In the last four articles, we explored two sides of the theist/atheist debate. On the atheist side, we explored Antony Flew’s argument that we start from a position of atheism, as a default, and that the onus is on the theist to persuade us otherwise. We also heard from William Kingdon Clifford, who contended for a strong evidentialist position—i.e., that we must have sufficient evidence for all of our beliefs, and that it is immoral to believe on insufficient evidence. On the theist side, William James took up Clifford’s contention, arguing that his evidentialism was motivated by an exaggerated fear of being mistaken; and finally from Alvin Plantinga, who argued that theism is entirely rational, independent of any argument or evidence, for the same reason that belief in other minds, an external reality, and other beliefs we can’t prove (but that are almost certainly true) are.
The discussion is, in other words, about who gets the benefit of the doubt in the theist/atheist debate. While science cannot settle the epistemic question, which ought to provide the starting point of our inquiry, it has plenty to say about the psychology of belief: about how, in fact, people acquire such beliefs. In the next two articles I’d like to explore the more empirical side of the question, because factual claims frequently arise in these discussions, and science has a lot of interesting things to say about it. And, as in the prior discussions, we’ll hear from both believers and skeptics.

According to Justin L. Barrett (in his book, Why Would Anyone Believe in God?), in order to get at the question of how belief, or disbelief, in God works, we first need to understand how belief, more generally, works. In order to do this, we need to distinguish between reflective and nonreflective belief. A reflective belief is one that we consider as a proposition: that 1,098 + 7 = 1,105; that Leonardo da Vinci painted the Mona Lisa; that Donald Trump is a bad president; and so on. An unreflective belief, on the other hand, is one that operates more or less continuously and unobtrusively in the background of our minds, whether or not we consider it as a proposition: that I can’t walk through walls; that if I throw a rock into the air it will fall back to the earth; that I ought to eat when I’m hungry; and so on. Of course, beliefs of the second class can be considered as propositions, but they don’t have to be.
We also need to appreciate that one of the major findings of cognitive psychology is that the mind possesses a number of task-specific tools that generate our nonreflective beliefs. They are responsible for identifying and categorizing objects in our environment. For instance, there is a face detector that distinguishes faces from, say, vegetables; and a living thing describer that categorizes, say, the faces of living humans from those of statues. When these tools operate, they automatically generate all sorts of nonreflective beliefs about the objects they have identified and classified. For instance, we automatically assume that living things require nutrition, grow, die, and have the ability to reproduce. Although we classify both our fellow humans and statues as both having faces, we attribute such needs only to the former class of faces, not to the latter. Again, all this happens automatically; we don’t have to sit and think about whether statues need to eat or not, we just do know that they don’t.
An especially important tool for this discussion is the agency detection device (ADD). It’s responsible for identifying active, volitional objects in our environment (such as animals or other people) and distinguishing them from inert ones (such as trees and rocks.) Clearly, this is a very important tool for us to have. Other living things could be sources of nutrition, opportunities for mating, members of our tribe, members of an enemy tribe, predators, or what have you. Organisms that can identify agency-possessing objects in their environment have a big advantage over ones that can’t. It stands to reason, then, that we acquired this ability quite early in our evolutionary history, and that, in consequence, it is part of the basic architecture of our minds. We don’t have to sit and consider whether an object before us has agency, we just automatically, nonreflectively, do or do not attribute agency.
The ADD is sometimes described as the hyperactive agency detection device (HADD) because in practice it can be a bit jumpy. In light of natural selection, it makes sense that it would be. Consider, for instance, two cavemen sitting alone around two separate camp fires. They each hear what sounds like a twig snapping somewhere in the darkness. In the first caveman, the ADD has a default of attributing agency to the noise. He doesn’t know what’s out there, but he assumes it’s another caveman or a jaguar or what have you, picks up his club, and goes out there to investigate. If it turns out it was another caveman or a jaguar or something like that, he’s ready to defend himself. But if it turns out to have been nothing, he breathes a sigh of relief, sits back down, and forgets all about it.
Now consider our second caveman, whose ADD is less sensitive. He hears the same twig snapping, but because his ADD is set to the default of not attributing agency to the noise, he ignores it. If it turns out to be a jaguar or what have you, he could get eaten before he even knows what’s happening. But if it turns out to be nothing, he spares himself the minimal trouble of having gotten up to investigate.
In light of these two examples, I think we can see why the ADD is a bit jumpy in us humans. Either caveman could be wrong in their default assumption about what caused the twig to snap, but the one who assumes it’s another agent expends a little bit of energy in order to defend against a catastrophic loss, where, conversely, the one who assumes it’s not another agent saves himself a little bit of energy, but leaves himself vulnerable to catastrophic loss. Put differently, if the second caveman decides to sneak up on the first, the one with the more sensitive ADD, the first will be ready for him; but if their roles are reversed, the second caveman is toast. So it makes sense for us to have not just an ADD, but an HADD. When in doubt, your best bet is to assume agency.
Nonreflective beliefs can become propositions if we decide that some uncertainty attaches to them. Our first cave-dwelling friend doesn’t actually know, on hearing that twig snap, that there’s an agent lurking out there in the darkness. What he has is more like a hunch, or an intuition. But even if he decides, upon investigation, that there was no agency at work—if, that is, he acquires the reflective belief “there is no agent prowling about my campfire”—the nonreflective attribution of agency may remain. He may still feel apprehensive or on the alert. After all, absence of evidence is not evidence of absence. Just because he didn’t *find* the jaguar or what have you prowling around in the night, doesn’t mean there isn’t one. It might just mean he couldn’t find it, and he’s about to get eaten despite his best efforts. Thus what we believe reflectively and what “our gut tells us” aren’t necessarily the same thing. They can agree or they can conflict, but basically they operate on two different tracks.
Many contemporary cognitive scientists of religion believe that the HADD accounts for the intuitive sense that many people have that there is some such being as God.
The relevance of the foregoing to the theist/atheist discussion is not far to find. Many contemporary cognitive scientists of religion believe that the HADD accounts for the intuitive sense that many people have that there is some such being as God. Because our minds are essentially the same across historical time and space, this also accounts for the extremely widespread belief in God, or some such being, across cultures. It also means that belief in God, or some such being, is almost inevitable for human beings (taken en masse, not necessarily as individuals), in consequence of the kinds of minds we have. There is at least one story about why people believe—because they are the blind followers of authority—that this theory of belief undermines. It would be more accurate to say that rituals, traditions, institutions, etc., of a religious character, influence and strengthen, but do not create, belief in God or some such being.
Now that we have a scientific account of belief—one that makes no reference to the supernatural or anything like that—can we then conclude that science has explained God away? Probably not. It certainly seems possible that belief in God is simply a trick of our over-active agency detector, a side effect of natural selection. Then again, if God does exist, it stands to reason that God would want us to be aware of that, and so might have set up the evolutionary process so as to produce such beliefs spontaneously and nonreflectively in the vast majority of people, which is what we in fact find when we look out across the diversity of cultures. “To know in a general and confused way that God exists is natural to us,” Thomas Aquinas said, and cognitive science of religion backs him up on that.
Philosophically, it’s not clear that the argument, “God-beliefs are produced by the HADD, and are therefore unlikely to be true,” stands up. The reason is that all beliefs without exception are the results of some kind of cognitive/neurological process. I believe that the sun is shining above me because photons strike my cornea, the information is relayed up my optic nerve, processed in my ganglia, and, by some as-yet-unknown process, translated into the mental experience of seeing the sun’s light. We have scientifically explained sight, but by no means explained away the real, external existence of the sun. Perhaps the HADD is the organ by which we perceive the Creator, just as the eyes are that by which we perceive light. Or, then again, perhaps we are justified in distrusting our HADD because it’s jumpy and unreliable. In either case, we can see that the intuition that there is a God, or some such being, needs to be distinguished from the reflective belief that there is, or is not. Being an atheist does not necessarily eliminate the underlying intuition that there may, after all, be a God, any more than theism does the worry that there may not.
However we decide the epistemic question, I think we can see now that belief in God, or some such being, really is quite natural for us. Whether, as an epistemic matter, we agree with Flew and his “presumption of atheism,” or with Plantinga in thinking that there ought to be a presumption of theism, contemporary cognitive science of religion makes it reasonably clear that theism does in fact provide the default position, for most of us, most of the time.
Daniel Halverson is in the PhD program at the Institute for the History and Philosophy of Science and Technology, University of Toronto. His research focuses on the history of evolutionary biology in the Victorian and WWII eras.
If you’re just beginning to follow this series, or would like a handy reference, here are links to the previous articles:
Part II: Ian Barbour—The Conflict Model
Part III: Ian Barbour—The Independence Model
Part IV: Ian Barbour—The Dialogue Model
Part V: Ian Barbour—The Synthesis Model
Part VI: John Hedley Brooke, Complexity Thesis
Part VII: Plato and the Geometric Model of Knowledge
Part VIII: Arthur O. Lovejoy, the Great Chain of Being
Part IX: Did Heliocentrism Knock Humanity off Its Perch?
Part X: Thomas Paine and the Controversy over Extraterrestrial Life
Part XI: Arthur O. Lovejoy, the Great Chain of Being and Pre-Darwinian Biology
Part XII: Michael Allen Gillespie, Theological Origins of Modernity
Part XIII: William of Ockham and the Origins of Nominalism
Part XIV: Nominalism, Petrarch, and the Renaissance Origins of Humanism
Part XV: A Fractured World: God, Humanity, and Nature
Part XVI: Did Medieval Islamic Theology Subvert Science?
Part XVII: Galileo Goes to Jail?
Part XVIII: Humanistic, Scientific, and Theistic Approaches to History
Part XIX: What Is Science? (Part A)
Part XX: What is Science? (Part B)
Part XXI: Charles Taylor: A Secular Age (Part A)
Part XXII: Charles Taylor—The Bulwarks of Belief (A Secular Age, Part B)
Part XXIII: Charles Taylor—Time, Space, and Self in the Enchanted World (Part A)
Part XXIV: Charles Taylor—Time, Space, and Self in the Enchanted World (Part B)
Part XXV: Charles Taylor—The Protestant Reformation and the Rise of the Disciplinary Society
Part XXVI: Charles Taylor — Providential Deism and the Impersonal Order
Part XXVII: Charles Taylor—The Malaise of Modernity
Part XXVIII: Charles Taylor—The Dark Abyss of Time
Part XXIX: Antony Flew—The Presumption of Atheism
Part XXX: William Kingdon Clifford—The Ethics of Belief
Great post, Daniel. Just to play a friendly devil’s advocate: (1) Is there the potential risk here that the HADD hypothesis is just so much scientism / just-so storytelling? (2) If not, will our explanatory space accommodate the insertion of Kahneman and Tversky’s ‘System 1/System 2’ analysis into the mix? (Meaning that, conceivably, both theistic and atheistic convictions could be based on reflective and/or non-reflective processing?)
https://en.wikipedia.org/wiki/Thinking,_Fast_and_Slow
Thanks for your kind words, Luke.
I don’t think that this line of inquiry runs the risk of scientism because to say that a belief has a scientific explanation is not to say that it is false. All beliefs of any description whatever have a scientific explanation, including those which constitute science itself, so there’s a real problem of self-refutation for the person who argues that if we can explain how people come to have beliefs about God, then we’ve shown why they’re false. It really just doesn’t follow. So this to me is not so much scientism as just a not-very-strong argument. A person could become an anti-realist about beliefs generally, I guess, but that’s a question more about the nature of belief just as such then about belief in God specifically, and it would leave one’s commitment to theism or atheism or what have you untouched.
I’m not sure about Kahneman, but the general thrust of contemporary Cognitive Science of Religion appears to be that our non-reflective, spontaneous beliefs are weighted in the direction of some kind of theism. I don’t think reflective belief is weighted in favor of skepticism, but it does seem to be a prerequisite. I.e. one has to think reflectively about one’s own unreflective beliefs before skepticism even becomes an option.
Good. And what about the basic human temptation or reflex to potentially think in tribal or self-serving terms (be that from religious conviction or otherwise)? Does non-reflective thinking incline us to intuitively ‘take sides’ in some epistemic or socially-freighted debate? Does it take us all more reflective thinking, for example, to default to a more ecumenical or open-minded disposition on whatever issue may be live (let’s just say, in this case, the existence of God)?
Or is my anecdotal observation here possibly more attributable to individuals’ personalities, rather than something that can be asserted as a general evolutionary trait? I appreciate that you are summarizing literature here, by the way, and not purporting to be cognitive scientist yourself. However, if the referenced text should speak to as much.
Thanks! Great essay! Not quite clear as to how HADD would lead us to suspect the existence of a God, as God makes no twigs snap, if I may continue the metaphor. Very engaging intro! Want to hear more!
Cheers, Bill
“…identifying active, volitional objects in our environment (such as animals or other people) and distinguishing them from inert ones (such as trees and rocks.)”
I am afraid that there is a truckload of evidence from botany supporting the view that trees are not “inert” objects. They react to their environment (temperature, sunlight, water levels), communicate with other trees via their roots and so on (http://www.ecology.com/2012/10/08/trees-communicate/).
Regarding religion, most skeptics see no difference between believing in God X (pick your flavour) and believing in alien abductions. Until solid evidence appears, they are just interesting mechanisms to help us deal with our existential anxieties.
“….perhaps we are justified in distrusting our HADD because it’s jumpy and unreliable. ”
Barrett has addressed this in a very interesting paper. You can find it here.
https://www.researchgate.net/publication/274132544/download
Those of us for whom belief has always (!) been counter-intuitive are left as graph blips
also see the work of Paul Bloom (Yale) whose work with small children suggests intuitive dualism, the ground floor for god.