[Previous] Explaining Popper on Fallible Scientific Knowledge | Home | [Next] Comments on An Eye for An I: Philosophies of Personal Power

Reply to David Stove on Popper

Popper and After: Four Modern Irrationalists, by David Stove criticizes Karl Popper's philosophy of knowledge.

But Stove's criticism doesn't focus on epistemology.

And Stove writes insults and other unserious statements. These are frequent and severe enough to stand out compared to other similar books. I give examples.

The book's organization is problematic as a criticism of Popper because it criticizes four authors at once. It only focuses on Popper for a few paragraphs at a time. It doesn't lay out Popper's position in detail with quotes and explanations of what problems Popper is trying to solve and how his ideas solve them.

First I discuss the book's approach and style. Then I address what I've identified as Stove's most important criticisms of Popperian philosophy.

My basic conclusion is that Stove doesn't understand Popper. His main criticisms amount to, "I don't understand it." Popper contradicts established philosophy ideas and some common sense; Stove doesn't know why and responds with ridicule. Stove is unable to present Popper's main ideas correctly (and doesn't really try, preferring instead to jump into details). And without a big-picture understanding of Popper, Stove doesn't know what to make of various detail statements.

Stove's Focus

Part 2, Ch. 3 begins:

Popper, Kuhn, Lakatos and Feyerabend have succeeded in making irrationalist philosophy of science acceptable to many readers who would reject it out of hand if it were presented to them without equivocation and consistently. It was thus that the question arose to which the first Part of this book was addressed: namely, how did they achieve this? My answer was, that they did so principally by means of two literary devices discussed in Part One. The question to which the present Part of this book is addressed is: how was irrationalist philosophy of science made acceptable to these authors themselves?

Stove says the first part discusses how Popper achieved influence. How did Popper convince readers? What literary devices did Popper use to fool people? And part two (of two) discusses the psychological issue of how Popper made irrationalism acceptable to himself.

By Stove's own account, he's not focusing on debating philosophy points. He does include epistemology arguments, but they aren't primary.

The problem Stove is trying to solve plays a major role in his thinking (as Popper would have said). And it's the wrong problem because it assumes Popper is an irrationalist and then analyzes implications, rather than focusing on analyzing epistemology. If Popper's philosophy is true, Stove's main topics don't matter.


Ch. 2:

It is just as well that Popper introduced this [methodological] rule. Otherwise we might have gone on indefinitely just neglecting extreme probabilities in our old bad way: that is, without his permission.

This is unserious and insulting. Popper's purpose was to discuss how to think well, not to give orders or permission.

To readers in whom the critical faculty is not entirely extinct, the episode has afforded a certain amount of hilarity.

This is mean.

I point out more examples of Stove's style as they come up.

Neutralizing Success Words

Ch. 1 discusses neutralizing success words. A success word like "knowledge" or "proof" implies an accomplishment. Compare "refuted" (a successful argument) to "denied" or "contradicted" (doesn't imply the denial has merit). Neutralizing knowledge yields idea – knowledge means a good idea, whereas an idea could be good or bad. Neutralizing proof yields argument – a proof is a type of successful argument, whereas a mere argument may not succeed.

Stove says Popper equivocates. Often, Popper uses success words with their normal meaning. But other times Popper changes the meaning.

It is the word "knowledge", however, which was the target of Popper's most remarkable feat of neutralization. This word bulks large in his philosophy of science (much larger than "discovery"), and in recent years, in particular, the phrase "the growth of knowledge" has been a favorite with him and with those he has influenced most. Some people have professed to find a difficulty, indeed, in understanding how there can be a growth-of-knowledge and yet no accumulation-of-knowledge.

There is accumulation-of-knowledge. Stove gives no cite, but I have a guess at what he's talking about. This quote is from C&R (Conjectures and Refutations) ch. 10 sec. 1, and there's a similar statement in LScD (The Logic of Scientific Discovery).

it is not the accumulation of observations which I have in mind when I speak of the growth of scientific knowledge, but the repeated overthrow of scientific theories and their replacement by better or more satisfactory ones.

The growth of knowledge doesn't consist of accumulating ever more observations (we need ideas). Nor are we simply accumulating more and more ideas, because scientific progress involves refuting, replacing and modifying ideas too. The growth of knowledge is more about quality than quantity.

Continuing the same Stove passage:

But then some people cannot or will not understand the simplest thing,

More ridicule.

and we cannot afford to pause over them. Let us just ask, how does Popper use the word "knowledge"?

Well, often enough, of course, like everyone else including our other authors, he uses it with its normal success-grammar. But when he wishes to give expression to his own philosophy of science he baldly neutralizes it. Scientific knowledge, he then tells us, is "conjectural knowledge". Nor is this shocking phrase a mere slip of the pen, which is what anywhere else it would be thought to be.

Expressing shock and talking about slips of the pen is not how one debates ideas seriously. But let's discuss conjectural knowledge.

Knowledge is good ideas. Sorting out good and bad ideas is one of the main problems in epistemology.

Conjectural serves two purposes. First, it indicates that knowledge is fallible (and lacks authority). Popper doesn't mean justified, true belief. He's not looking for perfect certainty or absolute guarantees against error.

Second, conjecture is the original source of the good ideas that constitute knowledge. Conjecture is, intentionally, an informal, tolerant, inclusive source. Even myths and superstitions can qualify as conjectures. There's no quality filter.

I think Stove's negative reaction has a thought process like this: No quality filter!? But we want good ideas. We need a quality filter or it's all just arbitrary! "Anything goes" can't achieve knowledge, it's irrationalism.

Popper has an answer:

Standard approaches do lots of quality filtering (sometimes all) based on the source of ideas.

Instead, all quality filtering should be done based on the content of ideas. This is done with criticism and human judgement, which lack authority but are good enough.

So we do have a quality filter, it's just designed differently and put in a different place.

For more, see Popper's introduction to C&R, On the Sources of Knowledge and of Ignorance. Excerpt from sec. XV:

The question about the sources of our knowledge can be replaced in a similar way [to the 'Who should rule?' issue]. It has always been asked in the spirit of: ‘What are the best sources of our knowledge—the most reliable ones, those which will not lead us into error, and those to which we can and must turn, in case of doubt, as the last court of appeal?’ I propose to assume, instead, that no such ideal sources exist—no more than ideal rulers—and that *all* ‘sources’ are liable to lead us into error at times. And I propose to replace, therefore, the question of the sources of our knowledge by the entirely different question: ‘*How can we hope to detect and eliminate error?*’

Continuing the same Stove passage:

On the contrary, no phrase is more central to Popper's philosophy of science, or more insisted upon by him. The phrase even furnishes, he believes, and as the title of one of his articles claims, nothing less than the "solution to the problem of induction" [28].

Note the lack of discussion of Popper's position.

In one way this is true, and must be true, because any problem clearly must yield before some one who is prepared to treat language in the way Popper does. What problem could there be so hard as not to dissolve in a sufficiently strong solution of nonsense? And nonsense is what the phrase "conjectural knowledge" is:

More insults.

just like say, the phrase "a drawn game which was won". To say that something is known, or is an object of knowledge, implies that it is true, and known to be true.

This is ambiguous on the key issue of fallibility.

Is Stove saying all knowledge must be infallible and known to be infallible? It must be the proven to be the perfect truth, with complete certainty, so that error is utterly impossible – or else it's not knowledge?

If that's Stove's view of knowledge, then I think he has a choice between irrationalism or skepticism. Because his demands cannot be met rationally.

Or if Stove's position is less perfectionist, then what is it? What allowances are made for fallibility and human limitations? How do they compare to Popper's allowances? And why is Popper mistaken?

(Of course only `knowledge that' is in question here). To say of something that it is conjectural, on the other hand, implies that it is not known to be true.

Does "known to be true" here mean infallibly proven? Or what?

And this is all that needs to be said on the celebrated subject of "conjectural knowledge"; and is a great deal more than should need to be said.

What's going on here is simple. Stove is scornful of a concept he doesn't understand. He doesn't appreciate or discuss the problems in the field. And he doesn't want to. He's unable to state a summary of Popper's view which a Popperian would agree with, and he wants the matter to be closed after three paragraphs.

Sabotaging Logical Expressions

Ch. 2:

What scientists do in such circumstances, Popper says, is to act on a methodological convention to neglect extreme probabilities

For example, how do you know a coin which flips 1000 heads in a row is unfair? Maybe it's a fair coin on a lucky streak.

Well, so what? I'm willing to risk a 2^-1000 chance of misjudging the coin. I'm far more likely to be struck be lightning than get the coin wrong. And the downside of misjudging the coin is small. If the downside were so large that I couldn't tolerate that much risk, I could flip the coin additional times to reduce the risk to my satisfaction (assuming I get more heads, that reduces the probability it's a fair coin).

So Popper offers: if you judge it's not a worthwhile issue to worry about, then don't worry about it. This judgement, like everything, could be a mistake, so it's always held open to criticism. That openness doesn't mean we think it's mistaken or spend our time searching for a mistake, it just means we recognize we have no infallible guarantee against error. We have to make fallible, criticizable judgements about what areas are problematic to focus attention on.

Stove dislikes this approach because he thinks you could do it to dismiss any problem. Stove fears arbitrarily creating a methodological convention to neglect any difficulty. The solution to this is criticizing bad methodological conventions. Stove (correctly) sees problems with some conventions that could be proposed, and those problems can be expressed as criticism.

The problem here is Stove's unfamiliarity with Popperian methods. Plus I think Stove wants methodological rules to guide thinking and reduce the scope for human judgement and creativity.

... Popper actually anticipated it. This is `the Quine-Duhem thesis': that "any statement can be held true come what may, if we make drastic enough adjustments elsewhere in the system [...]. Conversely, [...] no statement is immune to revision" [23].

There's an important logical point here. I wonder what Stove's answer to it is (he doesn't say). Popper offered some help with this issue, but not a full solution. That's OK because Popper's general approach of fallible judgement combined with error correction still works anyway.

Philosopher David Deutsch addressed the Quine-Duhem issue better. His two books offer refinements of Popper. (FoR ch. 1, 3, 7-8; BoI ch. 1-4, 10, 13.)

In short: You may try modifying whatever you want to rescue a statement, but those modifications have meaning and can be criticized. Ad hoc modifications commonly ruin the explanation which gave the idea value in the first place, or contradict vast amounts of existing knowledge without argument. If you can come up with a modification that survives immediate criticism, then it's a good contribution to the discussion (sometimes the error really is elsewhere in the system).

Other Thoughts

Ch. 3:

It is a favorite thesis with him that a scientific theory is, not only never certain, but never even probable, in relation to the evidence for it [3].

Right, because logically there's no such thing as evidence for a theory. There's only evidence which does or doesn't contradict a theory. And any finite set of evidence is logically compatible with (does not contradict) infinitely many theories, and those theories reach basically every conclusion.

What does Stove think of this?

These two theses [the one above and one other] will be acknowledged to be irrationalist enough; and they are ones upon which Popper repeatedly insists.

Stove doesn't present and discuss Popper's solution to the logical difficulties of positive support. Nor does Stove present his own solution. Instead he says it "will be acknowledged" that Popper's view is irrational, without argument. Stove treats it as if Popper only talked about this difficulty without also giving a solution. (The solution, in short, is that negative arguments don't face this difficulty.)

Ch. 3:

Scepticism about induction is an irrationalist thesis itself

Rather than present and discuss Popper's solution to the problem of induction, Stove simply asserts that the only alternative to induction is irrationalism. He goes on to discuss Hume at length rather than Popper.

Ch. 5:

One of these features, and one which is at first sight surprising in deductivists, is this: an extreme lack of rigor in matters of deductive logic.

Because Popper's main positions aren't about deduction. The technical reason that conjectures and refutations is able to create knowledge is that it's evolution, not deduction. The key to evolution is error correction, and that's also the key to Popper's philosophy, but Stove doesn't understand or discuss that. Stove only uses the word "evolution" once (in a Kuhn quote where it means gradual development rather than replication with variation and selection).

A core issue in Popper's philosophy is: "How can we hope to detect and eliminate error?" (as quoted earlier). Stove doesn't understand, present, or criticize Popper's answer to that question.

Note: My comments on Popperian thinking are summary material. There's more complexity. It's a big topic. There are books of details, and I can expand on particular points of interest if people ask questions.

Elliot Temple on July 4, 2017

Comments (1)

It's amazing how foolish one can appear when one judges something without understanding it. Stove illustrates this beautifully.

Larry at 8:53 AM on July 16, 2017 | #8834

What do you think?

(This is a free speech zone!)