ShowHide the Study Notes (2,071 words)
CARRIED ON THIS PAGE: TWELVE ayat. Carried before and now with named commentators or a named reading: 15:9, the participle said by Allah of the Reminder; 11:57, Hūd; 12:64, Yaʻqūb; 18:82, the wall, WHICH DOES NOT CONTAIN THE ROOT AND SAYS SO ON THE PAGE; 86:4; and 82:10–12.
ADDED: 34:21 and 42:6, WHICH WITH 11:57 ARE THE ONLY THREE PLACES THE WORD IS SAID OF ALLAH; 42:48, named in the entry since it was written and carried at no Name; 12:55, Yūsuf claiming the word for himself, also named and not carried; 12:12, THE BROTHERS SAYING THE EXACT WORDS OF 15:9; and 2:255, which the entry and the reflection both lean on while the page carried it nowhere of its own. All six are appended, the six existing sub-items keep their positions, and the distinct Qurʼan-page verse count in Notes on This Edition moves from 393 to 398. ADDED IN BUILD 182, BY THE AUTHOR’S DECISION (A50): 12:63, the brothers saying the same words of Binyāmīn, in his English.
NAMED HERE BUT NOT CARRIED ON THIS PAGE: 50:45, carried at Name 9; 4:80, 6:104, 11:86, 50:4, 50:32 and 85:22, all carried at NO Name; 6:107, which denies this Name and Name 52 in one breath, carried at Name 52.
CORRECTED ON THE RECORD, THE COUNTS, AND THE ENTRY HAD NONE: counted by hand from the Tanzil text with strip, ḥ-f-ẓ gives forty-four words in twenty-eight forms and no look-alike root falls inside the pattern. THE ADJECTIVE ḤAFĪẒ STANDS ELEVEN TIMES AND IS ALWAYS INDEFINITE — THE DEFINITE AL-ḤAFĪẒ IS NOT IN THE QURʼAN AT ALL, and the Name in its definite form comes through the enumeration at at-Tirmidhī 3507. The occurrence field stays attested, which is what the book uses for a word said of Allah without the article, as at Names 39 and 40.
ONLY THREE OF THE ELEVEN ARE SAID OF ALLAH: 11:57, 34:21 and 42:6.
FIVE ARE A DENIAL OR A DISCLAIMER: three tell the Prophet ﷺ he was not sent as a keeper over them, at 4:80, 6:107 and 42:48, and two are a speaker refusing the word for himself, the Prophet at 6:104 and Shuʻayb at 11:86. ONE IS A HUMAN BEING CLAIMING IT AND NOT BEING CORRECTED, Yūsuf at 12:55, in the same shape as a divine pairing. The last two go to the Book at 50:4 and to the penitent servant at 50:32. So the word is withheld from a creature five times for every three it is affirmed of Him, which is the shape this book has learned to look for and the entry had not looked for.
FOUND, AND THE BOOK HAD NOT NOTICED IT: the clause wa-innā lahu la-ḥāfiẓūn stands THREE times, not once. At 15:9 Allah says it of the Reminder. At 12:12 and 12:63 Yūsuf’s brothers say it, first of Yūsuf and then of Binyāmīn, and lose both. Ar-Rāghib, illustrating the root in the Mufradāt, quotes the words, and it is his editor who places them at the brothers’ line and not at the promise of Allah (corrected in build 162). 12:12 is now carried beside 15:9 and the note on each points at the other. CORRECTED, THE READING AT 12:64: the page printed khayrun ḥāfiẓan with no remark, and the readings divide. Aṭ-Ṭabarī records ḥifẓan from most of the readers of Madīnah and some of Kūfah and Baṣrah, and ḥāfiẓan from most of Kūfah and some of Makkah; az-Zabīdī adds that the Kūfans other than Abū Bakr read ḥāfiẓan. On the first, His keeping is the best keeping; on the second, He is the best of keepers. Ar-Rāghib gives both and chooses neither, and neither does this note.
CORRECTED, THE ROOT: glossBasis was desc and senseNote was null; rootRelation was already own and is unchanged, correctly, since no other counted Name shares ḥ-f-ẓ. READ, in one fetch of the Hawramani aggregator, which surfaced for this root with fourteen dictionaries: AL-KHALĪL’S KITĀB AL-ʻAYN, the earliest of them, ar-Rāghib’s Mufradāt, Lane, az-Zabīdī’s Tāj al-ʻArūs, Ibn Sīdah’s Muḥkam, al-Jawharī’s Ṣiḥāḥ, al-Fayyūmī’s Miṣbāḥ, al-Muṭarrizī’s Mughrib, ar-Rāzī’s Mukhtār, al-Jurjānī’s Taʻrīfāt, aṣ-Ṣāḥib ibn ʻAbbād, al-Fattinī, Dozy and Salmone. IBN FĀRIS WAS NOT ON THE AGGREGATOR FOR THIS ROOT and was read instead in two independent transcriptions agreeing word for word, one of them citing the printed pagination. NOT READ in a printed edition. AL-KHALĪL’S DEFINITION IS THE ONE THE PAGE NOW TURNS ON: a ḥafīẓ is the muwakkal, the one entrusted with a thing, who keeps it — and al-Bukhārī 2311 opens with the Prophet ﷺ making Abū Hurayrah muwakkal over the ḥifẓ of the zakāh, which the note had not noticed. Az-Zabīdī glosses the divine Name by quoting 2:255 for it, which is why that verse is now on the page.
REMOVED, AS UNVERIFIED: MEDS 38 attributed the two-motions reading of this Name to al-Ghazālī. AL-MAQṢAD AL-ASNĀ WAS NOT READ, not in a printed edition and not in a quotation with a page, so the name is struck rather than hedged and the sentence now reads that two motions have long been heard in this Name. The first motion, the keeping of the heavens and the earth in being, is independently in az-Zabīdī and in Lane; THE SECOND, THE HOLDING OF OPPOSITES APART, HAS NO SOURCE IN ANYTHING READ FOR THIS PASS and is left standing as a reading the book has carried, flagged here (both motions are al-Ghazālī’s, found in build 162).
CORRECTED, THE REF: it carried one long citation form the house forms forbid, Ṣaḥīḥ al-Bukhārī 2311; it named no lexicon; it gave no count; and it named 42:48 and 12:55 while the page carried neither. ONE duʻāʼ BECOMES TWO. MUSLIM 2708a HAD NO GRADE AT ALL, which the whole-book audit had been counting; it now reads a bare ṣaḥīḥ.
ADDED: Abū Dāwūd 5074 with Ibn Mājah 3871, which puts the Name’s own verb, iḥfaẓnī, on a page that had been carrying a supplication its own note admitted does not contain the root. It is SHARED WITH NAMES 6 AND 7 and keyed reciprocally. TWO aḥādīth BECOME THREE. THE GRADE AT AT-TIRMIDHĪ 2516 CARRIED TWO OF THE DATASET’S FIVE VERDICTS and rendered Darussalam’s as ḥasan where it is a verdict on the chain; all five are now given, and NAMES 7 AND 70, which carry the same report with the same string, are corrected with it.
ADDED: Muslim 809a, ten verses of al-Kahf held in the memory, which is the sense of the root the reflection and the poem are built on and the page had nothing for. POEMS 38 and CH 38 read and left; the seed-viability claim in the story, that a seed kept too long stops being a seed, is sound.
OPEN: no lexicon was read in a printed edition.
OPEN: whether 12:63, the brothers’ second use of the clause, should be carried as well as 12:12 (it is carried from build 182, in the author’s English, A50).
OPEN: the two-motions reading is al-Ghazālī’s in both halves, al-Maqṣad al-Asnā p. 110, read on Shamela’s transcription and not in print; whether the reflection names him is the author’s (A50).
COHERENCE: Names 33, 36, 62 and 63 gain a reciprocal quran link at their 2:255 and no change of text; NAME 37’S STUDY NOTE LISTED THE CARRIERS OF 2:255 AND IS UPDATED; Names 6 and 7 gain a reciprocal duʻāʼ link; Names 7 and 70 gain the corrected grade.
HYGIENE AFTER NAME 61: the Duʻāʼ note on Muslim 2708a narrated the page’s own editing — The grade was missing and is now given: Muslim publishes no grade rows, so the bare verdict is right — and the clause is removed with the fact it carried kept; the Ḥadīth note on al-Bukhārī 2311 narrated the page’s own editing — The printed wording had a silent splice and is repaired: al-Bukhārī runs from recite Āyat al-Kursī through the quoted opening of the verse and then to say the end of it, before fa-innaka lan yazāla, and the page had joined the two ends into a sentence that is not in the source. An ellipsis now marks the omission — and the clause is removed with the fact it carried kept; the Ḥadīth note on at-Tirmidhī 2516 narrated the page’s own editing — The grade recorded here had two of the dataset’s five verdicts and gave Darussalam’s as ḥasan where it is a judgement on the chain; all five are now given, and Names 7 and 70, which carry the same report, are brought into line — and the clause is removed with the fact it carried kept.
AUDIT AFTER NAME 70: SHARED_LOCI attached this page’s links for 2:255 (to Names 33 and 36, and as another portion to Names 62 and 63) to 34:21, so a reader saw them under the wrong āyah and none under 2:255; moved to 2:255. The note to Muslim 809a narrated the page’s own editing (the page had nothing for the sense of the root) and called memory both the root’s other sense and not a second meaning; rewritten, with the recensions in Muslim 809b named exactly: Shuʻbah has the end of al-Kahf, Hammām the beginning. AFTER THE CHAT-END AUDIT OF 81–85: the grade field of Abū Dāwūd 5074; Ibn Mājah 3871 now gives every grader in the dataset’s rows, each attributed to its record (the grade survey).
IN BUILD 144, AFTER THE PHASE 2 READING OF NAME 6: the supplication shared with Names 6 and 7 is cited Ibn Mājah 3871 alone, the record that carries the printed wording, its grade that record’s four rows; its note gives Abū Dāwūd 5074’s allāhumma ḥfaẓnī and four rows, and no longer says what this page lacked, a clause narrating the page’s own editing.
IN BUILD 162, FROM THE PHASE 2 READING: the 42:6 note gives the word withheld from the Prophet ﷺ as wakīl, with aṭ-Ṭabarī’s gloss, and no longer calls the verse the one where ḥafīẓ is affirmed and withheld; the 11:57 note gives Hūd the challenge he made and his people the madness they alleged; the 12:12 note gives aṭ-Ṭabarī’s closing gloss of the promise and three words for four; the 15:9 and 12:12 notes, the ref and this record give the placing of ar-Rāghib’s first example at 12:12 to his editor, the words standing at 12:12, 12:63 and 15:9 alike. The 12:64 note and the ref add aṭ-Ṭabarī’s verdict that either reading is right, the two close in meaning; the 42:48 and 2:255 notes no longer narrate the page’s editing; the at-Tirmidhī 2516 note names every Name carrying the report, 64 and 92 among them, and gives the riding to Musnad Aḥmad, at-Tirmidhī’s wording being behind him; the al-Bukhārī 2311 note says the report is in suspended form; the 86:4 note names the commentators on the keeper; the 18:82 note no longer says who built the wall and gives aṭ-Ṭabarī’s report of the seven fathers; the senseNote gives 85:22’s passive to the tablet in the reading printed and to the Qurʼan in Nāfiʻ’s; the ref cites the supplication by Ibn Mājah 3871, the record carrying the printed wording.
READ ON THE WEB, NOT PINNED: ar-Rāghib’s Mufradāt in two transcriptions of the Dāwūdī edition, whose verse numbers are the edition’s brackets; Musnad Aḥmad’s kuntu radīfa on dorar.net; al-Ghazālī’s al-Maqṣad al-Asnā on al-Ḥafīẓ, p. 110 of al-Jābī’s edition in Shamela’s transcription, where both motions of the reflection are his, the second the more apparent. The reflection’s 15:9, its once and his name wait on the author (A50).
BY THE AUTHOR’S DECISIONS OF 2026-09-25, IN BUILD 182: 12:63 is carried in his English, and the named line and the OPEN item say so in place; MEDS 38 quotes 15:9 in the book’s English and names al-Ghazālī for the two motions, and keeps Once, about one thing (A50); 2:255 takes Name 33’s His Kursi extends over and keeps this page’s preservation, which Names 33 and 36 now give, and the 2:255 note gives az-Zabīdī’s clause in that English (A23).
IN BUILD 207, FROM THE READING OF THE GRADES: Muḥammad Fuʼād ʻAbd al-Bāqī is no longer named as a grader on this page: the row the ḥadīth dataset gives in his name is al-Albānī’s ruling as it is printed with ʻAbd al-Bāqī’s edition of Ibn Mājah, read at Shamela through its page reader, not in print, and the edition gives no verdict of its own beyond quoting az-Zawāʼid; al-Albānī’s grade stands where it stood, and earlier records keep their words.