Mashrabaمشربة

About this text: The Wine Ode

الخمرية

Ibn al-Fāriḍ (576–632/1181–1235)

The wine drunk before the vine was created: 41 verses, with Nicholson's translation.

On the Platonic Current: al-Nābulusī · Ibn al-Fāriḍ

Synthesis of the readings

The arc

The ode opens with a drinking that took place, or whose drunkenness took place, before the vine was created. The commentators tie this to the souls' state before creation and to the Covenant of Alast. It then describes the wine through cosmic images (sun, full moon, crescent, stars) and notes that only a faint remnant of it survives among people. A long run of hyperbolic verses (5 to 20) follows, in which the wine's name, scent, seal, soil or strainer, short of any drinking, heals the sick, revives the dead, restores the senses, guides the lost and refines character. At verse 21 the poet accepts the request to describe the wine. In verses 22 to 30 he states its nature in metaphysical terms: qualities without elements, a tale older than creation, the ground on which things subsist, union with the spirit without indwelling, and a priority over Adam and over all time. The last part turns to the drinkers. The poet answers critics who call his drink sin, praises those intoxicated by aspiration alone, and claims a rapture that began early and outlasts the body. He urges the listener to drink it pure, or mixed only with the Beloved's saliva, in the tavern and to music. The ode closes by declaring the sober life no life and telling those without a share in the wine to weep for themselves.

Key moves

  1. The poem places the drinkers' intoxication before the creation of the vine, so from the start the wine belongs to an order prior to the created world.1
  2. The wine is pictured as a sun held in a moon-like cup and passed round by a crescent. Only a hidden remnant of it now survives, and in the end only its name.2, 3, 4, 6
  3. A long catalogue of wonders shows that the slightest contact with the wine is enough to intoxicate, heal, revive, guide and ennoble. Its mention, seal, scent, soil, strainer, name or covering each works without the wine being drunk.5, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20
  4. The poet accepts the request to describe the wine and claims knowledge of its qualities. He then lists qualities that belong to the elements while denying the elements themselves.21, 22
  5. The wine is set before all created things and all time. All things subsist through it, it is veiled by a divine wisdom, and it precedes Adam's era.23, 24, 29, 30
  6. The poet describes the relation of spirit and wine as a union that is not bodily mingling. Wine and vine, spirits and bodily forms, are two sides of what is at root one.25, 26, 27, 28
  7. The wine's own beauties guide its praisers. It moves even those who have not tasted it, and the poet turns the charge of sin back on his accusers by calling abandonment of the wine the real sin.31, 32, 33, 34
  8. The poet claims a lasting rapture, urges the listener to drink pure wine in the tavern to music, and promises mastery over Time. He ends by judging the sober life lost and bidding those without a share to weep.35, 36, 37, 38, 39, 40, 41

Where the sources agree

  • The wine is a spiritual wine. The sources name it variously as divine love, gnosis (maʿrifa) or a divine reality, and the vocabulary of wine poetry is taken in a Sufi technical sense.

    AI translational-Būrīnīal-Nābulusī

  • The wine precedes creation and time. It is placed before the vine, before created beings, before Adam and before the whole span of time.

    AI translationNicholsonal-Būrīnīal-Nābulusī

  • The wine acts through the slightest contact, without drinking. Its mention, seal, scent, name or covering is enough to intoxicate or transform.

    AI translationNicholsonal-Būrīnīal-Nābulusī

  • The wine is non-material. It has the qualities of the elements without their substance, and its union with the spirit is not one body permeating another, which guards against the charge of indwelling (ḥulūl).

    AI translationNicholsonal-Būrīnīal-Nābulusī

  • Knowledge and praise of the wine start from the wine itself. Its fragrance and radiance guide the seeker, and its beauties lead the praisers to describe it.

    AI translationNicholsonal-Būrīnīal-Nābulusī

  • A life without the wine is a lack. The sober person does not truly live, and failing to die drunk with it is a failure of sound judgment.

    AI translationNicholsonal-Būrīnīal-Nābulusī

Live divergences

  • How far the poem's images are to be read allegorically.

    • al-Būrīnī He often keeps to the literal and rhetorical level. He glosses the tavern as a wine shop, the healings as wonders (karāmāt) of the wine, and the strainer as an actual strainer, with attention to grammar and wordplay.3, 11, 14, 15, 16, 17, 37
    • al-Nābulusī He gives nearly every image a spiritual key. The tavern is the divine presences or all existing things, the strainer is the perfect human's intellect, the banner is the Sufi path, and the soil is the guiding shaykh.3, 11, 14, 15, 16, 17, 37
    • AI translation Its notes read the images as symbols of divine love and the Sufi path, such as the company of gnostics or the heart. They are usually less tied to a doctrine of the shaykh than al-Nābulusī's readings.3, 11, 14, 15, 37

  • The role of the perfect human being or spiritual guide in how the wine reaches others.

    • al-Nābulusī He repeatedly locates the wine's transmission in the perfect human (al-insān al-kāmil) or shaykh. This figure appears as the cup, the shade, the strainer, the charmer, the soil of the wine's land and the hand the disciple takes at the oath of allegiance.2, 10, 13, 14, 15, 16
    • AI translation It sometimes names masters as mediators, but it often has the wine act directly, as when divine breaths reach across any distance with no human intermediary.12, 13, 14
    • al-Būrīnī He speaks of the gnostic and his disciples passing on gnosis in the cup verse, but elsewhere he treats the wine's effects without a mediating figure.2, 13, 16

  • Whether the wine's hiddenness and vanishing come from human heedlessness or from the wine itself.

    • al-Nābulusī The wine seems faint and withdrawn because hearts are taken up with worldly adornment and lusts.4, 6
    • al-Būrīnī He reads khafāʾ as manifestness, with the wine hidden through excess of manifestness. He treats the vanishing as a general decline of gnoses.4, 6
    • AI translation The remnant reflects the wine's pre-eternity and its withdrawal beyond temporal perception, a return upward to its source.4, 6

  • Whether the wine's priority is temporal or a matter of rank and being.

    • AI translation It renders the precedence in temporal terms, as the wine coming before creation in the pre-eternal past and before any created vessel, while allowing an ontological sense.23, 26, 29
    • al-Nābulusī The precedence is essential and not temporal, since time is itself created. The two halves of the wine-and-vine verse are two attributions of one existence, not stages.23, 26, 29

  • What dying drunk and the lasting rapture refer to.

    • al-Būrīnī He includes actual physical death and reports that Ibn al-Fāriḍ himself died in longing. He places the rapture's start in early childhood, from the cradle to the grave.34, 35, 40
    • al-Nābulusī Dying drunk is a condition within life: spending every moment witnessing the True Existence until nothing else is perceived.40
    • AI translation Dying is passing away from the self (fanāʾ) in the Beloved. The rapture reaches back before bodily existence to the primordial covenant.35, 40

Left open

  • The wonders of healing, revival and restored senses may be pure poetic hyperbole, or they may make claims about the real powers of saints. Al-Nābulusī allows actual revival by God's leave, and the sources do not settle the matter.
  • It is unclear whether intoxication through mention or aspiration alone, without drinking, is a stage on the way to full attainment or a lasting state of its own.
  • The sources leave open how the wine relates to God. They do not say whether it is identical with a divine attribute or presence, or only its manifestation or subject.
  • It is unclear whether the ode's many hints at the poet's own condition, such as his dying in longing, are meant as autobiography or as a general teaching for the wayfarer.

AI synthesis (Claude), built from the syntheses of each passage and the commentators' introductions to the chapter; every point names its sources. Public sources only. Each passage's own synthesis is under its translation in the reader.

In the reader

41 verses · ~5 min · audio · commentaries · published translation · 1 Quran quotation

462 words

  • Audio: Arabic and English
  • Commentaries: aligned on 41 passages
  • Published translation: Nicholson (1921)
  • Quran quotations: 1
Read it →Grammar practiceStudy sheet (print)

Difficulty

some rare vocabulary

Lexical difficulty, from the words themselves: how many roots fall outside the 1,000 most frequent in the Quran, how long the words are, and how much of the vowelling had to be supplied by AI. It says how unfamiliar the vocabulary is, not how deep the ideas are.

Rare roots 18% · mean word length 3.9 letters · AI vowelling on 36% of words

Themes & resonances

What the text is about, and what it connects to across the history of thought. Every connection says what kind of claim it is: a documented lineage, a shared source, a resonance with no influence claimed, or a contrast. Mapped by AI; references are shown only once checked. Across the library

Themes

  • The wine drunk before the vine, and the drunkenness it givesسُكْرsukrroot سكر
  • Remembrance and the power of the mere nameذِكْرdhikrroot ذكر
  • A wine older than timeقِدَمqidamroot قدم
  • Mingling without indwellingاِمْتِزاجimtizājroot مزج
  • Spirit as wine, body as vineرُوحrūḥroot روح
  • Description by negation: subtlety without substanceلُطْفluṭfroot لطف
  • The written name that healsرَسْمrasmroot رسم
  • The wine's miracles of healingأَعْمىaʿmāroot عمي
  • The Beloved and the named belovedحَبيبḥabībroot حبب
What each theme is here
  • The wine drunk before the vine, and the drunkenness it gives. The ode opens with a drunkenness that comes before the vine was created (verse 1) and closes by calling the sober life no life at all (verse 40).
  • Remembrance and the power of the mere name. The wine is drunk 'in remembrance of the Beloved', and merely mentioning it wakes the encampment drunk and makes the mute speak (verses 1, 5, 11, 32).
  • A wine older than time. The wine's tale went before all created beings, it has no 'before', and its pressing came before the age of time and before Adam (verses 23, 29-30, 35).
  • Mingling without indwelling. The poet's spirit and the wine 'mingled in union' while no body entered another body, and in the end 'the whole is one' (verses 25, 28).
  • Spirit as wine, body as vine. The ode maps spirit and body onto wine and vine, and the wine revives the body of a dead man by returning his spirit (verses 9, 22, 28).
  • Description by negation: subtlety without substance. The wine is described in paired negations: clarity yet no water, subtlety yet no air, light yet no fire, spirit yet no body (verse 22).
  • The written name that heals. Only a name of the wine remains (verse 6), yet its letters traced on a madman's brow heal him and written on a banner make an army drunk (verses 16-17).
  • The wine's miracles of healing. A long catalogue (verses 9-16) gives the wine the power to raise the dead and cure the sick, the lame, the mute, the blind, the deaf and the envenomed.
  • The Beloved and the named beloved. The Beloved frames the ode, and the lover of Nuʿm, pining when her name is spoken, is the image for one moved by a wine he never tasted (verses 1, 32, 36).

Connections

Lineage 3

Documented transmission: a chain of reading, teaching, translation or citation, grounded in a verified reference or in the later text itself.

Resonance, not influence 9

Resonance, not influence: a structural or thematic likeness. No influence is claimed.

  • Ibn ʿArabī, The Interpreter of DesiresSufism · probable

    Two contemporaries wrote love poems in the classical Arabic manner that were read as mystical. Ibn ʿArabī calls out a roll of beloveds' names (Daʿd, Hind, Salmā) much as Ibn al-Fāriḍ invokes Nuʿm, and both make the Beloved present in the heart when named.

    Carried by“the beloved's name”

    In the library:Tarjumān al-ashwāq ¶26Tarjumān al-ashwāq ¶33Path: Love: the soul's longing back

  • Ibn Sīnā, The Ode on the Soul: the covenants of the sanctuaryFalsafa (Arabic philosophy) · probable

    Both poems give the soul a life before the body. Ibn Sīnā's dove has forgotten the covenants of the sanctuary and weeps for them, while Ibn al-Fāriḍ's drinker keeps an intoxication from before his coming-to-be that the decay of his bones cannot touch (verse 35).

    Carried by“pre-existence of the soul”

    In the library:al-ʿAyniyya ¶5al-ʿAyniyya ¶8Path: The fall and the return of the soul

  • Plotinus, Ennead VI.7.35: intellect 'drunk with nectar'Neoplatonism · probable

    Plotinus says that intellect in love with the Good becomes drunk with nectar, and that such drunkenness is better than solemnity. The ode likewise prefers drunkenness to sober life, with no known path between the two texts.

    Carried byμεθύωmethyō“to be drunk”

    In the library:Ennead VI.9, ch. 9

  • Philo of Alexandria: 'sober drunkenness' (methē nēphalios)Hebrew Bible and Judaism · probable

    Philo uses the paradox of a sober drunkenness for the soul possessed by God, a phrase later taken up by the Church Fathers. The ode's wine that brings no sin and refines character (verses 5, 18-19) sits in the same paradox by an independent route.

    Carried byμέθη νηφάλιοςmethē nēphalios“sober drunkenness”

  • Plato, Symposium 211e: the beautiful itself, 'pure, clean, unmixed'Platonism · suggestive

    Diotima's final object is beauty unmixed with flesh and colour. The ode urges the reader to take the wine pure and unmixed (ṣirf, verse 36) and gives it beauties that make those who praise it write beautifully (verse 31).

    Carried byصِرْفṣirfroot صرفἄμεικτοςameiktos“unmixed”

    In the library:Symposium 211e

  • al-Ghazālī, The Niche of Lights: the oil that shines 'though no fire touched it'Sufism · probable

    Al-Ghazālī expounds the Light Verse, where the oil almost shines untouched by fire, and the ode calls its wine 'light, yet no fire' (verse 22). Both describe a light with no earthly source.

    Carried byنورnūrroot نور“light without fire”

    In the library:Mishkāt al-anwār ¶36Path: Light and its veils

  • Ḥāfiẓ, Dīvān: the wine and the tavernPersian poetry · probable

    Ḥāfiẓ's ghazals also keep tavern, cup-bearer and wine in a register that readers take as both worldly and mystical. Both poets scorn the pious critic who calls the wine a sin (compare verse 33).

    Carried by“tavern (kharābāt, ḥān)”

  • Charles Baudelaire, 'Enivrez-vous' (Le Spleen de Paris)Modern literature · suggestive

    Baudelaire urges one always to be drunk, on wine, poetry or virtue, to escape the burden of Time. The ode's claim that there is no life for one who lives sober (verse 40) is an unrelated echo.

    Carried by“be drunk”

  • Li Bai, 'Drinking Alone by Moonlight'Daoism and Chinese thought · suggestive

    Li Bai drinks with the moon and his shadow as companions, while Ibn al-Fāriḍ makes the full moon the cup and the wine a sun passed round by a crescent (verse 2). Both place drinking among the heavenly bodies, with no connection between them.

    Carried byبَدْرbadrroot بدر“moon as companion or cup”

Contrast 2

An instructive opposite: the difference teaches something about both.

  • al-Ghazālī, The Niche of Lights: the drunken sayings of the gnosticsSufism · probable

    Al-Ghazālī treats the ecstatic sayings of the drunk as speech to be concealed and corrected once reason returns, and calls their union a union by metaphor, while Ibn al-Fāriḍ declares that there is no life for one who lives sober (verse 40). The two agree on one point, since the ode also denies that any body entered another (verse 25).

    Carried byسُكْرsukrroot سكر“intoxication and sobriety”

    In the library:Mishkāt al-anwār ¶52Mishkāt al-anwār ¶53Mishkāt al-anwār ¶54Path: Unification: the soul at peace

  • Edward FitzGerald, Rubáiyát of Omar Khayyám (1859)Modern literature · probable

    In FitzGerald's quatrains wine is a solace against time and the grave. Ibn al-Fāriḍ's wine precedes time, outlasts the decayed bone (verse 35), and turns Time into an obedient slave (verse 39).

    Carried byدَهْرdahrroot دهر“Time”

Claude (opus) via Claude Code, headless, no web access, 2026-09-28. A reference with a check was found and checked with a web search.

Versions

20 layers, each with its licence

The original

Sharḥ Dīwān Ibn al-Fāriḍ (Cairo 1289/1872)Rushayd ibn Ghālib al-Daḥdāḥ · 1289/1872Public domain
By
Rushayd ibn Ghālib al-Daḥdāḥ (compiler) · 1289/1872
Source:
Sharḥ Dīwān Ibn al-Fāriḍ: the commentaries of Ḥasan al-Būrīnī and ʿAbd al-Ghanī al-Nābulusī, compiled by Rushayd ibn Ghālib al-Daḥdāḥ (Cairo: al-Maṭbaʿa al-Miṣriyya, 1289/1872), part II, printed pp. 174–196 (PDF pp. 179–201), Bibliotheca Alexandrina scan · link ↗
In the reader:
In full: 41 verses, one per passage (= Nicholson's numbers), the hemistichs on two lines
Licence:
Public domain (1872)
Notes
Each verse printed in bold, fully vowelled, in round brackets, followed by the commentary. The compositor's margin carries a few notes on readings (vv. 7, 8, 11), kept as marginalia. Not used: the Toronto manuscript Qaṣāʾid ʿUmar ibn al-Fāriḍ (archive.org qasaidumaribnalf00unse; not read yet) and other Cairo prints (1306/1889 AAlexandrina-140352; 1888 AAlexandrina-209437; 1862 AAlexandrina-187876).

Our edition

Our transcription of the versesUNVERIFIED · 2026-09-28Our work
By
UNVERIFIED (transcriber (two passes; the verse lines located by an automatic line detector, then read)) · 2026-09-28
Source:
data/sources/khamriyya-base.json; text record §§3–6
In the reader:
In full, with the print's vowels
Licence:
Ours
Notes
Vowel marks read from 300 ppi page images (1,295 marks transcribed from the print); a person should spot-check them against the images. No reading of the print changed; the undotted final yāʾ read as ي; hamza seats as printed where legible. 4 doubtful readings listed (v. 11 first letter, v. 13 ضلّ/ظلّ, v. 18 بها لطريق). No Quran quotation. Verse count and order agree with Nicholson's translation; no second Arabic witness collated yet.
UNVERIFIED:
The text record does not say whether the verses and their vowels were read from the images by a person or by AI.

Published translations

Nicholson's translation (1921)R. A. Nicholson · 1921PD-US
By
R. A. Nicholson (translator) · 1921
Source:
R. A. Nicholson, Studies in Islamic Mysticism (Cambridge: University Press, 1921), ch. III, pp. 184–188: a complete "literal translation" of the 41 verses, with notes drawn from al-Nābulusī; Toronto copy; data/sources/khamriyya-nicholson.json · link ↗
In the reader:
Under each verse, with a link to its page
Licence:
Public domain in the US (1921; archive.org marks it NOT_IN_COPYRIGHT)
Notes
OCR of the Toronto scan proofread against the page images; Nicholson's footnotes and footnote numerals omitted. He cites the ode as "Dīwān, p. 472 foll." (another printing of the same commentary compilation), and notes after Nallino that vv. 23–30 are wanting in al-Būrīnī's commentary.
Arberry's translationArberry · 1952/56Copyrighted: cited only
By
Arberry (translator) · 1952/56
Source:
Arberry's translation of the Khamriyya (1952/56) (UNVERIFIED)
In the reader:
Not shown
Licence:
In copyright; not used
UNVERIFIED:
The title and publisher of Arberry's translation are not recorded in the repo.
Homerin's translationHomerin · UNVERIFIEDCopyrighted: cited only
By
Homerin (translator) · UNVERIFIED
Source:
Homerin's translation of the Khamriyya
In the reader:
Not shown
Licence:
In copyright; not used
UNVERIFIED:
The full name, title, publisher and date of Homerin's translation are not recorded in the repo.

AI layers

Vowelling (marks the transcription did not carry)Claude (AI) · 2026-09-28Our work
By
Claude (AI) (vowelling) · 2026-09-28
In the reader:
The print's marks are shown as the print's; the few AI marks in amber
Licence:
Ours; marked AI
Made by:
Claude (Opus; exact version not recorded), 2026-09-28 (first commit of app/src/data/khamriyya.json (scripts/recitals/prepare.py khamriyya: headless Claude Code, Opus; letters verified unchanged))
Notes
The verse lines are fully vowelled in the print. prepare.py (letters verified unchanged) kept the 1,295 marks transcribed from the print and added 194 AI marks, mostly the article's sukūn and long-vowel fatḥas; 17 ambiguous words listed.
English translation, passage by passageClaude (AI) · 2026-09-28Our work
By
Claude (AI) (translator) · 2026-09-28
In the reader:
Under each passage
Licence:
Ours; marked AI
Made by:
Claude (Opus; exact version not recorded), 2026-09-28 (first commit of app/src/content/ai/khamriyya.json (scripts/recitals/translate.py khamriyya opus: headless Claude Code, minimal context))
Notes
Written for this reader from the Arabic (our vowelled text), with the glossary in data/sources/khamriyya-meta.json; the model was told not to reproduce or imitate published translations (Corbin/Trask, Goichon, Heath, Thackston, Azal, Hussain, Beneito & Hirtenstein, Austin, Nicholson, Chittick) and was given none. Shown alongside Nicholson's public-domain translation.
Word glossesClaude (AI) · 2026-09-28Our work
By
Claude (AI) (glosses) · 2026-09-28
In the reader:
One gloss per word, in the sense the passage gives it (word panel, and under the words on request)
Licence:
Ours; marked AI
Made by:
Claude (Opus; exact version not recorded), 2026-09-28 (first commit of app/src/content/ai/khamriyya.json (scripts/recitals/translate.py khamriyya opus: headless Claude Code, minimal context))
Notes
Contextual glosses (1–5 words) for every word id, from the same run as the translation.
Term notesClaude (AI) · 2026-09-28Our work
By
Claude (AI) (notes) · 2026-09-28
In the reader:
Up to six technical terms per passage, each with a one-sentence note, on the word
Licence:
Ours; marked AI
Made by:
Claude (Opus; exact version not recorded), 2026-09-28 (first commit of app/src/content/ai/khamriyya.json (scripts/recitals/translate.py khamriyya opus: headless Claude Code, minimal context))
Notes
From the same run as the translation, with the glossary in data/sources/khamriyya-meta.json.
The commentary text, settled by AIClaude (AI) · 2026-09-28Our work
By
Claude (AI) (decided each disagreement between the two OCRs from the page image) · 2026-09-28
In the reader:
In the commentary text; the AI readings are recorded with both OCR readings
Licence:
Ours; marked AI
Made by:
Claude (Opus; exact version not recorded), 2026-09-28 (first commit of app/public/data/commentaries/khamriyya.json (headless Claude Code, Opus))
Notes
Two OCRs (macOS Vision on 300 dpi page images; archive.org's own OCR) collated letter by letter: 30.4% agreement (3,051 of 10,029 words). The 2,058 disagreements were decided page by page by Claude (Opus, headless, no tools or context) reading the framed page image: Vision chosen 864, archive.org 17, neither (read from the scan) 1,177. Agreed readings were not re-checked; some OCR slips remain. A person should proofread against the scan, starting with the 1,177. Scripts in scripts/khamriyya/; decisions in data/sources/khamriyya-commentaries/decisions/.
English of the commentariesClaude (AI) · 2026-09-28Our work
By
Claude (AI) (translator) · 2026-09-28
In the reader:
Under each comment
Licence:
Ours; marked AI
Made by:
Claude (Opus; exact version not recorded), 2026-09-28 (first commit of app/public/data/commentaries/khamriyya.json (scripts/khamriyya/translate.py))
Notes
Faithful and plain, from the collated text; technical terms transliterated with a gloss the first time in each comment; unreadable places marked […] (20 of the 217 comments). 217/217 translated.
Synthesis of the readingsClaude (AI) · 2026-09-28Our work
By
Claude (AI) (synthesis) · 2026-09-28
In the reader:
On the About page and per verse
Licence:
Ours; marked AI
Made by:
Claude (Opus; exact version not recorded), 2026-09-28 (generatedOn of content/synthesis/khamriyya.json (Claude Code))
Notes
From public sources only: our AI translation, Nicholson, al-Būrīnī and al-Nābulusī (the commentaries read with our AI English). No copyrighted translation was given to the model. Rerun with the two commentators on 2026-09-28.
Resonances: themes and connectionsClaude (AI) · 2026-09-28Our work
By
Claude (AI) (map of themes and connections, by docs/RESONANCES-METHOD.md) · 2026-09-28
In the reader:
For the whole text: themes and 14 connections, on the About page
Licence:
Ours; marked AI
Made by:
Claude (Opus; exact version not recorded), 2026-09-28 (generatedOn of content/resonances/khamriyya.json (Claude Code, headless, no web access))
Notes
Every connection states its kind; parallels are resonances, not claims of influence. External references were proposed by the model without web access and are shown only after an agent verified them with a web search.

Automatic analysis

Word analysis (CAMeL Tools)CAMeL Tools 1.6 (NYU Abu Dhabi) · 2026-09-28Our work
By
CAMeL Tools 1.6 (NYU Abu Dhabi) (analyser: calima-msa-r13 + MLE, constrained by the text's vowelling) · 2026-09-28
Source:
scripts/fusus/analyze.py → data/sources/khamriyya-analysis.json · link ↗
In the reader:
Lemma, root, part of speech, clitics and pattern, in the word panel
Licence:
Our analysis layer, made offline with CAMeL Tools (code MIT; calima-msa-r13 database GPL v2, not redistributed)
Notes
462 words, 280 with a root, 21 vowel conflicts with the vowelling. Automatic and non-generative; date = first commit of the analysis file.
Quran quotations found automaticallyscripts/quran_refs.py · 2026-09-28Our work
By
scripts/quran_refs.py (skeleton matcher against the Tanzil Uthmani text) · 2026-09-28
Source:
app/src/data/quran-refs/khamriyya.json
In the reader:
1 matches, on the words they cover
Licence:
Ours
Notes
Quotations verified when the text was established, plus an automatic match on the consonantal skeleton (vowels, Quranic marks, alif spellings and hamza seats ignored): a run of four or more words, or three words of at least 12 letters. No AI. v. 14 «تسمع الصم» (cf. 27:80, 30:52, 43:40, 10:42, inverted), recorded as an allusion.
Metre scansionscripts/metre/arud.py · 2026-10-03Our work
By
scripts/metre/arud.py (rule-based scanner (al-Khalīl's ʿarūḍ, as set out by Ibn Jinnī)) · 2026-10-03
Source:
app/src/content/metre/khamriyya.json (scripts/metre/build.py)
In the reader:
The metre line under each verse: al-ṭawīl, 40 of 41 verses scanned
Licence:
Ours
Notes
Reads the vowelled text (the print's, AI only where a mark was not legible); never guesses: a verse that does not fit is shown as 'not scanned'. v. 35 fits al-ṭawīl if ʿindī is read ʿindiya (a common licence), with a note written by hand. Date = first commit of the metre file.

Commentaries

al-Būrīnī's commentaryḤasan al-Būrīnī (963/1556 – 1024/1615) · d. 1024/1615; printed 1289/1872Public domain
By
Ḥasan al-Būrīnī (963/1556 – 1024/1615) (commentator) · d. 1024/1615; printed 1289/1872
Source:
Ḥasan al-Būrīnī, al-Baḥr al-fāʾiḍ fī sharḥ Dīwān Ibn al-Fāriḍ, in Sharḥ Dīwān Ibn al-Fāriḍ (Cairo 1289/1872), part II pp. 174–198 (PDF 179–203) · link ↗
In the reader:
70 comments, each anchored to the verse words it glosses, with AI English and a link to its page
Licence:
Public domain (the 1872 print)
Notes
Lexicon, iʿrāb, rhetoric (badīʿ), then the sense. Wanting for vv. 23–30 (Nallino, Nicholson), confirmed on the scan. vv. 19, 35, 38: his comment only. He reports readings at v. 13 (ضلّ/ظلّ) and v. 38 (rejects النِّعَم as a corruption). His comment on v. 41 is anchored at one late word. The text is our OCR collation settled by AI (see ai-commentary-readings); not proofread by a person.
al-Nābulusī's commentaryʿAbd al-Ghanī al-Nābulusī (1050/1641 – 1143/1731) · d. 1143/1731; printed 1289/1872Public domain
By
ʿAbd al-Ghanī al-Nābulusī (1050/1641 – 1143/1731) (commentator) · d. 1143/1731; printed 1289/1872
Source:
ʿAbd al-Ghanī al-Nābulusī, Kashf al-sirr al-ghāmiḍ fī sharḥ Dīwān Ibn al-Fāriḍ (his notes, marked (ن)), in Sharḥ Dīwān Ibn al-Fāriḍ (Cairo 1289/1872), part II pp. 174–198 (PDF 179–203) · link ↗
In the reader:
147 comments, each anchored to the verse words it glosses, with AI English and a link to its page
Licence:
Public domain (the 1872 print)
Notes
For vv. 23–30 the print gives one unsigned commentary in his manner (no siglum); attributed to him and flagged 'unsigned' on each card. The margin's note at v. 8 shows his copy read مِن دَنِّها ('from its jar') where the verse has مِن دُونِها. Vv. 25–26 and 40–41 are printed as pairs with one commentary. Not proofread by a person.

Audio

Arabic read aloud (machine voice)ar-SA-HamedNeural (Azure AI Speech) · 2026-09-28Our work
By
ar-SA-HamedNeural (Azure AI Speech) (machine voice) · 2026-09-28
In the reader:
Each passage, word by word highlighted
Licence:
Ours; generated on Azure's paid tier, whose terms allow the audio output to be published
Made by:
ar-SA-HamedNeural (Azure AI Speech), 2026-09-28 (first commit of app/src/data/khamriyya-audio.json)
Notes
Machine voice reading the vowelled text (including the AI vowels); a reading aid, not a recitation.
English read aloud (machine voice)en-GB-RyanNeural (Azure AI Speech) · 2026-09-28Our work
By
en-GB-RyanNeural (Azure AI Speech) (machine voice) · 2026-09-28
In the reader:
Each passage, word by word highlighted
Licence:
Ours; generated on Azure's paid tier, whose terms allow the audio output to be published
Made by:
en-GB-RyanNeural (Azure AI Speech), 2026-09-28 (first commit of app/src/data/khamriyya-audio-en.json)
Notes
Machine voice reading the AI English translation; a reading aid.

✦ marks AI that no person has reviewed yet; ✧ marks AI reviewed by a named person. Copyrighted works are cited, never shown. All sources · How texts are established

The text record

Our working notes on this edition: status, witnesses, the decisions taken and why, and what AI supplied. This is a working edition, not an authority.Show the notes
Work / author / datesal-Khamriyya (الخمرية), the Wine Ode, incipit «شربنا على ذكر الحبيب مدامة» · Sharaf al-Dīn ʿUmar ibn al-Fāriḍ (576–632 AH / 1181–1235 CE) · metre ṭawīl, rhyme -mu
Part addedWhole poem, 41 verses
Reader route/read/khamriyya · about and apparatus /read/khamriyya/about · commentaries side by side /read/khamriyya/commentaries
Statusworking edition established 28 Sept 2026 from one PD print (steps 2–9, 11) + a PD English translation layer + the classical commentary layer (al-Būrīnī, al-Nābulusī) · reviewed by: —

The ode praises a wine drunk "before the vine was created": divine love, prior to all things, whose mere name, seal, scent or memory heals, guides and intoxicates; it is described by negations ("pure, but not water…"), is one with the poet's spirit, and ends by pitying whoever dies sober.

1. Witnesses

SiglumWitnessKindVowelled?LicenceLocation
DSharḥ Dīwān Ibn al-Fāriḍ: the commentaries of Ḥasan al-Būrīnī (d. 1024/1615) and ʿAbd al-Ghanī al-Nābulusī (d. 1143/1731, his notes marked (ن)), compiled by Rushayd ibn Ghālib al-Daḥdāḥ; Cairo: al-Maṭbaʿa al-Miṣriyya, 1289/1872, part II pp. 174–196 (the commentary on v. 41 runs on to p. 198). Verses in bold, fully vowelled, in brackets, each followed by the commentary; the compositor's margin carries a few notes on readings.standard commentary printyes (verse lines)PD (1872)archive.org AAlexandrina-141241, PDF pp. 179–201 (commentary to PDF p. 203)
NR. A. Nicholson, Studies in Islamic Mysticism, Cambridge: University Press, 1921, ch. III, pp. 184–188: complete "literal translation" (41 verses, numbered as in D) with notes drawn from al-Nābulusī. Nicholson cites the ode as "Dīwān, p. 472 foll." (another printing of the same commentary compilation).translation—PD in the US (1921; archive.org marks it NOT_IN_COPYRIGHT)archive.org studiesinislamyst00nichuoft (Toronto); = the owner's private copy (identical OCR)
TQaṣāʾid ʿUmar ibn al-Fāriḍ, manuscript, Thomas Fisher Library, Toronto (archive.org qasaidumaribnalf00unse, 82 pp.).MSpartlyscan freely availablenot read yet (see §12)
—Other prints on archive.org not used: Cairo 1306/1889 (al-ʿĀmira al-Sharafiyya, same compilation, AAlexandrina-140352), Cairo 1888 (AAlexandrina-209437), the 1862 print AAlexandrina-187876; Nābulusī's Kashf al-sirr al-ghāmiḍ (modern uploads).

2. Base and why

D: public domain, fully vowelled in the verse lines, and it carries the two classical commentaries (Būrīnī, Nābulusī) that the Arabic tradition reads the ode with, and that Nicholson follows. Nicholson notes (after Nallino) that vv. 23–30 are wanting in al-Būrīnī's commentary; D prints them, with al-Nābulusī's comments (confirmed on the scan: for vv. 23–30 D gives one unsigned commentary in al-Nābulusī's manner, without «(ن)»; see khamriyya-commentaries.md §5).

3. Collation

Against N: same 41 verses in the same order (verse-for-verse alignment checked). No second Arabic witness collated yet.

4. Decisions

No reading of D changed. Mechanical: the undotted final yāʾ read as ي. Marginal notes of the print, kept in khamriyya-base.json → marginalia: v. 7 (the commentary's "joys and sorrows" vs the verse's الهم), v. 8 (reread 28 Sept: the note is «قوله وقوله من دنها …»: al-Nābulusī's copy read مِن دَنِّها "from its jar", where the verse has مِن دُونِها), v. 11 (the commentator's copy read وتنطق, the verse has وينطق). The commentators also report readings: v. 13 ضلّ/ظلّ, v. 38 النِّعَم (rejected by al-Būrīnī as a corruption) for النَّغَم. Doubtful readings in uncertain: v. 11 first letter, v. 13 ضلّ/ظلّ (both reported by the commentary), v. 18 بها لطريق (joined in print).

5. Quotations

No Quran quotation. The skeleton matcher finds one allusion, v. 14 «تسمع الصم» (cf. 27:80, 30:52, 43:40, 10:42: «تُسْمِعُ ٱلصُّمَّ», inverted), recorded as quran-allusion, left as printed.

6. Vowelling

From the print (verse lines fully vowelled), read from 300 ppi page images; AI supplies only marks not transcribed (prepare.py, letters verified unchanged; vowelMask 't' = transcribed from the print, 'a' = AI). Caveat: at this resolution small marks can be misread; a person should spot-check the vowelling against the images (image crops of each word, kept locally).

7. Segmentation and word IDs

One verse per passage (1–41 = Nicholson's numbers), hemistichs separated by a line break. Word IDs w1:<verse>:<token>.

8. Analysis layer

CAMeL Tools via scripts/fusus/analyze.py khamriyya.json khamriyya-analysis.json w1.

9. Commentaries

Added 28 Sept 2026: the commentary layer, al-Būrīnī and al-Nābulusī from D (printed pp. 174–198, PDF 179–203). Full record: docs/texts/khamriyya-commentaries.md.

  • Method: two OCRs (our macOS Vision OCR; archive.org's own OCR of the item) collated letter by letter. Agreement between the two OCRs: 30.4% (3,051 of 10,029 words), because the archive.org OCR is weak. The 2,058 disagreements were decided by Claude reading the page image (Vision 864 · archive.org 17 · read from the scan 1,177), and every decision is recorded as an AI reading.
  • 217 comments on all 41 verses (al-Būrīnī 70, al-Nābulusī 147), each anchored to the verse words it glosses, with AI English (marked AI) and a link to its archive.org page.
  • vv. 23–30: al-Būrīnī wanting (Nallino, Nicholson), confirmed; the print's unsigned commentary there is al-Nābulusī's. vv. 19, 35, 38: al-Būrīnī only. vv. 25–26 and 40–41 are printed as pairs with one commentary.
  • The synthesis (scripts/synthesis.py khamriyya) now reads both commentators.

10. Translation

  • Public translation layer: data/sources/khamriyya-nicholson.json = {verse: {"translation", "page"}} (Nicholson 1921, pp. 184–188; OCR of the Toronto scan proofread against the page images; footnotes and footnote numerals omitted; "(n)" OCR error → (11), "itsfiddm" → "its fidám"). Page link: https://archive.org/details/studiesinislamyst00nichuoft/page/<page>/mode/1up.
  • AI translation (translate.py khamriyya opus), marked AI, alongside.

11. Rights

D and N: public domain (N: US). Arberry's (1952/56) and Homerin's translations are ©, not used.

12. Open questions / to review

  • Spot-check the vowelling (above).
  • Collate against a second Arabic witness (T, the Toronto MS; or the 1889 Cairo reprint of D).
  • Proofread the commentary layer against the scan (see khamriyya-commentaries.md §9).

13. Change log

  • 2026-09-28: record created; D transcribed from page images (two passes; verse lines located by an automatic line detector, then read), khamriyya-base.json, khamriyya-meta.json, khamriyya-nicholson.json; prepare.py (41/41; 1,295 marks transcribed from the print kept, 194 AI marks added, mostly the article's sukūn and long-vowel fatḥas; 17 ambiguous words), analyze.py (462 words, 280 with root, 21 vowel conflicts), translate.py … opus (41/41).
  • 2026-09-28: commentary layer (al-Būrīnī, al-Nābulusī) added, see khamriyya-commentaries.md; v. 8 marginal note reread (من دنها); synthesis rerun with the two commentators.