I say: Its chain of transmission is of the highest authenticity, as al-Suyuti declared in al-Masabih. By itself this is enough to refute Ibn ʿAbd al-Barr’s claim; how much more so now that those other corroborations—whose collection, to my knowledge, no one before me attempted—have been added. Praise be to Allah for His enabling grace.

It is not established that ʿUmar performed twenty rakʿahs.

An examination of the relevant reports and a demonstration of their weakness

It is impermissible to oppose this sound narration with what ʿAbd al-Razzaq transmitted through another route from Muhammad ibn Yusuf with the wording “twenty-one,” for the error in that wording is obvious on two counts: first, it contradicts the reliable wording cited earlier, namely “eleven”; second, ʿAbd al-Razzaq stands alone in narrating it with this wording. Even if those between him and Muhammad ibn Yusuf were sound, the flaw would still lie with ʿAbd al-Razzaq himself. Although he was a trustworthy ḥāfiẓ and a well-known compiler, he became blind toward the end of his life and his memory changed, as al-Ḥāfiẓ states in al-Taqrib. For this reason Ibn al-Ṣalāḥ listed him among those who became confused in their later years, saying in the Introduction to the Sciences of Ḥadith:

“Aḥmad ibn Ḥanbal mentioned that he went blind near the end of his life, so he had to be prompted and would merely repeat what he was prompted with; therefore, whatever was heard from him after he became blind is worthless.” Al-Nasaʾi remarked: “There is room for criticism concerning what was written from him in his old age.”

He (Ibn al-Ṣalāḥ) further states in the introduction to the aforementioned section (p. 391):

“The ruling concerning such mixed-up transmitters is that the ḥadith they related before their confusion are accepted, while what was taken from them after the confusion is not accepted… ⦗56⦘ …or when the matter is unclear and it is not known whether one heard from them before or after the confusion.”

I say: This report falls into the third category—it is unknown whether it was related before or after the confusion—so it is not acceptable. And that is assuming it were free of oddity and contradiction; how, then, could it be accepted when these defects are also present?

If it be argued: al-Firyabi in al-Ṣiyām (1/76) and al-Bayhaqi in al-Sunan (2/496) transmitted, through the route of Yazid ibn Khuṣayfah from al-Saʾib ibn Yazid, that:

“They used to stand in prayer in the month of Ramadan during the time of ʿUmar ibn al-Khattab, may Allah be pleased with him, for twenty rakʿahs. He said: They would recite the miʾin, and during the time of ʿUthman, may Allah be pleased with him, they would lean on staffs because of the length of the standing.”

I reply: This transmission with the wording “twenty” is the main prop for those who deem twenty rakʿahs in the Tarawih prayer to be prescribed. Its chain appears sound and, on that basis, some scholars declared it authentic. Yet it bears a defect—indeed, several defects—that prevent it from being judged sound and render it weak and objectionable. That clarification is as follows:

First: Although Ibn Khuṣayfah is trustworthy, Imam Aḥmad said of him in one report, “His ḥadith is munkar (irregular).” Al-Dhahabi therefore listed him in al-Mizan. Aḥmad’s comment indicates that Ibn Khuṣayfah may at times transmit what he alone reports… ⦗57⦘