Analysis of Robert Spillane Quoting Roger Kimball

Rather than spend time constructing new sentences to respond to your questions and comments, I quote Roger Kimball (who made Americans aware of David Stove's essays).

'At the center of Popper's thinking about the philosophy of science is a profound skepticism, derived from David Hume, about the rationality of inductive reasoning...

This is incorrect. With Popper, I hold there is no inductive reasoning. Induction is a myth. No one has ever induced a conclusion. Since inductive reasoning doesn't exist, judging it as rational or irrational is beside the point. (This is why I try to ask questions about instructions for doing induction, or about which ideas to induce in a given situation. Without answers to these questions, then induction can't be done and can't reach any conclusion at all, rational or not.)

So the quote you're giving doesn't engage with the position I'm advocating.

Like the young Hume, Popper concluded from the fact that inductive reasoning was not logically valid - that inductive evidence does not yield absolute certainty - that it was therefore incapable of furnishing compelling reasons for belief.

This assumes one has induced some conclusion(s) and the issue is to debate whether we should accept those conclusions are rational, valid, practical, partially certain, etc.

But, as above, that isn't the issue. So this isn't engaging with the position I'm advocating.

Popper was a deductivist. He dreamt of constructing a philosophy of science based solely on the resources of logic.

No he didn't. Quote? Source? Conjecture – which played a huge role in Popper's epistemology – isn't deduction. Popper also emphasized explanation and problem solving, which aren't deduction.

He was also an empiricist: he admitted no source of knowledge beyond experience. As Stove shows, the combination of empiricism and deductivism - in Hume as well as in Popper - is a prescription for irrationalism and cognitive impotence. An empiricist says that no propositions other than propositions about the observed can be a reason to believe a contingent proposition about the unobserved; an empiricist who is also a deductivist is forced to conclude that there can be no reasons at all to believe any contingent proposition about the unobserved.

Popper explained what we can do instead of having positive reasons: we can make unjustified conjectures. We can then use criticism to improve our ideas and make progress. Error elimination, not justification, is the key to epistemology.

Whether Popper (and I) are correct or incorrect about this view, the quote isn't discussing it. The quote isn't a reply to us.

Hume himself, in his posthumously publishedDialogues Concerning Natural Religion, ridiculed this "pretended skepticism" as a juvenile affectation...

'Popper resuscitated Hume's brand of skepticism, dressing it up in a new vocabulary. In Popper's philosophy of science, we find the curious thought that falsifiability, not verifiability, is the distinguishing mark of scientific theories; this means that, for Popper, only theories that are disprovable are genuinely scientific...

What's wrong with that? And how can theories be verified? The quote doesn't say.

Popper denied that we can ever legitimately infer the unknown from the known;

Yes, but Popper says we can know about the unknown by methods other than inference. This quote doesn't discuss that.

audacity, not caution, was for him the essence in science; far from being certain, the conclusions of science, he said, were never more than guesswork...;

Right. What's wrong with that? Do you claim we can be certain – meaning we can have infallible knowledge? How?

and since for Popper "there are no such things as good positive reasons" to believe a scientific theory, no theory can ever be more probable than another; indeed, he says that the truth of any scientific proposition is exactly as improbable as the truth of a self-contradictory proposition - or, "in plain English", as Stove puts it, "it is impossible".

The quote isn't providing a criticism of Popper's view. It also, so far, hasn't said anything about the two questions I asked.

'...What was novel [about Popper's doctrine] was the amazing thought that positive instances do not - in principle cannot - act to confirm a proposition or theory. For Popper, if every raven anyone has ever seen is black, that fact gives no rational support for the belief that all ravens, in fact, are black.

Correct: if you want to argue that all ravens are black, you'll need a different argument that doesn't have a logical hole in it. This quote isn't stating what the logical problems with positive support are that Popper explained, nor answering Popper's arguments.

Scientific laws, he says, "can never be supported, or corroborated, or confirmed by empirical evidence". He goes even further: of two hypotheses "the one which can be better corroborated, is always less probable." Whatever else these statements may be, they are breathtakingly irrationalist...

Rather than provide a criticism of Popper's view, the author states the view then calls Popper irrationalist (without defining it).

What am I supposed to learn from this?

'It would be difficult to overstate the radical implications of Popper's irrationalist view of science. Popper was apparently found of referring to "the soaring edifice of science". But in fact his philosophy of science robbed that edifice of its foundation. Refracted through the lens of Popper's theories, the history of modern science is transformed from a dazzling string of successes into a series of "problems" or ... "conjectures and refutations". On the traditional view, scientific knowledge can be said to be cumulative: we know more now than we did in 1899, more then than in 1699. Popper's theory, which demotes scientific laws to mere guesses, denies this: in one of his most famous phrases, he speaks of science as "conjectural knowledge", an oxymoronic gem that, as Stove remarks, makes as much sense as "a drawn game that was won." (This paragraph contradicts your statement that Popper supports a cumulative view of science).

Overall, the quote is full of conclusion claims instead of arguments. It also doesn't speak to the two questions I asked.

(Roger Kimball, Against the Idols of the Age, Transaction, 1999, pp. xxi-xxiii).

I have answered your questions #1 and #2 directly and indirectly.

To repeat: in The Rationality of Induction, Stove has answered your questions.

I have the book but you didn't tell me which pages you believe answer the question. When you provide the page numbers which you claim answer me, then I'll read them.

Needless to say, I agree with him.

Specifically, Stove reduces inductive inferences to the inference from proportions in a population. As mathematician, James Franklin, writes: 'It is a purely mathematical fact that the great majority of large samples of a population are close to the population in composition'. In cases such as political polling the observed, if based on a large enough sample, is probably a fair sample of the unobserved. 'This applies equally in the case where the sample is of past observation, and the population includes future ones. The sample is probably still a fair one, and one can make a probable inference (unless, of course, one has further reason not to: probable inferences are always relative to the evidence at hand).' (J. Franklin, Corrupting the Youth: A History of Philosophy in Australia, 2003, p. 338).

For any finite set of data, there are multiple ways to infer from proportions in the population which contradict each other. So which inferences from which proportions is one to find (by what means?) and then accept?

Note that this is the same two questions I asked in my previous email. The questions were about (1) which ideas do you induce and (2) how much inductive support do they have (so, if there's more than one, which is accepted over the others for having more support?)?

Also, related, the future always resembles the past in some ways and not other ways. So how do you approach the issue of which proportions of populations will hold in the future and which won't?

If Stove answers this, simply provide a reference (page numbers) where I can find the answer.

A quick look at 'The Analytic-Synthetic Distinction' on Wikipedia gives a list of those philosophers who have, quite rightly, rejected Quine's criticism. Quine is routinely quoted by philosophers and psychologists who, I suspect, have never read Ayer, Quine or Strawson. I say this because they rarely, if ever, make clear exactly what Quine's arguments were.

I don't care about lists of people who took some position, I care about arguments.

Admittedly, your short rejection is not Quine's but the criticisms of Quine can be applied to you.

Which criticisms?

You will never convince me that the following two propositions are logically and empirically the same: 'All tall men are tall' and 'All tall men are blond'. By rejecting the a-s dichotomy, you deny the possibility of necessary truths. Do you accept, then, necessary falsity?

No I don't accept necessary falsity. It's the same issue. To judge if 1+1=3 you still have to sum 1 and 1 and compare the sum to 3. The arguments I gave about 1+1=2 apply to this too.

Elliot Temple | Permalink | Messages (8)

10 Robert Spillane Replies

Robert Spillane's latest email didn't directly reply to what I said previously. Here it is with my new comments which attempt to get discussion back on track:

1. '3 am in the morning' is a pleonasm and thus necessarily true.

2. '3 am in the afternoon' is an oxymoron and thus necessarily false.

We need to conclude our discussion of whether 1+1=2 is a necessary truth before opening a new, similar topic. My answer to the 3am issue is similar to my answer to 1+1=2, which is the easier case to discuss and which I already wrote an explanation of. I await your next reply about that.

If I end up conceding the point about 1+1=2, I expect I'll also concede about the 3am issue without any additional arguments. And if you concede about 1+1=2, then I think your reasoning will be relevant to the 3am case and make it easier.

3. 'Induction exists' cannot be falsified.

Why? My position (which is also Popper's) is that induction has never had any set of followable instructions (steps) with the properties claimed by inductivists. So no one has ever done induction since inductivists have never defined any set of possible steps someone could do that would constitute doing induction. There are also arguments for why no such set of steps could be invented in the future. This is why I've asked questions about how to do induction (what the steps are).

4. 'Inductive logic' can be rejected if one argues that 'inductive logic' is an oxymoron. But since you don't accept oxymora, you have to argue that you reject 'inductive logic' on empirical grounds. How do you do that without distorting the meaning of 'empirical'?

I can use logical arguments. There's nothing wrong with logic. I just said the laws of logic are based on the laws of computation which are based on the laws of physics, and physics is an empirical science.

5. If you can't reject it on empirical grounds, all that is left to you are your feelings - and they are irrelevant since one cannot argue with feelings.

I agree that feelings are irrelevant. I haven't brought them up.

6. It is a truism that inference from experience is not deductive. A proposition may imply another proposition, but an experience cannot imply another experience. But you deny that there can ever be an inference from experience? That is untenable. What do you think 'inference' means?

Inference means "a conclusion reached on the basis of evidence and reasoning."

Induction refers to some specific ways of learning using experience. CR says those are poorly defined and actually impossible to do, and there are other ways to learn from experience which work instead (conjectures and refutations – evolution).

7. If Popper rejected induction, he has to be a deductivist - what else could a philosopher who calls himself a (critical) rationalist be?

A person who thinks most arguments are neither inductive nor deductive. Both induction and deduction are pretty specific categories which most arguments don't fit into. More on this below. BTW this has been noticed by a lot of people – e.g. it's the issue "abduction" is intended to address.

8. In his Unended Quest (Fontana, 1977, p.79) Popper writes: '...I could apply my results concerning the method of trial and error in such a way as to replace the whole inductive methodology by a deductive one. The falsification or refutation of theories through the falsification or refutation of their deductive consequences was, clearly, a deductive inference (modus tollens)...

That doesn't say Popper could or did replace the whole of thinking or arguing with deduction. Popper is just saying that if you accept basic (observation) statements then you can deduce to reject theories which they contradict.

9. You repeatedly claim that I do not engage with your position. But what exactly is the position of a person who rejects necessary truths and falsehoods, rejects induction and yet claims not to be a deductivist?

Why don't you quote what I write and reply to quotes more? I have asked you direct questions – e.g. the two about induction – and you haven't replied in this email. I also asked, again, for criticism of my position regarding 1+1=2 not being a necessary truth, and you didn't reply to that.

I take specific things you say and reply directly to them. But you mostly don't use that method when you respond to me.

I attempted to explain my position about non-deductive, non-inductive arguments with the price controls and socialism example. You didn't discuss it. I tried again by commenting on your argument about "mental illness" which you claimed was deductive, and you stopped discussing that too. If you will continue discussing one of the issues – especially if you quote what I say and reply directly to it – then I think we could make progress. I don't think it's a good idea to open another, new attempt to discuss the matter instead of continuing one of the discussions we were already having.

10. Where do your conjectures come from, since you deny they come from experience? And how do you refute them if not by deduction?

Brainstorming involves generating random variants of existing ideas. This is like genetic evolution which generates random variants of existing genes.

Many ideas are interpretations of experience. Interpreting experience is different than being guided by experience. Observations are passive data which can't tell us what to think. Instead we think for ourselves and some of our reasoning references observations, e.g. by critically pointing out that an idea contradicts an observation, or more mundanely e.g. by saying "I'm not going to go that way because I saw a cliff over there and I don't want to fall."

Ideas are refuted (in the context of a particular CR-problem) by criticism. A criticism is an explanation of why an idea doesn't solve a CR-problem(s). A "CR-problem" is very broad and refers to any type of achieving a goal or purpose, answering a question, etc – accomplishing anything you'd want an idea to succeed at. (I prefixed the word "problem" because it's Popper's terminology, I don't know a better word, but you objected to it previously so I don't want CR-problems to be mixed up with "problems" in your terminology.)

Explanation is a key part of thinking and arguing which is covered by neither deduction nor induction. Explanations discuss why and how. Statements following a "because" are generally explanations.

If you carefully analyze the arguments from most thinkers, including Szasz and your own books, you'll find many of them don't follow the rules of deduction or induction, and involve explaining why some idea fails to solve a CR-problem(s).

This would involve carefully defining what qualifies as both induction and deduction. I've asked you questions about this regarding induction.

Regarding deduction, it's CR-problematic too. Deutsch discusses that some in FoR ch. 10, the chapter I referred you to previously. In short, people don't actually agree about what the rules of deduction are, and it's a very hard CR-problem to address. You may define "deduction" as only Aristotle's syllogisms, but then you'll find you can't prove much and you won't be able to classify very many arguments as deductive. If you want a broader deductive system, you'll have to specify it and address issues like Godel's incompleteness theorem.

You'll also have to face the CR-problem that you won't be able to rely on deduction to argue for your deductive system against rival deductive systems, or criticisms of why it's a poor system, or that'd be circular. My solution to that issue is that arguments about which deductive system is correct are regular critical arguments, just as people usually use. But since deduction and induction are your only tools, you will have a harder time figuring out how to make arguments regarding deduction itself without circularity.

Elliot Temple | Permalink | Messages (0)

More Induction Discussion With Robert Spillane

Robert Spillane thought this was particularly important and requested a direct answer. Here it is:

1. Two simple answers to #1 and #2 will suffice - yes or no.

2. 1+1=2 is a necessary truth; '1 pint of water + 1 pint of alcohol = 2 pints of the mixture' is not. Can you not see the difference between the two?

They have many differences and many similarities.

By "the" difference, I guess you mean: that "1+1=2" is a "necessary truth", while the other statement isn't. I don't agree with that because I don't think anything is a necessary truth.

Regarding induction, I've asked several times about a set of instructions someone could follow to do induction. I've been unable to get answers which address basic issues like telling you which ideas to induce and how much inductive support they have. Here's another failure to address the issue, and my comments. This is extremely typical of inductivists. They don't have answers to these questions and wouldn't be inductivists if they understood the questions.

You asked me for details about Stove's Rationality of Induction. Here is a very brief summary (pp. 3-5, 22) which addresses your concerns:

(1) 'That all the many observed ravens have been black is not a completely conclusive reason to believe that all ravens are black' is true and not contingent, even though it mentions two propositions which are contingent:

(2) 'All the many observed ravens have been black.'


(3) 'All ravens are black.'

But (1) is not contingent since it is enough to entail the truth of (1) that it is logically possible that (2) be true and (3) false, whereas something's being logically possible is not enough to entail the truth of any contingent proposition. Therefore, (1), being true and not contingent, is a necessary truth.

Another way of saying (1) is:

(4) 'The inference from (2) to (3) is fallible' and this is also a necessary truth.

The inference from (2) to (3) is an inductive one. So there is at least one inductive inference of which it is necessarily true that it is fallible.

This doesn't answer my question about how (2) and (3) were selected from the infinity of propositions which do not contradict the observation data under consideration. Why those statements instead of some other statements?

I asked about which statements to induce and for instructions someone could follow to do induction, but this description doesn't provide instructions for how to select or create statements (2) and (3) in the first place.

What are the rules of induction? Could one write any statements at all in place of (2) and (3), or what? (I'm familiar with many proposed rules of induction, but none of them work. You apparently think you know of some rules of induction that do work, so I'm asking what they are.)

(5) 'That all the many observed ravens have been black is a reason to believe that all ravens are black' is like (1) in that it is true but not contingent. Like (1) it mentions two contingent propositions, but it does not assert either of them. Its truth, therefore, does not depend on what their truth values happen to be.

Another way of saying (5) is:

(6) 'The inference from (2) to (3) is rational' and this, also, is a necessary truth (pp. 3-5).

Since induction is necessarily fallible, the validity of induction is a subject easily exhausted. 'And as to the truth of the conclusion of an induction, or whether the conclusion of an induction with true premises is true, or whether more of such conclusions are true than are false: well, these of course are all contingent matters, with which philosophers have nothing to do. The success rate among inductions is as little the concern of philosophers as the blackness rate among ravens. Hume, in particular, was as little concerned as the next philosopher with what the long-run success rate of induction might be, and of course he said nothing about this subject; and a fortiori, he said nothing discouraging about it. Yet there are philosophers who do not shrink from the absurdity of implying that in order to 'answer' what Hume said about induction, we would need to establish something encouraging about the long-run success rate of induction. Some people just like to make rope neckties for themselves. But, in general, it is scarcely possible to exaggerate the harm that has been done to the philosophy of induction by philosophers who drift from the success of induction to the rationality of induction, and back again, and all over the place. Squalor rules, OK?' (p. 22).

Now, you will probably reply that this is irrelevant to your concerns since it assumes induction and engages in arguments for and against its rationality. You, on the other hand, insist that induction is a myth. If by 'myth' you mean 'the presentation of facts belonging to one category in the idioms appropriate to another' (Ryle), this means that you accept that there are inductive arguments - from the observed to the unobserved - but believe they are inevitably invalid because the conclusions are not contained within the premises.

But this is not your position. You claim that by 'induction is a myth' you mean that there are NO inductive arguments - that there cannot be (and never have been) arguments from the observed to the unobserved. This is a much stronger claim than 'inductive arguments are invalid'. It is also a claim that is so obviously false that further argument should be unnecessary.

My position that induction is a "myth", in the sense I've described (no one has ever induced anything), is from Popper. Do you know that's Popper's published view and know his reasoning? You are calling Popper's position "so obviously false that further argument should be unnecessary".

I (following Popper again – see e.g. his discussion of manifest truth) don't think that's a reasonable thing to say about anyone's position. The truth isn't obvious, and argument is necessary for dealing with disagreements.

Elliot Temple | Permalink | Messages (0)