How I Solved Hume’s Problem and Why Nobody Will Believe Me

Discussion of articles that appear in the magazine.

Moderators: AMod, iMod

Philosophy Now
Posts: 1205
Joined: Sun Aug 29, 2010 8:49 am

How I Solved Hume’s Problem and Why Nobody Will Believe Me

Post by Philosophy Now »

Eugene Earnshaw saves Western philosophy.

https://philosophynow.org/issues/119/Ho ... Believe_Me
Impenitent
Posts: 4331
Joined: Wed Feb 10, 2010 2:04 pm

Re: How I Solved Hume’s Problem and Why Nobody Will Believe Me

Post by Impenitent »

definitely follows from is not "the conclusion is probably true."

nice article, but I'm still skeptical...

-Imp
Londoner
Posts: 783
Joined: Sun Sep 11, 2016 8:47 am

Re: How I Solved Hume’s Problem and Why Nobody Will Believe Me

Post by Londoner »

I do not think the writer's examples accurately represented the nature of induction.

The writer claims that Hume wants us to prove the truth of a statement like 'the future will always be like the past' to justify induction, whereas his contention is that what is required is just to show a certain sort of relationship between the premises and the conclusion, and that in this case the premises only have to make the conclusion probable. So, if we construct a deductive argument with a probable conclusion then induction becomes just a form of deduction.

I do not think this is the nature of induction. I think that induction rests on the belief that in the empirical world like gives rise to like. That is to say, that given the circumstances - all the circumstances - of the world in July 1914, the start of World War 1 was - not probable - but absolutely certain. That if the world could somehow be put back exactly to how it was at that date, exactly the same thing would happen again.

But we do not reason inductively that a World War will start every July, because the circumstances now are different. I think this is the case with all inductive reasoning; we never do claim that we have certainty that the future will be like the past - that is because we are aware the present is not like the past. Any piece of inductive reasoning always comes with the provision provided that... or ...all other things being equal.

To say that inductive reasoning is thought of as 'probable' does not seem quite right. The principle of induction, that like will give rise to like, is considered certain. What is uncertain is our knowledge of the state of the world - but that is true in any form of reasoning. A piece of deductive reasoning requires us to assume our premises are true; we say if this P and if that P, then C, but the truth of P must be derived from outside the deduction that uses it.

We may have more faith in a piece of inductive reasoning if we see an event repeated many times because we think this indicates that the state of the empirical world that made it happened once remains unchanged in some essential respects. But none of these repetitions were 'probable', each was certain to occur, given the circumstances. And if the circumstances change and something different happens next time, that would not be 'probable' either.

If we are doing science, we deliberately restrict what we understand by the circumstances of an event and the event itself. If I say 1cc of water weighs 1g this is not just 'probable', a conclusion gained from using induction on previous experiences of weighing water. It is certain, because my notions of 'water' and weighing have been made so precise that each event exactly duplicates the last, with no extraneous considerations. That being the case, then 'water' and 'the weight of 1cc of water' becomes a single fact, a certainty, such that if the water didn't weigh 1g we would know it could not be water.

So I would argue that the principle of induction is an axiom about the character of the empirical world, of the same character as those we use in deductive reasoning. Certainly this axiom cannot prove its own validity, but nor can any other axiom. True it cannot deliver specific knowledge about the empirical world with certainty, but (again) nor can any other axiom.
Eugene Earnshaw
Posts: 4
Joined: Sat Apr 01, 2017 4:40 pm

Re: How I Solved Hume’s Problem and Why Nobody Will Believe Me

Post by Eugene Earnshaw »

This is a very thoughtful response and I agree with a lot of it. I think the criticism of the article that Londoner provides can be mostly resolved by recognising that there are two relevant senses of induction at issue: the sense that Londoner describes in the positive remarks is (I think) real and important, but not the only correct sense. The sense of induction that refers strictly to arguments of a certain broad kind -- namely, those which attempt to establish a conclusion with some degree of probability rather than certainty -- is not the same as Londoner's sense. I think that these arguments are real and important, and the discussion of Hume's fork must address these arguments. This is because Hume's fork only works if it classifies all possible arguments into two kinds, and rules both kinds out. Therefore a sense of 'induction' which understands induction to be a kind of axiom, cannot be the relevant sense, since an axiom is not an argument.

That said, an axiom does play an important role in Hume's fork. Or, it would be an axiom if Hume would allow us to assume it: however, he does not. Londoner's axiom is another formulation of Hume's proposition "the course of nature continues uniform". What Londoner observes is that, if we help ourselves to this proposition, it makes our reasoning certain insofar as we can recognise where it applies. Yes, exactly. That is why by insisting we must justify that proposition, Hume is demanding we turn inductive arguments that provide probability into deductive arguments that provide certainty. However, Hume is not content to allow Londoner to assume that proposition as a rule of reasoning: Hume demands it be justified, and argues it cannot. So in the end, Londoner must either rebut Hume, or ignore him.

Now there is an aspect of the argument I did not really develop, on account of keeping the article as simple and direct as possible. Once we have shown that Hume's fork fails, and that there is no objection to probabilistic reasoning, then we have also rebutted his argument that the propositions such as "the course of nature continues uniform" or "like follows like" cannot possibly be justified. The rebuttal of Hume's fork relies on the recognition that such premises are not necessary for justifying inductive arguments generally, but it is the case (which the article does not discuss) that justifying such propositions would be quite useful as a basis for establishing a fair amount of the everyday knowledge that we actually rely on. And, since Hume's fork is indeed fallacious, it may be possible to do so, which would be nice. But even that may not be necessary: given that we know that these principles are true, and we also have no positive reason to think they are unjustifiable, we may perhaps help ourselves to them even before someone explains in detail how we might be justified in believing them.
Nathan Oseroff
Posts: 1
Joined: Wed Apr 12, 2017 11:24 am

Re: How I Solved Hume’s Problem and Why Nobody Will Believe Me

Post by Nathan Oseroff »

Dr Eugene Earnshaw,

I provide a brief response to your article, 'How I Solved Hume's Problem of Induction', published in Philosophy Now, here: https://nathanoseroff.blogspot.co.uk/20 ... nshaw.html. Your thoughts would be greatly appreciated.

Kind regards,

Nathan Oseroff
Eugene Earnshaw
Posts: 4
Joined: Sat Apr 01, 2017 4:40 pm

Re: How I Solved Hume’s Problem and Why Nobody Will Believe Me

Post by Eugene Earnshaw »

Hi Nathan,

Thanks for responding to my article. I've posted a detailed response at the blog you linked to. For those wanting the brief summary without having to click on the link (but do click the link): Nathan is critical of my deductive justification of induction and thinks that Stove and Williams and I should stop trying to win arguments by rhetorically berating our opponents; I think Nathan is missing that my main point was rebutting Hume's argument, and is responding more to what Stove and Williams said than what I actually wrote.
spike
Posts: 850
Joined: Thu Oct 22, 2009 3:29 pm

Re: How I Solved Hume’s Problem and Why Nobody Will Believe Me

Post by spike »

Londoner,

You are right. I don't think Hume was right in believing that 'future will always be like the past' . Kant had a similar belief, that people wouldn't move forward with mental ideas that improve politics.

In many respects Kant consider humans 'stupid'. I sure he would have believed 'human rights' stupid.
Last edited by spike on Sun Apr 16, 2017 1:51 pm, edited 1 time in total.
User avatar
Necromancer
Posts: 405
Joined: Thu Jul 30, 2015 12:30 am
Location: Metropolitan-Oslo, Norway, Europe
Contact:

Re: How I Solved Hume’s Problem and Why Nobody Will Believe Me

Post by Necromancer »

My take on Hume's:
Factors
1. Consistency and coherency (deduced from Descartes' Meditations)
2. The sum of natural laws (the usual consideration of our beings in nature)
3. Logics and mathematics are necessary aspects of nature and our minds (I hold the view that logics and mathematics are for real in a variety of senses)
4. The factors of 1. through 3. bridge our experience from one instant to the next and so on. Forever?
5. Point beside: 1. through 4. refute, in my opinion, "Hume's Custom or Habit", the problem of induction.

Comment on Lawfulness. With lawfulness, it's presumed that your set of observations complies and is exhaustive for all conditions this lawfulness concerns. This means that your set in your span of time, instance, is complete to the degree that information overall in reality is contained and is complete in this span of time, instance.
In the opposite, to have a confirmation bias to everything you experience, you have to believe that:
* there is no possibility for describing lawfulness in reality.
* there is no lawfulness in nature, simply!
This should be contradictive to intelligence to most people. I can't see any possible history of earth with a reality from those two beliefs if they are to represent reality.

For Hume on billiard, just watch the champs play it out in snooker on Eurosport! (Or wherever, billiard too if so be!) :D
bwstarke
Posts: 1
Joined: Tue Apr 18, 2017 12:23 am

Re: How I Solved Hume’s Problem and Why Nobody Will Believe Me

Post by bwstarke »

I read the article and shared it with my brother (a philosophy professor). He made the following point:

The conclusion "It is probable that the individual randomly selected from the population will blow the bugle beautifully" is only supported by a hidden third premise that future ability to blow the bugle will be like past ability to blow the bugle, and this is Hume's problem all over again.

Is he right? If not, why not?

Thanks.
User avatar
A_Seagull
Posts: 907
Joined: Thu Jun 05, 2014 11:09 pm

Re: How I Solved Hume’s Problem and Why Nobody Will Believe Me

Post by A_Seagull »

Eugene Earnshaw wrote:.
It seems that some of our posts have been deleted from this thread.

So I will try to pick up the main point anew.

You claim that : "Hume’s problem is that induction is unjustifiable." What do you mean by "unjustifiable"? And why is it a 'problem'?
spike
Posts: 850
Joined: Thu Oct 22, 2009 3:29 pm

Re: How I Solved Hume’s Problem and Why Nobody Will Believe Me

Post by spike »

I would really get the 'Hume' thing.
steven t johnson
Posts: 1
Joined: Fri May 05, 2017 3:47 pm

Re: How I Solved Hume’s Problem and Why Nobody Will Believe Me

Post by steven t johnson »

Professor Earnshaw wrote " Hume was demanding in a confused way that inductive arguments be turned into deductive arguments by a supplemental premise..." The supplemental premise was that "the future will be like the past, or that nature is uniform." There are I think two issues here. First, there is the notion that true justification demands deductive certainty. Absent that, there is no justification at all. It is obvious that there is no way to convince philosophers this is not a requirement. There is no power to demand that God provided a universe that can be irrefutably deduced from a priori first principles, that we can replicate the Mind of God as He created the universe. Nor is it clear why that kind of certitude is pragmatically necessary. Hume himself explicitly declared his daily life did not require it. And so far as I know, no one, not even Bertrand Russell, has actually demonstrated why it is impossible to accept the possibility of error. Also it is unclear that philosopher of knowledge can declare that a probability of error too small to calculate means that concluding a statement is true is unjustified, without taking upon themselves the power to legislate special meanings for words.

Second, there is Hume's unspoken assumption that a non-uniform nature can be an actual thing. I suggest (no doubt following upon many unsung predecessors,) there is no such thing. Descartes spoke much about "clear and distinct" ideas, by which he meant what modern philosophers call "coherent" ideas. (I think, therefore I am wrong?) Hume's notion of an unpredictable future is truly the same as the proposition that a divine Deceiver can at any moment rip away the veil. It is not clear to me that logical possibilities are always useful to reason. But illogical possibilities masked by simple words I think are clearly obstacles to reason. So far as I know there is not even the slightest attempt to sketch out how the universe could work without some sort of uniformity, and even less to explain away the experience that says it does.

There are numerous philosophical commitments to denying any status as knowledge to the conclusion that experience has in fact demonstrated no exceptions to the uniformity of nature, as scientism. For my part, I've come to conclude that any philosophy of science that cannot cope with induction is a failed philosophy of science. The thing that I think is overlooked, is that Hume's fork stabs in the heart any kind of science that isn't narrowly empirical, limiting itself to "facts." Hume's fork stabs not just philosophical metaphysics and sectarian theology (it's real target I think) but abstract concepts as well. Talk about energy and entropy and such like non-facts are to burn in Hume's bonfire too.

Earnshaw flirts with the notion this sort of thing is personally motivated. But it seems to me that epistemological skepticism is always associated with political reaction. Hume in his lifetime was primarily the great Tory historian. You may prefer to think of Kant as a moderation of the Enlightenment rather than a reactionary. But the example of the neo-Kantians I think refutes this. But perhaps these are issues too large for mere philosophical discussion. In many respects, rhetoric is not a tool of philosophy. Instead, philosophy is a branch of rhetoric, the branch that devotes itself to the appeal to reason. In practice of course it is inseparable from other branches of rhetoric like the appeal to emotion. As an exercise, philosophy is what someone (Descartes?) said it was, the considerate effort of gentlemen to explain and justify their opinions before the world, condescending to the intercourse of minds. But I am no gentleman. Sometimes I'm not sure that philosophy is of any more value than Sudoku, an entertainment by puzzle.
Eugene Earnshaw
Posts: 4
Joined: Sat Apr 01, 2017 4:40 pm

Re: How I Solved Hume’s Problem and Why Nobody Will Believe Me

Post by Eugene Earnshaw »

A_Seagull wrote:
Eugene Earnshaw wrote:.
It seems that some of our posts have been deleted from this thread.

So I will try to pick up the main point anew.

You claim that : "Hume’s problem is that induction is unjustifiable." What do you mean by "unjustifiable"? And why is it a 'problem'?
For Hume, unjustifiable means that you cannot establish using reason that inductions should be believed. That's based on the argument called Hume's fork in the article. And it is a problem because, if that is the case, much of our 'knowledge' is not actually known, such as what is likely to happen if I drive my car very fast into a brick wall..
Eugene Earnshaw
Posts: 4
Joined: Sat Apr 01, 2017 4:40 pm

Re: How I Solved Hume’s Problem and Why Nobody Will Believe Me

Post by Eugene Earnshaw »

steven t johnson wrote:Professor Earnshaw wrote " Hume was demanding in a confused way that inductive arguments be turned into deductive arguments by a supplemental premise..." The supplemental premise was that "the future will be like the past, or that nature is uniform." There are I think two issues here. First, there is the notion that true justification demands deductive certainty. Absent that, there is no justification at all. It is obvious that there is no way to convince philosophers this is not a requirement. There is no power to demand that God provided a universe that can be irrefutably deduced from a priori first principles, that we can replicate the Mind of God as He created the universe. Nor is it clear why that kind of certitude is pragmatically necessary. Hume himself explicitly declared his daily life did not require it. And so far as I know, no one, not even Bertrand Russell, has actually demonstrated why it is impossible to accept the possibility of error. Also it is unclear that philosopher of knowledge can declare that a probability of error too small to calculate means that concluding a statement is true is unjustified, without taking upon themselves the power to legislate special meanings for words.
I'm unclear on what the message is here. Hume's argument isn't that the possibility of error is unacceptable. It is that we cannot justify the idea that the future will be like the past, and hence inductive arguments are unjustified. So it's not that inductive arguments are bad because they are not totally conclusive. It is that they are bad because they are rationally totally useless. Now, I say that the idea we need to prove the statement 'the futre will be like the past' is wrong, because that only turns inductive arguments into deductive ones. But it doesn't sound like you are addressing that idea at all, so I am a little unclear what is going on in relation to the article.
Belinda
Posts: 8034
Joined: Fri Aug 26, 2016 10:13 am

Re: How I Solved Hume’s Problem and Why Nobody Will Believe Me

Post by Belinda »

Predicitions arrived at via induction are probabilistic.'Probabilistic' means that we don't know all the contributory causes of some predicted event.

The Morning Star doesn't cause the Evening Star despite that those have been constantly conjoined in sequence. The two effects, Morning Star and Evening Star have the same cause which is the Planet Venus .

The above exemplifies that we might surmise that the known material common causes of different perceptions might all eventually coalesce into a set of rules and then we can all be happy pantheists.
Post Reply