Three Lives open on this page, each introduced by a ONE-WORD NAME RUBRIC standing alone on its
own line and closed by a low point. Read from the plate, not inferred:
· line 3 of the page (= piece line 340) ܐܚܝܐ ܂ — AHIJAH
· line 11 (= piece line 348) ܝܘܐܕ ܂ — JOAD (Ἰωάδ / Iohad)
· line 17 (= piece line 354) ܥܙܪܝܐ ܂ — AZARIAH
Lines 1-2 close the PRECEDING Life, whose name stands on p. 102 and is therefore not on this
page; the close reads … ܠܒܪܗ ܂ / ܡܝܬ ܘܐܬܩܒܪ ܒܐܘܪܫܠܡ ܂ ("… to his son. He died and was buried
in Jerusalem."). ⚠ The Greek recension's order at this point is Nathan · Ahijah · Joad ·
Azariah · Elijah · Elisha · Zechariah b. Jehoiada, so the preceding Life is very probably
Nathan — but that is INFERENCE FROM THE ORDER, not a reading, and it is not recorded as fact.
⭐ MECHANICAL CUE FOR A LIFE BOUNDARY, confirmed on this page three times and reusable across the
piece: a short line that stops well short of the left margin and ends in a point, then a
ONE-WORD name rubric alone on the next line, then the Life proper opening with the SAME name
(line 4 here opens ܐܚܝܐ ܡܢ ܫܝܠܘ ܐܝܬܘܗܝ ܗܘܐ ܆ — "Ahijah was from Shiloh"). A short line on its
own is only a paragraph break; the one-word rubric is the boundary.
Nestle names no individual codex here. The piece as a whole is headed on p. 86
"(E tribus codicibus Musei Britannici.)" and he never says which of the three carries which
Life. There is no witness field in the schema; that gap is a known finding.
Verified on this page, not assumed: NO vowel signs anywhere on any of the 20 lines. What the
plate carries is (a) seyame on plurals (e.g. ܕܒܝܘܡ̈ܘܗܝ on line 1), (b) the dalath/rish points
(the ܕ of ܫܕܪܗ takes its dot BELOW, the ܪ its dot ABOVE — checked on the glyph, not assumed),
(c) further single letter points, including a dot over the final ܗ of ܠܒܪܗ. Constant with the
neighbouring pages of the shard (pp. 104-107) and with p. 86.
Th. Schermann, PROPHETARUM VITAE FABULOSAE, Teubner 1907 — archive.org prophetarumvita00schegoog. Latin version of the Syriac (cod. Sinait. syro 10) at pp. 105-106, leaf = page + 74 (calibrated on two running heads); Index nominum propriorum at printed p. 240. ⛔ The Latin Appendix is a two-page CAPITULUM (a summary list of 22 prophets with tribe, date and burial place), not the full Lives — SENSE-level control only, and a thin one. It catches a mis-keyed name or a meaningless sentence; it never adjudicates pointing and cannot be aligned page-for-page. See research/syriac-peshitta-editions.md
340, 345, 350, 355 — read off the plate in the RIGHT (outer) margin of this recto, standing against printed lines 3, 8, 13 and 18. The page therefore carries piece lines 338-357. Numbering is CONTINUOUS across the whole of Chrest. III and does NOT restart at a prophet: three Lives begin on this page and the count runs straight through all three.
Conventions, as ruled for this run:
· seyame → U+0308 combining diaeresis on the letter it sits over
· linea occultans → U+0304 combining macron over the elided letter
· single low point → ܂ · two-dot pause → ܆ · diamond/3- or 4-dot end mark → ܀
The qushshaya/rukkakha and participle points on individual letters are NOT transcribed. They ARE
on the plate; they are below the resolution at which this pass can be honest about them. That is
a KNOWN GAP, not a claim the plate lacks them. Likewise the dot over ܗ in ܠܒܪܗ and over the
final ܗ of ܫܕܪܗ is seen and deliberately not encoded.
⛔ No square brackets (Nestle's exemplar-gap marks) occur on this page.
⛔ Doubtful words are marked ⟨?…⟩; a line this pass could not read at all is emitted in position
as ⟨?LINE NOT READ⟩ so the line count stays right. NOTHING here was filled in from the Latin or
Greek control — that would be circular.
ܠܒܪܗ ܂ ܫܕܪܗ ܡܪܝܐ ⟨?⟩ ܕܒܝܘܡ̈ܘܗܝ ⟨?⟩ ܡܝܬ ܘܐܬܩܒܪ ܒܐܘܪܫܠܡ ܂ ܐܚܝܐ ܂ ܐܚܝܐ ܡܢ ܫܝܠܘ ܐܝܬܘܗܝ ܗܘܐ ܆ ⟨?⟩ ⟨?ܕܐܝܬܘܗܝ⟩ ܗܘܐ ⟨?⟩ ⟨?LINE NOT READ⟩ ⟨?LINE NOT READ⟩ ⟨?LINE NOT READ⟩ ⟨?LINE NOT READ⟩ ⟨?LINE NOT READ⟩ ܝܘܐܕ ܂ ⟨?LINE NOT READ⟩ ⟨?LINE NOT READ⟩ ⟨?LINE NOT READ⟩ ⟨?LINE NOT READ⟩ ⟨?LINE NOT READ⟩ ܥܙܪܝܐ ܂ ⟨?LINE NOT READ⟩ ⟨?LINE NOT READ⟩ ⟨?LINE NOT READ⟩
NONE on this page. There is no rule and no smaller type at the foot; the twentieth text line runs to the bottom of the type area. The only apparatus is the head numeral and the marginal line numbers.
keyed from the plate, PARTIAL and UNVERIFIED: 7 of 20 printed lines carry a reading, 13 are emitted as ⟨?LINE NOT READ⟩; within the 7, 5 words or stretches are flagged ⟨?…⟩
⚠ THIS IS THE FINDING, and it is a measurement, not an excuse. The master scan is 1586x2512 —
about 380 px across a quarter-page crop, ~5x when upscaled to 2000 px. At that scale an
UNVOCALIZED Estrangela word is legible only when the reader already has a hypothesis for it:
every word actually keyed above (ܡܝܬ, ܘܐܬܩܒܪ, ܐܘܪܫܠܡ, ܡܪܝܐ, ܗܘܐ, ܐܝܬܘܗܝ, ܐܚܝܐ, ܝܘܐܕ, ܥܙܪܝܐ,
ܕܒܝܘܡ̈ܘܗܝ, ܫܕܪܗ, ܠܒܪܗ) was confirmed against a guess, and every word with no obvious candidate
resisted three to five successive zooms. The observed cost is roughly 5-6 zoom crops per printed
line, i.e. ~500 crops for this shard's 98 lines, and even then the residual error would be
silent. ⛔ Ninety-percent-accurate ground truth with undetectable errors is worse than none for an
HTR corpus, so the unread lines are declared rather than guessed.
⭐ Two calibration aids that DO work and should be reused by anyone who takes this on again:
(1) Nestle's own alphabet table, p. 4 = leaf n21 — the BOLD glyph in each cell is Estrangela, the
light one Serto; ܠ is a tall ascender, ܥ the same wedge WITHOUT the ascender, ܣ a closed
double lobe, ܚ two humps on a flat base, ܩ a loop plus bar, ܡ a closed loop with a leg.
(2) p. 86 lines 1-3, whose text is already established, is a labelled specimen of ~15 letters.
No digital text exists: syriacgrammarwit00nestiala_djvu.txt is EMPTY (0 bytes) — the item has no
OCR at all, so there is nothing to diff a keying against even mechanically.
WHAT THE CONTROL CONFIRMED. Only the names, and only weakly. Schermann's Latin capitulum has
IV. Nathan ex Gabaon…, V. Achias porro fuit in diebus Ieroboam., VI. Ioad erat ex Samaria…
— so AHIJAH (ܐܚܝܐ) and JOAD (ܝܘܐܕ) are both in the Sinai list and both in this neighbourhood
of it, which is a real (if thin) check on two of the three rubrics I read.
WHAT IT CANNOT DO, on this page more than most. (a) AZARIAH (ܥܙܪܝܐ) is NOT in the Latin
capitulum's 22 at all — so the third rubric has no control and rests on the plate alone.
(b) The capitulum's ORDER is different from Nestle's (Latin runs Moyses · Samuel · David · Nathan
· Achias · Ioad · Elias …, Nestle's BM recension runs the Greek order beginning with Isaiah), so
the control cannot be used to predict what stands on a given page. (c) It is a summary list in
Latin against a full narrative in Syriac: nothing in the running text of this page has any
parallel in it, and it never adjudicates a letter, a point or a seyame.
⛔ Nothing above was read off the control. ܐܚܝܐ was decided on the glyphs (alaph · heth ·
yudh · alaph) BEFORE the Latin was consulted, and ܥܙܪܝܐ on 'e (low wedge, no ascender) · zayn ·
rish (dot above) · yudh · alaph.
UNCERTAIN READINGS ON THIS PAGE, listed as required:
· line 1, the word after ܡܪܝܐ — not resolved.
· line 1, the word after ܕܒܝܘܡ̈ܘܗܝ and the line's final word — not resolved (a lamadh-initial
cluster ending the line).
· line 4, the final word — not resolved.
· line 5, opening word read as ܕܐܝܬܘܗܝ but the initial dalath is not certain; the rest of the
line not resolved.
· lines 6-10, 12-16, 18-20 — not read at all.