agent definitions — philosophy

overview

philosophical definitions of agency foreground intention, reasons, and the explanatory role of mental states. the cluster spans analytic action theory (anscombe, davidson) through functional perspectives such as dennette’s intentional stance. these works treat agency as a bridge between mind and action, prioritizing how and why an actor counts as the originator of behavior.

signature traits

• intentional descriptions: actions are “under a description” (anscombe, 1957) or tied to an agent’s primary reasons (davidson, 1963).
• normative accountability: agency often implies responsibility; explanations trace back to the actor’s rational capacities.
• predictive stances: later philosophy (dennett, 1987) shifts toward pragmatic strategies—treating something as an agent if adopting the intentional stance yields good predictions.

illustrative definitions

• 1957 — anscombe, intention: action is known through practical knowledge; intention anchors agency even before outcomes occur.
• 1963 — davidson, “actions, reasons, and causes”: intentional action arises when primary reasons cause the movement, weaving causality into agency.
• 1987 — dennett, the intentional stance: anything predictable via attributed beliefs and desires earns agent status, extending discussion to artifacts and systems.

relation to other dimensions

• autonomy spectrum: philosophy roots autonomy in deliberation; even low-power actors are agents if they form intentions.
• entity frames: definitions here remain firmly human-centered but open conceptual doors for machine agents via the intentional stance.
• goal dynamics: negotiation is less explicit; focus stays on how goals emerge from rational deliberation rather than external mandates.
• persistence & embodiment: embodiment matters only insofar as it anchors intentional states—agents can be abstract if they support intentional explanation.

open questions

• how far can the intentional stance stretch before “as-if” agency dilutes normative responsibility?
• which philosophical criteria survive when applied to synthetic agents with opaque internal representations?
• do contemporary debates on moral agency for ai require new distinctions beyond intention and rational causes?