Questions about induction and God

jjn2 · 16-04-2012 03:31PM

King Mob wrote: »

Again you have misread what I wrote.
I stated clearly that:
"You might have strong reasons for making that prediction, but it would not be held as accurate or true until it can be verified to a high degree of certainty."
And yes, forming a predictive model that is testable is the fundamental way that science works.

As for what you use to choose your predictions you do use inductive logic and basing it on other facts that have already been confirmed.

But you don't hold your conclusion to be in any way true until it is confirmed to be.
The inductive logic only gets you to the point of making the model were the problem of induction might be an issue if and only if that's as far as you can go and you believed that the model was 100% true. But since we can make testable predictions and that science never accepts something to be 100% true, it's not a problem.

You think I'm misreading you, and I think you're misunderstanding me, but I'll try again!

You are using inductive reasoning if you try to use experiments or verification to assign a probability or degree of certainty to your theories and models. If you do this, you run into the problem that your use of induction is unjustified. Alternatively, you might say that you don't use inductive reasoning to analyze the results of your experiment. But then suppose that your experimental results don't match your predictions. Since you can't use inductive reasoning to estimate the probability that your various assumptions are true, you are left with the problem that you only know that at least one of your assumptions was incorrect, but you will have no way of establishing which one it was.

The conclusion is that it's impossible to avoid the problem of induction while doing science. You can choose between using inductive reasoning, and running into the problem of induction, or avoiding inductive reasoning, and never actually learning any generalities. If you've had a mandatory course in the philosophy of science, as many university science students are forced to endure, this would have been made clear to you.

sbsquarepants · 16-04-2012 03:50PM

jjn2 wrote: »

The claim that inductive reasoning is absolutely reliable is trivially false, but the problem I'm talking about isn't that we can't be absolutely certain of it. The problem is that there is absolutely justification at all for why you should have even the slightest bit of faith in your predictions.
.

You can't honestly believe that there is no justifiable reason to believe that the sun will rise tomorrow and not just skip a day or decide for itself not to set tonight and provide us all with eternal daylight from now on. Can you?

jjn2 wrote: »

The conclusion is that it's impossible to avoid the problem of induction while doing science. You can choose between using inductive reasoning, and running into the problem of induction, or avoiding inductive reasoning, and never actually learning any generalities. If you've had a mandatory course in the philosophy of science, as many university science students are forced to endure, this would have been made clear to you.

You keep using that word. It's not a problem, if it was it would be impossible to ever know anything. All predictions require "faith" for want of a better word, but it's not faith as in religious faith. It's not blind faith, it's faith based on previous observations. I have faith that the sun will rise tomorrow, i have perfectly valid and verifiable reasons for holding that opinion, reasons that i feel perfectly justified in calling evidence, induction is not an issue - i'll even put a bet on with you if you like:)!
You have faith in the existance of a supreme being - based on??
My faith and your faith are very different beasts.

jjn2 · 16-04-2012 03:56PM

Knasher wrote: »

Why would the selection of induction+Christ invalidate Occam's Razor? If Occam's Razor is invalidated, then under the induction+Christ system we can also include fairies and wizards and any other supernatural or miraculous explanation we care to postulate ad-inifitum and again we have sacrificed the scientific method because the scientific method simply cannot work in a supernatural universe. If Occam's Razor isn't invalidated, then under either system we are left with the problem of justifying the inclusion of the Christ assumption.

Whereas I think many of the things I have claimed in this thread are basic philosophy of science, and should be reasonably obvious, this is my personal opinion, and I could easily imagine that I am wrong (I actually started this thread hoping to discuss this claim).

If we assume that inductive reasoning is valid, we notice a pattern in which unnecessarily complicated explanations of phenomena are less likely to be good explanations than explanations which only take into account the relevant facts. It is a consequence of our assumption that inductive reasoning is valid. It can't apply to inductive reasoning, since we have assumed this to be true. If we had originally added the assumption that Christ was resurrected, we wouldn't invalidate Occam's razor, we would simply arrive at an Occam's razor which neither applies to inductive reasoning nor Christ's resurrection. Moderate Christians still manage to think that fundamentalists are a bit crazy.

Unfortunately, there appears to be no way of justifying our choices of assumptions about the world (for most people these aren't choices, we don't choose our beliefs). Therefore, while we might distrust people who have wildly different views of the world to us, but we can't rationally justify that our view is correct.

sbsquarepants wrote: »

You can't honestly believe that there is no justifiable reason to believe that the sun will rise tomorrow and not just skip a day or decide for itself not to set tonight and provide us all with eternal daylight from now on. Can you?

You keep using that word. It's not a problem, if it was it would be impossible to ever know anything. All predictions require "faith" for want of a better word, but it's not faith as in religious faith. It's not blind faith, it's faith based on previous observations. I have faith that the sun will rise tomorrow, i have perfectly valid and verifiable reasons for holding that opinion, reasons that i feel perfectly justified in calling evidence, induction is not an issue - i'll even put a bet on with you if you like!
You have faith in the existance of a supreme being - based on??
My faith and your faith are very different beasts.

I've explained my point of view in some of my earlier posts, but I'd like to repeat that this isn't my problem, it's generally accepted by everyone who has considered induction, that there is a problem of induction, and that it has the basic properties I've outlined. Even wikipedia has a problem with induction!

http://en.wikipedia.org/wiki/Problem_of_induction

Penn · 16-04-2012 04:00PM

jjn2 wrote: »

You think I'm misreading you, and I think you're misunderstanding me, but I'll try again!

You are using inductive reasoning if you try to use experiments or verification to assign a probability or degree of certainty to your theories and models. If you do this, you run into the problem that your use of induction is unjustified. Alternatively, you might say that you don't use inductive reasoning to analyze the results of your experiment. But then suppose that your experimental results don't match your predictions. Since you can't use inductive reasoning to estimate the probability that your various assumptions are true, you are left with the problem that you only know that at least one of your assumptions was incorrect, but you will have no way of establishing which one it was.

The conclusion is that it's impossible to avoid the problem of induction while doing science. You can choose between using inductive reasoning, and running into the problem of induction, or avoiding inductive reasoning, and never actually learning any generalities. If you've had a mandatory course in the philosophy of science, as many university science students are forced to endure, this would have been made clear to you.

Yes, but when one claim uses inductive reasoning, and another claim uses inductive reasoning, the differences between them are important, and this is where scientific skepticism and common sense come in. If we measure two things with inductive reasoning by the same barometer, one can still be more right than the other.

You're right, it's impossible to avoid the problem of induction. But even Hume himself realised that induction itself is not the problem when two claims are both made from inductive reasoning:

Hume's argument is often credited with raising the problem of induction in its modern form. For Hume himself the conclusion of the argument is not so much a problem as a principle of his account of induction: Inductive inference is not and could not be reasoning, either deductive or probabilistic, from premises to conclusion, so we must look elsewhere to understand it. Hume's positive account, discussed in sections 5.3 and 8.3 below, does much to alleviate the epistemological problem—how to distinguish good inductions from bad ones—without treating the metaphysical problem. His account is based on the principle that inductive inference is the work of association which forms a “habit of the mind” to anticipate the consequence, or effect, upon witnessing the premise, or cause.

http://plato.stanford.edu/entries/induction-problem/#IndJus

16-04-2012 04:09PM

jjn2 wrote: »

You think I'm misreading you, and I think you're misunderstanding me, but I'll try again!

No it's definitely you misreading me. None of the above points have anything to do with probability for instance..

jjn2 wrote: »

You are using inductive reasoning if you try to use experiments or verification to assign a probability or degree of certainty to your theories and models. If you do this, you run into the problem that your use of induction is unjustified. Alternatively, you might say that you don't use inductive reasoning to analyze the results of your experiment.

But you only use inductive reasoning to establish the model. The results of the test of the model are observed empirically.

For example you observe that gravity makes objects fall to the ground.
You deduce using inductive logic that this might be due to a force that attracts things to the Earth and you further deduce that this force would affect all objects equally.
You then propose the testable claim that if the above theory is true, then all objects would fall at the same rate regardless of their weight.

The experiment and the result of that experiment should be obvious.

Only in forming the theory and the testable model does inductive logic come into play.

jjn2 wrote: »

But then suppose that your experimental results don't match your predictions. Since you can't use inductive reasoning to estimate the probability that your various assumptions are true, you are left with the problem that you only know that at least one of your assumptions was incorrect, but you will have no way of establishing which one it was.

This is only if there is multiple assumptions you need to make to form your theory and you can't make up tests to exclude the various factors. And then if observation doesn't match your predictions it probably means your theory is not very sound.

Your not supposed to make assumptions that aren't supported and you don't use inductive logic to judge how likely they are.

jjn2 wrote: »

The conclusion is that it's impossible to avoid the problem of induction while doing science. You can choose between using inductive reasoning, and running into the problem of induction, or avoiding inductive reasoning, and never actually learning any generalities.

And again: you only use inductive reasoning to make your model, and this is not impeded by the problem of induction as you do not accept your model until it is verified by means other than inductive logic.

jjn2 wrote: »

If you've had a mandatory course in the philosophy of science, as many university science students are forced to endure, this would have been made clear to you.

I have a bachelor's in Physics and Philosophy, and yet your point is not a given I'm afraid.

jjn2 · 16-04-2012 04:15PM

Penn wrote: »

Yes, but when one claim uses inductive reasoning, and another claim uses inductive reasoning, the differences between them are important, and this is where scientific skepticism and common sense come in. If we measure two things with inductive reasoning by the same barometer, one can still be more right than the other.

It is of course possible to compare specific claims, if we agree on shared principles. If this wasn't the case, there really wouldn't be such a thing as science. The problem arises in justifying, or at least vindicating, the decision to use inductive reasoning at all.

You're right, it's impossible to avoid the problem of induction. But even Hume himself realised that induction itself is not the problem when two claims are both made from inductive reasoning:

http://plato.stanford.edu/entries/induction-problem/#IndJus

I think the most widely accepted "solution" is Reichenbach's pragmatic vindication of induction. The idea is that if any method of acquiring knowledge works, inductive reasoning can be used to discover this, and therefore will obtain the success of that other method. We would therefore be vindicated in choosing to use inductive reasoning. This certainly applies to something like astrology, but to me, it doesn't help in choosing induction, and induction with untestable modifications.

Penn · 16-04-2012 04:18PM

jjn2 wrote: »

It is of course possible to compare specific claims, if we agree on shared principles. If this wasn't the case, there really wouldn't be such a thing as science. The problem arises in justifying, or at least vindicating, the decision to use inductive reasoning at all.

I think the most widely accepted "solution" is Reichenbach's pragmatic vindication of induction. The idea is that if any method of acquiring knowledge works, inductive reasoning can be used to discover this, and therefore will obtain the success of that other method. We would therefore be vindicated in choosing to use inductive reasoning. This certainly applies to something like astrology, but to me, it doesn't help in choosing induction, and induction with untestable modifications.

Then what would you suggest?

jjn2 · 16-04-2012 04:33PM

King Mob wrote: »

And again: you only use inductive reasoning to make your model, and this is not impeded by the problem of induction as you do not accept your model until it is verified by means other than inductive logic.

This to me is the key point. You can't verify a general principle by any means except inductive reasoning. Without using inductive reasoning, how would you verify that the acceleration due to gravity of the earth at the surface is approximately 9.8 m/s^2? How would you verify any model? I think it's impossible. You can obtain a verification, but unless you apply inductive reasoning, you have no reason to believe that this verification is in any way representative of your model.

Penn wrote: »

Then what would you suggest?

I would suggest that there is no solution to the problem of induction, and that we simply ignore it and choose to use induction as a matter of faith.

16-04-2012 04:42PM

jjn2 wrote: »

This to me is the key point. You can't verify a general principle by any means except inductive reasoning. Without using inductive reasoning, how would you verify that the acceleration due to gravity of the earth at the surface is approximately 9.8 m/s^2? How would you verify any model? I think it's impossible. You can obtain a verification, but unless you apply inductive reasoning, you have no reason to believe that this verification is in any way representative of your model.

By directly measuring the acceleration due to gravity.
Inductive reasoning does not come into play when you are actually verifying the facts, only when you are forming your model with is conditional and not an absolute.

Knasher · 16-04-2012 04:48PM

jjn2 wrote: »

Whereas I think many of the things I have claimed in this thread are basic philosophy of science, and should be reasonably obvious, this is my personal opinion, and I could easily imagine that I am wrong (I actually started this thread hoping to discuss this claim).

There are certain assumptions underlying philosophy of science, I'm not denying that or am I downplaying its significance. Not only is it possible that those assumptions are invalid, but it is also impossible to make any measurement of probability that they are invalid because and test will ultimately rest on those assumptions and therefore the proof will be circular, this is why they are called assumptions. The justification for those assumptions rest entirely on the fact that, right now at least, we can't proceed without them.

jjn2 wrote: »

If we assume that inductive reasoning is valid, we notice a pattern in which unnecessarily complicated explanations of phenomena are less likely to be good explanations than explanations which only take into account the relevant facts. It is a consequence of our assumption that inductive reasoning is valid. It can't apply to inductive reasoning, since we have assumed this to be true. If we had originally added the assumption that Christ was resurrected, we wouldn't invalidate Occam's razor, we would simply arrive at an Occam's razor which neither applies to inductive reasoning nor Christ's resurrection. Moderate Christians still manage to think that fundamentalists are a bit crazy.

Occam's Razor simply states that the fewer assumptions you have to make the better (at least in the butchered form I'm stating it in) and it applies to all assumptions, be they the assumption of inductive reasoning or the assumption of Christs resurrection. If we come up with some way of discarding the induction assumption (though I can't currently imagine how) and disproving or proving it, then that is what we will do. But right now we can't operate without it so we have no other rational choice but to proceed with it.

What you seem to be getting at is, because science is built on a small set of justifiable (through necessity) assumptions, that it is the same as faith and that faith is therefor equally justified because it is also built on assumptions. But what is the necessity of the Christ assumption? I suspect it will be based on another assumption, and that on another assumption and so on until you get to a small set of assumptions with absolutely no justification, and should therefore be discarded.

jjn2 wrote: »

Unfortunately, there appears to be no way of justifying our choices of assumptions about the world (for most people these aren't choices, we don't choose our beliefs). Therefore, while we might distrust people who have wildly different views of the world to us, but we can't rationally justify that our view is correct.

My set of assumptions are simply a subset of yours, and given the problematic nature of assumptions I'd hope you'd agree that the fewer you make the better. So I don't think it is unreasonable that we would expect you to justify the additional assumptions you are making over ours.

jjn2 · 16-04-2012 04:51PM

King Mob wrote: »

By directly measuring the acceleration due to gravity.
Inductive reasoning does not come into play when you are actually verifying the facts, only when you are forming your model with is conditional and not an absolute.

Sorry, when first replying to your post, I didn't read your description of the experiment. It is clear that you are still forced to use inductive reasoning. If you carry out an experiment, and record the results, you have gained no general knowledge, or verification of your model. The verification comes, either from carrying out the experiment for every object in the universe, and from your experimental results deducing that your model was correct. Or you carry out the experiment for some suitable sample or subcollection, and induce that the general behaviour is represented by the behaviour observed during your experiment. Since the former is impossible, you haven't avoided the problem of induction!

jjn2 · 16-04-2012 05:09PM

Knasher wrote: »

Occam's Razor simply states that the fewer assumptions you have to make the better (at least in the butchered form I'm stating it in) and it applies to all assumptions, be they the assumption of inductive reasoning or the assumption of Christs resurrection. If we come up with some way of discarding the induction assumption (though I can't currently imagine how) and disproving or proving it, then that is what we will do. But right now we can't operate without it so we have no other rational choice but to proceed with it.

The problem is that Occam's Razor isn't inherently a good rule. We have learned through experience that Occam's razor is actually a good tool. It isn't necessarily true, like the claim that all bachelors are unmarried. We can therefore first apply Occam's razor, after we have chosen our initial assumptions about the universe. If our initial choice of assumptions is that inductive reasoning is justified, except where it contradicts the resurrection of Christ. Then the version of Occam's razor that we would learn through experience (since we have assumed that the resurrection of Christ is real), is that simpler explanations are more likely to be true, except where the simpler explanation contradicts the resurrection of Christ.

What you seem to be getting at is, because science is built on a small set of justifiable (through necessity) assumptions, that it is the same as faith and that faith is therefor equally justified because it is also built on assumptions. But what is the necessity of the Christ assumption? I suspect it will be based on another assumption, and that on another assumption and so on until you get to a small set of assumptions with absolutely no justification, and should therefore be discarded.

My set of assumptions are simply a subset of yours, and given the problematic nature of assumptions I'd hope you'd agree that the fewer you make the better. So I don't think it is unreasonable that we would expect you to justify the additional assumptions you are making over ours.

My point isn't that there is any justification for the assumption of Christ's resurrection, but that according to the above, the one reason (Occam's razor) that would tell us to get rid of excess assumptions, can't apply until after we have already chosen our most basic assumptions about the world. Indeed, if we choose sufficiently outlandish initial assumptions (like starting with counter induction), we would never actually manage to conclude that Occam's razor is useful.

16-04-2012 05:19PM

jjn2 wrote: »

Sorry, when first replying to your post, I didn't read your description of the experiment. It is clear that you are still forced to use inductive reasoning. If you carry out an experiment, and record the results, you have gained no general knowledge, or verification of your model. The verification comes, either from carrying out the experiment for every object in the universe, and from your experimental results deducing that your model was correct. Or you carry out the experiment for some suitable sample or subcollection, and induce that the general behaviour is represented by the behaviour observed during your experiment. Since the former is impossible, you haven't avoided the problem of induction!

Like say gravity on the Earth.
The experiment I described only deals with the gravity on Earth. It can be determined by experiment to apply to the whole Earth.
Induction plays no part.

jjn2 · 16-04-2012 05:30PM

King Mob wrote: »

Like say gravity on the Earth.
The experiment I described only deals with the gravity on Earth. It can be determined by experiment to apply to the whole Earth.
Induction plays no part.

Any general conclusion requires inductive reasoning. The actual experiment only verifies that certain objects fell at certain speeds, from certain places, at a certain time. If you want to conclude that the results of your experiment apply to certain types of objects in general, or to any place on the whole Earth, or in general, at any time, and not just when you carried out the experiment, induction is the only available tool.

Knasher · 16-04-2012 05:32PM

jjn2 wrote: »

The problem is that Occam's Razor isn't inherently a good rule. We have learned through experience that Occam's razor is actually a good tool. It isn't necessarily true, like the claim that all bachelors are unmarried. We can therefore first apply Occam's razor, after we have chosen our initial assumptions about the universe. If our initial choice of assumptions is that inductive reasoning is justified, except where it contradicts the resurrection of Christ. Then the version of Occam's razor that we would learn through experience (since we have assumed that the resurrection of Christ is real), is that simpler explanations are more likely to be true, except where the simpler explanation contradicts the resurrection of Christ.

My point isn't that there is any justification for the assumption of Christ's resurrection, but that according to the above, the one reason (Occam's razor) that would tell us to get rid of excess assumptions, can't apply until after we have already chosen our most basic assumptions about the world. Indeed, if we choose sufficiently outlandish initial assumptions (like starting with counter induction), we would never actually manage to conclude that Occam's razor is useful.

So if we start from your universe where the initial assumption is that inductive reasoning is justified, except where it contradicts the resurrection of Christ, derive Occam's razor that simpler explanations are more likely to be true, except where the simpler explanation contradicts the resurrection of Christ and then reapply it to the initial assumptions what do we get? The removal of Christ from the initial assumption doesn't contradict the resurrection of Christ, it simply doesn't state it as an assumption without proof and its removal would mean the simpler version is more likely to be true. We can then re-derive Occam's razor without the extra assumption and we are back to where we started.

Penn · 16-04-2012 05:48PM

jjn2 wrote: »

I would suggest that there is no solution to the problem of induction, and that we simply ignore it and choose to use induction as a matter of faith.

And that brings us back to square one with faith in science being a belief based on accuracy of evidence, trust in competent scientists and peer reviews, and faith in religion being based on belief in a book and 'what people think' as opposed to what can be proven to the best of our abilities and resources

jjn2 · 16-04-2012 06:21PM

Knasher wrote: »

So if we start from your universe where the initial assumption is that inductive reasoning is justified, except where it contradicts the resurrection of Christ, derive Occam's razor that simpler explanations are more likely to be true, except where the simpler explanation contradicts the resurrection of Christ and then reapply it to the initial assumptions what do we get? The removal of Christ from the initial assumption doesn't contradict the resurrection of Christ, it simply doesn't state it as an assumption without proof and its removal would mean the simpler version is more likely to be true. We can then re-derive Occam's razor without the extra assumption and we are back to where we started.

Everything that isn't bolded is a fair enough summary of what I've written. I would disagree with the bolded bits. If we were to simply reapply Occam's razor to our initial assumptions, we wouldn't just remove the Christ assumption, but we would also be forced to remove the validity of induction assumption.

Penn wrote: »

And that brings us back to square one with faith in science being a belief based on accuracy of evidence, trust in competent scientists and peer reviews, and faith in religion being based on nothing but a book and 'what people think' as opposed to what can be proven to the best of our abilities and resources

No, it would leave us with equally unjustified alternatives. We are therefore free to choose any or none of them, without it being irrational.

Knasher · 16-04-2012 06:47PM

jjn2 wrote: »

Everything that isn't bolded is a fair enough summary of what I've written. I would disagree with the bolded bits. If we were to simply reapply Occam's razor to our initial assumptions, we wouldn't just remove the Christ assumption, but we would also be forced to remove the validity of induction assumption.

Really, how do you suggest we proceed without the assumption of induction? The whole point was that we could remove the christ assumption because it was unnecessary under either set of rules, but the same cannot be said for induction, because we can't proceed without it, if it wasn't necessary then we would gladly discard it.

Look I really don't see where you are going with all this. I'm sure you could construct a nice little circular set of assumptions which couldn't be deconstructed so easily, and if that's what you want to do I'm not going to stand in your way. Anyone could shoehorn in any set of assumptions they want and nobody else could argue under their rules that their assumptions are invalid. Everyone can have their own little view of the universe.

Science however works by building consensus, and the only way it can do that is if everyone (believer and unbeliever alike) agrees on the same set of minimal but necessary assumptions. This set of assumptions doesn't preclude any supernatural belief that anyone likes to posit or any extra assumptions that you want to bring, it just doesn't validate them without evidence. Anyone who doesn't wan to play by those rules is free not to, they just can't expect to be taken seriously by the scientific community. I'd like to say the same about society in general, but lets be honest homeopathy is damn good evidence that you can have crazy beliefs and still be taken seriously by society.

jjn2 wrote: »

No, it would leave us with equally unjustified alternatives. We are therefore free to choose any or none of them, without it being irrational.

They are hardly equally unjustified. Science maintains an absolutely minimal set of assumptions, the philosophy of science expounds on why they are both necessary and unavoidable and the fruits of science verify that it is at least a somewhat rational way of exploring the universe. Religion (explicitly christainity here as its the one I suspect we are both most familiar with) on the other hand, maintains a huge amount of interconnected assumptions and demonizes the search of evidence on which to base rational thought, "blessed are those that believe without seeing --John 20:29" and is constantly having its conclusions externally torn down and invalidated by the expansion of science. At it's very core and in the words of the principle character it extols irrationality over reason, to claim equivalence is absolutely laughable.

jjn2 · 16-04-2012 07:46PM

Knasher wrote: »

Really, how do you suggest we proceed without the assumption of induction? The whole point was that we could remove the christ assumption because it was unnecessary under either set of rules, but the same cannot be said for induction, because we can't proceed without it, if it wasn't necessary then we would gladly discard it.

We seem to have meant slightly different things by the word unnecessary. Occam's razor encourages us to get rid of unnecessary assumptions, that is, assumptions that we aren't forced by the evidence to make. How would you propose that our assumption of induction wouldn't disappear if we apply Occam's razor to it? We certainly aren't forced by any evidence to assume induction.

Look I really don't see where you are going with all this. I'm sure you could construct a nice little circular set of assumptions which couldn't be deconstructed so easily, and if that's what you want to do I'm not going to stand in your way. Anyone could shoehorn in any set of assumptions they want and nobody else could argue under their rules that their assumptions are invalid. Everyone can have their own little view of the universe.

My point was in fact that everyone can have their own little view of the universe, and while we can have a low opinion of somebody else's crazy world view, we run into serious problems when trying to rationally justify the claim that ours is more likely to be true.

Knasher · 16-04-2012 08:44PM

jjn2 wrote: »

We seem to have meant slightly different things by the word unnecessary. Occam's razor encourages us to get rid of unnecessary assumptions, that is, assumptions that we aren't forced by the evidence to make. How would you propose that our assumption of induction wouldn't disappear if we apply Occam's razor to it? We certainly aren't forced by any evidence to assume induction.

Great, so you've excluded pragmatic considerations from the definition of the word necessary. I've already conceded several times that the only justification for induction is pragmatism and spent a good deal of time arguing that it is a justifiable and necessary assumption, and now you suddenly assert that the only acceptable justification for necessity is empirical.

Gee, thanks for that.

16-04-2012 09:33PM

jjn2 wrote: »

Any general conclusion requires inductive reasoning. The actual experiment only verifies that certain objects fell at certain speeds, from certain places, at a certain time. If you want to conclude that the results of your experiment apply to certain types of objects in general, or to any place on the whole Earth, or in general, at any time, and not just when you carried out the experiment, induction is the only available tool.

Unless of course we run the test several times at different random times and events and they always turn up the same answer.
So far every time the test is run, or that something based on that fact is used, it is has been verified.
This is all fact that doesn't rely on induction. And this is all that science claims about any of itself when we strip away the convenient short hand language we use.
It could be that someday we might find some instance that somehow contradicts the previous findings. This however is not a problem as the position is only ever held provisionally. And when we do find something that contradicts the accepted theory, they change to accept the new observation.

Again, induction is only a problem is you claim something is an absolute and/or claim it without verification.

jjn2 · 16-04-2012 10:12PM

Knasher wrote: »

Great, so you've excluded pragmatic considerations from the definition of the word necessary. I've already conceded several times that the only justification for induction is pragmatism and spent a good deal of time arguing that it is a justifiable and necessary assumption, and now you suddenly assert that the only acceptable justification for necessity is empirical.

Gee, thanks for that.

I have claimed all along that the pragmatic vindication of induction is unsatisfactory, since we can reformulate it as a pragmatic vindication of induction with additional assumptions. My subsequent claim was that Occam's razor can't differentiate between the two cases, since Occam's razor either has to be understood to not apply to the assumptions we used prior to obtaining Occam's razor, or it would get rid of all our assumptions, instead of just the additional ones.

King Mob wrote: »

Unless of course we run the test several times at different random times and events and they always turn up the same answer.
So far every time the test is run, or that something based on that fact is used, it is has been verified.
This is all fact that doesn't rely on induction. And this is all that science claims about any of itself when we strip away the convenient short hand language we use.
It could be that someday we might find some instance that somehow contradicts the previous findings. This however is not a problem as the position is only ever held provisionally. And when we do find something that contradicts the accepted theory, they change to accept the new observation.

Again, induction is only a problem is you claim something is an absolute and/or claim it without verification.

Your random test are only representative under the assumption of induction. We carry out experiments, and they strengthen our belief that a certain theory is correct. This is inductive reasoning. If we simply had experiments which verify nothing general, we could have no confidence in our scientific models and theories. Confidence in well validated theories is at the heart of the sciences, and you couldn't have any functioning science without it. Many theories in science are considered virtually certainly true, not just provisionally true because we haven't yet managed to falsify them.

robindch · 16-04-2012 10:16PM

jjn2 wrote: »

There isn't an immediate connection to the question of whether it's reasonable to believe in God.

I rather suspect there is, but more about that later.

jjn2 wrote: »

The question is, if there is no justification for the validity of induction, why is belief in induction reasonable?

First of all, I don't "believe in" what you refer to as "induction". On the contrary, as I said towards the beginning of this thread, I assume that the conjecture concerning the universality of physical law is true, and move on from there. The alternative, that physical law is, in some sense, "random" suggests an unreasonable universe which is not open to rational thought and explanation. Experience suggests the opposite, so as far as I'm concerned it's reasonable to assume the conjecture. There's also Occam's Razor and I'm sure somebody has mentioned that above.

jjn2 wrote: »

What makes belief in induction superior to belief in counter induction (which would lead us to conclude that because the sun has risen every day so far, we expect that it will not rise tomorrow), or belief in a modified principle of induction, [...]

For the reason I've just suggested -- that a non-inductive (in your meaning of the term) universe is a structurally unreasonable place, and both experience and history suggest that's not an accurate view.

I haven't read through the intervening posts, but so far as I can make out, you appear to be trying to argue that the "induction" argument is based on something called "faith". In the apparent expectation that you can then say that religion is also based on "faith" and therefore, that religion and science are structurally equivalent.

If this is your position, or your line of argument, then it's fairly obviously a sophistical argument which relies on two different, and almost entirely exclusive, meanings of the word "faith". That, and the fact that your argument can also be used to show that belief in the reality of the easter bunny rabbit is structurally identical to, say, belief in the reality of the catholic deity (which may not be something you wish to show).

jjn2 · 16-04-2012 10:33PM

robindch wrote: »

Whether you find what I have said even remotely convincing or not is up to you, but I have addressed all your points.

Zombrex · 16-04-2012 10:36PM

jjn2 wrote: »

These questions are for anyone who thinks that a belief in the existence of God is irrational, stupid or can't be reconciled with science.

Firstly, do you think there is a conclusive answer to the problem of induction? If so, what do you think it is?

Secondly, if you do think there is a real problem with induction, what makes believing in God different from believing in induction? What separates belief in induction, from belief in counter-induction, or from belief in a slightly modified version of induction, where, for example, induction holds whenever it isn't in contradiction with Christian theology?

I'm not claiming that this is in any way an argument for the existence of God, and I'm not trying to catch anyone out. I've just been curious about these questions for a while, and since I didn't study philosophy, I don't really know where to look for their answers.

The problem of induction only arises if one is attempting to make statements expressing absolute knowledge, such as "The Sun will rise tomorrow". The question then arises how can anyone know that, in the sense of knowing for certain that is the case.

Most people are not actually making a statement of absolute knowledge when they state something like that. The Sun will rise tomorrow is actually short hand for Based on my perception of reality and the conceptual model of the phenomena classified as the Sun, I am going to assume with high confidence that, based on the rules of the conceptual model, the Sun will continue to act as it has acted previously and as my model predicts it will.

But it is easier to say The Sun will rise tomorrow.

People can assume as well that God exists, but this leads back not to the assumption but the justification for the assumption. I can justify my assumption the Sun will rise, based on previous observations and the model that has been developed predicting it will. Both of these can be assessed and one can come to their own conclusions as to their validity.

Can the same be said of the belief in God?

jjn2 · 16-04-2012 10:46PM

Zombrex wrote: »

The problem of induction only arises if one is attempting to make statements expressing absolute knowledge, such as "The Sun will rise tomorrow". The question then arises how can anyone know that, in the sense of knowing for certain that is the case.

Most people are not actually making a statement of absolute knowledge when they state something like that. The Sun will rise tomorrow is actually short hand for Based on my perception of reality and the conceptual model of the phenomena classified as the Sun, I am going to assume with high confidence that, based on the rules of the conceptual model, the Sun will continue to act as it has acted previously and as my model predicts it will.

But it is easier to say The Sun will rise tomorrow.

People can assume as well that God exists, but this leads back not to the assumption but the justification for the assumption. I can justify my assumption the Sun will rise, based on previous observations and the model that has been developed predicting it will. Both of these can be assessed and one can come to their own conclusions as to their validity.

Can the same be said of the belief in God?

I'll probably get burnout if I don't start wasting less time on these forums, so won't repeat everything that's been said, but if you read the first page of the below link, you will see that what you have written above is wrong.

http://ls.poly.edu/~jbain/philsci/philscilectures/08.Induction.pdf

16-04-2012 11:11PM

jjn2 wrote: »

Your random test are only representative under the assumption of induction.
We carry out experiments, and they strengthen our belief that a certain theory is correct. This is inductive reasoning.
If we simply had experiments which verify nothing general, we could have no confidence in our scientific models and theories. Confidence in well validated theories is at the heart of the sciences, and you couldn't have any functioning science without it.

They are not made under the assumption of induction. Induction plays no part as the results are empirical. It would be inductive reasoning if we used these results to determine that the theory is an absolute.

jjn2 wrote: »

Many theories in science are considered virtually certainly true, not just provisionally true because we haven't yet managed to falsify them.

No that's simply not true.
All theories and models in science are provisional and based on empirical observation. All of them are subject to chance in the event of the discovery of new evidence regardless of how supported they are

Penn · 17-04-2012 07:50AM

jjn2 wrote: »

No, it would leave us with equally unjustified alternatives. We are therefore free to choose any or none of them, without it being irrational.

No, because again, once both methods of evaluating a claim are unjustified, the rationality behind each claim can be determined via scientific skepticism and common sense. Therefore, the claims are no longer equal, and one can be more irrational than the other based on scientific skepticism and common sense.

Otherwise, there is a purple donkey punching my throat. You can't see me, so you can't claim that what I said is any less rational than if I'd said "I have a sore throat" despite the fact that the purple donkey situation is clearly less rational then just having a sore throat.

If the method of evaluating two claims are both equally unjustified, then the claim itself must be evaluated.

robindch · 17-04-2012 09:53AM

jjn2 wrote: »

Whether you find what I have said even remotely convincing or not is up to you,

Well, to a certain extent, it's up to you too -- the more convincing you are, the more likely it is that people here will find it convincing

jjn2 wrote: »

[...] I have addressed all your points.

I don't believe you've said whether or not the aim of your argument is the claim that all viewpoints are based upon "faith" (for its multiple meanings) and therefore, that epistemologically, all viewpoints are equally plausible.

Is it?

jjn2 · 17-04-2012 10:13AM

King Mob wrote: »

They are not made under the assumption of induction. Induction plays no part as the results are empirical. It would be inductive reasoning if we used these results to determine that the theory is an absolute.

No that's simply not true.
All theories and models in science are provisional and based on empirical observation. All of them are subject to chance in the event of the discovery of new evidence regardless of how supported they are

We disagree fundamentally on how science is carried out. The way I see it though, your view of science is along the line of Popper's deductive model of science, which I believe has been discredited. The following link discusses it

http://ls.poly.edu/~jbain/scitechsoc/lectures/04.Popper.pdf

Penn wrote: »

No, because again, once both methods of evaluating a claim are unjustified, the rationality behind each claim can be determined via scientific skepticism and common sense. Therefore, the claims are no longer equal, and one can be more irrational than the other based on scientific skepticism and common sense.

Otherwise, there is a purple donkey punching my throat. You can't see me, so you can't claim that what I said is any less rational than if I'd said "I have a sore throat" despite the fact that the purple donkey situation is clearly less rational then just having a sore throat.

If the method of evaluating two claims are both equally unjustified, then the claim itself must be evaluated.

We are also disagreeing on a fairly fundamental point. You claim that the rationality behind each claim can be determined via scientific skepticism and common sense. If these are based on inductive reasoning, and inductive reasoning has no rational justification, how can they be used to evaluate the rationality of any claim?

robindch wrote: »

Well, to a certain extent, it's up to you too -- the more convincing you are, the more likely it is that people here will find it convincing I don't believe you've said whether or not the aim of your argument is the claim that all viewpoints are based upon "faith" (for its multiple meanings) and therefore, that epistemologically, all viewpoints are equally plausible.

Is it?

I have obviously tried to be convincing, but judging from the responses it is equally obvious that I haven't been very convincing! I actually think I explained what I was getting at in the third paragraph of my very first post! I wouldn't agree with your interpretation of my viewpoint. I have only used faith in one way (and it was probably as bad an idea as talking about Christian scientists), to mean holding a belief despite having no justification or vindication for holding it. My question came in 2 parts, firstly, do you agree that we aren't justified in trusting in inductive reasoning (problem of induction), and secondly, if we aren't, what separates an assumption of inductive validity from an assumption that God exists.

Questions about induction and God

Comments