Epistemology: Is knowledge knowable? If not, how do we know this? - Woody Allen
It's a very common claim. "Everybody has faith in something. You may not believe in God, but you have faith in science, or yourself! You just make a god of those things!" Some even say things like "If you turn the ignition key do you “believe” that the engine will come to life? When you enter a darkened room and flip the light switch, do you have faith the electricity will provide enough power to illuminate the bulb and brighten the room? How is it that you can have faith in the car starting or the light coming on?".
But 'having faith in' or 'believing in' takes a lot of different forms. There are big differences between 'believing in something based on evidence', 'believing in something without evidence', and 'believing in something despite evidence'. People 'have faith' in light switches in the sense of 'trust based on experience'. On the extreme other side, there are people like Kurt Wise who say "...if all the evidence in the universe turns against creationism, I would be the first to admit it, but I would still be a creationist because that is what the Word of God seems to indicate."
I'm perfectly willing to admit I 'have faith' in things based on evidence - like light switches and electricty. On the other hand, I try very hard not to 'have faith' in things in the face of evidence. I'll cheerfully deny that anybody has to 'have faith' in that sense.
But what about the third category - 'believing in something without evidence'? Do any of my beliefs fall into that category?
Among some theologians (and philosophers!), there's another common claim - that everyone has to take at least some things on faith, without good evidence - or even reasons that can be articulated beyond "I feel or hope this is true!" The argument goes something like, "If we should only believe in things based on evidence, what is the evidence for the proposition 'we should only believe in things based on evidence'?"
Now, I actually agree that beliefs 'bottom out' somewhere. It's true that people have fundamental assumptions that ground their worldview. 1 So in that sense I could be characterized as a "Foundationalist".
But, some theologians take it a step further. It's frequently used to get around the request for evidence for the existence of supernatural beings. The theist can effectively say, "I assume on faith that there's a God. And you can't say 'boo' about it; your faith in the idea that 'you should only accept the existence of entities because of evidence' has no better justification. Nyahh, nyahh!"
However, just because you can't use evidence to decide on fundamental axioms doesn't mean all possible axioms are equal, that there's no principled way to choose among them. There is at least one distinction that can be made among such 'candidate axioms'.
Some propositions fall into the category of "unfalsifiable... but useless".
Solipsism is the notion that we can only be sure that our own minds exist. The outside world, let alone other minds, might be illusions or delusions or just mistakes.
It's true that there's no way to disprove this. No evidence of any kind could possibly be mustered to contradict the idea... since any such evidence could just be more illusion. You can't prove that the outside world exists, that the sense-data coming in has some relation to an actual external reality. You have to take it 'on faith'.
Except... hold up. Let's assume the converse for a moment. Okay, fine, have it your way. We'll grant that nothing but your own mind is real and everything else is just a dream you're having.
You've just rendered everything pointless. Solipsism and related brain-in-a-vat models are internally consistent, but practically useless. If our senses don't correlate at all with an external world... then what? Assuming that sort of thing inevitably leads to futility. The alternative idea - that the senses do relay data that in some way informs us about an outside world - doesn't have that inevitable implication, and has the bonus of being at least potentially falsifiable.
Similarly, there's the issue with our ability to reason. If you assume that our reason can't ever be relied on... then what? Assuming the converse - that at least sometimes our reason can be effective - doesn't immediately run smack into a wall of futility. We can test it (with our reason) and see if we can get consistent answers.
And then there's Ockham's Razor. Sometimes it's phrased as "entities should not be multiplied without necessity" or "the simplest explanation should be preferred". The basic idea is that you can spin an unlimited number of explanations for any finite set of observations. Maybe things fall because of warped spacetime, or maybe God pushes them down, or maybe every atom is intelligent and 'wants' to run in the biggest herd around, or...
All of these explanations account for the phenomenon of gravity equally well, account for the data we see. In terms of effectiveness, they're all just as good. However, one of them is dramatically simpler than the others. Just for pragmatic reasons, it should be preferred - it's simpler to work with, and accounts for all the data. (By definition - if an explanation doesn't account for all the data, then it's not an 'explanation'!)
We've just seen that some propositions are futile to embrace. It's not that they couldn't be true. It's just that if they were true, they'd inevitably and automatically render everything else pointless.
So it's possible to have pragmatic grounds for selecting certain 'axioms', specific 'properly basic beleifs'. I can't prove fundamental notions like 'my reason has the potential to be effective' and 'my senses relay information correlated with an external reality' and 'the simplest explanation that covers the facts should be preferred'. And yet... it's not whimsy or prejudice that drives me to accept these ideas. It's the fact that not assuming them automatically means 'game over'.
And, interestingly, if you accept such 'non-defeatist' axioms, you get a coherent and demonstrably productive worldview. You get logic and science and medicine... and, yes, even love and all that. (Based on this, I'd probably best be categorized as a "Foundherentist" who leans to Foundationalism.)
In my view, truth isn't a binary value - it's a probability range. A few things have an effective probability of 1, other things have a probability of 0 (e.g. a 179-degree Euclidean right triangle). Everything else is somewhere in between - practical, provisional knowledge, but not certain knowledge. Some things have a probability of .999999-and-a-lot-more-nines (heliocentrism, for example) but still aren't at the level of "A is A". And then there are propositions which are not self-contradictory... but are self-defeating in the practical sense.
Humans don't get to have certainty of anything. Even in mathematics, we sometimes find flaws in proofs. I admit it's highly unlikely that anyone will find a problem with "2+2=4", but technically even that isn't completely certain.
There are definitely things that are 'judgement calls', and things that we simply can't be sure about (yet). But that doesn't mean everything is equally uncertain.
As noted above, there are things that I can take 'on faith'... in the sense of trust based on reason and experience. For example, that's the faith I have in my wife's love for me. (Note: everybody actually relates to love that way. As I rhetorically ask here, "how many songs are there about the difference between saying you love someone, and actually behaving as if you do?") That's entirely different from taking something 'on faith' because I kinda feel like it, or haven't happened to run across a 'defeater' for it yet.
And there are a few limited things that I take 'on faith' not because of evidence but because assuming anything else is automatically futile. However, I try to keep those to the smallest possible set - because one of those assumptions is Ockham's Razor.
But I don't (or, at least, I'm not aware that I) take anything 'on faith' in the sense of 'despite' evidence. So if you want to convince me to take God 'on faith', you're going to need evidence.
I don't know for sure that an 'infinite regress' of beliefs is actually impossible, though. (I'm not quite ready to argue for that, though some are.) Such abstract concepts would be 'outside of time' and wouldn't need to be 'traversed'; see proof by induction. In any case, I'm willing to concede that, in practice, everybody possesses such fundamental, a priori principles.