This is an open access article distributed under the terms of the Creative Commons
Attribution License (“Poles, get out of Lithuania” (Lenkai lauk iš Lietuvos), was written on a wall of the market-square in Šalčininkai town (Lithuania) one morning in April 2015. The media used this incident as an example of majority Lithuanians provoking hostility towards the Lithuanian Polish minority. Online articles about this episode received hundreds of comments. Anonymous Lithuanian commenters called this incident a “provocation of Poles by themselves”, because “Lithuanians have always been tolerant” or “have better writing skills than this” and therefore wouldn’t make such an inscription (15min, 2015).
With such comments Lithuanians attributed essentialist features (inherited tolerance, and having better writing skills than Poles) to themselves. Did this essentialist argumentation by Lithuanians result from facing negative accusations about their group as provoking hostility? Would Lithuanians have self-essentialised if the inscription had read “Poles, stay with us in Lithuania”?
Strategic use of psychological essentialism has received growing attention in social psychological research (e.g. Morton & Postmes, 2009; Plante et al., 2015; Verkuyten, 2003). Until now, little consideration has been paid to its dynamics in relation to facing positive or negative traits of the ingroup. It is not far-fetched to assume that essentialist ingroup and outgroup attributions of majority and minority members will depend on how they feel about their own group's standing, after being confronted with positive or negative attributes of their group.
The purpose of this paper is to address four important issues pertaining to the potential mutability of essentialist group attributions: (1) how essentialist ingroup and outgroup attributions differ among minority and majority groups (Studies 1 and 2); (2) how essentialist attributions differ if participants are primed with positive, neutral or negative ingroup traits (Studies 1 and 2); (3) how such group trait priming is related to essentialist attributions of minority and majority group members (Studies 1 and 2); and (4) how these differences emerge in typical and numerically reversed minority-majority settings (Study 2).
Essentialist Group Attributions
In philosophy, the term essence has been used over thousands of years to signify permanent, unalterable and eternal substance of things, in contrast to changing and ephemeral objects of reality (Wagner, Holtz, & Kashima, 2009). In psychology, Rothbart and Taylor (1992) describe essentialism as the attribution of group membership to a shared invisible essence among all members of a particular group (see Yzerbyt, Judd, & Corneille, 2004 for an overview). In general, lay people tend to attribute different essences to social groups in order to explain their differences (Leyens et al., 2001). By doing this, they see these groups as natural and immutable, meaningful and informative. The group essence determines identity and underlying similarity of group members (Haslam, Rothschild, & Ernst, 2000, 2002), and differentiates members of one group from another (Demoulin, Leyens, & Yzerbyt, 2006). People may attribute causal and inductive powers of a group to biological aspects (for example, ‘blood’, ‘genes’, or ‘race’ (Zeromskyte & Wagner, 2017), or to the essence of culture (Buhagiar, Sammut, Rochira, & Salvatore, 2018; Phillips, 2010). The usage of cultural (instead of biological) essentialism seems to be related to the culturalization of racial discourse in recent decades (Buhagiar et al., 2018). Following this, we focus on psychological essentialism in this paper henceforth.
Why do people attribute essences to social groups? On the one hand, essentialist thinking has a cognitive function which helps in comprehending complex social experiences (Demoulin et al., 2006). On the other hand, essentialist lay beliefs also obtain a moral function justifying and legitimizing resolute thoughts, feelings and actions (Wagner et al., 2009), social inequalities (Yzerbyt, Rogier, & Fiske, 1998), and serving political purposes (Verkuyten, 2003). Altogether, psychological essentialism has important social consequences including intergroup bias, conflict, and misunderstanding (Prentice & Miller, 2007). Figuring out how and when people from different groups engage in essentialist thinking may lead to a better understanding of these phenomena.
Essentialist attributions have been studied in relation to various groups, such as sexual minorities (Morton & Postmes, 2009) and stigmatized groups (Plante et al., 2015), with most attention focusing on ethnic minorities and majorities (Bastian & Haslam, 2008; Buhagiar et al., 2018; Morton, Hornsey, & Postmes, 2009; Verkuyten, 2003; Verkuyten & Brug, 2004; Zeromskyte & Wagner, 2017).
Minority and Majority Members Apply Essentialism Differently
In society, most social groups are situated in an asymmetrical relationship such that majorities frequently have higher social status, better access to resources and greater control over institutions than minorities (Staerklé, Sidanius, Green, & Molina, 2005). The members of majority groups experience lower levels of identity threat than members of minorities (Koomen & Van Der Pligt, 2015). Simultaneously, majorities often face normative pressures to avoid being seen as prejudiced or discriminating against others (Dovidio & Gaertner, 1986). Seeking to deflect an accusation of xenophobia, members of majority groups may justify otherness of outgroups by referring to inherent—that is to say: essentialist—differences in culture and ethnicity (Morton et al., 2009; Verkuyten, 2003). Therefore, ethnic majorities are likely to attribute a higher level of essence to minority outgroups than to the majority ingroup (Verkuyten & Brug, 2004).
Typically, members of minority groups face the common obstacle of prejudice and discrimination and less access to essential services (Phinney & Rotheram, 1987). Minority status enhances salience of the ingroup’s identity and may lead to the experience of a greater sense of vulnerability (Schaller & Abeysinghe, 2006) and spontaneously arouse ingroup identification based on the distinctiveness of the ingroup (Leonardelli & Brewer, 2001). In doing so, group members may employ psychological essentialism as a cognitive tool to 'harden' their identity and ethnic self-image (Zeromskyte & Wagner, 2017). Therefore, ethnic minorities are particularly likely to essentialize themselves (Verkuyten & Brug, 2004).
In the present paper we extend results of earlier studies by originally analysing essentialist ingroup and outgroup attributions of majority and minority groups simultaneously. Furthermore, we do so in two different contexts, and in traditional and reversed minority-majority settings. Besides exploring these understudied factors, we also aim to replicate the pattern found by Verkuyten and Brug (2004), according to which members of minority groups attribute a higher level of ingroup- than outgroup essence, and members of majority groups attribute more outgroup- than ingroup essence (H1.1 in Study 1; H2.1 in Study 2).
Essentialist Attributions May Serve Micro-Political Ends
People (re)construct essentialist attributions about groups in a context dependent manner (Morton et al., 2009; Verkuyten, 2003). The use of essentialist group ascriptions seems to be related to the group’s position, which, far from being neutral, can be more or less threatened, and ‘shift’ according to the circumstances. For instance, in a study of group discussions, Verkuyten (2003) found that both minority and majority group members attributed different levels of essence in the conversation depending on the context. Therefore, essence attributions can be used both for oppressive and progressive aims, depending on the situation and minority/majority group membership.
A number of further studies support the idea that essentializing strategies may be adopted to defend against group identity threats (e.g., Morton & Postmes, 2009; Morton et al., 2009; Plante et al., 2015). For example, Morton and Postmes (2009) found that members of sexual minorities attributed more essence onto themselves when facing marginalization than when feeling discriminated against by the majority. Hence, discrimination may lead to lower levels of ingroup-attributed essence than identity denial. Similarly, research by Plante and colleagues (2015) showed that high identifiers of a stigmatized furry community are the most likely to endorse essentialist ingroup-attributions when compared to a highly similar group (anime), but not when compared to a dissimilar group (sport). Thus, essentialist beliefs about the ingroup may be raised as a response to distinctiveness threats.
Morton and colleagues (2009) also found that prejudiced white majority participants endorsed essentialist notions in the condition of exclusion of a minority member, but de-essentialized ethnic groups when their ingroup (white majority) member was excluded. According to this research, prejudiced majority members may use essentialist notions to exclude others, but not themselves. Thus, experience of ingroup exclusion may lead to lower levels of essentialization.
Together, the findings of Morton and Postmes (2009), Morton and colleagues (2009), and Plante and colleagues (2015) suggest that raising the level of ingroup-essentialization may prove effective as a response to threats to the group’s existence, and lowering the level of ingroup-essentialization may be effective to threats which devalue the group.
One more study elaborating on the variable use of essentialist attributions was conducted by Bastian and Haslam (2008, Study 2). These authors examined essentialist beliefs and negative bias towards immigrants among Australian-born majority participants in relation to exclusive versus inclusive national identity. The results showed that although essentialist beliefs were generally associated with negative bias towards immigrants, this effect was amplified when participants were primed with an exclusive national identity. According to this research, not only negative but also positive aspects of one’s identity may have a specific effect on essentialist group ascriptions.
Together, all the studies described above demonstrate that essentialist attributions can be strategically adopted in line with circumstantial conditions. People (re)construct essentialist self- and other-attributions based on a group’s image represented by ingroups and outgroups (Wagner, Raudsepp, Holtz, & Sen, 2017). Therefore, in order to be able to use essentialism strategically, people have to consider positive and negative aspects of one’s group.
Calling Attention to Positive and Negative Group Traits
When people highlight the similarities between themselves and other ingroup members, they apply typical ingroup characteristics to the self (Turner, Hogg, Oakes, Reicher, & Wetherell, 1987). Researchers can direct attention to a social identity and increase its salience by mentioning its name, which is called 'priming' in experimentation. “Neutral” priming, such as asking individuals to identify their ethnicity, increases responses and behaviors that are stereotypically associated with that group (Steele, 2010). Priming increases related concern for identity-based interests (McLeish & Oxoby, 2011), including the interests to achieve and maintain a positive self-concept (Tajfel, & Turner, 1979) and collective self-esteem (Luhtanen & Crocker, 1992). Consequently, one can expect that in a condition where a positive self-concept is threatened by highlighting negative attributes of a group, or confirmed by highlighting positive attributes of a group, people may seek to enhance their collective self-esteem and dynamically re-construct their essentialist attributions.
When individuals are presented negative aspects about their group, like for example when they encounter negative feedback about the ingroup or have to deal with a threat against the ingroup’s interest or image, the group-based threat prime increases the salience of one’s group identity and interests (Flippen, Hornstein, Siegal, & Weitzman, 1996; Steele, 2010). In an effort to create and maintain a positive self-concept, people develop protective mechanisms to avoid or protect themselves against such threats (Steele, 1988; Tajfel & Turner, 1986). They may reinterpret and reappraise group-related events to avoid negative feelings (Doosje, Branscombe, Spears, & Manstead, 1998) and emphasize a high degree of variability among group members and the behaviors exhibited by them (Doosje, Spears, & Koomen, 1995). Consequently, attributing variability among group members and reinterpreting formerly identity related attributes, means to reduce ingroup-attributed essence and stability (see also Morton & Postmes, 2009; Morton et al., 2009).
When people face positive aspects of their group, they confirm a positive view of themselves. As a consequence, they experience positive intragroup emotions (Muldoon, McNamara, Devine, Trew, & Dowds, 2010) and may show a higher level of group identification (Smith, Seger, & Mackie, 2007) which can be reflected in attitudes and behaviours. For example, being subtly exposed to a picture of a female leader improves the speech length of female participants, compared to displaying a picture of a male leader or no picture (Latu, Mast, Lammers, & Bombari, 2013). Hence, priming positive ingroup traits has implications for essentialist group attributions. However, to our knowledge no research has yet explored essentialist ingroup-attributions in relation to positive versus negative trait priming.
In the present manuscript, we seek to clarify the relation between positive/neutral/negative ingroup trait priming and essentialist group attributions for minority and majority group members. As discussed above, minorities are more vulnerable and flexible in their social standing and self-concept than majorities (Adams, Johnson, & Evans, 1998; Schaller & Abeysinghe, 2006). Moreover, minorities draw on essentialism as a means to assert their identity (Morton & Postmes, 2009). Not only minority, but also majority group members attribute lower levels of ingroup essence when facing negative treatment of their group (Morton et al., 2009).
We hypothesize that, when primed with negative ingroup traits, both minority and majority group members should lower the level of ingroup-attributed essence compared to neutral priming (see H1.2 in Study 1; H2.2 in Study 2). When primed with positive ingroup traits and hence feeling safe in their positive self-concept, minority members should tend to differentiate themselves from the majority outgroup and thereby raise the level of ingroup-essentalization compared to neutral and negative group trait priming. We also propose that members of the ethnic majority, being more stable in their social standing and self-concept, should not change their level of ingroup-attributed essence with positive trait priming (see H1.3 in Study 1; H2.3 in Study 2).
Essentialist Attributions in “Reversed” Minority-Majority Settings
In some contexts, minorities and majorities can have an equal or even inverted power differential. For example, an ethnic group can numerically constitute a local minority, but still be the majority within a broader geographic region. The opposite can be true for a minority: being a majority by number in a small region. In such a “reversed” setting, members of each group can justifiably feel outnumbered by another group, depending on the geographical frame of reference and perceived balance of power (Schaller & Abeysinghe, 2006). The local inverse position may be marked by feelings of competition and tension of shifting power relations (Zeromskyte & Wagner, 2017).
Accordingly, all members of the population may perceive themselves to be in a threatened minority group and express attitudes typical of minority members. For example, members of a group that was supposedly outnumbered in one frame of reference exhibited more demonizing, negative and more malevolent stereotypes of a competing outgroup than when not outnumbered (Schaller & Abeysinghe, 2006).
The local inverse position may be marked by different essentialist attributions. For example, in group discussions in the “reversed” setting in Lithuania (Zeromskyte & Wagner, 2017), both ethnic minority members and ambiguous majority members (which were locally outnumbered by the minority) exhibited a stronger tendency to essentialize the ingroup than members of the unambiguous majority. However, to our knowledge, no quantitative research has addressed this issue until now.
In Study 2 we analyse essentialist attributions in relation to positive/neutral/negative ingroup trait priming in a “reversed” minority-majority setting. In this context, both local minority/national majority and local majority/national minority groups have an ambiguous minority/majority status. Here we hypothesize that both these groups should demonstrate rather “mixed” tendencies in group essentialization and thereby both will attribute similar levels of ingroup- and outgroup essence (see H2.3 in Study 2). We also propose that both “reversed” minority/majority groups should be less stable in their self-concept, and strategically adapt their essentialist attitudes with positive/negative group trait priming like minorities, due to potentially perceiving threat and shifting identities from the other group. Hence, we hypothesize that both ambiguous minority/majority groups should attribute a higher level of ingroup essence with positive than with neutral and negative group trait priming. With negative group trait priming, both ambiguous minority/majority groups should attribute a lower level of ingroup essence than with neutral group trait priming (see H2.4 in Study 2).
Study 1: Austrian Majority vs. Turkish Minority in Austria
Setting
Turks and people with Turkish-origins, both called Turks henceforth, are the second biggest immigrant minority group (after Germans) in Austria, constituting 1.9% of the population (Birner, 2015). The first wave of Turks came to Austria during the 1960s and early 1970s, through labour migration (Sayari, 1986). Until now, Turks are perceived as the “new” and “problematic” minority, who differ from the Austrian majority not only by religion and culture, but also by their low educational level, low percentage of female-employees, high level of unemployment and above-average number of children (Birner, 2015). Thus, Turks and Austrians are situated in a traditional minority-majority asymmetry and the differences between the two groups can be expected to be reflected in their ingroup identifications (according to Staerklé et al., 2005).
Hypotheses
In this study, we expected that:
(H1.1) Turkish minority respondents will attribute a higher level of ingroup- than outgroup essence. Austrian majority respondents will show the reverse pattern.
(H1.2) Both majority Austrians and minority Turks will attribute a lower level of ingroup essence with negative than with neutral ingroup trait priming.
(H1.3) Only minority Turks will attribute a higher level of ingroup essence with positive than with neutral and negative ingroup trait priming.
Method
Sample
We recruited 169 Turkish and 137 Austrian respondents in Upper Austria during local events (for example, in fairs, cinemas, festivals) using snowball sampling. Participants were asked to complete a questionnaire about “personal attitudes” and were fully debriefed on completion of the study.
Participants who self-identified as “not Austrian” (n = 5) or “not Turkish” (n = 7) were excluded prior to analysis. We also excluded participants who incorrectly performed a priming task (for example, wrote “I always feel good about my group” when the task was to list example traits of their group which make them feel bad, or who wrote “we also have negative features”, when the task was to list example traits of their group which make them feel good), n = 19. The final sample of 122 Austrian participants comprised 65 men and 57 women who ranged between 18 and 61 years of age, M = 38, SD = 11.49. The final sample of 153 Turkish participants comprised 72 men and 81 women who ranged between 18 and 61 years of age, M = 30.1, SD = 9.05. Participants were nearly equally divided between the three experimental conditions (see Table 1).
Materials
The experiment involved ramdomly assigning participants to one of three groups within each sample. Participants were asked to think and write down one or more example situations that made them feel either positively or negatively regarding some aspects of their ethnic group. In the neutral condition they were asked to fill in a statistical test about their group (for Austrians, statistics of Austria as a country; for Turks, statistics about the Turkish minority in Austria).
Before running the study, we tested different priming methods (including reading quasi-scientific publications and filling in a sentence scrambled test) with a group of students. The most effective priming method was reading examples of situations and then listing self-created attributes, as used by Morton and Postmes, 2009.
In the positive and negative group trait priming conditions, participants were instructed to read example sentences of having positive/negative feelings about some attributes of their ethnic group. The example sentences for the Austrian and Turkish groups were chosen following the most popular stereotypes (Biffl & Rössl, 2014) and after discussing them with representatives of the respective groups. Seeking to increase emotional effect of trait priming, with positive group traits we highlighted that the group was respected by others (according to Muldoon et al., 2010). With negative group traits, we highlighted that the group was devaluated by others (according to Morton & Postmes, 2009).
In the positive group trait priming condition, the example sentences for Austrians were “I feel good as an Austrian when we are described as sociable; when we are called educated; when our politics are perceived as tolerant”. The example sentences for Turks were “I feel good as a Turk when we are described as polite; when people are fascinated that we know many languages; when our culture and religion are respected”.
In the negative group trait priming condition, the example sentences for Austrians were “I feel bad as an Austrian when we are described as hostile; when we are called narrow-minded; when our politics are perceived as xenophobic”. The example sentences for Turks were “I feel bad as a Turk when we are described as badly integrated; when people say that we cannot speak any language well enough; when our culture and religion are not respected”. Participants were then invited to list any examples they could think of that made them feel similarly.
The dependent measure was a 5-item scale of essentialist beliefs adapted from the items in the Essentialist Beliefs Scale (EBS; Haslam et al., 2000, 2002). This scale was designed to cover different non-biological attributes of essentialist beliefs. The items were “A [group name] will always be a [group name]”, “[Group name] tend to be similar to each other and have a lot in common”, “Knowing that someone is [group name] tells you a lot about them”, “A person is either [group name] or not [group name]: there is no ‘in between’”, “Despite individual differences, all [group name] are essentially the same”.
Items of the scale were randomly ordered and rated on a 6-point scale from 1 (strongly disagree) to 6 (strongly agree). For the analysis we calculated the sum of all 5 items, yielding a score from 6 to 30. First participants responded to the essentialism scale concerning the ingroup, then the outgroup. Scale reliabilities for ingroup essentialism (α = .79 for Austrians and .81 for Turks) and outgroup essentialism (α = .82 for Austrians and .84 for Turks) were acceptable. Factor analysis confirmed that these items loaded onto one factor for ingroup versus outgroup separately (all loadings > .60).
Procedure
Respondents were asked to participate in a survey about their personal opinions about different groups. The questionnaire was answered in the presence of the experimenter within 20 minutes maximum. At the end the participants were debriefed.
The Turkish participants were offered the questionnaire in German or in Turkish, so they could answer in the language they are most comfortable with (84 participants filled in the German questionnaire, and 69 chose the Turkish version). No reward or incentive was offered.
Results
Manipulation Check
Prior to the main analyses, we checked participants’ self-created attributes (ideas they wrote down in the positive and negative group trait priming conditions). Responses in Turkish were translated to German. At first, the researcher sorted responses within each condition into a smaller number of subcategories. A second coder independently applied this coding scheme, resulting in 82% agreement on the classification of responses. Discrepancies were resolved by discussion. In the positive condition, 95% of Austrians and 85% of Turks reported their own attributes. The attributes of Austrians focused on the positive feelings about Austrian nature, culture, economic performance, politics and achievements in sports. The attributes of Turks focused on positive feelings about Turkish culture, religion, bilingualism, the feeling of togetherness with other Turks, positive features and feeling respected.
In the negative condition, 88% of Austrians and 72% of Turks listed self-created attributes. The attributes of Austrians focused on negative feelings about Austrian politics (past or present), confusing or negatively comparing Austria with other countries, negative stereotypes or descriptions of Austrians as racist, feeling ashamed that some Austrians express racist attitudes. The examples of Turks focused on disregard of their culture or religion, being viewed as a threat, being discriminated against, being ashamed of some Austrian Turks not knowing the German language.
Some participants did not write down self-created attributes in the positive or negative group trait priming conditions and proceeded straight to the following page of the questionnaire. We chose to retain these cases because independent samples t-tests confirmed that there were no differences in essentialist attributions between participants who provided and did not provide self-created attributes (p = .87 ingroup essentialism of Austrians, p = .96 outgroup essentialism of Austrians, p = .61 ingroup essentialism of Turks, p = .95 outgroup essentialism of Turks).
The examples listed by participants in response to the manipulation suggest that participants indeed understood and followed the instructions.
In Table 1 we present descriptive statistics for all the groups and conditions under Study 1.
Table 1
Means and Standard Error Values of Ingroup and Outgroup Essentialism of Austrian and Turkish Participants in Different Priming Conditions, Study 1
| Priming | Dependent variable | M | SE | n |
|---|---|---|---|---|
| Austrian | ||||
| positive | 36 | |||
| Ingroup essence | 15.12 | 1.01 | ||
| Outgroup essence | 16.58 | 1.07 | ||
| neutral | 45 | |||
| Ingroup essence | 16.73 | 0.87 | ||
| Outgroup essence | 17.50 | 0.93 | ||
| negative | 41 | |||
| Ingroup essence | 13.98 | 0.92 | ||
| Outgroup essence | 15.65 | 0.97 | ||
| Turk | ||||
| positive | 46 | |||
| Ingroup essence | 22.00 | 1.09 | ||
| Outgroup essence | 17.21 | 1.16 | ||
| neutral | 54 | |||
| Ingroup essence | 21.53 | 0.92 | ||
| Outgroup essence | 20.53 | 0.97 | ||
| negative | 53 | |||
| Ingroup essence | 18.71 | 0.90 | ||
| Outgroup essence | 18.54 | 0.96 | ||
Minority vs. Majority Respondents
We conducted a 2*3*2 mixed MANOVA, with group status (minority vs. majority), trait priming (positive vs. neutral vs. negative) and target (ingroup vs. outgroup, repeated) as independent variables and the sum scores of all five essence attributions for ingroup and outgroup separately as the dependent variables.
Hypothesis H1.1 postulates an interaction between type of group (minority vs. majority) and type of target essence attribution (ingroup vs. outgroup). This interaction was statiscally significant, Pillais’ Trace = 0.11, F(1, 221) = 26.59, p < .001, η2 = 0.11.
In line with our predictions, a paired-samples t-test confirmed that minority respondents showed a higher level of ingroup, M = 20.66, SD = 6.47, than outgroup essence attribution, M = 19.3, SD = 6.95, t(108) = 2.99, p = .003, d = 0.2. Majority respondents tended to have a higher level of outgroup, M = 16.61, SD = 5.39, than ingroup essence attribution, M = 15.43, SD = 5.14, t(116) = -3.45, p = .001, d = 0.22. This interaction can be seen in Figure 1.
Figure 1
Interaction between ethnic group (Austrians vs. Turks) and ethnic target essence attribution (ingroup vs. outgroup).
Moreover there was a main effect of group status, F(1, 221) = 26.58, p < .001, η2 = 0.11. An independent-samples t-test confirmed that minority respondents' essence attributions were generally higher, M = 19.75, SD = 6.11, than the majority's, M = 15.97, SD = 4.87, t(221) = -5.16, p < .001, d = 0.69.
Priming Effects
We predicted that both majority and minority members would attribute a lower level of ingroup essence with negative than with neutral ingroup trait priming (H1.2), and only minority group members would attribute a higher level of ingroup essence with positive than with neutral and negative ingroup trait priming (H1.3).
The MANOVA yielded a significant double interaction, multivariate Pillais’ Trace = 0.04, F(2, 221) = 4.62, p = .01, η2 = 0.04, indicating that Austrians and Turks reacted to the priming task differently by changing their ingroup and outgroup essence attributions (Figure 2).
Confirming H1.2, Tukey post hoc tests revealed that minority Turks expressed a significantly lower level of ingroup essence with negative, M = 18.71, SD = 6.96, compared to neutral, M = 21.53, SD = 5.05, p = .041, trait priming. Majority Austrians also demonstrated a significantly lower level of ingroup essence with negative, M = 13.98, SD = 5.11, than neutral, M = 16.73, SD = 5.8, p = .03, trait priming.
Confirming H1.3, minority Turks demonstrated a significantly lower level of ingroup essence with negative, M = 18.71, SD = 6.96, compared to positive, M = 22.00, SD = 7.09, p = .049, trait priming. There was no statistically significant difference between negative, M = 13.98, SD = 5.11, and positive, M = 15.12, SD = 4.92, trait priming for majority Austrians (Figure 2).
However, essentialist ingroup-attributions of minority Turks did not significantly differ in positive, M = 22.00, SD = 7.09, and neutral, M = 21.53, SD = 5.05, group trait priming conditions. Thus, H1.3 was only partly confirmed.
For both majority and minority respondents, the level of outgroup essence attribution was not significantly different across priming conditions.
Figure 2
Degree of essentialist attribution by priming (positive vs neutral control vs negative), group status (Austrian majority vs. Turkish minority) and target (ingroup vs outgroup).
Note. Points with different subscripts are significantly different (p ≤ .05) within each group's set of ingroup and outgroup measures.
Discussion
The results of Study 1 demonstrate the existence of different levels of ingroup and outgroup attributed essence among majority and minority group members. As expected, Turkish minority respondents showed a higher level of ingroup than outgroup essence, and Austrian majority respondents demonstrated the reverse pattern. This is congruent with prior research (Verkuyten & Brug, 2004).
Also, as expected, not only minority Turks but also majority Austrians decreased the level of ingroup-essentialization with negative ingroup trait priming (compared to neutral). This indicates that facing negative aspects about one’s group leads both minority and majority members to lower levels of ingroup-attributed essence. Probably, when encountering negative feedback about their ingroups, both majority and minority participants were experiencing group-based identity threat and negative emotions (according to Muldoon et al., 2010; Steele, 2010). Attributing less essence to the ingroup may work as a strategy to counter negative aspects of the group (according to Morton et al., 2009) and maintain a positive self-concept (according to Tajfel & Turner, 1986).
Further, confirming our expectations, only minority members expressed more ingroup essence with positive than negative group trait priming. We believe this occurs because minorities may be more vulnerable and flexible in their social standing and self-concept than majorities (Adams at al., 1998; Schaller & Abeysinghe, 2006), therefore drawing on essentialism as a means to assert their identity (Morton & Postmes, 2009), especially when this identity is seen as positive.
For minority Turks, the level of ingroup-attributed essence did not differ between the positive and neutral group trait priming conditions. This can be explained by members of this “typical” minority group having very high levels of generic ingroup-essentialism even when attention is only directed to their ethnic identifications (according to Verkuyten & Brug, 2004). This would mean a ceiling effect with positive group trait priming.
The results of Study 1 support our idea about majorities and minorities having different tendencies regarding ingroup-essentialization, especially after facing positive, neutral and negative aspects of the ingroup.
The minority-majority context of this study was rather typical, with the majority having more power and control over resources than the minority group. In the second study, we seek to confirm and expand the results of Study 1 in a different setting and with different minority-majority groups.
Study 2: Lithuanians and Lithuanian Poles in “Typical” and “Reversed” Minority-Majority Settings
In Study 2 we tested whether the results of Study 1 would hold in a different ethnic minority-majority setting, where ethnic minority-majority groups have a different (long-term) relationship, and the power and control over resources is more evenly balanced or even reversed between minority and majority groups in the local context.
Setting
We chose Lithuania to study this topic, by analyzing Lithuanians (ethnic majority) and Lithuanian Poles (the biggest minority group constituting 6.6% of the population). Contrary to the characteristics of the groups analysed in the previous study, Lithuanian Poles have a similar socioeconomic status to ethnic Lithuanians: they have similar levels of education and income (Statistics Lithuania, 2013). Lithuanians and Poles are tied not only by the same Roman Catholic religion, but also by hundreds of years of common history, intertwined with conflicts, hatred and tension (Kazėnas, Jakubauskas, Gaižauskaitė, Kacevičius, & Visockaitė, 2014).
Poles are unevenly distributed in Lithuania. In most areas they constitute a “typical” local minority. However, in some rural towns and villages, Poles make up 80% of the inhabitants, making them the local majority and Lithuanians the local minority (Statistics Lithuania, 2013). In these places, local Governmental Councils are 70–88% ruled by Polish political parties (Lithuanian Chief Electoral Committee, 2015). Thus, in these areas, the power dynamics between the ethnic majority and minority are “reversed” and dealing with daily feelings of competition and tension of shifting power relations, members of groups in “reversed” positions may use essentialist attributions differently than groups in unambiguous minority-majority positions (Zeromskyte & Wagner, 2017). We test these hypotheses in the present study.
Hypotheses
In the typical minority-majority setting, we expected to replicate the results of Study 1, so that:
(H2.1) Polish minority respondents will attribute a higher level of ingroup- than outgroup essence. Lithuanian majority respondents will show the reverse pattern.
(H2.2) Both majority Lithuanians and minority Poles will attribute a lower level of ingroup essence with negative than with neutral ingroup trait priming.
(H2.3) Only minority Poles will attribute a higher level of ingroup essence with positive than with neutral and negative ingroup trait priming.
In the “reversed” minority-majority setting:
(H2.4) Both Polish and Lithuanian respondents will attribute similar levels of ingroup- and outgroup essence.
(H2.5) Both “reversed” Lithuanians and Poles will attribute a higher level of ingroup essence with positive than with neutral and negative group trait priming, and a lower level of ingroup essence with negative than with neutral group trait priming.
Method
Sample
We recruited 415 Lithuanian and 413 Polish respondents during local events (for example, in libraries and school events, fairs, town festivals) and using snowball sampling in 2 locations: in the capital Vilnius area, where Poles constitute a minority of 15-30 percent of the population, and in a “reversed” area, where Poles constitute a local majority of 75-90 percent of the total population. Participants who self-identified as “not Lithuanian” (n = 15) or “not Polish” (n = 20) or who incorrectly performed the tasks (n = 65), were excluded prior to analysis. The final sample of 370 Lithuanians comprised 96 men and 274 women who ranged between 18 and 55 years of age (M = 37, SD = 9.76). The final sample of 358 Poles comprised 93 men and 265 women within the age range of 18 to 59 (M = 36, SD = 10.86). The gender imbalance of participants originated from the character of the events where we recruited participants. They were more attractive for female than male participants. Despite this, we chose to continue our research because the situation was similar between the different groups under analysis. A chi-square test of homogeneity revealed that the distribution of gender groups was similar between the three experimental conditions, p = .34 for Lithuanians and p = .72 for Poles. Importantly, independent samples t-tests confirmed that there were no differences in essentialist ingroup and outgroup attributions between male and female participants (for both ethnic groups and in both settings). Participants were near equally divided between the three experimental conditions (see Table 2).
Material
The experiment involved three groups. Within each sample, participants were ramdomly assigned to one of three experimental conditions, in which they were asked to think of and write down one or more example situations that made them feel either positively (positive condition) or negatively (negative condition) regarding some aspects of their ethnic group. In the neutral condition they were asked to answer a statistical test about their group (for Lithuanians, statistics of Lithuania as a country; for Poles, statistics about the Polish minority in Lithuania).
We applied the same introduction and the same questionnaire as in Study 1, translated into the Lithuanian and Polish languages. We used the same questionnaire for all groups under analysis.
In the positive and negative group trait priming conditions, participants were offered example sentences to read about having positive/negative feelings about some attributes of their ethnic group. The example sentences for the Lithuanian and Polish groups were chosen according to the most popular stereotypes (Kazėnas et al., 2014) and after discussing them with representatives of each group. In the positive group trait priming condition, example sentences for Lithuanians were “I feel good as a Lithuanian when we are described as friendly; when we are called hardworking; when our politics are described as tolerant”. Example sentences for Poles were “I feel good as a Pole when we are described as educated; when people are fascinated that we know many languages; when our ethnic uniqueness is respected”. In the negative group trait priming condition, the example sentences for Lithuanians were “I feel bad as a Lithuanian when we are described as hostile; when we are called not tolerant; when our politics are labelled xenophobic”. The example sentences for Poles were “I feel bad as a Pole when we are described as uneducated; when people say that we cannot speak any language well enough; when our ethnic uniqueness is not respected”. Participants were then invited to list any examples they could think of that caused similar feelings.
We used the same dependent measure (scale of essentialist beliefs) as in Study 1. First, participants responded to the essentialism scale concerning the ingroup, then the outgroup. Scale reliabilities for ingroup essentialism (α = .83 for Lithuanians and .86 for Poles) and outgroup essentialism (α = .91 for Lithuanians and .86 for Poles) were acceptable. Factor analysis confirmed that these items loaded onto one factor for ingroup and outgroup separately (all loadings > .70).
Procedure
Respondents were asked to participate in a survey about their personal opinions about different groups. The questionnaire was answered in the presence of the experimenter and took 20 minutes maximum to complete. At the end, participants were fully debriefed. The Polish participants were offered the questionnaire in Lithuanian and in Polish, so they could complete it in the language they felt most comfortable with (168 participants filled in the Lithuanian questionnaire, and 216 chose the Polish version). No reward or incentive was offered.
Results
Manipulation Check
Prior to the main analyses, we checked participants’ self-created attributes (ideas they wrote down in the positive and negative group trait priming conditions). Responses in the Polish language were translated into Lithuanian. At first, the first author sorted responses within each condition into a smaller number of subcategories. A second coder independently applied this coding scheme, resulting in 81% agreement on the classification of responses. Disagreements were resolved by discussion.
In the positive condition, 74% of Lithuanians and 83% of Poles listed attributes. The attributes of Lithuanians focused on the positive feelings about Lithuanian nature, culture, positive character traits, politics and language. The attributes of Poles addressed positive feelings about their bilingualism, Polish culture and positive character traits.
In the negative condition, 75% of Lithuanians and 81% of Poles listed their own ingroup attributes. The attributes of Lithuanians focused on the negative Lithuanian image abroad, bad economic situation and negative character traits, such as being depressive and jealous. The attributes of Poles addressed disregard of their culture, being viewed as a threat, and being discriminated against by the majority.
Some participants did not write their own attributes. As in Study 1, we chose to retain these cases because removing them did not substantively alter the pattern of results. Moreover, independent samples t-tests confirmed that there were no differences in essentialist attributions between participants who provided and did not provide self-created attributes (p = .96 ingroup essentialism of Lithuanians, p = .81 outgroup essentialism of Lithuanians, p = .95 ingroup essentialism of Poles, p = .91 outgroup essentialism of Poles).
As in Study 1, the different nature of the examples listed by participants in response to the manipulation suggests that participants indeed understood and followed the instructions. In Table 2 we present descriptive statistics for all the groups and conditions under Study 2.
Table 2
Means and Standard Error Values of Ingroup and Outgroup Essentialism of Lithuanian and Polish Participants in Different Priming Conditions in “Typical” and “Reversed” Settings, Study 2
| Priming / Setting | Dependent variable | M | SE | n |
|---|---|---|---|---|
| Lithuanians | ||||
| positive | ||||
| “typical“ | 86 | |||
| Ingroup essence | 18.06 | 0.60 | ||
| Outgroup essence | 20.07 | 0.71 | ||
| “reversed“ | 31 | |||
| Ingroup essence | 21.74 | 1.00 | ||
| Outgroup essence | 19.26 | 1.18 | ||
| neutral | ||||
| “typical“ | 89 | |||
| Ingroup essence | 19.45 | 0.59 | ||
| Outgroup essence | 21.18 | 0.69 | ||
| “reversed“ | 33 | |||
| Ingroup essence | 21.52 | 0.97 | ||
| Outgroup essence | 21.12 | 1.14 | ||
| negative | ||||
| “typical“ | 94 | |||
| Ingroup essence | 16.68 | 0.57 | ||
| Outgroup essence | 20.76 | 0.68 | ||
| “reversed“ | 37 | |||
| Ingroup essence | 17.49 | 0.91 | ||
| Outgroup essence | 19.57 | 1.08 | ||
| Poles | ||||
| positive | ||||
| “typical“ | 78 | |||
| Ingroup essence | 22.67 | 0.63 | ||
| Outgroup essence | 22.22 | 0.74 | ||
| “reversed“ | 38 | |||
| Ingroup essence | 21.05 | 0.90 | ||
| Outgroup essence | 18.55 | 1.06 | ||
| neutral | ||||
| “typical“ | 82 | |||
| Ingroup essence | 20.63 | 0.61 | ||
| Outgroup essence | 19.27 | 0.72 | ||
| “reversed“ | 39 | |||
| Ingroup essence | 18.92 | 0.89 | ||
| Outgroup essence | 17.18 | 1.05 | ||
| negative | ||||
| “typical“ | 82 | |||
| Ingroup essence | 18.91 | 0.61 | ||
| Outgroup essence | 18.67 | 0.72 | ||
| “reversed“ | 39 | |||
| Ingroup essence | 17.03 | 0.89 | ||
| Outgroup essence | 21.33 | 1.05 | ||
Typical Lithuanian Majority vs. Polish Minority Setting
Seeking to confirm H2.1, H2.2 and H2.3 we computed statistical analysis only for respondents in the “typical” setting. We conducted a 2*3*2 mixed MANOVA, with group status (Polish minority vs. Lithuanian majority), priming (positive vs. neutral vs. negative) and target (ingroup vs. outgroup, repeated) as independent variables and the sum scores of all five essence attributions for ingroup and outgroup separately as the dependent variables.
Hypothesis H2.1 postulates an interaction between type of group (Polish minority vs. Lithuanian majority) and type of target essence attribution (ingroup vs. outgroup). This interaction was significant, Pillais’ Trace = 0.118, F(1, 505) = 37.86, p < .001, η2 = 0.118.
In line with our predictions, a paired-samples t-test confirmed that minority respondents showed a higher level of ingroup, M = 20.71, SD = 6.53, than outgroup essence attribution, M = 20.01, SD = 6.49, t(241) = 2.60, p = .01, d = 0.11. Majority respondents attributed a higher level of outgroup, M = 20.67, SD = 6.70, than ingroup essence, M = 18.04, SD = 5.11, t(268) = -8.78, p < .001, d = 0.44. This interaction can be seen in Figure 5.
There was also a main effect of group status, F(1, 505) = 4.16, p = .04, η2 = 0.008. An independent-samples t-test confirmed that minority respondents' attributions were averagely higher, M = 20.49, SD = 6.17, than the majority's, M = 19.37, SD = 5.43, t(505) = -1.99, p = .05, d = 0.19.
Priming Effects for Typical Majority-Minority Members
Further we predicted that both “typical” Lithuanians and Poles would attribute a lower level of ingroup essence with negative than with neutral ingroup trait priming (H2.2), and only minority Poles would attribute a higher level of ingroup essence with positive than with neutral and negative ingroup trait priming (H2.3).
The MANOVA yielded a significant interaction of target and priming, Pillais’ Trace = 0.025, F(2, 505) = 6.601, p = .001, η2 = 0.025, indicating that in the “typical” setting, Lithuanians and Poles reacted to group trait priming by changing their ingroup and outgroup essence attributions (Figure 3). There was a significant between subject interaction of group status and priming, F(2, 505) = 5.36, p = .005, η2 = 0.021, indicating that “typical” Lithuanians and Poles in the different experimental groups attributed different levels of essence correspondingly.
As expected in H2.2, Tukey post hoc tests revealed that majority Lithuanians demonstrated a significantly lower level of ingroup essence with negative, M = 16.68, SD = 4.31, compared to neutral, M = 19.45, SD = 5.15, p < .001, trait priming. Minority Poles also demonstrated a significantly lower level of ingroup essence with negative, M = 18.91, SD = 6.29, than neutral, M = 20.63, SD = 6.47, p = .04, trait priming.
As expected in H2.3, a Tukey post hoc test revealed that minority Poles also attributed significantly higher level of ingroup essence with positive, M = 22.67, SD = 6.38, than neutral, M = 20.63, SD = 6.47, p = .03, and negative, M = 18.91, SD = 6.29, p < .001, trait priming.
Additionally, Tukey post hoc tests confirmed that minority Poles attributed more outgroup essence with positive, M = 22.22, SD = 6.75, than neutral, M = 19.27, SD = 6.3, p = .017, and negative, M = 18.67, SD = 5.99, p = .002, trait priming.
Thus, confirming our expectations, in the negative priming condition (compared to neutral) both Lithuanians and Poles attributed significantly lower levels of ingroup essence. Moreover, Poles attributed higher levels of ingroup and outgroup essence with positive than negative (and neutral) priming (Figure 3).
Figure 3
Degree of essentialist attribution by priming (positive vs neutral control vs negative) by group status (Lithuanian majority vs. Polish minority) by target (ingroup vs outgroup).
Note. Points with different subscripts are significantly different (p ≤ .05) within each group's set of ingroup and outgroup measures.
“Reversed” Minority-Majority Setting
Seeking to confirm H2.4 and H2.5, we computed statistical analysis only for respondents in the “reversed” minority-majority setting. We conducted a 2*3*2 mixed MANOVA, with group status (Poles vs. Lithuanians), priming (positive vs. neutral vs. negative) and target (ingroup vs. outgroup, repeated) as independent variables and the sum scores of all five essence attributions for ingroup and outgroup separately as the dependent variables.
According to Hypothesis H2.4, we predicted that both “reversed” Poles and Lithuanians would attribute similar levels of ingroup and outgroup essence. As expected, the MANOVA yielded no significant interaction of target and group status, Pillais’ Trace = 0.001, F(1, 211) = 0.46, p = .50, η2 = 0.001, indicating that “reversed” Lithuanians and Poles indeed attributed ingroup and outgroup essentialism similarly (Lithuanians: ingroup essence M = 20.11, SD = 5.82; outgroup essence M = 19.98, SD = 7.05; Poles: ingroup essence M = 19.00, SD = 5.17; outgroup essence M = 19.03, SD = 6.3, see Figure 5).
Priming Effects for “Reversed” Majority-Minority Members
In Hypothesis H2.5 we predicted that both “reversed” Lithuanians and Poles would attribute a lower level of ingroup essence with negative than with neutral group trait priming, and a higher level of ingroup essence with positive than with neutral and negative group trait priming. The MANOVA yielded a significant interaction of target, priming and group status, multivariate Pillais’ Trace = 0.028, F(2, 211) = 3.082, p = .048, η2 = 0.028, indicating that ingroup and outgroup essence was differently attributed by “reversed” Lithuanians and Poles across priming conditions.
Partly confirming H2.5, Tukey post hoc tests revealed that “reversed” Lithuanians demonstrated a significantly lower level of ingroup essence with negative, M = 17.49, SD = 5.4, compared to neutral, M = 21.52, SD = 5.95, p = .017, and positive, M = 21.74, SD = 5.1, p = .001, trait priming. Differently than expected, the level of ingroup essence among “reversed” Lithuanians was not significantly higher with positive than neutral trait priming.
Further, confirming H2.5, Tukey post hoc tests revealed that “reversed” Poles demonstrated a significantly lower level of ingroup essence with negative, M = 17.03, SD = 5.21, compared to neutral, M = 18.92, SD = 4.73, p = .03, and positive, M = 21.05, SD = 4.85, p = .001, trait priming, and a significantly higher level of ingroup essence with positive compared to neutral trait priming, p = .02.
Additionally, Tukey post hoc tests revealed that “reversed” Poles attributed more outgroup essence with negative (M = 21.33, SD = 6.71) compared to neutral (M = 17.18, SD = 5.02, p = .001) trait priming (Figure 4).
Figure 4
Degree of essentialist attribution by priming (positive vs neutral vs negative) by group status (Lithuanians vs. Poles) by target (ingroup vs outgroup).
Note. Points with different subscripts are significantly different (p ≤ .05) within each group's set of ingroup and outgroup measures.
Further Effects
Given that the results of Study 2 partly confirmed our expectations, we decided to do an a posteriori analysis for all of the participants in Study 2 together. We conducted a 2*3*2*2 mixed MANOVA, with group status (Poles vs. Lithuanians), priming (positive vs. neutral vs. negative), setting (typical minority-majority vs. “reversed” minority-majority) and target (ingroup vs. outgroup, repeated) as independent variables and the sum scores of all five essence attributions for ingroup and outgroup separately as the dependent variables. The MANOVA yielded a significant interaction of target, group status (Polish minority vs. Lithuanian majority) and setting, multivariate Pillais’ Trace = 0.033, F(1, 716) = 24.09, p < .001, η2 = 0.033.
We compared the level of essentialism in two settings according to ethnicity. An independent-samples t-test confirmed that the “reversed” Lithuanians attributed a higher level of ingroup essence, M = 20.11, SD = 5.14, than the “typical” Lithuanians, M = 18.04, SD = 5.14, t(364) = -3.34, p = .001, d = 0.38. The “reversed” Poles attributed a lower level of ingroup essence, M = 19.00, SD = 5.19, than the “typical” minority Poles, M = 20.71, SD = 6.57, t(352) = 3.09, p = .002, d = 0.46. There were no differences for outgroup essence attributions (Figure 5).
We also compared the level of essentialism according to the local majority-minority status. An independent-samples t-test confirmed that the typical majority (“typical” Lithuanians) attributed a higher level of outgroup essence, M = 20.67, SD = 6.67, than the ambiguous majority (“reversed” Poles), M = 19.03, SD = 6.34, t(391)= 2.27, p = .024, d = 0.24. The ambiguous minority (“reversed” Lithuanians) expressed similar levels of essence as typical minority Poles (Figure 5).
Figure 5
Interaction between ethnic group (Lithuanians vs. Poles) in two settings and ethnic target essence attribution (ingroup vs. outgroup).
Furthermore, the MANOVA yielded a significant interaction of target, group status, setting and priming, multivariate Pillais’ Trace = 0.012, F(2, 716) = 4.33, p = .014, η2 = 0.012, indicating that Lithuanians and Poles in “typical” and “reversed” settings attributed different levels of ingroup and outgroup essence across priming conditions. Independent-samples t-tests confirmed that with positive priming, “typical” majority Lithuanians attributed less ingroup essence, M = 18.06, SD = 5.57, than “reversed” Lithuanians, M = 21.74, SD = 5.1, t(125) = -3.53, p < .001, d = 0.3. With positive priming, “typical” minority Poles attributed more outgroup essence, M = 22.22, SD = 6.75, than “reversed” Poles, M = 18.55, SD = 6.53, t(116) = 2.59, p = .011, d = 0.23. With neutral priming, “typical” minority Poles attributed more outgroup essence, M = 19.27, SD = 6.29, than “reversed” Poles, M = 17.18, SD = 5.02, t(122) = 1.98, p = .049, d = 0.18. With negative priming, “typical” minority Poles attributed more ingroup essence, M = 18.91, SD = 6.23, than “reversed” Poles, M = 17.03, SD = 5.21, t(127) = 1.99, p = .049, d = 0.17, and less outgroup essence, M = 18.67, SD = 5.99, than “reversed” Poles, M = 21.33, SD = 6.7, t(121) = -2.45, p = .016, d = 0.22.
Discussion
The results of Study 2 demonstrate that majority and minority groups in “typical” and “reversed” minority-majority settings showed different tendencies in group essentialization. As expected, in the “typical” minority-majority setting, Polish minority respondents showed a higher level of ingroup than outgroup essence, and the Lithuanian majority respondents demonstrated the reverse pattern. This is congruent with prior research (Verkuyten & Brug, 2004). Further, confirming our expectations, Poles and Lithuanians in the “reversed” setting did not confirm this pattern and attributed similar levels of ingroup and outgroup essence. This effect could be explained by the fact that majority and minority groups in “reversed” settings have ambiguous minority-majority status, which is reflected in similar levels of self- and outgroup essentialization (Schaller & Abeysinghe, 2006; Zeromskyte & Wagner, 2017).
Further, confirming our expectations, both “typical” and “reversed” minority and majority groups reacted similarly to negative group trait priming: all groups across settings lowered their ingroup-attributed essence after encountering negative aspects of their ingroup (compared to neutral). This signifies that negative feedback about the ingroup may lead to the reduction of ingroup-attributed essence, both among ambiguous and unambiguous minority and majority members. Facing a negative ingroup image may be threatening to one’s self-concept and self-esteem. Thus, lowering the level of ingroup-essence in such a threat situation may serve as a strategy to counter negative emotions (according to Muldoon et al., 2010; Steele, 2010), retain positive self-concept (according to Tajfel &Turner, 1986) and enhance collective self-esteem (according to Luhtanen & Crocker, 1992).
Study 2 also demonstrated different levels of attributed essence in reaction to positive ingroup trait priming among minority and majority groups. As expected, unambiguous (“typical”) majority Lithuanians did not change the level of ingroup-essence with positive ingroup trait priming. Further, all groups that could claim minority status (Poles in both settings and “reversed” Lithuanians) raised their level of ingroup-essence after encountering positive ingroup traits: both “typical” and “reversed” Poles increased the level of ingroup-essentialization with positive ingroup trait priming (compared to neutral and negative). Also, Lithuanian participants in the “reversed” setting demonstrated higher levels of ingroup-essence with positive compared to negative (but not neutral, which is different than expected) priming. These results support the idea that minority members (both ambiguous and unambiguous) are more flexible in their self-concept and attributions than unambiguous majority members (Schaller & Abeysinghe, 2006).
General Discussion
The present paper examines how ethnic majority and minority group members in two countries essentialize their ingroup and outgroups in positive, neutral and negative ingroup trait priming conditions.
In Study 1, we demonstrated that minority members express a higher level of ingroup than outgroup essence, and majority members attribute more outgroup than ingroup essence. We also established that both minority and majority members decreased the level of ingroup-essentialization with negative ingroup trait priming (compared to neutral), and only minority members expressed more ingroup essence with positive than negative group trait priming.
The results of Study 2 confirm and broaden the pattern uncovered in Study 1. Foremost, in both studies we demonstrated that the members of ethnic minorities essentialize ingroups more than outgroups, and majorities show the reversed pattern. This is congruent with prior research (Verkuyten & Brug, 2004) and this pattern could be explained the fact that both minority and majority members balance their self-image and the image of the outgroup in a way that maximizes their political standing and the value they are attributed in a society (Wagner et al., 2017).
Next, in both studies we revealed that group trait priming is related to the participants’ level of ingroup-essentialization. In both studies we demonstrated that both ethnic minority and ethnic majority participants decrease the level of ingroup-essentialization when encountering negative ingroup traits. This is congruent with the study of Morton and colleagues (2009), indicating that majority members also employ de-essentialization when faced with ingroup devaluation. This can be explained by negative attributes of the ingroup challenging the members' positive self-image, when the need to keep a positive self-concept (according to Tajfel & Turner, 1979) is gratified by framing the negative group's attributes as ephemeral and malleable, thereby expressing a lower level of ingroup-essentialization. Moreover, group-based threats are uniquely effective in activating self-awareness that puts one's own and one's group interests into focus (Duval & Wicklund, 1972). Such activation of ingroup-interest seems to work not only for minority but also for majority members, who are generally more stable in their self-concept and attitudes (according to Schaller & Abeysinghe, 2006). Being faced with negative aspects of one’s group generates a negative affect (Fejfar & Hoyle, 2000), which may motivate both majority and minority members to lower the perceived ingroup essence, and in this way distance the self from the negative qualities of the ingroup.
Furthermore, in both studies we demonstrated that only minority participants attribute more ingroup essence with positive than with negative group trait priming. Presumably, minority members who are faced with positive attributes about their ingroup and, hence, feel safe in their positive self-concept, are motivated to differentiate themselves from other groups (Tajfel & Turner, 1979), thereby raising the level of ingroup-essentialization. Accordingly, for minority members, the positive ideas about one’s ethnic group contribute to a higher level of ingroup-essentialization. Such an essentialized image of themselves and of their culture may help to counteract assimilation, protect group-identity, bolster self-esteem and lend support to the groups’ values (Levy, Chiu, & Hong, 2006).
As we expected, the members of ethnic majorities did not show the same effect with positive group trait priming. Such less sensitive reactions of majority members can be explained by experiencing lower levels of identity threat (Adams et al., 1998; Schaller & Abeysinghe, 2006) and thereby being “less adjustable” in the use of strategic essentialism.
We also had an expectation that ingroup-essentialization would be higher with positive than neutral group trait priming for minority members. However, this was only true for “typical” and “reversed” Lithuanian Poles (Study 2). Austrian Turks (Study 1) and “reversed” Lithuanians (Study 2) did not confirm this pattern. We think this is due to Austrian Turks and “reversed” Lithuanians attributing rather high levels of ingroup-essence with neutral ingroup priming, and therefore quickly producing a ceiling effect. For these groups, solely neutral priming (making their ethnic identification salient) probably brought some group-related responses and attitudes to the forefront (according to Steele, 2010) and automatically enhanced the level of ingroup-essentialization. This may reflect the rather challenging minority position of these groups. In Study 1, Austrian Turks represent a typical minority position, because they are lower educated and have lower incomes than majority Austrians (Birner, 2015). They attribute a higher level of ingroup-essence than majority group members. In Study 2, “reversed” Lithuanians experience daily feelings of competition (Zeromskyte & Wagner, 2017) and discrimination by local Poles (Pileckas, 2004) and also express high essentialist ingroup-attributions in a neutral priming condition. This may signify that when ethnic identities of minority members are constantly threatened, solely mentioning neutral facts about the group (e.g. distribution in the country and region) leads to higher levels of ingroup-attributed essence. In this case, ingroup essentialization can be employed as a means to achieve a positive self-image, especially when combined with ingroup favouritism and outgroup disparagement (Zeromskyte & Wagner, 2017).
In Study 2, Lithuanian Poles, on the other hand, have a similar level of education and income to majority Lithuanians (Statistics Lithuania, 2013). They are represented in Lithuanian parliament by the Lithuanian Polish party, and Local Governmental Councils are 70–88% ruled by Polish political parties in “reversed” regions (Lithuanian Chief Electoral Committee, 2015). Putting all this together, Lithuanian Poles have a somewhat more secure position as a minority. In a neutral context, they have a rather low level of essentialist ingroup-attributions, more similar to the majority, which gives more leeway for increasing ingroup-attributed essence with positive group trait priming (compared to neutral). The more insecure minority groups (“reversed” Lithuanians and Austrian Turks) do not demonstrate such an effect.
Our results support the theoretical account that positive or negative presentation of group attributes stimulates the strategic use of essentialist attitudes. Thus, the level of essentialist attributions can be “adjusted” in relation to representations and perceptions of the ingroup (Bastian & Haslam, 2008; Morton et al., 2009; Morton & Postmes, 2009; Verkuyten, 2003; Wagner et al., 2009), and minority members seem to be more susceptible to positive group trait priming than majority members.
Outgroup Essentialization
We had no expectation regarding the impact of the study's priming format (positive vs. negative sentences vs. neutral test) about the ingroup on the outgroup’s image and, indeed, there was no significant effect of the experimental manipulation on outgroup essentialization in Study 1. However, the results of Study 2 revealed that ethnic minority members (Poles) changed not only their ingroup but also their outgroup essence attributions across priming conditions. The typical minority Poles increased the level of outgroup essence with positive priming, compared to neutral and negative, while “reversed” Poles attributed more outgroup essence with negative priming, compared to neutral.
Why did the manipulation of positive/neutral/negative group traits have an additional effect on outgroup essentialist attributions for minority Poles only? One of the possible explanations is, again, their relatively secure ethnic minority situation. When the identity of minority Poles is confirmed (with the positive aspects of the ingroup), the sense of security may raise the level both of their ingroup- and outgroup essence, thus strengthening the positive distinction between them and the ethnic majority. The “reversed” Poles probably seek a similar distinction in the context of ingroup devaluation (when facing negative aspects of the ingroup) by essentializing and disparaging the outgroup (Zeromskyte & Wagner, 2017). Such a reaction to group trait priming may reflect their specific ambivalent situation, being the local majority and ethnic minority at the same time.
“Reversed” Setting
The numerically reversed minority-majority groups in the second study enriched our data with new patterns. Although we expected both “reversed” Lithuanians and Poles to apply essentialism like minorities do, this was only partly confirmed. The “reversed” minority Lithuanians attributed essence similarly to the typical minority Poles. They expressed a higher level of ingroup essence than the typical majority Lithuanians. From this point of view, they applied essentialism more like a minority (Verkuyten & Brug, 2004). However, there were no differences for outgroup essence attributions.
Similarly, the “reversed” Poles attributed a lower level of ingroup essence than Poles in typical settings where they constitute a minority. Thus, compared to typical minority Poles, they applied essentialism more like a majority (Verkuyten & Brug, 2004). However, again, there were no differences for outgroup essence attributions.
Additionally, compared to typical majority Lithuanians, the “reversed” Poles expressed a lower level of outgroup essence, thus applied essentialism more like a minority. Nevertheless, they reacted to group trait priming by changing the level of ingroup-attributed essence in the same way as the typical minority Poles.
Thus, both “reversed” Lithuanians and Poles expressed the levels of essentialism more according to their regional minority-majority status. When experiencing positive versus negative group trait priming, only “reversed” Poles changed the level of ingroup essence, in accordance with their national ethnic minority status. This illustrates how flexibly group-essentialization can be “shifted” when participants have plural identities, for example, simultaneously being the members of a majority and a minority. The latter result is partly consistent with previous research using focus group discussions of Lithuanians and Lithuanian Poles in “typical” and “reversed” minority-majority contexts (Zeromskyte & Wagner, 2017), where the “reversed” Lithuanians essentialized and favoured their ingroup more often than the “typical” majority Lithuanians. Such motivation of “reversed” Lithuanians to be essentialist about themselves may be either a resistance to assimilation, constructing a positive self-image (Levy et al., 2006) or responding to a perceived permanent political threat by the local majority (Corenblum & Stephan, 2001; Esses, Dovidio, Jackson, & Armstrong, 2001).
However, in the group discussions (Zeromskyte & Wagner, 2017), the “reversed” Poles expressed a rather high (minority-like) level of ingroup-essentialization. In the current study, “reversed” Poles expressed lower levels of ingroup-essentialization than the “typical” Poles. This difference may be explained by group discussions being not only neutral but also creating confirming or devaluing emotional contexts for participants. This difference may also be related to changes in the political atmosphere: in 2015, when the present research was conducted, there was less political debate and tension between Lithuanians and Poles than in 2011, when the group discussions of Zeromskyte and Wagner (2017) were conducted.
The “reversed” Poles were the only group to raise the level of outgroup essence with the negative group trait priming, compared to neutral. As demonstrated by Zeromskyte and Wagner (2017), “reversed” Poles tend to boast about their well-off arrangement in terms of political and financial security which, according to their opinion, contrasts with the Lithuanians’ situation. Presumably, when the negative aspects of being a Pole were brought to focus, the local majority Poles protected their sense of security by attributing less essence to themselves and more essence to the outgroup (Lithuanians), disparaging the outgroup, instead of linking the threat content to themselves.
Consequently, the specific situation of numerically reversed minority-majority groups can reveal different effects on ingroup and outgroup attitudes than those of the typical minority-majority groups. The members of the “reversed” groups shift between patterns associated with either the minority or majority identities in the most convenient way to validate themselves, not only preserving and underlining their ethnic and linguistic identity but also positioning themselves as equal to the majority (Schaller & Abeysinghe, 2006; Zeromskyte & Wagner, 2017).
The described differences between “reversed” and other settings highlight the importance of local context and local group positions in the research of intergroup attitudes. Majority–minority relationships are an increasingly relevant issue in modern societies given global migration patterns and recent history across the world. Understanding the relationship between self-identification processes and group essentialization in “shifting” contexts may help us understand the processes of formation of such attitudes.
Limitations and Further Directions
The current research has some limitations. One weakness is the potential order effect resulting from the design of our study. We assumed that priming ingroup traits should have more effects on ingroup than outgroup essence attributions. We decided not to include a counterbalanced schema, because our research design was already complex, due to the inclusion of minority-majority and regional groups. That is why after the priming task all our participants first completed the scale of ingroup essence and after that the scale of outgroup essence. However, as results of Lithuanian Poles showed, ingroup trait priming may also be related to different levels of outgroup essence. Future research should test if this effect is stronger when the level of participants’ outgroup essence is evaluated directly after the positive or negative group trait priming.
One more limitation of our research is the fact that the neutral group trait priming condition may have been not neutral for some participants: solely seeing statistics about their ingroup (e.g. population density and distribution) may have made the ingroup identity salient and automatically raise the level of ingroup-attributed essence. Future research should test if different types of information about the ingroup (compared to no information at all) result in different levels of essentialist attributions by different majority-minority groups.
Also, we provided different positive and negative types of group trait priming to majority and minority groups, and these differences may be reflected in the results. However, before running this study, we tested other priming methods and found separate priming schemas for minority and majority members to be most effective. Nonetheless, it remains to be seen how the present results would generalize to some universal positive/neutral/negative trait priming for minorities and majorities.
To our knowledge, our research is the first one comparing levels of ingroup-attributed essence in positive, neutral and negative group trait conditions for members of ethnic majorities and minorities simultaneously. Our research groups were rather small and not perfectly balanced in every priming condition. Future research should test the generalizability of the current results to a variety of minority-majority constellations.
Conclusion
Our findings demonstrate that encountering positive, neutral or negative ingroup traits is related to different essentialist attributions among majority and minority members in Austria and Lithuania. This represents an important addition to the available research in this field.
When exploring the mutability of psychological essentialism, attention should be paid not only to negative, but also to positive aspects of group identity. We found evidence that with negative group trait priming, both minority and majority members lower their essentialist ingroup attributions and that with positive group trait priming, the members of ethnic minorities heighten their ingroup-attributed essence. In “reversed” settings the level of group essence may shift even more flexibly, according to the shifting minority-majority status.
According to our results, we have formulated an answer to the question stated at the beginning of this paper. Would Lithuanians have argued in an essentialist way if the inscription on the wall in Šalčininkai had read “Poles, stay with us in Lithuania”? Considering the results of this paper, our answer is rather “no”: the positive group trait condition (“we are tolerant if we write such inscripts”) would rather have no effect on the level of ingroup-attributed essence of majority Lithuanians.