Talk:Induction: Difference between revisions

This article does not present an objective presentation of the subject but the author's editorial opinion. First off, others would disagree that Quine "debunked" the analytic/sythetic distinction. Second, others would disagree that the Quine-Duhem thesis refutes Popper. Those who cite the Quine-Duhem thesis apparently go by second hand sources. In "The Library of Living Philosophers" series, in the two volumes devoted to Karl Popper, Quine and Popper were in large agreement. In Quine's paper, Quine brought up Quine-Duhem and Quine HIMSELF did NOT think it posed a big problem for Popper. Go read it for yourself! -63.98.140.237 Sep 13, 2003

Wikipedia dictates a NPOV policy, not an objectivity policy. Objectivity is a myth. The article does present a POV in POV fashion when it should present the POV in NPOV fashion. -B 22:18, Oct 23, 2003 (UTC)

I don't particularly like the way I've rephrased the opening paragraph in my last edit, but I think it is better than its predecessor. I'm just trying to shoot for more specificity here. -B 22:18, Oct 23, 2003 (UTC)

Well said, B. In point of fact, Popper expresses what later became known as the “Quine-Duhem thesis” in section 19 of The Logic of Scientific Discovery, preceding Quine’s Two Dogmas. Only a naive reading of Popper would lead to the conclusion given in this article Banno 19:53, 24 Oct 2003 (UTC)

---

Broke down the Quine/Hume/Popper paragraph. My changes at least have the advantage of not actually being wrong (I hope!!). Is there a purpose to retaining the discussion of the analytic/synthetic distinction? At the least, it needs something to point to its relevance. Personally, I liked the old introduction. Banno 22:59, 24 Oct 2003 (UTC)

I'm not stuck on my new intro and the previous one was actually a common way of introducing the subject. I think a better intro than the previous one nd mine can be written, but I don't object to reverting to the previous intro. Re: analytic/synthetic distinction, I see where the material is leading to...to Quine and the problem of induction...that material is actually more pertinent to the problem of induction article but it is in this article rather than the other. —B 04:36, Oct 25, 2003 (UTC)

---

Proposal to merge articles

Since there is no clear distinction between this article and problem of induction, Unless there is a loud protest, I'll move the stuff from inductive reasoning to here in a few days. Banno 10:54, 7 Dec 2003 (UTC)

My previous response to Banno here was in error as noted on the talk page to the problem of induction. I do not protest. B 21:54, Dec 17, 2003 (UTC)

---

Amateurish post

I removed/edited this garbage:

Some consider that the scientific method relies on inductive reasoning. However, many researchers use hypothetico-deductive approaches derived from the work of Popper and R. A. Fisher. The validity of some forms of such inductive and deductive reasoning is formally described by statistics.

The scientific method relies on both induction and deduction. "Hypothetico-deductive approaches" is merely a specialized definition of deduction. The validity or justificiation of deduction and induction are grounded in logic and philosophy respectively, not statistics. Statistics also rely on induction and deduction, but statistics is not in the business of justifying or validating induction and deduction. B 00:01, Dec 8, 2003 (UTC)

I quite agree – wonder who posted it.Banno 02:10, 8 Dec 2003 (UTC)

The Cunctator added the first and last line. Jfitzg (aka Trontonian) added the line about "...many researchers..." and modified the first and last line. It was Trontonian's edits that bugged me, but he no longer participates in wikipedia. B 18:43, Dec 17, 2003 (UTC)

---

Induction not a "problem" for science?

I have seen no justification for the claim that science depends on inductive reasoning, defined in such a way that it has a "problem". No reputable scientist should claim that a particular hypothesis or theory is the only explanation for a set of observations. It is merely the "best" explanation -- by such criteria as Ockham's Razor -- known to the scientist. For any set of observations, there are presumably an infinite set of potential explanations. You may argue that Ockham's Razor is derived by induction, but it is a probabilistic statement, not an absolute dictum of truth. Science does not predict that the sun will come up tomorrow; it states that we can expect such a result if the current theory is approximately true. Where is the problem? Fairandbalanced 21:12, 15 Dec 2003 (UTC)

It’s not a problem for science, but for the philosophy of science. Induction is not deductively justified - of course – so is there some other way of justifying it? Perhaps hypothetico-deductive method describes the problems with the method you describe more clearly?

Do you know of a scientific article in which someone dismisses a theory on the basis of a probabilistic statement derived in some way from Occam’s razor? I’d love to see one. Banno 20:07, 16 Dec 2003 (UTC)

Fairandbalanced, all scientific theory (and knowledge in general) relies on the principle of the uniformity of nature which is based on induction. The problem is that there is no philosophical justification for induction (and by implication for uniformity); it is, like Occam's Razor, merely a principle of preference. Induction is a problem not only for science, but knowledge in general. We may prefer that the universe be uniform, but there is no justification that the universe is uniform (nor that it is not uniform). All that we "know" is what we have experienced so far...and our future experience may give us reason to "know" otherwise. Notice "know" is always tentative...THAT is the problem; there is no absolute, certain knowledge. A far greater problem to knowledge is that posed by confirmation holism and its counter-part, ontological relativity. These seem to be related to the problem of induction, but go beyond it...but I'm not sure. B 21:54, Dec 17, 2003 (UTC)

---

Fine. Then it should not be touted as a problem for science, but as a problem for all knowledge. Science makes predictions based on the assumption that its theories are adequate representations of reality, and that assumption can be falsified by a contrary result. What field of science asserts knowledge of absolute truth?

Ockham's Razor apparently was used frequently in debates regarding continental drift during the first half of the twentieth century. It also rules out a lot of pseudoscience and paranormal rubbish. I'm sure it has appeared elsewhere. Fairandbalanced 21:50, 2 Jan 2004 (UTC)

Bayesian method

I reverted edits that defined induction as the making of an invalid inference, and then confuse Bayesian inference with induction. This article does need some material on Bayesian method, but it needs to be done in a competent fashion. Banno 23:06, Jul 6, 2004 (UTC)

Redirect

There is no point in the re-direct unless there is some content to go into induction; the re-direct disenfranchises the dozens of pages that link to induction. If someone wishes to keep the new definition, they also need to fix each and every one of the re-directs. It would have been courteous to at least suggest the re-direct in the discussion pages and wait for comment; the present state is just making work for others. I’ll allow a bit of time for discussion, but I think the re-direct should be reverted. Banno 07:51, Jul 8, 2004 (UTC)