Andy Skuce, in On and against method and process is (to me) bizarrely keen on Paul Feyerabend (though presumably he discards the numerous cites to Lenin, denigration of modern medicine, and all the really wacko stuff). I kinda tend to mix F up with the other out-of-their-depth French folk like Latour that Sokal and Bricmont shredded but that’s probably unfair; either way my suspicion is that if you’re interested in a criticism and analysis of F, you’d be better off with Sokal and Bricmont. But that’s rather long; who has the patience nowadays to carefully read reasoned expert critiques when you can read a short non-expert rant?

Anyway, F says stuff like

In this chapter I shall present more detailed arguments for the ‘counterrule’ that urges us to introduce hypotheses which are inconsistent with well-established theories. The arguments will be indirect. They will start with a criticism of the demand that new hypotheses must be consistent with such theories. This demand will be called the consistency condition.1

Which makes me go “WTF?” This appears to be a strawman. Note 1 is The consistency condition goes back to Aristotle at least. It plays an important part in Newton’s philosophy (though Newton himself constantly violated it). It is taken for granted by many 20th-century scientists and philosophers of science. Which also makes me go “WTF?”. Aristotle wrote any amount of bollox, since when has he been an arbiter of physics of philosophy – I thought they gave up that kind of hero-worship in the middle ages. Or perhaps the end thereof – Copernicus, perhaps. But who cares exactly when, its long gone, so why is F digging him up?

It appears especially stupid when laid against the great modern theories like relativity or QM. The founding hypotheses of relativity are inconsistent with Newtonian mechanics (by which I mean, in case I’m not clear, accepting the constancy of the speed of light and discarding One True Time), which is after all much of the point. Pretty well any “paradigm breaking” science is bound to do this. Only the “normative” infilling type science is consistent with what went before. Perhaps F should read Kuhn. The wiki article on F faithfully parrots One of the criteria for evaluating scientific theories that Feyerabend attacks is the consistency criterion but again makes no attempt to establish that this criterion exists.

Then we get

…Newton’s mechanics is inconsistent with Galileo’s law of free fall… Galileo’s law asserts that the acceleration of free fall is a constant, whereas application of Newton’s theory to the surface of the earth gives an acceleration that is not constant but decreases (although imperceptibly) with the distance from the centre of the earth.

Firstly, I think this is uninteresting nit-picking. But secondly, it is even true? One of the top google hits for G’s law tells me it is , in the absence of air resistance, all bodies fall with the same acceleration, independent of their mass. Which I would regard as a correct statement of G’s law, and which doesn’t suffer from the (nit-picking) flaw that F mentions (though it does suffer from another nit-picking flaw which the article does mention, but which F is too stupid to know, that if you measure the acceleration from the sfc of the Earth, since the Earth imperceptibly moves, the acceleration does slightly depend on the mass of the object falling).

Anyway, skipping over more stuff we come to

No theory ever agrees with all the facts in its domain

I’m not sure that’s true. GR is consistent with everything in its domain; not with QM, but that’s not in its domain. Or conversely, QM with everything in its domain. I might be wrong about that; if I am, do leave a comment.

But quickly we discover that F doesn’t mean facts when he says facts. No, he means “facts”. I don’t think he ever clarifies this point (there are lots and lots of words in F’s text; you can’t expect me to have read all of them), but what he says makes no sense without it. For we have Thus the Copernican view at the time of Galileo was inconsistent with facts so plain and obvious…; but of course it wasn’t. F doesn’t provide any examples of the “facts”, so its quite hard to guess what he’s going on about. Instead he skips on to Newton’s theory of gravitation was beset, from the very beginning, by difficulties serious enough to provide material for refutation. Notice how “facts” have silently become “difficulties”; the difficulty here alluded to was the possible instability of the solar system under the theory; but that’s different. And the stuff he then writes about (special) relativity in support of his thesis is silly.

There are then pages and pages of stuff I largely skipped about Galileo and rocks-fall-downwards-even-though-the-Earth moves. There must be something in there but its very thin; not worth the number of words spent on it. He seems overly obsessed by G; but other people have more interesting things to say about G.

The idea that there will be individual observations that contradict established theories certainly doesn’t appear interesting now.. Ditto the idea that people (most clearly Einstein) will wave away contradictory “facts” that “contradict” their beautiful theories. Perhaps these ideas were more exciting when this stuff was first written.


* More pointless “philosophy” is avilable if you want it; “Must science be testable? String wars among physicists have highlighted just how much science needs philosophy – and not just the amateur version”; Massimo Pigliucci. Pharyngula seems to like some of his earlier stuff, though.


  1. #1 Mark Pawelek
    Kent, UK

    I’m definitely more comfortable with your approach than Andy Skuce (Critical Angle). Don’t really want to talk about whether there’s “no scientific method” or F makes sense. There’s obviously method, and, I’d say rules of rule finding: A theory should be the simplest it can be while remaining consistent with the evidence. Beautiful (AKA symmetrical) seems better.

    Doh, … Me too thick. I thought the problem with Newton was instantaneous action at a distance. He knew it wasn’t “true” as did fellow natural philosophers. He justified it pragmatically as in:- if we assume this, then the theory has predictive power.

    [Yes, it is a problem with his theory. But of course it wasn’t a problem in the sense of being contrary to any facts anyone knew at the time -W]

    Getting back to Copernicus. It’s said his model was no better (worse?) than Ptolemy’s at prediction. Yet Copernicus’s new theory has beauty (symmetry) somewhat lacking in Ptolemy’s model.

    Which leads me to: Carlo Rovelli who criticizes scientists for giving up on philosophy of science (PoS) – in his Anaximander book. Too many scientists stop at Popper’s falsification. Popper is OK as far as he goes; he’s not wrong. There is more to PoS than him. I’m unhappy with Feyerabend, Latour, or even Kuhn telling me what the “more” is. I’ve rather proper scientists improved PoS.

    [Most scientists, I think correctly, have no time for PoS. Because it is dominated by Ps, and it doesn’t help anyone get their science done -W]

  2. #2 Andy Skuce

    I don’t think anyone should take Feyerabend completely seriously. He wrote in his preface:

    In 1 970 Imre Lakatos, one of the best friends I ever had, cornered me at a party. ‘Paul,’ he said, ‘you have such strange ideas. Why don’t you write them down? I shall write a reply, we publish the whole thing and I promise you – we shall have lots of fun.’ I liked the suggestion and started working. The manuscript of my part of the book was finished in 1972 and I sent it to London. There it disappeared under rather mysterious circumstances. lmre Lakatos, who loved dramatic gestures, notified Interpol and, indeed, Interpol found my manuscript and returned it to me. I reread it and made some final changes. In February 1974, only a few weeks after I had finished my revision, I was informed of Imre’s death. I published my part of our common enterprise without his response. A year later I published a second volume, Science in a Free Society, containing additional material and replies to criticism

    This history explains the form of the book. It is not a systematic treatise; it is a letter to a friend and addresses his idiosyncrasies. For example, Imre Lakatos was a rationalist, hence rationalism plays a large role in the book. He also admired Popper and therefore Popper occurs much more frequently than his ‘objective importance’ would warrant. Imre Lakatos, somewhat jokingly, called me an anarchist and I had no objection to putting on the anarchist’s mask. Finally, lmre Lakatos loved to embarrass serious opponents with jokes and irony and so I, too, occasionally wrote in a rather ironical vein. An example is the end of Chapter 1: ‘anything goes’ is not a ‘principle’ I hold – I do not think that ‘principles’ can be used and fruitfully discussed outside the concrete research situation they are supposed to affect – but the terrified exclamation of a rationalist who takes a closer look at history. Reading the many thorough, serious, longwinded and thoroughly misguided criticisms I received after publication of the first English edition I often recalled my exchanges with Imre; how we would both have laughed had we been able to read these effusions together.

    Perhaps I should have made it more clear that I enjoyed reading Feyerabend because he was an entertaining gadfly and a refreshment after reading the earnest, dull and dehumanized stuff that comprises so much of the writing on science methods. Admittedly, I’ve got a soft spot for crazy European Marxist/anarchist provocateur intellectuals (like Slavov Zizek), providing they are not pompous.

    [I think the “don’t take me seriously” bit is irresponsible. He simply uses it as license to be sloppy. Like Trump. Any time anyone calls him on his lies, he just says “ha ha that was a joke, have you no sense of humour?” F seems the same. When people you like do it, you excuse it, because you like them. When people you don’t like do it, its intolerable -W]

  3. #3 Andy Skuce

    Oops. That last paragraph was me, not PF, demonstrating my incompetence in closing HTML tags properly.

    [No problem, I fixed it up -W]

  4. #4 David B. Benson
    United States

    I am quite sure we can ignore whatever F wrote to our benifit.

    Dogun, on the other hand, …

  5. #5 ...and Then There's Physics

    Okay, this is horribly confused

    Galileo’s law asserts that the acceleration of free fall is a constant, whereas application of Newton’s theory to the surface of the earth gives an acceleration that is not constant but decreases (although imperceptibly) with the distance from the centre of the earth.

    As your link indicates, it’s that all objects would fall with the same acceleration in the absence of air resistance. It’s not that the acceleration itself has to be constant. Of course, for any realistic height above the surface of the Earth, it is close enough to being constant.

  6. #6 Raymond Arritt

    I’m glad somebody reads all this “philosophy of science” stuff so I don’t have to.

    According to Wikipedia Feyerabend was an idiot.

  7. #7 Raymond Arritt

    Oops, linked to the wrong section of WP. I meant this one. (Who’s the idiot now?)

    [Tee hee, yes you do look silly. I could fix it, but where would be the fun in that?

    On a more serious note, I’d missed the astrology bit. The numbers cites of Lenin are harder to miss, you might like those -W]

  8. #8 Tom C

    I think you are confused about what F wrote. He is referring to the dependence of velocity of free fall on the distance above the earth (r in Newton’s law of G). For all distances of practical importance in experimentation r is the same due to he value of distance to earth center.

    [I don’t think I’m confused about what F wrote. I suspect F is; he doesn’t appear to have any great reputation of respect for accuracy; even his supporters agree on that -W]

    Using Gallileo’s law, therefore, as a basis to understand planetary movements would led to misunderstanding, because the distances involved re on the order of, r larger than distance to earth center.

    So, a first step to developing an understanding of such problems would be to start with a hypothesis that contradicts Galileo (G depends on r) and see where it leads.

    William, why this bizarre antipathy to Aristotle? Ok, he got experimental physics wrong 1800 years before there were good measuring instruments. So what? Who was right about everything? Who of his contemporaries, anywhere in the world, were “right”?

    [Why is antipathy to A for the rules of science “bizarre”? Why would using him as a source make any sense? -W]

  9. #9 Eric Lund

    From these excerpts ISTM that F is confused (if not entirely ignorant) about the concept of successive approximations. The idea is quite ancient: the epicycles in Ptolemy’s model of the universe are an example. And it is an essential part of mathematics: anything involving an infinite series, e.g., computing π to arbitrarily many digits, uses this concept.

    The same goes for falling bodies. Galileo’s statement about equal falling rates, even under F’s interpretation, is a reasonable approximation for objects thrown off buildings: even the Burj Khalifa (let alone the Leaning Tower of Pisa, the tallest building Galileo had to work with) has a height roughly 10^-4 times the Earth’s radius. Newtonian mechanics is the next level of approximation, in that it explicitly includes the inverse square falloff of the force. It works quite well as long as you are dealing with objects of sufficient size, moving at a low speed compared to the speed of light, in a region of space with a metric not significantly different from g00=-1, g11=g22=g33=1, off-diagonal terms equal to 0. You need QM for the first situation and GR for the other two. It’s like adding more terms in a converging series: for all practical purposes you ignore anything too small for your apparatus to measure. When these effects are big enough for you to notice, e.g., relativistic frame dragging in GPS and other satnav systems, you include those terms in your calculation.

  10. #10 David B. Benson
    United States

    I once had a pleasant conversation with Patricia Churchland. Mostly we discussed Sir Karl, who had just retired at the time, and whether Patricia would take up his chair. Feyerabend never came up.

  11. #11 Kevin O'Neill

    I don’t think F was particularly concerned whether he got the details correct. His point was that science is constantly changing and that to teach it properly one should introduce theories that we now know as false, but that were once thought to be ‘true.’

    Others have taken F’s views and encapsulated them perhaps a little more succinctly. From The principles of teaching science based on the ideas of Feyerabend regarding the nature of science and the manner of its expansion

    The recommended principles include understanding scientific theories in the context of history, freedom of learner in scientific research beyond common beliefs and present theories and based on aesthetics, meta-physical and even religious tendencies, incommensurability and non-falsifiablity along with the testability of the available theories based on criteria such as coherence, the degree of adaption with the “posteriori” experiential conditions and adventurous approximations, conceptual teaching instead of algorithmic teaching and subject-based teaching.”

    Just as many economists flailed around in the aftermath of the financial crisis – recreating economic fallacies that had been dispensed with a hundred or more years ago – a lack of historical knowledge in a field can lead one down deadend research paths and/or be ruinous to one’s reputation.

    F would say that understanding the Aristotleian view and the Newtonian view and our current understanding should reinforce the idea that science constantly changes; scientific ‘truth’ is ephemeral, yet it is almost always building on the past.

    Among the various philosophers of science I find F the least annoying :)

    [If F had indeed said that, I would have been more sympathetic. But he didn’t. He said a pile of other things, many of which appear to be wrong. Are you sure you’re not just projecting your views onto him? -W]

  12. #12 Graeme Bird

    [Spammed -W]

  13. #13 Russell Seitz
    Squibnocket MA

    The Sokal- Latour debate in London was great fun, as the hecklers included John Maddox.

New comments have been temporarily disabled. Please check back soon.