Abstract
Thomas S Kuhn. The structure of scientific revolutions. 2nd edn, Chicago: University of Chicago Press, 1970
With hindsight, the 1970 second edition of Thomas Kuhn’s The Structure of Scientific Revolutions has had more influence in the social sciences than in the natural sciences to which it was directed. His notion of ‘paradigm’ is quoted frequently in the social sciences. Yet there are two Kuhns – the benign and the baleful.
He is one of those seminal thinkers who is often misunderstood despite his clarity of writing – principally, I suspect, because many of those who use the notion of paradigm may not have read Kuhn, let alone had a ‘second look’. For many, the notion of paradigm is associated with – is even an excuse for – relativism in both epistemology and ontology. Of course, Kuhn and more contemporary sources of relativism – the likes of post-modernism and post-structuralism – are worlds apart by intellectual origin. Yet a certain tendency within academe is so concerned to reject a straw-man version of positivism, and to over-privilege ethnographic knowledge of how individuals and cultures ‘understand', that relativism from whatever source ends up ruling.
Kuhn's concern by contrast was to explain scientific progress, not deny the basis for it. As a philosopher of science, Kuhn's idea was that scientific progress occurred not only when theory development occurred within a ‘paradigm', but when the very concepts which comprised a paradigm, and which theory used, were replaced or extended to create a new paradigm. Crucially, however, although socially or culturally influenced, this process was not relativist. The criterion for identifying progress was that the new paradigm did more in terms of both explaining the world and working with(in) it.
It makes sense to see Kuhn's approach as building upon the approach of Karl Popper 1 and Lakatos and Musgrave 2 rather than as an alternative. Popper rejected simple empirical induction – ‘positivism', against a background of common-sense realism – by proposing the attempt to falsify prevailing theory as the way to do science. Looking for either ‘the facts' or ‘theory' as the origin of the process creates a chicken-and-egg problem. Kuhn added a dimension – the paradigm as a constellation of concepts within which theory operated. But his was part of the same quest.
Kuhn made a throwaway remark in the Postscript to the book to the effect that, while paradigm change was not relativist in natural science, the social sciences were perhaps a different matter. This opened the floodgates in the social sciences to a relativism more nihilistic than that more traditional ethical relativism which is nevertheless based on epistemological realism. This nihilism has gone through the academic social sciences like a dose of salts, eventually reaching the applied social sciences in areas such as the sociology of health and health services as well as education, social work and management in general.
It is ironical that those who argue that even perception is relative are usually the most earnest, humanist and ethical people. But they kick away the stool upon which they wish to sit, and debunk the basis for the critical theory they usually profess.
In essence, the relativism which is at the core of epistemologies such as constructivism and interpretivism is often extended beyond the common-sense position that different people, cultures and classes in different spaces and times have different emotional and even physiological reactions to aspects of the social world round them. This common sense position may include the fact that people's actions and behaviour are affected by their emotional reactions, either unconsciously, sub-consciously, or ‘reflexively'.
But it does not follow from the fact that human behaviour is reflexive and/or meaningful to human actors that how it is perceived by others is epistemologically relativist. It may mean that one's own and others' behaviour and emotions are difficult to read – hence ethnography, which must indeed avoid relativism if it is to be more than a post-structuralist ‘text' which can be written or read any which way (and if that is all it is, why bother doing it?). However, one can still envisage an ideal observer seeing human interaction on this complex basis.
It is just that no human can stand outside the social world enough to do this fully, including observing the consequences of his or her own reflexivity upon him or herself and the rest of the social world. How this plays out – how the social world changes unpredictably – is thus not like (natural) science. But that is because social science is less possible as an explanatory and predictive enterprise than natural science – not because it is equally possible. Interpretivism is what we do when science does not work. Biographies, understanding people’s narratives and the like are worthwhile – but they are not science except in the widest sense of the term as the acquisition of knowledge and/or insight. To that extent the term social science is an oxymoron.
Back to Kuhn. His notion of what perception is (as unpacked in the Postscript) is important here. Unlike cod post-modernism, perception for Kuhn is not what happens when the receipt of sense-data external to the perceiver is interpreted by the perceiver. Perception is what happens when the external sense-data are neurologically processed, keeping in mind that such processing is involuntary and also influenced by past experience built into the DNA, as it were. Interpretation is what happens, if necessary, subsequently. Kuhn gives the example of seeing his mother in the store, only to realize later (through interpretation) that it could not have been his mother after all and how could he ever have thought it was. But seeing his mother was his perception.
Consider how one experiences the same degree of pain or some other external stimulus under different conditions. If one is happy and relaxed, a sudden bang on the head, say from a car door, may not be noticed much. If one is tense and depressed, it may seem almost unbearable. The physiological and neurological sensation is different. Later interpretation can rationalize or at least explain this perception and interpret it. The notion of external stimulus is not thereby invalidated, but perceived in a relativist way – the phenomenon itself is neither relative nor only given reality by the perceiver.
Social researchers should take care not to conflate different types of relativism. There is a view in much of applied sociology that ‘positivism' is an evil thing, and that hermeneutics means taking people's views as ‘their reality' which is as valid as any other. However laudably humanist this is, the epistemological stance it builds on denies humanism – which, according to epistemological relativism, is merely derived from one way of looking at the (social) world.
For post-structuralists, this relativism is linguistically based; for more general, eclectic post-modernists, the relativism is based on the human condition. But it is a strange arrogance to see this as (post) modern – only discovered by modern social science, as it were. I suspect this owes a lot to Foucault, who built a schema of history 3 which was almost an eschatology which concluded with man’s ‘disappearance’ while at the same time denying the philosophical possibility of so doing, or of such a schema being anything other than a narrative among many possibles. Like Foucault himself, (post) modern Foucauldians seek to have it both ways – using Foucault as a gadfly, relativist sceptic, yet also using him as an ethical critic of ‘power over’ and domination. The authors of The Order of Things3 (op. cit.) and of Discipline and Punish 4 struggle within the same skin.
Modern social science has a concern with the plight of the human being and not just with general ‘laws' about populations. Overlapping approaches such as hermeneutics, interpretivism and symbolic interactionism are cri-de-coeur against a view of human behaviour and its meaning, or lack of it, which is behaviourist. So far, so good. But these are means of augmenting what (little) positivism is possible in the social sciences, not denying it.
Like the professionals whom inter alia they study, epistemologists and research methodologists in the social sciences have professionalized themselves, whether consciously or not, out of self-interest as well as from their skill acquisition. John Gray once remarked that in aping the natural sciences the social sciences had merely succeeded in becoming unreadable. 5 One might add that knowledge acquisition itself has become unnecessarily top-heavy methodologically, and arguably illiterate epistemologically, in the social sciences.
In the search for respectability and attention, the wiles of Foucault have been preferred to the queries of Quine. 6 When we have different languages, the challenge is to translate, not wonder at the existence of different solipsistic worlds. Classical logical positivism was not enough. But (post) modern social science sadly misses having its own bullshit detector.
The danger with interpretivism is that it ignores power. If we take at face value people’s views of their own condition, their self-diagnosis may mean that they seek liberation through stories – through seeing themselves differently – rather than through action based on a realist account of their situation.
It is not politically correct to say so, but people may be brainwashed, stupid or uninterested. One needs a theory of power – a lens through which to interpret perceptions. We need Kuhn’s insights without his relativism, and the wrong turning which can be traced in part to his work, at least in giving Anglo-American thought an excuse for aping the Parisian relativist ‘post-' movements.
Applied areas – applications of different disciplines to different areas of health policy and health services research – are not necessarily in thrall to any one epistemological doctrine or tendency, and that is probably a good thing, but neither are they immune to the spirit of the age. Valuable qualitative work often means getting people’s opinions and world-view, but it says nothing about the status of these. For that, we need to stand outside narratives.
Even the word narrative – or its trendy sibling, ‘discourse', beloved of Foucauldians and their fellow-travellers – is a weasel-word. It may refer to a particular perspective or ‘ideology' within an area of enquiry or interest which requires translation between different perspectives before ‘undistorted communication' 7 can allow debate. But it may also refer to the whole debate, as if to imply (falsely) that the whole debate is trapped within language and not capable of rational pursuit. Foucauldians have it both ways, depending upon whether they are wearing their liberationist or pessimistic hat.
Kuhn's influence contributed to a fork in the road. One route led to a more sensitive realism; the other to a dead end relativism, in the name of Foucault and others. It would be good to go back and start again in 1970 to capture Kuhn’s original insight that the existence of paradigm shifts does not demonstrate that scientific knowledge is intrinsically relative. For social scientists, the challenge posed by Kuhn is to distinguish between what is science and what is not – not to relativize all knowledge. The use of the word paradigm in the non-scientific domain has reduced it to a buzz-word, usually meaning little more than a view or a perspective or – at best – a theory. And a theory is capable of testing, so it cannot in itself be a Kuhnian paradigm. If social science teachers must recommend Foucault and the ‘posties', let them first recommend Kuhn, whose work contains both insights and warnings.