The Case for Doubt

I want to argue for doubt as an intellectual virtue — not the paralysing kind, not the nihilistic kind, not the fashionable uncertainty that amounts to a refusal to think, but the active and disciplined kind: the habit of holding open questions open, of resisting the premature closure that passes, in much of public life, for conviction.

This argument is harder to make than it should be, because doubt has a bad reputation that it has done little to deserve.

The bad reputation comes from several sources. There is the long tradition that identifies doubt with weakness — the person who doubts is the person who lacks the courage of their convictions, who is not fully committed, who will not stand firm. There is the contemporary tendency to conflate doubt with relativism — the idea that to acknowledge uncertainty is to deny that anything is true, or that all positions are equally valid. There is the suspicion, particularly in political contexts, that expressed doubt is a form of evasion — a way of not having to choose, of not having to be held accountable for a position.

None of these identifications is correct, and I want to unpick each of them.

The identification of doubt with weakness confuses two different qualities: doubt as an epistemic condition and doubt as a psychological one. Epistemic doubt — the recognition that your beliefs might be wrong, that the evidence might be incomplete, that the matter is more complex than your current understanding — is not a weakness. It is an accurate representation of the actual situation, which is that most complex questions resist final resolution and that the people most confident of final answers are usually the people who have thought about them least.

Psychological doubt — the inability to act, the paralysis of perpetual indecision, the failure to commit to anything because everything might be wrong — is indeed a problem. But it is a psychological problem, not an epistemic one, and it has nothing to do with the intellectual virtue I am advocating. You can hold your views with full uncertainty about their ultimate truth while still acting on them with full commitment. These are independent dimensions. The general who doubts the infallibility of his strategy is not a less effective general than the one who is certain. He is, in most cases, more effective, because his doubt makes him adaptive.

The identification of doubt with relativism is a more sophisticated error, but an error nonetheless. To say "I might be wrong about X" is not to say "X is merely a matter of opinion" or "X has no correct answer." It is to say something much more modest: that my current belief about X is not guaranteed to be correct, and that I will revise it if I encounter sufficiently strong counter-evidence.

This is not relativism. Relativism is the position that there is no fact of the matter about X, or that all positions on X are equally valid. Doubt is compatible with a robust belief in objective truth. You can believe that there is a correct answer and also acknowledge that you have not yet found it, or that your current candidate for the correct answer is not certainly the right one. This is the normal epistemic condition of any honest reasoner confronting a hard question.

The suspicion that expressed doubt is evasion is perhaps the most practically significant of the errors. In public life, the politician who says "I'm not sure" is often treated as less trustworthy than the politician who is certain, regardless of whether the uncertainty is honest and the certainty is not. We have developed, collectively, a preference for confident wrongness over uncertain correctness, and this preference has consequences.

The consequences are visible in the quality of our public conversation about complex matters. The people most confident about the causes of economic cycles, about the solutions to social problems, about the correct responses to geopolitical crises, are rarely the people with the deepest understanding of these matters. The experts — the people who have studied these questions most carefully — tend, almost without exception, to be hedged, uncertain, and resistant to clean narratives. Their uncertainty is not a failure of confidence. It is a product of knowledge. The more you know about a complex system, the more you know about the limits of your knowledge of it.

This is an uncomfortable fact for a discourse that rewards certainty. But it is a fact.

Let me try to say something positive about what intellectual doubt looks like in practice, because so far I have been describing what it is not.

Intellectual doubt, as a virtue, looks like this: you hold your beliefs with a conviction proportional to the available evidence, and you are genuinely open to revising them when the evidence changes. You can argue vigorously for a position while simultaneously acknowledging its weakest points. You can commit to a course of action while remaining alert to signs that the commitment may need revision. You seek out the best counter-arguments to your own positions, not to win against them but to test your position against them — and if you can't answer them, you update.

This is harder than it sounds. The psychological tendency is in the opposite direction. Once a belief is formed, the mind tends to defend it — to seek confirming evidence, to discount contrary evidence, to reinterpret ambiguous evidence in the most favourable light. This is cognitive consistency bias, and it is very deep, and overcoming it requires active effort.

The active effort is what I mean by the virtue. Virtues are not natural states. They are cultivated dispositions — practices that have to be worked at because they go against the grain of the default tendencies. Courage goes against the grain of fear. Patience goes against the grain of urgency. Intellectual doubt goes against the grain of cognitive consistency.

You practice intellectual doubt the way you practice any virtue: by catching yourself in the failure mode and returning to the right disposition. The failure mode is the immediate defensive reaction — the feeling, when you encounter a strong counter-argument, of wanting to dismiss it rather than engage with it. You catch yourself in this feeling, and you ask: what would it mean if this argument were right? What would I have to revise? Is the revision small enough that my overall position survives, or does it require something more substantial?

These are not comfortable questions. They are, however, the questions that lead toward understanding, which is the only thing worth having in the long run.

I am aware that this argument can be read as a form of intellectual conservatism — as a counsel of caution that would, in practice, slow down necessary change or excuse the comfortable from the discomfort of conviction. I take this objection seriously. There are situations in which the case for certainty is strong: when the evidence is overwhelming and the stakes are high and the consequences of doubt are borne by people less privileged than the doubter.

But these situations are rarer than we usually pretend. Most of the contested questions in public life are contested because they are genuinely hard — because the evidence is mixed, because the systems involved are complex, because the values at stake are plural and in tension. In these cases, the person who says "I am certain" is almost always claiming more than the situation warrants.

Doubt, in these cases, is not a retreat from responsibility. It is the honest acknowledgement of a real complexity. And honesty about complexity, I would argue, is the beginning of actually solving it.I want to add something about what intellectual doubt looks like in practice — specifically, in the practice of changing your mind, which is the proof of the virtue and also its most difficult expression.

Changing your mind is hard in proportion to how publicly you have held the position you are changing. For private beliefs, it is relatively easy: you encounter new evidence, you update, you move on. The update is internal and the only witness is yourself. For public positions — things you have argued for, defended against challenge, staked some version of your reputation on — the change is much harder, because it involves not only the cognitive work of updating but the social and psychological work of revising a self-presentation.

This is, I think, the actual barrier. Not the intellectual barrier of updating a belief, but the social barrier of being seen to have been wrong. We treat changing our minds as a form of defeat — as evidence of having been weak, or uninformed, or gullible — and this treatment makes people reluctant to change their minds even when the intellectual case for doing so is overwhelming.

The consequence of this reluctance is that people defend positions long past the point at which they would update them if the positions were private. They seek counter-arguments to the new evidence rather than engaging with it honestly. They accuse the sources of the new evidence of bias or bad faith. They wait until the position is so obviously untenable that abandoning it can be framed as a concession to overwhelming pressure rather than a genuine change of view.

All of this is costly. It is costly to the individual, who maintains a false belief for longer than necessary and makes decisions on the basis of it. It is costly to the culture, which becomes increasingly unable to respond adaptively to new information because the social cost of updating is too high.

The alternative I am proposing is a reorientation of what changing one's mind means. Instead of treating it as defeat, we should treat it as a demonstration of the virtue I have been describing: the willingness to hold beliefs proportional to the evidence, and to revise them when the evidence changes. This is not weakness. It is intellectual integrity.

The people I most respect intellectually are, without exception, people who can describe in specific terms the things that would change their minds about their most important beliefs — and who have, at some point, actually changed their minds about something significant. This combination — the specified update condition and the demonstrated willingness to update — is the strongest available signal of honest reasoning.

The people I least trust intellectually are the people who cannot describe any evidence that would change their minds, and who have never publicly revised a significant position. This combination signals not conviction but rigidity — the defensive posture of someone who has confused the maintenance of their current beliefs with the practice of thinking.

I want to say something direct about the current moment, because I have been somewhat abstract and the current moment deserves directness.

We are living through a period of elevated certainty. The public conversation is increasingly populated by people who are very sure — about political questions, about social questions, about the nature of history and the demands of justice and the correct responses to complex crises. Some of this certainty is, perhaps, warranted. Some of the questions being contested do have clear answers that are worth defending with conviction. I am not arguing for a symmetrical doubt that treats all positions as equally uncertain.

But a significant portion of the certainty currently on display is not, I think, earned. It is performed. It is the certainty of people who have found that certainty is rewarded — that the social and professional rewards for confident declaration are higher than the rewards for honest uncertainty — and who have adjusted their self-presentation accordingly.

This is not a feature of any particular political or cultural position. It is a feature of the incentive structure of public discourse as it currently exists, which systematically rewards the confident and punishes the uncertain regardless of which has the better claim to accuracy.

The remedy is not more certainty from the other side. The remedy is a cultural revaluation of uncertainty itself — a shift in what we reward and what we trust. The honest acknowledgement of difficulty. The demonstrated willingness to update. The question asked with genuine openness rather than rhetorical purpose.

These are small things. They are also the things that keep the conversation honest, and honest conversation is the only kind worth having.