Version classiqueVersion mobile

The rules of Barat. Tribal documents from Yemen

 | 
Paul Dresch

Introduction

Entrées d'index

Mots clés :

imamat zaydite

Keywords :

Zaydi imam

To the tribes of Barat,
the “firebrand of the Arabs” as the chronicles call them, and these days an admirable group of people.
Also to the late Bob Serjeant,
an admirable person in his own right.

‏إهــداء خاص لقبائل برط
”جمرة العرب“ كما كانوا يدعونهم قديماً
والذين مازالوا كذلك إلى وقتنا الحاضر
وإلى الفقيد بوب سيرجنت،
الشخص الرائع بكل ما تحمله الكلمة من معنى.

Texte intégral

  • 1 Rashād al-ʿAlīmī: al-Qaḍā’ al-qabalī fī l-mujtamaʿ al-yamanī (1986) p. 43. We shall touch again on (...)
  • 2 Faḍl Abū Ghānim al-Bunyah al-qabalīyah fī l-yaman (1985) pp. 361 ff. A mention in the text of Arḥab (...)
  • 3 For the association of these phrases with Baraṭ, see R.B. Serjeant, The interplay between tribal af (...)

1The documents presented here are eighteenth-century agreements among tribesmen from Jabal Baraṭ, east of Ṣaʿdah and south of Najrān at the edge of the desert world. Some refer to the main text (texts A and B, below) as simply qawāʿid baraṭ, the Rules or the Code of Baraṭ. Others equate it with the legendary qawāʿid al‑sabʿīn. This latter phrase belongs perhaps with such names as that of Bin Zinbāʿ, whose house is pointed out in places as far apart as al-ʿUṣaymāt (near Shahārah) and Khawlān al-Ṭiyāl (east of Ṣanʿāʾ) and who himself was supposedly a jurist of note among Yemen's tribes but otherwise is lost in the mists of time.1 The phrase qawāʿid al‑sabʿīn is used by Faḍl Abū Ghānim of a different text, written in archaic script, of obscure provenance.2 The name is used sufficiently often of the present code, however, whether by those who know it well or by those who know only rumours of its existence, that provisionally I incline to accept the equation and translate the phrase qawāʿid al‑sabʿīn in as the Rules of the Seventy Shaykhs. A widespread story says that once, long ago, seventy shaykhs signed a document committing themselves to ancestral custom; an alternative tradition says qawāʿid al‑sabʿ or al-qawāʿid al‑sabʿ, for perhaps there were once only seven shaykhs or perhaps even seven codes.3 The documents at hand concern primarily the tribe of Dhū Muḥammad.

  • 4 ʿAbdullāh ʿAlī al-Wazīr Ṭabaq al-ḥalwā wa-ṣiḥāf al-mann wa-l-salwā (Muḥammad Jāzim ed. 1985) p. 223 (...)
  • 5 al-Nubdhah al-mushīrah ilā jumal min ʿuyūn al-sīrah fī... l-qāsim bin muḥammad (anon. n.d.) p. 406. (...)
  • 6 Muḥammad Zabārah Āʿimmat al-yaman bi-l-qarn al-rābiʿ ʿashar (1956) vol. 2 p. 8.

2It may be that behind these texts, and others from the Jawf as well as Baraṭ, lies a so-called “mother code” (al-qāʿidah al-umm) from which all derive, but this for the moment is scattered hearsay. The term umm or “mother” can be used of any original document (it occurs in lists of copies of the texts below) and the formulae in tribal documents of “renewing” ancestral practice suggest indefinite regression to a mythic past. Nor is the term qāʿidah (pl. qawāʿid) easy to pin down, despite how commonly one finds it used in historical sources. In the 1660s, for instance, a group of tribesmen from Baraṭ are reported to have pillaged a caravan in the territory of Sufyān, who demanded from them payment for the insult ʿalā qāʿidah-hum.4 To translate this “according to their usage” might not be wrong (a qāʿidah can be something as abstract as a rule of grammar); probably, however, there were written agreements. There certainly were further west, two generations earlier, when the people of Zafīr Ḥajjah refused to let the Imam al-Qāsim garrison the place in his fight against the Turks: it was hijrah, or protected, and this was expressed in qawāʿid, some parts of which, says the chronicler, were soundly Islamic and others not.5 Early in the twentieth century, at the start of his rising against the Ottomans, Imam Yaḥyā took qawāʿid and ḍawābit from the tribes to the effect that, for instance, spoils be divided fairly and that he have control of captured artillery.6

  • 7 For an 18th century example, Ḥusayn al-ʿAmrī Mi’at ʿām min tārīkh al-yaman al-ḥadīth (1988) p. 365. (...)

3In tribal usage the term qāʿidah would usually mean a written document, something like a pact or treaty. One might sign an agreement to recognize a paramount shaykh, for instance.7 Quite commonly in tribal circles we find qāʿidat or qawāʿid al-sūq or al‑ḥarām, which are documents providing for protection of a market (two examples are given in the present work); qawāʿid al-ḥudūd, meaning agreements or rules about tribal borders; and qawāʿid al-ṣaḥb, documents that affirm a tribeʼs solidarity. The main texts considered here (texts A and B, below) are referred to as qawāʿid al‑malāzim, roughly the Code or agreements setting out the tribe's responsibilities. They also comprise a pact of fellowship (ṣaḥb) or brotherhood (ikhāʾ). But why and how the first of them was drawn up on the date that it was remains a mystery. If this is indeed “the Rules of the Seventy”, or somehow derived from that mythic text, it may be that, rather than seventy shaykhs first convening to agree their custom, the meeting went on for seventy days or that seventy clauses were agreed. Further interpretations are possible and are sometimes the subject of speculation among tribesmen with a taste for the heroic past. Others, by contrast, may fear the whole subject is somehow ṭāghūt or wickedness.

Custom and (Imamic) law

  • 8 C. Rathjens, Ṭāghūt gegen Scherīʿah, Jahrbuch des Museums für Landes- und Völkerkunde (1951).
  • 9 Ibn al-Mujāwir Ṣifat bilād al-yaman wa-makkah wa-baʿḍ al-ḥijāz, ... tārīkh al-mustabṣir (Mamdūḥ Ḥas (...)

4These codes or rules form part of a world much affected by claims to exclusive truth. From the time of the first Zaydi Imam in Yemen (circa AD 900) we have rumours of tribal practice at odds with Islamic virtue, and through much of Yemen's written history we find mentions of texts and books of distinctively tribal law contradicting that of true Islam.8 In southern Yemen, where the term manʿ appears to have continued being used of tribal law, the two were perhaps seen as complementary. Ibn al-Mujāwir in the thirteenth century (AD), discussing the whole area, mentions the law of manʿ calmly enough although identifying it with the jāhilīyah;9 and R.B. Serjeant, whose experience lay mainly in the south, quotes a treatise from about the same period to the effect that arbitrators should “judge by the sharʿ in relation ... to the sharʿ, and by manʿ in that to which manʿ pertains”. In Upper Yemen, however, in the Zaydi sphere, such law was condemned quite often as ṭāghūt.

  • 10 See e.g. Ibn Hishām (d. AH 213 or 218), cited Dresch Tribes p. 196. The term ṭāghūt was used in sou (...)
  • 11 For such phrases as aḥkām al-ṭāghūt and ḥukkām ṭawāghīt see e.g. Ḥusām al-Dīn Muḥsin Ṭārīkh al‑yama (...)
  • 12 Dresch Tribes p. 199, from Nubdhah mushīrah p. 435. For the Khawlānī passage, Nubdhah mushīrah p. 2 (...)
  • 13 Zabārah Āʿimmat al-yaman vol. 3 p. 197. For the story of threatening execution, below, see Serjeant (...)

5The latter term itself is interesting, for although it occurs eight times in the Holy Qurʼān, and plainly refers to something reprehensible, scholars were already wondering what exactly it meant within 200 years of the Prophetʼs death.10 In many Yemeni formulations later the obvious association might be with tyranny, and most references cannot safely count as meaning ḥukm al-manʿ11. In the early seventeenth century, for instance, a shaykh of Khawlān is denounced as belonging to “the evil ones (ashrār) and the rulers of ṭāghūt”, which need hardly mean more than that the Imamʼs people thought him a wicked influence; one of the Imam al-Qāsimʼs appointees, however, is said more specifically to have taken from tribes their “books of ṭāghūt” and burned them."12 Imam Yaḥyā, before conquering al-ʿUṣaymāt in the 1920s, complained of “judgement by the ṭāghūt”, evidently meaning tribal custom,13 and in the 1930s is said to have threatened anyone owning books of ṭāghūt with execution.

  • 14 B. Messick The Calligraphic State: textual domination and history in a Muslim society (1993) pp. 14 (...)
  • 15 For a useful discussion of codification in Southeast Asia, W. Roff, Islam obscured? Some reflection (...)

6Such reports lead some authors to suppose there are two distinct codes of law –­even written law– inextricably at odds. Usually, in fact, this apparent conflict is a matter more of jurisdiction (and of course of political and moral rhetoric) than of textual contradiction or dissent from the Law of God. Some modern authors are concerned to show how compatible tribal custom is with sharīʿah or Islamic law; tribesmen for their part would not wish to be thought less than good Muslims. It has always been the case, more importantly for the present purpose, that Islamic law is free to use (indeed, by many reckonings obliged to use) what customary resources are to hand provided they are not at odds with sharīʿah principles, and that neither domain of law was categorized or bounded.14 No simple opposition of codes is possible. Indeed the idea of opposed and coherent codes is usually the product of Western empires. The Dutch in the East Indies thus turned adat or “custom” into adatrecht; the French in North Africa attempted to formalize Berber usage as an alternative to Islamic law and provoked widespread anger, not least among Berbers at the time.15 Why local productions should appear as “codes” to those who write them we shall touch on later.

  • 16 Muḥammad Jāzim (ed.) Nūr al-maʿārif fī nuẓum wa-qawānīn wa-aʿrāf al-yaman fī l-ʿahd al‑muẓaffarī l- (...)
  • 17 Ḥusayn al-Sayāghī (ed.) Qānūn Ṣanʿā’ fī l-qarn al-thānī ʿashar al-hijrī (Ṣanʿā’ n.d.), R.B. Serjean (...)
  • 18 Even this contrast between sharʿ and sharīʿah, although common in documents, should be treated care (...)

7The common form in Yemen, as elsewhere in Islam, would seem to have been a certain tolerance of subordinate legal patterns. In Rasūlid court documents, ably collected and edited by Muḥammad Jāzim,16 one finds hints of quite elaborate ʿurf or “custom” that governs such things as markets. Anyone familiar with modern-day disputes over grazing or water-rights knows how Islamic law uses local knowledge: for that matter the famous Statute of Ṣanʿāʾ (AD 1748), although scrutinized closely, was not denounced.17 But distrust of imposed authority is evident in tribal practice, which perhaps explains scholarsʼ enmity. In texts A and B, below, we find mention of “weak” neighbours, people of less than tribal status at Baraṭ who lived under the protection of the Baraṭ tribes, and text A section 19 (reproduced in part five, below) tells us that if they differ among themselves then tribesmen are not to take sides with them but drive them to a sharīʿah judge. Such people were known, in fact, as juhhāl al-sharīʿah, the sharīʿah being Islamic law and juhhāl being “children”, or more literally those who are “ignorant” and unable to care for themselves. To be a child of sharīʿah was to be a dependent person. Tribesmen, by contrast, were meant to resolve their own problems. If they had recourse to sharīʿah judges (as opposed to sharʿ, customary arbitration by those they knew),18 they did so ideally of their own volition.

  • 19 Ḥayīm Ḥabshūsh Ru’yat al-yaman (1992) pp. 37, 130, Yémen (Samia Naïm-Sanbar trans. and ed. 1995) pp (...)
  • 20 Rossi ‘Dirittoʼ p. 16, also p. 34. In Saudi Arabia, surprisingly, ṭāghūt may perhaps be a term used (...)

8It seems to be tribal custom, not the practice of peasants, traders or fishermen, that attracted the term ṭāghūt from learned Zaydīs, and Ḥayīm Ḥabshūsh in the late nineteenth century reports ṭāghūt as simply the name of tribal law; so too does Eduard Glaser.19 From manuscript material collected in the 1930s, however, Rossi was forced to conclude that the ḥukm al-ṭāghūt condemned by scholars corresponded “only in part” to what actually was tribal law, and the degree to which this depiction of tribal customary judgement (ḥukm qabalī, ʿurf qabalī) as ṭāghūt was internalized as self‑description remains unclear.20 What we can say is that a fairly coherent tradition, or set of traditions, existed that seems very different from those of the schoolmen and that claims its legitimacy in different style but matches the learned traditions in sheer longevity. The documents at hand represent what passes as “tribal” law quite well. The earliest of them claims to date from the equivalent of AD 1763.

  • 21 See e.g. M. Mundy, Womenʼs inheritance of land in Highland Yemen, Arabian Studies, vol. 5 (1979).
  • 22 Messick Calligraphic State pp. 31, 35, 61 and passim.

9Even Islamic law is dealt with too easily by writers on the Middle East. For instance, learned tradition within Islam often accuses the common people of denying inheritance to women, yet anthropologists and historians alike must surely be struck by how often in fact people try to follow the Qurʿānic rules in circumstances where evasion would be much easier: there is no real equivalent to the formalized ḥiyal (“devices”, “tricks”) by which merchants squared the letter of the law with practice.21 On the other hand, many usages, not least to do with kinship, are thought of as “Islamic” which in fact have no scriptural warrant. The ideal of such warrant recurs everywhere and with it an ideal of moral certainty. The jurisprudence (fiqh) which deduces law from scripture, and which its practitioners sometimes opposed rhetorically to the common peopleʼs ignorance, was itself meanwhile “open”, elliptically phrased and matter for debate among the learned.22 It was also, as scholars from many schools expressed this, like a sea without limits. In effect the sharīʿah or Islamic law is the whole of society conceived, and argued over, from a certain viewpoint: so too perhaps, more locally, was “custom”. But the practical uniformities of scholarly tradition are themselves not at all a mystery. Learned persons read each otherʼs work and travelled extensively to learn from each otherʼs teachers, and Islam in that sense is no more a question of a vague, insubstantial Geist than is contemporary physics or biology.

  • 23 The phrase histoire historisante comes of course from Blochʼs Métier d'historien.

10The geographical spread and historical continuity of non-formal law (of ʿurf or “custom”) is harder to explain. No obvious institution sustains it, and those who have absorbed the metaphysic of histoire historisante prefer to ignore the subject.23 Within the Arab world, however, repetitions of custom are often striking. Within what some would call “tribal” worlds the reach is greater still, and people interested in, let us say, the Berber fringe of the Maghreb find an odd sense of familiarity as they read of rural (Arab) Iraq, or those interested in the pastoral world of (Arab) Nejd find themselves not wholly on foreign ground as they read about hill-farmers in (Pakhtu) Afghanistan, particularly as regards protection, compensation, and forms of settlement. Certain principles of social organization (or better, of self-definition) recur widely. Indeed, one can reach well beyond what is now called the Middle East and find parallels between the customary law of many Arabic-speaking groups and the dooms (so called) of Anglo-Saxon England, law-texts from Celtic Ireland, or the legal codes of early Iceland (part seven, below).

  • 24 M. Meeker, Meaning and society in the Near East: examples from the Black Sea Turks and the Levantin (...)
  • 25 Compare Yemeni material with e.g. G. Murray Sons of Ishmael: a study of the Egyptian Bedouin (1935) (...)

11Structural principles recur. Rather widely within the Middle East, however, one also finds repetitions of key terms and of phrases with related meanings. Behind this, of course, lie such larger questions as to how and why the term nāmūs (presumably the classical Greek nomos, well known to philosophers) recurs so widely as denoting a vulnerable moral integrity, and why the Arabic ḥurmah should recur from North Africa to Afghanistan with the meaning of that which a man must protect from others.24 Yet the detail of repetition in the Arab world is impressive. If compensation for wounds is often calculated in formally the same way, we may suspect an Islamic model; but four-fold compensation for other wrongs finds echoes in Exodus, while minor resemblances of procedure and terminology link North and South Arabia, and the Maghreb too.25 Much work needs doing on which of these patterns are, so to speak, parasitic on the great tradition of Islamic justice and which of them suggest a lost common origin of the kind that is likely with “Greek” medicine or with magic. To speak in each instance simply of ʿurf or “custom” is to skirt large problems.

  • 26 F.H. Stewart, Tribal law in the Arab world: a review of the literature, International Journal of Mi (...)
  • 27 R.B. Serjeant, Materials for South Arabian History, Bulletin of SOAS vol. 13 (1950), Two tribal law (...)
  • 28 Serjeantʼs copy differs somewhat from Rossiʼs account and suggests perhaps access to further copies (...)

12Recent studies of tribal customary law in the Arab world are summarized by Frank Stewart, who himself has produced an excellent scholarly edition of disputes from Sinai.26 But one has to say as a Yemenist that disappointingly little is said of Yemen, where related forms occur from Ḥaḍramawt to the far north-west and the reports that we have suggest something very like a corpus. Admittedly, we are not making rapid progress unravelling its connections. Indeed, the history of studies of tribal law in Yemen is one of frustration and delay. The late Professor R.B. Serjeant, for instance, who published several pieces on tribal law and often mentioned a collection of documentary material,27 gave me many years ago a photocopy of what seems to be part of Kitāb al-tabyīn fī l-manʿ, which he had written out for his own use but not found time to edit fully or to publish; and such texts had been lent him or given him, many years before, by Ettore Rossi, who cites the work in his classic article on diritto consuetudinario but never found time to publish the text either.28 What has become of other texts Rossi cites I do not know.

  • 29 Ḥabshūsh Ru’yah, Yémen; Glaser Ma’rib, ‘My journey through Arḥab and Ḥāshid’ (1884, 1993). Carlo La (...)

13Rossiʼs achievements were remarkable, not least that he acquired manuscript copies of important works on tribal law at a time, in the late 1930s, when possession of such works was supposed in Ṣanʿāʾ to be punishable by death. Perhaps we shall never be sure where for instance Kitāb al-tabyīn fī l-manʿ came from. As important as the demonstration such texts really did exist, however, was Rossiʼs summary of the principles apparent in tribal custom. It is these which pull together in his article several earlier reports from Ḥabshūsh, Glaser and others, and a good deal of linguistic detail to be found in, for instance, Landberg.29 These principles, some of which had been noted by Glaser, also inform much that has since been written on ʿurf or custom among Yemenʼs tribes. Let us single out three for the present purpose.

  • 30 Ḥabshūsh Ru’yah p. 83. The spelling of Arabic in what follows is that in the original.

14First, offences are recuperated by vengeance or compensation, and the stated scales of recompense imply formal equality among tribesmen. Secondly, asymmetric relations, marked by such terms as jār and mujawwir (the protégé and one protecting), are reversible: the protector on one occasion can, on another, be the one protected. Thirdly, this reversibility applies only among tribesmen. It does not apply to “weak” people, to slaves or servitors, to women or minors, who are always dependants and protégés. These people, said Ḥabshūsh in the nineteenth century, are not subject to retaliation because “they have no price”.30

سبب أن شرع قبايل ذو حسين وهمدان الجوف واشراف الجوف عادتهم اذا قتل العبد او المرة او الودل او الفراري او اليهودي فلا قصاص علية الا من اهله او من اصحابه سبب ان ما له ثمن واذا قتل القبيلي من هو لا فحجّته كبيرة ولذلك في محاربهم اذا غلب جهد واحد طرف في الحرب قدموا نساهم واولادهم للحرب فلا يمكن يجاهدوهم ابدا فيختسروا وعيب كبير عندهم قتل هو لا سوا حق او باطل.

  • 31 Material damages (for instance blood-money) are nominally the same for all males. The multiple here (...)
  • 32 Serjeant ‘Sunnah jāmiʿah’ p. 14. The order of terms has been reversed for euphony. The term manīʿ d (...)

15The tribesman does have a “price”, by contrast. Indeed, apart from wounds to himself, multiple amends are due him for violating his protection whether the person he protects is somehow of higher than tribal status or of lower –a “judge”, for instance, or a lowly servitor may equally require amends four-fold.31 Serjeant, following Rossi and on the basis of his own wide experience in Ḥaḍramawt and the Aden hinterland, as well as in northern Yemen, summarized the tribesman’s status as sharīf, qawī, manīʿ, that is, honourable, strong, and able to protect others.32 In the summaries below (part three, sections a and c) I set out the implications and try to explain how distinctive a moral world may be at issue.

  • 33 Overlaps are found even within Rossiʼs material. If Rossiʼs (untitled) second manuscript is indeed (...)

16Obviously, in doing this one is scratching the surface. Whatever else may be found, the Rossi and Serjeant material, which seems to span a half millenium, needs collecting and editing, and what connections there may be among even these few sources remains unclear. As Rossi suggests, the Kitāb al-sinnah wa-l-ṣā’ibah, a manuscript dating to 1059 hijrī (AD 1649) which deals with Ḥāshid and Bakīl, and the (undated) Kitāb al-tabyīn attributed to an author from al-Bayāḍ (near Najrān) seem part of overlapping traditions; more surprisingly, so does Kitāb al-ādāb wa-l-lawāzim (perhaps early fourteenth century), collected by Serjeant in Ḥaḍramawt."33 In places one finds common terms and, less expectedly, closely parallel formulations.

17Where Kitāb al-tabyīn fī l-manʿ is a treatise, the qawāʿid examined here are practical documents which not only specify certain points of law but also express a political formation among the signatories. It is partly for this reason that they appear so much like a “code”. Although few general principles are set out, and no claim is made to be exhaustive, the points agreed are each applicable to everyone bound by the pact. As tribesmen at Baraṭ say, the qawāʿid are like a constitution (dustūr). One of documents’ essential features, indeed, is their long lists of guarantors (ḍumanā’) who commit themselves to uphold the agreement on their own behalf and that of their men, all acting together to maintain a set of rules, such that lawyers might be tempted to use the phrase legal sovereignty. But the pact at Baraṭ was not a state in the making. It was a political association of a specific kind, which in one degree or another must have recurred endlessly in Yemen and which the histories of states or rulers and of formal learning tend strongly to obscure.

History of the project

  • 34 P. Dresch, The Northern Tribes of Yemen: their nature and their place in the present republic (D.Ph (...)

18I was first shown a version of the main text (text B, below) near Sūq al-ʿInān, Jabal Baraṭ, in 1979 and was allowed both to photograph this and to copy it by hand. By late 1980 I felt I had a sense of the document overall, but there remained several passages I did not feel confident of paraphrasing or translating. A typed version was provided as an appendix to my doctoral thesis in 1982, which was deposited at the Yemeni Centre for Research and Studies. A few passages were translated in the thesis for illustration, and some were later used in articles and in a monograph on the northern tribes.34 Others copies came to light through the years, but only recently have I sought them out and tried to reconcile them.

  • 35 Abū Ghānim Bunyah qabalīyah pp. 385 ff. Neither he nor I took proper note of Ḥamūd al-ʿŪdī’s earlie (...)
  • 36 al-ʿAlīmī Qaḍā’qabalī pp. 118-40. Neither the preamble to the text nor the list at the end of guara (...)
  • 37 Rashād al-ʿAlīmī is at present Minister of the Interior and has more pressing concerns than digging (...)
  • 38 al-ʿAlīmī Qaḍā’qabalī p. 117, cf. al-ʿŪdī Turāth shaʿbī p. 219. What is not clear is whether the ch (...)

19The text itself appeared also in Faḍl Abū Ghānim's work on the tribal system in 1985, this time, very helpfully, as a photocopy of a handwritten version but without any commentary or explanation.35 A printed version in Arabic then appeared in 1986 as part of Rashād al-ʿAlīmī useful study of tribal justice.36 This is plainly the same text (text B in part six, below) of which I copied a version at Baraṭ and Abū Ghānim presents a closely related copy. Unfortunately the author gives no indication of where his own version came from37, and at several points either he or the person who wrote for him a transcript has distorted the text in directions standard Arabic grammar suggests might clarify the sense or allow one to parse the sentences. As al-ʿAlīmī himself says, the language of the document presents difficulties.38 Not all his readings and interpretations are sound, I think, but this was the first real attempt, so far as I know, to provide an explanation of all the documentʼs substantive passages, and the clarity of al-ʿAlīmī’s effort is admirable.

  • 39 Cf. an editor’s comments on Anglo-Saxon laws: “... it is not in all cases possible to expand their (...)

20Text A, a shorter but related set of qawāʿid al-malāzim, was first shown me at Baraṭ in 1983, and I copied this by hand in less than ideal circumstances. As far as I know, it has not been published. Two clean copies were found me in 2003, and a third at the turn of 2004; a fourth soon followed in March that year. At first, in 1983, it seemed this shorter text might simply be a reduced version of text B, but as copies of B accumulated it became more likely that A and B were related yet distinct documents, and further copies of text A have strengthened that supposition. The history of the copies will be summarized below. Even with the clearest, however, one has trouble on occasion with the script and often with divining those ideas which the original scribes took for granted as common sense.39 The documents are written almost in a kind of short-hand, particularly the market-guaranty procured for me in 2003.

  • 40 Vocabulary differs from place to place, and so does use of the same vocabulary (the way rifles as p (...)
  • 41 One does not wish to suggest that nothing at all changed until recently and that Baraṭ lived in a t (...)

21Finding well-informed help with interpreting the texts among tribesmen elsewhere than Baraṭ, let alone finding help in Ṣanʿā’, is difficult. Time and again one finds oneself in need of local knowledge from people who live where a document was written,40 and even they, we should always remember, are separated from the texts at hand by more than two centuries. The problem is intensified by the way tribal judges or arbitrators use contemporary documents often as aides-mémoire. The thrust of intention (maqṣūd) is what matters, not the detail of grammatical meaning (maʿnā lughawī); and most people steeped in “custom” are not used to parsing sentences. The ideal situation is thus to have a young person present who probably has no clue about custom but knows roughly how grammar works and can suggest, with due respect to his uncle or grandfather, that a word or particle is missing, and grandfather who has little idea of formal grammar but once lived in the kind of world to which the document refers.41 This is one of those subjects that, unless handled carefully, leaves everyone exhausted with the other’s dim-wittedness. The problem is compounded by my own poor knowledge of classical Arabic, in which many supposedly “dialect” words might be found if one only had the breadth of learning.

  • 42 ʿAbdullāh Nājī al-Ḍuwaynī of al-Maʿāṭirah was also enormously helpful, and so were friends such as (...)

22In 1983 I had the chance to work on text B in Yemen with Gerd-Rüdiger Puin of the University at Saarbrücken, who was then engaged on a manuscript-restoration project in Ṣanʿā’. Singly and together we visited Baraṭ several times. Dr. Puin transcribed a further copy of text B, which he pursued with his own tribal contacts. I received particular help at the time from Muḥammad Ḥasan Dāris, Ḥamūd Muḥammad Abū Ra’s, ʿAbdullāh ʿĀ’id Yaʿqūb of al-Maʿāṭirah, and Ṣāliḥ Muqbil Juzaylān, the latter two figures being widely-known experts in ʿurf.42 Dr. Puin and I had intended to pursue this effort to a published article. In 1985 I sent him draft translations and what notes I had on tribal law, and he sent to me a hand-written version of text B which represented the work he had done resolving doubtful readings. Unfortunately, pressures of teaching and of family meant he could not pursue this further and the effort lapsed. From time to time others expressed an interest. The main result was to remind one how little sense the text made without the principles informing it. In 1999 a serious interest was expressed by Laila al-Zwaini, then a doctoral student at Leiden, who typed up an Arabic draft of text B in 2000, but her own work had of course to come first and again the chance was lost of a joint project.

  • 43 The term naqīb is used as a title by certain shaykhly families, mostly from Bakīl tribes (Dresch Tr (...)

23Discussions with Aḥmad ʿAbduh Thawābah and al-Naqīb43 ʿAbdullāh Muḥsin Thawābah of Dhū Muḥammad in late 2000 revived my personal interest, and they kindly found me a further copy of the main text (text B, below). François Burgat of the Centre Français in Ṣanʿā’ (CFEY, now CEFAS) expressed enthusiasm the following year for publishing text B. His organizational and financial help in 2002–3 were invaluable and without the technical assistance of the Centre staff the project would not have been completed. It is a pleasure to be able to thank François at last for his encouragement: Jean Lambert, his successor, extended the same hospitality in 2003–4 although the project had grown more complex. I should also like to thank Dhikrā Muṭahhar and Amīrah ʿĀ’id of CEFAS for their secretarial help, and on the financial front, St Johnʼs College, Oxford, generously paid travel-expenses for several short trips to Yemen.

  • 44 These thanks are much more than formal. Although, for instance, I have known Ḥamūd Muḥammad Abū Ra’ (...)

24The greatest possible thanks are due to ʿAbdullāh Muḥsin Thawābah (himself a justly noted arbitrator and expert in ʿurf for the patience he showed in taking me through text B again and attempting to riddle out the details. Shaykh ʿAbdullāh b. Nājī Dāris then kindly gave me the run of his dīwān, where tribesmen from Dhū Muḥammad congregate at al-Rawḍah, out by Ṣanʿā’ airport. Naṣr b. ʿAbdullāh Dāris and Ṣāliḥ Muḥammad al-Farajī provided enormous help in the summer of 2003; Nājī ʿAbdullāh and Nūr al-Dīn ʿAbdullāh did so later, and so too did Ḥamīd Muḥammad Jarād.44 ʿAbd al-Karīm ʿAbdullāh al-ʿAnsī, a very old friend, mobilized endless friends and relations from Baraṭ and the Jawf. Special mention should be made of Muḥsin Muḥammad al-Ghuraybī of Wādī Silbah, who not only knows tribal custom well (his father was a respected arbitrator) but showed an admirable knack for comparing texts and spotting clues in the script and grammar. As with most foreign work on Yemen, the bulk of useful ideas has come from Yemenis.

  • 45 E.g. CEFAS Le Règlement des conflits tribaux au Yémen (2003), B. Dupret, Systèmes coutumiers, centr (...)

25I can only apologize for the long delay in publishing the main texts and commentary, and apologize for the obvious shortcomings. Even certain key-terms, explained as best I can below, resist translation. Certain passages remain obscure. One hopes that historians, anthropologists, linguists, and others concerned with Yemen will be encouraged to publish what they have on customary law and thus allow a fuller reading than I can manage here. There are signs, at least, of a fresh interest in custom and arbitration if not yet in the detail of texts.45 It seems important to precipitate criticism and some solid work.

Document 1: Dhū Muḥammad and Dhū Ḥusayn (1251/1836)

Document 1: Dhū Muḥammad and Dhū Ḥusayn (1251/1836)

The different versions of the texts

  • 46 I do not have a linguistʼs ear, and asking people to write vowels, where they know how to do this, (...)

26All the texts available are recent copies, and we do not have means to verify dates. Nonetheless the postscripts give dates on which earlier copies were made, and provisionally we seem to be dealing with successive versions of the Code. These in turn rest on earlier agreements. Both texts A and B make mention of existing market­-rules (qawāʿid al-sūq) or qāʿidat al-ḥaram, a document that defines the protected space of the market and what should be done to uphold the market-peace at Sūq al‑ʿInān, Dhū Muḥammadʼs main centre at Jabal Baraṭ. Text A begins, “The first thing [concerns] the market”; text A section 21 says that later agreements do not erase (lā timḥaq seems to be the pronunciation)46 the existing market-pact. People in the area insist that the original market-guaranty must still exist, and the status of the market is referred to with great confidence, but despite extensive efforts by friends from Dhū Muḥammad we never found quite the document referred to.

  • 47 See H. Steffen et al. Final Report of the Airphoto Interpretation Project (1978) p. II/183-4. The a (...)

27When a market was first established at al-ʿInān is unclear, and to pin down the date might in any case not help since markets often lapse for a time and are then renewed. Nor does geographical logic explain the distribution of markets. The other main markets in the area, apart from Ḥarf Sufyān (a long dayʼs walk and more to the south-west), are Rajūzah, just a few kilometres away from al-ʿInān in Dhū Ḥusayn, and al‑Maḥjal, 60 km. or so further east in Dhū Ḥusayn’s territory at Wādī Khabb.47 Āl Sālim and al-ʿAmālisah, tribes to the north and north-west of Baraṭ, seem to have no weekly markets of their own. The late Turkī Khursān of Dhū Ḥusayn was said to have a qāʿidah for Rajūzah, dated 1206 hijrī (AD 1791–2), which placed it somehow under the joint guaranty of Dhū Ḥusayn and Dhū Muḥammad; Bayt al-Baḥr of Dhū Zayd in Dhū Muḥammad are said to hold further documents. Whether a reciprocal agreement was ever made that involved Dhū Ḥusayn in al-ʿInān’s protection is unclear, but it deserves remembering that tribal agreements often overlap and may supplement or contradict each other.

28From the texts of qawāʿid al-malāzim (texts A and B) one can take a fair guess at what must have been in documents concerning al-ʿInān pre-1763, that is, before the date claimed for first drawing up text A. They must at least have specified a ḥaram, or protected area, and stated who was responsible for upholding the market-peace; they would also have specified amends for violating that peace. Unfortunately the only text of the guaranty we have, which begins with the claim that Dhū Muḥammad have decided to establish the market, dates its own first copying to 1185 hijrī (AD 1771–2), a decade later than would be ideal. The explicit additions dated 1225 and 1275 hijrī (AD 1811, 1859) are interesting in themselves. But an obvious contradiction arises in that text 1, the market-guaranty, continues to specify amends as four-fold, while text B section 16 (1211 hijrī, AD 1796) says eleven-fold when the dates of renewal claim that text 1 was still in force. To get much further with understanding the original system is probably not possible for the moment, not least because in recent years, as the centre of events has shifted towards Ṣanʿā’, many people have declared themselves quit of responsibility for upholding the Baraṭ market-peace (ibtaraw min al-ḍamānah).

  • 48 The version given here in photocopy was kindly found me by Ṣāliḥ Muḥammad al-Farajī. Nūr al‑Dīn ʿAb (...)
  • 49 Leaving off the copyist’s name means also leaving off an official stamp from the ibtidā’ī law‑court (...)

29The version of the guaranty we have to work with (text 1, below) is a curious document in that various attempted classicisms occur of a kind not usually seen in scripts like this (e.g. nāqūn for those seeking vengeance or compensation) yet simple things like the names of Islamic months are mis-spelled. Although the scribe who made the final copy in 1998 notes scrupulously where something is missing (hunā kalimāt maṭmūsah) and says at the end that the crossings-out are internal to the text (al-shaṭb alladhī fī bāṭin), I am all too familiar with his work from other texts, not least an illegible copy of text A. I provide (in part four, below) a photocopy of the guaranty, which gives an idea of what one is sometimes up against.48 But the standard of hand-writing is so poor in places as to be embarrassing, and I have therefore left off the name of the final copyist.49

  • 50 Perhaps I should say lent rather than given. Even where people have offered hand-written versions i (...)
  • 51 The document begins with the common formulae of “renewing” or “clarifying” existing custom. What th (...)

30We then have two versions of qawāʿid al-malāzim. First there is a short version (text A), which I transcribed by hand in 1983 from a copy in the possession of Ḥamūd Muḥammad Abū Ra’s. This corresponds quite closely to hand-written notes shown me by ʿAbdullāh Muḥammad al-Rudaymī and referred to him simply as “the custom of the Jawf” (the provenance was unclear and the guarantors not listed). A clean copy was later given me by ʿAbdullāh Nājī Dāris in 2003 (reproduced here, as part five, in photocopy), and another by Ṣāliḥ Muḥammad al-Farajī; a further, rather splendid, copy was given me at the turn of 2004 at al-Qāʿidah near Taʿizz by Yayqar Muḥammad Nājī of Bayt Abū Ra’s, and another in March that year by Nājī ʿAbdullāh Dāris.50 The manuscript copies of text A all carry the date Muḥarram 1177 hijrī (AD 1763) and comprise an agreement solely among the sections (the five akhmās) of Dhū Muḥammad.51 Al-Rudaymī’s version differs in certain details from the others, and these differ sometimes among themselves, though mainly in simple omissions and the ordering of adjacent names or phrases. Where the differences seem important I note them in parentheses under text A. Apart from this, and discussing the few passages that differ markedly from text B, I present text A simply in Arabic and in summary translation, working out from the Abū Ra's text I was shown first and keeping notes to a minimum.

  • 52 A copy from Muḥammad al-Sudumī, made for the Ministry of Justice in the early 1970s, was passed on (...)

31Second is a much longer and fuller document (text B: part six, below) which corresponds to the text, first shown me by Ḥamūd b. Nājī Dāris, that I gave in my doctoral dissertation in 1982. This comprises an agreement among Dhū Muḥammad, Āl Muḥammad (al-Maʿāṭirah), and Dhū ʿAmr. Abū Ghānim reproduces a closely related copy of the same text as I photographed; another (copied for al-Maʿālim of Āl Ṣalāḥ at al‑Marāshī in the early 1980s) was provided by friends from Dhū Ḥusayn, and Nājī ʿAbdullāh Dāris found me a further copy from one held by al-Maʿāṭirah. With a version given me by Aḥmad ʿAbduh Thawābah in 2000 this makes five manuscript copies.52 The photocopy is of the Thawābah text, the one with the clearest script, written out (like text A) by perhaps the best copyist at Baraṭ, Yaḥyā Aḥmad ʿAtīq. The typed version represents an edition compiled from the different copies and from discussions with experts on ʿurf at Baraṭ. This is followed by an English translation and explanatory notes. Most of the explanatory material relevant to text A as well will be found here or in the introduction (part three). The first date mentioned on all the copies we have of text B is Muḥarram 1211 hijrī (AD  1796). Despite the obvious resemblance to text A, none mentions the date 1177 (AD 1763) or a previous agreement.

  • 53 This is taken from text B. For text A see the photocopy in part five, below. Some explain the repet (...)

32Odder still, the list of guarantors is identical in texts A and B, except that signatories for al-Maʿāṭirah and Dhū ʿAmr are added on to those for Dhū Muḥammad’s “fifths”. The original witnesses (shuhūd) are the same too, though in one document they are dated an implausible thirty-three years later than in the other:53

هذا الذي تواجبوا عليه بحضر الشهود النقيب منصور بن مهدىوالنقيب حسن بن ناصر جزيلان والنقيب مرشد بن حسين طشان والشيخ حسين بن قاسم دمينة والشيخ علي بن قاسم دمينة والشيخ أحمد بن صالح أبو رأس والشيخ صالح البحر وهادي بن علي فرج وعلي بن راشد بن عاطف والشهود من المعاطرة ناصر بن حسن الضويني وناصر بن جابر وحسن القاضى علي بن محمد العنسي وعدة من الناس (sic) بن أحمد بن محمد والولد

Fig. 1: Contents of Qawāʿid al-Malāzim

Fig. 1: Contents of Qawāʿid al-Malāzim

33The obvious conclusion, particularly since both lists of witnesses are dated to the same month, Muḥarram, would be that at some stage an error was made in copying the dates (they are only in numerals, not written out) and we have, so to speak, a corrupt text. However, the first provision of text A, on the market, is omitted from all the versions of B, as too are the details of rights that derive from sharing food or drink (text A section 17) and provisions at the end on a market-tax (text A sections 21, 22); on the other hand, text B contains several passages that are not in A. Readers must judge the relation for themselves, but the copies of A are consistent and the copies of B are consistent (lists of contents are provided as Figure 1). Whatever part mistakes in transcription may or may not have played, it seems to me that one cannot reduce the relation between A and B to scribal error. People kept, reproduced, and perhaps referred to these texts in parallel.

  • 54 The earlier date seems marginally the more likely, appearing clearly in the better-written versions (...)
  • 55 Bayt Ṭuhayf were part of the qādī family, Bayt al-ʿAnsī (of whom more in parts two and three). The (...)

34In three of our copies of text A the lists of transcriptions run strictly in parallel until 1300 or 1311 hijrī (although there is no solid means to choose, this means the 1880s or early 1890s).54 All end the text, after listing the guarantors, with the witnesses and the name of the original scribe, Muḥammad Aḥmad ʿAlī, and mention copies being made in Rabīʿ Awwal 1214 (AD 1799) from a version in Dhū Mūsā’s possesion, the scribe being ʿAlī Ḥasan Ṭuhayf.55 A copy was then made by Muḥammad Yaḥyā ʿAbd al‑Raḥmān in 1218 (AD 1803), at the request of al-Naqīb Mufliḥ b. Ṣāliḥ Abū Ra’s, from a version of this owned by Dhū Zayd. The Abū Ra’s family of shaykhs are not from either Dhū Mūsā or Dhū Zayd but from Āl Aḥmad b. Kawl (these, to anticipate, are all separate “fifths” or akhmās of Dhū Muḥammad). Ḥusayn b. Mufliḥ Abū Ra’s apparently had a copy made in 1232 (AD 1817); from this a further copy was made for Muḥammad ʿAbdullāh al-Baḥr of Dhū Zayd in 1883 or 1894 (Shawwāl 1300 or 1311), and this in turn comes down to us via recent copies for Bayt Juzaylān (Dhū Mūsā), Bayt al-Baḥr (Dhū Zayd), and Bayt Dāris (Āl Dumaynah). Plainly these documents are not simply passed on in families or even in tribal sections such as Dhū Muḥammad’s fifths.

  • 56 Resolving this would need tracking down the names in Lower Yemen, which I have not had a chance to (...)

35Two other copies of text A, shown me by different branches of Bayt Abū Ra’s, claim a different transmission. Muḥammad Aḥmad ʿAlī’s version was again copied by Muḥammad Yaḥyā ʿAbd al-Raḥmān, though the only date given is Rabīʿ Awwal 1214 (1218 is not mentioned; nor does ʿAlī Ṭuhayf appear). The next date is unclear, al‑Qaʿdah 1249 (AD 1834, which seems more likely) or 1349 (AD 1931), when apparently a fresh copy was made by ʿAbd al-Khayr al-Maḥaysī (?) from one in the possession of Aḥmad ʿĪḍah Abū ʿUrūq of Āl Aḥmad b. Kawl, who was living at Jabal Jaḥḥāf in Lower Yemen. At least one was copied from a certain Aḥmad Ḥusayn ʿAlī Abū ʿUrūq, but the links between these two copies and the others of text A are not clear.56 In one of these two and in the Dāris copy a reflux action of text B on A seems evident, where the phrase ahl al-malāzim appears though it properly belongs in B.

  • 57 For a genealogy of Bayt Abū Ra’s, including Ḥusayn b. Mufliḥ, see Muḥammad al-Yāzilī, Min al‑thawra (...)

36Text B lists the same witnesses (although the date is 1211 hijrī, not 1177) and also mentions Muḥammad Yaḥyā copying from Muḥammad Aḥmad ʿAlī, who himself apparently wrote multiple copies (qawāʿid marqūmah mutanāẓirah) at the start. Again copies are made in 1214 hijrī, but in Jamād Awwal (Jumādā Ūlā), two months later than for text A. Ḥusayn b. Mufliḥ Abū Ra’s, presumably the man mentioned in the later copying of text A,57 seems to have made a copy of text B, as does Qadi Muḥammad ʿAlī Muḥammad. The text is unclear on the details, but it appears that a comparison was made among different copies in writing the full version, which in all the recensions I have seen comes down to us through a copy made in 1238 hijrī (AD 1822) for Āl al-Ḍuwaynī, who are shaykhs of al-Maʿāṭirah at Baraṭ, One copy, belonging to Āl Yaʿqūb of al‑Maʿāṭirah, looks to have been taken from this about fifty years ago but is not dated. The others, made for various shaykhs at Baraṭ, are more recent. Unlike the case with text A, unfortunately, we do not have clearly separate routes of transmission.

  • 58 Unfortunately I never managed to get the story from Muḥsin Ismāʿīl of how the copy was made and how (...)

37The version of text B that I transcribed in 1979 and the version Abū Ghānim published in 1985, though the final stage of transmission differs, both carry the same first paragraph at the top of the scroll, explaining that a copy was made by Muḥsin Ismāʿīl Aḥmar al-Shaʿr of Āl Ṣalāḥ in 1387 hijrī (AD 1967) because the original was in poor condition.58

هذا نقل قاعدة الملازم كان من المحاولة لنقل صورتها سبب تهدمة الاصل واعاننا الله على ذلك والمُملي لي هو الحاج حسين بن محمد دومان اللفظة باللفظة لا زيادة ولا نقصان وكتبت نقلها بخطى وانا الحقير الراجي عفو ربه محسن بن اسماعيل احمر الشعر وفقه الله ونقله بتاريخ جماد أول سنة 1387

38Yet the list of transcriptions at the end of the document in others I have seen with the same small distinctive features jumps from 1238 (AD 1822), when a fairly important note was made about the status of witnesses and the payment of ʿayb (amends for shame or disgraceful acts), to the latest transcription without mention of Muḥsin Ismāʿīl. It is not impossible, although I think it unlikely, that all but one of the extant copies of text B in fact come from the 1967 version.

  • 59 Credit for this explanation, which I think very likely to be right, goes to Muḥsin al-Ghuraybī. I h (...)
  • 60 The postscript to the Thawābah copy is given in full with the list of guarantors in part six, below (...)

39Plainly there is much more one wants to know. Text A describes itself as “five equal codes” (khamsah qawāʿid mutanāẓirah) and text B as three such, but neither text divides at all obviously into anything of the kind. The repeated phrase qawāʿid mutanāẓirah can hardly mean a division of the text into chapters or its derivation from existing documents. Probably what is meant is that five master-copies were made of text A (one for each of the fifths of Dhū Muḥammad, the main parties to the first agreement) and three such copies of text B (one for each of Dhū Muḥammad, al‑Maʿāṭirah and Dhū ʿAmr, the parties to the later pact).59 Support for this is found in the postscript to text B which mentions al-Maʿāṭirah and Dhū ʿAmr asking for their own copies.60 But how, and quite when, the two texts diverged remains unclear in detail. Plainly copies of each were made and were passed back and forth regardless of the tribes’ divisions.

The market and the tribes involved

40In the documents, as at present, the fifths or akhmās (sing. khamīs) of Dhū Muḥammad are Āl Dumaynah, Āl Aḥmad bin Kawl, Āl Ṣalāḥ, Dhū Zayd, and Dhū Mūsā, Guarantors for each are listed in order. But the signatories for Dhū Faraj (these days called Āl Faraj) and Dhū ʿĀṭif, both of which now fall within Āl Dumaynah, are listed separately. In text A section 20 we find a distinction between al-Maḥlaf (meaning Dhū Zayd and Dhū Mūsā, as is still the case) and a compound identity referred to as Āl Aḥmad bin Swaydān and Ibn ʿĀṭif. Nowadays Āl Aḥmad b. Swaydān comprises Āl Dumaynah, Āl Aḥmad b. Kawl, and Āl Ṣalāḥ (see Figure 2).

  • 61 Cf. Muḥammad al-Ḥajrī Buldān al-yaman wa-qabāʿil-hā (1984) pp. 109-11, which agrees with what membe (...)

41Dhū Mūsā and Dhū Zayd (al-Maḥlaf) are supposed to descend from Swaydān b. Muḥammad (b. ʿAmr) b. Ghaylān. People agree that Ṣalāḥ and Dumaynah, as well as Aḥmad, were all sons of Kawl b. Aḥmad b. Swaydān b. Muḥammad, offset from al‑Maḥlaf by two generations; but Faraj, the eponym of Dhū Faraj, now in Āl Dumaynah, is said to have been son of Aḥmad Swaydān. ʿĀṭif, on the other hand, is said to have been son of Muḥammad Ghaylān, three generations back, and thus “half the tribe”.61 But were there originally only five “fifths”? Or did Dhū Faraj and Dhū ʿĀṭif have some other status? The answer –or part of the answer– appears in the market guaranty (text 1 section 1), where responsibility for the market-peace circulates month by month among the fifths and we find Dhū Faraj and Dhū ʿĀṭif though named explicitly, listed with Āl Dumaynah as they would be now. In text A section 8, purportedly from AD 1763, however, where the obligations of “blood and flesh” are spelled out (primarily these are obligations that derive from marriage), Dhū Faraj are said to be separately answerable for Dhū Faraj and Dhū ʿĀṭif for Dhū ʿĀṭif.

Fig. 2: The fifths of Dhū Muḥammad

Fig. 2: The fifths of Dhū Muḥammad
  • 62 Swaydān is a rather mythic name, and where Bāb Swaydān may have been is a mystery. Many at Baraṭ sa (...)

42Each fifth –or in one case, perhaps, a fifth and two associated units– was responsible in turn for the ḥaram of the market, serving in rotation as ahl al-shahr, “the people of the month”. The guarantors of each unit within the other fifths were responsible to ahl al-shahr for their fellow tribesmen paying what amends they themselves incurred (text 1 sections 2, 3, 4) and for supporting the “people of the month” in the event of claims against outsiders (text 1 sections 8, 16). The market also figures as one of those issues where intestine disputes among the signatories are suspended and responsibility is definitively shared (text A sections 11, 15, 16; text B sections 11, 25). How far geographically such responsibility extended is unclear. But a recurrent theme in discussion is that Dhū Muḥammadʼs writ once ran on all the market-roads, beyond their own territory, even “from Bāb Swaydān”.62 Acceptance or rejection of outsiders at the market devolves on the parties to the pact as a single unit (text 1 section 11, text A section 18, text B sections 23, 25).

  • 63 M. Kister, The market of the Prophet, Journal of Economic and Social History of the Orient vol. 8 ( (...)
  • 64 Ḥabshūsh Ru’yah pp. 107-8. For a discussion of the Ṣuraymī document referred to below (Document 2) (...)
  • 65 P. Dresch, A pact of brotherhood from Sufyān (North Yemen), Chroniques Yéménites vol. 10 (2002). Fo (...)
  • 66 M. Mauss, Essai sur le don: forme et raison de lʼéchange dans les sociétés archaïques, Année Sociol (...)

43That protection of the market might provide the nucleus around which grows something larger comes as no surprise. The beginnings of Islam itself, turning as they did on control of trading towns, seem to have involved a protection pact of a kind familiar later in Yemen’s history.63 References to Yemeni markets are numerous, and in Upper Yemen, as Ḥabshūsh says, markets are hijrah or muhajjarah, set aside and protected.64 A document from Banī Ṣuraym (reproduced here as Document 2) shows a typical hijrah system of fines and guaranty agreed in AD 1831. Near our own time, in the 1920s, we can point to how appropriating tribal market-guaranty was part of how Imam Yaḥyā established control over Upper Yemen, and how towards the end of a civil war in the 1960s –republican against royalist, but also a proxy-war between Egypt and Saudi Arabia for control of the Arab world– tribes aiming to preserve their formal cohesion did so by protecting a common market.65 To reduce this to “economic” considerations would miss the point. (The nearest market geographically may not always be protected by one’s own tribe). As anthropologists would expect from the works of Mauss or Hocart, the market has a moral value determined by its contrast with the dispersed forms of everyday life.66 Briefly put, the market on market‑day is the one shared space in the moral system, and the one place recognized by all where violence of all kinds is forbidden. It is also where agreements are formed on collective matters and where announcements must be made of public claims.

  • 67 Things may have been different elsewhere in Yemen, and the rule may not have held throughout the no (...)
  • 68 The same system of monthly responsibility applied to protection of a learned family at Baraṭ that w (...)

44In Upper Yemen, on the northern plateau, markets seem not to have been a direct source of income for local chiefs as often they were in Europe.67 If a tax was imposed it usually went to non-tribal learned persons. The system at al-ʿInān is somewhat out of the ordinary in that a tax (majbā, jibāyah) was levied on for instance livestock-traders (text A sections 21, 22) and this went to maintaining a guard at the market-place throughout the week drawn from the protecting tribe. More than this, responsibility for maintaining the market-peace on market-day rotated, as it still did until recently, among the fifths, who each provided the standing guard for a month in turn.68 Outside the ḥaram, or protected space, the “people of the month” were not responsible for disputes among their fellow tribesmen (text 1 section 16); if blood were spilled inside the ḥaram on market-day, however, they were to take vengeance or extract compensation four-fold (text 1 section 10). The other fifths were obliged to support them if required, and were responsible for the wider space of tribal territory and the roads to market. This broader responsibility was shared by the market-guarantors together and hence by all the tribesmen for whom they signed.

Document 2: Market guaranty from Banī Ṣuraym (1246/1831)

Document 2: Market guaranty from Banī Ṣuraym (1246/1831)

Document 2: Market guaranty from Banī Ṣuraym (1246/1831) verso

Document 2: Market guaranty from Banī Ṣuraym (1246/1831) verso
  • 69 In other words, they are not expected to fight. While Dhū Muḥammad speak of them often as a “sixth (...)

45Whatever the history of the texts, we seem provisionally to have a political history of the region. Supporting evidence is thin to non-existent. Internal evidence, however, suggests strongly a sequence from an agreement among the “fifths” (akhmās) of Dhū Muḥammad to uphold their market-peace to a wider agreement embracing their immediate neighbours. Text A (AD 1763), as we saw, like the market-guaranty binds only the fifths of Dhū Muḥammad; text B (AD 1796) comprises an agreement among Dhū Muḥammad, Āl Muḥammad (al-Maʿāṭirah), and Dhū ʿAmr. In certain copies of text B the conjunction “and” appears between Āl Muḥammad and al-Maʿāṭirah, which is simply wrong: Āl Muḥammad (as opposed to Dhū Muḥammad) and al-Maʿāṭirah are the same. They are not one of Dhū Muḥammad's fifths but live interspersed among them. Although members of Dhū Muḥammad sometimes claim al-Maʿāṭirah have no ḥudūd, no distinct “borders” or “territory” of their own, they in fact have distinct locations and rights of territory at Wādī Tamr, in the upper reaches of Wādī Silbah, at the border area of al-Rahā’iṭ, and at three locations on Baraṭ itself –­at Wādī Khanaq (around Bayt al-Qayl), Ḥajjān, and Wādī al-Balasah. It is said that al-Maʿāṭirah do not fight alongside Dhū Muḥammad against the neighbouring tribe of Dhū Ḥusayn.69

  • 70 Āl Ḥuṭbān, who are not at all numerous, seem to live mainly to the east of Baraṭ around Wādī Khabb, (...)

46In Dhū Ḥusayn one finds a similarly anomalous section, al-Ḥuṭbān, who form part of Āl Ḥusayn70 but apparently are not part of Dhū Ḥusayn’s “eighths”. Al-Ḥuṭbān do fight alongside the eighths of Dhū Ḥusayn against Dhū Muḥammad, or in theory should do. It is said that long ago they held much of what is now Dhū Ḥusayn's territory and part was held by Banī Nawf, a tribe now associated with the Upper Jawf. Banī Nawf themselves have supposedly lost territory to Hamdān and Dhū Ḥusayn; but their name remains in Wādī Banī Nawf at Baraṭ, south-east of Rajūzah. Āl Sulaymān, whose centre is now al-Haḍabāt and part of whose border with Dhū Muḥammad lies at Wādī al-ʿArīshah just east of Raḥūb, are supposed once to have held Jabal Āl Ṣalāḥ, now squarely within Dhū Muḥammadʼs borders. The scattered distribution of al‑Maʿāṭirah’s territory at and around Baraṭ suggests perhaps they too once held more of the region than they now do. Mere geography or supposed genealogy, however, do not determine politics and text B section 36 precisely in assimilating al‑Maʿāṭirah to an alliance with Dhū Muḥammad severs certain forms of joint responsibility between them. Where al‑Maʿāṭirah (Āl Muḥammad) may have stood at which date with regard to Dhū Muḥammad remains obscure. In two documents from Qadi Ismāʿīl al-Akwaʿ concerning the protection of Bayt al-ʿAnsī at Baraṭ, between the mid-seventeenth and nineteenth centuries, they appear as an independent unit of the same structural order as Dhū Muḥammadʼs fifths.

  • 71 ʿAmr is nearly always spelled as if it were ʿUmar. Although the classical waw is added in books, an (...)
  • 72 There are very few plainly ʿAmrī names among the signatories to text B. In one of al-Akwaʿʼs docume (...)

47The third major party to the pact in text B are Dhū ʿAmr.71 Local tradition has it that they were once a major tribe of the region but there are now only scattered remnants: their territory was mainly in the low-lying area of al-Marāshī, on the route west and south from Baraṭ to Sufyān and from there to Ṣanʿā’. In the early 1980s they were still to be found in some numbers at the village of Shaqrat bin Ḥaydān, at al-Marāshī, and some thirty men lived at al-Jaljalah in the territory of Dhū Mūsā (Dhū Muḥammad) near al-Kharāb.72 Bin Ḥaydān, a shaykh of Dhū ʿAmr, is said to have dominated long ago the whole region, not least the trade route. A proverbial saying was that kānat arḍ al-marāshī ʿashīrah min dūn shaqrat bin ḥaydān, meaning that if it were not for Bin Ḥaydān, who extorted taxes from the area and from caravans passing through, al‑Marāshī and its surroundings would only have paid the Islamic tithe.

Document 3: Agreement among Dahm (1231/1814)

Document 3: Agreement among Dahm (1231/1814)

Fig. 3: Divisions of Shākir and Dahm

Fig. 3: Divisions of Shākir and Dahm
  • 73 See e.g. text B section 23. One might have expected the pact to spread further and involve Dhū Ḥusa (...)
  • 74 Steffen et al, Final Report p. II/119, have Shaʿbān and ʿAmr transposed, making Āl Sulaymān more di (...)

48ʿAmr is also, it seems, a collective name for Dhū Muḥammad, Dhū Ḥusayn, and al‑ʿAmālisah. The collective name for Dhū Ghaylān (that is, Dhū Muḥammad and Dhū Ḥusayn together), al-ʿAmālisah, and Āl Sālim is perhaps Nasr. Despite this ranking of proximity, we should note, the market guaranty (text 1 section 7) draws no distinction of kind between al-ʿAmālisah and Āl Sālim; more strikingly, none of our documents mentions Dhū Ḥusayn, supposedly Dhū Muḥammadʼs “brother”, except to exclude them.73 Different views of the relations among tribes of the area are common, but Steffen’s team in the 1970s were offered a sketch of the main genealogy, and ʿAbdullāh ʿA’id Yaʿqūb of al‑Maʿāṭirah some few years later gave me a related sketch of local detail (see Figure 3).74 ʿAbdullāh Nājī Dāris has on the wall of one of his reception rooms at al-Rawḍah, just outside Ṣanʿā’, a personal genealogy which gives a similar picture: ʿAbdullāh b. Nājī b. ʿA’id b. ʿAlī b. ʿAlī b. Dāris b. Ṣāliḥ b. Qāsim b. Ṣāliḥ b. Muḥammad b. Dāwud b. Dumaynah b. Kawl b. Aḥmad b. Swaydān b. ʿAmr(w) b. Muḥammad b. Ghaylān b. Muḥammad b. Shaʿbān b. Nasr b. ʿAmr(w) b. Duhmah. The essential point to note among these different versions is the repetition of the names ʿAmr and Muḥammad.

  • 75 Though it is hardly by itself evidence, one finds mention in the early 17th century of someone name (...)
  • 76 According to some versions (e.g. Steffen Final Report p. II/119), Ghaylān includes also ʿIyāl Yazīd (...)
  • 77 Dresch Tribes p. 325. Serjeant’s reference to a mention in al-Rasūlīʼs Ṭurfat al-asḥāb (Serjeant ‘T (...)

49It is widely accepted that al‑Maʿāṭirah descend from the older Muḥammad, as uncle, so to speak, of Muḥammad and Ḥusayn.75 The idea that Dhū Muḥammad and Dhū Ḥusayn both descend from Ghaylān Muḥammad, and that repetition of the name Muḥammad suggests a possible historical priority of one over the other, is current even among shaykhs of Dhū Ḥusayn. For other purposes Dhū Muḥammad and Dhū Ḥusayn are spoken of as brothers, two parts of Dhū Ghaylān.76 The name Ghaylān itself is old, although attached a millennium ago to a location nearer Ṣaʿdah than to Baraṭ; Dahm (or Duhmah) and Shākir, higher up in the genealogy, provide larger, fairly stable conceptual shapes, already associated with the relevant geographical space north of the Jawf as early as the tenth century. The names Dhū Muḥammad and Dhū Ḥusayn, however, seem not to be attested until about AD 1600.77

50ʿAmr, as we have seen, had at least two meanings. In its lesser sense, lower down the supposed genealogy, Dhū ʿAmr was a unit of the order of al‑Maʿāṭirah, Text B of the qawāʿid thus has Dhū Muḥammad, Āl Muḥammad (al‑Maʿāṭirah), and Dhū ʿAmr as three distinct parties; one of al-Akwaʿʼs documents from Baraṭ, dating to the early nineteenth century, binds Dhū ʿAmr, al‑Maʿāṭirah, and Āl Ṣalāḥ (one of Dhū Muḥammadʼs fifths). In its larger sense, by contrast, higher up an imagined genealogy, Dhū ʿAmr would include Dhū Muḥammad and the others. But the texts here are far from clear.

  • 78 al-Akwaʿ Madkhal pp. 27 ff. The al-ʿAnsīs will be discussed in part two and the text itself in part (...)

51Text A binds the fifths of Dhū Muḥammad. Section 13 of that document distinguishes Dhū Muḥammad and Dhū ʿAmr, section 8 mentions both but then lists only Dhū Muḥammadʼs fifths as if the names were equivalent, and section 10 says that anyone committing a wrong at a meeting of Dhū Muḥammad or of its sections can call on no obligation at further meetings anywhere “in the territory of Dhū ʿAmr, east or west” or even at market in al-ʿInān. Text B binds Dhū Muḥammad, al‑Maʿāṭirah, and Dhū ʿAmr in the lesser sense; but someone committing a wrong at a meeting has no obligation to call on at further meetings among “Dhū Muḥammad, east or west...” (section 10), as if Dhū Muḥammad (in text B) and Dhū ʿAmr (in text A) were identical. In one of al-Akwaʿʼs earlier texts, whose first renewal is dated to 1112 hijrī (AD 1700), we find the same implication that the names are synonymous.78 Dhū ʿAmr promise to protect the descendants of ʿAlī Qāsim al-ʿAnsī, but the fifths of Dhū Muḥammad are then mentioned as if they comprise the whole of the people involved, and all the signatories look to be from Dhū Muḥammad. Under each of these tribal names, of course, sections may have views of their own worth and history.

ذو محمد عدتْ يا نماراَ بدتْ
خصها بن حمد بن سويدانِ
لَحمدي لا بدا في نحــور العِدا
تنزع الروح من بين الأبدانِ
كُبّ تالي بكيل كب كم من جعيل
كُبّ نهمي وجبري وخولاني
حاشدي وأرحبي لا بسين العبي
والمفارس خراطيُم غربانِ
كُبّ كم من بليد بندقه من حديد
بندقه شغل محبوب صنعاني

52(from al-Ḥajrī Buldān p. 115)

  • 79 al-ʿAlīmī (Qaḍā’ qabalī p. 134) gets this badly wrong, guessing al-manūʿ might be “representatives (...)

53Domiciled among the sections of Dhū Muḥammad one finds people called manūʿ, who are spoken of as having been agricultural workers or sharecroppers (ḥurrāth, shurakā’) and who, unlike the groups discussed so far, were not of full tribal status.79 In the early 1980s there seemed to be about four or five hundred men of al-manūʿ, scattered among several villages. They shared ghurm wa-jurm with Dhū Muḥammad, in other words they contributed to joint debts and expenses in tribal conflicts; they could not and cannot, however, make undertakings to escort or protect others in their own name. Nor could they shelter one Muḥammadī from another.

  • 80 The carrying of arms is the usual test of tribal status, but how reliable it is in the present case (...)

54Although they would fight alongside Dhū Muḥammad in battles with other tribes, if one of them were killed in circumstances other than war then recompense was demanded four-fold, a common sign of subordinate or protected status despite the fact al-manūʿ bore arms.80 Their position is now a sensitive issue and hard to enquire about. On several occasions I have blundered by asking about the status of al-manūʿ in a qat-chew where it turned out one of al-manūʿ was present. On each occasion people changed the subject and quietly told me to do so.

55Some practical ambiguity, also, can attach to this status. Dhū ʿAlī, who live around Khays, Malāḥah and Zafq al-Ṭawīl, and were reckoned manūʿ by most people, seem to have been attached partly to Dhū Mūsā and partly to Āl Dumaynah. After a difference with Dhū Mūsā, in the late 1970s, they found Āl Dumaynah would only help them on the clear admission that they were dependants. They went to Dhū Zayd for help and concluded a pact of brotherhood (ṣaḥb, ikhāʼ). This implies equality. The rest of Dhū Muḥammad appear to have declined to recognize its validity, and through the 1980s Dhū ʿAlī acted as a quasi-independent section (about 70 rifles strong), neither clearly manūʿ nor clearly part of Dhū Muḥammad,

  • 81 al-Akwaʿ (Madkhal p. 40) suggests ʿurwah means gifts between affines. In fact it means simply the b (...)
  • 82 The term ʿurwah and the related phrase ʿurūq al-laḥm wa-l-dam are said by many at Baraṭ to include (...)

56For the moment we can only suspend judgement on what the status of al-manūʿ may originally have been in detail. The qawāʿid are unclear. Text B section 28 says that if any of them offends against a tribesman of Dhū Muḥammad then he has only one dayʼs right of escort, and presumably has to leave Baraṭ; but if the next line also refers to them (one cannot be entirely sure) then the manūʿ are nonetheless “brothers” with Dhū Muḥammad in affairs that concern the whole tribe. Most intriguing of all is the association of the term manūʿ with the term ʿurwah (text 1 section 14), which usually means specifically the bond produced by marriage.81 Whether that linguistic connection denotes habitual kinship, whereby al-manūʿ would structurally have been sistersʼ sons to the tribesmen as mothersʼ brothers, or denotes a mere parallel of subordinate status is obscure.82

57Tribesmen in discussion now tend to deny inequality in marriage between different types of people and say simply the manūʿ are attached to land (yuḥibbū al-­arḍ). The manūʿ themselves sometimes claim they are Baraṭʼs original inhabitants and perhaps descendants of the legendary Banī Hilāl, left behind in Yemen at the great migration which followed the collapse of the Ma’rib dam. The story that al-manūʿ were there first at Baraṭ is widely accepted in Dhū Muḥammad, and many suppose that al- manūʿ until fairly recently made up a third of Baraṭʼs settled population: they are said to be found also in Dhū Ḥusayn, Āl Sālim, and al-ʿAmālisah, It may be worth noting that none attaches to al-Maʿāṭirah,

  • 83 Most of these “weak” people appear to have been concentrated around al-ʿInān and Kharāb al‑Marāshī: (...)

58Though also of subordinate status and protected by the tribes, the jīrān should be distinguished from the manūʿ. The term jīrān (pl. of jār) means “neighbours”. It can apply to tribesmen, as it plainly does in the market guaranty (text 1 section 8) where the phrase ḍumanā’ jīrān means guarantors who act together to uphold the market-peace. But it also applies, quite differently, to non-tribal protected persons, not least the bayyaʿ, the traders or market-folk.83 These are the “people of Baraṭ” referred to in text A section 16 (text B section 24). They had no “face” (wajh), that is, no ability to make undertakings, to protect people or provide surety, and were spoken of as “weak” (ḍuʿafā’). Although again this is a delicate subject, it seems fairly clear that tribal families took wives from them but were unwilling to give them wives. Unlike the manūʿ, the jīrān did not receive or contribute to collective payments in disputes with other tribes, except with regard to losses at the market (text 1 section 14), and certainly they did not bear arms. All these people, like the dawāshīn or non-tribal heralds at Baraṭ, were settled villagers, described as hajarī or qarawī.

  • 84 See, for instance, the passage from Ḥabshūsh quoted earlier in Arabic, Ru’yah p. 83. Also op. cit. (...)

59In Dhū Ḥusayn and Banī Nawf, and further south near Ma’rib, one hears the term qarār used to mean “weak” non-tribesmen or jīrān. Ḥabshūsh in the late nineteenth century gives the singular qarārī,84 which I have not heard in Dhū Muḥammad. The word qarawī (pl. qurā), literally a “villager”, can sometimes be used of the manūʿ as opposed to tribesmen, as it is in text A section 8 (the logic perhaps is that they are assumed to have been there before the tribes of Dhū Ghaylān arrived), but usually in Dhū Muḥammad it just means a settled person as opposed to a nomad or badawī (pl. badu) and refers to someone of full tribal status. Text A section 6 contrasts badawī and qarawī, text B section 5 says badawī and hajarī; but whether one lives in a house or a tent makes no difference. What does make a difference, as tribesmen see things, is descent, and although tribal status may perhaps have been lost in the distant past by those later spoken of as “weak”, it cannot explicitly be gained. Supposedly tribesmen have been there in effect forever.

Notes

1 Rashād al-ʿAlīmī: al-Qaḍā’ al-qabalī fī l-mujtamaʿ al-yamanī (1986) p. 43. We shall touch again on Bin Zinbāʿ in footnotes later. For the phrase qawāʿid al-sabʿīn, al-ʿAlīmī: op. cit. pp. 55-6.

2 Faḍl Abū Ghānim al-Bunyah al-qabalīyah fī l-yaman (1985) pp. 361 ff. A mention in the text of Arḥab al-Maṣʿabī (ibid. p. 383) makes it almost certain this is the second of four manuscripts described by Rossi, cf. E. Rossi, Il diritto consuetudinario delle tribu arabe del Yemen, Rivista degli Studi Orientale vol. 23 (1948) pp. 18-25.

3 For the association of these phrases with Baraṭ, see R.B. Serjeant, The interplay between tribal affinities and religious (Zaydi) authority in the Yemen, al-Abḥāth vol. 30 (1982) pp. 43-4.

4 ʿAbdullāh ʿAlī al-Wazīr Ṭabaq al-ḥalwā wa-ṣiḥāf al-mann wa-l-salwā (Muḥammad Jāzim ed. 1985) p. 223. For simplicity, most terms in the present work will not be inflected more than they are in colloquial pronunciation. The letter ʿayn is dropped from transliteration if followed by an apostrophe.

5 al-Nubdhah al-mushīrah ilā jumal min ʿuyūn al-sīrah fī... l-qāsim bin muḥammad (anon. n.d.) p. 406. For the institution of hijrah or tribal protection, Serjeant ‘Tribal affinities’ pp. 25-8, P. Dresch Tribes, Government, and History in Yemen (1989) pp. 124, 135-6, 138-51.

6 Muḥammad Zabārah Āʿimmat al-yaman bi-l-qarn al-rābiʿ ʿashar (1956) vol. 2 p. 8.

7 For an 18th century example, Ḥusayn al-ʿAmrī Mi’at ʿām min tārīkh al-yaman al-ḥadīth (1988) p. 365. For a further selection of tribal documents see Abū Ghānim Bunyah qabalīyah (1985) and al‑Qabīlah wa-l-dawlah fī l-yaman (1990).

8 C. Rathjens, Ṭāghūt gegen Scherīʿah, Jahrbuch des Museums für Landes- und Völkerkunde (1951).

9 Ibn al-Mujāwir Ṣifat bilād al-yaman wa-makkah wa-baʿḍ al-ḥijāz, ... tārīkh al-mustabṣir (Mamdūḥ Ḥasan Muḥammad ed. 1996) p. 117. It has to be said that Ibn al-Mujāwir mentions many odd things quite calmly. But it is interesting that he feels no need to explain the term manʿ to his readers elsewhere. For the following quotation see Serjeant, Sunnah, qur’ān, ʿurf in C. Toll and J. Skovgaard-Petersen (eds.) Law and the Islamic World (1995) p. 45.

10 See e.g. Ibn Hishām (d. AH 213 or 218), cited Dresch Tribes p. 196. The term ṭāghūt was used in southern Yemen as well as the north, but the case of this Serjeant cites (Sunnah, qur’ān, pp. 46-7) is from a late 19th century author of Salafī inclination. Though it is often difficult to be sure from Serjeantʼs work which periods are at issue, the impression he gives is that the term manʿ was current fairly recently in southern Yemen.

11 For such phrases as aḥkām al-ṭāghūt and ḥukkām ṭawāghīt see e.g. Ḥusām al-Dīn Muḥsin Ṭārīkh al‑yaman ʿaṣr al-istiqlāl ʿan al-ʿḥukm al-ʿuthmānī al-awwal 1056–1160 hijrī (ʿAbdullāh al-Ḥibshī ed. 1990) pp. 24, 35. In the “time of corruption” in the 19th century the word is extremely common. When marauding tribes are driven away from Ṣanʿā’, and the roads made safe, a Ṣanʿānī author says simply “the ṭāghūt was effaced”: Ḥusayn al-Sayāghī (ed.) Ṣafaḥāt majhūlah min tārīkh al-yaman (1978) p. 74.

12 Dresch Tribes p. 199, from Nubdhah mushīrah p. 435. For the Khawlānī passage, Nubdhah mushīrah p. 276. Sources are patchy and my own knowledge very limited, but this use of ṭāghūt seems more common from about al-Qāsimʼs time.

13 Zabārah Āʿimmat al-yaman vol. 3 p. 197. For the story of threatening execution, below, see Serjeant ‘Tribal affinitiesʼ p. 44.

14 B. Messick The Calligraphic State: textual domination and history in a Muslim society (1993) pp. 140, 182-4 and passim. Also Y. Linant de Bellefonds, Un problème de sociologie juridique. Les terres ‘communesʼ en pays d'Islam, Studia Islamica vol. 10 (1959).

15 For a useful discussion of codification in Southeast Asia, W. Roff, Islam obscured? Some reflections on studies of Islam and society in Southeast Asia, Archipel vol. 29 (1985). For North Africa, e.g. K. Brown, The impact of the dâhir berbère in Salé in E. Gellner and C. Micaud (eds.) Arabs and Berbers (1973). For the beginnings of the idea of distinctive Berber law, A. Bernard and L. Milliot, Les qânouns kabyles dans l'ouvrage de Hanoteau et Letourneux, Revue des études Islamiques vol. 7 (1933).

16 Muḥammad Jāzim (ed.) Nūr al-maʿārif fī nuẓum wa-qawānīn wa-aʿrāf al-yaman fī l-ʿahd al‑muẓaffarī l-wārif parts one and two (2003, 2005).

17 Ḥusayn al-Sayāghī (ed.) Qānūn Ṣanʿā’ fī l-qarn al-thānī ʿashar al-hijrī (Ṣanʿā’ n.d.), R.B. Serjeant, The statute of Ṣanʿā’ in Serjeant and Lewcock (eds.) Ṣanʿā’, an Arabian Islamic city (1983).

18 Even this contrast between sharʿ and sharīʿah, although common in documents, should be treated carefully. On occasion the two terms are synonyms for Islamic law, contrasted with other custom (Rossi ‘Dirittoʼ pp. 24, 33). On the following point, that sharīʿah judgement should be freely chosen, see P. Dresch, Keeping the imam's peace: a response to tribal disorder in the late 1950s, Peuples Méditerranéens vol. 46 (1989). Tribesmen might well agree to mandatory sharīʿah judgement, but usually within the bounds of some specific place or with reference to specific persons, cf. Ismaʿīl al‑Akwaʿ al‑Madkhal ilā maʿrifat hijar al-ʿilm wa-maʿāqil-hu fī l-yaman (1995) pp. 41-4.

19 Ḥayīm Ḥabshūsh Ru’yat al-yaman (1992) pp. 37, 130, Yémen (Samia Naïm-Sanbar trans. and ed. 1995) pp. 68, 197, E. Glaser Reise nach Ma’rib (1913) p. 5.

20 Rossi ‘Dirittoʼ p. 16, also p. 34. In Saudi Arabia, surprisingly, ṭāghūt may perhaps be a term used by tribesmen themselves of tribal law. But I have not heard this in Yemen, and the slippage between a term of denunciation and more substantive self-description needs treating cautiously: cf. F.H. Stewart, Ṭāghūt in New Encyclopaedia of Islam vol. 10 (2000) p. 94.
Whether ṭāghūt and ṭāgh(in) are really cognates seems doubtful. But whenever I have asked tribesmen the meaning of the former term I have been given examples of what in English one would call tyrannical conduct: e.g. arbitrary demands for wealth or labour. I do not think a tribesman has ever spontaneously described ḥukm ʿurfī to me as ṭāghūt.

21 See e.g. M. Mundy, Womenʼs inheritance of land in Highland Yemen, Arabian Studies, vol. 5 (1979).

22 Messick Calligraphic State pp. 31, 35, 61 and passim.

23 The phrase histoire historisante comes of course from Blochʼs Métier d'historien.

24 M. Meeker, Meaning and society in the Near East: examples from the Black Sea Turks and the Levantine Arabs, International Journal of Middle East Studies vol. 7 (1976), P. Bourdieu, The sentiment of honour in Kabyle society in J.G. Peristiany (ed.) Honour and Shame (1966), J. Anderson Doing Pakhtu: social organization of the Ghilzai Pakhtun (Ph.D. thesis 1979).

25 Compare Yemeni material with e.g. G. Murray Sons of Ishmael: a study of the Egyptian Bedouin (1935) or K. al-Hilw and S. Darwish Customary Law in Northern Sinai (1989). Also e.g. R. Montagne, Documents pour servir à l'étude du droit coutumier du sud­-marocain, Hespéris vol. 7 (1927).

26 F.H. Stewart, Tribal law in the Arab world: a review of the literature, International Journal of Middle East Studies vol. 19/4 (1987), Texts in Sinai Bedouin Law parts one and two (1988, 1990).

27 R.B. Serjeant, Materials for South Arabian History, Bulletin of SOAS vol. 13 (1950), Two tribal law cases (documents): Wāḥidī Sultanate, Journal of the Royal Asiatic Society (1951), Dawlah, tribal shaykhs, the manṣab of the Waliyyah Saʿīdah, qasāmah in the Faḍlī sultanate in M. Ibrahim (ed.) Arabian Studies in Honour of Mahmoud Ghul (1989).

28 Serjeantʼs copy differs somewhat from Rossiʼs account and suggests perhaps access to further copies. He also gave me part of Rossiʼs second text (cf. Abū Ghānim Bunyah qabalīyah pp. 363‑84: again there are differences) and several sections of Kitāb al-ādāb wa-l-lawāzim, on which he had plainly done much work. Unfortunately these copies are not very clear and have, for instance, missing margins. Serjeantʼs papers were given by his widow, Marion, to the University of Edinburgh, where presumably the originals now are.

29 Ḥabshūsh Ru’yah, Yémen; Glaser Ma’rib, ‘My journey through Arḥab and Ḥāshid’ (1884, 1993). Carlo Landberg published a great deal, but the most convenient source for the present purpose is his wonderful Glossaire daṯinois (1920-42).

30 Ḥabshūsh Ru’yah p. 83. The spelling of Arabic in what follows is that in the original.

31 Material damages (for instance blood-money) are nominally the same for all males. The multiple here is paid partly to the victim, partly to his protector, sometimes wholly to the latter. While offences against judges (quḍāh) and “weak” market-folk are each supposed to be compensated four-fold at Baraṭ, those against descendants of the Prophet (sādah) and heralds (dawāshīn) are supposed each to be eleven-fold. For notes on the dawāshīn see Dresch Tribes pp. 120-1, 406-7.

32 Serjeant ‘Sunnah jāmiʿah’ p. 14. The order of terms has been reversed for euphony. The term manīʿ derives from manʿ or manʿah), which itself has associations with defence and protection, Rossi ‘Diritto’ p. 29, Serjeant ‘Sunnah, qur’ān’ p. 45.

33 Overlaps are found even within Rossiʼs material. If Rossiʼs (untitled) second manuscript is indeed Abū Ghānimʼs text, then certain passages recur out of sequence in Kitāb aI-tabyīn (cf. Rossi ‘Diritto’ pp. 23, 25‑6, Abū Ghānim Bunyah qabalīyah pp. 370, 372). Resemblances extend beyond this. To take an obvious case, tribesmen will claim, quite wrongly, that the present documents (texts A and B) provide even for what happens if one kills someone’s dog. They are often quoting almost verbatim from Serjeantʼs manuscripts although none has heard of these, far less read them. It is difficult not to feel that there is, or was until recently, a fund of shared formulae extending through much of Yemen.

34 P. Dresch, The Northern Tribes of Yemen: their nature and their place in the present republic (D.Phil. thesis, 1982); Placing the blame: a means of enforcing obligations in Upper Yemen, Anthropos vol. 82 (1987); Tribes (1989); Guaranty of the market at Ḥūth, Arabian Studies vol. 8 (1990).

35 Abū Ghānim Bunyah qabalīyah pp. 385 ff. Neither he nor I took proper note of Ḥamūd al-ʿŪdī’s earlier effort at fragments of this text, al-Turāth al-shaʿbī wa-ʿalaqāt-hu bi-l­-tanmīyah (1980) pp. 220-7. The last transcription date of al-ʿŪdī’s copy suggests this is very much the same version that Dresch and Abū Ghānim used.

36 al-ʿAlīmī Qaḍā’qabalī pp. 118-40. Neither the preamble to the text nor the list at the end of guarantors and recopyings is given, and the order of the clauses is changed, which makes the document more difficult to read. Also al-ʿAlīmī (op. cit. pp. 42, 117) dates the text to 1127 hijrī. This is much earlier than other versions, and I assume is a mistake: the figure 1211 has been miscopied as 127 and the assumption made that it must mean 1127. (For how easy this would be to do see the copy in Abū Ghānimʼs Bunyah qabalīyah).

37 Rashād al-ʿAlīmī is at present Minister of the Interior and has more pressing concerns than digging out old notes from his student days. But his recollection is that a copy was given him by the late Qadi ʿAlī al-Kuhālī.

38 al-ʿAlīmī Qaḍā’qabalī p. 117, cf. al-ʿŪdī Turāth shaʿbī p. 219. What is not clear is whether the changed order of certain sentences and sections in the published versions derives from a manuscript copy or from an editorʼs attempt to make sense of the text.

39 Cf. an editor’s comments on Anglo-Saxon laws: “... it is not in all cases possible to expand their laconic statements without committing oneself to an interpretation which was not necessarily the author’s meaning. There are many places also where there can be doubt as to what antecedent a pronoun refers [to]”, D. Whitelock (ed.) English Historical Documents c. 500–1042 (1979) p. 365.

40 Vocabulary differs from place to place, and so does use of the same vocabulary (the way rifles as pledges are described would fill a pamphlet). But so, on occasion, do more profound assumptions. Men from Dhū Muḥammad, for instance, are very cautious about fighting groups that include their in-laws: al‑ʿAmālisah, more or less next door, treat the in­laws as responsible for opting out of the dispute.

41 One does not wish to suggest that nothing at all changed until recently and that Baraṭ lived in a timeless world until the 1962 revolution. But there is no doubt things have changed very fast in the last twenty years or so. The foreign ethnographer thus finds himself these days sharing terms and common sense with a few old men, while younger people, often raised primarily in towns, do not recognize even the terms.

42 ʿAbdullāh Nājī al-Ḍuwaynī of al-Maʿāṭirah was also enormously helpful, and so were friends such as the late ʿAbdullāh Dhaybān of Sufyān, who shared much of Baraṭʼs vocabulary. It has been a long time. A lot of old friends have died.

43 The term naqīb is used as a title by certain shaykhly families, mostly from Bakīl tribes (Dresch Tribes p. 405). Presumably it was given them by some ruler or set of rulers, but I cannot tell which set of rulers or when. Although naqīb was already current as a tribal title, 18th century sources mainly use it of what look to be slave commanders.

44 These thanks are much more than formal. Although, for instance, I have known Ḥamūd Muḥammad Abū Ra’s half a lifetime, it was well beyond the call of friendship for his son Fahd to track down on his own initiative and have copied for me a version of the market-­guaranty. Working in the Dāris dīwān was especially interesting. Although anthropologists as a matter of course distrust the idea of “disappearing cultures”, there was no doubt what was happening: those furthest from the world of Ṣanʿā’, particularly badu living east of Baraṭ, were by far the most familiar with the terms and phrases in the older documents.

45 E.g. CEFAS Le Règlement des conflits tribaux au Yémen (2003), B. Dupret, Systèmes coutumiers, centralisme juridique de l'état et usages du droit, Chroniques Yéménites vol. 8, Dupret (ed.) Le Shaykh et le procureur: systèmes coutumiers et pratiques juridiques au Yémen et en égypte (2005), particularly al‑Zwaini’s useful piece.

46 I do not have a linguistʼs ear, and asking people to write vowels, where they know how to do this, runs the risk of them reverting to classical pronunciation. But this occurrence of initial kasrah in imperfect verbs, active voice, is common. It also occurs in many people’s pronunciation of certain verbs in the perfect tense, e.g. fisala for fasala, to fail in one’s duty: al-Akwaʿ (Madkhal p. 32) goes so far as to suggest fisila, which I doubt.

47 See H. Steffen et al. Final Report of the Airphoto Interpretation Project (1978) p. II/183-4. The aerial photograph on p. II/171 gives an excellent idea of times and distances.

48 The version given here in photocopy was kindly found me by Ṣāliḥ Muḥammad al-Farajī. Nūr al‑Dīn ʿAbdullāh Dāris later found me a copy that belonged to Aḥmad ʿAlī Fāḍil. The latter is helpful at some points to check readings of the former but lacks many details. Fahd Ḥamūd Abū Ra’s later had made for me a further copy and acquired a very faded (no less illegible) photocopy of a version dated 1275 hijrī. Unfortunately none of these gives a chain of transmission.
The translation (part four, below) suggests how less “finished” a text this seems than texts A and B despite the few isolated classicisms. Readers from Baraṭ, just as much as the translator, are left guessing. But comparison with other original documents (see e.g. Abū Ghānim al-Qabīlah wa‑l‑dawlah) gives no safe stylistic grounds for deciding the documentʼs likely history.

49 Leaving off the copyist’s name means also leaving off an official stamp from the ibtidā’ī law‑court at al-ʿInān, which some would seize on as evidence of rapprochement among idealized versions of ʿurf, qānūn, and sharīʿah. In fact all the court does in a case like this is to stamp the copy as being written on the date claimed, rather as a notary or commissioner of oaths might do: no legal opinion is offered on the contents.

50 Perhaps I should say lent rather than given. Even where people have offered hand-written versions it has always seemed best to make a copy and return the manuscript. Yayqar Muḥammad’s copy is written in beautiful script, with many spellings and many points of orthography corrected to classical form, but at 65 x 35 cm. is simply too large to be reproduced here.

51 The document begins with the common formulae of “renewing” or “clarifying” existing custom. What there may have been earlier is unclear. Serjeant (‘Tribal affinities’ p. 43) suggests that Ṭabaq al‑ḥalwā, a century before, contains references to “the laws of Baraṭ”, but I cannot find this in the published version. The copy given me by Nājī ʿAbdullāh Dāris is dated 1171, but the others are all very clear and I am inclined to think this a miscopying.

52 A copy from Muḥammad al-Sudumī, made for the Ministry of Justice in the early 1970s, was passed on to me by Laila al-Zwaini, but neither al-Sudumī’s copy nor his notes adds anything to what we have. Even the copies made at Baraṭ or in Lower Yemen simply pile up. I have 8 copies of text A, now, and 9 of text B, but they only repeat what is in the copies cited or differ in such obvious ways that one can strike out minor transcription errors.

53 This is taken from text B. For text A see the photocopy in part five, below. Some explain the repetition of names by saying guaranty is “inherited”, which ideally it is. But comparison with other documents of this sort (see e.g. al-Akwaʿ Madkhal) suggests that changes of the order apparent here should be freshly witnessed.

54 The earlier date seems marginally the more likely, appearing clearly in the better-written versions by copyists who are usually careful to note if something is illegible.

55 Bayt Ṭuhayf were part of the qādī family, Bayt al-ʿAnsī (of whom more in parts two and three). The name is no longer current, but ʿAbduh Ismāʿīl Aḥmar al-Shaʿr remembers the story. A ṭahfah is a small white moth of the kind that flutters around lamps at night. The ancestor of the family is supposed to have worn a particularly extravagant white turban whose ends flapped about, hence Ṭuhayf, the little moth.

56 Resolving this would need tracking down the names in Lower Yemen, which I have not had a chance to do. The later date (1349 hijrī) appears in Yayqar Muḥammad’s copy which, despite the beauty of the script, plainly miscopies one of the earlier dates. All one can do for the present is suspend judgement.

57 For a genealogy of Bayt Abū Ra’s, including Ḥusayn b. Mufliḥ, see Muḥammad al-Yāzilī, Min al‑thawrah al-bikr ilā l-thawrah al-umm (2002) pp. 30-31.

58 Unfortunately I never managed to get the story from Muḥsin Ismāʿīl of how the copy was made and how it got into wide circulation, as it did quite quickly. The copy I photographed at Baraṭ was made for Ḥamūd Nājī Dāris exactly a year later; Abū Ghānim reproduces one made for ʿAbdullāh Nājī Thawābah just nine months after that. Al-ʿŪdī’s copy was made in Jamād Awwal 1389.
Muḥsin Ismāʿīl’s preface says also that the person who dictated or passed on (mumlī) the text was Ḥājj Ḥusayn Dawmān. Bayt Dawmān are the main dawāshīn or heralds of Dhū Muḥammad. Traditionally the dawshān often held important documents, and would not show them to others without a formal request from his shaykhly sponsors.

59 Credit for this explanation, which I think very likely to be right, goes to Muḥsin al-Ghuraybī. I had fretted over the phrase, in a vague sort of way, for two decades and not spotted it. For a parallel usage see the 1231/1816 agreement given later as Document 3.

60 The postscript to the Thawābah copy is given in full with the list of guarantors in part six, below. See also Abū Ghānim Bunyah qabalīyah p. 400.

61 Cf. Muḥammad al-Ḥajrī Buldān al-yaman wa-qabāʿil-hā (1984) pp. 109-11, which agrees with what members of Āl Faraj say nowadays. For higher-order connections in Dhū Muḥammad see also Steffen et al. Final Report part four. An error in transliterating someone’s notes means Āl Aḥmad bin Kawl came out as Aḥmad Dunkal, but otherwise Steffen’s account is good on local detail.

62 Swaydān is a rather mythic name, and where Bāb Swaydān may have been is a mystery. Many at Baraṭ say it was a gate at Ṣaʿdah, but it seems not to be listed as such in standard gazetteers.

63 M. Kister, The market of the Prophet, Journal of Economic and Social History of the Orient vol. 8 (1965), R.B. Serjeant, The sunnah jāmiʿah: pacts with the Yathrib Jews and the taḥrīm of Yathrib, Bulletin of the School of Oriental and African Studies vol. 41/1 (1978).

64 Ḥabshūsh Ru’yah pp. 107-8. For a discussion of the Ṣuraymī document referred to below (Document 2) see Dresch ‘Ḥūth’ pp. 68-70. For markets more generally see W. Dostal, The special features of the Yemeni weekly market system, New Arabian Studies vol. 3 (1996).

65 P. Dresch, A pact of brotherhood from Sufyān (North Yemen), Chroniques Yéménites vol. 10 (2002). For Imam Yaḥyāʼs politics in the 1920s, Dresch ‘Ḥūth’.

66 M. Mauss, Essai sur le don: forme et raison de lʼéchange dans les sociétés archaïques, Année Sociologique 1925, pp. 180-1, M. Hocart, Kings and Councillors: an essay in the comparative anatomy of human society (1936, 1970), pp. 257-8.

67 Things may have been different elsewhere in Yemen, and the rule may not have held throughout the north. For references to market-tax see Ḥusayn al-ʿAmrī: Yemen in the 18th and 19th Centuries (1985) p. 121.

68 The same system of monthly responsibility applied to protection of a learned family at Baraṭ that we shall look at later: al-Akwaʿ Madkhal p. 36. One cannot be entirely sure of dates from the layout of texts, but probably this provision dates to AD 1851. How old the system of rotating market-guaranty may be is unknown.

69 In other words, they are not expected to fight. While Dhū Muḥammad speak of them often as a “sixth fifth”, al-Maʿāṭirah consider themselves a tribe apart. In the course of disputes in the 1970s and 1980s which we shall sketch in the footnotes to part two, they found themselves at odds with Dhū Muḥammad as a whole and attempted to establish a market of their own at Ḥajjān.

70 Āl Ḥuṭbān, who are not at all numerous, seem to live mainly to the east of Baraṭ around Wādī Khabb, cf. al-Ḥajrī Buldān p. 113. The category Āl Ḥusayn, though I have never succeeded in pinning this down, is supposed to include a number of ashrāf, or groups descended from the Prophet.

71 ʿAmr is nearly always spelled as if it were ʿUmar. Although the classical waw is added in books, and in modern renderings such as that on the wall of ʿAbdullāh Dāris’s reception room (below), most tribesmen insist it is simply wrong: some have even said the name should be pronounced ʿAmir. In transcribing the texts I have followed the local preference.

72 There are very few plainly ʿAmrī names among the signatories to text B. In one of al-Akwaʿʼs documents (al-Akwaʿ Madkhal pp. 47-8), which clearly names Dhū ʿAmr, I am not sure one can find any. Nowadays Āl Baṭḥān and Āl Daḥwā are still thought of in Dhū Mūsā as ʿAmrī. While people “remember” al-Marāshī as once all controlled by Dhū ʿAmr, the trail peters out in fragmented folk-tales.

73 See e.g. text B section 23. One might have expected the pact to spread further and involve Dhū Ḥusayn, but apparently it never did. Repeatedly people have said they know of a joint Muḥammadī-Ḥusaynī document and then produced Abū Ghānim’s copy of text B, which he labelled misleadingly (Bunyah qabalīyah p. 385). Other agreements between the two neighbours of course exist (see e.g. Document 3) but not qawāʿid al-malāzim.

74 Steffen et al, Final Report p. II/119, have Shaʿbān and ʿAmr transposed, making Āl Sulaymān more distant than al-ʿAmālisah from Dhū Ghaylān. I followed suit, wrongly (Dresch Tribes p. 347). The scheme is here corrected on the basis of information from shaykhs of Āl Sulaymān. The spelling of names is corrected also. But, despite the unanimity of written accounts, the few people I have known from Āl Sālim all insist the order of al-ʿAmālisah and Āl Sālim should be as Steffen has it, and with it the order of ʿAmr and Nasr.
In the lower reaches of the genealogy some give ʿAmr(w) b. Muḥammad b. Ghaylān b. Muḥammad b. Shaʿbān, others (e.g. al-Yāzilī Thawrah p. 30) give Muḥammad b. ʿAmr(w) b. Ghaylān b. Muḥammad b. Shaʿbān.

75 Though it is hardly by itself evidence, one finds mention in the early 17th century of someone named ʿAlī Ṣāliḥ al-Maʿāṭirī al-Muḥammadī as if al-Maʿāṭirah were a sub-set of Dhū Muḥammad . See Nubdhah mushīrah p. 256. This is about as early as one hears of Dhū Muḥammad and Dhū Ḥusayn themselves.

76 According to some versions (e.g. Steffen Final Report p. II/119), Ghaylān includes also ʿIyāl Yazīd, miles away to the south and west in the area near Raydah, but this has no significance in practice. Few people at Baraṭ imagine any connection. The name Ghaylān is still recognized, however, as a toponym (I think in Āl Sālim’s territory) towards Ṣaʿdah.

77 Dresch Tribes p. 325. Serjeant’s reference to a mention in al-Rasūlīʼs Ṭurfat al-asḥāb (Serjeant ‘Tribal affinities’ p. 28) is surely wrong here. For the larger categories of Shākir and Dahm in the area a thousand years ago see al-Ḥasan al-Hamdānī, Ṣifat jazīrat al-ʿarab (D.H. Muller ed. 1891, 1968) pp. 167, 168, 194.

78 al-Akwaʿ Madkhal pp. 27 ff. The al-ʿAnsīs will be discussed in part two and the text itself in part three, below.

79 al-ʿAlīmī (Qaḍā’ qabalī p. 134) gets this badly wrong, guessing al-manūʿ might be “representatives of the state in the area”. There were none. The manūʿ were surely indigenous, and their status seems to have been like that of people referred to elsewhere in Arabia as khuḍayr. The manūʿ are probably the same stratum referred to in Dhū Ḥusayn as qabā’il al-wasaṭ, and wasaṭ in tribal usage near Baraṭ means not “in the middle” but “under the care of”.

80 The carrying of arms is the usual test of tribal status, but how reliable it is in the present case is doubtful. In the Jawf and at Baraṭ in the 19th century even some of the Jews went armed: Ḥabshūsh Ruʼyah 89, 112, Yémen p. 137, 170-1.

81 al-Akwaʿ (Madkhal p. 40) suggests ʿurwah means gifts between affines. In fact it means simply the bond of affinity itself. The term manʿ is also used of the relation produced by marriage. Dhū Ḥusayn at odds with Dhū Muḥammad in the 1980s ibtaraw min al-ḥilf wa-l­-manʿ (Dresch Tribes p. 86), meaning they repudiated all their obligations deriving from sworn alliance and from affinity.

82 The term ʿurwah and the related phrase ʿurūq al-laḥm wa-l-dam are said by many at Baraṭ to include not only affinity but the relation of being namesake (samī) to someone: a bond can be formed by naming one’s son for somebody of equal or higher status. Cf. Rossi ‘Diritto’ pp. 21‑2, al-Sudumī Naẓarīyat al-ʿuqūbah fī l-sharīʿah wa-l-aʿrāf al-qabalīyah fī aḥkām al‑qabāʿil al‑yamanīyah (1993) pp. 135-6. But how far the resulting obligation extended at Baraṭ is unclear: it seems not to have bound whole tribes.

83 Most of these “weak” people appear to have been concentrated around al-ʿInān and Kharāb al‑Marāshī: indeed al-ʿInān in the strictest sense, excluding what were thought of as adjacent villages, contained only market-folk. In tribal eyes the bayyaʿ were identified with weak servitors such as mazāyinah, small numbers of whom were scattered about tribal settlements. For a general account of their status see e.g. Dresch Tribes pp. 118-23.
One of the oddities of the present documents is that we find no mention of the Jews, who, though obviously different from the (Muslim) “market-folk”, were also jīrān or protected neighbours and must surely have been present at Baraṭ. Ḥabshūsh, a century later, has a good deal to say about them.

84 See, for instance, the passage from Ḥabshūsh quoted earlier in Arabic, Ru’yah p. 83. Also op. cit. pp. 52, 86. Several references to this family of terms are given by Rossi, ‘Diritto’ pp. 2-3, 8. In Sufyān one hears the plural qarwān used to mean simply settled folk. In Dhū Ḥusayn, on the other hand, qurā or qarā seems always to mean weak, non-tribal people or the manūʿ.

Table des illustrations

Titre Document 1: Dhū Muḥammad and Dhū Ḥusayn (1251/1836)
URL http://books.openedition.org/cefas/docannexe/image/920/img-1.jpg
Fichier image/jpeg, 428k
Titre Fig. 1: Contents of Qawāʿid al-Malāzim
URL http://books.openedition.org/cefas/docannexe/image/920/img-2.png
Fichier image/png, 223k
Titre Fig. 2: The fifths of Dhū Muḥammad
URL http://books.openedition.org/cefas/docannexe/image/920/img-3.png
Fichier image/png, 26k
Titre Document 2: Market guaranty from Banī Ṣuraym (1246/1831)
URL http://books.openedition.org/cefas/docannexe/image/920/img-4.jpg
Fichier image/jpeg, 388k
Titre Document 2: Market guaranty from Banī Ṣuraym (1246/1831) verso
URL http://books.openedition.org/cefas/docannexe/image/920/img-5.jpg
Fichier image/jpeg, 352k
Titre Document 3: Agreement among Dahm (1231/1814)
URL http://books.openedition.org/cefas/docannexe/image/920/img-6.png
Fichier image/png, 1,3M
Titre Fig. 3: Divisions of Shākir and Dahm
URL http://books.openedition.org/cefas/docannexe/image/920/img-7.jpg
Fichier image/jpeg, 428k

© Centre français de recherche de la péninsule Arabique, 2006

Conditions d’utilisation : http://www.openedition.org/6540

Cette publication numérique est issue d’un traitement automatique par reconnaissance optique de caractères.

Acheter

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search