Franklin, J.  (2000)  'Thomas Kuhn's irrationalism', in The New Criterion, Vol 18, No 10  [online]

[This is a witty and ironic piece criticising Kuhn's central arguments and the popular caricatures of them, and trying to explain why they are so popular among people who don't understand science. I'm going to reduce the wit and irony to a few basic points as usual. It is also worth noting that a mirror image of this critique could be levelled at Franklin himself -- Kuhn ignores realist arguments, Franklin ignores contructivist ones; Kuhnians ignore actual science, Franklin ignores actual sociology and history -- and so on]

 Kuhn's work turns out to be the most frequently cited book in the US Arts and Humanities Citation Index. It is popular with politicians from Bush to Gore. This can be understood simply by asking  'What would the humanities crowd want said about science?' (no page references on this downloaded copy). They would want science to be no different from humanities, certainly no better or more rigorously grounded:  'Kuhn declared logic outmoded and replaced it with history'.

 He actually did this as a teaching device initially, having been given the task of teaching non-scientists  (policy makers in particular) about science at Harvard. His sponsor, the President of Harvard at the time, realise that tasters of natural science were not effective. Discussing earlier episodes of scientific development would prove more exciting and illustrate the processes of discovery. Kuhn's teaching did not prove very successful, and he lost tenure -- the famous book can be seen as a kind of revenge.

 The caricatured version says that an established paradigm leads to normal science until anomalies accrue which  'eventually make the paradigm unsustainable'. A revolutionary phase ensues, but the new paradigm does not prevail really until the old adherence to die out. Advocates of the caricature, who include Fukuyama, believe that this means there can simply be no scientific knowledge of nature at all. Kuhn himself denies that he said any such thing, but this leaves only two further inconsistencies and means that  'the caricatured has a historical career considerably more vigorous than the original, whose qualifications would have lessened its appeal'. The caricature renders the history of science as a simple literary plot,  'the story of the Morte d'Arthur, of the peace and order and its ageing King, the virtue undermined by internal corruption, falling to the challenge of the vigorous and bloodthirsty young challenger'. It also reveals a structure known as  'theomachy... [where]... what was previously thought to be a continuous and an interesting succession of random events is discovered to be a conflict of a finite number of hidden gods... who manipulate the flux of appearances to their own advantage, but whose imaginations may be uncovered by the elect to whom the key has been revealed' [Freud and Marx are cited here -- it could be a perfect description of British Cultural Studies as well].

 Kuhn's work used selective examples, ignoring sciences where there was a progressive accumulation of established results, such as  'ophthalmology, oceanography, operations research, and ornithology, to keep to just one letter of the alphabet'. It appealed to social scientists in particular who had been embarrassed earlier by the theoretical fights in the subject, but now thought these lay at the very heart of science. Even natural scientists could see themselves as potential revolutionary heroes. The thesis appealed perfectly to the spirit of the Sixties.

 There are ambiguities, of course. Unsustainability may be a matter of logic or psychology. The first option delivers us back to conventional views of science: the second option fails to explain why psychological rejections actually changed scientific theories. There are left-wing criticisms, such as those offered by Fuller -- these point out that Kuhn failed to connect the progress of science with the military-industrial complex and other social processes. Kuhn's work offered a harmless diversion, according to Fuller. This points to one theme in Kuhn at least --  'its talk of  "revolution"  is very harmless; the revolution is in the past, against the previous paradigm, and no present entities have anything to fear from it. Like the violence and horror film, it's all virtual and it's over when you come out into the light'. However, Franklin admits that dominant paradigms do attract better funding and resources.

 Fuller also connects the notion of paradigm to a lot of earlier cultural work, including that of Piaget, but he still does not refer to the actual practice of science. As a result, he fails to realise how logical actual science is. This is also ignored by much of science studies, where Kuhn's work prevails. There has been a backlash on the constructivism and relativism of the approach, but the core argument remains --  'that usually different scientists will come to agree on what they say about reality, but that the reason for this is not in reality but in the scientists, in particular in their social relations'. Franklin points out that this simply ignores the observable effects of scientific arguments  [I am not sure it does]. Fuller argues that knowledge must be embodied in linguistic and social practices, and that it is this that reliably transmits knowledge at the least.

 Franklin says this is a bad argument because it simply rules out the possibility that we can know things as they are in themselves  [this is the crux of the debate of course -- Franklin is wrong to assert that Kuhnian positions are only a product of the Sixties or whatever. Constructivism does have a philosophical argument too.] Franklin illustrates his point by quoting a caricature: "we have our eyes, therefore we cannot see". There is a further analogy  [shouldn't we be suspicious of caricatures and analogies being used at this point?] -- a calculator successfully performs arithmetic because its circuitry is assembled to deliver the right result, but successful performance also depends on more abstract arithmetical facts such as that numbers do indeed add together  [this seems dubious grounds to me, assuming that arithmetic simply mirrors some real-world equivalents]. Franklin admits that  'there needs to be some philosophical story about why causes cooperate with reasons' as in this case, but says that Kuhn simply denies that this requirement is necessary. As a result,  'the worst effect of Kuhn... has been the frivolous discarding of the way things are as a constraint on theory about the way things are'.

 What is needed is a proper account of science, showing how evidence is related to conclusion, how measurement works and so on.