Because we can only know in relation to something else, this discussion of the Politics of Ontology gets to the heart of the anthropological project. Ontology provides a relational view of method. Every ethnographic description is equally a description of the anthropology producing it.

Anthropology's engagements with the political have been turned inside out over recent years. Any distinction or definition between textual representation and political representation has been collapsed. Speaking about can now be heard as speaking for. As much as what an ethnographic text or description might say, even the act of ethnographic description itself can make a political statement. But this roundtable is important for it provides an opportunity to separate out again these twinned politics of representation. And it also provides a space therefore, in which to leave aside the question of whether a discussion of anthropological method should be political or non-political.

My book, Exchanging Skin (2007), derives from research in Bolivip village in Papua New Guinea. The book takes up the Min Problem—a long-standing analytical impasse—and argues that the problem all along was one of Anthropology's own making. Intriguingly, the very peoples and places that, through Fredrik Barth's work on the Baktaman, came to stand for and exemplify secrecy and knowledge, have provided the discipline with one of its most critically demanding tests. Although analyses based on Euro-American conceptualizations of secrecy and knowledge were produced, they did not stack up with the ethnography in Bolivip.

In Bolivip, “knowledge” implicates people in a double life by affording and bringing together divergent gendered perspectives: not so much revealing to a viewer their position in the composition of a field of knowledge, as newly revealing the composition of the knower and the subtleties of their personal capacities and relational supports. This is not so much being in the world (a figuring out of positions) as world in the being (a figuring of internal capacities). Revelations have the dual life effect of revealing that there is always more to things than one knows—and so it creates a relation that carefully positions a person in those new possibilities.

Knowledge practices in Bolivip employ the imagery of relative positions on a tree: the muddled confusion of junior cultists is likened to the multiplicity of branches and leaves, whilst very senior cultists display their solidifying grasp of things in the way that the ever-branching stories of juniors seem always to come down to the same thing. There is a double-ontology in Bolivip: for juniors in the crown, words from seniors at the base appear to branch into multiple possibilities.

Clearly, ontology is no one thing. As we've already heard, “ontology” can serve to describe an all-encompassing world view, and to describe an all-encompassing anthropological method. That “ontology” foregrounds and highlights this isomorphism between ethnographic object and anthropological method is its most important virtue.

Of course, anthropologists are adept at discerning the wider cultural histories and metaphysical concerns at work in world views, and thus it is possible to discern contemporary Euro-American conceptual collapses of nature and culture such that things seem to have “micro-ontologies” (so every thing has a world, and a worldview, of its own), and to discern emergent Christian and process theologies which refashion the position and the mathematics of the Godhead (so that God and his believers are part of, and can pass through, each other).

Ontology provides a relational view of method, and reminds us that a critical test for ethnographic knowledge-practices is the faithfulness with which they acknowledge that they are both enabled and constrained by the knowledge practices of our ethnographic subjects. For too long, the pretense of scrupulously separating data from theory had anthropology barking up the wrong tree, and afforded a privileged analytical position as if narrating from outside the ethnographic relation. I take it that looking for theory in the same place we look for data provides a crucial disciplinary and decolonizing turn.

Every ethnographic description is equally a description of the anthropology producing it, then. Ontology is useful because it foregrounds our part in the relational and conceptual scheme, and reminds us of three important lessons:

(1) Roy Wagner's (1981) enduring insight about our invention of culture—that is, the efficacy and contingency of using our concepts (such as “culture” or “ontology”) to apprehend, apportion to and describe the concerns of our ethnographic subjects.

(2) Marilyn Strathern's (2011, 92) insights into exchanges between knowledge practices—that is, “to be perspectivalist acts out Euro-American pluralism, ontologically grounded in one world and many viewpoints; perspectivism implies an ontology of many worlds and one capacity to take a viewpoint.”

(3) As I understand Viveiros de Castro's (2004, 3) “comparison between anthropologies,” it is neither multiple natures nor singular cultures that require analysis, but a description of the metaphysics, potentials and affordances that find manifestation and expression in different forms.

Any methodological insistence on these three lessons carries political force for the reproduction and transformation of the disciple. It may even save us from being dazzled and taken in by the effects of our own creativity, and allow the creativity of ethnographic subjects to further expand our understandings of being human.