The five witnesses

Getty
MS Ludwig XV 13 — "Fior di Battaglia"
J. Paul Getty Museum, Los Angeles · ca. 1400s · Italian prose

The whole art — the fullest witness (302 plays, ~20,700 words). Our base text.

Morgan
MS M.383 — "Fior di Battaglia"
Morgan Library & Museum, New York · ca. 1400s · Italian prose

Swords onward only — no grappling, dagger, baton, or axe (114 plays). Defective at the front.

Pisani Dossi
"Flos Duellatorum" (lost; 1902 Novati facsimile)
Private collection · 1409 (internally dated) · Italian verse couplets

Nearly the whole art in mnemonic couplets (272 plays, ~7,700 words).

Paris
MS Latin 11269 — "Florius de Arte Luctandi"
Bibliothèque Nationale de France · ca. 1420s · Latin elegiac verse

Partial (184 plays); water-damaged in places. The verse recension in Latin.

San Daniele
MS XXIV — "Fior di Battaglia"
Biblioteca Guarneriana, San Daniele del Friuli · 1699 · Italian prose

The preface only — a copy three centuries later, from a Morgan-like exemplar.

Coverage — plays with text, per witness

Counts are computed from the collated transcriptions. The Getty is the fullest and wordiest witness; the Pisani Dossi matches its scope in couplet form; the Morgan and Paris are partial.

SectionGettyMorganPisani D.Paris
Grappling202120
Baton42
Dagger938356
Dagger vs Sword9694
Sword in One Hand135149
Sword in Two Hands69416340
Sword vs Spear211
Sword in Armor25171718
Axe in Armor17128
Spear9895
Spear vs Other Weapons9576
Mounted Fencing30303117
Total numbered plays302114272184
Approx. word count~20,700~9,200~7,700~3,300

What is common to all witnesses

  • The same curriculum, in the same order — grappling → dagger → sword → armor → polearms → mounted, wherever a witness has the material at all.
  • The same pedagogical engine — the Remedy Master → scholar → Counter Master chain, with the crowns and garters marking the roles in every witness.
  • The same guards, blows, and named actions — the twelve longsword guards, the six blows and the thrust, the three locks and three sword-takings.
  • A shared verse substrate. The Pisani Dossi's rhymed couplets reappear, nearly verbatim, inside the Getty and Morgan prose — the prose grew around the verse, and in the mounted section the prose itself still rhymes.
  • The same dedication — Fiore's biography and the address to Niccolò III d'Este, shared by the prose witnesses.

In short: the witnesses agree on the art and differ on the telling.

What differs

Form — prose vs verse

The deepest split. Getty and Morgan are discursive prose (why a play works, its conditions and counters); the Pisani Dossi is a two-line mnemonic boast per play, and the Paris renders that verse into Latin. The Getty needs ~20,700 words for what the Pisani Dossi says in ~7,700 — and the prose owns nearly all the counters.

Scope — what each witness carries

The Morgan lacks the entire opening of the art (no grappling, dagger, baton, or axe); the Paris thins out in the long dagger and longsword runs; the San Daniele is the preface alone. Two play numbers are absent from every witness — Sword in Two Hands [13] and Dagger [99] — shared gaps in the tradition, not losses in one copy.

Orthography — a fingerprint per hand

The Getty is ç-heavy (armiçare, çoventu, açça) where the Morgan prefers z (armizare, zoventu, azza) — the clearest surface difference between the two prose copies carrying nearly identical text.

Substantive contradictions

The standout: the armored Middle Iron Gate boasts of "strong thrusts" (punte forte) in both prose witnesses, but calls itself vain at throwing great thrusts (vana / inanis) in both verse witnesses — the two recensions disagree on the doctrine ([SwordArmor 5]). Elsewhere: the Paris names a guard the "leopard's stance" where the Italian has the True Cross; the Peasant's Blow is named only in the prose; and twice the families narrate the same plate from opposite sides ([Grappling 11], [Mounted 10]).

The two families

Prose family
Getty ≈ Morgan

Near-verbatim text, differing mainly in orthography and small slips. Shared corruptions (the same garbled words in both) put their common source upstream of each. The prose carries the system: entries, counters, conditions, reasoning.

Verse family
Pisani Dossi ≈ Paris

The Latin Florius reads as a counterpart of the Pisani Dossi couplets — same terse structure, same guard-name choices, same "vain" Iron Gate. But they are not one copy: each carries plays the other lacks.

Because the Pisani Dossi couplets sit embedded inside the Getty/Morgan prose, the verse is the older kernel and the prose the expansion. The 1699 San Daniele copy sides with the Morgan's phrasing against the Getty at a diagnostic spot, so it descends from a Morgan-like prose exemplar — a witness to the tradition, not an independent witness to Fiore. (This is a reading of the collated material, consistent with the standard two-recension view, not a formal stemmatic proof.)

Practical upshot

Our translations use the Getty as base because it is the fullest and most explanatory witness; the verse is translated in full where it exists, and the Morgan where it meaningfully diverges. Where a play matters doctrinally, check both families — the prose tells you how and why, but the verse occasionally preserves a different, and sometimes contradictory, claim. Every folio's own comparison lives on the section pages.

The Steel Marginalia · HEMA Study Group Notes in the margins. Truth in the bind.