Repository path: workshop/translations/kalila-labwa/collation.md · rendered 2026-09-09

Copy-text and collation — «باب اللبؤة والإسوار والشغبر»

1. The witnesses

Same pair of witnesses as ../kalila-saih/collation.md, ../kalila-nasik/collation.md and ../kalila-ibn-urs/collation.md. As with kalila-ibn-urs, the witness-2 chapter page is a bare transclusion ({{صفحات|فهرس=كليلة ودمنة.pdf|من=296|إلى=300}}) and returns an empty extract; the text was taken from the five صفحة: pages and cut at the chapter headings.

The witness-2 transcription for this chapter is pagequality level="1" — not proofread — and that is a real limit on this collation, stated here rather than discovered later. The recognition damage is dense and systematic: غ→ع (الشغبر → الشعبر throughout), ل↔ه (اللبؤة → البوة / الهبوة / القبوة), ث→ن or س (الثمار → النمار / السمار), ض→ح (تضجّي → تضحى), ظ→ر (ينظر → يُرْ), plus stray marginal footnote digits. A reading is adopted from witness 2 only where the underlying Arabic is unambiguous through the damage. Where the damage could be hiding a genuine variant, that is recorded in §3.3 and nothing is adopted.

2. Sentence numbering

The copy-text carries no numbering. Sentences are cut at ., ؟ and !, as in the three earlier chapters. The split yields 37 sentences, numbered [1]–[37] in copy-text-adopted.txt, and every citation on the translation page and in the two figure inventories uses these numbers.

One boundary is a corruption and is not repaired by renumbering. The copy-text prints a full stop after «وكان عيشه من الثمار» at the end of [31], orphaning a فلما clause with no apodosis; [31] and [32] are one syntactic period («فلما رأى ذلك ورشان … قال لها: …»). Witness 2 has a stray ) at the same place, so the damage is in both. The numbering is kept as split — renumbering would break citation against the raw copy-text — and [31]+[32] are rendered as one period.

3. The collation

Read whole, both witnesses, in Arabic.

3.1 Substantive adoptions — five, all from witness 2

3.2 Orthography

The witness-2 chapter heading spells اللبؤة with hamza; the copy-text page title spells it اللبوة while its own body has لبؤة at [7] and اللبؤة at [13], [19], [21], [23], [25], [30], [35], [36]. اللبؤة is used throughout this project's pages, including in the work title. Witness 2 glosses it in the margin: «الأسدة وهى مهموزة وغير مهموزة» — the lioness, spelt with or without hamza — so the variation is the tradition's, not an error in either witness. Vocalisation is witness 2's alone and is not carried into the adopted text.

3.3 Recorded and not adopted — six places

Each is a difference where witness 2's reading is either equally good or too damaged to trust.

3.4 What witness 2 supplies that the copy-text does not

Four marginal glosses, printed in the 1937 edition and absent from the copy-text. Two of them settled a rendering decision (see the translation's log, D2 and D6):

• اللبؤة — «الأسدة وهى مهموزة وغير مهموزة» (the lioness; spelt with or without hamza)
• الإسوار — «قائد الفرس» (a commander of horse) — the word is Persian aswār, a mounted noble, not a generic hunter
• العراء — «الفضاء لا يستر فيه شيء» (open ground where nothing is covered)
• الورشان — «طائر شبه الحمامة والأنثى ورشانة» (a bird like a dove; the female a warshāna)

4. Result

copy-text-adopted.txt is the copy-text with the five adoptions of §3.1 and nothing else. 599 words of body in 37 sentences; the adoptions change no word count. It, and not the raw copy-text, is what was translated.