The many theories for the origin and persistence of religion fall into two classes: those that think that religion piggybacks on some aspect of human nature, usually evolved (credulity, need to attribute agency to natural phenomena, tendency to accept what parents or elders tell you, and so on), and those claiming that religion is “hardwired,” that is, we have genes that directly produce in us a propensity to apprehend and/or worship God. That claim is often the one adduced by religious people, since it feeds into the notion (e.g., Plantinga’s “sensus divinitatus”) that God instilled in us the need and desire to find Him.
I find the first class of theories more credible, but secular studies of religious belief have been bogged down by the fact that none of these theories are obviously testable by science. I have suggested several times on this site, though, that the “hardwired” theory is in principle testable: all you have to do is bring up children in an environment where they’re completely free of religious knowledge or influence, and see if they spontaneously come to conceive of (and maybe worship) a God. Unfortunately, that’s impossible, because we can’t do experiments with humans. And there’s virtually nowhere that one can raise a child without some exposure to religion.
But a new paper in Trends in Cognitive Science by Konika Banerjee and Paul Bloom, psychologists at Yale University, claim that the evidence is largely in—and doesn’t support the “hardwired” hypothesis. (The paper is just a two-page review, and is free at the link; reference is at bottom. The “Tarzan” reference in the title refers to the question of whether Tarzan, raised by apes, would come to believe in God.)
Their “evidence,” however, is pretty thin—so thin that I don’t think it shows anything.
Here’s their first assertion:
Consider belief in a divine creator. Young children are prone to generate purpose-based explanations of the origins of natural objects and biological kinds. They believe, for example,that clouds are ‘for raining’ and animals are ‘to go in the zoo’ . However, there is no evidence that children spontaneously come to believe in one or more divine creators. It is one thing, after all, to think about natural entities as intentionally designed artifacts of a sort; it is quite another to generate an enduring belief in invisible agents who have created these artifacts. Indeed, other studies ﬁnd that young children are not committed creationists; they are equally likely to provide explanations of species origins that involve spontaneous generation .
But what I see here is not evidence against hardwiring, but an absence of any evidence. And creationism is not equivalent to belief in a supernatural being that affects the world: one can be religious without being creationist.
Later, however, the kids do become creationists, but that’s imputed to cultural or parental indoctrination:
Older children, by contrast, do exclusively endorse creationist explanations. This shift to a robust creation is preference arises in part because older children are more adept at grasping the existential themes invoked by the question of species origins (e.g., existence and ﬁnal cause) and also because the notion of a divine creator of nature meshes well with their early-emerging teleological biases . However, these older children do not spontaneously propose novel divine creators. Instead, they adopt the particular creationist account that their culture supplies. This might be a singular God or multiple gods; it might be alien visitors or Mother Earth. If children are not exposed to such cultural beliefs,the explicit notionof an intentional creator might never arise.
Likewise, where’s the evidence that their newfound creationism comes from absorbing it from their culture, rather than appearing spontaneously as a product of their genes at a later age? No reference is cited.
The authors cite two more pieces of “evidence.” The first is this:
Some, such as Barrett , take children’s readiness to reason about life after death as evidence that they are ‘born believers’ in an afterlife.
This conclusion is probably too strong, however.There is no evidence that belief in the afterlife arises spontaneously in the absence of cultural support.For instance, research in rural Madagascar, where there is widespread belief in ancestral spirits, ﬁnds that the conception of an afterlife emerges in the course of development. Even if children are ‘natural-born dualists’ , this initial stance need not directly give rise to the afterlife beliefs that are characteristic of many of the world’s religions.
Again, this isn’t evidence for a lack of “hardwiring” as opposed to cultural inculcation. After all, beliefs or behaviors generated by genes needn’t appear at the outset of development. Interest in sex, for instance, is surely “hardwired,” but doesn’t appear till near puberty.
The authors’ final claim again appears to rest on an absence of evidence, not positive evidence:
Consider, as a test case,belief in multiple deities.This is the historically foundational religious stance, with monotheism a more recent invention . It would be striking support for the generativity position if children raised in monotheistic societies declared their belief in multiple gods. However, to our knowledge, they never do. They come to believe instead in the same singular omnipotent deity that everyone else believes in.
But this shows either that children are inculcated in one omnipotent deity—OR are genetically predisposed to believe in a single deity. (The thesis, after all, is that children believe in supernatural agents, not whether it’s one or many, and even “genetic” belief in a single deity can be altered by culture, just as a genetic predisposition to have sex can be altered by the availability of pornography.)
I have’t read the 14 papers cited by Banerjee and Bloom, but they don’t give citations for their major empirical conclusion. I am sympathetic to their idea that religion piggybacks on other evolved tendencies, simply because an individual reproductive advantage of believing in a God isn’t obvious to me, but the evidence so far is thin. Too thin, at least, to support the authors’ conclusion, from the abstract:
Drawing on evidence from developmental psychology, we argue here that the answer is no: children lack spontaneous theistic views and the emergence of religion is crucially dependent on culture.
Well, “crucially dependent on culture” can mean several things, and the “genetic” hypothesis might depend on cultural exposure too—just not cultural exposure to religion. One might simply need exposure to society and the environment.
I’m not sure how to discriminate among the many theories for the origin of religion, since it occurred in the distant past and has been culturally transmitted ever since. But at least one is testable in principle: that belief in a god or gods is hardwired, and will arise spontaneously—and in a similar form—in people who are never exposed to religion.
Yes, that’s testable in principle, but not in practice. We still remain profoundly ignorant of how religion came to dominate our species.
Banerjee, K., and P. Bloom. 2013. Would Tarzan believe in God?: Conditions for the emergence of religious belief. Trends Cognitive Sci. 17:7-8.