⁂ Concepts · Mind · No. 94
Agency

Alien hand: ownership of the limb intact, authorship of the act gone.

Agency

The capacity to act rather than merely to move, and the sense that a movement is one's own doing.

Theme
Mind, self and consciousness
Tradition
Cross-tradition
Term fixed
c. 340 BC

In plain terms

Your arm going up and you raising your arm are not the same event described twice. One can happen to you; the other is something you do. Everyone can tell the difference from the inside, immediately and without evidence.

What that difference consists in is much harder. It is not effort - the easiest actions are still actions. It is not deliberation - most of what you do is not deliberated. It is not physical origin - a reflex and a movement start in the same body.

And the sense of being the author of an action is separable from the action itself. It can be present when you did not cause the movement, and absent when you did, which suggests it is something the system computes rather than something it simply has.

As a philosopher would put it

Anscombe's criterion is that actions are the movements to which a certain sense of the question why applies - the sense answered by a reason rather than a cause. Davidson's reply is that reasons are causes, and that an action is a bodily movement caused in the right way by a belief-desire pair, which raises the problem of deviant causal chains: the right cause producing the movement by the wrong route is not an action.

The comparator model is the standard empirical account. The motor system predicts the sensory consequences of its own commands and compares them with what arrives; a match yields the sense of agency, a mismatch yields the sense that something happened to you. This explains why you cannot tickle yourself and why the sense of agency can be manipulated in the laboratory.

Frankfurt's contribution shifts the question from causation to structure: what matters for agency is whether one identifies with the desire one acts on, which is a hierarchical rather than a causal criterion. The disability and situated-cognition literatures add that agency is distributed - it depends on tools, support and circumstance, and treating it as a property of an isolated individual misdescribes most of it.

Where you meet it

Alien hand syndrome follows damage to the corpus callosum, the medial frontal cortex or the parietal lobe. The affected hand performs coherent, purposive actions - unbuttoning a shirt that the other hand has just buttoned, reaching for objects, in some reported cases interfering with the patient's own intended movements.

The patients are not confused and not psychotic. They can see the hand is theirs, will say so, and are frequently distressed by it. What is missing is not ownership of the limb but authorship of what it does; several describe restraining it with the other hand, or sitting on it.

So the two things ordinary experience delivers together - this is my hand, this is my doing - come apart cleanly, and the second can fail while the first holds. Whatever produces the sense of agency is a specific mechanism with a specific anatomy, and it is not simply the fact of having caused the movement.

What is actually disputed

Whether agency requires a special kind of causation.

The causal theory makes actions ordinary events with mental causes, which fits them into the natural world at the price of the deviant-chain problem, never fully solved. Agent-causation holds that a person causes an action directly rather than an event causing it, which preserves the intuition and is hard to make intelligible.

A separate argument concerns how much the sense of agency tracks the fact of it. The comparator model and a body of experimental work show the sense can be induced when the subject did not act and removed when they did, which some read as showing that agency is a construction imposed after the fact. Others reply that a fallible detector is still a detector, and that showing a sense can be fooled establishes nothing about whether the thing it detects is real.

Commonly confused with

  • Free will c. 397 Free will asks whether one could have done otherwise. Agency asks what makes something a doing at all, and a determinist can accept the second while denying the first.
  • Autonomy 1785 Autonomy is governing oneself by one's own considered principles. Agency is the capacity to act, which a thoroughly unfree person still has.
  • agency in the sociological sense That is about the scope for action within structures. This is about what an action is.

Defined against

  • Determinism Whether agency survives a world in which every event is fixed by prior conditions is the oldest question about it

Philosophers

  • Aristotle The voluntary and the involuntary, and where responsibility attaches
  • Immanuel Kant Acting from a maxim rather than merely in accordance with one
  • Jean-Paul Sartre Action as the exercise of a freedom that cannot be evaded
  • Hannah Arendt Action as beginning something in a world with others
  • John Dewey Agency as transaction with an environment rather than an internal event

Works

Schools

Related concepts

Aristotle's distinction between voluntary and involuntary action sets the terms; Anscombe's Intention in 1957 restarts the modern discussion by asking what makes something an action at all.