# Cognitive Science of Religion ## Intro Cognitive Science of Religion (CSR) is a young field, about 30 years old, that asks a simple question with a surprising answer: why do humans, everywhere on earth, naturally form religious beliefs? The surprise is that you do not have to teach children to believe in unseen agents, in life after death, or in a Big Mind behind nature. They show up that way. Researchers like Justin Barrett, Deborah Kelemen, and Olivera Petrovich have documented this in study after study. Preschoolers explain natural objects in terms of purpose ("clouds are for raining"). Even Japanese kids raised in non-theistic homes attribute the origin of natural things to a "God-like" agent. Even atheists, when their guard is down, show implicit intuitions that the dead "still exist somewhere." The interpretation, of course, varies. The atheist version goes: religion is a *cognitive byproduct.* Evolution wired us to detect predators (so we over-detect agents), to read minds (so we project minds onto storms), and to fear death (so we invent afterlives). Religion is a misfire of useful mental tools. The Christian version goes: this is exactly what the Bible predicted. [Romans 1:19-20](/codex/romans-1-19-20/) says *"what may be known about God is plain to them, because God has made it plain to them."* Calvin called it the *sensus divinitatis*, a sense of the divine built into every human mind. CSR's data fits that picture cleanly. Either way, the popular New Atheist claim that *"atheism is the natural default and religion is a culturally-imposed weirdness"* does not survive the data. The default is religious. Atheism takes work. This page lays out the research and the rival interpretations. ## In full The interdisciplinary research program investigating how ordinary human cognitive architecture produces religious belief. CSR studies how evolved cognitive faculties, agency detection, teleological reasoning, mind-body dualism intuitions, and theory of mind, generate religious concepts without requiring special revelation or cultural indoctrination. The field emerged in the 1990s from developmental psychology, evolutionary psychology, and anthropology. CSR is apologetically load-bearing because its central finding, that religious belief is cognitively natural and developmentally early, directly inverts the New Atheist claim that atheism is the cognitive default. Christianity predicts this finding ([Romans 1:19-21](/codex/romans-1-18-21/): "what may be known about God is plain to them, because God has made it plain to them"); atheism does not. ## Core findings ### 1. Children are intuitive theists Young children spontaneously infer purposive design and supernatural agency cross-culturally, prior to formal religious instruction: - **[Justin Barrett](/codex/justin-barrett/)** (*Born Believers*, 2012; *Why Would Anyone Believe in God?*, 2004), children naturally form God-concepts; "natural religion" precedes enculturation. Barrett's research shows children default to understanding objects as designed-for-a-purpose and events as agent-caused. - **Deborah Kelemen** ("Are Children 'Intuitive Theists'?", *Psychological Science* 15:5, 2004), documents *promiscuous teleology*: preschoolers default to purpose-based explanations for natural objects ("clouds are for raining," "rocks are pointy so animals don't sit on them") cross-culturally, including in secular Western samples where parents actively discourage religious belief. - **Olivera Petrovich** (Oxford developmental research, 1990s-2000s), Japanese preschoolers in a culturally non-theistic context spontaneously attribute origins of natural kinds to a "God-like" agent rather than to people or natural processes. ### 2. Agency detection is hyperactive, not deficient - **Stewart Guthrie** (*Faces in the Clouds*, 1993), humans naturally anthropomorphize the world; supernatural-agency intuitions are cognitively default. - **HADD (Hyperactive Agency Detection Device)**, Barrett's term for the cognitive module that over-detects intentional agents in the environment. From an evolutionary standpoint, false positives (seeing a predator that isn't there) are less costly than false negatives (missing one that is). The result: humans are "wired" to detect agency. ### 3. Afterlife belief is cognitively natural - **Jesse Bering** (*The Belief Instinct*, 2011; "The Folk Psychology of Souls," *Behavioral and Brain Sciences* 29, 2006), even young children and self-identified atheists show implicit afterlife-persistence intuitions; the mind naturally resists the concept of total personal cessation. ### 4. Religion is cross-culturally universal - **Pascal Boyer** (*Religion Explained*, 2001; *The Fracture of an Illusion*, 2010), develops the cognitive-evolutionary account of religion's universality; religious concepts are "minimally counterintuitive" agents that exploit normal cognitive architecture. - Convergent ethnographic evidence: no documented human culture lacks some form of supernatural belief (Schmidt's *Urmonotheismus* / High-God pattern; Eliade's sacred-profane universal; Richardson's pre-contact monotheistic concepts). ## The atheist interpretation vs. the Christian interpretation CSR researchers (many of whom are atheists or agnostics) typically interpret their findings as evidence that religion is a *cognitive byproduct*, an evolutionary spandrel produced by agency detection and teleological reasoning that was adaptive for other reasons. On this reading, religious belief is "explained away" as a glitch. The Christian interpretation treats the same data as confirmation of design: | Data point | Atheist reading | Christian reading | |---|---|---| | Children default to theism | Cognitive bug; HADD misfires | *Sensus divinitatis* functioning as designed (Calvin, Plantinga) | | Cross-cultural universality | Shared cognitive architecture produces shared errors | God "made it plain to them" (Rom 1:19) | | Teleological reasoning | False-purpose attribution | Genuine perception of real design | | Afterlife intuition | Mind-body dualism cognitive error | Ecclesiastes 3:11, "He has set eternity in their hearts" | The data underdetermines between these interpretations. The apologetic point is narrower: **CSR refutes the claim that atheism is the cognitive default.** Whatever the ultimate explanation, proto-theism is closer to the human cognitive baseline than atheism. ## Apologetic deployment ### 1. Inverts the "atheism is the default" claim The New Atheist rhetorical move, "atheism is simply the absence of belief, the default position from which theism must argue", is empirically falsified by CSR. Children do not start as blank-slate agnostics who must be indoctrinated into theism. They start with theistic-type intuitions that must be *argued out of* to produce atheism. The developmental trajectory runs theism → atheism, not the reverse. ### 2. [Romans 1:19-21](/codex/romans-1-19-21/) predicts the CSR finding Paul's claim that "what may be known about God is plain to them, because God has made it plain to them" maps onto the empirical observation that theistic intuition is innate, not culturally imposed. Christianity generates the prediction; the data confirms it. ### 3. The HADD/byproduct deflection is self-defeating The atheist move of invoking "religion is just an evolutionary byproduct via hyperactive agency detection" to neutralize the CSR data is performatively self-contradicting. The byproduct theory *presupposes* that theism is the cognitive default that needs explaining away, which is exactly the default-inversion the atheist was trying to dodge. Deploying the byproduct theory concedes the point. Furthermore, the byproduct objection proves too much: if cognitive origins debunk the belief, then *every* belief produced by evolved cognitive faculties is equally debunked, including the belief that religion is a cognitive byproduct. This is structurally identical to [Alvin Plantinga](/codex/alvin-plantinga/)'s Evolutionary Argument Against Naturalism (EAAN). ### 4. Five-pronged refutation of "atheism is the default" The CSR data is prong 4 in a five-pronged refutation deployed in [Atheism is a Belief](/codex/argument-from-science-as-faith-based-guillen/): 1. The neutral default is agnosticism, not denial 2. Babies-and-rocks reductio makes lacktheism vacuous 3. Inter-theistic reductio, every theist "lacks belief" in other gods 4. **CSR empirical inversion**, proto-theism is the cognitive default 5. Even if atheism were the default, defaults are not truth-conducive (geocentrism was the default too) ## Key scholars | Scholar | Key work | Contribution | |---|---|---| | **Justin Barrett** | *Born Believers* (2012) | Children are "natural believers"; HADD framework | | **Deborah Kelemen** | "Are Children 'Intuitive Theists'?" (2004) | Promiscuous teleology in preschoolers | | **Jesse Bering** | *The Belief Instinct* (2011) | Afterlife-persistence intuitions | | **Pascal Boyer** | *Religion Explained* (2001) | Cognitive-evolutionary universality of religion | | **Stewart Guthrie** | *Faces in the Clouds* (1993) | Anthropomorphism as cognitive default | | **Olivera Petrovich** | Oxford developmental studies | Japanese-preschooler God-attribution | | **Alvin Plantinga** | *Warranted Christian Belief* (2000) | Philosophical articulation of *sensus divinitatis* as proper cognitive faculty | Barrett is the scholar who most explicitly bridges CSR empirical findings to Christian theology. See Justin Barrett. ## See also - [Innate Knowledge of God](/codex/innate-knowledge-of-god/), the theological doctrine CSR empirically confirms - [Reformed Epistemology](/codex/reformed-epistemology/), Plantinga's epistemological framework grounding the *sensus divinitatis* - [Suppression of God Thesis](/codex/suppression-of-god-thesis/), [Romans 1:18-21](/codex/romans-1-18-21/) applied philosophically - [Atheism is a Belief](/codex/argument-from-science-as-faith-based-guillen/), deploys CSR as P5 (empirical inversion of "default" claim) - [Atheism](/codex/atheism/), the worldview whose "default position" rhetoric CSR inverts - [Argument from Religious Experience](/codex/argument-from-religious-experience/), companion evidential argument from the phenomenology of religious belief - [Romans 1.18-21](/codex/romans-1-18-21/), locus classicus predicting CSR findings - [Ecclesiastes 3.11](/codex/ecclesiastes-3-11/), "He has set eternity in their hearts" ## Common questions this page answers **Q: What is Cognitive Science of Religion (CSR)?** CSR is an interdisciplinary research program (developmental psychology + evolutionary psychology + anthropology) that investigates how ordinary human cognitive architecture produces religious belief. Its central finding: religious belief, in some form, is cognitively natural and developmentally early across cultures. See [Justin Barrett](/codex/justin-barrett/)'s *Born Believers* and Deborah Kelemen on promiscuous teleology. **Q: Are children naturally religious?** The CSR research says yes. Children spontaneously infer purpose, design, and supernatural agency across cultures, prior to formal religious instruction. Kelemen documents preschoolers defaulting to teleological explanations ("clouds are for raining"); Petrovich found Japanese preschoolers in non-theistic contexts spontaneously attributing natural origins to a "God-like" agent. The developmental trajectory runs theism to atheism, not the reverse. **Q: Does CSR prove God exists?** No, the data underdetermines the metaphysical question. Atheist CSR researchers read the findings as evidence that religion is a *cognitive byproduct*, an evolutionary spandrel of agency-detection and teleological reasoning. Christian readers interpret the same data as the *sensus divinitatis* (Calvin, Plantinga) operating as designed. The narrower apologetic point is that CSR refutes the claim that atheism is the cognitive default. See [Innate Knowledge of God](/codex/innate-knowledge-of-god/). **Q: How does CSR relate to [Romans 1](/codex/romans-1/)?** [Romans 1.19-21](/codex/romans-1-19-21/) claims "what may be known about God is plain to them, because God has made it plain to them." Christianity therefore predicts exactly what CSR finds: theistic-type intuition is innate and cross-culturally universal, not a cultural overlay. The biblical text generates the prediction; the empirical data confirms it. See [Suppression of God Thesis](/codex/suppression-of-god-thesis/). **Q: What is the Hyperactive Agency Detection Device?** HADD is [Justin Barrett](/codex/justin-barrett/)'s term for the cognitive module that over-detects intentional agents in the environment. False positives (seeing a predator that is not there) are evolutionarily cheaper than false negatives (missing one that is), so the system is biased toward over-attribution. Atheists deploy HADD to explain religion away; Christians note that invoking HADD presupposes theism is the cognitive default that needs explaining, conceding the default-inversion point.