This is the the tenth post in my Theories of Knowledge and Reality series. Follow the link for more on the series and for links to other entries as they appear.
Reliabilism responds to skepticism by challenging one of Descartes' key methodological assumptions. He claims that you can't know anything unless you have absolute certainty of that thing, not just having a subjective sense of feeling certain but an absolute, objective understanding of why that thing must be true. No one, of course, has such certainty about most things we believe. Reliabilists just argue that we don't need to. Knowledge just doesn't involve that kind of certainty. Reliabilists simply deny the premise that everyone else seems to assume, that knowledge requires this idea of absolute certainty. Reliabilists consider such a notion ridiculous.
We know all sorts of things without being able to prove them to ourselves and without being able to rule out all the alternatives. How do we ordinarily use the word 'know'? We say we know all sorts of things. How do we find out what words generally mean? We see how people use them. In this case, people use the word 'know' when they don't have certainty.
One way to take this is to say that this is correct but only in ordinary contexts. In higher-standards contexts, the skeptic is right. The contextualist thus thinks the word changes its meaning from context to context. This explains why we feel the pull of skepticism in the philosophy room but then feel fine speaking of knowing things when the standards aren't so high. The reliabilist wants to resist this variability, at least at the level the contextualist thinks. It's correct to say we know things, not just in the ordinary context but even in the philosophy room. When someone raises the stakes by asking if we really know, all they're doing is illegitimately trying to change the meaning of the term to something that it simply doesn't mean. We can know without having to prove that we know, and when the skeptic tries to get us to treat it otherwise it's like the spoiler who says that a group of people isn't a family just because their kids are adopted or that Harry Potter isn't fantasy simply because it's in modern times. That's just not what the word means.
The reliabilist wants to say there's a better way to look at the evidence about how we use the word 'know'. The word just doesn't require the kind of certainty Descartes assumed it requires. When you ask in a philosophy class if we really know or if we know for certain, that adds a new shade of meaning to the word, but it's not the fundamental meaning of the word. That's a new use of the word but not the most common. So we are literally telling the truth when we say we know, and we would even be literally telling the truth if we use the word as we ordinarily use it but in a philosophical context. What brings us to the new meaning isn't being in a new context, as if the very existence of someone else's higher standards changes what the word means. That's what contextualism requires.
According to standard reliabilism, you can't just shift contexts and change the meaning of 'know' that drastically. Maybe you can express the higher-standards technical meaning of the word in a philosophical context, but you need to clarify that you're using the word is a specialized and normally illegitimate sense, or you've violated the norms of conversation, because the word simply doesn't mean that sort of thing. The reason we aren't so sure in the philosophy room is that we're asked about a stronger kind of knowledge, absolute knowledge. This would be like saying the table is flat and being right but then being asked if it's absolutely flat, only to say no. But it could be flat without being absolutely flat, according to this view. So we can know without knowing in a completely guaranteed and certain way, and that's what the word 'know' just means.
So what then is required for knowledge if it's not absolute and objective certainty? There are three things the reliabilist would require:
A. Knowledge requires a true belief. You can't know something that isn't even true. Maybe you haven't proved it to be true, but it has to be true.
B. Knowledge requires that you arrived at your belief through a reliable method of getting true beliefs. This is where the name 'reliabilism' comes from. The method doesn't have to lead to true beliefs in every case, as long as it's the sort of thing that would have a good track record. If it will generally give true beliefs, it's reliable and can lead to knowledge. If the skeptical scenarios are false, then our senses would be an example of this, and we could come to know things by following what our senses seem to be telling us.
C. Knowledge requires that your belief is caused in the right way. If you believe there's a sheep in the field, your belief better have been caused somehow by that sheep, or you may have arrived at a true belief accidentally (if, say, there's a real sheep in the field behind the sheep-shaped rock that you saw). Accidentally arriving at true beliefs by a reliable method doesn't lead to knowledge. Your belief needs to be connected in a reliable way with what it's about.
Note: a reliabilist hasn't tried to prove that our senses are reliable. Descartes would insist that he has to prove those very things, or he hasn't proved that he has hands. That's why he's so intent on giving a priori arguments for everything, because those are the only things he can have that level of certainty about.
A reliabilist is well aware of this objection but doesn't consider it to be worth worrying about. A reliabilist doesn't think you need to prove those things. Our senses need to be reliable. You do need to have hands to know that you have them, and your senses need to be a reliable way of getting that information to you. You don't need to prove that all these things are true. They just need to be true. If our senses are reliable, and the things we think we see exist, then we know about them. We don't need to have a proof that our senses are reliable or a proof that there's no evil genius deceiving us. It just needs to be true that our senses are reliable. If so, we can come to know things by using that reliable method.
This view can work for any reliable method, not just the senses. This will be important when we look at the responses to other kinds of skepticism, particularly skepticism about science and skepticism about God. If there are reliable ways to get true beliefs about things that are genuinely caused by that thing, then it can be knowledge, even if we can't prove that we know it and even if we have no access to the reason why our belief is true. Descartes simply wanted too much and thus built into knowledge more than it really requires. Maybe it's worth distinguishing which things we have absolute certainty about, but if we restrict knowledge to be just about those things then we're really pretending that we don't know things that we really do know.
Next: Skepticism About Science