Sunday, 3 January 2016

Popper, Art and Science.

Observe the two images below:

"La Maja Desnuda" by Francisco de Goya, c. 1795–1800. Museo del Prado. [A]
"Femme nue assise" by Pablo Picasso, c. 1909-10, Tate Modern. [B]

Call me artistically challenged, but I find both images hardly comparable. Mind you, there are similarities: (1) their authors were renowned Spanish artists; (2) both works are pictorial representations of naked female bodies, (3) held in the collections of respected European institutions.

For all these superficial shared features, there is something else much deeper those works share: they are works of art. Mind you, I am no authority in art and my guess is that it would be hard for anyone -- certainly, it would be impossible for me -- to adjudicate which one has the greatest artistic value. How to compare works so evidently different?

And if a judgement between two paintings is difficult, one might think it safe to conclude that the difficulty is only compounded if one were to compare works in different art forms: what is more valuable, "La Maja Desnuda" or Miguel de Cervantes' "Don Quixote"? What about Richard Wagner, Akira KurosawaOscar Niemeyer?

There may be a simple test, which you answer objectively with either a "Yes" or a "No" to decide, but if there is, I've never heard of it. In my ignorance, I believe there is no silver bullet to judge the merit of art works: liking it or not, one must rely on experts in the field -- a.k.a. artists -- for that (assuming they actually can produce a unanimously valid opinion).

And, if there is no fool-proof rule to judge the relative merits of two works of art, then there is no absolute criterion to say this is a work of art, this is not. Who establishes the acceptance threshold? On what grounds?

To pretend otherwise would be not only infantile, it would impose a destructive and arbitrary Procrustean bed on artists: Cubism does not conform to the artistic canon Goya established -- for instance -- therefore, Picasso's work is not art.

Or, following the Popperian Joan Robinson: I establish the threshold. Art is another elephant and I know one when I see one; only this is Art; that ain't. Add this and chop that off and maybe I'll let you into the nightclub.

Sounds absurd and arbitrary, right?


Bouncer in Antwerp, 2010. [C]

One should imagine this argument could easily extend with little controversy to any science (science in its older, humbler sense: organised corpus of academic knowledge, including, among others, mathematics, logic, law, history, and, yes, philosophy).

Ah, but when it comes to the "Sciences" (in its more modern and exclusive meaning, involving experimentation) there is a wonderfully simple, fool-proof silver bullet to decide who goes to the party: Popperian falsificationism. What is absurd for science, suddenly becomes wisdom for "Science". Lucky us.

Don't get me wrong, falsificationism is just great; it allows philosophers of science speak in abstract with great authority about things they don't understand: this is Science, this ain't. And if you are a philosopher, what's not to like? The Philosopher of Science promoted to Big Brawny Bouncer at the door of the Science party.

It also compels mainstream economists to misappropriate the language of science, with their own self-interested re-definitions. Like underage teenagers with false IDs what matters is going to the party: check this box. Far from being an antidote to scientism, falsificationism enables it, gives it extra-strength.

Falsificationism has only two little problems: neither real scientists apply falsificationism, nor should they. If the very concrete 19th century astronomers John Couch Adams and Urbain Le Verrier had applied abstract, manichean falsificationism, (1) Newtonian mechanics and Kepler's laws of planetary motion would have been unfairly falsified, and (2) Neptune would not have been discovered. One can only imagine their dilemma: stick to the recently "falsified" Newton/Kepler or go back to the also falsified Aristotle/Ptolemy?

Nobody told Popper -- and in his philosophising, he never learned of the fact by himself -- but there was no large population of alternative hypotheses in a quasi-Darwinian struggle for survival: there was no other "physics", warming up in the bench, ready to replace Newton/Kepler in the soccer match. The demise of Newtonian astronomy would have meant the "extinction" of astronomy, at least until another theory appeared to replace it.

Luckily for Le Verrier, Popper was born in 1902. We don't have the same luck. We have to put up with the Popperian Procrustean bed (incidentally, and to be fair, not all philosophers of science or physicists agree with the Privileged Kibbitzer).


Okay, you don't care that there is an ideological agenda behind Popper's "abstract" falsificationism: all you care about is Science. At best, mine are antiquarian concerns. Yes?

Know this: the future of physics, for example, may well hinge on a very similar Dark problem existing today (more on this in the near future). That's the problem with silver bullets shot in the dark: anyone can be hit, not only those one doesn't like.

Image Credits:
[A] "La maja desnuda". Image in the public domain. Source: Wikimedia.
[B] "Feme nue assise". Image in the public domain. Source: Wikimedia.
[C] Bouncer giving the thumbs up. Author: Roger Price (04-09-2010). My usage of the image does not imply Price's endorsement in any way. File licensed under the Creative Commons Attribution 2.0 Generic license. Source: Wikimedia.


  1. Good Day

    This is very interesting, one of the best posts on the topic.

    Popper has an ideological agenda which does infiltrate his philosophy more so than most commentators either realise or speak about.

    Questions for you:

    1) Do you think there is a frame-work for science, even if it uses a mixture of philosophies including falcification and verification?

    2) Have you heard about Bayesian probablity and it's application to the philosophy of science (I don't know much about it).

    3) How does a realist philosophy of science integrate with any of the other philosophies particularly falcification.

    Some of those questions are a bit broad but I wanted to start a discussion by knowing your postions

    Kind Regards

  2. Hi D.K.

    Thanks for your thoughtful comment.

    I'm afraid I will be of little help.

    "1) Do you think there is a frame-work for science, even if it uses a mixture of philosophies including falcification and verification?"

    Frankly, I'm not sure I understand. Assuming you mean whether there is a common strategy to find things out in science, I'd say that chances are small. But I really don't know.

    "2) Have you heard about Bayesian probablity and it's application to the philosophy of science (I don't know much about it)."

    Yes, I have heard of it. My general impression is positive, but I know too little to form a strong opinion, one way or the other.

    "3) How does a realist philosophy of science integrate with any of the other philosophies particularly falcification."

    That's a really good question, and one exceeding by far my knowledge.

  3. The importance of falsification is widely recognized these days, and there's overlap between Popper's critical rationalism and a thoroughgoing realism.

    An adequate summary of Popper would not end at falsificationism. Also important is anti-justificationism; i.e., attempts to justify knowledge involve other knowledge that must also be justified, ad infinitum. Therefore, all knowledge is contingent and fallible. So, theories cannot be justified absolutely, but they can be falsified. Popper also argues the development of science is conjectural -- a scientist makes creative, daring hypotheses and then subjects them to testing and possible refutation.

    These are by and large compatible with realism. (For space's sake I'm leaving aside other aspects, such as his "three world" ontology, metaphysical research programs, etc.) However, they are also open to criticism. E.g., Feyerabend, in Against Method, gives examples of theories that were by all appearances falsified, yet still advanced our understanding (and in some cases were later vindicated). In other words, falsification must be flexible and open to calibration rather than naively absolute; no theory is consistent with all facts. So, "justification as absolute" is still out, and "falsification as absolute" is also not above criticism.

    Another problem with Popper is his use of falsifiability as demarcation of what is or is not science. Apart from Feyerabend's point that this dogmatically discourages some avenues of inquiry, Magpie also correctly noted that this was a political cudgel -- often aimed at Marx, because Popper mistook a method (historical materialism) for a theory. Ironically, despite Popper's own objections to positivism, this move supported the broader positivist/neoliberal weltanschauung.

    Before we go on, allow me a brief tangent. Let's make a few quick abstractions: The domain of the "Real" (Dr) contains all that is -- structures, mechanisms, events, phenomena. We might distinguish this from the domain of the "Actual" (Da), which only includes events and phenomena. Finally, we'll distinguish a third domain, "Empirical" (De), which consists solely of phenomena.

    Each of these domains is a subset of the previous one. There are mechanisms that go unactualized, but are nevertheless real and retain the potential to activate. Similarly, there are events that are not experienced. However, we can bring non-empirical denizens of the Actual into the Empirical (i.e., experience events, even if it requires special measurement equipment). Similarly, we can experimentally actualize mechanisms.

    We can say Dr ≥ Da ≥ De. But there is a special case in which, for a given object of study, Dr = Da = De -- a mechanism is actualized and experienced directly. This case is scientific experiment under conditions of experimental closure, rendering other possible causes absent or negligible.

    How does this relate? Because humanity is both subject and object in science experimental closure is impossible in the social sciences. I believe Popper recognized this, per his anti-historicist arguments. But if falsification of a theorized mechanism can only be conclusive under closure, then by this metric no study of open systems can be "scientific." This led Popper to an instrumentalist position (prioritizing prediction over explanation), effectively landing him within the positivist camp he himself opposed. This can be seen, e.g., in his (eventually retracted) characterization of Darwinian evolution as tautological. (Incidentally, evolutionary biology is a good example of a science that is more explanatory than predictive.) For more, see this SEP entry.

    So, while there is overlap, critical rationalism ultimately parts ways with realism. I would argue that critical realism/dialectical materialism has the good aspects of Popper without the weaknesses.

  4. Hello again to both of you.

    Thanks for the responses. I was going to comment but Hedlund not only mentioned what I would have said: "The importance of falsification is widely recognized these days" but expanded on it in such a way as to make me see things in profoundly different ways.

    I started my education very much in the empiricist camp, perhaps you could say "positivist" without even realising it. I don't think this was a terrible mis-step on my part, afterall the education system is geared towards it. Anyway, just added that to flesh out why I never thought of things in a certain way:

    "Evolutionary biology is a good example of a science that is more explanatory than predictive".

    I had only mildly stumbled upon the idea of the "real" and that was via Lacan (Is it related or only shares a name?)

    As for the terms "Dr" "Da" "De" not so much... Yet it does make sense.

    Much to think about. I have taken note of the books you have recommended by the way. Will start them in due course.

    Kind Regards

  5. Hey again I just thought of another question:

    How do you demarcate whether or not a theory is more predictive than descriptive and visa versa?

    Secondly: evolutionary biology, in a more evolved state (pun intended!) can very much form the basis of predictive practice. An example is Dawkin's work with evolutionary stable strategies. A game theory application to how certain biological traits are likely to be expressed.

    I guess my question is, if a theory can move from the descriptive to the predicitve state of affairs, can any theory do so?

    Without tripping up over the elephant in the room: Marxism is predicitive of certain things, it is also very descriptive of certain things. How can we say it is more so one, than the other. And if it is can it move from one arena into the next? You would surely think yes, since once you understand a description of something, you should be able to determine outcomes from a certain start point. What are is the evolution of Marxist theory? I guess it is hard to say because it is lacking a coherent progressive research programe at present.

    (Emphasis on: I am a noob (Newbie) so if I have completely mischarecterised something, let me know)

    Kind Regards

    1. @D.K. and Hedlund.

      For some reason Blogger wrongly classified Hedlund's January 8, 2016 at 11:45 AM as Spam.

      I just found out and approved it. It's at the bottom of the thread.

      Apologies for the inconvenience.

  6. Thanks, Hedlund for the long comment. Give me some time to digest it instead of replying now (I might have to write a post on it).

    I'm hardly qualified to dispute you on this, and, besides, I am sure you are right on all the elaboration you added (most of which is post-Popper).

    Having said that, I however think you may have missed something important: the use given to Popper's falsificationism and justificationism (which he intended and encouraged) as a simple rule to determine what is and what is not science. Once out of the bottle, the genie cannot be easily forced back by any amount of elaboration.

    By definition, added elaboration transforms a simple rule into a complex one.

    Anyway, I'll answer later. Cheers.

    1. @Hedlund

      My reply to your comment

      Reply to Hedlund.

      Sorry for the delay. It was a hard week for me.

  7. @DK

    I don't know a lot about Lacan, so I can't answer. I use "real" in sense of actual existence — that there is an objective mind-independent reality that may contradict appearance, which is capable of being known, though in practice knowledge must be treated as a fallible work in progress.

    I apologize for the awkwardness of the Dr/Da/De thing; those probably would have looked better with subscripts, but the tag wouldn't fly. Chapter 2 of the Collier book (here, for any other curious parties) explains it much more clearly than I did.

    Re: Predictive vs. Descriptive: Correct explanation implies a better ability to predict, and the distinction blurs in those natural sciences in which we can effect closure to our satisfaction. However, the point is more that some discount the importance of explanation and attempt to jump right to prediction. This is generally an explicit point; see, for example, Milton Friedman's 1953 essay, "The Methodology of Positive Economics" (linked in this post, which discusses the related issue of "as if" arguments), which famously argued against the importance of realistic assumptions.

    Your point about prediction in genetics is well taken. My point was simply that we can hardly know what changes a given species will undergo in the future, given random mutation, environmental effects on gene expression, etc. Yet at the same time, we can express confidence in Darwin's theory as an explanation of the reality of the development of life on Earth.

    I would argue any theory cannot become as predictive as, say, physics, since prediction is flatly impossible in some domains. If we could accurately predict our behavior under a given set of circumstances, then our foresight effectively generates a different set of circumstances that negates the prediction. If we could credibly predict where asset bubbles would form, people would take precautions against that specific course of action and the bubble would instead form elsewhere. Thus, the prediction's credibility is destroyed. Etc.

    I can say Marxian economics is inherently more descriptive because the goal of the materialist method is explicitly to grasp reality. But as I said, correct understanding implies a certain capacity for prediction. Given the impossibility expressed above, this means Marxian economics can predict "broad strokes" laws of motion to the economy — that the class struggle will fuel technical change, that said changes and the growth of capital will tendentially pull the rate of profit downwards, that this will generate a rise in debt and and increasing systemic instability, and ultimately the interconnectedness of the finance-capital-dominated economy will lead a shock to kick off a crisis, which restores the rate of profit through depressing wages and destroying capital, and then it begins again. But when it comes to predicting this or that bubble, we're not that much better off than the next guy.

    Some of the most important ways Marxist theory has evolved include theories of imperialism (developed by, e.g., Bukharin, Lenin, Luxemburg, et al.) and neocolonialism (via Kwame Nkrumah, Walter Rodney, et al.), if that is what you mean.


    You say you’re "hardly qualified to dispute" me, but I find that a tad problematic; I'm hardly qualified not to be disputed. :)

    On simplicity, you’re right. Arguing against Popper is an uphill battle, since his ideas are very well and simply communicated, and they have definite value in the course of development of the philosophy of science to boot.

    If I ever get around to starting a blog, I'm going to bust my hump to keep my discussions simpler than the wordy dissertations I tend to toss around.

  8. Magpie, I comment here (on Jason's blog) about the timing of the recent announcement of a hypothesized "9th planet" and your use of the discovery of Neptune as an example here.

    1. @Tom Brown.

      Thanks, Tom. By coincidence, I was about to post about the 9th planet.

    2. And while it's on my mind, I discuss with Jason here a 2012 article by another physicist (David Deutsch). Why it comes to mind is Deutsch's advocacy of Popper (not central to the article, but he brings Popper up more than once in a favorable light).