As for a compiler’s splitting a single ḥadith and distributing its parts under separate chapter headings, the view closest to correctness is that it is permissible. The Shaykh said that such a practice is not devoid of reprehensibility, but I do not think anyone will side with him on that point.

Sixth. One should not transmit a ḥadith when it has been read aloud by a laḥḥān—someone prone to grammatical error—or by a muṣaḥḥif, a reader given to taṣḥīf (graphic misreadings). A student of ḥadith must acquire enough grammar and vocabulary to protect himself from laḥn (faulty grammar) and taṣḥīf. The surest safeguard against such corruptions is to receive ḥadith orally from scholars who are both knowledgeable and scrupulous.

If a grammatical slip or distortion does creep into his narration, Ibn Sīrīn and Ibn Sukhbara held that he should report it exactly as he heard it, but the correct opinion—endorsed by the majority—is that he should give it in its sound form. As for amending the written text, some have allowed it outright; the sound procedure, however, is to leave the original wording in place, mark it, and record the correct form in the margin. At the time of public recitation it is best to read the corrected wording and then say, “In our transmission,” or “According to our teacher,” or “Through the authority of so-and-so it goes thus.” Alternatively, he may read what appears in the archetype and then supply the correction. The most commendable emendation is that which is corroborated by another narration or by a different ḥadith. Allah knows best.

When the emendation consists of restoring an omitted word, then—so long as its insertion does not change the sense—the previous ruling stands. If it does alter the meaning, the obligation is stronger: the original must be quoted and the clarification appended. Should he know that a single transmitter alone dropped the word, he may insert it in the text itself together with the explanatory term “that is (yaʿnī).” This applies when he is certain that his own shaykh recited it incorrectly. If, on the other hand, he finds the error in his own notebook and is confident that it stemmed from his copy rather than from his teacher, he should correct both his book and his narration—just as, when a portion of the chain of transmission or the wording has worn away in his manuscript, he may replace it from someone else’s book, provided he is satisfied of its authenticity and that it is indeed what had fallen out. That is the teaching of the meticulous authorities, although some scholars forbade it; in any case, explaining the matter at the moment of narration is preferable. The same rule applies when a ḥāfiẓ seeks confirmation of something over which he harbors doubt, whether by consulting another person’s book or by reviewing his own memory; and if he discovers in his own volume—