Symmetries and asymmetries in the belief in a just world. Mariia Kaliuzhna. Personality and Individual Differences, Volume 161, July 15 2020, 109940.
https://doi.org/10.1016/j.paid.2020.109940
Abstract: Despite evidence to the contrary, people have a need to represent the world as a just and fair place where prosocial behaviour is rewarded and negative acts are punished. This cognitive bias is termed the Belief in a Just World (BJW). Previous research assumes the BJW to be symmetrical, i.e., one believes to the same extent that a negative action will be punished and that negative events in one's life are punishments for previous negative actions. Similarly, good deeds are expected to be rewarded and positive events are interpreted as rewards for previous positive actions. The present work tests this symmetry assumption. We show multiple asymmetries in the way people endorse positive and negative BJW statements for the past and the future that concern the self or other people. The results are discussed in terms of the interaction of BJW with other cognitive biases.
Keywords: Belief in a Just WorldJust dessertsKarmaOptimism biasSelf-serving bias
11. Discussion
One robust finding across all the surveys is the asymmetry between negative prospective and retrospective questions. Thus, participants more readily made the connection between doing a bad deed and getting retribution than between a negative event in one's life and a previous bad deed. Previous research shows that recollected memories are more detailed than imagined events (
D'Argembeau & Van der Linden, 2004;
Johnson et al., 1988), and that imagining the future requires a constructive and, possibly effortful, process (
Schacter & Addis, 2007;
Szpunar et al., 2007). One could thus hypothesise that the retrospective link would be easier to construct and we would observe the opposite pattern. However, when making the prospective link (i.e., do a bad deed – be punished) one does not necessarily have to imagine specific details for the future (i.e., what kind of punishment one would incur as a result of a negative deed), thus no additional cognitive effort is solicited. When in the retrospective condition, on the other hand, one needs to make a link between a negative event and a
specific negative act that could have caused it (i.e. go through memories of one's behaviour). It is then possible that the cognitive load of recollecting instances of past behaviour and their differing accessibility resulted in participants’ judging the retrospective condition as less plausible. Indeed, as mentioned in the introduction, previous research shows that ease of information retrieval and processing fluency lead to higher statement endorsement and confidence in them (
Bonnefond et al., 2014;
Cummins, 1995;
Oppenheimer, 2008;
Weingarten & Hutchinson, 2018).
The retrospective/prospective asymmetry for negative events could additionally reflect a self-serving bias. Previous research shows that positive autobiographical memories are reported twice as often as negative autobiographical memories (
Berntsen et al., 2011;
Seidlitz & Diener, 1993;
Walker et al., 2003). This memory bias could result in a representation of oneself as a positive person, making it difficult to recollect (or imagine, as in our task) a negative deed in the past. Thus, one would more readily concede that a negative deed could bring about a negative event, than that negative events represent
one's past negative acts, as the person represents themselves as a good person. The impact of the self-serving bias is particularly evident when comparing baseline positive and negative questions (questions 5 & 6 in Experiment 1 & 3, and questions 3 in Experiment 2). Indeed, participants rated the probability of positive events as higher than that of negative events, and this distinction was even more pronounced when the questions were framed in terms of deservingness – a result consistent with the optimism bias (
Sharot, 2011;
Sharot & Garrett, 2016).
These interpretations, however, do not appear plausible for our Experiment 4, where specific negative deeds and events were presented to participants, precluding memory search, and we still observed the same asymmetry. Similarly, Experiment 3 which tapped into BJW for others, should not have triggered a self-serving bias. However, the wording of the questions in Experiment 3 (“somebody”, “this person”) might have led participants to include themselves in the interpretation, engaging the self-serving bias (thus the only “clean”
other BJW condition would be in Experiment 4). Similar ambiguity has been reported for BJW-general vs BJW-self questionnaires, as the general BJW includes both oneself and other people (
Bartholomaeus & Strelan, 2019;
Lipkusa et al., 1996;
Wenzel et al., 2017).
The second result of the present experiments concerns the symmetry between positive prospective and retrospective conditions. Similar to the negative asymmetry, across experiments participants tended to give higher scores in the prospective than in the retrospective condition. This result, however, was less pronounced and consistent than for negative asymmetry: a significant difference in the mentioned direction was only observed in Experiment 2 and 3. A similar explanation as for the negative asymmetry could be invoked: the cognitive load of retrieving past instances of good behaviour that could be rewarded resulted in lower endorsement ratings for the retrospective condition, compared to the prospective condition. This result might be less pronounced than for the negative questions because the self-serving bias is not involved – that is, the questions did not provoke a need to foster a positive image of oneself (
Bradley, 1978;
Mezulis et al., 2004).
The third asymmetry we observed showed that participants tended to give higher scores for positive than negative statements (especially retrospective questions), a result that could again be due to the self-serving bias (i.e., the desire for a positive representation of oneself (
Bradley, 1978;
Mezulis et al., 2004), as well as the optimism bias – the overestimation of the probability of encountering positive events (
Morewedge et al., 2005;
Sharot, 2011;
Sharot & Garrett, 2016) and biased recollection of positive events (
Baumeister & Cairns, 1992;
Mischel et al., 1976;
Seidlitz & Diener, 1993). Interestingly, this result was absent in Experiment 4. We believe this result might highlight the implicit (irrational, preconscious)/explicit (rational, conscious) distinction in the BJW literature.
Lerner (1998) draws this distinction emphasising that most adults do not actually believe that the world is just, but have the need to reduce the anxiety of living in an unpredictable world by re-establishing justice through cognitive scripts (
Lerner, 1998). Thus, when the impact of the self-serving bias (implicit effect) is reduced (due to the use of situations not involving the self as in our Experiment 4) it is the prospective negative situation (do a bad deed – get punished) that is endorsed the most. In line with previous studies showing increased cognitive processing and significance of negative events ( HYPERLINK \l "bib6" Baumeister et al., 2001 ), the negative statements in Experiment 4 would thus appear more representative of the way participants perceive the world (i.e. as harbouring an increased probability for negative events).
Interestingly, the above asymmetries are represented in folk psychology by the notion of “karma” (that appears to have a predominantly negative connotation) and prospectively formulated proverbs (also appearing predominantly negative): he who digs a pit will fall into it; old sins cast long shadows; you reap what you sow; before you begin the journey of revenge, dig two graves; every man's sin falls on his own head; harm set, harm get; what goes around comes around; etc. It is unclear whether participants’ responses in our surveys were driven by the exposure frequency to such sayings, resulting in higher endorsement of similarly phrased prospective statements, or whether a common mechanism explains the preference for prospective moral representations. Importantly, although previous research shows that karma beliefs and BJW are distinct phenomena, it has also been reported that karma beliefs were specifically associated with perceived causality for negative events following bad behaviour ( HYPERLINK \l "bib74" White, Norenzayan & Schaller, 2019 ), i.e. our prospective condition. Although we did not explicitly measure karma beliefs in our participants, this could explain why in this experiment they scored the highest on this prospective condition.
Finally, the asymmetry for BJW between self and other is difficult to interpret based on our results. Previous work shows a pronounced asymmetry between self and other BJW, which relate to different aspects of behaviour and personality (
Bartholomaeus & Strelan, 2019;
Lipkusa et al., 1996;
Schindler et al., 2019;
Wenzel et al., 2017). As mentioned above, Experiment 3 might not be considered as solely representing judgements for other people, but also includes implicit judgement about the self. Comparing Experiment 1 and Experiment 3, we find that participants scored higher in Experiment 3 (other) in the deservingness survey for positive prospective, retrospective and baseline questions. Interestingly, Experiment 4, which we consider to be a “cleaner”
other condition, yielded the lowest endorsement rates for all the questions, as compared to the other three experiments. This is consistent with previous work that shows that participants report higher scores for self BJW than for general BJW (reviewed in
Bartholomaeus & Strelan, 2019;
Lipkusa et al., 1996;
Wenzel et al., 2017), meaning that self-relevant questions are endorsed to a higher degree than questions concerning others.
In accordance with previous research, our results indicate an association between BJW and religiosity (
Dalbert & Katona-Sallay, 1996;
Kunst, Bjorck & Tan, 2000;
Szmajke, 1991;
White et al., 2019): most questions in our surveys positively correlated with the belief in God. Disentangling the relationship between BJW and religiosity was not the primary aim of the present research; however, it confirms the previously reported association between the two.
Our results yield several implications. First, the assumed symmetry of the BJW, reported since Lerner first introduced the phenomenon, and up to the most recent papers in 2019 (
Armstrong, 2019;
Schindler et al., 2019;
Westfall et al., 2019), does not hold. Participants appear to endorse the prospective BJW “do -> get” more than the retrospective “got -> did” BJW. It would be of interest to examine what influences this asymmetry, and how it interacts with other domains, affected by the BJW, for example, life satisfaction (
Ucar, Hasta & Malatyali, 2019), antisocial behaviour (
Schindler et al., 2019;
Wenzel et al., 2017) or social judgements of value (
Alves, Pereira, Sutton & Correia, 2019).
Specifically, the finding that the retrospective negative condition was endorsed the lowest among the other target conditions may have some hope for the adverse effects of BJW, such as victim blaming. Making people's retrospective BJW explicit might help reduce their perceptions that someone else's apparently arbitrary misfortunes are caused by some previous misdeeds (
Landström et al., 2016;
Vonderhaar & Carmody, 2015).
Second, the above has larger implications for the information processing theories, which invoke implicational molecules (
Kruglanski, 2013;
Loken & Weyer, 1983;
Wyer, 2019;
Wyer & Carlston, 2018;
Wyer, 2006). Our direct survey approach shows that beliefs bound together by psychological implication are not necessarily all endorsed to the same degree, thus calling for a refinement of the theory.
Third, BJW seems to be asymmetrically affected by religiosity and the belief in karma (i.e. the prospective negative condition is more endorsed than others). Future research could test whether the asymmetries we observed are present across different cultures that endorse the karma belief to a different extent.
Finally, BJW is a complex belief that interacts with other cognitive biases and personality traits. Previous research in the latter domains has used computational modelling to identify the circumstances under which participants’ behaviour departs from optimality (
Eil & Rao, 2011;
Mobius, Niederle, Niehaus & Rosenblat, 2011). A similar normative approach could be beneficial to represent the complexity of BJW, an approach that surprisingly has not been applied in this field. Hierarchical Bayesian models could be used to formalise participants’ predictions about future events based on their prior knowledge, and how new information (confirming or disconfirming the priors) is integrated. Such an approach will shed new light on the debate about whether belief formation and change in neurotypical individuals follows the rules of Bayesian optimality (
Tappin & Gadsby, 2019).
Posts
