Talk:Paul E. Meehl

GA criteria

The article is currently quick fail eligible due to verifiability issues. For instance, there are statements which the source isn't clear and other information is cited to very large, 80+ page ranges which makes it nearly impossible to verify the information. buidhe 05:12, 29 March 2020 (UTC)

GA Review

This review is transcluded from Talk:Paul E. Meehl/GA1. The edit link for this section can be used to add comments to the review.

Reviewer: Buidhe (talk · contribs) 05:32, 29 March 2020 (UTC)

The article is currently quick fail eligible due to verifiability issues. For instance, there are statements which the source isn't clear and other information is cited to very large, 80+ page ranges which makes it nearly impossible to verify the information. If the issues are not fixed, the article will have to be failed. buidhe 05:32, 29 March 2020 (UTC)

Since it's now been a week without a response from the nominator, I'm going to close the review. Anyone is of course welcome to fix the issues discussed above and renominate. buidhe 18:33, 4 April 2020 (UTC)

GA again?

There's a lot of action on this page, which is great to see. Any interest in moving it to GA status? See immediately above on why it quickly failed the last time I nominated it. Vrie0006 (talk) 19:55, 26 January 2022 (UTC)

I noticed that the article still needs work to address the verifiability issues. I didn't bother to tag all the spots where citations are needed, but it definitely needs more citations (and more precise page ranges for some existing citations) before it's ready for a GA nomination. I think the article is currently properly classified as B-class except perhaps for the referencing. Biogeographist (talk) 20:40, 26 January 2022 (UTC)

Meehl and Popper

This edit suggests that the relation between Meehl and Popper is important. Certainly, Meehl found it important. However, I feel it is not adequate to enter into this subject in a superficial manner. It is a complex subject that deserves a lot of research by us, the editors of Wikipedia. Of course, we must rely on reliable sources, but especially in that case, we cannot just pick one source and certainly not focus on selected excerpts in one or even more sources. If this subject is covered, we must make sure the readers have the big picture and we must do some work to cover well the subject.

It's clear that Meehl is responding to accusations or rumours that he was a follower of Popper. For example in this "interview of Meehl"., at 23:30, he says (emphasis mine) "I am not, contrary to rumours, a disciple of Popper". This rumor or accusation in itself is weird, because, as far as I can tell, Popper did not have any follower, certainly not a disciple, among philosophers. Certainly, none of his students would qualify as a follower. Yet all of them admit the value of Popper's philosophy. Kuhn for example emphasized how much he agreed on Popper on many points. In the same way. in the same interview, Meehl recommends the reading of LoSD. So, what concretely is the rumor to which Meehl feels he must respond? It cannot be just that he appreciates Popper's philosophy, because he admits it—he would not refer to this as a rumour. Can we find out what it was or should we infer it from Meehl's reply only? This kind of context is important. Also, if we cover the subject, we must make sure that the big picture is covered. In particular, on what major points they agreed and disagreed should be covered. I don't think it is appropriate to only say "Meehl said that he was not a Popperian". It sounds a bit like superficial gossiping. It also creates a form of undue weight by omission of the context, i.e., omission of other information that would relativize the meaning of this statement. Dominic Mayers (talk) 23:51, 27 January 2022 (UTC)

It is more than "superficial" rumor. There is plenty of evidence that Meehl was influenced by Popper but also differed from him in substantive ways. Some relevant quotations (references are to the works listed in the article, and see the rest of the quoted works for more detail):

• "There are revisions (as I think, necessary) of the classic Popperian position urged on us by his heretical exstudents P. K. Feyerabend and the late Imre Lakatos, but psychologists must reach at least the stage of Bayes and Popper before they can profitably go on to the refinements and criticisms of these gentlemen." (1978, in Waller et al. 2006, p. 73)
• "I am not persuaded from his writings nor from conversations that I have had with him that Sir Karl adequately appreciates the degree to which this theory and auxiliary problem permeate research in the inexact sciences, especially the social sciences in their soft areas. Whether it presents a general problem for the Popperian formulation of scientific method is beyond the scope of this article and my competence. It is perhaps worth saying, however, for the benefit of philosophically oriented readers, that the above described situation—certainly no rarity in our field or in biology—may represent a social fact about the way science works that presents grave difficulties for the Popperian reconstruction." (1978, in Waller et al. 2006, p. 76)
• "I am incidentally replying to Serlin and Lapsley (1985), who advanced discussion of the issue by correcting my overly Popperian stance ('strict falsificationism') and pointing out that it is more realistic to think of theories as being 'good enough [even if, literally, false]' than to set up a rigid true/false dichotomy in the way I did in 1967 and 1978. I cheerfully accept their criticism, as well as their 'good enough' principle, although I am not convinced that their specific statistical implementation of the principle is as helpful as they think. ... As to Serlin and Lapsley's complaint that, although I cited him in my 1978 article, I did not integrate his views with my neo-Popperian critique of significance testing, the reasons for that were (a) space and (b) doubts as to whether I could do it. I think now that I can, but I'm not sure. Moving from Popper to Lakatos does not appreciably soften the blow of my 1967 attack, and here I shall try to show that a proper interpretation of Serlin and Lapsley's "good enough" principle must rely on two other principles, both Popperian in spirit although not 'orthodox Popper.'" (1990, in Waller et al. 2006, p. 92)
• "Because not all psychologists subscribe to a Popperian or Lakatosian metatheory, I must emphasize that one need not subscribe to Popper's anti-inductivism, nor to his emphasis on falsification, to accept the notion of risky test, perhaps expressed in other, less committed language. Working scientists who never heard of Popper, and who have no interest in philosophy of science, have for at least three centuries adopted the position that a theory predicting observations 'in detail,' 'very specifically,' or 'very precisely' gains plausibility from its ability to do this. I have not met any scientist, in any field, who didn't think this way, whether or not he had ever heard of Karl Popper. If my meteorological theory successfully predicts that it will rain sometime next April, and that prediction pans out, the scientific community will not be much impressed. If my theory enables me to correctly predict which of 5 days in April it rains, they will be more impressed. And if I predict how many millimeters of rainfall there will be on each of these 5 days, they will begin to take my theory very seriously indeed. That is just scientific common sense, part of the post-Galilean empirical tradition that does not hinge on being a disciple of Popper or Lakatos." (1990, in Waller et al. 2006, p. 95)
• "How does a theory get money in the bank—how does it earn an impressive track record? We rely on the basic epistemological principle that 'If your aim is a causal understanding of the world, do not attribute orderliness to a damn strange coincidence.' We could label this 'Reichenbach's maxim,' because in his famous justification of the straight rule of induction he says that, although we can have no guarantee it will work, it will work if anything works. Or we might label it 'Novalis's maxim,' remembering the epigraph of Popper's great 1935 book, quoted from Novalis, 'Theories are nets: Only he who casts will catch.' We apply this maxim to formulate Salmon's principle: that the way a theory gets money in the bank is by predicting observations that, absent the theory, would constitute damn strange coincidences. I don't label this 'Popper's principle,' because accepting the Serlin-Lapsley critique of my overly Popperian earlier statements, I am here emphasizing that a theory can get a lot of money in the bank, and hence warrant us in conducting a Lakatosian defense, despite its being falsified. It does this by achieving a mixture of risky successes (passing strong Popperian tests) and near-misses, either of these being Salmonian damn strange coincidences." (1990, in Waller et al. 2006, pp. 119–120)
• "This is similar to Popper's original emphasis on corroboration being a function of risk, except that here again it is not yes-or-no falsification but Salmon's principle that we wish to numerify. The revised methodology retains the Popperian emphasis on riskiness, but now instead of asking 'Did I pass the test, which was stiff?' we ask, 'How close did I come?'" (1990, in Waller et al. 2006, p. 133)
• "Theories are inscription products of the human mind, having a physical and psychological existence in Popper's Worlds I and II (I do not understand his World III, so I say nothing about it)." (1990, in Waller et al. 2006, p. 142)
• "I have tried to provide a reformulation of Serlin and Lapsley's (1985) 'good enough' principle that preserves the Popperian emphasis on strong corroboration. Accepting their criticism of my overly strict Popperian formulations, and moving from Popper to Lakatos as a metatheoretical guide, we ask not,'Is the theory literally true?' but instead,'Does the theory have sufficient verisimilitude to warrant our continuing to test it and amend it?' This revised appraisal in terms of verisimilitude rather than strict truth leads to adopting a strategy of Lakatosian defense by strategic retreat, provided the ad hockery is 'honest' at all stages (i.e., not ad hoc in any of Lakatos's three senses). The warrant for conducting a Lakatosian defense is the theory's track record. A good track record consists of successful and almost-successful risky predictions, of 'hits' and 'near misses' for point or interval predictions of low tolerance, and predictions of function forms. It is crucial in my argument that this low tolerance is not best judged by traditional significance testing, whether of the strong or weak kind, or even by confidence-interval estimation, but by comparing the theory's intolerance, and the nearness of the 'miss' when there is a miss, with a reasonable a priori range of possible values, the antecedent Spielraum. Whether my specific proposals for quantitative indexes of corroboration are acceptable is not the main point. The big qualitative point is Salmon's principle." (1990, in Waller et al. 2006, p. 152)
• "As to Humphreys's preference for having good data before embarking on theories, here is one of several places that I am not strongly Popperian, as I agree with Humphreys. But in agreeing I mean good data, not necessarily a lot of good data. Small amounts of good data, especially if qualitatively diverse, suffice to warrant embarking on bold conjectures." (1990, in Waller et al. 2006, p. 155)
• "Lakatos's amendments aside, I have never accepted the original Popper doctrine that antecedently improbable theories are to be preferred. Here, at least, I have always been a Bayesian. The big puzzle here lies in the difference between the theory's prior probability (which, like the Bayesians and the nonphilosophical working scientist, I prefer to be high) and the prior (absent theory) predictions, which I prefer to be low. I believe the logician's ready identification of content with consequence class is what causes the trouble. Someone more competent than I will have to fix that up. But one reason why I prefer the theory-properties list in my Figure 5.1 as an approach to comparing two theories' contents is that it avoids the consequence-class business, which is what killed Popper's attempt to define verisimilitude. I am more concerned with Lakatos's acceptability₃, as they say." (1990, in Waller et al. 2006, p. 159)
• "I can show, in my work, where I was helped by Feigl, Carnap, Reichenbach, Salmon, Popper, Hempel, up to 2 dozen I recently listed (when wife challenged my "2 dozen"). Lakatos helps re diachronic emphasis, but not much otherwise. Feyerabend is a brilliant nut. Kuhn never helped me one bit in thinking about psy matters, and his current effect in 'soft' areas is positively cancerous. People now defend their theories by invoking his name (in ways I'm pretty sure he would dislike—he says early in his book that he doubts social science even has a paradigm yet!) The important revisions of logical empiricist line came mostly from themselves, such examples as Carnap, Feigl, Hempel. ... Even my post-Lakatos theory testing formula is due to Feigl (center period 1950+) as much as Lakatos. And Lakatos doesn't stress a big Feigl (anti-Kuhn, anti-Quine, anti-Duhem) point, stressed recently by physicist Franklin, that we can usually test the auxiliaries separately, i.e., without regard to theory T." (1991, in Peterson 2005, p. 11)
• "Much of the positivist-bashing distorts what the Vienna Circle said or takes a single mistake by a single member—even if corrected shortly thereafter—as a group dogma. The worst mistake was Schlick's, 'The meaning of a sentence is the method of its verification,' held briefly by some of the group. Similar errors were made by a few Americans, but under the influence of Bridgman's operationism, quite independent of Vienna. Herbert Feigl was the one who in 1924 first urged leader Schlick to hold regular, organized meetings. He co-authored the first paper in English on the Vienna position (Blumberg & Feigl, 1931) five years before Ayer's (1936) bombshell book Language, Truth, and Logic. He invented the term 'logical positivism' and was the leading expositor of the doctrine in the United States. He came to Minnesota in 1940 and was my teacher, then co-author and co-instructor, and personal friend for 30 years. I know what the positivists believed, where they disagreed among themselves, and how their views evolved. Much of what is attributed to them by the positivist-bashers is simply incorrect. When psychologists in the 'soft' areas write about their liberation by Kuhn and Co., I want to inquire of them 'What was the ingenious experiment you devised but never performed because of the wicked positivists? What is the exciting theory you conceived but feared to publish lest the positivists make fun of you?' I have done theoretical and empirical work in several areas of psychology, including latent learning, MMPI scale construction and profile analysis, psychometric theory, clinical prediction, interview assessment, political behavior, forensic psychology, theory of schizophrenia, behavior genetics, and taxometric methods. All of these topics present complex conceptual problems involving what Arthur Pap (1953, 1958, Chapter 11, 1962, Chapter 3) called 'open concepts,' for which any strict operational definition is a fake. But I have never felt the least bit constrained, cramped, or inhibited by the logical positivists or their intellectual descendants. On the contrary, my thinking about these deep and difficult matters has been facilitated by my reading and discussion with such first-class intellects as Feigl, Carnap, and Hempel. I say the same for Reichenbach who in his classic 1938 Experience and Prediction uses the term 'positivists' to denote his Viennese opponents—hence the positivist-bashers who label him a positivist thereby reveal that they have not even read his book! I have also learned much from anti-positivist Popper and his erstwhile disciples Feyerabend and Lakatos. The way to treat philosophers of science is to pick and choose, take what you can use, and in the process acquire philosophical sophistication from their disputes." (1993, in Waller et al. 2006, pp. 415–416)
• "While severity of tests is associated with the views of Sir Karl Popper, inductivists like Wesley Salmon have a place for it also from a different perspective, one in which Bayes' Theorem plays a part and the successful prediction would constitute a 'damn strange coincidence' if the theory had no verisimilitude. It is obvious from the history of science that working scientists put a great weight upon this matter of riskiness or strange coincidence even if they do not engage in formal philosophy of science (Meehl, 1978, 1990a, 1990e)." (1993, in Waller et al. 2006, p. 419)
• "Conceiving metatheory as the empirical theory of scientific theorizing, we know that formal logic, probability theory, and basic, unproblematic armchair epistemology do not suffice to provide principles of scientific method which aim to generalize the practices of successful science. But when we cull a list of such principles from the writings of scientists, philosophers and historians of science, we are disconcerted by the texts. Most of the principles are incommensurable, involving properties or relations that neither entail nor contradict one another. Examples: 'We prefer theories that correctly imply numerical observational values that are precise, that take a high risk.' 'We prefer a theory that forecasts novel facts rather than one that only explains facts already known and used to concoct the theory.' 'We prefer theories that imply qualitatively diverse facts, illuminating widely different realms of experience.' This independence of metapredicates produces conflict between principles when appraising theories and highlights the point that we deal with principles, not rules in the strict sense. The so-called 'scientific method' is not a set of rules such as reading the mercury meniscus or summing degrees of freedom in an analysis of variance. Second, there is not even qualitative consensus. For example, the preference for forecasting is crucial for Popper, a mild preference for most scientists, and irrelevant for Carnap. ('But Meehl,' Carnap asked me, 'how could the mere date of a fact affect its logical relevance?') Since the principles can countervail each other, and none of them is a guaranteer of truth, their relation to epistemic success is probabilistic. Now, how does one investigate probabilistic relations? By actuarial methods. Hence my thesis: 'Metatheory should be conducted by random sampling of episodes in the history of science, applying formal statistical and psychometric methods to analyze the results.' This Strong Actuarial Thesis I also call the Faust-Meehl thesis, since David Faust and I seem to be its only proponents (Faust, 1984; Faust & Meehl, 1992; Meehl, 1990a, 1992a)." (1993, in Waller et al. 2006, p. 424)
• "Popper is worse here, because (for some reason, what?) he harps on there being no rational prescription for discovery. Makes it seem all spooky, all 'non-rational' psychological processes, like Kekulé's hoopsnake. But that's plainly incorrect. Kekulé asked, 'How hook the 6 Hs and 6 Cs together? Can't be a string, like methane, ethane, etc.' Such constraints on the possibilities were rational constraints." (1998, in Peterson 2005, p. 70)
• "For psychologists lacking intrinsic philosophical interest but seeking help in their scientific theorizing, the 'received view'—a tolerant, amended form of logical empiricism—is the best prescription, and an excellent treatment of it is by Suppe (1977, pp. 7–118). It is still 90% sound, and the 10% error will do little or no harm to the working psychologist. I also recommend Carnap (1966), Hempel (1952), Nagel (1961), and Pap (1962), and there are excellent collections of readings by Boyd, Gasper, and Trout (1991), Brody and Grandy (1989), and Feigl and Brodbeck (1953)." (1993, in Waller et al. 2006, p. 431)
• "Because of my emphasis upon severe tests, which refuting H₀ is not and cannot become, some perceive me as a staunch disciple of Sir Karl Popper. This is incorrect. I count approximately 16 major theses which Popper defended throughout his life, and I agree with only three of them." (1998, in Waller et al. 2006, p. 439)
• Footnote to the previous quotation: "Meehl continued to think about the extent to which he could be called a disciple Popperian, and in May of 2000 made the following list of 36 positions, agreeing with 36% of them, disagreeing with the other 64%. He agreed with Popper's positions: Good science is good because it uses severe tests; Realism has been historically the aim of scientists and is a valid and valuable goal; Verisimilitude is an indispensable meta-concept whether quantitatively rigorizable or not; Science is often advanced by daring conjectures rather than cautious 'inductive' statements sticking close to the facts; The open society is the best and its utopian enemies are dangerous; Tarski's semantic conception of truth has relevance for empirical science and informalized language, and essentially solves the problem of realism; The world of science is a world of propensities, and the propensity concept should be allowed as a primitive notion; The primary aim of social organization and action should be to reduce suffering, rather than to bring about idealized states of affairs; Some form of democracy is the least evil kind of government so far devised; Society is better improved through piecemeal social engineering by incremental steps than by violent revolutions or grand schemes; Contemporary art is mostly of low value or fraudulent; Contemporary music is mostly of low value or fraudulent; The view of creative art, music, and literature as 'valuable self-expression' arising during the Romantic period is an undesirable view, both inherently and in terms of the product. He disagreed with Popper's positions: There is no such thing as the scientific method; a sharp demarcation exists between science and metaphysics; a scientific theory gets zero support from explanation without prediction; Freud's theories are untestable; Adler's theories are untestable; Marxism is untestable; Darwinism is untestable; there is no such rational process as induction; there is no such psychological process as induction; there exists an abstract world 3 which acts back upon worlds 1 and 2; it is impossible to speak of the probability of theories on their evidence; 'historicists' like Spengler and Toynbee (presumably Popper would today include Quigley, Kennedy, James Q. Wilson, Fukiyama et al.) have nothing correct or important to teach us; instrumentalism as a view of science is indefensible; all observational statements are theory laden; Hume's problem has been solved, or dissolved, by Popper; psychophysical determinism has been shown to be false; if determinism were true, human rationality would be precluded; if determinism were true, creativity in art, music, and literature would be impossible; theology shows lack of faith; when ethical propositions are properly understood, there is no unsolved philosophical problem as to their objective status or the bearing of facts on them; the problems of Platonism in mathematics, and why mathematics 'works' in empirical science, have been solved, or dissolved; the two-slit experiment is not paradoxical when properly formulated in terms of propensities of ensembles; accepting a protocol is wholly a matter of sheer decision, not warrantable or debatable.—LJY" (Waller et al. 2006, pp. 500–501)

Biogeographist (talk) 02:50, 28 January 2022 (UTC)

Upon further thought, I can see that the bit about Popper arguably belongs chronologically later in the section, since Meehl didn't read and discuss Popper until after LSD was published in English, which would have been later than some of the other stuff mentioned in the section. So I moved the part about Popper, but there's still no question in my mind that it is due weight to mention it. Biogeographist (talk)

You wrote It is more than "superficial" rumor. There is plenty of evidence that Meehl was influenced by Popper but also differed from him in substantive ways. The rumour was not at all about the fact that Meehl differed from Popper. The rumour, as described by Meehl, is that he was supposedly a disciple of Popper. Moreover, the concern is not about the veracity or falsity of the rumour. There is no doubt that the rumour makes no sense. It's very clear that Popper has no disciples. Even his best students criticized him. This is exactly why it will be very useful to find sources that clarify what exactly was the rumour and whether there is evidence, not for the veracity or falsity of the rumour, but for its existence. In other words, what evidence Meehl had to speak of the existence of the rumours in "I am not, contrary to rumours, a disciple of Popper"? Can we find this evidence. It's important, because it's the context that explains why Meehl says that he is not a disciple of Popper. We cannot use a content without explaining the context. Regarding relevance and encyclopedic quality, it is useful to explain how Meehl's view compare to Popper's view with the goal of explaining the central points of Meehl's philosophy. It's less useful to argue that Meehl's was not a follower of Popper just to make this point. Any insistence on the second objective would have undue weight. In the first objective, the similarities become as important as the differences and it naturally gets less POV pushing. In the second objective, there might be a tendency to select content only because Meehl criticizes Poppers in that content, not because it helps explains Meehl's philosophy or Popper's philosophy. Dominic Mayers (talk) 05:02, 28 January 2022 (UTC)

Regarding undue weight, consider how much time Meehl spent in the interview to explain that he is not a disciple of Popper. May be 10 seconds while the interview lasted more than one hour. The importance that is attributed to this 10 seconds is very subjective. I insist that the fact that very often Meehl criticizes Popper is a completely different thing. Consider the article about Thomas Kuhn. He has criticized Popper in more fundamental ways than Meehl and yet there is nothing in his article saying that he is not a disciple of Popper. Some differences in their respective philosophy are mentioned, but it's all within the overall goal of explaining Kuhn's philosophy. There is no intention at all in the article to make the point that Kuhn is not a follower of Popper only to make that point. Dominic Mayers (talk) 05:02, 28 January 2022 (UTC)

there is nothing in his article saying that he is not a disciple of Popper. Probably because he didn't say it multiple times like Meehl did. Since Meehl did say it multiple times, I don't think it's POV pushing to mention it. Biogeographist (talk) 12:56, 28 January 2022 (UTC)

Thinking more about this, it's very likely that the context is that Meehl actually often refers to Popper to explain his points, i.e., that the influence of Popper on Meehl is apparent and that his goal was only to clarify that he remains critical of Popper. So, the context could be that in reality Meehl is the closest among philosophers of being a disciple of Popper, though clearly he was not, because Popper did not invite that at all. In fact, his very philosophy invites the opposite and it was applied around him and toward him. Popper created a paradox. Perhaps no philosopher was more criticized by "his followers" than Popper, but yet he is also considered as one of the most influential philosophers of the 20th century. Dominic Mayers (talk) 10:56, 28 January 2022 (UTC)

Probably because he didn't say it multiple times like Meehl did. Since Meehl did say it multiple times, I don't think it's "POV pushing" to mention it. This argument suggests that the context, which determines the actual meaning and significance of a content, is not important. It suggests that relevance and due weight can be determined mechanically by counting the number of occurrences of a content without considering the context and the big picture, which depends on every thing else that Meehl and others wrote. I disagree with that. I haven't done the required readings to have a definitive judgment, but it seems very likely that this content taken out of context conveys the opposite of what a more thorough understanding of the sources would reveal. I know that many editors conflate trying to understand what is really going on in the sources with OR, but I see it as simply a required encyclopedic work. Moreover, undue weight is relative to a context. This content is relevant only if we decide that the connection Popper-Meehl is important. I think it is, but if it is, we must present the big picture, not only the differences, but also the similarities and only in that big picture the content in question becomes relevant and can respect due weight. Only in that context, we can consider that the readers are correctly informed and can make their own judgment adequately. Dominic Mayers (talk) 13:48, 28 January 2022 (UTC)

I'm really tired of you saying the same thing over and over to me, and implying that I don't understand the context and that I am being superficial. That's not true. I have been reading Meehl for over 20 years, and I know what he says. I don't see any problem with what I added to the article. If you have a positive contribution to make, edit the damn article. Biogeographist (talk) 14:05, 28 January 2022 (UTC)

I have no intention to seriously edit this article and many experienced editors that intervene in a talk page have the same attitude, you included for some articles. I do not want to say that you are superficial. I am very sorry that you feel this way. I am just sharing my understanding. For example, I only did a single Google search with "Popper Meehl connection" and the first result was "Heads I win, tails you lose? Meehl and many Popperians get this wrong". There could be a subsection on the subject with different views being presented—there might be others than Mayo. In that context, only in that context, the content could deserve a mention, but even then it should not have too much weight, not because Meehl is actually a Popperian or things like that, but because in the actual discussion, it's a secondary point. Mayo did not write to say that Meehl was a Popperian. This is not what she discusses. Similarly, when Meehl says he is not a Popperian it is not his central point, but only a clarification that he adds. In the interview, he spoke a lot to explain Popper, and only took 2 minutes to explain in which way he is not a Popperian. I agree that it makes a lot of sense to mention it, but only in a similar proportion in an appropriate section. It's more anecdotal than anything else. Besides, in these two minutes, Meehl has a lot of kind words for Popper. Within these two minutes, the excerpt taken is biased. Anyway, if a good job is done and we focus on the real important concepts, it should appear anecdotal in comparison. I do understand that Meehl really reacted to people saying that he is a Popperian, but then the context must be provided and it should be done in its own subsection. Dominic Mayers (talk) 14:25, 28 January 2022 (UTC)

Here is my transcript of the excerpt:

About the same time [in the middle 60s], a philosopher [...], a friend of mine, gave me a gift copy of Sir Karl Popper's book, The logic of Scientific Discovery, which I strongly recommend to you and I read that and I was impressed by it, but I had my philosopher's hat. Believe it or not, it took me three to four years and arguments with [...], which office was near me at the medical school, to see the deep connection. So, I want to tell you about Sir Karl Popper. I am not, contrary to rumours, a Popper disciple. Popper had a lot of opinions, ranging from arts, music to ethics, history and quantum mechanics. I got so tired of people saying 'Well, you know, Meehl he is a Popper disciple and that's why he has this crazy view about significance tests' that I made a list of Popper's opinions, 33 in numbers and I agree with only 11. So you can't be a guy disciple if you think he is 67% mistaken. For example, Popper detested [the name of an author]. Unlike Mark Twain, I think [the author] is not as bad as it sounds. Finally, I made the connection. Let me tell you a little bit about the guy. I am very please to see how often Sir Karl Popper's books are now being cited in the psychology literature and I take credit for being the one that [introduced] it into the profession. Though I am not a disciple, I think he is importantly right about a couple of very big things. On a lot of things, I think he is wrong.

— Paul E. Meehl, 2002

So, the key point is that he was tired of hearing the fallacy that his argument is wrong, because he agreed with Popper on many very big things. This is not at all what is being conveyed in the little excerpt that we see currently in the article. The real intention of Meehl cannot be understood from it. Dominic Mayers (talk) 17:52, 28 January 2022 (UTC)

First, a somewhat off-topic comment: I looked at that blog post by Mayo. Mayo quotes Meehl 1978 and responds: "No, there is a confusion of logic." I agree that she is making a point here about error that Meehl ignores in the quoted passage, but in Error and the Growth of Experimental Knowledge Mayo approvingly cites Meehl 1990c as a good argument from error. It looks like Meehl was more attuned to Mayo's point by the 1990s. But what's funny is that Mayo's refutation of Meehl—"No, there is a confusion of logic"—is what Meehl himself so often emphasized in other contexts. On some occasions Meehl used "Popperian" or "somewhat Popperian" as shorthand for this, but then he had to clarify that this didn't mean that he endorsed all of Popper's philosophy, e.g.: "The important point is that this state of affairs is not a matter of one's preferred philosophy of science, but it is a matter of formal logic" (1997, p. 400).

Anyway, even though I think the quotation is fine, I just changed the quotation at issue to a more general summary statement, which should make due weight even less questionable. Biogeographist (talk) 18:21, 28 January 2022 (UTC)

And thanks for writing the transcript of the Meehl interview; I appreciate it. Biogeographist (talk) 18:48, 28 January 2022 (UTC)

In 1978, Meehl was simply stating the Duhemian problem: we do not know if it is the theory $T$ or the initial condition $C$ or the auxiliary hypothesis $A$ that is incorrect given an observation $\neg P$ that differs from a prediction $P$. He wrote $(T,A,C)\Rightarrow P$ and $\neg P$ allows us to infer $\neg T\vee \neg A\vee \neg C$, only. Therefore, we cannot falsify $T$. I could not understand the details, but he seems to argue that this problem is more important in soft science. Meehl is perhaps aware that Popper theory unfold in a formal world where $A$ does not exist. So, the logic can be rewritten as $T\Rightarrow (C\Rightarrow P)$ and we observe $C\wedge \neg P$ then $\neg T$. Popper explains clearly that the potential falsifier is $C\wedge \neg P$. In the case of $T$ = "All swans are white", we have $C$ = "It's a swan" and $P$ = "It's white". So, it works perfectly in the formal world, i.e., on the logical side. Again, Popper is fully aware of the methodological problems that arises when we must introduce $A$ and also of the fact that even if we falsify $T$, it does not tell us what is the correct theory. However, he says that this is not a problem that can be solved logically. His answer is simply practical in terms of severe tests. He has not succeeded to formally justify why severe tests are useful. He observes that it's what we do in science and refers to an analogy with natural selection in the evolution of life. He tried to have some kind of explanation in terms of verisimilitude, but even that failed. As far as I can tell Mayo has misinterpreted Meehl. I suspect that it's the "playing heads I win, tails you lose" that confused her. Meehl cleary says in the article (before the part that is quoted) that if the theory is not rejected, it can count as a corroboration of $(T,A,C)$. They both say the same thing. The part that Mayo does not discuss is that the corroboration of $(T,A,C)$, which Meehl also acknowledge as valuable, has no logical value: a corroboration proves nothing (without required assumptions). The edifice of science is build over a swamp—there is no solid rock. Additional assumptions (often made implicitly) are needed, as when we apply Bayes theorem. However, when we use additional assumptions or conjectures and use them with deductive logic, the whole edifice remain hypothetico-deductive: no induction here, we are simply adding piles that seem to solidify the structure into the swamp, but actually the whole thing globally has still no foundation, no justification, no solid rock. From where the edifice of science seems to get its solidity is not answered logically. Many philosophers refused that situation, in particular Russell and Lakatos, but nobody said anything better than Popper as far as a logical justification is concerned. There are conjectures that some day we might find something and some ideas why it's possible (e.g. Nelson Goodman), but that day is yet to come. This paradoxical situation, i.e., the fact that science seems so solid and yet has no logical justification globally, is the source of all confusions. The reason is that, practically, we cannot keep saying that there is no foundation, no logic, etc. Even Popper is against that. Therefore, he claims that there is a rationality in using the hypothetico-deductive approach. He is against the sceptics. He viewed Kuhn and Feyerabend as sceptics, because they kind of denied the rationality of science and a methodology to support it. Dominic Mayers (talk) 22:33, 28 January 2022 (UTC)

Meehl mentioned Popper's swamp at least once (he could have mentioned Peirce too): "The arch positivist Otto Neurath (1932–1933/1959) spoke of 'repairing the raft you are floating on,' and Popper (1935/1959) made the analogy to 'sinking piles into a swamp.' Unfortunately in the social sciences, the situation is more like standing on sand while you are shoveling sand (MacCorquodale & Meehl, 1954, pp. 232–234), and, alas, in soft psychology the sand is frequently quicksand" (1990, in Waller et al. 2006, p. 130). Ian Hacking, in his chapter that I mentioned at Talk:Falsificationism, suggested that where the edifice of science seems to get its solidity is from the interplay between adjusting theory and adjusting instrumentation (Hacking deliberately bracketed out people from his analysis, but the long-term evolutionary adjustment of people's living bodies is part of that interplay as well): "The process of modifying the workings of instruments—both materially (we fix them up) and intellectually (we redescribe what they do)—furnishes the glue that keeps our intellectual and material world together. It is what stabilizes science." (Hacking 1992, p. 58). Biogeographist (talk) 04:43, 29 January 2022 (UTC)

After having written that, it occurred more clearly to me that the distinction between the logical and methodological sides and the role of the logical side in the rationality of science is very much related to the distinction between fallibilism and scepticism, as seen by Popper. There might be other ways to distinguish these two terms, but I feel Popper's way is very fundamental. Popper had no issue to refer to himself as a fallibilist, but he was against scepticism. The fact that there is no solid rock corresponds to the fallibility of science. It's only when we do not accept the rationality of the hypothetico-deductive approach with severe tests that we are a sceptic. The distinction is not easy to grasp. So, it's not surprising that many considered Popper as a sceptic. The distinction made by Popper between fallibilism and scepticism corresponds to the usual one, except that his fallibilism is extreme at the logical level and rejects any form of logical induction. He explains the rationality (i.e., the non scepticism) of the fallibilist in a different manner than with justifiable induction. Here I cannot stop myself from mentioning the joke that Meehl attributed to Morris Raphael Cohen:

All logic texts are divided into two parts. In the first part, on deductive logic, the fallacies are explained; in the second part, on inductive logic, they are committed.

— Attributed to Morris Raphael Cohen

If my understanding is correct, Meehl was not suggesting that we should commit these logical fallacies. If he did, he was certainly not a Popperian. Dominic Mayers (talk) 00:16, 29 January 2022 (UTC)

Meehl mentions the Morris Cohen quote in print too (1990, in Waller et al. 2006, p. 94), in the same article where he mentions Popper's swamp. Of course, as you noted, and consistent with Popper, it's a joke, not a prescription! Biogeographist (talk) 04:43, 29 January 2022 (UTC)

There is an important distinction that Meehl sees between soft and hard sciences. In hard sciences, the substantive theory is the null hypothesis and you are happy if it is corroborated, i.e., if the test is not significant and there was no falsification. On the other hand, in soft science, the falsification means an effect. So, it's the same mathematics, but the goal is turned upside-down: we want a falsification of the null hypothesis. In that case, we have the problem that we cannot tell for sure that the cause of the falsification is the substantive theory. I only mention that for completeness. It changes nothing regarding Mayo's comment, because she discussed the case of hard science as indicated by her reference to Einstein’s General Theory of Relativity. Perhaps she has not realized the distinction made by Meehl between soft and hard sciences, but then, if that is the case, she really missed the point. Dominic Mayers (talk) 03:55, 29 January 2022 (UTC)

Those are interesting thoughts about Meehl and Mayo. You may be right that Mayo misinterpreted that passage from Meehl; that would be another explanation for the contradiction with her approving citation elsewhere of Meehl's later text despite both texts covering similar territory. My impression upon reading Mayo's blog post was that she was right that Meehl didn't directly address her point "that a test might have had little chance to detect the effect, even if it existed". But that is not, strictly speaking, a "logical" issue, so calling it "a confusion of logic" isn't quite right. Biogeographist (talk) 04:43, 29 January 2022 (UTC)

Ian Hacking, in his chapter that I mentioned at Talk:Falsificationism, suggested that where the edifice of science seems to get its solidity is from the interplay between adjusting theory and adjusting instrumentation (Hacking deliberately bracketed out people from his analysis, but the long-term evolutionary adjustment of people's living bodies is part of that interplay as well): "The process of modifying the workings of instruments—both materially (we fix them up) and intellectually (we redescribe what they do)—furnishes the glue that keeps our intellectual and material world together. It is what stabilizes science." This is very interesting. I had a similar thought by myself without being able to link it to any particular text in the literature, perhaps because I read it and forgot it. Now that I think about it, the importance of the interplay between theory and technology is a lot emphasized in ancient Indian philosophy. Dominic Mayers (talk) 11:09, 29 January 2022 (UTC)