Recently I was thinking about the concept of culture again, and I think I had an insight into why it's been so troubling for anthropologists. I think there is a confusion between two very different ways of thinking about culture - the holistic and the particular. We want to be a holistic discipline, and some definitions of culture reinforce that - think of Morgan's definition as "that complex whole which includes...." or Leslie White's definition as "man's extrasomatic means of adaptation." These are all fine, but then there are other definitons which indicated that culture is just one part of a system among many other parts - politics, economics, society, etc. This is the case whenever we talk about cultural factors influencing outcomes, as these cultural factors conflict with various policies. It is also the case with cognitive theories of culture, and many other approaches to measuring culture.
The problem, I think, is that these two approaches to culture get used interchangeably and are often confused with one another. When you say that something is cultural, to I take you to mean that it is reflective of a whole system or do I take you to mean that it is one factor among many in a larger complex system? There's often no way to tell unless you spell it out explicitly. But it cannot be both. If culture is a complex whole, then it subsumes economics and politics. If it's a part, then politics and economics are external to it. (I suppose Levi Bryant's strange mereology could explain this, but I'm not sure I buy into that yet).
I suspect that these two concepts may map onto colonial discourse. When we talk about culture as a whole, what comes to mind are "primitive" groups whose lives are depicted as being subsumed by traditional (and irrational) beliefs and values. When we talk about ourselves, on the other hand, we tend to think of our political and economic institutions as having risen out of the culturl mire to a more rational form. Thus, culture in Western societies is one part among many.
Basically, I think we have to pick one or the other, stick with it no matter what group we're talking about, and be explicit about which definition we're using. I'm starting to think that the particular approach is the way to go. It's more practical, since it's very difficult, if not impossible, to know a whole system, and, as I've argued recently, there's not much you can do with a whole system even if you do know it. Instead, I'm thinking of culture as that which politics, society, and economics leaves out: a kind of foundational set of influences which interact with economics, politics, and society in complex ways - the reason none of them or even all of them taken together can ever be compete. Culture is why things don't go as expected when people enact legislation or change the economy - it is the confounding remainder.