Elsevier

Cognition

Volume 120, Issue 2, August 2011, Pages 177-185
Cognition

Religion and action control: Faith-specific modulation of the Simon effect but not Stop-Signal performance

https://doi.org/10.1016/j.cognition.2011.04.003Get rights and content

Abstract

Previous findings suggest that religion has a specific impact on attentional processes. Here we show that religion also affects action control. Experiment 1 compared Dutch Calvinists and Dutch atheists, matched for age, sex, intelligence, education, and cultural and socio-economic background, and Experiment 2 compared Italian Catholics with matched Italian seculars. As expected, Calvinists showed a smaller and Catholics a larger Simon effect than nonbelievers, while performance of the groups was comparable in the Stop-Signal task. This pattern suggests that religions emphasizing individualism or collectivism affects action control in specific ways, presumably by inducing chronic biases towards a more “exclusive” or “inclusive” style of decision-making. Interestingly, there was no evidence that religious practice affects inhibitory skills.

Introduction

Humans do not passively await external stimuli to execute reflex-like responses to them but, rather, actively explore their environment and carry out intentional actions to reach their goals. The recent two decades have shed considerable light on the control processes that adapt and configure the necessary cognitive systems to the goals and tasks at hand. The common idea is that control processes parameterize lower-level perceptual, memory, and action-related processes in such a way that relevant stimulus events are attended to and appropriate actions are performed (e.g., Logan and Gordon, 2001, Monsell, 1996). What is not yet known, and commonly not even discussed, is where the necessary parameter values are coming from and how control processes “know” which values to pick. In other words, it remains unclear who or what is controlling, or at least informing cognitive-control processes.

An obvious possibility is that learning processes play an important role in shaping cognitive control and, perhaps, in creating individual control profiles. For instance, there is increasing evidence that individualistic and collectivistic cultures lead to specific biases of visual attention towards local versus global aspects of visual displays and scenes, respectively (e.g., Boduroglu, Shah, & Nisbett, 2009). However, even though this might be taken to imply a systematic impact of culture-specific learning on cognitive control, the concept of culture is very general (e.g., Heine, 2008) and it is practically impossible to specify which behavioral rules are to be acquired in order to count as a member of a given culture. This renders it notoriously difficult to predict which control parameters might be affected and how. As argued by Colzato, van Beest, et al. (2010), religions are much better suited for that purpose. The rules defining them are often laid down in sacred, publicly accessible writings and interpreted to meet everyday-life requirements by experts and mediators (like rabbis, priests, and vicars), who are also actively involved in practicing procedures and rituals that are teaching the rules to active believers.

Given that religions differ with respect to the behavioral rules they developed, they would be expected to affect cognitive-control parameters in different, religion-specific ways (Hommel & Colzato, 2010). In particular, one would expect that practicing a religion biases the preferred values of control parameters to a range that is likely to generate behavior that is approved by one’s religious environment. First evidence that this might indeed be the case was reported by Colzato, van den Wildenberg, and Hommel (2008), who compared well-matched Dutch neo-Calvinists and Dutch atheists brought up in the same country and culture. A hallmark of Dutch neo-Calvinism is the concept of sphere sovereignty, which emphasizes that each sphere or sector of life has its own responsibilities and authority, and stands equal to other spheres. This concept has penetrated Dutch culture and caused a considerable “pillarization” (segregation) of Dutch society. Applying the concept to everyday life has established the idea that, in a nutshell, everyone should “mind his or her own business”. Colzato et al. speculated that shaping one’s behavior to fit with the sphere sovereignty concept would lead to an increased attentional focus on detail rather than the broader context. If so, neo-Calvinists should focus more on the local (rather than the global) aspects of perceived events than atheists—a hypothesis that was confirmed by means of Navon’s (1977) global–local task. Consistent with the idea that religion biases individuals towards particular attentional control parameters, Colzato, van Beest, et al. (2010) compared Italian Roman Catholics with Italian seculars and Israeli Orthodox Jews with Israeli seculars. Given that Catholicism and Judaism emphasize social solidarity, Catholics and Orthodox Jews were expected to show a larger global precedence effect than otherwise comparable nonbelievers—which they did. Along the same lines, Taiwanese Buddhists, notorious for their emphasis on “compassion” with the physical and social context, showed a larger global precedence effect than Taiwanese Atheists (Colzato, Hommel, van den Wildenberg, & Hsieh, 2010) while Dutch neo-Calvinists exhibited greater difficulty to attend to two successive visual events than Dutch Atheists (Colzato, Hommel & Shapiro, 2010).

Taken together, these studies show that particular religions seem to exert rather specific effects on information processing by biasing attention towards local or global stimulus aspects. In the present study, we asked whether this is a specific perceptual-attentional effect or a more general bias of cognitive control. Control processes should be able to target several processing stages but not just the selection of a particular level of visual stimulus features. This is why we were interested to see whether religious practice is also associated with differences related to rather “late” processing stages, like response selection and response inhibition—processes that are separable from input-selection processes targeted by the global–local task (Hommel, 1997, Johnston et al., 1995).

A particularly well-suited task to tap into response selection is the Simon task, which can be assumed to provide a rather pure measure of the emergence and resolution of response conflict (for overviews, see Hommel, 2011, Proctor, 2011). In this task, participants carry out spatially defined (e.g., left and right) responses to non-spatial stimulus features (e.g., blue and green color patches) presented on randomly varying locations (e.g., left and right). Performance is better with spatial stimulus–response correspondence, that is, if the stimulus location happens to match with location of the correct response (Simon & Rudell, 1967)—the so-called Simon effect. The Simon task, and conditions with stimulus–response non-correspondence in particular, introduce a high degree of response conflict, as evident from frequent response errors and electrophysiological observations (e.g., Hommel, 1996). Let us now consider how this uncertainty might be resolved and how this resolution might be modulated by religious practice.

Fig. 1 indicates a common way neural decision-making is modeled. Making a decision between the alternative responses A and B (such as a left or right response) is commonly assumed to involve competition between the representations of the alternatives, as indicated by the mutual inhibitory links in the figure (connection 2), and some sort of top-down support for the alternative that fits the current goals best (connection 1; for a review, see Bogacz, 2007). If stimulus and response correspond in terms of space, there is no conflict whatsoever and the representation of the correct response should be the most activated response alternative. In the case of non-correspondence, however, both response alternatives will be activated, one through the “intentional” route (connection 1) and the other through the spatial stimulus code (Kornblum, Hasbroucq, & Osman, 1990). Mutual inhibition (connection 2) will reduce the total activation of the correct response alternative, at least as long as the incorrect alternative is active, and thereby delay the correct response or even induce an error. This problem is resolved the earlier and the more likely the more top-down support of the correct alternative (connection 1) is provided.

Let us now consider the possibility that individuals can differ with respect to the strength of both top-down support and mutual (local) inhibition. Evidence for this possibility comes from research on the cognitive effects of bilingualism. Other than monolinguals, bi- and multilinguals are continuously facing the problem of cognitively keeping the languages they master apart, so to prevent mixing up words from different languages in the same sentence or utterance (Bialystok & Craik, 2010). Colzato, Bajo, et al. (2008) suggested that this might lead to a relatively “exclusive” configuration of attentional control parameters, which amounts to a relatively strong top-down support for words from the currently spoken language (connection 1) and relatively strong local inhibition between response alternatives (connection 2), such as words from different languages. If that configuration would become chronic, so the suggestion of Colzato, Bajo, et al. (2008), it might generalize to other, non-language tasks. Indeed, bilinguals have been reported to show a reduced Simon effect (Bialystok, Craik, Klein, & Viswanathan, 2004)—which fits with the idea that bilinguals benefit from stronger top-down support and/or more local inhibition between response alternatives.

Applying these considerations to Calvinists results in exactly the same prediction: If we consider that Calvinism is associated with or even favors a local attentional focus over the processing of the broader context (Colzato, van den Wildberg, et al., 2008), and if we assume that this induces a chronic, generalizable bias towards a more “exclusive” response control configuration, Calvinists should have acquired a comparatively strong top-down support route (connection 1) and more pronounced local inhibition (connection 2).1 If so, Calvinists should show a smaller Simon effect than Atheists. In Experiment 1, we tested this prediction by comparing Dutch Calvinists and Atheists in a standard Simon task.

We also considered an alternative theoretical account of conflict resolution in the Simon task. Even though the top-down bias represented by connection 1 in the figure could explain why and how people are able to pick the correct response even in the face of response competition, alternative accounts have been suggested (e.g., Dempster, 1992, Harnishfeger, 1995). In particular, Ridderinkhof (2002) has considered the possibility that incorrect responses are directly suppressed. Hence, rather than supporting and selectively strengthening appropriate responses, an inhibitory system might prevent inappropriate responses from interfering and being executed (connection 3). If so, a possible difference in the size of the Simon effect between Calvinists and Atheists might not be due to stronger/more selective top-down support but rather reflect more efficient response inhibition. Given that religions like Calvinism provide rather strict rules to regulate everyday behavior, including the prohibition of numerous activities, it is not unreasonable to consider that practicing these religions might lead to a general improvement of inhibition skills. Indeed, McCullough and Willoughby (2009) have claimed that religious people may be better in inhibiting “wrong thoughts” than nonbelievers and are thus less likely to commit crimes and misconduct. To assess this possibility, we also included a comparison of Calvinists and Atheists in the Stop-Signal task developed by Logan and Cowan (1984). This task requires the intentional suppression of already planned actions and can thus be considered a relatively direct measure of inhibitory abilities (as indicated by connection 3). If we would find an effect of religion on the Simon effect and if this effect would be accompanied by a similar effect in the Stop-Signal task, this would suggest an account of the former in terms of inhibition skills (connection 3). In contrast, if he would find an effect on the Simon task but no effect in the Stop-Signal task, this would suggest an account in terms of selective top-down support (connections 1 and 2).

Section snippets

Experiment 1

In Experiment 1, we compared the performance of Dutch neo-Calvinists and (non-baptized) Atheists, brought up in the same country and in the same cultural setting, in the Simon task (Simon & Rudell, 1967) and the Stop-Signal task (Logan & Cowan, 1984). If Calvinists are having a chronic bias towards a relatively “exclusive” configuration of attentional control parameters, we would expect them to show a less pronounced Simon effect than Atheists. If they do, a similar effect pattern in the

Experiment 2

The observation of a reduced Simon effect in Calvinists suggests that religious practice helps managing interference from competing response alternatives. We have argued that this effect should be specific to Calvinists, because their believe system emphasizes individualism and focusing attention on one’s own societal pillar, which in turn should induce a rather “exclusive” cognitive-control configuration. Other interpretations are possible, however. For instance, it might be that religious

General discussion

The aim of our study was to test whether religious practice can not only affect spatial (Colzato et al., 2010, Colzato et al., 2008) and temporal (Colzato, Hommel & Shapiro, 2010) characteristics of stimulus selection but also control processes devoted to action regulation, such as response selection and response inhibition. We expected that the emphasis of Calvinism on segregation and detail would induce a particular configuration of control parameters that does not only favor the processing

Acknowledgments

We thank Marlies van Bochove, Elize Vlainic and Maud Grol for their enthusiasm and invaluable assistance in recruiting, testing the participants of this study and collecting the data. The research of LSC and WPMvdW is supported by the Netherlands Organization for Scientific Research (NWO).

References (34)

  • E. Bialystok et al.

    Bilingualism, aging, and cognitive control: Evidence from the Simon task

    Psychology and Aging

    (2004)
  • A. Boduroglu et al.

    Cultural differences in allocation of attention in visual information processing

    Journal of Cross-Cultural Psychology

    (2009)
  • L.S. Colzato et al.

    How does bilingualism improve executive control? A comparison of active and reactive inhibition mechanisms

    Journal of Experimental Psychology: Learning, Memory, and Cognition

    (2008)
  • L.S. Colzato et al.

    Buddha as an eye opener: A link between prosocial attitude and attentional control

    Frontiers in Psychology

    (2010)
  • L.S. Colzato et al.

    Religion and the attentional blink: Depth of faith predicts depth of the blink

    Frontiers in Psychology

    (2010)
  • L.S. Colzato et al.

    Impaired inhibitory control in recreational cocaine users

    PLoS ONE

    (2007)
  • L.S. Colzato et al.

    Losing the big picture: How religion controls visual attention

    PLoS ONE

    (2008)
  • Cited by (32)

    • How metacontrol biases and adaptivity impact performance in cognitive search tasks

      2019, Cognition
      Citation Excerpt :

      Neuroscientific evidence suggests that such metacontrol biases (to use the terminology suggested by Hommel & Colzato, 2017) are regulated through the interplay of frontal and striatal dopaminergic pathways (Cools & D’Esposito, 2011; Cools, 2015) and/or the interplay between dopaminergic D1-receptors and D2-receptors (Durstewitz & Seamans, 2008). Humans do not only show intraindividual variability in their metacontrol state bias, but they also differ systematically in their individual metacontrol default or trait bias: Differences in both genetic setup (in genes relevant for dopaminergic processing) and cultural background have shown to be associated with particular biases towards persistence or flexibility (Hommel, Colzato, Scorolli, Borghi, & van den Wildenberg, 2011; for a review, see Hommel & Colzato, 2017). The persistence/flexibility tradeoff is thus likely to emerge from some interplay between inter- and intraindividual (i.e., trait and state) metacontrol biases, and it was this interplay that we aimed to investigate in the present study.

    • The social transmission of metacontrol policies: Mechanisms underlying the interpersonal transfer of persistence and flexibility

      2017, Neuroscience and Biobehavioral Reviews
      Citation Excerpt :

      The metacontrol state model suggests that the degree to which goals and stimuli contribute to action control is dependent on the metacontrol setting, and we have discussed evidence showing that religion indeed determines the degree of the Simon effect. Interestingly, the mentioned study comparing neo-Calvinists, atheists, and Roman Catholics on the Simon task revealed that neo-Calvinists no longer showed a significant Simon effect (Hommel et al., 2011). This suggests that the presence and size of the Simon effect depends on metacontrol settings, which in turn fits with the observation that electrophysiological evidence of stimulus-induced response tendencies in the Simon task disappear if the stimulus is presented before the stimulus-response mapping (Valle-Inclán and Redondo, 1998).

    • Between Persistence and Flexibility: The Yin and Yang of Action Control

      2015, Advances in Motivation Science
      Citation Excerpt :

      According to the MSM, Calvinists would be expected to show particularly low sensitivity to task-irrelevant information while Catholics should show particularly high sensitivity. Hommel, Colzato, Scorolli, Borghi, and van den Wildenberg (2011) tested this prediction in a Simon task that assesses the degree to which response selection is affected by irrelevant stimulus information. While Dutch Calvinists showed a less pronounced Simon effect than matched Dutch atheists, Italian Catholics showed a more pronounced Simon effect than Italian atheists—which confirms expectations.

    • A neurocognitive model of ideological thinking

      2021, Politics and the Life Sciences
    View all citing articles on Scopus
    View full text