Talk:Falsificationism

Idea how to proceed for the article

@Biogeographist: Given that you are the person that wants the most to expand beyond what is currently in the section Falsifiability#Lakatos'_falsificationism, which I think is fine but difficult, you should take the lead. My idea would be that at first you do not mention Popper when you refer to the view point of Bunge, Rescher, etc. The reason is that there an issue of due weight, precision and also of not implying anything in Wikipedia's voice when we consider the confusion that exists around Popper. Perhaps, Rescher found it useful to refer to a common view on Popper's philosophy, especially the view that it does not have any inductive or similar component, which is clearly false, because Popper himself referred to his methodology as quasi-induction. It is also perhaps perfectly fine that he did that in his publications. As you say, he was an excellent philosopher and he knew what he was doing. It's not for us to judge. Indeed, as you say, he simply referred to a common view on Popper in a way that helped explain his work. However, this does not mean at all that it is fine that we include that content in the article even it is well attributed. For the article, we need to understand what is being said by Rescher, all aspects, including his view on Popper's philosophy, and decide whether it is relevant to the article, consider due weight, etc. and be precise. What I mean is that, if it is not appropriate to say "Rescher says that Popper's view is XYZ" in a precise manner, for any reason, due weight, not relevant to the context, etc., then it's not better, perhaps even worst, to imply it in an ambiguous manner. Again, I do not mind that we say "X said that Popper's view is YZ", even if Popper said the opposite of YZ, as long as it said in the right context, where it is relevant and that we respect due weight, etc. Given that this is not an article about Popper, I do not think it should be a big issue to remove content that implicitly present the view of philosophers on Popper's philosophy. This is not going to resolve all possible differences between us, but I think it would help a lot. Dominic Mayers (talk) 02:11, 29 December 2021 (UTC)

One thing I want to do first is find good secondary sources that argue that "falsificationism" shouldn't be associated with Popper. We already have Popper's own words in the notes to Falsifiability § Lakatos' falsificationism, and those sources are good (part IV of his 1982 introduction to Realism and the Aim of Science is especially interesting, and suggests that Kuhn played a key role in Popper's repudiation of the term), but I want to see some secondary sources (at least one, but more if possible). I am thinking that the secondary sources can be cited in the lead section, to counter the abundant secondary and tertiary sources that associate "falsificationism" with Popper. If you already know of such sources and can present them here, please do. I will do some searching in the next few days. It would be really interesting if there is already a source that compares a view like Miller's embrace of the term with Popper's repudiation of the term. Such a source would be pure gold.

Just so you know, I am not approaching this with any agenda of including Bunge, Rescher, etc. I mentioned Rescher in the discussion at Talk:Falsifiability because it was an interesting source that I happened to know and that contained a perspective a little like (but not exactly the same as) the view that "falsificationism" is a label for Popper's philosophy. It was information for helping to make the decision about whether to split off an article about falsificationism, but at this point in the discussion I don't see it as an important source for this article. It can be moved from the "Further reading" section of Falsifiability to this article's "Further reading" section, but I am not planning to try to work it into the article.

How are you thinking of dealing with the content in Falsifiability § Lakatos' falsificationism? Should we start by moving that whole section into this article, and just leaving a short summary in Falsifiability? Biogeographist (talk) 03:24, 29 December 2021 (UTC)

You want secondary sources that explain the confusion about Popper's philosophy like Popper did in part IV of his 1982 introduction to Realism and the Aim of Science. As you most likely noticed, there is Thornton that says that Popper emphasized the distinction between logic and methodology, which was the object of the confusion, but he is more polite: he does not refer to a confusion in the literature. I have seen another source, but it was in very specific context and I am not sure how notable is the author. However, though I agree that it will be nice to have more secondary sources that explicitly verify Popper's point, I would like to emphasizes there is no need for that, because there are many ways by which reliable sources can make a point of view important. When sources in general are not interested in the view point of an author, it's fine to ignore it. For example, I have no problem that we ignore the point of view of Faurisson in the article about the Holocaust. It's clearly relevant to the subject and it might look like a criticism that we should not hide, but most sources ignore him, so we can ignore him. However, if the so called reliable sources were to refer to him all the times while discussing the fact there is no way to deny the holocaust, then I would find it unacceptable that we ignore Faurisson's actual view, especially if he expressed his view very clearly so that all reasonable editors agree about what he wrote. This is the situation that we have with Popper. Reliable sources have not just written the opposite of Popper's view, because that would not be sufficient at all. No, they have attributed to Popper a view that is very relevant to the distinction issue. In this manner, they have demonstrated without any doubt that Popper's view on this issue is highly important. Also, irrespective of what we think of Popper's view, his actual point of view, which can be obtained when we read him directly (and Thornton, etc.) is a non controversial fact. We don't need secondary sources in that case to interpret correctly Popper. So, we can comfortably move ahead, especially given that we have Thornton as a secondary source anyway, but I would go as far as saying that it was not even needed, because the interpretation is easy, not controversial and shown highly relevant by "reliable" sources that act as secondary sources for notability, though not at all for accuracy. However, on the terminological side, I would say that the "reliable" sources that appear to clearly misinterpret Popper given the non controversial interpretation of a direct reading of Popper (and of Thornton and perhaps a few others) should not be considered secondary sources in this situation, because this would imply that their view point should prime over the opposite view expressed by Popper up to the point that we should exclude Popper's view point and Popper as a source, which of course would be totally against WP:NPOV, especially WP:DUE, using a text of Jimmy Wales himself. Dominic Mayers (talk) 06:13, 29 December 2021 (UTC)

Do you think Miller has a "correct" interpretation of Popper when Miller writes about falsificationism in his books Critical Rationalism and Out of Error? I think you may be begging the question about a correct interpretation of Popper. You imply that you and Thornton have a true "direct reading" of Popper and others such as Miller are misinterpreting him. To successfully make such a case without relying on questionable presumptions would require very detailed argumentation about the sources: for example, you may incorrectly presuppose that Popper was perfectly consistent and unambiguous over many decades of writing. But if Popper was not perfectly consistent and unambiguous in everything he wrote, then multiple valid interpretations are possible (although still not all interpretations would be valid—not everything goes). Biogeographist (talk) 14:48, 29 December 2021 (UTC)

Of course, all depend on our understanding of the sources. This argument of yours is relevant all the times, but can and must be often dismissed. If it could not be dismissed, the basic principle that we must use the point of view of the main partisans given in WP:due would become obsolete, because one could always use your argument to dismiss the point of view. So, it is important that editors read the sources and understand them. There is no way out of this. So, it is perfectly legitimate that I argue given a reading of Popper that his position was not controversial. He explained it in details in 1982, as you know. He also give a very enlightening reply to Lakatos in 1974 in Schilpp about this issue. He never wrote anything that could give any reason to believe that he changed his view. I am not talking about what other so called reliable sources say— I am referring to our direct understanding of Popper. Also, given an understanding of Popper's fundamental view regarding the distinction between logic and methodology, it's not imaginable that he would have changed his view about that. If he did, he would have made it very clear. Don't do cyclic reasoning and decide that Popper's view should be dismissed from the start, because this is exactly what would be opposed to the text of Jimmy Wales in WP:due. Besides, we have Thornton, but as a general principle, I insist that a direct reading of Popper is important. But, OK, I already admitted that if editors disagree on their understanding of a source, then they should use other sources to help them getting a consensus, but this does not mean at all that the original sources should be ignored in the discussion. That would be a misinterpretation of the role of the secondary sources in the context of WP:due, which says that we must give the point of view of the main partisan. Don't you see that this misinterpretation could always be used to dismiss the point of view of a main partisan, by saying that it's the point of views of others that must be used? Honestly, I am getting tired of this kind of arguments. Are you seriously saying that personally you have read Popper in 1959, Popper in 1974 in Schilpp, Popper in his 1982 Introduction of the Aim of Science, Thornton recently and you personally think that Popper's point of view is not clear on this issue. If not, then I don't want to hear anything else, because if others think differently, they can come and we will discuss with them. And, if yes, you think that Popper was not clear about the importance of a distinction, then I am very sorry, but I don't think I can continue this discussion anyway. I say that as kindly as I can. Dominic Mayers (talk) 15:45, 29 December 2021 (UTC)

(edit conflict) OK, I must admit that it is very legitimate to expect that our reading of other sources, Lakatos, Kuhn, Watkins, Miller, etc. is consistent with our interpretation of Popper. Of course, this consistency is very necessary. But, do not forget that a key point in Popper's point of view and easy to see in his writing is that Lakatos and an entire literature failed to understand the importance of this distinction. So, on the contrary, the fact that Popper is interpreted differently than in our direct reading is perfectly consistent with our reading of Popper. Not only that, but when you read in which way Kuhn, Watkins, etc. criticized Popper, you can see that it is actually consistent with Popper's position. For example, Watkins in Science and Scepticism made fun of Popper because of his way to emphasize the logical (and thus imaginary) side. Similarly, Kuhn in Logic of Discovery or Psychology of Research?, mentions that Popper only relate sentences with sentences (as is the case in logic) and he criticizes that. So, when you look at the details of the controversy between Popper and other philosophers, it only confirms very clearly that Popper did emphasize the importance of the logical side. It's all over the place. There is no possible misinterpretation here. Dominic Mayers (talk) 16:22, 29 December 2021 (UTC)

I didn't say that Popper's writings should be dismissed; in fact, I said in my first comment those sources are good ... but I want to see some secondary sources. When I said "I want to", I was speaking only of my own contribution to the article; I was not saying that other editors need to find secondary sources. I have not read Thornton (or if I read it, I have forgotten what he said), so I will look at that in my review of secondary sources. The problem with Popper is that, if I am not mistaken, he only engages with Lakatos and Kuhn, and not with Miller. Also, Popper says that it's wrong to call him a "naive falsificationist" (and I get the impression that Popper objects not only to the fact that "naive falsificationism" refers to a philosophy that is not his but also that the word "naive" is insulting), but he does not say that it is wrong to call him a falsificationist (with no qualifier), only that he does not use the term himself. That may be why Miller feels free to proudly use the term falsificationism. If Popper had asked people not to use it, I do not doubt that Miller would have refrained from using it. Biogeographist (talk) 16:20, 29 December 2021 (UTC)

(edit conflict) Popper apparently applied his critical approach with other philosophers' view, but this means that he tried to criticize them on fundamental issues that could help the field to progress. So, if he has not criticized Miller, it's because he could not find anything important to criticize. You are correct that Popper does not care about the use of the term "Falsificationism". Popper says of himself that he does not quarrel about terminology as long as the terminology is clear in the context where it is used. You are right that his rejection of the term "Falsificationism" was only to dissociate himself from Lakatos's many falsificationisms. Otherwise, he would have complained to Miller and Miller would not have used the term. I agree. But, I am not sure why you always come back to this terminology issue, especially after I have accepted that we can use "Falsificationism" as a very broad term that includes Popper's philosophy. However, this is not at all saying that Popper would have been so flexible. It's one thing to agree that in a specific context "Falsificationism" is used to refer to Popper's philosophy and not Lakatos's philosophy and, in another context, used with a different meaning. It's a different thing to agree that "Falsificationism" can be used to mean both in the context of one article, which is what I finally accepted to do in Falsificationism, but under the condition that whenever possible we use a more precise name to refer to Popper's philosophy and are very careful not to create or infer any confusion by the use of such a broad term, which, in a way, starts to mean nothing when interpreted in such a broad manner. After I wrote this, i.e., now for me, I start to wonder if your concern is that Miller might have endorsed Lakatos by using the term "Falsificationism". I can assure you that this would be a complete misinterpretation of Miller. Miller did not consider Lakatos's philosophy as a part of his view on falsificationism. He actually wrote that Lakatos's philosophy was a "falsification of falsificationism". In any case, it's clear that Lakatos explicitly tried to find an inductive procedure, whereas Miller is even more than Popper opposed to any form of induction, that is, he was more than Popper centred on the logical side. For example, I need to read Miller more carefully, but I don't think that he says a lot about Popper's evolutionary perspective. Dominic Mayers (talk) 17:04, 29 December 2021 (UTC)

OK, I think I finally understand your position. I agree with all this. Biogeographist (talk) 17:12, 29 December 2021 (UTC)

And when I wrote about multiple interpretations, I was especially thinking of the difference between Miller and Watkins about the importance of corroborations in Popper. If Watkins is right, then I think that Miller's use of the term falsificationism may be a bad idea, since it may imply that falsification matters but corroboration doesn't matter (Miller literally says the latter). But I don't know that Watkins is right; as far as I know, there may be two valid interpretations here. Biogeographist (talk) 16:28, 29 December 2021 (UTC)

You wrote : I think that Miller's use of the term falsificationism may be a bad idea. I don't know what you are trying to do. If you are trying to see a single meaning or a central point common to all use of "falsificationism" in the literature, I don't think it's a good idea and it might even create confusion by suggesting that the term was used in a consistent manner in the literature. In fact, this is why I say it was a big concession on my side, perhaps a mistake, to accept that we use "falsificationism" in a very broad manner. Well, when I made this concession, I somehow give the meaning that I attach to this very broad term, which is a generalization perhaps not used by Miller nor Lakatos:

This page is about different approaches to face the falsification problems in science. For the approach in Popper's philosophy, see Falsifiability.

As you see, this is very broad. It certainly does not exclude the use of corroborations. It does not exclude Popper's philosophy, because his logical criterion and his methodology can certainly be seen as a way to approach the problems of falsifications, by using the fact that the logical side is free from the experimental issues and the problems that cannot be covered logically are covered within rational critical discussions. Dominic Mayers (talk) 17:13, 29 December 2021 (UTC)

(edit conflict) Regarding Miller's position, when you read him, keep in mind that he focuses a lot more on the logical side. This is consistent with the fact that he says that corroborations are useless. Indeed, in the purely mathematical situations that were considered by Popper and Miller in their joint papers, corroborations are useless or rather make a negligible difference. I know that you say that I am too generous with him, but I disagree. It's very natural to consider that Miller makes reference to the contexts that he considered in his joint work with Popper. I mean, it's very clear that Popper's point that the "probability" of any law is zero is not true under many assumptions. This is the whole point: Popper and even more Miller are regularly arguing on the logical side where unless stated otherwise, there is no assumptions. You can also see that Miller is grounded on the logical side in the way he responds to criticisms that refer to the limitations of the logical side. In my view, he evades the question by responding that his use of the logical side is not in contradiction with usual practice in science. He is like saying that the burden of the proof to show a problem in Popper's philosophy is on your side. I find his position unsatisfactory. Popper was expressing a greater understanding of the need to address the limitations of the logical side and this is why he discussed his evolutionary perspective and it is also why he insisted that the logical criterion must be complemented by a methodology and the important role of critical discussions. Dominic Mayers (talk) 17:53, 29 December 2021 (UTC)

Thanks, this is helpful and explains a lot. Biogeographist (talk) 18:03, 29 December 2021 (UTC)

(edit conflict) When I said Miller's use of the term falsificationism may be a bad idea, that was a bad way of phrasing it. More neutrally stated: Can the conflict between Miller and Watkins about whether Miller's falsificationism is correctly associated with Popper be clearly resolved in favor of one or the other, or are both interpretations valid? This is an interesting question that may have a place in the article, especially if the question is addressed in another source.

If you are trying to see a single meaning or a central point common to all use of "falsificationism" in the literature, I don't think it's a good idea and it might even create confusion by suggesting that the term was used in a consistent manner in the literature. I agree; that's not my point.

In fact, this is why I say it was a big concession on my side, perhaps a mistake, to accept that we use "falsificationism" in a very broad manner. I am not suggesting using "falsificationism" in a broad manner. That's why I didn't understand your argument about general and specific cases at Talk:Falsifiability. I am in favor of clarifying the different uses of the term in historical context. Biogeographist (talk) 17:29, 29 December 2021 (UTC)

So, I interpret your question simply as whether Miller was fairly exposing Popper's philosophy? I think that it is certainly a very valid question, which we should clearly have in mind when we read him and Popper, etc. This is basic. However, once we have our answer, I don't see that we need to expand on this in the article. Instead, we use our understanding, as we have no choice as to do, to write the article. This is different than addressing the question as a subject in itself in the article. I already give my answer to the question. Miller is quite close to Popper. Otherwise, Popper would have criticized him. The key difference is that he is less interested than Popper to address the limitations of the logical side. Because of that, I find him much less interesting than Popper. It's logical that Popper would not have criticized Miller on that, because Popper himself considered that this part of his writing was important to help people accept his philosophy, but was nevertheless complementary to his philosophy, not what epistemology should be rigorously about. Dominic Mayers (talk) 18:11, 29 December 2021 (UTC)

I will keep that in mind. Biogeographist (talk) 18:17, 29 December 2021 (UTC)

History

Now I am thinking that trying to start by writing the lead section is approaching the problem backward. It would be better to try to start writing the body in a draft article first. I will probably start with a "History" section. For now, I would start with Popper's LSD and its translation into English in 1959. As far as I know currently, falsificationism as a term didn't start to be used until after LSD was translated into English, although the problems of falsification date back to the original publication of LSD in the mid-1930s. The first appearance of the term that I have found is in: Buchdahl, Gerd (March 1965). "A revolution in historiography of science". History of Science. 4 (1): 55–69. doi:10.1177/007327536500400103. Buchdahl's article is a review of key books by Kuhn and Agassi. This is very significant since both Buchdahl and Lakatos are using the term "falsificationism" to talk about issues that arise in the confrontation between Kuhn and Popper, and Popper's 1982 discussion of the issue also focuses on this. Biogeographist (talk) 18:17, 29 December 2021 (UTC)

(edit conflict) I had written a long text with many concerns, but I decided that I should wait and see instead. I just kept the final paragraph, which follows. There is not only one way of doing things. I could imagine that in the historical context, we have consecutive sentences such "As early as 1965, the term falsificationism was used by Buchdahl to refer to a popperian approach that faces serious problems of falsifications. Later, Lakatos, used the terms dogmatic and naive falsificationism for a similar purpose. In this context, Popper insisted that his falsifiability solution was always presented as a logical criterion used in science without having to face any of these problems of falsifications, which are better addressed within rational critical discussions that refer to severe (statistical) tests." The point is that it's better to say things clearly rather than letting them be implied. If they cannot be stated clearly, say Buchdahl was not so clear, then we avoid any phrasing that can suggest them. Dominic Mayers (talk) 21:22, 29 December 2021 (UTC)

(edit conflict) Unfortunately because you responded, I need to keep the other paragraphs as well: What I don't understand is that you say that you agree that "Falsificationism" should not be a broad term, but it appears to me that you are taken a very broad perspective that includes Popper's philosophy. Perhaps, you only want to provide a context for a more specific perspective. Or that you consider that the subject is more "FalsificationismS" than "Falsicicationism". As you see, I am very sensitive and concerned when we use terminology that can support a confusion. Well, I already agreed that we give to the title "Falsificationism" a very broad meaning that cover all possible kind of falsificationisms in the literature, including Miller's falsificationism. I do not change my mind, but, again, if we do that, first let us agree between us that our "falsificationism" is not the falsificationism of Miller nor the "falsificationism" of Lakatos. It is just our generalization of the term and, second, let us agree to be very careful to use a more specific term for Popper's philosophy so that it's clear that it is a very special case that has nothing to do with Lakatos-like falsificationism. We must look carefully at every use of the term falsificationism in the literature, i.e., look at how the content relate to the problems of falsifications, not what is claimed by the author, to determine if it is about Popper's falsificationism or not. My understanding is that almost all the times, Miller is one of the exceptions, it's about a Lakatos-like falsificationism, not about Popper's philosophy, despite what is being claimed. By Lakatos-like falsificationism I mean, either something like the non-sophisticated falsificationism of Lakatos, i.e., a falsificationism that suffers from the problems of falsifications or something like the sophisticated falsificationism of Lakatos that tries to provide a rational inductive methodology with some form of confirmationism. Most of the times, the Lakatos-like falsificationisms that face the problems of falsification (not the sophisticated falsificationism) are somehow attributed to Popper. This is exactly the confusion that I refuse to present in Wikipedia's voice. In particular, Buchdahl clearly refers to a Lakatos-like falsificationism and, I don't remember for sure, but I think he attach it a lot to Popper, but this is confusing. The distinction between these different falsificationisms is very important and it is not because there is no awareness of this in some review that we can just report what the review says. We need to look at the content and refer to it in a coherent manner as a specific kind of falsificationism. For example, if we detect a Lakatos like falsificationism, say because there is a reference to the problems of falsification as being problems for the philosophy instead of the role of the distinction between logic and methodology as a way to address the problems, then we cannot imply that it's Popper's philosophy, no matter what is implied or stated by the source. You might argue that we will simply state what the source says, but that's not a solution. We can and must use our understanding of the sources to say more precisely what kind of falsificationism is being discussed. Again, there is a due weight aspect here. I am not against saying that Buchdalh attributed a problematic falsificationism (facing problems of falsifications without any mention of the role of a distinction between logic and methodology) to Popper, but we must be aware that we are doing that, make sure that it's not done in Wikipedia's voice, and more importantly, we must make sure that it's relevant in the context. In particular, it's not relevant if there is no understanding of Popper's philosophy in the context. I hope you see that I am not interested in hiding any point of view, but they must be stated in a proper context in such a way that the readers can make the best judgment, that is, we should not present a view point that is implicit in the source, not even clearly stated, without providing the proper context to make a judgment. A good encyclopedic work must organize the content in view of a good understanding of the sources in a way that helps the readers. Verifiability is very very far from being sufficient, especially for this subject. Dominic Mayers (talk) 20:27, 29 December 2021 (UTC)

I understand your concerns, and I think they can all be addressed. I think of the article as a history of points of view on falsificationism and related problems of falsification and corroboration.

• Before anyone uses the term falsificationism, Popper's POV on problems of falsification sets the stage for the whole show with LSD, translated into English in 1959.
• Then comes Kuhn's POV in SSR in 1962.
• Popper's Conjectures and Refutations comes out in 1962, and he uses the term "falsificationist" to refer to a group that implicitly includes himself.
• Buchdahl reviews Kuhn's SSR in 1965, using the term falsificationism with his own POV.
• Then there's Lakatos' POV with CMSRP in 1968 and later versions. (There may be other relevant material in Criticism and the Growth of Knowledge from 1970, where a new version of Lakatos's paper joins papers from Kuhn, Popper, and others.)
• Schilpp's The Philosophy of Karl Popper in 1974 brings more from Kuhn & Lakatos, and Watkins, with responses from Popper.
• Chalmers writes about falsificationism in What Is This Thing Called Science in 1976.
• Popper responds again to Kuhn in his 1982 introduction to Realism and the Aim of Science.
• Watkins says some stuff in Science and Scepticism in 1984.
• Miller responds to Watkins in Critical Rationalism in 1994.

Those are the primary sources I know off the top of my head. Are there any important interventions that are missing?

This is a very straightforward approach. There is no building of an original argument that could possibly be called WP:OR. It just lays out the facts of who said what and when, in temporal order. Later interventions correct earlier interventions when they are wrong. Secondary sources can be used when they accurately reflect what the facts of the primary sources show.

After all that has been written, it can be summarized in the lead section. The major misunderstandings have to be conveyed in the history, because they are part of the history, but the lead section can leave out the messiness of the history and focus on what was the outcome of the debates. Biogeographist (talk) 21:59, 29 December 2021 (UTC) and 23:08, 29 December 2021 (UTC)

Well, I am not saying that mentioning the early use of the term falsificationism is irrelevant, but I am saying that these reference to usage of the term are often not useful to convey the important notions and can even be confusing. For example, I just making this up — it's most likely not the case, it could be that "General Relativity" was used by physicists before Einstein in a way that is only very remotely related to Einstein's theory. In that case, a reference to that historical context in an article on General Relativity can be confusing. It's not the same situation with Buchdahl, but it creates a similar issue if we present the connection with Popper that is implied by Buchdahl before there is a context to be able to appreciate the controversial aspect of that connection. I am just saying that this connection should be removed and there is nothing wrong with that: not every thing that is verifiable must be included. Besides, I am not against stating the connection, but if we do it, we do it clearly, not in Wikipedia's voice and in a relevant context, i.e., only after the controversial aspect can be appreciated given an understanding of Popper's philosophy. Dominic Mayers (talk) 20:50, 29 December 2021 (UTC)

I see what you mean, but general relativity is not a good analogy since Popper never used the term "falsificationism" to refer to his philosophy the way Einstein used "general theory of relativity". All we have to do is explain Popper well enough before Kuhn and Buchdahl and Lakatos enter the scene, and that solves the problem. Biogeographist (talk) 21:59, 29 December 2021 (UTC)

I was aware that the analogy was broken in that respect, but it was not entirely broken. The analogy still convey the general concern that an historical aspect can be beside the point. I know that anything written anywhere can be beside the point. But, the difference is that if we are motivated by history instead of the actual issues that are even today important, then it has more chance to be beside the point. Dominic Mayers (talk) 22:20, 29 December 2021 (UTC)

Well, there is not only one way of doing things. I could imagine that in the historical context, we have consecutive sentences such "As early as 1965, the term falsificationism was used by Buchdahl to refer to a popperian approach that faces serious problems of falsifications. Later, Lakatos, used the terms dogmatic and naive falsificationism for a similar purpose. In this context, Popper insisted that his falsifiability solution was always presented as a logical criterion used in science without having to face any of these problems of falsifications, which are better addressed within rational critical discussions that refer to severe (statistical) tests." The point is that it's better to say things clearly rather than letting them be implied. If they cannot be stated clearly, say Buchdahl was not so clear, then we avoid any phrasing that can suggest them. Dominic Mayers (talk) 21:22, 29 December 2021 (UTC)

@Biogeographist: You wrote I think of the article as a history of points of view on falsificationism and related problems of falsification and corroboration. This taken together with the key history points that you proposed looks good. My only concern is that, even though Miller is close to Popper in his respect of the importance of the logic, he does not present well how Popper motivates his view using the evolutionary perspective, etc. I would say that how critical discussions can work is not as clear with Miller than with Popper, because Miller sticks more on the logical side. In that sense, his replies are not truly representing Popper's view and we can understand Watkins's position in that light. So, it will be good to use other sources in addition to Miller at the end. Dominic Mayers (talk) 22:13, 29 December 2021 (UTC)

Yes, that sounds good. I am preparing to write a draft. Biogeographist (talk) 17:30, 5 January 2022 (UTC)

Link with Quine-Duhem

@Biogeographist:, I had a long discussion with Loew_Galitz in Talk:Falsifiability#False_statement_in_the_lede_(highlighted) and Talk:Falsifiability#Extreme_request_for_mechanical_verification_in_the_lead.. This discussion sheds light on the difficulties that some people might have to understand the subject, but above all it made me realize the importance of the Quine-Duhem thesis in Falsificationism. When you read the Introduction in Harding 1976, it appears as essentially the same subject. I would suggest that the history begins with the Quine-Duhem thesis.

We should try to see what's the conceptual difference between the two articles Falsificationism and Quine-Duhem thesis. Falsificationism can be seen as a study of the Quine-Duhem thesis: a search for a way to work with it, to make it less problematic. Popper proposed Falsifiability as a logical criterion in the light of the Quine-Duhem thesis. It's sad that we always explain Falsifiability in the light of Hume's problem when it could as well be explained, perhaps even better explained, in the light of the Duhemian problem. They are both problems in the growth of science, which we can refer as the general problem of induction. Hume focused on verifications in the general notion of induction whereas Duhem focused on falsifications also in the general problem of induction. It's very naive to think that Popper was only concerned by Hume's problem (the verification aspect) and did not address the Duhemian problem (the falsification aspect). Popper did not call his philosophy "falsificationism", but it nevertheless fits within the above view on falsificationism. With Lakatos, falsificationism has developped as a distinct inductivist approach, opposed to Popper's philosophy, but it certainly fits within this view on falsificationism also. The term "falsificationism" was originally used by Buchdahl, Lakatos and others in the context of a criticism or misunderstanding of Falsifiability. It's paradoxical, but yet fundamental, that they turned the Quine-Duhem thesis against Popper's philosophy. The point is that the Quine-Duhem thesis seems central.

We have to bring these views on Popper's philosophy, because it's a reality in the literature, but these point of views must be clearly attributed so that Popper's philosophy is never misrepresented. I mean, clearly many authors will present Popper's philosophy in a way that allow them to explain their contribution, but often it's biased and it cannot be stated in Wikipedia's voice. Also, these specific (biased) views on Popper's philosophy must only be stated when they are relevant, not implied in a confusing way there and there in a way that violates WP:due. Dominic Mayers (talk) 12:36, 13 January 2022 (UTC)

In general that is an important issue, but to be precise, it is anachronistic to say that the history begins with the Quine-Duhem thesis, since LSD was originally published long before Quine intervened (but, interestingly, LSD was translated into English after Quine intervened, so LSD was written in a pre-Quinean context but read in English in a post-Quinean context). It's certainly an issue that should be addressed in the proper historical order with attention to the primary and secondary sources (e.g., Duhem and Quine were not saying exactly the same thing, and they were writing in different contexts). One secondary source is Thornton 2007. Biogeographist (talk) 16:18, 13 January 2022 (UTC)

Also, Otto Neurath is an important part of that history. Neurath influenced Quine, and Quine popularized Neurath's boat, and Neurath had already written an important review of Popper's LSD in 1935 very soon after it was first published that addressed the problems of falsification (Neurath 1983). See, for example, Rutte 1991, p. 87:

In the case of Duhem, it was of course the holism and conventionalism which Neurath adapted early on. (Mach, incidentally, viewed Duhem's main work as a continuation of his own efforts.) If today one speaks of the Duhem-Quine thesis, one really should speak of the Duhem-Neurath-Quine thesis in order to indicate the historical development of these ideas.

Biogeographist (talk) 16:46, 13 January 2022 (UTC)

Even if we think of the article in a historical perspective, the subject is not the history of falsificationism and we do not have to claim an order in time. So, if the best order to explain the concepts is not time-ordered, we can still use it, i.e., make an exception to our perspective as editors. Besides, I claim that, in a way, it's not an exception. First, let me say what my argument is not. It's not that Popper considered the Duhemian problem in the original 1934 German version. He did, but that's not my argument. Popper faithfully translated his book in 1959 and all additions occur in footnotes marked with an asterisk. The following excerpt is from the body written in 1934 :

Let p be a conclusion of a system t of statements which may consist of theories and initial conditions. [...] By means of this mode of inference we falsify the whole system (the theory as well as the initial conditions) which was required for the deduction of the statement p, i.e. of the falsified statement. Thus it cannot be asserted of any one statement of the system that it is, or is not, specifically upset by the falsification. Only if p is independent of some part of the system can we say that this part is not involved in the falsification.

Here auxiliary hypotheses and the background knowledge may be seen as a part of the initial conditions. i.e., initial conditions may include every thing else in the system besides the theory itself. For example, in "All swans are white", "Being a swan" is a very complex "initial condition": the conclusion that the bird (or the thing) is white depends on this initial condition, which includes a lot of background knowledge and auxiliary hypotheses. This is also Popper's interpretation of himself: in the index of his 1959 translation, the entry "Quine-Duhem thesis" asks to see "Systems", which refers to this text. My argument could also have been that the Duhemian problem was described by Duhem (1861–1916) way before 1934. Here is an excerpt from Ariew 2020:

Against this dogma Quine suggests that “our statements about the external world face the tribunal of sense experience not individually, but only as a corporate body” (1953, 41), and, in a footnote of the reprinted article in his collected essays, From a Logical Point of View, says that the doctrine was well argued by Pierre Duhem.

So my argument could have been that the Duhemian problem preceded Popper's work, but it's more fundamental than that. My argument is that not only the Duhemian problem preceded Popper's work, but it was also well known by the Vienna circle before Popper's work. Ariew 2020 also says:

Duhem’s work was important for members of the Vienna Circle, including Otto Neurath and Philipp Frank, as it had been for Ernst Mach. Despite Duhem’s conservative beliefs, his work was also taken up by participants in the Viennese political scene, such as Friedrich Adler, who had translated La théorie physique into German in 1908. The Duhem thesis surfaced fully in Anglo-American philosophy in the 1950s through the work of W. V. O. Quine.

The key point is that Duhem's problem was background knowledge in Popper's environment when falsificationism (in the form of criticism) surfaced. It was well understood very early by Popper, Lakatos and most likely Buchdahl as well. The issue in falsificationism was not a misunderstanding of Duhem's problem, but a misunderstanding of the distinction between the logic side and methodological side of science and its relevance to the general induction problem, i.e, the problem of how to explain the growth of knowledge. So, this background knowledge should be explained first and then the real issue described in the light of this background knowledge.

I emphasize the relation between concepts over time-ordering here. Again, let me emphasize that it's not sufficient for inclusion to argue that we quoted philosophers or even paraphrased them. We need to make sure that what they say is well understood by the readers given the context and that it is relevant in the context. For example, if we consider to include that Lakatos wrote that Popper was partly a naive falsificationist, we must make sure that it is relevant and respect WP:due and, more importantly, that Popper's actual position, which is misrepresented by Lakatos here, was clearly explained before. It's not clear at all in Lakatos's text, but Wikipedia cannot present this confusion in its own voice. On the contrary, it must provide the best information so that the readers can perfectly understand Popper's view and Lakato's view on Popper's view and allow them to easily see the difference, i.e., the confusion, by themselves. The readers should see the confusion, but not be themselves confused. Dominic Mayers (talk) 20:29, 13 January 2022 (UTC)

I still maintain that it's anachronistic to say that the history begins with the Quine-Duhem thesis; I hope you would agree. Your source, Ariew 2020, says that Duhem and Quine didn't say exactly the same thing. The same could probably be said of Neurath. All of this needs to be laid out in proper historical context. Here is a copy of the first English edition of LSD in the Internet Archive. I searched for "Quine" in it; Quine's name doesn't appear anywhere, not even in the index. In a later edition, some indexer added "Duhem-Quine thesis" to the index, but it's an anachronistic index entry.

Popper does refer explicitly to Duhem; on page 78 Popper says that against Duhem he asserts "the possibility of crucial falsifying experiments". So that part is certainly a relevant issue. Biogeographist (talk) 20:55, 13 January 2022 (UTC)

(edit conflict with the added section and the last paragraph above) The subject of the article is not the history of anything. If we start to think of the scope of the article in terms of history, then the scope change, because then one could argue that a point of view or some nuance was important historically. For example, I am sure that many people consider that the distinction between the Quine-Duhem thesis and the Duhemian problem is very important historically. I liked your historical perspective and I still like it, but it should be used to explain the main concepts that are relevant. We can help us with this historical perspective and I think it can also help the readers, but history is not the subject.

Moreover, regarding history, we must distinguish two things: misrepresenting the history and not presenting the concepts in a time-order. For example, saying "I went to bed at 9pm, because I was tired by my work that I finished at 8pm" does not say things in a time-order, but it does describe the correct flow in time of the events. In our case, we can present the Duhemian problem first and say in the same section that Quine expanded on it much later. This would not misrepresent the history at all.

Actually, I don't care that much about the order in the article as well. The important is that the relationship between concepts and their respective role is well explained. It's the situation when falsificationism surfaced in the 60's that matters. The situation was that Popper understood very well the Quine-Duhem thesis and his philosophy was more a solution to this negative thesis than the other way around. Nevertheless, some presented the thesis against Popper's philosophy. In this case, the real issue, the source of the confusion, is the misunderstanding of the distinction between the logic side and the methodological side and its importance in Popper's philosophy and a misunderstanding that the Duhemian's problem was well understood and addressed by Popper. Wikipedia cannot suggest in its own voice that Popper presented his philosophy and next it was criticized in the light of the Duhem-Quine thesis. It's very tempting to present things in this wrong way, because it creates some simplicity. In this wrong way, we present Popper's philosophy without insisting on the distinction and how the logical criterion was still practical even though it is purely at the level of statements, including imaginary potential falsifiers. Next, after having presented Popper's philosophy in a limited manner and out of its Duhemian context, we presents the Quine-Duhem thesis as a problem for Popper's philosophy. This is very wrong and is making Wikipedia presents confusion in its own voice. But, there are many ways to present the correct view and I don't care about ordering details.

In that context, the distinction between the Duhemian problem and the Quine-Duhem thesis seems not to respect WP:due. You will need to argue in which way it is relevant to the main focus of the article, if you want to say more than a small historical remark about it. Given that falsificationism emerged after the Quine-Duhem thesis, the point could only be that it influenced Popper's position after 1934. The debate around Quine's version did that in a way, because Popper definitively added footnotes, etc. in the English version in view of his realization that he was accused of not taking it into account and I am sure that the debate around the Quine version helped. However, I don't think that we should cover the history at this level of details. If we did, we would have to present the Duhemian version and how it was very early background knowledge at the least for Popper, but apparently many others in the Vienna circle. Then explain the difference with the Quine's version. All of this is not so relevant. It's a distraction away from the key issue: a failure to understand distinction between the logic and the methodology and its importance in Popper's philosophy. The distinction between Duhem and Quine-Duhem is a small detail in comparison. Again, an emphasis on this would easily fail to respect WP:due. Dominic Mayers (talk) 22:27, 13 January 2022 (UTC)

To be clear, I am not saying that the article should reach a stable form that is strictly historical. It just seems that the historical approach is a good way to start organizing the material, and in the process to answer the kinds of questions you have raised.

My current position is that the distinction between Duhem and Quine is relevant in relation to LSD. Quine shouldn't appear in the article before LSD. But again this is just my attempt to organize the material in advance and it may turn out later that some other organization will work better. Biogeographist (talk) 22:45, 13 January 2022 (UTC)

Popper uses the term "falsificationist"!

Wow! I was just looking for other places where Popper mentions Quine, and on pages 238 and 239 of Conjectures and Refutations (1962), the same pages where Popper mentions both Duhem and Quine, Popper also uses the term "falsificationist" twice, implicitly refering to a group that includes himself! So perhaps the first person to use the term "falsificationist" was Popper himself, to refer to himself! Amazing! I'm adding Conjectures and Refutations to the list of primary sources above! Biogeographist (talk) 21:29, 13 January 2022 (UTC)

On page 228 it's not even implicit, it's fully explicit: "Falsificationists (the group of fallibilists to which I belong) believe..."! Biogeographist (talk) 21:48, 13 January 2022 (UTC)

(edit conflict with the paragraph below) I don't see the importance of the use of the term "falsificationism" or "falsificationism" by Popper or others. The term is so unclear in the literature that the only thing that can be conveyed to the readers by such an emphasis is confusion. Again, if there is any thing that is conveyed by an author by the fact that he used "falsificationism" in a sentence, then we say it explicitly not ambiguously (not only implying it by a use of the term "falsificationism"), verify that it is consistent with other sources (including from the same author) and make sure that it's relevant in the context where it is included and that it respects WP:due. Otherwise, it's like picking a sentence out of context. It's the actual point of view of the author, as can be seen by considering many texts that matters. And it must be relevant and respect WP:due. I thought that we had an understanding on this. Dominic Mayers (talk) 22:47, 13 January 2022 (UTC)

I'm not advocating picking a sentence out of context! That's exactly the opposite of what I'm advocating! Biogeographist (talk) 22:51, 13 January 2022 (UTC)

How can we clear up confusions about how the term is used if we don't mention and explain such uses? That has to be part of it. Biogeographist (talk) 22:59, 13 January 2022 (UTC)

By the way, this is a good example of what I said above: But if Popper was not perfectly consistent and unambiguous in everything he wrote, then multiple valid interpretations are possible (although still not all interpretations would be valid—not everything goes). Someone who only reads Conjectures and Refutations could argue from the evidence in that book that Popper considers himself a falsificationist. Someone who only reads the 1982 Introduction to Realism and the Aim of Science, where Popper wrote "my views on science (sometimes, but not by me, called 'falsificationism')", could argue from the evidence in that introduction that Popper does not consider himself a falsificationist. But a historical contextualization of both texts would show: In 1962 Popper wrote that he counted himself among the falsificationists, and twenty years later he wrote that he did not call his views falsificationism. And a historical contextualization of both texts would show what happened in those intervening twenty years that may have caused Popper to change his position. That is an example of why the historical approach is so important: it gives a deeper view. Biogeographist (talk) 22:37, 13 January 2022 (UTC)

(edit conflict with the two paragraphs added on top of the preceding one) I disagree. It could be important if we were interested in details of the history. It's dealing with the confusion around the terms "falsificationism" and "falsificationists". One should not conflate this with a fundamental change in Popper's perspective. If you want to argue for a change in Popper's philosophy, you cannot do this by using the use of a term that has so different meanings. You would have to do that by explicitly explaining Popper's philosophy in 1959, not by cherry picking sentences there and there, then do the same in 1982. The only thing that you will see is that Popper adapted himself to the arguments presented by others. Of course, Popper's way to present his philosophy changed in view of what others wrote. However, you will not see a fundamental change in Popper's philosophy. Many philosophers fundamentally changed their position over time. For example, we speak of the earlier Russell and the later Russell. The same is true for Wittgenstein. They are actually proud of their change of position and it's very clear that they fundamentally changed their philosophical position. You don't have this in the case of Popper. What you have is some people such as Lakatos that speak of earlier Popper and later Popper, but it matches with a lack of understanding of the distinction between the logical side and the methodological side. When you don't appreciate the distinction, then Popper's evolution in his way to explain his philosophy appears to you as fundamental change. The key point here is that, no, absolutely not, you cannot speak of change in Popper's position in Wikipedia's voice. If it is relevant in the context, you could say that Lakatos spoke of Popper0, Popper1 and Popper2, but it has to be relevant and respect WP:due. Dominic Mayers (talk) 23:11, 13 January 2022 (UTC)

You wrote: I'm not advocating picking a sentence out of context! That's exactly the opposite of what I'm advocating! and How can we clear up confusions about how the term is used if we don't mention and explain such uses? That has to be part of it. As I said, it has to be relevant in the context. The context where it would be relevant is a section that deals specifically about the different, often conflicting, uses of the terms falsificationism and falsificationist. I see what might be your position. You might be thinking that people are perhaps confused by the different uses of these terms and, if we don't cover that, then they might misunderstand what we say and even think that we misrepresent the literature. However, there is a danger that we remain at the level of confusion when we deal with confusion. I think that the most important is to present the important concepts in a way that is verifiable without getting caught in terminological issues. It is important to primarily put our attention in communicating the key point of views as clearly as possible using terms that are not ambiguous given the context. This is what will help the most to remove confusion. Dominic Mayers (talk) 23:38, 13 January 2022 (UTC)

I'm glad you added that last paragraph, because after I read your first couple of sentences in this section I was thinking that you were an incredible hypocrite for deciding that Popper's use of the term was important enough for you to mention in the main text of Falsifiability § Falsificationism but completely unimportant to this article. Biogeographist (talk) 23:54, 13 January 2022 (UTC)

Now that you brought out that apparently Popper's forgot that he used the term 20 years earlier, then I think I should remove the line or few lines that discuss that. Honestly, I never considered this important. I kept it, despite the fact that I was uncomfortable with the fact that it discusses terminology, because it added some support to the view that Popper rejected Lakatos's way to describe his philosophy (which Lakatos did using the terms dogmatic, naive, etc. falsificationism). It was that rejection of an actual view that I considered important, not the rejection of a terminology. I was really uncomfortable with this, especially given that Popper is known to have said that he does not quarrel about terminology. This last point is in my opinion a greater contradiction. Dominic Mayers (talk) 00:15, 14 January 2022 (UTC)

Reusing the section Falsificationism in Falsifiability

This section was written the best I could many years ago. Falsifiability was in an horrible state when I started to work on it, especially regarding falsificationism. I believe my section was an improvement, but it's difficult to present in a single section what Lakatos wrote in a complete article. When, I started to translate it in French (which I did not do finally), I saw that the match with Lakatos description was not accurate. It's not bad, but still not exactly how Lakatos describes these different falsificationisms. I don't remember exactly what was the issue and whether I corrected it. Anyway, most likely, if I had to write it again, it would be different. Or may be I would not focus too much on the specific way Lakatos describes these different falsificationisms. The way I see it now is that Lakatos's falsificationisms (dogmatic, naive, sophisticated) present Lakatos misunderstanding of Popper's philosophy. For example, the dogmatic falsificationist of Lakatos is very much like the logical falsificationist in Popper's philosophy. However, in Popper's philosophy, this dogmatic or logical falsificationist still exists, whereas somehow Lakatos consider that Popper got rid of him. The naive falsificationist is the falsificationist that only consider the logical use of the logical falsificationist (given an empirical basis) on the methodological side. This is Popper when he says that epistemology should only consider the logical aspect, i.e., the Popper that says that biological or evolutionary aspects are not a part of epistemology. Again Lakatos considers that Popper is getting rid of the naive falsificationist when he discusses the role of metaphysical theories and metaphysical research programs and other notions that cannot be justififed logically. However, Popper always considered the distinction important. This naive falsificationist still exists and it was always there. The sophisticated falsificationist that Lakatos saw partially in Popper philosophy is the falsificationist that accepts the importance of metaphysical theories and research programs. This sophisticated falsificationist was there very early in Popper's philosophy. Certainly, the role of metaphysics was there in 1934. But, even metaphysical research program is as old as 1949. He wrote a footnote in Schilpp p. 175 to explain that he used the concept in lectures as early as 1949. The reason why Lakatos says that it's only a partial sophisticated falsificationism is because of the metaphysical part. He wanted very much a verifiable sophisticated falsificationism, verifiable in accordance with some new general inductive principle. He explains that very clearly. This is why, he explains, he uses the term methodological research program, not metaphysical research program. He proudly says that it distinguish him from Popper's limited sophisticated falsificationist. However, he did not succeed. So, Lakatos was never more than a sophisticated falsificationist that promotes the use of metaphysical research programs. Anyway, the key point is that Lakatos saw an evolution in time in Popper's philosophy, which was not there at all. They just represent different focuses of Lakatos on different aspects of Popper's philosophy that were quite stable over his entire carreer. In other words, for Popper, it was always the case that a scientist must be like the dogmatic falsificationist some times, like the naive falsificationist other times and as the (partial) sophisticated falsificationist yet other times. Popper did not explain that explicitly. Instead, Popper wrote that he did not want to enter into these terminological distinctions made by Lakatos and that Lakatos used them to misrepresent his intellectual history. Nevertheless, this analysis helps an understanding of the sources and comes very naturally when we do understand the sources. Dominic Mayers (talk) 01:18, 14 January 2022 (UTC)

I am not proposing to use this in the article. Just as a consideration of a source's sources has no place in an article, but is encouraged in a talk page to evaluate the source, this analysis is only offered to evaluate Lakatos paper and help us decide how to present it and see how relevant it is. Dominic Mayers (talk) 03:24, 14 January 2022 (UTC)

@Biogeographist: The context of my analysis of Lakatos's falsificationisms is that editors must use their understanding of a source to evaluate its relevance and the weight that it must have in different contexts. An example of this, different from the above, but still an example of the same general principle, is when editors consider the source's sources. They cannot refer to source's sources in the article, unless the author does it explicitly, but certainly editors can and should often consider the source's sources to better understand the author. Some people, even experienced editors, refer to this as "allowed original research in the talk page". I disagree with the use of this expression, because "original research" has a technical negative meaning in Wikipedia and anything that helps understand the source is best seen as opposed to original research, because when you don't understand the source as deeply as needed, you are very likely to do some form of original research. My point is also that original research should not be associated with people that claim or seem to believe that they know better. Original research can also and perhaps more often be done by editors who claim to have the most common view on a subject. If it is really the most common view and it is presented as such with a good understanding of the sources in respect of WP:due, then it is fine and it is not original research. Yes, but only when we can identify the sources for this most common view and have a good understanding of it in terms of these sources. My point is that the valid argument that we must write for a large audience cannot justify that we do not need to understand the sources and only present our limited understanding under the umbrella of the "most common view". Anyway, perhaps you agree with all of this and it was not necessary to expand on this. In terms of our specific case here, my point is that it is useful to understand that a lot in Lakatos's falsificationism is only Lakatos's perspective on Popper's philosophy and it cannot be presented in Wikipedia's voice and if we are in a very general context, as you wish to be, then I don't think it should have a lot of weight. I still think that an article that focuses on Lakatos's falsificationism would be justified and then, in that context, going into the details of Lakatos's falsificationism would be justified without violating WP:due. Dominic Mayers (talk) 11:24, 15 January 2022 (UTC)

Re: if we are in a very general context, as you wish to be, then I don't think it should have a lot of weight. I agree with that. Thanks also for everything else you said in this section, which is useful to think about, although I don't have any specific comment about it right now. I personally am not very interested in contributing to an article only about Lakatos, so I will continue to focus on preparing for a more general article. If you want to write an article on Lakatos, I would repeat that an article about his paper "Falsification and the Methodology of Scientific Research Programmes" would be more than justifiable and very welcome, just like there is an article about Popper's The Logic of Scientific Discovery, and indeed I think the latter article should also be expanded to summarize Popper's book in greater detail. Biogeographist (talk) 22:59, 15 January 2022 (UTC)

Difference between general falsificationism and Quine-Duhem thesis

As soon as we want to be general, we need to ask what is the difference between a large perspective on falsificationism and a broad coverage of the Quine-Duhem thesis, eventually larger than what is done in Quine-Duhem thesis? My point is that the Quine-Duhem thesis (the two versions as described by Lakatos) covers most of the problems of falsification. Both subjects are attached to the Vienna Circle. There is an important overlap. If you remenmber our previous discussions, you will see that I am not against overlap. I always argued that it's fine that almost the same subject is covered using different perspectives in different articles as long as it is not used to do POV fork, which is taken care of when the two articles refer to each other adequately, i.e., the readers can easily see the big picture. So, I am not asking the question in a negative manner. I do think, however, that we need to clarify what is the perspective specific to this article and how it is different from a perspective that would be natural for a broad coverage in Quine-Duhem thesis. One way to create a distinction would be to use a focus on terminology, but then I would disagree. Dominic Mayers (talk) 23:35, 15 January 2022 (UTC)

This brings us back to a previous discussion regarding the role of the Quine-Duhem thesis and the Duhemian problem in this article. At the conceptual level, I consider that it's a background knowledge that was essentially known by all parties when falsificationism was discussed in the 60's and also essentially known through the Duhemian problem even earlier when LSD was first written in 1934. In particular, I would oppose a description of Popper's philosophy that does not address the Duhemian problem followed by a criticism that is based on the Duhemian problem. Dominic Mayers (talk) 23:54, 15 January 2022 (UTC)

Do we even need an article on falsificationism?

@Dominic Mayers: Above you wondered how this article would be different from Quine-Duhem thesis. That got me thinking about how this article would be different from other articles as well, and I concluded that this article is not necessary. I would be happy with a disambiguation page at Falsificationism together with greater detail in the other relevant articles like critical rationalism, Quine-Duhem thesis, falsifiability, The Logic of Scientific Discovery, and a new article on Lakatos's "Falsification and the Methodology of Scientific Research Programmes". Biogeographist (talk) 13:00, 16 January 2022 (UTC)

I don't see the need for a disambiguation page if we have only a single article that uses the name "falsificationism". Disambiguation pages are not meant to help decide among conceptually related articles or to distinguish the use of a name in Wikipedia from other uses of this name outside Wikipedia. The purpose of a disambiguation page is lexical within Wikipedia pages, not conceptual within these pages, and it has to be used for Wikipedia pages. For example, once I wanted to add David Kaye, the Distinguished Professor of Law Emeritus at Penn State Law to distinguish it from another important lawyer under that name: the Special Rapporteur at the UN nation. It was reverted because the former did not have an article in Wikipedia. I was invited to create that article, if I wanted, and then add it in the disambiguation page. I know that you worry about the use of "falsificationism" in David Miller's book and elsewhere in the literature that presents it as Popper's philosophy, but we can use an About template at the top of Falsificationism to say that the article is about Lakatos's falsificationism and that for Popper's philosophy one should see Falsifiability and Critical Rationalism. Dominic Mayers (talk) 14:29, 16 January 2022 (UTC)

The Kaye example doesn't fit this situation; we saw above that falsificationism is used by Popper himself and by Miller (not to mention others) as basically equivalent to critical rationalism. Not yet mentioned is the fact that the 1989 edition of Conjectures and Refutations lists falsificationism in the index with a reference to "rationalism, critical". So it's perfectly appropriate that this disambiguation page points to Critical rationalism; all that is needed is to mention the Popperian usage of falsificationism (and related issues) in Critical rationalism. Biogeographist (talk) 15:11, 16 January 2022 (UTC)

Besides, the Falsificationism article should whenever possible in the light of verifiability cover attributions of falsificationism to Popper that are done in the same line of thought as Lakatos. Let's keep in mind that Lakatos's falsificationism was attributed to Popper by Lakatos himself. Therefore, the simple fact that an author attributes a falsificationism philosophy to Popper does not mean that it's not Lakatos's falsificationism. This is definitively a situation where there is a need to understand the source and not attach too much importance to the use of the term "falsificationism", because the term was used to refer to almost opposite concepts. Dominic Mayers (talk) 14:40, 16 January 2022 (UTC)

But the fact that Lakatos's falsificationisms are not Popper's is already explained in Falsifiability § Falsificationism and properly linked in this disambiguation page. So currently there is no problem. Biogeographist (talk) 15:11, 16 January 2022 (UTC)

I said currently there is no problem, but it would be better to also have a separate article on "Falsification and the Methodology of Scientific Research Programmes" that could be properly linked from elsewhere in the encyclopedia. I just don't think that Falsificationism should be about Lakatos. Biogeographist (talk) 15:27, 16 January 2022 (UTC)

Except in Miller's book or short entries in dictionaries that do not cover criticisms, Popper's philosophy is usually referred to as falsificationism to criticize it because it does not explain progress in science, it lacks a confirmation principle, etc., which are criticisms that apply to Lakatos's naive falsificationism. In the context of these criticisms, the fact that Popper's philosophy is a coexistence of three falsificationisms, a logical falsficationism, a form of naive falsificationism (when it separates what can be done logically from what must be done metaphysically) and a sophiticated falsificationism, is often not what is meant by Popper's falsificationism. Instead, there is a focus on the fact that naive falsificationism has no confirmation principle, etc., ignoring the fact that Popper discussed the importance of quasi-induction (supported by critical discussions), methaphysical principles and methaphysical research programs. Dominic Mayers (talk) 15:07, 16 January 2022 (UTC)

All of that should be covered in Critical rationalism, no? Biogeographist (talk) 15:11, 16 January 2022 (UTC)

(conflic edit with last reply) Most philosophers have criticized Popper by misrepresenting him and ignoring important aspects of his philosophy. I am not saying that Popper's covered every thing and cannot be criticized to "come short" on some aspects like Deborah Mayo pointed out. I would say that he has not explained enough the nature and role of metaphysical research programs. When we read Popper, we don't realize enough that the use of statistic in science is a part of metaphysical research programs: the construction of the edifice of science, all these interrelated statistical results that bring confidence in a theory are part of metaphysical research programs. Deborah Mayo says that science must let the data speak and it must do it through statistical theories and models and complain that Popper did not cover this aspect sufficiently. She wrote

the tests Popper sets out are white-glove affairs of logical analysis . . . it is little wonder that they seem to tell us only that there is an error somewhere and that they are silent about its source. We have to become shrewd inquisitors of errors, interact with them, simulate them (with models and computers), amplify them: we have to learn to make them talk.

Dominic Mayers (talk) 15:41, 16 January 2022 (UTC)

You wrote I just don't think that Falsificationism should be about Lakatos. If we only have an article that covers Lakatos's falsificationism (and name it "Falsificationism"), then it does not matter what we think, it would be a fact and, if there is a single article that uses the name in the title, there is no justification for a disambiguation page. In particular, having Critical Rationalism as an entry in a Falsificationism disambiguation page does not respect the normal use of disambiguation pages in Wikipedia, because a disambiguation page is not used to help readers decide among articles that are conceptually related. It must be used when lexically similar titles are used by different articles. I would not be so surprise that you can find examples where a disambiguation page is used for conceptually related pages with entirely different titles, but it's not the intended purpose and I am opposed to an abuse of disambiguation pages. If, as you suggest, no article uses the name "falsificationism" in its title, then we even less need a disambiguation page. It's better to present the results of the search engine or do a redirect to a page such as Critical rationalism or Falsification and the Methodology of Scientific Research Programmes and use an About template at the top of this article to present other uses. You seem to disagree with me that Falsificationism is a lot associated with Lakatos and in this I include the fact that, most of the times, when it is associated with Popper it is in the same line of thought as Lakatos, so it is still Lakatos's falsificationism. I know that you might feel that, nevertheless, most people still associate falsificationism with Popper, but most people make this association in a way that is similar to Lakatos's way in the following sense that they then view the Quine-Duhem thesis as a criticism of falsificationism. In other words, the term "falsificationism" is irreversibly compromised in the literature and I refuse to present this confusion, even if it is wide spread, in Wikipedia's voice. Creating a disambiguation page only to associate Critical Rationalism as a possible meaning of Falsificationism is already too much. It's way better to use an About template, because then it's part of the article. I would really prefer a redirect to Falsification and the Methodology of Scientific Research Programmes with an about template. This would do the maximum to compensate for the confusion that exists in the literature regarding falsificationism. Dominic Mayers (talk) 16:19, 16 January 2022 (UTC)

I have been doing a lot of research, and based on what I have found, which is a lot, I don't believe that falsificationism usually refers to Lakatos. I would be fine with redirecting Falsificationism to Critical rationalism, but of course there would be a lot of explaining to do in Critical rationalism for all the reasons you have outlined. Biogeographist (talk) 16:23, 16 January 2022 (UTC)

Please provide the most representative sources that made you think that "Falsificationism" is not used in Lakatos's way. It's not sufficient that Lakatos's is not mentioned and that Popper is. The only criterion that I will consider is what is the actual concept that is being described, especially by looking at criticisms. If the criticisms ignore the coexistence of different falsificationisms in Popper's philosophy, then it's exactly what Lakatos did and it's Lakatos's falsificationism. Dominic Mayers (talk) 16:30, 16 January 2022 (UTC)

You don't need to provide Miller's book. I consider it as a rare exception that uses falsificationism in Popper's way. Actually, even Miller, as I said before, tend to neglect the sophisticated falsificationist in Popper. The difference is that he does not use it to criticize him. Dominic Mayers (talk) 16:35, 16 January 2022 (UTC)

OK, I will provide sources that show that falsificationism is widely used in a non-Lakatosian way. It won't happen today, because I have to collect all the files on my computer, and today I have other things to do. Meanwhile, will you please provide sources that substantiate your claim that there are different falsificationisms in Popper's philosophy? It sounds to me like you just made up that idea, so I would like to see your sources that explicitly say that.

WP:D2D says "Disambiguation is required whenever, for a given word or phrase on which a reader might search, there is more than one existing Wikipedia article to which that word or phrase might be expected to lead." If there is no primary topic, then the term should be a disambiguation page. I can show that Falsificationism could be expected to lead to Critical rationalism, so a Lakatosian topic is not primary. However, if it can also be shown that Critical rationalism is not primary, then there needs to be a disambiguation page. Biogeographist (talk) 16:42, 16 January 2022 (UTC)

Ok, reading formally WP:D2D, you are correct that a disambiguation page can list articles with completely different names. However, almost all examples provided in WP:D2D are not like that. In fact, even in the rare examples provided where an article of a completely different name is listed it's not like in our case. In our case, the confusion exists even conceptually, whereas in the few example provided of a completely different name being used in a disambiguation page there is no ambiguity at the conceptual level: the problem is lexical. I really don't think a disambiguation page is appropriate when the confusion is conceptual. It's not the place to address indirectly a conceptual confusion. In fact, it can contribute to exacerbate the confusion, which I feel is what will happen in our case. Dominic Mayers (talk) 17:22, 16 January 2022 (UTC)

I asked myself if there is another example that is closer to this one, and the word Universalism popped into my head as a possible example. It is a weird article. In Universalism § Philosophy there are two summary sections with links to main articles, then there is a long Universalism § Religion section with descriptions of very different ideas, and then there is Universalism (disambiguation) with still more links to other articles: note that Open individualism is listed there, which has no lexical similarity to "Universalism". Biogeographist (talk) 17:43, 16 January 2022 (UTC)

You wrote Meanwhile, will you please provide sources that substantiate your claim that there are "different falsificationisms in Popper's philosophy"? I already explained in one or two paragraphs above that I consider this as part of an understanding of the source, not something that I plan to include in any article. Please read Lakatos description of his different falsificationisms, especially dogmatic vs naive, and then read Popper insistence that we must separate logic and methodology, if you don't see that the dogmatic falsificationist described by Lakatos with

If a theory is falsified [in the usual sense], it is proven false; if it is falsified [in the technical sense], it may still be true.

— Imre Lakatos, Lakatos 1978, p. 24

match pretty well with the always existing logical side in Popper's philosophy, then I am at a lost. The issue, if you understand that there is a good match, is that Lakatos says that this dogmatic falsificationism is only an early stage of Popper's philosophy, whereas Popper insists that it is an essential part of his philosophy used to define falsifiability as a purely logical criterion. Dominic Mayers (talk) 17:39, 16 January 2022 (UTC)

But you're conflating Lakatos's ideas with Popper's ideas, which is exactly what I thought you didn't want to do? Popper never said there were different falsificationisms in his philosophy. Biogeographist (talk) 17:43, 16 January 2022 (UTC)

Yes, Popper did not say that there are different falsificationisms in his philosophy. I never claimed that he said so. However, Popper description of his philosophy does include different sides or aspects. He certainly said that there is a logical side and a methodological side. I am just saying that the logical side matches very well with the dogmatic falsificationism of Lakatos–please read the source it's very clear that Lakatos refers to Popper's logical sidse, but a key difference is that in Lakatos it is only an early stage whereas in Popper philosophy, it is fundamental and always needed. The terminology does not matter as long as you understand what is meant. Well, the terminology does matter as a way to communicate an understanding of the source, but not more than that, because there is no plan to use this terminology in the article. In fact, I don't think it is a good terminology, because we do not normally use "falsificationism" (or any "ism" for that matter) to refer to an aspect or side of a philosophy. Sorry, if this confused you a bit. It was only a way to say that the different falsificationisms in Lakatos correspond to aspects of Popper's philosophy. Dominic Mayers (talk) 17:52, 16 January 2022 (UTC)

I understand what you meant now. Biogeographist (talk) 18:27, 16 January 2022 (UTC)

You wrote I asked myself if there is another example that is closer to this one, and the word Universalism popped into my head as a possible example. I looked at Universalism (disambiguation) and indeed Open individualism is an example of a completely different name with no lexical connection with Universalism. But my point went further than that. Unless, there is a possible confusion between Open individualism and another article in the disambiguation page at the conceptual level–i.e., an incorrect conflation of two distinct concepts, it's not the same thing as in our case. The existence of a conceptual confusion makes an important difference. I don't think a disambiguation page does a good job in such a case. Besides, even if you found a few isolated example that matches very well with our case, including the conceptual confusion, I don't think it would be a strong argument to justify that we do the same. It would still not be the most standard use of a disambiguation page. Dominic Mayers (talk) 18:15, 16 January 2022 (UTC)

I don't think a disambiguation page does a good job in such a case. I could agree with that, but that would be just as true of disambiguation in a hatnote, so more generally no simple disambiguation would do a good job in this case. So the answer to the question in the heading of this section would be: Yes, we need a general article on falsificationism to clear up conceptual confusion. Biogeographist (talk) 18:27, 16 January 2022 (UTC)

I feel a hatnote has less chance to exacerbate the confusion, because it creates a bigger contrast between the different concepts and also, especially if we use Falsificationism vs Critical rationalism, often the reader will directly get to the Falsificationism article instead of a page on Popper's philosophy. If we use Falsification and the Methodology of Scientific Research Programmes instead of falsificationism, it will still happen very often that the user will get to this article instead of Critical rationalism or Falsifiability and then the hatnote will clearly explain the distinction. What I have in mind is the following. I consider that many people are mislead by the attribution of a Lakatos-like falsificationism to Popper. Having Wikipedia attach the term "falsificationism" to Popper by the use of a disambiguation page can exacerbate the confusion. I know that you said that the confusion will be removed when the readers will read the articles, but meanwhile a deliberate attachment of falsificationism to Critical rationalism only exacerbate the confusion. In contrast, having the readers get to the article on Lakatos's falsificationism when they use the search term "falsificationism" will immediately bring them to question themselves about this confusion and, because we provide a hatnote, we do not mislead them in any way. We only help them, even before they read the articles. It's better when it's done through the organization of the articles (hatnotes, redirect, etc.) as well as through the content of the articles. Dominic Mayers (talk) 18:56, 16 January 2022 (UTC)

OK, I think I can better explain my position. It's not a symmetrical situation of the following kind: some people associate falsificationism to Popper with a correct understanding whereas others associate it with one of Lakatos's falsificationism with a correct understanding. If it was symmetrical like that, then my argument could be reversed by symmetry (by considering the other people). The fact that Popper's philosophy requires a coexistence of many aspects or sides makes it conceptually more challenging. The understanding that a typical reader has of falsificationism is very unlikely to be Popper's falsificationism with all its required coexisting aspects and the distinctions between them. Without these distinctions between logical, methodological, etc. which are coexisting aspects, Popper's philosophy is not so different than Lakatos's falsificationism. Therefore, it is safe to assume that most readers associate implicitly, but incorrectly, a Lakatos-like falsificationism to Popper. I want an organization of the articles that challenges this misunderstanding in the most immediate manner while providing all the information (with hatnotes, etc.) so that the readers are well informed and not mislead. Dominic Mayers (talk) 19:25, 16 January 2022 (UTC)

In contrast, having the readers get to the article on Lakatos's falsificationism when they use the search term "falsificationism" will immediately bring them to question themselves about this confusion: I am not sure that is true, but we could be empirical about it and test that option. (Falsificationism had basically 8 average daily pageviews before the end of December, so the stakes are pretty low here; we can afford to experiment. The most likely outcome here is that nobody actually uses the Falsificationism redirect anyway.) Like I said, I am not very interested in contributing to an article only about Lakatos, but I guess I would be willing to accept redirecting Falsificationism to "Falsification and the Methodology of Scientific Research Programmes" as an experiment, and watch the page and see what happens over a few months. Biogeographist (talk) 19:28, 16 January 2022 (UTC) and 19:57, 16 January 2022 (UTC)

Or we adopt the idea of "general" article on "Falsificationism", but with a removal of any confusion regarding Popper's falsificationism, Lakatos's falsificationism, etc. as its main emphasis that distinguishes it from other articles such as Quine-Duhem thesis. My only concern is that directly dealing with confusion might not succeed. This is why I feel it can be simpler to present plainly Lakatos's falsificationism that makes it clear that he considers one falsificationism at a time. Ideally, we would have a source that says that these falsificationisms correspond to different aspects of Popper's philosophy. It might not exist, but what might be possible is that when Lakatos describes his dogmatic falsificationism, he actually refers explicitly to Popper's description of it's logical side. We would just have to let the readers see the obvious by themselves. Anyway, the idea is to not deal with confusion by talking about it, but simply remove it by describing each view point so that the differences show up clearly and then use hat notes in a way that challenges in the most immediate manner the most likely confusions. Dominic Mayers (talk) 19:53, 16 January 2022 (UTC)

OK, I see two options that I consider possible:

• Direct to a general page on falsificationism (with that name) with an emphasis on removing confusion, which distinguishes it from other articles that cover broadly issues with falsifications.
• Redirect to a page or a section on Lakatos's falsificationism

I have no problem with the first option, except that dealing explicitly with confusion within an article might not be so easy and efficient. The idea is to challenge in the most direct manner the most likely confusions using redirects, hatnotes and articles that present point of views in the clearest manner. Most people that are redirected to a page on falsificationism that is clearly not Popper's philosophy will be challenged, but it's good. We want that. The point is that it's often not true that they actually want to see Popper's philosophy. More often, they might want to know more about falsificationism as presented in some texts and it's not truly Popper's falsificationism, even if the author claims or suggests it. We want to make sure the misunderstanding that it's Popper's philosophy is being challenged. The hatnote will make sure that the readers are not mislead and are well informed. I understand that we could put a hatnote at the top of an article on Popper's philosophy and redirect there, but I find it a bit artificial. Normally, someone that searches for falsifiability or critical rationalism knows what he wants. It's clear and there should be no need for a hatnote. The hatnote is way more natural on top of the article about falsificationism, because this is the term that is confusing. Because this is where the hatnote is natural, it is also where the redirect should go. Dominic Mayers (talk) 20:41, 16 January 2022 (UTC)

Another related claim that I think is true is that Lakatos is generally associated with research programs as opposed to falsificationism. This is what the current article Research program says. Paul Feyerabend, for example, also said this explicitly in Against Method where he contrasted falsificationism versus inductivism versus research-programmism, where research-programmism corresponds to Lakatos's view. Nobody but Feyerabend uses the term research-programmism, but his term reflects a widespread association of Lakatos with research programs as opposed to falsificationism. The best option is still a general page on falsificationism focused on removing confusions. Biogeographist (talk) 17:04, 18 January 2022 (UTC)

Given that terms have different meanings for different people, we can question the usefulness of focusing on these associations. The strategy I prefer is to make sure the different point of views are presented as clearly as possible while making sure that the terminology that we use is clear from the context. In this appraoch, we avoid the general question of what term is associated to a philosopher. We don't try to address the question of terminology globally. We present the concepts, which exist beyond terminology, and only make sure the terminology used is clear in the local context. Yes, we want to react to a confusion around a term, but by challenging the most likely confusion with redirect and hatnotes, not explicitly in a text. Dominic Mayers (talk) 19:10, 18 January 2022 (UTC)

To be clear, my previous comment was aimed at the question in the heading of this section, not a proposal for the content of the article. It was a reason why Falsificationism should not redirect to an article on Lakatos. I still intend to present sources that show that Falsificationism would not be expected to redirect to an article on Lakatos, i.e., the topic is not primarily Lakatosian as per WP:D2D. Biogeographist (talk) 19:55, 18 January 2022 (UTC)

If your objective is that falsificationisms redirects to an article with a scope that matches with all meanings of the term, then this article will lack a clear focus. I know that we considered the idea that the scope would be the confusion as a way to get a clean scope, but I feel it's giving too much importance to the confusion and worst it might not succeed. It's not the standard Wikipedia's approach. In Wikipedia, we simply present the different point of views as clearly as possible in respect of due weight. So, we explain falsifiability (mentioning lakatos's criticism in the form of falsificationism while respecting WP:due). We also explain Critical Rationalism. We also explain what Lakatos did. There is no POV fork here, because the articles will be linked and there will be some overlap as needed so that due weight is respected and the readers will get the big picture. The task of redirecting "falsificationism" should not determine how we will organize the articles. We should not create an article only to resolve our problem where to redirect falsificationism. My approach is instead to organize the articles as if the term "falsificationism" did not exists, i.e., we organize the articles in terms of the actual point of views, which exist independently of the term "falsificationism". Here I imply that point of views around terminology are not important. I understand that, from your perspective, this seems to create a big problem for the redirection of "falsificationism", because we would not know where to redirect it. It seems that no article would be a good match. Again, a disambiguation page is not meant to address this kind of issue. My approach, and I know you don't like it, is to redirect Falsificationism to one of the many perspectives around it and use an hatnote so that the readers can readjust themselves. My idea is that this perspective is the one that was used by Lakatos. It's a perspective that covered different kind of falsificationisms, but it still give a clear focus on the article. By the way, when considering Lakatos's perspective, we should not put too much attention on whether or not he supported falsificationism. This would be too much focus on the term "falsificationism", because it's a question that cannot be answered without considering the different meanings of the term. This is why I say that it's much simpler to simply explain Lakatos's perspective without worrying too much about terminology globally. It's sufficient to make sure that locally the meaning of the term and other terms are clear from the context. Terminology is a tool to discuss concepts that exist independently of any particular choice of terminology. We should not create new concepts or questions around terminology. Dominic Mayers (talk) 02:39, 19 January 2022 (UTC)

So the task is the following. We must find a scope for an article where falsificationism will redirect, but the scope of the article should make sense irrespective of the term falsicationism. We don't create an article, raise questions to be answered, find confusions to be removed, etc. around the term falsificationism. I don't think this article should be falsifiability or critical rationalism. The choice that comes to my mind is the article around Lakatos's perspective on the role of falsifications, which is a lot about Popper's philosophy. It's an offspring of Falsifiability and critical rationalism. Even the notion of research programs is attributed to Popper by Lakatos. His goal was to make the notion more rigorous. Popper realized the importance of research programs, but for him we don't have much to say about them, because they are metaphysical. Most people blame Popper for focusing too much on falsfications, but at the same time nobody succeeded to say anything more rigorous than Popper said about a confirmative methodology. Popper's confirmative position was an evolutionary one, as he explains in "All life is problem solving". It's not very rigorous. It's only saying that life progresses through what we call an evolutionary process and science progress in a similar manner. Falsifications play the role of problems in life. In the same way that, of course, the problems in life cannot explain progress of life in themselves, falsifications also, of course, cannot explain progress in science by themselves. However, Deborah Mayo has a good point that there is perhaps more to be said about how the logical side is constructed using statistical laws, data models, etc. It does not contradict Popper's philosophy, but it says that it's not the end of the story. I fell that the link of Popper's philosophy with AI is also particularly challenging. It's not studied much. Seeing a contradiction there would also be a mistake. Dominic Mayers (talk) 03:20, 19 January 2022 (UTC)

OK, what you have said sounds reasonable enough, so why don't you take over and do whatever you want to do to resolve this issue. If I think there is a problem with whatever you decide to do, I will say something, but whatever you decide to do will probably be fine. Biogeographist (talk) 17:35, 20 January 2022 (UTC)

Another confusing expression is "auxiliary hypothesis"

I was looking at Popper's view on the Duhemian problem. This problem is often explained in terms of auxiliary hypotheses defined as the hypotheses that are required for the experiment to work as planned. An example of auxiliary hypothesis that uses this definition is the hypothesis that the telescope is properly designed and the associated theory of observation is correct. This is the way the expression "auxiliary hypothesis" is used here and here. With this definition, the Quine-Duhem thesis becomes the statement that a required auxiliary hypothesis can be false instead of the theory. However, this is not the way the expression "auxiliary hypothesis" was usually used by Popper and Lakatos and perhaps many others in the Vienna circle. Lakatos used the expression 'ceteris paribus clause' instead of auxiliary hypothesis to refer to these required hypotheses. For Lakatos and Popper, most of the times, the expression auxiliary hypothesis refers instead to a modification of the theory or of the background knowledge, but this is different. Let us use "explanatory auxiliary hypothesis" to indicate that we refer to this other meaning (I just coined this expression for the purpose of this discussion). In this case, the explanatory auxiliary hypothesis is added to protect the theory against falsification. It's different, because the explanatory auxiliary hypothesis must be true to save the theory (by explaining away the falsification), whereas a ceteris paribus clause must be false to save the theory. Perhaps the idea is to turn the negation of a ceteris paribus clause into an explanatory auxiliary hypothesis. For example, an explanatory auxiliary hypothesis would be that the theory of the telescope is incorrect and must be modified. The explanatory auxiliary hypothesis would even go further and provide the new theory of the telescope. The situation gets even more confusing when we realize that there are two categories of ceteris paribus clauses and thus two corresponding categories of explanatory auxiliary hypotheses. A ceteris paribus clause might not be about any theory of observation, but only about the initial condition. For example, the clause that the telescope is not defective is not the same as the clause that the theory of the telescope is correct. Correspondingly, the explanatory auxiliary hypothesis that the given telescope was defective is not the same as a new theory for how telescopes fundamentally work. Dominic Mayers (talk) 12:40, 18 January 2022 (UTC)

There are various views on how the structure of theories is best modeled, and there may be different optimal models of theories for different domains. For example, Ryan Samaroo wrote about this in a recent chapter about Michael Friedman's philosophy of science, arguing (in the context of general relativity) that while Friedman tried to create a position distinct from Quine's holism, Friedman's position wasn't distinct enough from Quine:

But in spite of Friedman's work to restore the idea that conceptual frameworks of physics are stratified, his inclusion of mathematical principles in the category of constitutive principles is a step in the direction of Quine's centrality: it undermines the application of the factual–nonfactual distinction to different components of our conceptual frameworks. I have argued that those principles that define and interpret basic theoretical concepts should be distinguished from the formal prerequisites or auxiliaries that the principles presuppose, and all of these principles and prerequisites, which together constitute frameworks of empirical investigation, should be distinguished from the empirical hypotheses whose formulation they permit. This allows us to better recognize the salient differences in methodological character. In particular, separating mathematical auxiliaries, on the one hand, from coordinating principles and empirical hypotheses, on the other, allows us to distinguish the factual from the nonfactual components of our theoretical frameworks.

— Ryan Samaroo in Samaroo 2020

I think it is very helpful when talking about the structure of theories to use some kind of schematic diagram (graphic) to illustrate the relevant concepts. Samaroo does not give a diagrammatic illustration of his proposed structure, but such a diagram would have been helpful. For example, Boris Grozdanoff constructed a table of the concepts in Michael Friedman's philosophy of science that helps clarify the structure of the concepts, although even that table is not detailed enough in my opinion. I am often astonished that philosophers can write long books expounding complex conceptual structures and fail to provide a single schematic graphic illustration of the concepts. (By the way, one thing I like about Mario Bunge's books is that he provides plenty of diagrams.) Biogeographist (talk) 18:25, 18 January 2022 (UTC) and 18:27, 18 January 2022 (UTC)

I haven't read the papers that you mention, but, even without having read them, I agree that there are different ways to take an abstract perspective on scientific theories. The reason I agree is that my personal research these days involve a new kind of logic to explain paradoxes in quantum mechanics, which is, of course, a very (almost considered obsolete) old idea, but I do it differently. Anyway, there are so many kinds of logics, different kinds of type theories, etc. and this provide a useful context to describe different ways to abstract the process of doing science and focus on some important issues rather than others in that process. However, despite my agreement with the general idea, I don't see the tight connection between this and the specific confusion around the term "auxiliary hypothesis" that I mention. I know that the term is used in the context of the Quine-Duhem thesis or more generally the Duhemian problem and that we can take different perspectives, but it's making us lose focus to respond to a problem mentioned in one perspective by looking away from the perspective. It could be that a new perspective bring some light to an existing problem, but it does not make the original perspective wrong and by itself it does not address the valid issue as seen in this valid original perspective. The difference between these perspectives often only corresponds to an abstraction of different aspects. Of course, an abstraction, i.e., the theory that results from the abstraction could be wrong, but I don't think it's the case with the basic Duhemian thesis. (The extension that removes the analytic-synthetic distinction is more questionable.) Besides, since these theses are not physical theories, but theories about how we represent or should represent physical theories, the "empirical basis", if we can call it this way, is even more like a swamp than in the case of the physical empirical basis and therefore it's even more difficult to claim that it could be wrong. So, I see your comment as a diversion from the point that I raised, though I insist that what you wrote is interesting, perhaps even more interesting, but still a diversion. Dominic Mayers (talk) 12:55, 19 January 2022 (UTC)

OK I have read the first two sections of Samaroo 2020. I cannot judge whether the entire philosophy of Friedman is similar to Popper's philosophy, but I immediately noticed a very strong resemblance in terms of the structure of scientific theories and their relation with observations. The three levels in Friedman (as described by Samaroo) correspond respectively to the logical side, the material requirement and the methodological side. The resemblance was so striking that I was surprise to see that neither Samaroo or Friedman mention Popper. So, I decided to search the literature to see if others have noticed a connection. It is actually a subject of debate in the literature. For example, see Hoyningen-Huene 2017. So, I am not alone to see a strong connection between Popper and Friedman. However, the resemblance noticed in the literature, at the least in the result of my search, is not the one that I noticed. In the debate, the main argument of those who reject the connection seems to be around the notion that Popper is against instrumentalism, but not Friedman. I don't understand this argument because the kind of instrumentalism that Popper rejected is very unlikely to be the kind of instrumentalism that Friedman endorsed. It might be a terminology issue. Popper could have used "instrumentalism" in different contexts at different times with different meanings. The context that I consider is the instrumentalism that he discusses in section 3 of LoSD. It makes no sense to me that Friedman would be that kind of instrumentalist. Anyway, it's another example of a possible confusion around terminology and this is the point of this subsection. BTW, I am still a bit at a lost regarding your objective in mentioning this paper of Samaroo. Dominic Mayers (talk) 20:21, 19 January 2022 (UTC)

In Caldwell 1994, Preface to revised edition Caldwell discusses Friedman's instrumentalism and I don't see it as an instrumentalism that Popper would have rejected: Popper would not have liked the notion that metaphysical statements could be false and yet be useful, but only because rigorously speaking we cannot know that the metaphysical statement is false. Except for that, the instrumentalism of Friedman, as described by Caldwell, is very much in line with Popper's philosophy. Dominic Mayers (talk) 15:11, 20 January 2022 (UTC)

That article by Paul Hoyningen-Huene that you found is about the economist Milton Friedman, a very different guy, but an easy mistake to make. I didn't look at the Bruce Caldwell book, but that is probably Milton Friedman too, since it is about economics and it was published before Michael Friedman published his major works.

When I brought up Samaroo's comments on Michael Friedman, I was free-associating to something I had recently read. Carl Hempel was one of Friedman's teachers, and Friedman's analysis of scientific theory-change is very much in the tradition of the 20th-century philosophers (both logical empiricists and Popperians) who were doing what Rudolf Carnap called "the rational reconstruction of science". I was wondering how Friedman's account might be useful for putting auxiliary hypotheses in a larger analytical framework. Friedman is really insistent that Quine's view is insufficient. But I agree that I was probably "barking up the wrong tree" there.

A more useful reference for putting auxiliary hypotheses in a larger analytical framework (not the only way to do it, but one way) may be an article I once read by Ian Hacking (Hacking 1992). In that article, Hacking classifies laboratory science into 15 elements, in three groups (1–5 are "ideas", 6–10 are "things", 11–15 are "marks"). I have to list all the elements below so that his subsequent remarks on auxiliary hypotheses will make sense:

1. Questions
2. Background knowledge: "In what is so often called theory we should distinguish at least three distinct kinds of knowledge about the subject matter of the experiment. The divisions (2, 3, and 4) that I propose are sharp in some disciplines and vague or almost nonexistent in others. First is the background knowledge and expectations that are not systematized and which play little part in writing up an experiment, in part because they are taken for granted." (Hacking 1992, pp. 44–45)
3. Systematic theory
4. Topical hypotheses: "are part of what in physics is commonly named phenomenology. Because that term has another meaning in philosophy, and because it can also be used for (5), we want another name. We are concerned with what connects systematic theory to phenomena. Logical empiricism, with its strong emphasis on language, spoke of bridge principles (Hempel 1966, 72–5). The name is attractive, although 'principles' suggests something that cannot readily be revised, whereas we are concerned with what is revised all the time in laboratory work." (Hacking 1992, pp. 45)
5. Modeling of the apparatus: "There are theories, or at least background lore, about the instruments and equipment listed below as (6–8). To avoid ambiguity I shall speak of the (theoretical) modeling of the apparatus, an account of how it works and what, in theory, it is like. We are concerned with phenomenological theory that enables us to design instruments and to calculate how they behave." (Hacking 1992, pp. 45)
6. Target: "This together with elements (7)–(10) comprises the matériel of the experiment." (Hacking 1992, pp. 46)
7. Source of modification: "There is usually apparatus that in some way alters or interferes with the target." (Hacking 1992, pp. 46)
8. Detectors "determine or measure the result of the interference or modification of the target." (Hacking 1992, pp. 47)
9. Tools
10. Data generators
11. Data
12. Data assessment
13. Data reduction
14. Data analysis
15. Interpretation "of the data demands theory at least at the level of background knowledge (2), and often at every other level, including systematic theory (3), topical theory (4), and apparatus modeling (5). Pulsars provide an easy example of data interpretation requiring theory: once a theory of pulsars was in place, it was possible to go back over the data of radio astronomers and find ample evidence of pulsars that could not have been interpreted as such until there was theory." (Hacking 1992, pp. 49)

After making those 15 distinctions, Hacking then has a section titled "Extending Duhem's Thesis" that may be worth quoting at length:

Duhem (1906) observed that if an experiment or observation was persistently inconsistent with theory, one could modify theory in two ways: either revise the systematic theory (3) or revise the auxiliary hypotheses (in which we include both topical hypotheses [4] and modeling of the apparatus [5]). His classic example was astronomy, not a laboratory science, but the message was clear. Should a theory about the heavens be inconsistent with data, he said, we may revise astronomy, or modify either the theory of the transmission of light in space or the theory of telescope (5). But that is only the beginning of the malleability of my fifteen elements. For example, we can try to modify the telescope or build a different kind of telescope. That is, try to save the systematic hypothesis by adapting the detector (8).

Several recent contributions help to enlarge the Duhemian vision.... Duhem, Pickering, and Ackermann point to interplay among several subsets of the elements (1)–(15). Pickering attends to the modeling of the apparatus and the working of the instruments: we acknowledge data as data only after we have gotten handmade apparatus to work in ways that we understand. Duhem emphasized the intellectual elements (1)–(5). Ackermann, observing that data can be understood in many ways or not at all, put the emphasis on a dialectic involving theories and interpretation, regarding instruments and the data that they produced as fixed points. We should learn from all these authors. Let us extend Duhem's thesis to the entire set of elements (1)–(15). Since these are different in kind, they are plastic resources in different ways. We can (1) change questions; more commonly we modify them in midexperiment. Data (11) can be abandoned or selected without fraud; we consider data secure when we can interpret them in the light of, among other things, systematic theory (3). But it is not just Ackermann's interpretation of data by theory that is in play. Data processing is embarrassingly plastic. That has long been familiar to students of statistical inference in the case of data assessment and reduction, (12) and (13). Because statistics is a metascience, statistical methodologies are seldom called into question inside a laboratory, but a consultant may well advise that they be. Data analysis is plastic in itself; in addition any change in topical hypotheses (4) or modeling of the apparatus (5) will lead to the introduction of new programs of data analysis....

The truth is that there is a play between theory and observation, but that is a miserly quarter-truth. There is a play between many things: data, theory, experiment, phenomenology, equipment, data processing.

— Hacking 1992, pp. 52–55

Hacking says some other relevant points, but that is more than enough already. The point here is that the different kinds of "auxiliary hypotheses" are very easy to understand when they are (1) clearly internally differentiated/classified and (2) located in a larger analytical framework. And all of this would be even clearer if it were illustrated with a graphical diagram. Biogeographist (talk) 17:35, 20 January 2022 (UTC)

I understand that many things might explain the falsification besides the falsity of the "systematic theory" and that it might be useful to classify them. I would say, though, that separating "systematic theory" from "topical hypotheses" is a bit artificial. It would be like saying that "All swans are white" might be correct despite the existence of an observation of what is usually called a black swan, because "white" could be interpreted to mean what is usually referred to as "black or white". In other words, the systematic theory usually includes the topical hypotheses required to interpret it concretely in terms of "phenomena". By the way, a systematic theory corresponds to a theory in the logical side in Popper's philosophy and the need for topical hypotheses correspond to Popper's material requirement. This distinction in 15 elements might seem much more complete than Popper's way to analyze science, but it's superficial. For example, these distinctions fail to bring out the importance of the theory-ladden nature of observations. In Popper's philosophy, nothing "connects systematic theory to phenomena", unless by "phenomena" we means other theories: Popper accepted, as did many other philosophers before him, that statements can only be connected with statements. This is why Popper says the empirical basis is like a swamp: we have an edifice of statements pilling up into a swamp and we never reach any solid rock. For Popper, the material requirement implies that, though we accept by conventions not to go deeper into the swamp, it should always possible to do so, if requested. I don't think he meant that we always know the theory needed to go deeper into the swamp. He meant that it's understood that the theory exists in principle, the point being that we never need to connect statements with anything else than with statements, but simply we stop going deeper into this edifice of statements by convention. I do not mean that Hacking is not aware that statements can only be connected with statements. I am only referring to the danger of seeing these distinctions as providing the big picture. Popper's philosophy is that it's the fact that this edifice does not fell apart despite our attempts to falsifications that is "convincing", but it's not convincing logically, because there is no foundation. Logically speaking, the edifice is absolutely baseless and not verifiable at all. The part that Popper does not explain and that nobody else explain is why the "swamp" makes the edifice very solid, as we experience it. Popper does not even suggest what might be this magical swamp in which somehow the edifice is grounded. Instead, he refers to our acceptation of the evolutionary nature of life. OK, but all of this, not only my last point regarding the theory-ladden nature of observation, is again a diversion if the goal is to understand the confusion regarding "auxiliary hypothesis" vs "ceteris parebus clause" that I raised, this confusion being not very deep in comparison and understandable irrespective of all these distinctions. Dominic Mayers (talk) 19:28, 20 January 2022 (UTC)

(conflict with the reply below) But I believe that I now understand that you took my point as a pretext to bring a related point that concerned you regarding the Duhemian problem. Your point being that if we try to say more precisely what can explain away the falsification, there are many distinctions to take into account. Dominic Mayers (talk) 19:40, 20 January 2022 (UTC)

Re: these distinctions fail to bring out the importance of the theory-laden nature of observations. That impression is probably due to what I have excerpted from Hacking's chapter. Notice that none of the elements is "observation". Element 15 is another name for theory-ladenness, for example. I could say more but it would probably be a diversion as you said. Biogeographist (talk) 19:38, 20 January 2022 (UTC)

As I said, I did not meant that he was not aware of it. I will not argue against the fact that it is mentioned only in the "marks" category when it is most pertinent when he explains the concept of "topical hypothesis". I have a more serious concern here. Putting too much attention to all these distinctions defeat the purpose of abstracting from a concrete problem at hand to focus on some aspects and not others. For example, consider mathematical logic, which only concerns itself with the meaning of logical connectives. It usefully abstracts away from the details of mathematics to ask questions at this level of abstraction. It will be weird to reply that many aspects are not being considered. In the same way, I believe the Duhemian problem is best stated at some level of abstraction. I don't think that the distinctions that are raised by Hacking are useful to better explain the Duhemian problem. They are more like raising issues about the level of abstraction at which the problem is explained, but that is true for any use of abstraction. It's not specific to the Duhemian problem. Dominic Mayers (talk) 20:09, 20 January 2022 (UTC)

Hacking used the heading "Extending Duhem's Thesis", so I assume he would grant that he is not addressing exactly the same problem as Duhem. So I see now how this is not the problem you wanted to point out. Your heading Another confusing expression is "auxiliary hypothesis" meant something broader to me than it does to you. I don't find the "ceteris paribus clause" concept to be confusing. Sorry that I brought up something that is not relevant. Biogeographist (talk) 20:37, 20 January 2022 (UTC)

I am glad that you are not confused by two completely opposite meanings of "auxiliary hypothesis" in the literature: its use in Popper and Lakatos and its use here and here. In one case, it means something that must be true to save the theory. In the other case, it means something that must be false to save the theory. It's not a deep point at all. It's a terminological point. Perhaps in this case, the two meanings are so opposite logically, that it could not be the basis for a serious confusion. In that sense, I admit it was a bit out of context, not related to the confusion that exists around "falsificationism". I felt compelled to mention it anyway, because until after I realized these two different uses in the literature, I would not understand why I was uncomfortable with the actual meaning of "auxiliary hypothesis". Yet, I knew in the back of my mind that it was a bit out of context, unrelated to the confusion regarding "falsificationism". So, I don't blame you to have taken this as an opportunity to discuss something else. However, even if you had discussed this other point in its own section, I would still have made the point that it's best to explain the Duhemian problem by abstracting away from these detailed distinctions. The distinction between the "Systematic theory" (taken with its interpretation, which must exists by the material requirement) and every other assumptions needed for a test of the theory to work as planned is sufficient to explain the Duhemian problem. These auxiliary hypothesis (in the sense used here and here) can be further distinguished in different categories, but unless there is a specific point that you want to make that needs more precision, we can abstract away from these distinctions in our discussion. This is why I say that I do not know where you want to go with this. I also took this opportunity to explain that these distinctions have not been ignored by Popper and others. He and others made these distinctions when needed without referring to them explicitly and, in fact, often went much further than these distinctions suggest. Dominic Mayers (talk) 21:45, 20 January 2022 (UTC)

Re: unless there is a specific point that you want to make that needs more precision, we can abstract away from these distinctions and He and others made these distinctions when needed without referring to them explicitly and, in fact, often went much further than these distinctions suggest: Yes, to construct an adequate analytical framework one first has to clarify the problem or question at hand. (I imagine that is why "questions" is number 1 in Hacking's list of elements above.) Hacking considered his framework to be good enough for his questions or purposes in that chapter, and no doubt other thinkers, such as Popper and Friedman, thought their analytical frameworks were good enough for whatever questions and purposes they were pursuing in a given work. I am certainly not suggesting that Hacking's framework would be adequate for all purposes.

I think we're done on this talk page, since above I invited you to redirect Falsificationism to wherever you want. Since there is no longer going to be an article here, there is no longer any problem or question at hand here. Biogeographist (talk) 00:38, 21 January 2022 (UTC)

I do not question whether these distinctions were used by Hacking. I asked what was the purpose of this here? It was not a negative question, but I am fine with not pursuing this. However, I want to clarify one thing. Having you given up on working on your ideas is not my objective. This is not at all the expected outcome of this discussion. So, I need to step back and see what is happening here. I want to understand what is the knowledge in the literature that you feel should be included somehow in Wikipedia and see how it can be harmonious with the content that I feel should be included. I feel it must be done in a modular manner, each article having a clear focus not based on terminology, but on concepts, with proper links and required overlaps so that the readers have the big picture and there is no POV fork. Dominic Mayers (talk) 02:39, 21 January 2022 (UTC)

Notes

Link with the distinction logic vs methodology

On a related subject, in the specific context of an opposition to an excessive conventionalism, Popper warned against explanatory auxiliary hypotheses. This was misinterpreted by Lakatos as if Popper was fundamentally against explanatory auxiliary hypotheses. On the contrary, Popper's insistence on the importance of the logical side is a way of saying that the ceteris paribus clauses are important and we must use them to do science. Working on the logical side is similar to accepting the ceteris paribus clauses. To put it in another way, the logical side contains by default, without the need to state them, the ceteris paribus clauses. At the same time, Popper was fully aware of the methodological issues, of the Duhemian problem and of the problem of induction. He was also aware that he needed an evolutionary perspective, metaphysical research programs, quasi-induction, etc. to explain the growth of knowledge. Dominic Mayers (talk) 12:40, 18 January 2022 (UTC)

References

• Caldwell, Bruce J. (1994) [First published 1982]. Beyond Positivism Economic Methodology in the Twentieth Century (Revised ed.). London and New York: Routledge.
• Hoyningen-Huene, P. (2017). "Revisiting Friedman's F53. Popper, Knight, and Weber" (PDF). Hannover: Leibniz Universitat.
• Samaroo, Ryan (2020). "Friedman and some of his critics on the foundations of general relativity" (PDF). In Beisbart, Claus; Sauer, Tilman; Wüthrich, Christian (eds.). Thinking about space and time: 100 years of applying and interpreting general relativity. Einstein studies. Vol. 15. Cham: Birkhäuser. pp. 133–151. doi:10.1007/978-3-030-47782-0_7. ISBN 9783030477813. OCLC 1222781621.
• Harding, Sandra (1976). "Introduction". Can theories be refuted?: essays on the Duhem-Quine thesis. Springer Science & Business Media. pp. IX–XXI. ISBN 978-90-277-0630-0.
• Lakatos, Imre (1978). Worrall, John; Curry, Gregory (eds.). The Methodology of Scientific Research Programmes: Volume 1: Philosophical Papers (1980 ed.). Cambridge, UK: Cambridge University Press. ISBN 0-521-28031-1.
• Neurath, Otto (1983) [1935]. "Pseudorationalism of falsification". In Cohen, Robert S.; Neurath, Marie; Fawcett, Carolyn R. (eds.). Philosophical papers, 1913–1946. Vienna Circle collection. Vol. 16. Dordrecht; Boston: D. Reidel. pp. 121–131. doi:10.1007/978-94-009-6995-7_10. ISBN 9027714835. OCLC 8952476.
• Rutte, Heiner (1991). "The philosopher Otto Neurath". In Uebel, Thomas E. (ed.). Rediscovering the forgotten Vienna Circle: Austrian studies on Otto Neurath and the Vienna Circle. Boston studies in the philosophy of science. Vol. 133. Dordrecht; Boston: Kluwer Academic Publishers. pp. 81–94. doi:10.1007/978-94-011-3182-7_6. ISBN 0792312767. OCLC 23654059.
• Hacking, Ian (1992) [1988]. "The self-vindication of the laboratory sciences". In Pickering, Andrew (ed.). Science as practice and culture. Chicago: University of Chicago Press. pp. 29–64. doi:10.7208/9780226668208-003. ISBN 0226668002. OCLC 24246745.
• Thornton, Stephen (2007). "Popper, basic statements and the Quine-Duhem thesis". Yearbook of the Irish Philosophical Society. 9.
• Ariew, Roger (2020). "Pierre Duhem". In Zalta, Edward N. (ed.). The Stanford Encyclopedia of Philosophy. Metaphysics Research Lab, Stanford University.

Tentative title: Naive Falsificationism

The title needs to be specific enough so that it's clear that the article does not cover all meanings of falsificationism. I would have used falsificationism, but it would have been an invitation to cover all meanings of the term in the literature and it's better to have a clear focus. Yet, the title should not present a too narrow scope such as only one specific work of Lakatos. The concept that I wish to cover is the notion of falsificationism that is often criticized and, unfortunately, also often inadequately attributed to Popper. This naive notion is not referred to by Lakatos only. Even Lakatos says that Kuhn refers to it. The title "Naive Falsificationism" is not perfect, because Lakatos used this term in a very narrow manner as different from dogmatic falsificationism. On the other hand, it's only Lakatos that used it in this very narrow manner. For many authors, naive falsificationism includes Dogmatic falsificationism. In fact, for some authors, naive falsificationism means what Lakatos calls dogmatic falsificationism and they do not discuss what Lakatos calls naive falsificationism. For example, I do not believe that Kuhn used the distinctions that Lakatos created. "Naive falsificationism" as a title should be taken in its broadest sense. It would still be narrow enough. I created a draft in Draft: Naive Falsificationism. Currently, it only contains a cut and paste of the section that exists in Falsifiability. It has to be adapted and extended to make it less specific to one work of Lakatos. I might even use the historical approach that Biogeographist suggested, but addding the Duhemian problem as the primary context that precedes every thing. The only difference is that the objective is to clarify the notion of Naive falsificationism in its context, certainly including Popper's philosophy. Dominic Mayers (talk) 16:09, 21 January 2022 (UTC)