Randomness, Attributions of Arousal, and Belief in God

Beliefs in God, or similar spiritual forces, have permeated every culture the world has seen, past or present (Atran & Norenzayan, 2004). Although there are likely many reasons why such beliefs are so strongly held (Kirkpatrick, 1998; Norenzayan & Hansen, 2006), attempts to cope with perceptions of randomness may be a key factor. Randomness is presumed to be highly aversive (Pennebaker & Stone, 2004), and people will go to considerable lengths to reaffirm order in the face of evidence to the contrary (e.g., by blaming victims of random misfortune or seeing patterns in random arrays; Jost, Banaji, & Nosek, 2004; Lerner, 1980; Whitson & Galinsky, 2008; also see Heine, Proulx, & Vohs, 2006). Affirming the existence of a controlling God, therefore, may provide an excellent means for insulating oneself from the aversive arousal associated with randomness. However, no experimental test of this hypothesis exists. Park (2005) has suggested that traumatic events can strengthen belief in God because of the threat they pose to nonrandomness, but this correlational research (also see Laurin, Kay, & Moscovitch, 2008) focused only on negative events and assumed (rather than directly assessed) the role of randomness. Some research has manipulated self-conceptions that may be related to preserving beliefs in order (e.g., Kay, Gaucher, Napier, Callan, & Laurin, 2008; McGregor, Haji, Nash, & Teper, 2008), but none has investigated randomness directly, and none, crucially, has assessed the role of arousal in generating between-condition differences in belief in God. In this study, we employed a novel paradigm to test (a) whether direct manipulations designed to prime thoughts of randomness cause increased beliefs in supernatural sources of control (even when controlling for negative valence) and (b) whether this effect is due to arousal generated by thoughts of randomness. To heighten thoughts of randomness, we supraliminally primed half the participants with randomness-related words; the other half were primed with words matched in negative valence. To assess the role of arousal, we employed a misattribution paradigm (Zanna & Cooper, 1974), which involved requiring all participants to swallow a pill ostensibly containing an herbal supplement. Half the participants were told that the pill sometimes induces arousal as a side effect, and half were told that the pill has no side effects. Previous work has shown that the side-effect condition leads participants to attribute the cause of any experienced arousal to this salient source (Proulx & Heine, 2008; Zanna & Cooper, 1974). Hypothesizing that beliefs in supernatural control function, at least in part, to down-regulate the aversive arousal associated with randomness, we expected the randomness primes to increase beliefs in God, but only for those participants not given the opportunity to attribute the cause of their arousal to the ingested pill.