then, presumably, it does not accommodate the conception of the body that is supposed to rise to the problem; but in that case, it presumably does not accommodate the distinction between basic and non-basic actions either. This is a point to which I will return.
5 What Is an Agent?
Danto does not go so far as to equate the agent with its body; that is, he does not say I am my body. Rather, he equates the agent with its powers to perform basic actions – which leads him to say, rather poignantly, that the life of an agent diminishes as these powers diminish, and when they cease altogether that constitutes the death of the agent. The point that I want to emphasize, however, is that these powers are not separable from the agent’s first-person point of view – that is, these powers exist only insofar as the agent who possesses them knows that it possesses them in a first-person way. It follows that the life of an agent is inseparable from its first-person point of view.
When we adopt a normative orientation in the philosophy of action, we will agree with this last point. But we will conceive the agent’s first-person point of view in normative terms, as the body of commitments from which an agent is committed to deliberating as it strives to meet the normative requirements that define individual rationality. The most general such requirement is a requirement to arrive at and act upon all-things-considered judgments about what it would be best to think and do, where the “all” ranges over all and only the agent’s own commitments.
I take it to be a definitional point that an individual agent ought to meet the requirements that define individual rationality. Somewhat more controversially, I take it to be a corollary of this definition that whenever, and wherever, there is a commitment to meeting these requirements, there is an individual agent who reasons and acts as one. What makes this a normative analysis is that it makes no reference to any non-normative condition of agent identity, such as the living human organism. If we were to suppose that an agent just is a human being, then we would have to suppose that an individual human being ought to strive to meet the normative requirements that define individual rationality within the whole of its human life. This would mean that it ought to arrive at and act upon all-things-considered judgments that take into account all and only the thoughts that arise within its human life. However, a human life is not necessarily the site of a commitment to doing this – a commitment to reasoning and acting as one – within the whole of that human life. This commitment may also arise within different parts of a human life, and it may arise within groups of human lives. In all of these cases, there is an individual agent who reasons and acts as one, from a body of commitments that constitute its own first-person point of view. The upshot is that while there can be agents of human size, there can also be multiple agents within a single human being and group agents comprising many human beings.
This is not the place to try to elaborate or defend these admittedly controversial claims about multiple and group agency. What I can do is clarify why they follow upon a thoroughgoingly normative orientation in the philosophical study of agency. Once I have clarified this, I can offer some reflections on what it means to say that agency is embodied, and compare the resulting conception to Danto’s.
The quickest route to seeing the normative point of the normative analysis of agent identity lies in the case of group agency. Insofar as a group of human beings is able to reason and act as one, it generally manages this by coordinating its efforts through overt discussion. But these discussions are not carried out in the spirit of arguing, or other techniques of persuasion that distinct agents exert over one another from their separate points of view. They are carried out rather in the spirit of working out what it would be best to think and do in the light of a body of pooled deliberative considerations. One example I often use in order to illustrate this highlights the difference between a philosophy department that “decides” matters by arguing and then voting, vs. a philosophy department that truly deliberates, by working out the all-things-considered significance of all of the considerations that have been brought to the table to consider. In the latter case, the discussions that occur in a department meeting have the same normative form that we find in the silent deliberations of an individual agent of human size: one thing is said, and the next thing that is said registers something that rationally follows from what is said, and that next thing is a step in further discussion that continues in the same spirit of working out what follows from the entire pooled set of deliberative considerations. This pooled set then constitutes the single point of view from which the philosophy department reasons and acts as one. Obviously, the boundaries of this point of view are not metaphysically given but are forged through effort and will.
According to the normative analysis, this is always true, no matter what boundaries get set. That is, the existence of an individual agent with its own point of view is always a product of effort and will. It follows that a human being is not born as an individual agent with a single, first-person point of view in the normative sense. A human being becomes the site of such a single point of view when it becomes the site of commitment to reasoning and acting as one within the biological boundaries of its life. But human lives may support other commitments as well – commitments to reasoning and acting as one within less than that whole human life, as well as across many human lives.
It might seem paradoxical to say that agents are products of effort and will: how can there be any such effort and will unless there is an agent already there to exert it? But there is no paradox, if we allow that the life of an agent consists in the intentional activities – the thinkings and doings – through which its life is constituted. It is the thinkings and doings that do the constituting work so as to yield a recognizable individual who reasons and acts as one. This, in my view, is the correct interpretation of, and the deep truth in, the claim that agents are self-constituting. The life of an individual agent just is its rational activities, which are directed at meeting the normative requirements that define individual rationality.
The boundaries that circumscribe a given agent’s life are always forged for reasons. These reasons concern the worth of what can be done by constituting a particular agent, with its own point of view, within those particular boundaries. It is obvious that what a group agent like a philosophy department can do, is not the same as what an agent of human size can do, which in turn is not the same as what can be done by multiple agents within a human life. This is another respect in which the normative analysis of agent identity is thoroughgoingly normative – not only does it say that no other metaphysical condition is required for the existence of an agent, besides the presence of a commitment to meeting the requirements that define individual rationality within a given boundary; but furthermore, it says that the very reasons for an agent’s existence are likewise normative, insofar as they have to do with the fact that reasoning and acting as one within that given boundary makes it possible to do something worth doing.
What, then, does the agent’s body emerge to be on the normative analysis of agent identity? To some extent, we can model it on Danto’s definition of an agent’s body, as that over which it exerts direct intentional control by performing basic actions, only we must substitute the normative definition of an agent’s powers that I offered above in my friendly amendment to Danto and Levi. What an agent can do in this normative sense is what the agent believes it will do if it recognizes reasons to do it. The agent’s body then emerges to be the domain of intentional control that is referenced by such first-person conditional beliefs about its own powers. Note that, just as the very existence of an agent is not a metaphysical given according to the normative analysis but a product of effort and will, the same holds for an agent’s body. This may seem paradoxical in just the way that the idea of self-constitution may seem paradoxical. But the point should be easy to see if we consider what lies within the power of a philosophy department to do, and how a particular domain of intentional control comes to be. This point about self-constitution makes for a sharp contrast
