The Humean Problem of Induction

[freejroll]

Hello, I was curious about what you all thought about the problem of induction, as introduced by David Hume. Whether or not Hume posed this problem isn't really the question, but the implications that were raised later on caused serious epistemological problems.

We use induction in most reasoning, but we've never really justified using it. And can we even? If we can't, why do we use it?

Just an example as to the importance of induction. When we claim that the sun will rise tomorrow morning, we are using induction. When we claim that gravity will stop me from floating away in two seconds to mars, we are using induction. When we claim that tomorrow our bodies won't start rejecting oxygen and we'll all die, we are using induction.

Most people seem to argue that since we used induction in the past and it has worked, then we should continue using it in the future. However, this is a circular argument. This is in fact saying that since future events resembled past events in the past, past events ought to resemble the future, in the future. The argument tries to use it's own conclusion and this won't work.

There are other attempts to answer, but none of them work. In fact, most answers to it have stopped and the problem has just been avoided.

I was wondering what you guys thing because we all use induction.

 

[Solitude_within]

I'll use one of your examples - the one about the sun rising.

Since the sun has risen every morning from the very beginning of recorded history, it's safe to believe that it will happen again tomorrow. No comment is being made on what will definitely happen. Only on what is most likely to happen. Given the odds, it is probable that the sun will indeed rise tomorrow.

This process of gathering and analyzing large amounts of information forms the base of what we call science. Regardless of whether it is a circular argument, or whether it "tries to use its own conclusions", it gives veritable results.

 

[freejroll]

I'll use one of your examples - the one about the sun rising.

Since the sun has risen every morning from the very beginning of recorded history, it's safe to believe that it will happen again tomorrow. No comment is being made on what will definitely happen. Only on what is most likely to happen. Given the odds, it is probable that the sun will indeed rise tomorrow.

What do you mean, given the odds, it is probable that the sun will indeed rise tomorrow? How does the fact that the sun has risen every day in the past suggest that it will probably rise daily in the future? The fact of the matter is, it doesn't. It is only information about something that has happened. Stating that it gives us probability, suggests that:

(1) all past and future occurances have been looked at, examined, and then analyzed. Since we don't have knowledge of unobserved past occurances, or of future occurances, both of which we have no idea the amount, we simply cannot make this suggestion. After all, it is possible that there are billions and billions of years full of days where the sun did not rise before we came into being, and before we could observe it. It is also possible that starting tomorrow for a time period longer than we have observed in the past the sun will simply quit rising in the morning. Because of this, we can't say it's "probable" or the "odds suggest" unless

(2) we are indeed assuming induction once again. The very fact that we are looking at past events claiming that they lead to a "probable" conclusion (though probable is an error) is text book inductive reasoning. Here we are stuck at using induction to prove induction. As stated in my previous thread, we cannot do this.

This process of gathering and analyzing large amounts of information forms the base of what we call science. Regardless of whether it is a circular argument, or whether it "tries to use its own conclusions", it gives veritable results.

I agree that it gives us a base for science. However, we haven't justified using this method of reasoning (I'm not sure if we even can). Claiming that it gives us results therefore we should continue using it, is stating that because induction worked in the past, we should use induction in the future, or to put it more simply, because in the past the future resembled the past, in the future the future will resemble the past. This is again an obvious circle and we must rule out this method of reasoning.

This is actually quite a big deal in philosophy. No philosopher has been able to solve it. Many have tried, but eventually most people gave up. Induction forms the basis for what we call science. If we can't justify using it, then what is our justification for science? This is a huge claim!

 

[Solitude_within]

However, we haven't justified using this method of reasoning (I'm not sure if we even can). Claiming that it gives us results therefore we should continue using it, is stating that because induction worked in the past, we should use induction in the future,....This is again an obvious circle and we must rule out this method of reasoning.

Answer me this:
"Please explain the justification of using words to communicate."

In trying to reply, you will inevitably use words. Since words are the very thing you are trying to justify, (by your reasoning) you are not allowed to use them. Does this make words unjustifiable? If we are using your line of logic, they are, and we all should stop talking immediately.

Will you stop using words, though? I doubt it, because they WORK for you. Just like science and induction does.

 

[freejroll]

Answer me this:
"Please explain the justification of using words to communicate."

In trying to reply, you will inevitably use words. Since words are the very thing you are trying to justify, (by your reasoning) you are not allowed to use them. Does this make words unjustifiable? If we are using your line of logic, they are, and we all should stop talking immediately.

Will you stop using words, though? I doubt it, because they WORK for you. Just like science and induction does.

But this isn't a good argument. With this reasoning we can lead to a number of conclusions that we wouldn't want to lead to.

Besides, trying to justify induction by saying that "it works" is again, providing a circular argument. In the past, induction worked for us, so therefore in the future induction will work. Or, in other words, in the past, the future was like the past, so in the future, the future will be like the past. We can't argue like this.

I understand that it is something that just seems like the right thing to accept, but I'm only asking why. Why does it seem like the right thing to accept? In order to justify this, as you have shown, you end up using a circular argument which proves to be invalid. So because of this, no matter how much we want to think that we should accept it, we cannot. Because it sets a precedent that we can prove anything with this type of reasoning. I could argue for counter-induction, which might seem absurd, but it really would be using the same type of reasoning.

This really isn't an easy thing to accept. No one wants to believe this. The fact of the matter is though, hundreds of years of philosophy later, no one can justify using it, though we seem to use it all the time. It even forms the basis for our science. Scary thought!

 

[Rated E]

Or, in other words, in the past, the future was like the past, so in the future, the future will be like the past. We can't argue like this.

Why not? Aren't you arguing somewhat circularly in saying that we can't argue this way because it falls under this label? Or, we can't argue this way because it is this kind of argument.

Maybe the only thing I can be actually completely certain of is that I exist, since I am experiencing, but where does this get me? Sure it's interesting to think about, but it doesn't help me much. It seems like a matter of semantics whether you claim that we know things "for certain" or whether we know them "to varying levels of probability". We can use science, while knowing that everything we know is based on probability.

In almost any scientific experiment to do with human behaviour that bases it's findings on statistical tests, the inductions made are dealing with statistical probabilities. What's wrong with this? It makes sense to me that we base things on probability if we have nothing else to base it on.

 

[ebola?]

What if we abandon the project of attaining absolutely certain knowledge?

ebola

 

[neonads]

I don't know about you guys, but my body subconsciously uses induction to learn motor skills. Seems to work for me.

 

[freejroll]

Why not? Aren't you arguing somewhat circularly in saying that we can't argue this way because it falls under this label? Or, we can't argue this way because it is this kind of argument.

I don't think so. Let me explain my argument more clearly.

When you say "in the past, the future was like the past, therefore in the future the future will be like the past" you are making an inductive argument. The argument itself assumes induction since it's process entails it. The problem is that this argument is trying to prove induction. So you are saying induction is good because induction is good, basically. Do you think this is a good way of reasoning?

Maybe the only thing I can be actually completely certain of is that I exist, since I am experiencing, but where does this get me? Sure it's interesting to think about, but it doesn't help me much. It seems like a matter of semantics whether you claim that we know things "for certain" or whether we know them "to varying levels of probability". We can use science, while knowing that everything we know is based on probability.

Well, don't assume that you can be certain that you exist. Experiences doesn't entail your existence, necessarily. Having experiences does, but it assumes there is a you to have them. And we can't just assume this. This is a different topic entirely really, so I won't delve into it.

I would say that it is definately a manner of semantics to know things for certain or with varying levels of probability. But since we haven't came up with any good arguments as to why induction is a good method of reasoning, we can't come up with probability. Probability assumes as least two things, as I've explained in my post above. It is that

(1) all past and future occurances have been looked at, examined, and then analyzed. Since we don't have knowledge of unobserved past occurances, or of future occurances, both of which we have no idea the amount, we simply cannot make this suggestion. After all, it is possible that there are billions and billions of years full of days where the sun did not rise before we came into being, and before we could observe it. It is also possible that starting tomorrow for a time period longer than we have observed in the past the sun will simply quit rising in the morning. Because of this, we can't say it's "probable" or the "odds suggest" unless

(2) we are indeed assuming induction once again. The very fact that we are looking at past events claiming that they lead to a "probable" conclusion (though probable is an error) is text book inductive reasoning. Here we are stuck at using induction to prove induction. As stated in my previous thread, we cannot do this.

We can use science, but unless we can come up with a good reason to use induction should we? We can't assume probability, as that's only ignoring the question (and actually raising more).

In almost any scientific experiment to do with human behaviour that bases it's findings on statistical tests, the inductions made are dealing with statistical probabilities. What's wrong with this? It makes sense to me that we base things on probability if we have nothing else to base it on.

Ah, two things. One, there are no statistical probabilities, as I've stated previously. Well, at least we haven't found a good reason to use them, yet (maybe we will some day). Two, you assume that since we don't have anything to base it on, we should just choose something. Why should we choose the way we do it now, with induction? Why not choose some other method. Why not counter-induction? Well, because (a) in the past this way has worked. This however leads us to our initial problem and only ignores the question. (b) in the past counter-induction has not worked. This however leads us to our inital problem as well and ignores the question.

 

[freejroll]

What if we abandon the project of attaining absolutely certain knowledge?

ebola

Induction doesn't try to discover absolutely certain knowledge. It tries to discover what will "probably" happen. This is why questioning it leads to such a conundrum.

 

[freejroll]

I don't know about you guys, but my body subconsciously uses induction to learn motor skills. Seems to work for me.

But you haven't read my posts. It has worked for you in the past. Fine. The problem is when we try to say that it will work in the future because it worked in the past. Oddly enough, this is the only important part of the argument. We can say that things have worked for us in the past all we want. This makes no argumentative claim whatsoever. It only states something that has happened. Now relating that to something else (the future or the present) is where it gets tricky. Why would you use it now? Because it worked that way in the past? Well, that's not a good method of reasoning.

We can't say that it seems to work, or that it will probably work. Well, we can say that, but we aren't making a valid argument. It's the same as saying "In the future the sun will not rise because every day we observed in the past, the sun has risen." They aren't valid arguments. We just tend to use the former in so many things we do. The truth is, we haven't really justified doing so. I'm not sure that we even can.

It's a huge claim, and it seems outlandish even, but there is no flaw to it's reasoning that we can tell. This is why it's such an important problem. It's just also a very difficult one.

 

[Rated E]

What if we abandon the project of attaining absolutely certain knowledge?

ebola

Induction doesn't try to discover absolutely certain knowledge. It tries to discover what will "probably" happen. This is why questioning it leads to such a conundrum.

Yes, so what?

Like I was saying... semantics. It doesn't bother me that most knowledge is based on probabilities. If I am given two choices, I tend to choose the one that seems most probable. What alternative is there to induction? What alternative can you offer to basing things on a complex system of varying degrees of probabilities and at least one assumption (that induction is useful).

So, in line with what ebola was saying, I ask you freejoll, when was it assumed that we must seek out absolutely certain knowledge?

I instead "assume" that we don't need to attain it.

 

[Solitude_within]

So because of this, no matter how much we want to think that we should accept it, we cannot.

What do you propose we do in order to free ourselves from the evils of unjustifiable induction?

 

[neonads]

We can't say that it seems to work, or that it will probably work. Well, we can say that, but we aren't making a valid argument. It's the same as saying "In the future the sun will not rise because every day we observed in the past, the sun has risen." They aren't valid arguments. We just tend to use the former in so many things we do. The truth is, we haven't really justified doing so. I'm not sure that we even can.

Ya, it's a paradox fo sho. Pure justification cannot arise out of a representative system. What to do now?

 

[Cthulhu]

But you haven't read my posts. It has worked for you in the past. Fine. The problem is when we try to say that it will work in the future because it worked in the past. Oddly enough, this is the only important part of the argument. We can say that things have worked for us in the past all we want. This makes no argumentative claim whatsoever. It only states something that has happened. Now relating that to something else (the future or the present) is where it gets tricky. Why would you use it now? Because it worked that way in the past? Well, that's not a good method of reasoning.

We can't say that it seems to work, or that it will probably work. Well, we can say that, but we aren't making a valid argument. It's the same as saying "In the future the sun will not rise because every day we observed in the past, the sun has risen." They aren't valid arguments. We just tend to use the former in so many things we do. The truth is, we haven't really justified doing so. I'm not sure that we even can.

It's a huge claim, and it seems outlandish even, but there is no flaw to it's reasoning that we can tell. This is why it's such an important problem. It's just also a very difficult one.

I like neonad's argument - that we "subconsciously" use induction (his example being muscle memory). Induction is intrinsic, apparently, to humanity. Even the average man, even primitive man, without a scientific background, uses induction - if obtaining food worked by bringing down buffalo with spears in the
past, it's worth trying in the future, and if eating that buffalo satisfied a biological desire (hunger) in the past, then it probably will in the future. That biological need is like a primary drive come about by evolution - evolution is a purely logical phenomenon which has preserved the drive of hunger (logically) because it sustains us biomachines/animals by urging us to act.

Not all of us, though - some of us start questioning induction, and it leads us in paradoxical loops (like questioning whether anything I just said has actually been proven or can be extrapolated beyond limited observed cases, thus questioning what I
am saying right not as I write this word and the next word werd).

This person goes insane when he take the questioning far enough, so far that he overrides the drive of hunger and just sits motionless in one spot frozen like a deer in the headlights, stuck without a logical reason to do anything. Then, in an epiphany, he realizes that he doesn't have a reason to not do anything, either, so he gets up and randomly jerks around like a crazy person. There he is, just bopping and jerking, muscles contracting randomly popping and stopping, when suddenly - he has another epiphany and realizes that all these muscle contractions do have a reason, and a natural, logical progression has taken him from living a normal life to questioning induction to sitting motionless to getting up to... now.

Unfortunately, by now he has neglected his basic needs such as food because he did not see a reason to believe in them, and he is now so weak and starved that he collapses, to exhausted to do anything, and dies. Evolution preserves those who logically are preserved (get it?), and those who aren't aren't. That means you.

 

[swilow]

What if we abandon the project of attaining absolutely certain knowledge?

ebola

Now that is a good thought. It allows room for thinkig about things that are important.

Theres no way words can ever be written to actually prove that something will happen is a place as nebulous as the future....so why waste time with wondering, when the present moment is the one thing we actually do have, clear as crystal....?

 

[freejroll]

What do you propose we do in order to free ourselves from the evils of unjustifiable induction?

I don't propose anything. I just think it's interesting.

Ya, it's a paradox fo sho. Pure justification cannot arise out of a representative system. What to do now?

I don't know about it being a paradox, but regardless, we definately can't justify using this method of reasoning (at least we haven't so far). I don't know what to do now.

I like neonad's argument - that we "subconsciously" use induction (his example being muscle memory). Induction is intrinsic, apparently, to humanity. Even the average man, even primitive man, without a scientific background, uses induction - if obtaining food worked by bringing down buffalo with spears in the
past, it's worth trying in the future, and if eating that buffalo satisfied a biological desire (hunger) in the past, then it probably will in the future. That biological need is like a primary drive come about by evolution - evolution is a purely logical phenomenon which has preserved the drive of hunger (logically) because it sustains us biomachines/animals by urging us to act.

But the same problems are here. Even though we subconsicously use induction, this doesn't mean that it is the right thing to do. Unless of course you really want the claim that anything done subconciously is a good enough reason to be the correct thing to do. And then again I would say "how do you know this" and you might say because it seemed to work that way in the past. Again, you come to the same initial problems.

Primative men did use induction. We use it in most things we do, someone might argue! However, we don't seem able to justify using it. I'm not saying that it doesn't seem the "right thing to do" because it certainly does. I can't seem to help using induction typing this response to you. However, we haven't came up with a justisification for doing so yet. This is a problem, unless we can claim that somethings don't need justification and some things do, which leads us to a slew of more problems.

Using the argument about food is the same as those posted in previous posts. Saying that it worked in the past so it should work in the future, is not a valid argument for induction.

Not all of us, though - some of us start questioning induction, and it leads us in paradoxical loops (like questioning whether anything I just said has actually been proven or can be extrapolated beyond limited observed cases, thus questioning what I
am saying right not as I write this word and the next word werd).

This person goes insane when he take the questioning far enough, so far that he overrides the drive of hunger and just sits motionless in one spot frozen like a deer in the headlights, stuck without a logical reason to do anything. Then, in an epiphany, he realizes that he doesn't have a reason to not do anything, either, so he gets up and randomly jerks around like a crazy person. There he is, just bopping and jerking, muscles contracting randomly popping and stopping, when suddenly - he has another epiphany and realizes that all these muscle contractions do have a reason, and a natural, logical progression has taken him from living a normal life to questioning induction to sitting motionless to getting up to... now.

Unfortunately, by now he has neglected his basic needs such as food because he did not see a reason to believe in them, and he is now so weak and starved that he collapses, to exhausted to do anything, and dies. Evolution preserves those who logically are preserved (get it?), and those who aren't aren't. That means you.

This is just a personal attack, lol I don't know why. I have made no claims of my beliefs or anything crazy like that. I'm only pointing out a philosophical problem that has never been solved (as far as I know).

You haven't really addressed the problems with using induction at all. I'm not sure that we can justify using it. You seem to want to suggest that we don't have to justify using things in order to use them. This is a fine suggestion, but it will definately lead to chaos, and I definately don't believe that you want to bear the consequences of this claim. Instead we should decide if we can justify induction.

 

[Jamshyd]

I'll give you a justification: because God wills it!!

In all seriousness, it is rather presumptuous of us to think we can solve all the puzzles of the universe.

 

[freejroll]

If we can't justify using it, though, why should we use it? I can't justify that we should not use induction. Does this mean that we should start not using induction? Do you see the problem? AAH!!

 

[Rated E]

But I can justify using it. I can't justify that it will provide absolutely certain knowledge, but I can justify using it simply because it has worked in the past, so it's probable that it will work in the future. Or at least more probable that it will provide more desirable results than randomly acting as if there are no rules or trends in the universe.

The results it has produced in the past are justification for its use in the future, even if it only provides outcomes that are "better than chance".