Tuesday, April 24, 2012
Matias Vernengo— Shift happens indeed
Matias Vernengo contrasts Popper's premise that falsification of hypotheses is paramount in science, and Kuhn's view that it is the explanatory and predictive power of theories that is basic. I see them as complementary. Normal science is conducted along Popperian lines, while Kuhnian paradigm shift, in which the normal paradigm is replaced by another, is institutional and pragmatic rather than being based chiefly on testing specific truth-value.
Read it at Naked Keynesianism
Shift happens indeed
by Matias Vernengo
Subscribe to:
Post Comments (Atom)
11 comments:
Tom, I have a few problems with this one, and will quote various portions of your post to illuminate them:
"Popper's premise that falsification of hypotheses is paramount in science, . . . "
I don't recall Popper ever stating that premise. He did say that falsifiability, in a rather precise sense of logical relations among statements of a theory or conjecture (See pp. xix -- xx of Realism and the Aim of Science) is the criterion we should use to distinguish science from metaphysics. But he didn't say that falsification is "paramount" in the sense that it is more important than other aspects of science.
Next,
". . . and Kuhn's view that it is the explanatory and predictive power of theories that is basic."
I think Popper also believed that the explanatory and predictive power of theories is basic. See, for example, his essay called "The Bucket and the Searchlight" an appendix to his Objective Knowledge (1972) where he says: (p. 349)
"What is the task of science? . . . .
"The task of science is partly theoretical-explanation-and partly practical-prediction and technical application. . . . "
and he then goes on to give his views on each. "The Bucket and the Searchlight" first appeared in the form of a talk Popper gave in German in 1948. It wasn't published in English until 1972, but it's likely Kuhn had access to it or similar thoughts when he attended Popper's William James lectures at Harvard in 1950.
Evidently, Popper agrees with you that falsification, explanation, and prediction are all complementary and important aspects of science, though I'm not sure you mean the same thing by "falsification" as Popper.
On this:
"Normal science is conducted along Popperian lines, while Kuhnian paradigm shift, in which the normal paradigm is replaced by another, is institutional and pragmatic rather than being based chiefly on testing specific truth-value."
I'm not sure this is borne out by historical studies in science. Kuhn's theory of scientific change is, essentially, a theory of scientific revolutions that grows out of Polanyi's fideism and Wittgenstein's philosophy of language. and, I don't believe that Kuhn's theory of change stands up very well to accounts like those of Paul Thagard, especially in Conceptual Revolutions and How Scientists Explain Disease or to Joseph Agassi's The Continuing Revolution, Steve Fuller provides a pretty comprehensive review and evaluation of Kuhn's work in his Thomas Kuhn: A Philosophical History of Our Times (2000). In my view Kuhn's work doesn't stand up very well to Fuller's examination, and I think the distinction between normal and revolutionary science with its emphasis on the non-rationality of paradigm change doesn't really hold up. I don't think this means the distinction between institutional science and more creative and critical science is invalid, but rather that institutional science frequently doesn't follow the critical orientation of Popper, constantly and insulates its knowledge claims against falsification in a very anti-scientific "unpopperian" way. While it is often non-institutional science that is more Popperian, Fallibilist, and critical in its outlook.
Next, on Matias Vernengo's piece itself he says:
"It is worth remembering that the official methodology in economics remains Popperian."
It seems to me, however, that if Critical Rationalism is "the official methodology in economics" then it is honored only in the breach, since if it were practiced, neoliberalism would long since have been falsified by the profession. In fact, I think the economic profession is more Kuhnian, than Popperian since it seems to place community norms and conformity above the need to rigorously test its accepted doctrines.
Don't disgree with your more in depth analysis, Joe. I should perhaps have said "Popperian" and "Kuhnian," in the popular distillation of their views are basically objective (falsification of hypotheses, empirically determined) v. subjective (institutional and pragmatic, socially determined).
The basic idea is that falsification of hypotheses through which normal science operates in testing should lead to paradigm change when key hypotheses fail, since this amounts to paradigm failure. But, has history shows, normal science responds on an ad hoc basis or by denial, even in a hard science like physics. Old ideas die hard and, as they say, progress is funeral by funeral.
What we are seeing with Neoliberalism is that falsification of hypotheses is meet with either ad hoc repairs or just plain denial, and the failure of the paradigm itself to predict the crisis in the first place, and then to explain it or offer a way out, hasn't shaken mainstream confidence in it and heterodox economists are still marginalized.
BTW, I think that Paul Feyerabend got philosophy of science much more correct.
Thanks Tom, Feyerabend is fun reading and certainly does brilliant historical analysis. But I still think Popper is better and more important.
What happened to Popper over the years is strange. LSD was never published in English until 1959, so not very many philosophers had exposure to it before then. A small number of people, Popper's students and colleagues and others who could read the German version were familiar with it. But that was only a few years before Kuhn's book hit, so the community didn't have the opportunity to grasp Popper's early thinking before Kuhn's stuff appeared. I remember logical empiricist writing in the 1960s spreading the view that Popper's position was close to theirs except that he used the falsifiability demarcation criterion rather than the verifiability criterion. I think that Popper never quite overcame this very superficial reading of LSD.
During the late 1940s and early 1950s Popper wrote his Postscript to LSD, but before it could be published he developed eye problems and for awhile couldn't work on the book. By the time he recovered he became immersed in so much other work, that he couldn't complete the editing of his ms. His student Bill Bartley, picked up the editing work in 1978, and completed it for Popper. Finally the Postscript was published in three volumes in 1982, close to 30 years after it was written. So, two of Popper's major works LSD and the Postscript, never hit the street until more than a quarter of a century after they were written, a real problem for people during the 1960s and 70s to get a good view of the development of his views over time.
Of course, his students, including Lakatos and Feyerabend, has access to LSD, the Postscript, and Conjectures and refutations way before anyone else. But it was hard for external observers to see how his writings affected theirs, or Kuhn's for that matter, because the written record of those likely influences wasn't available.
In fact, Kuhn, Lakatos and Feyerabend, all seemed much more aware of the role of the metaphysical underpinnings of science than Popper. Yet, when you read the Postscript and become aware of Popper's account of metaphysical research programs, it's hard not to see the influence of Popper on all three of the other philosophers who had early access to his postscript thinking. I'll continue this discussion in another comment, since I must be running out of space.
Popper's original ideas
To continue my discussion on Popper, his work always spanned philosophy of physics and also the philosophy of the social sciences. Remember his Ph.D. came out of Karl Buhler's rogram in cognitive psychology at the University of Vienna, the same program that also produced Paul Lazarsfeld, Else Frenkel-Brunswik, and a number of other top scholars. It was a wonderful program eventually killed by the Nazis.
Anyway, apart from the Social Science elements in his thesis, Popper completed two major works during WW II, The Poverty of Historicism and then the two volumes of The Open Society and its Enemies. Both are major works in the Philosophy of social science and the latter in Democratic Theory as well.
During the mid and late 1950s Popper again returned to his social concerns and in Conjectures and Refutations published in 1962, just before Structure he has essays on epistemology, philosophy of physics, and philosophy of the social sciences. But, during the 1960s Popper broadens his view still further, and expands some early developments in LSD into the first major work in evolutionary epistemology. he also backs off his view that Darwinian is metaphysics and not science, and becomes part of the neo-Darwinian synthesis in Biology making contributions to the philosophy of biology. This work culminates in Objective Knowledge (1972), a terrific book, which hasn't been read by anywhere near the number of people who've read LSD or CR. Not only is Popper's thinking about evolutionary epistemology reflected in this book, but also his social, and cultural turns as well.
The 1970s were also a busy time for him. The Library of Living Philosophers did a festschrift on him in 1974, and he got some hard shots from many people, but he also wrote an intellectual autobiography called Unended Quest as part of the project and delivered some very illuminating replies to his critics and also to his many supporters who contributed to the collection including Donald T. Campbell who became critical to the contnued development of evolutionary epistemology.
Later in the 1970s, Popper also wrote a fascinating book collaborating with Nobel Prize winner in Biology John C. Eccles called The Self and Its Brain (1977). That book is a great one on the Body-Mind problem and the dialogues between Popper and Eccles, which are a good portion of that long book are very rich.
During the 1980s Popper continued to publish. As I said earlier, his postscript to LSD was published in 1982, but thereafter there a number of smaller books comprised of essays and talks that he continued to give. There were some major articles co-authored with David Miller and some contributions to the Bartley/Radnitzky volume on Evolutionary Epistemology.
Though his work in the 80s and early 90s consists of minor contributions, they do fill out parts of the picture of his complex and very rich philosophy. His notion of a world of propensities characterized by indeterminism and freedom is very important. So is his work in the book with Eccles on emergence which fits into complexity theory very well.
In my view, Popper is the major philosopher of the 20th century. Greater than Wittgenstein, greater than Kuhn, or Feyerabend, more important than Russell or Carnap or Quine or Dennett or Foucault or the Existentialists or Habermas.
It may take another 50 years for this to be fully realized. But I think there will be a turn away from constructivism, and relativism eventually, and when that happens Popper and his objectivist realism built on a foundation of constructivism will be there waiting to pick up the pieces.
Regarding Wittgenstein and Feyerabend argued v. Popper, Lakatos and Kuhn that what is most interesting and important occurs at the margin rather than at the core. The common mistake is to generalize about the core, ignoring the significance of the margin, which is where "shift happens." At the core, subjective bias is often conflated with objectivity. Wittgenstein sought to demonstrate this based on a logical analysis rather than a scientific or historical one.
Tom, when did Popper argue that what is most interesting occurs at the core or words to that effect? And when did Wittgenstein argue "that what is most interesting and important occurs at the margin rather than at the core?"
I don't recall either of them using Quine's metaphor.
That is my interpretation, Joe. Again, it would be more correct to say "Popperian" and "Kuhnian" in that this is how Popper and Kuhn are generally understood v. Feyerabend and . PF and LW are all about the margin, e.g., PF's "against method" and LW's notion of family resemblance v. essential definition. It's a difference of degree and emphasis, and the common understanding of their chief contributions reflects this.
The general understanding of Popper and Kuhn is within structuralism broadly speaking, whereas Feyerabend and LW are clearly post-structural — although there is no generally agreed upon interpretation of LW and I would argue that most interpretations are wrong.
Tom, I don't accept the validity of the structural/post-structural distinctions. Nor do I accept the "general understanding" of Popper and Kuhn, since many people who offer up these understandings have apparently never read much Popper and also haven't read Kuhn.
Perhaps an analogy would be apropos here. The way American loosely pragmatist analytical philosophers read Popper is the way the neo-liberals read MMT. Either they don't, or they read and write about him in such a way as to safeguard their paychecks. Their whole interpretation of what he had to had to say is easily dismissed, usually with some quotes from LSD, and as far any of them knowing the Postscript, Objective Knowledge, Unended Quest, The Self and Its Brain, Knowledge and the Body-Mind Problem, All Life is problem Solving, A World Of propensities, forget about it. They're not Popper scholars.
They're anti-Popper propagandists protecting an academic establishment. Anyway, the purpose of my comments is not to talk about generally accepted understandings of Popper, Kuhn, Wittgenstein, or Feyerabend, but my understanding of what they had to say and how they relate to one another. Also, your comments and Matias's interpretation refer to Popper's positions and to Kuhn's and not to generally accepted understandings.
Sorry to be so persistent about this, but I'm sure you see red when you read the persistent misconstruals of MMT we see from the likes of Krugman. I share those feelings. But I also have similar feelings when I see people doing the same thing to Popper.
I get this, Joe, because I deal with it in every subject that is near and dear to me. The misunderstanding is thick, and virtually impossible to correct, other than very locally. In fact, one can explain it, get agreement, and then see it immediately contradicted.
But I am resigned to accepting that this is the way the world works when things are labels, which is way Taoists reject all labeling as misleading if not false. The way out is to state exactly what you mean. It's always safer to argue specifics.
However especially in blogs we tend to rely on labels as shorthand. That is a weakness that seems pretty much inherent in blogging.
But getting back to Popper, as I said, my favorite work is The Logic of Discovery, one of his early works, actually. In it he looks as the logical processes called deduction (reasoning from rule to instance), induction (reasoning from instances to rule), and abduction (discovery of a rule heuristically). It is the latter that has proved most significant. and it is why the margin is more significant than the core. It is where innovation arises through an essentially intuitive process. As Feynman said of Einstein's discovery of general relativity, "I still can't see how he thought of it," i.e., there is no clear logical process involved.
BTW, there is a good article about C. S. Peirce's pioneering contribution to understanding deduction, induction, and adduction entitled, What is Abductive Inference? by Uwe Wirth.
The essence of what Feyerabend and Wittgenstein assert, somewhat radically, is that a fixed notion of method and logic inhibit discovery and innovation, and bias understanding. The germ of this wrt science is in Popper's Logic of Discovery. Thus is it PF and LW who are regarded as the radicals.
Tom,
LSD was one of Popper's earliest works. As I indicated earlier it was originally published in German in 1934, and in English in 1959. In LSD, of course, Popper famously rejected induction as a logical method, taking and extending Hume's position. This was always a constant in his later work. I don't recall Popper ever discussing abduction. In fact, I don't think he became familiar with Peirce until the 1960s.
As for abductive inference, some would view it as similar to Popper's "Conjecture." But I don't think that Popper really believed this. Since he viewed as Conjectures as guesses, not as inferences.
I understand what you're saying about the germ of a method or set of rules being present in LSD but Popper never advocated "a fixed notion of method." In fact, in Realism and the Aim of Science he says: (pp. 5-7) he says:
"As a rule I begin my lectures in Scientific Method by telling my students that scientific method doesn't exist. I add that I ought to know, having been, for a time at least, the one and only professor of this non-existent subject within the British Commonwealth. . . .
"I assert that no scientific method exists in any of these three senses. To put it in a more direct way:
(1) There is no method of discovering a scientific theory;
(2) There is no method of ascertaining the truth of a scientific hypothesis, i.e., no method of verification.
(3) There is no method of ascertaining whether a hypothesis is 'probable', or probably true."
then he goes on to say:
". . . The only things which the partners in an argument must share are the wish to know, and the readiness to learn from the other fellow by severely criticizing his views -- in the strongest possible version that can be given to his views -- and hearing what he has to say in reply.
"I believe that the so-called method of science consists in this kind of criticism. Scientific theories are distinguished from myths merely in being criticizable, and in being open to modifications in the light of criticism. They can be neither verified nor probabilified."
These statements seem pretty clear to me. From at least 1956 on, when they were written, Popper did not believe in a scientific method in the normal sense of this term, whether fixed or otherwise. He only believed in the best critical analysis of competing theories and hypotheses one can muster.
Continuing due to limited space:
Now, I suggest to you that Wittgenstein was less radical than this. He believed that language communities set their own rules, and the implication of this, developed at length by numerous adherents of Wittgenstein was that each scientific discipline was free to form its rules that were valid for that community, and, of course, the implication of that is that there are many scientific methods, as many as there are disciplines of scientific communities. And, btw, Wittgenstein believed that language communities including scientific communities were closed and not open with respect to their basic categories and ideas. Indeed, it is from this that Kuhn hit upon his famous incommensurability claims.
So, on the one hand, we have Popper saying, there is no method, while Wittgenstein says there are many methods each held beyond criticism by closed scientific communities. Forgive me, but I think Popper's view is far more radical than Wittgenstein's, whatever the mainstream says. In fact, Wittgenstein's seems more like Bronislaw Malinowski's view of communities and culture than anything else, a kind of solidarist relatvism. While Popper's view is a pretty thoroughgoing skeptical fallibilism that, however, doesn't fall into relativism, since it still provides for the search for truth as the goal of inquiry with at least the possibility of finding it, even if we can never be certain that we have.
Finally, Feyerabend's work, in my view, falls back into justificationist individualist relativism, and in the end provides no possibility for science. You may consider this radical. But I see it as just a modern re-statement of extreme Classical Greek pyrrhomism, and, what, after all, is "radical" about that.
Btw, I do that Popper's philosophy allows for assertions about scientific method, even if he did not. It just seems to me that theories of scientific method would be like other conjectures, subject to vigorous and thoroughgoing critical evaluation, and like other theories they would have to survive criticism.
This position, btw, allows for theories of scientific method and for the growth of knowledge about them over time.
Post a Comment