Thread: Problem of Induction

I was wondering what philosophy of science people here hold, and why.

In particular, what do you think about Hume's problem of induction: that there is no basis for thinking that if certain things are repeatedly seen to act in certain ways, then they will continue to act in certain ways? I think it's clear that science and induction can never be proof, but what is our justification for even suspecting that, for instance, all swans are white, or if you drop an object it will fall to Earth? Does science even imply that we should suspect these things? If not, what use is science?

It's worth noting that 'it has tended to work in the past' is itself an inductive argument.

I think it's clear that science and induction can never be proof, but what is our justification for even suspecting that, for instance, all swans are white, or if you drop an object it will fall to Earth?

Let's say for the sake of argument that there is no justification. This means we have to reject essentially all knowledge. You can't even make observations because, say, [all evidence of a bottle existing and being dropped towards the ground] means you actually infer the bottle's existence and that it moved in a certain manner from a certain position via the input from your senses, and your experiences of what the input means with relation to events.

The justification for inductive reasoning is that you have to accept it to do anything in the world. Without it you can't observe anything, or apply knowledge.

Does science even imply that we should suspect these things?

Yes, because the entire domain is built upon empirical observation, and consequentially it already grants the premise that we can assume facts that backed up with enough observation and evidence will continue to be true.

Needless to say, the results of accepting this have been astonishingly successful.

If you are going on just observation then it is possible to be wrong. Such the example with all swans being white. However science goes much further than that. It makes observations but then also tests them, and seeks to explain why things happen. Things like gravity should continue and always be true. With our understanding of gravity we can say with certain that if you drop something it will fall to earth, unless there is a force stronger pushing or pulling it upwards.

If something works over and over in the past, then it should work in the future. If it didn't work, then something has changed. There should never be a situation that is repeatable, that always has the same result, and then later in an identical situation doesn't work. If that happens either it isn't an identical situation, or you didn't test it well enough to start with. Obviously if you do something twice, that isn't adequate to judge how something will react in all future situations.

All swans being white is a horrible example anyway. Knowledge isn't just knowing what happens, it's knowing why or how it happens. Saying that all swans are white is not a scientific statement because it lacks the most crucial part, which is the explanation.

I agree with photolysis. To say that induction is invalid is a self-detonating statement. You can't say the words "induction is invalid" without making inductive assumptions about how your lips will move (or fingers in this case). Therefore, as unintuitive it may seem to you, induction fundamentally can't be invalid.

Originally Posted by Xei

Does science even imply that we should suspect these things?

No but as near as I can determine, it does require it.

When we observe that something has happened many times in the past, we can infer that, in the past, something was causing that thing to happen as it did. So it's reasonable to think, given no other information, that the cause of it happening many times in the past isn't going to disappear, and it's going to happen again. Of course, it might happen differently next time, but but it would be too big a coincidence for the cause to change now.

I think what Hume was trying to get at here is that we do not experience one thing causing another. We only experience an event followed by a subsequent event. Since according to Hume all knowledge comes from experience, we cannot deduce that one thing "caused" another. He argues that we have innate ideas and causation is not a feature of the physical world but an association that our minds impose onto our experience of the physical world.

Originally Posted by Xei

I think it's clear that science and induction can never be proof, but what is our justification for even suspecting that, for instance, all swans are white, or if you drop an object it will fall to Earth? Does science even imply that we should suspect these things? If not, what use is science?

The scientific method is setup in a way that allows for inductive statements to be falsified in the future. For example "all swans are white" just means all swans are white until we observe a swan that isn't white. Likewise the hydrogen atom is the lightest and most common element in the universe until we observe otherwise.

I think I agree with it, there seems to always be the possibility that any pattern or structure we once thought fundamental or absolute could be dynamic and change, with the emergent product being that in time any humanly observable phenomena could seize to produce the same outcomes in a variety of ways.

The 'possibility' of a pattern changing is obvious. There's a possibility of almost anything. But that isn't what Hume's Problem of Induction was about. He said that we have no reason to even think it's more likely than not that any given pattern will continue to its next step.

Imagine that some event has happened 10508325 times in the past, which we have witnessed, and we now must decide the likelihood of it happening yet again. The Problem of Induction states that we have no reason to think it's more likely to happen again than to not.

I'd say that, if it happened in the past that many times, there must have been something causing it in the past. And it's safe to assume that it cause will very likely persist. Or more like, we have no reason to think it should stop. But the same problem arises.

I'm really thinking on the spot here. What if we just consider some completely physical situation, like the sun rising (wow I just realized that was Hume's original example after I decided to use it). With the assumption that all known physical laws are valid - newton's laws, etc - we can say that, since we don't know of any comet coming to interfere with the sun or anything like that, the sun will rise again because it's a physical object following those physical laws.

But the same problem still arises - we have no reason to think that the physical laws themselves are any more likely to remain constant a moment from now than to change.

I understood the argument as soon as I first heard it, and thought I had come up with a way around it. I don't know why I thought I had. I suppose I have to agree, for now. Although it seems like there must be some way around it.

Originally Posted by Dianeva

The 'possibility' of a pattern changing is obvious. There's a possibility of almost anything. But that isn't what Hume's Problem of Induction was about. He said that we have no reason to even think it's more likely than not that any given pattern will continue to its next step.

Imagine that some event has happened 10508325 times in the past, which we have witnessed, and we now must decide the likelihood of it happening yet again. The Problem of Induction states that we have no reason to think it's more likely to happen again than to not.

I'd say that, if it happened in the past that many times, there must have been something causing it in the past. And it's safe to assume that it cause will very likely persist. Or more like, we have no reason to think it should stop. But the same problem arises.

I'm really thinking on the spot here. What if we just consider some completely physical situation, like the sun rising (wow I just realized that was Hume's original example after I decided to use it). With the assumption that all known physical laws are valid - newton's laws, etc - we can say that, since we don't know of any comet coming to interfere with the sun or anything like that, the sun will rise again because it's a physical object following those physical laws.

But the same problem still arises - we have no reason to think that the physical laws themselves are any more likely to remain constant a moment from now than to change.

I understood the argument as soon as I first heard it, and thought I had come up with a way around it. I don't know why I thought I had. I suppose I have to agree, for now. Although it seems like there must be some way around it.

The law of large numbers is a standard answer.

What I think is the tricky question is determining what is the relevant zone of existence (spacetime, I suppose) that this applies to. It's really quite subjective.

For instance, the notion of Euclidean space. This seems to work for everything on the surface of the Earth. But how far does that give us the right to extend it? Of course, we now know that we are allowed to extend it a long way down, but not very much further up (because of the curvature of spacetime). So before this knowledge, would we have had to restrict our 'stuff will probably keep happening in the same way' argument to a pretty small realm, between, say, a millimetre and a few kilometres?

The stickier problem is time, I guess partly because we can't see much of it. How long can we say that Saturn has been around for; and hence, how long will it be before we have no logical basis for saying that it won't spontaneously disappear? What does such an argument look like?

Originally Posted by Xei

The stickier problem is time, I guess partly because we can't see much of it. How long can we say that Saturn has been around for; and hence, how long will it be before we have no logical basis for saying that it won't spontaneously disappear? What does such an argument look like?

The spirit of science has only ever rightly been in degrees of certainty, so there's nothing really new being proposed. However, personally, the two things I lean toward being most certain are the unity of nature and hence, it's conservation, so the chance of Saturn spontaneously disappearing is a little on the extreme side, although not absolutely impossible in principle I suppose, we would just seek a new logical structure to encompass it.

Perhaps a more likely example might be that we can't expect the proportions in our equations of gravity to stay constant between Saturn and the Sun. Maybe it is possible that the strength of universal or even local gravity can be dynamical over time according to some unknown and so far unnoticed mechanism and one day it may make the gravity between Saturn and the Sun too weak for Saturn to orbit it, or make it possible for us to walk comfortably on the moon like we do on earth. Can we expect this never to happen just because our current equations for gravity have produced consistent patterns for our solar system so far? It may seem like the continuation of the familiar pattern could be said to be more likely than such a change based on past experience, but how sure can we be about how likely it's hypothetical dynamical nature will emerge in this point in time, tomorrow may be coming into an era where the change becomes more frequent or accelerates.

The goal of science is the empirical knowledge of all phenomena, in other wards, deterministic certainty and absolute truth.

Everything we know could be wrong (Heard this before, I know), but it makes sense to form some fundamental laws rather than disregard everything.

I think the key phrase here is "safe to assume."

There's no reason to give a shit if a pattern is going to repeat itself or not except when making decisions. When we make decisions, we use patterns to predict outcomes. That doesn't mean a pattern enforces a particular outcome. It's just a good way to make decisions. And that is the purpose of Science, finding solid ground to base decisions upon.

But the problem is that there's no actual reason to use science. Science only works under the assumption that what has happened in the past is likely to continue into the future. Once you realize you have no basis for that assumption, everything falls apart.

Of course, we have to make that assumption. It's ingrained into our brains. We wouldn't be able to live without it. But it is strange to realize that we can't actually find a rational basis for it at all (or none that anyone's been able to come up with).

What are you talking about? Science is a fantastic tool to judge likely outcomes. Just because what is likely is not definitive doesn't mean science is all malarky.

Again, it comes down to decision making. I make decisions mostly based off what can be repeatedly observed in the most objective sense I can discover. I've argued this extensively in my thread in R/S about scientific study of spirituality using metaphors such as Gambling. If Science had no value, then Casinos would be out of business. They create games that statistically slant the outcome in their favor and let time do the rest. That's all it comes down to, making decisions that, based on your knowledge, will most likely give you the desired result.

I don't see why just because life has no guarantees that means you shouldn't attempt to discover the most likely outcome.

Originally Posted by Omnis Dei

Just because what is likely is not definitive

Originally Posted by Omnis Dei

just because life has no guarantees

The problem isn't merely that nothing is guaranteed. It's that nothing is even likely.

I'm not saying that science doesn't work. Of course it works. It has in the past. It's responsible for much of our technology and discovery.

But the problem is that we have no reason to think that science will continue to work as it has in the past.

Why do you think that a pattern that has occurred in the past is likely to continue into the future? You cannot answer this question without first assuming its truth.

Originally Posted by Dianeva

But the problem is that there's no actual reason to use science. Science only works under the assumption that what has happened in the past is likely to continue into the future. Once you realize you have no basis for that assumption, everything falls apart.

Of course, we have to make that assumption. It's ingrained into our brains. We wouldn't be able to live without it. But it is strange to realize that we can't actually find a rational basis for it at all (or none that anyone's been able to come up with).

As Xei mentioned in the OP, we cannot use induction to justify induction. We cannot use deduction to justify induction either because well inductive statements are inductive because their conclusions don't necessarily follow from the premise. Sometimes I have trouble reading Hume but from what I have deciphered, induction cannot be justified through reason or experience but is an innate feature of our minds that is imposed on our experience.

I think we can still do science with the problem of induction looming overhead. The problem of induction implies that we cannot be certain of knowledge derived through inductive reasoning but IMO the pragmatist theory of knowledge does away with the need for certainty.

Truth has no value at all to me. It's just a word. I fail to see how something is no longer likely just because of induction if past activity has so far confirmed the judgment. A judgment is only as good as the action it leads you to. There is no point in clinging to it for any other purpose. I plan for the sun to rise tomorrow because it's more practical to plan on another sunrise than to plan on no sunrise.

I'm not sure I understand what you're saying. It seems like you're saying you agree with the problem, but you don't care. Is that right?

Originally Posted by Dianeva

I'm not sure I understand what you're saying. It seems like you're saying you agree with the problem, but you don't care. Is that right?

If it does not effect decision making, what is the problem?

Xei's head asploding with NAWLEGE!

But why is the choice one or the other? I see it more like a grade, where on one polarity of the grade we have phenomena that is more sound and predictable while on the other side we have more exploratory concepts that hold less value for decision making but also provide new venues of thought.

Absolute truth is unnecessary. Furthermore, it's an obstacle in discovery.

And by the way empirical knowledge is an oxymoron

Originally Posted by Omnis Dei

But why is the choice one or the other? I see it more like a grade, where on one polarity of the grade we have phenomena that is more sound and predictable while on the other side we have more exploratory concepts that hold less value for decision making but also provide new venues of thought.

Absolute truth is unnecessary. Furthermore, it's an obstacle in discovery.

And by the way empirical knowledge is an oxymoron

?

I use the word empirical as it relates to chemistry, the most elementary form or understanding of something. But even if you associate it with its other denotations it just means experimentally or observable sound knowledge... I don't see how that is an oxymoron.

You give me so much heat for all my posts, haha.

Originally Posted by Dreams4free

?

I use the word empirical as it relates to chemistry, the most elementary form or understanding of something. But even if you associate it with its other denotations it just means experimentally or observable sound knowledge... I don't see how that is an oxymoron.

You give me so much heat for all my posts, haha.

I merely challenge the paradigm in every possible way. Please do not take it personally.

Originally Posted by Omnis Dei

I merely challenge the paradigm in every possible way. Please do not take it personally.

I take a lot of things too personally, I don't know why. I'm very much an artist so I guess I could surmise that it is a manifestation of my emotional sensitivity and humanistic assimilation of life to my self. Double edged sword.