How Forests Think. Eduardo Kohn
That which ought to be delimited as something unique becomes instead the bedrock for our assumptions about representation.
We anthropologists tend to view representation as a strictly human affair. And we tend to focus only on symbolic representation—that uniquely human semiotic modality.16 Symbolic representation, manifested most clearly in language, is conventional, “arbitrary,” and embedded in a system of other such symbols, which, in turn, is sustained in social, cultural, and political contexts that have similar systemic and conventional properties. As I mentioned earlier, the representational system associated with Saussure, which is the implicit one that underlies so much of contemporary social theory, concerns itself only with this kind of arbitrary, conventional sign.
There is another reason why we need to provincialize language: we conflate language with representation even when we don’t explicitly draw on language or the symbolic for our theoretical tools. This conflation is most evident in our assumptions about ethnographic context. Just as we know that words only acquire meanings in terms of the greater context of other such words to which they systemically relate, it is an anthropological axiom that social facts can’t be understood except by virtue of their place in a context made up of other such facts. And the same applies for the webs of cultural meanings or for the network of contingent discursive truths as revealed by a Foucauldian genealogy.
Context understood in this way, however, is a property of human conventional symbolic reference, which creates the linguistic cultural and social realities that make us distinctively human. It doesn’t fully apply in domains such as human-animal relations that are not completely circumscribed by the symbolic but are nevertheless semiotic. The kinds of representational modalities shared by all forms of life—modalities that are iconic and indexical—are not context-dependent the way symbolic modalities are. That is, such representational modalities do not function by means of a contingent system of sign relations—a context—the way symbolic modalities do. So in certain semiotic domains context doesn’t apply, and even in those domains such as human ones where it does, such contexts, as we can see by attending to that which lies beyond the human, are, as I will show, permeable. In short, complex wholes are also open wholes—hence this chapter’s title. And open wholes reach beyond the human—hence this anthropology beyond the human.
This conflation of representation with language—the assumption that all representational phenomena have symbolic properties—holds even for those kinds of projects that are explicitly critical of cultural, symbolic, or linguistic approaches. It is apparent in classical materialist critiques of the symbolic and the cultural. It is also apparent in more contemporary phenomenological approaches that turn to the bodily experiences we also share with nonhuman beings as a way to avoid anthropocentric mind talk (see Ingold 2000; Csordas 1999; Stoller 1997). It is also, I should note, apparent in Eduardo Viveiros de Castro’s multinaturalism (discussed in detail in chapter 2). When Viveiros de Castro writes that “a perspective is not a representation because representations are a property of the mind or spirit, whereas the point of view is located in the body” (1998: 478), he is assuming that attention to bodies (and their natures) can allow us to side step the thorny issues raised by representation.
The alignment between humans, culture, the mind, and representation, on the one hand, and nonhumans, nature, bodies, and matter, on the other, remains stable even in posthuman approaches that seek to dissolve the boundaries that have been erected to construe humans as separate from the rest of the world. This is true of Deleuzian approaches, as exemplified, for example, by Jane Bennett (2010), that deny the analytical purchase of representation and telos altogether—since these are seen, at best, as exclusively human mental affairs.
This alignment is also evident in attempts in science and technology studies (STS), especially those associated with Bruno Latour, to equalize the imbalance between unfeeling matter and desiring humans by depriving humans of a bit of their intentionality and symbolic omnipotence at the same time that they confer on things a bit more agency. In his image of “speech impediments,” for example, Latour attempts to find an idiom that might bridge the analytical gap between speaking scientists and their supposedly silent objects of study. “Better to have marbles in one’s mouth, when speaking about scientists,” he writes, “then to slip absent-mindedly from mute things to the indisputable word of the expert” (2004: 67). Because Latour conflates representation and human language his only hope to get humans and nonhumans in the same frame is to literally mix language and things—to speak with marbles in his mouth. But this solution perpetuates Cartesian dualism because the atomic elements remain either human mind or unfeeling matter, despite the fact that these are more thoroughly mixed than Descartes would have ever dreamed, and even if one claims that their mixture precedes their realization. This analytic of mixture creates little homunculi at all levels. The hyphen in Latour’s (1993: 106) “natures-cultures” is the new pineal gland in the little Cartesian heads that this analytic unwittingly engenders at all scales. An anthropology beyond the human seeks to find ways to move beyond this analytic of mixture.
Erasing the divide between the human mind and the rest of the world, or, alternatively, striving for some symmetrical mixing between mind and matter, only encourages this gap to emerge again elsewhere. An important claim I make in this chapter, and an important foundation for the arguments to be developed in this book, is that the most productive way to overcome this dualism is not to do away with representation (and by extension telos, intentionality, “aboutness,” and selfhood), or simply project human kinds of representation elsewhere, but to radically rethink what it is that we take representation to be. To do this, we need first to provincialize language. We need, in Viveiros de Castro’s words, to “decolonize thought,” in order to see that thinking is not necessarily circumscribed by language, the symbolic, or the human.
This involves reconsidering who in this world represents, as well as what it is that counts as representation. It also involves understanding how different kinds of representation work and how these different kinds of representation variously interact with each other. What sort of life does semiosis take beyond the trappings of internal human minds, beyond specifically human propensities, such as the ability to use language, and beyond those specifically human concerns that those propensities engender? An anthropology beyond the human encourages us to explore what signs look like beyond the human.
Is such an exploration possible? Or do the all-too-human contexts in which we live bar us from such an endeavor? Are we forever trapped inside our linguistically and culturally mediated ways of thinking? My answer is no: a more complete understanding of representation, which can account for the ways in which that exceptionally human kind of semiosis grows out of and is constantly in interplay with other kinds of more widely distributed representational modalities, can show us a more productive and analytically robust way out of this persistent dualism.
We humans are not the only ones who do things for the sake of a future by re-presenting it in the present. All living selves do this in some way or another. Representation, purpose, and future are in the world—and not just in that part of the world that we delimit as human mind. This is why it is appropriate to say that there is agency in the living world that extends beyond the human. And yet reducing agency to cause and effect—to “affect”—side steps the fact that it is human and nonhuman ways of “thinking” that confer agency. Reducing agency to some sort of generic propensity shared by humans and nonhumans (which in such approaches includes objects) thanks to the fact that these entities can all equally be represented (or that they can confound these representations), and that they then participate by virtue of this in some sort of very humanlike narrative, trivializes this thinking by failing to distinguish among ways of thinking and by indiscriminately applying distinctively human ways of thinking (based on symbolic representation) to any entity.
The challenge is to defamiliarize the arbitrary sign whose peculiar properties are so natural to us because they seem to pervade everything that is in any way human and anything else about which humans can hope to know. That you can feel tsupu without knowing Quichua makes language appear strange. It reveals that not all the signs with which we traffic are symbols and that those nonsymbolic signs can in important ways break out of bounded symbolic contexts like language. This explains not only why we can come to feel tsupu without speaking Quichua but also why Hilario can communicate with a nonsymbolic being. Indeed, the startled