Wednesday, September 16, 2020

Agent causation

I have long identified as having an agent-causal theory of free will. But I have just realized that my Aristotelian take on agent causation is far enough from the most common agent-causal theories—those of Clare and O’Connor—that it may be misleading to talk of myself as accepting an agent-causal theory of free will.

Standard agent-causal theories distinguish between agent causation and event causation as two distinct and real things in the world. For instance, they hold that I agent cause my writing of this post but there is an event causal relation between my being in this armchair and the cushion being squished. Agent causation has the agent as the cause and event causation has an event as the cause. But I think in both cases the cause is the same: it is a substance, namely myself. I cause the writing of this post and I cause the cushion to be squashed. On the standard view on which agent causation is distinguished solely by the fact that the cause is the agent, both are cases of agent causation. But that would be misleading to say.

So, if we take seriously the Aristotelian account of causation as substance causation, we shouldn’t distinguish agent causation from other kinds of causation by whether the cause is an agent or something else. But we can still make the distinction. My writing this post is an actualization of a power of my will (or my practical rationality, if you prefer). My squashing the cushion is an actualization of the power of my weight. Agent causation is distinguished from other kinds of causation not by what does the causing, but how it does the causing. In agent causation, the substance causes by actualizing its will (and any substance with a will is an agent). In other kinds of causation, the substance causes by actualizing a different power.

So, I think the fundamental relation underlying causation is actually at least ternary: agent X causes event E by actualizing power P.

This neatly integrates agent causation with reasons causation. There are more and less proximate powers. I now have a nearly proximate power to speak Polish (some motivation might be needed to make it proximate). When I was an infant, I had a remote power to speak Polish, by having a nearly proximate power to learn Polish. When the power P in the causal relation is specified as the maximally proximate power, in the agential case, it is a maximally proximate power of the will. And a maximally proximate powers of the will are tied to reasons: it is only by having a reason to do something that I am able to will it (one wills under the guise of the good). So, my reasons for action supervene on the maximally proximate power for action. Thus, the ternary description of agent causation neatly includes the reasons.


Peter said...

O’Connor and Jacobs have a chapter called “Agent Causation in a Neo-Aristotelian Metaphysics” in Mental Causation and Ontology (ed. S. C. Gibb, E. J. Lowe, and R. D. Ingthorsson) that explores and briefly refines (but falls short of endorsing) this sort of view. I was always inclined to accept this sort of unified view of causality. (They do not distinguish two types of substance causation by pointing to the will, but instead claim “What is distinctive about agent causation among other varieties of substance causation on this view is merely that the cause is conscious, intentional, and freely chooses the ends for which it will act. As we see it, each of these distinctive sorts of capacity is ontologically fundamental.”) In any case, I always thought this alternative view was preferable to the one O’Connor develops elsewhere.

One question about your own view. You write; “So, I think the fundamental relation underlying causation is actually at least ternary: agent X causes event E by actualizing power P.” I am curious about the relationship between the causation and the actualization. Does the actualization of the power constitute the agent’s causing something, or do you think of this as a process with something prior and something posterior (perhaps the actualization of the power is somehow prior to the agent’s causing the event)?

Alexander R Pruss said...

I think the actualizing is the causing.

A Counter Rebel said...

Let's say Didi agent-causes A (Didi-->a). In another universe (inaccessible from the other), eveything runs the same until the point of the agent-causing. In this world, Didi* agent-causes B (Didi*-->b). (I don't pretend I'm about to make a new or uncontroversial argument). Why did Didi* agent-cause differently? You might say because of how they exercise their free will, but the exercise of free will is the agent-causing, so that would just be saying they agent-caused differently because they agent-caused differently. It seems like it is a matter of luck that the agent agent-caused A rather than B. I define randomness as (1) without definite [synonym] plan, purpose, or pattern (Webster's) (2) event with no cause (3) an event where a different event could've occurred with the cause being the same. Agent causality fits all of these definitions. The agent is torn -- its considered reasons and desires don't definitively lead to a certain outcome--1. The complex event "Agent causes event" has no cause--2. And the event that is the object of the overall event could've been different with every aspect of the agent being identical--3.

I distinguish between causation and control. All control implies causation, but not all causation implies control. To control X means X was intended. Also, control implies method. An agent does not get to choose whether it controls its actions in a deterministic style or indeterministic style. So even if agents control choices, they control them in a manner that they didn't consent to. (They have no control over their control.) Furthermore, agents didn't choose to be the agents they are, so choices are determined by something unchosen. Why should I be responsible for choices just because I am the cause, rather than anyone else? Either way, its cause is alien (something not-up-to-me).

Take the event of "Agent causing an intention"? Did the agent intend for this event to happen, or did it just happen? It just happened. You might say it is impossible to cause an agent-caused event. Say it's true. Even so, it doesn't contravene the fact that there is no cause for it.

Alexander R Pruss said...

On the ternary relation proposal in my post, it is not correct to say that in the two universes the cause is the same. For it is a different proximate power in the agent that is involved in the causation in the two worlds.

A Counter Rebel said...

They're the same in every relevant sense. Obviously they're not token-same.

Alexander R Pruss said...

Let's say that I choose between eating ice cream and playing a game. In one world, the cause is me + my power to eat ice cream, and in the other case the cause is me + my power to play a game. These are different causes.

A Counter Rebel said...

I thought the whole point of agent-causal free will is that free will is a single two-way power. In any case, why did the power to eat ice cream take ascendance over the power to play a game? Does the agent first cause a power to take ascendance, and then the agent and power co-cause the event of intending to eat ice cream?