Levi has a recent post on this subject discussing these 3 models which he terms poststructuralist, contemporary, and Deleuzian. The challenge here is figuring out how to create space for a subject with agency who can undertake political change. Thus he ends this way:
When we talk about resistance we want something approximating decision, choice, reflexivity, self-reflexivity, or, in short, agency. Yet when we adopt an ontological perspective on these issues of political emancipation and resistance, we seem to embrace a perspective where things happen of their own accord. Tornadoes don’t choose to come into existence, they just do come into existence when the requisite gradients of barometric presence are present. This is exactly the sort of thing we don’t want to argue when we discuss subjugation because we acutely sense that resistance might not occur in these circumstances. Decision seems to be required. This is the problem with the idea of a political physics. It somehow misses the dimension of subjective engagement and decision. I don’t know how to get this dimension into the framework I’m trying to think through, but then again I’ve seen no other position that’s able to– though they all assert it –either.
The poststructuralist position with its ideological overdetermination of the subject doesn’t offer that. The contemporary view, where the subject is always non-identical to itself and to those determinations. Though as Levi points out ” it’s not at all clear– and maybe I’m just dense –how a void or emptiness can be a seat of agency.” The Deleuzian model is a model of gradients and thermodynamics, but as noted above, it’s hard to see where decision arises here.
For me subject and agency are separate issues. I tend toward viewing the subject as being roughly analogous to the tornado. It’s an emergent phenomenon. We don’t control our experience of subjectivity. I don’t get to decide what I sense or feel. There’s a feedback loop between working memory, what you’re consciously thinking, and what comes next into the mind. But do I get to decide what I think? Not exactly. I can shape it. I can try to focus on one thing rather than another. I can try zazen meditation and let thoughts slip by. The conscious mind, the site of the subject, is part of the mechanism that shapes subjective experience. As for the rest of the mechanism, some of it is part of the body and brain, and part is not. Symbolic behavior is the best example of this. What is human subjectivity without language?
Agency, at least in this conversation, is about acting out of subjective states. In the most banal example of making a selection from a menu: if I have agency then it is the capacity to choose an item from the list, but I certainly don’t have the agency to control my subjective preference for one item over another, whether I feel hungry or not, etc. Of course without the restaurant, the menu, and the wait staff, I can’t order lunch. As Levi points out, the idea that agency is some magical ability should be viewed with skepticism. I can’t magically order lunch. We all experience what we call decision-making. As subjects it certainly feels like we are making a selection from the menu. However since we can only make one order, there’s no evidence that we could have chosen something else. It’s the same thing with politics or other social values. We all have them. Did we choose them? Do you recall sitting down and deciding whether you are a liberal or conservative? Do you have the freedom to switch views?
What is it that we want agency to be able to do? It seems to me that the desire for agency is the desire to have the potential to act other than the way we do. Assuming that we largely agree with our own actions, even though the results are often imperfect, I would have to guess that our belief in agency is our hope that others have the potential to act differently so that there is at least some chance that in the future we might persuade them to do so. But here’s the thing. We already do that. It’s called rhetoric, right? I choose from the menu, but the menu also persuades me. So let’s say that agency is a force. It is directed by thinking, which is an activity over which we have limited input, an activity that is shaped by our relation with objects both within and without the body. Our agency includes our capacity to affect others’ thinking and decision-making. But agency isn’t something that we have as a ontological quality.
One reply on “Levi Bryant's 3 models of the subject”
“But agency isn’t something that we have as a ontological quality.”
Ontologically, it might be represented as a difference in the type of relation between controller and agent. Neither object is different per se, but their relation is. An object with no agency is controlled and driven, an object with agency might take requests.
In programming it’s common to recognize different types of relationships formally, and that relations do exist in both modelling and in practice. Changing the relation would be an ontological change if you accept that relations _exist_.
Even if you don’t accept the existence of relations, major structural differences can occur to account for the differences in nature of relations.
In programming, there is a type of system architecture we call a transactional system. Say a request for a report is packaged into a transaction, and passed around internally to all kinds of different objects. Some objects like a logger, might record seeing the report request, other objects might answer the report request and fill in the needed data, still another reporting object might actually send a filled in report to a printer. The original asker, has no control over what objects do to the request, it cannot ask to be excluded by the logger for example. The logger decides that. In fact, a year after, you could add another object that would send a message to a pager every time the report was requested, and the original asker would never know that pager service existed. A transactional system has objects with agency.
Ontologically, this is represented very differently than another system which wasn’t transactional. Notably the existence of a central message passer (server controller), which other architectures generally lack. Sometimes the entire architecture is framed to create the agency in question even if not identified directly.