Abstract
We comment on work by Ginges, Hansen, and Norenzayan (2009), in which they compare two hypotheses for predicting individual support for suicide terrorism: the religious-belief hypothesis and the coalitional-commitment hypothesis. Although we appreciate the evidence provided in support of the coalitional-commitment hypothesis, we argue that their method of testing the religious-belief hypothesis is conceptually flawed, thus calling into question their conclusion that the religious-belief hypothesis has been disconfirmed. In addition to critiquing the methodology implemented by Ginges et al., we provide suggestions on how the religious-belief hypothesis may be properly tested. It is possible that the premature and unwarranted conclusions reached by Ginges et al. may deter researchers from examining the effect of specific religious beliefs on support for terrorism, and we hope that our comments can mitigate this possibility.
What role do religious beliefs play in supporting suicide terrorism? Ginges, Hansen, and Norenzayan (2009) suggest that “the relationship between religion and support for suicide attacks is real, but is orthogonal to devotion to particular religious belief, or indeed religious belief in general.” (p. 230). Ginges et al. argue that the relationship between religion and suicide terrorism is best explained by the role that religious collective rituals play in enhancing within-group cooperation and out-group hostility. The results they present confirm the predictive power of religious attendance vis-à-vis support for suicide terrorism. However, the authors' dismissal of religious beliefs themselves as an important factor in suicide terrorism is premature and unwarranted.
The “religious-belief hypothesis” (p. 224), as tested by Ginges et al., has little to do with actual beliefs and is concerned only with level of devotion to religious beliefs, whatever those beliefs are. This is a misrepresentation of the religious-belief hypothesis. The importance of devotion (a term which is not explicitly defined in the article) is introduced along with a citation to Harris (2004), although Harris clearly articulates the argument that the relationship between religion and suicide terrorism is a result of specific religious beliefs. For example, Harris (p. 52, italics in original) argues that “Beliefs are principles of action: … they are processes by which our understanding (and misunderstanding) of the world is represented and made available to guide our behavior.” The devotion to which Harris refers is to specific religious beliefs, namely a literal interpretation of the Koran (p. 45).
Ginges et al. argue that belief in the afterlife and in martyrdom fall conceptually within the religious-belief hypothesis, yet they inexplicably rely only on religious devotion, as measured by prayer frequency, to test the religious-belief hypothesis. Granted, the null relationship between prayer frequency and support for suicide terrorism does not support the religious-belief hypothesis, but to conclude that this provides a “disconfirmation of the religious-belief hypothesis” (p. 230) is unjustified. Regardless of the level of devotion, one who believes there is a moral obligation to kill infidels is likely to have a different attitude toward suicide terrorism than one who does not hold these specific beliefs (even if both subscribe to the same religion). This is just one example of a prediction that can be derived from the religious-belief hypothesis, and predictions like this should be empirically tested before concluding that the religious-belief hypothesis has been confirmed or disconfirmed.
A proper test of the religious-belief hypothesis would assess the actual beliefs held by participants. Ginges et al. assessed beliefs in Study 2, but erroneously used these data as a measure of personal support for suicide attacks. Participants were asked what they believed Islam's position was regarding suicide terrorism (i.e., whether Islam forbids, allows, encourages, or requires suicide attacks), failing to acknowledge that Islam is a religion and, therefore, not independent of religious beliefs. Ginges et al. assumed that those who responded “requires” were themselves supporters of suicide attacks. They assumed that one's religious beliefs determined their personal support of suicide terrorism – the very hypothesis they argue was disconfirmed. If responses to this question had been used as a predictor variable, Ginges et al. could have asked participants whether they personally support martyrdom attacks (as was done in Study 1), and then examined the relationship between one's religious beliefs and support for suicide terrorism. This type of examination would serve as a justifiable confirmation or disconfirmation of the religious-belief hypothesis.
Finally, there may be an important difference between willingness to engage in suicide terrorism and support for suicide terrorism. Ginges et al. are often clear that they are investigating the latter, but in concluding their article they write as if they have investigated the former. They conclude that “…the association between religion and suicide attacks is a function of collective religious activities that facilitate popular support for suicide attacks and parochial altruism more generally” (p. 230, italics added). The results they present do not address the association between religion and suicide attacks; instead, these results address the association between religion and support for suicide attacks. It is troubling that Ginges et al. make this terminological error in a sentence that concludes the article, as well as in the abstract: “Implications for understanding the role of religion in suicide attacks are discussed.” (p. 224, italics added).
In summary, although we appreciate Ginges et al.'s contribution, which provides evidence of the relationship between religious collective rituals and support for suicide terrorism, we do not agree with their assessment of the religious-belief hypothesis. One's specific religious beliefs may be related to support for – and willingness to engage in—suicide terrorism. The religious-belief hypothesis has not yet been disconfirmed, despite remarks to the contrary by Ginges et al. Prayer frequency is unrelated to support for suicide terrorism, but this does not address whether belief in the afterlife, or any other specific religious belief, is related to support for suicide terrorism. We hope that the unwarranted and unjustified conclusions reached by Ginges et al. do not discourage researchers from investigating the role that religious beliefs themselves may play in encouraging or supporting suicide terrorism.