6 Skepticism and Its Challenges
Examine skeptical challenges to knowledge of the external world, other minds, the future, and the structure of justification, along with influential responses.
What challenges
questions whether—and how—we know. It can target a particular kind of claim, such as predictions about the future, or challenge knowledge more broadly. A skeptical argument typically introduces a possibility that would make an ordinary belief false or unknowable, then argues that our evidence cannot rule out that possibility. is a challenge to examine, not simply the claim that nothing is known.
about the external world
We ordinarily believe that objects exist independently of our minds: for example, that a book is on a desk even when nobody looks at it. But perception can mislead us, and dreams can seem vivid while they last. Descartes sharpened this concern with the : imagine a powerful deceiver producing experiences just like ours while the world we seem to perceive does not exist. A is a contemporary version in which experiences are artificially produced.
A skeptical argument about whether we know that we have hands can be stated as follows:
If I know that I have hands, then I know that I am not a deceived .
I do not know that I am not a deceived .
Therefore, I do not know that I have hands.
The first premise relies on : if you know one claim and know that it entails another, you can know the second. Skeptics argue that ordinary evidence, such as seeing and touching one’s hands, does not exclude every deceptive scenario. The argument’s force depends on whether its premises and the closure principle are right.
Responses to external-world
Several responses challenge different parts of the skeptical argument. reasons from ordinary knowledge: if he knows that he has hands, then he knows that the skeptical scenario is false. This reverses the skeptic’s argument, though critics say it does not explain how we know that the skeptical alternative is false.
Contextualists propose that standards for saying “knows” rise when skeptical possibilities become salient. In ordinary conversation, someone may know that there is a desk nearby; in a philosophical discussion that demands the exclusion of far-fetched scenarios, the same claim may not meet the heightened standard. explains some shifts in how “know” is used, but it does not settle whether the view is correct.
Fallibilists deny that knowledge must eliminate every conceivable possibility of error. Externalist theories, such as , allow a belief to count as knowledge when it is produced by a reliable process, even if the knower cannot independently prove that the process is reliable. A takes another route: it claims that some facts questioned by the skeptic are necessary conditions for coherent experience or thought. These approaches each reject or revise a different part of the skeptical argument.
Knowing other minds
We experience our own thoughts and feelings directly, but other people’s mental lives are not open to us in the same way. We observe their words, expressions, and actions, and can ask whether these are accompanied by genuine thought or feeling rather than produced by mindless beings that merely behave as people do.
The is one traditional reply: one’s own pain is often associated with certain behavior, and other people behave similarly, so their behavior is evidence that they feel pain too. Another reply treats other minds as the best explanation of people’s flexible, complex behavior. Testimony and ordinary interaction may also provide knowledge of what others feel or think. These replies do not give direct access to another person’s experience, but they show how knowledge of minds might rest on more than a single inference from outward appearance.
and expectations about the future
uses observed cases to support conclusions about unobserved cases. For example, after seeing that a medicine has helped many patients, we may expect it to help a new patient. Inductive reasoning is essential in everyday life and science, but its conclusions are not guaranteed by its evidence: previous cases do not logically entail that the next case will be similar.
David Hume’s asks what justifies projecting past regularities into the future. The projection cannot be proved deductively, because the future could differ without contradiction. Nor can we justify by arguing that it has worked in the past and will therefore work again: that argument already assumes the reliability of that it is meant to establish. Hume did not deny that people make inductive inferences; he questioned whether reason can provide a non-circular foundation for them.
Responses include accepting as a practical or basic method of reasoning rather than demanding a proof that it must succeed, using probability to measure how evidence supports competing hypotheses, and defending . These approaches offer ways to understand or defend inductive practice, though philosophers disagree about whether they fully answer Hume’s challenge.
The structure of justification
A general challenge asks whether any belief is justified. If a belief needs support from another belief, that supporting belief may need support in turn. This raises the problem of : does justification end, loop, or continue without end?
Three major theories offer different accounts of the structure of justification. says that some beliefs—perhaps beliefs based directly on experience—can provide a foundation for other beliefs without depending on further beliefs. says that beliefs are justified through mutual support within a coherent system rather than by resting on a single foundation. allows justification to involve an endless chain of reasons. These are theories about how justification is structured, not simple proofs that is false.
What skeptical challenges reveal
Skeptical arguments remain useful even when their conclusions are rejected. They reveal assumptions behind claims to knowledge: that perception is trustworthy, that other people have minds, that patterns persist, and that reasons can support belief without circularity. The central question is whether knowledge requires certainty or whether well-grounded but fallible belief is enough.