When Past Group Events and Identities Define the Present: Effect of Perceived Collective Continuity on Defensive Behaviors of the French In-Group

Abstract

Previous studies have highlighted the benefits of perceived collective continuity. However, to the authors’ knowledge, none have considered the negative effects of such perceptions when they concern a negative past or take the form of a break with a positive past. The authors therefore conducted three studies to examine the influence of perceived continuity (or a break) with positive versus negative events or identities on French in-groups’ defensive behaviors (i.e., perceptions of and attitudes toward refugees and intention to engage in collective actions). They expected to observe the positive impact of a positive (rather than negative) past continuity and a negative (rather than positive) past break. The results of Experiment 1 partially confirmed this hypothesis, as individuals who identified strongly expressed a greater intention to engage in collective actions when they perceived continuity with positive past events. Similarly, participants were more opposed to the reception of refugees when they perceived continuity with a positive past French identity (Experiment 2). Finally, high-identifier participants who perceived a break with a negative past identity expressed greater opposition to the reception of refugees and saw them more as a threat (Experiment 3). The authors discuss the importance of considering the emotional valence of past group memories for the continuity literature and a better comprehension of actual in- and between-group dynamics.

Keywords

perceived collective continuity collective memory national identification in-group defensive behaviors

The most important events in the memory of a group are often those that the current group members have not experienced (Paez & Gonzalez, 1997). Both the repression and the recall of the memory of a past event can influence attitudes toward society (Hirst & Manier, 2008). We can assume that this is made possible by the perception of continuity. Continuity can be defined here as the perception of a link between the past, present, and future (Rutt & Löckenhoff, 2016). For a decade now, this diachronic (i.e., concerning changes over time) perspective has been studied in social psychology (Smeekes & Verkuyten, 2013) in terms of the concept of perceived collective continuity (PCC; Sani et al., 2007).

PCC—namely, the perception of a strong link between a group’s past, present, and future (Sani et al., 2007)—can be broken down into two dimensions: (1) the perception of a solid, permanent, and causal link between different historical periods or events, giving rise to a coherent group history (historical or narrative continuity), and (2) the perception of a strong and rooted group culture, incorporating norms, values, beliefs, attitudes, and traditions (cultural or essentialist continuity; Sani et al., 2007; Smeekes & Verkuyten, 2015). As the elements on which individual continuity rests are held in autobiographical memory (Bluck, 2015), we can, by analogy, assume that the elements on which collective continuity rests are held in collective memory.

Collective memory is a form of memory that is shared by a group and is of central importance for the social identity of its members (Halbwachs, 1950; Jetten & Wohl, 2012; Licata & Klein, 2005; Tavani et al., 2017; Wertsch & Roediger, 2008). First described by Halbwachs (1950), it gives human memory a social framework that makes sense in the life of a group (Romero et al., 2012). Collective memory plays an important role in defining the individual or collective self (Jetten & Wohl, 2012; Romero et al., 2012; Tavani et al., 2017), as it helps individuals and groups to retrieve past experiences for identity purposes or to put the present into perspective (e.g., judging out-groups; Doosje et al., 1999; Liu & Hilton, 2005). Presumably, therefore, collective memory influences the decisions, attitudes, and behaviors of the members of a group with a salient social identity.

Social identity can be defined as that part of the individual’s concept of self which arises from the awareness that the individual belongs to one (or more) social group(s), as well as the value and emotional meaning linked to this belonging (Tajfel, 1978). Social identity theory is based on three assumptions: (1) individuals strive to maintain or achieve a positive social identity; (2) a positive social identity is largely based on favorable comparisons that can be drawn between individuals’ own group (in-group) and a relevant external group (out-group); and (3) when they are dissatisfied with their social identity, individuals either try to leave their current in-group and join a more positively distinct group or attempt to give their current in-group a more positive image according to the group’s permeability (Tajfel & Turner, 1979).

Research has shown that social identity affects many group processes, including decision-making, performance, and negotiation (e.g., Terry et al., 2000). Collective memory influences social identity and the processes associated with it, such that individuals who share a common collective memory are perceived of as being similar, and thus as potential members of the same in-group (Tavani et al., 2017). Collective past behaviors and events stored in collective memory may, among other things, feed, strengthen, inhibit, or preserve individuals’ social identity. Collective memory can therefore serve to define a social identity, category, or group (e.g., Licata & Klein, 2005; Tavani et al., 2017), and maintain a positive identity, as well as be used as a group justification (Licata & Klein, 2005; Rimé et al., 2015). We can assume that it is through PCC that collective memory influences individuals, thereby helping to maintain a positive social identity.

The results of early studies of PCC supported this assumption, highlighting its benefits in terms of identification, a sense of entitativity, collective self-esteem (Sani et al., 2007), and social well-being (Sani et al., 2008). This initial research mostly adopted an intragroup perspective, but more recent studies have extended their focus to intergroup dynamics. Their results show that PCC increases defensive reactions such as rejecting social change and opposing stigmatized out-groups (Smeekes & Verkuyten, 2014a, 2014b).

To date, researchers have seldom considered the negative effects of PCC when it concerns a negatively valenced past (event, or identity) or when it takes the form of a break or rupture with a positively valenced past (event, or identity), even though emotion is key to understanding memory. Group-based emotional preferences are linked to emotional experience (Porat et al., 2016), and group history is made up of traumatic and glorious experiences in equal measure. A recent study showed that PCC, as opposed to perceived collective discontinuity or break (PCB), only reduces perceptions of identity threat when it is perceived to concern a positive past (Roth et al., 2017). This result calls into question the notion that PCC plays a consistently beneficial role, and suggests that in-groups should distance themselves from past wrongdoings, and feel that they have changed for the better, in order to satisfy their members’ need for a positive identity (Sahdra & Ross, 2007). As yet, scholars have had to make artificial inductions, giving neutral events an emotional valence (Roth et al., 2017), as it is impossible to find two different historical episodes (positive vs. negative) with the same time lines (i.e., temporal distance can be a confounding variable). We, however, had the opportunity to conduct our work in a national context (France) that possesses just such an ecological event: World War II.

World War II is a particularly salient event in most western countries’ (like France) collective memory (Giner-Sorolla et al., 2020; Hanke et al., 2015; Liu et al., 2005). For the French, this period can be seen as either positive—when it refers to the French Resistance and the people who fought against Nazism—or negative—when it refers to French collaboration with this totalitarian regime (Christofferson & Christofferson, 2006). They are very aware of these two sides as a result of both commemoration (Keith, 2012) and education (French Ministry of Education, 2015, 2019a, 2019b, 2020). Consequently, references to this period may simultaneously induce a negative and a positive past, depending on which facet is made salient.

We used this unique event to test the impact of PCC or PCB (i.e., the perception that there is no link between the past, present, and future of a group) on both intragroup and intergroup dynamics, in line with Smeekes and Verkuyten’s (2013, 2014a, 2014b, 2015) research on in-groups’ defensive attitudes. Our first objective was to test their impact on the intention to engage in collective actions—that is, actions aimed at improving the conditions of the entire group, be they normative (i.e., moderate acts in line with social norms, such as peaceful demonstrations) or nonnormative (i.e., radical acts such as violent behavior or terrorism; Jiménez-Moya et al., 2015)—bearing in mind the current climate of political protest in France with the yellow vest movement (i.e., demonstrations and actions protesting against government policies to gain social improvements). Several analogies have been drawn between this movement and great protest movements of the past, such as the French Revolution (e.g., Bourquin, 2019) and the Resistance (e.g., Leclerc, 2019). We also wanted to examine how these perceptions affected opposition to the reception of refugees and the perceived threat they pose, in the context of the refugee crisis that has beset European Union member states in recent years (Badea et al., 2017), as analogies have also been drawn between the treatment of refugees today and that of Jews in World War II (e.g., Lorriaux, 2015; Sorman, 2015). These analogies, together with the fact that continuity with a positive past reduces the social identity threat (Roth et al., 2017), led us to think that perception of continuity with the Resistance might lead people to act according to this movement’s fighting spirit by (1) engaging more in collective actions and (2) accepting refugees.

Present Research

We conducted this study during the first semester of 2019, when France was witnessing the beginning of the yellow vest movement and was still in the throes of the refugee crisis.

In the first of three experiments, we aimed to show that narrative PCC with the Resistance had a positive impact on the in-group’s defensive behaviors, as reflected in (1) decreased opposition to refugees, (2) a lower perceived threat from refugees, and a greater intention to engage in (3) moderate and (4) radical collective actions. We expected to observe the opposite pattern for narrative PCC with collaboration. This effect would be greater when participants identified strongly with their in-group, as weak identifiers would derive fewer benefits from the stability afforded by PCC (Jetten & Wohl, 2012; Roth et al., 2017), and would then feel less threatened by PCB.

In Experiment 2, we tested essentialist PCC on the same variables as in Experiment 1. We hypothesized that essentialist PCC with Resistance fighters would have an impact on the in-group’s defensive behaviors, decreasing (1) opposition to refugees and (2) their perceived threat, and increasing the intention to engage in (3) moderate and (4) radical collective actions. We expected to observe the opposite pattern for PCC with collaborators.

Finally, as scholars have mostly studied PCC (e.g., Sani et al., 2007; Smeekes & Verkuyten, 2015), we asked ourselves whether, in certain cases, individuals may benefit from PCB with their in-group’s past. Based on the results for Experiments 1 and 2, we decided to examine (Experiment 3) whether essentialist PCB with Resistance fighters had an impact on the in-group’s defensive behaviors. We expected to observe (1) decreased opposition to refugees and (2) lower threat perception, but no impact on the intention to engage in (3) moderate and (4) radical collective actions. We expected to observe the opposite pattern for essentialist PCB with collaborators. We then discussed the contribution of this research to the PCC literature and current understanding of the role of PCC in collective-memory functioning.

Experiment 1: Effect of Narrative PCC With a Positive Versus Negative Past on the In-Group’s Defensive Behaviors

In Experiment 1, we decided to focus on the narrative aspect of history, as it is frequently used when individuals (mostly politicians) recall the past by referring to glorious (e.g., the French Revolution and the “Trente Glorieuses” postwar prosperity) or darker periods of history (e.g., the World Wars and the recession; Mols & Jetten, 2014). We therefore induced PCC with the Resistance or collaboration, and examined its impact on the in-group’s defensive attitudes.

Method

Participants and Procedures

G*Power analysis (Faul et al., 2007) indicated that we would need a minimum sample size of 88 to ensure satisfactory power. We ran the analysis based on an estimated small effect size of F² = 0.10, α = .05, 1 − β = .90, for a linear regression analysis with two predictors. This analysis served for each of the three experiments. The initial total sample included 216 participants. We only retained data from those who stated that they were French (controlling for nationality, places of birth and residence, and native language) and over 18 years of age, and who did not deviate by more than one standard deviation from the mean response time. The final sample therefore included 169 participants: 93 in the Resistance condition (70.97% women, M_age = 32.10, SD = 13.30) and 76 in the collaboration condition (64.47% women, M_age = 32.70, SD = 13.10). The participants were recruited on social media and asked to complete an online questionnaire. More specifically, after giving their informed consent, the participants provided sociodemographic information (sex, age, education level, and socioeconomic status) and read a fictional article on “The history of France” linking the Resistance (see Appendix 1) or collaboration (see Appendix 2) to recent historical events. The participants were invited to illustrate with two examples this continuity, and then to respond to scales measuring continuity and emotions. To gauge the in-group’s defensive behaviors, they successively responded to scales measuring the intention to engage in moderate and radical collective actions, opposition to the reception of refugees, and the perceived threat posed by refugees. The participants also completed a measure of identification with the national in-group. Finally, they read a debriefing document that explained the purpose of this study.

Measures

We performed a separate principal component analysis (with parallel analysis as a criterion; Horn, 1965) for each measure after running general and item measures of sample adequacy (Hutcheson & Sofroniou, 1999) and Bartlett’s sphericity test. A 7-point Likert-type scale ranging from 1 = totally disagree to 7 = strongly agree was used for each measure. Every scale in the present study was subjected to these analyses and schemas.¹

Emotions About the Past

The participants indicated the intensity of their feelings toward France and the French during World War II on a list of 12 positive and negative emotions. The first component grouped together four positive emotional states: pride, tenderness, joy, and recognition (α = .827). The second brought together five negative emotions: shame, regret, disgust, fear, and ungratefulness (α = .803). The mixed, nostalgia, and melancholy items were removed because they did not load sufficiently (i.e., they were less than .30) on either component.

Collective Actions

The participants had to complete an adapted French version of the 10-item scale measuring collective intentions (Becker et al., 2011; Tausch et al., 2011) on two dimensions: moderate actions (seven statements) and radical actions (three statements). The higher their score, the greater the participants’ intention to engage in moderate (α = .856) or radical (α = .824) actions.

Opposition to Reception of Refugees

The participants were asked to complete an adapted (for the refugee context) and abridged (from 11 to 5 items) version of the Opposition to Immigration measure (Jetten & Wohl, 2012; e.g., “The presence of refugees in France will increase unemployment”). The higher their score, the more the participants opposed the reception of refugees in France (α = .949).

Perceived Threat From Refugees

The participants indicated their degree of agreement with nine statements relating to the presence of refugees and its impact on French lives. The higher the score, the greater the perceived threat posed by refugees in France (α = .953).

National Identification

Social identification with the national in-group was assessed with five items (e.g., “I am proud to be French”) inspired by previous studies (e.g., Doosje et al., 1999). The higher the score, the more strongly the participants identified with the French in-group (α = .913).

Analysis

First, we conducted descriptive statistics and correlation analysis to evaluate the strengths of the linear relationships among all the variables. Then, we tested the interaction between the conditions and national identification using regression analysis (controlling the participants’ age and sex) on each dependent variable. When the moderating effect was significant, the statistical significance of the simple effects (Aiken et al., 1991) of the predictor variable was assessed. The analyses followed the same format in each experiment.

Results

Manipulation Check

The participants felt more negative emotions, d = 0.972, t(167) = 4.100, p < .001, and less positive emotions, d = −0.923, t(167) = −4.191, p < .001, in the collaboration condition than in the Resistance condition.

Intention to Engage in Moderate Collective Actions

There was an interaction between condition and identification with the French nation, b = 0.244, t(162) = 2.040, p = .022, 95% CI [0.008, 0.480], R² = .331 (Figure 1). When participants in the Resistance PCC condition identified strongly with their in-group, they had a greater intention to engage in moderate collective actions, b = 0.612, p = .013. We observed the opposite effect for weak identifiers, but this effect was not significant, b = −0.135, p = .602.

Intention to Engage in Radical Collective Actions

We did not observe any interaction between conditions and identification with the French nation, b = 0.185, t(162) = 1.091, p = .277, 95% CI [−0.150, 0.521], R² = .103. There was no main effect of the condition, b = −0.589, t(162) = −0.665, p = .507, 95% CI [−2.337, 1.159], R² = .103. There was, however, a main effect of the level of identification on the intention to engage in radical collective actions, b = 0.187, t(162) = 2.181, p = .031, 95% CI [0.018, 0.356], R² = .103, such that the more strongly the participants identified with their national in-group, the greater their intention to engage in radical collective actions.

Opposition to Reception of Refugees

There was no interaction between conditions and identification with the French nation, b = −0.081, t(162) = −0.406, p = .686, 95% CI [−0.474, 0.312], R² = .081. Likewise, there was no main effect of the condition, b = 0.950, t(162) = 0.916, p = .361, 95% CI [−1.098, 2.998], R² = .081. The effect of the level of identification on opposition to refugees was not significant either, b = 0.173, t(162) = 1.724, p = .081, 95% CI [−0.025, 0.371], R² = .081.

Perceived Threat From Refugees

We did not observe any interaction between conditions and identification with the French nation, b = −0.110, t(162) = −0.618, p = .259, 95% CI [−0.461, 0.241], R² = .065. Likewise, we did not observe a main effect of the condition, b = 1.124, t(162) = 1.215, p = .226, 95% CI [−0.703, 2.951], R² = .065. The effect of the level of identification on the perceived threat from refugees tended toward significance, b = 0.166, t(162) = 1.851, p = .066, 95% CI [−0.011, 0.342], R² = .065, such that the more strongly the participants identified with their national in-group, the more they felt threatened by refugees.

Discussion

We partially validated our hypothesis, as narrative PCC with the Resistance (a positive past) led to a greater intention to engage in moderate collective actions among strong identifiers. However, we did not find this effect on other variables. We therefore decided to go further by conducting a second experiment to explore the possible role of essentialist PCC in intragroup and intergroup dynamics, in line with previous research (e.g., Smeekes & Verkuyten, 2015).

Experiment 2: Effect of Essentialist PCC With a Positive Versus Negative Past on the In-Group’s Defensive Behaviors

To pursue our investigation on the impact of positive versus negative PCC on group dynamics, we used the same design as in Experiment 1 but this time focused on essentialist PCC. As a reminder, essentialist continuity refers to the perception of a link between the past, present, and future values, traditions, attitudes, and behaviors of a given group. We therefore adapted the continuity induction such that the participants perceived continuity with Resistance fighters (rather than the Resistance) or collaborators (rather than collaboration).

Method

Participants and Procedure

The total sample included 356 participants. The participants were selected using the same criteria as in Study 1. The final sample included 289 participants: 141 in the Resistance fighters PCC condition (81.56% women, M_age = 32.50, SD = 14.10) and 148 in the collaborators PCC condition (79.05% women, M_age = 27.80, SD = 11.80). The participants were also recruited on social media and gave the same consent and information as in the previous experiment. Following this, they read a fictional article linking Resistance fighters (see Appendix 3) or collaborators (see Appendix 4) to the behavior of French people today. The participants were then invited to illustrate with examples this continuity and to complete the same scales as in the previous study.

Measures

The participants completed the same measures as in Experiment 1. We then performed the same analysis. Positive (α = .821) and negative (α = .805) emotions about the past and national identification (α = .898) were assessed using the same scale. Moderate (α = .815) or radical (α = .825) collective actions were measured using the same 10-item measure. The items for opposition to the reception of refugees (α = .926) or perceived threat from refugees (α = .967) were similar to those used in Experiment 1.

Results

Manipulation Check

The participants felt more negative emotions, d = 0.062, t(287) = 0.366, p = .714, and less positive emotions, d = −0.296, t(287) = −1.178, p = .076, in the collaborators condition than in the Resistance fighters condition, but the results were near significant or non-significant. However, when we focused only on certain feelings—like recognition, d = −0.559, t(287) = −2,512, p < . 001—we observed a significant difference. We decided to pursue our exploration based on this.

Intention to Engage in Moderate Collective Actions

There was no interaction between conditions and identification with the French nation, b = 0.013, t(282) = 0.131, p = .896, 95% CI [−0.164, 0.188], R² = .269. We did not observe a main effect of the condition, b = 0.381, t(282) = 0.829, p = .408, 95% CI [−0.524, 1.287], R² = .269. There was, however, a main effect of the level of identification on the intention to engage in moderate collective actions, b = 0.372, t(282) = 8.062, p < .001, 95% CI [0.281, 0.462], R² = .269, such that the more strongly the participants identified with their national in-group, the greater their intention to engage in moderate collective actions.

Intention to Engage in Radical Collective Actions

We did not observe any interaction between conditions and identification with the French nation, b = −0.035, t(282) = −0.251, p = .802, 95% CI [−0.311, 0.240], R² = .074. We did not observe a main effect of either condition, b = 0.338, t(282) = 0.469, p = .639, 95% CI [−1.079, 1.755], R² = .074, or the level of identification with the French in-group, b = 0.077, t(282) = 1.060, p = .290, 95% CI [−0.066, 0.219], R² = .074.

Opposition to Reception of Refugees

There was no interaction between conditions and identification with the French nation, b = −0.209, t(284) = −1.507, p = .133, 95% CI [−0.481, 0.064], R² = .076. We did, however, observe a main effect of the condition, b = 1.456, t(282) = 2.045, p = .042, 95% CI [0.054, 2.859], R² = .076 (Figure 2). The French were more opposed to refugees when continuity was perceived with the Resistance fighters. Moreover, the effect of the level of identification with the French in-group tended toward significance, b = 0.133, t(282) = 1.859, p = .064, 95% CI [−0.008, 0.273], R² =. 076, such that the more strongly the participants identified with their national in-group, the more they opposed the arrival of refugees.

Perceived Threat From Refugees

Again, we did not observe any interaction between conditions and identification with the French nation, b = −0.122, t(282) = −0.971, p = .332, 95% CI [−0.371, 0.126], R² = .047. Likewise, we did not observe a main effect of the condition, b = 0.839, t(282) = 1.294, p = .197, 95% CI [−0.438, 2.116], R² = .047. There was, however, a main effect of the level of identification on the perceived threat from refugees, b = 0.140, t(282) = 2.147, p = .033, 95% CI [0.012, 0.267], R² = .038, such that the more strongly the participants identified with their national in-group, the more they felt threatened by refugees.

Discussion

We observed a main effect of essentialist PCC on opposition to the reception of refugees, but the results did not confirm our hypothesis, as essentialist PCC with Resistance fighters increased rather than decreased opposition. This result can be explained in the light of social identity theory (Tajfel & Turner, 1979). Insofar as the participants were seeking to maintain a positive identity, they may have rejected any relevant out-group members rather than showing them acceptance. This rejection could be explained, first, by the fact that the participants may have felt that their in-group values were threatened by these new invaders. Another explanation is that the analogy between the Resistance fighters’ behaviors and the acceptance of refugees was not made, even if this link has been a popular one spread in the media (e.g., Lorriaux, 2015; Sorman, 2015). Moreover, we only observed an impact of essentialist PCC on opposition to refugees, whereas, in Experiment 1, narrative PCC influenced the participants’ intention to engage in moderate collective actions. On the strength of these results, we surmised that narrative PCC (or PCB) mainly affects intragroup dynamics, whereas essentialist PCC (or PCB) may have a greater impact on intergroup dynamics.

Experiment 3: Effect of Essentialist PCB With a Positive Versus Negative Past on the In-Group’s Defensive Behaviors

We focused our third experiment on PCB and considered solely the essentialist aspect of collective memory. If PCC could have a positive impact for individuals or groups (e.g., decrease negative attitudes toward stigmatized groups; Licata et al., 2012; Smeekes & Verkuyten, 2014a), we surmised that PCB could also have this potential, if it involved a break with the collaborators’ (negative) past. Distancing themselves from collaborators’ behaviors may help current in-group members to preserve a positive social identity. We therefore predicted that we would observe more rejection of refugees but no effect on collective action, as the first two experiments had shown that the essentialist aspect of the past had an impact on intergroup dynamics (opposition to the reception of refugees), while the narrative aspect had more of an impact on intragroup dynamics (the intention to engage in collective actions).

Method

Participants and Procedure

The total sample included 146 participants. The participants were selected according to the same criteria as in Experiments 1 and 2. The final sample included 124 participants: 63 in the Resistance fighters PCB condition (73.02% women, M_age = 28.60, SD = 13.60) and 61 in the collaborators PCB condition (68.85% women, M_age = 27.50, SD = 11.00). The participants were recruited through the same process as in the previous experiments. After reading a fictional article highlighting the absence of a link between Resistance fighters (or collaborators; see Appendices 5 and 6) and the behavior of French people today, the participants were invited to illustrate this break with examples, and to complete the same measures as in the first two experiments.

Measures

The participants completed the same measures as in Experiments 1 and 2. Positive (α = .841) and negative (α = .750) emotions about the past and national identification (α = .884) were assessed using the same scale. Moderate (α = .841) or radical (α = .876) collective actions were measured using the same 10-item measure. The items for opposition to the reception of refugees (α = .927) or perceived threat from refugees (α = .948) were also similar to the ones used in Experiments 1 and 2.

Results

Manipulation Check

In line with Experiment 2, the participants felt less recognition, d = −0.679, t(122) = −1.902, p = .060, in the collaborators condition than in the Resistance fighters condition.

Intention to Engage in Moderate Collective Actions

There was no interaction between conditions and identification with the French nation, b = −0.016, t(118) = −0.104, p = .918, 95% CI [−0.321, 0.289], R² = .222. We did not observe a main effect of the condition, b = 0.039, t(118) = 0.048, p = .962, 95% CI [−1.553, 1.631], R² = .222, but there was a main effect of the level of identification on the intention to engage in moderate collective actions, b = 0.401, t(118) = 5.211, p < .001, 95% CI [0.248, 0.553], R² = .222, such that the more strongly the participants identified with their national in-group, the greater their intention to engage in moderate collective actions.

Intention to Engage in Radical Collective Actions

There was no interaction between conditions and identification with the French nation, b = 0.179, t(118) = 0.781, p = .436, 95% CI [−0.275, 0.633], R² = .166. We did not observe a main effect of the condition, b = −1.503, t(118) = −1.255, p = .212, 95% CI [−3.874, 0.868], R² = .166, but there was a main effect of the level of identification on the intention to engage in radical collective actions, b = 0.378, t(118) = 2.429, p = .017, 95% CI [0.051, 0.505], R² = .166, such that the more strongly the participants identified with their national in-group, the greater their intention to engage in radical collective actions.

Opposition to Reception of Refugees

There was an interaction between conditions and identification with the French nation, b = −0.585, t(118) = −2.687, p = .008, 95% CI [−1.016, −0.154], R² = .218 (Figure 3). Thus, in the Resistance fighters PCB condition, there was no significant effect, b = 0.103, p = .461, whereas in the collaborators PCB condition, the more strongly the participants identified with the French nation, the more they opposed the reception of refugees, b = 0.688, p < .001.

Perceived Threat From Refugees

There was an interaction between conditions and identification with the French nation, b = −0.577, t(118) = −2.665, p = .009, 95% CI [−1.005, −0.148], R² = .150 (Figure 4). Thus, in the Resistance fighters PCB condition, there was no significant effect, b = 0.017, p = .902, whereas in the collaborators PCB condition, the more strongly the participants identified with the French nation, the more they perceived refugees to be a threat, b = 0.594, p < .001.

Discussion

Confirming the assumptions of Experiment 2, essentialist PCB only had an impact on intergroup dynamics when it interacted with national identification. Only a negative past had a significant effect in that the more the participants perceived a break with a collaborationist past, the more they rejected refugees. However, these results did not support our hypothesis, as we had predicted that we would observe these results when participants perceived a break with a positive past. One possible explanation is that a break with a negative past can also threaten a positive social identity. Thus, as suggested in the first two experiments, people benefit from continuity because it is the fundamental basis on which they construct their self (Smeekes & Verkuyten, 2015; Vignoles, 2011). Breaking with the past, even if it is a negative one, may therefore be perceived as a threat, such that to maintain a positive identity, individuals feel compelled to reject relevant out-group members.

General Discussion

In a series of three experiments, we used an ecological event to test how emotional valence influences the effect of PCC on both intragroup and intergroup dynamics. Taking advantage of the uniquely two-sided status of World War II in French history, we distinguished between (1) the narrative versus essentialist dimensions of collective memories of World War II; (2) the positive versus negative emotional valence of these memories; and (3) PCC versus PCB.

We began by focusing on the impact of narrative PCC (the Resistance vs. collaboration) on the in-group’s defensive behaviors. The participants had a greater intention to engage in moderate collective actions in the Resistance PCC condition. However, we only observed this pattern among stronger identifiers, and the results were not significant for weaker identifiers. Although there was no effect of the condition on intention to engage in radical actions, we did observe that the more people identified with their in-group, the greater their intention to engage in radical collective actions and the more they tended (a trend to significance) to perceive refugees as a threat. We decided to pursue our investigation of the impact of PCC by focusing on the essentialist aspect, developing a new hypothesis consistent with the results of this first experiment.

In the second experiment, we observed no main or interaction effect of essentialist PCC on the intention to engage in moderate or radical collective actions, and no effect of the level of identification. However, the participants were more opposed to the reception of refugees in the Resistance fighters PCC condition than in the collaborators PCC condition. This contradicted our assumption that people behave in a schema-coherent way—for instance, by helping a currently stigmatized group (refugees) just as past group members (Resistance fighters) assisted a previously stigmatized group (Jews), rather than rejecting the out-group. One possible explanation with regard to social identity theory is that to preserve past values (here, a positive link with the Resistance/Resistance fighters), in-group members will reject any out-group that is perceived to pose a threat to them (Smeekes & Verkuyten, 2013). Our results are therefore consistent with this study, as strong identifiers in the Resistance fighters PCC condition were more opposed to the reception of refugees. Furthermore, essentialist PCB (Experiment 3) had no impact on the intention to engage in collective actions. We did, however, find that stronger identifiers in the collaborators PCB condition perceived a greater threat from refugees and were more opposed to their reception. Distancing themselves from a negative past may have activated social identity differentiation processes (Ellemers, 1993), whereby people preserve and maintain their social identity by rejecting the relevant out-group (here, refugees). We did not find any impact of positive essentialist PCB. We can draw several general conclusions from these results.

First, PCC did not seem to follow a past–present directional attitude analogy perspective (e.g., helping a group that was stigmatized in the past helps a group that is stigmatized in the present) and instead supported social identity theory assumptions. Thus, to maintain or reinforce a positive social identity, people distinguish themselves from the out-group, and one of the strategies they use is to reject the members of that out-group (Smeekes & Verkuyten, 2013). Previous research (Roth et al., 2017) had led us to think that a negative break might be beneficial for social identity. We therefore predicted that it would result in lower perceived threat, but this prediction was not confirmed. It may be that this situation involved the perception of a perpetrator versus victim role (see Roth et al., 2017). Accordingly, if the in-group is viewed as the perpetrator of past wrongdoing, a perceived break with a positive element of the in-group’s past (or continuity with a negative element of the in-group’s past) may threaten its members’ identity and result in their turning against the out-group (Schori-Eyal et al., 2017). Future studies will therefore need to control the assessment of the group’s past roles, as well as the identity threat level.

Second, these results lead us to think that narrative PCC (and the events that support it) has a greater impact on intragroup dynamics (e.g., the intention to engage in moderate or radical actions), while essentialist PCC (and the norms and traditions that support it) has a greater impact on intergroup dynamics (e.g., opposition to the arrival of refugees or the perception of threat). This schema also needs to be explored further to understand the process behind it. For now, we can say that the distinction between these two aspects of PCC might be key to a better understanding of how the past shapes the present.

Third, social identity benefits from positive (narrative or essentialist) continuity and a negative break, leading in-group members to have a greater intention to engage in defensive in-group behaviors, particularly when they identify strongly with their in-group.

Limitations and Further Research

Researchers in the collective continuity field have often justified in-group treatment of out-group members by advancing that it maintains a positive identity (e.g., Roth et al., 2017). Nevertheless, there may be another explanation, drawing on both social identity theory and self-categorization theory. People who see their group status as both illegitimate and unstable may want to change it by adopting an individual or group mobility strategy, depending on the group’s (im)permeability (Bettencourt et al., 2001). Accordingly, when individuals perceive continuity with a positive past (or discontinuity with a negative past), it may lead them to discern a difference between their past and present status that they want to challenge by engaging in collective actions and/or rejecting refugees. Consequently, it seems important to consider status in further PCC research.

We therefore need to replicate our study results with other groups. To do so, however, our induction process needs to be improved by (1) removing everything that does not refer to World War II and (2) referring directly to the treatment of minorities in the Resistance (the protection of ethnic, sexual, and other minorities) and collaboration (persecution, insults, and the denunciation of minority members) conditions. These elements could make the connection between past and present contexts more salient, and might increase validity. In addition, a study considering both the valence and continuity of memories in a crossed 2 (past event: positive or negative) × 2 (continuity: break or continuity) design, and including new variables (e.g., social identity threat), could help us understand how perceived continuity impacts group processes. Furthermore, we are aware that using social media to collect data is an issue, as individuals who use social media may not be representative of the general French population. If possible, we recommend using survey institutes. Finally, we presume here that there is a collective consciousness developed through education and commemoration of World War II events. However, we did not consider the differences between participants that can arise from family history (Dresler-Hawke, 2005) or within-group differences such as gender or social class, which may lead to intersectional consciousness (Nair & Vollhardt, 2019). Both will need attention in the future. Moreover, we need to go further by finding another central two-sided historical memory in another country (either as old as France or with a younger history), another in-group context (e.g., women’s history, for which the role has changed across time), or another period of French history (e.g., the 18th century). The French in-group has a strong heritage of diversity owing to its long history of migration and its former colonial empire. Recent research in social psychology has looked at the consequences of colonialism, as this period has clearly shaped current intergroup relations and in-group identity and function among both the former colonized and colonizers (Bobowik et al., 2018). Two visions of colonial history can be identified (i.e., a positive vision highlighting the development aspect and a negative vision highlighting the exploitation aspect), and African and European groups have different social representations of colonialism (Licata et al., 2018). Initial research has already been undertaken among Belgian Congolese immigrants in order to identify acculturation dynamics (Figueiredo et al., 2018), but we could go further and look at the impact of collective memories of colonialism among African descendants in studies, adding a collective continuity perspective.

Conclusion

Our research opens up a new perspective on the theory of collective memory by explaining its effects on current attitudes according to the notion of PCC. To judge their current circumstances, individuals clearly have to bear in mind a time line and the history of their in-group (Jetten & Wohl, 2012). PCC is thus the process whereby these memories allow a positive social identity to be defined and maintained (Ellemers, 1993; Tajfel, 1978). Whereas early research suggested that PCC is always beneficial, in line with more recent studies, we have shown that scholars must consider the past elements to which the in-group refers in order to understand the positive or negative impact that this memory may have on its functioning. It is also important to distinguish the narrative aspect of the past from the essentialist aspect, as they seem to play different roles in group dynamics. The former has a greater impact on intragroup dynamics, while the latter has a greater impact on intergroup dynamics. We have demonstrated this by using memories of World War II, confirming that this historical event enjoys a unique status in French collective memory, such that it can be viewed either positively or negatively by current members of the French in-group. The impact of this two-sided memory shows that we need to find out which side people perceive more continuity with in order to better understand their perception, acceptance, or rejection of a stigmatized out-group, and their involvement in collective actions. Nevertheless, these conclusions can only be extended to other groups with a similar duality (i.e., a positive vs. negative aspect of the same past) in their collective memory. A cross-cultural study could be carried out by looking at colonial history or the history of women, who may have experienced this duality in many societies, both western and eastern.

Figure 1.

Interaction Effect Between Narrative Continuity (Positive vs. Negative) and National Identification on the Intention to Engage in Moderate Collective Actions.

Figure 2.

Impact of Essentialist Continuity on Opposition to the Reception of Refugees.

Figure 3.

Interaction Effect Between Essentialist Break (Positive vs. Negative) and National Identification on Opposition to the Reception of Refugees.

Figure 4.

Interaction Effect Between Essentialist Break (Positive vs. Negative) and National Identification on Perceived Threat From Refugees.

Footnotes

Acknowledgments

We are grateful to the undergraduates who helped to publicize these studies on social media and to the people who completed our questionnaires.

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 authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Haifat Maoulida

Jean Louis Tavani

Note

Appendix 1

Appendix 2

Appendix 3

Appendix 4

Appendix 5

Appendix 6

References

Aiken

L. S.

West

S. G.

Reno

R. R.

(1991). Multiple regression: Testing and interpreting interactions. SAGE.

Badea

Tavani

J. L.

Rubin

Meyer

(2017). Self-affirmation, political value congruence, and support for refugees. Journal of Applied Social Psychology, 47(7), 355–365. https://doi.org/10.1111/jasp.12441

Becker

J. C.

Tausch

Spears

Christ

(2011). Committed dis(s)idents: Participation in radical collective action fosters disidentification with the broader in-group but enhances political identification. Personality and Social Psychology Bulletin, 37(8), 1104–1116. https://doi.org/10.1177/0146167211407076

Bettencourt

B. A.

Charlton

Dorr

Hume

D. L.

(2001). Status differences and in-group bias: A meta-analytic examination of the effects of status stability, status legitimacy, and group permeability. Psychological Bulletin, 127(4), 520–542. https://doi.org/10.1037/0033-2909.127.4.520

Bluck

(2015). Going global: Functions of autobiographical remembering world tour. Memory, 23(1), 111–118. https://doi.org/10.1080/09658211.2014.937721

Bobowik

Valentim

J. P.

Licata

(2018). Introduction to the Special Issue: Colonial past and intercultural relations. International Journal of Intercultural Relations, 62, 1–12. https://doi.org/10.1016/j.ijintrel.2017.10.003

Bourquin

(2019, June 3). Avec les gilets jaunes, la Révolution française toujours aussi présente? [With the yellow vests, is the French Revolution still with us?]. France Inter. https://www.franceinter.fr/histoire/avec-les-gilets-jaunes-la-revolution-francaise-toujours-aussi-presente

Christofferson

T. R.

Christofferson

M. S.

(2006). France during World War II: From defeat to liberation. Fordham University Press.

Doosje

Spears

Ellemers

Koomen

(1999). Perceived group variability in intergroup relations: The distinctive role of social identity. European Review of Social Psychology, 10(1), 41–74. https://doi.org/10.1080/14792779943000017

10.

Dresler-Hawke, E. (2005). Reconstructing the Past and Attributing the Responsibility for the Holocaust. Social Behavior and Personality, 33(2), 133–148. https://doi.org/10.2224/sbp.2005.33.2.133

11.

Ellemers

(1993). The influence of socio-structural variables on identity management strategies. European Review of Social Psychology, 4(1), 27–57. https://doi.org/10.1080/14792779343000013

12.

Faul, F., Erdfelder, E., Lang, A. G., & Buchner, A. (2007). G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39(2), 175–191. https://doi.org/10.3758/BF03193146

13.

Figueiredo

Oldenhove

Licata

(2018). Collective memories of colonialism and acculturation dynamics among Congolese immigrants living in Belgium. International Journal of Intercultural Relations, 62, 80–92. https://doi.org/10.1016/j.ijintrel.2017.03.004

14.

French Ministry of Education. (2015). Bulletin officiel spécial n^o 11 du 26 novembre 2015 [Special official bulletin no. 11 of 26 November 2015]. https://www.education.gouv.fr/pid285/bulletin_officiel.html?pid_bo=33400

15.

French Ministry of Education. (2019a). Bulletin officiel spécial n^o 1 du 22 janvier 2019 [Special official bulletin no. 1 of 22 January 2019]. https://www.education.gouv.fr/pid285/bulletin_officiel.html?pid_bo=38502

16.

French Ministry of Education. (2019b). Bulletin officiel spécial n^o 8 du 25 juillet 2019 [Special official bulletin no. 8 of 25 July 2019]. https://www.education.gouv.fr/pid285/bulletin_officiel.html?pid_bo=39051

17.

French Ministry of Education. (2020). Bulletin officiel spécial n^o 1 du 6 février 2020 [Special official bulletin no. 1 of 6 February 2020]. https://www.education.gouv.fr/bo/20/Special1/MENE2003002A.htm?cid_bo=149192

18.

Giner-Sorolla

Hilton

Erb

H.-P.

Durante

Flaßbeck

Fülöp

Mari

Petrović

Sekerdej

Studzinska

Skitka

L. J.

Washburn

A. N.

Zadora

(2020). 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. Advance online publication. https://doi.org/10.1111/ajsp.12450

19.

Halbwachs, M. (1950). La mémoire collective (2nd ed.). Les classiques des sciences sociales. https://doi.org/https://dx-doi-org.web.bisu.edu.cn/doi:10.1522/cla.ham.mem1

20.

Hanke

Liu

J. H.

Sibley

C. G.

Paez

Gaines

S. O.

Moloney

Leong

C.-H.

Wagner

Licata

Klein

Garber

Böhm

Hilton

D. J.

Valchev

Khan

S. S.

Cabecinhas

(2015). “ Heroes” and “villains” of world history across cultures. PLOS ONE, 10(2), Article e0115641. https://doi.org/10.1371/journal.pone.0115641

21.

Hayton, J. C., Allen, D. G., & Scarpello, V. (2004). Factor Retention Decisions in Exploratory Factor Analysis: A Tutorial on Parallel Analysis. Organizational Research Methods, 7(2), 191–205. https://doi.org/10.1177/1094428104263675

22.

Hirst

Manier

(2008). Towards a psychology of collective memory. Memory, 16(3), 183–200. https://doi.org/10.1080/09658210701811912

23.

Horn, J. L. (1965). A rationale and test for the number of factors in factor analysis. Psychometrika, 30(2), 179–185. https://doi.org/10.1007/BF02289447

24.

Hutcheson, G. D., & Sofroniou, N. (1999). The Multivariate Social Scientist: Introductory Statistics Using Generalized Linear Models. SAGE Publications.

25.

Jetten

Wohl

M. J. A.

(2012). The past as a determinant of the present: Historical continuity, collective angst, and opposition to immigration. European Journal of Social Psychology, 42(4), 442–450. https://doi.org/10.1002/ejsp.865

26.

Jiménez-Moya

Spears

Rodríguez-Bailón

Lemus

S. de

(2015). By any means necessary? When and why low group identification paradoxically predicts radical collective action. Journal of Social Issues, 71(3), 517–535. https://doi.org/10.1111/josi.12126

27.

Keith

(2012). Forgetting the last big war: Collective memory and liberation images in an off-year anniversary. American Behavioral Scientist, 56(2), 204–222. https://doi.org/10.1177/0002764211419356

28.

Leclerc

(2019, September 7). Les gilets jaunes et le gaullisme [The yellow vests and Gaullism]. La TVA sociale. http://www.tva-sociale.fr/les-gilets-jaunes-et-le-gaullisme/#.XrVIv6gzY2w

29.

Licata

Khan

S. S.

Lastrego

Cabecinhas

Valentim

J. P.

Liu

J. H.

(2018). Social representations of colonialism in Africa and in Europe: Structure and relevance for contemporary intergroup relations. International Journal of Intercultural Relations, 62, 68–79. https://doi.org/10.1016/j.ijintrel.2017.05.004

30.

Licata

Klein

(2005). Regards croisés sur un passé commun: Anciens colonisés et anciens coloniaux face à l’action belge au Congo. In Sanchez-Mazas

L. Licata (Eds.), L’Autre: Regards psychosociaux (pp. 241–278). Presses Universitaires de Grenoble.

31.

Licata

Klein

Saade

Azzi

A. E.

Branscombe

N. R.

(2012). Perceived out-group (dis)continuity and attribution of responsibility for the Lebanese Civil War mediate effects of national and religious subgroup identification on intergroup attitudes. Group Processes and Intergroup Relations, 15(2), 179–192. https://doi.org/10.1177/1368430211414445

32.

Liu

J. H.

Goldstein-Hawes

Hilton

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 12 cultures. Journal of Cross-Cultural Psychology, 36(2), 171–191. https://doi.org/10.1177/0022022104272900

33.

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. https://doi.org/10.1348/014466605X27162

34.

Lorriaux

(2015, September 6). Crise des réfugiés et Shoah: Un utile et contestable point Godwin [Refugee crisis and the Holocaust: A useful and questionable example of Godwin’s law]. Slate. https://www.slate.fr/story/106463/refugies-comparaison-juifs-migrants-point-godwin?amp

35.

Nair, R., & Vollhardt, J. R. (2019). Intersectional Consciousness in Collective Victim Beliefs: Perceived Intragroup Differences Among Disadvantaged Groups. Political Psychology, 40(5). https://doi.org/10.1111/pops.12593

36.

Mols

Jetten

(2014). No guts, no glory: How framing the collective past paves the way for anti-immigrant sentiments. International Journal of Intercultural Relations, 43(A), 74–86. https://doi.org/10.1016/j.ijintrel.2014.08.014

37.

Paez

Gonzalez

J. L.

(1997). Social processes and collective memory: A cross-cultural approach to remembering political events. In Pennebaker

J. W.

Paez

Rimé

(Eds.), Collective memory of political events (pp. 147–174). Psychology Press.

38.

Porat

Halperin

Tamir

(2016). What we want is what we get: Group-based emotional preferences and conflict resolution. Journal of Personality and Social Psychology, 110(2), 167–190. https://doi.org/10.1037/pspa0000043

39.

Rimé

Bouchat

Klein

Licata

(2015). When collective memories of victimhood fade: Generational evolution of intergroup attitudes and political aspirations in Belgium. European Journal of Social Psychology, 45(4), 515–532. https://doi.org/10.1002/ejsp.2104

40.

Romero

J. J.

Mendoza

Bernal

S. A.

(2012). Mémoire des origines, identité du Mexique et identité du Mexicain. Bulletin de Psychologie, 65(1), 45–54. https://doi.org/10.3917/bupsy.517.0045

41.

Roth

Huber

Juenger

Liu

J. H.

(2017). It’s about valence: Historical continuity or historical discontinuity as a threat to social identity. Journal of Social and Political Psychology, 5(2), 320–341. https://doi.org/10.5964/jspp.v5i2.677

42.

Rutt

J. L.

Löckenhoff

C. E.

(2016). From past to future: Temporal self-continuity across the life span. Psychology and Aging, 31(6), 631–639. https://doi.org/10.1037/pag0000090

43.

Sahdra

Ross

(2007). Group identification and historical memory. Personality and Social Psychology Bulletin, 33(3), 384–395. https://doi.org/10.1177/0146167206296103

44.

Sani

Bowe

Herrera

(2008). Perceived collective continuity and social well-being: Exploring the connections. European Journal of Social Psychology, 38(2), 365–374. https://doi.org/10.1002/ejsp.461

45.

Sani

Bowe

Herrera

Manna

Cossa

Miao

Zhou

(2007). Perceived collective continuity: Seeing groups as entities that move through time. European Journal of Social Psychology, 37(6), 1118–1134. https://doi.org/10.1002/ejsp.430

46.

Schori-Eyal

Klar

Roccas

McNeill

(2017). The shadows of the past: Effects of historical group trauma on current intergroup conflicts. Personality and Social Psychology Bulletin, 43(4). 538–554. https://doi.org/10.1177/0146167216689063

47.

Smeekes

Verkuyten

(2013). Collective self-continuity, group identification and in-group defense. Journal of Experimental Social Psychology, 49(6), 984–994. https://doi.org/10.1016/j.jesp.2013.06.004

48.

Smeekes

Verkuyten

(2014a). Perceived group continuity, collective self-continuity, and in-group identification. Self and Identity, 13(6), 663–680. https://doi.org/10.1080/15298868.2014.898685

49.

Smeekes

Verkuyten

(2014b). When national culture is disrupted: Cultural continuity and resistance to Muslim immigrants. Group Processes and Intergroup Relations, 17(1), 45–66. https://doi.org/10.1177/1368430213486208

50.

Smeekes

Verkuyten

(2015). The presence of the past: Identity continuity and group dynamics. European Review of Social Psychology, 26(1), 162–202. https://doi.org/10.1080/10463283.2015.1112653

51.

Sorman

(2015, September 2). Les réfugiés d’aujourd’hui me rappellent mon père fuyant le nazisme [Today’s refugees remind me of my father fleeing Nazism]. Le Monde. https://www.lemonde.fr/idees/article/2015/09/07/les-refugies-d-aujourd-hui-me-rappellent-mon-pere-fuyant-le-nazisme_4744078_3232.html

52.

Tajfel

(1978). Interindividual behaviour and intergroup behaviour. European Monographs in Social Psychology, 14, 23–76.

53.

Tajfel, H., & Turner, J. (1979). An Integrative Theory of Intergroup Conflict. In W. Austin & S. Worchel (Eds.), The social psychology of intergroup relations (pp. 33–47). Brooks/Cole Pub. Co.

54.

Tausch

Becker

J. C.

Spears

Christ

Saab

Singh

Siddiqui

R. N.

(2011). Explaining radical group behavior: Developing emotion and efficacy routes to normative and nonnormative collective action. Journal of Personality and Social Psychology, 101(1), 129–148. https://doi.org/10.1037/a0022728

55.

Tavani

J. L.

Collange

Rateau

Rouquette

M.-L.

Sanitioso

B. R.

(2017). Tell me what you remember and I will know who you are: The link between collective memory and social categorization. Group Processes and Intergroup Relations, 20(1), 91–108. https://doi.org/10.1177/1368430215596076

56.

Terry

D. J.

Hogg

M. A.

White

K. M.

(2000). Attitude–behavior relations: Social identity and group membership. In Terry

D. J.

Hogg

M. A.

(Eds.), Attitudes, behavior, and social context: The role of norms and group membership (pp. 67–93). Lawrence Erlbaum Associates.

57.

Vignoles

V. L.

(2011). Identity motives. In S. Schwartz, K. Luyckx, & V. Vignoles (Eds.), Handbook of identity theory and research (pp. 403–432). Springer. https://doi.org/10.1007/978-1-4419-7988-9_18

58.

Wertsch

J. V

Roediger

H. L.

(2008). Collective memory: Conceptual foundations and theoretical approaches. Memory, 16(3), 318–326. https://doi.org/10.1080/09658210701801434