The weakening of the twenty-rakʿah report from ʿUmar by Imām al-Shāfiʿī and al-Tirmidhī

Al-Tirmidhī, in his Sunan (2/74), alluded to the fact that the tradition attributing twenty rakʿahs to ʿUmar—indeed, to any of the Companions—has no firm basis. He wrote:
“It has been narrated from ʿAlī, ʿUmar, and others among the Companions of the Prophet ﷺ.”

Al-Shāfiʿī said the same concerning the twenty rakʿahs reported from ʿUmar, as transmitted by his student al-Muzanī in the Mukhtaṣar (1/107).

Their shared wording—“it has been narrated (ruwiya)”—is, as hadith scholars know, a deliberate signal of weakness in the report. It is well established that Imām al-Shāfiʿī and al-Tirmidhī belong to the class of meticulous scholars whom al-Nawawī—may Allah have mercy on him—meant when he wrote in al-Majmūʿ (1/63):

The rigorous scholars from among the specialists of hadith and others say: when a hadith is weak, one must not state categorically, “The Messenger of Allah ﷺ said, did, commanded, forbade, ruled,” and the like. Nor should one say, “Abū Hurayrah narrated, said, reported, informed, related, transmitted, issued a fatwa,” or similar expressions. The same applies to the Successors and those after them when the report is weak: nothing should be phrased with certainty. Instead one says in every such case: “It is narrated from him,” “It is reported from him,” “It has reached us from him,” “It is said,” “It is mentioned,” “It is related,” “It is attributed,” and the like—all tentative forms, not definitive ones.
⦗64⦘

They add: categorical expressions are reserved for sound or acceptable (ḥasan) reports, while tentative phrases are for everything else. Using a definitive form implies authenticity on the part of the one to whom the statement is ascribed; it must therefore be restricted to what is in fact authentic, otherwise a person borders on lying about him. This etiquette has been neglected by the present author and by most jurists—indeed, by the majority of scholars in every discipline except the adept hadith critics. It is a blameworthy laxity, for they often use “it is narrated from so-and-so” for a sound report, and “he said” or “so-and-so narrated” for a weak one, which is a departure from what is correct.

These narrations do not strengthen one another

Someone might object: “Granted, each individual isnād is weak, but do they not reinforce one another through their sheer number?”

I reply: Certainly not, for two reasons.

First: that numerical ‘abundance’ may well be only superficial. In reality we possess nothing more than the connected narration of al-Sāʾib ibn Yazīd, together with the disconnected reports of Yazīd ibn Rumān and Yaḥyā ibn Saʿīd al-Anṣārī. It is quite possible that all of these ultimately go back to a single transmitter among those who carried the first narration—or to someone else altogether, as will be shown—and once such a possibility exists, the inference collapses. ⦗65⦘