About

Purpose, posture, colophon

Purpose

This site is a public explainer of discursive closure — a bug in how we argue: the question-begging move of putting the question-at-issue outside the frame of the conversation — as a rhetorical system, using the rhetoric of Donald Trump as the primary corpus. It aims to do for closure what a natural-history museum does for evolution: make a technical concept self-evident through specimens, mechanism, and lineage. It is not a news site, hosts no advocacy, takes no electoral positions, and updates on a fixed monthly cadence rather than the news cycle.

Authorship posture

The work may be published pseudonymously. The Method page exists so that the work is checkable independent of authorship: verifiability, not authority, carries the credibility. The dataset can be deposited under an organizational identity, so citation does not force de-anonymization.

Corrections policy

Corrections are public and visible: corrected text remains struck through for 90 days with the correction adjacent, and material corrections are announced in the changelog feed. Visible self-correction is on-thesis — this site models the epistemic behavior its subject forecloses. To submit a correction or a counterexample, use the adversarial-collaboration intake described on the Method page.

License

Text: CC BY-NC-ND. Graphics: CC BY-NC with attribution. Data: CC BY 4.0 — hostile reanalysis explicitly invited.

Colophon

Built with Astro; static output; no server, no database, no accounts. Typography: Libre Franklin (display and labels), Newsreader (body text), IBM Plex Mono (provenance lines and the codebook). No analytics, no trackers, no cookies. One caveat while in preview: fonts are served from Google Fonts, which is a third-party request; production self-hosts them. No photographs of any politician appear on this site, and no audio of the subject’s voice — prosody carries the affect the rule exists to exclude; where sound would aid analysis, a neutral narrated reading is used, with the primary video one click away. Dossiers are print-styled to read as clean two-page briefs for the seminar table.

Preview-build status

This deployment is a preview build. Every individually dated quotation has been checked against a primary transcript or archive (verification pass of July 2026); corpus-level frequency claims (the “witch hunt” count, the epithet table, superlative densities) still carry verification pending badges until the documented archive queries are run and published. Also staged but not yet running: the F1 reliability pilot, the coded-corpus explorer aggregates, the annotated full speech (legal review pending), the public changelog and corrections feed, and the Zenodo dataset deposit are staged but not yet running. The scaffolding shows where each lands.