dotfiles/claude/memory/project-gloss-vs-commentary-distinction.md

permalink

permalink
claude-memory/project-gloss-vs-commentary-distinction

---

name: Gloss vs Readings as distinct ARC content types — distinction discovered 2026-05-13, name locked 2026-05-14 description: The Brown piece "The Obligating Word" (committed 82c2ba1 as class:gloss) is structurally Readings, not Gloss. Diagnostic: gloss illuminates a passage (passage = subject); a Reading composes its account of a work (work = subject, passages = evidence). Steward chose name "Readings" (plural) over "Commentary" — register family matches Glimpse/Fragment/Observation; the plural admits "these are my readings, not the readings." Genette's Seuils ratifies the structural frontier (paratexte vs métatexte; "forme latérale de la critique"). Remaining work: spec amendment + Brown post reclassification. type: project originSessionId: a5411fd1-4a96-43e7-aa1a-1a50d8951a8f

Context (how we got here)

2026-04-29 — First Gloss published: Berger · Ways of Seeing, p. 21. Form was single-anchor, apparatus-as-title. The §2 row of content-typology-specification.md noted "URL-Gloss form backfills when the form arrives."

2026-05-13 evening — Second piece prepared: "The Obligating Word" on Mandy Brown's Into the Gap. Composed by steward in Cardona 2026-05-09. Committed as 82c2ba1 with class: gloss. Two formal extensions from Berger noted in commit body:

• Named-thesis title ("The Obligating Word") distinct from apparatus locator
• Double-anchor with H2 scene markers between quoted passages

Steward then asked: "We may need to extend the gloss definition slightly: I wanted to have a way to comment on things — passages in my reading be it books, articles, online, offline…"

Claude proposed unified gloss-for-passages with apparatus as variable. Steward then asked, sharper: "Is what we just published pushing the limits of a gloss too far?"

That catch resolved the form question.

The structural diagnostic

• Gloss: "this passage means..." — the passage is the subject.
• Commentary: "this work argues X, and here is why..." — the work is the subject; passages are evidence.

The Brown piece's opening sentence — "Mandy Brown's Into the Gap is a Woolfian manifesto for refusal — and a quiet elegy for the word that once made refusal binding." — makes a claim about the essay as a whole. The two anchored sections elaborate that thesis using passages as evidence. The title "The Obligating Word" is a thesis-name about what the essay is and what it fails to make binding. That's commentary form, not gloss form.

The Berger piece, by contrast, illuminates the meaning of a specific paragraph on p. 21. It makes no claims about Ways of Seeing as a whole. The instrument-case extension follows directly from what the paragraph says. That is gloss form.

What the distinction implies for §V Content Types

• Two rows, not one.
• Gloss — passage as subject. Apparatus carries the locator inside the work (page, chapter, paragraph, timestamp).
• Commentary — work as subject. Apparatus carries the work-locator in full (URL, citation, archive locator). Multi-anchor with passages-as-evidence is the norm, not the exception.
• The "single vs multiple anchors" question is mostly resolved by this distinction: gloss tends single-anchor (illuminating one passage); commentary tends multi-anchor (evidencing a thesis across passages). But neither is strictly bounded — a long gloss could still be passage-focused; a tight commentary could turn on one quoted move.
• The steward's "comment on passages across media" framing remains correct for gloss — but doesn't subsume commentary, which has different structural commitments.

Three resolution paths (Claude's framing; steward leaning A)

A. Amend the Brown post to class: commentary and spec the type properly.

• Cleanest: honest classification, no temporary state, the form distinction announces itself paired with Berger as Gloss inaugural.
• Requires: spec amendment (content-typology §1 adds Commentary row; §2 adds its apparatus matrix row; AldineXXI §V Content Types adds Commentary entry; possibly a _commentary-source.scss partial paired with _gloss-source.scss).
• Re-commit the Brown post with corrected class once spec lands.
• Form questions in the original commit body resolve cleanly: not "extend Gloss to multi-anchor" but "introduce Commentary as a distinct content type."

B. Keep class: gloss temporarily, ship spec, reclassify on landing.

• Less clean: leaves the post mis-classified during interim.
• Buys nothing significant over A.

C. Argue the Brown piece really is gloss.

• Would require treating the opening sentence as transition rather than thesis work.
• Claude's read: the opening sentence is doing thesis work (claims about the essay as a whole). C doesn't hold structurally. But the steward composed the piece and knows what he meant.

Claude leans A. Steward leaned A but parked the decision pending fresh thinking.

What specs would need to be touched on A

File	Touch
docs/AldineXXI-Codex/content-typology-specification.md §1	Add Commentary as a 15th content type (currently locked at 14)
docs/AldineXXI-Codex/content-typology-specification.md §2	Add Commentary apparatus matrix row
docs/AldineXXI-Codex/AldineXXI-specification.md §V	Add Commentary entry to Content Types table
docs/AldineXXI-Codex/AldineXXI-specification.md §VI	Decide whether Commentary takes its own Reading Compass sigil or shares one
scss/AldineXXI/_gloss-source.scss	Possibly companion _commentary-source.scss if the visual treatment diverges; possibly shared partial if apparatus form is the only variable
site.hs	Eligibility lists per recurring pattern (per the 2026-05-07 code-shape audit)
content/posts/2026-05-09-the-obligating-word.md	Change class: gloss → class: commentary after spec lands
templates/partials/footer-compass.html	Add Commentary route if it gets its own sigil (per the Phase C spec consistency pass noted in MEMORY.md)

Open questions for the fresh-thinking session

1. Reading Compass sigil for Commentary — does it share with Gloss (both are "engaged response to a source") or get its own?
2. Apparatus form diversity within Commentary — book, online article, offline article, letter, manuscript, archival. Does each get its own apparatus subform, or is there a unified Commentary apparatus that accepts any locator?
3. Naming — "Commentary" is the working name. Other candidates: Response, Engagement, Reading. The lex-method approach would gather recognitions before locking. Worth a Round 1 chavruta if the steward wants it.
4. The §V "14 of 14 locked" status — adding a 15th content type touches the spec-stability commitment. Worth framing as an amendment per the recognized lex-method discipline rather than as a quiet addition.
5. Backfill rule for the URL-Gloss spec note in the Berger commit — does that note get retired (no URL-Gloss subform needed because Commentary covers URL sources) or refined (URL-Gloss as a Gloss subform for single-passage commenting on online articles)?
6. The steward's "I wanted to comment on things" framing still stands as the intent. The distinction we discovered tonight refines how to honor that intent across two forms rather than collapsing it into one.

Pickup notes for the next session

• The Brown post is committed but not deployed. Holding pattern: don't make build/rebuild until the spec question settles, OR ship the post as-is with class: gloss knowing it will be reclassified soon. Steward's call.
• The first move in the fresh session is probably the chavruta or direct-spec move on naming the second content type. The structural distinction is clear; the name and the matrix entries follow.
• This memory and the Brown post (file: content/posts/2026-05-09-the-obligating-word.md, commit 82c2ba1) are the materials. The Berger post (content/posts/2026-04-29-on-berger-ways-of-seeing-p21.md) is the comparison anchor.
• The form question recurred on the second instance, not the third or fourth. The lex-method principle ("the form discovers itself across instances") landed faster than the original Berger commit deferred for.

---

RESOLUTION (2026-05-14 morning)

Genette findings from Seuils (1987)

Steward identified Genette's Seuils in ~/Desktop/New texts may 2026/. Claude converted via pandoc (/tmp/seuils_genette.md, 4234 lines, 145 H1 headings, structure visible without repair_epub_headings pass — sufficient for reading-survey). Three load-bearing findings:

1. Genette names our exact frontier explicitly. From the chapter on the preface allographe:

Reste que la dimension critique et théorique de la préface allographe l'entraîne manifestement vers la frontière qui sépare (ou plutôt vers l'absence de frontière qui ne sépare pas nettement) le paratexte du métatexte, et plus concrètement la préface de l'essai critique.

His working terms in French:

• commentaire — the live unmarked word he uses throughout
• métatexte — the formal category for commenting-on-the-work-from-outside
• essai critique — the concrete form

Examples he cites of the form-crossing: Valéry's prefaces for Montesquieu, Stendhal, Flaubert have "tous les caractères d'essais critiques" and are now "fort légitimement, et indiscernablement, mêlées à des essais autonomes." Sartre's prefaces collected in Situations. Barthes' in Essais critiques. Sartre's Baudelaire and Saint Genet — "ci-devant préfaces," formerly prefaces, now standalone critical works. This is exactly the form the Brown piece is.

2. Borges, quoted by Genette — the line worth holding:

« Une préface, quand elle est réussie, n'est pas une manière de toast ; c'est une forme latérale de la critique. »

"When successful, a preface is not a kind of toast; it is a lateral form of criticism." The phrase "forme latérale de la critique" is the geometric image: the Reading runs alongside the work — neither subsuming it nor merely appraising it. Worth holding as the lex anchor for what a Reading is.

3. Genette ratifies the structural distinction without specifying its English name. He confirms paratext ≠ métatexte; he names commentary as the practice; he doesn't pick "Commentary" vs "Readings" in English. The English name choice belongs to ARC's lex, not to Genette.

Naming decision: Readings (plural)

Steward locked the name 2026-05-14: "Agreed in readings."

Three reasons that converged:

1. The plural admits epistemic humility built into the name. A "Commentary" sounds like Aquinas-on-Aristotle — authoritative voice over a work. "Readings" admits these are my readings, not the readings. Matches the steward's voice in the Brown piece itself ("she doesn't press it hard enough" is offered as a reading, not the verdict).

2. Register family. ARC's other content types are personal/partial: Glimpse, Fragment, Observation, Hearth, Hic, Adsum, Annals. "Readings" sits in this family; "Commentary" sits in the scholastic register (commentarius), at odds with the rest.

3. Pairs cleanly with Gloss. Gloss illuminates a passage (singular point of light). A Reading composes its account of a work (plural moves over time). The pair tells the story: I noticed this passage / I read this work.

Remaining work (now scoped)

The structural decision and the naming are settled. Outstanding:

• Spec amendment (the §V "14 of 14 locked" amendment): add Readings as 15th content type
  • docs/AldineXXI-Codex/content-typology-specification.md §1 (Readings as type) + §2 (apparatus matrix row)
  • docs/AldineXXI-Codex/AldineXXI-specification.md §V Content Types table; §VI Reading Compass sigil decision
• Reading Compass sigil for Readings — open question (own sigil, or share with Gloss?)
• Apparatus form for Readings — supports any source type (book / online article / offline article / letter / manuscript / archival); the apparatus carries the work-locator in full (URL, citation, archive locator); multi-anchor is the norm. Concrete form to be worked through on at least one more instance before locking the apparatus matrix; the lex-method principle says "the form discovers itself across instances."
• SCSS partial — _readings-source.scss paired with _gloss-source.scss; possibly shared if the visual treatment is identical and only the apparatus form differs
• site.hs eligibility lists (per recurring code-shape audit pattern)
• templates/partials/footer-compass.html — Readings route
• Brown post reclassification: content/posts/2026-05-09-the-obligating-word.md from class: gloss → class: readings once spec lands
• Lex-method recognition cluster for Readings apparatus across source-types — when the steward wants to gather instances

The Borges line ("forme latérale de la critique") and Genette's framework belong in the spec amendment as the lineage anchor. Readings is recovery, not invention.