If reason is self-correcting, why do the communities built most explicitly on it become dogmatic?

Reason is supposed to be self-correcting. Yet the communities that organize themselves around it most explicitly — movement atheism, the rationalist subculture, technocratic policy circles — produce, with remarkable regularity, beliefs that are fantastical, reality-denying, and held with dogmatic fervor. Meanwhile, communities organized around inherited doctrine sometimes hold rational lines that the reason-communities abandon.

The Enlightenment era thinkers considered it inevitable that reason would self-correct once it is freed from tradition, custom, and doctrine, and take humanity to heights it has never seen before. They were not wrong, but the cases above offer some evidence that they were not wholly right. It seems to me that reason needs something (or things) it cannot provide to itself by itself; the question is what those things might be.

Main thesis. firmly held The Enlightenment thinkers were right that dogma obstructs reason (creating what I will refer to as a “ceiling”), but they were wrong that reason would benefit from being totally unfettered. Reason corresponds to truth only when disciplined by friction it cannot generate from within itself — an external constraint that, directly or indirectly, forces it to meet the demands of reality — what I am calling a “floor.” 1 1 This is an expansion of the model I first proposed in early 2025. As religious convictions shape decision-making, we might assume that if a faith has coexisted with society for a length of time, it is unlikely to include or otherwise require its adherents to commit to ideas that lead them to catastrophically destructive consequences in the near-term. And as these religions are competing with other ideological systems, we might also expect them to include some degree of the inverse: ideas that act as a prophylactic against the most self-destructive bad ideas that humanity has shown itself to be vulnerable to. This is what I propose constitutes the “floor”.

Unfloored reason does not automatically converge on truth. Instead, it is pulled towards other external incentives (like status), or towards the psychological needs or desires of the self or specific communities. In other words, without the forced encounter with some reality-enforcing mechanism, reason begins to serve other masters.

Reason-based communities fail because of their lack of floors.

The taxonomy. moderately held Many different “floors” can exist, differentiated here primarily by who or what enforces them.

Traditional floors. Tradition is the “cached learning” of a thousand multi-generational experiments. It is a cultural adaptation made necessary due to the finite nature of both our cognition and lifespan. Tradition is the result of countless experiments without the explanations for those results (and it is this lack of explanation that makes it appear “irrational”) — but it does not need legibility in order to be effective. 2 2 I am indebted to James C. Scott here. This is my articulation, but it is his idea.

Cultural evolution will favor beneficial (and thus, relatively reality-conforming) norms, and communities that hold onto them (for any reason) will persevere over communities which do not.

Humans enforce observance to tradition (ex: believers punish deviation), while the environment enforces content (abandon the “cached learning” and whatever hazards it has been deterring arrive eventually). Tradition is the only floor that recruits its own enforcers, and the only one that imposes a ceiling (what the Enlightenment thinkers correctly identified), because it cannot distinguish falsehood from heresy.

However, the cultural cache takes time to construct, and in moments of rapid environmental change, tradition begins to fail. Tradition dominated in slow worlds (our history up until the modern era), but is losing its footing today, as the rate of drift outpaces what tradition can absorb and encode. 3 3 Worth clarifying: tradition isn’t dying because champions of reason or science “debunked” it, or reasoned our way out of it. The reasons behind tradition were always opaque to us, but now it is also not useful. It is dying because its value is genuinely in question in a fast-changing world. If the rate of change slows again, I argue that the value of tradition would again re-assert itself.

Methodological floors. The discipline of a practice with external standards, like falsification, replication, or proof, can also create the necessary tension — pulling us back onto more truth-oriented ground. Unlike doctrinal floors, this can be a floor without a ceiling: it can block what doesn’t correspond to reality, and lets what is more true pass. But this floor is limited in range — you can apply the scientific method to the laboratory, less easily to the bigger questions of life.

Environmental floors. The floor most directly aligned with “reality”: the frontier. On a frontier, reason serves the purposes of survival — reality-denying means death. This too is a floor without a ceiling, requires no adoption, and is incapable of being socially captured or incentivized away (as the most vital incentive is to survive), as it is the only fully non-human enforcer.

The frontier is the only floor whose enforcer is not human at all, which is why it alone cannot be socially captured. Reality cannot be argued with, you cannot socially pressure starvation. The frontier can only be abandoned.

Follow-up. moderately held Communities that identify as committed to certain methods (rationalists, movement atheists, and so on) are wish-casting: outside of a narrow range of questions, these communities are bound and restricted not by any methodological floor, but by group consensus. Members are sanctioned for deviation from what the group believes, not for failing to meet any methodological standard. While identifying as “methodological,” they in fact operate as doctrinal communities without the inherited cache of tradition. 4 4 In other words, they contain nearly all the negative hallmarks of traditional communities (heresy hunting, in-group dynamics, status chasing, charismatic leadership) but lack — and in the case of atheists, actively reject — the redeeming feature of traditional communities: the norms encoded by environmental survival over generations. They fill the “dogma-shaped hole” without the time-tested value of inherited dogma. This is why these communities can fail in far more devastating fashion than tradition.

Just one, but a big one: how can secular-based communities incorporate “floors”?

• The environmental floor is the subject of its own inquiry: Liberal Order Requires a Frontier.
• Atheism without Reason, and the response to critics and Richard Dawkins.