jost_j_t_kay_a_c_thorisdottir_n_social_and_psychological_bas

sometimes defined in terms of their basis in faith. In this respect, religious thinking is challenging to sociology and psychology, for its epistemological basis is usually thought of as different from that of other belief systems. If religious belief is not typically based on direct empirical evidence, why are religious beliefs so strong and so widespread? Can mechanisms used for understanding other aspects of human nature be useful in understanding religious belief?

It is very likely that religious beliefs are multiply determined, having bases in many social and psychological mechanisms. Sociological scholars of religion have focused on the social functions of shared religious beliefs in promoting community and solidarity (Durkheim, 1995 [1912]). Indeed, religion has historically rivaled, if not bested, all other bases of group identity and membership in strength and influence (Christiano et al., 2002). Psychological researchers have also posited several factors that may drive various religious beliefs (e.g., Batson & Ventis, 1982; Gilbert, Brown, Pinel, & Wilson, 2000; James, 1902). More recently, evolutionary theorists have joined the fray, offering explanations of the possible biological evolution of predispositions toward religious thinking (e.g., Atran & Norenzayan, 2004; Boyer, 2001; Dennett, 2006; Sapolsky, 1997; Wilson, 2003).

Consistent with Voltaire’s famous claim, religious beliefs may to a large extent be the products of motivational factors. In this article, I argue that religious beliefs can be understood in part as a motivated psychological response to fear of death. Specifically, I argue that fear of death increases belief in an afterlife, and religious belief in general, as a result of a motivated psychological process. In turn, increased afterlife belief should serve to mollify mortality concerns. In the sections that follow, I first present two competing theoretical accounts of this relationship. Then, I review past correlational and experimental research on the theorized bidirectional relationship between fear of death and afterlife belief. I then present the results of two studies investigating the effects of fear of death on afterlife belief. Finally, I conclude with a discussion of these results, their implications, and limitations.

THEORY

Motivated Bases of Religious Belief

Recent research on political ideology has primarily explained the formation of political attitudes in motivational terms (e.g., Jost, Banaji, & Nosek, 2004; Jost et al., 2007). For example, politically conservative attitudes have been recently explained in terms of a variety of individual psychological motivations, such as needs for certainty, cognitive closure, unambiguous and just worldviews, and the need to manage mortality concerns (Jost, Glaser, Kruglanski, & Sulloway,

2003). Although this approach has increasingly come to dominate the explanation of political belief, comparatively less research has used the same approach to explain religious belief. Nonetheless, such an approach seems potentially very fruitful, and is presaged by the speculations of scholars for centuries. To address this gap, I present and test a theoretical explanation of afterlife belief based on individuals’ motivations to address their fear of death.

Humans are unusual among animals in that their self-consciousness allows them awareness of the inevitability of their own death. Research and theory suggest that thoughts related to one’s mortality are a significant source of anxiety (Becker, 1973; Wahl, 1959). If death awareness tends to promote fear and anxiety, then individuals will have strong motivations to adopt beliefs that serve to neutralize or moderate these responses, such as belief in life after death.

A multitude of scholars have cited fear of death as central to the existence of religious belief. For example, Malinowski (1965 [1935]) proposed, “Death, which of all human events is the most upsetting and disorganizing to man’s calculations, is perhaps the main source of religious belief.” William Durant (1954) reported that Schopenhauer described death anxiety as the “beginning of philosophy and the cause of religion.” Gordon Allport (1950) described religious belief as functioning to reduce the concerns brought about from anxiety over the knowledge of one’s own death. Lucretius (1947 [50 B.C.E.]) put it simply, “fear begets Gods.” Perhaps this function is why afterlife beliefs are so fundamental to religions, so much so as to become even part of the definition: “most religions have taught in one form or another that the ‘soul’ or ‘spirit’ of the individual does not perish when his body dies, but goes on living in another world” (Ducasse, 1961, p. 14).

Here, I present two theoretical accounts of the relationship between fear of death and afterlife belief. Although both accounts rely on psychological motivations to explain afterlife belief, they are based on distinct theoretical rationales. One account is based on a relatively simple and parsimonious application of motivated reasoning, the other on terror management theory (TMT). Although the simpler motivated reasoning account argues that fear of death promotes afterlife belief directly (because afterlife belief neutralizes the finality of death), TMT offers a more nuanced theoretical logic for the same prediction based on “cultural worldview defense.”

Motivated reasoning describes the tendency of individuals to come to conclusions that are desirable to them (Kunda, 1990). Motivational influences on human reasoning have been observed in a variety of forms, including motivated self-characterization (Dunning, Leuenberger, & Sherman, 1995), motivated group affiliation (Cialdini et al., 1976), motivated recall (McFarland & Buehler, 1997), motivated stereotyping (Sinclair & Kunda,

1999), and motivated political reasoning (Westen, Blagov, Harenski, Kilts, & Hamman, 2006). Despite these and other examples of judgments being biased by preferences, it is also worth noting that the effects of motivation have limits. People prefer plausible, justifiable conclusions and are often motivated by judgmental accuracy.

The logic of the parsimonious motivated reasoning account is as follows: Fear of death creates significant anxiety in individuals. Individuals are motivated to come to conclusions that avoid negative arousal states. Therefore, greater fear of death should lead to greater belief in ideas that offer an escape from death anxiety, such as belief in an afterlife. Further, afterlife belief should reduce fear of death. This reciprocal relationship between fear of death and afterlife belief is portrayed in Figure 10.1.

But, would motivated reasoning predict that fear of death is related to other religious beliefs, such as belief in God? Here, I see little in the way of a direct motivational relationship between mortality concerns and other religious beliefs (but see also Norenzayan & Hansen, 2006). However, belief in an afterlife is itself strongly associated with other religious beliefs, such as belief in God or other supernatural agents. As a result of cognitive consistency mechanisms (Festinger, 1954), individuals should generally hold associated religious beliefs at comparable levels. Thus, my application of motivated reasoning predicts that fear of death leads not only to greater afterlife belief, but also in turn to greater religious beliefs in general (e.g., belief in God). This relationship is also given in Figure 10.1.

Figure 10.1 Theorized bidirectional relationship between fear of death and belief in an afterlife predicted by basic motivated reasoning account.

An alternate theory that also predicts that fear of death leads to greater afterlife belief, although for different reasons, is TMT (Greenberg et al., 1990; Greenberg, Solomon, & Pyszczynski, 1997; Pyszczynski, Greenberg, & Solomon, 1999; Rosenblatt, Greenberg, Solomon, Pyszczynski, & Lyon, 1989). Briefly, TMT asserts that humans are strongly motivated toward selfpreservation. As a result, thoughts of death produce mortal terror. However, a person’s “individualized version of the cultural worldview” and self-esteem help to buffer against this anxiety by providing individuals a sense of value, order, and symbolic immortality (Greenberg et al., 1994). Because one’s cultural worldview provides an anxiety buffer against mortality concerns, the salience of mortal thoughts tends to lead to worldview defense (Rosenblatt et al., 1989). Examples include the tendency of individuals for whom thoughts of their own death have been made salient to exhibit increased ingroup affiliation and biases (Harmon-Jones, Greenberg, Solomon, & Simon, 1996), hostility and aggression toward outgroup members (McGregor et al., 1998), systemjustifying stereotypes (Schimel et al., 1999), and increased liking of fellow ingroup members who conform to cultural expectations (Greenberg et al., 1990).

Thus, TMT would also predict that fear of death leads to increased belief in an afterlife. To the extent that afterlife belief is part of an individual’s cultural worldview, this belief should be strengthened following mortality salience because an individual’s cultural worldview serves as a buffer against death anxiety. The theory’s argument for the relationship between fear of death and afterlife belief is also based on individuals’ psychological motivations to reduce fear of death. But, the theoretical argument is different because of the theorized mediating role of worldview defense.

PREVIOUS RESEARCH

Correlational

Perhaps because of the limitations of studying reciprocal relationships with cross-sectional research designs, correlational studies of the relationship between fear of death and afterlife belief are mixed in their results, variously suggesting a positive, negative, curvilinear, or no relationship between these variables. Several studies have shown that lower fear of death or death anxiety is correlated with religious belief or attendance (Alexander & Alderstein, 1960; Jeffers, Nichols, & Eisdorfer, 1961; Swenson, 1961; Martin & Wrightsman, 1965; Richardson et al., 1983). In a meta-analysis, Spilka, Hood, and Gorsuch (1985) found that in 24 of 36 studies, fear of death was lower among religious individuals.

Despite this, a variety of correlational studies are suggestive of a positive relationship between mortal fear and afterlife belief. For instance, research

shows that those with greater “death experience” (defined as the number and recency of deceased loved ones) exhibit greater levels of religious behavior and stronger religious views (Peterson & Greil, 1990). Also, a study of gay men with and without acquired immune deficiency syndrome (AIDS) showed that, among the men with AIDS, greater death anxiety was correlated with greater church attendance (Franks, Templer, Cappellety, & Kauffman, 1990). Studies of people reporting near-death experiences show that these individuals afterward tend to report greater belief in God, the afterlife, and reduced fear of death, although this research is based on small samples (Wulff, 1991). Research on Israeli soldiers in the 1982 Lebanon War shows that nonreligious soldiers reported greater fear of death, and greater responsiveness to higher death-risk experiences in war, than did nonreligious soldiers (Florian & Mikulincer, 1993). Anecdotal reports suggest greater religious observance and rates of baptism among U.S. soldiers awaiting deployment to Iraq (Finer & Baker, 2003).

Results of some studies are ambiguous in the direction of the observed effect or find no effect. In a cross-sectional survey study, Hoelter and Epley found that seven of eight dimensions of fear of death correlated with at least one religiosity measure, but that the correlation was at times positive and other times negative (1979). Other cross-sectional research has failed to find any correlation between religious belief and death anxiety (Aday, 1984; Christ, 1961; Lester, 1970). Also, although it is generally assumed that religious beliefs increase with age, a relationship possibly attributable to greater fear of death among the elderly, evidence also exists that age and afterlife belief are unrelated (Harley & Firebaugh, 1993).

Still other research has shown a more complex relationship moderated by the nature of individuals’ religious beliefs. Spilka and colleagues found that the intrinsically religious tended toward a benevolent view of the afterlife, whereas the extrinsically religious had a more negative view (1977). Others have found a similar pattern, including research showing that the intrinsically religious report less death concern (Kahoe & Dunn, 1975; Wulff, 1997), and greater fear of death has been reported among the extrinsically religious (Wulff, 1997). Leming (1980) found a curvilinear relationship between fear of death and religious belief, concluding that “religiosity may serve the dual function of afflicting the comforted and comforting the afflicted.”

One way to understand these diverse and apparently contradictory findings is to look closer at the characteristics of the studies. Based on the model presented in Figure 10.1, we would expect individuals high in fear of death to adopt greater afterlife beliefs, and then subsequently fear death less. This model would predict that longitudinal and experimental research will find a positive relationship (as fear of death leads to afterlife

belief), whereas cross-sectional studies should find a negative relationship (as adopted afterlife beliefs leave the individual less afraid of death).

This model is generally consistent with the past research presented here. The Spilka and colleagues’ (1985) meta-analysis indicates that cross-sectional research tends to find a negative relationship between religiosity and fear of death. Meanwhile, several of the studies that have found a positive relationship involve quasi-longitudinal designs in which individuals who should be higher in fear of death because of some shared, common factor are surveyed with respect to their afterlife or general religious beliefs.1 Research cited earlier, studying the attitudes of servicemen anticipating deployment, gay men with AIDS, and people with recently deceased loved ones, has found that members of these groups report greater afterlife belief than comparison groups. However, because these designs are nonexperimental, significant alternative explanations are possible for the findings. Thus, I next review relevant experimental research on this relationship.

Experimental

The research reviewed in the prior section was nonexperimental and therefore is limited in its ability to speak to the causal relationship(s) between individuals’ fear of death and belief in an afterlife. Because of the possibly bidirectional relationship between religious belief and fear of death, experimental approaches offer more promise for understanding how these concepts are related. Researchers have employed experimental designs to study both sides of the theorized reciprocal relationship between fear of death and afterlife belief portrayed in Figure 10.1.

Most applicable to the current investigation is experimental research on the effects of fear of death on afterlife belief. In a seminal study, Osarchuk and Tatz (1973) first classified participants as high or low in afterlife belief based on responses to one of two forms of an afterlife belief survey. Participants were then assigned to one of three conditions: (a) a Death treatment condition wherein participants watched a presentation of death-related imagery (e.g., photos of corpses and auto wrecks) while a tape played an exaggerated presentation on the high likelihood of death for individuals aged 18–22 set to dirge-like music; (b) a Shock threat treatment condition designed to increase

1 Note, however, that not all research finding a positive correlation involves quasi-longitu- dinal designs (e.g., Berman & Hays, 1973). Kurlychek found that afterlife belief was greater among those with greater fear of death of others, but not fear of one’s own death (1976). A classic interview study found greater reported fear of death among religious individuals (Faunce & Fulton, 1958).

fear, but not mortal fear, wherein participants spent a comparable amount of time anticipating participation in a study involving electric shocks; or (c) a Control condition in which participants played with a child’s toy for a comparable amount of time. Following the manipulation, all participants completed the other form of the afterlife belief survey.

Osarchuk and Tatz found the greatest change in afterlife belief among participants initially high in afterlife belief who were exposed to the Death threat condition. No other conditions of the study produced notable results. This is the only known experimental study to show an effect of fear of death on afterlife belief. However, Osarchuk and Tatz’s finding has not yet been replicated. In one subsequent study, Ochsmann (1984) had groups of theology students and students of other subjects fill out a questionnaire on death and dying (or not). Following this manipulation, all participants filled out a belief in afterlife scale. Participants did not show increased afterlife belief as a result of the mortality salience manipulation.

Other research suggests that fear of death may lead to greater religious beliefs besides afterlife belief, consistent with the motivated reasoning account given in Figure 10.1. Norenzayan and Hansen (2006) found that a mortality salience manipulation led to greater general religiosity and belief in God. Follow-up studies showed that study participants expressed greater belief in supernatural agents that were outside of their own cultural worldview (e.g., Buddha, Shamanic spirits), apparently contrary to the predictions of TMT (see Anson et al. this volume).

Additionally, Greenberg and colleagues (1995) showed that participants who wrote short essays about their death were more reluctant to make inappropriate use of a crucifix (for hammering) on a task, an effect suggestive of greater religiosity. Greenberg and colleagues interpret their results as supportive of TMT, since mortality salience led to apparently more respectful treatment of culturally significant objects in general, but this result could also be interpreted as consistent with the parsimonious motivated reasoning account presented here.2

Experimental research has also investigated other aspects of the motivated reasoning account of the relationship between fear of death and afterlife belief, specifically the possibility that afterlife belief serves to decrease fear of death. Friedman and Rholes (2007) tested the idea that religious belief functions to suppress thoughts of death among believers. They presented evidence of inconsistencies in the Bible to both fundamentalist and non-fundamentalist

2 Still other research has failed to find an effect of experimentally manipulating mortality sali-

ence on reported religious belief (Burling, 1993).

Christians. They then measured posttreatment accessibility of death-related thoughts via a word-stem completion task. Fundamentalist Christians presented with evidence of Biblical inconsistency showed greater death thought accessibility, but non-fundamentalist Christians were unaffected by the manipulation. The findings suggest that fundamentalist Christianity’s teachings function to keep death-related thinking out of consciousness.

Working in the TMT tradition, Dechesne and colleagues (2003) conducted a series of studies in which they successfully manipulated participants’ belief in an afterlife via an essay arguing that scientific research on near-death experiences supports (or contradicts) the existence of life after death. The result of the manipulation was to nullify the previously discovered link between mortality salience and self-esteem striving. Participants tended to positively distort negative personality feedback under mortality salience, but not if they had previously been assured that scientific research supports the existence of an afterlife. This research shows how belief in an afterlife can suppress effects from thoughts of death.

Taken together, experimental research relating fear of death and religious beliefs has produced results generally consistent with the parsimonious motivated reasoning account, although with some exceptions and ambiguities. Research has shown a positive effect of fear of death on afterlife belief (Osarchuk & Tatz, 1973), as well as belief in God and general religiosity (Norenzayan & Hansen, 2006), yet many questions remain unanswered. In particular, Osarchuk and Tatz’s effects were only observed among participants initially high in afterlife belief, and alternative explanations of their findings based on the accessibility of religious thoughts in the Death threat condition are possible. Most problematic, Osarchuk and Tatz’s findings have not been replicated, and have even been contradicted by subsequent research (Ochsmann, 1984). Thus, further, controlled experimental research is called for to more carefully assess the validity of the claim that fear of death increases afterlife belief.

Additionally, the theoretical explanation for the relationship between fear of death and religious belief remains ambiguous. For example, Norenzayan and Hansen found increased belief in culturally foreign supernatural agents, a contradiction of TMT’s predictions. On the other hand, Osarchuk and Tatz’s finding that only those initially high in belief in an afterlife responded to mortality salience with increased afterlife belief is consistent with TMT’s prediction of cultural worldview defense in reaction to mortality salience. The present studies are intended to address these ambiguities and extend research on the effects of fear of death on religious beliefs, especially belief in an afterlife, while testing a motivated reasoning account of the relationship against one based on TMT.

asking participants how religious they were on a 10-point scale ranging from “Extremely Religious” to “Not Religious at All.”4

After completing the demographic questionnaire, participants were asked to write short essays in response to two essay prompts on a computer. Half of the participants were randomly assigned to respond to essay prompts regarding death, and half were assigned to respond to essay prompts regarding watching television (Greenberg et al., 1994). The exact wording of the prompts was:

•Briefly describe the feelings and emotions that the thought of your own death (watching television) arouses in you.

•Please describe in as much detail as possible what your thoughts would be as you physically die (watch television).

After participants finished writing the short essays, they were given a survey with miscellaneous questions, including one asking “How likely do you think it is that there is life after death?” Participants responded on a 10-point scale ranging from “Extremely Unlikely” to “Extremely Likely.” The survey also asked participants to answer a series of 10 questions regarding characteristics of the afterlife adapted from the General Social Survey; for example, participants were asked how likely it was that the afterlife would be “A paradise of pleasure and delights.”

Finally, participants were debriefed regarding the true purpose of the study, paid, and thanked for their participation.

Results

Participants who wrote essays regarding death rated the existence of an afterlife as more likely (M = 7.30, SD = 2.53) than did participants who were assigned to write essays about television (M = 5.62, SD = 3.51), although this effect was only marginally significant (t = 1.84, p < .08). I next conducted a regression analysis of the effect of the essay manipulation on reported belief in the afterlife while controlling for participants’ pre-manipulation levels of reported religious belief. This analysis showed a highly significant effect of initial religiosity (b= .558, p < .001) and a significant effect of condition (b = .329, p = .01) on participants’ stated belief in an afterlife. The

4 Random assignment was successful in creating two groups that were roughly equal in their initial levels of religious belief. There was no significant difference in the reported level of religious belief, as measured on the demographic questionnaire prior to the manipulation, between participants assigned to the Death essay and those assigned to the Television essay condition ( p > .50).