The Acceptability of H.adīth Narration by Meaning

vs. by Literal Wording

[H.adīth Narration ad Sensum (riwāya bil-ma‘nā) vs. ad Litteram (bil-lafz.)]

by Dr. G.F. Haddad



The H.anafī h.adīth Master Murtad.ā al-Zabīdī began his great commen­tary on the Ih.yā’ with an explanation that al-Ghazzālī’s method of h.adīth citation – conveying the general mean­ing without ascertaining the exact wording – had a basis in the practice of the Com­panions and Salaf:


<<The verification of the wording of narrations was not an obligation for al-Ghazzālī – Allāh have mercy on him! He would convey the general meaning, conscious of the different significations of the words and their mutual conflict with one another avoiding what would constitute interpolation or arbitrary rendering of one term with another.


<<A number of the Companions have permitted the conveyance of Prophetic h.adīths in their meanings (riwāya bil-ma‘nā) rather than their very wordings (riwāya bil-alfāz.). Among them: ‘Alī, Ibn ‘Abbās, Anas ibn Mālik, Abū al-Dardā’, Wāthila ibn al-Asqa‘, and Abū Hurayra  .[1] Also, a greater number of the Successors, among them: the Imām of Imāms al-H.asan al-Bas., al-Sha‘bī, ‘Amr ibn Dīnār, Ibrāhīm al-Nakha‘ī, Mujāhid, and ‘Ikrima…. Ibn Sīrīn said: “I would hear a h.adīth from ten different people, the mean­ing remaining one but the wordings differing.”[2] Similarly, the Companions’ wordings in their narrations from the Prophet have differed one from another. Some of them, for example, will narrate a complete version; others will narrate the gist of the meaning; others will narrate an abridged version; others yet replace cer­tain words with their synonyms, deeming that they have consider­able leeway as long as they do not contradict the original meaning. None of them intends a lie, and all of them aim for truthfulness and the report of what he has heard: that is why they had leeway. They used to say: “Mendacity is only when one deliberately in­tends to lie.”[3]


<<‘Imrān ibn Muslim  [al-Qas.īr] narrated that a man said to al-H.asan [al-Bas.rī]: “O Abū Sa‘īd! When you narrate a h.adīth you put it in better and more eloquent terms than when one of us narrates it.” He replied: “There is no harm in that as long as you have fully expressed its meaning.”[4] Al-Nadr ibn Shumayl (d. 208) said: “Hushaym (d. 183) used to make a lot of mistakes in Arabic, so I adorned his narra­tions for you with a fine garment” meaning, he arabized it, since al-Nad.r was a philologist (nah.wī).[5] Sufyān [al-Thawrī] used to say: “When you see a man show strictness in the wordings of h.adīth, know that he is advertising himself.” He narrated that a certain man began to question Yah.yā ibn Sa‘īd al-Qat.t.ān (d. 198) about a specific wording inside a h.adīth. Yah.yā said to him: “Yā Fulān! There is not in the whole world anything more sublime than the Book of Allāh, yet He has permitted that its words be recited in seven different dia­lects. So do not be so strict!”[6]


<<In the h.adīth Master al-Suyūt.ī ’s commentary on [al-Nawawī’s] al-Taqrīb, in the fourth part of the twenty-sixth heading,[7] the gist of what he said is as follows:

<<If a narrator is not an expert in the wordings and in what shifts their meanings to something else, there is no permission for him to narrate what he heard in terms of meaning only. There is no disagreement concerning this. He must relate the exact wording he has heard. If he is an expert in the matter, [opinions have differed:] a large group of the experts of h.adīth, fiqh, and us.ūl said that it is not permitted for him to narrate in other than the exact same words. This is the position of Ibn Sīrīn , Tha‘lab, and Abū Bakr al-Rāzī the H.a­nafī scholar.[8] It is also narrated as Ibn ‘Umar’s position [as illustrated in the reports of ‘Ubayd ibn ‘Umayr al-Marwazī quoted above].

<<At any rate, the vast majority of the Salaf and Kha­laf from the various groups, among them the Four Imāms, per­mit narration in terms of meaning in all the above cases provided one adduces the meaning.[9] This dispensation is wit­nessed to by the practice of the Companions and Salaf, and shown by their narrating a single report in differ­ent wordings.

<<There is a h.adīth of the Prophet relevant to the issue narrated by Ibn Mandah in Ma‘rifat al-S.ah.āba and al-T.abarānī in al-Kabīr from ‘Abd Allāh ibn Sulaymān ibn Aktham[10] al-Laythī [= ‘Abd Allāh ibn Sulaym ibn Ukayma][11] who said: “I said: ‘Messenger of Allāh! Verily, when I hear a h.adīth from you I am unable to narrate it again just as I heard it from you.’” That is, he adds or omits something. The Prophet replied: “As long as you do not make licit the illicit or make illicit the licit, and as long as you adduce the meaning, there is no harm in that.”[12] When this was mentioned to al-H.asan  he said: “Were it not for this, we would never narrate anything.”[13]

<<Al-Shāfi‘ī [14] adduced as his proof [for the same position] the h.adīth “The Qur’ān was revealed in seven dialects.”[15]

<<Al-Bayhaqī narrated from Makh.ūl that he and Abū al-Azhar went to see Wāthila [or Wā’ila] ibn al-Asqa‘ and said to him: “Narrate to us a h.adīth of the Prophet in which there is no omission, no addition, and nothing forgotten.” He replied: “Has any of you recited anything from the Qur’ān?” They said: “Yes, but we have not memorized it very well. We sometimes add ‘and’ or the letter alif, or omit something.” He said: “If you cannot memorize the Qur’ān which is written down before you, adding and omitting something from it, then how about narrations which we heard from the Prophet , some of them only once? Suffice yourself, whenever we narrate them to you, with the general meaning!”[16] He narrated something similar from Jābir ibn ‘Abd Allāh in al-Madkhal: “H.udhayfa said to us: ‘We are Bedouin Arabs, we may cite a saying with­out its proper order.’” He also narrated from Shu‘ayb ibn al-Hajjab: “I visited al-H.asan together with ‘Abdān. We said to him: ‘Abū Sa‘īd! Someone may narrate a h.adīth in which he adds or from which he omits some­thing.’ He replied: ‘Lying is only when someone deliberately intends this.’”[17] [He also narrated something similar from Ibrāhīm al-Nakha‘ī,[18] al-Sha‘bī,[19] al-Zuhrī,[20] Sufyān,[21]Amr ibn Dīnār,[22] and Wakī‘.[23]]>> End of quotation from al-Suyūt.ī and here also end H.āfiz. al-Zabīdī’s text.[24]


    The Imāms of h.adīth are unanimous in accepting the “narration in meaning” only on condition that the narrator masters the Arabic language and his narration does not present an aberration or anomaly (shudhūdh), among other conditions.[25] Al-Zabīdī’s documentation of the majority position that it is permissible to narrate the h.adīths of the Prophet in their meanings rather than their wordings is also the position of Ibn al-S.alāh. in his Muqaddima, but the latter avers that the dispensation no longer applies at a time when the h.adīths are available to all in published books.[26] Shaykh Nūr al-Dīn ‘Itr adopts the latter position: “The last word on this subject is to prohibit h.adīth narration in the sense of meaning only, because the narrations have all been compiled in the manuals of h.adīth, eliminating the need for such a dispensation.”[27]

–  – Rajab 1423 / September 2002

[1]Al-Khat.īb in al-Jāmi‘ fī Akhlāq al-Rāwī (2:24, 2:26-28) mentions Ibn Mas‘ūd, Abū al-Dardā’, Anas, ‘Ā’isha, ‘Amr ibn Dīnār, ‘Amir al-Sha‘bī, Ibrāhīm al-Nakha‘ī, Ibn Abī Nujayh, ‘Amr ibn Murra, Ja‘far ibn Muh.ammad ibn ‘Alī, Sufyān ibn ‘Uyayna, and Yah.yā ibn Sa‘īd al-Qat.t.ān as allowing the narration of Prophetic h.adīth other than in its precise original wording. He narrates examples from Ibn Mas‘ūd (#1113), Abū al-Dardā’ (#1114-1115), and Anas (#1116-1117) to that effect. He also narrates the prohi­bition of narrating Prophetic h.adīths other than in their precise original wording from Wakī‘ (2:24 #11108), Mālik (2:25 #1110-1111). Al-Khat.īb documents this subject at length in al-Kifāya (p. 203-211).

[2]Also narrated from Abū al-Ah.was. Muh.ammad ibn al-Haytham by al-Khat.īb in al-Jāmi‘ li Akhlāq al-Rāwī (2:21 #1099).

[3]See on this chapter, al-Khat.īb, al-Kifāya (1986 ed. p. 239-247=Madīna ed. p. 204-211).

[4]Narrated by al-Khat.īb in al-Kifāya (1986 ed. p. 243=Madīna ed. p. 207) and al-Jāmi‘ (2:22 #1101-1102). Cf. al-Shāfi‘ī  – without naming al-H.asan or al-Zuhrī – in al-Risāla (p. 275).

[5]Ismā‘īl ibn Umayya said: “We used to correct Nāfi‘ [‘Umar’s freedman] if he made mistakes of lan­guage [in his narrations] but he would refuse and say: ‘Nothing but exactly what I heard.’” Cited by al-Dhahabī in the Siyar (5:567).

[6]Cf. al-Shāfi‘ī, al-Risāla (p. 274).

[7]Al-Suyūt.ī, Tadrīb al-Rāwī fī Sharh. Taqrīb al-Nawawī (1:532-539).

[8]Cf. al-Khat.īb in al-Kifāya (1986 ed. p. 242=Madīna ed. p. 207) who also names Ibrāhīm ibn Maysara, al-Qāsim ibn Muh.ammad, Raja’ ibn Haywa, and Ibn Tawus.

[9]Al-Suyūt.ī, Tadrīb al-Rāwī (1:532-533, cf. Taqrīb p. 77-78). Al-Nawawī continues in his Taqrīb (p. 78 = Tadrīb 1:538): “This holds true in other than h.adīth compila­tions (musannafāt). The alteration of a h.adīth compilation is impermissible, even if in the same sense. Also, it is imperative for the one who narrates in terms of meaning to say, at the conclusion of his narration: ‘or something near it’ aw kamā qāl, aw nah.wahu, aw shibhahu – or other such expressions.” Al-Suyūt.ī adduces proofs that this was the practice of Ibn Mas‘ūd, Abū al-Dardā’, and Anas ibn Mālik. Further proofs to this effect are adduced by al-Tirmidhī in his book al-‘Ilal al-Kabīr and its commentary by Ibn Rajab entitled Sharh. ‘Ilal al-Tirmidhī (1:145-152), al-Khat.īb in al-Kifāya (1986 ed. p. 232-247 = Madīna ed. p. 198-211), and al-Qād.ī ‘Iyād. in al-Ilma‘ (p. 174-178). See also Ibn H.ajar’s discussion and its commentary by al-Qārī in Sharh. Sharh. Nukhbat al-Fikar (p. 497-502).

[10]This is a misspelling in al-Zabīdī’s text.

[11]As stated by Ibn H.ajar in al-Is.āba and Ta‘jīl al-Manfa‘a. Al-Husayni erred in al-Ikmal (p. 565 #1211) when he identified the Ibn Ukayma cited in Ah.mad’s Musnad as ‘Abd Allāh ibn Sulaym ibn Ukayma, as the Ibn Ukayma of the Sunan, the Muwat.t.a’, and Ah.mad’s Musnad is named by al-Tirmidhī in the Sunan – and others – as ‘Umara or ‘Ammār ibn Ukayma al-Laythī. Muslim in his S.ah.īh. (3:1566), Ibn H.ibbān (5:158, 13:238-239), Abū Ya‘lā in his Musnad (12:348), and Ibn ‘Abd al-Barr in al-Tamhīd (17:237) further identify him as ‘Amr ibn Muslim ibn ‘Ammār ibn Ukayma al-Laythī, all agreeing that he is not a Companion, but a Successor who narrated from both Abū Hurayra and Sa‘īd ibn al-Musayyab. As for ‘Abd Allāh ibn Sulaym(an) ibn Ukayma – al-T.abarānī’s narrator – he is unknown.

[12]Narrated from ‘Abd Allāh ibn Sulaymān ibn Ukayma by al-T.abarānī in al-Kabīr (7:100 #6491, 117) and Ibn Qāni‘ (d. 351) in Mu‘jam al-S.ah.āba (3:17), both with a chain containing two unknown narrators – Ya‘qūb ibn ‘Abd Allāh ibn Sulaym(an) ibn Ukayma and his father ‘Abd Allāh ibn Sulaym(an) – as stated by al-Haythamī (1:154), cf. al-Sakhāwī, Fath. al-Mughīth (3:145). Also narrated by al-Jawraqānī (d. 543) in al-Abāt.īl (1:90-97) where he said: “This h.adīth is null and void (bā, and there is confusion (id.t.irāb) in its chain.” Still, al-Khat.īb adduced it through two similar chains in his discussion of the permissibility of narration in terms of meaning in al-Kifāya (1986 ed. p. 234 = Madīna ed. p. 198), as well as al-Qārī in Sharh. Sharh. Nukhbat al-Fikar (p. 498). Also narrated from Salama ibn al-Akwa‘ by Ibn ‘Asākir as stated by Ibn H.amza al-H.usaynī in al-Bayān wal-Ta‘rif (2:77-78). Ibn H.ajar narrates it in al-Is.āba (3:166 #3436, 6:341 #8532) and says: “Ibn al-Jawzī included it among the forgeries, blaming al-Walīd ibn Salama for it, but it is not as he claimed. For Ibn Mandah narrated it [in Ma‘rifat al-S.ah.āba] through another way from ‘Umar  ibn Ibrāhīm, from Muh.ammad ibn Ish.āq ibn Ukayma, from his father, from his grandfa­ther, in similar terms. However, ‘Umar is a contemporary of al-Walīd. Ibn Mandah narrated it through another way from ‘Umar ibn Ibrāhīm, saying: ‘from Muh.ammad ibn Ish.āq ibn ‘Abd Allāh ibn Sulaym.’ He added ‘Abd Allāh in his genealogy. Then he cited it under ‘Abd Allāh’s entry with this chain. It was also narrated by Abū al-Qāsim ibn Mandah in his book al-Was.iyya through two chains going back to al-Walīd ibn Salama, ‘from Ish.āq ibn Ya‘qūb ibn ‘Abd Allāh ibn Ukayma, from his father, from his grandfather.’ There are other discrepancies…. Abū Mūsā in al-Dhayl and Ibn Mardūyah also narrated it in Kitāb al-‘Ilm, both through ‘Abdān al-Marwazī…. I believe some re­shuffling took place and that the correct chain is: Muh.ammad ibn Ish.āq, from ‘Abd Allāh ibn Sulaym ibn Ukayma, from his father, from his grandfa­ther.” In Ta‘jīl al-Manfa‘a (p. 531 #1440) Ibn H.ajar declares Ibn Madah’s chains “flimsy” (wāhiya). Thus he considers the h.adīth weak, but not forged. Its content is confirmed by two other h.adīths of the Prophet adduced by al-Khat.īb, the first: “As long as one adduces the meaning, let him narrate it,” and the second: “I did not mean to prohibit that [one should narrate verbatim], but only that whoever falsely claims that I said something which I did not say, and his purpose is to shame me and smear Islām – or: to smear me and shame Islām.” Narrated respectively from Ibn Mas‘ūd and an unnamed Companion by al-Khat.īb in al-Kifāya (1986 ed. p. 234-235 = Madīna ed. p. 198). From Abū Hurayra: “The Prophet was asked about a man who nar­rates something he said while interchanging the position of clauses or words, and the Prophet replied: ‘There is no harm in it as long as he adduces the meaning.’” Nar­rated by al-H.akīm al-Tirmidhī in Nawādir al-Us.ūl (p. 389). So the mass-transmitted h.adīth narrated in unconditional terms from Salama ibn al-Akwa‘ by al-Bukhārī in his S.ah.īh. (book of ‘Ilm): “Whoever says that I said something which I did not say, let him prepare himself for his seat in the Fire” must be understood in terms of those other h.adīths. This is confirmed by the comments of the Companions and Successors related by al-Zabīdī and the practice of the Salaf as demonstrated by al-H.akīm al-Tirmidhī (p. 389-390, As.l #268) as quoted in full by al-Qāsimī in Qawā‘id al-Tah.dīth (p. 223-224), and Allāh knows best.

[13]Narrated by al-Khat.īb: Jāmi‘ (2:21-22 #1100), al-Kifāya (Madīna ed. p. 207).

[14]In al-Risāla (p. 274).

[15]Narrated from ‘Umar and Ibn ‘Abbās by al-Bukhārī, Muslim, and Ah.mad, and also from Ubay ibn Ka‘b in the Sunan.

[16]Narrated by al-Khat.īb in al-Jāmi‘ (2:20-21 #1098) and al-Kifāya (1986 ed. p. 239= Madīna ed. p. 204). Al-Khat.īb also nar­rates something identical from Qutayba. In al-H.akīm al-Tirmidhī’s version in Nawādir al-Us.ūl (p. 389) Makh.ūl asks: “Has any of you stood in prayer at length at night?”

[17]Narrated by al-Khat.īb in al-Kifāya (1986 ed. p. 244 = Madīna ed. p. 208).

[18]See n. 1.

[19]See n. 1.

[20]Al-Khat.īb, al-Jāmi‘ (2:22 #1103).

[21]Al-Khat.īb, al-Jāmi‘ (2:23 #1104-1106).

[22]See n. 1.

[23]Also H.ammād ibn Zayd as narrated in al-Khat.īb, al-Jāmi‘ (2:23-24 #1107). However, the reports indicate that Wakī‘, like Mālik, forbade al-riwāya bi al-lafz. and insisted on the precise original wording, cf. n. 1.

[24]Al-Zabīdī, Ith.āf al-Sādat al-Muttaqīn (1:48-49).

[25]‘Itr, Manhaj al-Naqd (p. 227-230).

[26]Ibn al-S.alāh., ‘Ulūm al-H.adīth (p. 214).

[27]Nūr al-Dīn ‘Itr, ed., Ibn H.ajar, Sharh. al-Nukhba Nuzhat fī Tawdīh. Nukhbat al-Fikar (p. 95 n. 1). Cf. al-Qāsimī’s Qawā‘id al-Tah.dith (p. 223-225) and T.āhir al-Jazā’irī’s Tawjīh (p. 298-312).