myself as being Scottish; I am pleased to be Scottish; I am a person who feels strong ties with the Scottish; I identify with other Scottish people; α = .83).
Preliminary Analyses
To check participants used the dial as instructed, we printed each participant’s trace and gave it to two independent judges who were blind as to the nature of the study. The judges were given the same instructions as the participants (when they were trained in the use of the dial) and were asked to identify any departing from these guidelines. Both judges identified one such case which was excluded from analysis. One further participant was excluded because he/she lived on a farm which could impact upon their evaluations for reasons unconnected with their Scottish identity. This resulted in a sample of 55 (civic n = 28, ethnic n = 27).
For statistical analysis we obtained mean levels of endorsement for each of the four phases of the speech. Specifically, we obtained mean levels of endorsement through averaging each individual’s dial-rating over the last 20 seconds of each phase (and then averaging these to obtain condition means). The logic to the selection of this 20 second period for statistical analysis was simple. Speeches unfold over time and it therefore takes a while for audiences to discern the shape, form and direction to a speaker’s argument. Moreover, our instructions on use of the dial asked participants to show their ongoing evaluation of the speech which means changes in the evaluation of the speech would emerge gradually. We therefore reasoned that selecting the last 20 seconds of each phase was appropriate as it would allow time for participants to evaluate the emerging argument.ii
Results
Again participants’ Scottish identification was heavily skewed towards the upper end (mode = 7, Mdn = 6.5, M = 6.34, SD = .70, skewness = -1.12, kurtosis = .80). Accordingly, we again report two sets of analyses. The first is for all, regardless of identification-level. The second is for the higher identifiers (as in Studies 1 and 2 we employed a criterion of over 5.5 on a 7-point scale). This reduced sample comprised forty-six Scots (ethnic n = 22, civic n = 24) with Scottish identification being equally high in both conditions (civic M = 6.70, mode = 7.00, Mdn = 6.87; ethnic M = 6.47, mode = 6.50, Mdn = 6.50).
Evaluation of Speech
The average evaluations of the speech during the last 20 seconds of each phase of the speech are presented by condition in Table 1. The upper panel (A) provides these data for the whole sample. The lower panel (B) provides these data for the more highly identifying sample. The most basic test of our prediction (a harsher response to the counter-normative response in the ethnic condition) is to compare the Phase 4 cell means. For the full sample this comparison showed no effect of condition (ethnic M = 1854, SD = 1273; civic M = 2313, SD = 1432), t(53) = 1.25, p = .215. However, for the sample that excluded lower identifiers we found a stronger trend for the ethnic claimant (M = 1682, SD = 1278) to receive a more negative evaluation than the civic claimant (M = 2322, SD = 1425), t(44) = 1.60, p = .06 one-tailed, Cohen’s d = .47). Accordingly, our subsequent analyses focused on this sample‘s data (panel B of Table 1).
Table 1
Study 3: Audience Response During the Last 20 Seconds of Each Phase
| Condition | Phase
|
||||
|---|---|---|---|---|---|
| 1. | 2. | 3. | 4. | ||
| A. All participants | |||||
| Civic (n = 28) | M | 3110 | 2532 | 3089 | 2313 |
| SD | 858 | 1067 | 954 | 1432 | |
| Ethnic (n = 27) | M | 3165 | 2717 | 2948 | 1854 |
| (SD) | 802 | 1034 | 849 | 1273 | |
| B. Higher identifiers only | |||||
| Civic (n = 22) | M | 3038 | 2442 | 2957 | 2322 |
| SD | 884 | 1091 | 966 | 1425 | |
| Ethnic (n = 24) | M | 3251 | 2836 | 3022 | 1682 |
| SD | 839 | 1064 | 803 | 1278 | |
We investigated the data in the lower panel (B) of Table 1 with analyses that focused on Phase 3 and 4 (corresponding to the normative and counter-normative phases respectively). We conducted two analyses. Our first analysis compared the evaluations in Phase 4 whilst taking into account the evaluations made by the same participants in Phase 3. This was done through using the Phase 3 evaluations as a covariate (which, as it allows us to take into account participants’ different usage of the dials, is statistically more powerful than simply comparing the Phase 4 cell means). When this was done we found the covariate had a significant effect, F(1, 43) = 19.25, p <.001, ηp2 = .31, as did condition, F(1, 43) = 4.26, p = .045, ηp2 = .09. That is, there was a more negative response in the ethnic condition (estimated marginal means and standard errors: ethnic M = 1653, SE = 243, civic M = 2348, SE = 233).
The second analysis compared the Phase 3 and Phase 4 cells in a 2 (Condition: ethnic, civic) x 2 (Phase: Phase 3, Phase 4) repeated-measures ANOVA. As should be apparent, as the speech moved from Phase 3 to Phase 4 (in which counter-normative proposals were espoused) we expected agreement with the speech to decline. However, following the logic of the black sheep effect we predicted a steeper decline in the ethnic than civic condition. The decline in agreement from Phase 3 to Phase 4 was apparent in both conditions (civic M Phase 3 = 2957, SD = 966; civic M Phase 4 = 2322, SD = 1425, t = 2.86, p = .009; ethnic M Phase 3 = 3022, SD = 803, ethnic M Phase 4 = 1682, SD 1278, t = 5.30, p = .001). However, what is most relevant here is our finding of a significant interaction, F (1, 44) = 4.42, p = .041, ηp2 = .091. Inspection of the means in the lower panel (B) of Table 1 shows this is because relative to Phase 3, evaluations in Phase 4 fell more steeply in the ethnic condition. This interpretation is supported from inspecting Figure 1 which presents the (mean) continuous evaluation of the speaker throughout the speech (higher identifiers only). For information, when these analyses were conducted with the full sample, none of the above effects were significant.
Figure 1
Study 3: Continuous (mean) evaluations of the speaker in the ethnic and civic conditions throughout the four phases of the speech (higher identifiers only).
The Speaker’s Scottishness
Judgments of the speaker’s claim to be Scottish did not vary according to condition for the full sample (M ethnic = 3.23, M civic = 3.02), t(51) = .50, p = .62. Nor did it differ for the more highly identified sample (M ethnic = 3.30, M civic = 2.80), t(43) = 1.06, p = .30.
Discussion
Whereas the logic to the intergroup sensitivity effect (used in Study 2) is that ingroup critics should receive a more positive reception than outgroup critics, there are some situations in which ingroup members receive a more negative response. Specifically, in a context where a position is established as normative and identity-defining, deviance elicits a more negative response when it comes from an ingroup rather than out-group member (the black sheep effect). Following this logic, in Study 3 we predicted more negative responses to the speaker’s proposals for land access when the speaker had a stronger claim to Scottishness. In support of this we found that for participants with a stronger national identification, the Phase 4 evaluations of the speaker’s proposals tended to be more negative in the ethnic than civic condition. More importantly, we found that the decline in agreement from Phase 3 to Phase 4 (corresponding to the transition from normative to counter-normative arguments) was steeper in the ethnic than in the civic condition. This confirms that different claims elicited different responses. Moreover, following the logic of the black-sheep effect these data confirm a greater acceptance of the ethnic over the civic claim.
More generally, it is appropriate to note these results complement existing research into the black sheep effect. Typically, investigations of group members’ reactions to ingroup deviance employ between-subjects designs in which the group membership of the target individual is manipulated. As we recorded participants’ reactions across both normative and counter-normative phases of a speech we have a repeated-measures design allowing us to document the temporal course to participants’ responses and this suggests that the effect really does emerge when ingroup members advocate identity-incongruent positions.
General Discussion
Our research addressed the issue of who is judged as a co-national. We were particularly interested in immigrants who identify strongly with their nation of destination and whether this identification allowed them to be heard as full members of the national community.
Our first study was simple and showed that a claim to national inclusion in terms of place of birth and ancestry was judged stronger than a claim based on subjective identification. Studies 2 and 3 showed that these different claims elicited different responses. Study 2 tested predictions derived from the literature concerning group members’ responses to criticism of their group (in which an ingroup member’s criticisms are better received than an outgroup member’s). We found that a claim invoking ancestry and place of birth resulted in a more sympathetic reception of the claimant’s criticisms than a claim invoking subjective identification. Specifically, the levels of agreement with the criticisms and endorsement of the need for behavioral change were greater when ancestry and place of birth were invoked. So too, the affective response to the criticisms was better when the claim to national inclusion was couched in more ethnic terms.
Study 3 complemented this finding through employing one of the few effects in which ingroup members are treated worse than outgroup members – the black sheep effect (Abrams, Marques, et al., 2000). This allowed us to show that the more positive response to the ethnic claimant in Study 2 reflected their categorization as ingroup (rather than something else – e.g., interpersonal liking). Amongst high identifiers, we again found that agreement with the claimant depended on the nature of the identity claim with the views of the identity claimant invoking ethnic heritage being judged more problematic than the views of the claimant invoking civic commitment. At first sight this might appear to imply that the civic claimant was advantaged. However, the dynamics to the black sheep effect reflect the idea that the degree to which one’s views matter is a function of one’s identity. When there is uncertainty about what the group thinks on a particular issue, ingroup members are more influential (Abrams, Wetherell, et al., 1990; Turner et al., 1987). However, when there is a consensus (such that an attitudinal position is normative), then ingroup members who deviate from that norm are judged more negatively, and again this is because they and their opinions count (Abrams et al., 2000). That is, counter-normative opinions can be tolerated when coming from outgroup members rather than ingroup members because the opinions of the former do not matter in the same way as those of the latter; it is only when counter-normative positions are adopted by ingroup members that we care enough to take them seriously, experience them as threatening ingroup identity, and respond accordingly (i.e., negatively). Thus, Study 3’s results (showing less negative judgment of the civic claimant’s proposals) underline the marginality of the civic claimant. Studies 2 and 3 (when we focus on higher identifiers) therefore provide clear, complementary and consistent support for the hypothesis that in the Scottish context, claims to national inclusion are more potent when couched in terms of place of birth and ancestry rather than choice and subjective identification, and that this impacts on the degree to which the identity claimant’s opinions are taken seriously.
At the outset to this paper we explained our motivation for this research was to begin to explore the degree to which different claims were accepted, and whether such acceptance was socially consequential. Given research shows that ethnic definitions of the ingroup can result in hostility to migrants (e.g., Pehrson, Vignoles, & Brown, 2009), the outcomes investigated here may seem trivial. However, work focusing on the commission of negative acts of discrimination needs complementing with studies of how individuals may miss out on the benefits of community membership (Wakefield et al., 2011). One of these benefits – of obvious significance for everyday citizenship – is of being heard as a community member and of having one’s opinions taken seriously. For high identifiers we found a coherent pattern in which a claim based on ancestry and place of birth was stronger and impacted upon the degree to which the claimant was taken seriously.
Limitations and Directions for Future Research
With regards to the limitations of our research, the restricted variation in participants’ own national identification meant we could not properly investigate whether higher levels of identification moderated the effects we found. Some research suggests high identifiers might be inclined to be more restrictive as to who they accept as ingroup (Castano, Yzerbyt, Bourguignon, & Seron, 2002; Meeus, Duriez, Vanbeselaere, & Boen, 2010). Yet, other work suggests that there may be no simple relationship between national identification and exclusionary attitudes. For example, Pehrson, Brown, and Zagefka (2009) found that the relationship between English people’s level of national identification and prejudice towards asylum seekers was moderated by the criteria participants employed in group definition: high identifiers were only more negative if they held more essentialist (i.e., ethnic) conceptions of Englishness. This implies that those identifying strongly with their nation can be identifying strongly with an inclusive conception of the criteria for belonging and that this could be addressed in future studies of how identification level impacts on the findings reported here.
It should also be clear that our analyses do not mean ethnic claims will always trump civic claims to group membership, whatever the nation. Nations differ in the degree to which they are defined in ethnic terms and such differences are consequential (Pehrson & Green, 2010; Pehrson, Vignoles, & Brown, 2009). Moreover, it is important to note that what we have termed ‘civic’ criteria can themselves involve a mix of elements and some of these may actually be quite restrictive. For example, Reijerse, Van Acker, Vanbeselaere, Phalet, and Duriez (2013) note that non-ethnic definitions of belonging which emphasize a common cultural tradition can be bases for exclusion rather than (as has often been assumed) bases for inclusion. Future research should therefore consider in more detail the different ways in which non-ethnic conceptions of the nation can have mixed effects.
More generally, future research concerning social identity processes could complement the emphasis on subjective self-definition as a group member (Turner et al., 1987) through exploring the experience of how one’s claims to identity are received. Indeed, there is a sense in which if we are to speak of a person as ‘having an identity’ it is important that others treat that person as having that identity (Jenkins, 1996) and social identity research could usefully consider how the experience of an identity is bound up with its social recognition by others (see Hopkins & Greenwood, 2013). With regard to what we have termed ‘everyday citizenship’, future research concerning the making and receiving of claims to national group membership should consider the acculturation attitudes of majority group and minority group members and how these are shaped by official (state) policies on citizenship (Bourhis, Moïse, Perreault, & Senécal, 1997). Indeed, much may be gained though considering the ways in which notions of citizenship are communicated and received in official programmes such as ‘citizenship tests’ (see Gray & Griffin, 2014). Building upon this, citizenship research could also consider minority group members’ experiences of how their attempts to be heard as co-nationals are received. Certainly, rejection can be painful. For example, British Muslims sometimes report feeling that their criticisms of British foreign policy in Afganistan, Iraq and Palestine are disregarded because these criticisms are heard as motivated by Muslim rather than British identity commitments (Hopkins & Blackwood, 2011). Obviously, this can be experienced as denying these individuals’ Britishness. In addition, as such reactions imply that Islam is an alien faith that is ‘out of place’ Muslims may also feel that their own understanding of their faith (as a universal religion capable of taking on a local hue) is denied (Hopkins, 2011).
Future research could also examine the everyday encounters in which identity claims are made. Sporting events and festivals (e.g., St Patrick’s Day) are obvious occasions for the claiming of national identity, and, despite the appearance of fun and care-free solidarity can be experienced by ethnic minorities as contexts where they are judged as outsiders (Pehrson, Stevenson, Muldoon, & Reicher, 2014). Other contexts include those where surveillance is routine (e.g., airports) and where the pain of misrecognition may be particularly high because it is perceived as carrying the weight of authority (e.g., the police; Blackwood, Hopkins, & Reicher, 2013a, 2013b).
Conclusion
The work reported here suggests that contemporary analyses of citizenship can benefit from exploring the factors impacting upon everyday identity claims-making. It is in the claiming of identities and in the reception of such claims that the right to be heard and have one’s views considered is adjudicated.