The various sceptical arguments try to prove that we cannot know anything, or at least that we cannot know something. There’s always some possible scenario in which we are wrong about just about everything. A simple example is that if you were in fact dreaming right now, then likely pretty much nothing that you think that is going on around you is really happening.
Of course, we know what dreams are like, and we can see that the ongoing life around us is not such a delusion. It’s too permanent and consistent. But this does not suffice to remove the hypothetical reason to doubt, because it will always remain possible that we are under some greater delusion that permeates our entire life and/or prevents us from seeing how things really are. Maybe we never have been awake, so we cannot make comparisons. More generally, it can be said that maybe the world is so different from the way we imagine it that everything that we imagine we know about it is delusion.
Such possibilities cannot be ruled out for certain, but they’re not too worrying as such, provided that we can study the world. We have already found it to be very different from what we thought in many ways. It turns out that it’s the Earth that goes around the Sun, matter is made up of weird elementary particles, and so on. Even though we may be wrong about things now, we may be able to find out how they really are at some point in the future. In addition, what has happened in practice is that we don’t need to say that we have been wrong about everything, but rather that we have only known about some sides of how things actually are. Even though we previously had no idea about the elementary particles composing the objects that we know, the resultant objects are nevertheless in many ways as we imaged.
A greater threat would be if the world were not reasoned and ordered, in which case nothing could ever be known about it. If there were no general natural laws of any sort, anything could happen at any moment. For example, I could not rely on the chair on which I am sitting now to stay where it is, because there would be no reason for it not to just disappear. But in truth, the situation would be much more dramatic still; I could not, in fact, even exist, because even if there were anything at all in the universe that did not obey rules, everything would be in a state of total chaos. It may seem as if just one small object not acting according to the laws of nature would not mess up everything. But if it were limited in such a way as to be just one small object and not to mess up the laws of nature everywhere at once, it would obey some rules. If it were an object at all, or limited in any way, it would have rules. Rulelessness taking place in a limited area doesn’t yet make the world incomprehensible, because some parts of the world are outside its sphere of influence. For example, the randomness apparently inherent in quantum physics is handily limited to random choice between certain options, and also fuzzes out when moving from elementary particles to larger objects.
So, the world needs to be somehow rule-bound and understandable by reason in order for us to understand it. However, we cannot prove that it is. Fundamentally, this is about the problem of induction, though I suppose it’s not usually presented like this. Induction is the inference from things having worked in some way to the conclusion that they will continue to work that way. In other words, it derives rules that are assumed to apply to the world. For example, if water has so far always removed thirst, one can assume that it will go on doing so in the future, and one can apply this assumption to practice by drinking water when one gets thirsty.
However, we cannot prove that induction is a reliable way of obtaining knowledge. Not even probable knowledge. How could we prove that it always holds that there are regularities to be observed in the world so that information about the past can be used for inferences about the future? One might be tempted to say that it has worked that way before, but that would itself be induction and therefore circular — it would mean saying that we can rely on inference from past events because inference from past events proves it to be reliable.
The continuing lawfulness of the world cannot be proven, but it can at least be said that it’s rational to assume it. We could not know anything, nor exist, if the universe were not regular within at least some bounds. (From this, of course, it follows that since we exist, there is some regularity in the world. However, we cannot know whether it holds of, say, the future, and if it were not to, that would collapse everything.) Therefore, we lose nothing in assuming the world to be regular, because in the positive scenario we will be right and in the negative one all is already lost anyway.
There remains one more thing that can be doubted. Those speaking of the problem of induction often mention that induction is not a form of valid inference. Its conclusions do not necessarily follow from its premises. Things could turn out differently from the way induction suggests them to be, though it might prove it to be likely that they are that way. And then, one might mention also the fact that induction can only be “proven” using itself. This is ironic in that the last also applies to valid, logical, flawless deductive inference — the very thing that induction is supposed to be so inferior to.
We could not understand the world without laws, but we could not even think anything without the laws of logic applying. We know that some things follow from other things and that a thing cannot simultaneously be some way and not be that way. (The last is different from being something in between. There could be a gray ball or a ball with black and white stripes, but not a ball that would be black all over and white all over at the same time.) Such things seem perfectly obvious, and it’s not possible to even properly imagine something going against the laws of logic. However, we cannot prove that they are right. After all, we would have to apply them in the proof, just as in all other thought, but then the proof would be wrong if they were themselves wrong, so, just as with induction, we’d be stuck.
One might say, as with induction, that we’ll just have to assume the applicability of logic, because that’s the only way we’ll get anywhere. And in fact, we do have to. This doesn’t work the way it did before, however. If logic were not to hold, we would not even know this. This thought is also a result of inference. Everything said here is the result of thinking and inference. If logic were not to hold, then even if we knew how something was, we would not be able to know that it isn’t the opposite way, much less know anything that follows from it. Therefore, scepticism about logic renders void everything said here, this included, but also the fact that it counters itself. Any counterargument whatsoever against this will automatically be rendered void because all of them will require logic to work.
To summarise, it could be said that we have discussed three kinds of scepticism:
- About beliefs. What if we are wrong about things?
- About lawfulness. What if the world isn’t comprehensible at all?
- About logic. What if ????
The first can be answered by saying that there’s nothing else to it than to try to find out how things really are. The second can be answered by saying that it’s rational to assume the comprehensibility of the world even though it cannot be proven. The third cannot be answered in any way whatsoever, but it’s also possible to care about in practice (although, considering the argument, we cannot even know this), so you can only say, “Oh, well, can’t do anything about that, let’s just ignore it.” Fortunately, this answer seems to work fairly well.