identity / hostile edition

THIS WORK
Copyright (C) 2026 the project contributors.
Code: AGPL-3.0-or-later. The complete license is served at
../otherwise/LICENSE.txt. The complete, unminified application source is
served beside this notice.

HOW IT WAS MADE
The artist directed this work. The poster compositions, experimental
application and home-page integration were written with Codex, an OpenAI
language model, in October 2026. This is a separate edition of the earlier
identity poster study, with a new design vocabulary and protocol.

The c0unt typeface in ../f/ was made for this repository with Claude at the
artist's direction in the initial otherwise session. Its attribution is
retained. The scrawled and photocopied appearances here are computer-drawn
treatments of c0unt, not undisclosed human handwriting or archival flyers.

This study uses no Strudel code, third-party runtime libraries, recordings,
external models or provider credentials. The older otherwise experience
remains available at ../otherwise/ with its own credits and license notice.

LOCAL RECORDING
Recording begins when a participant selects Begin after the local-recording
disclosure. It includes pointer movement, clicks, control-key categories,
touch gestures, viewing time and deliberate responses. Arbitrary typed text
is not recorded. Records remain on the device; the application uploads no
participant data. JSON exports contain the retained session and events.
Researcher imports combine only the files the operator chooses locally.
The design preview creates no participant record and captures no interaction
data. Preview exposure can still influence a later participant's responses.

CLAIM BOUNDARY
Protocol identity-irreverent-1 records experimental assignments and explicit
self-report. Cursor movement is not eye tracking. Viewing time cannot prove
belief or identity. Interpretive feedback is an experimental suggestion, not
a validated diagnosis. The assessment follows the participant's reported
position. The opposite presentation retains the same recorded evidence.

The eight visual families are compound design treatments. Palette and print
finish are crossed within each family. Statistical differences do not isolate
every individual shape or typographic decision. The requested matched-then-
opposite reveal order is fixed and is confounded with repetition and order.
The earlier identity study's data are not pooled into this protocol.