from himself, independent of the ḥadīth itself. For this reason Ibn al-Qayyim, d. 751 AH, says in his Nūniyyah:
“Abū Hurayrah said it out of his own pocket,
And men of insight can at once discern it.”

How is idrāj (interpolation) detected? By three means:

1. An explicit report: another chain arrives and states clearly that the wording is interpolated.
2. The impossibility that the Prophet ﷺ could have uttered it—either because it contains an obvious error or because a contextual clue shows it cannot be part of his speech.
3. A plain declaration by one of the master ḥadīth critics that the passage is interpolated.

What, then, is the ruling on interpolation?
If the interpolation alters the meaning, it is impermissible to include it without flagging it as such.
If it leaves the meaning intact—as in al-Zuhrī’s gloss, “al-taḥannuth means worship”—then there is no harm, for it does not conflict with the marfūʿ ḥadīth. When no conflict exists, it may be cited by way of explanation and clarification.

Once interpolation is established, it carries no probative force, because it is not the Prophet’s speech and therefore cannot be used as evidence.

As for the author’s phrase, “min baʿḍi alfāẓ al-ruwāt ittasalat” (“some of the transmitters’ words became attached”), the word “ittasalat” functions as a circumstantial clause (ḥāl) linked to the subject of “atat,” meaning that the phrasing came attached to the ḥadīth with no indication of separation.