Coda · after stage nine

Putting it all together.

The nine stages are not a staircase; they are a ring. Reflection at the end is a spark for the next turn. This page shows the whole loop on one page, then follows the running example — a contemplative reader — one full revolution around, to show how the concept evolves once the loop has been walked.

The loop, one page

Click any node to open that stage. The ring closes: reflection is the next spark. A full accessible list of the nine stages follows below.

The running example, all nine at once

An interactive daf — A contemplative reader that keeps the argument — traced through all nine stages, in progress.

01✺
Spark
The spark is a refusal. Most interactive humanities pieces choose one side: beautiful and thin, or rigorous and ugly. The daf refuses that trade — the page itself is the argument. I want a reader that feels like immersion and keeps the citation, the counter-voice, the gloss. The question I am carrying: can a page be contemplative and load-bearing at the same time?
02◐
Research
Three lineages, held in parallel. First, the daf itself — the Vilna page, its typographic choreography of Mishnah, Gemara, Rashi, Tosafot; why the eye trusts it. Second, contemplative interactives that set the bar: Bret Victor's explorable explanations, the Pudding's long-form, Are.na channels as slow reading. Third, the Grammar of Meaning corpus itself — what it can already do that a static page cannot. I am reading, not deciding.
03◑
Concept
One sentence: a reader where the primary text breathes, and the grammar surrounds it like commentary — visible, weighted, walkable. Success looks like a stranger spending fifteen minutes on one passage without noticing the time, and leaving with the citation. Failure looks like a beautiful shell that hides the argument, or a citation apparatus that kills the reading. The both-and is the whole point.
04◒
Form
The daf is the composition. Centre column: the text, generous line-height, one voice. Margins: the grammar — glosses, counter-voices, source links — set smaller, in a second face, so the hierarchy is felt before it is read. Colour does the argument's work: the primary voice in ink, the commentary in a muted second tone, disagreement marked, not shouted. Negative space is not empty; it is where the reader thinks.
05◓
Material
Web, not print — the medium has to move. HTML and CSS for the page, because the daf's logic is typographic before it is interactive. The grammar layer is data (JSON over the corpus), rendered as marginalia that can be dimmed, expanded, or followed. No framework yet; the prototype should be legible in view-source. Fonts are the real material choice — a serif that carries argument, a second face that recedes without disappearing.
06◔
Prototype
Three bad ones, cheap. (1) A single passage, static HTML, daf layout, no interaction — does the page hold? (2) The same passage with the grammar as hover-glosses — does the commentary help or interrupt? (3) The same passage with the grammar as a permanent margin — does the reader still read the centre? Line them up, show three people, kill two. The Thurman water mock is the ancestor of #1 and taught me that beauty alone is not enough.
07●
Make
The finished unit is one daf — one passage, fully realised — not the whole reader. Build it the way a printer sets a page: type first, then margins, then ornament, then interaction. Resist the urge to generalise before the single page is right. If the one daf works, the system follows; if it does not, no framework will save it. Keep the file small enough that the argument is legible in the source.
08◉
Present
Publish the single daf as a standalone page on grammarofmeaning.org, with a short essay on why the daf, not a demo reel. The frame matters: this is a reading, not a product. Invite three kinds of reader — a Talmudist, a designer, a stranger — and ask each the same question: did the page keep the argument? Their answers are the presentation's real output. The page itself is the artist statement.
09✦
Reflect
The honest questions, asked after: did the grammar earn its place on the page, or was it decoration? Did anyone read the centre column to the end? Where did the eye go that I did not intend? What would I cut if I could only keep half the marginalia? The next spark is already visible from here — either a second daf (does the system generalise?) or a retreat to the single page (was one enough?). Choose deliberately.

One loop later — the concept, evolved

From an interactive daf to a contemplative reader with a visible spine.

"A visible spine" — the citations, sources, and counter-voices are not hidden in footnotes; they run down the page as part of the composition, so the reader can see what the argument is standing on.

Walking the nine stages once changes the question. The first spark was a refusal — the false trade between beauty and rigour. After one turn, the refusal has hardened into a design principle: the citation apparatus is not a footer; it is the spine of the page. The commentary is not decoration around the text; the text and commentary are one composition, and the reader's eye is the third voice.

What survives from the original sketch: the centred primary text, the typographic hierarchy that is felt before it is read, the insistence on one passage done fully before generalising. What the loop revised: the marginalia are no longer optional gloss to be dimmed; they are load-bearing, and the interaction is not hover to reveal but read to unfold — the counter-voice appears as the reader arrives at the claim it answers. The daf metaphor stays as scaffolding, but the public name drops the jargon: a contemplative reader is what a stranger meets first; the Talmudic lineage is what a Talmudist recognises second.

The next spark is already visible: does the system generalise to a second passage? If yes, the reader becomes a form. If no, the single page was the work, and the honest move is to publish it as a page and stop.

← Back to the path