Is Grid-Group cultural theory really a theory?

It was a trick of course. Yesterday I used Grid-Group cultural theory to ‘predict’ the Fatalist viewpoint of Nicholas Taleb, author of The Black Swan. But like the magician who successfully predicted the lottery numbers, it’s more about sleight of hand than about actual magic…

Despite the name, cultural theory isn’t really a theory at all. It’s a conceptual scheme – an heuristic we use, it might be argued, because we are cognitive misers and like making short cuts in our thinking. The world is big and hard to understand, so we make biased assumptions about what ‘usually’ happens, what ‘must’ happen, or what ‘really’ happens, what ‘the facts of the matter’ are, and so on. (to give just one example from millions, some more, some less trivial: Nick Cave’s deep-voiced assertion, ‘People they ain’t no good’). Moreover, we don’t just get these ideas out of our own heads somehow. They are embodied in the institutions in which we participate, from the family mealtime to the Copenhagen Climate summit, so that they are accounts of actual reality as we experience it. They really do explain how the world is – at least parts of it – so that our ideas seem like common sense. It is argued that these cultural biases or cultural solidarities coalesce into four basic ‘ideal types’, the four cultures for Grid-Group Cultural theory. But can we actually measure this? And can we really use the result to predict anything?

The question goes right back to Durkheim and the birth of modern sociology. In his groundbreaking study of suicide, Durkheim identified four different socially-conditioned types of suicide. It is one thing to label these heuristically but quite another to demonstrate their empirical existence. (On a coroner’s report “Cause of death: anomic suicide” would look a lot like a subjective value judgement rather than a matter of ‘fact’, and the significance of context is precisely the issue here – sociologists can say what they like, but not if they want to be recognized as scientific). Long before Mary Douglas originated the grid-group schema, George C. Homans criticised Talcott Parsons for doing ‘theoretical work’ that isn’t actually theory. Whereas Parsons claimed his classifications were ‘exhaustive of the relevant logical possibilities’ (Parsons 1951:66), Homans objected as follows:

‘But anyone can always set up a logically exhaustive conceptual scheme in this sense. What he does is to define a class X and then say that everything else falls into the class Non-X, giving the classes a name, and then he has a logically exhaustive conceptual scheme. He can readily compicate the scheme, making it a fourfold or an n-fold one, by intersecting the first two classes with new ones, like Y and non-Y and calling the result a paradigm if he likes. But none of this work makes the conceptual scheme a theory, for it imples no contingent relationships between properties of nature. Only if, for instance, actual instances of X turned out also to be instances of Y, or instances of Non-X, instances of Non-Y would such a relationship be implied. You can make a conceptual scheme as exhaustive as you like and in any way you like, but it still remains only a conceptual scheme.’

George C. Homans 1984 Certainties and Doubts: Collected Papers. Chapter 2, Contemporary theory in sociology. New Brunswick, NJ & Oxford: Transaction Books. p. 27.

Cultural theory doesn’t actually amount to a theory in the sense Popper meant that term or in the sense which Homans used contra Parsons and also Robert Merton. Their idea of a theory as ‘logically independent generalized concepts’ (Parsons) or ‘logically interconnected conceptions’ (Merton) isn’t falsifiable and it isn’t truly predictive (certainly not in the sense that, say, gravity, relativity and quantum mechanics have been).For Homans, theory had to mean ‘a deductive system’, the same in sociology as in the natural sciences. Grid-Group cultural theory is not deductive. So what is it that makes it seem to work? Is it that having found a rather neat scheme, everything starts to fit it – that ‘to a person equipped with a hammer everything looks like a nail’?

This website could certainly be accused of that: to a blogger equipped with a fourfold conceptual scheme, everything comes in fours.

Or, more helpfully, is it that Cultural theory should stick with the modest claim to be a neat but ‘unscientific’ typology – a useful but arbitrary scheme for the work of categorising, but no more than that? There would of course be no shame in this. The academic world is stuffed full of more or less arbitrary fourfold schemes. People are allowed to get away with it all the time and often these schemes do actually help to organise otherwise disparate data. As Homans said, ‘None of this means that a conceptual scheme is useless. A scheme with clear criteria may allow investigators to identify instances of significant properties of nature, and the elaboration of such a scheme may be a useful preliminary to the construction of a theory.’ (p. 26), but as it stood, much of sociological theory he claimed provided ‘the dictionary of a language that posesses no sentences’ (ibid).

These were the kind of debates taking place in the 1950s and 60s. The social sciences generally have a long-running conflict about quantitative research versus qualitiative research. There is an interesting very recent comment on this division in the field of cultural anthropology at the decasia blog.

An illuminating but controversial contemporary attempt to overcome this divide is Bjent Flyvbjorg’s study Making Social Science Matter. (2001). He argues for a recasting of the social sciences not as an imitation of the natural sciences, nor as an art, but as a form of phronesis – ‘practical wisdom’ (an alternative term would be ‘prudence’, as in ‘jurisprudence’).

I think Grid-group cultural theory can in practice swing both ways – it can be depicted as empirical or as interpretive – and this is a dangerous, though perhaps inevitable,  thing because it can leave an unhelpful measure of ambivalence about what kind of intellectual lifting work it is actually supposed to be doing. However, this is hardly unusual for theories and quasi-theories in the social sciences. The reality of the model and the model of reality can easily become confused (Bourdieu, cited by Jenkins 2001: 56). Richard Jenkins quotes a great example of this in the case of methodological individualism, of which he says

‘how quickly a mode of explanation becomes a statement about the nature of the phenomena being explained, and methodology becomes ontology (2002: 158) .

I would add it is seen at its absolute crudest in the case of populist economics writers who, enamoured of the idea of ‘perfect markets’, procede exactly as though such an entity really existed (Harford 2006).

Numerous writers, Michael Thompson among them, but I don’t mean to single him out particularly, resist the quantification of grid-group as a scientific theory in the Popperian sense and more or less assert that it is a theory and has been ‘proven’. This approach derives, I think, from an anthropological resistance to quantification in favour of interpretation, or hermeneutics, and seems to be based on the idea that a scheme that can be shown to be logically consistent can indeed be termed a theory. In other words, because the grid-group categorisation covers all possible social configurations (every relationship must sit somewhere on the chart, must be high or low grid, high or low group) therefore all possible social configurations must fit within it. This is a circular argument, it seems to me. At the very least it indicates a different use of the word theory from that used in much of the natural sciences.

On the other hand, a number of researchers take grid-group theory as inspiting a collection of testable hypotheses, and try to use it in organising testable research agendas.

One such attempt is the Cultural Cognition project based at Yale Law School. This has taken US National Values survey data and sought to find correlations with the axes of grid-group theory, modified for various ways by the team. This team has found quite good correlations and certainly holds out the promise that grid-group cultural theory, or something very much like it could be demonstrated in an empirical sense. Dan Kahan says:

‘Cultural cognition refers to the tendency of people to conform their view of facts to their identity-defining group commitments.’

This, to me, seems eminently testable, provided we can pin down what ‘identity defining group commitments’ are.

One of the matters for such work to resolve is where exactly the object of study is believed to reside. This is an issue common to social studies in general: what exactly is the social. Where is it? How do we define something we can’t see, touch, smell, hear or taste? And once we’ve defined it, does this then mean we actually can now sense it? Since the dominant paradigm is Individualist, it’s not surprising that Individualist assumptions are present in some of this work. For instance: The idea that the cultural biases of cultural theory reside in the words or even brains of individual respondents. This is a tendency this website is also prone to – depicting Grid-group cultural theory as though it was entirely encompassed within the views of one person. Compared with previous generations of social science, we are much more at home within methodological individualism, and I think the cultural cognition project is doing a good job of ‘translating’ grid-group cultural theory into that language.

Can we really use a theory like this to understand what individuals are up to? I think this has yet to be proved decisively, but I’d claim my ‘trick’ made a bit more sense of Nicholas Taleb’s political position than those puzzled British journalists seemed to be able to. It was somewhat trivial, but the work of the Cultural Cognition project could have rather large implications.