THE USE AND ABUSE OF EVIDENCE: THE QUESTION OF PROVINCIAL AND ROMAN INFLUENCES ON EARLY ISLAMIC LAW*

In her monograph Roman, Provincial and Islamic Law, Patricia Crone argues that provincial legal practice in Syria, the capital of the nascent Islamic state, influenced the SharPa no less than did Roman law.

I

Although the problem of foreign influences on Islamic law has engaged the minds of Islamicists for well over a century, it appears that the solution to this problem is no less elusive now than it was when Alfred von Kremer commented on it as early as 1875. The contributors to this issue have ranged over a wide spectrum, from those who affirmed a debt to Mosaic law to those who saw in Roman law the material which the emerging Shari^ca was to appropriate. Their estimation of the degree of this debt has also varied from “systematic borrowing” from Jewish and/or Roman law to more modest claims of indebtedness.¹ Dr. Patricia Crone’s work, Roman, Provincial and Islamic Law, represents the most recent contribution to this problematic issue.

While postulating that Jewish law “manifestly did contribute to the formation of the Shari^ca” (p. 2), Crone rejects an unqualified assertion of Roman influence. She argues that if Roman elements had entered the Shari^ca they would have done so only through provincial law, which in her definition “refers to the non-Roman law practiced in the provinces of the Roman empire, especially the provinces formerly ruled by Greeks” (p. 1). And since nascent Islamic culture took shape in the Roman Fertile Crescent, the geographical area of influence could have only been Syria and not Iraq as Schacht had argued.

rest of her work, with the exception of chapter 2, which provides non-specialists with the basics of Islamic law in order to facilitate their path throughout the book. The third chapter treats the legal effects of the Islamic patronate in its two forms wala^J al-^citq (a legal tie arising from manumission) and wala^J al· muwalat (contractual clientage). Here, Crone declares her intention to prove that pre-Islamic Arabia “sup­plied the general context of wala^J” but “did not supply the institution itself. The crucial features of waler* derive from Roman and provincial law”(p. 41). In the fourth chapter, she sets out to prove the first thesis, namely, that wala^J did not originate in pre- Islamic Arabia.

Before we proceed to examine Crone’s evidence and arguments in support of her thesis, we must address the underlying assumptions of the book, particularly the author’s notion of provincial law.

Aside from the definition of provincial law already given, Crone attempts no investigation of what the sources of provincial law may have been, nor does she explore who the people applying this law were. But it seems that in Crone’s mind these people are asso­ciated with the lofty Hellenic culture, and not with their uncivilized southern neighbors in the Peninsula who were responsible for the “culturally destructive” and “barbarian conquests.”³ Needless to say, the determination of the actual occurrence and the extent of borrowing are intimately related to the manner in which one conceives of the general levels of culture of those who borrow and those who lend. Once a culture is seen as ‘sophisticated’ it becomes incapable, in the mind of those who see it as such, of borrowing from another, ‘primitive’ culture; whereas ‘desolate’ cultures are a priori given to appropriating from more de­veloped ones. Crone’s pre-Islamic Arabia is “sparsely inhabited and uniformly impoverished,” a place where the miserable bedouin goes “to battle for the sake of honour and excitement.”⁴ This Arabia, which had no history but only primitive events to record, possessed no real culture,⁵ and the Islam that it produced makes for no more than a historical mischance.

When the culturally deprived peninsular Arabs con­quered Syria, Crone tells us, they found a high culture including a developed provincial legal tradition which was Romanized to some extent. There is scarcely any attempt at revealing the identity of those Syrian pro­vincials, though again we suspect that in Crone’s mind they are not Arab, for, after all, they are cultured. Crone’s Arabs of Provincia Arabia do not fare any better than their southern brothers. They are subser-

5

vient and disorganized, and when the Yemeni South provided them with a script, all they could do with it was to record the “details of their genealogies and sheep.”⁸

Viewing the Arabs thus can only result in assigning them the passive role of recipients of provincial culture and institutions. By insisting on the polarity between the ‘primitive’ Arabs and cultured provincials, Crone has ignored the findings of an entire field of scholarship without which no study of foreign influences on Islamic law can be conducted. The works of Briinnow and von Domaszewski,⁹ Altheim and Stiehl,¹⁰ Noldeke,¹¹ Mommsen,¹² Dussaud,¹³ Bowersock and a score of others¹⁴ have long been recognized as authoritative sources on the demographic and cultural composition of ethnic groups that existed in Syria and the Syrian desert throughout the millennium before Islam.¹⁵ The best of this massive scholarship has recently been conjoined with the fruits of another scholarly career in the voluminous work of I.

The findings of these works make it clear that the social and cultural map of West Asia during the few centuries preceding Islam cannot be seen in terms of black and white, as Crone seems to envisage it in this book. The massive migrations from the Peninsula to the north, the trade routes between Syria and western Arabia, the Arab foederati who constituted a cultural and commercial bridge between Romans and Byzan­tines on the one hand and the southern Arabs on the other,¹⁷ are among the many factors that clearly cre­ated a constant state of flux between North and South, thus making any rigid or well-defined classification of cultural and social institutions suspect, if not impos­sible. At the same time, and despite the Hellenic im­print that was left on the society of the Fertile Crescent, the distinctive Semitic character of this society was never lost, a fact that constantly impresses itself on the historians of this period.¹⁸ The Arabs of Syria and the Syrian desert preserved their Semitic rites, language, and above all their traditional laws, old tribal struc­tures and ancestral customs.¹⁹ These fundamental facts alone cast grave doubts on any scholarly work which operates on the assumption that geographical units and ethnic and cultural groups were neatly separated one from the other, and that southern Arabs were, to any significant extent, isolated from the ethnic and cultural groups of the Syrian North. In this respect Crone’s notion of the international relations existing in West Asia before Islam seems dangerously archaic— at least insofar as it bears upon the question of legal influences.²⁰

The fact that Crone failed to take into considera­tion the significant findings in the fields of Romano- and Byzantino-Arabica has gravely weakened her thesis, which is based on the erroneous assumption of the non-existence of any meaningful relation between what she sees as the arid South and the flourishing North.

7 twisted or suppressed. As we shall see, in Crone’s work it is the hypothesis that shapes evidence, not the other way around.

II

As we have mentioned, Crone concerns herself with the institution of patronage (a good part of which, as we shall see, was not recognized by three Sunni legal schools; and for those who did admit it, the institu­tion, juridically speaking, was of little importance).²¹ In the third chapter, which begins the real stuff of the book, she opens with a statement to the effect that every society must have a way of incorporating out­siders. All non-Arab newcomers to Islam, she tells us, were affiliated to individual Muslims with whom they were placed in an unequal relationship. This relation­ship turns out to be wala^J, a general term employed in two types of legal institutions known to Muslim lawyers as walcf al-^citq and wala^J al-muwalat. We must emphasize from the outset that while the latter may, but does not necessarily, involve converts, the former is a legal institution exclusively related to the manumission of slaves and has nothing to do with the incorporation of outsiders into Muslim society;²² for it does not, as such, bring slaves into Muslim society. It only changes their legal status; it is an instrument for freeing slaves within Islamic society. Be that as it may, our author goes on to enumerate the rights arising from wala^D al-^citq. Slavery, needless to say, can hardly constitute a relationship of equals. The manumitter, Crone argues, acquires a right to the freedman’s estate, but he does so only as the last agnate (according to the great majority of jurists). The freedman, however, does not possess the same rights to the patron’s estate. The manumitter must pay blood-money on behalf of his freedman, but the

8

freedman is not charged with the same obligation toward his former master. This last specification loses importance when the reader is told that a great many jurists do assign such a duty to the freedman (p. 37).

When Crone comes to walcP al-muwaldt. it is obvious that she has misread the legal texts. These texts specify that inheritance and blood money are mutual rights and obligations, provided that both parties agree to them. Crone presents these rights in a manner that gives the false impression of the patron possessing rights superior to those of his counterpart. She claims, without any support from the sources, that “[u]sually succession is assumed to be unilateral” (p. 39). One would like to know whose assumption that is, since it is commonly held by the jurists that inheritance is mutual (wa-yatawarathani min al· janibayni)²³ Although she characterizes waler* al· muwalat as “contractual clientage,” she fails to bring out clearly the fact that waler* al-muwalat rests on contractual obligations and has nothing to do with waler* al-^citq. which arises from slavery. In Crone’s account the contractual elements are missing. Any contract in Islamic law must be concluded by two persons of full capacity, persons who are presumed, legally speaking, to be equals. Furthermore, wala^J al- muwdlat (rejected by the Shafi^cTs, Malikis and Han- balis) does not necessarily involve conversion, as Crone herself admits in passing (p. 38). In fact, one may convert to Islam “at the hands” of one person and simultaneously enter into a wala^J al-muwalat contract with another. If wala^J al-muwalat takes effect simultaneously with conversion, it would be valid but certainly not as a consequence of conversion.²⁴ Such a contract may be concluded between a dhimmi and a Muslim, without conversion being involved.²⁵ It could even be concluded between two dhimmis²⁶ In fact, it does not have to be concluded at all, for it is

9 commonly held to be ja^Jiz (permissible), not lazim (obligatory).²⁷ As Crone knows, the only school which insists on wala^J al-muwaldt as the instrument of conversion is that of the Qasimis, but they allow this only in dar al-harb, not in Muslim territories. Most important of all is the fact that the raison d'etre of waler* al-muwalat is nothing but mutual support be­tween the two parties in matters financial and moral. As the jurists put it: wa-sababu walcr*i al-muwalati al- ^caqdu wal-matlubu minhu al-tanasuru²* The term tandsur (or other derivatives of n-s-r, connoting help and mutual support) seems to figure prominently in the language of offer and acceptance in this contract: e.g., ^caqadtuka ^cala an tansurani wa-ansuraka.²⁹ This contract by itself can hardly be said to create a relationship of dependence. It is important to remem­ber this, for many later arguments in the book rest on Crone’s misunderstanding of the exact nature of this contract.

Having discussed what she thinks to be the legal rights and duties in these two (very different) wala^J relationships, Crone concludes that “[a]ll non-Arab newcomers in early Islamic society thus paid for their membership of this society by acceptance of a private relationship of dependence” (p. 40). The reader is left to wonder about the relevance of this statement to the two legal institutions which have in themselves almost nothing to do with incorporating converts into Islamic society.

Crone proceeds to build another argument on what she claims to have demonstrated. She sees a gulf between doctrines of Islamic law concerning wala^J and the social reality of the “homeland of the con­querors,” and she sets out to prove the existence of this gulf, aiming ultimately to show that if wala^J was not taken from Arabia, then it must have been ap­propriated from another source. The gulf, as Crone sees it, is represented by four points (pp. 43-44): (i) “The Islamic patronage is individual. It binds one person to another in a relationship of dependence.” In pre-Islamic Arabia such relationships existed between groups, not individuals, (ii) “The Islamic patronage detaches the client from his natal group and incorpor­ates him in that of the patron as a passive member.” In pre-Islamic Arabia clients remained in their own groups, (iii) “The Islamic patronate is assimilative. Having been detached from his natal group, the client sooner or later acquires full membership of the new group.” “Collective clientage, by contrast, is not neces­sarily assimilative, and assimilation is effected col­lectively if it occurs at all.” (iv) The Muslim patron must pay “blood-money on his client’s behalf in return for a title to his estate. No such rights and duties can be assigned to patrons where clientage is collective, nor can they be documented for pre-Islamic Arabia.”

Let us address (ii) and (iii) first. We have seen that wald³ al-muwaldt does not necessarily involve conver­sion. Nor does conversion, in and by itself, give rise to this wald^J. The client is under no obligation to enter into the contract, and can at any time (as long as the patron has not paid blood-money on his behalf) withdraw without incurring any penalty.³⁰ The rela­tionship established by the contract is clearly a ficti­tious kinship which is capable of dissolution. The client, according to many jurists, is not supposed to have a natal group. The doctrines of the jurists give one to understand that the contract of wald^J al- muwalat is specifically designed to provide for those non-Arabs who have no lineage or relatives. This is the reason why certain jurists insist that for such a contract to be valid the client must have no existing relationship (wa-kawnuhu majhiila al-nasabi shartun li-sihhati ^caqdi al-muwalati).^u Abu Hanifa and the Twelver Shi^cis moreover hold this contract to be valid only when both parties have no nasab.³² There is nothing in this contract which obliges the client but does not oblige the patron. (Incidentally, the terms “client” and “patron” do not reflect the legal status of the parties in the contract but indicate, respectively, who made the offer and who accepted it.) Rights and duties are mutual between the two parties. At the same time, assimilation does not necessarily follow from these doctrines. How can, for example, a dhimmi (let alone a Muslim) entering into such contract with a Muslim or another dhimmi be assimilated into Muslim society more than he already is? These con­siderations certainly confute Crone’s point (iii) as well.

Thus, (i) and (iv) remain to be proven. Crone begins by examining modern Arabian society in the nineteenth and what seems to be the first third of the twentieth century. She realizes that what she finds in the modern period “is not of course in itself evidence for pre-Islamic conditions; but in its light the frag­mentary pre-Islamic evidence becomes more informa­tive than when it is taken on its own” (p. 44). There is, however, no justification for taking into account situations removed from the main subject of study by more than twelve centuries; the evidential value of such situations is obviously too suspect to be seriously considered. And if it be true that pre-Islamic evidence “becomes more informative” in the light of nineteenth and twentieth century evidence, then the pre-Islamic evidence must be fundamentally problematic.

The first difficulty Crone encounters in establishing (i) and (iv) is that the evidence she needs from the pre-Islamic period is not only meager but also hope­lessly confused. As anyone who is familiar with such books as Isbahani’s Aghani and Tabari’s Tarikh will admit, the term mawla and its usages are diverse and lack any fixed connotation. When Tabari reports that x was the mawld of y, there is simply no way for us to know what the exact social or legal relationship between the two was, unless of course we are provided with detailed information about the persons involved, which is, to be sure, a rare luxury. The other difficulty is Crone’s own confusion concerning wala^D al-^citq and wald^J al-muwalat. When she proceeds to try to prove the absence of their equivalents in pre-lslamic Arabia, she insists that they are two items of one and the same set. This, as we have seen, rests on a misunderstand­ing of the legal texts.

Crone argues that although the pre-lslamic halif, who is an ally or a confederate mawla, enjoyed full protection and was frequently avenged when killed, he was not incorporated into the host tribe, unless he changed his name to that of the tribe. Such was the case of the jar (protege) (pp. 51 -52). “Some sources claim that the compact between two halifs established, or could establish, mutual succession between the two: whoever died first would leave a sixth of his estate to the other” (p. 53). But this, Crone tells us, is merely later Islamic information (Tabari’s Tafsir) which is “typically volunteered in explanation of Sura 4:37.” Typically, she quickly dismisses this informa­tion as apocryphal, without giving a good reason for doing so. But she finds it a little harder to dismiss other sources on this point, especially the fact that “classical Muslim scholars sometimes identify the contractual client of Islamic law with the pre-lslamic halif and modern scholars [Nallino, Goldziher and Juda] also tend to seek the origins of wala^J in hilf But though hilf might establish a claim to succession, it does not in fact provide a prototype for wala^J of Islamic law”(pp. 53, 134).

The first reason Crone gives for rejecting the hilf origin of. Islamic wald^J is that hilf resembles only wald^J al-muwdlai, not wala^J al-^citq. But we may. well ask why it should resemble both? The latter wala^J could have scarcely been in question, since it is an entirely different type of relationship from the former; in form and content the two types of wald^J possess nothing in common except the first part of their compound names. Neither the pre-Islamic Arabs nor the Muslims could ever have confused the two, but Crone obviously does. The fact that the clients of wala^J al-muwdldt were not known as halifs (p. 53) is no proof that wala^J al-muwdldt did not derive from hilf. What is important is that the legal features of the latter are in many ways identical with the former; hilf. like wala^J al-muwdldt, was a compact/contract in­tended for mutual support, and neither, as we have seen, required any sort of detachment of the parties from their natal groups (assuming they had such groups). Moreover, mawld and halif were often inter­changeable terms, as in the poet’s words, al-mawla al-tdbi^c al-muhalif— >quoted by Crone herself in a different context (p. 55). A few lines later Crone quotes another poet who makes a distinction between mawld by hilf and mawld by kinship (the latter probably referring to wala^J al-^citq). Had the terms mawld and halif signified distinctly different mean­ings, the poet’s need to draw the distinction would not have arisen. In yet another context (p. 137, n. 160), Crone tells us that the term mawld was used to indicate halif It would thus seem highly likely (at least more likely than Crone’s alternative) that wattP al-muwdldt represented a modified continuity of hilf al-wala^J that consisted, as Tyan put it, of the “admis­sion of an individual to a clan, by an agreement with one of the members of this clan or by collective assent. This individual, known as mawld. is generally accorded the same social and juridical position, from the standpoint both of rights and of obligations, as the original members of the tribe. There is even established a right of succession between him and the members of the tribe with whom the agreement is made” (my italics).³³ That the majority of Muslim lawyers rejected wala^J al-muwalat altogether and that those who accepted it called it waler³ al-muwalat and not, for example, wala^J al-hilf may well have been due to the notion that the Prophet prohibited hilf in Islam (la hilfa fi al-Islami). As testified in an early tradi­tion, however, Muslims still thought of wa/a³in terms of hilf³⁴

Notwithstanding the considerable body of evidence opposing her thesis Crone remains persistent. Contrary evidence in Arabic poetry is rejected on purely infer­ential grounds (pp. 55-56). This evidence is dismissed also because the mawali are “almost always mentioned in connection with the theme of mutual help,” a theme which does not fit Crone’s image of the mawali as people who must be abused and scorned in order to qualify as mawali proper. “Thus,” Crone concludes, the evidence of poetry is “irrelevant to our concerns” (p. 55). The judgment that this evidence is irrelevant is not surprising in light of Crone’s misinterpretation of the legal effects of waler* al-muwalat. From this point in the book the reader begins to entertain many doubts about Crone’s use and abuse of evidence. Evidence is stricken, not after it has been carefully weighed and compared with other more conclusive evidence, but on the basis of long chains of inference in which the transition from one stage to another can easily be disputed, if not altogether invalidated.

Although the issue of halif is poorly resolved, our author proceeds to prove that the mawali in pre- Islamic Arabia stood in a collective relationship of dependence with their Arab patrons. Here, she cites the case of those Jewish tribes who were collectively dependent on other tribes (pp. 56-57). We do not dispute this fact, since a mere glance at the chronicles shows its truth. But collective affiliation is not the issue here. What we need is proof that wala^J between individuals did not exist in pre-Islamic Arabia, a proof Crone does not adduce. Again, and to the great frustration of the reader, Crone proceeds to ask the question, “What happened to the slave on manumis­sion?” (p. 58) —without settling, or even attempting to settle, the question of collective dependence. The failure to deal with this matter certainly demolishes Crone’s hypothesis number (i), especially since we have no reason to abandon the conclusions of such scholars as W. R. Smith, Goldziher, Tyan and others concerning the existence of individual waliP and hilf In fact a glance at such works as that of Tabari assures us of the existence of wn/d³ (which kind we do not always know) and hilf between individuals; we know, for instance, that Marthad b. Abi Marthad al-Ghanawi was the halif of Hamza b. ^cAbd al-Muttalib,³⁵ just as Barrad was the halif of Harb b. Umayya.³⁶ That hilf was concluded between two individuals is also obvious from the Constitution of Medina, where it is stated that a believer must not take as his halif the mawla of another believer to the exclusion of the latter.³⁷

In this context we should observe that the pre- Islamic hilf and wala^J, on both the collective and the individual levels, acquired a parallel in Islamic law. The Hanafi and sectarian schools that accepted wald^D al-mu w a Id t allowed a continuity of the pre-Islamic individual hilf whereas the remaining three Sunni schools endorsed collective wala^J and hilf in the form of the Islamic dictum al-wald Hl-Muslimin.^ The admission by the Hanafi and the sectarian schools of the individual wala^J should not preclude the likelihood that they also endorsed al-wala^J lil-Muslimin.

To the aforestated question about what happens to the slave on manumission, Crone’s answer is that there is the “implication that freedmen were not incorporated in their manumitter’s kin” (p. 60). Here, Crone at least admits that this is only an implication and that “none of the passages” used as evidence “explicitly mentions freedmen” (p. 60). The argument on this point, however, continues for one more para­graph, but is quite thin. The reader is then suddenly confronted with the statement: “Having established this point, we may proceed to an examination of the precise manner in which the freedman’s status was regulated” (p. 60)—when, in fact, the point has not been established at all.

And Crone continues: “The two questions of inter­est to us are, first, who was responsible for the freedman’s blood and blood spilt by him? And sec­ondly, who had a claim to his estate?” (p. 61). Her answer to the first, as might be expected, is that the manumitters “might well set out to avenge the freed­man, but it would of course be their own honour, not that of the freedman, that they would try to vindicate” (p. 61). It is the freedman’s own agnates, Crone tells us, who are responsible for his blood as well as the blood spilt by him. But since the manumitters often avenged the blood of their dead freedman, how do we know that it was always their own honor which they were vindicating? Crone does not explain. Who had a claim to the freedman’s estate? “The answer,” our author says, “is presumably his agnate... not the manumitter.” Here Crone concedes that “[t]here is no direct evidence on this point but it is hard to see on what ground the manumitter could have made such claim” (p. 62). To support the entire argument Crone offers three footnotes, none of which in any way constitutes evidence about the pre-Islamic Arabs. They refer exclusively to freedmen in Roman and Jewish law. Nevertheless, Crone concludes the argu­ment as follows: “In short, there is no reason to assume that the manumitter acquired a right to suc­cession by virtue of manumission itself” (p. 62). She follows this by a more general statement concluding the fourth chapter: “It will be admitted that there is little in the evidence to suggest that the legal institu­tion of wala^D came out of pre-Islamic Arabia” (p. 63).

The absence of a reason to establish the existence of x is insufficient to prove that x is non-existent. If 1 have little evidence to show that x exists in y, this does not mean that x does not exist in y, nor does it imply that x exists elsewhere (unless of course ‘little evidence’ is used rhetorically). Even if we accept Crone’s argument and set aside for a moment the fact that it is based on a fallacy, we cannot consider it conclusive. Therefore, it is pointless to proceed to build further arguments on it, something which Crone nevertheless does, and with confidence at that. She opens chapter 5 by asking: “If the personal clientage of Islamic law did not come from Arabia, where could it have come from?” (p. 64).

It is important to note that at this stage in the book wala^ al-muwalat drops out, never to appear again. Not only does Crone fail, as we have seen, to establish its absence in pre-Islamic Arabia, but she does not even make a claim of its having been borrowed. We are left then with the claim that the Islamic institution of personal clientage (wald^J al-^citq), and more speci­fically kitaba (manumission contract), has its origins in the Roman Near East. This Near Eastern institu­tion, we are told, is the paramone. It is a service contract by which a slave is manumitted, provided that he render certain services to his patron for a specific period of time during which he “remains” with his master. Crone does find some evidence which shows that “pre-Islamic Arabs were familiar with the institution of paramone, and it may have been known not only to the settled Arabs, but also the bedouin....

But it should be stressed that what they knew was the institution in its ancient Near Eastern form... not more sophisticated versions current in the Greco- Roman Near East” (p. 68). The evidence that Crone presents is clearly partial; a few indirect examples are not enough to establish that the institution in Arabia represents an ancient form. Positive evidence is needed to convince the reader that this is the case; without this Crone cannot legitimately proceed to argue that the “evidence relating to paramonar manumission thus confirms that the conquerors were influenced by the legal practice of their subjects” (p. 68).

But she does. In order to prove the influence, she sets out to show that the conquerors were familiar (in the period of pre-classical law) with a wide variety of methods of manumission that existed in the Roman Near East. She cites seven early traditions to the effect that manumission would be granted in return for (i) an obligation to remain; (ii) service for a specified number of years; (iii) payment of alimony; (iv) a share in the freedman’s estate; (v) celibacy; (vi) enslavement of the freedman’s or the freedwoman’s children; and finally (vii) manumission by premature apolysis. Now, (i) clearly existed in pre-Islamic Arabia as testified to in the case (quoted by Crone, p. 67) of Suhayb who made a hilf with Ibn Jud^can and stayed with him until he died. The last phrase “until he died,” which occurs in at least three other examples quoted by Crone from pre-Islamic Arabia, must be taken to indicate that the duration of the contract was an important factor. Thus, it is highly likely that (ii) did exist in Arabia. As for (vii), it also seems to have existed in pre-Islamic Arabia in one form of tasyib or another (pp. 67, 68). Tasyib, as Crone points out, is manumission without service or payment. What remains then is (iii-vi). While the cases of manumission mentioned in these four traditions have not in any way been proven to have been absent in pre-Islamic Arabia, they have not been admitted to Islamic law because they are of the resolutive type. Manumission of this type is one in which the slave is freed now provided he render service after his manumission. Islamic law recognizes only suspensive manumission, in which the slave pays for his freedom and only after doing so acquires the status of a freedman. Thus, none of these cases proves any borrowing. Even if we assume that they did not originate in pre-Islamic Arabia, their being known to Muslims did not entail their admission into Islamic law.

To be sure, suspensive manumission constitutes only one of many differences between Islamic and Roman provincial law. Before. we come to these differences, we should draw attention to the difference in names, which Crone dismisses as a negligible point (p. 74). We recall that one of the reasons for dismissing the institution of hilf as the progenitor of wala³ al-muwalat is that the two institutions have different names. The linguistic difference between ki~ taba and paramone, however, is deemed by Crone to be unimportant because this time it obviously serves her purposes to do so.

In a similarly easy and arbitrary way Crone also dis­misses the kitaba mentioned in the Qur³an (S. 24:33), holding it to refer to a marriage, not a manumission, contract. The discussion of this crucial point is rele­gated to a footnote (p. 145, n. 100). Sura 24:1-35, she explains, “is concerned with sexual morality.... It covers subjects such as zina, qadhf li^can, and female modesty, not manumission.” The textual sequence of the Sura and the general context dictate, Crone insists, that the QuPanic expression fa-katibuhum should be taken as meaning a marriage contract, just like the Jewish matrimonial contract known as ketu- bah. Thus, for Crone, the thematic continuity of the Sura is sufficient to preclude the possibility that the Qur³an meant manumission, although elsewhere she (together with M. Cook) rebukes the Qur^Dan for being so “perfunctory in its linking of disparate materials.”³⁹ At one time the Qur^Dan is perfunctory, at another coherent, depending on when it is of service to Crone. She does not account for the fact that the Qur^an never employs terms derived from k-t-b for marriage, and that it consistently refers to marriage with words derived from n-k-h and z-w-y.⁴⁰ Indeed, one must be unduly confident, especially on the basis of such extremely poor reasoning, to state that the “exegetes and lawyers alike were unfamiliar with the Qur^Danic use of the word kitab" (p. 145).

Aside from the “crucial difference” of kitaba being suspensive and paramone being resolutive, we must look at other differences between Islamic and Roman provincial law, at least those acknowledged by Crone. They are as follows: (i) as a rule, paramone involved service; kitaba involved payment; (ii) kitaba allows the freedman to leave the residence of his master during the fulfillment of the contract; paramone does not; (iii) kitaba for the lifetime of the manumitter is strictly forbidden; paramone allows it. The first differ­ence represents another major difficulty in Crone’s thesis, for it opens the door to a yet more serious problem: the raison d'etre of kitaba is manumission; i.e., without manumission there is no kitaba. Para­mone, on the other hand, does not necessarily involve manumission, and its raison d'etre is providing ser­vice. It is a contract of labor that covers “wide ranges of arrangement from virtual self-hire to hire of ser­vices from specific jobs and periods with terms of protection for the worker as well as the employer.”⁴¹ In paramone what all the parties to the contract are interested in is labor,⁴² not manumission. It should not therefore be surprising that in S. Treggiari’s study of Roman contracts for manumission, paramone plays no role whatsoever.⁴³ It is the reduction, or rather the virtual suppression, of these differences—as well as the overstretching of similarities between provincial and Islamic law—that is characteristic of Crone’s mode of argument.

Thus far, Crone has not proven any single influence upon Islamic law from Roman provincial law. She has merely shown that Muslims, like the provincials before them, freed slaves for some kind of compensa­tion. Does the fact that both engaged in the same practice force us to conclude that Muslims borrowed from Roman provincials the idea of freeing slaves for a compensation? Such a conclusion is necessary only if we operate on the assumption that Muslims needed the example of the Greeks to learn how to be charitable.

Crone then discusses five features of manumission reflected in early traditions. Her aim is to “demon­strate” that early Islamic lawyers were familiar with provincial methods of manumission and that “the legal behaviour of the late Roman patronate coincides with that of pre-classical wala^D in certain respects: there was actual continuity in terms of law” (p. 78). We need not go into these features since Crone herself does not make the claim that any of these succeeded in entering Islamic law. Immediately thereafter, how­ever, she introduces four other features that are alleged to constitute similarities between the Roman patronate and pre-classical wala^J (pp. 84-86). They are as follows: (i) both paramone and wala^J, “con­ferred a title of succession in the absence of certain heirs, not a fixed share.” This similarity, however, must not be attributed to provincial influence because the manumitter stands before law as the kin of the mukatab freedman, and therefore he inherits as a relative (al-wala^J yajri majra al-nasab) in accordance with the Islamic rules of inheritance which were regulated to a considerable extent by the Qur³an. The patron could not inherit in any other way. Insisting on provincial borrowing in this case is tantamount to claiming that the Islamic law of inheritance was influenced by, or borrowed from, the Roman pro­vinces—a proposition which has not been proven. Another alleged similarity under this category is “that the patron was entitled to the entire estate of a freedman who died leaving neither relatives nor will.” It is only normal that this should be the case. In a slave-master relationship, it stands to reason that in the absence of any relatives on the side of the slave, the master, his original owner, should inherit from him. How could this be otherwise? (ii) “[M]ost of the rights vested in the Roman patronate over freedmen passed to the manumitter’s heirs in accordance with the ordinary rules regarding the devolution of prop­erty. On the Muslim side succession to the patronate would likewise seem to have followed the ordinary rules in pre-classical law” (my italics). But all we have from Crone concerning the Muslim side is specula­tion, backed by not a shred of evidence, (iii) “Both could be renounced.” There is, however, the equally valid possibility that the Islamic renunciation of the patronate could have been an extension or a modified form of the pre-Islamic tasyib, which we have already discussed. At any rate, the evidence Crone adduces with regard to this practice in provincial law is not convincing and cannot be taken seriously, (iv) “Both failed to arise if the freedman had... purchased both manumission and freedom from further obligations.” Crone does not in fact establish that this law survived after 212 a.d. The argument here is shaky and is couched in highly speculative terms (e.g., “it stands to reason that...,” or “we may thus take it that...”). But even if we assume for the sake of argument that the law did survive and that nothing like it developed

23

in pre-Islamic Arabia—a pojnt Crone never settles— we are.still left with the fact that it was rejected by all jurists except some Zaydis, some Ibadis and the Imamis (p. 86).

In summing up the last chapter Crone admits that “several aspects of the Roman patronate fail to re­appear oh the Muslim side.” But, she adds, “it should be abundantly clear that the Arab conquerors soon came to be familiar with a cluster of provincial practices in which the predominant element was the paramone of the Hellenized Near East.... In both positive and negative terms, the evidence thus points to the conclusion that wala^J has its origin in the Roman Near East” (pp. 86-88). These quotations, in fact, are not summaries of the findings of the book but rather are unwarranted leaps from one premise to another. As we have seen, not a single law regulating the institution of kitaba was conclusively proven to have infiltrated the Shari^ca. Furthermore, the differ­ences between the two legal institutions (even to the extent acknowledged by Crone herself) are sufficiently numerous as to cast serious doubts on any alleged borrowing. The fact that Muslims became familiar with the provincial legal tradition does not in itself constitute evidence of borrowing, and the value of any evidence based on this familiarity is further dimin­ished, if not eliminated, when one gives full credit to the ingenious processes of legal systematization and cultural synthesis that took place in the first two centuries of Islam. But what is most striking about Crone’s book is that while making categorical claims of Islamic indebtedness to Roman provincial (and Jewish) law/ it rarely attempts, much less succeeds, in showing influences that survived into what has be­come the Shari^ca.

The inability to demonstrate conclusively influences is not in the least surprising. It is commonly acknowl­edged by students of Islamic and comparative law that the Shari^ca as a “finished product” resembles none of the legal systems that it came into contact with during the period of its formation. Even if some similarities appear, they are quickly frustrated by serious differ­ences. The ingenious process of assimilation, systema­tization and Islamicization managed to dissipate all the indigenous features of legal institutions and to recast them in a fashion that is not in the least reminiscent of the older institutions. Systematization in particular had a powerful effect on reshaping and reformulating whatever raw legal material Islam encountered. This systematization was given sharp expression in the profound desire of Muslim scholars for logical co­herence, while at the same time they took into full consideration what they deemed to be divinely inspired propositions. The law which Muslims encountered in the conquered territories had, as a rule, to be modified to accord with the laws laid down in the Qur³an. It is no wonder then that attempts to prove some specific debt of Islamic law to other legal systems in conclusive and clear-cut terms have resulted in failure. Crone’s attempt shares this failure, for she obviously does not appreciate the Shari^ca’s inner logical structure that gives Islam its unique law. The fundamental structural differences between Islamic and ancient Near Eastern legal systems makes the identification of influences on the former by the latter virtually impossible; for though there persists a nagging suspicion of borrowing, this suspicion can never be confirmed, at least not with the present state of documentation. All that we, in fact, have are some vague similarities between indigenous laws and the Shari^ca that can never conclusively establish borrowings.

In the concluding chapter of the book, Crone takes an unexpected turn that makes nonsense of the thesis she has tried to defend throughout the book. At this point serious contradictions appear. We recall that a consistent element in Crone’s argument was the theme that pre-classical and classical Islamic law could not have been derived in any significant way from pre- Islamic Arabia, and that none of the similarities she was forced to acknowledge could, in her opinion, account for the institution as it appeared in Islam (“... if the Arab conquests had not included Roman provinces, there would not have been an Islamic patronate of the type we know” [p. 64]). More spe­cifically we recall that Crone rejected the institution of hilf as the progenitor of wala^J al-muwalat. However, in the last pages of the book, when speaking about the widespread practice of provincial law in the Near East, Crone acknowledges, in a preliminary manner, that this law was “shared by many or most of the inhabitants of the Near East... indeed even, it would

seem, by pre-lslamic Arabs themselves to some ex­tent” (p. 92). But then she goes further: “As has been

seen, manumission with paramone was known to the pre-lslamic Arabs too, as was the paramonar service contract.” “[W]e can see that the halif who arranged for mutual succession between himself and his ally was in fact including a succession pact of the type common in the provinces; and when the Prophet paired off Ansaris and Muhajirun as brothers in Medina, he was undoubtedly practicing the Near Eastern adoptio in fratrem, a special form of the succession pact whereby two men would adopt each others (sic) as brothers and institute both common property and mutual succession” (p. 93). Now, if the pre-lslamic Arabs were familiar, even “to some ex­tent,” with provincial law, why does one fail to see this fact noted in Crone’s work, except in the conclusion and then only in passing? We must ask why she is so adamant, even at the expense of twisting and ignoring evidence, in trying to prove that the Islamic patronate was a provincial and not a peninsular institution. Although the halif provides an excellent possibility of demonstrating legal continuity between pre-Islamic Arabia and Islamic law, she has dismissed it and instead opted for provincial legal precedents. None­theless, in her concluding remarks, she concedes that hilf could have been the source of wala^J al-muwaldt. What are we to make of these contradictions?

In the last few pages of the concluding chapter, it becomes evident that Crone is not, after all, entirely ignorant of the significant extent to which the Islamic patronate finds its origins in pre-Islamic Arabia. However, had she acknowledged the existence and developed nature of the pre-Islamic Arab institution in the beginning of her work, she would have dealt her project a fatal blow, for the entirety of her thesis rests on the assumption that the peninsular Arabs were not in any meaningful way in possession of the institution as it came to be known in Islam. To mention the undeniable influence of the pre-Islamic peninsular institution only toward the end of her work marks a serious defect in the structure of the work, and leaves her thesis in complete disarray.

Confusion increases with yet another concluding remark offered “on the basis of the material reviewed in this book”; namely, that the “tribal legacy of the invaders in conjunction with Jewish concepts provided the Muslims above all with the capacity to reshape, though Jewish law certainly and tribal law possibly contributed raw material too” (p. 99). Assuming that the ‘sader knows what the grammatical object of “to reshape” is, he or she will remain puzzled as to how this statement follows from Crone’s work when Jewish law has not even been discussed. But she goes even further. Provincial practice, she says, “contained elements of Roman law in Syria and Egypt, just as it contained elements of Sasanid law in Iraq; and Roman law

27

certainly, and Sasanid law probably, entered the Shari^ca as a result” (p. 99).

Ill

We shall not argue against Crone’s conclusions, except to say that none of them follows from the evidence or the arguments presented in the book. It was precisely this speculative scholarship that Crone severely criticized when she dismissed the writings of Goldziher, Schacht and others on the grounds that they merely affirm rather than demonstrate what they set out to prove. Her own arguments are far-fetched, her use of evidence is tortuous, and she certainly approaches the subject with a predetermined mind. There is little doubt that Crone writes—here and elsewhere—with the underlying assumption that the Muslim conquerors were culturally impoverished. Like the barbarians who invaded Roman lands (and be­tween whom and the Muslims Crone is fond of drawing analogies), the Muslims could have possessed only little culture, and it follows, therefore, that the civilization or empire they built must have been due to other, ‘higher’ sources. When she comes to Islamic law, Crone is blinded by her enthusiasm for establishing provincial, Roman, and other influences. Although she knows that the institution of.paramone was first attested as early as the second millennium b.c. in Babylonia and that it took until the fourth century b.c. before it surfaced in Greek law, and still another five or six centuries to make an appearance in Roman law (p. 65), Crone never associates this law with the Semitic peoples who were the preponderant inhabi­tants of the area. If the institution of paramone was present in the Near East long before Greek rule, and if the institution of slavery in the pre-Islamic Near East was characteristically Semitic, as von Kremer rightly

28

asserted,⁴⁴ then one must be reluctant, when simi­larities urge the notion of cultural influences, to claim borrowing from non-Semitic sources.

To sum up, although it is admittedly written in an engaging and stimulating style,⁴⁵ Crone’s book says nothing of value that has not been said better before. For one, von Kremer’s statement on the subject in his Culturgeschichte, written over a century ago, remains superior to that of Crone. There is no important argument she adduces which was not stated more succinctly and more prudently by von Kremer. Fur­ther, his conclusion about the entire legal institution of slavery being of Semitic origin⁴⁶ is more in accord with the known facts of the Semitic history of the ancient Near East. Crone’s work, in contrast, is confused, methodologically deficient, and definitely a step in the wrong direction. Her thesis does not even meet the minimal requirements of “a working hy­pothesis,” something which she remembers to point out only in the last two sentences of the conclusion. Everyone will acknowledge that history can never aspire to the same degree of certainty as science, and to that extent we have no choice but to live with the problem; but when history is made to mock the most elementary principles of scientific thought, historians should be the first to resist.

* A review of Roman, Provincial and Islamic Law: The Origins of the Islamic Patronate. By Patricia Crone, Cam­bridge: Cambridge University Press, 1987. Pp. 178. Four appendices, bibliography, index. $39.50.

I should like to record my thanks to my colleagues Charles Adams, Donald Little, and Paul Walker for their valuable comments.

¹ Two essays representing the disparate claims for indebted­ness are: J. R. Wegner, “Islamic and Talmudic Jurispru­dence: The Four Roots of Islamic Law and their Talmudic Counterparts,” The American Journal of Legal History, 26,

I (1982): 25-71, and H. J. Liebesny, “Comparative Legal History: Its Role in the Analysis of Islamic and Modern Near Eastern Legal Institutions,” The American Journal of Comparative Law 20 (1972): 38-52. For other sources see Crone, Roman, 1-17, appendices 2-3, and the bibliography.

² Crone’s thesis concerning the role of provincial law in solving the problem of legal influences was alluded to a century ago by von Kremer and vaguely repeated in 1950 by Schacht himself. See Alfred von Kremer, Culturgeschichte des Orients unter den Chalifen, 2 vols. (Wien: Wilhelm Braumiiller, 1875-77), 534f.; Joseph Schacht, “Foreign Ele­ments in Ancient Islamic Law,” Journal of Comparative Legislation and International Law 32 (1950): 10 (11. 18-19),

II (1. 36). At the same time, we have no reason to preclude Iraq as a possible source of influence. Crone’s almost dog­matic insistence on Syria stems from her reconstruction in God's Caliph (with M. Hinds) (Cambridge: Cambridge Uni­versity Press, 1986). See Norman Calder’s useful review of this book in the Journal of Semitic Studies 32.2 (1987): 375-78.

³ Crone, Slaves on Horses (Cambridge: Cambridge Uni­versity Press, 1980), 80, as well as Part 1 for similar state­ments; Crone & M. Cook, Hagarism. The Making of the Islamic World (Cambridge: Cambridge University Press, 1977), 73.

⁴ Crone, Slaves, 23.

⁵ Ibid., chapter II; Crone, Meccan Trade and the Rise of Islam (Princeton: Princeton University Press, 1987), 240.

⁶ Crone, Slaves, 23, 25, 221 (n. 163).

⁷ Crone, Meccan Trade, 244. Crone’s intense hostility toward her subject has been observed by a number of Islamicists. See, e.g., Fred Donner’s review of her Slaves on Horses in the Journal of the American Oriental Society 102.2 (1982): 367-71; also Richard Bulliet’s review of the same work in the Middle East Studies Association Bulletin 16.1 (1982): 88-89; Norman Daniel’s review of Hagarism (by Crone and M. Cook) in the Journal of Semitic Studies 24.2 (1979): 296-304; Donald Little’s review of Meccan Trade in the Canadian Journal of History (forthcoming).

⁸ Crone, Slaves, 24.

⁹ R. E. Briinnow and A. von Domaszewski, Die Provincia Arabia, 3 vols. (Strassburg: K. J. Triibner, 1904-1909).

¹⁰ Franz Altheim and Ruth Stiehl, Die Araber in der alien Welt, 5 vols. (Berlin: Walter de Gruyter & Co., 1964-68), including significant contributions by others.

¹¹ Th. Noldeke’s important contributions to this field are too numerous to list here. For the most relevant ones see the bibliography of I. Shahid in Byzantium and the Arabs in the Fourth Century (Washington, D.C.: Dumbarton Oaks Re­search Library and Collection, 1984).

¹² Th. Mommsen, The History of Rome, trans. W. P. Dickson, 4 vols. (New York: Charles Scribner’s Sons, 1886).

¹³ Rene Dussaud, La Penetration des arabes en Syrie avant Plslam (Paris: Paul Geuthner, 1955).

¹⁴ See G. Bowersock, Roman Arabia (Cambridge, Mass.: Harvard University Press, 1983), including a number of his articles listed in the bibliography. Other important contri­butions were made by scholars such as A. A. Vasiliev, D. Graf, Franz Cumont, Trimingham and others. See notes below.

¹⁵ But the contributions of Altheim and Stiehl, Bowersock, and Noldeke do not seem to be totally unknown in such studies as Meccan Trade and Hagarism.

¹⁶ 1. Shahid, Rome and the Arabs (Washington, D.C.: Dumbarton Oaks Research Library and Collection, 1984), and see note 11 above. The other volumes dealing with Arab-Byzantine relations in the fifth and sixth centuries are to appear in the near future.

¹⁷ It is well known that when the southern Arabs deepened their penetration into Syria after the decline of the Seleucids, there were already other Arabs who had established their own dynasties in the area. When the Romans appeared on the Near Eastern scene, Arab groups existed south of -the Taurus, in Antioch, Edessa, Chalcidice, Palmyra, Emesa, and Arethusa. Three of these groups, the Ituraeans, the Nabataeans and the Idumaeans, were in possession of Syria (Shahid, Rome, 4-5; Bowersock, Roman Arabia, chs. 1-11; Mommsen, History, IV:63ff., 162ff., passim). Most of these Arabs had succeeded in maintaining their identity despite the influence of the Greco-Roman culture. Dussaud has shown the impressive extent to which they conserved their char­acter; the Nabataeans, for instance, kept their Semitic rites, Arabic language, as well as their traditional laws and an­cestral customs (see his La Penetration des arabes, 205, passim; see also Shahid, Rome, ch. I, especially p. 10; Bowersock, “A Report on Arabia Provincia,” Journal of Roman Studies 61 [1971]: 219-42). “[U]nder the native princes,” Fr. Cumont asserts, the Arabs “kept their own alphabet, their own religion and traditional laws” {Cambridge Ancient History [Cambridge, 1936], XI:616). No less impor­tant is the fact that the old tribal organization of the Semites remained very powerful, as is testified to in the case of the Nabataeans (Bowersock, Roman Arabia, 154; Cumont, CAH XL622-24). Perhaps the most important factor that con­tributed to the preservation of the identity of these Arabs was the constant waves of penetrators and immigrants from the Peninsula (Cumont, CAH XL614; J. S. Trimingham, Christianity among the Arabs in Pre-Islamic Times [London & New York: Longman, 1979], 3-4). These unceasing incur­sions must explain both the ongoing cultural replenishment of the northern Arabs and Roman need for the oriental limes (Shahid, Rome, 8—12; D. F. Graf, “The Saracens and the Defense of the Arabian Frontier,” Bulletin of the American School of Oriental Research 229 [1978]: 13). Other external factors also contributed to the preservation of the Semitic character of the Syrians. The Seleucids never attempted the Hellenization of the peoples they ruled, and their admini­stration which took a network of cities as its basis failed to touch Syrian society at large (Trimingham, Christianity, 22). Hellenization could reach the municipal middle class only after the Roman conquest, at which time this relatively small class could only supply the provincial administrators. (See G. H. Stevenson, Roman Provincial Administration [Ox­ford: Basil Blackwell, 1939], 85, where he speaks of the absence of a Roman class of civil servants in the provinces.) Again, the Semitic population, which was to survive the Seleucids, the Romans and the Byzantines, was debarred from entering the inner layers of Hellenic culture (Cumont, CAH XI:613, 620, 639-41). Syria was no more Hellenic than Mesopotamian Hatra was Iranian. On Hatra’s conservation of its Semitic character under the Persians, see J. Segal, “Arabs in Syriac Literature,” Jerusalem Studies in Arabic and Islam 4 (1984): 91.

Commercial interests constituted the backbone of Rome’s and Byzantium’s relations with the Arabs of the Syrian desert and North West Arabia. On the basis of the Rawwafa inscriptions (and, partly, the Notitia Dignitatum), D. Craf has established that in the second and beginning of the third centuries Rome was on good terms with the Thamud tribal federation, whose elements encompassed up to one hundred tribes extending from al-^cUia in the south (about 250 km. south of Tabuk) to the Syrian desert in the north (Graf, “Defense,” 3, 10—12; see also Bowersock, Roman Arabia, 13If. On the Arabs in the Notitia Dignitatum, see Shahid, Rome, 51-63). In these inscriptions, Graf argues, “Rome is presented as the major contributor to the erection of the tribal temple of the Thamudic federation and an esteemed intermediary in their intratribal disputes. This sanctuary and its inscriptions also reveal that such tribal structures were neither primitive nor culturally isolated; for centuries before their encounter with Rome they had been in contact with the major Near Eastern civilizations and involved in the inter­national traffic of the Orient” (Graf, “Defense,” 20). The Thamud tribes were known even before this time to have cultivated land with sophisticated systems of irrigation and to have possessed a boat-building industry on the coast of Midian (Trimingham, Christianity, 15-16). On the eve of the Islamic conquests, the Judham tribal confederation, which was related to ^cAmila and Lakhm and which seems to have supplanted Thamud, was regarded as the guardian of the Syrian limes. Their area of control extended from the Dead Sea in the north down to the southern parts of Wadi al- Qura, less than one-hundred miles north of Yathrib (Triming­ham, Christianity, 120-23). These Arabs, with their hulafcC— the Bahra⁰, Bali, Wa^il, Bakr, Ghass&n and Tanukh—were the commercial intermediaries and cultural transmitters be­tween Mecca and Byzantium (Shahid, Byzantium, 15, 18; on the role of Arab foederati in Byzantine-Arab relations, see Shahid, Rome, 29-32, 128f., passim; idem, Byzantium, pas­sim). Syrian and Mesopotamian Arab merchants operated in Mecca, and the Ghassanids, as the hulafcP of Banu Asad, had an establishment in the neighborhood of the Ka^cba. Representatives of the Syrian (and Mesopotamian) North were to be found in Mecca. The Meccans were constantly in touch with their northern brothers and, as the Qur°an shows (S. XXX: I -3), they were fully aware of the wars between Byzantium and Persia (Trimingham, Christianity, 258-60, passim). Abu Talib, Muhammad’s uncle, could not have been the only Meccan to frequent Bostra, and Abu Sufyan himself must have been doing what other Hijazi merchants had done when he purchased a farm near Damascus before Islam (Trimingham, Christianity, 258; Shaban, Islamic His­tory [Cambridge: Cambridge University Press, 1971], 1:43). And of particular interest is what the sources tell us not only about the thriving cities of the Hijaz but also about the large slave market at Mecca. It is no coincidence that Zayd b. Harith, the adopted son of Muhammad, was a Syrian slave. See Trimingham, Christianity, 260; E. R. Wolf, “The Social Organization of Mecca and the Origins of Islam,” South­western Journal of Anthropology 7 A (1951): 334; S. Smith, “Events in Arabia in the 6th Century a.d.,” The Bulletin of

34

the School of Oriental and African Studies 16 (1954): 467.

Furthermore, when the Muslims occupied Syria, the local Arabs already formed the principal part of the population; many of the Greeks who inhabited the area left when the Muslims arrived. (A. A. Vasiliev, “Notes on Some Episodes Concerning the Relations Between the Arabs and the Byzan­tine Empire From the Fourth to the Sixth Century,” Dum­barton Oaks Papers 9-10 [1956]: 309; Shaban, Islamic History, 41.) Almost every Arab tribe or sub-tribe was, to a gieater or lesser extent, represented in Syria. These tribes had long been established in the area either as half-nomads or as full sedentaries. (See the article of A. Fischer “Kais- ^cAilan,” and that of C. Brockelmann “Kalb B. Wabara,” in the Encyclopaedia of Islam, II:ii, 653, 688-90. Also W. Madelung, “Apocalyptic Prophecies in Hirns in the Umayyad Age,” Journal of Semitic Studies 31.2 [1986]: 142.) Indeed, it is these people who undertook the preliminary tasks of building an empire, and it is these people, with the Arab immigrants from the south, who claimed to be the experts on hadith', in Hirns, for instance, a strong school of hadith arose, “drawing its inspiration” from the almost five hundred Companions who came to the city. See Madelung, “Apoca­lyptic,” 143.

¹⁸ See note 17 above, and Dussaud, La Penetration des arabes, passim.

¹⁹ See note 17 above.

²⁰ Although in her Meccan Trade, 153, “southern Syria and Hira formed part of Arabia in all respects except one.” The exception is unclear, but I take it to be political domination. Further, on Byzantine trade and influence in the Peninsula, see ibid., 115ff., 138-40, 151, 246. Crone’s thesis in Roman, Provincial and Islamic Law obviously requires the suppres­sion of the facts about Byzantine-Arab relations which she admits in her Meccan Trade.

²¹ If the space allotted to the treatment of a subject is any indication of its importance, then walcP stands as one of the least important of legal subjects expounded in the works of

furu^c. For example, the space given to the discussion of walcP al-muwdlat, accepted only by the Hanafi and sectarian schools, is less than one page in the more than one thousand pages of Marghinani’s Hiddya. In al-Fatdwd al-Hindiyya it occupies a page and a half out of about 3200 pages. It is interesting to note that the legal status of hermaphrodites normally takes more than double the space devoted to wala^ al-muwdlat.

²² This is clearly understood from Al-Fatdwd al-Hindiyya, 6 vols., 3rd ed. (Beirut: Dar Ihya³ al-Turath al-^cArabi, 1980), V:26, 32; Abu Ja^cfar Muhammad b. ^cAli al-Tusi, al-Khilaffi al-Fiqh, 2 vols. (Tehran: Matba^cat Rangin, 1377-1382 h.), II:648ff.; Burhan al-Din al-Marghinani, al-Hidaya, 4 vols. (Cairo: Matba^cat al-Babi al-Halabi, 1980), 111:271.

²³ Al-Fatawa al-Hindiyya, V:32 (11. 29-30); Tusi, Khilaf, 11:652; Muhammad ^cAli al-Turi, al-Bahr al-RdPiq, 8 vols. (Cairo: al-Matba^cat al-^cHmiyya, n.d.), VIII:77 (11. 26-27).

²⁴ Shams al-Din al-Sarakhsi, al-Mabsut, 30 vols. (Cairo: Matba^cat al-Sa^cada, 1324-1331 h.), VI11:91 (11. 13-14), 92, 96 (1. -2).

²⁵ Al-Fatdwd al-Hindiyya, V:32.

²⁶ Sarakhsi, Mabsut, V1II:96 (1. 12).

²⁷ Al-Fatawa al-Hindiyya, V:32 (1. —4).

^{28 c}Abd Allah b. Mawdud al-Musili, al-Ikhtiyar li-Ta^clil al-Mukhtar, 5 vols. (Cairo: Mustafa Babi al-Halabi, 1951), IV:44 (11. 17-18); Sarakhsi, al-Mabsut, VIII:95 (11. 9, -4). See also Tusi, al-Khilaf, 11:652 (mas^Jala 9).

²⁹ Tusi, al-Khilaf, 11:652. The language of offer continues: wa-tadfa^ca ^canni wa-adfa^ca ^canka wa-ta^cqila ^cani wa-a^cqila ^canka wa-tarithani wa-arithaka (11:652, 1. 4).

³⁰ Al-Fatawa al-Hindiyya, V:33; Tusi, al-Khilaf, 11:652 (11. 9 -10).

³¹ Al-Fatawa al-Hindiyya, V:32 ( — 9); Turi, al-Bahr al- Ra^iq, VIII:77. See also ^cAla³ al-Din al-Samarqandi, Tuhfat al-Fuqaha³, ed. M. Kattani and W. Zuhayli, 3 vols. (Damas­cus: Dar al-Fikr, 1964), 11:403, who adds the condition that the party making the offer must not have Muslim heirs in existence.

36

¹² Tusi, al-Khilaf, 11:652 (11. 7-9).

³⁵ E. Tyan, “Hilf,” Encyclopaedia of Islam, 2nd ed., 111:388-89; see also the references to his Institutions de droit public musulman', W. R. Smith, Kinship and Marriage in Early Arabia (Cambridge: Cambridge University Press, 1855), 45ff. who argues much the same.

^{34 c}Abd al-Razz^q al-San^cani, al-Musannaf, 11 vols. (Bei­rut: Dar al-Qalam, 1970-1972), IX, no. 16,140 (al-waltPu bi- manzilati al-hilfi... ).

³⁵ Abu Ja^cfar al-Tabari, Tankh al-Umam wal-Muluk, 13 vols. (Beirut: Maktabat al-Bayan, n.d.), 111:29-30.

³⁶ Cited by Crone, Meccan Trade, 145.

³⁷ Crone, Roman, 32; The Constitution of Medina, ed. and trans. M. Hamidullah (Lahore: Ashraf Press, 1968), 57 (12b), and 62 (37b) where hilf is clearly between individuals.

³⁸ Abu Ishaq al-Fazari, Kitab al-Siyar, ed. F. Hamada (Beirut: Mu^assasat al-Risala, 1987), 173 (no. 212); Crone,, Roman, 87(11. 14-15).

³⁹ Crone & Cook, Hagarism, 18.

⁴⁰ Cf. under the relevant terms in M. Fu³ad ^cAbd al-Baqi, al-Mu^cjam al-Mufahras (Cairo: Dar al-Kutub al-Misriyya, 1945).

⁴¹ J. A. Crook, Law and Life of Rome (London: Thames and Hudson, 1967), 192-93.

⁴² Ibid., 193; see also pp. 201, 246 for particular services rendered under this contract.

³ S. Treggiari, Roman Freedmen (Oxford: Clarendon Press, 1969), 20-31.

⁴⁴ von Kremer, Culturgeschichte, 1:546-47.

⁴⁵ The book is nonetheless marred by a good number of typographical and diacritical errors. Some of these errors have occurred at the following pages: 7 (11. 6, 36), 11 (1. —6), 19 (1. 15), 20 (1. -14), 35 (1. -7), 37 (1. 3), 40 (1. -19), 43 (11. 16-17), 45 (1. 3), 46 (1. -14), 53 (1. 17), 56 (1. 26), 59 (11. 2, 8, 22), 60 (1. 15), 62 (1. 5), 69 (1. 3), 71 (1. 8), 81 (1. 18), 85 (1.2), 86 (I. 23), 87 (11. -14, -12, -9), 103 (1. 22), 107 (1. -4), 109(1. 14), 138 (1. -18).

⁴⁶ von Kremer, Culturgeschichte, 1:546-47; also 524ff.