Something I think doesn’t get enough attention in discussions about God is how much of the “evidence” people point to lines up neatly with what we already know about human cognition. In particular, three ideas keep cropping up for me: hyperactive agency detection, “third person” or sensed-presence experiences, and Julian Jaynes’ bicameral mind. Taken together, they offer a surprisingly clean way of understanding why gods feel so real to people, without needing to assume there actually is one.
Humans are extremely good at detecting agency—so good, in fact, that we routinely see intention where there is none. This makes sense evolutionarily: it’s safer to assume the rustling in the bushes is something alive than to assume it’s nothing. But this tendency doesn’t stop at predators. When we face uncertainty, coincidence, or events we don’t yet understand, our minds naturally lean toward “someone did this” rather than “this just happened.” Once you start from that posture, invisible agents become a very intuitive explanation.
Layered on top of that are experiences people often describe as a “presence.” In moments of stress, isolation, fear, or exhaustion, many people report feeling accompanied, guided, or watched over. Neurologically, this seems tied to how the brain models itself and others, and sometimes misattributes its own internal monitoring or emotional regulation as something external. The important part isn’t whether the experience is vivid (it often is), but how easily it gets interpreted as another mind nearby rather than as something arising within.
This is where Julian Jaynes’ idea of the bicameral mind becomes interesting, even if you don’t buy his full historical thesis. The core insight is that parts of our cognition that generate commands, judgments, or moral imperatives can be experienced as “not me.” When those internal signals feel authoritative or directive, they’re especially prone to being framed as coming from outside the self. In a culture already primed to expect gods, spirits, or divine voices, that framing becomes almost automatic.
Put together, these ideas suggest that a lot of what people take to be encounters with God may actually be encounters with their own cognitive machinery, filtered through agency detection and cultural expectations. That doesn’t mean people are lying, foolish, or mentally unwell. It means the experiences are real, but the explanation layered on top of them may not be. We mistake phenomenology for ontology.
To me, this ends up being a cleaner way of understanding the facts. It explains why these experiences are so common, why they feel personal and meaningful, and why different cultures report radically different “gods” with similar psychological textures. You don’t need to posit an undefined supernatural agent when we already have a well-supported account grounded in how human minds work. At the very least, these ideas deserve serious consideration before concluding that the best explanation is that a god is actually there.