Shame onMe ? Shame on You ! Emotional Reactions to Cinematic Portrayals of the Holocaust

The media are playing an increasingly important role in teaching the public about the history of the Holocaust. In Germany, however, Holocaust documentaries have been criticized for eliciting unintended, adverse reactions among the viewers, such as distancing from the victims or calling for closing the books on the past. This criticism stems from the concern that such reactions pose an obstacle to critical-constructive engagement and coming to terms with history. This study examines the interplay between cinematic representation of the Holocaust, film-induced defensive strategies, and group-based emotions of shame. Based on a content analysis of six different film excerpts, we investigated the mediating effects of four defensiveness strategies (distancing from victims, victim blaming, closeness to perpetrators, and rejection of the relevance of the Holocaust) on group-based shame in a sample of 224 pupils from Germany’s third post-war generation in a quasi-experimental field study. The results reveal the complexity of film-portrayals which can foster as well as hinder group-based shame and thus, a constructive dealing with past injustice.

The number of Holocaust survivors is constantly decreasing.Consequently, the task of teaching the history of the Third Reich and the destruction of the Jews has been shifting increasingly to the media's domain of responsibility.
Over the course of the last decades, German television in particular has focused strongly on the topic (Köppen, 2007).The media engagement covers a broad spectrum ranging from the refusal of some filmmakers to even consider the Holocaust as a past event with a definite conclusion (Lanzmann, 2007, p. 35) to the goal of structuring the "presentation of history not just in an investigative manner, but rather […] to make it […] exciting and moving and at the same time authentic" (Knopp, 1999, p. 311, translation by author).These self-set media commitments to teaching the events of the Holocaust often conflict with the wishes and motivations of the German public whose majority refuses to hear "again and again of German crimes against the Jews" (Heyder, Iser, & Schmidt, 2005, p. 151).As the media attention to historical injustice like the Holocaust can evoke a threat to one's social identity (Branscombe, Ellemers, Spears, & Doosje, 2000), German media audiences may seek opportunities for defense and distancing (see also Brown, González, Zagefka, Manzi, & Cehajic, 2008;Iyer, Schmader, & Lickel, 2007;Johns, Schmader, & Lickel, 2005).Thus, strongly conflicting phenomena can be observed: intensified historical engagement with the Holocaust in films and on television on the one hand, and a rejection of engagement by many recipients on the other hand.This raises the question as to how TV-documentaries influence different forms of dealing with the Holocaust and emotional reactions in response to the issue.Of particular interest are the effects of these cinematic testimonials on the third post-war generation in Germany, as they represent a significant source of information about the history of their country.
For this generation, it is essential to adapt a constructive and critical way to deal with their country's past during the Holocaust-that is, by accepting rather than neglecting, dismissing or downplaying the past crimes against Jews.On an individual level, both the lack of certain emotions-such as group-based shame and associated prosocial consequences (Brown & Cehajic, 2008;Brown et al., 2008;Dresler-Hawke & Liu, 2006)-and dysfunctional defense strategies may be opposed to this learning outcome.
This current field study aims to examine the effects of cinematic Holocaust representations broadcasted in Germany.
It tests to what extent these representations provide the ground for different dysfunctional coping strategies among the third post-war generation in Germany, and in what way they affect emotional responses.In contrast to former studies on group-based emotions in the context of the Holocaust that used artificial stimulus material or vignettes (Imhoff, Bilewicz, & Erb, 2012;Peetz, Gunn, & Wilson, 2010), this field study focuses on film material that was actually broadcasted and critically discussed in Germany.Starting from the critique of the influencing mechanisms of specific films and TV documentaries, the current study is carried out in two major steps.In a preliminary study, the real-life stimuli are content analyzed.In the main study, based on the content analysis, we test specific hypotheses in a quasi-experimental field-study investigating how certain approaches to portray the Holocaust influence film-induced ways of dealing with it and group-based shame.

Criticism of Media Treatment of the Holocaust
Since the media's first effort in engaging with the Holocaust, the media's treatment has been accompanied with public scrutiny and criticism about its adequacy, especially in Germany.The TV-series "Holocaust" of 1978 can be seen as the date of founding a medial discourse about the genocide against the European Jews (Stiglegger, 2015).It represented the first fictional accounting, which was watched by a mass audience and harshly criticized for trivializing and emotionalizing history (Hickethier, 1998).Ever since, cinematic representations of the Holocaust have been objects of public discourse and often blamed for producing non-intended adverse responses, and thus hindering constructive engagement with history.These concerns have arisen both in regards to the portrayal of the victims and the perpetrators.
In terms of the victims, Loose (2009) rejected the use of historical pictorial material of the Holocaust to create a sense of authenticity, because the presentation of anti-Semitic stereotypes in National Socialist propaganda videos could reproduce prejudice in viewers.Brink (2003) and Krings (2006) similarly argued that graphic representations of Holocaust victims (e.g., mutilated corpses or dehumanized Jewish victims) could encourage emotional distancing.
In an older study, Hormuth and Stephan (1981) found that viewers of the TV mini-series "Holocaust" who identified with the Nazis were most likely to blame the Jews for their own persecution.They were more likely to do this than those who had identified with the Nazis but had not seen the series, and also more likely to do so than those who identified with the Jewish victims.
Public debates on how to deal with the Nazi past reappear in Germany at regular intervals.Regarding the perpetrator side, criticism recently emerged against films such as Unsere Mütter, unsere Väter [Our mothers, our fathers] (Kadelbach, 2013), which were accused of neglecting the enthusiasm of the majority of the German population for National Socialism at the time (Breuer, 2015).According to this critique, the film belittles the role of the average citizen in the rise and continued existence of the Third Reich.As a consequence, such films follow a perpetrator narrative to a certain extent.Such perspectives submit that films can create an understanding, closeness, and implicit justification for the behavior of the perpetrator side, which is opposed to a reconciliation-oriented approach to dealing with the national past.
The aforementioned refusal to deal with the Holocaust at all (Heyder et al., 2005) is another recurring object of such public debates in Germany (Frindte, 2006).The call for historical closure (Imhoff, Wohl, & Erb, 2013), or Schlussstrichdebatte, is used as a mean of protecting the national group-image (Hanke et al., 2013).This sort of historical distancing represents one facet of secondary anti-Semitism (Frindte, 2006; see also Bergmann & Erb, 1991;Schönbach, 1961), which further includes relativizing and denying the Holocaust (Heyder et al., 2005), negative attitudes against Jews as a result of the confrontation with the Holocaust (Imhoff & Banse, 2009), victim blaming (Zick & Küpper, 2007), and the denial of responsibility (Frindte, 2006).The Walser debate in 1998 was one of the most prominent instances (see Funke, 2004) of such a refusal to deal with the Holocaust.The wellknown writer and public person Martin Walser called for shifting the remembrance of the Holocaust to the private sphere.This displacement from the public was harshly criticized.The central accusation was that Walser would personalize and privatize the systematic failure of the political system and society in Germany in the 1930s.This accusation was also voiced against Holocaust documentaries produced by Guido Knopp.In his TV documentaries, Knopp strongly relied on eyewitnesses and their emotionalized portrayals to illustrate individualized suffering.
Nevertheless, he was also criticized for the fact that such an individualization of history would neglect the political circumstances that generated this suffering (Breuer, 2015).With such neglect, the relevance of the past for current generations would be ignored, which again might indirectly facilitate the demand to close the books on the past.

Group-Based Emotions in the Context of the Holocaust
The way individuals deal with their nations' history is closely linked with emotional experiences.Past research revealed the emergence of group-based emotions such as shame, guilt or regret in various countries and referring to different cases of historical injustices (Brown et al., 2008;Doosje, Branscombe, Spears, & Manstead, 1998;Gausel, Leach, Vignoles, & Brown, 2012;Imhoff et al., 2012).Referring to the Holocaust in particular, Peetz et al. (2010) showed the appearance of group-based guilt regarding German atrocities during the Holocaust as a function of how the events are subjectively situated in time.Imhoff et al. (2012) explored the occurrence of guilt and regret in the context of the Holocaust and their effects on inter-group relations for succeeding generations.Rees, Allpress, and Brown (2013) investigated the consequences of different forms of group-based shame on intergroup-attitudes that are not directly related to the Holocaust.The link between different strategies to deal with their nation's history and group-based shame among Germans was examined by Dresler-Hawke and Liu (2006), who interpreted the occurrence of this emotion as a form of identity positioning.
These findings suggest that group membership and identity are the basic preconditions for the appearance of group-based emotions in general, and the ways in which media engagement with Holocaust exerts influence among contemporary Germans in particular.According to inter-group emotion theory (Mackie, Devos, & Smith, 2000;Smith, 1993), individuals are able to experience emotions on behalf of their group (i.e., emotions that are relevant to their group).Since individuals internalize their group membership and the ascribed characteristics (Tajfel & Turner, 1986;Turner, 1987), the group and its members become part of one's sense of self and influence individual members' experiences, perceptions, cognitions, emotions and actions (Branscombe et al., 2000;Iyer & Leach, 2009).Based on the group's importance for the individual, normative integration and the behaviors of in-group members are essential for a member's positive self-perception (Leach, Ellemers, & Barreto, 2007).Further preconditions for the emergence of group-based emotions such as shame are the perception of relevant events such as the Holocaust as unjustified and the assignment of (collective) responsibility for these relevant events to one's in-group (here, "the Germans") (Lickel, Schmader, & Barquissau, 2004).If individuals are exposed to information on perceived moral transgressions and ascribe it to their in-group, this can threaten the moral image of their group (Branscombe et al., 2000) and elicit an immediate emotional response such as shame or guilt (Doosje et al., 1998;Mackie et al., 2000;Smith, 1993).
As the short summary of the most relevant research on emotions in the context of the Holocaust shows, shame and guilt can be considered as the most prominent forms of emotional responses.While guilt is described as individual-focused (see among others Tangney & Dearing, 2002) and as an internal phenomenon that addresses the questions of agency and responsibility (see Niedenthal, Tangney, & Gavanski, 1994), shame is considered an emotion that affects the core of the self.Shame is linked with the perception of a global worthlessness (Rees et al., 2013;Tangney & Dearing, 2002) and worries about social exclusion because of a damaged social image and condemnation by others (Gausel et al., 2012).Thus, shame is more closely associated with social relations.
According to Dresler-Hawke and Liu (2006), dealing with past injustice is rather about identity position than "commission or omission" (p.134), that is, rather about social images and social relations than the question of who bears the blame.Because of the temporal distance, a missing direct involvement of current generations and thus missing individual component, guilt is less often expressed in the context of past injustice (see Allpress, Brown, Giner-Sorolla, Deonna, & Teroni, 2014).The lower incidence rate of guilt and the association of shame with identity positioning (Dresler-Hawke & Liu, 2006) make emotions of group-based shame especially relevant in the context of the effects of identity threatening media contents about the Holocaust for German recipients.Therefore, they were in the focus of the current research.

Specification of Defensiveness
In order to reduce self-relevant threat and immediately associated emotions of shame (Brown et al., 2008;Iyer et al., 2007;Johns et al., 2005), individuals are motivated to use different defense strategies.Thus, they seek to avoid (subjective) fear of rejection and social isolation caused by attributions of negative characteristics (Lewis, 1971;MacDonald & Leary, 2005).They show avoidance or distancing from the involved protagonists and the threatening events themselves or other forms of defensiveness (Iyer et al., 2007;Lickel, Schmader, Curtis, Scarnier, & Ames, 2005) in order to cope with this self-threatening situation.
Former research on group-based emotions in the context of the Holocaust mainly used artificial stimulus material or vignettes (Imhoff et al., 2012;Peetz et al., 2010; see also Doosje et al., 1998).In contrast, this study focused on the influences of real-life stimuli, TV-documentaries and films which were actually broadcasted and critically discussed in Germany.For this approach, we specified defensiveness strategies following the criticism mentioned above.
Referring to the victim side and derived from the critique by Loose (2009), Brink (2003), and Krings (2006), we identified (1) film-induced distancing from victims as one strategy to reduce group-based emotions of shame.The work by Hormuth and Stephan (1981) suggests that (2) film-induced victim blaming might be another possible reaction against threatening information of Holocaust-documentaries for German recipients.Therefore, it was assumed as a further defensiveness strategy.
Referring to the perpetrator side, we assumed that (3) film-induced closeness to the perpetrators through a perpetrator perspective on the historical events (see the criticism by Breuer, 2015) would inhibit or prevent image threat and group-based shame.
Based on the findings by Heyder et al. (2005) and referring to public discourses like the Walser debate (Funke, 2004), we presumed a (4) film-induced rejection of the relevance of the Holocaust as a generalized distancing, not directed at a specific protagonist like victims or perpetrators, but instead at the event itself.
Thus, assuming that complex real-life stimuli about the Holocaust do not result in simple shame reactions, but rather in multilayered responses, this research addresses the following questions: a) how certain Holocaust portrayals offer the basis for different defense strategies and ways of dealing with the issue and b) how these again hinder or facilitate the emergence of film-induced group-based shame.In line with the criticism mentioned above, we argue that the use of defensiveness strategies is contingent on film contents and thus a film-initiated process.
Therefore, we hypothesize that film effects on shame are mediated by film-induced defense strategies.Due to the complexity of the selected film material in this study, a detailed content analysis of relevant aspects for the emergence of certain defense strategies and group-based emotions is required before formulating hypotheses about film effects.

Preliminary Study: A Content Analysis of Films About the Holocaust
The current study was carried out in two major steps.In a preliminary study, all film sequences we used in the main study were content analyzed for the absence or presence of relevant content aspects in order to predict effects of the film clips on film-induced defense strategies and film-induced group-based shame.Thus, the initial content analysis represents the basis for formulating hypotheses for the main study, in which we tested the hypothesized indirect effects on film-induced group-based shame through the mediator variables film-induced distancing from victims, film-induced closeness to perpetrators, film-induced blaming the victims, and film-induced rejection of the relevance of the Holocaust.

Description of the Coding Manual
The film clips used in this study were chosen because they were broadcasted countrywide and illustrate different ways of portraying the Holocaust.Therefore, they should represent an externally valid selection of cinematic trueto-life stimuli.Even though they are more complex and multilayered than artificially produced experimental stimuli, they can be characterized by certain core features.
We examined the selected film excerpts for key aspects in order to better understand the influencing mechanisms that the single film clips use to arouse threats to the image of the German national identity and offer connecting points for specific film-induced defense strategies to reduce image threat and thus, group-based shame: (1) filminduced distancing from victims, (2) film-induced blaming of the victims, (3) film-induced closeness to perpetrators, and (4) film-induced rejection of the relevance of the Holocaust.
For this purpose, a coding manual was developed using a theory-based approach (Kempf, 2008;Mayring, 2000).
In contrast to a material-based approach, the development of the coding categories in the current research was based on theoretical criticism of certain strategies of portraying the Holocaust voiced in the literature by Loose (2009), Brink (2003), and Krings (2006;see Breuer, 2015), on empirical findings (Heyder et al., 2005;Hormuth & Stephan, 1981) and social psychological theory (e.g., Lickel et al., 2004).This deductive strategy offered the opportunity to narrow down the complexity of the film stimuli to the relevant contents and variables without loosing pertinent information.
Starting from the criticism by Loose (2009), Brink (2003), and Krings (2006) that a stereotypical portrayal of Jews and a dehumanization of the victims could result in a devaluation of Jews and victim blaming, we coded for stereotypical portrayals of Jews (Content Analytical Variable 1 [CAV1]), the extent of Jewish suffering (CAV2), the filmic individualization (CAV3), humanization (CAV4), and exclusive victimization (CAV5) of Jews in the footage.
In order to reveal portrayed closeness to the German perpetrator side, we coded the share of any scenes featuring Germans (CAV6), their portrayed role (CAV7), and their filmic individualization (CAV8), humanization (CAV9), victimization (CAV10), and objectification (CAV11).The latter was expected to reduce the level of potential identification and a devaluation of the German side.Regarding factors that might affect the perceived relevance of the Holocaust for current Germany, we identified and coded for the use of foreign languages (CAV12), a foreign film setting (CAV13), and the usage of historical footage (CAV14).Two raters (male and female) were trained in using the manual.The film excerpts were divided into scenes and the raters independently coded the presence or absence of each category in each scene (three films per rater).The calculation of the interrater reliability was based on a random sample of 68 double coded scenes (corresponds to 33% of all 205 coded scenes, see Riffe, Lacy, & Fico, 2005, p. 144).The mean interrater reliability estimated across all variables (Cohen's κ = .84;Cohen, 1960; percentage agreement = .95)can be considered as very good (Neuendorf, 2002).For a detailed presentation of the coding categories, examples, explanations and their frequency of appearance see Table 1.-tests revealed a significant relation of all CAVs and the excerpts (see Table 1).For a more detailed description of the development of the coding manual and definition of the coding categories see Kopf-Beck (2013).

Description of the Film Excerpts' Contents
The six film excerpts were taken from previously released films or TV documentaries, in order to secure true-tolife stimulus material.The selection of the excerpts (see Table 2) was intended to cover the broadest possible spectrum of representational strategies (Dengler, 2010).The excerpts vary with regard to their stimulus qualities, which are relevant not only for the arousal of group-based shame, but also for the possibilities of connecting different defense strategies.In the following, we briefly introduce the excerpts based on the results of the detailed content analysis and label them according to their main features.For a more detailed description of the film excerpts see Kopf-Beck (2013).

Clip 1: Historical Stereotypes
The film excerpt Ghetto belongs to the genre of docutainment (Dengler, 2010).In an emotionally evocative manner, it describes the situation of the Jewish population in the Warsaw Ghetto in 1942 (Dohle, Wirth, & Vorderer, 2003).
The unique characteristic of the film is the use of historical, pictorial material from National Socialist sources, among others from the anti-Semitic propaganda film Der ewige Jude [The eternal Jew] (Hippler, 1941).Ghetto is marked by a focus on the victims, who are partly portrayed from the National Socialist perspective (percentage of historical footage [CAV14]: 60.3% of all scenes), which emphasizes stereotypes (by far the highest percentage of all excerpts [CAV1]: 12.0%) and denigrates the Jews.In contrast, German perpetrators were presented in an individualized (CAV8: 40.3%) and humanized manner (CAV9: 15.0%), and occured almost as often as Jewish victims (CAV6: 63.8% vs. CAV2: 69.8%).The historical insertions alternate with interviews of Holocaust victims and perpetrators.For all details of the content analytical variables, their frequencies and χ 2 -statistics, see Table 1.

Clip 2: Emotionalized Closeness to Victims and Perpetrators
The second excerpt, Befreiung [liberation], addresses the liberation of the concentration camps by Allied troops, their "re-education" (de-Nazification) policy, as well as the question of collective German guilt.Only Befreiung makes use of historical film clips from Allied sources.In this excerpt, the Jewish suffering is omnipresent (CAV2: 96.3% of all scenes) and portrayed in a dramatic and graphic manner by showing, for example, vast heaps of corpses.The excerpt is characterized by its portrayal of an exclusive and generalizing victimization of Jewish life (CAV5: 53.1%).Similar to the excerpt Ghetto, the clip uses music, speakers and a suitable choice of eyewitnesses and depicts an emotionally evocative form of representation.This gives the perpetrator side much room for attempts to explain what happened from their viewpoint and to voice self-justifications, such as the externalization of responsibility to the societal elites or the political system.Despite some parallelisms, the excerpt Befreiung can be differentiated from Ghetto by the dispensation of stereotypes, the additional focus on Germans as victims from a post-war perspective and a higher relevance for current Germany because of the clear link to a within-Germany setting (see CAV13 in Table 1).

Clip 3: Individualized Victims (Historical Footage)
The excerpt Free Fall.taken from a film by Peter Forgács, exclusively uses historical amateur films by a Jewish-Hungarian businessman, György Petö, from Szeged (Hungary), showing the everyday life of the Petö family.
Thereby, the images in the excerpt focus exclusively on the Jewish side and their victim role (CAV5: 69.9% of all scenes), but in contrast to the excerpts Befreiung and Ghetto, they do so in a way very much detached from the victim role and in a strongly individualized and humanized form (CAV3: 73.6%; CAV4: 60.6%).The pictures contrast with readings of anti-Jewish legislation from Hungary in the early 1940s.Petö's amateur videos resemble "filmdiaries" (Fisher, 2008, p. 240) and acquire their documentary value not through the filmmaker's intentions, but rather subsequently from the producer, Peter Forgács.Free Fall does not try to explain the persecution of the Jewish population but rather leaves the contrasts between individual victims and anonymous legislation to the recipient's interpretation.
Due to the exclusive focus on Jewish life in Hungary (CAV13: 100%), the film only vaguely refers to a German setting.It describes the completion of the disenfranchisement process that ends in the expulsion and destruction of the Jewish population of Hungary.It reveals this above all in the subtitles (percentage of scenes in which foreign language was used [CAV12]: 100%) and in part contrasts it with the visual humanization of the Jews in the film footage.In addition to a geographical and cultural distance through setting and language, the film stresses the temporal distance to current Germany by the exclusive use of historical footage (see CAV14 in Table 1).

Clip 4: Individualized Victims (non-Historical Footage)
In his film Shoah (1985), French filmmaker Claude Lanzmann, similar to Peter Forgács in Free Fall, engages with the "dilemma of representation" of the Holocaust (Young, 2001, p. 44).Thus, Lanzmann contrasts long interviews of victims with contemporary film clips showing the original sites of the reported events in Warsaw (percentage of scenes including foreign settings: 100%).With its extended interview sequences, not held in German, Shoah hardly accommodates the viewing habits of the young study participants and attenuates the relevance for current Germany.The unique characteristic of this film excerpt is the eschewal of any historical pictorial footage (CAV14: 0.0%), as well as the clear individualization and humanization of the Jewish victims (CAV3: 45.7%; CAV4: 58.8%), stressing their active role in the resistance movement during the Warsaw Ghetto Uprising, in contrast to the other film excerpts.This evidence is contrasted with contemporary interviews at the original sites of the events with German residents who deny the connivance or refuse to seriously engage with the past.Thus, especially the perpetrator side (CAV6: 90.7%) is shown in an individualized (CAV8: 40.8%), but also strongly objectifying (CAV11: 53.4%) manner, when it depicts the perpetrator side's strategies of coping with history and presents them in an accusatory way.Even though the representation of the victims is formally high (CAV2: 100%), the victim side plays a secondary role since the clip only refers to it in the guilt context of the perpetrators.One of the main characteristics of this film excerpt is its confrontational character: Self-serving excuses are systematically deconstructed, and in this manner, National Socialism and the destruction of the Jews are made more comprehensible as mass phenomena.Because of its comparatively rare use of historical footage (CAV14: 23.3%), foreign settings (CAV13: 2.7%), and foreign language (CAV12: 4.8%), the relevance for current Germany is the excerpt's most salient feature.

Clip 6: Control Scenario
As a control scenario, we used the first part of the film Free Fall by Peter Forgács (Free Fall 1).Since this sequence immediately precedes Clip 3, the control scenario deals with similar contents and displays similar formal features such as the use of foreign language (CAV12: 100%), foreign setting (CAV13: 100%), and historical footage (CAV14: 100%).Nevertheless, since this first part of Free Fall describes the start of the gradual disenfranchisement of the Jews, the portrayed suffering is less extensive and the Holocaust itself less clearly present than in all the other film excerpts (CAV2: 41.7%; see Table 1).Therefore, this first part of Free Fall was used as a cinematic reference condition for the other five experimental groups, because it most closely corresponded to a classical control condition (no treatment), without completely detaching itself from the issue i .

Defensiveness and Group-Based Shame Hypotheses
The following development of hypotheses regarding the tested model is based on the content analysis of the film material.We derived hypotheses about the effects of film excerpts on film-induced group-based shame and filminduced defense strategies from the frequency distribution of the single coding categories.Since the major research focus of the current study is the effect of the broadcasted real-life stimuli in their entire complexity, we formulated hypotheses regarding the full film-excerpts.

Direct Effects of the Film Clips on Group-Based Shame
According to Lickel et al. (2004), the arousal of group-based shame depends on the extent to which an image threat is present for the in-group.For the Holocaust, this further depends on the extent to which viewers hold the German population accountable for the historical events and experience the events as unjust.The more clearly responsibility and injustice are marked in the excerpts, the more likely is the occurrence of film-induced groupbased shame among the audience.The level of attribution of responsibility for past injustices induced by the films should be greater if Germans are shown in the various films frequently (CAV6) and function in a perpetrator or accomplice role (CAV7).A role as a victim or resistance fighter, for example, should reduce this ascription of responsibility.The perceived injustice of the event should, in the case of a high degree of portrayed Jewish suffering (CAV2), rise with the increased objectivation of the perpetrators (CAV11) and humanization of the Jewish victims (CAV4).Considering all these aspects together, we expected a direct shame-inducing effect of the film clip contemporary relevance (Panorama) only (H1).

Effects of Films on Defense Strategies
The results of the content analysis of all six film clips were used to develop hypotheses on the effects of the single excerpts on the different film-induced defense strategies.Two of them refer to the victim side (distancing from victims and blaming the victims), one to the perpetrator side (closeness), and one to the contemporary relevance of the Holocaust (rejection of relevance).

Film-induced distancing from victims -
We assumed a low level of film-induced distancing from the Jewish victim side, if the film excerpt contained a high degree of individualization (CAV3) and humanization (CAV4) of victims (Bandura, 1999), and a low degree of victimization of the Jews (CAV5) or dramatic portrayals of Jewish suffering (CAV2; see also Brink, 2003), and devaluation through the non-contextualized use of anti-Semitic stereotypes (CAV1) in historical film material (see Loose, 2009).These four elements appear more extensively in the film excerpts emotionalized closeness to victims and perpetrators (Befreiung), individualized victims in historical footage (Free Fall), and non-historical footage (Shoah).Therefore, we assumed a low level of film-induced distancing from victims among the recipients of these three film clips (H2a).
Film-induced blaming the victims -Blaming the victims of an injustice represents one strategy in the process of moral disengagement (Bandura, 1999).Such disengagement from moral standards can be justified by otherness, inferiority of the victim side, and a lack of resistance against the perpetrator side.If a film used representational strategies that facilitated the devaluation of the Jewish victims through a lack of humanization (CAV4; see Bandura, 1999), dehumanization through victimization (CAV5; Brink, 2003) and the presentation of non-contextualized stereotypes (CAV1; Loose, 2009), we assumed a positive (heightening) effect on blaming the victim tendencies.
The results of the content analysis indicate a positive effect on film-induced blaming the victims of the clip featured by historical stereotypes (Ghetto; H3a).
Film-induced closeness to perpetrators -Analogous to the victim side, there are many ways in which closeness to the perpetrator side can be constructed in films.We assumed that a high degree of individualization (CAV8), humanization (CAV9), victimization (CAV10; Bandura, 1999) and a low degree of objectification of the perpetrator's side (CAV6) result in greater perceived closeness of the recipients with the German perpetrator side.Deduced from the content analysis, we presumed such an effect for the clip emotionalized closeness to victims and perpetrators (Befreiung; H4a).

Film-induced rejection of the relevance of the Holocaust -
The rejection of the relevance of the Holocaust for contemporary Germany represents a form of distancing from the events.We assumed the following cinematic elements to facilitate rejection of the relevance of the Holocaust: the use of non-German-speaking eyewitnesses (CAV12), a lack of cinematic references to Germany through non-German settings in films (CAV13), restricted reference to the present through the use of historical film material (CAV14) (Kopf-Beck, Gaisbauer, & Dengler, 2013), as well as the share of scenes featuring Germans and their portrayed role as offenders.Based on these assumptions, we expected a positive effect of the excerpts featured by individualized victims (historical footage; Free Fall) and a negative effect of the clip featured by contemporary relevance (Panorama) (H5a).For an overview of all hypothesized effects of film-clips on film-induced defense strategies, see Table 3.

Effects of Defense Strategies on Film-Induced Group-Based Shame
Distancing strategies are expressed in the escape from or avoidance of situations which might evoke the threatand shame-inducing event (Brown et al., 2008;Lewis, 1971).For the following four film-induced defense strategies we assumed a shame-reduction.
Film-induced distancing from victims -We defined distancing from the victims as the rejection of compassion, sympathy and understanding for Jewish Holocaust victims.Distancing oneself from the victims diminishes the moral failure of the perpetrator side, and thus closes the gap between the in-group's moral standards and its behavior (Bandura, 1999; see also Aguiar, Vala, Correia, & Pereira, 2008).Experiencing the injustice of the Holocaust as less serious serves to reduce image threat, thereby also mitigating group-based shame (Hypothesis H2b).
Film-induced blaming the victims -Claiming that victims are themselves to blame for their persecution is a strategy in the process of moral disengagement (Bandura, 1999).Moral disengagement justifies in-group behavior by circumventing moral standards.Similar to distancing from the victims, it reduces the moral failure.The intended effect of victim blaming is reduced image threat and therefore a lower level of group-based shame (H3b).
Film-induced closeness to perpetrators -Analogous to the victim side, closeness to the perpetrators is operationalized as the compassion, sympathy and understanding for the German side.Interpreted as a form of identification, it facilitates or even represents defensive reactions, which in turn reduce group-based emotions (Klein, Licata, & Pierucci, 2011) such as shame.Thus, we hypothesized that closeness to the perpetrators predicts a lower level of film-induced group-based shame (H4b).

Film-induced rejection of the relevance of the Holocaust -
The demand to close the books on history, often voiced in Germany (Heyder et al., 2005), can represent a temporal distancing from and avoidance of engagement with the historical crimes.The more vigorously the relevance of the Holocaust is rejected, the less group-based shame should be experienced (Peetz et al., 2010;H5b).

Method Data Analysis Strategy
In the main study, we tested if the film excerpts have indirect effects on film-induced group-based shame through the mediator variables film-induced distancing from victims, film-induced closeness to perpetrators, film-induced blaming the victims, and film-induced rejection of the relevance of the Holocaust.To do so, we calculated a multiple mediation model in SPSS (IBM Corporation, 2012) using the product of coefficients approach by Hayes and Preacher (2014; for a detailed discussion about advantages of this approach see Cerin & MacKinnon, 2009;Hayes, 2009;MacKinnon, Lockwood, Hoffman, West, & Sheets, 2002;Preacher & Hayes, 2004, 2008).

Recruitment and Procedure
To answer the question about the effects of Holocaust documentaries on viewers, the third German post-war generation was the study's target group.For this segment of the population, Holocaust engagement can be assumed to occur primarily through media or in a school context.In order to answer the research question on emotional reactions, we selected a quasi-experimental design in the frame of secondary school instruction in German and History classes as the most true-to-life setting.Data was collected at two time points.
Taking into account the former division of Germany, secondary schools in Baden-Württemberg (former FRG) and Thuringia (former GDR) were contacted and asked for participation.Six school classes (9 th and 10 th grade) from each state were randomly assigned to one of the six film sequences.In this way, the individual participants were not artificially separated (i.e., through random assignment) from the classroom group in which they otherwise engaged with historical topics.
Initially, the pupils filled in a pre-test questionnaire.Three weeks later, in the classroom group, they watched one of six film excerpts on the Holocaust and immediately afterwards completed a post-test questionnaire.We obtained written declarations of informed consent from the legal guardians, as well as from the pupils themselves.The participants did not receive any kind of compensation or reward.

Measures
The pre-test questionnaire contained items on demographic information and on national identification (Maes, Schmitt, & Schmal, 1996).The participants were asked to indicate to which degree the following statements were applicable to them personally on a six-point Likert-scale, ranging from 1 = not at all to 6 = strongly: (1) "I identify myself with Germany as a nation"; (2) "I am pleased when I see the German flag"; (3) "I think it is important to care for one's home country's customs"; (4) "If German sportsmen participate in an international competition, I support them"; (5) "I am pleased when I hear the German anthem"; (6) "I am proud to be German"; M = 4.19, SD = 1.00;Cronbach's α = .87.Since emotional approaches to one's nation and its history are of special interest in this study, the items cover not just identification with the country and its representatives, but especially emotional attachment to the country and its symbols (see Items 2, 5, & 6).They do so by taking into account German idiosyncrasies such as the cultural component of national identity (see Item 3; Ditlmann, Purdie-Vaughns, & Eibach, 2011) and the social (un-) desirability of national attachment by using sensitive and unsuspicious items (see Item 4).Overall, the scale covers what can be considered as ingroup attachment (see also Roccas, Klar, & Liviatan, 2006) and thus refers to the basic operationalization of national identification as it was used in former research on historical injustice and group-based emotions (Doosje, Branscombe, Spears, & Manstead, 1998, 2004).
The post-test questionnaire captured items on four different film-induced defense strategies and film-induced group-based shame.Due to film-specificity, it was not possible to use pre-formulated items.Therefore, and in order to cover group-based shame and defense strategies in the context of the watched movies, we developed new items assessed on five-point scales ranging from 1 = not true at all to 5 = absolutely true: (1) Film-induced distancing from victims: "The portrayed Jews aroused my compassion" (reversed); "The portrayed Jews gave a likeable impression to me" (reversed); "I understood the portrayed Jews' situation" (reversed) (M = 2.32; SD = 0 .81;Cronbach's α = .61).
(2) Film-induced blaming of the victims: "The film gave the impression to me that the Jews were persecuted because they were somehow strange and different"; "…because they were somehow inferior"; "…because they were weak and did not resist" (M = 2.45; SD = 1.10;Cronbach's α = .80). ( 3) Film-induced closeness to perpetrators: "The portrayed Germans aroused my compassion"; "The portrayed Germans gave a likeable impression to me"; "I understood the portrayed Germans' situation" (M = 2.00; SD = 0.79; Cronbach's α = .74).As can be seen from the items, the defense strategies (1) film-induced distancing from victims and (3) filminduced closeness to perpetrators describe the two poles of one dimension of an emotional relation (emotional distancing vs. closeness) of the recipient towards the relevant conflict groups.For reasons of readability, the defense strategies were reverse-coded and labeled according to their hypothesized effects on group-based shame (see below).( 4) Film-induced rejection of the relevance of the Holocaust: "The film gave the impression to me that the Holocaust was a dark chapter in German history that has nothing to do with the presence" (M = 3.19; SD = 1.26).
Additionally, the participants indicated their agreement with film-induced group-based shame, assessed with the following self-designed items: "The film gave the impression to me that one has to be ashamed of Germany because many people were holding prejudices against Jews"; "…because many people participated in the persecution of Jews"; "…because many people supported National Socialism"; "The film gave the impression to me that one has not to be ashamed of Germany because the German people were misled and shamelessly used by the Nazis" (reversed); "…because the offenders had no choice than obey the orders -otherwise they would have brought themselves in danger" (reversed); "…because Adolf Hitler and a few Nazis bear the blame for the persecution and killing of the Jews" (reversed).(M = 3.12; SD = 0 .77;Cronbach's α = .77).For detailed descriptions of the post-test measures and their distribution between film sequences see Table A1 in the Appendix.

Sample
According to our theoretical framework, all participants eligible for this study should hold German citizenship and should identify with being German at least to a minimum extent (Mackie et al., 2000;Smith, 1993).The random sample recruited for the study consisted of 258 pupils from twelve school classes of grades 9 and 10 from Baden-Württemberg and Thuringia.
Because of lack of identification with Germany or incomplete questionnaires, 34 participants were excluded from the analysis, so that the final sample consisted of 224 individuals.Of these, 112 (50%) were male, 122 (54%) lived in Thuringia, and the average age was 15.26 years (SD = 0.66).Thirty-four participants in the first scenario watched the sequence "Historical stereotypes" (Ghetto), 43 students in the second scenario saw the clip "Emotionalized closeness to victims and perpetrators" (Befreiung), 44 and 37 participants, respectively, in the third and fourth scenario were assigned to the clips "Individualized victims" using historical footage (Free Fall 2) or nonhistorical footage (Shoah), and 34 people in the fifth scenario watched "Contemporary relevance" (Panorama).
Thirty-two participants watched the control scenario (Free Fall 1).
The primary aim of the study was to investigate the different effects of Holocaust documentaries on the recipients' emotions of group-based shame and potential interaction of films and defense strategies.Therefore, power analysis was conducted setting the alpha error at .05 and power at 0.80 in order to estimate sample sizes that are able to detect medium effect sizes, that is f 2 =0.15 regarding the R 2 increase in the total effect regression model and f = 0.25 in ANOVAs regarding main and interaction effects of film excerpts and defense strategies (Cohen, 1992).The ultimate sample size of N = 224 clearly exceeded the computed required sample size of N = 92 (total effect regression model) and fitted the requirements for ANOVAs (N = 211) in order to identify such effects.
The participants in the different experimental conditions did not differ across conditions in regard to their origin (Thuringia vs. Baden-Württemberg; χ In the further analysis, beside age and gender, we controlled for origin (East-/West-Germany) and national identification in order to identify potential main effects of different memorial cultures in the former divided country and potential main effects of national attachment ii .Intercorrelations of all measures are reported in Table 4.

Discussion
The interplay between historical injustice and group-based emotions, and the ways of coping with such emotions, are of a complex nature.Past research has shown that in the context of past injustice, as for example the Holocaust, group-based shame plays an extraordinary role because of its prosocial implications (see among others Brown & Cehajic, 2008;Brown et al., 2008;Gausel et al., 2012).
Multilayered cinematic representations, which represent a key source for today's generations when addressing such injustices, not only elicit specific emotions like shame, but also simultaneously offer the basis for dealing with or defending against them.Thus, they may represent a starting point of a constructive way to deal with history as well as be opposed to it (Brown & Cehajic, 2008;Dresler-Hawke & Liu, 2006).
The present study aimed to investigate influential mechanisms of cinematic representations of the Holocaust in the third German post-war generation and disentangle their ambivalent effects on reconciliation processes.We postulated specific ways in which media presentation affects group-based shame, and examined the mediating influence of various film-induced defense strategies on this emotion.We selected a quasi-experimental design and used films previously broadcasted on German television in order to assure the external validity of our findings.
Thus, the current research was oriented towards the reality of the teaching and knowledge transfer about the history by the German media and schools as it takes place about 75 years after the Holocaust and the Second World War.
Starting from earlier empirical findings (Hormuth & Stephan, 1981) and theoretical criticism of different strategies to portray the Holocaust (Breuer, 2015;Brink, 2003;Krings, 2006;Loose, 2009), we identified different filmic features which, mediated by certain ways of dealing with the Holocaust, elicited or reduced group-based shame.

The Victim Side
Regarding the victim's side, we were able to show that an emotionalized portrayal of victims, which focused on Jewish suffering during the Holocaust without exclusive victimization, increased group-based shame among members of the third post-war generation.This effect, which was evoked by drastic footage from concentration camps and interviews with contemporary witnesses in the clip Befreiung, was mediated by less distancing of the German recipients from the Jewish victims.This could be ascribed to the fact that the German audience developed more compassion, sympathy, and understanding for the Jewish side, and thus reported a higher level of film-induced group-based shame.
The criticism by Brink (2003) and Krings (2006), according to which a drastic and dehumanizing portrayal of victims would result in increased distance between viewers and the victim group, could not be completely disconfirmed because of the complexity of the real life stimuli we used.Nevertheless, the results suggested that the audience needed to be confronted with a minimum of explicit suffering of the victim's side in order to report film-induced shame.The rather demanding excerpts with an individualized victim focus (Free Fall and Shoah), which visualized the suffering in a less striking and more abstract way, were not able to do so.Moreover, the results did not confirm From a theoretical perspective, the results showed that formal criteria of the excerpts we used, such as historical footage, language etc. (see Table 1) as well as the way of portraying the victims (stereotypes, victimization, and individualization) played a subordinate role for the emergence of group-based shame.The reason for this might be the greater social distance between the recipients on the one hand and the victim out-group on the other hand.

The Perpetrator Side
In contrast to these rather distal factors concerning the victims, that is the out-group, the more proximal factors portraying the perpetrator in-group and perceived relevance had a stronger impact on the development of filminduced group-based shame.
In particular, the emotionalized way of portraying the perpetrator side in the Holocaust established emotional closeness to the perpetrators (compassion, sympathy, and understanding), which again had a strong shame-reducing effect.It gave room for a perpetrator perspective in parallel to a victim's point of view (Dengler, 2010).This strategy of parallelization (Befreiung) lowered the severity of the moral failure through a partly victim status of the German side (see Table 1), that is through externalizing the responsibility of the individual German to social elites and the political system.Thus, it decreased the necessity to be ashamed for the historical injustice.Beside the emotional closeness to the perpetrators, the subjectively perceived relevance of the Holocaust had the strongest shame-reducing impact on group-based shame (see path 4b in Figure 1).The excerpt Panorama "immunized" against this powerful defense strategy by presenting the complicity of the German population in the Holocaust and the denial of today's generations to accept this involvement.Thus, the clip systematically deconstructed the self-serving claims of eyewitnesses and made National Socialism and the destruction of the Jews understandable as a mass phenomenon.By making clear references to contemporary Germany, the clip lowered the opportunity to downplay the relevance of the Holocaust for today and thus had a shame-enhancing effect.

Limitations
more toward the individual recipients rather than cinematic intentions as the current shame-items might suggest, and therefore they might reveal different findings.Such a wider understanding of group-based shame would additionally offer the opportunity to measure long-term effects of films on emotions of shame and complement the results of the current study on immediate responses, and would allow for pre-testing these emotions.In the present study, it was not possible to pretest the items before watching the films due to their specific reference to films.
Real-life film clips comprise complex sets of stimuli and thus come with a higher risk of missing relevant variables compared to artificially created stimuli for lab settings.This can threaten the internal validity of the postulated relationships.While this study cannot rule out with certainty that the observed effects are not (partly) triggered by not analyzed aspects of the films, the theory-based selection of the content analytical variables, the deductive approach using a well-elaborated coding procedure, and the high intercoder reliability should alleviate this threat.
Besides considering long-term effects and a content free category of reference, future research should integrate further developments in the field such as the differentiation of different forms of moral shame and image shame (Allpress et al., 2014) or between appraisals and emotions (Gausel et al., 2012).The distinction between victimand perpetrator-focused ways of dealing with past injustice might reveal interesting differences in specific facets of these emotions.In order to validate the results of the current study, the research question should be applied to another case of historical injustice.The Holocaust implies specific and unique characteristics such as the discourse in current Germany, which for example affect victim-blaming or a perpetrator-victim reversal.The generalizability of the findings should be verified in different contexts which are debated more controversially in the public.
In sum, in the specific case of the debate about the Holocaust, this study was able to show the effects of influencing mechanisms of the cinematic stimulus qualities on different ways of dealing with the issue and their partly mediating effects on group-based shame.The partly counter-intended effects regarding film-induced emotions point out the great significance of which portraying strategies are chosen in the media, especially of the perpetrator in-group.
The study closes a gap in our knowledge about the influencing mechanisms of history transmission by the media, which is gaining in importance with the increasing temporal distance from the historical events.

Notes
i) Still, this film does not represent a neutral reference category; thus we gave preference to effect coding of the film variables instead of dummy coding.
ii) The question if national identification moderates the relationship between experienced social identity threat and group-based shame (see among others Doosje et al., 1998;Klein et al., 2011;Roccas, Klar, & Liviatan, 2006) was not central in this paper.Nevertheless, we evaluated this effect, controlling for gender and age.The explanatory power of the regression model did not significantly increase when adding moderator terms (film * national identity), F increase (5, 210) = 1.36, p = .24.In both models, the extent of shame was lower when the participants identified more strongly with Germany, as postulated by Lickel et al. (2004).These findings are further in line with Doosje et al. (1998), who only found differences when individuals were faced with ambivalent, that is positive and negative consequences of their in-group's wrongdoing.Due to its clear and one-sided division into perpetrators and victims, probably the singularity of the Holocaust is only to a limited extent suitable for testing the moderating function of national identification.
iii) The total direct effect of the influence of all the films on the extent of group-based shame, holding constant all control and mediator variables, was significant (R 2 direct = .25,adjusted R 2 direct = .21,F total (12, 211) = 5.85, p direct < .001).For details see Table A3 in Appendix.

.
in a victim role only to distinguish them from a role for example as German citizen.Percentage of scenes in which German characters, such as politicians or citizens, were portrayed.CAV6: Occurrence of Germans c portrayed role of Germans.CAV7: Role of Germans Portrayal of Germans as individuals to distinguish them from a uniform collective.as undifferentiated and heteronomous characters like a people receiving orders from elites.Percentages refer to the total amount of scenes of a film.a As the only category, CAV2 refers only to the number of scenes in which Jews are actually portrayed.b Special case Free Fall 1: Jews were portrayed as victims only to a low degree (see CAV2).c Non-Jewish. d Of contemporary witnesses.e df = 5; f In case of missing frequencies in some excerpts, Monte-Carloprocedure was applied: Number of replications = 2400; g Monte-Carlo-procedure was applied: Number of replications = 2470.***p ≤ .001.Chi 2 The contribution Holocaust -Die Lüge von den ahnungslosen Deutschen (Holocaust -The lie of the unsuspecting Germans) from the political TV magazine Panorama was shown in its full length.It addresses the complicity in and active support of the Holocaust by ordinary Germans using historical documents and expert witness testimony.
Loose's criticism (2009)  postulating that the use of historical footage from NS propaganda sources would reproduce stereotypes and lead to a devaluation of Jews.We found neither a direct effect of the clip historical stereotypes on victim blaming or distancing nor an indirect effect on film-induced shame.The missing effect of victim blaming on group-based shame might be ascribed to the special context of the Holocaust.The narrative of the Holocaust offers a clear assignment of perpetrator and victim roles, which made a perpetrator-victim reversal less likely in comparison to other, more complex cases of historical injustice.

Table 2
Sources of Stimulus Material a In minutes.b Northern German Broadcast (Norddeutscher Rundfunk).

Table 3
Hypothesized Effects of Film Excerpts on Film-Induced Defensiveness Strategies

Table 4
Intercorrelations of Dependent, Mediator, Film, and Control Variables