Origins of the ghiyar.
This conclusion rests primarily upon two claims respecting the evidence for the first promulgation of the ghiyar: its unanimous attribution to Umar [b. Abd al-Aziz] by the sources, and their consistency regarding its contents" (p. 92). In what follows I will try to show that the sources in fact contain many plausible attributions of the ghiyiir to rulers both earlier and later than Umar II and that reports of a ghiyar edict under him are inconsistent in most of their details. Consequently, I will argue that the evidence does not support the proposition that Umar II formulated and first implemented the ghiyar. Indeed, at the present time a more defensible position might be that the evidence for the origins of the ghiyar is, like that for much else in the early history of Islam, intractable.
Levy-Rubin repeatedly informs her readers that all of the Muslim sources that mention the origins of the ghiyar attach them to Umar II. For example, "the Muslim sources are correct in attributing the first code regarding the attire and behaviour of non-Muslims in Muslim society to the caliph Umar b. Abd al-Aziz" (p. 88); "[t]he sources all point in one direction: they all attribute the creation of the ghiyar to the caliph Umar b. 'Abd al-Aziz" (p.89); "an examination of the sources supports the traditional claim that it was Umar b. (Abd al-cAziz who institutionalized the use of ghiyar" (p. 97); "according to Muslim tradition itself, it is only in the days of cUmar b. 'Abd al-'Aziz that demands for distinctive dress began" (p.127); "[t]he process of adoption [...] began with cUmar b. cAbd a1-Azlz, who took the initial steps and issued the first edict listing a set of demands regarding the appearance of non-Muslims in public, the ghiyar" (p. 168).
It is quite true that several sources credit Umar II with measures of this sort (without claiming, however, that he was the first to promulgate the ghiyar). (4) The statements quoted above are nevertheless problematic. The countervailing evidence includes that presented by Abii Hilal al-cAskari (d. 395/1005), who cites in Kitab al-Awa'il:
The first to order dhimmis to distinguish (taghyir) their . dress was al-Mutawalckil. Abu Ahmad informed us, from a1-5511, saying: Al-Mutawakkil ordered dhimmis to wear honey-colored clothing, and to ride upon [wooden saddles1,5 and to place a button (jrr) at the front of the saddle and at its back, and upon the conical cap (yalansitwa), and upon the outer garment patches before and behind, and upon their doors wooden likenesses (movar). (6)
Here al-Askari states explicitly that it was the 'Abbasid caliph al-Mutawakkil (r. 23247/847-61) who first instituted the ghiyar. The express purpose of the awed genre, in the most famous example of which we find this statement, was to identify the first person to have said or done a particular thing. In this the awei'd evidence differs from the other testimonia relevant to the question of origins. It might thus be granted a certain weight on the matter of who instituted the ghiyar.
However, there is a much greater quantity of evidence that would lead us to date the "ghiyar code- to the reign of (Umar II's maternal grandfather Tmar b. al-Khattab (Umar I), the second "rightly guided" caliph who ruled nearly eighty years before. The ghiyar is ascribed to cUmar I in numerous and diverse sources, quite apart from the "Pact of Umar," which purports to describe ghiyar restrictions in his day but which in the form we have it must date from a considerably later time. In Kitab al-Amwal of Abu '1Jbayd al-Qasim b. Sallam (d. 224/838) is found:
Abd al-Rahman related to us, from Abd Allah b. Umar, from Nafi, from Aslam, that Umar commanded concerning dhimmis that they dock their forelocks, ride upon pack saddles, ride side-saddle, not ride as the Muslims ride, and be sure to wear the belts (manatiq; or "girdles"). Abu cUbayd said: That means the zananir [sg. zumnar "belt"]. Al-Nacir b. Isma'11, from (Abd al-Rahman b. Istfaq, from Khalifa b. Qays, [who] said [that] (Umar said: 0 Yarfa', write to the people in the garrison cities (ahl al-amtvar) concerning the People of the Book: that they clip their orelocks, and fasten belts (kustijan; or "girdles") around their waists, so that their clothing is known from that of Muslims (ahl al-islam). (7)
There can be no doubt that the Umar here is Umar I--Aslam and Yarfa' were his associates, cAbd Allah b. (Umar (d. 73/693) his son. Nor can it be questioned that we have here to do with the principle of ghiyar--the visible differentiation of non-Muslims--in its three major areas of concern as expressed also in the "code" ascribed elsewhere to Umar 11: hair, dress, and manner of riding animals. As Abu. Yusuf (d. 182/798) wrote in Kitab al-Kharal after laying out his own ghiyar prescriptions, -in this manner Umar b. al-Khattab commanded his agents to require dhimmis to dress this way, saying: In order that their clothing be distinguished from that of the Muslims." (8) The later Ibn Zanjawayh (d. ca. 251/865) concurs in Kirab al-Amtv(71: "Al-Hushaym b. 'As reported: Muhriz Aba Raja reported to us, from Makhal, that (Umar b. al-Khattab ordered dhimmis to clip their forelocks, fasten on their belts (awsiit), and not to resemble the Muslims in any of their affairs." (9)
Similar measures are attributed to 'Umar I in Fatah Misr of 'Abd al-Rahman Ibn Abd al-Hakam (d. 257/871):
Then (Umar b. al-Khattiib wrote--as related to us by (Abd al-Malik b. Maslama, from al-Q5sim b. cAbd Allith, from CAbd Al1h b. Dinar, from Abd AllAh b. cUmar--that.the necks of dhimmis be sealed (yukhtam) with leadm (10) and their girdles worn in plain view (yazharumanatiquhum), that they dock their forelocks and ride side-saddle upon pack saddles [. 1 and let them not resemble the Muslims in their clothing (yatashabbahana bi-l-mustimina fi labusihim). (11)
Iconic later works of Islamic political and legal thought also credit the ghiyeir to Umar I. From Siraj al-mulak of al-Turtilshi (d. 520/1126):
Nafi narrated, from Salim [sic] the mawta of Tmar b. al-Khattab, that (Umar wrote to the [Muslims] of al-Sham concerning the Christians, that their stirrups be severed (? an yucgda rukubuhum), and that they ride upon pack saddles, ride side-saddle (bi-shiqq), and dress differently than the Muslims dress, that they may be known. (12)
Ibn Taymiyya (d. 728/1328) also knew (Umar I as having issued the ghiyar, deriving this information from the (now lost?) Shurat ahl al-dhimma of Aba 1-Shaykh al-Isbahani (d. 369/979):
As the 134fi: Aba I-Shaykh al-Isbahani narrated with his chain of transmission (isnad) in Shuria ahl al-dhimma, from KbMid b. (Urfuta: 13 tirriar [...] wrote to the garrison cities (arnsa r) that their forelocks be docked, meaning the Christians. and that they not wear the dress of Muslims. so that they might be known I. . .1. And AbCi I-Shaykh al-Isbahani narrated with his isnad that Umar b. al-Khattab wrote: And order the dhimmi women to fasten their belts (zunnii rat), and let down (yurkhina) their forelocks [...] that their clothes may be known from [those of] Muslim women. (14)
Less well-known later medieval works confirm this ascription. A passage long considered to be part of al-Madhamma ft isti'mal ahl al-dhimma of the eighth/fourteenth-century writer Ibn al-Naqqash, but which probably represents a later addition to that work, (15) is one example, most likely derived from the work of al-Turpshi above. (16) A similar (if anonymous) eighth/fourteenth-century work--probably our single richest source on the ghiyiir--concurs, quoting (in addition to the passages from Ibn Sallam and Ibn Zanjawayh above and many reports about (Umar II) from what it calls Kitab al-Sunna of al-Lalakai (d. 418/1027):
Abu 1-asim al-Tabari said in Kirab al-Sunna: cUmar wrote to the garrison cities (anksar) that they dock their forelocks and not wear the Muslims' dress, in order that they be known. And he said something similar about cUmar b. (Abd a1-cAz1z. (17)
Such examples could doubtless be multiplied. Yet these few suffice to show that the case for a unanimous attribution of the ghiyar to (Umar II in the Muslim sources cannot be sustained. If the cumulative testimony of the Muslim tradition is to be summed up, we might leave the job to Ibn Taymiyya, who in still another work attributes the ghiyar to 'Umar I and immediately thereafter describes Umar H as a renewer of it. (18)
The sources are not, then, unanimous. But what of the other proposed reason to accept as historically accurate reports that (Umar H issued an early edict--perhaps even the earliest--about the ghiyar, viz., their purported consistency? Do the sources at least give consistent testimony to the contents of the decree that he is said to have published? Are the "variations [...] minor, and in any case not contradictory"? (19)
If we are willing to tolerate pervasive disagreement on the details so long as a few sources indicate that the caliph decreed something about non-Muslims' clothing, hair, manner of riding, or a combination thereof (as well as assorted other things), then the evidence is consistent enough. But if we expect an imperial edict to produce uniform recorded memories of its language and scope--agreement on what exactly was enjoined or proscribed, or what other issues the same edict was also concerned with--we fail to find the evidence for this putative edict consistent.
Take dress, for example. What did cUmar II want non-Muslims to wear, or not to wear? According to a source whose testimony on this issue has so far been neglected, al-Baladhuri's (d. ca. 280/892) Ansa al-ashreif the caliph's concerns were negative: non-Muslims were not to wear turbans Cantei'im) or dress like (yatashabbahuna hi-) the Muslims at al1. (20) Al-Baladhuri reports elsewhere in the same work, in the context of a rather different edict (against non-Muslim officials), that (Umar H ordered non-Muslims to fasten their belts (maneitiq), just as we found cUmar I commanding earlier. (21) In another early work, Sirat 'Umar b. (Abd al-'Aziz attributed to (Abd Allah b.Abd al-klakam (d. 214/829)--there is what seems to be an edict textually related to the latter one from al-Baladhuri's Anseib, but it omits all mention of dress (of ghiyar restrictions it mentions only riding). (22) However, Ibn (Abd al-Hakam cites a separate text that does concern dress, both negatively and posi-tively. (23) This text is also attributed to (Umar H in two other sources, one dependent on the other: Shuriit of Ibn Zabr (24) (d. 329/940) whence in Ta'rikh Dimashq of Ibn cAsakir (d. 571/1176). (25) Here there are no (amieim or maneitiq, (26) but rather leather girdles (zunnar min jildljulild), which Christians must wear, and robes (qaber), Persian mantles (taylasan, "a cowl worn over the turban"), trousers with anklets (sardwil dheit khadama), and shoes with straps (ndl laha cadhaba), which they must not wear. (27) The manatiq return in a detailed and strikingly different text cited by the hadith specialist cAbd al-Razzaq al-Sancani (d. 211/827), (28) who also mentions turbans, using a different term (cisb). But 'Abd al-Razzaq relates nothing about robes, cowls, trousers, or shoes. Abli Yusuf credits to 'Umar II the most encompassing review of forbidden and required clothing in the form of a text with still another wording and set of concerns; the robes are here, joined now by silk clothing (oddly, since Muslim men are not supposed to wear silk either), as are both terms for the turban as well as the manatiq (but not zunnar, cowls, trousers, or shoes). Ibn Abd Rabbihi agrees that (amii'im were forbidden, adding the unparalleled detail that certain garments (aksiya) were required. (29) If we add the unique testimony of al-Qawl al-mukhtar, the picture becomes still more complex: one report, a letter from Tmar II to Khurasan with a full transmission chain, informs us that non-Muslims were to "split their shirts from both sides to the sleeve" (an tushaqqa qumusuhum min al-janibayni jamican ila l-kumm). (30) To sum up: no item of clothing, by any name, is found in all the early testimonia, nor do any of the testimonia include all the items of clothing found in the others. The most extensive list of restrictions on riding does not mention clothing at all.
The case for consistency is most difficult to sustain when it comes to hair. The two pieces of evidence from the work of al-Baladhuri both state that Umar H wanted non-Muslims to "shave the middle of their heads," i.e., tonsure. No one other than al-Baladhuri mentions this requirement. Ibn 'Abd al-I-Jakam reports quite clearly that non-Muslims were to part their forelocks (mafray a1-na,.siya). (31) The parallel text in the work of Ibn Zabr (whence that of Ibn (Asakir) agrees with the stipulation (and wording) that the forelocks be parted. (32) The text of 'Abd al-Razzaq says that their forelocks were to be trimmed (yajuzza nawaghim), but, using the same verb, that hair was not to be parted (yunhaw an yafruqa ru'asahum). Can clipping forelocks be seen as consistent with parting them, or being commanded to part one's forelocks with being forbidden to part one's hair? Abu Yusuf, for his part, uses a different term for clip-ping--tagyi,s--and adds that the caliph also objected to two particular hairstyles--al-jamam and al-wafr--both of which involved letting hair hang down. Again the sources fail to agree upon any one thing that Tmar II decreed regarding hair. None of them includes more than one of the three clearly delineated requirements: tonsuring, parting, and clipping. It seems probable that a few of these stipulations are in fact contradictory, and it cannot be said that the reports of cUmar Ws policy on non-Muslims' hair are characterized by consistency.
The evidence is most consistent, by contrast, when it comes to riding animals. One of the two notices provided by al-Baladhuri in Ansab al-ashraf indicates that cUmar H, perhaps like his grandfather Umar I, ordered non-Muslims to ride on pack saddles (ukuf, the other does not mention riding). According also to the "document" against non-Muslim officials produced by Ibn (Abd al-iJakam, both male and female Christians were to ride on pack saddles, being forbidden to ride on saddles. They were moreover not to straddle riding animals but to ride with both their legs to one side. Abu Yasuf mentions all of these requirements save for the last. cAbd al-Razzaq and Ibn Zabr (and following him Ibn 'Asir) report that Christians were not to ride on saddles. But they know nothing of pack saddles, which are in fact the only riding regulation of cUmar II known to Ibn Sanaa'. (Abd al-Razzaq alone among the three knows of instructions about women (no saddles). There are no hard contradictions here, but neither is there universal agreement on any single point. Ibn cAbd al-I-Jalcam's first text, against non-Muslim officials, does, however, list all of the components found elsewhere.
Finally, we should note the assorted additional stipulations that one finds attached to the ghiyar as attributed to (Umar II in these sources; these do not involve dress, hair, or riding per se. They matter to us, however, because the sources tend to present the ghiyar evidence as transcripts of actual documents. Thus one expects--but fails to find--a high degree of uniformity in the details (how many edicts could 'Umar II have issued on the matter in his two-and-a-half-year reign?). Al-Baladhuri and Ibn (Abd al-Hakam alone place the ghiyar in letters that inveigh against the employment of non-Muslim state officials; their versions appear to bear a distant textual relationship to each other, but when it comes to the ghiyar they share only pack saddles. Ibn Abd Rabbih reports that Umar II commanded the reverse: that non-Muslims might not employ Muslims. Abu Yrisuf says nothing about employment, but his evidence, unlike that of al-Bandhuri and 1bn (Abd al-Hakarn, adds that publicly displayed crosses were to be broken. 'Abd al-Razzaq's report forbids the public display of crosses but does not recommend breaking them; it adds a prohibition on striking the seman-tron (neicias). Ibn Zabr (whence Ibn 'Asakir) and the second report of Ibn eAbd al-Hakam contain nothing about crosses, employment, or the semantron in connection with Umar II, but add that non-Muslims were not to possess weapons. Here again, only a very weak consistency can be ascribed to the evidence.
Read closely, the reports about the ghiyar often display additional problems. Consider, for example, one of the weightiest testimonia that connects promulgation of the ghiyar to 'Umar II; it is found in the work of 'Abd al-Razzaq, where it is supplied with an impressive chain of transmission:
(Abd al-Razzaq reported to us, saying (gala): Macmar r eported to us, from ('an) 'Amr b. Mayman b. Mihran. (33) saying (gala): 'Umar b. 'Abd al-'Aziz wrote (kataba) that the Christians of Syria be prevented from striking a semantron. He said (gala): And they should be forbidden (wa-yunhaw) to part their hair, and they should clip (yazuzza) their forelocks, and tighten their belts (rnanatig) [several more ghiyar stipulations follow] for if [the Muslims] have authority over anyone of them who does one of these things after being informed [of them], any [Muslim] who finds him may despoil him. He said (gala): And he wrote (kataba) that their women should be prevented from riding on saddles.
The verb gala ("he said") appears four times in this passage. In three instances its subject is indisputably a transmitter: once cabd al-Razzaq, twice one of his sources. Most of the ghiyeir stipulations are, however, spoken by the subject in the remaining instance: the third. The identity of this subject is unclear. For Levy-Rubin (p. 89), it is the caliph 'Umar II, and thus he who set forth these stipulations. But it is likely that the grammatical subject of gala is in fact a transmitter. Twice in this passage, immediately before and immediately after the speech in question, a decree of the caliph is clearly marked as such by the verb kataba ("he wrote"). Between these two instances, the ghiyar stipulations are bracketed only by qala 'not wa-qala, which could imply a continuation of the previous subject), which in all other instances here marks the speech of a transmitter. The report concludes as follows: "'Amr b. Maymiln said (qa/a): And 'Umar consulted me about destroying their churches, so I said (fa-quitu): They are not to be destroyed; this is what was agreed upon with them. So 'Umar left them." The actions of the caliph 'Umar are thus interspersed here with the opinions of a transmitter. It would have been entirely natural for the latter's opinions to be ascribed eventually to the caliph himself; when al-Khallal cited this report via (Abd al-Razzaq a century later, the ambiguity was gone. (34) This and similar problems affecting the evidence for the ghiyezr must be recognized and, where possible, resolved before historical conclusions may be drawn.
In sum, the case for unanimous attribution to 'Umar II is untenable, while that for consistency is tenuous. (35) The origins of the ghiyar should therefore be considered in question. Future scholarship must either affirm or deny that the ghiyar was instituted prior to the reign of 'Umar II, as many sources report on the authority of multiple early transmission chains. If in fact it was, then 'Umar II was at most reviving or elaborating upon earlier regulations, which is the view of Albrecht Noth.36 Whatever connections might be drawn between the ghiyar and any broader ideology discerned in the welter of reports about 'Umar H must in this case also be related to the ideology of earlier rulers. Moreover, if the ghiyar originated shortly after the conquests and Levy-Rubin's novel and intriguing argument for its Sasanian roots is correct (chapter 5, esp. pp. 143-44), then Sasanian influences must already have shaped religious policies at the very start of Arab Muslim rule in former Sasanian territories. If, on the other hand, the ghiyar was not in fact such an early institution, then the reports that say it was are unreliable as evidence of what they describe, in which case we are dealing with a milieu in which transmitters and/or compilers were actively originating reports about the ghiyar along with their chains of transmission,37 presumably because they lacked sufficient authoritative precedent. With such a milieu it is far more difficult to accept the historicity of the eclectic, prescriptively charged reports concerning 'Umar II that we find in literary sources compiled many decades after his death.
A variety of hypothetical intermediate solutions could preserve for these reports some historical facticity. For example, the reports concerning 'Umar II might have a historical basis, however difficult to spell out in detail, while those concerning 'Umar I might have been originated to furnish authoritative precedent for the policies of his Umayyad grandson. Such hypotheses remain to be tested (requiring the development of analytical tools for evaluating the akhbeir evidence that are more robust than those we currently possess).
However much we might prefer a satisfying narrative, we are compelled to acknowledge the problems that affect the evidence on which any such narrative must be based. Given the thinness of that evidence and our current methodological limitations in assessing it, it will be difficult to cut through the fog of obscurity that shrouds both the contents and the authorship of the early ghiyar reports. For now, at least, the evidence for the origins of the ghiyar, as for so much else in early Islam, is inconclusive. Where there are few corroborating transmission chains and ample incentive for the pseudepigraphical creation of prescriptive precedents, we have no way conclusively to distinguish many reports recalling historical Umayyad and pre-Umayyad events from instrumentalized retrojections of the 'Abbasid period. In the present case, for example, it is eminently possible that learned elites of the 'Abbasid period, inspired by Sasanian precedents in the new metropolis of Baghdad and by the growing need to distinguish non-Arab Muslims from non-Muslims, ascribed a variety of sought-after ghiyar measures to earlier caliphs. The preponderance of Persian vocabulary in these reports suggests eastern origins. (38) The possibility that reports connecting the ghiyar to the two cUmars were originated diffusely by 'Abbasid-period tradents in need of precedents cannot be lightly discarded; the reign of Abil Yusuf's patron Haran al-Rashid, for example, reportedly saw attempts to impose the ghiyar. (39) This alternative account could explain why the various reports fail to agree on any single thing that 'Umar II decreed about the dress, hair, or riding habits of non-Muslims, why the verifiably eighth-century non-Muslim sources are uniformly positive in their assessment of cUmar 11, (40) and why the reports attributed to both cUmar I and Umar II look oddly similar to the better-documented measures taken by 'Abbasid caliphs. (41) cUmar b. 'Abd al-'Aziz and/or
'Umar b. al-Khattab might really have issued ghiyar edicts--as 1bn 'Asakir wrote after transcribing one such edict, -only God, may He be exalted, knows." The historicity and contents of these putative edicts remain to be systematically demonstrated against viable competing accounts of the evidence.
(1.) Compare, e.g., the elaborately stratified sumptuary restrictions ascribed to the Aztec state (Fray Diego Duran, The History of the Indies of New Spain, tr. D. Heyden [Norman, Okla.: Univ. of Oklahoma Press, 19941, 208-10) or the Qing requirement that the Chinese adopt the Manchu queue.
(2.) Non-Muslims in the Early Is/antic Empire: From Surrender to Coexistence (New York: Cambridge Univ. Press, 2011).
(3.) Scholars will also want to consult Ilse Lichtenstadter, "The Distinctive Dress of Non-Muslims in Islamic Countries," Historia Judaica 5,1 (1943): 35-52 and literature cited there. allel version of the first report. see al-Khallal, Ahkam rthl ed. I. b. H. b. Sultan. 2 vols. (Riyadh: Niaktabat al-Macarif. 1996), 2: 429 no. 995; the isnad diverges after (Abd Allah b. 'Liman
(4.) Levy-Rubin (p. 207 n. 32) repeats Antoine Fattal's claim (Le statut legal des non-musultnans en pays d'islam, 2nd ed. [Beirut: Dar el-Machreq, 19951, 98) that the jurist al-Kasani (d. 587/1189) "also claimed that Umar b. 'Abd al-cAziz was the initiator of the ghiyar." But Fattal misconstrued al-Kiisiinr s statement--al-aslfi marks a ratio legis, not a historical claim for originality--and the reader finds no other evidence that any source claims explicitly that (Umar II instituted the ghiyar.
(5.) Read khashab for hasan.
(6.) Al-Askari, nab al-Awii'il, ed. M. al-Misri and W. Qassab, 2 vols. (Damascus: Wizarat al-Thaqafa wa-1-Irshad, 1975), 1: 395. Other evidence of this edict is abundant; see the citations in Levy-Rubin. Non-Muslims, 103-11. For al-tAskarrs claim repeated, see al-Suyati, ila mdrifat al-awa'il. ed. A. A. (Abd al-Q5dir (Kuwait: Dar 1bn Qutayba, 1990), 104-6.
(7.) Ibn Sallarn. Kitab al-Aniwal, ed. M. Kb. Harras (Cairo: Maktabat al-Kulliyy5t al-Azhariyya, 1969), 130. Levy-Rubin cites the first report as evidence of an edict of Um& II (pp. 89-90. 206 n. 10). She mentions neither the second nor a shorter one that follows, which associates Umar LI with the requirement to ride on pack saddles and clip forelocks. But in fact the first report concerns cUmar I. It is thus incorrect (p. 90) to connect (Umar II with the belts called manatiq on the basis of this passage; it is cUmar I who mandates manatiq here. For a corroborating
(8.) Abu Yasuf. Kiteib al-Kharaj (n.p.: Dar al-Islab, 1981), 262. Noted already by Lichtenstadter, "Distinctive Dress," 42-43, and cited by Levy-Rubin (p. 206 n. 10) without mention of cUmar 1.
(9.) Ibn Zarkjawayh, Kitab al-Amwal, ed. Sh. Fayyad. 3 vols. (Riyadh: King Faisal Center for Research and Islamic Studies. 1987), 1: 186.
(10.) Of this practice we have evidence both early (late seventh century) and firm in the form of surviving lead seals. See Chase Robinson, "Neck Sealing in Early Islam," Journal of the Economic and Social History of the Ori- ent 48,3 (2005): 401-41.
(11.) The History of the Conquest of Egypt, North Africa and Spain, Known as the Futah Misr of Ibn 'Abd al-Hakarn, ed. Charles C. Torrey (New Haven: Yale Univ. Press, 1922). 151. The passage omitted in my translation relates to the poll tax.
(12.) Al-Turtashi, Sireij al-mrdak (Alexandria: al-Matbaca al-Wataniyya, 1289/1872-3), 136. "Salim" must be a mistake for "Aslam."
(13.) Khalid (d. 61/680-1) is a minor narrator from both the Prophet and 'Umar I who lived his later life in Kafa, where he had some political authority; see al-Mizzi, Tahdhib al-kamal fi astraii) al-rijal, ed. B. A. Macraf, 35 vols. (Beirut: Mu)assasat al-Risala. 1992), 8: 128-30.
(14.) Ibn Taymiyya, lqtida al-sirat al-mustaqim, ed. N. al-cAql, 4th ed., 2 vols. (Riyadh: Maktabat al-Rushd, 1994), 1: 327-28. I elide clauses not related directly to the ghiyar. Al-Qawl al-mukhtar (n. 17, below) provides the entire isnad of this report to Abu I-Shaykh (p. 130).
(15.) On the work, author, and probable addition, see Luke Yarbrough, -Ibn al-Naqqash," in Christian-Muslim Relations: A Bibliographical History, vol. 5: 1200-1350. ed. D. Thomas and A. Mallett (Leiden: Brill, 2013), 123-29.
(16.) F. A. Belin, "Fetoua relatif a la condition des zimmis," Journal asiatique, 4e ser., 18 (1851): 499-500.
(17.) Al-Qawl al-mukhtar .fi* l-man' 'an takhyir al-kuficar, in lithographed collection; uniform title (Uddat al-umare wa-l-bukkant li-ihanat al-kafara wa-(abadat al-asnant (Cairo: Matbdat tiajar, 1856-57). 128. See (Sharb usui ictiqad ahl al-sunna wa-l-jamda. ed. A. b. S. Bin Hamdan, 5 vols. [Riyadh: Dar Tayba, 1995]). in which I have not located this report.
(18.) Ibn Taymiyya, Mafia' al-fatawa, ed. A. b. M. b. Qa.'sim and M. b. A. b. M. b. Qiisim, 37 vols. (Medina: Wizarat al-Shu'an al-Islamiyya, 2004), 28: 654. His follower Ibn Qayyim al-Jawziyya credits the ghiyar to (Umar I in high rhetorical fashion, adding that all the Companions approved of cUmar's policy (Ahkam ahl al-dhimma, 4th ed., ed. S. Salih, 2 vols. [Beirut: Dar al-'11m li-l-malayin, 19941, 2: 756).
(19.) Levy-Rubin. Non-Muslims, 90.
(20.) Al-Baladhuri, Ansab al-ashraf. ed. M. al-F al-cAztn et al., 25 vols. (Damascus: Dar al-Yaqza, 1996-2010), 7: 91.
(21.) Ibid., 7: 138.
(22.) Ibn 'Abd al-Hakam, Sirat 'Llmar b. 'Abd al-'Aziz. ed. A. cUbayd, 2nd ed. (Cairo: Maktabat Wahba, 1983), 135-36.
(23.) Ibid., 136.
(24.) See Mark R. Cohen, "What Was the Pact of %mar? A Literary-Historical Study," Jerusalem Studies in Arabic and Islam 23 (1999): 146-47. Page 140 of this edition and the corresponding pages in the dependent passage by Ibn 'Asakir (Ta'rikh madinat Dimashq, ed. A. Shia, 80 vols. [Beirut: Dar al-Fikr, 1995-981 2: 179-80) contain not an attribution of the ghlyeir to cUmar II. as Levy-Rubin claims (p. 90, where for "MasriIk b. cAbd al-Rahman b. Ghanam," read "Masroq from rani cAbd al-Ratiman b. Ghanm"), but rather reference to a version of the Pact of %mar Ill that attributes precisely these words to the elder Tatar. See further n. 25, below. Many versions of the Pact include very similar measures, with similar wording, and might be cited as additional attributions of the ghiyar to Umar I.
(25.) Ibn cAsakir, Ta'rikh inadinat Dimashq, 2: 185. It can be seen from the isnads that Ibn (Asfilcir is quoting directly from the earlier work by Ibn Zabr; thus, they cannot corroborate one another, a fact that Levy-Rubin does not, in my view, make sufficiently clear. She translates Ibn Zabr's crucial comment as follows: "I did not see this addition [by which this passage about the ghiyar was included by a certain author known to Ibn Zabr in a version of the Pact of %mar] in what has been passed down to us of Shwa! (Umar [b. al-Khattab]; rather, I found it related from 9.1mar b. (Abd al-(Aziz," adding, "unlike many other cases regarding regulations governing the non-Muslims, there is no confusion between %mar b. al-Khattab and Umar b. Abd al-Aziz here." In fact, this passage is evidence that such confusion did exist; and it is precisely this confusion that induced Ibn Zabr to make this comment in reaction to the attribution of a textually identical edict concerning the ghiyar to Umar I in the book by his colleague (wa-reaytu hadha l-haditha hi-kitabi rajulin min ashabina bi-Dimashq). See also n. 24, above.
(26.) Pace Levy-Rubin, who states (p. 90) that Ibn 'Abd al-klakam uses the term manatiq.
(27.) I adopt many of Levy-Rubin's translations (p. 90). Since the term zunnar is clearly used here, her later remark that "Umar H's edict does not use the term ztirmar" (p. 154) is mistaken, also contradicting, a page later, "Tmar's edict indicates that the requirement of the zunniir [by which belts in general must be meant] was established at the beginning of the eighth century." Quite apart from the contradiction, it seems hazardous to make any statements about what Tmar might have declared because there is no stable text.
(28.) Al-Sancani, al-Alu.yannafji 1-hadith, ed. I-j. al-R. al-A'?.ami, 10 vols. (Beirut: al-Majlis aJcIlnii, 1970-), 6: 61.
(29.) Cited by Fatal (Le swat legal, 96 n. 48), who attached the report erroneously to (Umar I.
(30.) Al-Qawl al-mukhgir (supra, n. 17), 130.
(31.) Levy-Rubin consistently translates this phrase as having the "forelocks trimmed," arguing that the obvious meaning (i.e., parting) "is not possible in this case, as it is well established that the demand was to cut the forelocks and not to part them" (p. 206 n. 20). With the evidence so scanty and so disparate, it is hard to see how this could be considered well established. Restrictions on hair parting are widely attested in versions of the Pact of Umar.
(32.) Cohen, "What Was the Pact of cUmar?" 146-47; Ibn Asdkir, Ta'rikh mad mat Dimashq, 185.
(33.) On the historicity of Macmar's tranmission to (Abd al-RuzAq, see Harald Motzki, "The Mutscumaf of (Abd al-Razzati al-Sancani as a Source of Authentic agidith of the First Islamic Century," Journal of Near Eastern Studies 50 (1991): 1-21. For (Amr b. Maymiin's reports about (Umar 11, see idem, "The Prophet and the Debtors: A Haditlz Analysis under Scrutiny," in idem, Analysing Muslim Traditions: Studies in Legal, Exegetical, and Maghazi Hadith (Leiden: Brill, 2010), 197, and references there.
(34.) Al-Khallal, Ahkarn ahl al-milal, 2: 428 no. 996; cf. 426 no. 986. For an unambiguous, grammatically parallel case, see al-Sancani, Mtqannal: 6: 59 no. 9999: ''Abd al-Razziki reported to us, saying (gala): My uncle Wahb b. Nafic reported to us, saying (qiila): 'Umar b. 'Abd al-(Aziz wrote (kataba) to 'Urwa b. Muhammad that he should destroy the churches in the garrison cities (am.yar) of the Muslims. He said (gala): I saw 'Urwa b. Muhammad ride [...] then I witnessed to the letter of 'Umar, and 'Urwa's destruction of them." Here the subject of the third qa/a is undoubtedly the transmitter. Wahb.
(35.) I have not addressed a third aspect of Levy-Rubin's argument, that the link to cUmar b. Abd al-Aziz is "supported by a consistent ideology which was zealously proclaimed by 'Umar b. 'Abd al-Aziz himself" (p. 92). This assertion ascribes reliability to a broad selection of literary evidence--including, apparently, the caliph's entire sira by Ibn 'Abd al-Hakam--in order to establish that of the ghiya r "code." But we have no Archimedean point from which to survey most of this literary evidence, the contemporary documentary record being relatively thin. Thus a "consistent ideology" of %mar 11 is no more firmly established than the origins of the ghiyar itself; the pronounced inconsistencies long known to characterize his putative ideology stubbornly persist 'see W. W. Barthold, "Caliph %mar II and the Conflicting Reports on His Personality," Islamic Quarterly 15,2/3 [19711: 69-95; and, for a brilliant recent conspectus, Antoine Borrut, "Entre tradition et histoire: Genese et diffusion de l' image de Tmar B," Melanges de l'Universitd Saint-Joseph 58 [20051: 325-78). In any case, other leading candidates for the initiatory role with respect to the ghiyar 'e.g., Tr= I, al-Mutawakkil) are credited with ideologies that are equally consistent and zealous. For the non-Muslim and documentary evidence on the reign of <Umar II, see Luke Yarbrough. "Did Tmar b. Abd al-cAziz Enact a Religious Criterion for State Employment?" (to appear in Christians and Others in the Umayyad State [tentative title], ed. A. Borrut and E Donner)--in brief, the earliest non-Muslim source to refer to a ghiyar measure of 'Umar II (riding on saddles) was probably that of Dionysius of TelImahre (d. 845), whose lost chronicle was a source for extant twelfth- and thirteenth-century Syriac histories.
36. Noth ("Problems of Differentiation between Muslims and Non-Muslims," tr. M. Muehlhaeusler, in Muslims and Others in Early Islamic Society, ed. R. Hoyland [London: Ashgate, 20041, 103-25, esp. n. 99) saw regulations concerning dress as products of the immediate post-conquest period, and any measures of 'Umar II ("supposing that this tradition is authentic") as renewals.
37. Because the isniids in several cases go back to companions of <Umar I, the ascriptions cannot be brushed off as mistakes caused by homonymy. as much scholarship on the origins of regulations pertaining to non-Muslims has done.
(38.) Cf. C. E. Bosworth, The Protected Peoples' (Christians and Jews) in Medieval Egypt and Syria," Bulletin of the John Rylands University Library of Manchester 62 (1979): 11-36, esp. 18. Yet Levy-Rubin's effort to demonstrate the popularity of Persian fashions in Umayyad Syria (Non-Muslims, 97. 130-35) must not be ignored.
(39.) For measures touching upon non-Muslim clothing and riding habits under Hariin, see al-Tabari, Annales quos scripsit . . . ,ed. M. J. de Goeje (Leiden: E. J. Brill, 1879-1901), 3,2: 712-13; [Mari b. Sulayman], Maris Amri et Slibae De patriarchis Nestorianorum commentaria. ed. H. Gismondi. 2 vols. (Rome: F. de Luigi, 1896-99), I: 73 11. 11-13 (Arabic). Neither of these important works mentions anything of the sort under cilmar H.
(40.) On the non-Muslim sources, see Yarbrough, "Religious Criterion?" (supra, n. 35).
(41.) Recall the explicit awed evidence that al-Mutawakkil was in fact the first to issue such a decree. Al-Jatri?.'s Radii 'ala 1-nasara, written shortly before al-Mutawaldcil's decree, confirms that the issue was a pressing one at the time, and that distinctive belts were worn. But it does not refer to a law promulgated by the state: "Many of [the Christians] have ceased wearing belts, while others wear them underneath their outer clothing- (wa-taraka kathirun minhum caqda 1-zananiri wa-(aqadaha akhariina dam thiyaihirn); Thalath rasa'il Tatman (Amr b. Bahr al-Jahiz, ed. J. Finkel (Cairo: al-Matbda al-Salatiyya, 1344/1926), 18; Joshua Finkel. "A Risala of al-Jabi?.," JADS 47 (1927): 329. It may also be inferred that Christians' clothing (thiyab) was in practice not clearly differentiated from that of Muslims.
SAINT LOUIS UNIVERSITY
|Printer friendly Cite/link Email Feedback|
|Publication:||The Journal of the American Oriental Society|
|Date:||Jan 1, 2014|
|Previous Article:||Beginnings of indian and chinese calendrical astronomy.|
|Next Article:||Epistemic modality in old babylonian.|