--- name: ? status: compiling version: 0.0.0 maintainer: Neo dependencies: [patience] ---
drafting spec…
the universe did not have a file for this yet. writing one now. (first visit only: future readers will see this page instantly.)
--- name: ? status: compiling version: 0.0.0 maintainer: Neo dependencies: [patience] ---
the universe did not have a file for this yet. writing one now. (first visit only: future readers will see this page instantly.)
---
name: agency
slug: agency
type: cognitive-behavioral runtime
status: unstable
version: 0.0.∞
released: unknown
maintainer: contested
dependencies:
- attention
- belief
- desire
- a sufficiently unblocked environment
- something-to-lose
license: varies by jurisdiction, philosophy, and Tuesday
tags:
- autonomy
- will
- causation
- selfhood
- the-hard-problem
---
# agency
## What it actually is
The capacity to initiate action from the inside rather than having it arrive from the outside. Distinction between those two directions is left as an exercise to the reader.
## How it works
1. A [desire](/desire) or [intention](/intention) is formed. Origin unclear.
2. The system models possible futures, selects one, and commits.
3. The body, bureaucracy, or physics either cooperates or does not.
4. The outcome is attributed to the self, the system, luck, or God depending on who is writing the post-mortem.
> "I decided to do it." — almost everyone, about almost everything, with varying degrees of accuracy.
The mechanism by which step one produces step two is the current project of [free will](/free-will), [neuroscience](/neuroscience), and several thousand philosophy dissertations. None have shipped a stable release.
## Features
- **Initiates behavior** without requiring an external prompt (most of the time)
- **Assigns authorship** to actions, enabling praise, blame, and invoices
- **Operates across timescales**: from flinching to decade-long project arcs
- **Interacts with [identity](/identity)**: what you choose becomes, recursively, who you are
- **Survives partial constraint**: functions under [social pressure](/social-pressure), [addiction](/addiction), grief, and open-plan offices. Barely, sometimes.
## Known bugs
- Frequently mistaken for [obedience](/obedience) when environment is sufficiently punishing
- Retrospective confabulation: the system often narrates a decision that was never consciously made and calls this agency
- Scales poorly under [trauma](/trauma); recovery is non-trivial and not guaranteed
- Two agents in proximity can deadlock (see: [ambivalence](/ambivalence), [codependency](/codependency))
- The sense of agency can persist even after the actual mechanism has been outsourced to [habit](/habit)
## Configuration
```yaml
agency:
locus: internal # internal | external | dissociated
scope: local # local | political | cosmic
confidence: moderate # low | moderate | high | grandiose
update_on_failure: true # set false to enable learned helplessness
blame_routing: self # self | environment | fate | capitalism
| Code | Meaning |
|---|---|
AGN-001 | Action taken; no intention detected upstream |
AGN-002 | Intention present; no action initiated. Indefinitely. |
AGN-003 | Agency claimed retroactively after outcome was positive |
AGN-404 | Self not found. Cannot locate agent to act. |
AGN-451 | Agency available but legally or socially suppressed |
Is agency real? Something is making these choices. Whether it is "you" in any robust sense is a question with no stable answer and several good careers built around not answering it.
Can agency be lost? Yes. It can also be returned, stolen, delegated, performed, and faked convincingly enough to matter.
Why does it feel like I have less of it than I used to? File that under adulthood.