Positive Essentialization Reduces Prejudice : Reminding Participants of a Positive Human Nature Alleviates the Stigma of Indonesian Communist Party ( PKI ) Descent

This study aims to demonstrate and change negative perceptions of descendants of members of the Indonesian Communist Party (Partai Komunis Indonesia/PKI), a stigmatized social group in Indonesia. In Studies 1 and 2, participants were given positive descriptions of an adult (Study 1) and a child (Study 2), and were asked to evaluate them twice, before and after reading information about the target’s family background. In Study 1, targets were described either as descendants of PKI members, members of another Indonesian party or criminals. In Study 2, the target was presented as a descendant of PKI members, of members of another Indonesian party or without information on family background (control condition). The studies showed that whenever people were ‘revealed’ to be descendants of PKI members, the respondents’ judgments became more negative, and their assumptions about commonly shared views of these people became more negative as well. In Studies 3 and 4, participants were again given descriptions of an adult (Study 3) and a child (Study 4), which were both described as descendants of PKI members. Half of the participants were reminded afterwards with a written statement that every human is by nature good and unique (the experimental condition), while the other half did not get any additional information (control group). By making salient a shared and positively valued human ‘essence’, it was possible to alleviate the stigma that still is attached to PKI-affiliations in Indonesian society. We end the study with a discussion of our findings’ political and societal implications.

the respondents answered that they had learned about the 1965 events from the film, and most of them (87%) had watched "Pengkhianatan G30S/PKI" more than once.A recent assault on a group accused of holding a communist meeting (Sutari, 2017) and a recent rally against the resurrection of PKI (Bayhaqi, 2017) are evidence that everything related to PKI is still perceived as dangerous by a large part of the Indonesian public.
Taken together, communism is still perceived to be a major threat by large parts of the Indonesian population as a consequence of ensuing anti-communist propaganda.Hence, any ever so vague association with the PKI leads to a devaluation of a person in the eyes of most fellow Indonesians even in the absence of any form of deviant behavior.This phenomenon is known as stigmatization in social psychological research.

Stigma, Prejudice, and Discrimination
Stigmatization means that a person or a group of people are devalued by members of the wider society because they are perceived to differ from other members of the society in a certain regard.Stigmatization can be based, for example, upon visible differences between the stigmatized and the stigmatizing group such as skin color, upon deviating personal traits or behavior patterns in the stigmatized group, or upon mere membership in a stigmatized social group which is for some reason associated with undesirable traits and behaviors (Campbell & Deacon, 2006).In the words of Major and O'Brien (2005): "People who are stigmatized have (or are believed to have) an attribute that marks them as different and leads them to be devalued in the eye of others." (p. 395).In all cases, the stigmatized group is a social construct (Howarth, Nicholson, & Whitney, 2013) and stigmatization always occurs within a particular socio-historical context (Major & O'Brien, 2005).
In Indonesia, the stigmatization of PKI affiliates is in part based upon an ascribed inhumanity and atheism of PKI affiliates (Heryanto, 2013).This is often contrasted in Indonesia with the view that religious people are good people (Putra, 2016).As we have described in the introduction and in the section on the history of communism in Indonesia, the PKI stigma can impact the lives of people of various ages and in different settings in a very negative way.In our empirical studies we will focus on those expressions of prejudice that could occur nowadays in Indonesia (Putra, 2014;Putra & Wagner, 2017), such as a withdrawal of support for a political candidate or the devaluation of a student's scholarly achievements.In both cases, we assume that merely alleging PKI descent is enough to solicit prejudice towards a person.
Whenever stigmatization is based on mere group membership or visibly different characteristics of an individual -as is the case with Indonesian PKI descendants -questions arise as to how and why a person is stigmatized for having some sort of an invisible and intangible essence that sets him or her apart from other members of the society, and that apparently has been passed on to the person from his or her parents and/or ancestors, at least in the eyes of the stigmatizers.A belief in such essences underlying differences between social groups is known as psychological essentialism.

Psychological Essentialism
Projecting essence onto a social category, or essentialization, means to "think, talk, and act as if the category were a discrete natural kind and as if its members were all endowed with the same immutable attributes determined by the category's essence" (Wagner, Holtz, & Kashima, 2009, p. 21).Bastian and Haslam (2006) based their operationalization of essentialist beliefs on three characteristics: a) a belief in a biological basis of differences between members and non-members of the essentialized group, b) a belief in the discreteness of the two groups, and c) a belief in the informativeness of membership in an essentialized category in view of relevant human attributes such as traits and behavior patterns.Tawa (2017) also mentions a belief that differences between groups "are rooted in nature" (p. 1) and considers the assumption that the stereotyped characteristics are "innate/inherent" (p. 1) to the discriminated groups to be one of the the main characteristics of psychological essentialism.However, although the idea that people believe that things which look similar share the same set of hidden essential features (Medin & Ortony, 1989) is a widespread phenomenon in human cognition in general (Gelman, 2003), not all social categories are equally prone to essentialization processes.
Easily visible attributes such as skin color or sexual characteristics are apparently most prone to essentialist categorizations (Rangel & Keller, 2011;Wagner et al., 2010).Essentialist beliefs about race (Chao, Chen, Roisman, & Hong, 2007;Holtz & Wagner, 2009) and gender (Morton, Postmes, Haslam, & Hornsey, 2009) in particular have been found in empirical studies to be strongly related to stereotyping and prejudice (Haslam, Bastian, Bain, & Kashima, 2006).However, social categories based on similar ideology, socialization, parents' upbringing, peers, or social status can be subject to psychological essentialization as well.Wagner, Holtz, and Kashima (2009) discuss the examples of members of the US Marine Corps, who discursively undergo a physical transformation of their essence after becoming a marine (in the sense of rhetorically becoming irreversibly a different kind of human being; once a marine -always a marine) as well as discourses on hereditary nobility in Europe.In fact, even nonsensical category labels such as 'zav' can induce essentialist thinking among children (Heyman & Gelman, 2000); among adults as well, arbitrary and nonsensical category labels could be shown to affect inductive processes very much like the supposed 'natural kind categories' such as gender and ethnicity that are most often studied in research on psychological essentialism (Yamauchi & Yu, 2008).Taking into account that even a vague reference to family members' ties to the PKI can have detrimental consequences for Indonesians as pointed out beforehand, it can be assumed that PKI-affiliation is to some degree an essentialized category in so far as the stigma of PKIaffiliation can apparently be transmitted across generations in the same way that bodily characteristics are passed on from parents to their offspring.The essentialization of PKI-affiliations can hence be hypothesized to further contribute to stigmatization of and prejudice against supposed PKI-affiliates.
However, despite the ample evidence that psychological essentialism can lead to prejudice (for a summary, see Haslam, Bastian, Bain, & Kashima, 2006), there is also empirical evidence that under certain conditions essentialization can be a means of countering prejudice.For example, Haslam and Levy (2006) found that on the one hand, essentialist beliefs lead to the perception that homosexuality is an abnormality and that homosexuals and heterosexuals are indeed two separate categories; this way of thinking is associated with anti-gay attitudes.On the other hand, people with pro-gay attitudes frequently hold the essentialist belief that there is a normal, biological basis to homosexuality.Hence, essentialist discourse can be used both to justify anti-gay prejudice and to try to overcome it as well.
So far, social psychological research on hereditary stigma and essentialization has focused mostly on prejudiceprone social categories such as race or ethnicity.Whereas essentialist beliefs about these kinds of categories are often related to prejudice, is it possible for other forms of essentialization to have predominantly positive consequences?What would happen, for example, if 'human nature' itself were essentialized?Could this be a way to alleviate the potentially negative effects of the essentialism-based stigmatization of PKI descendants in Indonesia?

Human Nature and Attempts to Reduce the Negative Effects of Stigma
Over the last centuries, a number of philosophers have proposed the idea that human nature is 'essentially' good-an idea that could potentially be used to counter the stigma of negatively connotated social categories such as PKI-descendants in Indonesia.Philosopher Jean-Jacques Rousseau (1712-1778), for example, believed that man is good by nature, although this goodness can be contaminated through outside forces.Other philosophers such as Thomas Hobbes (1588-1679) held the opposite view, that human nature is evil: Man's natural desire is to acquire more and more power.In the field of psychology, B. F. Skinner and other behaviorists proposed that there are in fact no inborn human qualities: All human characteristics are acquired through reward and punishment during socialization and enculturation through cultural practices (see Waller, 2004).More currently, Staub (2007) has shared Rousseau's belief that humans are basically good.
While the discussion on human nature is ongoing, there is evidence that the belief in a good human nature can have positive effects.Studies conducted by McFarland and colleagues (e.g., McFarland & Brown, 2008;McFarland, Brown, & Webb, 2013;McFarland, Webb, & Brown, 2012) reiterated this view by showing that global humanitarian concerns and international altruism are related to identification with all of humanity.When people have a sense of belonging to the one human family, they will display "a deep caring for all human beings regardless of their race, religion, or nationality" (McFarland et al., 2013, p. 194).There is empirical evidence that when a common identity (i.e., being human) is highlighted, intergroup bias can be reduced or intergroup boundaries can be made less salient (e.g., Gaertner & Dovidio, 2000;Maoz, 2004;Rosenthal & Crisp, 2006).
In intergroup contact programs with Jewish and Arab participants, activities promoting mutual understanding, such as those which emphasize similarities on the personal level and reveal cultural and language commonalities, lead to a reduction of stereotypes and more positive intergroup attitudes (Maoz, 2012).Moghaddam (2012) argues that particularly in the age of globalization, the "omnicultural imperative" (p.304) should guide all kinds of intergroup situations: Instead of focusing on supposed differences between social groups, it is necessary to acknowledge first and foremost human commonalities (McFarland et al., 2012).
Our studies focus on the question as to whether reminding participants that they have a positive human nature in common can help to overcome the PKI stigma among Indonesian participants.It also may make sense here to differentiate between participants' own views towards a person and what they perceive to be the majority perception of the person.It is often assumed (at least in Western countries) that people hesitate to state that they personally are prejudiced.Because of such (presumed) social desirability concerns people often are asked about what others, or society as a whole, thinks about a group (see for example Fiske, Cuddy, Glick, & Xu, 2002, p. 884).Thus, in particular in Indonesia's very collectivist culture (e.g., Jetten, Postmes, & McAuliffe, 2002), such perceived majority views can be of importance insofar as they can in turn influence personal views as well.Elcheroth, Doise, and Reicher (2011) could demonstrate the importance of supposed knowledge of other people's perceptions of intergroup conflicts for the emergence of violent conflicts in Yugoslavia during the early 1990s even in regions where there actually had been only rare instances of intergroup conflicts in the past.Hence, assumptions about other people's attitudes and beliefs can be an important factor in understanding intergroup relations.Given that there is plenty of evidence of social exclusion of PKI members and their sympathizers in Indonesia, and given that most Indonesians who were born before 1995 have probably watched movies such as "Penumpasan Pengkhianatan G 30 S PKI", we assume that Indonesians are able to differentiate between a) their own judgment of supposed PKI affiliates and b) the majority's judgment of the PKI in Indonesia.

Overview of the Present Studies
Our studies addressed two main research questions in this Indonesian context.First, is 'communist' an essentialized social category in Indonesian society?If so, the mere fact of being the offspring of a PKI member should lead to stigmatization, even in cases where the person in question is not described as a PKI member or sympathizer and does not display any behavior or personality characteristics that are related to PKI affiliations.Second, can reminding participants of a common, intrinsically good human nature help to overcome this stigma?
The first two experiments examined the possible stigma of being related to a PKI-member for an adult (Study 1) and a child (Study 2).We wanted to see how detrimental the effect of PKI stigma is in general.Participants evaluated the target person twice (repeated design): in a first step we asked participants to evaluate a person that was described in predominantly positive terms; in a second step we asked participants to reevaluate the person after we provided in the experimental conditions information regarding the person's family background.In Study 1, target was described either as descendants of PKI members, of members of another Indonesian party or criminals.In Study 2, target was a descendant of PKI members, or of members of another Indonesian party or had no information on family background (control condition).Thus, in the control condition of Study 2, participants evaluated the target person a second time without additional information.
Study 1 examined how the fact that an expert in urban planning had a blood relationship with an accused PKImember affected both the participants' own personal judgments and their opinions about judgments of the 'average' Indonesian to support the nomination of this expert as a presidential candidate.Study 2 examined how the participants' judgments as well as an 'average' school teacher's judgments about having such a student were affected by the fact that an excellent primary school student had a blood relationship with an accused PKI-member.We expected that a supposed PKI-descent would lead to more negative judgments for the adult person as well as for the child (compared to other conditions).Because of the aforementioned social desirability concerns and in view of Indonesia's collectivist culture, we complimented participants' own views towards a person with what they perceive to be the majority's assessment of the person.
The next two experiments took the research one step further.In Studies 3 and 4 we tried to overcome the PKI stigma by reminding participants of a positive human nature and the fact that every human being is basically kind and unique.Participants were again given descriptions of an adult (Study 3) and a child (Study 4), which were both described as descendants of PKI members.Half of the participants were reminded afterwards with a written statement that every human is by nature good and unique (the experimental condition), while the other half did not get any additional information (control group).Both groups evaluated the described person twice (in the case of the experimental condition, before and after receiving additional information).
Study 3 used a design similar to Study 1 and Study 4 resembled Study 2. We expected that by explaining that every human is basically kind and unique, the negative effect of the PKI stigma could be reduced.

Hypotheses
As 'communist' appears to be an essentialized negative social category in Indonesian society, the mere fact of being the offspring of a PKI member should lead to prejudice (Hypothesis 1).This prejudice can take the form of participants' own judgments towards supposed PKI affiliates (H1a) as well as what they perceive to be the majority's judgment towards the PKI descent (H1b) in Indonesia.
We furthermore assume that reminding participants of a good and unique human nature would reduce the prejudice that comes along with the PKI stigma (Hypothesis 2).

Study 1
We conducted an experiment to address the question of whether PKI stigma would lead to prejudice in a political setting.We expected that when participants learned that the target person is of a PKI descent, they would appraise this person more negatively as a potential presidential candidate than they would evaluate a descendant of another political party or of criminals.We used the Indonesian National Party (PNI) for comparison.Both the PNI and PKI do not exist anymore, but they were among the five largest parties in the 1955 national election.

Method Participants
One hundred and ten university students from Jakarta and Bogor participated in this experiment.Two were eliminated for incomplete data and one was excluded for having a PKI family history.This left us with 33 males and 74 females between 18 and 44 years of age (M age = 22.51).Among the participants were 62 Muslims, 38 Christians (i.e., Catholics and Protestants), 2 Buddhists and 5 who did not report their religion i .

Procedure and Design
Participants were addressed by a research assistant in their classrooms either before or after class, asking them to participate in a brief paper-and-pencil study.The participants were informed that the study was about personal and societal perceptions.After reading the purpose of the study, participants were asked to sign a consent form.
Participants were randomly assigned to one of three groups: PKI group (N = 37), PNI group (N = 35), or Criminal group (N = 35).They were then asked to read the following vignette: There is a person who is an expert in public welfare, has good experience in area development, and has a great understanding of the condition of society in Indonesia.
After the participants read the paragraph, they were asked to rate in percentage points (0% = oppose, 100% = support) how supportive they would be of such a person and how supportive the average Indonesian would be in responding to the issue that the expert would be nominated as a presidential candidate (i.e.Time 1, or first rating in response to the vignette).
Next, participants read some information explaining the expert's family history; he was described as being either of PKI descent (PKI group), of PNI descent (PNI group), or of criminal descent (criminal group).We then asked again how supportive they would be and how supportive the average Indonesian would be regarding the expert's presidential candidacy (i.e.Time 2, or second rating after reading descriptive information).
Finally, participants were asked to provide some demographic information about themselves, including gender, age, ethnicity, institution, and whether they had anyone in their family who had been accused of being a PKI member.

Measures
We adapted a single item thermometer scale for prejudice to assess personal judgments and perceived majority judgments.Personal judgment was measured with the item: "How supportive would you be if the expert were nominated as a presidential candidate?"Answers ranged from 0% = oppose to 100% = support (M Time 1 = 73.27,SD = 15.59;M Time 2 = 54.07SD = 23.52).Perceived majority judgment was measured with the item: "How supportive would the average Indonesian be if the expert were nominated as a presidential candidate?"Answers for perceived majority judgments again ranged from 0% = oppose to 100% = support (M Time 1 = 72.34,SD = 13.88;M Time 2 = 48.04,SD = 19.81).There was a significant correlation (r Time 1 = .51,p < .001;r Time 2 = 54, p < .001) between personal and majority judgment.

Preliminary Analysis
The results of independent-sample t-tests revealed no significant gender differences in the dependent variables.Moreover, correlations between participants' age and the dependent variables were all non-significant, except for the correlation between age and Time 1 of perceived majority judgment (r = -.27,p = .004).In view of this inconsistent pattern, we did not use age or gender as control variables.

Main Analysis: Personal and Perceived Majority Judgment
To assess whether levels of personal and perceived majority judgment varied across the three groups, two separated split-plot ANOVAs (i.e., Repeated Measures ANOVA) for personal judgment and perceived majority judgment were computed, using the target group (PKI, PNI, and Criminal) as a between-subjects factor and either personal judgment or perceived majority judgment as a within-subjects factor (i.e.change in judgment from Time 1 to Time 2; repeated).We found a significant main effect of time on personal judgment (F(1, 104) = 67.31,p < .001η 2 p = .39)in the sense that the personal judgment scores decreased across the two time points and became more negative.We also found a significant effect (F(2, 104) = 4.14, p = .019,η 2 p = .07) of the between-subjects factor for the target group (i.e., PKI, PNI, and Criminal); for Time 1, we found no significant difference of the target group (F(2, 104) = .81,p = .449),whereas for Time 2, we found a significant difference (F(2, 104) = 4.59, p = .012):Tukey post hoc tests revealed a statistically significant difference for PKI vs. PNI group (p = .009),but not for PKI vs. Criminal group (p = .402)and PNI vs. Criminal group (p = .208)(please see the paragraph below for Ms and SEs of personal judgments under different conditions for Time 2).
However, there was no significant interaction between the 'target group" (i.e.PKI, PNI members, or a criminal) and the repeated factor (F(2, 104) = 2.14, p = .123,η 2 p = .04).Since the significance value indicated a trend, we still performed a post-hoc comparison.A Tukey post hoc test revealed a statistically significant difference for PKI vs. PNI group, (p = .013)but not for PKI vs. Criminal group (p = .366)or Criminal vs. PNI group (p = .296).These With regard to perceived majority judgment (see Figure 1), we found a significant main effect of the repeated factor (F(1, 104) = 158.59,p < .001,η 2 p = .69)in the sense that there was a decrease in the perceived majority judgment scores across the two time points (judgments became more negative).We also found a significant (F(2, 104) = 6.50, p = .002,η 2 p = .11)effect of the between-subjects factor target group (i.e., PKI, PNI, and Criminal); for Time 1, we did not find a significant difference of the target group (F(2, 104) = .88,p = .418),whereas in Time  We also found a significant interaction between the 'target group" (i.e.PKI, PNI members, or criminal) and partic-

Discussion
Information about PKI descent affected the participants' own judgments (i.e.personal judgment) and their view of the supposed average Indonesian's judgment (i.e.perceived majority judgment) of a potential presidential candidate.Being labeled as someone of criminal or of PKI descent both had a negative effect, but PKI descent had a significantly more negative effect than PNI descent.Hence, both, being of criminal descent and being of PKI descent, represent negatively laden and prejudice prone categories in Indonesian society.Still, whereas in the case of criminal descent implicit assumptions about social milieus and other social structures could easily be at play (Nye, 1976), the fact that a tie through 'blood' alone to a no longer existing political party is enough to create a stigma, regardless of any positive personal characteristics or behavior in the provided description, supports the assumption that in Indonesia, PKI labeling is an essentialized category (Wagner, Holtz, & Kashima, 2009).

Study 2
Study 1 indicated that PKI affiliation led to stigmatization.The goal of Experiment 2 was to test the very same effect in a primary school setting: How detrimental would the label of assumed PKI-descent be for a primary school student?Even though PKI labeling had a negative effect, no difference between PKI and the criminal descent group could be found in the split-plot analysis of Study 1. Apparently, PKI descent is not overly different from attachment to other very negatively laden categories (criminal), even though in the case of PKI descent, an 'inheritence' of negative traits (via 'blood' or via nurture) is arguably less plausible.In Study 2, we were interested in comparing the effects of PKI labeling with another more neutral control condition to further test whether the discrimination of PKI descent is in fact caused by stigmatization of PKI affiliation or if the effect can be explained in more general terms, for example, through the mere fact that people revised their previous judgment.We asked participants in Study 2 to rate their feelings of shame or pride with regard to having a student with or without PKI descent.In research on intergroup emotions (e.g., Fiske & Taylor, 2013), pride is commonly attributed to ingroup or reference groups.We expected hence that less pride is expressed towards supposed PKI affiliates in comparison to other groups.

Method Participants
One hundred and thirty university students from Jakarta and Depok participated in the study.Ten were eliminated for incomplete data.This left us with 40 male and 75 female participants (five did not report their gender) which were between 17 and 40 years old (M age = 19.74).Among the participants were 98 Muslims, ten Christians (i.e., Catholics and Protestants), four Hindus, two Buddhists, and six who did not report their religion.

Procedure and Design
The procedure was the same as in Experiment 1 with a few adjustments.Participants were randomly assigned to one of three groups: PKI group (n = 38), PNI group (n = 41), or Control group (n = 41).Afterwards, participants read the following vignette describing a school student: There is a primary school student who has many achievements and is always at the top of the class.The student often represents the school in competitions, and has received many awards.
Participants were then asked to rate in percentage points (0% = shame, 100% = proud) how proud they would be to become the student's teacher and how proud an average Indonesian teacher would be (Time 1, first-step response).In the next step, participants read information on the student's family history; the student was either of PKI descent (PKI group) or of PNI descent (PNI group).In the control group condition, participants were only asked to re-think their previous judgments.We then asked again how proud they would be to become the student's teacher and how proud an average teacher would be (Time 2, second-step response).
Finally, participants were asked to fill out some demographic information including gender, age, ethnicity, institution, and whether they were blood-related to anyone accused of being a PKI member.

Measures
Similar to Study 1, we used a single item thermometer scale to assess personal judgments and perceived majority judgments.Personal judgment was measured with the item: "How proud would you be to become the student's teacher?" Answers ranged from 0% = Shame to 100% = Proud (M Time 1 = 89.54,SD = 11.69;M Time 2 = 78.79SD = 20.03).We used shame-to-proud as a bipolar item because shame and proud are considered to be direct opposites of each other in research on emotions.Shame tends to be strongly associated with negative feelings and pride tends to be strongly associated with positive feelings (Tracy & Robins, 2007;Vujadinović, 2011).Perceived majority judgment was measured with the item: "How proud would the average teacher be …?".Just as for answers about one's personal judgment, answers for perceived majority judgment ranged from 0% = shame to 100% = proud (M Time 1 = 86.78,SD = 15.10;M Time 2 = 75.89,SD = 20.08).There was a significant correlation (r Time 1 = .47,p < .001;r Time 2 = .82,p < .001) between personal and majority judgment.

Preliminary Analysis
Independent-sample t-tests showed that there were significant gender differences at Time 1 for personal judgment (t(113) = 2.15, p = .034,M Male = 86.37 (SD = 14.42),M female = 91.20 (SD = 9.54)) and perceived majority judgment (t(113) = 2.40, p = .018,M Male = 82.17(SD = 16.85),M female = 89.12(SD = 13.51)).We found no significant differences at Time 2 for personal judgment' as well as perceived majority judgment.There were no significant correlations between age and the dependent variables.In view of this inconsistent pattern (gender affected only Time 1 of personal judgments and perceived majority judgment), we decided against using age and gender as control variables.

Main Analysis: Personal and Perceived Majority Judgment
To assess whether levels of personal and perceived majority judgment varied across the three groups and the two time points, again two separated split-plot ANOVAs for personal judgment and perceived majority judgment were computed, using as between-subjects factor the target group (PKI, PNI, or Control) and as within-subjects factor the change between Time 1 and Time 2 of either personal judgment or perceived majority judgment (i.e.change in judgment from Time 1 to Time 2; repeated).We found a significant main effect for change in the personal judgments (see Figure 2); (F 1, 117) = 44.46,p < .001,η 2 p = .27),showing a decrease in the personal judgment scores across the two time points.We also found a significant effect (F(2, 117) = 6.39,p = .002,η 2 p = .10)for the between-subjects factor target group (i.e., PKI, PNI, and Control); for Time 1, we found no significant difference between the target groups (F(2, 117) = .11,p = .897),whereas for Time 2, we found a significant difference (F(2, 117) = 10.94,p < .Subsequently, we used perceived majority judgment as the repeated factor (see Figure 2).Here again we found that the main effect of perceived majority judgment was significant (F(1, 116) = 43.85,p < .001,η 2 p = .27)showing a decrease in the perceived majority judgment scores across the two time points.There was a significant effect of the target group (F(2, 116) = 6.97, p = .001,η 2 p = .11);for Time 1, we found no significant differences between the target groups (F(2, 116) = .20,p = .817),whereas for Time 2, we found a significant difference (F(2, 117) = 19.199,p < .001):Tukey post hoc tests revealed statistically significant differences for PKI vs.Control group (p < .001),PKI vs. PNI group (p = .006),and for PNI vs.Control group (p = .029)(please see the paragraph below for Ms and SEs of majority judgments under different conditions for Time 2).
We also found a significant interaction of target group vs. the repeated factor (F(2, 116) = 18.47, p < .001,η 2 p = .24).A Tukey post hoc test showed a statistically significant difference for PKI vs. PNI group, (p = .042)and PKI In the next step, in contrast to Study 1, participants were randomly assigned to either the control group (N = 49) or experimental group (N = 50).In the control group, participants were told that the candidate was of PKI-descent.
In the experimental group, along with the information telling them that the candidate was of PKI descent, participants received additional statements reminding them that human nature is good and kind.

Control group:
The expert is of PKI descent.

Experimental group:
The expert is of PKI descent.
There is a view in society that every human is basically good.Moreover, every human is seen as always changing and unique, different from each other, even from the ones which are related through blood.This condition is considered natural.
After the experimental manipulation, we then asked again how supportive they would be of the candidate and how supportive the average Indonesian would be towards the person's presidential candidacy (Time 2).Finally, participants were asked to fill out some demographic information including their gender, age, ethnicity, institution, and whether they were blood-related to anyone accused of being a PKI member.

Preliminary Analysis
Using independent-sample t-tests, we found no significant gender differences for Time 1 of the dependent variables, and only a significant difference for Time 2 of perceived majority judgment (t(94) = 3.09, p = .003,M male = 26.72 (SD = 19.65),M female = 41.96(SD = 21.74)).Moreover, we only found a significant correlation between age and personal judgments at Time 1 (r = -.37,p < .001).Again, with these results we decided not to use age or gender as control variables.

Main Analysis: Personal and Perceived Majority Judgment
To assess whether levels of personal and perceived majority judgment varied across the three groups, again two separated split-plot ANOVAs for personal judgment and perceived majority judgment were computed, using the target group (PKI vs. PKI + Humanity info) as between-subjects factor and either personal judgment or perceived majority judgment as within-subjects factor (i.e.change in judgment from Time 1 to Time 2; repeated).

Discussion
This was our first attempt at testing whether reminding the participants that every human is kind and unique could have any effect on the PKI stigma.Study 1 showed that the stigma of PKI-descent had a negative effect on the participants' own judgments (i.e.personal judgment), as well as on their view of the supposed average Indonesian's judgments (i.e.perceived majority judgment) of an expert nominated as a presidential candidate.However, the negative effect of the PKI stigma was indeed alleviated (at least for the initially positive judgments) by reminding the participants of a positive shared human nature.7.95; M = 90.55,SD = 12.15, respectively) and negative in Time 2 (M = 65.92,SD = 20.02;M = 62.53,SD = 17.88, respectively).Moreover, there was a significant correlation (r Time 1 = .37,p < .001;r Time 2 = .57,p < .001) between personal and majority judgment.

Preliminary Analysis
In the independent-sample t-tests, except a significant difference for Time 1 of majority judgment (t(85) = 2.06, p = .043,M male = 83.64(SD = 25.80),M female = 91.55(SD = 8.51)), we found no significant gender differences for Time 1 and Time 2. We also found no significant correlations between age, personal judgments, and majority judgment in Time 1 and Time 2. With these results, we decided not to use age or gender as control variables.

Main Analysis: Personal and Perceived Majority Judgment
For personal judgment as repeated factor (see Figure 4), we found a significant main effect (F(1, 85) = 159.25,p < .001,η 2 p = .65)in the form of a general decrease in the personal judgment scores (judgments become more negative) across the two time points.We found no significant effect of the target group (F(1, 85) = .58,p = .449,η 2 p = .01).However, we found a significant interaction between the 'target group' and the repeated factor (F(1, 85) = 7.15, p = .009,η  For perceived majority judgment as dependent variable (see Figure 4), we again found a significant main effect (F(1, 85) = 175.34,p < .001,η 2 p = .67)in the form of a decrease in the perceived majority judgment scores across the two time points.Nonetheless, we found only a marginally significant effect of the target group (F(1, 85) = 2.87, p = .094,η 2 p = .03):for Time 1, we found no significant difference of the target group (t(85) = .01,p = .990),whereas for Time 2, we found a significant difference (t(85) = 2.22, p = .029),showing that the target was evalu-ated more positively in the PKI + Humanity info condition than in the PKI condition.We also found a marginally significant interaction (F(1, 85) = 3.92, p = .051,η 2 p = .04).This indicates that participants in the 'PKI + Humanity info group (M Time 1 = 90.53,SE = 1.86;M Time 2 = 66.74,SE = 2.67) showed a smaller decrease in perceived majority judgments compared to participants 'PKI' group (M Time 1 = 90.57,SE = 1.84;M Time 2 = 58.41,SE = 2.64).

Discussion
Study 4 showed as well that the PKI stigma can be reduced by reminding the participants that the human being is basically kind and unique.Here, the negative effect of a school student's PKI stigma on personal judgments and views of average judgments (i.e.perceived majority judgment) could be alleviated.

General Discussion
In the first two experiments, participants first read positive descriptions of an adult as a potential presidential candidate (Study 1) or a child who was described as a good student (Study 2).After learning that the target person was of PKI descent, their initially positive judgments became more negative.Moreover, they rated the average Indonesian's positive judgments of the described persons more negatively as well.
In Study 1, being of PKI-descent reduced positive personal judgments towards an adult as well as perceived majority judgments more strongly than affiliation with another political party (PNI).However, we found no significant difference between the negative effect of PKI affiliation and of being related to a criminal.It should be noted that labeling persons as criminal was found to be strongly related to stigmatization in a number of earlier studies as well (e.g., Nye, 1976).Hence, criminals as well as PKI-Descendants were seem to represent stigmatized and prejudice prone social categories in Indonesian society.
In Study 2, we dropped the criminal category and used "being of PNI descent" as a more neutral comparison group.The information that a student is of PKI descent affected the participants' perceived majority judgments towards this student significantly more negatively than affiliation with another political party (PNI); however, for the participants' personal judgments, we could not find a significant difference between PKI-and PNI-affiliation.
This could be the result of participants' self-censorship of their own prejudice in contrast to statements regarding prejudice of the supposed majority (e.g., Loury, 1994).In Study 2, we also added a more neutral control condition where participants reconsidered their initial judgments of the target person with no additional information on their background.Our results showed that that only the PKI condition significantly differed from the control condition in both personal and majority judgments.This suggests that learning about PKI descent is what drives increasingly negative judgments of participants at Time 2, rather than participants' general tendency toward negative evaluation.
Overall, all these findings show that PKI affiliation is a negatively laden stigmatized social category in Indonesian society.
In view of the fact that vaguely mentioning descent from a PKI affiliate could have a negative effect in absence of any suspicious behavior, we assume that to some degree PKI affiliation is an essentialized category (in the sense of Wagner et al., 2009) in Indonesian society insofar as the PKI stigma can be transmitted 'by blood' in an hereditary way like biological characteristics.Such an essentialization of a political category has rarely been documented in research on psychological essentialism, which instead has usually addressed categories such as ethnicity, race, or gender.Our studies show how a complex phenomenon such as political ideology, which is ob-tained and adopted by individuals as a consequence of a long educational and socialization process, can also be essentialized if essentializing discourse is widespread and constructed purposefully by societal forces.
The question is then how to tackle the negative effect of stigma?In Studies 3 and 4 we could show that whenever participants were reminded that human beings are by nature kind and unique, the effects of the PKI stigma could be alleviated.Two psychological processes may be at play here: On the one hand, positive framing of the human 'essence' could help to overcome negative attitudes toward a person which arise from a single characteristic feature, such as being of PKI-descent or not.On the other hand, thinking of 'humans' as a unified category in a positive and inclusive way could help to overcome the 'us' vs. 'them' thinking which is typical for stereotypes and prejudice.McFarland and Brown (2008) found that people tend to believe that identification with all humanity is an expression of a high degree of maturity and morality; identification with all of humanity was shown to have positive consequences in terms of more prosocial behavior and fewer discriminatory tendencies.We argue that reminding participants that the nature of human beings is basically good may also be a means of rehumanizing a devalued -and hence, dehumanized -outgroup (Major & O'Brien, 2005).
There have been many efforts by human rights activists in Indonesia to ask the government to apologize for what the accused PKI-members and their descendants have had to endure -so far to no avail.Even today, many PKI descendants are still excluded from society, banned from becoming teachers, and have difficulties getting a job altogether (Sukanta, 2011(Sukanta, , 2013)).We argue that reminding stakeholders of a good human nature could be used as a first step in an effort to deal with the reparation of the maltreatment of accused PKI members and their descendants.Our argument is supported by Moghaddam's (2012) omnicultural imperative: In order to build intergroup harmony, it is necessary to first and foremost acknowledge human similarities, before discussing differences.We assume that whenever the stigmatized group is seen as fellow human beings by the ingroup (McFarland et al., 2012), apologies for past maltreatment may become more likely.
Whereas we did not know for sure whether the exposure to a positive view of human nature would either affect personal judgments or perceptions of majority judgments (i.e.beliefs of average Indonesians), our findings show that bringing up the fact that every human being is innately good and unique seems to affect both.This finding lends hope to the conviction that by influencing the perceptions of individuals, the collective may change as well.
Our research contributes to the growing body of empirical studies on stigmatization and psychological essentialism (Howell, Weikum, & Dyck, 2011) by focusing on a specific case where affiliation with an opinion based group, the PKI, works as a hereditary stigma like racial or ethnic categories do in other contexts (e.g.Verkuyten, 2003;Young, Sanchez, & Wilton, 2013).Such processes have been described from a theoretical point of view (e.g., in Wagner, Holtz, & Kashima, 2009), but empirical studies on such cases of political or other societal categories becoming "essentialized" are still rare.Our studies also contribute to the growing body of empirical research on ways of attenuating stigmatization processes (Campbell & Deacon, 2006;Stathi, Tsantila, & Crisp, 2012).
Although we think that our study offers new insights, many questions remain.First, our study focused on the stigma of one particular group in Indonesia.Of course it would be a positive outcome to see our study being replicated with other stigmatized groups in different societies.For example, Islamophobia seems to be on the rise in most of the western world.In many non-Muslim countries, Muslims are regarded as threatening.In this case as well, designing reflective studies which would remind participants of human uniqueness and a basically good human nature might help to overcome the stigmatization of Muslims in non-Muslim countries.
Second, we used an experimental approach to test our assumptions and the effect sizes are small.The question remains as to whether the processes we created in the laboratory are relevant and applicable to real world contexts as well.It would be great to be able to complement our research with qualitative research into 'real' instances during which human nature and human uniqueness are brought up in discussions of discrimination and stigmatization.
Third, our experiments do not solve the question as to whether the same processes would apply to people who were themselves accused of PKI-membership, that is, whether the victims of PKI stigmatization would be more willing to accept apologies, and whether in turn the same processes could lead the Indonesian government to apologize for their maltreatment of them.However, reparation for historical wrongdoing as a means of moving ahead as a society (Brown, Zagefka, González, Manzi, & Čehajić, 2008) is undoubtedly important.
Fourth, we only used one-item-measures as dependent variables.However, although some findings were only marginally significant (e.g. in Study 4), we obtained rather consistent results in favor of our hypotheses across all four experiments.
Fifth, it must be taken into account that besides reminding the participants that the human nature is good, we also reminded them that human beings can change as well.Hence, we cannot know with certainty whether the participants were affected primarily by reminding them of the good and kind human nature or by emphasizing the malleability of human beings.Previous research on stigma reduction (Pennington, Campbell, Monk, & Heim, 2016) showed that asking participants to imagine positive social contact with stigmatized individuals, in this case mentally ill, could influence their attitudes towards mentally ill persons in a positive way as well.Bilewicz and Jaworska (2013) found that bringing up examples of Poles who helped Jews during the holocaust facilitated reconciliation in Polish-Jewish encounters.Hence, we still believe that the main factor behind our findings is that participants were reminded of similarities between them and the target persons and that these similarities were described in positive terms.However, further research is needed to differentiate the respective effects of the different reminders in reducing stigma.
In conclusion, this study suggests that there are ways to reduce negative feelings and perceptions toward a stigmatized group.We hope that our study not only opens up new perspectives into stigmatization and essentialization processes, but also contributes to the question as to how prejudice and discrimination resulting from stigmatization can be attenuated.

Notes
i) In Java island, where we collected our data, Islam is the largest religion (85%) whereas Christianity is second (10%).Around the years 1965-66, exclusions and executions of supposed PKI affiliates and their friends and relatives occurred in almost all Indonesian provinces.In Bali, where the majority of population is Hindu (80%), about 100.000 people were killed (Pratomo, 2015).In the Eastern part of Indonesia, where Christians are the majority, such exclusions and executions also happened.Moreover, in North Sumatra province in the Western part of Indonesia, extreme nationalists worked hand in hand with religious groups against supposed PKI affiliates.In view of the fact that all major religions in Indonesia played a role in the atrocities against supposed PKI affiliates, we decided not to use the religious background as a control variable.
results indicate that for participants in the PKI group (M Time 1 = 70.81,SE = 2.57; M Time 2 = 46.49,SE = 3.74), the experimental manipulation led to a stronger decrease in the level of personal judgment compared to participants in the PNI group (M Time 1 = 75.43,SE = 2.64; M Time 2 = 62.71,SE = 3.85), which was statistically significant.No significant difference was found either between the criminal group (M Time 1 = 73.71,SE = 2.64; M Time 2 = 53.43,SE = 3.85) vs. the PNI group, nor the criminal group vs. the PKI group.
2, we found a significant difference (F(2, 104) = 12.75, p < .001):Tukey post hoc tests revealed a statistically significant difference for PKI vs. PNI group (p < .001)and PNI vs. Criminal group (p < .001),but not for PKI vs. Criminal group (p = .973)(please see the paragraph below for Ms and SEs of majority judgments under different conditions for Time 2).

Figure 1 .
Figure 1.Mean estimates of positive judgments in Study 1.
001): Tukey post hoc tests revealed a statistically significant difference for PKI vs.Control group (p < .001)and PNI vs.Control group (p = .006),but not for PKI vs. PNI group (p = .318)(please see the paragraph below for Ms and SEs of personal judgments under different conditions for Time 2).

Figure 2 .
Figure 2. Mean estimates of positive judgments in Study 2.

Figure 3 .
Figure 3. Mean estimates of positive judgments in Study 3.

Figure 4 .
Figure 4. Mean estimates of positive judgments in Study 4.