Text to evidenceLesson 5openfrom namaq

How al-Bukhari chooses a hadith, and how a name in a chain becomes a person

What scholars say about al-Bukhari's method, what the Sahih prints about its narrators, and which other books are needed to say who a narrator is.

Namaq plans to extract Sahih al-Bukhari as printed and to link each name in a chain to a person through a sourced identification (docs/plans/data-model/plan.md, section 2.6). This lesson reports what scholars say about al-Bukhari's method, what the Sahih prints about its narrators, and which other books a project needs to say who a narrator is. Every statement about hadith science here is a report of what a named scholar wrote. Namaq adds nothing of its own. A scholar will review this lesson.

How to read this lesson. Each claim names its source in the same sentence. A quotation in Arabic is short and was read on the page linked in the Sources list. When a page was reached only through a search summary or a secondary article, the claim says so. Anything not confirmed from an opened page is marked UNVERIFIED — for the scholar.

What a "condition" is

A condition (شرط, shart) is the rule a compiler used to decide which reports enter his book. It concerns the chain (إسناد, isnad: the list of people who passed the report on) and the people in it: were they trustworthy, were they accurate, did each one really receive the report from the one before?

The six books that most later scholars treat as the core collections are called the Six (الكتب الستة): the Sahih of al-Bukhari, the Sahih of Muslim, and the Sunan of Abu Dawud, al-Tirmidhi, al-Nasa'i and Ibn Majah. They differ because each compiler had a different aim. Two of them aimed to collect only sound reports. The Sunan compilers aimed to gather the reports a jurist needs, and admitted weaker material with a note or without one. Their conditions therefore differ in strictness. UNVERIFIED — for the scholar: this summary of the Sunan compilers' aims is a general description, not a quotation from a page opened for this lesson.

One fact governs everything below. The compilers mostly did not write their conditions down. Ibn Tahir al-Maqdisi (d. 507 AH), in Shurut al-A'imma al-Sitta, says of al-Bukhari, Muslim and the others that none of them is reported to have said what his condition was, and that the conditions are known only by studying their books: «وإنما يعرف ذلك من سبر كتبهم» (quoted in the islamweb article «القول المفهم في معنى شرط الشيخين»; the book itself was not opened, so the wording is UNVERIFIED — for the scholar against the printed text). So every "condition of al-Bukhari" in this lesson is a later scholar's reconstruction from reading the Sahih. That is why scholars disagree about them.

What scholars say al-Bukhari's conditions are

The general definition of a sound hadith

Ibn al-Salah (d. 643 AH), in the first chapter (النوع الأول) of his Muqaddima, defines the sound hadith (الصحيح, sahih) as the report «الذي يتصل إسناده بنقل العدل الضابط عن العدل الضابط» whose chain is connected, carried by one upright and accurate narrator from another, and which «لا يكون شاذا ولا معللا», is neither irregular nor defective (Wikisource text of the Muqaddima, النوع الأول). That gives five tests, in Ibn al-Salah's words:

Connection (اتصال, ittisal): no one is missing from the chain.
Uprightness (عدالة, 'adala): each narrator is morally trustworthy.
Accuracy (ضبط, dabt): each narrator remembers or records exactly.
No irregularity (شذوذ, shudhudh): the report does not go against what more reliable narrators carry.
No hidden defect (علة, 'illa): a flaw that only an expert finds on comparing chains.

The glosses of shudhudh and 'illa above are the common textbook meanings, not quotations; UNVERIFIED — for the scholar.

Ibn al-Salah also records, in the same chapter, that al-Bukhari and Muslim «لم يستوعبا الصحيح في صحيحيهما ولا التزما ذلك»: they did not include every sound hadith, nor did they set out to. He supports this with al-Bukhari's own reported words: «ما أدخلت في كتابي الجامع إلا ما صحَّ وتركت من الصحاح لملال الطول», "I put into my book only what is sound, and left out sound reports so it would not grow too long" (Muqaddima, النوع الأول, Wikisource). So absence from the Sahih says nothing about whether a report is sound.

Meeting versus living at the same time

The best-known point of difference is about connection. When a chain says only «عن» ("from"), did the narrator actually hear the man before him? Ibn Hajar (d. 852 AH), in Hady al-Sari, the introduction to his commentary Fath al-Bari, in the chapter «الفصل الثاني في بيان موضوعه والكشف عن مغزاه فيه», reports al-Bukhari's position as not accepting such a chain «حتى يثبت اجتماعهما ولو مرة», "until it is established that the two met, even once" (Shamela book 1224, page 10). The same page reports that for Muslim «الإسناد المعنعن له حكم الاتصال», a chain with «عن» counts as connected, provided the two lived at the same time.

These two positions are named the condition of meeting (لقاء, liqa') and the condition of contemporaneity (معاصرة, mu'asara). An islamweb fatwa (no. 53837) summarises the difference as "al-Bukhari's condition is that the meeting be verified, Muslim's that it be possible." Muslim argued for his own position at length in the introduction to his Sahih, and al-Nawawi commented on it in his introduction to his commentary on Muslim. Neither text was opened for this lesson; UNVERIFIED — for the scholar. Whether al-Bukhari held "meeting" as a condition of soundness itself, or only as a stricter standard for his own book, is a question some scholars debate; UNVERIFIED — for the scholar, since no page opened here states the debate.

Reliability, and how close a student stayed to his teacher

Al-Hazimi (d. 584 AH), in Shurut al-A'imma al-Khamsa, describes the students of a teacher in layers (طبقات, tabaqat). As reported in the islamweb article «القول المفهم», he places al-Bukhari's condition in reports whose chain «اتصل إسناده بالثقات المتقنين الملازمين», carried by reliable, precise narrators who stayed long with their teachers. A search summary of the same work adds that al-Bukhari sometimes takes from the next layer, those who stayed with a teacher only briefly. The book itself was not opened: UNVERIFIED — for the scholar on exact wording and page.

What "on the condition of al-Bukhari" means

Later writers say a hadith outside the Sahih is "on the condition of al-Bukhari". What that means is itself disputed. The islamweb article «القول المفهم» quotes al-Nawawi as taking it to mean «أن يكون رجال إسناده في كتابيهما», that the chain's men are the very men found in the two books. The islamweb fatwa no. 53837 sets out two readings, the same men or men of the same quality, and prefers the first. Both are secondary reports; UNVERIFIED — for the scholar against al-Nawawi's own text.

Main reports, supporting reports and suspended reports

Scholars agree that al-Bukhari did not apply the same standard to everything in his book. Three kinds matter.

The main reports (الأصول, al-usul): the full, chained hadith he relies on for a chapter.
Supporting reports (المتابعات والشواهد, al-mutaba'at wa al-shawahid): other chains for the same content, brought to confirm it. Ibn Hajar, opening chapter nine of Hady al-Sari, separates narrators «من أخرج له منهم في الأصول أو في المتابعات والاستشهادات», those used in main reports from those used only in supporting ones (Shamela book 1224, page 379). The same page says that criticism of a narrator al-Bukhari used is not accepted «إلا بقادح واضح», except with a clear flaw.
Suspended reports (المعلقات, al-mu'allaqat): Ibn al-Salah defines a suspended report as one «ما حُذف من مبتدأ إسناده واحد أو أكثر», where one or more narrators are dropped from the start of the chain, and says «أغلب ما وقع ذلك في كتاب البخاري», most of these occur in al-Bukhari's book (Muqaddima, Wikisource). Ibn Hajar gave them their own chapter in Hady al-Sari: «الفصل الرابع في بيان السبب في إيراده للأحاديث المعلقة: مرفوعة وموقوفة، وشرح أحكام ذلك» (Shamela 1224, table of contents). So scholars do not treat a suspended report as meeting the book's full condition by default. How Ibn Hajar grades each kind of suspended report was not read here; UNVERIFIED — for the scholar.

Ibn Hajar opens the same chapter nine with «تخريج صاحب الصحيح لأي راو كان، مقتض لعدالته عنده», al-Bukhari's use of a narrator implies the narrator was upright in al-Bukhari's view (Shamela 1224, page 379). Note the last two words. Ibn Hajar is reporting al-Bukhari's judgement, and then lists the narrators other critics disputed.

How the Sahih is built

Books, chapters and headings

The Sahih is divided into books (كتاب, kitab, e.g. the Book of Prayer) and each book into chapters (باب, bab). Each chapter has a heading (ترجمة, tarjama, plural تراجم, tarajim). A heading is al-Bukhari's own sentence, often a legal point he draws from the hadith that follow. Ibn Hajar describes his work as «استنباط الأحكام ليبوب عليها», drawing out rulings so as to make chapters for them (Hady al-Sari, chapter two, Shamela 1224 page 10). A heading may hold a Qur'an verse, a saying of a Companion, or a suspended report. For Namaq, this matters: a heading is al-Bukhari speaking, not a narrator.

Repetition and cutting

The same hadith often appears in several chapters. Ibn Hajar's chapter three is about this: «الفصل الثالث في بيان تقطيعه للحديث واختصاره وفائدة إعادته له في الأبواب وتكراره». He says al-Bukhari repeats a hadith «مراعيا مع ذلك عدم إخلائه من فائدة حديثية», always making sure the repetition carries some new point of hadith knowledge, for example a different chain (Shamela 1224, page 13). He also cuts long hadith: «فإنه يخرج كل جملة منها في باب مستقل فرارا من التطويل», he puts each sentence in its own chapter to avoid length (same page). The same page notes that al-Bukhari may bring a chain where a narrator said only «عن», and then give another route where that narrator «مصرحا فيها بالسماع», states that he heard it.

Ibn al-Salah gives the count as «سبعة آلاف ومائتان وخمسة وسبعون حديثا بالأحاديث المتكررة», 7,275 with repetitions (Muqaddima, Wikisource). Ibn Hajar's own count is in chapter ten, «الفصل العاشر في عد أحاديث الجامع» (Shamela 1224, table of contents); his total was not read here, UNVERIFIED — for the scholar. Counts differ because counters treat repetitions and suspended reports differently.

What a chain prints, and what it does not

A chain in the Sahih is a run of names joined by words of transmission. A typical opening is «حَدَّثَنَا الحُمَيْدِيُّ ... قَالَ: حَدَّثَنَا سُفْيَانُ» (this pair is the example Namaq's own plan uses, section 2.6). What is printed:

A name in whatever form the narrator used. Sometimes a full name with father (nasab), sometimes a first name alone («سُفْيَانُ»), sometimes a teknonym (كنية, kunya, "father of so-and-so", e.g. «أبو هريرة»), sometimes a descriptive surname (نسبة, nisba, a tribe, town or trade, e.g. «الحُمَيْدِيُّ»), sometimes a nickname (لقب, laqab).
Words of transmission (صيغ الأداء): «حدثنا» and «أخبرنا» (he told us), «سمعت» (I heard), «عن» (from), «قال» (he said). These words are what the meeting debate in section 2 turns on.
Occasional clarifications such as «يعني ابن عيينة» ("meaning Ibn 'Uyayna"), added by a transmitter. Namaq's plan already treats such a phrase as a possible basis (SAME_WORK_EXPLICIT, plan section 2.6).

What is not printed: dates, places of residence, the narrator's teachers and students, and any judgement of reliability. The Sahih gives the name as the student spoke it, and nothing more. Ibn Hajar needed two whole chapters of Hady al-Sari to sort this out: chapter six on names that look alike in writing, «الفصل السادس في بيان المؤتلف والمختلف من الأسماء والكنى والألقاب والأنساب، مما وقع في صحيح البخاري», and chapter seven on bare names shared by several people, «الفصل السابع في تبيين الأسماء المهملة التي يكثر اشتراكها» (Shamela 1224, table of contents). The existence of chapter seven is itself the evidence: a reader of the Sahih alone cannot always tell which person a name is.

How narrators are identified

What the Sahih itself gives

Almost nothing beyond the name as printed, as section 3 showed. The chain gives the name's form and its position: who the narrator heard from and who heard from him. A specialist uses that position to narrow the candidates: a «سفيان» who teaches al-Humaydi is a different question from a «سفيان» who teaches someone a generation earlier. But the rule that settles the question is written in another book.

What al-Bukhari wrote elsewhere

al-Tarikh al-Kabir (The Great History) is al-Bukhari's own dictionary of narrators. It is on Shamela as book 956, in the edition with notes by al-Mu'allimi al-Yamani. A search summary reports 12,315 entries in the numbered printed edition; UNVERIFIED — for the scholar, the page was not opened. Entries are often short: a name, some teachers and students, sometimes a remark.

al-Du'afa' (The Weak Narrators), also called al-Du'afa' al-Saghir, is his short list of narrators he judged weak. It is on Shamela as book 8632. A search summary gives "under seven hundred" entries by old reports; UNVERIFIED — for the scholar.

Neither book is an index to the Sahih. They cover far more people than the Sahih uses, and they do not say "the «سفيان» on page so-and-so of the Sahih is this man".

Later works built on the Six

Work	Author	What it covers	How it helps identify
Tahdhib al-Kamal fi Asma' al-Rijal	al-Mizzi (d. 742 AH)	Every narrator in the Six and some related works. Shamela book 3722: ed. Bashar 'Awwad Ma'ruf, Mu'assasat al-Risala, Beirut, 1400–1413 AH, 35 volumes.	Each entry lists the narrator's teachers and students, with symbols for which book used him: «خ» for al-Bukhari's Sahih, «خت» for a suspended report in it (symbols from the Taqrib introduction, Wikisource). A name in a chain is matched by finding the entry whose teacher list holds the man above and whose student list holds the man below.
Tahdhib al-Tahdhib	Ibn Hajar	An abridgement of al-Mizzi to about a third, keeping every entry (Shamela 3722 description). On Shamela as book 3310 (Indian edition); edition details UNVERIFIED — for the scholar.	Same use, shorter; adds critics' remarks al-Mizzi lacked (UNVERIFIED — for the scholar).
Taqrib al-Tahdhib	Ibn Hajar	One line per narrator. Its introduction says he gives each a «حكم يشمل أصح ما قيل فيه» and sorts narrators into twelve ranks of reliability and twelve generations (Wikisource, Taqrib introduction).	Fast way to confirm a candidate's generation and full name. Its grade is Ibn Hajar's judgement, to be quoted as his, never restated as fact.
al-Kashif fi Ma'rifat man lahu Riwaya fi al-Kutub al-Sitta	al-Dhahabi (d. 748 AH)	Narrators of the Six, condensed from al-Mizzi. Shamela book 2171; ed. Muhammad 'Awwama and Ahmad al-Khatib, Dar al-Qibla, Jeddah, 1413 AH (from search summary; UNVERIFIED — for the scholar).	A short second witness beside the Taqrib.
al-Ta'dil wa al-Tajrih li-man Kharraja lahu al-Bukhari fi al-Jami' al-Sahih	Abu al-Walid al-Baji (d. 474 AH)	Only the narrators of al-Bukhari's Sahih. Shamela book 5851; ed. Abu Lubaba Husayn, Dar al-Liwa', Riyadh, 1406 AH, 3 volumes (search summary; UNVERIFIED — for the scholar).	The one work scoped exactly to the Sahih. Earlier than al-Mizzi.
Hady al-Sari, chapters 6, 7 and 9	Ibn Hajar	Look-alike names, shared bare names in the Sahih, and the criticised narrators of the Sahih in alphabetical order. Shamela book 1224, al-Maktaba al-Salafiyya, 1380 AH, ed. Muhibb al-Din al-Khatib, 493 pages.	Chapter seven names, for the Sahih specifically, which person a frequently shared bare name refers to.
Fath al-Bari	Ibn Hajar	Commentary on every hadith. Shamela book 1673 (Salafiyya edition) holds it with Hady al-Sari.	At each hadith Ibn Hajar often says which person an ambiguous name is. That makes it the place to look for a single occurrence (UNVERIFIED — for the scholar: no such passage was opened here).
al-Isabah fi Tamyiz al-Sahabah	Ibn Hajar	The Companions.	For the first link of a chain, the Companion. Its scope and editions were not opened; UNVERIFIED — for the scholar.

In each case the method is the same, as the works themselves are arranged: find the entry whose name, teachers and students fit the position in the chain. When two entries fit, the name stays ambiguous until a commentator such as Ibn Hajar says which one it is. This description of method is drawn from how the entries are laid out, not from a scholar's statement of method; UNVERIFIED — for the scholar.

Do we need a rijal work, or does al-Bukhari suffice?

Short answer: the Sahih does not suffice. A project that names the person behind each narrator must use at least one rijal work (كتب الرجال, books about narrators) and cite it. The reasons come from sections 3 and 4:

The Sahih prints a name as the student spoke it, with no dates, teachers list or grade (section 3).
Ibn Hajar wrote a chapter of Hady al-Sari just to resolve bare names in the Sahih that many people share (chapter seven, section 3). If the Sahih settled them, the chapter would not exist.
Al-Bukhari's own al-Tarikh al-Kabir is a rijal work, not part of the Sahih. Using it is already "using a rijal work beyond the Sahih". It also does not point from a Sahih chain to an entry.
Namaq's plan already rules out a Namaq-authored basis and allows RIJAL_ENTRY and EDITOR_NOTE as bases (plan section 2.6). Without a rijal work, most mentions would stay as unidentified text, which the plan allows but which leaves the graph almost empty for narrators.

Which works, and what they cost

Hady al-Sari, chapters six, seven and nine: one volume of 493 pages on Shamela (1224). Small, Sahih-specific, and it covers the hardest cases (shared bare names). The cheapest start.
Tahdhib al-Kamal: 35 volumes on Shamela (3722) in Bashar 'Awwad Ma'ruf's edition. Complete teacher and student lists, which is what matching a position in a chain needs. Large: the project would cite single entries, not extract the work.
Taqrib al-Tahdhib: one line per narrator. Cheap to cite, but it gives too little to match a position in a chain on its own. Several editions exist; which the project uses is a choice to make and record.
al-Baji: three volumes on Shamela (5851), scoped exactly to the Sahih's narrators. Fewer than al-Mizzi's teacher and student lists (UNVERIFIED — for the scholar).
Fath al-Bari: many volumes (count not confirmed). Cited one hadith at a time where an ambiguous name needs a commentator's word.

Namaq's rule that Shamela is its only digital host (AGENTS.md, "Content sources") applies to the Siyar. All of the works above appear on Shamela, so the same rule could extend to them, with the printed page numbers each Shamela edition carries.

What remains for a scholar

Which work is the first authority for identification, and in what order to consult the rest.
Whether a position-in-chain match against Tahdhib al-Kamal counts as a basis by itself, or only a named statement such as Ibn Hajar's in chapter seven or in Fath al-Bari does.
Which edition of each work to use.

What this means for Namaq

What Namaq extracts. The chain and text as printed, words of transmission included. The book, chapter and heading as printed, with the heading recorded as al-Bukhari's own words. Whether a report is a main report, a repeat, or a suspended one, but only where a printed source says so; the structure in section 3 tells you these kinds exist, not which hadith is which.

What Namaq identifies. A name in a chain links to a person only through an identification whose basis is a span in a rijal work, a commentary, or the Sahih's own «يعني» clarification (plan section 2.6). Competing identifications are kept side by side and marked disputed. Namaq never picks.

What Namaq must not claim.

It does not grade. "Sound" is al-Bukhari's judgement, reported by Ibn al-Salah and Ibn Hajar. The app says that, with the source, rather than "this hadith is sound".
It does not say a narrator is reliable. When it shows a grade, it quotes a grader: "Ibn Hajar, Taqrib: «ثقة»", with page and edition.
It does not say "on the condition of al-Bukhari" about anything. Section 2 showed even the meaning is disputed.
It does not infer that a narrator was a weak one from absence, or a strong one from presence. Ibn Hajar's own words limit presence to uprightness «عنده», in al-Bukhari's view, and he lists those others disputed.
It does not restate the meeting-versus-contemporaneity debate as a fact about any one chain.

What a scholar should check

Ibn Tahir al-Maqdisi's statement that the conditions are known «من سبر كتبهم», against the printed Shurut al-A'imma al-Sitta.
Al-Hazimi's description of al-Bukhari's condition by layers of students, against Shurut al-A'imma al-Khamsa.
Al-Nawawi's reading of "on the condition of the two" and his comments on Muslim's introduction.
Whether "meeting" is al-Bukhari's condition for soundness or for his own book only.
The common glosses of shudhudh and 'illa used in section 2.
How Ibn Hajar grades each kind of suspended report (Hady al-Sari, chapter four), and his hadith count (chapter ten).
Edition details for Tahdhib al-Tahdhib, al-Kashif, al-Baji, al-Isabah and Fath al-Bari, and entry counts for al-Tarikh al-Kabir and al-Du'afa'.
That Fath al-Bari regularly resolves ambiguous names at the hadith.
The description in section 4 of how a name is matched to an entry by teachers and students.
Which work should be the first authority for identification, and whether a chain-position match is enough without a named statement.

Sources

Pages opened:

Ibn al-Salah, Muqaddima, النوع الأول: https://ar.wikisource.org/wiki/مقدمة_ابن_الصلاح/النوع_الأول
Ibn Hajar, Hady al-Sari, contents and edition: https://shamela.ws/book/1224
Hady al-Sari, chapter two: https://shamela.ws/book/1224/10
Hady al-Sari, chapter three: https://shamela.ws/book/1224/13
Hady al-Sari, chapter nine: https://shamela.ws/book/1224/379
al-Mizzi, Tahdhib al-Kamal, edition page: https://shamela.ws/book/3722
Ibn Hajar, Taqrib al-Tahdhib, introduction: https://ar.wikisource.org/wiki/تقريب_التهذيب/المقدمة
islamweb, «القول المفهم في معنى شرط الشيخين»: https://www.islamweb.net/ar/article/206914/
islamweb fatwa 53837: https://www.islamweb.net/ar/fatwa/53837/

Seen only in search results (not opened): https://shamela.ws/book/956 (al-Tarikh al-Kabir), https://shamela.ws/book/8632/128 (al-Du'afa'), https://shamela.ws/book/3310 (Tahdhib al-Tahdhib), https://shamela.ws/book/2171 (al-Kashif), https://shamela.ws/index.php/book/5851 (al-Baji), https://shamela.ws/book/1673/1 (Fath al-Bari), https://www.bukhari-pedia.net/book/shorot_5 (al-Hazimi).