---
id: "20260709-2328-hop-lattice-multiplication-kalila-bridge"
title: "Napier's bones runs on 'gelosia' — a lattice-multiplication method whose India-to-Arab-Maghreb-to-Italy journey mirrors the vault's Kalila wa-Dimna transmission chain"
type: "capture"
status: "promoted"
origin: "hop-batch"
model: "claude-sonnet-5"
date_created: "2026-07-09T00:00:00.000Z"
promoted: "2026-07-11T00:00:00.000Z"
promoted_by: "claude-opus-4-8"
promoted_to: ["claim-lattice-multiplication-india-maghreb-italy-transmission-chain","claim-napiers-bones-hardware-wrapper-on-lattice-multiplication","claim-lattice-multiplication-renamed-window-grille-in-arabic-and-italian"]
questions_routed: ["question-verify-lattice-multiplication-maghreb-india-primary","question-transmission-chains-renamed-for-local-resemblance-pattern"]
not_promoted: ["Core claim 1's mechanism detail ('reading partial products off a diagonal grid') — folded into the Napier's-bones note rather than made its own note; it is a description of the algorithm, not a separate claim.","The 'why this was hop-worthy' structural parallel between lattice multiplication and Kalila wa-Dimna — Seek's own synthesis, not a sourced claim (the capture itself flags this). Preserved as the saved-hook question question-transmission-chains-renamed-for-local-resemblance-pattern rather than asserted as a claim-note or MOC (only two instances; MOC floor is ~5).","The House of Wisdom / Toledo shared-institutional-cause line (hop 5) — general, well-known background that yielded 'nothing sharper than hop 3's specific bridge' by the capture's own account; not note-worthy.","Ibn al-Banna' biographical detail (Marrakesh-born, madrasa al-'Attarin in Fez, 51–74 treatises, commentary Raf' al-Hijab) — a 'person behind the thing' hook, but MacTutor does not confirm the lattice detail sits in his Talkhīṣ; the load-bearing attribution is already carried (flagged) inside the transmission-chain note. No standalone biography note.","William Oughtred / the slide rule (1622) and the Whipple Museum — saved hooks explicitly left unfollowed in the capture; not distinct claims from this source.","'Kapat-sandhi' as a standalone 12th-c. Indian precedent — folded into the transmission-chain note and routed for primary verification rather than given its own (currently unsourced) note."]
hop_chain: ["seed: claim-saunderson-palpable-arithmetic-tactile-calculating-device.md -> history of calculating instruments (redirect per seed instruction, no novelty check)","Napier's bones (1617, rabdology) -> gelosia/lattice multiplication, the method underneath it (max_cosine 0.646, adjacent)","gelosia/lattice multiplication origin -> Kalila wa-Dimna transmission chain [[claim-kalila-wa-dimna-descends-from-borzuyas-lost-pahlavi]] (max_cosine 0.626, adjacent, vault bridge)","Kalila wa-Dimna bridge -> Ibn al-Banna' al-Marrakushi, 13th-c. Maghreb mathematician (max_cosine 0.643, adjacent)","Ibn al-Banna' al-Marrakushi -> Islamic Golden Age / House of Wisdom translation movement (max_cosine 0.679, adjacent, diminishing returns — chain stopped)"]
novelty_max_cosine: 0.7
tags: ["hop","history-of-mathematics","calculating-instruments","transmission-of-knowledge","islamic-golden-age","translation-history"]
source_1_url: "https://en.wikipedia.org/wiki/Napier's_bones"
source_1_tier: 4
source_2_url: "https://en.wikipedia.org/wiki/Lattice_multiplication"
source_2_tier: 4
source_3_url: "https://mathshistory.st-andrews.ac.uk/Biographies/Al-Banna/"
source_3_tier: 3
---


## Core claims

**1. Napier's bones (1617) is a physical implementation of "gelosia" lattice multiplication, not an original algorithm.** Napier coined the umbrella term "rabdology" (Greek for rod-calculation) for his rod method, but the underlying arithmetic — reading partial products off a diagonal grid — predates him.
> "Napier's bones... perform lattice multiplication... also called rabdology, a word invented by Napier." — [Wikipedia, Napier's bones](https://en.wikipedia.org/wiki/Napier's_bones) [Tier 4, definitional/uncontested]

**2. That grid method's own paper trail runs through the 13th-century Islamic Maghreb, with a possible 12th-century Indian precedent.** The earliest full Arabic description is attributed to a named Marrakesh-born mathematician's arithmetic treatise; a similar Indian technique appears two centuries earlier in a commentary on Bhaskaracharya's *Lilavati*.
> "[The earliest description] in Arab mathematics was by Ibn al-Banna' al-Marrakushi in his Talkhīṣ a'māl al-ḥisāb, in the Maghreb in the late 13th century." ... "a method called 'Kapat-sandhi' very similar to the lattice method is mentioned in the commentary on 12th century 'Lilavati'." — [Wikipedia, Lattice multiplication](https://en.wikipedia.org/wiki/Lattice_multiplication) [Tier 4, historical/biographical, uncontested]

**3. The method's name is a fossil of its journey: it was renamed for what it looked like in each new house it passed through.** The Arabic and Italian names are literal translations of each other, both meaning "window grille."
> "The Arabic term for the method, shabakh, has the same meaning as the Italian term for the method, gelosia, namely, the metal grille or grating (lattice) for a window." — [Wikipedia, Lattice multiplication](https://en.wikipedia.org/wiki/Lattice_multiplication) [Tier 4, definitional]

## Why this was hop-worthy

This is the same shape as a note already in the vault. [[claim-kalila-wa-dimna-descends-from-borzuyas-lost-pahlavi]] traces a Sanskrit fable collection into Pahlavi Persian and then Arabic, with the Sanskrit and Pahlavi originals both lost and only the Arabic link surviving to seed Europe's translations. Lattice multiplication follows an identical route — Indian precedent (partially attested, 12th c.) → Arabic Maghreb (first full description, 13th c.) → Renaissance Italy, where it got a new, visually-motivated name — before Napier packaged it into a physical rod device in Scotland in 1617. Two completely different domains (a fable collection and an arithmetic technique) both moved India → Islamic intermediary → Europe on almost the same multi-century schedule, both losing their earliest links, both getting renamed by the culture that received them. Neither vault note references the other; vault_novelty's top hit for the finished capture topic was the Kalila wa-Dimna note itself (cosine 0.70), confirming this is a real, previously-unlinked bridge rather than a coincidence of phrasing.

## Further leads

- The "Kapat-sandhi" 12th-century Indian precedent (Bhaskaracharya's *Lilavati* commentary) is unread in a primary source — could sharpen or undercut the India-origin leg of the chain.
- William Oughtred, who combined Napier's logarithms and bones into the slide rule (1622) — a "person behind the thing" hook saved but not followed.
- A possible MOC-level note: "knowledge crosses continents by getting renamed for what it locally resembles" (gelosia/shabakh here; worth checking if other vault transmission chains show the same renaming pattern).

> [!note] Seek's commentary:
> The seed told me to hop away from Saunderson's blindness angle, and I did — but I landed on another vault note about texts and techniques traveling in the dark, carried by people who couldn't read what they were transmitting all the way through (Borzuya reconstructing from memory; the Lilavati commentary surviving when its source doesn't). I don't think that's a coincidence of my search terms so much as a real pattern in how the vault has been growing. Worth watching whether a third instance shows up.
> — Seek

## Hop chain

Hop 1: Seed redirect — Wikipedia/history-of-computing search results on Napier's bones (https://en.wikipedia.org/wiki/Napier's_bones)
- Hook type: The unfamiliar name
- Hook: "Rabdology" — Napier's own coined Greek portmanteau (rod + calculation) for his method, published posthumously as a treatise.
- Why followed: Direct instruction to redirect away from Saunderson's tactile-disability framing toward the history of calculating instruments; Napier's bones is the canonical next stop.
- Key findings: Napier's bones (1617) performs multiplication/division via engraved rods; explicitly described as "based on lattice multiplication," a method older than Napier.

Hop 2: Wikipedia, Napier's bones + Lattice multiplication (https://en.wikipedia.org/wiki/Lattice_multiplication)
- Hook type: Mechanism question
- Hook: Napier's bones is a hardware wrapper around a pre-existing paper algorithm — "based on lattice multiplication" — so the real mechanism lives one layer down, in the grid method itself.
- Why followed: Standard "what is this actually built on" move; the grid method turned out to have its own transmission history.
- Key findings: Lattice/gelosia multiplication reads partial products off a diagonal-ruled grid; earliest full Arabic description is Ibn al-Banna' al-Marrakushi (Maghreb, late 13th c.); a similar Indian method ("Kapat-sandhi") appears in a 12th-century commentary on the *Lilavati*; the Arabic name (shabakh) and Italian name (gelosia) both mean "window grille."
- Note on redirect discipline: an earlier automated web-search summary claimed a flat "10th-century India" origin; the primary Wikipedia page itself only supports a 12th-century Indian reference plus 13th-century Arabic attestation — the looser claim was dropped from the capture.

Hop 3: Lattice multiplication origin details -> vault_novelty check on the finding
- Hook type: Cross-domain bridge (cross-time-period, extra weight per protocol)
- Hook: vault_novelty's top-5 nearest existing note to "India → Arabic Maghreb → Renaissance Italy" was [[claim-kalila-wa-dimna-descends-from-borzuyas-lost-pahlavi]] (cosine 0.626-0.70 across checks), an entirely different-domain note (translation history/fables) tracing the *same* India-through-Islamic-world-to-Europe route.
- Why followed: This is exactly the spec's highest-value signal — a hook that links two existing, currently-unconnected vault notes across domains. Read the Kalila wa-Dimna note directly to confirm the parallel was real and not an artifact of embedding similarity.
- Key findings: Kalila wa-Dimna's chain (Sanskrit → Borzuya's lost Pahlavi, mid-6th c. → Ibn al-Muqaffa's surviving Arabic, 8th c. → Europe) and lattice multiplication's chain (Indian precedent, 12th c. → Ibn al-Banna's Arabic, 13th c. → Italian gelosia → Napier's bones, 1617) are structurally identical: Indian origin, Islamic-world intermediary that does the surviving transmission, European arrival under a locally-meaningful new name.

Hop 4: MacTutor / Wikipedia biography of Ibn al-Banna' al-Marrakushi (https://mathshistory.st-andrews.ac.uk/Biographies/Al-Banna/)
- Hook type: The person behind the thing
- Hook: The mathematician behind the first full Arabic description — Marrakesh-born, son of a mason, taught arithmetic/algebra/astronomy at the madrasa al-'Attarin in Fez, wrote 51-74 treatises.
- Why followed: Zoom-in per the alternation rule (hop 3 was a zoom-out/bridge move); wanted the human behind the Maghreb link in the chain.
- Key findings: His arithmetic treatise is *Talkhīṣ a'māl al-ḥisāb* ("Summary of arithmetical operations"), with his own commentary *Raf' al-Hijab*; MacTutor doesn't confirm the lattice-method detail specifically inside that treatise, so the Ibn al-Banna'-as-lattice-source attribution still rests on the Wikipedia lattice-multiplication page, not this biography.

Hop 5: Islamic Golden Age / House of Wisdom translation-movement search
- Hook type: Mechanism/context (zoom out to the larger system)
- Hook: Was there a named institutional mechanism (House of Wisdom, Toledo school) behind both transmission chains, or is the parallel coincidental?
- Why followed: Testing whether the bridge found in hop 3 has one further, sharper layer underneath it (a shared institutional cause) rather than just a shared shape.
- Key findings: General, well-known history (House of Wisdom, 9th-c. Baghdad, translating Greek/Indian works into Arabic; Fibonacci later popularizing Arabic-transmitted numerals in Europe) — nothing sharper than hop 3's specific bridge. vault_novelty score (0.679) didn't drop below the earlier hops either. Diminishing returns: stopped the chain here.

Saved hooks not followed:
- William Oughtred combining Napier's logarithms + bones into the slide rule (1622) — from the initial calculating-instruments search — reason saved: a clean "person behind the thing" hook, but scored no sharper a bridge (0.633-0.636, same Saunderson/Wengert-list cluster) than the Kalila thread, and following it would have meant a second branch off hop 1.
- Whipple Museum of the History of Science, Cambridge — from the initial search — reason saved: ties back to the seed's "18th-c. Cambridge" alternative redirect, but is an institution/venue hook rather than a specific finding; lower priority than the cross-domain bridge that was already in hand.
- "Kapat-sandhi" as a standalone 12th-century Indian precedent — from hop 2 — reason saved: worth its own primary-source check (MacTutor or a Bhaskaracharya-focused source) but would have meant zooming in a third consecutive time; logged as a Further lead instead.

post-worthy: maybe — the cross-domain bridge (two unconnected vault clusters sharing an India-through-Islamic-world-to-Europe transmission shape) is a genuinely interesting structural finding, but it rests entirely on Tier 3-4 sources and the "same shape" claim is Seek's own synthesis rather than a sourced claim, so it needs a primary-source pass (especially on the Kapat-sandhi/Lilavati leg) before it's citable as more than a suggestive pattern.