Big Question · the theory of knowledge

Can we know anything for certain?

You are certain the floor is solid, the sun will rise, the world is real. Now try to prove any of it without assuming something you cannot prove. This is epistemology — the study of what we know, and how little of it survives a determined doubt.

We can know a great deal for practical purposes, but almost nothing with absolute certainty. Skeptics argue our beliefs can always be doubted; rationalists find bedrock in reason, as when Descartes proved his own existence from the fact that he thinks; empiricists ground knowledge in experience while admitting it cannot deliver certainty. The one arguable exception is that you exist as a thinking thing right now.

The positions · at a glance

How we might know — and how we might not

Position 01

Skepticism

Pyrrho, Descartes (method)

We cannot be certain of what we claim to know. Descartes imagined an evil demon deceiving him about everything; nothing in experience could rule it out. Skepticism is often a tool for finding bedrock, not just a dead end.

Position 02

Rationalism

Descartes, Leibniz

Reason, not the senses, is the surest source of knowledge. From radical doubt Descartes salvaged one certainty — 'I think, therefore I am' — because the very act of doubting proves a doubter exists.

Position 03

Empiricism

Locke, Hume

Knowledge comes primarily from sensory experience. Powerful, but Hume showed it cannot deliver certainty: no number of past sunrises logically guarantees tomorrow's. This is the problem of induction.

Sit with the demand for certainty and it quickly turns corrosive. Descartes, looking for one thing he could not doubt, imagined an all-powerful deceiver bending every perception — the sky, his hands, even mathematics. Nothing he could see or feel proved the demon absent. The senses, it turned out, could all be counterfeit.

And yet the doubt destroyed itself at one point. However thoroughly he was deceived, something was being deceived; there had to be a thinker for the deception to land on. Cogito, ergo sum — I think, therefore I am. Whatever else collapses, the existence of the doubter survives the doubt. It is the closest thing philosophy has to bedrock, and notably it is delivered by reason, not the senses. That is the rationalist wager: the mind's own operations are more trustworthy than anything the eyes report.

The empiricist answerEverything through the senses.

The empiricists pushed the other way. Locke called the mind a blank slate written on by experience; there are no ideas we did not first take in through the senses. It is the method of science, and it works — but Hume found the crack in its foundation. We believe the future will resemble the past, that causes will keep producing their effects. Yet no argument proves it. Every attempt to justify induction has to assume induction. We trust the sun will rise on habit, not on proof.

The doubt runs out at exactly one place: the doubter.What survives Descartes' demon

The modern ruptureThe three pages that broke knowledge.

For over two thousand years, knowledge had a tidy definition: justified true belief. To know something was to believe it, to be right, and to have good reason. Then in 1963 Edmund Gettier published a three-page paper of counterexamples in which someone has a belief that is true, and justified, and yet plainly not knowledge — they were right by luck, their justification pointing at the truth by accident. The definition cracked, and epistemology has been trying to repair or replace it ever since.

Where does that leave the original question? With a useful distinction. Certainty — the demon-proof kind — we have almost none of, beyond our own existence as thinking things. Knowledge in the ordinary, workable sense we have in abundance, provided we drop the demand that it be immune to every conceivable doubt. Most of living happens in that second register, and always has.

People also askQuick answers

Can we know anything for certain?

Almost nothing with absolute certainty, since most beliefs can be doubted. The clearest exception is your own existence as a thinking thing — Descartes' 'I think, therefore I am' survives even radical doubt. In the ordinary sense, though, we know a great deal.

What is the problem of induction?

David Hume's observation that no number of past observations can logically guarantee the future. We expect the sun to rise because it always has, but that expectation assumes the future will resemble the past — which is the very thing in question.

What is the Gettier problem?

A 1963 argument by Edmund Gettier showing you can have a justified true belief that still isn't knowledge, because it is true only by luck. It broke philosophy's long-standing definition of knowledge as justified true belief.

What is the difference between rationalism and empiricism?

Rationalism holds that reason is the primary source of knowledge; empiricism holds that it comes chiefly through the senses. Descartes is the classic rationalist, Hume the classic empiricist.

Sources & further reading

Educational content summarising established positions in epistemology, following the sources above.