Is religion anthropomorphism?

Connor Wood

Emperor Moth

The verdict is in: we are our brains, roughly speaking. That is, according to modern neuroscience and cognitive science, our personalities, dreams, and experiences are all products of intensely complex interactions of the neurons in our craniums. You can disagree or agree with this claim, but nearly all experts who study the brain and mind are convinced of it. When it comes to things spiritual, the cognitive science of religion (CSR) is a field that tries to understand religious beliefs from within this naturalistic framework. And recently, one of the founding thinkers in CSR outlined a central claim in the field: religion is essentially about anthropomorphism, or the tendency for our brains to see persons in the world around us.

You’ve probably had an experience like this one: you’re walking down a dark street and you hear a noise behind you. Nervously, you turn your head to look over your shoulder, thinking it might be a denizen of the night who means to do you harm. But there’s no one there – the street is empty. Or this: you’re letting your mind wander one lazy afternoon at home, looking at nothing in particular, when you realize that the design of your living-room carpet has an old man’s face in it, bearded and wrinkle-browed – and he’s looking right at you.

These types of experiences, in which we erroneously perceive human-like beings in the environment around us, are universals. Everyone, regardless of culture, gender, or historical era, is vulnerable to these perceptual mistakes. And of course, every culture has some form of religion or belief in spiritual beings. Stewart Guthrie, a professor emeritus of anthropology at Fordham University and one of the de facto founders of the cognitive science of religion, famously put these two universals together in a theory of religion published in 1993 as the book Faces in the Clouds. In it, Guthrie argued that religious beliefs fundamentally arise from humans’ ubiquitous tendency to overestimate the amount of agency in the world.

I happened to be presenting alongside Guthrie at a recent, and extremely rewarding, conference in Boston (plug: two fellow researchers covered the conference in a nice blog post at This View of Life). I was inspired to relay Guthrie’s recent insights into religion’s cognitive foundations to readers, in part because Science On Religion tends to emphasize the Durkheimian, or group-oriented, aspects of religion. This gives me a lot of leeway to tackle subjects I find interesting (such as the fiery debates between kin and group selection theories), but it leaves out big swathes of what’s going on elsewhere in the scientific study of religion. Guthrie’s well-developed theories are an excellent complement to this blog’s usual focus on religion’s group-bonding effects.

Guthrie lays out his argument for how our evolved minds generate religion in seven clear steps. First, he points out that perceiving is interpretive. That is, we don’t experience raw sense data without attempting to make sense of it. You hear a noise, and immediately your brain is trying to figure out what caused it: a gunshot? A firecracker? A backfiring car? Second, the way that we make sense of raw sense data is by comparing it with cognitive models. For instance, we hear a low, rumbling noise near our ears, and immediately interpret that sensation as coming from our cat, who is purring (awww). We do this automatically; the instant we become aware of a perception, we’re already automatically searching for models that might explain it.

Third, we’re fine-tuned to interpret sensations based on our survival interests. For example, from a survival perspective it’s in our interest to know whether there’s a spider crawling up our leg. So if we feel a tickle on our ankle, we’re more likely, at least in the heat of the moment, to interpret that sensation as matching the model of “spider” than the model of “loose thread tickling my skin.” Think of the smoke detector in your home: your smoke detector is calibrated to be over-sensitive to hints of smoke (at least it is in my house, where we experience near-weekly aggravating episodes of shrieking false alarms), because a smoke detector calibrated the other way – one that was under-sensitive to signs of smoke – would be worse than useless. In other words, avoiding false negatives is worth occasional false positives when it comes to matters of survival.

Guthrie cites our tendency to see faces and face-like patterns everywhere as an example of interpreting sense data in ways that are relevant to survival. For early humans – as well as modern ones – the most important elements in the environment have tended to be other humans. Other humans are where we get our resources, knowledge, care, affection, vital information, and most other goods. They’re also the sources of most significant threats: physical aggression, social ostracism, bullying, and competition. So it makes sense for our brains to be finely tuned to over-perceive human agents in our environment.

Guthrie’s fourth step is that, in his words, “perceptual uncertainty is chronic.” That is, it’s hard to always know for certain what we’re seeing or experiencing, and even harder to know what might be causing that experience; our senses are unreliable, and our ability to tell causal stories about the world even more so. This unreliability, combined with our human-oriented social brains, leads us to err on the side of perceiving events as having been caused by humans or human-like agents: the leaves didn’t move because of the wind, but because someone moved them.

Tree bark face


Fifth, we’re led to perceive agency in our environment by formal characteristics – such as eyespots and body shape – as well as by the behavior of things. This explains why we’re so apt to see faces everywhere: lots of natural patterns in nature boast formal similarities to faces, such as two spots laterally aligned above a third spot in the bark of a tree (see right, as if you haven’t already looked). But, Guthrie stresses, it’s actually behavioral cues that awaken our agency-senses the quickest. For example, seeing rocks move in the California desert suggests to many people that somehow the rocks are being moved by conscious agents – or are moving themselves.

Sixth, plenty of research over the past decades has suggested that humans are intuitive mind/body dualists, sensing at a gut level that our minds are somehow separate from, and independent of, our bodies. Guthrie’s not talking here about metaphysical, Cartesian-style dualism; instead, he’s referencing our general tendency to feel that emotions, inner states, dreams, and so forth belong to our “minds,” while physical sensations and actions belong to our bodies. This mind-body dualism allows us to perceive minds in places where there are no bodies: for example, in wind that blows our hats off, or in the gurgling of water in a stream.

Finally, our minds are constantly making the anthropomorphic equivalent of Pascal’s wager: “betting” that it’s most worthwhile to use models of human agency in interpreting perceptions. If we’re wrong, we don’t lose much: a moment of distraction. But if we’re wrong in the other direction, we stand to lose a lot: we could get ambushed, killed, or excluded from social relationships. So when choosing which models to apply to our perceptual experience, we tend to err on the side of choosing the model labeled “human mind.”

Guthrie’s model, when added up, presents a picture of humans as intelligent, socially aware animals whose evolutionary history has pressured us to be hyperalert to signals of agency and intelligence in our environment. Since we’re intuitive dualists, this intelligence doesn’t always have to be connected to a body, which means our minds are free to detect agency in the shapes of clouds, in meaningful coincidences, and in experiences we interpret as answered prayers. Together, these proclivities lay the cognitive foundation for the universal human tendency to believe in gods and spirits – the core of religion, according to most cognitive scientists of religion.

An interesting consequence of Guthrie’s theory – which in the years since 1993 has become almost universally accepted among CSR researchers – is that it may help explain why there’s such an overlap between the autism spectrum and irreligion. People with autism-spectrum disorders are generally less socially oriented than neurotypicals, and tend to be poor with social cues, body language, and imagining others’ mental states. Such people are also often less interested in imaginative play or storytelling as children than their peers. Together, these traits make it unsurprising that autistics tend to anthropomorphize less than neurotypicals.

On the other hand, many autistics are high systemizers, showing interest in impersonal systems with regular, predictable features. Interestingly, there’s evidence to suggest that the networks of the brain that underlie systemizing thought are distinct from, and may even inhibit, those that drive social cognition. Guthrie points out that the brain’s so-called “default mode network” is likely oriented toward social events and relationships. The fact that dozens of brain-imaging studies have found this network to light up when subjects had no tasks to attend to implies that, whenever humans aren’t actively engaged in a focused task, they tend to revert to daydreaming about what matters most: other humans. (Anecdotal corroboration: this is certainly true for me, for example when I win my recent arguments in the shower.)

So if some people tend to be higher systemizers, and to use social cognition less than most folks, then according to Guthrie’s theory you’d expect such people to be less likely to anthropomorphize, and therefore be less religious than average.* And, in fact, this is exactly what studies have found: people on the autism spectrum tend to be less religious than normal.

But is anthropomorphization all of what religion’s about? Personally, I’m skeptical – there’s too much evidence that religion intrudes into other territories, including emotion regulation, morality, social organization, and experiences of transcendence. However, many of these other spheres of religion are also inextricable from sociality, which in turn is driven by the same neural circuitry that underlies anthropomorphization. And the fact remains that individualism and low levels of interest in personal relationships are two of the best predictors of religious nonbelief. So Guthrie’s theory may not be all-encompassing, but it certainly sheds light on many of the basic features of the religious landscape. Religion may not be exclusively social. It may not be solely our brains’ tendency to anthropomorphize reality. But there is something deeply social and anthropomorphic about much of what we call “religion,” and Guthrie’s lifetime of work forces us to take that fact seriously.


* Note that I’ve once again brought the discussion back to the social aspects of religion, despite saying at the beginning that Guthrie’s work was a nice opportunity to switch over to an individualistic, cognitive take on the subject. I swear, this isn’t what I intended to do when I set out to write this essay. I guess what they say is true: if you’ve got a hammer, every problem does, in fact, look suspiciously like a nail.

Animals evolve. People evolve. Can groups evolve?
The culture wars come for public higher education
Religion and evolution, part deux
Is religion evolutionarily adaptive?