Ross Douthat on Why Religion Makes More Sense Than You Think
In his third appearance on the show, New York Times columnist Ross Douthat defends the argument of his book Believe: Why Everyone Should Be Religious against Tyler Cowen’s Bayesian scepticism, ranging across the simulation hypothesis, the hard problem of consciousness, UAPs, near-death experience, and the Antichrist.
Key ideas
- Secular alternatives to religion already smuggle in gods. Douthat argues the simulation hypothesis and multiverse theories are functionally doing the work of polytheistic or Gnostic religion — positing powers who create and sustain a world — without their proponents following the logic to its religious conclusion: that such ‘simulators’ would themselves be small-g gods demanding a relationship.
- Cultural inheritance of belief is evidence for, not against, religion. Against Cowen’s objection that religious belief simply tracks where one is born (Hindu children in Sri Lanka, Muslim children elsewhere), Douthat concedes the point but reframes it: the fact that most cultures independently arrive at some religious response to reality is evidence of a real, convergent signal, not proof of arbitrariness — and, unlike inherited soap preferences, people do change religions through argument and conversion.
- The book’s argument moves from probabilistic deism to ‘probabilistic supernaturalism.’ Douthat’s case does not try to prove a specific creed. It argues that the balance of evidence — cosmic fine-tuning, the diversity yet cross-cultural consistency of religious and near-death experience, the unsolved problem of consciousness — points past bare deism toward a cosmos with intermediate powers, a soul distinct from but related to the body, and life after death; the further step to Christianity specifically is a separate, personal argument made at the book’s end.
- Convergent religious experience is his central evidentiary category. Rather than treating visions, near-death experiences, and psychic phenomena as an embarrassment to be explained away, Douthat treats their persistence — cross-culturally patterned in both variation and consistency — as data that any adequate theory of reality, materialist or religious, has to account for, not dismiss as fraud, insanity, or drug-induced delusion.
- The evidentiary bar he sets for the supernatural will be tested by the AI era. Cowen presses him repeatedly for reproducible, video-grade evidence of the miraculous; Douthat concedes the asymmetry but notes that reliable video technology has coincided almost exactly with the most secular decades in human history, and — more unsettlingly — that AI-generated video will soon make even a literal recording of an angel unpersuasive, collapsing the very evidentiary standard Cowen is demanding.
Content
What counts as religious: the simulation hypothesis test
Cowen opens by pressing on definitions: if he believes Nick Bostrom’s simulation hypothesis, how does that differ from being religious? Douthat’s answer is that it does not differ much at all — functionally, the simulation hypothesis performs the work of a polytheistic or Gnostic religion, in which ‘whoever is running the simulation’ occupies the position of a small-g god, or the Gnostic demiurge, an intermediate power responsible for the world’s structure and suffering. The view crosses fully from analysis into religion, he argues, the moment someone starts asking what the simulators want from them and whether to align with it — at which point prayer becomes a rational response. Cowen pushes further: if a chain of creators is plausible, why not infinite layers of gods creating gods, stretching in both directions? Douthat replies that parsimony has to work with the evidence of religious and spiritual experience as actually reported, not against it — and that evidence points to intermediate powers (angels, demons, small-g gods) beneath one higher power, a pattern nearly universal across the major religious traditions rather than a purely Christian idiosyncrasy.
The heritability objection and probabilistic supernaturalism
Cowen’s sharpest challenge is sociological: children raised Hindu in Sri Lanka become Hindu, children raised Muslim become Muslim, and this correlation with upbringing should make anyone suspicious of their own particular religious convictions — pointing, at most, toward a generic, low-commitment deism. Douthat calls this ‘my most liberal book in a certain way’: he grants that the diversity of religious traditions suggests a connection to God is available through many cultural paths (citing this as post-Vatican II Catholic teaching, not a fringe position), and that defaulting to one’s own culture’s tradition, absent better information, is not irrational. But he turns the heritability point back on Cowen: the same inherited, culturally conditioned quality applies to Cowen’s own economic and political convictions, which he does not treat as grounds for abandoning them. And religious belief, unlike a soap brand, changes through conversion and argument all the time — which is why the argument in his book is worth having at all. The destination he argues for is not a single decisive proof but ‘probabilistic supernaturalism’: not just that some deist first cause is likely, but that there is a cosmic order in which humans matter, that supernatural impingements on reality are real, and that a soul distinct from — but related to — the body likely survives death. Getting from there to Christianity specifically, he is explicit, is a further and separate argument made only at the book’s end.
Consciousness, the soul, and the quantum measurement problem
The conversation’s most philosophically dense stretch runs through consciousness. Cowen asks whether psychedelic-induced religious experience simply has a mundane neurological explanation; Douthat resists the word ‘obviously,’ arguing that neuroscience lacks even a working theory of ordinary consciousness, let alone the altered kind, and that disrupting the brain’s normal operation — as hermits, vision-questers, and drug users have long done deliberately — is no evidence against what is disclosed by the disruption. Pressed on where a soul might enter the evolutionary chain (are monkeys ensouled? trilobites?), Douthat declines to specify a transition point but reframes Cowen’s ‘interaction problem’ objection: the same unresolved gap exists in the transition from non-life to life and from non-conscious to conscious life, which materialism cannot explain either. He presses the quantum-mechanics measurement problem directly: on one reasonable interpretation, physical reality requires some act of observation to collapse possibility into actuality, and a measuring instrument only functions as an instrument of measurement because a conscious mind reads it — ‘the measuring stick without a process of consciousness is itself just a ruler.’ When Cowen describes his own decision-making as a mostly unconscious process with consciousness as an after-the-fact narrative — closer to what he calls the Colin McGinn ‘mysterian’ view that consciousness is real but permanently beyond intuitive human understanding — Douthat rejects the deflationary move, arguing that a theory of mind as mere aggregation does not do justice to the direct, undeniable experience of being a person who writes books and hosts podcasts.
Alien life, UAPs, and the Fermi question
Asked whether abundant alien life would diminish the significance of Jesus Christ, Douthat first questions the premise: he takes the Fermi paradox’s cosmic silence seriously even granting many apparently habitable planets, since either humanity is unusually early or intelligent life is genuinely rare — both readings leave Earth ‘fairly special.’ On theology, he sketches two live Christian responses to a confirmed extraterrestrial civilisation: a missionary model (Jesuits evangelising other species, per science-fiction precedent) or a C.S. Lewis/Madeleine L’Engle-style model in which other species were never subject to the Fall and so do not share humanity’s need for redemption. On UAPs specifically, Douthat treats reported abduction and encounter narratives as strikingly continuous with pre-modern folklore about fairies and the ‘Unseelie Court’ (an argument he credits to ufologist Jacques Vallee) — evidence, he suggests, either of a persistent Jungian dream-structure in human minds, or of genuinely mischievous non-human agents operating across eras under different guises. He resists collapsing this into a single tidy account, and is candid that the physical craft reported by military witnesses do not fit comfortably into either a purely psychological or a purely demonic explanation.
Evidentiary standards: near-death experience, psi, and the miracle at Fatima
Cowen repeatedly asks what evidentiary bar would actually move him: reproducible, recorded, laboratory-grade proof, the kind available for Winston Churchill’s recorded voice but never for the miraculous. Douthat’s response builds from cross-cultural near-death experience research: reports vary by tradition (Tibetan Buddhists report seeing the Buddha, Catholics an archangel or saint) yet share a consistent structural core — tunnel, light, resuscitation narrative — that is hard to write off wholesale as culturally manufactured hallucination. He argues that supernatural phenomena, almost by definition, cannot submit to laboratory reproducibility without undermining the coherence of an otherwise law-governed physical world — you cannot do physics if an angel might move a planet on demand — while psychic phenomena, tied more to human agency than to divine will, are in principle more testable (he cites Freeman Dyson’s view that psi correlates with bursts of creativity under duress, which resist laboratory conditions for the same reason falling in love does). Pressed on the total absence of decisive photographic or video evidence, Douthat cites Carlos Eire’s history They Flew, on multiply-witnessed accounts of levitating saints in the 1500s–1600s, and the 1917 miracle of the sun at Fátima — witnessed by a large public crowd and reported in contemporary newspapers, including The New York Times — as historical evidence that meets ordinary standards even without cameras. He concedes the absence of modern video is a fair objection but only a partial one: reliable video technology has existed for only a few decades, coinciding almost exactly with the modern West’s most secular period, too short a window to generalise from.
The Antichrist, AI, and spiritual intermediaries
Asked about Peter Thiel’s fascination with the Antichrist, Douthat positions himself as less apocalyptic: defining the Antichrist loosely as a coming global rule under a single false view of reality, he argues that the twenty-first century has produced only brief flares resembling that pattern — Barack Obama’s global popularity in 2008, or ‘woke’ institutional dominance over technology and speech — each of which collapsed under competing pressure within a few years. A serious Antichrist theory, he says, needs a power ‘not yet fully in evidence’; a maximalist AI-doomer view of a machine-god would qualify, as would ancient legends of returning gods reasserting control in a new guise, but neither currently meets the bar. On Catholicism itself, Douthat — a practising but self-described ‘normal’ Catholic, not a member of Opus Dei — discusses Mexican folk Catholicism’s comparative comfort with spiritual intermediaries against mainstream American Catholic suspicion of the same, citing devotion to Santa Muerte as an example of the genuine risk (dabbling with what the Church treats as illicit spirits) alongside the genuine insight (a fuller reckoning with intermediary spiritual reality) that folk practice contains. On papal authority, he notes his own priors have shifted since the start of the Francis pontificate: infallibility in narrowly defined dogmatic pronouncements (the Immaculate Conception) still makes sense to him, but the boundaries of when the pope is and is not speaking with that authority on moral questions have grown murkier in his view.
Writing fantasy in an AI age
Douthat closes on his Substack-serialised fantasy novel, The Falcon’s Children, tracing its roots to a Tolkien-soaked adolescence (versus the Ayn Rand path some future conservatives take instead) and a return to an abandoned draft during a period of illness with Lyme disease, when writing fiction was one of the few things his ‘imprisoned’ consciousness could still do. He reads fantasy’s recurring narrative interest in the passage from an enchanted to a disenchanted world — present in both Tolkien and George R. R. Martin — as unusually timely for what he calls the culture’s current ‘psychedelic and UFO-haunted moment.’ Where Cowen advises writing for an AI readership, Douthat draws a line between his journalism, which he half-accepts may increasingly be read by AI systems, and his fiction: he doubts AI can yet match human-level creative writing and wants his own work in the world before it can, on the conviction that art made by an actual human consciousness is inherently better regardless of what future audiences might think. Closing notes cover his slow, non-linear recovery from chronic Lyme disease, and his next project — a less apocalyptic counterpoint to Thiel’s Antichrist thesis, framing demographic collapse and the unravelling of long-standing institutions as a civilisational ‘bottleneck’ rather than an ending.
See also
- Ross Douthat — speaker
- Tyler Cowen — host
- Science and Religion — the wiki’s existing science-versus-religion concept page; Douthat’s evidentiary, experience-based case for the supernatural extends it with a new argumentative mechanism
- Consciousness — Cowen’s self-identification with Colin McGinn’s mysterian position, and Douthat’s dualist rebuttal via the quantum measurement problem, engage this page’s ‘Main positions’ directly