| Work / author / dates | Tarjumān al-ashwāq (ترجمان الأشواق, "The Interpreter of Desires"), a dīwān of 61 love poems · Muḥyī al-Dīn Ibn ʿArabī (560–638 AH / 1165–1240 CE) |
| Part added | Poems I–X (64 abyāt), one library text id tarjuman; later poems to be appended as further sections |
| Reader route | /read/tarjuman (not yet created: needs a page under app/) |
| Status | working edition, transcribed and proofread twice (28 Sept 2026) · reviewed by: — |
1. Witnesses
| Siglum | Witness | Kind | Vowelled? | Licence | Location |
|---|---|---|---|---|---|
| Nich | R. A. Nicholson (ed. & tr.), The Tarjumán al-Ashwáq: a collection of mystical odes by Muhyi'ddín Ibn al-'Arabí, London: Royal Asiatic Society, 1911 (Oriental Translation Fund, NS XX). Arabic text pp. 10–47 (preface pp. 10–14; poems I–X pp. 15–18); "Translation and Commentary" pp. 48–155 (I–X pp. 48–66). | standard ed. | yes: fully vowelled by the editor (the MSS are not vocalized, Nicholson p. 9) | public domain | archive.org tarjumanalashwaq00ibnaiala (University of California copy, 400 ppi, NOT_IN_COPYRIGHT). Downloaded locally (PDF, jp2 page images, djvu text). Scan leaf = printed page + 11 (.../page/n{leaf}); jp2/png file number = printed page + 12. |
| N | Nicholson's own MS, dated 1029 AH, poems (in red) + commentary; twice collated. Base of his text. | MS | no | — | via Nich |
| L | Leiden, Cod. 875(2) Warner, 992 AH: poems only, with the preface of the first recension. | MS | no | — | variants in Nich's apparatus |
| M | Leiden, Cod. 641 Warner, 984 AH: poems + commentary (fuller than N's, with excerpts from the Futūḥāt). | MS | no | — | variants in Nich's apparatus |
| — | Other copies of the scan: archive.org tarjumnalashwqco0000ibna, tarjumanalashwaq029354mbp, DLI in.ernet.dli.2015.213238 (same edition). | not used |
Not consulted (to do): the Beirut Dār Ṣādir edition of the Tarjumān (1961 and reprints) and of the Dhakhāʾir al-aʿlāq; M. ʿAbd al-Raḥmān al-Kurdī's Cairo edition of the Dhakhāʾir (1968); modern translations (Sells, Stations of Desire, 2000; Ruiz/Álvarez, etc.) are copyrighted: private comparison only.
2. Base and why
Nicholson 1911 is the only critical edition that is openly usable, it is based on a good twice-collated MS with the variants of two Leiden MSS, and it is fully vowelled. Its Arabic text is our base, transcribed from the page images; its English is published alongside as a public-domain layer.
3. Transcription and proofreading
- Transcribed bayt by bayt from the 400-ppi images: each line cut into enlarged right/middle/left thirds (scripts in the session scratchpad:
lines.py,stack.py,zoom.py), close-ups for doubtful marks. The djvu text layer is useless for the Arabic. - Pass 2: every line re-read against fresh crops (full re-read of I, II, VIII; targeted close-ups for the rest). 8 transcription corrections (listed in
tarjuman-base.json→proofreading), none to the print. An independent vowel-mark proofread is still recommended before review. - The print's conventions kept: final yāʾ undotted (ى), waṣla (ٱ), madda over alif before hamza (ٱلْعَزَآءُ), ي for hamza in plurals (سحايب, غداير, عقايل), sukūn on the rhyme letter in V, tanwīn on diptotes metri causa.
- Changes to the print: none. Doubtful readings (print clear, sense or grammar odd) are listed in
uncertain(10): II.10 عبّيت, II.12 شرّتها, V.1 ٱلغُرام (ḍamma in print; gharām expected), V rhyme sukūn, VII.1/IX.5 tanwīn on diptotes, VII.2 عَنَ ٱنوار, VII.8 ٱلشُّعْور, VIII.6 ٱلرُّوَيْدا, IX.2 hamza spellings, VII footnote 8 keying. - Nicholson's apparatus (variants of L and M against N) is recorded verse by verse: 32 entries in
apparatus.
4. Decisions (substantive)
None: we reproduce Nicholson's text. His own decisions (N preferred to LM) are visible in apparatus.
5. Quotations and allusions
Verified against the Tanzil Uthmani text (matched on the consonantal skeleton, verse words quoted in quotes[].verseText): 27:23 (Bilqīs's throne, II.2), 27:44 (the glass palace II.3; sāqayhā II.5; aslamat II.12), 19:57 (Idrīs, II.3), 3:49 (Jesus quickening the dead, II.4), 7:145 (the tablets of Moses, II.5), 75:26 (the soul at the collar-bones, II.11), 3:185 (zuḥziḥa, II.12, verbal echo), 22:18 (stock formula a-lam tara anna, VII.5). None is a verbatim quotation: all are allusions, so the text is left as printed. Poetry: IV.3 echoes Imruʾ al-Qays's Muʿallaqa ("the night let down its curtains"); hadith: X.2 perhaps "the believer is the mirror of the believer": both noted from knowledge, not yet linked to editions. scripts/quran_refs.py has not been run for this text (its TEXTS table would need a tarjuman entry; shared script, not edited).
6. Vowelling
- Source: the print (Nicholson).
scripts/recitals/prepare.py tarjuman(Claude opus, 3 workers) supplied marks only where the print leaves a letter bare: 436 AI marks (mask 'a'), letters verified unchanged in all 64 passages. - Two AI interventions against the print were found by checking every printed mark and reverted by hand in
tarjuman.json: VI.1 madda of ٱلْعَزَآءُ turned into plain alif by prepare.py's Mehren rule; V.1 ḍamma of ٱلْغُرامْ changed to fatḥa. After the fix, every printed mark is preserved.
7. Segmentation and word IDs
One passage per bayt, hemistichs separated by "\n" (the reader shows two-line verse). Passage n = 1…64 running through the poems; each passage carries poem, line (Nicholson's numbering) and label ("II.5"). sections in the base lists the poems. Word IDs j1:<passage>:<token>.
8. Analysis layer
scripts/fusus/analyze.py tarjuman.json tarjuman-analysis.json j1 (CAMeL Tools): 619 words, 384 with root, 246 distinct roots, 45 vowel conflicts.
9. Commentaries
Ibn ʿArabī's own commentary, Dhakhāʾir al-aʿlāq, is available here only in Nicholson's abridged English (data/sources/tarjuman-nicholson.json, per passage). The Arabic of the Dhakhāʾir is not yet on disk (a public-domain print: Beirut 1312 AH lithograph? to check on archive.org/HathiTrust).
10. Translation
- Published (public domain): Nicholson 1911, verse translation + abridged commentary, per passage with page references:
data/sources/tarjuman-nicholson.json(64 verses, 59 with commentary). Transcribed from the page images, re-read, and collated word by word with the archive.org OCR as an independent reading (residual differences: OCR noise only). - AI:
scripts/recitals/translate.py tarjuman opus→app/src/content/ai/tarjuman.json(64/64 passages, glosses and term notes), glossary intarjuman-meta.json(nasīb, aṭlāl, Niẓām, rīḥ al-ṣabā, place-names, pilgrimage rites, Christian vocabulary of II).
Context
Niẓām. In the preface to the first recension Ibn ʿArabī tells that on arriving in Mecca in 598/1202 he met the Persian scholar Makīn al-Dīn Abū Shujāʿ Zāhir b. Rustam al-Iṣfahānī and his aged sister Fakhr al-Nisāʾ, and Makīn al-Dīn's young daughter Niẓām, called ʿAyn al-Shams wa-l-Bahāʾ, famed for beauty, learning, asceticism and eloquence. He celebrated her in these poems in the language of love, and says that every name he mentions alludes to her, every abode he mourns is hers, and all of it signifies divine influences and spiritual realities (Nicholson pp. 3–4). The Dhakhāʾir. At Aleppo a theologian doubted that claim, suggesting the mystical preface was cover for erotic verse. At the request of Badr al-Ḥabashī and Ismāʿīl b. Sawdakīn, Ibn ʿArabī wrote the commentary, Dhakhāʾir al-aʿlāq, reading every image as a figure of divine knowledge; after hearing it the critic withdrew his doubt (Nicholson pp. 4–5). The second recension dates the poems to Rajab–Ramaḍān 611/1214; Dozy thought the author post-dated them; Nicholson rejects the charge (pp. 5–8).
12. Open questions
- Independent proofread of the vowel marks of III–VII and IX–X (pass 2 was partly targeted).
- Obtain a public-domain Arabic text of the Dhakhāʾir al-aʿlāq; align it with Nicholson's abridgement.
- Compare with the Dār Ṣādir text (private) and record differences.
- Link the Imruʾ al-Qays echo and the mirror hadith to editions; run
quran_refs.pyonce the text is registered. - Poems XI–LXI (Arabic pp. 18–47; English pp. 66–155) remain.
translate.pywrote an emptyapp/public/data/commentaries/tarjuman.json(no commentators); the Nicholson commentary is a separate public layer, not in that file.
13. Change log
- 2026-09-28: poems I–X established from Nicholson 1911 (64 abyāt); Nicholson's English layer; prepare, analysis (j1), AI translation.
Proofreading status (2026-09-28)
All ten poems have now had a full second reading against the page images (third pass): 31 corrections to our transcription (19 marks the print has that had been left out, 12 marks the print does not have), each logged in tarjuman-base.json → proofreading.correctionsInPass3. Seven readings remain undecided (see uncertain), notably VI.1 ٱلْعَزآءُ, VII.4 أَعْلام, VII.8 ٱلظُّلُماتِ and the fatḥa with the shadda in تَحِيّة (III.6, IV.2).