Amira Henare, Martin Holbraad, and Sari Wastell propose

a radical constructivism not dissimilar to that envisaged by Deleuze. . . . Discourse can have effects not because it ‘over-determines reality,’ but because no ontological distinction between ‘discourse’ and ‘reality’ pertains in the first place. In other words, concepts can bring about things because concepts and things just are one and the same.¹

As David Graeber put it: “Apparently there’s virtually nothing, no matter how obviously crazy, a contemporary academic can’t get away with if they find some way to attribute it to Gilles Deleuze.”²

James Elkins, writing about the body in Mesoamerican art¹ articulates a particularly sensible perspective on material objects, the contingency of analytical categories, and theoretical predilections for reflexive self-awareness:

Perhaps the most common formula [for dealing with the effect of analysts’ biases on their ability to understand the past] could be stated as such:  What we see in the past is certainly influenced by our own thinking, but we will be steered toward true accounts by remaining faithful to the historical evidence.  This stance, or something like it, has been used as a reason to eschew the philosophy of history and “theory” more generally.  If the historical facts do not support whatever biases or theories we have — so this argument runs — then we will automatically tend toward more accurate accounts.  There are many reasons to doubt this notion, and they all proceed from one general claim that was first articulated by Hegel:  theories, he thought, are what guide our apprehension of “facts,” so that there is no such thing as a fact free of theory.  (113)

… the most interesting scholarship is the product of informed study of the artifacts together with reflective inquiry on our own cultural situation.  Without the archaeological detail, Mesoamerica becomes speculation; but without cultural self-awareness, it becomes the unnoticed projection of our own ideas.  In this form, the problem is applicable to any historical research that finds itself engaging with alien material and at the same time spurning historiographic inquiry [by which he means, as he later clarifies, an attempt to understand the Western etymologies of our critical terms/analytical categories].  This is a condition not only of historical knowledge but of knowledge in general. It is more corrosive than this description suggests, since “artifacts” and “archaeological detail” are also names for unnoticed projection, so that ultimately there is no contrast between a firm base of facts and the endlessly cycling groundless self-awareness.  Any object, qua object, can set this groundless specularity in motion, and no object, qua object, can settle the difference between seeing mind and seen object.  But I am not concerned with this kind of speculation on speculation, because it quickly makes history impossible; it is the particular balances and decisions of historical writing that are of interest. In order to remain within history, and to continue to speak about history, we need to agree on questions that cannot be asked; and it seems to me that this is one such moment. It can be helpful to draw attention back to the self-awareness of the historian, but only if the material object can remain as such: perhaps inaccessible, always difficult, but potentially engaging.  (113)

What is needed is a detailed, inch-by-inch look at the images, paired with an historiographic inquiry that tries to find and understand the Western etymologies of our critical terms.  Neither strategy alone is sufficient.  A theoretical inquiry into the origins and meaning of concepts such as fragmentation, wholeness, or pain cannot make contact with the works, and a myopic but untheoretical analysis will only rehearse unnoticed Western categories of thought.   (124)

Issue 7 of Anthropology of this Century¹ (May 2013) features an interesting interview with Maurice Bloch, in which he distinguishes sharply

…between what I would call anthropology and what I would call ethnography.  These are two quite different enterprises.

Anthropology is about developing theories concerning the human species…  Ethnography is a different business. It’s about getting to know certain people in certain places, getting to know what makes them tick and their own way of thinking. So these are really very, very different enterprises.
Most people doing a PhD in anthropology do ethnography and then they’re asked to stick in some theory. But the relationship is always very uncomfortable.  It’s in the tradition of anthropology to try to combine these two different enterprises. I think there has been a point in trying to combine them, but it’s always difficult because one is talking in general and scientific terms when doing anthropology and one is doing interpretation when doing ethnography, i.e. trying to situate people in the contexts in which they live, trying to get at what makes them act in the way they do.

These are completely different things.

Something of the sort could be said about archaeology as well:  the fit between theory and the analysis and interpretation of particular material data often seems very uncomfortable.  The reason, I think, is most often that the theories being mobilized have little to do with the data.  It is no doubt true that theories constrain our perception of data, but that does not mean that we cannot select theories in terms of their likely relevance to actual data, rather than struggle to link data to avant-garde theories.  This would make data somehow primary, which would be very passé.

“Theoretical discourse,” so popular in anthropology and archaeology and in the academy generally, is probably very absorbing, and it is certainly well rewarded.  Inverting Bloch’s characterization of Anthropology PhDs, graduate students in archaeology, like many of their mentors, more and more do theory and stick in some data.  Happily, a few archaeologists cling to the view that theory is interesting and useful only insofar as it elucidates some particular part of the past.