November 2014

Sun Mon Tue Wed Thu Fri Sat
            1
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29
30            
Blog powered by Typepad

« Which Monetary Policy Rule Suffers from the Fatal Conceit? | Main | What Monetary Policy Cannot Do »

TrackBack

TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d83451eb0069e2011570394f50970c

Listed below are links to weblogs that reference 21st Century Economic Methodology:

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

Most people do not think they're positivists, because they understand that induction is only probable. But that's like saying you're not a radical Muslim, because they do not think every infidel should be slaughtered.

Almost everyone, not just economists, uncritically accepts most of the assumptions and goals of the old positivist philosophers. In fact, most people have never even considered an alternative, because to do economics, science, or philosophy is, for them, to hold those assumptions and pursue those goals.

W. W. Bartley nailed it in his critique of justificationism. Goals are almost important as assumptions. I realised some time ago, that what philosophers called knowledge was something I was not particularly interested in, because it is a distraction from, and not an aid to, the pursuit of truth.

This became clear in the philosophy of biology in the 1970s -- Ruse's utter failure to put Darwinian natural selection into the deductive-nomological model was a landmark here.

Rosenberg's relentless effort to vindicate the deductive-nomological model of "science" -- and his Popper-like conclusion that biology was therefor not ontologically grounded in nomological-deductive reality was another landmark (this is a rough version of Rosenberg. His position is incoherent, he believes in biology and he believes in the nomological deductive model, and the two are incompatible).

The key finding however was David Hull's identification of a many-many problem in Darwinian biology blocking "reduction" in the Nagel sense.

This is the same sort of "many-many" problem blocking reduction that Hayek had already identified in global brain theory in his two 1952 books ... Hayek's result, of course, is more fundamental, as is the many-many problem identified by Wittenstein in the domain of language / conceptualization.

Boettke writes:

"Expression in a clear language, quantitative where possible, is crucial to good science, but the ideal of a full deductive system of axioms and theorems is often unattainable, and not, as far as one can see, actually sought by many scientific subcommunities that are nevertheless thriving."

The key causal explanans in economic science in entrepreneurial learning in the context changing relative prices and changing local conditions.

The key empirical explanandum in economic science is the price -> cost pattern, the division of labor, want satisfaction, and overall global economic order.

Neither the causal explanans nor the empirical patter to be explained in economics are can be reduced to mathematical "modeling" or quantitative specification -- these things can give us insight but the don't give us the explanans or explanadum, which are blocked from "reduction" because of multiple many-many problems, and because of the open-endedness of the causal element of individual learning in local contexts and globally changing social contexts (i.e. constant prices changes, taste changes, relative quantity/value changes, etc.)

I should note that this is Friedrich Hayek on the non-reductive / non-mathematizable but utterly scientific / causal nature of economic explanation in a nutshell:

I wrote:

"The key causal explanans in economic science in entrepreneurial learning in the context changing relative prices and changing local conditions.

The key empirical explanandum in economic science is the price -> cost pattern, the division of labor, want satisfaction, and overall global economic order.

Neither the causal explanans nor the empirical patter to be explained in economics are can be reduced to mathematical "modeling" or quantitative specification -- these things can give us insight but the don't give us the explanans or explanadum, which are blocked from "reduction" because of multiple many-many problems, and because of the open-endedness of the causal element of individual learning in local contexts and globally changing social contexts (i.e. constant prices changes, taste changes, relative quantity/value changes, etc.)"

Greg post, Pete.

Lee is dead on here:

"W. W. Bartley nailed it in his critique of justificationism."

And note well that the failure of the reductive model is one of the nails in the coffin of the justificationist tradition -- Wittgenstein, Hayek, Hanson, Popper, and Kuhn were daggers in the heart to the justificationist tradition in part because they exploded the reductionist part of that program.

And note well that the reductionist / justificationist tradition goes back to Euclide's geometry model of knowledge -- and the traditions of Plato, Aristotle, Descartes, Kant, Hume, etc. who took that foundational / justificationalist tradition as "given".

Hayek responded to Bartley's attack on justificationalism because he was already 95% of the way they with his work in neuroscience and in the philosophy of social science.

Greg,

Hayek's [i]The Fatal Conceit[/i] borrows heavily from Bartley and Popper, and their critique of justificationism, in its chapter by the same name.

Unfortunately, in my opinion, its critique in the 20th Century had almost no impact. It is ubiquitous among intellectuals, and even laymen, from self-styled rationalists to post-modernists, empiricists to solipsists, and everyone else in-between.

The meta-contextual shift of Hayek is especially needed in the realm of economics. But in the meantime, most economists will continue to criticise and reject his views for not satisfying justificationist criteria.

Pete;

"Economics is what economists do, and what economists do is build parsimonious models and subject them to statistical tests."

Um, isn't this what every scientist/philosopher/thinker/person does? [you can remove "statistical" for generality]. How can one deviate from this and still provide something of value?

For example, any of the basic patterns that one calls "economics" will be a simplified (parsimonious) version (model) of reality, which is subjected to logic, our knowledge of human nature, outcomes we see, etc. (subject to ... tests). For example, "demand curves slope downward", is a simple distillation of human nature - one that can be tested, one that is logical, one that we can see evidence of all the time.

Perhaps part of the model is that, obviously, complex systems such of the economy cannot be generally modeled with a simple framework - this is something that is hard to test for, and is a stance. But one can, and should, make statements on what we can know, and one must look for strong tests - otherwise one is simply talking. If you cannot actually test your claim, how is your claim meaningful? Doesn't that suggest the claim has no meaningful effect? Yes, negative claims (such as "complex systems are complex - and do not fall into simple models") are important (and it seems to me that each person has a bundle of these claims that they take as working hypothesis, with informed faith), but behind this claim surely there is a positive framework (for example "human action", or the like) which CAN and MUST be tested.

"Deviations from this methodology of inquiry are viewed with skepticism."

Isn't the skepticism here fully justified? Please provide some example where someone does not follow this methodology of inquiry, and yet provides a body of work that is useful.

I am unsure what your claim is, "the model and measure program" is the program that everyone lives by, no? We look at the world, see the negative realities and go from there: it is complex, so we can't figure out much, but we get some understanding of the nature of the complex situation, and we tackle what we can figure out (model). Our gains are largely from these simple models (demand curves downward, etc.), as well as plain experience (were we get an understanding of how complex and open things are, and sometimes, if lucky, we see some surprising simplicity).

I'm pretty sure most people have this "the model and measure program" as their stance to life (at least successful people, who take care to measure), and understand that both aspects are very hard (for example, in testing the model one must bring in various assumptions and tools that prevent simple clean tests) - many (most?) people beyond their teens reach this insight, except for the people who literally don't care or work very hard to fool themselves. I cannot see any other way of going about things, other than the "the model and measure program" - i.e. do what we can, and learn.

Note that the non-reducibility of Darwinian biology to physics/math is due to the direct perceptual nature of much of the explanadum in Darwinian biology -- directly perceived teleological / functional adaptations. (See L. Wrigth's _Teleological Explanation_, work which Alex Rosenberg describes as the most important work on teleology since Aristotle.)

This is another direct parallel with economics -- the teleological structure of our minds/brains allows us to direct perceive teleological behavior in nature and in other human beings, and teleological structures are by nature multiply instantiated and open-ended (see Hayek's "Scientism and the Study of Society" papers and my own "Insuperable Limits to Reduction in Biology" paper.)

Causal explanations in Darwian biology ire "open-ended" because adaptive structures and selective causes are open-ended -- and in the same way causal economic explanations are open-ended because learning in open-ended and is not circumscribable in nomological or physical or mathematical terms.

Arare, note well. This is NOT what Darwin did:

>>"Economics is what economists do, and what economists do is build parsimonious models and subject them to statistical tests."

Um, isn't this what every scientist/philosopher/thinker/person does?<<

Math models and statistics can help, but it's a mistake even in Darwinian biology to mistake the math models and the statistical work for the causal explanation -- Which is what Darwin gave us and which is something significantly more than mere math and stats.

Pete,

Very perceptive analysis! I am going to think about your comments.

Greg,

I believe the Darwinian evolution example is a good one, and one that supports my claim:

Biology is a very complex system, but evolution is a very simple model that collapses a lot of the complexity and gives a constructive framework for thinking with. Evolution can be tested - and the idea is dead simple and logical (which is the first test). Based on this simple logical structure one can go further, and look at details - what is the effect of mutations, versus crossovers, etc - and can go look at actual DNA to test various theories one can come up with. Further, one can do quantitative genetics, homology analysis, or other detailed positive work.

The key point is - everyone in the field understands that the system is insanely complex, that can't model all (or even most, or really much) of it with simple models, and they instead look to what they can do in a positive manner. Scientists, indeed everyone, does the same - life is complex, we can't understand most of it, we look to what we can productively do.

"Which is what Darwin gave us and which is something significantly more than mere math and stats."

*Everything* is more than mere math & stats (other than pure math & stats that is).

"Math models and statistics can help, but it's a mistake even in Darwinian biology to mistake the math models and the statistical work for the causal explanation -"

I don't think many people make this mistake, certainly the majority of scientists who do quantitative evolutionary work do not.

"Arare, note well. This is NOT what Darwin did:"

I highly disagree - Darwin considered the evidence, and came up with a parsimonious model that explained the complexity. The model is testable on pure logical grounds, and Darwin tested it against animal breeding evidence as well as the evidence he collected on his voyage. Evolution has continued to be tested, and is now widely studied and applied. It was pure positive science from the beginning, and continues to be studied in detail due both the nice *simplicity* of the model, and ability to make positive statements and undertake positive science with it.

To me economics is in basically the same boat - a complex system, but one were we can find simple models that we can leverage to make predictions and take actions.

Greg,

"Note that the non-reducibility of Darwinian biology to ... math"

Perhaps we have different definitions of math, I use the classic version: the precise symbolic study of patterns. Can not Darwininan evolution be described as an algorithm, albeit one with some random inputs? An algorithm that is precise, and surprisingly simple/accurate (i.e. parsimonious)?

Greg,

Hm, I realize the last comment might come off sounding bad - I do not intend this; I literally mean what I say (e.g. you may be describing differential models, or generic "full system", mathematics - i.e. you are describing a widely used subset of mathematics, and a subset that is often used in economics; I take a more inclusive description of what math is - e.g. simple logic is included, algorithms, etc.).

Arare -- I'm not going to argue this here.

The whole "model" picture is post-Darwin, and cashing out what you mean by "model" would tell a big part of this tale. Is a "model" something like what Nagel & Hempel had in mind? Is it something like what you averaged MIT economists has in mind?

Looks like we are on a freight train to scientism here ..

Arare -- If you'd like to discuss this, read some of Alex Rosenberg's work on this topic and then email me at my email address on my Hayek blog.

"Can not Darwininan evolution be described as an algorithm .. "

"Arare -- I'm not going to argue this here."

Indeed, what can you say?

"what you mean by "model" would tell a big part of this tale."

By "model" I mean model, a simplified representation of a system/structure/phenomenon/entity. i.e. the normal usage of model.

Arare -- there are hundreds of accounts of what a "model" is, and hundreds of different accounts of what role these different things can have in science.

Greg,

"there are hundreds of accounts of what a "model" is, and hundreds of different accounts of what role these different things can have in science."

Could you provide 3 very distinct accounts, that demonstrate the variation? Perhaps one that is very common, and 2 others that demonstrate just how different people see models?

From what it sounds like, Kincaid and Ross are just pointing out the last decade's natural experiment fetishism.

What we really need is a solid critique of that. I don't think natural experiments should be jettisoned, but they lead to bad generalities and bad policy because the particular connections the empirical work illustrates are uncritically taken to have general significance.

Adam, Pete, Mario, Mast-headers of this blog, and other Austrians:

"What we really need is a solid critique of [natural experiments] ... they lead to bad generalities and bad policy [natural experiments] are uncritically taken to have general significance."

Is this proposed critique of natural experiments, or simply of uncritical thinking and acceptance? One can hardly fault a given study/person/process if *others* misuse/misinterpret it. One must decouple these aspects - is the critique of politicians, decision makers who do not carefully consider evidence, the field in general, particular studies?

Could you give an example of the said fetishism? I'd argue that natural experiments are simply one of they few areas one can constructively investigate - and that study of them by someone implies little negative about their conceptual framework.

I find it surprising that much of discussion in Austrian circles seems to be "stylistically critical" - sweeping generalizations (such as this one I'm making of Austrians...) and assumptions about OTHER peoples beliefs and intents and assumptions. The criticisms often seem hollow, hyperbolic, and empty. It is fun to make such arguments, and of some use (hey, I'm doing it right now...), but it seems like many Austrians *really do* have a surprisingly negative and bitter and ungenerous view of others - given that the base stance of Austrians is that the world is a complex and rich place it is somewhat ironic that such a simple and barren view of other people seems to be seriously taken.

Yes, yes, I see the irony that I appear to be taking the exact same view. But I say this because this is the strong sense one (at least me, and the few people who I meet who know about Austrians) gets in first (and next, and always) gets in reading Austrians - an almost victim mindset, with bitterness towards, and ungenerous claims about, others.

Do some people misuse math? Sure. But most people realize that it doesn't fully capture reality, and is simply a way to precisely state a model in order to follow the consequences. One could, if one was more generous and respectful of peoples intents and intelligence, say that they are creating a "Platonic form" in order to study it, to gain insight, and hopefully to find more or less accurate niches to apply this form to in reality. When I read science/economics/etc. this is what I see, honest attempts to gain insight - not overly brittle and simplistic assumption that models accurately *capture* reality, by instead a look for sketches that bring out beauty and features.

I say this because reading Austrians is often a bizarre mixture of insight with petty and simplistic assumptions about others.

If we take the "80:20" rule as a given and in a negative manner we can say 80% of all intellectuals are basically turning the crank and simply publishing in order to sustain their funding, and are not doing anything productive: but that leaves 20% who are honestly toiling to understand reality. There is little point in criticizing the 80% who "don't get it", who think a differential equation fully captures reality, who make inane statements and spin insignificant results with verbage and name dropping (I personally do not think that the number is that high, but for the sake of argument...). The 20% who are honestly toiling and thinking do NOT have simple views of the world - they look, very hard and very honestly, for the simplicity in it.

Perhaps this is simply a misunderstanding - but it should be addressed, as this is the public view of Austrians. And yes, Austrians are far from heterogeneous, many - especially Pete - push for constructive work, but this does seem to be a cultural trait, and one that has serious consequences in terms of trying to get other people to take you seriously. "Austrians, you mean the cult?" is not too far off from the response that people have when the topic comes up (which is almost never).

Of course one only gets the tip of the iceberg in publications, on blogs, etc. - and someone from GMU who actually sees a productive "strain" of Austrianism (I use the -ism with intent) up close would have to comment (Zac?), or someone who went the the Mises Inst. etc.

Why do people write Austrians off? Likely because Austrians *seem to* write off others, appear to make little effort to address others, do not seem to make critical and hard analysis of their work - instead falling back on the "intrinsic complexity" to excuse testing, and in general, show signs of an -ism or cult: closed in, owners of special truth, victims of the stupid and evil hordes.

Pete -

to continue on my arc of trying to understand the Austrian perspective, I call on you again to clarify what method of inquiry OTHER than the scientific method is available - i.e. model & measure, or simply and in the plainest and most accurate language: pose and ask questions. How else can you get answers? The scientific method is brutally simple and is basically a truism, the rub is in actually performing this in a fruitful manner - simple, not easy.

To me, and others I suspect, Austrians seem to be critical of the scientific method - yet do not clearly state what other method can exist. This contributes to the "bad PR" of Austrians, as most scientists/economists are likely like me - they take a very simple view of science: we apply the scientific method, it is the only tool we have - luckily it is a good one, one that is enjoyable and fruitful to learn to handle, and the world/life is amazingly beautiful so it is nice to ask questions of it.

Why other "methods of inquiry" are there? There really appears to be a miscommunictation between Austrians and others here.

Arare Litus and others,

Before I start this, let me say that I don't believe blogs are the best place for this. As Tyler has pointed out, philosophical argument is too subtle for blogging and thus there are no really satisfactory philosophical blogs, while there are many very interesting blogs concerning economic policy and politics.

But let me try to offer an answer knowing that there is a great than 90% chance I will be misunderstood. The "scientific method" that you are describing is not applicable to the science of economics (or science of complex phenomena in general). There is a methodological dualism (by degree) that is in order, and the more we have access to "knowledge from within" the more the methodology appropraite for simple phenomena with access to only "knowledge from outside" are inapproriate for the scientific task.

This position is not "strange" to Austrians, but one that is recognized by several philosophers of science. But for Austrian statements that are easily understood by someone wedded to the image of science found in the "scientific method" see Hayek's essays in Studies and New Studies, and Machlup's essays --- especially his essay "If Matter Could Talk".

Deirdre McCloskey has argued at length that the 3x5 card approach to science that many economists engage in represents a major confusion of the nature of science in the human sciences.

As for the scientific process in general, I highly recommend reading Michael Polanyi's The Study of Man, but also Personal Knowledge.

The bottom line: scientific methods are chosen to be appropriate to the discipline being studied. This is an argument that Aristotle made, so it is not new!

As for your issue about testing --- lets separate out claims. Do we submit our ideas to critical evaluation? Yes, that is all we can do. Can theoretical ideas be refuted? Yes, you can demonstrate logical errors, and you can demonstrate inapplicability of subsidary assumptions to the case under examination. Can theories be refuted by econometrics? NO, testing is more subtle than that. Is that true just for economics? NO, read Duhem-Quine, let alone Kuhn, Polanyi, etc.

The application of notions of science which are dubious even for physics to the sciences of man are inappropriate and in fact have been dangerous --- read anything on Otto Neurath and the unity of science movement.

Now you ask a question about bad "PR" and "cult", etc. This is a very complicated question that requires a lot more nuance and distinctions between thinkers than is advisable for a blog. But put it this way -- is Hayek a cult figure?; how about Machlup?; Kirzner? Why not judge Austrian economics by its best practioners, rather than its marginal members?

Nobody here is denying that economics is a science, but it is a philosophical science (see not only Mises on this but Knight's What is Truth in Economics). A science different from the natural sciences --- what if matter could talk? --- and thus adjustments must be made to the way we do our science. To blindly follow a method that is appropriate for one set of intellectual activity in an field of inquiry which requires different methods would not only be wrong-headed, but produce a corrupting of scientific knowledge in that field.

Does this help at all?

You write:

"Why do people write Austrians off? Likely because Austrians *seem to* write off others, appear to make little effort to address others, do not seem to make critical and hard analysis of their work - instead falling back on the "intrinsic complexity" to excuse testing, and in general, show signs of an -ism or cult: closed in, owners of special truth, victims of the stupid and evil hordes."

Who says that Austrians are written off? I would argue that quality arguments find the appropriate home. In other words, the filter mechanisms in scientific inquiry in economics --- the job market, the journal rankings, and citation studies, etc. while clearly imperfect do not communicate zero information. Go to your university data base, do an EconLit search, you will find that the most productive members of the Austrian school as picked up by that search engine do have the "best" appointments. Compare them to the top names in the field today (e.g. Shleifer). You will see we are not as productive as Shliefer. But look at someone like James Buchanan as picked up on EconLit, or Vernon Smith.

The Austrians who don't get a hearing in the profession are those that limit their writing to internet discussion groups and blogs, or publish in non-traditional outlets exclusively. In other words, the only Austrians who are crowded out of the profession are those who are consciously trying to side-step the profession.

If you never knock on the door, how do you expect to be welcomed in. Those who knock, are often invited in.

So I deny your premise that Austrians don't get a hearing?

Arare:

"Pete;
"Economics is what economists do, and what economists do is build parsimonious models and subject them to statistical tests."

Um, isn't this what every scientist/philosopher/thinker/person does? [you can remove "statistical" for generality]. How can one deviate from this and still provide something of value?"

For the most part it is only scientists who build parsimonious models (simplified to contain as few variables as possible). The only instance where models like these work is where you can control for or otherwise zero in on specific variables to be tested. The reason this is problematic in the social sciences is that there are no(?) controls to test against, nor is there an assurance of uniformity among people. Consequently economic models cannot be definitively proven or disproven by statistics.

Contrast that with chemistry where you isolate a specific liquid compound and induce a reaction that can be repeated until your sure that your theory about that liquid is correct. This is not to say experiments always hold up, but its impossible in economics to isolate a group of people and observe there reactions in given economic situations, and then find another extremely similar group and do the same to them. There are all kinds of variables that cannot be controlled for in economics. I'm reminded of a debate between Walter Block and some guy who had 'proved' that higher minimum wages don't cause unemployment. It is axiomatically true that minimum wages will cause unemployment if there are people who's marginal productivity is lower than the minimum wage. However, there are scenarios where we could show statistically that the minimum wage does not cause unemployment due to other circumstances (e.g. forced labor, a minimum wage of $0.05/hr, and charitable employment).

As to value, even a vaguely true model can be much better for informing action than a situationally correct statistical model, when its applicability to current circumstances is questionable. You can see the worst examples of this kind coming from the Fed. They now focus on managing 'market expectations' which means they have to model the psychology of market participants, who are all trying to guess what model the Fed is using, because it will allow them to anticipate and profit from changes in Fed policy, changes which will eminate from the aforementioned model of models regress.

Pete,

I'll break up my responses: (1) and (2) address the basic scientific method:

(1) "The "scientific method" that you are describing is not applicable to the science of economics (or science of complex phenomena in general)."

I do not follow why this is so - all the "scientific method" that I describe is is thinking of, and posing questions. We ask questions, and look for answers. We shift through the evidence and learn. Why the scare quotes, and what other specific methods are there?

(2) "Can theoretical ideas be refuted? Yes, you can demonstrate logical errors, and you can demonstrate inapplicability of subsidary assumptions to the case under examination. Can theories be refuted by econometrics? NO, testing is more subtle than that."

But doesn't EVERYONE agree to this? One can find internal structural issues and show a theory is incorrect. Otherwise one can check the predictions or in inputs. But the Duhem-Quine issue, which afflicts these two empirical means of testing, is a straw "man" - everyone in science agrees with this and sees the basic issue: as I said "in testing the model one must bring in various assumptions and tools that prevent simple clean tests". This is why a single study is insufficient to topple a theory, but by testing from various angles one builds up evidence and we learn. Sure things are "subtle", but we accrue evidence and change our thinking based on this - this is the whole point of empirics. I doubt anyone, at least any critical thinker, believes that you can collect a few stats and conclusively say anything - this overly brittle opponent doesn't seem to exist, at least not in any of the scientists I speak with.

Unless economics is vastly different than it appears from the outside, there is no overly brittle "opponent" as set up. I find that in general when I talk to people - biologists, physicists, doctors, medical researchers, mathematicians, etc. that they understand the basic structure and limitations of the scientific method. The Duhem-Quine problem, while often not called that, is well understood by critical thinkers. This is not a failing of the scientific method, but a basic feature of the nature of things, and I do not find that many critical thinkers do not reconize the Duhem-Quine issue. It is a basic knowledge problem.

Now to conclude with (3), a discussion of social/natural sciences:

(3) "A science different from the natural sciences --- what if matter could talk? --- and thus adjustments must be made to the way we do our science. To blindly follow a method that is appropriate for one set of intellectual activity in an field of inquiry which requires different methods would not only be wrong-headed, but produce a corrupting of scientific knowledge in that field."

I agree that the social sciences are more difficult and one is more limited in what one can do, but I disagree that there is actually a significant difference in approach: you ask questions, you look at the answers, you learn. The questions you ask are different, yes, but the generic scienfific method is still what one uses.

Perhaps economics is in poor philosophical shape - I personally find that essentially everyone I come into contact with has a feel for the limits and nature of the scientific method, and as I describe it I do not see how this does not apply to any aspect of reality. I do not see a need for "dualism" (nor is it clear to me what the dual is that you are trying to get across). Your comments help, but there is still a gap here...

Pete,

"The Austrians who don't get a hearing in the profession are those that limit their writing to internet discussion groups and blogs, or publish in non-traditional outlets exclusively. In other words, the only Austrians who are crowded out of the profession are those who are consciously trying to side-step the profession."

I fully agree, I am playing the devils advocate well past the point of credibility here, as I believe that your perspective on this is (1) accurate, (2) fruitful, and (3) in need of being heard in order to counteract the negativity that people interested in Austrianism are exposed to when seeking out information on Austrians.

Pete,

"Why not judge Austrian economics by its best practioners, rather than its marginal members?"

This is my 80:20 point - and yes, I am discussing mainly the 80% marginal members. Why? Because it seems that the 20% have a overly brittle view of their "opposition", i.e. Austrians are complaining about the marginal 80% of the "orthodoxers". i.e. the important section of Austrians seem to think that others are hugely misguided.

Again, maybe economics really is in a very poor state of affairs - but I find this surprising if accurate, as even basic critical thought leads people to understanding the limits of our knowledge. Perhaps I'm being overly generous of thinkers in general, and I'm just assuming that this is obvious and understood by most people. But it seems to me, that when I discuss ideas with anyone seriously interested in ideas they "get" our limitations.

Pete,

I just saw the comment on my blog -

Ah... I think I'm seeing your basic stance more clearly now - I'd say there is also an undercurrent of similiar problems in theoretical physics, with some basic misunderstandings of philosophical issues (e.g. "theory of everything", much of string theory). But, and here is the key point for me, there may be a fraction of people who have misguided and strange thoughts on what science can do and how it should be done, but that this is a basic problem with people. It is not an artifact of a particular field being off track, but the general nature of thinking - it is hard.

And it's important that the public knows that much -- perhaps most -- of what economists do isn't science, it's publish or perish make-work, math games for the sake of math games, pseudo-scientic "modeling" & "testing", and "empirical" statistics for God knows what reason.

The problem with the economics isn't so much what they produce -- if that stayed simply in the un-read journals. The problem is the role the economists play in public life.

The public which knows the truth about the economists is one step ahead in the effort to make sense of the world around them.

Pete writes:

"Economics is what economists do"

To Arare Litus,

To claim that the scientific method is the common-sensical idea of asking questions, you seek answers and your learn is to misunderstand the scientific method as understood more formally. It is not a question of economics being off-track, it is a question of the philosophy of science that justifies demarcation as being mis-specified.

In my essay for SAGE, my main point is that economists have stopped making appeal to the philosophical arguments they once did, but now just adopt the justification stance they settled on in the past but without any strong philosophical reasons. We do what we do because our professors told us this was the way we do it, and the work in the journals conforms to that style of presentation. In fact, my biggest complaint about modern economics relates to how the substantive propositions that used to form the core of what it meant to understand the economic way of thinking have now been abandoned provided that the argument takes a certain form.

Your 80:20 argument I think is highly dubious as a charge against the Austrians who are intimately engaged professionally and attempting to relate their work to the broader research community in economics. My goal, as a research scientist in economics, is to provide contributions that can be seen as productive inputs into the research production of others (not restricted to Austrian economists). If you want to do that, you cannot mischaracterize the work of your peers. I encourage you to read the work of Roger Koppl as an example of someone who is trying to relate the Austrian program to the work of others within the profession. Or look at the work of Pete Leeson, or Chris Coyne. The "brittle" charge is wrong to make on these people, and I would say that would be true for Hayek and Kirzner as well. Look also at George Selgin, Larry White and Steve Horwitz and their engagement with their peers in monetary theory. Or Mario Rizzo and his historical engagement with the Law and Economics community, and more recently with the Behavioral Law and Economics crowd.

You cannot be engaged unless you appreciate the arguments --- and you cannot appreciate if you have an overly brittle view as you say.

I think there are significant miscommunication going on. In an effort to be clear, let me state the methodological position I at least hold.

1. Economics is a way of thinking, and our job as economic teachers is to offer an invitation to inquiry.

2. This invitation to inquiry captures the same excitement of learning as Richard Feynman describes in his The Pleasure of Finding Things Out.

3. Unfortunately, while economics has the same ontological status as physics, it requires epistemological procedures due to the different subject matter (that is why I kept asking Machlup's question "what if matter could talk?"). The term methodological dualism is not mine, but has a deep history that goes back to the late 19th century.

4. Attempts to apply the epistemological procedures appropriate for the natural sciences to the social sciences leads not to progress but disciplinary retrogression. This is the argument in Hayek's Counter-Revolution of Science (1952) as well as Mises's Theory and History (1957) and Ultimate Foundations of Economic Science (1962).

5. Since the 1970s, the philosophy of science literature has moved more in the Austrian direction, then away from it. Even physics doesn't conform to the strict philosophy of science model of high modernism.

Anyway, as I said when I started this, blogs are bad vehicles for philosophical debate. However, if the common-sense meaning of inquiry is what is on the table, then you should know that Austrians are on-board with you. It is only when we attempt to translate that common-sense into formal rules of scientific procedure that exalt the form and argument takes rather than assess the substance of an argument that the objections start to take on a sense of urgency.


Pete,

"To claim that the scientific method is the common-sensical idea of asking questions, you seek answers and your learn is to misunderstand the scientific method as understood more formally."

Is it a misunderstanding? The "ask questions, seek answer, learn" can be formalized easily. The formalized version doesn't seem to change the essence, to me at least: for example, one can take the classic hypothetico-deductive variant: (a) take ones current understanding as a starting point, (b) create a conjecture, (c) deduct implications, (d) gather evidence to see if implications are borne out (or not). Or: ask (informed) questions, seek answer, learn. You can go even more formal - but each ramp up of formalism does not seem to change anything, other than becoming more precise. I see no *qualitative* difference that can account for misunderstanding.

My understanding of economics is pretty biased - I was introduced via Jane Jacobs, read "Economics in One Lesson" next, I listen to Russ Robert's excellent podcast, read Caplan's book, read some Julian Simon, follow your blog, look at GMU types, etc. So perhaps I am simple exposed mostly to people who do have a nonbrittle view on economics and I'm missing the big picture of the state of the field. But your comment on brittleness and engaging the general community, as well as looking at various people in economics, seems to indicate that there is a clear understanding of the limits of knowledge - at least in the subset of people who are doing interesting things.

RE: brittle charge.

"In fact, my biggest complaint about modern economics relates to how the substantive propositions that used to form the core of what it meant to understand the economic way of thinking have now been abandoned provided that the argument takes a certain form."

Is this not viewing modern economists as being brittle? i.e. - not having a clear understanding of the situation?

"Anyway, as I said when I started this, blogs are bad vehicles for philosophical debate. "

I agree - but while communication is inefficient, it is not negligible.

"that exalt the form and argument takes rather than assess the substance of an argument that the objections start to take on a sense of urgency."

I fully agree - and I am being overly strong in my arguments simply to clarify positions quickly (in addition, come midnight tonight "Arare Litus", goes back to the ether from which he came, so my time is short! : http://arare-litus.blogspot.com/2009/06/utility-of-this-blog_20.html ).

Greg,

"it's publish or perish make-work, math games for the sake of math games, pseudo-scientic "modeling" & "testing", and "empirical" statistics for God knows what reason."

I agree that a subset of people do this, and this is true in every field of science, but in general this sort of "effort" is simply ignored. The majority of publications have zero (!) citations, an imperfect measure that indicates that most work is useless. But who cares? Yes, this is a waste, yes this is unfortunate, but the harm is fairly contained. A waste of resources, yes, but the impact is zero in most cases. When people do work that has implications and impact, then there is most often counter balancing studies.

Much of science/economics/etc. research is unproductive - of course. And some people game the system their entire career - but to take a positive spin: most research will turn up empty, after all to figure out something new is hard, and this work is published (as publication is the metric used to see if one survives, or not), but I'd guess that most scientists - even the ones with lots of "null" work - will hit an interesting result once in a while.

The question is not if there is garbage produced, but if the garbage clogs the system, and if the "hits" outweigh the stikeouts. Most of the garbage is simply ignored - one could almost make the claim that low tier journals exist simply to publish stuff that should be ignored by scientists, but which gives administrators at Universities a number to make them happy. Scientists learn to publish garbage in journals that flag it as garbage, but which enables them to keep getting funding, until they get good results which they publish in good journals. Other scientists look to the good journals and ignore the rest.

The safety value is also in place - if something is not in a top journal, policy cannot be made on the "findings". Politicians can at least recognize prestige.

I fully agree that much of the publications are poor - but that is due to the incentives and the nature of work that intellectuals face. The question is not if there is garbage out there, but if the garbage is clogging everything up and dilution out the good stuff. Yes, there are examples of horrible gaming - this will exist under any system - but in general it seems, to me at least, that the system provides proper flagging of what is garbage (published in "noname journal z") and what is good (published in top journal A).

Evan,

"Consequently economic models cannot be definitively proven or disproven by statistics.

Contrast that with chemistry where you isolate a specific liquid compound and induce a reaction that can be repeated until your sure that your theory about that liquid is correct."

This is a matter of degree, not kind. Even in chemistry we have had a long history of incorrect views. In general all our understanding is very wrong, and only partially accurate.

All knowledge is partial, imperfect, and provisional. The distinction between the hard and soft sciences is only a matter of degree.

FWIW, I think Pete is quite right on methodological dualism. In physics we see rocks and stars and imagine atoms. In social science, we see human actions and imagine general equilibrium. That difference matters, I think. Add to the impossibility of describing human action in "physical" terms, and you end up depending on something more or less like "introspection." The details matter here and, as Pete says, a blog is a bad place to handle such details. But the basic differences between social sciences and physics are there, I think, and it is not "scientific" to wish them away out of a desire to be "rigorous."

Roger, at least we can SEE individual atoms with electron microscopes.

Arare,
"The distinction between the hard and soft sciences is only a matter of degree."

Yea, a read that before. The text that said it concluded that economics is essentially a midterm between literary criticism and physics. I think that Smith, Menger and Hayek would disagree...

What about how about quarks, mesons, and strings, Rafael?

Well, we have physical evidence of the existence of quarks, so they have a different nature .

While strings are pretty much abstract concepts, like general equilibrium. So I think that the analogy works better with strings.

The Ross, Kincaid stuff is on my priority reading list, but note Don Ross also has a book on economic theory and cognitive science that may be worth taking a look at.

Ah! I was aiming at a different point, Rafael. In physics we observe the systemic consequences of the interactions of many small elements. We observe the overall order and infer the existence of the constituting elements. In economics, we observe the "atoms" and draw inferences regarding the existence and nature of the overall order. This is what Hayek called "Menger's compositive method." In this sense economics is "synthetic" and undergraduate physics is "analytical." Combine that difference with our "inner" knowledge of the nature of the "atoms" of social science, and you may decide that the methods of the social science need not, and in some ways should not, be the methods of physical sciences such as physics and chemistry. In both cases, however, we are talking about how the elements combine to form "complex" phenomena. Thus, complexity theory is going to have something to say to economists and chemists alike. But matter does not talk, as Pete reminds us, and that means some methods appropriate to social science have no application to physics or chemistry. (I am carefully skirting the interesting question of infra-human ethology.) In other words, in spite of my greater zeal for "scientific" methods such as complexity models and nonparametric statistics, I think Pete is completely right to defend the old-time religion of methodological dualism and "literary" economics.

Yes, Rafael, it´s all in Hayek´s _The Counter-Revolution of Science_...

Yes, Ludwig, it is all in Hayek's The Counter-Revolution, but also the essays in Studies and New Studies. Mises's "best" methodological work is in Epistemological Problems and Human Action in my opinion, not as well put in his other statements (though once you have read Hayek you can see what he is getting out, especially in Theory and History).

Mario Rizzo and Bruce Caldwell have written the "best" methodological work among the modern Austrians. Rizzo's paper on Mises and Lakatos is a classic, and Caldwell's Hayek's Challenge discusses at length the Hayek contribution to modern methodology of economics.

The interesting thing is that the philosophy of science literature caught up to the Austrians, not the other way around, as philosophers of science came to understand the difficulties of testing, and applying the model of the natural sciences to the social sciences, and even more that perhaps the picture of the natural sciences drawn by philosophers did not conform to the real world of scientific practice. Einstein's practice, for example, was different than Popper's depiction as Polanyi pointed out and so on.

Pete

Pete wrote: "The interesting thing is that the philosophy of science literature caught up to the Austrians, not the other way around,..."

You should know that I am familiar with Studies, New Studies, Epistemological Problems etc...
But what you say here is a serious exaggeration. A good survey of how these ideas evolved in the philosophy of science is still contained in _The Structure of Scientific Theories_ ed. by Frederick Suppe. The intellectual background here is very different from that of the Austrians...

Ludwig,

I don't want to get into a debate on this, but Suppe's book was published in 1977, and I am talking about the way the literature moved in the 1970s, 1980s, 1990s, and beyond.

Also, and I am thinking more along the lines of Polanyi and others following up on his Personal Knowledge, as well as the critique from Rorty, etc.

The real issue for the context of my essay, is that economists stopped paying attention to the philosophy of science literature. Justification and demarcation are no longer as important to them as these topics once were. Justification and demarcation is assumed to have already been settled. So now what keeps economics moving along as a "scientific enterprise" is (a) intellectual curiosity, (b) intrinsic interest of economic topics, and (c) a methodological conventionalism that is communicated through the professionalization process. (a) and (b) give us great reasons for hope of progress, but (c) keeps getting in the way --- or at least that is my position.

And in fact my position is the hard-line one, that (c) not only gets in the way, but distracts from (a) and (b).

OK, but Caldwell´s book was published in 1982, and Rizzo´s paper is not a particularly recent one either. (I don´t exactly remember the year right now.) Lakatos has been déjà vu for quite some time now. Maybe his most enduring work is that in the philosophy of mathematics. After Lakatos we still had Laudan, but the debate on theory choice/appraisal has evolved very little since then, except perhaps for the fact that Bayesianism has now become much more influential (not per se a bad thing imho...)
You may be right that most of the things that had to be said on theory choice have been said...
The work of Caldwell and Rizzo is important because it made certain developments in the philosophy of science accessible to (Austrian) economists. This being said, when I need to check out something about Popper or Lakatos, I look directly into Popper or Lakatos etc.; I don´t go to look into Caldwell or Rizzo...
Also partly because of the disrepute into which "positivism" has fallen very few people realize today how great the work was of earlier authors such as Nagel, Hempel, Reichenbach etc.
But Mises´s point that all this philosophizing is primarily inspired by the situation in the physical sciences has still much relevance...
Lots of success with your essay, and keep us posted...


OK, but Caldwell´s book was published in 1982, and Rizzo´s paper is not a particularly recent one either. (I don´t exactly remember the year right now.) Lakatos has been déjà vu for quite some time now. Maybe his most enduring work is that in the philosophy of mathematics. After Lakatos we still had Laudan, but the debate on theory choice/appraisal has evolved very little since then, except perhaps for the fact that Bayesianism has now become much more influential (not per se a bad thing imho...)
You may be right that most of the things that had to be said on theory choice have been said...
The work of Caldwell and Rizzo is important because it made certain developments in the philosophy of science accessible to (Austrian) economists. This being said, when I need to check out something about Popper or Lakatos, I look directly into Popper or Lakatos etc.; I don´t go to look into Caldwell or Rizzo...
Also partly because of the disrepute into which "positivism" has fallen very few people realize today how great the work was of earlier authors such as Nagel, Hempel, Reichenbach etc.
But Mises´s point that all this philosophizing is primarily inspired by the situation in the physical sciences has still much relevance...
Lots of success with your essay, and keep us posted...


Rizzo's paper is from the 1970s.

I applaud you for going to original sources, but if the relationship is between these thinkers and the Austrian school, then I think consulting Rizzo and Caldwell is the place to look --- and of course Hayek.

As for my essay, it is coming out in a Russell Sage Foundation volume on 21st Century Economics.

Pete

Well, yes, I will/would perhaps take a look at Caldwell and/or Rizzo but I would almost certainly disagree on some fine points. It´s very unlikely that I would simply copy them, even with respect to the relationship between these authors and the Austrian school. Interpretative issues in the history of thought are not an exact science. I have been reading the philosophers of science as well as Hayek at least since my student days; I tend to make up my own mind on these things.
It´s good to see that their papers are still read after a few decades. Much of the secondary literature is forgotten much faster. Cheers for Caldwell and Rizzo!!!

Caldwell was good when he caught up with later developments in Popper's work, especially the theory of metaphysical research programs, but imho the best folks on this topic are Jack Birner and Larry Boland.
For Birner on Popper and Menger http://www.the-rathouse.com/2008/Birner-Menger-Popper.html
For Boland on the Popper-Hayek program. http://www.the-rathouse.com/shortreviews/revBoland.html

Pete's claim that the example of Einstein refutes Popper can be disputed but he is correct to talk about philosophy catching up with the Austrians. It seems that the positivism of the 20th century represented a regress from the philosophy of science that Weber knew about, and it required the conjectural turn that Popper introduced in the 1930s to get the philosophy of science back on track. This process has some way to go, as indicated by LVDH.

Popper made too much out of the demarcation issue (the positivists obsession) and that hindered both the development of his best ideas and also the reception and interpretation of his work.

Well, that was my round, who's buying next?

I missed that remark of Pete on Einstein and Popper. Whatever the matter, Einstein remains an interesting test case for philosophers of science; a bit like the Great Depression for economists...

Pete wrote: "Rizzo's paper is from the 1970s."

My files indicate that Rizzo´s paper was published in 1983 in a volume edited by Kirzner in honor of Ludwig von Mises.

In fact the problems related to falsificationism were already pointed out by the French physicist Duhem.
The analogy between Duhem´s position and that of the praxeologists is only a superficial one, however. See on this, the excellent paper by Ludwig van den Hauwe:

http://philsci-archive.pitt.edu/archive/00003677/01/Falsificationism.pdf

Someone wrote: "The Duhem-Quine problem, while often not called that, is well understood by critical thinkers. This is not a failing of the scientific method, but a basic feature of the nature of things, and I do not find that many critical thinkers do not recognize the Duhem-Quine issue. It is a basic knowledge problem."

Maybe, but imagine that someone claims to come up with an alternative methodological framework that simply sidesteps the whole issue. One can make a case that this is precisely what Austrian methodology does. It´s indeed related to the point Roger pointed out, namely that in economics, one starts "from the inside" so to speak, not "from the outside"... We have privileged cognitive access to the building blocks of the system... This is surely a thing Popper didn´t understand very well...

And then, I am not sure, say, Popper (or even Lakatos for that matter) has a satisfactory answer to the Duhem problem. He starts the discussion of falsificationism by saying that what we have here is a _logical_ problem, asymmetry etc... But later on he becomes much softer on the issue...

For a short thesis on the Duhem problem, supervised by Alan Chalmers, with a a review of the Popperian strance, the Lakatosian contribution, the Bayesian response and the "new experimentalism". http://www.the-rathouse.com/Theses/Duhem-QuineIntroPopperians.html

It seems that there is no logical or philosophical fix to the problem, as "Someone" said (cited above). Popper made a strong point of the distinction between the logic of falsifibility (what you can logically achieve with a correct statement of fact) and the practice of falsificatiaon (irretrievably conjectural). The issues have to be worked out on a case by case basis in actual scientific situations. Sometimes falsifications can be achieved rapidly in the eyes of the scientific community, sometimes decades are required to get a decision.

It is amusing to see how quickly Quine retracted his strong restatement of the problem.

The Bayesian "solution" to this "problem" is now the mainstream view of this matter, and rightly so. See e.g. the well-known article by Dorling, and the other cites in Ludwig van den Hauwe´s article, such as the book by Howson and Urbach.

From the Austrian (deductivist, rationalist...) perspective, there _is indeed_ a philosophical fix to the problem, in the sense that the problem does not/need not arise.

Again, and sorry for the Popper fans, Popper is here at his worst...


Except I actually read The Logic of Scientific Discovery, and I highlighted a dozen paragraphs where Popper pre-empted and addressed the Duhem-Quine problem.

In short, it's not a problem for a fallibilist like Popper, because the logic of scientific discovery is a seperate issue from the psychology of scientific discovery.

Kelly,
No, no, you must have misunderstood the issues here. In fact it is rather unfortunate that Popper himself presented the rule of modus tollens in a misleading fashion in LScD. In his terms, “p” is a conclusion or prediction logically entailed by a set of statements “t” consisting of both “theories and initial conditions” such that “t entails p” is logically true (Popper [1959] 1980, 76). He then argues that “-t” can be concluded from “(t entails p) . –p” by modus tollens, such that empirical evidence for “-p” refutes “t”. This is misleading, however, in the sense that empirical evidence conflicting with the prediction “p” does not falsify the theory; it merely falsifies the conjunction of the theory and initial conditions: if the initial conditions are false, the theory may yet be true. To see how this works, let us consider Popper´s own example:

“If we analyze this causal explanation we shall find several constituent parts. On the one hand there is the hypothesis: “Whenever a thread is loaded with a weight exceeding that which characterizes the tensile strength of the thread, then it will break”; a statement which has the character of a universal law of nature. On the other hand we have singular statements (in this case two) which apply to only the specific event in question: “The weight characteristic for this thread is 1 lb.”, and “The weight put on this thread was 2 lbs.”.
We have thus two different kinds of statement, both of which are necessary ingredients of a complete causal explanation. They are (1) universal statements, i.e. hypotheses of the character of natural laws, and (2) singular statements, which apply to the specific events in question and which I shall call “initial conditions”. It is from universal statements in conjunction with initial conditions that we deduce the singular statement, “This thread will break”. We call this statement a specific or singular prediction.”(Popper [1959] 1980, 60)
We can schematize Popper´s little example (in slightly abbreviated form by collapsing his two initial conditions into one combined statement) as follows: Let “i” mean “this thread”, and translate “Lx” as “x is loaded with a weight exceeding its tensile strength” and “Bx” as “x breaks”. Then, the hypothesis is “(x)(LxBx)”, the initial condition is “Li” and the singular prediction is “Bi”. If we put this example into Popper´s form for modus tollens as “(tp).-p-t” we get:

“[(x)(LxBx). LiBi].-Bi-[(x)(LxBx). Li]”,

but if our purpose is to show what empirical evidence would contradict/refute Popper´s hypothesis in the “falsifying mode of inference,” this format will not do, because empirical evidence for “-Bi” does not contradict the hypothesis, but merely the conjunction of “Li” and “(x)(LxBx)”. It may well be that “Li” is false and the hypothesis true. To refute the hypothesis, we need to show that “Li.-Bi” is true.
The correct way to apply modus tollens as the falsifying mode of inference is to follow a different pattern and to treat predictions as substitution instances. Thus, with “LiBi” as the predicted substitution-instance of hypothesis “(x)(LxBx)”, the latter would be contradicted by “-(LiBr)” which is equivalent to “Li.-Bi”. Using modus tollens, the falsifying pattern would be instead:

“[(x)(LxBx)(LiBi)].-(LiBi)-(x)(LxBx)”.

The essential point, however, is the following: Insofar as “(x)(LxBx)” is to be deemed an empirically testable and refutable hypothesis, two independent pieces of empirical data are required for the same test object, one bearing upon the antecedent clause “Lx” and one bearing upon the consequent clause “Bx”.

Exactly.

Jonathan,

You missed the point. Using your example, Popper is just assuming that "Li" is true. Once that assumption is made, Popper's simple characterisation works fine. But this does not mean that Popper did not realise that "Li" might be false; indeed, his entire discussion of "conventionalist stratagems" would hardly make any sense if he did not.

I can hardly believe that anyone would seriously what you wrote important. It seems to me that Popper discussed the issue so casually in LoSD that most people don't realise he discussed it at all. For Popper it was never important that the initial conditions *might* be wrong, because he was a fallibilist. The decision to falsify a theory was exactly that: *a decision* predicated upon the potentially mistaken categorisation of something else as true.

It's impossible to falsify something unless you first "truify" something else--that's just basic logic. Your criticism of Popper can be boiled down to the fact that the decision to "truify" something might be mistaken, and therefore the falsification may be mistaken too. But Popper is a man who claimed that all knowledge is guesswork and conjecture, a man who knew fine well the situation you described and never saw a problem to solve.

The logic of testing that Popper described works fine, so long as you have the guts to "truify" something first--and while your decision may be mistaken, that's just something we have to live with. For Popper, there are no guaruntees that we will decide correctly, because logic offers only limited guidance.

[Unfortunately, many of the characters you used on your formula have not displayed correctly on my computer, but I believe that I am familiar with the point you are trying to make well enough].

To clarify:

To derive "~Ex[Lx->Bx]" it is necessary to presume that something else is true. However, the premises of the following inference are insufficient.

~Bi |≠ ~Ex[Lx->Bx]

The problem above is that no "initial conditions" are being assumed, and so for the inference to work, some must be introduced.

Li,~Bi |= ~Ex[Lx->Bx]

But it is important to remember that Popper is discussing testing in the context of assuming that "Li" is true. As a fallibilist, he is offering no guaruntees that "Li" is true, and remarks in many passages that we should be careful about such assumptions, but also understands that testing is impossible without them.

You have to understand that Popper, especially later, rejected the goal of justified true belief, or even justified (probably true) belief. He was perfectly comfortable with just guessing (i.e. conjecturing) that "Li" is true, because for him all knowledge was ultimately guesswork anyway. The problem here is that many of Popper's critics had other goals, but they largely ignored Popper's discussion and critique of those goals, and then rejected his other ideas for not satisfying them.

One final note:

Why is the Duhem-Quine problem a problem? Many people seem to consider Popper's falsificationism problematic, because it does not unambiguously declare a single result. That is, one can always reassign truth-values and protect a theory from falsification, and, moreover, sometimes they are even correct to do so!

Consider the two implicit standards by which falsificationism is being judged, 1) it should never leave any room for choice, and 2) it should lead to no wrong results. While these two standards are quite ridiculous, they are the implicit assumptions behind the Duhem-Quine "problem".

Intellectuals don't like making choices: they want a rule to resolve every dilemma. They look at Popper's rules for testing and declare it problematic, because it doesn't tell them what to think.

Kelly,

The problem is, for some individuals, Popperism is a cult, so the great saint Popper should not be criticized... In the same manner, for many Austrians, Austrian economics is a cult, so the great saint Mises etc. should not be criticized.

It cannot be conceived of course that the great Popper wrote some questionable things like that?
Well, no...

I have no hope of convincing any such individuals by rational argument...

Still professional scientific activity, professional philosophy go on... And very few still take Popper´s falsificationism very seriously today... The majority is not always right, but it certainly is not always wrong either...

Jonathan,

I do not agree with Popper about everything, but I was fortunate enough to read LoSD before I became aware of its critics, so I did not read it with the expectation of finding the mistake you think Popper made. In fact, as I previously noted, Popper's entire discussion of "conventionalist stratagems" would have made no sense unless he recognised the "problem" you are describing.

It seems to me that professional philosophers are simply averse to making conjectures. Their basic objection to Popper's falsificationism is that it does not unambiguously declare what to believe in the face of ralcicitrant evidence, since it is alwas possible to claim, to use your example, that "Li" is false. But for Popper this was not a problem, because his goal was not to provide a set of rules for what to believe, but rather to analyse the logic of experimental testing. (This is why I made a distiction between the logic and psychology of science).

Popper explicitly compares his "logic of scientific investigation" to the "logic of chess", because I think he was trying to avoid a discussion about pure logic. To state that "Li" might be false is, therefore, just irrelevant to Popper's goal, because it focuses upon mere logical possibilities. Any method of science must introduce extra-logical rules, and even then, Popper just refers to them as "rules of thumb".

To suggest that I am somehow part of a Popper of Austrian cult is just ridiculous, because all of this has been obvious to me since the very first time I read Popper. If anything, discovering the mainstream critique of Popper was extremely disappointing, and severely reduced my respect for the entire profession of philosophy.

Jonathan,

With regard to whether I can be convinced by rational argument, or merely a mindless member of a Popperian cult, I actually agree with your logic.

Except for the fact that your presentation was needlessly complicated, (because "(t&i)->p" would have been sufficient), I have no objection to your logic. The "problem" you described was one of the things I considered when first reading Popper. But while I was quite satisfied with his later discussion of it, I have since learned that almost nobody else was paying attention. Popper was not making an error, because he did not share the same goals as most of his critics, and, in fact, explicitly critiqued and rejected them--though the philosophical profession, by and large, doesn't seem to have paid any attention.

Lee Kelly,

Apparently you´ve understood nothing since it isn´t "needlessly complicated". It´s just exactly sufficient.

Popper made plenty of mistakes.

In any technical field he touched - e.g. quantum mechanics - he was a great amateur.

Moreover he didn´t deny it. He always admitted that he would never have been good enough to become a real scientist...

Jonathan,

My point was that you needn't have shifted into quantificational logic to make your argument, since merely introducing a conjunction as the antecedent in Popper's original formulation is sufficient. In other words, where "t" is a theory, "i" is a set of initial conditions, and "p" is a prediction, the following inference is invalid:

(t&i)->p, ~p |≠ ~t

If we correctly apply modus tollens to the above premises, then we get:

(t&i)->p, ~p |= ~(t&i)

Since there are three possible assignments of truth values where "~(t&i)" is true, including one where "t" is true, such an application of modus tollens does not unambiguously declare a theory false.

I do not object to your logic, but rather your claim that it is somehow Popper's mistake. To restate, if Popper did not recognise the logical situation you describe, then his entire discussion of "conventionlist stratagems" would have been impossible. Moreveor, because he had different goals from the majority of his contemporaries, the "problem" you identify was only ever in the eyes of his critics.

Popper was a nonjustificationist. For critical rationalists, an unjustified belief is not a problem, because justification is neither attainable nor desirable. Since different standards are being used to recognise problems and solutions, much of what Popper wrote has been severely misunderstood. He has been accused of offering solutions where he never saw a problem, and of being mistaken for not offering solutions to problems he never intended to solve.

The whole Duhem-Quine thesis, when used as a critique of Popper's falsificationism, is a lazy, thoughtless critique, and one that serves to butcher everything Popper wrote on the matter. There are many naive fans of Popper and falsificationism out there, and they may have something to learn from the Duhem-Quine thesis, but to suggest that it shows Popper "at his worst" is simply ignorant.

And I do not agree with Popper about everything. I think he made mistakes, (and so did Popper). What on earth would give you the impression that I am suggesting anything otherwise? I really think you need to spend more time thinking about these matters, and less time throwing around vague accusations.

If a scholar makes (too many) mistakes, this is a reason to weaken our trust in what he says. This is an inductive argument, but it may be a correct one.

"The problem is, for some individuals, Popperism is a cult, so the great saint Popper should not be criticized... In the same manner, for many Austrians, Austrian economics is a cult, so the great saint Mises etc. should not be criticized."

Examples?

Kelly wrote: "The whole Duhem-Quine thesis, when used as a critique of Popper's falsificationism, is a lazy, thoughtless critique, and one that serves to butcher everything Popper wrote on the matter."

Here you seem to be saying: if there exists a single big knockdown argument against a particular position, it´s inappropriate to use it when you are facing a philosopher of Popper´s caliber... This is what I meant by the cult-like spirit surrounding much of Popper.

But the question is rather: if you cannot satisfactorily answer a single simple knockdown argument, why should anyone expect you to be able to satisfactorily answer more sophisticated arguments?

Popper is at his best in his work on the history of philosophy and science, and his arguments in favour of the critical attitude, fallibilism etc.
The latter argument is of course much less original.

Jonathan,

The only reason to not trust what a scholar *says* is because you think what he says is false, because if it is not false, then it is true, and we do well to trust a truth regardles of who said it.

In any case, you haven't offered a coherent explantion of why Popper made the mistake you accuse him of, and have dismissed my objections with an invalid argument (i.e. induction). Did you read Popper's discussion of conventionalist stratagems? Or have you seen his work on the theory-ladenness of observation? Both are discussed in LoSD, and each depends on a prior understanding of the Duhem-Quine thesis, (although Popper never gave it that name).

I challenged you on this matter because I learned the Duhem-Quine thesis from Popper when reading LoSD. Of course, Popper doesn't call it the "Duhem-Quine thesis", and doesn't offer it a formal treatment, but it is there, implicitly acknowledged in so much of what he wrote. Only afterward did I discover what Duhem, Quine, and others had written, and I was astonished to learn that professional philosophers actually considered it a critique of Popper's falsificationism. It seemed to me that either they had poor reading comprehension, or just weren't that interested in an honest discussion. So you might be able to imagine my response to people who make claims like you.

As for Popper's "too many" mistakes, what are you talking about? I suspect he did err in his claims about quantum mechanics, (though I am not really in a position to evaluate them satisfactorily), but so too do many professional physicists, and it was outside of his field. There are also a number of instances where he admits his error, and seemed to have learned from his mistakes. But Popper is not special in that regard; it is as though you expect a fallibilist to be infallible.

I have read most of Popper - and also much of the secondary literature - as a philosophy student many years ago. I also read Lakatos, Kuhn, Feyerabend, Laudan, Putnam, Dummett and others, also the older empiricists like Hempel, Nagel, Reichenbach etc.

I was insufficiently infatuated with Popper, however, to go on specializing in it, in particular when considering there are a number of other, more interesting giants at the philosophical firmament... Perhaps it has been an error of mine not to pursue that interest. (I finally turned myself into an economist. But anyway I hold a master-level degree in philosophy so I do not think my reading comprehension is inadequate.)

I may want to reread what he wrote about conventionalist strategems. If you have written something yourself on the subject, please do not hesitate to communicate it.

My guess is that Popper is famous for reasons different from the fact that he was so original and deep etc.

He almost has the stature of a mythical figure; as a young man he had known personally some of the great scientists of the Vienna of the beginning of the twentieth century etc. This fact alone already predestined him for a great career as a philosopher...

Also he explicitly writes for the layman - also the layman within academia. All those individuals who do not understand anything of modern science now all at once, when reading Popper, experience the excitement of understanding a little bit of modern science... Popper gives you the impression that it is fairly easy to understand the deep secrets of science. This certainly also explains some of his success... Individuals who are not professional philosophers but who nevertheless know one single philosopher, know Popper...


I still add that it is not necessarily problematic to make mistakes, to the extent that it may encourage others to correct them, and pursue research along similar lines. Popper´s work on verisimilitude is a good example.

Note that he considered quantum mechanics his field, since he wrote an entire book on it.
I am only an amateur quantum physicist myself but if I understand well the remarks of Giancarlo Ghirardi in this respect, Popper made rather elementary mistakes in his interpretation of the Einstein-Podolski-Rosen experiment. The problem is not merely a differing interpretation, but an inability to correctly apply the principles of quantum mechanics to a simple physical situation.
Popper was also the kind of guy who told the audience that of course "he didn´t master the formalism" but who nevertheless went on telling the physicists in the audience he knew and understood better than them... He was not particularly modest.

Jonathan,

My point is that the Duhem-Quine thesis is not a "single big knockdown" critique of Popper's falsificationism. There are, I think, some simple-minded fans of Popper who do not appreciate the ramifications of the Duhem-Quine thesis, but Popper was not one of them.

Consider one of Popper's objections to induction that observation is theory-laden. His basic point is that our sense experiences do not entail propositions, but must be interpreted into propositions before a logical inference can be performed. Therefore, what is induced from experience depends on how those exeperiences are interpreted, and since there are infinitely many possible ways of doing this, nothing specifically can be induced from experience.

Obviously, if observation is theory-laden, as Popper describes, then so is any falsification. An experience is interpreted in the context innumerable assumptions, which we might lump together as the "initial conditions", so when an experience falsifies a theory, it is logically possible that our interpretation is false instead.

All this was implicit in Popper's writings on falsification, and he stressed that each person needed to decide when to falsify a theory for themselves, (i.e. when to accept the initial conditions and the experimental result). In other words, a falsification is not a logical necessity which everyone is forced to accept, but a decision to accept a set of propositions as true, and a fallible one at that!

Popper's discussion of "conventionalist stratagems" concerned the fact that one can always wriggle out of a falsification (in the manner described in the Duhem-Quine thesis). Although he acknowledges that sometimes people might be right to reject an apparent falsification, (see, for example, the discovery of Neptune), he considers it bad form and is critical of those who use these "stratagems" frequently.

Jonathan,

For the record, I don't think I would have liked Popper as a person, and I do not much like his style, but I think he was right about a lot, especially about philosophy. I have never read his work on QM, but few have been as enlightening to me philosophically.

Except in the cases where I think something should be obvious, I make it a rule to never argue about the beliefs of another, especially when they are dead philosophers. But, in this case, I actually learned the Duhem-Quine thesis (although not be that name) from Popper when reading LoSD, so when someone claims that it is a "knockdown" critique of falsificationism, I am quite vocal in my objection.

Jonathan,

By the way, as a nonjustificationist, I believe the vast majority of philosophers make elemetary errors of logic frequently. In my opinion, they are unable to apply the principles of logic coherently, and yet they write books about it all of the time. This is not a position I relish or take lightly, but one that just follows from my elementary objections to justificationism.

I may, of course, be wrong, but since I do not willingly take wrong positions, and working on the assumption that I am right, it is difficult for me to take seriously the work of many professional philosophers--after all, if they can make such an elementary mistake, how can I trust them? The answer, of course, is that I can't, because they are no less fallible than me. Instead I should simply consider what they are *saying*, eveluate it for truth, and consider whether *that* is worth trusting.

I don´t know as far as Jonathan is concerned, but as to your next to last post, all this sounds extremely familiar to me, especially from a Duhemian viewpoint.

The problem is, however, that Popper _does say very explicitly_ that his falsificationism brings us closer to a _logical_ solution to this conundrum. He does say very explicitly that the grist of falsificationism is the _logical_ asymmetry as compared to verificationism, in the sense that it is possible to argue from the truth of a singular statement to the falsity of a universal statement. As your own remarks bring out clearly, this is typically _not_ the case.

So Popper´s initial suggestion/claim that he introduced a Copernican revolution into the philosophy of science via his falsificationism is just not credible.

We now also know that within a Bayesian framework most of these problems can be circumvented so any defense of Popper should start with a refutation of the alternative view.

It is not correct that all the auxiliary assumptions can be lumped together under the label "initial conditions". The Duhem point relates to auxiliary assumptions, ceteris paribus conditions etc. which are needed to make an unambiguous prediction. Per se it does not refer to disagreement about whether the theory applies.

We all make a first-order intuitive distinction between a theory not being applicable and a theory being false, but Popper´s falsificationism cannot even make that distinction since in the typical case of a false prediction either of both can be the case.

Indeed, Kelly cannot distinguish between cases in which the theory fails the test, and cases in which there simply is no test because the theory does not apply. There is only a test of the theory, if it has been ascertained that the theory applies. If the theory does not apply, of course there can be no test. Complicated?

As to the theory-ladenness claim, there are reasons to believe the imporantce of it has been over-rated, but I see it as a circularity problem of sorts: how is it that facts can be supposed to test the theory if it is theory that decides what counts as fact?

Note that Mises also adhered to a variant of the theory-ladenness claim but that his apriorism allows him a way out of the circularity: _correct_ theory (which is decided aprioristically) is a pre-condition of _correct_ observation.

So it is not quite correct to claim - as some have done - that his position is analogous to that of Duhem-Quine. If there is any analogy, it is a very superficial one.

LVDH,

I believe you are confusing the logic with the application. While Popper's falsificationism does solve the *logical* problem of testing, (i.e. it avoids the invaldity of alternatives), it does not offer any guaruntees that our auxiliary assumptions are correct. And I do not believe it was ever intended to, because Popper was, first and foremost, a fallibilist who thought no such guaruntees exist. With regard to Bayesianism, Popper and Miller spent a considerable amount of time critiquing it. For me, Bayesianism is flawed because it seems to be jsut a fancy form of inductivism, but with most of the same problems.

Jonathan,

I can disinguish between cases when a theory fails a test and cases where the theory has not been tested, and do so frequently. That is, sometimes I categorise the experiment sound and the theory false, and other times I categorise the experiment flawed and the theory untested (whether I consider it true or false is a seperate issue). But you're right! I can't offer any logical rule for what course should be taken. In other words, I guess, conjecture, go with something and see where it leads.

It is as though you desire a set of rules which tell you what to think in every circumstance, but neither I nor Popper were attempting to offer any such rules--that is not a problem I am attempting to solve.

Today only a few cranks still contend that induction does not exist. What have you ever read about Bayesianism, apart from Popper and Miller?

Your distinction between "logic" and "application" is a nice example of "immunizing stratagem". If the whole process of testing is mediated via decisions that have to be made, where is the fundamental asymmetry as compared with verificationism/confirmationism?

When you have observed a long series of white swans, why not decide, perhaps not logically but by way of application, that all swans are white? A black swan is not really a swan after all...

Popperians´ lack of insight is amazing...

LVDH is right; by the way Popper´s degree of corroboration is, in the broad sense, a form of induction. It is nonsensical to totally reject the idea of induction since it would mean that our assessment of theories is totally insensitive to positive, supportive evidence, which is absurd.

Note also that in many cases there is independent evidence for a theory being applicable. For instance in economics, we can ascertain whether or not the quantity of money has been increased etc. So the problem is not with respect to initial conditions...

LVDH,

The comments of a blog are not a good place to hold discussions on these matters, and even less so when participants are disposed to give others an ungenerous reading. Since I do not think you are interested in even trying to understand my views, I will spare no further effort trying to explain them.

Jonathan,

Although I disagree that Popper's concept of corroboration is a form of induction, it is something in Popper's work of which I am quite critical. It is, in my opinion, rather redundant and misleading, (and mostly because it does *seem* like induction).

My understanding of Popper on corroboration is like this. Suppose that we have two theories called A and B, where A is more testable, (i.e. more falsifiable), than B, (B might even be a tautology), and let e be some evidence. Suppose that e is consistent with both A and B, but not-e is only consistent with B. In such case, e may be a test of A but not B. I think Popper might say, even though neither A nor B have been falsified, that A has been more severely tested, i.e. corroborated. In any case, I think Popper himself downplayed the concept of corroboration later in life, because it wasn't very important.

With regard to induction, I do not think there is any *logic* to induction. It is invalid, arbitrary, and, in my opinion, serves no purpose and has no place in reasoning. Of course, this is not to say that people do not generalise, nor that seeing a repetitive pattern might not be a reason for such a generalisation. But I make a sharp distinction between reasons in the causal sense, and reasons in the sense of premises. The psychological reasons, or causes, of what someone believes, should not be conflated with the premises of a logical inference.

The point is: your distinction between "logic" and "practice" here is ad hoc, since when we talk scientific testing, there is only practice, there is nothing outside practice. If the simple modus tollens falsification model is not what is used, or what happens in the practice of testing, then it is an error to bring it up, and Popper shouldn´t have brought it up.
I am genuinely interested in your views. If there is any published evidence of your competence in these matters, please provide references to your publications, and I will give them the most generous reading I can.

The point is: your distinction between "logic" and "practice" here is ad hoc, since when we talk scientific testing, there is only practice, there is nothing outside practice. If the simple modus tollens falsification model is not what is used, or what happens in the practice of testing, then it is an error to bring it up, and Popper shouldn´t have brought it up.
I am genuinely interested in your views. If there is any published evidence of your competence in these matters, please provide references to your publications, and I will give them the most generous reading I can.

This discussion is starting to get interesting, although at this stage there probably are only three or four readers. Let me offer a few debating points in haste, then I will defend them at leisure if anyone in addition to the major protagonists are interested.

1. Popper's decisive move was not falsification but the conjectural turn. Falsification was just the most obvious point of difference from the verificationism of the positivists.

2. He developed that move with his critique of the authoritarian (justificationist) structure of western philosophy (Introduction to Conjectures and Refutations). Incidentally the same structure of argument infects both epistemology and political philosophy as well.

3. David Miller and Bill Bartley took on board non-justificationism in the strongest form, though they did not have Popper's blessing in that venture.

4. Miller's forte is probability theory, so Bayesians and inductivists need to be sure that they have disposed of his objections. Some papers can be found on his website and I will ask him what is his best on Bayesianism.

5. There are serious problems with the Bayesian approach and readers of the relevant chapter of my thesis will be aware of them, and also that this approach (as exemplified by Dorling's well-known paper) lets us down when we need it most, that is, when there are serious rival theories in competition.

6. Bartley ran with non-justificationism in a big way and I have recently put several of the relevant papers on line as pdf files (sorry, slow loading). Links can be found on this page http://www.the-rathouse.com/writingsonbartley.html

7. Good scientists practice the logic of falsification whether they have heard about Popper or not because the logic is strong and it is the most cost-effective way to use evidence.

8. Scientists who attack their problems in a critical and imaginative way probably don't need to know about Popperism unless they need to be emancipated from unhelpful ideas like positivism. Einstein escaped from his early positivism, quite likely without Popper's assistancec. They agreed on such matters, but not on mataphysics because Popper did not succeed in talking Einstein out of determinism.

9. Be careful when you mock Popper as a physicist. Admittedly his formal qualifications were in school-teaching and cabinet-making but he was taken seriously by some physicists who matter.

10. In the social sciences, Ian Jarvie drew up a longish list of major contributions that Popper made in "The Open Society and its Enemies". http://www.the-rathouse.com/2009/Jarvie-rat-sit.html

Not a bad effort when you consider how few people manage even one or two major contributions. Done in his spare time as well.

10. Jarvie also pointed out that Popper anticipated the "social turn" in the philosophy of science, he argued that institutions matter (with an analogy between institutions and traditions) and he pre-emptively refuted the strong school of the sociology of knowledge in chapter 23 of The Open Society.

11. Bayesian subjectivism can be seen as the latest gasp of "justified true belief" theories of knowledge. There is an alternative approach which is concerned with forming and testing critical preferences between contending theories, policies, or goods in the marketplace. This is a point of contact between Popperism and praxeology. It is helpful to note that Popper's theory of "world 3" objective knowledge in no way conflicts with Austrian subjectivism because there is a whole world (2) for subjective thoughts, judgements and beliefs.

12. It is interesting to see how sloppy arguments and sleazy ad hominems have so far been used effectively to marginalise Popper's place in the profession of philosophers. About 25 years ago I decided that the game of protracted arguments with philosophers was not worth the candle but in the interests of Austrian economics and classical liberalism it may be necessary to saddle up and go the distance.

To better evaluate Popper's falsificationism, I think it worth considering the problem he was trying to solve.

When Popper was a young man, Marxist sociology and Freudian psychology were in ascendancy. Although both were presented as scientific disciplines, Popper felt that something important separated them from other sciences like physics and chemistry. When arguing with Marxists and Freudians, he realised that no matter what the evidence might be, it was always taken by his opponents to further confirm their views. In contrast, physicists were performing dramatic experiments where results could go for or against a theory. What separated Marxist sociology and Freudian psychology from the other sciences, thought Popper, was their unfalsifiable theories.

The problem for Popper was that contemporary philosophy did not distinguish between these two forms of science. So far as Marxists and Freudians were concerned, their theories were as scientific as those of physics and chemistry, and Popper thought this situation unsatisfactory. There was no real test, Popper said, unless it was possible for a theory to fail, and so he went about developing a new vision of science where falsifiability took precedence.

When developing his views, Popper discovered that he could dispense with induction altogether. Although he had solved the problem of induction, it was not in a manner that would satisfy inductivists. By not concerning himself with where theories come from, the old problem of how to derive, and thereby justify, theories from the evidence simply never arises. Science could progess, Popper thought, by a process of conjecture and refutation, and without the need for any logic but the deductive kind.

The nonjustificationist character of Popper's falsificationism was implicit in his early writing; it was primarily evident in his disinterest in the source of a theory and confort with making assumptions. What was important for Popper was that, with sufficient assumptions, a theory be falsifiable. Of course, this testing procedure would be fallible, and sometimes our decision to falsify a theory may be mistaken. However, for Popper, scientific testing and rational criticism could go on indefinitely, and at no point was their any guaruntee against error.

In the world of critical rationalism, there is neither a solid nor weak foundation to knowledge--we are floating.

Very bad publicity for Popper! The examples of Marxism and Freudianism only illustrate how naïve falsificationism really is. Marxism is can be refuted by economic reasoning, and psychoanalysis still works fine, falsifiable or not...

By the way, don´t forget Mises, in his philosophy of natural scence, was also an inductivist...

I believe Popper´s starting point, namely that traditional epistemology was about justificationism, authoritarianism etc. is simply mistaken.

The general idea behind inductivism is that our assessment of theories should not be insensitive to positive, supportive evidence. One could just as well argue that the neglect of positive, supportive, inductive evidence is, or can be, in certain circumstances, a kind of authoritarianism, justificationism...

Jonathan,

I think you should be more careful with your objections. My previous comment was simply intended to illustrate Popper's original problem-situation.

There is far more wrong with Marxism than just being unfalsifiable, (or, at least, being very difficult to test), and Popper recognised this; part II of his "Open Society and Its Enemies" is an explicit critique of Marxism. It is also worth considering Popper's work on metaphysical research programs, because there he makes it very clear how important and useful unfalsifiable ideas can be.

And as for "justificationism", I suspect, from your previous comment, that you have yet to grasp what I mean by the term. Logic was only cleaved in two--induction and deduction--because philosophers wanted to justify their theories with evidence. Most of the traditional problems of philosophy, follow inexorably from the implicit ethical presuppositon that one should not believe what one has not justified, and the vast majority of philosophical discourse concerns what is justified and how it is justified.

Jonathan,

As for good or bad publicity, I suspect few enough people are still reading this thread for it to matter. Though I am inclined to think that your snide and dismissive remarks are doing your reputation more harm than Popper's.

It would be difficult to still lower Popper´s reputation... If Popper is the only philosopher you have read - as is the case for many laypeople - you should be more modest. First also read some real giants (van Fraassen, Putnam etc.)...

Again if there is any written evidence of your competence I will give it the most generous reading. Please provide references.

The Marx-Freud-story about how Popper came to his idea is general knowledge, déjà vu, it has been recounted over and over again... It´s very pedantic to behave as if you are the only person on this list who knows this... You are simply rehearsing things every educated person has heard hundreds of times...

You should take the challenge by considering a number of texts - e.g. the books by Howson and Urbach and that by Jeffrey, as well as the article of Dorling - and then demonstrate in detail where there is any sloppy reasoning going on...

I do not question that Popper is an important philosopher but there is no justification for all this cult-like bullsh´t

Ladies and Gentlemen:

POPPER IS AN IMPORTANT PHILOSOPHER!!!

Jonathan,

I haven't actually read much Popper--The Open Society and Its Enemies Parts I & II, The Logic of Scientific Discovery, Conjectures and Refutations, and some articles online. Although I have tried to read the works of many philosophers, usually they bore me, because I do not share their goals or problems. Popper is just one of the few philosophers who has impressed me. My knowledge about Popper comes more from some pieces online written by his students.

With regard to my story about Popper, yes, it is common knowledge, which makes it all the more strange that it has been ignored by his critics. My primary point during this thread is that Popper has been accused of failing to answer a problem which he never intended to solve, i.e. the Duhem-Quine problem. It is a problem in the eyes of his critics, because they want something more than Popper from science, and this difference of goals has its root in Popper's nonjustificationism.

As for this "cult-like bullshit" ... I am a critical rationalist, so I agree with Popper more than most, and I was challenging what I believe is a false accusation. Which part of that is "cult-like bullshit"?

Why is professional philosophy so intolerant of any positive mention of Popper? Whenever I try and discuss these matters online, sensible argument is replaced with rude dismissals, though only by people who have studied philosophy in an academic setting. Although I am sure that I will somehow learn to live with your opinion of me, it is strange that anyone would react like this at all.

I never said that Popper intended to solve the Duhem-Quine problem - and he certainly did not solve it as a matter of fact. In fact, I am not very much interested in the historical context of Popper´s thinking or in the concrete problems he did or did not intend to solve, but I do consider actual accomplishments more important than intentions.

I do believe, however, that the DQ problem and also the (different) problem of under-determination of theory choice by data objectively constitute important problems in the philosophy of science, as perceived from a present-day perspective.

I also believe that authors such as Popper are extremely important since obviously philosophy should not merely be an ivory-tower business.
It´s not true I have a low opinion of Popper fans; I only want to encourage them to go beyond Popper....

I enjoyed Popper enormously as a beginning philosophy student, and only later on became rather bored by it. One reason is that Popper is better in formulating problems than in solving them. Another is that he likes talking about his own person too much. In particular I don´t think his proclaimed modesty is very sincere...

Lee Kelly

I still add that I don´t believe there is any kind of conspiracy against Popper.

Popper has received all the attention from his peers a professional philosopher can wish. Adolf Grünbaum´s well-known 1976 paper is a good place to start...

The point is that most of the challenges he launched are now perceived as having been settled but often in ways that leave little intact of his own contributions...

Lee Kelly:

You wrote: "Although I am sure that I will somehow learn to live with your opinion of me, it is strange that anyone would react like this at all."

Why strange? The strangeness is in (1) the way Popper handles certain issues; and (2) the way Popper is handled by his fans.

Both exemplify a certain lack of maturity.

As to (1): I would for sure never write a book on quantum physics if I would not be sure I also master the formalism. As to (2): I would not publicly defend Popper if I had not also read his critics. These are among the best of professional philosophers...

So the strangeness, anomaly etc. - if any - is not with me, it is with Popper and his fans...

Giancarlo Ghirardi´s critique of Popper´s experiment can be found here:

http://arxiv.org/PS_cache/quant-ph/pdf/0702/0702242v1.pdf

Getting back to Pete's main points, especially the persistence of formalism and positivism. Mark Blaug wrote pungently on the topic in a paper which regrettably did not cite Pete's 1995 (?) Critical Review contribution. Blaug's paper is on line at Questia, as is his wonderful account of his career.
[Disturbing Currents in Modern Economics by Mark Blaug; Challenge, Vol. 41, 1998. The Problems with Formalism: Interview with Mark Blaugh Journal article; Challenge, Vol. 41, 1998. Not Only an Economist - Autobiographical Reflections of a Historian of Economic Thought by Mark Blaug; American Economist, Vol. 38, 1994.]

The first place to explore an issue with philosophy is to check what students of philosophy are being taught. In 1989 I surveyed the undergraduate courses and reading lists in Philosophy, Politics and Sociology in the (then) 21 Australian universities. The objective was to find what they were being told about Popper and Hayek who I regarded as the twin pillars of anti-scientism and classical liberalism. The short answer is that you had to be very lucky to get more than a passing reference to Popper and the situation with Hayek was worse.

Round about 1998 I wanted to repeat the survey but there were more than 50 universities thanks to the recruitment of a heap of minor colleges who used to award diplomas in teaching, arts and crafts, rural studies etc. So instead I searched 200+ websites of philosophy schools, mostly in the US but also a few others like Cambridge. The story was the same, Popper rated a mention as a part of the convulsion in the field that involved Kuhn, Lakatos, Feyerabend and the sociology of science. No indication that his contribution was more robust or extended into other fields where he did first order work - evolutionary epistemology, logic and probabililty theory, metaphysics, philosophy of mind, history of ideas, politics and the social sciences. Cambridge at that time was arguable the worst course in the world in terms of keeping positivism alive!

Other indicators: there is now a thriving genre of introductory books on philosophy written by the likes of de Botton and Greyling. These are mostly either silent or misleading on Popper's contribution.
http://en.wikipedia.org/wiki/A._C._Grayling

"The Open Society and its Enemies", one of the great classical liberal books, is kept in print by a lay readership. See if you can find a uni course on Plato with vol 1 of the Open Society on the reading list. Given that Plato is arguably the most influential promoter of statism, this is a serious matter.

What is to be done? Probably the least effective approach is to challenge philosophers on their interpretation of Popper. If you want to try that, good luck! They are deeply entrenched behind the ramparts of justificationism, so the attack has to proceed at a deeper level, starting with the work of Bartley on non-justificationism, now providentially on line (as of last week).

More helpful would be to take on important problems in other areas and demonstrate how Popperian/critical rationalist ideas can be helpful in making progress with them. The primary focus would be the substantive problem and the credibility of the exercise would depend on the progress (if any) with that problem.

Another approach, which I don't seriously recommend, is to write essays in the form of blog comments. Sometime it helps to dash off some quick "first draft" thoughts in this kind of venue, then they can be worked up for some other destination.

Jonathan wrote "I am not very much interested in the historical context of Popper´s thinking or in the concrete problems he did or did not intend to solve, but I do consider actual accomplishments more important than intentions."

This statement is very helpful because it probably indicates the standard approach to Popper's thoughts (and to the work of the Austrian economists as well).

It is very dangerous to try to interpret the work of a major thinker without taking some trouble to place their problems in historical context. So first of all you have to ask "What is your problem?". When young Alan Chalmers turned up in Sydney and started asking this question at parties and pubs he got a bad reaction so he had to use a different form of words to find out what people were working on.

There can be two aspects of this problem-question, one is to identify the problems that concerned the great philosophers. Popper wrote an essay on the way that philosophical problems often arose elsewhere, in science or maths of religion or social reform. He suggested that the standard manner of teaching philosophy often hides these "roots" of philosophical problem in a most unfortunate way. For a summary of this paper. http://www.the-rathouse.com/CRNatureofproblems2.html

Second, there may be presuppositional issues in play. The philosopher may be challenging some unstated assumption of a foundational or metaphysical nature. This can result in a reformulated question but if the reformulation is not accepted by the profession then the thinker may find that he and the profession are talking past each other.

Alfred North Whitehead and also Hayek wrote that the most influential (and possibly the most dangerous) ideas are those which are accepted uncritically (and often enough unconsciously) by all the major schools of thought at the time. An example is the idea that epistemology is about the justification of beliefs. Another example is the cluster of ideas that Hayek called scientism.

One of the great negative achievements of logical positivism was to banish metaphysics from polite society, so the presuppositions were placed out of bounds for criticism. They didn't go away, as von Mises and others pointed out they continued to operate but without scrutiny by positivists. So thinkers like Popper who operated much of the time at the presuppositional level flew over (or under) the radar of the positivists.

That is probably enough to write in a comment that hardly anyone will read, the point is that great thinkers make a difference at a level that calls for close attention to the formulation of their problems, this in turns calls for a historical and situational analysis to locate the presuppositions that frame the questions and influence the kind of answers that deemed to be adequate.

OK, but note that in quantum mechanics the formalism is essential to understanding the very substance of the theory. But it´s correct the situation may be different in economics.
Note also that Mises himself apparently dreamt of praxeology as a science developed deductively with the same rigour as in logic, mathematics etc.
But I don´t believe it is correct to view Mises´ apriorism as a kind of metaphysics. As Rothbard saw clearly, it´s only empiricism of a different sort.

Rafe wrote: "It is very dangerous to try to interpret the work of a major thinker without taking some trouble to place their problems in historical context. So first of all you have to ask "What is your problem?"."

I am sufficiently aware of the historical context of Popper´s work. What I meant is that Popper´s lasting influence will ultimately depend upon whether the particular solutions he offered transcend this historical context, upon whether they exhibit any universal significance, beyond the particular context of place and time in which they arose in the first place.

The comments to this entry are closed.