How would any of us live if we believed that the past doesn't predict the future? — Shane

As I see it, Hume himself continued to believe. He knew that we all must do so. What he noticed was our inability to justify this trust deductively or in terms of pure reason. It is almost sanity itself to predict the future in terms of the past, so what Hume arguably revealed was that sanity wasn't logical in some pure non-empircal sense. The 'uniformity of nature' is apparently a sort of hard-wired implicit assumption.

It's impressive that Hume could become conscious of this. It's not easy to understand Hume's point. I've tried to explain it to friends and they couldn't grasp it. I think it's like trying to point out water to fish. As profound as the thought is in one sense, it's trivial in another. It's like the denial of the possibility of presuppositionless thought. Only philosophers have dreamed of such a thing in the first place. Being perfectly rational is some fully explicit way is, in my view, a quasi-religious goal. I suspect that it is somewhat inspired by monotheism.

askfgbaksjfgbaskjfgasnfkiuj

He believes that even if something happens literally over and over again, every day, it's not "more probable" that this thing will happen again tomorrow. My mind is absolutely and utterly blown. I can't comprehend it. It completely goes against everything I know and always took pride in. — Shane

If memory serves, Hume couldn't think of an argument and didn't expect anyone else to either. The temptation is to say that the future will resemble the past because it always has (which is of course circular). This is not to deny that we do think probabilistically or inductively, but only that such thinking is a kind of primitive without deductive foundation. This damages the position of the rationalist-type of philosopher who takes mathematics as a model. Induction is absolutely essential to us. If it is 'outside' of deduction, then deduction can only be or tell part of the story.

I believe that the past definitely predicts the future. — Shane

And you were right. So far, the past has definitely predicted the future. Whether or not the past will continue to predict the future, it remains to to be seen. Supposedly there is no way of knowing.

But a few years ago, I came across the Wiki article on Hume's Problem of Induction, and it basically says Hume disbelieved in this probability idea. He believes that even if something happens literally over and over again, every day, it's not "more probable" that this thing will happen again tomorrow. My mind is absolutely and utterly blown. I can't comprehend it. It completely goes against everything I know and always took pride in. — Shane

You're not alone. The belief that the same regularities that happened to a great extant in the past don't imply that it would probably continue in the future is mind blowing. It means that preparing for the future is a totally irrational endeavor.

How would any of us live if we believed that the past doesn't predict the future? We'd commit the same mistakes over and over. How would you learn anything? How would science progress? — Shane

Your absolutely right. Without the belief that the past will probably predict the future we'd be totally lost. Everything you learned about the past falls out the window, including science. They say those who don't study history are doomed to repeat it. Hume says that's rubbish.

After all, the proof seems to be that if nothing has changed, why would something new happen? — Shane

That's no proof. It's just a prejudice of yours. And it comes from your believe in the uniformity of nature. The other side of the question is equally valid: Why shouldn't something new happen?

I just want someone to explain to me how I am wrong. Maybe I don't understand the problem or his stance well enough. — Shane

I've been told that I haven't grasped the problem of induction many, so I'm the wrong person to ask.

Well, firstly nobody seems to agree on what Hume's arguments establish, but in my opinion:

Hume's argument against a logical justification for induction appears at first to be rather trivial. For he stated that the effects of a cause are not made true by the definition that constitutes what is meant by the cause.

Taking his snooker ball example, if the 'cause' is a cue-ball travelling towards the black ball, then there is nothing present in the definition of a 'travelling cue ball' that necessitates a particular movement of the black ball upon contact. But of course, if we saw the cue ball pass through the black ball we would likely revise our identification of the cue-ball, to say a ghost cue-ball. Hence our identification of types expresses our expectation of effects.

Hence Hume's argument against logical justification for induction might be rephrased by saying that one only 'believes' that one is seeing a travelling cue-ball as opposed to say a ghost-cue ball, where the certainty of this belief cannot be established purely via appeals to logic.

Hume also gave an argument against an empirical justification for induction, which boils down to the fact that in our repeated observation of causes and effects we do not in a literal sense see anything called 'necessity'. Hence we are under no purely empirical obligation to alter our beliefs in response to seeing anything.

But this shouldn't be taken to mean that Hume implies that induction is a misguided principle, or that there is no such thing as 'necessity' or that 'necessity' is meaningless, but only that what we mean by necessity cannot be grounded purely in terms of rules of logic and observations. In other words, when we speak of necessity we aren't merely appealing to stated rules of logic and observations, but also to our behavioural inclinations that includes what we think, say and do in response to what we call 'evidence'.

Hence Hume is merely forcing us to revise our understanding of what 'necessity' means so that it includes more things, namely our thoughts, our psychological preferences, our choices and our customs.

And the resulting picture should feel holistically positive in the sense of including our intuition and customs as integral components of the meaning of causation and determinism , whereby it is seen to make no sense to speak of 'free will' being an illusion or of being incompatible with determinism.

My life has always been guided by the fact that I take pride in living by way of induction. For example, if something has happened over and over in the past, I've felt it were perfectly rational to believe this thing has a high probability of happening again in the same way (say 98%), and if anyone believes contrary to that - say they believe something new will happen - I've looked down on them. I believe that the past definitely predicts the future.

But a few years ago, I came across the Wiki article on Hume's Problem of Induction, and it basically says Hume disbelieved in this probability idea. He believes that even if something happens literally over and over again, every day, it's not "more probable" that this thing will happen again tomorrow. My mind is absolutely and utterly blown. I can't comprehend it. It completely goes against everything I know and always took pride in. — Shane

Something happening over and over again isn't the cause of it happening again. There is another cause that leads us to experience some effect. If that cause is happening, then of course the same effect will happen over and over again.

When any prediction fails, we end up coming up with a new explanation that doesn't contradict or go against, rather it compliments, existing theories about other things.

To say that something might not happen the same way again is also derived from our experiences of making failed predictions. Again, we would be applying induction in order to make an argument against induction - that the effect might not occur based on past experiences of making failed predictions. We simply make predictions - some of which we expect to fail and some of which we don't.

Sure, it makes sense that it's not 100% probable (I think of Russell's chicken story), but to say it isn't any more probable at all... that seems wrong. How would any of us live if we believed that the past doesn't predict the future? We'd commit the same mistakes over and over. How would you learn anything? How would science progress? To me, if something happens over and over again, day by day, it's totally safe to assume that the probability is higher that it will happen the same way tomorrow than not. After all, the proof seems to be that if nothing has changed, why would something new happen? — Shane

Exactly. Failure, or making mistakes is how we learn. I would love to hear those arguing that induction isn't a good means of acquiring knowledge, especially new knowledge, explain how we can learn anything without induction. Do we not need to verify claims of knowledge by testing them?

Warning: idiosyncratic opinion bombastically expressed incoming. Really this is just the vigorously expressed opinion of an amateur, FWIW. :)

I wouldn't worry about it, Hume's problem of induction is way overblown and really falls out of the hoary old Empiricist stance on "impressions," or the "the Way of Ideas." If you believe that we are in immediate contact only with our "impressions," and only mediately in contact with things (or not), then and only then does Hume's problem (or any modern "sceptical" philosophical problem really) have any bite. Another way of putting the "Way of Ideas" would be that if we know only seemings-to-be, then we are blocked off from knowing actually-ises (or: we have a problem getting to is-es via seemings-to-be). But of course that's nonsense, because things can both seem to be and actually be as they seem to be; we're not permanently cut off from the way things really are by an opaque inner phenomenological screen of seemings-to-be ("impressions").

Induction works on the basis of things having definable identities, natures or essences and behaving consistently according to their nature. So if you peg a thing's nature - IOW: if either (if it's a new thing) you posit an identity for it that pegs the nature it actually has, or (if it's an already-known thing) you identify it correctly - then you can expect behaviour from it that's consistent with that nature under given circumstances. (In particular, you can expect certain consequences for experience if you poke and prod it, or "interrogate" it, scientifically. This is the same generate-and-test method as science generally, hence science is induction.)

That's really all induction is. If a thing has identity A, then (under given circumstances) it necessarily behaved in accordance with identity A yesterday, and it necessarily will behave in accordance with identity A today and tomorrow. Percentage certainties arise just because practically speaking there's always room for error in identification, IOW, you can't be fully certain that you've pegged anything's identity properly (this is really what was highlighted by Nelson Goodman's "new problem of induction" - the stuff about "grue" and "bleen" and all that). IOW, all other circumstances being equal, if the thing you're calling "sun" really is the sun (as we understand it now: the star at the center of our solar system, but earlier, more approximate identifications had the same implications for it rising and setting), then it will rise tomorrow with certainty. But it might be something else - an alien construct, a mischievous projection on everyone's retina by a mad scientist, etc., etc., etc. Or all things might not be equal (i.e. circumstances might have changed, unbeknownst to you - e.g. the law of gravity might have changed). But everything being equal, if it's truly the sun, then it cannot possibly not rise tomorrow.

Another way of saying this would be: logical necessity is a feature of things (or of semantics), not of how we talk about things (or of syntax). That was the classical understanding, which changed with "modern" philosophy, which made logical necessity a feature of talk about things (or of syntax); but that was a mistake.

Aristotle said "induction is easy" - again, all an induction is, is a deduction arising from the positing of an identity/nature/essence (which is then - the identification - confirmed/disconfirmed by experience), or from the correct identification of something whose identity has already been settled. The reason why induction has seemed to be problematic is because of a series of confusions arising during the Scholastic period, when some comments about induction in a particular logical context by Aristotle were taken out of context and misinterpreted (it's to do with something called "induction by enumeration", not sure I fully understand the confusion myself, it's quite complicated). The empiricist discussions culminating in Hume's problem only further muddied the waters.

Really, "modern" philosophy was a complete mess that we're only just now (after the later Wittgenstein and a few others) starting to recover from. The only modern philosopher worth a toss, ultimately, was Thomas Reid, and maybe (on a certain minority interpretation) Kant (but since, if that interpretation is correct, he's been largely misinterpreted, it hardly matters, in that the real Kant wasn't the famous Kant in philosophy books). That's not to say modern philosophy wasn't wonderful in the sense of being an intellectual adventure, or that its thinkers weren't great and profound and sometimes said important, true things; also a lot has been learned from it (mainly how not to go about doing philosophy). But ultimately it's been a giant detour that's led to no end of trouble in both philosophy and politics (particularly the megadeaths of Communism and Fascism) - and a lot of crappy 20th century art to boot.

I can't comprehend it. It completely goes against everything I know and always took pride in. — Shane

That's because you, and all self-sufficient humans, have evolved to believe it in an unshakeable, instinctive way. There is no escaping the belief. To escape it you'd have to be non-human.

Hume is not suggesting that anybody should disbelieve it. And he said that he continued to believe it, or at least to act as though he did. But he is pointing out that there is no logical support for the belief. Many people are shocked at the realisation that we hold beliefs that have no logical support. But one gets used to it, and continues to live with the belief just as one did before.

But he is pointing out that there is no logical support for the belief. — andrewk

You mean deductive logical support. Inferential and probabilistic reasons can be given for it. It's seems perfectly rational to me to infer that we have this habit of mind because causality exists, which permits an evolutionary account.

You mean deductive logical support. Inferential and probabilistic reasons can be given for it. It's seems perfectly rational to me to infer that we have this habit of mind because causality exists, which permits an evolutionary account. — Marchesk

So now we need three unjustifiable principles to certify our knowledge as true: Causality, the future will be the same as the past, and that the unexperienced is the same as the experienced.

What a mess!

So now we need three unjustifiable principles to certify our knowledge as true: Causality, the future will be the same as the past, and that the unexperienced is the same as the experienced. — tom

The problem has always been certifying our knowledge. If all we have is deductive logic, then we'll never know much. If the demand is certainty, Descartes got as far as the self before he had to bring God in to save the day.

Inferential and probabilistic reasons can be given for it. — Marchesk

I don't know what you mean by an 'inferential reason'. If an inference is not inductive or deductive, I don't know what it means. There's 'abductive' but since that involves measuring a conclusion against a set of pre-determined criteria, it just begs the question of the justification for the criteria.

Neither does 'probabilistic' provide a viable alternative. To make probabilistic arguments, we first need a basis on which to assign probabilities to events, and none of the ways I know of to do that would survive Hume's guillotine.

The only justification I have come across for induction is Reichenbach's, which is essentially: 'we have no better alternatives, so we might as well use induction'. Which is fine, but unnecessary, since we will use induction anyway, as we could not do otherwise, being the sort of creatures we are.

The only justification I have come across for induction is Reichenbach's, which is essentially: 'we have no better alternatives, so we might as well use induction'. Which is fine, but unnecessary, since we will use induction anyway, as we could not do otherwise, being the sort of creatures we are. — andrewk

We don't use induction, mostly because there is no such logic, or method.

Sure, we may generalize, but we do so from data that is already reported in a manner amenable to generalization, and there is no method or logic or principle that can help us to know which of our generalizations will be successful.

Unfortunately people tend to refer to this conjecturing of generalizations, of ideas or theories, as "induction" for some reason. Some even claim it is part of the scientific method!

The problem of induction is most clearly stated by Sextus Empiricus imho, and he was writing well before Hume:

"It is also easy, I think, to find fault with the inductive mode of inference. For when the Dogmatists attempt to lend credence to a universal by induction from the particulars, in doing this they will consider either all the particulars or only some of them. But if they consider only some, the induction will not
be firm, since some of the particulars omitted in the induction may refute the universal; while if they consider all, they will be working at an impossible task, since the particulars are infinite in number and unbounded. So that either way, I think, the induction turns out to be shaky."
-Sextus Empiricus, "Outlines of Pyrrhonism" book 2, ch. 15

As for how science progresses in light of the problem of induction, well, that's investigated in "Conjectures and Refutations" by Popper.

Wow, I am really grateful for all the helpful responses. I thank each of you! I just now got around to checking this post. I shall work through reading the responses soon enough!