the narrator in the form of a report traced back to the Prophet, then issues a legal ruling citing it as his own opinion, after which it is conveyed from him as a report that stops with him. In that case the two transmissions are not mutually contradictory, for both can be reconciled.

Our answer is that this line of argument can neither be accepted unconditionally nor rejected unconditionally: in some situations it must be accepted (1), in others it must be rejected, and in still others one must withhold judgment.

• If the reliable, firmly established Imams have transmitted the ḥadīth as traced back to the Prophet or with a continuous chain, and someone who is not their peer contradicts them, or one of them on his own renders it as stopping with the Companion or as broken, that does not constitute a flaw in the report and does not undermine it; the ruling belongs to those who narrated it as traced and linked.

• But if the reverse is true—as with this report of Sufyān b. Ḥusayn and others like it—no attention is paid to it (2) or to those who deviate by presenting it as mawqūf or mursal. It is given no weight whatsoever, and by virtue of it the ḥadīth can never be deemed traced back to the Prophet or fully linked. All the experts of ḥadīth are agreed on this.

For when the trustworthy, exacting Imams among al-Zuhrī’s students have always transmitted it from him as stopping with Saʿīd, and not one of them has ever once raised it to the Prophet—despite their preservation of al-Zuhrī’s narrations, their precision in recording them, their intense concern for them (3), and their clear differentiation between what he narrated as traced, stopped, broken, and fully linked—and then someone appears who has not run in their track, who falls well short of them in memorization, accuracy, and personal association with al-Zuhrī,