It Must Have Something to do with 'Logic': A Rejoinder to Aaron Wildavsky

Believe it or not, I have taken more time on my article concerning cultural theory than on any other article I have ever written. Why? Because I was (and still am) searching for something new that was theoretically stimulating, i.e. able to dig deeply into what politics is all about, and at the same time works empirically. I do not feel comfortable within either the dominant 'left-right tradition' or alternatively within 'modern' theories of post-materialism. My article on cultural theory should be seen as part of a process in which I am trying to figure out whether or not cultural theory will become a basis for my own political thinking. From the mid-1980s on, the attempt to grasp what cultural theory is all about has more than anything else improved my own understanding of politics.

Even so, I see fundamental theoretical problems that, in my opinion, have to be 'solved' before cultural theory is really able to compete with the dominant approaches within our field. Some of these problems appear quite clearly from Aaron Wildavsky's response to my article, and I shall concentrate on two topics that seem to me to constitute the main intellectual challenges: (1) whether it is quite clear what the dimensions 'grid' and 'group' are and-related to this-the question of 'logical' vs. social explanations, (2) the unit of analysis and problem of aggregation.

Aaron Wildavsky does not comment upon my underlining that the dimensionsare Not only do I stress that it is empirically difficult to separate 'group' from 'grid', but that, theoretically at least, cultural theory tells us that these are independent dimensions relating to different aspects of social or cultural life. Yet as I argue in my article (p. 107), if a group with very strong boundaries (say a caste in India) prescribes all kinds of behaviour for its members, the grid dimension is by definition empty. This

is not an understanding of 'grid' and 'group' based on 'old' versions of the theory. On page 5 in Cultural Theory Wildavsky and his associates define 'group' as the extent to which an individual is incorporated into bounded units. The greater the incorporation, the more individual choice is subject to group determination. 'Grid', they say, denotes the degree to which the individual's life is circumscribed by externally imposed prescriptions. If you define 'grid' as being externally imposed prescriptions, then the grid dimension under certain circumstances becomes empty. On the other hand, if you include internally imposed prescriptions, it becomes very difficult to separate 'group' from 'grid', i.e. the dimensions are not independent and measure partly the same thing. Where one is placed on grid becomes a function of group. This would not do.

Not only do the dimensions remain unclear, but we are not really told very much about them. In Cultural Theory we find surprisingly little about the dimensions themselves, i.e. about the theoretical basis of the theory. For instance, in the very interesting discussion on the social construction of nature (ch. 1) which emphasizes the five myths of nature that suit the different cultures, these myths are not deduced from the dimensions, they only go together with the cultures. The truth is that most of what is written within the tradition starts out from the four (or five) cultures and not from the dimensions as such. Instead of trying to specify how to move from the dimensions towards the different combinations of cultures - the most important and the most difficult question - most of the literature is mainly about the cultures, trying to relate preferences (and biases) to the different cultures, but decoupled from the dimensions as such. This won't do if cultural theory is to remain the kind of theory we are told it is.

Could the above-mentioned difficulties be one reason why Aaron Wildavsky arguing that cultural theory has nothing to do with logic? But it must have, because if not, we are not dealing with coherent cultures that are deduced from theoretically based dimensions, i.e. a typological approach grounded in theoretically based dimensions as opposed to categories.

Even if for the moment we take it for granted that the dimensions are independent and clearly defined, it would still be disturbing if we find combinations of 'grid' and 'group' that theoretically should 'produce' a certain culture, but what we find rather is a preference structure similar to that of another culture. In Cultural Theory Wildavsky and his associates tell us that most disturbing for the theory would be if one could demonstrate that values are little constrained by institutional relationships. I agree, but this is a statement on a very general level and not all that controversial.

More important here, however, because it is directly related to our
discussion about the dimensions, is their statement that 'if the same cultural
biases thrived in dissimilar social contexts or, conversely, if dissimilar biases

existed in similar social contexts, then our faith in cultural theory would be greatly weakened' (p. 273). My example from Bremnes (pp. 112-113) would be one such example. What should have been a hierarchical culture looks more like an egalitarian one, and, frankly speaking, I repeat that I cannot see why strong group-strong grid 'has to' produce a hierarchy. Another example can be taken from South Africa. These days I am reading Allister Spark's monumental book The Mind of South Africa. As Spark describes it, in the old and traditional South Africa (before the arrival of the white man) both 'group' and 'grid' were high. The culture itself, however, is described as being fatalistic. Could it be that the cultures are not really deduced from the dimensions at all and that difficulties here at least to some extent explain why so little is said about the dimensions?

One of the more unclear or at least diffuse parts of cultural theory has always been the question of coherence and that of the unit of analysis. On the most general level, Wildavsky tells us that the unit of analysis is the socialized individual i.e. 'the individual in context'. At this general level I fully agree. In the early phase of cultural theory, however, the coherent individual was emphasized, not just coherent cultures. Yet since most of us have to relate to different environments, in later versions of the theory, including Aaron Wildavsky's response to my article, one takes as a startingpoint more realistic assumption that none of us is located in just one culture at a time, but possibly several. My view is that this move has made cultural theory theoretically more unclear while at the same time it has failed to 'solve' the great measurement problems involved.

In Aaron Wildavsky's response to my article the once very interesting and new way of putting together voluntarism and determinism has disappeared. What is left is mainly determinism or pure contextualism, because the individual, as I read Wildavsky's response, is purely a product of the different contexts of which he takes part. I am no longer able to see the contours of any individual personality, no real individuality. Furthermore, we are told nothing about different types of socialization, i.e. qualitative statements emphasizing that certain types of contexts must have a greater influence upon the individual than others. What is lacking is any real discussion of what is most important to me as an individual, i.e. what is my main culture - how I want to relate to other people and how I want other people to relate to me, which originally was what cultural theory was mainly about.

Wildavsky's statement that you change cultures if you are surprised enough, moreover, cannot work within such a contextual approach, because if you do not have any cultural basis - i.e. a deep-structure that is yours and is able to define you as an individual - how can you then figure out that something is wrong? If 'anybody' is to find out that something is wrong it

must be the culture itself and not the individual, which is something quite different and theoretically very problematic. Additionally, not only is this theoretically unconvincing, but in my opinion it makes the measurement problems unsolvable when trying to measure the culture of large organizations,not mention entire nations. What it means is that the problem of aggregation becomes unsolvable, making only illustrative use of empirical data and no real empirical analysis possible.

Cultural theory now has to concentrate on the challenges concerning the dimensions and the unit of analysis. If not, it is very unlikely that the perspective will ever really be able to compete with the dominant traditions within our field. However, even as the theory now stands there is more to learn about politics than most other places.