Concept

Cartesian Skepticism

Intro

How do you know you are not dreaming right now? Or hooked into a computer simulation? Or being deceived by a powerful evil being feeding fake experiences into your mind? Whatever you point to as evidence that this is real, the dreamer or the simulated person could point to the same thing. Their experience would look exactly like yours.

That is the basic move of Cartesian skepticism. It comes from Rene Descartes, who in 1641 published the Meditations on First Philosophy. Descartes himself was not trying to defend skepticism. He was trying to find something he could not doubt, then rebuild knowledge on that foundation. To do that, he first dug down to bedrock by doubting everything he possibly could. He invented the famous "evil demon" who might be feeding him false sensations about the entire external world.

Modern versions of the argument do not bother with the original goal. They just take the doubt and run with it. The brain in a vat. The Matrix. The dream argument. They all work the same way: invent a scenario where your experience looks just like it does now, but where your beliefs about the external world are completely false. If you cannot rule out the scenario, the argument says, then you cannot really claim to know what you think you know.

The argument has shaped modern philosophy more than almost any other single problem. It pushes on what it means to "know" anything at all. Most epistemologists today treat it as a problem to solve, not a doctrine to refute.

This page lays out Descartes's original three doubts, the modern brain-in-a-vat version, the closure principle that makes the argument bite, and the main responses to it.

In full

The species of skepticism that proceeds via skeptical hypotheses, vivid scenarios (the dream, the evil demon, the brain in a vat, the Matrix) in which one's experience is just as it is now while one's beliefs about the external world are radically false. Cartesian skepticism is named for Rene Descartes, who deployed it methodologically in Meditations on First Philosophy (1641), but the contemporary argument runs free of Descartes's specific reconstruction project: most current epistemologists treat the argument as a problem to solve, not a doctrine to refute. The argument's structural form, leveraging the closure principle to derive doubt about ordinary knowledge from doubt about a skeptical hypothesis, has shaped late-20th- and early-21st-c. analytic epistemology more than any other single problem.

Descartes's three doubts

In the First Meditation, Descartes escalates three doubts:

Sense deception, "the senses sometimes deceive." Examples: the bent oar in water, distant towers that look round but are square. Local doubt, easily contained.
The dream argument, "I am sometimes deceived in dreams about what I take to be perceptions; there is no certain mark distinguishing waking from dreaming, so I cannot be certain that what I take to be perception is not dream." This generalizes the sense-deception worry to any perceptual belief.
The evil demon (or, in a more secular version, deception by God or systematic mathematical error), "Some malignant genius, no less powerful than deceitful, has employed all his energies to deceive me." This generalizes further: even a priori beliefs (mathematical, logical) might be products of cognitive distortion.

The three are escalating in scope: (1) doubts particular perceptions; (2) doubts the perceptual faculty as a class; (3) doubts even reasoning itself.

Descartes's purpose is methodological, find what survives radical doubt and rebuild knowledge from there. Cogito, ergo sum, I think, therefore I am, is the unshakeable Archimedean point. From the cogito he tries to prove God's existence and goodness, then trust clear-and-distinct ideas as veridical, then re-establish external-world knowledge via God's non-deception.

This reconstruction project has fared poorly in the philosophical tradition; the Cartesian Circle objection (we use clear-and-distinct ideas to prove God, then use God's veracity to validate clear-and-distinct ideas) is widely accepted as a fatal flaw. So what remains in contemporary philosophy is Descartes's destructive phase, treated as a problem the rest of epistemology has to solve.

The contemporary reformulation

Modern Cartesian skepticism abstracts from Descartes's specific scenarios. The structural argument, in its cleanest form (often attributed in this form to Stroud's The Significance of Philosophical Scepticism, 1984):

(P1) I do not know that I am not a brain in a vat (BIV) being electrically stimulated to have exactly the experiences I am having now.
(P2) If I know that I have hands, then I know that I am not a brain in a vat. (Closure: knowledge is closed under known logical entailment, see Closure Principle.)
(C) Therefore I do not know that I have hands.

The argument generalizes: replace "have hands" with any external-world proposition. Replace "BIV" with any skeptical hypothesis (Descartes's evil demon, Putnam's BIV, Nozick's "I'm dreaming," Bostrom's simulation hypothesis, the Matrix, etc.). The form of the argument is what matters.

Why P1 is intuitive

The brain-in-a-vat scenario is empirically equivalent to the actual situation by stipulation, the BIV has exactly the experiences I am having now. Whatever evidence I might cite for my having hands (visual experience, kinesthetic feedback, others reporting that I have hands) would be exactly the same in the BIV scenario. So I have no evidence that distinguishes the two. Without distinguishing evidence, I lack the kind of ground that knowledge requires, or so the skeptical thought goes.

Why P2 is intuitive

Closure under known entailment is a near-universal feature of how we think about knowledge. If you know p, and you know that p entails q, then you should know q (or at least be in a position to know q). Otherwise knowledge would be epistemically inert, knowing things wouldn't license drawing conclusions. P2 is closure applied to the negation of skeptical hypotheses: I-have-hands entails I-am-not-a-handless-BIV; so if I know I have hands, I know I'm not a handless BIV.

Why the conclusion is unwelcome

If the argument is sound, ordinary perceptual knowledge is impossible. We don't know we have hands, that there's a chair we're sitting in, that the sun rises tomorrow, etc. This conflicts radically with the most secure-seeming beliefs we have. The modus tollens version (Mooreanism) runs the argument in reverse: since I clearly do know I have hands, and the closure step is sound, I must know I'm not a BIV, even though I have no evidence that distinguishes my situation from a BIV's. (See Mooreanism.)

The geography of responses

The argument's three load-bearing components, P1 (no skeptical-hypothesis-knowledge), P2 (closure), and the implicit assumption that knowledge requires evidence-discrimination, give three principal places to push back. Sub-hubs:

Response	Pushes back at	Hub
Mooreanism	P1 (we do know we're not BIVs because we know we have hands)	Mooreanism
Closure denial / nonclosure (Dretske, Nozick)	The closure step in P2	Closure Principle
Contextualism (DeRose, Lewis, Cohen)	The univocity of "know" between contexts	Epistemic Contextualism
Externalism / reliabilism	The implicit internalist requirement that we have evidence-discrimination	(covered in Epistemology)
Reformed Epistemology	The implicit demand for inferential support	Reformed Epistemology
Presuppositionalism	The skeptic's right to use the very rational standards she questions	Presuppositionalism

Key contemporary works

Hilary Putnam, "Brains in a Vat" in Reason, Truth and History (1981), the BIV terminology entered analytic philosophy here. Putnam offered a semantic anti-skeptical argument ("I am a BIV" is self-defeating because the words don't refer to brains and vats in the BIV scenario). The argument is much-discussed but not widely accepted as decisive.
Barry Stroud, The Significance of Philosophical Scepticism (1984), the major book-length statement of why the skeptical argument is so hard.
Robert Nozick, Philosophical Explanations (1981), counterfactual / tracking theory of knowledge; denies closure as the price of accommodating ordinary knowledge.
Fred Dretske, "Epistemic Operators" (1970) and Knowledge and the Flow of Information (1981), the original modern denial of closure.
Keith DeRose, "Solving the Skeptical Problem" (1995), contextualist solution.
Ernest Sosa, A Virtue Epistemology (2007), virtue-theoretic Mooreanism; reflective vs animal knowledge.
Duncan Pritchard, Epistemic Angst (2015), recent synthesis; identifies two distinct skeptical problems and gives a hybrid solution.
James Pryor, "The Skeptic and the Dogmatist" (2000), contemporary defense of immediate justification (the dogmatist line behind Mooreanism).

Christian-apologetic engagement

The Cartesian-skeptical argument is the implicit horizon against which contemporary Christian epistemology operates. Three principal Christian responses:

Reformed Epistemology (Plantinga, Wolterstorff, Alston)

Reformed Epistemology argues belief in God is properly basic, it has the same epistemic status as belief in other minds, the past, and the external world. None of these are inferentially supported by arguments that defeat skepticism; all are formed by reliable, non-inferential cognitive faculties (perception, memory, the sensus divinitatis). The Cartesian argument fails because it presupposes an internalist evidentialism that it cannot justify, and on a properly externalist epistemology, our cognitive faculties' actual reliability is what matters, not our internal ability to rule out remote possibilities.

This is structurally parallel to Mooreanism: belief in the external world is more secure than the premises of the skeptical argument; therefore the skeptical argument fails. Plantinga makes the parallel explicit (see Warranted Christian Belief, 2000).

Presuppositionalism (Van Til, Bahnsen, Frame)

Presuppositionalism argues skepticism is self-refuting. The skeptic uses logic, induction, and the assumption of meaningful linguistic communication to advance the skeptical argument; but the very rational order she uses cannot be justified except on the supposition of the Christian God who is the metaphysical ground of intelligibility. The skeptic is "stealing from God", using a worldview-dependent set of tools while denying the worldview that grounds them. The transcendental anti-skeptical argument of Stealing from God Argument is the developed form of this move.

Classical evidentialism / Reidian common sense

Classical apologists (Aquinas, Reid, modern: William Lane Craig, J. P. Moreland, Doug Geivett) accept some version of Mooreanism for ordinary perceptual knowledge, the senses are reliable enough to ground action and inference, and the existence of God is then argued via the cumulative natural-theological arguments. The Cartesian challenge is "more clever than serious"; it is taken seriously enough to be answered (typically by some combination of reliabilism + the natural-theology arguments) but not allowed to set the agenda.

The simulation hypothesis as contemporary BIV

Bostrom's simulation argument (2003) and the wider cultural awareness of the Matrix scenario have made BIV-style skepticism accessible to a non-specialist audience. The structural epistemology is the same: if we are simulated, our experiences are systematically misleading about the underlying reality. The apologetic-relevant point: the simulation hypothesis is not a defeater for theism; if anything it raises the question of the simulator, a designer who is, by hypothesis, vastly intelligent, powerful, and creative, and whose own existence is unsimulated. (One can pose this as a backdoor cosmological argument, though it has technical wrinkles.)

A meta-comment on the argument

The Cartesian argument is seductively difficult. Three observations that have shaped recent literature:

It is much easier to state the argument than to block it. Every published anti-skeptical strategy has serious open objections.
The strongest reply is not philosophical but psychological: virtually no one, no matter how committed to skepticism in writing, lives as a skeptic. This is Hume's own position. Whether this implies the argument is unsound or merely psychologically unbearable is itself a contested question.
The argument changes character depending on the standards of "knowledge" in play. The contextualist insight (see Epistemic Contextualism) is that this is not a vagueness or imprecision in our concept of knowledge but a feature: "know" picks out different relations in different conversational settings.