Jordan W. Moon, Adam B. Cohen, Kristin Laurin, David P. MacKinnon
5 min
Abstract
Religion makes unique claims (e.g., the existence of supernatural agents) not found in other belief systems, but is religion itself psychologically special? Furthermore, religion is related to many domains of psychological interest, such as morality, health and well-being, self-control, meaning, and death anxiety. Does religion act on these domains via special mechanisms that are unlike secular mechanisms? These could include mechanisms such as beliefs in supernatural agents, providing ultimate meaning, and providing literal immortality. We apply a critical eye to these questions of specialness and conclude that although it is clear that religion is psychologically important, there is not yet strong evidence that it is psychologically special, with the possible exception of its effects on health. We highlight what would be required of future research aimed at convincingly demonstrating that religion is indeed psychologically special, including careful definitions of religion and careful attention to experimental design and causal inference.
Sam: Health is the outlier, and it's worth being precise about why. Some longitudinal studies find that religious attendance predicts mortality outcomes even after adjusting for social support, health behaviors, and other known mediators. The effect doesn't fully disappear. The authors' interpretation is cautious—they suggest this might reflect the way religion bundles mechanisms at a scale and duration that secular equivalents haven't matched, or that it provides a form of sustained meaning-making that we don't yet have good instruments to measure.
Alex: So the residual health effect might not be evidence of something theologically unique—it might just be evidence that our secular control conditions are underpowered.
Sam: That's the more parsimonious read. We don't have a robust secular analogue for a lifetime of ritualized, community-embedded meaning-making. Until we can construct a control condition that mirrors that longitudinal depth, we can't cleanly separate "the sacred does something irreplaceable" from "we haven't built the right comparison yet." The health finding is real, but its interpretation is genuinely underdetermined by the current evidence.
Alex: That's a meaningful distinction for how you'd design the next generation of studies. You're not just adding a control group—you're trying to match the full social and temporal architecture of religious participation. [[RP_SECTION:comparative-functionalism-approach|Comparative functionalism approach]]
Sam: Exactly, and that's where the methodological bottleneck sits. The field has been comparing religious groups to secular individuals, which confounds group membership with belief content. The informative comparison is between a church and a secular organization offering equivalent longitudinal engagement, equivalent community obligations, equivalent ritual structure. When those comparisons exist, the religious premium tends to shrink. The agenda the authors are pushing is what they call comparative functionalism—stop asking whether religion is special in the abstract, and start asking how specific mechanisms perform across sacred and secular containers.
Alex: Which reframes the whole enterprise. Instead of "does religion do something unique," you're asking "which component of this bundle is doing how much work, and can it be replicated." [[RP_SECTION:future-research-directions|Future research directions]]
Sam: And that question is tractable in a way the original one isn't. Isolate reputation effects, test them. Isolate meaning-making, develop better instruments for it, test that too. The current literature's weakness is treating religion as a black box and then being surprised when the box contains ordinary parts. Even if religion turns out not to be special in the strong sense, it's clearly a highly optimized system—one that has been refined over millennia for sustaining large-scale human cooperation. Studying it as a model system for cultural evolution is likely to be more productive than continuing to chase the question of uniqueness.
Alex: It's a more honest framing, too. The question isn't whether religion matters—it clearly does—but whether the mechanisms behind it are exclusive to it.
Sam: Precisely. And the current evidence says: probably not, with health as the one domain where the jury is still out, and where better-designed studies could genuinely shift the answer. That's the work that remains.
Alex: Thanks for listening to ResearchPod.