| Work / author / dates | Kitāb khatm al-awliyāʾ (كتاب ختم الأولياء), "The Seal of the Saints"; called Sīrat al-awliyāʾ, "The Life of the Friends of God", by Radtke, who holds that to be its proper title and «khatm al-awliyāʾ» a phrase taken from inside it (§0). Abū ʿAbd Allāh Muḥammad b. ʿAlī al-Ḥakīm al-Tirmidhī, of Tirmidh on the Oxus, d. c. 295–300/c. 905–910 (dates differ; the file's own catalogue line says «after 318»). |
| Part added | The first third (the session's budget rule: the whole book is about 23,900 words): the author's preface and chapters 1–8 of the edition, 70 passages, 7,099 words, OpenITI lines 66–765, pp. 5–31 of the 1999 print. Chapter 4 holds the whole questionnaire (157 questions). Chapters 9–29 (about 16,700 words, among them the later Seal chapters 13 and 25) are queued (§12). |
| Reader route | /read/khatm-awliya · about page /read/khatm-awliya/about |
| Kind | Sufism (sufism), first in date of the group (before al-Qushayrī) |
| Status | One digital text of a modern popular edition, not collated · AI-vowelled throughout (the edition is bare) · AI English ✦ unreviewed · CAMeL word analysis · not yet checked by a person · reviewed by: — |
Structure (of the edition). A short preface: the author answers a correspondent who asked about sainthood, and dismisses those who speak of it «by way of knowledge … by analogy and conjecture» and stop at ṣidq (the servant's own sincerity), not reaching the minan (God's favours) (¶1–2). Ch. 1–3: the friends of God are of two kinds. The walī ḥaqq Allāh, the friend of God's due, guards his seven members, then finds his self full of desire, cuts the tree's branches and then its root, and strives by his own sincerity until it fails him; in his helplessness he cries out, and mercy lifts his heart to the place of nearness by the Throne (Q 27:62); the walī Allāh, the friend of God, is drawn there by God (¶3–32). Ch. 4: the masāʾil rūḥāniyya, 157 questions put to anyone who claims to speak of sainthood (the stations of the friends, the people of nearness and of converse, who deserves the Seal of the saints as Muḥammad deserved the Seal of the prophets, the knowledge of the beginning, the letters, the Mother of the Book …), with no answers given (¶33–45). Ch. 5–7: this is the knowledge of the prophets and the friends; the friend of God; the ten marks of friendship with God (¶46–62). Ch. 8 (titled by the editor «the Seal of the saints and the Seal of the prophets»): asked about the master of the friends of God who holds «the seal of friendship with God» (already named at ¶59), the author turns to the Seal of prophethood: the seal between the Prophet's shoulders, why the sealed is guarded, khātam and khātim, and the hadith of the prophets gathered in Jerusalem, where Muḥammad says God «made me a seal and an opener» (¶63–70). The Seal of the saints himself is the subject of chapters 13 and 25 (queued).
The chapter titles are the editor's, printed in parentheses (e.g. «(الفصل الثامن) (خاتم الأولياء وخاتم الأنبياء)»). Our section headings give them in English with the Arabic kept in khatm-awliya-base.json → sections. The editor's numbering of the questions is kept in the text as printed («(السؤال الأول) …»).
0. The title, and the English translation that does exist
The scoping note (docs/research/text-niche-scope.md on research/text-niche, rank 3) found no English translation of the Khatm, only of «two other works» (Radtke & O'Kane 1996). That is corrected here (ledger docs/research/khatm-awliya-sources.md, 1a–1b, verified): the BnF record of Radtke & O'Kane lists its two works as the autobiography and Kitāb Sīrat al-awliyāʾ, and Radtke's own article ("A Forerunner of Ibn al-ʿArabī: Ḥakīm Tirmidhī on Sainthood", Ibn ʿArabī Society, link) says that Sīrat al-awliyāʾ is the proper title of the work known until then as Khatm al-awliyāʾ, first edited by Yaḥyā, and that he was preparing a new edition from newly found manuscripts. So the English translation exists, in copyright, of Radtke's recension, not of the text shown here. Whether it is complete is not established. It is not held; our English was made without it (§10). In the reader the work keeps the name its edition and Ibn ʿArabī's readers use, Khatm al-awliyāʾ, with Sīrat al-awliyāʾ given on the About page and in this record.
1. Witnesses
| Siglum | Witness | Kind | Vowelled? | Licence | Location |
|---|---|---|---|---|---|
| W | ed. al-Shaykh ʿAbd al-Wārith Muḥammad ʿAlī, Beirut: Dār al-Kutub al-ʿIlmiyya, 1st ed. 1420/1999, via OpenITI 0320HakimTirmidhi.KhatmAwliya.ShamAY0033732-ara1 (from al-Maktaba al-Shāmila; header «موافق للمطبوع», paged to the print) | modern edition, digital | no | CC BY-NC-SA 4.0 (digital text) | committed whole: data/sources/openiti/ (SHA-256 9b705e4a…297874b, README there); GitHub, commit 089e665b; every passage links its lines |
| Y | O. I. Yaḥyā (ed.), Kitāb khatm al-awliyāʾ, Recherches de l'Institut de Lettres Orientales de Beyrouth XIX (Beirut: Imprimerie Catholique, [1965]), viii + 586 pp. (BSOAS 30.3 (1967) 688–689, Lings's review, DOI; UPenn OCLC 22680669) | critical edition | — | copyrighted | not seen; cited only (ACQUISITIONS.md) |
| R | B. Radtke (ed.), Drei Schriften des Theosophen von Tirmiḏ, Teil 1, Die arabischen Texte (Beirut/Stuttgart 1992; Bibliotheca Islamica 35), with Sīrat al-awliyāʾ; Teil 2, Übersetzung und Kommentar (1996) | critical edition (other manuscripts) | — | copyrighted | not seen; cited only (ACQUISITIONS.md) |
What W is. The OpenITI version note says all paratext (introductions, notes, indices) was removed in its 2023 clean-up, so W carries none of the editor's apparatus. It does carry his chapter titles and his parentheses, which mark words he supplied («(عز وجل!)», «(لا)», «(هو)», «(الرسول عليه الصلاة والسلام)»). Whether he took these from Yaḥyā or a manuscript is not stated in the file or its metadata (ledger item 6, unresolved). We keep them as printed and do not cite any reading as Yaḥyā's. The Shāmila header's author field wrongly joins al-Wāḥidī's biography to al-Tirmidhī's; a catalogue slip. Two user-uploaded scans said to be Yaḥyā's edition are on archive.org (KhatmAwliyaTarmeziOsmanYahya, KhatmAwliyaTarmezi) with no licence; the edition is in copyright, so they were not fetched. No public-domain printing was found (one method, archive.org and the web; absence not established).
2. Base and why
W: the only openly licensed text of the work. It is a modern popular edition and not the critical text scholars cite (Y), and a third witness (R) exists from other manuscripts. Our text is therefore a reading edition of W, said to be so on every page through the version record (app/src/content/versions/khatm-awliya.json), and answers to Y and R once they are acquired.
3. Collation
None: one witness held. Mechanical clean-up only (scripts/recitals/khatm_base.py, logged in khatm-awliya-base.json → mechanical and changes): page tags and milestones removed; OpenITI's automatic header marks («### | AUTO», mostly not headings) read as paragraph starts; the nine chapter-heading lines moved out of the running text into the section headings; the edition's curly braces round Quran quotations removed.
4. Doubtful readings (kept as printed; AI-proposed, for a person to check against Y and R)
Reported by the vowelling readers and kept in the text exactly as W prints it (the vowels follow the letters; where the sense needs another reading the English says so in brackets, and the word is listed in the passage's ambiguous). None has been emended.
- ¶6 «وحسب إن رفضها إماتتها»: «إن» where «أن» is expected.
- ¶8 «فلم فتح له الطريق»: does not construe; almost certainly «فلما» (English: «when»).
- ¶8 «فاحذرك هذا الباب»: no hamza; the sense is «I warn you against this door».
- ¶12 «إلا يقضى»: probably «ألّا» (hamza seat kept as printed).
- ¶15 «لأ تقرب»: stray space for «لأتقرب» (kept); «أحسب هذا» read as «did this man suppose» (or «I suppose»).
- ¶17 «تعطي» printed with ي, read as passive «is given».
- ¶18 «الأخوان» read as al-ikhwān; «روح القربة» read rawḥ (relief) not rūḥ.
- ¶19 «تبقيا» where «تبقية» is expected.
- ¶21 «منعهما»: dual suffix where the singular self is meant; probably «منعها».
- ¶23 «الأخوان»: hamza above alif; al-ikhwān meant.
- ¶25 «أن وله قلبك إلى صدق نفسك وجهدك يكشف السوء عنك»: the next clause contradicts it; a negative («لا يكشف») has probably dropped out. English has «[does not]».
- ¶25 «تنزليه»: misprint for «تنزيله».
- ¶35 (Q34) «الربع والثلاثون»: misprint for «الرابع» in the edition's question number; letters kept as printed.
- ¶36 (Q45) «أوهم بالسجود له»: looks corrupt (perhaps «أُمِروا»); vowelled as printed (passive أُوهِمَ), English marked [text uncertain].
- ¶38 (Q72) hadith opens «((» with no closing «))» in the print.
- ¶38 (Q72) «فيذهل أهل الجنان»: intransitive يَذْهَلُ أَهْلُ or causative يُذْهِلُ أَهْلَ; AI chose intransitive.
- ¶40 (Q94) «من يكون محقا»: مَحْقًا (effacement, as Q93) or مُحِقًّا (in the right); AI chose مَحْقًا.
- ¶40 (Q91) «بماذا وكل»: passive وُكِّلَ or active وَكَّلَ; AI chose passive.
- ¶41 «أمين» printed with hamza-alif (āmīn, Amen).
- ¶46 «محارصة»: doubtful word (perhaps «مخالصة»); English marks it.
- ¶48 «ما زجه»: probably «مازجه»; the sentence is obscure, English tentative and flagged.
- ¶51 «ونصحاؤه»: nominative in apposition to a genitive.
- ¶53, ¶54 «عقله»: perhaps «عقاله» (its restraint).
- ¶56 «يطمأنوا»: hamza seat irregular (for «يطمئنوا»).
- ¶60 «لا يدري»: passive «لا يُدرى» probably meant (English: «is not known»).
- ¶63 «يعمي»: read يُعْمِي.
- ¶69 «المحتوم»: probably «المختوم» (the sealed [book]).
- ¶69 «ينبؤك»: hamza on wāw (for ينبئك).
- ¶70 «الربيع بن أبي العالية»: probably al-Rabīʿ [b. Anas] from Abū l-ʿĀliya (a chain slip).
- ¶63 «(الأصل: محمد)»: an editor's note that survived OpenITI's clean-up («the original has Muḥammad», i.e. the nominative where the accusative is needed); kept, and said so in the English.
5. Quotations verified
17 places in the edition's Quran braces. 15 match the Tanzil Uthmani text exactly on the consonantal skeleton (searched against the whole Quran, not a cited verse: W gives no references) and are replaced by it verbatim, then restored after vowelling (restore_quran.py): Q 29:69 (twice), 8:24, 27:62, 22:78, 10:14, 28:88, 29:45, 2:152, 2:257, 10:62, 13:28, 10:2 (twice), and «وبشر الذين آمنوا وعملوا الصالحات» (2:25, also found elsewhere in the same words). 2 are allusions left as printed: «الظاهر والباطن والأول والآخر» and «هو الظاهر والباطن» (Q 57:3, reordered and cut). Hadith in the edition's double parentheses «(( ))» are kept as printed and not yet identified (¶70: the prophets gathered in Jerusalem, «He made me a seal and an opener»). scripts/quran_refs.py khatm-awliya adds the automatic matches.
6. Vowelling
W is unvowelled, so every mark is AI. Made inside this session by Claude Sonnet subagents (seven batches of ten passages; no headless claude -p run), from fixed instructions: full tashkīl, letters, spacing and punctuation unchanged, Uthmani Quran text untouched, two-reading words listed per passage (ambiguous). scripts/recitals/khatm_assemble.py rejects any passage whose letters changed (none rejected) and sets the per-character mask ('a' = AI). 142 words with two defensible vowellings are listed (khatm-awliya.json → passages[].ambiguous). CAMeL disagrees with the AI vowelling on 249 words (§8). Not compared with any vowelled edition (none known).
7. Segmentation and word IDs
70 passages, following the book's own divisions: every chapter opens a passage (with the preface as passage 1); prose in whole OpenITI paragraphs of about 80–150 words; the questionnaire in groups of twelve whole questions (the last thirteen), each headed «4 · The spiritual questions, nos. a–b». Word IDs ka:<passage>:<word>. Pages of the 1999 print from the file's page tags.
8. Analysis layer
CAMeL Tools 1.6 (calima-msa-r13 + MLE) via scripts/fusus/analyze_ch.py khatm-awliya.json khatm-awliya-analysis.json ka, run in this session (pip camel-tools, data by camel_data -i light; the raw QAC morphology and Maqāyīs files fetched by the lines in scripts/fetch_sources.sh, SHA-256 of QAC as recorded): 7,077 words, 4,479 with a root, 659 distinct roots, 249 vowel conflicts. Reader data written by scripts/recitals/khatm_reader_data.py (the EXTRA-texts step of build_data.py and the per-word files of build_client_data.py, for this text alone, since the full build needs the lexicons a clone lacks). 30 roots have no root page yet (e.g. أزف، بجل، ختل، كفح، وله): their links resolve after the next full build_data.py on the owner's machine.
9. Commentaries aligned
None exists as a commentary. Ibn ʿArabī answered the questionnaire: a separate short treatise (al-Jawāb al-mustaqīm, per Radtke's article) and at length in al-Futūḥāt al-makkiyya ch. 73 (Ibn ʿArabī Society article anqa.co.uk; one Spanish article says ch. 72: discrepancy noted, ch. 73 adopted pending a check in the Futūḥāt file). Not in the reader yet; aligning his answers to the 157 questions is the natural ring for this text (§12).
10. Translation
AI (Claude Sonnet subagents in this session, with the glossary in data/sources/khatm-awliya-meta.json: walī "friend of God", walāya "friendship with God", khatm al-awliyāʾ "the Seal of the saints", walī ḥaqq Allāh "the friend of God's due", ṣidq "sincerity", minna "favour", qurba "nearness", majlis "session", mulk al-mulk "the Kingdom of the Kingdom", muḥaddath "one addressed [by God]", jadhb "drawing" …). Told not to reproduce or imitate Radtke & O'Kane and given none. The questions are numbered in the English as in the edition. Glosses for every word id and term notes from the same run. All ✦ unreviewed.
10b. Translation triangulation
- Held: no specialist translation (Radtke & O'Kane 1996 is the only English, of Radtke's recension, in copyright, not held:
ACQUISITIONS.md).scripts/triangulate.pywas not run (nothing to compare). - Partial check against open scholarship: Radtke's article "A Forerunner of Ibn al-ʿArabī" (Ibn ʿArabī Society, read 11 Oct 2026 in full by an agent) turned out to contain no translated passage of the work, only paraphrase. Four touchpoints compared in sense: the two kinds of friends (our ¶3 and ¶52: the friend of God's due repents, resolves on faithfulness, keeps the obligations and guards the seven members) agrees; the forty (¶28) consistent; «seal of friendship with God» as al-Tirmidhī's coinage agrees with our terms; no Arabic differences between the recensions are shown there. No change to our English followed.
- Fixed by hand after the run: ¶34 «Seal of the friends of God» → «Seal of the saints» (the glossary's set phrase); ¶63 the surviving editor's note rendered as such.
11. Rights and private layers
W: CC BY-NC-SA 4.0 (OpenITI), committed with attribution. Y, R and Radtke & O'Kane: in copyright, cited only, none held, none quoted; their wording is not in this reader. Audio: none made.
12. Open questions / to review
- The rest of the book: chapters 9–29, OpenITI ll. 766–2160 (about 16,700 words), with the two later Seal chapters (13, «خاتم الأولياء»; 25, the same title). Rerun
khatm_base.pywithLASTmoved (word IDs of ¶1–70 stay stable), then the same batch method. - Acquire Y and R (
ACQUISITIONS.md) and collate; decide which of W's parentheses are Yaḥyā's; check the doubtful readings in §4. - Acquire Radtke & O'Kane privately and run
scripts/triangulate.py khatm-awliya. scripts/resonances.py whole --only khatm-awliya(a paid run, on the owner's machine), and the companions sweep.- Ibn ʿArabī's answers (Futūḥāt ch. 73) as a ring on chapter 4, once the Futūḥāt passages are in the reader.
- A person should review the vowelling and the English, above all the Seal chapter (¶63–70) and the Seal questions (¶34).
13. Links
- Faṣṣ Shīth (
docs/texts/fusus-02-shith.md, "Links"): the Fuṣūṣ chapter's resonance «al-Ḥakīm al-Tirmidhī, Khatm al-awliyāʾ» (lineage, theme prophecy-and-sainthood) now links inside the reader to ¶63 (ch. 8: the master of the friends of God and his seal, then the Seal of prophethood) and ¶34 (questions 13–15: who deserves the Seal of the saints, by what attribute, and what the seal is and means) (app/src/content/resonances/fusus-2.json→connections[0].internal, edited by hand; a rerun ofresonances.pymust keep it). The reverse link is in the reader's links bar (lib/library.ts). Faṣṣ Shīth's Seal passages are its ¶14–19 and ¶24 (the gold and silver bricks, «the Seal of the saints … follows in the law what the Seal of the messengers brought»). - Ibn ʿArabī's Futūḥāt, ch. 73 (his answers to these questions): not in the reader yet; linked in the links bar to the Futūḥāt's opening.
14. Change log
- 2026-10-11: first third (preface, ch. 1–8) built from W; Quran restored (15); vowelled, translated and glossed by AI inside the session; CAMeL analysis; registered at
/read/khatm-awliyaunder Sufism; linked from Faṣṣ Shīth; sources ledger and acquisitions.