Note: I wrote the following essay in my Epistemology and Metaphysics exam earlier this year. So it lacks footnotes/references and may be a bit scruffy in places (we were only given an hour per essay, after all). But I quite like it still, so thought I'd type it up here - with some minor editing, including the addition of hyperlinks. I don't recall the exact question, but it was something along the lines of "What is skepticism, and what rational responses can be made to it?"
Skepticism is the thesis that we cannot obtain genuine knowledge - that even our most basic beliefs about the world are epistemically unjustified. The basis for such an extreme claim is found in the apparent impossibility of ruling out various skeptical scenarios or hypotheses which are inconsistent with our everyday beliefs. For example, it seems possible that we could be brains in vats (BIVs), being 'fed' experiences by a mad scientist electrically stimulating our brains in such a way as to give rise to the exact same phenomenal experiences that we usually assume are caused by reality. The skeptic argues that because we are unable to distinguish the BIV scenario from Realism, we are not justified in assuming either one to be the case. Instead, the skeptic argues, the rational thing to do is suspend belief.
In everyday life, this strikes us as a rather implausible suggestion. As the Humean response goes, we are 'natural-born believers'. No matter the philosophical force of skepticism, it is of little practical import. Many philosophers, however, have thought that we can actually beat skepticism on its own terms. Various rational responses have been made to try to refute the skeptic's argument.
As the skeptic's argument is based on the empirical indiscernibility of Realism from Anti-Realism, the most obvious response would be to try to find an a priori reason for preferring the former. Most of these attempts fail - for example, we cannot assign probabilities to these cases. Nor can we say Realism is necessarily any simpler or more parsimonious.
Some have suggested a semantic response, however. Hilary Putnam argues that when an envatted person says "there is a tree", they speak the truth - given what they mean by 'there' and 'tree' (something like "a tree in-the-image").
Putnam advocates a form of anti-Realism here, but we need not follow him in that respect. David Chalmers has a similar response which sees the "matrix" (the world the envatted people think they occupy) as a genuinely real world in itself. He offers a complete metaphysical analysis of it, arguing that it is a world whereby:
(1) It is computational at the most fundamental level of reality;
(2) Inhabitants' minds exist outside of physical space-time, but interact with physical processes;
(3) Physical reality was created by beings outside of (the matrix's) space-time.
This nullifies the skeptical power of the BIV hypothesis, for we can say that envatted people still have largely true beliefs about their world.
Alternatively, if one is dissatisfied by the a priori responses, one must look elsewhere than foundationalist internalism for epistemic justification.
One possible alternative is coherentism. Here, a belief is said to be justified to the extent that it coheres with our other beliefs. Since we are 'natural born believers', and perception causes (though does not itself justify) us to have beliefs about the natural world, it follows that Realism - as a general framework - coheres better with our existing beliefs, than does anti-Realism. We are thus justified in rejecting the BIV hypothesis, due to the incoherence it would bring to our wider belief system.
However, coherentism cannot judge between equally-coherent alternative systems. So the BIV-believer could always insist that his own beliefs are perfectly coherent (and thus justified) - even though they may clash with ours. Coherentism forces a sort of neutrality, then. If we want to go on the offensive, and argue that common-sense / Realist beliefs are not merely justified, but epistemically obligatory, then we will require the philosophical power of epistemic externalism.
According to externalism, justification arises in part from conditions external to the agent. That is, a belief could be justified without the agent subjectively realising that this was so. This instantly overcomes the skeptic's arguments, for the indiscernibility problem becomes irrelevant.
Reliabilism, for example, merely requires that a justified belief be one which was formed by a reliable process. Now, we may not be aware of whether perception is reliable. But, nevertheless, there is some fact of the matter. If it is reliable, then my beliefs about the world will (according to Reliabilism) be justified.
Contextualism is another popular response to skepticism. It suggests that the standards required for knowledge vary according to context - so although we may not know we are not BIVs according to 'high standards' contexts, we can, contextualists claim, know our everyday beliefs based on the 'lower standards' appropriate in those contexts.
The problem with such a response is that if we are foundationalist internalists, then there is no possible evidence to recommend common-sense Realism over the BIV hypothesis. This means that no matter how low the standards are set, we can still never have justification for our everyday beliefs. For contextualism to be any use at all, it must work in tandem with one of the previously discussed solutions - most usually, externalism. (A common answr is the 'relevant alternatives' view, whereby we only need to rule out alternatives that are relevant to the present context. This is at heart an externalist theory though, since 'relevance' is an appeal to factors external to the agent.)
Perhaps the most impressive such fusion is Keith DeRose's mix of contextualism and possible-worlds externalism. The core idea is that S's belief that P is justified to the extent to which it satisfies the condition S believes that P iff P is true in close possible worlds. The justification is stronger, the more close possible worlds there are where this condition holds; that is, where S's belief "tracks the truth" (as Nozick put it) through those worlds.
DeRose's contextualist twist is that the 'strength' of justification required for knowledge will vary according to context. Our belief that we are not BIVs is very strong, for it is accurate in all nearby possible worlds. The closest world in which we are mistaken (i.e. the BIV world) is a very distant one indeed!
So there are several different rational responses one could make to the skeptic. Skepticism demonstrates that an a posteriori, foundationalist, internalist view of knowledge about reality is untenable. We can thus overcome the objection by denying any one of the three conditions. We need a response which is either a priori, coherentist, or externalist in nature.