Also called mawsul; defined as what every one of its narrators heard from the one above him, all the way to its end, whether elevated to the Prophet or stopped at a Companion. Itr adds that hearing here also covers other recognized modes of reception, such as being read back to the shaykh or a sound written authorization, since hearing is named in the definition only because it is the most common mode.
What Malik related from Nafi from Abdullah ibn Umar, that the Messenger of Allah ﷺ said whoever misses the afternoon prayer is as if he lost his family and wealth. Itr notes in his own footnote that this exact chain, Malik from Nafi from Ibn Umar, is known among the scholars as the golden chain, silsilat al-dhahab.
Also from Malik, from Nafi, who heard Ibn Umar say whoever gives a loan should stipulate nothing in it but its repayment.
A maqtu report, one attributed to a Successor, is technically connected too if its own chain has no gaps, and no one disputes it falls under this category. But most scholars held it should not be called simply mawsul or muttasil without qualification. It should be paired with whatever marks it off from the two cases just given, the connected marfu and the connected mawquf, so that one says: this is connected up to Said ibn al-Musayyab, for example.
The standard, most relied-upon definition, which Itr adopts: a report whose chain is connected and raised all the way to the Prophet ﷺ. On this definition a mawquf or maqtu report does not qualify even if its own chain is fully connected, and a munqati report does not qualify even if it is elevated to the Prophet.
Itr singles out and pushes back on Ibn Hajar's Nukhba gloss, which held that musnad only requires an apparently connected chain, so a hidden gap such as an unproven tadlis-style an-chain or two contemporaries who never actually met would not disqualify it. Itr says he found no earlier scholar who said this, and found evidence against it: surveying the first two hundred hadiths of Ahmad's Musnad, he counts thirty that the hadith scholar Ahmad Shakir (d. 1377H) had already flagged as genuinely disconnected. He adds that al-Hakim (d. 405H), whom Ibn Hajar (d. 852H) claims to be following, explicitly denied any tadlis in the musnad category, so Ibn Hajar's claim to be matching Hakim's own position does not actually hold up.
Itr flags two looser uses worth knowing: some scholars, following Ibn Abd al-Barr (d. 463H), used musnad for anything raised to the Prophet regardless of whether the chain is connected; and musnad is also simply the title convention for hadith collections arranged by Companion-narrator, such as al-Shihab's Musnad or al-Firdaws's Musnad, meaning there just those books' own chains.
Mu'an'an is a chain where it is said so-and-so from so-and-so, without stating outright that direct narration, informing, or hearing took place. Mu'annan is its close cousin, a chain where it is said so-and-so that so-and-so said, using anna instead of an. Itr says the sound majority position treats the two exactly alike; the specific word does not matter, only whether the men actually met, sat together, and heard from each other. Itr first mentions the an-from chain among the modes of narration, then returns to it here in full.
For an an-from report to be judged connected: first, that the narrator's actual meeting with the one he narrates from is established; second, that he is clear of any taint of concealment, tadlis. Once both hold, saying from so-and-so counts the same as saying he told me or I heard him.
Muslim ibn al-Hajjaj broke from the majority requirement. In his Sahih's introduction he argued that requiring proof of an actual meeting was an invented position with no earlier precedent, and that the long-agreed practice was that it is enough for the two narrators to be shown as contemporaries, even without proof they ever actually met. Itr reports the majority's answer: this claimed consensus does not hold, since major critics such as Ali ibn al-Madini and al-Bukhari (d. 256H) held the stricter view; where they did accept an unproven an-from chain, it was because of other contextual indicators establishing the meeting, not mere contemporaneity on its own.
Itr sides with the stricter majority position as the more cautious one, since it establishes connectedness more firmly, and notes this exact disagreement is one of the reasons scholars have long ranked Bukhari's Sahih above Muslim's.
With the hadith scholars: what its chain's narrators, one after another, share a single quality or state in, whether the quality belongs to the narrators themselves or to their act of narrating. Itr places it as the fifth of eight connectivity-related sciences, right after the two anna-wording chains and before highness and lowness.
Chain-linked by a spoken state shared across every narrator. Example: the Prophet's ﷺ hadith to Muadh ibn Jabal, O Muadh, I love you, so say after every prayer: O Allah, help me to remember You, thank You, and worship You well; each narrator down the chain adds and I love you too, so say, before relaying it, so the very words of transmission chain-link along with the report itself.
Chain-linked by a shared physical act. Example: Abu Hurayra's report, Abu al-Qasim ﷺ interlaced his fingers with mine and said, Allah created the earth on Saturday; every narrator down the chain likewise interlaced fingers with the one he was narrating to while passing it on. Itr notes the same pattern with a hand placed on the shoulder, or on the head.
Both a word and an act chain-link together. Example: Anas's report that the Prophet ﷺ said a servant does not taste the sweetness of faith until he believes in decree, its good and its bad, while grasping his own beard and saying I believe in decree, its good and its bad; each narrator repeated both the saying and the beard-grasping in turn.
Here the shared thread is not something each narrator says or does while transmitting, but a lasting trait of the narrators as people. The fourth kind, verbal traits, Itr says is in practice near-identical to the first kind above. The fifth, physical or biographical traits, covers chains where every narrator happens to share a name, such as a chain linked entirely by men named Muhammad, or shares a description, such as a chain of jurists, or memorizers, or centenarians, or Sufis.
The shared thread is a feature of the act of transmission itself, not of the people or of what the hadith says. By wording of transmission: every narrator using the same set formula, such as I heard so-and-so or so-and-so informed us. By timing: a chain narrated only on the day of Eid. By place: a chain in which each narrator's supplication was answered at the Multazam, the wall between the Black Stone and the Kaaba's door. What links the chain is the answered supplication, not merely that the hadith was passed on there.
Itr's own assessment: chain-linking shows that the links of the chain are genuinely connected, each bound to the next by a specific shared quality or state, which reinforces the very meaning of connectedness in the hadith. He cites al-Hakim (d. 405H) calling it a form of open, undisputed hearing, and Ibn al-Salah (d. 643H) calling the best of it that which points to connected hearing and freedom from tadlis, adding that part of chain-linking's virtue is the extra precision it shows in its narrators.
Highness is of two kinds: highness of distance through fewer intermediaries, and highness of attribute; lowness (nuzul) is the increase of intermediaries, and highness is always sought.
That a narrator adds, in a connected chain, a man whom others did not mention; known by gathering and comparing the routes.
The author follows the broad view (Ibn Abd al-Barr (d. 463H)): what has one narrator or more dropped at any point of the chain, not the narrower one-narrator one-spot wording.
What a Successor raises to the Prophet, by saying: the Messenger of Allah ﷺ said, whether the Successor is senior or junior. This is the meaning to which the later scholars confined the term.
Scholars differed a great deal over using a mursal report as proof. Itr sets out the three most important and best known positions, in this order.
First position, held by most hadith scholars and by many jurists and legal theorists: the mursal is weak and is not used as proof.
Second position, al-Shafii's (d. 204H), as he set it out in the Risala: he accepts the mursal of the senior Successors only, and subject to corroboration (itibar) in both the report and the narrator who sends it.
Third position, that of Abu Hanifa (d. 150H), Malik (d. 179H) and their companions: a reliable narrator's mursal is sound and is used as proof.
Itr's own position: a mursal report hangs between soundness and weakness, so when indications attend it that strengthen it, it should be acted on and used as proof. He regards that as the furthest the leading jurists go on this question.
Al-Shafii's condition of itibar, meaning corroboration, is met for the report itself by any one of four things: it is also narrated with a connected chain through another route; or narrated as mursal with the same meaning by a different narrator who did not take from the first narrator's teachers, showing the report has more than one point of origin; or a Companion's statement agrees with it; or most scholars have held to it. Corroboration in the mursal narrator himself means that when he does name who he narrated from, that person is not unknown or someone whose narration is rejected.
What a Companion relates from the Prophet ﷺ without having heard it from him, whether for his young age, late acceptance of Islam, or absence from witnessing it; it is accepted.
A hadith whose chain has had its beginning removed, whether the removed portion is one narrator or more in unbroken succession, even if the removal runs all the way to the end of the chain. This is the opposite end of the chain from where mursal drops its narrator; mu'allaq is missing from the start, mursal from the end.
Hadith scholars do this often, especially in their own compiled works, aiming either for brevity in citing a hadith, or to strengthen an argument for a chapter's topic using material that does not otherwise meet that book's own stated conditions.
Named mu'allaq, suspended, because by having its start removed it becomes like something cut off from the ground below but still attached from above, the way something might hang suspended from a ceiling.
Rejected, on the same footing as munqati, since the state of whoever was removed is unknown, unless it occurs in a book that has committed itself to including only sound material, such as Sahih al-Bukhari (d. 256H) and Sahih Muslim; scholars studied the suspended reports in those two books specifically and reached a special scholarly conclusion just for them.
For al-Bukhari (d. 256H) specifically: his wording when suspending a report is either decisive, such as so-and-so said, or he narrated, or he related, or it is a wording that does not commit him, such as it is narrated from so-and-so, or it is said. Scholars call this second kind the wording of illness, since it signals doubt rather than commitment. Decisive wording counts as Bukhari's own ruling that the report is sound up to whoever he names it from; the hedged wording carries no such commitment from him.
With the hadith scholars: what has two or more narrators dropped from its chain at a single place, whether at the start of the chain, its middle, or its end.
Scholars have proposed many sub-classifications of tadlis, but Itr notes they all reduce to two main kinds: tadlis of the chain (isnad) and tadlis of the shaykhs.
What is depended on: the hadith a narrator relates from a contemporary he neither heard from nor met; outwardly connected, inwardly severed.