Loading field map
Skepticism, in epistemology, is the critical examination of the nature, scope, and limits of human knowledge. It is not merely the disposition to doubt, but a structured philosophical practice that asks how knowledge is possible, what its conditions are, and whether our claims to know can withstand rigorous scrutiny. The term derives from the ancient Greek skepsis, meaning inquiry or examination, and the skeptical tradition has always understood itself less as a doctrine of denial than as a method of ongoing investigation.
At its core, skepticism asks a deceptively simple question: What, if anything, can we know? This question branches into several distinct but related inquiries. First, there is the question of possibility: is knowledge attainable at all, or is it an ideal we can approach but never reach? Second, there is the question of scope: even if some knowledge is possible, are there domains—such as the external world, other minds, the past, or the future—that lie permanently beyond our reach? Third, there is the question of justification: what would count as adequate grounds for a knowledge claim, and do our ordinary beliefs actually possess such grounds?
The stakes are considerable. If skepticism is correct that knowledge is impossible or severely limited, then many of our everyday assumptions—that we know the sun will rise tomorrow, that we know our own hands exist, that we know other people have conscious experiences—are called into question. This is not merely an academic exercise. Skeptical arguments challenge the foundations of science, morality, religion, and law, all of which depend on the possibility of reliable knowledge. At the same time, skepticism serves a constructive function: by exposing the weaknesses in our claims to certainty, it disciplines inquiry, reminding us of the difference between what we believe and what we can defend.
The earliest systematic skeptical traditions emerged in the Hellenistic period, and they established the two fundamental postures that have shaped the field ever since. The first is Academic skepticism, associated with the Platonic Academy under Arcesilaus and Carneades. The Academics argued that knowledge is impossible—that no belief can be justified with certainty—and therefore the wise person should withhold assent from all claims. However, they recognized that action requires some basis for decision, so they developed the notion of the probable or persuasive impression: a belief that, while not certain, is sufficiently well-supported to guide practical life. The Academics thus combined a strict epistemological standard with a pragmatic fallback.
The second tradition is Pyrrhonian skepticism, named after Pyrrho of Elis and systematized by Sextus Empiricus. The Pyrrhonists did not assert that knowledge is impossible; rather, they claimed to have achieved a state of epochē—suspension of judgment—by setting arguments for and against every claim into opposition. Their goal was not to prove that knowledge fails, but to show that for every argument, an equally compelling counterargument can be offered, leaving the inquirer with no decisive reason to affirm or deny. This suspension of judgment was said to yield ataraxia, or tranquility, by freeing the mind from the anxiety of dogmatic commitment. The Pyrrhonists famously organized their arguments into the "Ten Modes" of skepticism, which catalogued the ways in which appearances vary across species, individuals, senses, circumstances, and cultural conventions, thereby undermining confidence in any single appearance as the truth.
The distinction between these two traditions is crucial. Academic skepticism is negative dogmatism: it asserts that knowledge is impossible. Pyrrhonian skepticism is zetetic: it asserts only that we have not yet found a way to decide, and it refuses even to affirm that decision is impossible. This difference matters because the Pyrrhonian position is more resilient: it cannot be refuted by producing a single piece of knowledge, since the Pyrrhonist can simply add that piece to the scales and seek a counterweight. The Academic position, by contrast, is vulnerable to any successful knowledge claim.
After a long period in which skepticism was known primarily through the writings of Sextus Empiricus, whose works were rediscovered in the sixteenth century, the early modern period transformed skepticism from a school of thought into a methodological challenge. The pivotal figure is René Descartes, whose Meditations (1641) used radical doubt as a tool for foundational inquiry. Descartes did not aim to remain in doubt but to use it as a filter: by doubting everything that could conceivably be false—including the evidence of the senses, the existence of his own body, and even the truths of mathematics—he hoped to find a bedrock of certainty on which to rebuild knowledge. His famous conclusion, cogito ergo sum ("I think, therefore I am"), was meant to be such a bedrock.
Descartes's innovation was to turn skepticism into a method rather than a destination. The skeptical arguments he deployed—the dream argument, the evil demon hypothesis—were not intended as genuine threats to knowledge but as instruments for testing the reliability of our cognitive faculties. This methodological skepticism proved enormously influential, but it also created a problem that would haunt subsequent philosophy. By setting the standard for knowledge at the level of indubitable certainty, Descartes made it difficult to see how any empirical claim could qualify. If the only beliefs that survive radical doubt are those whose denial is self-contradictory, then knowledge of the external world, other minds, and the future seems impossible.
David Hume radicalized this challenge in the eighteenth century. Hume accepted the empiricist premise that all knowledge derives from experience, but he argued that experience itself cannot justify many of our most basic beliefs. His most famous skeptical argument concerns causation and induction. We believe that the future will resemble the past—that bread will nourish, that fire will burn—but this belief, Hume argued, cannot be justified by reason, since it is not contradictory to suppose that the future will differ. Nor can it be justified by experience, since any appeal to past experience presupposes the very principle that the future will resemble the past. Hume concluded that our belief in causation is not rational but instinctive: a habit of mind that we cannot defend but also cannot abandon. He extended this skeptical analysis to the self (which he claimed we never actually perceive), to the external world (which we infer from impressions but never observe directly), and to miracles (which are always less probable than the error of those who report them).
Hume's skepticism is distinctive because it is mitigated: he did not conclude that we should suspend judgment, but that we should recognize the limits of reason and rely on custom, sentiment, and practical life. This position—sometimes called "skepticism with a wink"—acknowledges that our beliefs cannot meet the standard of rational justification while insisting that we cannot help but hold them. The result is a form of naturalism: our cognitive practices are explained by human nature rather than justified by rational argument.
Immanuel Kant attempted to answer both Descartes and Hume by redefining the terms of the debate. In the Critique of Pure Reason (1781), Kant argued that skepticism arises from a misunderstanding of the relationship between the mind and the world. He distinguished between phenomena—things as they appear to us, structured by the mind's own forms of intuition (space and time) and categories of understanding (causality, substance, unity)—and noumena—things as they are in themselves, independent of our cognitive apparatus. Knowledge, Kant argued, is possible only of phenomena. We can know with certainty that every event has a cause, not because we have observed this to be true, but because causality is a condition of experience itself: without it, we could not have coherent experiences at all.
Kant's response to skepticism is thus to limit knowledge to make room for faith, as he put it. By showing that the mind actively structures experience, he argued that the skeptic's demand for knowledge of things-in-themselves is incoherent: we cannot know what it would even mean for something to be knowable outside the conditions of possible experience. This "transcendental" approach—asking what must be true for experience to be possible—became the template for much subsequent anti-skeptical philosophy. However, it also left a troubling residue: the noumenal realm remains forever unknowable, and Kant's own account of it is notoriously difficult to state without contradiction.
The Kantian legacy is visible in the phenomenological tradition of Edmund Husserl and Maurice Merleau-Ponty, who sought to describe the structures of experience without presupposing the existence of an external world, and in the pragmatist tradition of Charles Sanders Peirce, William James, and John Dewey, who argued that the skeptic's demand for certainty rests on a mistaken conception of inquiry. For the pragmatists, knowledge is not a mirror of reality but a tool for coping with problems; a belief is justified if it works in practice, not if it survives every possible doubt. This approach dissolves skepticism by changing the subject: instead of asking whether we can know the world as it is in itself, we ask whether our beliefs are reliable guides to action.
In the twentieth and twenty-first centuries, skepticism has remained a central problem in epistemology, but it has fragmented into several distinct research programmes.
Ordinary language philosophy, associated with the later Ludwig Wittgenstein and J. L. Austin, argued that skeptical doubts are a kind of linguistic confusion. When we say "I know" in ordinary life, we are not making a claim that requires certainty; we are expressing a degree of confidence that is appropriate to the context. The skeptic's demand for certainty—that I know I am not a brain in a vat, that I know the world did not begin five minutes ago—is not a deeper version of ordinary knowledge claims but a misuse of language. Wittgenstein's metaphor of the "hinge" is instructive: some propositions (that the earth has existed for a long time, that I have two hands) are not beliefs we hold on the basis of evidence but the framework within which evidence can be assessed at all. To doubt them is not to be more rigorous but to step outside the language game altogether.
Contextualism formalizes this insight within the framework of formal epistemology. According to contextualists such as Keith DeRose and Stewart Cohen, the truth conditions of "S knows that p" vary with the context of the speaker. In ordinary contexts, the standards for knowledge are low: I know that I parked my car on Main Street because I remember doing so. In philosophical contexts, the standards are raised: I do not know that I parked on Main Street unless I can rule out the possibility that I am dreaming or that my memory has been manipulated. Contextualism thus explains the appeal of skepticism without conceding its conclusion: the skeptic is right that, by the standards of the philosophy seminar, we know very little; but this does not show that, by the standards of everyday life, we do not know a great deal.
Relevant alternatives theory, developed by Fred Dretske and Alvin Goldman, offers a different response. According to this view, to know that p, I must be able to rule out the relevant alternatives to p, not every conceivable alternative. When I look at a zebra in the zoo, I know it is a zebra because I can rule out the relevant alternative that it is a horse painted with stripes. But I do not need to rule out the irrelevant alternative that it is a cleverly disguised robot, because that possibility is not live in the context. The skeptic's error is to treat every alternative as relevant, thereby setting an impossible standard. This approach has the virtue of explaining why skeptical scenarios (dreams, evil demons, brains in vats) seem compelling in the abstract but irrelevant in practice.
Externalism in epistemology, associated with Alvin Goldman, Robert Nozick, and others, rejects the internalist assumption that a knower must have access to the grounds of her knowledge. According to externalism, a belief counts as knowledge if it is produced by a reliable process—one that tends to produce true beliefs—whether or not the believer can articulate or defend that reliability. On this view, my belief that there is a tree before me is knowledge if it is caused by the tree in the right way, through a properly functioning perceptual system, even if I cannot rule out the possibility that I am dreaming. Externalism thus sidesteps the skeptic's demand for justification: knowledge does not require that I be able to defeat every skeptical hypothesis, only that my belief-forming mechanisms are in fact reliable.
Skeptical invariantism, defended by Peter Unger and others, takes the opposite tack: it accepts the skeptic's conclusion and argues that we do not, strictly speaking, know very much at all. On this view, knowledge requires certainty, and certainty is rarely if ever achieved. Our ordinary claims to knowledge are therefore false, though they may be useful or excusable. This position is unpopular but logically resilient: it concedes the skeptic's premises and conclusion, and its main burden is to explain why we so confidently make claims that are, on its account, false.
A persistent structural problem in skepticism is the problem of the criterion, first articulated by the ancient skeptics and revived in the early modern period. To know anything, we must have a criterion for distinguishing true from false beliefs. But to establish that our criterion is reliable, we need a further criterion, and so on indefinitely. This regress—sometimes called the Agrippan trilemma after the Pyrrhonian philosopher Agrippa—offers three possibilities: we can stop at an unjustified assumption (dogmatism), we can appeal to a circular justification (the criterion is justified by the very beliefs it is used to assess), or we can regress indefinitely. None of these options seems to provide a satisfactory foundation for knowledge.
Responses to the trilemma define much of the contemporary landscape. Foundationalism accepts the first horn: there are basic beliefs that are self-justifying or non-inferentially justified, and all other knowledge is built upon them. Coherentism accepts the second horn: beliefs are justified by their mutual support within a system, and the system as a whole is justified by its coherence. Infinitism, defended by Peter Klein, accepts the third horn: justification is an infinite chain of reasons, and this is not a defect but the nature of justification itself. Each position has its difficulties, and the debate among them remains unresolved.
Contemporary skepticism is not a single doctrine but a family of problems and responses. The most active areas of research include the skepticism of the external world (can we know that there is a world outside our minds?), skepticism about other minds (can we know that others have conscious experiences?), skepticism about induction (can we know that the future will resemble the past?), and skepticism about moral and religious knowledge (can we know what is right or whether God exists?). These are not always treated as separate problems; advances in one area often have implications for others.
The field is also increasingly informed by formal epistemology, which uses probability theory and decision theory to model skeptical arguments. For example, the preface paradox and the lottery paradox show that there are cases where it seems rational to hold a belief while also acknowledging that it is probably false, challenging the idea that knowledge requires high probability. Bayesian approaches to confirmation theory have been used to assess the force of skeptical hypotheses: if the hypothesis that I am a brain in a vat has very low prior probability, then the evidence of my senses may be sufficient to confirm the existence of the external world, even if it cannot refute the skeptical hypothesis with certainty.
At the same time, there is a growing interest in the history of skepticism as a resource for contemporary philosophy. The rediscovery of Sextus Empiricus in the sixteenth century is now recognized as a pivotal event in the development of modern philosophy, and scholars have traced the influence of Pyrrhonian arguments on Descartes, Hume, Kant, and their successors. This historical work has complicated the standard narrative that skepticism is a problem to be solved, suggesting instead that it is a permanent feature of the philosophical landscape—a challenge that can be managed but not eliminated.
The durable lesson of skepticism is that the demand for certainty is not the same as the demand for knowledge. A mature epistemology must distinguish between the two, recognizing that knowledge may be possible even where certainty is not, and that the skeptic's scenarios, while they cannot be refuted, may not be relevant to the ordinary conduct of inquiry. Whether this constitutes a refutation of skepticism or a concession to it remains, appropriately, an open question.