A canna’ change the laws of physics

Scotty, The Naked Time, stardate 1704.3, Episode 7

Lost In Translation? Part I: What Is Incommensurability And Why Should I Care?

Posted by apgaylard on March 1, 2008

I have some family and friends who are well disposed towards CAM in general and homeopathy in particular.  I’ve noticed that discussions of the relationship between their views and science are fraught.  At times it’s because they don’t really understand the CAM ‘therapy’ or science; at other times it’s like we’re talking in different languages.

The latter can be thought of as incommensurability: the lack of common units of measure shared by concepts that we’d like to compare.  Apparantly, this idea reaches back in time to Pythagorean geometers who had the notion that any two lengths were measurable by multiples of some common unit, hence are “commensurable“. One of their number subsequently discovered that this is not so, legend has it that the discoverer of “incommensurable” quantities (irrational numbers) was killed by his fellows.  Incommensurability is only a little less controversial today!

The cry of “incommensurable” is often heard when CAM modalities are threatened with a fair test of their claims, it’s become a standard ‘defensive’ gambit.  How valid is this defence along with the common invocation of the work of philosophers like Kuhn?

I thought that it might be interesting to explore this idea in the context of my on-going investigation of Thomas Kuhn’s “The Structure of Scientific Revolutions” (SSR, 1996).  To try and keep the discussion manageable, I’m going to present it in two parts.  In this instalment I’ll try to illustrate the abuse of the concept.  In the second, I’ll explore what Kuhn really said and tease out some (hopefully) useful lessons.

So, why bother spending time thinking about an academic concept from the philosophy of science?  The short answer alluded to above, is that people are using it in an attempt to exempt their favoured therapies (and even social policies) from proper scrutiny: it’s become a post-modern excuse.

Let’s just review a few of the many examples where the incommensurability gambit is deployed.  For instance, commenting on the implications of Evidence Based Medicine (EBM) for Traditional Chinese Medicine (TCM), Wainwright Churchill (136kb pdf) offers up the common CAM version of incommensurability:

Chinese medicine does offer an explanation of how acupuncture works, but this is within a separate and incommensurable paradigm from the biomedical [] Many scientists believe that the scientific paradigm has absolute truth value, but philosophers such as Thomas Kuhn reject this. If no paradigm does have absolute value, there is no absolute basis with which to judge another paradigm. Any paradigm will appear limited or incorrect from the perspective of a different paradigm, so Chinese medicine will seem incorrect from a biomedical point of view, and vice versa

[…] different paradigms are incommensurable, meaning that their frames of reference are different so that the world depicted in one paradigm cannot be translated into another, just as not everything that can be said in one language can be said in another.”

“[…] Because of the incommensurability of paradigms, any CAM practised its original way cannot be the same as its biomedical version – they must remain two separate worlds. If the biomedical paradigm is adopted, the system will have the characteristics of that paradigm – materialistic, mechanistic, reductionist, linear-causal, and deterministic (many of the characteristics that inspire people to use complementary medical therapies). Biomedical research into that discipline will reflect the biomedical worldview. Statements such as ‘Yin and Yang do not exist’ will follow as unavoidably as night follows day.”

Is he right to state that Kuhn’s incommensurability thesis means that their can be no translation between ‘paradigms’?  Are there really no ways to compare TCM interventions with proper medicine?

(It should be noted that this quote demonstrates a fatal inconsistency: Churchill states that the world described by one paradigm cannot be translated into another: “just as not everything that can be said in one language can be said in another”.  Here he mistakenly equates complete incommensurability with partial translation.)

Sometimes I wonder whether incommensurability is just used to obscure the wishful thinking present in CAM ‘therapies’.  For instance, Coulter and Willis invoke incommensurability rather than the downright crankiness present in homeopathy as the reason for its incompatibility with modern medicine.

In sociological terms, the issue is one of commensurability of paradigms. To argue for complementarity or integrativeness implies that the knowledge bases of the paradigms are commensurable – that is, they are not logically inconsistent. For example, in the paradigm that we now call conventional scientific medicine, dilution of a therapeutic substance weakens its potency. However, in the homoeopathic paradigm, dilution – even multiple times so that few molecules of the original substance remain – actually increases its potency. Presumably dilution can’t do both. The paradigms are incommensurable, and so the possibilities for combining treatments based on the two paradigms must be limited.

On other occasions incommensurability seems to be used as a fig leaf to cover the embarrassingly primitive nature of CAM beliefs.  Take this extract from a discussion of the merits of Kampo (a Japanese version of Chinese herbal medicine):

I would like to point out one of the biggest of such obstacles before I describe the details of Kampo medicine. Stated in a way that may astonish Western readers, with Kampo it is simply impossible to design a clinical trial which asks the question ‘is such-and-such Kampo prescription effective or not for chronic hepatitis?’ Out of ignorance, this approach has sometimes been taken in the West for the testing of Chinese herbal medicines, and one of the main themes of this lecture series is to persuade the readers that this kind of approach is not appropriate. As the readers may anticipate, the reason is the incommensurability of paradigms. The concept of chronic hepatitis simply does not exist in the Kampo paradigm.”

Is it really so that Kampo cannot be tested in a scientific (fair) trial?  Does this mean that chronic hepatitis is just called something else in the ‘Kampo paradigm’?  If not, shouldn’t this incompleteness be taken as an indication of the inferiority of this ‘paradigm’?

Hepatitis is an inflammation of the liver and can be caused by viruses, toxic agents and auto-immune disorders; are these agents really controlled by our beliefs?  If adherents of the Kampo paradigm were injected with the hepatitis B virus (HBV), for example, would they be immune?  How come infection rates for hepatitis B are up to 15% in South East Asia and China where Kampo (or Chinese herbal medicine) has at least some influence, but less than 1% in the non-Kampo UK?  Taking Japan, the home of Kampo, according to the CDC (quoted in this report (1430kb pdf) ) it has an intermediate rate of HBV infection (2-8%), whilst the UK’s is rated low (<2%): clearly no correlation between not having the infection as part of a ‘paradigm’ and not having the infection!

(Please note that I’m not refering to a necessarily general view of Kampo here.  I’m just discussing this author’s use of the incommensurability gambit.)

To subordinate the facts to a philosophical stance is clearly perverse.

It’s when incommensurability is used (267kb pdf) as a Trojan horse for introducing an ‘anything goes‘ relativism into state funded healthcare that the dangers become most obvious. (My thanks to the Freeborn John blog for the reference, contained in a fascinating discussion of the perversion of the word qualitative by some social ‘scientists’.)

What is striking about the plea of incommensurability is that it is frequently used to justify the existence of a barrier than cannot be crossed:  science can’t prove my ideas to be wrong (or even judge them) because they are part of an incommensurable paradigm.

This total barrier to communication and comparison is clearly self-refuting: if I can’t compare your theory with mine and decide which is best; how come you can reject mine?  Or, if our ‘paradigms’ are truly incommensurable how can you tell me exactly what it is that I ‘cannot’ understand?

This type of incommensurability also seems, at times, to be defined as one-way: science cannot evaluate my views as they are part of an incommensurable paradigm, but they are nevertheless right!  Here is a particularly outrageous example, courtesy of a homeopath who inhabits the halls of quackademe.

This is clearly a vacuous attempt to avoid the self-refuting nature of complete incommensurability: who gets to decide which way the one-way traffic flows?  Why can’t it be one-sided in favour of science rather than CAM, for example?  This ends up as nothing more than a partisan incommensurability.

So if the CAM version of incommensurability is so empty, did a man as bright and accomplished as Thomas Kuhn really advocate it?  If not, what was he really trying to get at?  When allegedly Kuhnian incommensurability is invoked – or disparaged – what are some useful points to bear in mind?

Those questions will be explored in part II.

18 Responses to “Lost In Translation? Part I: What Is Incommensurability And Why Should I Care?”

  1. caisersoze said

    This is the usual claim for woo proponents: “science cannot evaluate my woo”. It is ridiculous. If something has any effect at all then by default it is possible to test it with science. Otherwise, it makes no difference anyway and it is just an unfalsifiable and useless hypothesis. We may as well ignore it.

    “the scientific paradigm has [no] absolute truth value”. No scientist ever said that the scientific method is absolute or gives you absolute results. All scientific results are tentative. Science has proved the best method to lead us towards the truth. If someone can devise a better method then he is more than welcomed to teach it to us and show us why it is better. Presumably such different method would make limited use of logic and rational thinking?


  2. […] apgaylard wrote this today. I think it is worth reading. Here is a little snippet:“In sociological terms, the issue is one of commensurability of paradigms. To argue for complementarity or integrativeness implies that the knowledge bases of the paradigms are commensurable – that is, they are not logically … […]

  3. draust said

    I love the phrase “the incommensurability gambit”, AP.

    As a former chess geek, Can I term your response to it the

    “Kuhn’s Gambit Declined”?

  4. […] Grrrrrrl! – Vox has something worth reading today (Lost In Translation? Part I: What Is Incommensurability And Why…)Here’s a brief bit, but follow the link for the rest.Out of ignorance, this approach has sometimes been taken in the West for the testing of Chinese herbal medicines, and one of the main themes… […]

  5. apgaylard said

    Couldn’t agree more. The assertion about “absolute truth value” is nothing more than a straw man; as you say no sensible scientist would make such claims.

    Wish I’d thought of that, it would have made a much snappier title. I should have played more chess!

    Thanks both for taking the time to comment.

  6. […] Original post by apgaylard […]

  7. dvnutrix said

    Nicely done. Good examples, thought-provoking and an enjoyable read.

    I look forward to part II.

  8. Andysnat said

    I’ve come up against this very thing today, in the Hpathy forum, except now I have a name for the problem.

    They keep telling me to learn their science if I want to ask them how homeopathy works, and unfortunately their belief system isn’t a science at all.

    It is all very frustrating, and quite nice to see some sound thinking going on here.

  9. pleick said

    Between science and (most of ) CAM, the really incommensurable parts are the language and the ways of thinking.

    Since I too am a (still active) chess geek, I have to follow the gambit idea a little bit… Many gambits are known to be unsound and to backfire against a skilled opponent.
    Here, declining “Kuhns Gambit” (as offered by proponents of CAM) is the right antidote.

    Coulter & Willis nearly get to the heart of the matter when they write that “Presumably dilution can’t do both. The paradigms are incommensurable […]” This, of course, begs the question whether homeopathy and evidence-based medicine (“conventional” medicine) can both be true. Regarding this, there seems to be a lot of confusion, but also some double-talk among homeopaths.
    If homeopaths want homeopathy and conventional treatment to complement each other, both need – in some way – to be true. Generally, this is what they publicly state… (although some homeopaths openly reject conventional medicine). But then, both methods cannot meaningfully be incommensurable.

    Correct me if Kuhn says something different (haven’t read his work…), but if two scientific theories are incommensurable, then:
    – if they share common subject matters (as homeopathy and conventional medicine certainly do!), this situation should only be temporary or partial. The optimal description of the relevant phenomena has not been found yet. But eventually, one of the two theories (or a different one altogether) will become the dominant paradigm, and the alternatives will be discarded.
    – they do not treat the same subject. An example from physics would be general relativity and quantum mechanics: combining the two theories is exceedingly difficult because of their radically different mathematical structures. While both are certainly “true” (in the scientific sense of the word), they describe very different classes of phenomena.

    I don’t see any other kind of scientific incommensurability. So, is “Kuhns Gambit” anything more than an outright rejection of the scientic method, nicely wrapped up in inconspicuous language?

  10. pleick said

    … or do are we waiting for the GUT (Grand Unified Theory) of homeopathy, physics, chemistry and biology?

  11. apgaylard said

    dvnutrix: Thanks, part two will be ready next week!

    Andysnat: I’ve had some similar circular discussions. Some homeopaths or fellow-travellers can’t seem to make their minds up. The invocation of explicitly Kuhnian incommensurability implies that we are comapring two scientific theories. If they don’t think they’re doing science (or something very similar) then this concept doesn’t really apply and they can’t use it as an excuse!

    Thanks to both for your comments.

  12. apgaylard said

    Thanks for your interesting comments. Kuhn’s view of incommensurability (which I’ll explore in part two) was always partial, never total. It developed a local character as well. (He worked on defining this idea for the rest of his life; eventually defining incommensurability as existing between branches of hierarchical lexicons underlying scientific ‘language’)

    Of course, he was concerned with comparing scientific theories. His ideas come from a historical analysis, looking at the classic examples from physics and chemistry (Aristotle, Ptolomy, Copernicus, Newton, Dalton etc.). I think that to apply them outside of the physical sciences requires extensive justification. The body of knowledge would need to be shown to progress in a very similar manner and share many of the features of science.

    Homeopathy clearly doesn’t meet the criteria for being science, but if it’s advocates want to make claims about being scientific, having ‘paradigms’, having theories that are subject to Kuhnian incommensurability than I’m happy to evaluate those claims.

    Given all that, Kuhn also thought that competition between different schools of thought in science was quickly settled. His revolutions, where a new ‘paradigm’ supplanted an old one, seem to be thought of as fairly swift as well.

    Also Kuhn’s analysis was of competing paradigms [disciplinary matrices as he later called them]. So homeopathy (or elements of it) are, as you say, in direct competition with parts of medicine (pharmacology, germ theory, etc.). In fact homeopathy pre-dates virtually all scientific medicine. They competed in the past (Germany, US particularly) and homeoapthy lost. I think in Kuhnian terms the competition is over. Homeopathy has lost to the challenges presented by dose-response relationships, germ theory, the massive success of antibiotics, vaccination and many, many more.

    To quote Kuhn [p.204], “as argument piles on argument and as challenge after challenge is sucessfully met, only blind stubbornness can at the end account for continued resistance”

    On the other hand General Relativity and Quantum Mechanics don’t really compete, hence the percieved need for ‘Unification’.

  13. pv said

    So, basically no-one is deluded or ill-informed, or ignorant, or dishonest because these things just aren’t possible. One isn’t useless at science subjects, rather it’s a paradigm not commensurable with the subject one is good at – namely homeopathy.
    Must try that defence if ever I’m up in front of the Beak having been done for fraud.

  14. draust said

    I think most scientists are very impatient with the way they are typically presented with Kuhnian ideas, and particularly with that group of the science studies lot who essentially seem to use Kuhn to justify across-the-board relativism as a great received truth. (Which is also where the homeopaths and their ilk stand).

    As far as most science “practitioners” think of it, a lot of the time science really does work by gradual improvements in understanding based on empiricism – experiments underpinned by falsifiability / veriability (though few can be bothered with the philosophical arguments).

    Most would accept parts of Kuhns’s argument, I think, in the sense that (i) the science-proceeds-by-infinitesimal-gradualism view struggles to take in “big jumps” in understanding, which one MIGHT call paradigm shifts; and (ii) Kuhn correctly identifies that there is a certain inbuilt resistance to junking theories that are dominant in the field. (Also known as “human nature”)

    In this sense you could simply view Kuhn as adding a new element to the “gradual accretion of better ideas” paradigm.

    However, the way Kuhn’s thinking is interpreted by some sociologists to say “science is all a construct of a belief system” is rejected by every single working scientist I have ever met. Kuhn is quoted somewhere as saying that he didn’t believe in the idea of one theory being “a better representation of how the world really works” than another. The standard response to this among scientists is eye-rolling, hoots of derision and perhaps a quoting of Richard Dawkins’ line about “show me a relativist [in an airplane] at 35,000 feet and I’ll show you a hypocrite”

  15. Nice intro – waiting for part 2. Though in some respects (the alternative practitioners) citing Kuhn regarding philosophy of science is similar to citing Newton on physics. He represents an important idea, but one that has been largely superseded. But as later ideas are less easy to mis-represent I don’t expect the purveyors of woo to understand this.

    A couple of observations based on the excerpts you have highlighted. Firstly the people using the term incommensuarable clearly don’t understand it. As you say it seems to be a rhetorical gambit, one that means “I refuse to play by your rules, so you aren’t allowed to judge me by them”, rather than a position with any logical force.

    Secondly they seem to confuse Kuhn with a strong sociological programme in the study of science, that is as some sort of strong relativist. This couldn’t be further from the truth (as you clearly knwo based on your comments).

  16. […] Sinosplice: Life in China, Learning Chinese always has something good to say. I like this one posted earlier today. Follow the link for the whole thing. Excellent post from Gaylard.  bookmark this on del.icio.us – posted by bengoldacre and saved by  others […]

  17. metherton said

    It’s also rather strange to use as an analogy that “not everything that can be said in one language can be said in another”. If this means that there are some things that are difficult to translate, that’s trivially obvious. But if it is supposed to mean that there are statements in languages that cannot be conveyed in another by any means, then a moment’s reflection will demonstrate that this is nonsense. In any case the analogy demonstrates the partial/complete incommensurablity point, as apgaylard points out; for the analogy to have any meaning, there would have to be a language that was utterly untranslatable.

  18. apgaylard said

    Thanks for all the comments. Part II will address some of the points made; in the meantime just a few observations.

    draust: Completely agree with your position. The big problems with Kuhn are his ambiguity, which have opened the door to wild interpreations that he disavowed, and his aversion to the idea that improved theories are a better approximation to the truth. The latter view makes him seem to be a relativist, a label he rejected.

    His idea of resistance to change is a bit more interesting than the “scientists are stick-in-the-mud’s” interpretation some put on his work. In his scheme resistance stops most scientists wasting time on wild goose chases and enables them to get some useful work done.

Sorry, the comment form is closed at this time.

%d bloggers like this: