Despite its title, this chapter is primarily about rethinking the notion of agency, and more than anything else that is what holds this chapter together. In order to understand this different notion of agency, we'll have to understand two philosophical terms: The first, the affective bodies of Spinoza, and the second the notion of assemblage from Deleuze and Guattari.
Understanding the affective body seems, at first gloss, to be a rather easy concept to understand. It is any body that is capable of affecting another body. Of course, the affective body is also a body that is capable of being affected upon. Because the nature of the affective body is as being both affected and able to affect means it is neither subject nor object, but rather enters into relations with other affective bodies that form a mode (there is no way in this understanding to conceive of an affective body in a vacuum). Now, it is important to remember the discussion of conatus from chapter one (see pp. 2-3, plus the interesting footnotes). Every affective body, which also means every level of these plural affective bodies, has its own conatus. And what these bodies all strive for (in their own different ways) is the enactment of their own power (Spinoza calls this joy, but this word is missing from Bennett's vocabulary). This means that affective bodies enter into alliances and relationships with other affective bodies in order to increase their power. In order to think these complex relations, we need to turn to Deleuze and Guattari's notion of the assemblage (agencement). Bennett's short, four paragraph long section on "What is an Assemblage" should almost be quoted in toto (and What Is An Assemblage would be a great title for a book). Bennett explains: "Assemblages are ad hoc groupings of diverse elements, of vibrant materials of all sorts." (p. 23) She then goes on to continue to explain their external and internal structure in several clear, concise, and dense sentences. For our purposes, assemblages are composed of various affective bodies that come and work together. They form together and break apart. They exist therefore for only particular times and places, that do not necessarily entail the same finitude as their component parts. Their structure resembles a dialogic structure from Bakhtin, except for the fact that assemblages do not exist exclusively or even necessarily of human components (which might also be true of certain readings of Bakhtin).
Bennett goes on to give a political and policy-oriented example how assemblages function with a great blackout in August 2003. The section does lend itself to summary, so I suggest just reading the narrative itself.
After the explicating the blackout, Bennett goes on to distinguish the assemblage from a repressive structure on a willing or self-intentional subject. To explain, it is not uncommon for people to accept that a material non-human object can change the actions of a human. But what is usually understood by this point is that these material structures or systems limit human agency, it is seldom understood as these structures having agency themselves. To give a personal examples, I have one of those desks where you have to open the center drawer in order to open the three draws on the side of the desk. Every time I go to take things out or put things into these draws the desk and I engage in this little dance. Now, one way to understand this is that the desk is a mute object merely reducing my agency of opening up the drawers in any damn order I please. Another way to understand it though it is to see how everything changes. What books and papers get put where (in drawers or on the top of the desk or stacked next to me on the floor) have shifted due to this interaction. This changes the way I interact with materials, find them, draw connections in my work. All of this means that a particular writing and research assemblage including books, shelves, drawers, me, papers, surfaces, all get changed. The desk actively produces a different writing assemblage of which I am only one component of.
Bennett then takes up the three traditional criteria to determine agency: efficacy, trajectory, causality. Bennett doesn't set out to argue for different terms to understand agency, but rather seeks for us to understand these terms outside of intentionality. Efficacy comes to be seen as a swarm of different actants, trajectory comes to take on a derridian notion of messianicity, and causality comes to mean something emergent and fractal.
Agency of assemblages can be perhaps understood by the Chinese notion of shi. Shi comes originally from military strategy (which doesn't surprise me at all. If you want to hear people talk in terms of swarms and emergence and fractals talk to some military think tank types), and it is "the style, energy, propensity, trajectory, or elan inherent to a specific arrangement of things". (p. 35) Shi, therefore, is not about any specific element, but rather the composition of forces, allied elements being able to move together.
In the final section of the chapter Bennett confronts head on that this notion of distributed agency, of the agency of assemblage, exists in tension with a desire to hold people accountable. Where energy speculators guilty for the blackout? Are BP and the MMS guilty for the current oil spill? Bennett doesn't give a definitive answers, but at least raises directly the question. If we take seriously that agency lies in assemblages rather than intentional human subjects, holding people accountable makes very little sense. And moreover, such morality plays can help hide to us proper policy solutions to problems. On the other hand, she admits that outrage and shock are frequently the proper and just response to situations. So, this chapter ends with a sort of stutter, a sort of stammer. I don't think this stammer will satisfy, but I am at least sympathetic to it.
To see my commentary on this chapter, click here.