If experience is usually assumed to be of things, action is usually assumed to be done by agents.
Someone decides. Someone intends. Someone acts. The world changes as a result.
This picture feels so natural that it is rarely examined. Even attempts to soften it — by appealing to unconscious processes, neural mechanisms, or social conditioning — tend to preserve the basic architecture. There is still an inner locus from which action issues, even if it is more crowded or less transparent than we once thought.
This post asks what happens if we suspend that architecture.
Not to deny action, and not to dissolve responsibility, but to ask a prior question: what if action does not begin with agents at all?
Why Agents Feel Necessary
Agents promise explanation.
They allow us to say why something happened: because she chose to do it. They give action a source, a centre, a point of origin that seems to make sense of deliberation, intention, and control.
They also promise moral clarity. If there is an agent behind an action, then praise and blame appear to have somewhere to land. Remove the agent, and it can feel as though action becomes a kind of impersonal drift — things merely happening, with no one accountable.
These pressures are strong, and they are not irrational. But they are pressures toward a particular explanatory grammar, not evidence that the grammar is unavoidable.
Acting Before Deciding
Consider a familiar case: reaching for a cup while speaking.
You do not pause the conversation, form an explicit intention, issue a command to your arm, and then resume talking. The movement happens within the ongoing activity. It is coordinated with posture, gaze, rhythm of speech, and the location of the cup — all without being planned as a discrete act.
Or consider correcting yourself mid‑sentence. The words are already leaving your mouth when the correction occurs. There is no temporal gap in which a central agent intervenes. The action re‑patterns itself as it unfolds.
In skilled activities — driving, typing, playing music — this is even clearer. If you attempt to locate the agent that is “in charge,” performance degrades. Action improves when agency recedes.
These are not marginal cases. They are the norm.
From Agents to Coordination
The ontology developed on this blog treats action not as something issued by an inner controller, but as a construal of coordinated processes.
Bodies, environments, histories, social norms, and immediate affordances participate together in what we later describe as “someone acting.” The description is not false — but it is compressed. It packages a complex relational event into a single grammatical subject.
From this perspective, an agent is not the origin of action, but a retrospective stabilisation within it.
This mirrors what we saw with objects in experience. Just as objects emerge as stable patterns within perceptual relations, agents emerge as stable patterns within practical relations. Neither needs to be primitive.
Intention Without an Intender
A common objection arises here: surely intentions must belong to someone.
But notice how intentions actually function. An intention is not a private object consulted before acting. It is a constraint on how action unfolds. It shapes timing, sensitivity to interruption, responsiveness to outcomes.
Saying “I intend to finish this paragraph” does not describe a mental thing pushing events forward. It marks a trajectory as held open against distraction, revision, or abandonment.
Intentions, in other words, are features of ongoing coordination, not possessions of an inner agent.
Responsibility Revisited
If action does not originate in agents, what becomes of responsibility?
What disappears is not accountability, but the fantasy of absolute authorship.
Responsibility, on this view, tracks participation in patterns of coordination. It concerns how one’s history, responsiveness, and capacities enter into what happened — not whether an isolated self could have done otherwise in a metaphysical vacuum.
This is not a weakening of responsibility. It is a relocation of it, from an imagined inner sovereign to the real structures of practice in which actions take shape.
What Becomes Visible
Once agents are no longer treated as foundational, several features of action become clearer:
Action is continuous, not a sequence of discrete acts.
Deliberation is embedded, not preparatory.
Control is distributed, not centralised.
Failure and success are relational, not personal essences.
Most importantly, the familiar opposition between free will and determinism loses its grip. Both presuppose an agent standing apart from the world, either constrained by it or mysteriously exempt from it.
When action is understood as relational coordination, that standing‑apart never arises.
Acting Without Erasure
To say that action does not begin with agents is not to say that persons do not matter. Persons matter precisely because patterns of coordination stabilise around them — around their histories, sensitivities, skills, and commitments.
But these patterns are maintained, not owned.
Action, like experience, does not need a hidden core to get started. It is already underway.
In the next post, we will turn from action to meaning, and ask what becomes of language and understanding once representation is no longer taken as the default model.
For now, it is enough to see this: