Manifest Ontology of Agency

    "My strategy is to focus attention on the distributive and composite nature of agency," states Joan Bennett in her account of a electricity blackout that effected North America in 2003, "Are there not human, biological, vegetal, pharmaceutical, and viral agents?"  The answer is mu (無).  To respond to this question otherwise would be to reproduce what might be termed a compatibilist realism, whereby something akin to a will (although for Bennett, perhaps not as strong a nominee as "free will") is deemed to inhere in an otherwise historically determined assemblage, or else be forced (by way of the inclusion of the "human" in her query) to abandon ourselves to a nihilist solipsism, in which we can not even admit to our own ineffable experiences as agents.  Although Bennett would rightly move the loci of agency from subjects, relocating it to assemblage, nonetheless she retains an idea of agency as inhering in such loci.  Agents, here, are still things-in-the-world, even if those things aren't specifically human bodies.  It is telling that she adopts Latour's term for such nonhuman agents, actants, a term taken up from narratology.  Narrative actants are not found in the world, they are found in the stories we tell about a world.  If it has become my tired habit to parrot Korzybski, this would be the time to do it:  "The map is not the territory."

    Bennett's constellation of efficacy, directionality, and causality, can help explode this notion of inherent agency.  For efficacy, Bennett would "locate intentions within an assemblage".  Here, theory of mind rears its head:  for how exactly do we "locate intentions", whether within a human, a nonhuman, or an assemblage constituted of any number of both, except through a working theory of that presumes such intentions?  Are intentions more observable in an assemblage than in a human?  If yes, then we have empirical access to an assemblage that is deemed impossible with a human:  which would cast a human as ontologically not an assemblage!  If no, then we "locate" such intentionality only insofar as we tell ourselves a story about it.  As for directionality, Bennett's elaboration on Derrida's account clearly undermines the essentialism of inherent agency:  "things appear to us only because they tantalize and hold us in suspense" and, quoting Derrida now, "someone or something that, in order to happen . . . must exceed and surprise every determinate expectation".  Directionality here "appear[s] to us", "hold[s] us in suspense", "surprise[s] every determinate expectation":  directionality as experiential, phenomena as an existential encounter.  Finally, in her movement from "efficient causality" (ringing back to efficacy) to "emergent causality" to Arendt's sources (falling back to "intentionality"), agentic processes are revealed as distinguishable from deterministic processes only insofar as the agentic effect "can never be deduced" from its originating source.  To draw upon Hacking's historical ontology, the agency of assemblage comes down to our inability to deductively "make up" their cause.  Would Laplace's demon (with omniscient "degrees of possibility") find our world so populated by agents?

    It is Hacking's dynamic nominalism, whereby he differentiates between a realist account of horses and planets and a nominalist account of gloves and multiple personality, that provides a way to approach Bennett's agency of assemblage.  To do this, however, we must explode the dichotomy that Hacking likewise takes as a given:  "a contrast between people and things".  For, Hacking tells us, in contrast to horses and camels, and other such things (when not interfered with by us), "some of the things that we ourselves do are intimately connected to our descriptions".  But this is our critique of Bennett's constellation!  Where we can "locate intentions", when "things appear to us" that "surprise every determinate expectation", that is, their effect "can never be deduced" from an originating source, then we name such things agents, or actors, or actants.  Bennett's shi then is a property not of an assemblage, but of our encounter with and description of said assemblage.  However, we can be more exacting here, for an encounter is not a description, and indeed, it can be argued that the latter is predicated on the former.  For both Bennett and Hacking implicitly admit into their arguments a raw ontological distinction:  as much as the world may contain real things in which we do not "locate intentions" and which do not "surprise every determinate expectation", so too the world is peopled by actants, ourselves among them, "intimately connected to our descriptions" of them.  It is on this basis that I would augment Hacking's dynamism with a manifest nominalism, whereby, in order for our descriptions of agentic people "to happen", there must be the surprise encounter, holding us in suspense.  Such a nominalism is manifest, as opposed to say, evident, in that such agency, whether of people or assemblages, is perceived by the surprised senses, felt by the suspense-bound body, before any conclusion is drawn, any description formulated, any intention located.  Agency, by this argument, is not a real thing-in-the-world, not an empirically accessible essential property of actants, but rather the perceptual predicate for our naming of agents as such.