Ask someone how they know what they claim to know, and watch how quickly the ground softens. We move through the day confident in a thousand things — that the sun will rise, that our memories are roughly accurate, that the people around us have minds like ours — and almost none of it could we actually prove if pressed. Epistemology, the study of knowledge, begins by taking that softness seriously. It asks the deceptively simple question: what do we actually know, and how could we ever be sure?
More than true belief
The classical starting point defines knowledge as justified true belief. To know something, on this view, three conditions must hold: you must believe it, it must be true, and you must have good reason — justification — for believing it. The three are each necessary. A lucky guess that happens to be true is not knowledge, because it lacks justification. A well-reasoned belief that turns out false is not knowledge either, because it lacks truth.
For two thousand years this seemed close enough. Then, in a three-page paper in 1963, the philosopher Edmund Gettier produced cases where all three conditions are met and yet we still hesitate to call it knowledge — situations where a belief is true and justified, but true by a kind of accident that the justification did not actually track. The Gettier problem has never been cleanly solved, and its lesson endures: even our best definition of knowledge has seams. The concept we lean on so heavily is harder to pin down than its everyday confidence suggests.
We move through the day certain of a thousand things, almost none of which we could prove if pressed.
The two great sources, and their quarrel
Where does justification ultimately come from? Two traditions give rival answers, and the tension between them runs through the whole history of thought.
- Rationalism holds that the deepest knowledge comes from reason itself — that some truths, like those of mathematics and logic, are grasped by the mind independently of experience and are all the more certain for it.
- Empiricism holds that knowledge is built from experience — that the mind begins as something close to a blank page, and that observation, not pure reason, is the court of final appeal.
Each captures a real strength and runs into a real wall. Pure reason can prove the theorems of geometry but cannot, by itself, tell you whether it rained last night. Pure experience can report what your senses deliver but cannot guarantee that the senses are not deceiving you, nor justify the leap from “the sun has always risen” to “the sun will rise tomorrow.” Immanuel Kant’s great move was to argue that knowledge requires both — that the mind actively structures raw experience through categories it brings to the encounter, so that what we know is always a collaboration between the world and the knower.
The sceptic at the door
No figure has done more to sharpen epistemology than the sceptic, who refuses to let any claim pass unchallenged. Descartes, seeking an unshakeable foundation, imagined an evil demon devoted to deceiving him about everything, and asked what could survive such doubt. His answer — that even a deceived mind must exist to be deceived, cogito ergo sum — gave him one fixed point. But the demon has modern descendants. Perhaps you are a brain in a vat, fed convincing signals. Perhaps you are dreaming now. The unsettling feature of these scenarios is not that they are likely but that they are, strictly, impossible to rule out from the inside.
The sceptic’s value is not in convincing us — almost no one actually believes they are dreaming this sentence — but in showing how little of our knowledge rests on the bedrock we imagine. Most of what we know rests on trust: in our senses, in memory, in testimony, in the basic reliability of the world. The sceptic reveals these as assumptions rather than proofs. The mature response is not to defeat the sceptic on their own terms, which may be impossible, but to notice that a life requires standing somewhere, and that some assumptions are more reasonable to inhabit than others.
I think, therefore I am. — and almost everything else, it turns out, requires a measure of trust.
After Descartes
Most of what we know, we were told
Here is a fact that epistemology long underrated and that our moment has made urgent: the overwhelming majority of what any of us knows, we did not discover. We were told. That the earth orbits the sun, that a distant country exists, that a vaccine works — almost no one has verified these personally. We know them through testimony, through a vast web of trust in other people and institutions. Knowledge is not, in practice, a lone mind confronting the world. It is profoundly social.
This makes the health of our information ecosystem an epistemological matter, not merely a political one. When the chains of testimony degrade — when sources multiply, authority fragments, and confident falsehoods travel faster than careful truths — it is our actual capacity to know that suffers. The old image of the heroic doubter, trusting nothing they have not personally proven, was always a fiction. The real task is the harder one of trusting well: knowing whom to believe, and why.
Living with fallibility
If certainty is mostly unavailable, how should we hold our beliefs? The answer that emerges from epistemology is fallibilism: the stance that we can have genuine knowledge while remaining open to the possibility that we are wrong. This is not wishy-washy. It is how science actually works and how careful thinking proceeds everywhere. A few commitments follow from taking it seriously.
- Hold beliefs in proportion to the evidence, with confidence that rises and falls rather than snapping between certainty and denial.
- Seek what would change your mind. A belief you cannot imagine revising is a belief you are not really holding rationally.
- Distinguish your sources. Treat direct evidence, expert testimony, and mere rumour as the different things they are.
Fallibilism asks us to live in the space between dogmatism and despair — neither clutching our beliefs as infallible nor abandoning the very idea of knowing. It is an uncomfortable middle, and it is where honest minds spend most of their time.
Knowing how, not only knowing that
Epistemology has historically obsessed over propositional knowledge — knowing that something is the case. But the philosopher Gilbert Ryle pointed out that much of what we know is not like this at all. A cyclist knows how to balance without being able to state the physics; a fluent speaker knows how to form sentences without consulting the grammar; a doctor recognises an illness through pattern and experience that resist full articulation. This is knowing how, and it cannot be reduced to a list of propositions you could write down and hand to someone else.
The distinction matters because our culture systematically overvalues the kind of knowledge that can be written, tested, and transmitted in words, while undervaluing the embodied, practical, hard-won knowing that actually runs most of life. Skills, crafts, and the tacit judgement of the experienced practitioner are knowledge in the fullest sense, even though they live in the hands and the trained eye rather than in statements. Any account of knowing that ignores them describes only a thin slice of what humans actually know.
The epistemology of a connected age
If knowledge is largely social, resting on chains of testimony and trust, then the structure of our information environment shapes what we are able to know — and ours has been transformed almost beyond recognition. We now have instant access to more information than any previous generation could dream of, and, simultaneously, an unprecedented difficulty in telling the reliable from the worthless. The gatekeepers who once filtered claims, for all their faults, have been replaced by feeds optimised for engagement rather than truth, where a confident falsehood routinely outruns a careful correction.
This raises genuinely new epistemological puzzles. How should a non-expert adjudicate between duelling experts? When everyone can find sources agreeing with them, how do we distinguish justified belief from a comfortable echo? The old sceptic asked whether we could trust our senses; the contemporary version asks whether we can trust our sources, and the answer requires skills our schooling rarely taught — evaluating credibility, tracing claims to their origins, noticing when a feeling of certainty has been manufactured. The classical questions about knowledge have not been retired by the internet. They have been made, for ordinary citizens, more urgent and more practical than at any time in history.
The humility that knows something
The strange gift of epistemology is that its deepest lesson is a form of humility that does not collapse into nihilism. We know less, and less securely, than our everyday confidence implies; our knowledge rests on trust, structured by minds that shape what they receive, vulnerable to doubts we cannot fully answer. And yet we know a great deal — enough to build, heal, predict, and understand — provided we hold it honestly.
To ask “what do we actually know?” is not to court paralysis. It is to learn the difference between the things worth our confidence and the things we have merely never questioned. In an age drowning in claims, that discernment — knowing how we know, and how surely — may be less a philosopher’s luxury than a basic condition of thinking clearly at all.
None of this should be mistaken for an argument that knowledge is hopeless. The point of epistemology’s doubts is not to leave us paralysed but to make us better calibrated — to match our confidence to our actual grounds, and to know the difference between what we have examined and what we have merely absorbed. A person who has genuinely internalised these lessons does not believe less; they believe more carefully, and they are far harder to deceive.
That, perhaps, is the practical pay-off of asking how we know. In a world engineered to manufacture certainty and sell it back to us, the rarest and most valuable habit of mind is the willingness to ask, of any confident claim including our own, the quiet question the whole discipline is built around: how, exactly, do we know that? The person who keeps asking it will not arrive at perfect certainty. They will arrive at something better suited to the world as it is — a kind of knowledge that holds itself accountable, and is all the sturdier for it.
To ask how we know, then, is not an idle exercise for philosophers with time on their hands. It is among the most practical questions a person can carry, because the alternative to asking it is not certainty but credulity — believing whatever is most confidently asserted or most comfortable to accept. The examined belief and the unexamined one can feel identical from the inside. Telling them apart is the entire work, and it never quite ends.
Leave a Reply