1. The falsifiability criterion is about meaning.
2. Failure to draw the distinction between falsifiability (a matter of logic and the form of statements) and falsification (a practical matter).
3. Scientists don’t practice falsification.
4 Falsificationism is refuted by the history of science.
5 Popper was subjected to effective criticism by Lakatos/Kuhn/Feyerabend.
6. The failure of Popper’s theory of verisimilitude casts doubt on his whole program.
7. There is no getting away from induction/justificcationism.
8. From Habermas: Popperism is a form of positivism, it is analytical and provides no dialectic or effective theory of criticism.
9. From Habermas: The dualism of things and values, is/ought or propositions and proposals, provides no leverage for criticism of the status quo.
Rafe,
That’s a nice list. But let me give a sample of what I have in mind and see what you think. Maybe, if posted here, others may also provide criticisms that will be helpful.
My idea was not to put up a list of misrepresentations but a list of criticisms. The way I see it, a misrepresentation is not a criticism. It’s just a mistake. People might rather get tired of hearing for the millionth time that Popper’s criterion of falsifiability is not a criterion of meaning. A mistake like that would likely be corrected in the course of answering substantial criticisms anyway. Perhaps more importantly, a list of misrepresentations somewhat begs the question of who or what is right. That itself does not strike me as consistent with critical rationalism. I thought we could add value and improve critical rationalism by practicing severe criticism on our own philosophy.
What this means is taking a survey of the literature and assembling a list of criticisms, and then formulating them as sharply and crisply as we can, possibly revising if we can make the criticism tougher. As far as I can see, there can be only one sensible place to start, because only one critical rationalist has attempted such a daunting task and managed to assemble the results in a readable and helpful format. That would be David Miller’s first book, Critical Rationalism – A Restatement and Defense. Perhaps you know of some book or source where this has been attempted. If so, I should get it immediately. Anyway, David touched only some of the most outstanding criticisms. Great as that book is, there’s still much more to say.
So far I’ve learned a lot by taking a shot at this. This is perhaps not surprising, since it’s exactly what critical rationalism suggests one ought to do. Now I don’t say that these summary criticisms cannot be improved. I’m sure they can. So nothing is written in stone. Each criticism is extracted from a column in a spreadsheet where one may find verbatim quoted text, a specific reference to the work or article from which the criticism comes, and, where applicable, attempted solutions in the published literature. There’s more where this came from, but here’s a start:
1) Critical rationalism fails to tell us why we should believe or trust our sciences (Blackburn)
2) If we can hold a theory true come what may, then we can falsify it come what may as well. Critical rationalism fails to explain why we should not do this. (Caldwell)
3) Falsification tells us nothing about the future. Falsification alone is therefore insufficient. (Hesse)
4) The fundamental problem which Popper fails to answer is this: why is it rational to base one’s practical decision upon the best-tested theory, if there are no good reasons for expecting that it will be a successful choice?” Why is the best corroborated theory the best theory. (Niiniluoto)
5) Like verification, falsification too must ultimately depend upon justification. (Hubner)
6) Corroboration is either empty or it’s a form of induction. (Salmon)
7) Critical rationalism is like the admonition to live the good life. It’s vacuous. (Hands)
8) Probability theory is inductive logic, and can be applied with the assistance of Bayes’ Theorem. Thus Popper’s assertion that there can be no induction collapses. (Cox)
9) The Duhem-Quine thesis implies that falsification is either useless or impossible. (Cross)
10) Critical rationalism implies that there has been no growth of knowledge. (Stove)
11) Empirical investigation cannot reduce the number of false statements implied by our hypotheses. Critical rationalism does nothing to improve this situation. (Hattiangadi)
12) Induction cannot be eliminated because any coherent account of reality requires that we assume a stable order in the world. (O’Hear)
13) On the basis of logic and philosophy alone, it is not possible to say what the best method would be in all circumstances. (Feyerabend)
14) Background knowledge must be used inductively in determining the severity of a test. (O’Hear)
15) Falsifying a proposition is always equivalent to verifying its negation. So falsificationism is just verificationism turned on its head. Critical rationalism is nothing but an upside down way of talking, a jargon. (Gardner, et. al.)
So there’s a start. This list can easily be quadrupled in length. I’ve got several more almost ready for publication. This should be enough to get the idea across.
k
Rafe,
Here’s another interesting one that I just discovered:
16) Popper is inconsistent about methodological monism, because he says that the application of situational logic is the unique method of the social sciences. (Hands)
Figuring out what’s going on there with that is enlightening, because it requires, I think, a large perspective on Popper’s whole project.
Besides, many more standard criticisms, there are lots of mathematical criticisms of the Popper-Miller proof. I have a large, black binder full of these, and some of the responses are available to the general public only in the Popper papers at the Hoover Archive. Many of these criticisms and rejoinders, especially those by Miller, require a considerable understanding of probability theory. But after all, Popper and Miller too made large and important contributions in that area, and they are both very good at bringing that work to bear upon more general philosophical questions. The possibility of inductive inference is one area where the work in probability theory is very illuminating.
After working on this for a while I can only say that I have a new appreciation for the amount of work done by the author of any substantial book on the subject.
k
Repeating my comment on the Giddens post, I have two games going on here, or maybe three. What you are suggesting here is in (2).
(1) One is not intellectually challenging, it is a matter of putting on the record a moderately comprehensive record of the extent to which Popperism has been mangled and misrepresented by supposedly reputable scholars who have had their work accepted and printed by supposedly reputable publishers.
(2) This is the task you suggest, to identify serious criticism of CR that calls for an answer.
(3) This is a part of my work on Popper and the Austrians, where I want to look at alternative programs in the social sciences and see what difference it would make if they understood Popperism and CR better. Hence my interest in Giddens and also the program of Jeffrey C Alexander which will be my next post.
Kenneth,
Your list is very helpful, but it targets particular versions of CR, versions I do not think have stood up to criticism. Going down the list,
1. This criticism applies to all other alternatives: either they fail to provide good (positive) reasons or they beg the question. I do not think any good (positive) reasons exist, but there are good (negative) reasons for not believing or trusting other particular modes of inquiry.
2. I do not think I understand this objection fully, but I will attempt to give a response. Since any theory, no matter how convoluted, could be true, I do not see truth as a guiding principle in philosophy of science other than as a regulatory principle — we want theories to have a great deal of verisimilitude. But there is no way to measure it, so it is not too important. However, if we focus on empirical adequacy, explanatory content, or exactitude in predictions, preference between theories becomes far easier. I see no compelling reason to intentionally make our background assumptions or scientific theories contradict empirical evidence, only that we seek out contradictions.
3. This problem is the same as the first: we can know nothing about the future, in the sense of having something grounded, justified, or warranted. Any program that claims that it can provide grounds, justification, or warrant claims to possess something that it does not have; any other program that recognizes the hand that we’ve been dealt is being realistic.
4. I think Popper was mistaken on this point, and van Fraassen and Miller give a much stronger negative formulation: we must be epistemic voluntarists, and reject scientific theories that do not stand up to criticism. All theories that remain, even those that use ill-behaved predicates, are permissible, in that they are not ruled out by the evidence, but are not chosen because they do not cohere with background assumptions (they are ruled out by conflicting with implicit criticisms or assumptions).
5. I don’t understand the strength behind this objection in light of Bartley and Miller. Could you elaborate?
6. This objection amounts to little — years ago, for instance, I came to the same conclusion and rejected any notions of corroboration as providing an objective probability assessment. So much for corroboration.
7. Could you elaborate further on what Hands says? For instance, there are actions that are either permitted or prohibited under CR, but few that are positively obligated. Is this lack of positive obligations the problem Hands sees?
8. I do not see this objection as holding much water. Even a few proud Bayesians acknowledge that due to its epistemic or subjective interpretation of the probability calculus, it does little but describe changes in assigned probabilities. In other words, the Bayesian Newtonian is still confounded by Eddington in 1919.
9. In his later years Quine backed away from his stronger formulation in a letter to Grünbaum. I think the objection raised by Grünbaum that it was vacuous stands. Yes, it is possible to make any number of adjustments to scientific theories in order to save them from refutation, but there reaches a point where one either speaks of entirely different scientific theories with different predictions as before, or have so many ceretis paribus clauses that alternative theories are of more use.
10. I don’t think Stove’s objection stands unless one assumes that replacement scientific theories must increase in objective verisimilitude most of the time and be known to increase in its objective verisimilitude over its predecessors. The second is impossible; the first happens some of the time. But replacement scientific theories, more often than not, are not known to be empirically inadequate while their predecessors were known to be empirically inadequate, and that is sufficient according to other accounts.
11. If the first is true, then nothing can improve this situation.
12. Induction is not the same as assuming a particular distribution of properties. The first takes empirical evidence as grounding such an assumption, but this is false; the second is — admittedly — a conjecture.
13. Feyerabend’s objection rings true if one takes the aim of science as that of approaching truth. But that is not necessary in philosophy of science. If we seek only coherence of a system + searching for failures in the system, it may lead to falsehood, but it is a method that is, at least so far, preferable to alternatives if empirical adequacy is taken to be the central goal in science. It may be the best of a bad lot, but such a possibility does not undermine it if one never argued that the method was best in the first place.
14. I think O’Hear is using the term ‘induction’ in a way that does not fall in line with its traditional usage. I, for one, see no difference in the sentence and its modification “Background knowledge must be used in determining the severity of a test.” How do the two differ?
15. We learn that “Not all ravens are black” is true if “There is a raven and it is not black” is true; we do not learn that “All ravens are black” is true if “There is a raven and it is black” is true. The first is basic modus tollens, the second is modus ponens. Therefore, the two differ.
d,
I hope I did not create the impression that I thought the criticisms were especially worthy. That was not my intention. I am only trying to catalog them at the moment. Nevertheless, I am baffled by some of your responses. In the last one, 15, for example … Your explication of the criticism is incorrect. Nobody suggests that you learn ‘All ravens are black’.
Suppose you have ‘All ravens are black’. You encounter a non-black raven. Everything is acceptable, so you falsify the theory and conclude that not all ravens are black.
Suppose you have ‘There is a non-black raven’. You encounter a non-black raven. Everything is acceptable, so you verify the theory and conclude that not all ravens are black.
You learn exactly the same thing in either case. That’s the whole point of the criticism: there’s nothing you can do with falsification that you cannot do just as well with verification. In their result, the two do not differ. Popper even says in LSD that the logic is symmetrical. So what’s all this malarkey about falsification? There is a satisfactory answer, I believe, but it isn’t that the two differ logically. They don’t.
Of course, there are many more criticisms than the ones I’ve listed so far. I am not of the opinion that we can pop through a list of “misrepresentations”, read off a one or two sentence answer, and end up with anything useful.
d,
To follow through just a little more .. the critics conclude, just by examining the symmetrical logic, which Popper himself admits, that Critical Rationalism is a big con game. It’s just an upside down way of talking. In the Critical Cafe, Larry Tapper calls it Popperese. The contention is that critical rationalists have been duped by Popper’s linguistic flim flam. The peanut gallery finds this whole thing delightful, because it reveals that CR is a lot of hot air. So .. if you don’t defeat this criticism utterly, you’re in trouble.
k
Kenneth,
I would disagree — there is the idea that or one should assign a high probably to the statement ‘all ravens are black’, or it is rational to believe that ‘all ravens are black’, and so on. But this does not work, because the inference is ampliative.
The only thing one can learn from empirical evidence is the result of a deductive inference: a strictly universal statements + background assumptions conflict with the outcome of a test.
I don’t see how that’s linguistic flim-flam.
d,
I don’t agree either. But how do you answer the criticism that inspires this conclusion. You’re not addressing that.
k
Ken:
>To follow through just a little more .. the critics conclude, just by examining the symmetrical logic, which Popper himself admits, that Critical Rationalism is a big con game. It’s just an upside down way of talking.
Surely the first answer to that criticism is that theirs is the “upside down” way of talking. They’ve obviously verified an existential statement, not a universal law. Is “There is a non-black raven” their definition of an interesting scientific theory? I suppose if you want to define science as the quest to verify statements such as “there is a green tree” and “there is a non-green tree” or “it is raining” and “it is not raining” then good luck and God speed. Clearly Hume’s problem does not apply either, so no ticking time bomb under this vision of science either. Brilliant – science as, almost literally, Trivial Pursuit. However, if these critics ever tire of this and feel like getting back in the game of searching for the laws of the universe, they can always get back in touch.
Daniel,
Nice idea, and I wish that was right. That would certainly make this criticism easy to knock down. But I’m afraid that it is a little more tenacious than that. The thing you have to keep in mind is that it doesn’t matter if ‘There is a non-black raven’ seems trivial compared to ‘All ravens are black’. The point is that, whether you talk one way or the other, falsifying ‘All ravens are black’ is equivalent to verifying its negation. So, talk in whatever terms you want. Consider ‘All ravens are black’ your law, if that’s what you prefer. It remains the case that, when you test it, you can replace the law with its negation and verify its negation. So it is quite true that there’s nothing you can do with falsification that you cannot do just as well with verification. That much is undeniable. This is actually not a verbal issue, though the critics want to claim that critical rationalists are just playing word games. In other words, the critics DO treat this as a verbal issue. That’s really part of their criticism. They’re saying .. look! .. you critical rationalists think you’ve got something but you’ve really got nothing. You’re just playing word games. Martin Gardner made a big issue out of this, and really became very nasty, insulting and contemptuous about it. So I feel a tiny bit guilty about this because I am playing devil’s advocate here. The jokes on him, though .. because it is really a cool problem. It’s like Fermat’s Last Theorem in a childish sort of way .. really easy to state but damn frustrating to solve. Thinking about it for the last several years has been very helpful for me, because when it starts unraveling, it changes your conception of critical rationalism. I think I’ve gotten a lot out of that, and I don’t want to spoil it for anyone else. Popper was right .. focusing on criticizing your own philosophy really is the way to go.
As a little aside here, I’d like to put in a plug here for Lawrence Boland’s book, Critical Economic Methodology. I am just re-reading parts of it. Lawrence highlights in that book some of what I’m getting at about changing your conception of critical rationalism. He talks about the Socratic Popper as opposed to the Popper who is seen largely through the reinterpretation given by Imre Lakatos. Though I’m not down on the term ‘falsificationism’ like he is, I understand what he’s getting at when he rejects it. It can easily give the impression of a non-critical kind of critical rationalism. Really, critical rationalists should be their own biggest critics, if they’re going to practice what they preach. Critical rationalists should, among themselves, criticize to an extent that would make other critics pale in comparison. That would do more than almost anything else to make critical rationalism resilient and easily able to handle what any other critic has to say.
Daniel,
In fact, you ARE verifying the negation of your hypothesis when you falsify it. Logically, the distinction between verification and falsification is illusory. There’s no difference. So … that means ……… 😉
k
OK .. no takers. I’ll say it in one sentence: don’t confuse verification with verificationism.
Does anyone else see this? .. cuz I really want someone else to explain it.
Popper said in his book Open Society & its Enemies Ed. 2 Page 13 “‘In so far as scientific statements refer to the world of experience, they must be refutable … in so far as they are irrefutable, they do not refer to the world of experience'” .
In so far as such statement is concerned, isn’t it self-defeating to assume that “ALL” scientific statements are falsifiable while at the same time to admit that Singular Existential propositions are only verifiable?? After all, our corpus of scientific knowledge requires to have at least some of such existential propositions (like statements that are logically implied by the universal conjectures)??
Kenneth – I’m new to the blog, and therefore may very well be about to demonstrate my extreme ignorance, but here goes:
It seems to me that the linguistic flim-flam is on the side of those suggesting that the equivalence to which they point is toxic to falsificationism. It strikes me as plain obvious that the refutation of a theory *must* entail the verification of its negation. Isn’t that precisely what a testable theory *means*? It also seems rather plain that the opposite is not true (the equivalence doesn’t work both ways): I can’t refute the theory “there is a non-black raven” by confirming its negation (i.e. — “all ravens are black”), for the obvious reasons….
If I’m entirely missing some sophisticated nuance to their critique of CR, I apologize for wasting everyone’s time 🙂
Parker Whittle
To answer Gardner’s criticism:
Negating “All swans are black” gives you “There is a swan that is not black”, which is completely different sort of statement. That is, it’s not a universal statement; it’s a singular existential statement. That said, verification, in this case, is about verifying singular existential statements, which is far less than what verificationism is – verification of universal statements through verification of singular existential statements.
Falsificationism, on the other hand, is about refutation of universal statements through verification of singular existential statements.
Perhaps it might be better to say that falsificationism is about refutation of universal statements by possibly true singular existential statements. Popper would not hesitate to state that not even falsification can claim to be certain as the basic statements may not be actually true.
The asymmetry is that if the basic statements are true they may falsify a universal law whereas they could never verify a universal law.