This is the the twelfth 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.
So far I've looked at the problem Hume raises for scientific knowledge and some less helpful attempts to respond. This post looks at some responses with more promise. Each corresponds to a style of responding to skepticism that I've already discussed in earlier posts.
Hume didn't intend his argument to lead to skepticism any more than Descartes intended his own to lead to skepticism. Hume's response wasn't anything like Descartes', however. It's actually a good example of Forced Pragmatism (see the pragmatism post) with respect to this very specific brand of skepticism. He admits that we can't know anything about scientific laws. He insists that we have to pretend we do. We have no choice. That's simply how we're constructed. We form habits or customs, and then we rely on those as we observe uniformities in nature. As things consistently happen together, we conclude that something must be making it that way. We have no rational reason to believe such a thing, but we're brought to believe it by our need to have simple explanations for things in order for our existence to be livable. This explains why we do believe in the results of science, but does it answer why we should?
The other kinds of pragmatism above might come in here as well. Maybe we ought to believe in scientific laws for moral reasons. Maybe it's simply in our best interests to do so. Of course, this probably will fall prey to the objection that it may only have been in our best interests to do so in the past. We don't know that it's in our best interests unless we know the future will be like the past, and that's the issue in question. Even so, "Why Not?" Pragmatism may be as good here as with Cartesian skepticism.
The A Priori Knowledge Response:
As the A Priori Responses to Skepticism post explains, a priori knowledge is the sort of knowledge you can come to without looking into the world or having certain experiences (e.g. "2+2=4" or "nothing can be red and green all over in exactly the same way at the same time and place" or "all bachelors are unmarried" or "every triangle has three sides"). You can know these sorts of things just by thinking about them. You can know these things from before you even look into the world to discover anything about them.
The Principle of Sufficient Reason (or PSR) provides an objection to Hume's argument. PSR states that there's an explanation for everything that happens. You can explain my existence from my parents and their interactions before my conception. You can explain lots of things similarly. That's what science does. PSR doesn't mean we will always be able to get the explanation we look for, but it does say that there must be one, or the thing would never have happened. PSR seems plausible enough that Immanuel Kant said we can know it a priori -- just by thinking about it. If so, then laws of nature do affect what happens. We may not always get it right, but there's something to get right. Hume disagrees, thinking that we can't know PSR (he even says he's sure it’s false), but it seems like a plausible enough principle to many people. We do seem to rely on it all the time.
One problem with this approach is that even if we know about PSR a priori, we don't know scientific laws a priori. We have to investigate them, which means we don't have certainty about which laws are right. But we might be able to say we know that some laws must be true to explain what we see in our experiments. Then it comes down to finding what seem to be the most reasonable laws to believe, and science is worth doing once again. But this all assumes that laws that have been true in the past will continue to be true in the future. All we have been able to argue so far is that we can infer that there are laws governing what has happened. For all we know, the laws could change tomorrow. We still haven't solved Hume's basic problem about knowing that the future will be like the past, even though we have solved the problem about whether there are any scientific laws at all.
Reliabilism can come alongside the a priori response to solve that last problem. If we don't need to establish to ourselves or to anyone else what methods can bring us knowledge, but if they only have to be reliable at bringing true beliefs to us, then we may be ok. If PSR is true, and nature is reliable, with laws governing how things go, and if this is true of the future, then our trusting in them is using a reliable method of getting to true beliefs. According to reliabilism, we can then know things about future scientific experiments, because we do know the laws if we happen to get them right. Even if we only get them approximately right, we have knowledge about the things the laws we come up with are true about. So Newton did know about what would happen when he dropped an apple, but he didn't know about what would happen if you accelerated the apple so that it was close to the speed of light, since his theory would have led to false beliefs about that. But it was reliable enough about most things we encounter every day, so it leads to knowledge. People who have objections to reliablism won't find it any more plausible for this problem, but it has the nice feature of solving this problem if it's correct.
Next in the series will be five posts on No-Evidence Arguments, starting with the Divine Silence Argument.