This Ain’t No Place for No Hero: Prevalence and Correlates of Representations of Victims,Helpers,and Perpetrators During the Time of National Socialism in German Families

Abstract

The transmission of national history in general and family narratives in particular is prone to censorship and bias, protecting or enhancing social identities. The authors propose that, as has been shown for national groups, families also create and pass on representations about their roles and behaviors through history. In a representative survey, 1000 German respondents estimated the percentages of victims, perpetrators, and those who helped potential victims during the time of National Socialism to be 35%, 34%, and 16%, respectively. For family representations, the percentages shifted toward helping (29%) and away from complicity (20%), while representations of victimhood were as prevalent (36%) as estimates for the general population. Systematic differences suggested an alignment of general social representations of history with family representations. Participants reporting a perpetrator family representation held more positive attitudes toward refugees coming to Germany today than participants who did not report such a representation. This link was mediated through differences in societal representations. The authors discuss family representations as an intermediate, more proximate prescriptive background and points of reference, according to which more general historical representations on a national level may be aligned, and individual present-day political attitudes and behaviors oriented.

Keywords

collective memories family representations historical representations in-group serving bias transgenerational remembrance political attitudes

Waitress: “Is this a story or did this really happen?”

The Doctor: “Every story ever told really happened. Stories are where memories go when they’re forgotten.”

(Doctor Who, Series 9, Episode 12, Moffat & Talalay (2015))

Introduction

Personal memories tend to be highly subjective. They will always be shaped by what we paid attention to in the very moment we experienced what we remember. We cannot personally remember what we did not perceive in the first place. Memories may also be adapted based on information we obtain afterwards and with whom we share our memories. For some audiences, we might leave out some details while taking the liberty of adding others. And we may do so to make the memories we share sound more appealing—or to make ourselves look better in the eyes of our audience. Of course, all of this is not meant to challenge the fact that there is an objective reality or historical truth to be acknowledged when remembering world history. Neither is it our aim to suggest that we cannot form memories about events that we did not witness in person. Memories, however, as The Doctor suitably describes in the brief dialogue above, quickly become stories. They turn out to be essentially social as we share them with others (see Moscovici, 1981) and particularly prone to bias as we are motivated to make ourselves look good, or less bad (see Baumeister, 1982, 1986). All these features of personal memories seem intuitive—why should they not apply similarly to collective memories transmitted in groups such as families and even nations?

Historical knowledge about socially relevant groups can be a burden, especially if such knowledge implies horrible atrocities involving those close to us (Doosje et al., 1998; Rees et al., 2013; Welzer et al., 2002). In the current article, we argue that, from a social-psychological perspective, the transmission of national history in general and representations of history within families (i.e., family representations) in particular should be prone to censorship and bias, protecting or enhancing our social identities (Tajfel & Turner, 1979). We propose that, as has been theorized and shown for national groups (Liu & Hilton, 2005; Liu & László, 2007), families also create and pass on representations about their roles and behaviors through history, and that individual family members will align their general historical representations and these family representations in a systematic way. Family representations may therefore further bridge the gap between the structure and consequences of social representations of history on a national level and individual motivations to remember some aspects and forget others (see Bilewicz et al., 2019; Klar & Bilewicz, 2017).

Our study is set in the context of German collective memory of the era of National Socialism—family representations of victims and heroes (Welzer et al., 2002). However, general psychological processes such as self- and group-serving bias, self-enhancement, and cognitive consistency are well established and have been widely studied in other contexts. In the first part of the article, we therefore briefly review the general theoretical framework in which the current research is placed. It is one that was proposed and developed by Liu and his colleagues (e.g., Hilton & Liu, 2017; Liu, 1998; Liu & László, 2007; Liu & Liu, 1997; Liu et al., 1999), where “general psychological processes are examined within the context of a specific and temporal case study” (Liu et al., 1999, p. 1022). Historical facts merely provide content that will be the object of operation of these general processes. The resulting stories will certainly not be accurate descriptions of historical facts but biased in a predictable fashion (e.g., Bilewicz et al., 2019).

From Social Representations to Family Representations of History

In developing Durkheim’s notion of individual and collective representations, Moscovici (1961) first introduced the concept of social representations. A social representation is “a system of values, ideas and practices” (Moscovici, 1973, p. xiii). These systems relate to social or cultural objects such as persons or events that are anchored within familiar systems in order to help individuals orientate themselves in a complex world and to provide shared means of communication (e.g., Sammut & Howarth, 2014).

Liu and Hilton (2005) famously introduced the concept of social representations of history into the social-psychological study of collective memories. The general aim of this approach is to “understand how groups’ representations of their history influence their collective attitudes and behaviour” (Hilton & Liu, 2017, p. 298). Social representations contain descriptive components such as historical events that are usually regarded as important across and within countries (Liu et al., 2005), but may also be contested by subgroups (Moscovici, 1988). Social representations, however, are also prescriptive in helping group members define who they are and how they ought to behave when faced with present-day political decisions (Bilewicz et al., 2019; Figueiredo et al., 2017).

The interpretation of specific historical facts, in other words, will differ across groups and even between individuals as a function of not only prior knowledge, but also social identification and motivation to remember certain aspects and not others (Tajfel & Turner, 1979; Wishart, 1997). “Motivated histories” should therefore exist not only on the national level (Bilewicz et al., 2019; Liu & Hilton, 2005), but also—and probably even more so—on the family level. As this particular social group will be more relevant to most, motivation to (intentionally or unintentionally) distort such family representations of history may be even more pronounced than for social representations on the national level. Family representations are usually also less public and constrained than historical representations on the national level, which are strongly shaped by official or hegemonic social representations, the state, or interstate relationships (see Brown et al., 2009; Giner-Sorolla et al., in press). At the same time, individuals will strive for cognitive consistency (Gawronski & Strack, 2012). Both representations should therefore not be independent of each other, but aligned to form a consistent whole. It seems plausible that representations on both levels—more general social representations as well as more personal family representations—should influence each other in a bidirectional fashion: we may grow up with a certain representation of our family’s role through history and align our general social representations accordingly. However, family representations may be challenged when faced with an incongruent broader social representation, for example, in the form of historical facts taught in school. We argue that individuals will have difficulties with incongruent family representations, on the one hand, and more general social representations, on the other. This cognitive dissonance (Festinger, 1957) may be resolved by challenging either the wider social representation or the family representation. In any case, the result should be an alignment of family and social representations of history.

Paralleling more general social representations, family representations of history as we conceptualize them contain descriptive components about family members’ roles and behaviors through history (Hilton & Liu, 2017; Liu & Hilton, 2005). Just as social representations, family representations hence offer a frame of reference or structure “through a process of selective interpretation, biased attribution, … and by privileging certain historically warranted social categories and category systems above other alternatives” (Liu & László, 2007, p. 87). However, instead of national groups, family representations of history refer to the family and therefore contain family members’ roles and behaviors, and the “temporal case study” (Liu et al., 1999) will be somewhat more personal for most individuals. In this sense, our notion of family representations of history is not novel but may be considered a mere application of Moscovici’s (1963) original definition of social representation to the context of family history. However, we believe that family representations may serve as an important bridging concept in understanding the interplay of general social representations of history and present-day political attitudes and behaviors (Bilewicz et al., 2019; Klar & Bilewicz, 2017).

German Family Representations of Heroes and Helpers

According to Liu and Hilton (2005), World War II and the era of National Socialism in Germany is one of the most influential examples, if not the single most influential example, of representation of world history in terms of importance and evaluation, and also its prescriptive nature. It should therefore be worthwhile examining more closely the interplay of highly personal family representations and more general historical representations in the German context.

In contemporary Germany, the social representation of National Socialism and the systematic persecution and murder of social groups from 1933 to 1945 seems to fall right in between what Assmann (1992) termed cultural and communicative collective memory. Somewhat similar to the descriptive components in Liu and Hilton’s (2005) conceptualization, according to Assmann’s (1992) distinction, cultural memory contains a group’s experience and behavior, which are transmitted through generations via traditions and informal and formal practices, such as institutionalized education or visits to memorial sites. Communicative memory, on the other hand, contains the active and interactive form of remembrance in daily interactions by individuals and groups. In terms of Assmann’s (1992) conceptualization, family representations as we understand them span cultural and communicative memories and connect both to autobiographical memories.

It is generally assumed that communicative memory depends on individuals such as contemporary witnesses and relies on their personal and potentially biased memory within a timespan of about three to four generations, before it is eventually transformed into cultural memory—or lost and forgotten. Assmann’s (1992) distinction seems particularly relevant for the study of historical representations in Germany today as memory of the Nazi era now falls precisely within communicative and cultural memory. Now is the time, in other words, when social representations of history might be renegotiated, and some have connected the recent increase in far-right electoral support in Germany with its challenges of German memory culture (e.g., Rees, Rees, et al., 2019; Stanley, 2018). Age should therefore be an important demographic feature to be considered when studying collective memory in Germany and elsewhere.

The critical analyses of current developments in German memory culture by Stanley (2018) and others can also be framed in terms of social representations theory. Liu and Hilton (2005, p. 542) argued that a historical symbol “loses representational status to the extent that competing views undermine its claims to authenticity or when the things it symbolizes become irrelevant to the current political context.” This means that, three to four generations after the era of National Socialism, social representations of that era may now be particularly prone to competing views as contemporary witnesses disappear and social representation of that time becomes increasingly contested (Moscovici, 1988).

Reinterpretations of the German past have already been documented in a study by Welzer et al. (2002), which was discussed widely in Germany when it was first published. In their study on the intergenerational transmission of collective memory of the Nazi era in German families, Welzer et al. (2002) conducted 40 family interviews and 142 individual interviews. They found that about two-thirds of the family representations were about victimhood and heroism, whereas wrongdoing and involvement in Nazi crimes were strikingly rare. In other words, most of the participants reported knowledge of ancestors who were victims or helped potential victims, but hardly anyone reported knowledge of perpetrators among their ancestors. This bias in the prevalence of German family representations was due not only to selective reporting within the study itself (i.e., social desirability), but also to the participants’ biased perception, reshaping representations into more positive ones, or omission to ask about acts of wrongdoing in the first place (see Liu & László, 2007).

Welzer et al. (2002) argued that remembrance within German families faces the challenge of bringing together a cognitive perspective based on knowledge from institutionalized contexts such as education and the media, on the one hand, and a highly emotional perspective shaped by personal family representations, on the other (see also Hirst et al., 2009). Similar to our conceptualization of family representations as an intermediate level between the individual and national level and bidirectional links between social and family representations, Welzer et al. (2002) argued that the cognitive “encyclopedic” perspective may contain factual knowledge about events and societal conditions. The emotional “family album” perspective, however, comprises personal representations of heroism, suffering, and sacrifice in German families during that time, and the two perspectives may be difficult to reconcile.

The Current Study

Similar to the notion of social representations of history for national groups (e.g., Liu & Hilton, 2005), we propose and investigate family representations as “foundational myths” for individuals and their ancestors. Such family representations contain descriptive components, the most notable of which for the current study are the roles ancestors played during important historical eras (Giner-Sorolla et al., in press; Liu et al., 2005). While family representations may be part of a larger and more complex “family narrative” (e.g., László, 2008), we focus on mere descriptive components in the current study. We do so to illustrate how family representations may help us to understand more fully the interplay of general social representations of history and present-day political attitudes and behaviors.

As an intermediate, more emotional and proximate prescriptive background, we argue that family representations may serve as points of reference to which, on the one hand, more general historical representations on the national level will be aligned and, on the other, individual present-day political attitudes and behavior will be oriented. While causal claims are beyond the scope of this article, our general underlying argument here is that all three concepts—family representations, more general social representations of history, and present-day political attitudes—will be systematically linked. Previous work has provided partial support for the latter half of this chain of thought. Licata and Klein (2010), for example, have shown that Belgian participants’ representations of the national past were associated with support for reparative action mediated by group-based guilt. In the German context, Rees et al. (2013) linked group-based guilt and the shame participants experienced with regard to the Holocaust to attitudes toward Turkish people living in Germany today. Previous work has therefore linked social representations of history with present-day political attitudes in a way that is consistent with basic premises of social representations theory (Hilton & Liu, 2017; Liu & Hilton, 2005; Liu & László, 2007). We argue that the concept of family representations can complement such work by offering insights into the sources of systematic bias in collective memory, and open up new research perspectives. It may serve as a conceptual bridge and help us to understand some of the underlying psychological motivations for individuals to remember some facts and forget others (e.g., Bilewicz et al., 2019). Both historical and family representations should be biased, with the latter being even more prone to bias than the former due to their personal relevance (e.g., Campbell & Sedikides, 1999; Miller & Ross, 1975; Zuckerman, 1979). While there is some preliminary evidence for an overemphasis on family representations of victims and helpers over perpetrators (Welzer et al., 2002), their prevalence and correlates have not yet been documented in a representative German sample as far as we know. Similar group-serving biases should exist on the national level but may be less pronounced (Bilewicz et al., 2019).

If our motivational account of self-serving bias in social and family representations of history is correct, age should be a particularly relevant demographic variable for the current study. Older participants should be highly motivated to produce historical representations of history—on both the national and the family level—that are biased in a self- or group-serving manner, as they should be more directly implicated by such representations than younger participants. While we did not have any specific predictions with regard to the participants’ age based on the previous literature, we generally assumed that older participants would tend to report more positive representations than younger participants.

Specifically, our predictions were that (1) family representations, as well as more general societal representations embedded in social representations of history, would be systematically biased in a way that overestimated positive and underestimated negative behaviors, and that this tendency may be more pronounced for older participants; (2) family representations would be systematically linked with more general historical representations in such a way that positive family representations coincided with more general positive societal representations; and (3) attitudes toward present-day political issues would be a function of the two.

Methods

Participants

A representative sample of 1000 German participants was surveyed in standardized telephone interviews, which were conducted by a professional survey institute between November and December 2018. The survey was part of a larger project on the culture of remembrance in Germany (for a descriptive overview and technical details, see Rees, Zick, et al., 2019; for work based on the same research program, see also Papendick et al., in press) and covered a broad range of topics. The most relevant for the current article were social representations of history, including societal representations, family representations, and present-day political issues, which we will focus on and describe in more detail below. To ascertain the representativeness of the sample, telephone numbers were randomly generated and the last-birthday method was used to select participants randomly within households. One-third of the participants were contacted via their mobile phone numbers.

Both subsamples (landline only vs. mobile only) differed with regard to age, with a younger mean age in the mobile-only subsample than in the landline-only subsample: M = 42.84 (SD = 15.68) versus M = 53.92 (SD = 19.19), t(803) = 9.77, p < .001. The analyses will be based on the overall sample, which was M = 50.19 (SD = 18.82), with slightly more female (51.1%) than male participants.

Measures

Social Representations

In an open format at the beginning of the interview, the participants were asked to name “the event from German history you think future generations in Germany should remember.” The responses were later coded into the categories “National Socialism” (including general references to World War II, the Holocaust, or specific related historical events such as the “end of World War II” or “seizure of power”); “Reunification” (including references to the division of Germany); and “Other Historical Events.” As an indicator of a societal representation specific to the Nazi era, the participants were later asked to give estimates of the proportion of perpetrators, victims, and those helping potential victims among the German population during the time of National Socialism. Reponses for these three questions ranged from 0% to 100%.

Family Representations

Three questions—“Were ancestors of yours among the perpetrators during the time of National Socialism?”; “Were ancestors of yours among the victims during the time of National Socialism?”; and “Did any of your ancestors help potential victims during the time of National Socialism?”—(coded 0 = “no” and 1 = “yes”) were used to assess family representations for that time. If affirmed, they were followed up with more specific categories to describe exactly how the ancestors had been perpetrators, victims, or helpers. Multiple responses were possible, which is why some of the numbers in Table 2 add up to more than the superordinate category total.

Present-Day Political Issues

The final part of the interview covered a range of political issues. In order to show a generalization from the remembrance of the Nazi era to a seemingly unrelated outgroup (Hypothesis 3), we assessed attitudes toward refugees seeking protection in Germany today. These were measured with the two items “Germany should receive more refugees from regions of crisis around the world” and “Germany should have an upper limit for the number of refugees admitted to the country” (reverse scored) on a scale ranging from 1 = strongly disagree to 5 = strongly agree. Both items were correlated, r(979) = .48, p < .001, and averaged to form an index of attitudes toward refugees.

Results

In the first part of this section, we will give an overview of the distribution and correlations of the social, societal, and family representations, as well as present-day political attitudes, in our sample. From the mere descriptive analyses, we can derive some conclusions regarding the selective transmission of both national and family history. In the second part, we will present some more detailed analyses supporting our proposition that family representations are not only selectively transmitted, but also systematically correlated with general historical representations and present-day political attitudes.

Descriptive Analyses and Demographic Correlates

Social Representations

Events related to the time of National Socialism were nominated most frequently as those that future generations in Germany should remember: 44.3% of our participants nominated events from that category. The second most frequent category was “Reunification” (36.7%). In addition, 8.2% of the respondents also explicitly nominated “both World Wars” and 9.6% nominated other events that were too unspecific to fall within one of the previous categories and too infrequent to form their own categories (e.g., foundation of the German Republic, introduction of women’s voting rights, and reformation of the church). For the analyses reported below, we will focus on the two most frequent categories—“National Socialism” and “Reunification”—as these turned out to be the two most prominent and potentially competing social representations.

With regard to German society during the time of National Socialism, the mean percentage estimates were highest for victims (M = 34.71%) and perpetrators (M = 33.98%) and lowest for helpers (M = 15.77%). While the first two did not differ significantly, t(897) < 1, p = .46, the mean percentage estimates for helpers were significantly lower than for both victims, t(895) = 25.59, and perpetrators, t(914) = 19.53, both p < .001, according to paired-samples t tests. Only estimates for perpetrators were correlated with participants’ age in the sense that older participants gave lower estimates, r(937) = −.17, p < .001. As can be seen from Table 1, the societal victim and helper representations were moderately correlated with each other, r(895) = .43, p < .001, while the perpetrator representation was only weakly correlated or unrelated with the other two.

Table 1.

Distribution and Correlations of Measures.

	M (SD)	(2)	(3)	(4)	(5)	(6)	(7)
Family representations
(1) Perpetrators	19.6	.08*	.08*	.19**	−.06	−.10*	.11**
(2) Victims	35.9		.25**	.04	−.03	−.01	.00
(3) Helpers	28.8			−.02	.17**	.19**	.08*
Societal representations
(4) Perpetrators	33.98 (25.27)				.09**	.06	.20**
(5) Victims	34.71 (23.73)					.43**	−.06
(6) Helpers	15.77 (15.26)						−.15**
(7) Attitudes toward refugees	3.07 (1.11)

Note. (1) to (3) were coded 1 = “yes” and 0 = “no” and the overall percentage of “yes” responses is displayed. (4) to (6) were recorded as estimates from 0% to 100%. (7) ranged from 1 = strongly disagree to 5 = strongly agree.

*p < .05. **p < .01.

Table 2.

Representations in German Families (%).

	Yes	No	Don’t know
Perpetrators	19.6	69.8	10.6
Direct involvement (e.g., member of the SS, soldier, or policeman)	12.7
Indirect involvement (e.g., as an employee of a public authority)	4.3
Victims	35.9	58.8	5.3
Civilian victims of war (e.g., in air raids)	18.5
Refugees or displaced persons	17.4
Members of a persecuted group (e.g., Jews, Sinti, and Roma, victims of political persecution)	8.1
Helpers	28.8	42.2	29.0
Supporting them in everyday needs (e.g., slipping them food or medicine)	22.4
Hiding potential victims	9.6
Helping them to flee	8.2
Supporting them as a public servant (e.g., issuing false papers)	1.7

Family Representations

Table 2 gives a more detailed overview of the representations passed on in our respondents’ families. The distribution of family representations was as expected, with the most frequent family representation being one of victimization—mostly civilian victims of war (18.5%) and refugees or displaced persons among the ancestors (17.4%). The second most frequent family representation was one of helping—mostly support in everyday needs (22.4%), but also more dangerous forms of helping, such as hiding potential victims (9.6%) and helping them to flee (8.2%). Family representations of perpetrators were the least frequent. However, if the participants reported perpetrators among their ancestors, these usually fell within the most immediate category of “direct involvement” (12.7%).

The participants who reported helpers and victims among their ancestors were slightly older, t(708) = 2.33, p = .02 and t(863) = 1.89, p = .06, respectively, with the latter approaching conventional levels of statistical significance. Family representations of perpetrators, on the other hand, tended to be more prevalent among younger participants, but not significantly so, t(892) = 1.56, p = .11. The demographic backgrounds of the participants reporting different family representations are compared in more detail in Table 3 (for additional comparisons of the subgroups, see Rees, Zick, et al., 2019). As can be seen from Table 1, family representations of victims and helpers tended to co-occur within families, χ² (N = 691, df = 1) = 42.33, p < .001. Of those respondents who did not know of victims among their ancestors, 31.5% reported helpers, whereas of those respondents who knew of victims among their ancestors, this ratio increased to 57.0%. A perpetrator family representation, on the other hand, was only weakly correlated with family representations of helpers, χ² (N = 673, df = 1) = 4.14, p = .042, and victims, χ² (N = 876, df = 1) = 5.62, p = .022.

Table 3.

Age Differences of Participants Reporting Different Family Representations.

Representation	Yes	No	Don’t know	Test statistic
Perpetrators
M_age (SD)	48.71 (18.04)_ab	51.13 (19.12)_a	46.77 (17.76)_b	F(2,997) = 3.24, p = .04
Victims
M_age (SD)	52.49 (16.34)_a	50.25 (19.80)_a	34.04 (15.35)_b	F(2,997) = 23.38, p < .01
Helpers
M_age (SD)	52.42 (19.41)_a	48.99 (19.19)_b	49.74 (17.50)_ab	F(2,997) = 2.97, p = .05

Note. The different subscripts indicate significant mean differences per line according to post-hoc LSD tests.

It also seems noteworthy that there was a substantial number of participants who responded “don’t know” when asked about their ancestors, especially with regard to helpers, where 29% chose that response option (see Table 2). In terms of demographic background, the respondents choosing that category versus either of the other two were younger in the context of perpetrators, t(998) = 1.98, p = .048, and in the context of victims, t(61) = 7.81, p < .001. In the former case, however, this difference was driven by age difference between the “no” and “don’t know” response categories (see Table 3).

Present-Day Political Issues

Overall, the participants did not have a clear tendency in their attitudes toward refugees seeking protection in Germany today, with a mean close to the scale midpoint, M = 3.07 (SD = 1.11). There was a tendency for younger participants to hold more positive attitudes toward refugees, r(990) = .10, p = .002.

Alignment of Family Representations, Social Representations, and Present-Day Political Attitudes

While we have analyzed them separately so far, family representations, more general historical representations, and present-day political attitudes were not independent of each other. Recall that we did not predict any specific direction or causality; our general assumption was that family representations and broader social representations of history will be aligned, and previous work has linked the latter with present-day political attitudes. We speculated that individuals may use their more proximate family representations as points of reference to align larger national representations because family representations will likely be the first, most emotional and most relevant source of historical knowledge for many. Based on this logic, we ran the following set of analyses.

Those participants who reported knowledge of perpetrators among their own ancestors also estimated the percentage of perpetrators in the general population to be higher, M = 43.04 (SD = 27.68) versus M = 31.29 (SD = 24.32), t(263) = 5.21, p < .001, and the percentage of helpers to be lower than participants who did not report such knowledge, M = 13.01 (SD = 14.07) versus M = 16.81 (SD = 15.60), t(320) = 3.17, p = .002. When controlling for participant age in analyses of covariance, victim estimates also differed by groups, with lower estimates for victims in the general population by those participants reporting perpetrators among their ancestors, M = 32.11 (SD = 22.63) versus M = 36.49 (SD = 22.95), F(1,746) = 5.03, p = .03. The two other effects held when controlling for age.

Participants reporting knowledge of perpetrators among their ancestors also held more positive attitudes toward refugees seeking asylum in Germany today, M = 3.31 (SD = 1.17) versus M = 3.00 (SD = 1.08), t(292) = 3.27, p < .001. This effect held when controlling for age.

In a final step, we tested for mediation of the link between a perpetrator family representation and present-day political attitudes via differences in more general societal representations. This was based on our assumption that individuals will use their more proximal family representations as points of reference to align more general social representations of history and, ultimately, present-day political attitudes. These analyses are summarized in Figure 1. The direct link between perpetrator family representation and attitudes toward refugees was indeed rendered insignificant when estimates for percentages of perpetrators, victims, and helpers among the general population during the Nazi era were entered into the regression model. Bootstrap analyses using Hayes’ (2018) PROCESS macro with 5000 bootstrap resamples and 95% confidence intervals confirmed two independent significant mediation effects of a perpetrator family representation through higher estimates of perpetrators, CI = (.06; .17), and lower estimates of helpers, CI = (.01; .07), but independent of estimates of victims among the general population, CI = (−.01; .01).

Figure 1.

Mediation of the Link Between Perpetrator Family Representation and Present-Day Political Attitudes via Societal Representations.

Discussion

In the current study, we set out to investigate family representations as a concept spanning cultural and communicative collective memories, and connecting both to autobiographical memories. We have also argued that family representations may serve as an intermediate or more proximate and personal prescriptive backdrop for present-day political attitudes and behavior than general historical representations at the national level. To use Welzer et al.’s (2002) image, abstract “German Nazis” of the past will be far less threatening to the self than a family member who may have been involved in atrocities—therefore “Grandpa was not a Nazi” for many Germans (for a reflection on the Goldhagen controversy, see Rensmann, 1999). While the mere numbers reported in our study are clearly incompatible with historical facts, such biases can be particularly problematic if, as we suggest, family representations are then used as points of reference to align more general societal representations and, ultimately, may also inform present-day political attitudes. In other words, what is regularly referred to as historical moral responsibility for Germans because of the Holocaust (for a recent example, see Steinmeier, 2020) may be less emotional and less obligatory the more family representations of perpetrators are censored from German collective memory (Bilewicz et al., 2019; Licata & Klein, 2010; Rees et al., 2013).

Two Competing Representations in German Collective Memory Today

As expected, we found an overemphasis of positive and underemphasis of negative behaviors—that is, representations of victims and helpers as compared to perpetrators on the levels of both the overall German population during the Nazi era and family representations in Germany today. On the latter level, there was an even clearer tendency, with family representations of victimization being the most prevalent followed by representations of helping potential victims, and representations of perpetrators being the least prevalent. As far as we know, ours is the first study to document such biases systematically in a representative German sample.

Family representations and more general societal representations were not independent of each other but linked in a consistent way. Those participants reporting perpetrators among their ancestors also estimated the percentage of perpetrators in the general population to be higher and the percentage of helpers to be lower. Knowledge of perpetrators in one’s family seems to be linked with a more pessimistic—or realistic—view of the behavior of the German population during the Nazi era more generally. Also, it seems that “historical responsibility” for current political decisions, such as welcoming refugees from regions of crisis in Germany today, may be more pronounced for those who know of ancestors who were perpetrators during the Nazi era (Licata & Klein, 2010; Rees et al., 2013).

It seems noteworthy that family representations of victims and helpers tended to co-occur within families while a perpetrator representation was independent of the other two. We argue that this pattern indicates two competing representations in German collective memory today, by which present-day political attitudes are shaped. It is through such alignments of more general social representations of history with family representations, we believe, that national historical charters may be filled with quite different content—historical moral responsibility because Germans have been perpetrators or rejection of responsibility because Germans have suffered too (Liu & Hilton, 2005; Noor et al., 2012; Young & Sullivan, 2016).

Sources of Bias in German Family Representations

There are certainly different processes at play in producing the pattern of results we found, and disentangling them was beyond the scope of the current study. We want to emphasize that we do not think that our participants’ ancestors actively and knowingly lied to them, or that our participants lied during the interviews. However, there were systematic differences by age that seem to be indicative of self- and group-serving biases (see Campbell & Sedikides, 1999; Miller & Ross, 1975). More specifically, older participants gave lower estimates for perpetrators during the Nazi era and tended to report more helpers and victims in their own families. Interestingly, family representations of victims during the time of National Socialism seemed clearer among the older participants—those participants replying “yes” or “no” to the respective question were older, while younger participants “didn’t know.”

Motivational processes (Zuckerman, 1979) and relational dynamics (Sedikides et al., 1998) have been shown to play a role in this context. In this sense, the information that was provided to our participants within their families had probably been censored already. Another important process is the selection, reinterpretation, and bias introduced by individuals themselves (Bilewicz et al., 2019; Klar & Bilewicz, 2017). For example, they may only ask for some information in the first place or adapt their questions to the alleged family climate, introducing in the family what Noelle-Neumann (1974) called a “spiral of silence” in the context of public opinion expression (see also Moy et al., 2001). Future research may find it fruitful to investigate in more detail what kind of bias is introduced and at what stage. We also hope that the concept of family representations of history that we outline in the current article will inspire future work in the context of social representations of history. Indeed, other study designs may prove more effective to that end than the instrument of a representative survey that we used in this case (e.g., Welzer et al., 2002).

Limitations and Future Directions

An obvious limitation of the current study is its cross-sectional correlational design, which bars any causal claims. We have argued that it seems plausible for individuals to align general societal representations according to their family representations. However, another reading of the results we have presented may suggest the reverse: contemporary German educational curricula (i.e., national representations) seem to emphasize that responsibility for the Nazi crimes was distributed among the population. The fact that younger participants were more likely to rate members of their own family as having been perpetrators may therefore reflect the influence of this wider social representation. In other words, national representations drive family representations, and not the other way round.

We assume that both of the above notions are true. Given their dynamic, interactive relationship in reality, it seems highly likely that the relationship between family representations, on the one hand, and structural social representations, on the other, will be bidirectional. We hear the views of our family about a past event, which shapes how we think and feel about that event. However, our family’s version of the past may be challenged in history classes at school. Next time at the dinner table, when we hear our family’s version, we may challenge it by referring to the alternative interpretation we learnt from our history classes, and the family representation may be revised. For the current study, we used meditation analyses not to infer causal claims, but to illustrate how all three measures—family representations, more general social representations, and present-day political attitudes—are linked in a consistent manner.

Longitudinal or experimental data could address issues of causality more adequately, and we are currently preparing a panel study on historical representations among German adolescents. Our research group is also currently establishing a monitoring instrument to examine potential changes in the overall prevalence of family representations in Germany over the years. It should be interesting to investigate such changes for the time of National Socialism in German collective memory between communicative and cultural memory (Assmann, 1992), but also other events and eras that are collectively remembered in Germany and elsewhere. We strongly encourage large-scale studies on family representations in other cultural contexts and how they are transmitted in families (for a study in the Dutch context, see Coopmans et al., 2017) or in historical education (for a collection of psychological obstacles in the German and Polish context, see Bilewicz et al., 2017).

Conclusion

In a historical and national context where it is particularly difficult to reconcile the need for a narrative of a morally superior, just, and justified in-group (Leach et al., 2007; Liu & Hilton, 2005), construing this in-group as victims may be psychologically tempting (Giner-Sorolla et al., in press; Klar, 2016; Klar & Bilewicz, 2017). Historically and politically, however, it seems highly problematic if a group who have caused unprecedented historical harm self-construe as a group of victims and helpers in retrospect (Welzer et al., 2002). It seems particularly relevant to study such processes of collective reinterpretation of the past at a time when far-right parties advocating fundamental changes in German memory culture are gaining electoral support (e.g., Stanley, 2018). Understanding exactly how societal processes and dynamics function for regimes to normalize inhuman ideology in the general public seems one of the core lessons of the past. We believe that biased family representations and more general social representations of history may be social-psychological barriers to really grasping this lesson (see also Papendick et al., in press).

To come back to The Doctor’s words at the beginning of this article, and in light of the high prevalence of family representations of helpers during the time of National Socialism, we have to conclude that some stories unfortunately never happened—not enough of them anyhow.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The MEMO project is supported by a research grant from the foundation “Remembrance, Responsibility and Future” (EVZ Foundation). Preparation of this manuscript was facilitated by a grant from the German Federal Ministry of Education and Research (01UG2050JY).

References

Assmann

(1992). Das kulturelle Gedächtnis: Schrift, Erinnerung und politische Identität in frühen Hochkulturen. Beck’sche.

Baumeister

R. F.

(1982). A self-presentational view of social phenomena. Psychological Bulletin, 91(1), 3–26.

Baumeister

R. F.

(1986). Public and private self. Springer.

Bilewicz

Witkowska

Fritsche

Barth

Babinska

(2019). Motivated histories: How visions of national past reflect human motivations in Poland and Germany. In Mukherjee

Salter

P. S.

(Eds.), History and collective memory from the margins (pp. 167–188). Nova Science.

Bilewicz

Witkowska

Stubig

Beneda

Imhoff

(2017). How to teach about the Holocaust? Psychological obstacles in historical education in Poland and Germany. In Psaltis

Carretero

Čehajić-Clancy

(Eds.), History education and conflict transformation: Social psychological theories, history teaching and reconciliation (pp. 169–197). Palgrave Macmillan.

Brown

N. R.

Lee

P. J.

Krslak

Conrad

F. G.

Hansen

T. G. B.

Havelka

Reddon

J. R.

(2009). Living in history: How war, terrorism, and natural disaster affect the organization of autobiographical memory. Psychological Science, 20(4), 399–405.

Campbell

W. K.

Sedikides

(1999). Self-threat magnifies the self-serving bias: A meta-analytic integration. Review of General Psychology, 3(1), 23–43.

Coopmans

Lippe

T. van der

Lubbers

(2017). What is “needed” to keep remembering? War‐specific communication, parental exemplar behaviour and participation in national commemorations. Nations and Nationalism, 23(4), 746–769.

Doosje

Branscombe

N. R.

Spears

Manstead

A. S. R.

(1998). Guilty by association: When one’s group has a negative history. Journal of Personality and Social Psychology, 75(4), 872–886.

10.

Festinger

(1957). A theory of cognitive dissonance. Stanford University Press.

11.

Figueiredo

Martinovic

Rees

J. H.

Licata

(2017). Collective memories and present-day intergroup relations: Introduction to the special thematic section. Journal of Social and Political Psychology, 5(2), 694–706.

12.

Gawronski

Strack

(2012). Cognitive consistency: A fundamental principle in social cognition. Guilford Press.

13.

Giner-Sorolla

, Hilton, D., Erb, H.-P., Durante, F., Flaßbeck, C., Fülöp, E., Mari, S., Petrović, N., Sekerdej, M., Studzinska, A., Skitka, L. J., Washburn, A. N., & Zadora, A. (in press). Assigning moral roles within the Second World War in Europe: National similarities, differences, and implications for group-level moral representations. Asian Journal of Social Psychology.

14.

Hilton

Liu

(2017). History as the narrative of a people: From function to structure and content. Memory Studies, 10(3), 297–309.

15.

Hirst

Phelps

E. A.

Buckner

R. L.

Budson

A. E.

Cuc

Gabrieli

J. D.

Johnson

M. K.

Lyle

K. B.

Lustig

Mather

Meksin

Mitchell

K. J.

Ochsner

K. N.

Schacter

D. L.

Simons

J. S.

Vaidya

C. J.

(2009). Long-term memory for the terrorist attack of September 11: Flashbulb memories, event memories, and the factors that influence their retention. Journal of Experimental Psychology: General, 138(2), 161–176.

16.

Klar

(2016). Four moral obligations in the aftermath of historical ingroup victimization. Current Opinion in Psychology, 11, 54–58.

17.

Klar

Bilewicz

(2017). From socially motivated lay historians to lay censors: Epistemic conformity and defensive group identification. Memory Studies, 10(3), 334–346.

18.

László, J. (2008). The science of stories: An introduction to narrative psychology. Routledge/Taylor & Francis Group.

19.

Leach

C. W.

Ellemers

Barreto

(2007). Group virtue: The importance of morality (vs. competence and sociability) in the positive evaluation of in-groups. Journal of Personality and Social Psychology, 93(2), 234–249.

20.

Licata

Klein

(2010). Holocaust or benevolent paternalism? Intergenerational comparisons on collective memories and emotions about Belgium’s colonial past. International Journal of Conflict and Violence, 4(1), 45–57.

21.

Liu

J. H.

(1998). Social representations of history: Preliminary notes on content and consequences around the Pacific Rim. International Journal of Intercultural Relations, 23(2), 215–236.

22.

Liu

J. H.

Goldstein-Hawes

Hilton

D. J.

Huang

L.-L.

Gastardo-Conaco

Dresler-Hawke

Pittolo

Hong

Y.-Y.

Ward

Abraham

Kashima

Ohashi

M. M.

Yuki

Hidaka

(2005). Social representations of events and people in world history across twelve cultures. Journal of Cross-Cultural Psychology, 36(2), 171–191.

23.

Liu

J. H.

Hilton

D. J.

(2005). How the past weighs on the present: Social representations of history and their role in identity politics. British Journal of Social Psychology, 44(4), 537–556.

24.

Liu

J. H.

László

(2007). A narrative theory of history and identity. In Moloney

Walker

(Eds.), Social representations and identity: Content, processes, and power (pp. 85–107). Palgrave Macmillan.

25.

Liu

J. H.

Liu

S. H.

(1997). Modernism, postmodernism, and neo-Confucian thinking: A critical history of paradigm shifts and values in psychology. New Ideas in Psychology, 15(2), 159–177.

26.

Liu

J. H.

Wilson

M. S.

McClure

Higgins

T. R.

(1999). Social identity and the perception of history: Cultural representations of Aotearoa/New Zealand. European Journal of Social Psychology, 29(8), 1021–1047.

27.

Miller

D. T

Ross

(1975). Self-serving biases in attribution of causality: Fact or fiction? Psychological Bulletin, 82(2), 213–225.

28.

Moffat, S. (Writer), & Talalay, R. (Director). (2015, 5 December). Hell Bent. [Television series episode] In P. Bennett (Producer), Doctor Who. BBC Studios.

29.

Moscovici

(1961). La psychanalyse, son image et son public. Presses Universitaires de France.

30.

Moscovici

(1963). Attitudes and opinions. Annual Review of Psychology, 14, 231–260.

31.

Moscovici

(1973). Foreword. In Herzlich

(Ed.), Health and illness: A social psychological analysis. Academic Press.

32.

Moscovici

(1981). On social representations. In Forgas

J. P.

(Ed.), Social cognition: Perspectives on everyday understanding (pp. 181–209). Academic Press.

33.

Moscovici

(1988). Notes towards a description of social representations. European Journal of Social Psychology, 18(3), 211–250.

34.

Moy

Domke

Stamm

(2001). The spiral of silence and public opinion on affirmative action. Journalism and Mass Communication Quarterly, 78(1), 7–25.

35.

Noelle-Neumann

(1974). The spiral of silence: A theory of public opinion. Journal of Communication, 24(2), 43–51.

36.

Noor

Shnabel

Halabi

Nadler

(2012). When suffering begets suffering: The psychology of competitive victimhood between adversarial groups in violent conflicts. Personality and Social Psychology Review, 16(4), 351–374.

37.

Papendick

Rees

J. H.

Zick

(in press). Unknowing, indifferent, or committed: Relations between age and the assessment of the German population’s involvement and inaction during the time of National Socialism. Asian Journal of Social Psychology.

38.

Rees

J. H.

Allpress

J. A.

Brown

(2013). Nie wieder: Group-based emotions for ingroup wrongdoing affect attitudes toward unrelated minorities. Political Psychology, 34(3), 387–407.

39.

Rees

J. H.

Rees

Y. P. M.

Hellmann

J. H.

Zick

(2019). Climate of hate: Similar correlates of far right electoral support and right-wing hate crimes in Germany. Frontiers in Psychology, Article 10:2328.

40.

Rees

J. H.

Zick

Papendick

Wäschle

(2019). Multidimensional Remembrance Monitor: Study II. Stiftung Erinnerung, Verantwortung und Zukunft. https://www.stiftung-evz.de/service/publikationen/studien.html

41.

Rensmann

(1999). Holocaust memory and mass media in contemporary Germany: Reflections on the Goldhagen debate. Patterns of Prejudice, 33(1), 59–76.

42.

Sammut

Howarth

(2014). Social representations. In Teo

(Ed.), Encyclopedia of critical psychology (pp. 1799–1802). Springer Science.

43.

Sedikides, C., Campbell, W. K., Reeder, G. D., & Elliot, A. J. (1998). The Self-Serving Bias in Relational Context. Journal of Personality and Social Psychology, 74, 378--386.

44.

Stanley

(2018, September 10). Germany’s Nazi past is still present. The New York Times. https://www.nytimes.com/2018/09/10/opinion/germanys-nazi-past-is-still-present.html

45.

Steinmeier

F.-W.

(2020, May 8). 75th anniversary of the end of the 2nd World War. Der Bundespräsident. https://www.bundespraesident.de/SharedDocs/Reden/EN/Frank-Walter-Steinmeier/Reden/2020/05/200508-75th-anniversary-World-War-II.html

46.

Tajfel

Turner

J. C.

(1979). The social identity theory of intergroup behaviour. In Worchel

Austin

(Eds.), Psychology of intergroup relations (pp. 33–48). Nelson-Hall.

47.

Welzer

Moller

Tschuggnall

(2002). “Opa war kein Nazi”: Nationalsozialismus und Holocaust im Familiengedächtnis. Fischer.

48.

Wishart

(1997). The selectivity of historical representation. Journal of Historical Geography, 23(2), 111–118.

49.

Young

I. F.

Sullivan

(2016). Competitive victimhood: A review of the theoretical and empirical literature. Current Opinion in Psychology, 11, 30–34.

50.

Zuckerman

(1979). Attribution of success and failure revisited, or: The motivational bias is alive and well in attribution theory. Journal of Personality, 47(2), 245–287.