Al-Shāfiʿī Against the Kufan School

One of the short works of al-Shāfiʿī (d. 204/820) is Ikhtilāf ʿAlī wa-ʿAbd Allāh ibn Masʿūd (“the disagreements of ʿAlī and ʿAbd Allāh ibn Masʿūd”). It comprises first quotations of the Companions ʿAlī and Ibn Masʿūd advocating rules that the Kufans reject, secondly counter-reports supporting the rules that al-Shāfiʿī advocates. It seems to be one of the earliest of al-Shāfiʿī’s works. It argues mainly by authority against an undifferentiated Kufan school. It testifies to a time when regional schools were predominant and legal reasoning primitive, also when divisions within Kufa, especially between adherents of raʾy and hadith, were less important than they seemed later. Its argument tends to imply that al-Shāfiʿī’s doctrine is better than the Kufans’ because it is eclectically based on the learning of all centers, suggesting that loyalty to a regional tradition was becoming an embarrassment.

One of the short works appended to al-Shāfiʿī (d.204/820), Kitāb al-Umm, is Ikhtilāf ʿAlī wa-ʿAbd Allāh ibn Masʿūd ("the disagreements of ʿAlī and ʿAbd Allāh ibn Masʿūd").1 Organized topically, it comprises quotations of the Companions ʿAlī and Ibn Masʿūd advocating rules that the Kufans reject, followed by counter-reports supporting the rules that al-Shāfiʿī advocates.By my count, it includes 148 items with isnāds from ʿAlī (46%), 80 from Ibn Masʿūd (24%), 96 counter-reports with isnāds from others (30%).2ʿAlī and Ibn Masʿūd are the Companions most often quoted by Kufan jurisprudents.Joseph Schacht thought that whereas Ibn Masʿūd was usually adduced in support of the majority Kufan position, ʿAlī would often be adduced in support of a minority position, although it also happened that ʿAlī would endorse the majority position, Ibn Masʿūd that of the opposition.3If there was such a dynamic of minority and majority factions within the regional schools, it cannot surprise that al-Shāfiʿī should have found Kufans rejecting the reported positions of ʿAlī and Ibn Masʿūd, and those of ʿAlī more frequently.Unfortunately, we have no literature purporting to represent a Kufan regional school (as perhaps we may say that the Muwaṭṭaʾ purports to represent a Medinese school, periodically justifying this or that rule by appeal to local consensus4).Therefore, we cannot say how emphatically Kufans defended their positions as what ʿAlī and Ibn Masʿūd had taught them.However, al-Shāfiʿī must have thought that dependence on ʿAlī and Ibn Masʿūd was a sufficiently well-advertised position for it to be worth his while to point out where the Kufans went against it, or, conversely, should have gone against it in favor of better opinions from elsewhere.
The evidence for the existence of regional schools at all is first continual references in the sources to the Basrans, the Kufans, the Medinese, and so on.As Schacht observed, al-Shāfiʿī identified certain local Followers (tābiʿūn) whom the jurisprudents of each center characteristically cited as their leading authorities.5The introduction to the Mukhtaṣar of Abū Muṣʿab al-Zuhrī (d.242/857?) is entirely concerned with defending "the people of Medina," not Mālik (d.179/795).6As late as the end of the ninth century, al-Nasāʾī (d.303/915?)Then Norman Calder argued that the Umm was altogether pseudonymous, the product of gradual accumulation of material over the course of the third century H. His principal arguments for redating (which he proposed also for early works in the Mālikī and Ḥanafī traditions) were incoherence in detail, as between early and later segments of particular discussions, and systematic preference for Prophet hadith, which he thought must have prevailed only gradually.13Stylistic inconsistency has been adduced against the direct attribution of the Risāla to al-Shāfiʿī by, among others, Jonathan E. Brockopp and Mohyddin Yahia.14However, they detect reworking nowhere near so extensive as what Calder alleged.Ahmed El Shamsy has made the strongest case for trusting attributions directly to al-Shāfiʿī.His evidence for promulgation of the Umm just as we know it, mainly quotations in other works, is much stronger for the second half of the third century H. than for the first.Whatever room his case leaves for posthumous reworking, however, is likewise nowhere near so extensive as what Calder alleged.15 The organization of the Ikhtilāf is slightly confused.It proceeds from disagreements over ritual ablutions to the ritual prayer, fasting, the alms tax, marriage and divorce, and so on-more or less the usual sequence of law books.But then its last three books treat the alms tax, fasting, and the pilgrimage, seemingly out of place.It is as though some material was retrieved from different notes and attached at the end, perhaps by a later hand than al-Rabīʿ's.There are also small variations in the form of isnāds: they usually begin with akhbaranā l-Rabīʿ qāla akhbaranā l-Shāfiʿī, but they sometimes begin rather with qāla l-Shāfiʿī or directly quoting al-Shāfiʿī's shaykh.Some evident interpolations in the Ikhtilāf are expressly identified as coming from someone later than al-Shāfiʿī, presumably al-Rabīʿ.Following a report that ʿAlī stoned a lūṭī (male homosexual), al-Shāfiʿī says, "We hold to this, stoning the lūṭī whether muḥṣan or not.This is the position of Ibn ʿAbbās and Saʿīd ibn al-Musayyab, who says that the sunna is to stone the lūṭī whether muḥṣan or not" (471-2).16Muḥṣan (lit."safeguarded") refers to the status of a free Muslim who has been married to another free Muslim.Islamic law famously distinguishes between the muḥṣan adulterer, who is liable to be stoned, and the non-muḥṣan, who suffers only flogging and temporary banishment.17After this statement from al-Shāfiʿī, someone else says, "Al-Shāfiʿī went back on this, saying 'He is not stoned unless he has become muḥṣan.' Once, al-Shāfiʿī's opinion is introduced by "Al-Shāfiʿī goes by this," then a restatement of the Kufans' rule (concerning the refund of a nuptial gift), with which they all reportedly agree against ʿAlī (424).After three hadith reports banning the temporary mutʿa marriage, someone says, "Al-Shāfiʿī advocates this" (434).One gloss is explicitly attributed to al-Rabīʿ: "Al-Shāfiʿī's saying 'the least [amount of time] in which a waiting period may expire' applies to one whose menstrual period is 35 days, since the minimum length of menstruation is a day and a night, the minimum length of time between periods fifteen days" (430).19 Al-Rabīʿ reports a contrary account of al-Shāfiʿī's opinion concerning atonement for an unfulfilled vow to perform the pilgrimage walking.According to al-Shāfiʿī by a Basran isnād, ʿAlī allowed riding and an atoning sacrifice if one was unable to walk.The Kufans allow riding and an atoning sacrifice whether one is able or not.Al-Shāfiʿī says that he allows an atoning sacrifice only for someone unable to walk.But then al-Rabīʿ (expressly identified) says, "Al-Shāfiʿī [also] said other than this, mainly that he is to atone for his vow" (421).20 ʿAbd al-Razzāq relates a similar hadith report from ʿAlī by a Kufan isnād according to which he called for riding with a sacrifice for one who was unable.21However, Ibn Abī Shayba relates two versions of ʿAlī's opinion, one by a Kufan isnād, with no condition of inability.22Probably, then, the Kufans did not think they were going against ʿAlī's opinion.Al-Rabīʿ's comment suggests that al-Shāfiʿī himself changed his mind.question, saying he must start the prayer over again if he suffers nosebleed" (Shāfiʿī, Umm, 2:66).19 In the Umm, al-Shāfiʿī says that the shortest menstrual period known to him is a day but expressly allows that a divorced woman who swears that she has passed less than fifteen days between periods is to be believed if she was known to have such frequent periods before her divorce as well: Shāfiʿī, Umm, 6:533.Al-Muzanī reports that al-Shāfiʿī said, in one place, a day, in another, a day and a night, possibly alluding to exactly this difference between the Ikhtilāf and Umm: al-Muzanī, al-Mukhtaṣar, on margin of vols.1-5 of al-Shāfiʿī, al-Umm, 7 vols.in 4 (Bulaq: al-Maṭbaʿa al-Kubrā l-Amīriyya, 1321-5), 5:5.
In the main body of Kitāb al-Umm also there are occasional references to multiple versions of al-Shāfiʿī's opinion known to al-Rabīʿ.23On the whole, however, the Ikhtilāf is stylistically more uniform than other books of the Umm.Consistently, it introduces a quotation with isnād of ʿAlī or Ibn Masʿūd, then quotes the Kufans' rejection of it ("They dislike this and do not advocate it," "In their opinion this is speech that renders the ritual prayer defective," "They do not advocate this, saying one recites the Qur'an only when praying by oneself," and so on), then quotes one or more counter-reports, usually with an isnād, confuting the Kufan position.Unlike the Umm, the Ikhtilāf never switches to dialogue (qāla . . .qultu or qulnā . ..), nor proposes hypothetical objections (in qīla or in qāla qāʾil . . .qīla . ..).Its rudimentary legal argumentation (more on this to come) seems consistent with its predating the rest of the Umm as we know it.Altogether, it seems safe to ascribe it to al-Shāfiʿī except for a few comments added in dictation and allowing for such minor alterations over time as were normal in the manuscript age.

Al-Shāfiʿī's Quotations
Since ʿAlī and Ibn Masʿūd are the Companions most often quoted in support of Kufan positions, al-Shāfiʿī necessarily relies mainly on Kufan traditionists for what they say.At the Follower level, the reports he relates of ʿAlī and Ibn Masʿūd are 11% Basran, 3% Meccan, almost all the rest Kufan.The shaykhs he directly cites most often in this book are, in ascending order, Abū Muʿāwiya (ʿAbbād ibn ʿAbbād, Basran, d.Baghdad, 179/795?), Ismāʿīl ibn ʿUlayya (Kufan, d.Baghdad, 193/809), ʿAbd al-Raḥman ibn Mahdī (Basran, d. 198/814), and Hushaym ibn Bashīr (Wasiti, d.Baghdad, 183/799).For the learning of Kufa, al-Shāfiʿī thus relied heavily on traditionists who collected hadith from more than one center.Such trans-regional movement became common in the Abbasid period as it had not been in the Umayyad.24G. H. A. Juynboll has remarked the particular importance of Wasiti traditionists in the later eighth century for hadith from Basra and Kufa.25 Ismāʿīl ibn ʿUlayya was originally Kufan, married a Basran and transmitted hadith mainly in Basra, finally took up supervision of the maẓālim court in Baghdad and died there.26ʿAbd al-Raḥmān ibn Mahdī was born and died in Basra but evidently traveled, being remembered by Aḥmad ibn Ḥanbal for a sojourn in Baghdad, for his expertise in the hadith of the Kufan Sufyān al-Thawrī (d.161/777?), and for inclining as to law toward Medinese raʾy.27Indeed, twenty-eight of thirty-one quotations of ʿAlī and Ibn Masʿūd through ʿAbd al-Raḥmān ibn Mahdī are related by him from Sufyān.Hushaym, by contrast, quotes in turn twenty-three different shaykhs.Al-Shāfiʿī may well have taken dictation from all four in Baghdad.
A substantial number of hadith reports in the Umm are formally defective on account of anonymous links (not counting quotations of the Prophet without isnād, about a quarter of the total).28An example in Ikhtilāf ʿAlī wa-ʿAbd Allāh ibn Masʿūd is his relation < Sufyān (ibn ʿUyayna) < someone < al-Shaʿbī (ʿĀmir ibn Sharāḥīl, Kufan, d. 104/723-4?):when the remaining heirs are a grandfather, a mother, and a sister, the sister gets three shares, the mother one, and the grandfather two.Al-Shāfiʿī says the Kufans actually go by Zayd's position (i.e., that of the Companion Zayd ibn Thābit), making them nine: to the mother three shares (i.e., a third), the grandfather four, and the sister two (457).This counter-report is anomalous not only for the anonymous link but for not being traced back all the way to either ʿAlī or Ibn Masʿūd.There is a close parallel in the Muṣannafs of ʿAbd al-Razzāq and Ibn Abī Shayba.ʿAbd al-Razzāq reports < someone < al-Shaʿbī, Ibn Abī Shayba < Sufyān < someone < al-Shaʿbī, that he related the opinions of a series of Companions concerning this set of heirs.ʿAlī said the sister gets half, the mother a third, and the grandfather a sixth.Ibn Masʿūd said the sister gets half, the mother a sixth, and the grandfather a third.ʿUthmān said the mother gets a third, the sister a third, and the grandfather a third.Zayd said there are nine shares: to the mother three, to the grandfather two-thirds of what is left, and to the sister a third.Ibn ʿAbbās said the mother gets a third, the remainder going to the grandfather, nothing going to the sister.29 ʿAbd al-Razzāq and Ibn Abī Shayba also relate the same summary from Ibrāhīm al-Nakhaʿī (d.96/714), reputedly al-Shaʿbī's great rival in Kufa.30Ibn Abī Shayba's version may be seen as confirmation of al-Shāfiʿī's: that Sufyān ibn ʿUyayna indeed did not name his source for al-Shaʿbī's summary.If so, the puzzle becomes ʿAbd al-Razzāq's apparently hearing this from someone he does not name who heard it directly from al-Shaʿbī.31It also remains a puzzle why al-Shāfiʿī traces ʿAlī's proposed division only to al-Shaʿbī, not ʿAlī (but this may be a scribal error).32 Schacht draws heavily on the Ikhtilāf in his treatment of ʿAlī in the early development of Islamic law.
Most of the opinions advanced under the authority of ʿAlī remained unsuccessful, but some succeeded in gaining recognition . . . .Generally speaking, traditions from ʿAlī are as typical of unsuccessful opinions of the Iraqian opposition as those from Ibn Masʿūd are of the normal doctrine of the school of Kufa; this appears from the contents of Tr. ii, compared with those of Āthār A. Y. and Āthār Shaib.
The "unsuccessful" ʿAlī traditions in Tr. ii show often a rigorous and meticulous tendency, obviously inspired by religious and ethical considerations . . ..33Such a division of labor between Ibn Masʿūd and ʿAlī would account for al-Shāfiʿī's naming many more reports from ʿAlī than Ibn Masʿūd: he wants to pile up instances of Kufan disagreement with these two Companions, and naturally this is easier with the Companion commonly cited for minority opinions.
However, it is difficult to generalize about the rigorism of ʿAlī's opinions.Sometimes they do seem harsh or scrupulous.The examples of the restrictive version of a pronouncement about atoning for an unfulfilled vow and ʿAlī's calling for all convicted homosexuals to be stoned have been mentioned already.ʿAlī is said to have required the witnesses themselves to cut off a convicted thief's hand, apparently to discourage false testimony (465).Through Sufyān al-Thawrī it is said that ʿAlī prescribed, for someone who confessed to ejaculating while dreaming about someone else's wife, that he be made to stand in the sun and his shadow beaten (468-9).
By contrast, others of ʿAlī's preferences seem relatively lenient.Through a Basran isnād, he reportedly named just four chapters of the Qur'an requiring prostration at some point (Q.32, 40, 53, and 96), which al-Shāfiʿī rejects, saying there are about twice this many (415).34By a Kufan isnād going through al-Shaʿbī, he says, concerning the Christian whose wife converts to Islam, "He has a greater claim to her so long as he does not take her out of dār al-hijra" (423-4).Al-Shāfiʿī rejects this opinion, too, saying that no one goes by it.(Fairly widespread support for it is indicated by ʿAbd al-Razzāq and Ibn Abī Shayba.The latter relates a similar pronouncement from ʿAlī by a mostly Basran isnād, both of them similar pronouncements by Kufan isnāds from both al-Shaʿbī and Ibrāhīm al-Nakhaʿī.This was probably at one time the majority opinion in Kufa.35)

Al-Shāfiʿī's Opponents
Al-Shāfiʿī complains continually about what opponents uphold or reject without ever naming them.I have been referring to them as the Kufans; that is, the regional school of his time.This is to disagree with Ahmed El Shamsy, who calls Ikhtilāf ʿAlī wa-ʿAbd Allāh ibn Masʿūd "his own original attack on Ḥanafism" (as opposed to al-Radd ʿalā Muḥammad ibn al-Ḥasan, an attack in the form of a commentary on an earlier Ḥanafi work for identifying al-Shāfiʿī's opponents as specifically the Ḥanafiyya.However, after pointing out a case in which al-Shāfiʿī and his opponents agree contrary to the reported opinion of ʿAlī, El Shamsy says, "Al-Shāfiʿī criticizes the Ḥanafīs' divergence from ʿAlī's precedent: given that they provide no countervailing evidence to refute the report from ʿAlī, he argues, 'if they consider it authentic then they are bound by it.'"37 The passage in question (like the whole work) does not actually mention the Ḥanafiyya in particular.Actually, there are two passages in question, El Shamsy having read the beginning of one and the conclusion of the other (a regrettable mistake but not materially affecting his argument).The first question, expressly described by El Shamsy, is about whether a concubine who has given birth to her master's child (umm al-walad) is freed on his death.Ibn Abī Shayba confirms al-Shāfiʿī's version of the report from ʿAlī that he decided, against ʿUmar, that she is not automatically freed.38But Ibrāhīm al-Nakhaʿī, who usually represents the majority opinion of the Kufan regional school, said she is freed, according to al-Shaybānī,39 or endorsed ʿUmar's rule that she is freed, according to ʿAbd al-Razzāq, Ibn Abī Shayba, Abū Yūsuf, and (in a second version) al-Shaybānī.40So al-Shāfiʿī seems to misrepresent the basis of the Kufans' preference for freedom; that is, they presumably thought they were following Companion opinion, not just raʾy.Be it observed also that the Ḥanafiyya are doubtfully the only Kufans who rejected ʿAlī's opinion.
The second question, to which El Shamsy's quotation about being bound directly applies, concerns the sale of an animal with the condition that some part be returned to the seller after slaughter.Al-Shāfiʿī relates ʿAlī's rejected opinion ("They say it is an invalid sale") through ʿAbd al-Raḥmān ibn Mahdī from Sufyān al-Thawrī.ʿAbd al-Razzāq relates a similarly worded report directly from Sufyān al-Thawrī.ʿAbd al-Razzāq's next item is another description of such a sale, followed by "Al-Thawrī said, 'We say the sale is invalid.'"41Here and elsewhere, al-Shāfiʿī's attack on his opponents is not specific to the Ḥanafiyya seems more likely that the later Ḥanafī tradition tended to suppress discrepant reports than that Abū Ḥanīfa and the two shaykhs were singularly resistant to changing their opinions and to being quoted by conjecture.Similarly, al-Shāfiʿī says that his opponents find fault with their master (ṣāḥibuhum) as to whether a fatal assault with a large rock or a wooden club may be considered manslaughter, not murder (445).According to al-Shaybānī, al-Āthār, Abū Ḥanīfa held that fatal assault with a weapon was a matter for qiṣāṣ (retaliation, potentially fatal to the perpetrator) but with anything else a matter for the kin group (that is, to pay the wergild for manslaughter)."We also go by this," comments al-Shaybānī, "except for one characteristic: if he strikes him with something else than a weapon that takes the place of a weapon, it requires retaliation (qawad), this being Abū Yūsuf's position and ours."47This might be taken as the fault-finding al-Shāfiʿī refers to.However, the major Kufan authorities al-Shaʿbī, al-Ḥakam ibn ʿUtayba (d.113/731-2?),Ḥammād ibn Abī Sulaymān (d.120/737-8?),and, most importantly, the usual exponent of the Kufan position, Ibrāhīm al-Nakhaʿī, are likewise quoted as requiring retaliation.Therefore, it seems likely that those who disagreed with Abū Ḥanīfa on this point included more than Abū Yūsuf and al-Shaybānī.48"Their master" is said to hold that a surviving grandfather excludes surviving brothers from inheriting, contrary to a report that ʿAlī would allot him only a sixth, whereas al-Shāfiʿī agrees with those Kufans who divide an estate between grandfather and brothers such that the grandfather gets no less than a third (453)."Abū Bakr's view that the grandfather completely excludes all brothers and sisters of the deceased was endorsed by Abū Ḥanīfa and subsequently became the authoritative doctrine of the Ḥanafī school."49This endorsement is surprisingly hard to document from earliest Ḥanafī sources.Finally, Abū Ḥanīfa apparently comes up in relation to the question mentioned above of how to punish homosexual adultery, concerning which early Shāfiʿī and Ḥanafī literature both show some instability.As observed above, al-Shāfiʿī advocates stoning whether the culprit is muḥṣan or not.By contrast, he says, Their master (ṣāḥibuhum) says the lūṭī does not suffer the ḥadd penalty.If he engages in homosexual sex when in a sacral state, it does not violate his sacral He need not perform the major ritual ablution if he does not ejaculate.One of his followers (baʿḍ aṣḥābihī) disagrees with him on this, for he says the lūṭī is like an adulterer, to be stoned if he is muḥṣan and flogged if not muḥṣan.The lūṭī is in no worse condition than an adulterer (472).
Al-Jāmiʿ al-ṣaghīr confirms that Abū Ḥanīfa's opinion was at least close to what al-Shāfiʿī ascribes to "their master": "A man does the work of the people of Lot: he is chastised (yuʿazzaru) and imprisoned.Abū Yūsuf and Muḥammad held that he suffers the ḥadd."53The classical Ḥanafī position is evidently that homosexual sex is not adultery (zinā), either because it does not threaten confusion of genealogy (ikhtilāṭ al-nasab) or because it does not meet the precise definition of adultery, namely genital sex between two persons for whom it would be permissible if properly related by marriage or ownership.54The  56 However, the allusions to Abū Ḥanīfa and even less often to his most famous followers, reviewed above, are the only acknowledgement in Ikhtilāf ʿAlī wa-ʿAbd Allāh ibn Masʿūd of any smaller unit than the usual anonymous third-person plural.They are reminiscent of the way al-Shaybānī refers to Mālik in al-Ḥujja ʿalā ahl al-Madīna as "their master" without ever implying that his polemic is actually against a Mālikī school.In Ikhtilāf ʿAlī wa-ʿAbd Allāh ibn Masʿūd, al-Shāfiʿī continually refers to "them" as a homogeneous bloc.
Al-Shāfiʿī does not consistently call attention to disagreement between Abū Ḥanīfa and his followers.For example, by a Kufan isnād immediately through Sufyān (ibn ʿUyayna), he quotes ʿAlī as calling for the dawn prayer as soon as the darkest part of the night is past, called taghlīs."These two reports from ʿAlī," he says, "show that he practiced taghlīs of the most extreme kind.They disagree with him and advocate waiting till the dawn is bright," called isfār (397).In the Umm, al-Shāfiʿī quotes the Prophet through Mālik calling for prayer at ghalas (the dark of the night).57ʿAbd al-Razzāq indicates that ʿAlī could be quoted through Sufyān al-Thawrī in favor of waiting for light.58Ibn Abī Shayba strongly suggests that this was the majority Kufan position, reporting among other things a pronouncement from Ibrāhīm al-Nakhaʿī that the Companions agreed on nothing so strongly as waiting for light.59According to the Ḥanafi tradition, Abū Yūsuf said the dawn prayer is permissible anywhere in the second half of the night, in agreement with al-Shāfiʿī.60But al-Shāfiʿī does not point out this internal disagreement.(Possibly, Ḥanafī doctrine was still in flux across the ninth century.Al-Ṭaḥāwī reports no such disagreement between Abū Ḥanīfa and Abū Yūsuf, saying only that the best time for the dawn prayer somehow combines taghlīs and isfār and that Abū Yūsuf, contra Abū Ḥanīfa and al-Shaybānī, allowed the dawn adhān before the time of the dawn prayer.61) Similarly, al-Shāfiʿī reports by three isnāds, the first all-Kufan, that ʿAlī required a muʾlī, someone who has sworn not to have sex with his wife for four months or longer, to be "suspended" (yūqafu), meaning that after four months he is not automatically divorced but made to choose expressly between taking back his wife and divorcing her.He comments at the end, "They disagree with him, saying that he is not suspended: when four months have passed, she is separated from him" (426-7).As in the cases of an unfulfilled vow and praying well before dawn, there is disagreement over ʿAlī's position.62It is indeed the Ḥanafī rule that the wife is automatically divorced.63ʿAbd al-Razzāq reports that Sufyān al-Thawrī required a separate pronouncement of divorce, against the Ḥanafī position, but al-Shāfiʿī makes nothing of it.64 Al-Shāfiʿī often proves an unreliable witness to Kufan opinion.For example, he reports by a Basran/Kufan isnād that when a man came to Ibn Masʿūd with a sealed purse he had found, which he had advertised without finding anyone who recognized it, Ibn Masʿūd said, "Enjoy it."Al-Shāfiʿī says, "We go by this but they go against it" (452).Indeed, early Ḥanafī literature quotes ʿAlī as calling for giving away property as alms after no one has claimed it for a period.65Sufyān al-Thawrī is quoted as saying, "If you are afraid of its perishing, give it away as alms."66However, al-Shaybānī also quotes Ibrāhīm al-Nakhaʿī as saying, "Giving Another example concerns the salutation at the end of the ritual prayer.Al-Shāfiʿī relates by two isnāds that ʿAlī saluted to left and right salāmun ʿalaykum, salāmun ʿalaykum.Al-Shāfiʿī comments, "They do not go by it, adding wa-raḥmatu Llāhi wa-barakātuh" (399-400).Al-Shāfiʿī himself does not go by it, calling for al-salāmu ʿalaykum as the minimum, one particle less requiring repetition.69ʿAbd al-Razzāq and Ibn Abī Shayba both relate that ʿAlī said al-salāmu ʿalaykum wa-raḥmatu Llāh but not salāmun ʿalaykum without alif-lām.70(The argument here is not that their reports are more reliable than al-Shāfiʿī's, only that Kufan practice did not obviously go against ʿAlī's.)ʿAbd al-Razzāq also mentions a report that Ibn Masʿūd saluted al-salāmu ʿalaykum wa-raḥmatu Llāhi wa-barakātuh, a precedent for the Kufan practice that al-Shāfiʿī alleges.71Abū Dāwūd relates by a Kufan isnād that the Prophet ended his prayer with wa-raḥmatu Llāhi wa-barakātuh.72However, the Ḥanafiyya evidently called for only al-salāmu ʿalaykum wa-raḥmatu Llāh.73 Al-Tirmidhī identifies this as the position of Sufyān al-Thawrī, too.74Again, the Kufans' preference seems to have been exactly what al-Shāfiʿī says they reject (unless the Kufans he has in mind here are other than the Ḥanafiyya and Sufyān al-Thawrī).
The Kufans often had different versions of what ʿAlī and Ibn Masʿūd stood for.Several examples have come up already, concerning the timing of the dawn prayer, an unfulfilled vow to make the pilgrimage on foot, the īlāʾ divorce for ʿAlī, and the wording of the final salutation.Another example concerns coitus interruptus: according to al-Shāfiʿī, Ibn Masʿūd condemned it as al-waʾd al-khafī, meaning comparable to (if less violent than) the pre-Islamic custom of burying alive an unwanted child (430-1).ʿAbd al-Razzāq relates something similar from both Ibn Masʿūd and ʿAlī, Ibn Abī Shayba from ʿAlī.75However, ʿAbd al-Razzāq also relates through both Sufyān al-Thawrī and Abū Ḥanīfa that Ibn Masʿūd saw no harm in coitus interruptus.76Ibn Abī Shayba relates the opinion from Ibrāhīm al-Nakhaʿī and Ibn Masʿūd that it is permissible with a concubine but requires the permission of a free wife.77Al-Shaybānī bases his and Abū Ḥanīfa's acceptance expressly on Ibn Masʿūd's saying that one may practice it or not as one wishes.78Thus, al-Shāfiʿī was right that the Kufans allowed coitus interruptus but doubtfully that they were doing so in deliberate defiance of Ibn Masʿūd's opinion.

Al-Shāfiʿī's Counter Arguments
As related above, a little over two-thirds of Ikhtilāf ʿAlī wa-ʿAbd Allāh ibn Masʿūd comprises pronouncements from ʿAlī and Ibn Masʿūd in favor of rules not followed by the Kufans, a little under a third counter arguments.Al-Shāfiʿī fairly often adduces Companion opinion in his favor.Against ʿAlī, he remarks without an isnād that Ibn ʿUmar held, along with Ibn Masʿūd, that kissing requires renewal of ritual ablutions (394)79 and that ʿUmar gave the Friday khuṭba only after the zenith (406).As to whether the muʾlī's wife is automatically separated from him after four months, al-Shāfīʿī adduces along with his three reports from ʿAlī the reported opinions of ʿUmar, Ibn ʿUmar, ʿĀʾisha, ʿUthmān, Zayd ibn Thābit, "and ten-odd Companions of the Messenger of God" (427).As for the geography of his counter-reports, a few have such mixed isnāds that it is doubtfully wise to identify them with any particular region.However, about two-fifths may be characterized as Medinese in their upper reaches (earlier in time), about a third as Kufan, an eighth as Meccan, fewer still Basran.An exact count is difficult, but it seems that al-Shāfīʿī most often cites against the Kufans the Prophet together with one or more Companions, almost as often the Prophet by himself, then almost as often one or more Companions without the Prophet.He sometimes calls for following one Companion opinion instead of another because it agrees with the Prophet's precept or example.Concerning an opinion of ʿAlī's rejected by the Kufans, he says, "We go by it because it agrees with the established sunna of the Messenger of God" (462).Concerning a rule supported by an opinion of Ibn Masʿūd's, he says, "We encourage this and advocate it, since it agrees with what is related of the Prophet . . ., whereas they strongly discourage it" (499).
Only occasionally does al-Shāfiʿī touch on legal theory.Concerning whether penetration without ejaculation requires the major ritual ablution, al-Shāfiʿī quotes Ibn Masʿūd as holding it does not, a position neither he nor the Kufans go by.He goes on to say, "This position [obtained] at the beginning of Islam, then was abrogated" (394-5).He seldom adduces analogy (qiyās).At one point, apropos of what nullifies a ritual prayer, he says, "If they affirm this narration [of ʿAlī's opinion], then they must say about rumbling of the intestines (rizz) what they say about nosebleed, since I know of nobody else of the Prophet's Companions who contradicts him as to rumbling of the intestines" (397).But this objection is about following a hadith report more than logical consistency between different rules.Al-Shāfiʿī once observes that the Kufans themselves invoke qiyās, pretending to extend the law of inheritance shares (farāʾiḍ) by analogy from Zayd ibn Thābit's scheme when in the present instance they do something else (454).
Al-Shāfiʿī's strongest appeal to analogy has to do with a story by which, after ʿAlī had ordered a convicted thief's hand cut off, the witnesses identified someone else as the thief.ʿAlī fined them but added, "If I knew you had done this deliberately, I would cut off your [hands]."80Al-Shāfiʿī says he agrees with this.The Kufans say of the false witnesses, "We will not cut off two hands for one."Al-Shāfiʿī says that his position is the better analogy (ashbah bi-l-qiyās), meaning the more consistent: "If it is permissible that two be killed for one, why should not two hands be cut off for one?"(465-6).(He uses qiyās here as a student of al-Shaybānī, not Mālik.81)The Ikhtilāf can yet be useful to us for illustrating the state of Islamic legal thought in its time; that is, at the turn of the ninth century ce.For one thing, the tradition of regional schools was still sufficiently vital that it made sense for al-Shāfiʿī to polemicize against the Kufans, not individuals whether named or not.This is to disagree with Wael Hallaq, who argues by reductio ad absurdum that there cannot have been regional schools in the eighth century because the term implies that "doctrine was anonymous (for there was no personal allegiance), homogeneous, and closely affiliated with a particular region."86This insistence on anonymity and homogeneity means that Hallaq is mostly trying to knock down a straw man of his own invention, not Schacht's or anyone else's actual description of the eighth-century regional schools.Still, anonymity, homogeneity, and close affiliation with a particular region do usually characterize the Kufans as al-Shāfiʿī represents them in Ikhtilāf ʿAlī wa-ʿAbd Allāh ibn Masʿūd.I should also say precisely where my interpretation differs from Ahmed El Shamsy's.This has to do first with the connection of the emerging personal schools with the regional schools that preceded them.According to El Shamsy, "Ḥanafī thought . . .retained a regional specificity, rooted in the Kufan tradition, that precisely mirrored the Medina-centrism of Mālik."87It is certainly true that Ḥanafī positions tend to preserve earlier Kufan traditions, as Mālikī positions tend to preserve earlier Medinese.88In the late eighth century, however, al-Shaybānī is much more inclined than Mālik to adduce opinions from other centres.Al-Ḥujja ʿalā ahl al-Madīna often and al-Shaybānī's edition of al-Muwaṭṭaʾ usually assert that Abū Ḥanīfa's views are supported by those of Followers and Companions not identified with Kufa, whereas Mālik almost never bothers with Followers and Companions not identified with Medina.89In short, El Shamsy's expression "precisely mirrored" goes too far.
Secondly, of course, El Shamsy characterizes Ikhtilāf ʿAlī wa-ʿAbd Allāh ibn Masʿūd as a polemic specifically against the Ḥanafiyya, not Kufans more generally.There are three dozen passages in the Umm headed al-khilāf fī . . .("disagreement concerning" some part of the law) that record debates (presumably idealized) between al-Shāfiʿī and baʿḍ al-nās ("a certain person"), evidently al-Shaybānī.90But whereas El Shamsy considers Ikhtilāf ʿAlī wa-ʿAbd Allāh ibn Masʿūd to be in continuity with those passages, I would say only that it sometimes anticipates them.We are dealing with a transition period, out of the regional stage and into the personal, so in the nature of things sharp lines 87 El Shamsy, Canonization, 47. 88 "The bulk of the ancient school of Kufa transformed itself into the school of the Ḥanafīs, and the ancient school of Medina into the school of the Mālikīs": Joseph Schacht, An Introduction to Islamic Law (Oxford: Clarendon Press, 1964), 57.89 See further Christopher Melchert, "Kitāb al-Ḥujjah ʿalá ahl al-Madīnah and the Transition from Regional Schools to Personal," forthcoming in Studia Islamica.90 Al-Rabīʿ once comments, "When he says 'a certain person,' they are Easterners (mashriqiyyīn).
2 ʿAbd al-Muṭṭalib counts over 100 more items than I do, 346 in the whole work, but the discrepancy is due mainly to his including items without isnād.3Joseph Schacht, The Origins of Muhammadan Jurisprudence (Oxford: Clarendon Press, 1950), 240-2.4Medinese practice (ʿamal) is stressed by Yasin Dutton, The Origins of Islamic Law: The Qur'an, the Muwaṭṭaʾ and Madinan ʿAmal, Culture and Civilization in the Middle East (Richmond, Surrey: Curzon, 1999).Strictly speaking, however, the Muwaṭṭaʾ refers to agreement and disagreement more often than practice.Consider also the contemporary testimony of al-Shāfiʿī, K. Ikhtilāf Mālik wa-l-Shāfiʿī, where al-Rabīʿ apparently describes the Muwaṭṭaʾ as the book they were used to following "in which is al-amr al-mujtamaʿ ʿindanā and al-amr ʿindanā," not mentioning ʿamal ahl al-Madīna.Al-Shāfiʿī's rebuttal stresses disagreement within Medina (in Umm, 8:771).5 Schacht, Origins, 7-8; Shāfiʿī, Umm, 8:763.Recently stressed anew by Sohail Hanif, "A Tale of Two Kufans: Abū Yūsuf's Ikhtilāf Abī Ḥanīfa wa-Ibn Abī Laylā and Schacht's Ancient Schools," Islamic Law and Society 25 (2018), 173-211, esp.188-92, defining the regional schools by their peculiar sets of authorities although not limiting these to Followers.Tasmiyat fuqahāʾ al-amṣār min aṣḥāb Rasūl Allāh wa-man baʿdahum classifying jurisprudents and traditionists (scarcely distinguished by him) geographically.7 Al-Shāfiʿī and his followers appear there among the people of Mecca.Al-Shāfiʿī himself respects Mālik (d.179/795) but occasionally complains that he fails to represent Medinese consensus as he says he does; that is, al-Shāfiʿī himself is still largely writing from a regional perspective.8Further evidence for the existence of regional schools is the tendency of majorities in each center to form behind opposed rules, with some tendency for Basran and Medinese opinion to agree, also more weakly for Kufan and Meccan.This evidence has become much more accessible with the publication of the Muṣannafs of ʿAbd al-Razzāq (Yemeni, d. 211/827) and Ibn Abī Shayba (Kufan, d. 235/849), not available to Schacht.9The present article tends to confirm the Schacht identified it as one of the oldest of the short works, brought to Egypt from Iraq, along with al-Radd ʿalā Muḥammad ibn al-Ḥasan, apparently a section of al-Ḥujja ʿalā ahl al-Madīna by al-Shaybānī (d.189/804-5) overlaid by counter-arguments from al-Shāfiʿī, and Ikhtilāf al-ʿIrāqiyyayn, apparently an assemblage by Abū Yūsuf (d.182/798) of the positions of Abū Ḥanīfa (d.150/767) and Ibn Abī Laylā (d.148/765) on various questions with commentary by him, overlaid by more commentary from al-Shāfiʿī.12 Attributions to al-Shāfiʿī have become controversial.Schacht identified the longest of the short works, Ikhtilāf al-ḥadīth, as a posthumous assemblage.
This is another question on which the later Ḥanafī school appears to have departed from the earlier Kufan consensus.Al-Shaybānī expressly rejects Ibrāhīm's distinction between kissing those whom one is forbidden to marry and those whom one is allowed to: Shaybānī, Āthār, 1:59-60.Abū Yūsuf quotes Ibn ʿUmar, along with al-Ḥasan al-Baṣrī and Ibn ʿAbbās, against requiring the minor ritual ablution: Abū Yūsuf, Āthār, 5, nos.16-18.
80 Al-Shāfiʿī's isnād is Kufan, going back to al-Shaʿbī through Sufyān ibn ʿUyayna.ʿAbd al-Razzāq quotes the same story by a wholly different isnād, Muṣannaf, 10:88.81 On Paul Gledhill's showing that Mālik had no use for the concept, for which see Paul .84 Contemporary Sunni hadith critics (although even more of the generation after) proposed to sort the more and less reliable reports by isnād comparison.Al-Shāfīʿī advocates hadith criticism by isnād comparison in the Risāla, but in Ikhtilāf ʿAlī wa-ʿAbd Allāh ibn Masʿūd he hints that consensus should decide which hadith to follow.Shāfiʿī's legal theory is much more logical and formally consistent than that of his predecessors whom he blames continually for what appears to him as a mass of inconsistencies.Explicit legal reasoning occupies a much more prominent place in Shāfiʿī's doctrine than in that of any of the earlier lawyers, even if we take differences of style and literary form into account.85His generalization does not apply to Ikhtilāf ʿAlī wa-ʿAbd Allāh ibn Masʿūd, which includes very little explicit legal reasoning.On examination, it may be the least impressive of al-Shāfiʿī's works, crudely confronting report with report, not attempting to demonstrate its superior authority (except inasmuch as the sunna of the Prophet trumps pronouncements of Companions), continually misrepresenting the Kufan position, and continually ignoring Kufan alternatives to the Kufan hadith he complains that they fail to follow.