Induction

[Dooga Aetrus Blackrazor]

http://en.wikipedia.org/wiki/David_Hume#Th...em_of_induction (http://en.wikipedia.org/wiki/David_Hume#The_problem_of_induction)
http://en.wikipedia.org/wiki/Problem_of_induction

Is induction invalid? Should we use it even thought it is not reliable? What does everyone think?

[apathy maybe]

As Hume said, it is built into us. We have to use it, 'cause you can't learn about the world deductively. And if we know nothing about the world, then we'll starve.

[Angry Young Man]

I think we know bugger all and that empiricism (including induction, even though it involves coherent logic) is the only way we can explain things in a tolerable way. From the little I know of Hume, his epistemology makes alot of sense.

[Hegemonicretribution]

Induction is problematic, and I have maintained this for a while. However it does seem to work, and does aid us in our continued existence.

If the choice is the possibility of doubt or complete solipsism I think I know what works best. We must simply accept that we could be wrong, but whilst our theories work we should also continue using them. It is through this continuous testing that we can uncover problems and areas for improvement.

[Floyce White]

The problem isn't induction/deduction. The problem is when people discover that their pet theories are false according to induction/deduction, but assert that their theories are nonetheless "true" because they are not disproven by dialectics.

[Bretty123]

I believe Hume in the fact that we always have to be inductive to discover new things, however I don't believe it is built into us in any metaphysical or epistemic way. It's natural, going on observations of humans, to inductively search and test hypotheses.

I've always assumed that science is not a practice of only discovering new things through induction but also one of falsifying old ideas. I guess you could say searching for facts of nature and reality is a practice of falsification.

[apathy maybe]

This is a Popperian idea, and while I used to agree with you and think it made sense, I have since seen (after some study) that it has problems. In fact all the theories of science have problems.

The idea of falsification (using deduction rather then induction) is that it relies on induction. If you disprove a theory, you use induction to assume that every time you repeat the experiment you will also disprove the theory.

Popper accept ideas around "theory-ladenness" that also can cause problems for falsification. I wrote an essay on the subject which can be found ...
http://ids.org.au/harrismw/writing

I am currently (well ...) writing a piece on other theories of science, basically Kuhn's idea of paradigms does not actually distinguish what science is and what is not, astrology can claim to be a science because it fits a paradigm. Same with Marxism. Feyerabend rejects rationality and the objectivity of science altogether.

[Hegemonicretribution]

I know where you are coming from apathy maybe, I also think that the Duhem-Quine thesis has some relevence here, as falsification becomes much more problematic even without induction.

As far as science goes Kuhn seems to make sense, even if the conclusion isn't that we know everything. To be honest that is where I think we are at, once we do "know" everything science becomes obsolete and we are a long way from there. So in a way the doubt that is provided for by induction prevents stagnation in our sciences.

[apathy maybe]

I don't really like Kuhn, because he doesn't really draw a line that says what is on one side is science and on the other is not. As I have said previously, astrology is a Kuhnian paradigm.

Personally I think that science is about attempting to make closer and closer approximations to reality. We use whatever methods (preferably rational), create theories, then compare these theories to what we perceive.

If enough people see that the theories fit, then we have a good approximation to reality.


(I haven't actually really read anything on the Duhem-Quine thesis or by either of the two characters. I guess I should?)

[Hegemonicretribution]

Newton was an alchemist, and his scientific method in physics would be thrown out today, yet I don't know a "scientist" that does not recognise the impact of him. For Kuhn, the evolution of science should be led by thye best scientific minds (those that spend their lives doing science), and the concepts accepted should be those best able to deal with anomaly.

On the one hand it might be hard accepting as science that which has been disproven, yet at least this allows for changes. Other views often exhalt current theories whilst ridiculing that which was a prequisite for current thinking because it has been replaced by an improved version.

As for the Duhem-Quine or Quine-Duhem thesis it is a faily simple idea that is key in philosophy of science: It is a response to falsification theory. Basically it states that because all theories are so interlinked that you can only falsify a theory when it is taken with all the other theories as one...for example. A theory of gravity may predict an orbit, and be off by quite some way. Now if it turns out that another undiscovered planet wsa responsible for this, then this anomaly may even help prove the theory. Falsificationists should dismiss it for being wrong, but really it was contemporary astronomy that was out. Hypothesis upon which theories depend are called auxhiliary hypothesis; the problem then becomes establishing whether or not the fault in an experiment is within the new theory or an established hypothesis upon which it relies.

Popper said Marxism was falsified, and is pseudo science for still proclaiming itself. The Duhem-Quine thesis would state that it may well have been another issue that caused Marx's predictions to be off, and that it is the auxhilliary hypotheses that we should focus on (perhaps) in the many forms they appear. Of course this can be taken to more or less any extreme, however it does appeal in a strange way....that is a basic outline anyway.

[apathy maybe]

Aha. I have heard of the ideas, I obviously just didn't remember them ...

It does have some appeal, I can see that.


As I said, I think science is about providing approximations to the world. Any method is possible, but I like the method that Popper puts forward. The one method of philosophical enquiry (in fact "the one method of all rational discussion") is "that of stating one's problem clearly and of examining its various proposed solutions critically". This is from Popper's Logic of Scientific Discovery. And while falsification does have problems, what I have quoted here is not falsification. It is simply a way to make sure that you have not missed anything.

[rouchambeau]

I think we know bugger all and that empiricism (including induction, even though it involves coherent logic) is the only way we can explain things in a tolerable way. From the little I know of Hume, his epistemology makes alot of sense.

So things that are not sensible cannot be explained?

[hoopla]

I am reading Hume at the moment. Enquiries is alot easier if anyone is interested, apparently skips alot though. Anyway, Hume did think that induction was "reasonable" even if not derived from reason - it generally matches the shape of the world etc.

Not that convinced by Hume as a complete philosophy, tbh.

Can Rosa explain Wittgenstein's argument against extreme empiricism - I don't understand any I've come across? I am fairly sure this is my own fault.

[hoopla]

Rosa did not help, which is odd considering the number of empiricists here.

I would guess that its something about how in extreme empiricism everything muct come prepackaged, a hammer is the impression of a hammer, a car the impression of the car? That might be totally wrong.

How could pure impression lead us to seek out food (which we do do)? It would be a fantastic prepackaged impression that caused goal directed behaviour if there was no beginning potential for it in the individual.

If we don't already know that we are looking for food, how would we know when we had found it? We do look for and find food.

Please explain, Rosa.

[hoopla]

Please Rosa, can you explain Witts argument against empiricism. I think it would help many comrades here.

[gilhyle]

The problem is not with induction - which clearly exists and patently 'works' in some significant sense. The problem is with the ideological impulse to have a general theory abstracted from content which wants to legitimate forms of arguments (as distinct from fomal systems). Equally the problem is the desire to have a theory of science which dictates what is valid science and what isnt. It is history that determines the nature of science and not any uniquely valid method. The scientific debate between Popper, Kuhn Feyerabend and Lakatos resolved this issue forty years ago - thats not to say there arent still significant philosophy of science questions and its not to say that its wrong to criticise something as pseudo science - I think it is practical to do so on the basis of a detailed comparison of working methods with closely comparable sciences.

But to do that yields a false sense of security that we have grasped science. Science has an ideological and a sociological/economic/institutional aspect to it which cant be ignored - just as the context, genesis and motive of inductive arguments must contribute to any assessment of them.