Field of Science

Supernatural explanations just don't occur to kids - they need to be taught them

It's pretty much taken as an assumption these days that human beings are 'natural-born believers'. Ask a cognitive scientist who specializes in religion, and they will tell you that our brains are predisposed to all sorts of supernatural concepts.

One consequence of this consensus is a vast outpouring of articles and books pondering over what the evolutionary advantages of religion are. A lot of these explanations are pretty tendentious, and to me it has never seemed likely that this was the whole story.

One popular way to investigate the 'naturalness' of religion problem is to see if supernatural concepts are hardwired into children - as you would expect if religious ideas are intuitive and naturalistic ideas have to be learned. Perhaps surprisingly there are very few studies to support this idea - the same 'classic' studies keep getting recycled in each new article or book.

And when independent researchers outside the core groups test the hypothesis, they often get results that don't fit the story. That's the case with a new study by Jacqui Woolley, a psychologist at the University of Texas.

She and her colleagues read some short tales to a bunch of kids (67 in total) aged 8, 10 or 12, and also 22 adults. All the stories illustrated a 'difficult to explain' event.

For example, one featured a guy who steals a little money regularly, until he has enough money to buy a really fast car - which he promptly crashes. Another featured a terminal cancer patient whose cancer went away 'miraculously'. And another featured a woman who jogged regularly, and yet on her wedding day she tripped and hurt her leg badly - thus making her miss her wedding and also stopped her running. The stories were designed to illustrate events that could be ascribed to moral justice, divine intervention, or luck/fate.

So they read these stories and then asked the listener how the event could be explained. The surprising thing was that the kids hardly ever offered up supernatural explanations. Instead, they would say that maybe the cancer patient slept a lot, which helped her get better. Or, for the athletic woman who tripped on her wedding day, “because she tripped over a rock while she was walking. People usually trip over stuff and fall.”

Adults, on the other hand, readily offered up supernatural explanations. There was a clear trend, too, as you can see in the graph - the older the child, the more likely they were to explain these strange happenings by recourse to the supernatural

Then Woolley and Co. put some suggested explanations to them. The kids tended to agree that god or other supernatural explanations were plausible (although they felt that god explanations were more likely for stories with happy outcomes). So it's not that they aren't aware of the concepts - it just that they don't occur to them spontaneously.

Crucially, even the religious kids were more likely to provide naturalistic explanations than supernatural ones. They were, it's true, more likely to give supernatural explanations than the non-religious, and they were also more likely to give god-based explanations - but according to Woolley even this relationship only becomes significant at around age 12.

This doesn't of course, mean that humans are not predisposed to think supernaturally. Clearly, in some circumstances we are - and it seems likely that some people are more predisposed to think supernaturally than others.

But what this, along with other evidence, does show is that it is far too simplistic to argue that we are 'born believers'. In fact, we are born with a wide range of tools with which to understand the world around us, and culture is critical for shaping how those predispositions are shaped into beliefs.

ResearchBlogging.orgWoolley, J., Cornelius, C., & Lacy, W. (2011). Developmental Changes in the Use of Supernatural Explanations for Unusual Events Journal of Cognition and Culture, 11 (3), 311-337 DOI: 10.1163/156853711X591279

Creative Commons License This article by Tom Rees was first published on Epiphenom. It is licensed under Creative Commons.


  1. Religion has a social component. I'll bet adults, being socialised into a religious system, come up with answers that seem plausible to the religious 'tribe'. Kids, less so.

    Wonder how we could test that.

  2. Adults have much more experience of life than kids. If, therefore, adults are more likely to attribute "supernatural" explanations, does that not carry much more weight?

  3. I'm not sure if this experiment was the right way to test it.

    I don't think religion is hardwired, but I think thoughts that lead to religion are, like a twig snapping leading to an automatic assumption that something stepped on it, rather than say gravity pulling it off a tree.

  4. Good stuff.

    Darren Sherkat often complains of the community studying sociology of religion being pretty stacked with agenda-laden Christians, and your comments about default assumptions and "retesting outside the core group" seem to support that idea.

    I partially agree with the second Anonymous (11:08), that there are other ways of testing this that would likely get different results. I disagree that this is the wrong way, but I do think there are other ways to look at this. Certainly this seems to suggest that it's not a clear-cut as all that.

    Talking out of my ass here, I might speculate that what we are seeing here is the idea of a just/karmic god is not natural. I'd point to the capriciousness of the ancient Greek gods as an example. I'd also point to the seemingly irreconcilable contradiction in modern Christianity between praying to god asking for favors vs. believing in an omnipotent omniscient omnibenevolent god. That's of course a logical absurdity -- but it makes sense if you think of omnnibenevolence (or even just regular ol' not-so-omni benevolence) as a relatively recent addition. The idea of begging god to please sir be nice to me has a long tradition that predates the idea that god would actually want to be nice to you of her own accord.

    Just some thoughts. We may still be predisposed to supernatural (or at least teleological reasons), but need to be taught ideas of cosmic justice.

  5. I've got a couple of random thoughts that I'll throw out in no particular order.

    First, let me say that those are some very interesting results. I definitely think there's some good work going on here.

    Second, I think a crucial piece here is that adults seek meaning in life much more than children do. That could a factor in these results.

    I also think a key piece missing from this study was whether the parent or guardian of the children was themselves prone to using supernatural explanations or not. Although, that information would lend weight to the hypothesis that the supernatural is taught, not inherent.

  6. This doesn't of course, mean that humans are not predisposed to think naturally.

    There's a missing "super" there, Dr Rees.

  7. Well yeah but I think you missed the point. No one is saying we are "born with religion" in the sense that we innately contain or create supernatural explanations. We are born predisposed to accept supernatural explanations. That the kids increasingly attribute superatural explanations as they age is evidence that they have more exposure to such.

  8. I think that the take home from this study is that the answer to the question 'Is religion hardwired' is very nuanced - you'll get different answers depending on how you test the problem.

    Second anon there (and yes, Sabio, maybe I'll have to stop anonymous posts!) says that the hardwiring is more a predisposition to attribute agency. Which is true, but then this study allowed for that, because any explanation that wasn't natural was labelled as supernatural.

    Supernatural explanations given by younger children typically don't mention a god, but instead talk about 'moral justice', e.g,

    A typical response was given by a 10-year-old, who explained a man winning a contest by saying that it was “because he helped the animals at the animal shelter.”

  9. Peter, well perhaps but arguably we're predisposed to accept natural and supernatural explanations. But what comes most readily to the mind of an 8-year old are the naturalistic ones. Based on other studies, many cognitive scientists of religion would tell you the opposite - that the supernatural explanations (or rather, explanations that are contrary to accepted ideas about how the world actually works) are the ones that children resort to most readily.

  10. Interesting study.

    Just to add another thing:
    Maybe you also have to take into account, that children may not really distinguish between natural and supernatural.

    I.e. the example with cancer. If an adult would say the cancer was cured by sleep it maybe would seem more like a supernatural explanation to us (or just silly). A child just doesn't know how hard it is to cure cancer.
    The other to examples are working with probability. I think you need more live-experience to consider an event as that unlikely, that you prefer a supernatural explanation.

    "You have to know the laws of nature to notice the supernatural."

  11. OK, but any answer that attempts a naturalistic explanation - no matter whether or not an adult thinks it credible, counts as 'natural' in this study. So the sleep and cancer thing counts as a naturalistic explanation. Whereas an attempt to say that some random event occurred because the person 'deserved' it is supernatural (this was the most common kind of supernatural explanation given by the younger kids). Adults also tended to say that these occurrences were a 'sign' of something - also a supernatural interpretation.

    Of course, some people also said outright that god did it - a clear cut supernatural explanation. But the point is that the study was broader than that.

  12. OK, I understand that.

    My main point is another one, I am sorry if it was not clear enough:

    Did the children consider the events really as "difficult to explain"?

    From the answers it seems like for them those events were just normal. "People get cured or fall all day."

    (The study still would say that we are not predisposed to think supernaturally. If we think the natural explanation is good enough.)

  13. It might be that during the time we are still discovering the basic world around us, "naturalistic" explanations seem more interesting, which we at some point get bored with. Or perhaps the brain isn't developed enough in childhood to appeal to invent fantastic explanations.

  14. I noticed when I was in primary school and I was 'taught' about God and prophets and miracles I completely believed them, it was hard-wired in my brain but when I went to secondary school instead of being told what to believe we were to to think with no bias towards God or atheism. When I found my natural independence age 11 and I thought about it I was scared that I would go to hell but couldn't help having a lot of doubts. Infact out my entire class I recall stating I was a firm believer in God because I was scared of hell and if my parents found out, I was really an agnostic, I was the only one who believed everyone else where agnostic or atheist. Not to sound cocky but these were privately well educated kids, smart in top set and I found it quite alarming that what we thought was different to what society wanted to teach us.

  15. For anyone who have kids between 6-11 (best result when they are 7 years old)

    Try this ...

    Place a laptop or something very catchy to the child and tell the child do not touch this. You can also try this anytime a child lies to you.

    After the child disobeys ...

    In a sweet nice voice, now ask the child ..

    "Before you touched the object I told you not to touch you heard 2 voices .. what did the first voice tell you to do (keep asking, what else did they tell you) ... after listening. Now ask, what did the second voice tell you to do".

    Make sure you put no fear on the child throughout this whole process pleasssssse.

    Would love to hear your results... and please kindly share.

  16. I like this article but another explaination can be that developmentally the kids up to age 12 may have been concrete vs abstract thinkers. The supernatural aspect is very abstract.

  17. Give me an explanation of the universe that isn't supernatural. Yes, the relative truths within this temporal nowhere realm are "natural" but that's where the circle ends and the jerk begins.

  18. How would they expect children to respopnd well to moral principles. children start out as a amoral(2year old are the most violent part of population). Moral is a growing aspect, not a premade slip we live by from day one.

    A bit banal way to try to test things aswell. They read a story, and expect that religious answers would come by default.

    Various myths occured after real life experiences, not books, probably mostly for the individual that experienced it. If this should spread to others I suppose it would depend on the charismatic nature of the storyteller.


Markup Key:
- <b>bold</b> = bold
- <i>italic</i> = italic
- <a href="">FoS</a> = FoS