Skepticism: From Pyrrho and Sextus Empiricus to Descartes, Hume, and the Brain in a Vat

Can we know anything? Skepticism is the current that takes this question seriously — and, in its most radical versions, answers no, or at least that we have no way of knowing. More than an attitude of casual doubt, it is a position (and, in antiquity, a way of life) that became the great engine of epistemology: much of the theory of knowledge was born of the attempt to answer the skeptical challenge. This article surveys its ancient and modern forms and the main responses it has received.

1. Ancient skepticism: Pyrrhonism

The most influential skeptical tradition goes back to Pyrrho of Elis (c. 360–270 BCE), who left no writings and reaches us above all through Sextus Empiricus (2nd–3rd c. CE), whose Outlines of Pyrrhonism systematize the school. Pyrrhonism does not dogmatically assert that “nothing can be known” — that would itself be a dogma. It operates through modes (or tropes): for every thesis it sets out an opposing thesis of equal force, and from this equipollence follows epoché, the suspension of judgment.

Agrippa’s tropes condense the strategy into what became known as the Agrippan trilemma (or Münchhausen trilemma): every attempt to justify a belief ends either in infinite regress (each reason requires another reason), or in circularity (the proof presupposes what it was to prove), or in an arbitrary assumption (the chain is broken off at an unjustified point). The practical aim, however, is not despair: by suspending judgment about what is not evident, the skeptic was said to attain ataraxia, tranquility of soul. See also the article on the epoché.

In parallel, the New Academy of Arcesilaus and Carneades developed an “Academic” skepticism aimed against the Stoics: it denied that there is any “cataleptic impression” (a self-certifying perception, a guarantee of truth) and proposed being guided by the probable (pithanón), since certainty is unattainable.

2. The modern revival: Montaigne and Descartes

Rediscovered in the Renaissance, skepticism reappears in Montaigne (“Que sais-je?” — “What do I know?”), who uses it as an antidote to dogmatic arrogance (see the article on Montaigne’s skepticism). But it is Descartes who gives it its decisive modern form — not to surrender to it, but to defeat it. In his Meditations (1641), he submits everything to methodic doubt: the senses deceive; perhaps I am dreaming; and, at the limit, an evil demon could be falsifying even my mathematical certainties. From this hyperbolic skepticism Descartes nonetheless extracts an unshakable point: even if deceived, I think — therefore I exist. The cogito becomes the foundation on which he aims to rebuild knowledge (see the cogito). Descartes, then, is not a skeptic: he is the philosopher who uses doubt as a method.

3. Hume and mitigated skepticism

The skeptical challenge takes its sharpest form with David Hume. Two points are decisive. First, the problem of induction: there is no rational justification for expecting the future to resemble the past — only habit makes us predict that the sun will rise tomorrow. Second, causation: we never perceive a “necessary connection” between cause and effect, only the constant conjunction of events; the necessity we ascribe to them is a projection of custom, not a datum of reason (see the article on Hume).

Hume extends the doubt to the external world and to the self. But he notes that radical doubt is unsustainable in practice: nature, through custom and feeling, tears us from it the moment we leave the study. Hence his mitigated skepticism: in order to live, we accept what we cannot prove — an epistemic humility, not paralysis.

4. The modern skeptical argument: the brain in a vat

Contemporary epistemology reformulates the challenge with a thought experiment: what if I were a brain in a vat, stimulated by a computer to have exactly the experiences I am having now? (It is the philosophical version of the Matrix scenario, heir to the Cartesian evil demon.) The argument has logical force: I cannot rule out that hypothesis; and if I cannot rule it out, then — by the closure principle (if I know that p and that p entails q, I know that q) — I do not even know that I have hands. Global skepticism rests on this impossibility of excluding the deceiving scenario.

5. The responses

Philosophy has offered several ways out:

• Moore’s common sense. G. E. Moore, in “Proof of an External World” (1939), raised his hand and said: “here is one hand.” His point is strategic: the certainty that I have hands is stronger than that of any skeptical premise; it is therefore more reasonable to reject the premise than to deny the obvious.
• Putnam’s semantic anti-skepticism. Hilary Putnam argued, in Reason, Truth and History (1981), that the brain-in-a-vat hypothesis is self-refuting: if I had always been a brain in a vat, my word “vat” could not refer to real vats (I never had causal contact with them); so the sentence “I am a brain in a vat” would be false if uttered by a brain in a vat.
• Contextualism. Keith DeRose and David Lewis hold that “to know” is context-sensitive: in ordinary conversation the standards for knowledge are low and I do know that I have hands; when the skeptic raises the standards, the word’s demand changes. There is no contradiction — there is a shift of context.
• Externalism/reliabilism. If to know is to have a true belief produced by a reliable process, then I need not rule out every deceiving scenario in order to know: it is enough that my cognitive mechanism be, in fact, reliable.
• Wittgenstein’s “hinges.” In On Certainty (1969), Wittgenstein suggests that certain certainties (“the world has existed for a long time,” “I have a body”) are not pieces of knowledge to be justified, but the hinges on which doubt itself turns: to doubt everything is impossible, because doubt presupposes a background that is not questioned.

6. Local skepticisms

Besides global skepticism (about all knowledge), there are local skepticisms, restricted to one domain: about the external world, about other minds (how do I know that others have experiences?), about induction, about moral values, or about religion. Even those who reject global skepticism usually take seriously some of these particular challenges — and the Gettier problem shows that even the definition of knowledge remains open.

7. Assessment

Skepticism is rarely “refuted” once and for all; it is, rather, the permanent adversary that forces every theory of knowledge to justify itself. Its forms should be distinguished: the ancient Pyrrhonist sought tranquility by suspending judgment, whereas the modern skeptic poses a theoretical problem about the limits of knowledge. Perhaps the most durable lesson is Hume’s: we may have no decisive refutation of radical doubt, yet life — and science — go on, guided by a knowledge that is fallible, revisable, and nonetheless sufficient for action. The question “what can we know?” remains open — and that is precisely why it stays at the center of philosophy.

Essential reading

• Sextus Empiricus, Outlines of Pyrrhonism (2nd–3rd c.).
• René Descartes, Meditations on First Philosophy (1641).
• David Hume, An Enquiry concerning Human Understanding (1748).
• G. E. Moore, “Proof of an External World” (1939).
• Ludwig Wittgenstein, On Certainty (1969).
• Hilary Putnam, Reason, Truth and History (1981) — the brain-in-a-vat argument.

← Articles