Version classiqueVersion mobile

The rules of Barat. Tribal documents from Yemen

 | 
Paul Dresch

Part 1. The detailed setting

Tribes and borders

Texte intégral

  • 1 The inhabitants seldom present a unified front, however, and Baraṭ falls repeatedly to deception an (...)

1Discussion of ʿurf or “custom” makes only abstract sense without attention to the populations involved and the details of where they live. In this case the landscape and provisions for its use are distinctive. Jabal Baraṭ is a steep-sided granite plateau (altitude about 2,000 metres), the approaches to which from Ṣanʿā’, up Wādī al‑Nīl from Kharāb al‑Marāshī, are difficult; and those from the north and east are not a great deal easier. Elaborate plans are now afoot for tarmac roads, but even the bulldozed dirt roads of 20 and 30 years ago left the approaches to Baraṭ among those mythic places where a few men with rifles could delay an army.1 The Jabal, about 500 metres higher than al‑Marāshī, is classically said to be two days across in each direction. A more practicable estimate for modern times might be 40 km. north to south and 60 km. west to east, where (at around 1600 metres) the Jabal falls away into broken terrain before the desert begins in earnest.

  • 2 al‑Ḥasan al‑Hamdānī Ṣifat jazīrat al‑ʿarab (1968) pp. 194-5. Between the two parts of this flatteri (...)

2The plateau itself is sandy, with outcrops of darkly weathered rock. Some of these form spires, others boulder-strewn hills; and in many places weathering has left isolated large black rocks in the midst of nowhere which look as if they dropped on the sand from heaven. Between these outcrops the run-off (on those rare occasions the plateau gets decent rain) is such that trees grow, grazing can sometimes be very good, and fields are established. Deep wells provide some insurance against drought, and although field-crops suffer immediately that rain fails, small flocks can be watered for a year and more. Apart from the usual grain crops, mainly sorghum, there are walled grape-gardens and groves of date palms. Al‑Hamdānī a millennium ago described the plateau as “wide, with numerous different settled areas (buldān) and many crops... ; the summit of Baraṭ is one of the healthiest parts of Yemen, one of the most agreeable, and one of those with the most equable climate”.2 Were there only more reliable rain, Baraṭ would be paradise, but drought is frequent and in recent times periods of six or seven years without rainfall have been common.

3Along the wādīs on the plateau are scattered houses, which cluster rather loosely as settlements. Steffen’s demographic survey map from the 1970s, reproduced later as Map 2, gives a good idea of the lay-out, though the density of settlement has since increased. The looseness of the settlement pattern is suggested by the fact Muḥammadīs gave me a list of about 120 place-names just for part of the area shown by the map, and even for that small area my list and Steffen’s are not quite identical. From what documentary clues we have, the major settlement names seem quite stable; less expectedly, so too are the names of families and tribal sub-sections. The population itself, however, is less constant than the names suggest and, as we shall see later, many people have connections elsewhere in Yemen which are hugely important in times of famine. The houses, clustered around wells and along the wādīs, are often four or five stories high, some even six stories. In the late 1970s money from migrant labour and political disbursements to Baraṭ’s shaykhs funded a good deal of building. But tall mud houses are nothing new. Steffen’s team provide excellent photographs and drawings.

  • 3 H. Steffen et al. Final Report of the Airphoto Interpretation Project (1978) p. II/198.
    At a very ro (...)
  • 4 A good deal of detail is given in Muḥammad al‑Ḥajrī Buldān al‑yaman pp. 109-11. See also Ibrāhīm al (...)

4When Steffen and his colleagues were at Baraṭ in 1976 they reckoned that 60‑‑80% of Dhū Zayd may have been nomadic, and 30‑‑40% of both Āl Aḥmad b. Kawl and Āl Ṣalāḥ.3 Although the overall numbers have perhaps not fallen, the percentages would be lower now. One imagines that in the eighteenth century the proportion of Dhū Muḥammad who lived in tents and moved with flocks might have been quite high (at least as high as when Steffen visited), but the qawāʿid, for their part, seem to concentrate on settled tribesmen, hence for instance the passage on breaking and entering houses or firing from the houses of enemies (text B section 26). The qaṭīr who “places his house alongside another’s” (text B sections 6, 7) might be a nomad. But when the documents speak of getting a man “out of his house” by agreeing to truce (text B section 22), they almost certainly mean mud buildings not the badu’s “houses of hair”. This is despite the fact that many sections mentioned in the lists of signatories have been largely nomadic in recent times, and one imagines they were largely nomadic two hundred years ago.4

  • 5 The location of al‑Faraḥ (cf. Document 1, above) is surprisingly difficult to pin down.
    Many have s (...)
  • 6 Dresch Tribes pp. 343-4. When we come to mention areas further east again, the inadequacy of the st (...)

5Besides the many villages at Baraṭ, there are two small market-towns, which nowadays are centres of administration. Sūq al‑ʿInān is in Dhū Muḥammad; Rajūzah, about 10 km. to the east, is in Dhū Ḥusayn. Between these is a clear border-line which runs from near the head of Wādī al‑Nīl across the plateau and probably goes northwards then eastwards uninterrupted, separating the two tribes along the length of Baraṭ. Provisions on protecting the market (text B section 16) mention al‑Faraḥ, saying any offence committed by Dhū Ḥusayn west of there (by implication, in Dhū Muḥammad’s territory) is governed by the market-rules.5 Al‑Faraḥ is well south and east of Baraṭ: a line from al‑Faraḥ would join up readily with that on the plateau, and tribesmen often insist that a ḥadd or khaṭṭ fāṣil (a border or dividing line) also separates the two tribes east of the Jabal, as if territory were everywhere contiguous and exclusive. Between Baraṭ and the end of Wādī Silbah, 60‑‑70 km. to the north-east of al‑ʿInān, however, Dhū Muḥammad and Dhū Ḥusayn do not have a simple line between them. Rather, each claims specific areas within what appears to be simply Dahm’s territory.6 Dahm, a major subdivision of Shākir (part one, above), includes also al‑Maʿāṭirah and Āl Sulaymān, among others, both of which have similar restricted patches of territory along the wādīs east and north of Baraṭ,

  • 7 Presumably this would come as news to the tribes of Wādī Dawāsir or the Gulf hinterland. But the te (...)
  • 8 Dresch Tribes pp. 338-43. Much further north and east, at al‑Jāhimayn, was an area where Dahm, Wā’i (...)

6Further eastward still, Dhū Ḥusayn make extravagant claims about what the territory of Dahm as a whole might cover and if one asks where the furthest borders (aqṣā ḥudūd) of Dahm are, then one is sometimes told Qaṭar.7 More realistically, the meeting point of Dahm as a whole with al‑Ṣayʿar and Kurab is near Sharūrah, just north of the recently agreed Saudi-Yemeni border at the longtitude of central Ḥaḍramawt. A transect from west to east, from the main Ṣanʿā’-Ṣaʿdah road to the desert, by way of Baraṭ, in fact connects very different systems of territory and borders. Briefly, that to the west, in Sufyān, can be seen as related to systems in more densely settled areas further south, where the territory of subordinate units exhausts that of the larger set comprising them; systems to the east of Baraṭ, by contrast, merge with nomadic arrangements further east and north in which rights to resources may be determined by time or season as much as space, and tribal territory is not made up of exhaustive and exclusive section-territories.8

  • 9 ʿUlah is a very ancient name, mentioned in al‑Hamdānī, and formed part of Madhḥaj not of Shākir. It (...)
  • 10 Sundry other names are used. Ḥadrānī, for instance, seems to be used of Wā’ilah and Āl Salīm togeth (...)

7Some way north of Baraṭ lies Wā’ilah, which is spoken of as a quite separate branch of the tribal scheme. Dhū Muḥammad, Dhū Ḥusayn, Āl Sālim, Āl Sulaymān, and sundry others all think themselves part of Dahm. The opposite of Dahm, within the larger category of Shākir, is said to be ʿUlah, and Wā’ilīs may apparently refer to themselves or be referred to by Dahmīs as rijāl ʿulah, although the name seems only to refer to Wā’ilah: so far as I can gather, it does not align Wā’ilah with anyone else.9 Dhū Muḥammad and Dhū Ḥusayn together make up Dhū Ghaylān, and the collective name for Dhū Ghaylān and al‑ʿAmālisah is said to be ʿAmrī, as we mentioned in part one; the collective name for Dhū Ghaylān, al‑ʿAmālisah, and Āl Sālim together may perhaps be Nasrī.10 Tribesmen at Baraṭ will insist there are clear border-lines between each tribe and the next within this scheme as there are between Wā’ilah and Dahm. But quite where the line between Āl Sālim and Dhū Muḥammad runs, for instance, if there is a simple line, I am not sure. The upper reaches of Wādī Jawār, shown on the excellent Swiss map (Map 2, below) to the west of Sūq al‑ʿInān, belong to Āl Aḥmad b. Kawl of Dhū Muḥammad; the tributaries of this wādī, both from Qaṣabah Bin ʿĪshān and from Maqām Dhū Mahdī, appear to belong to Āl Sālim.

8On the Sufyān side of Dhū Muḥammad (to the west and south of Baraṭ) is a clear border-line, most of which coincides with the crest of Asḥar. Even this is complicated, however, by the presence of an outlying division of Sufyān, named al‑Marānāt, which lies well within Dhū Muḥammad’s borders near Kharāb al‑Marāshī. In Wādī Madhāb one routinely finds groups from Dhū Ḥusayn as well as Dhū Muḥammad. Al‑Jabal al‑Maflūq (Split Mountain, which seen from the south and west as one goes up the main road to Ṣaʿdah looks like a pair of elephants standing head to head) is divided at least three ways: Āl ʿAmmār to the north, Sufyān to the west, and Dhū Muḥammad to the east. What the division of the mountain itself may be, and for what purposes, is not certain. To the east and north of Baraṭ one’s attempts to map territory become still more difficult, or perhaps, in extreme cases, misdirected.

  • 11 I am grateful to Aḥmad ʿAbduh Thawābah for a copy of the document, and to many for discussing it. S (...)

9Some idea of the issues is given by an agreement drawn up among Dhū Muḥammad and their immediate relatives or allies in Rajab 1377 (AD 1958), probably in the course of a dispute with Wā’ilah.11 A copy is given below as Document 4.

  • 12 The last of these is a tiny grouping which seems to have minor settlements near Ṣaʿdah and in the l (...)

10Guarantors are first listed for four of Dhū Muḥammad’s “fifths” (the last, Dhū Mūsā, seems only to have joined when the dispute flared up again 25 years later). The pact can be joined, it says, by “the rest of their tribes” (bāqī qabā’il-hum), presumably meaning the rest of Dhū Muḥammad, and three other groups are then mentioned: not Dhū Ḥusayn, who are Dhū Muḥammad’s closest relatives genealogically, but al‑Maʿāṭirah, Āl Sulaymān, and al‑Kharashah.12 Dhū Muḥammad provide the leading names:

حضروا ذو محمد وتراضوا بما يرضي الله سبحانه وتعلى بجمع رأيهم علي حفظ حدودهم وقدمت الوجيه منهم فيما ذكر... وتواجبوا على رضاء الله سبحانه فيما بدا عليهم في حدودهم المعروفة لديهم اولا مِن جهة القبلة مجرا الوادي في العَطْف مصبابه الحَجر مِن مضيق خزام ومشرق المعروف بينهم واعشار القُبل القِبْلية ومِن جهة الشرق حيث نقيلة القِطَار جايزة بينهم وبين الاعشار وذلك غاوية الرباع الجامعة لهم حدّ دَهْم و الهَضْبِة وأم الشداد

Document 4: Agreement on borders involving Dhū Muḥammad (1377/1958)

Document 4: Agreement on borders involving Dhū Muḥammad (1377/1958)
  • 13 Relations with Wā’ilah seem seldom to have been good. In the mid-1970s, for reasons connected with (...)
  • 14 Connections south and east would be particularly interesting to know more about. Āl al‑ʿUkām’s hijr (...)

11The phrasing is obscure, but the main points to note are these. A border between Dahm as a whole and Wā’ilah (probably part of Yām also) coincides with Wādī al‑ʿAṭfayn, which runs north-east towards the modern international border-post of al‑Buqʿ. The document, however, specifies the “bed of the wādī at al‑ʿAṭf where it runs down to al‑Ḥajr from Maḍīq Khizām”: this appears to be some way short of the line between Dahm as a whole and Wā’ilah.13 Most probably the areas of shared concern to Dhū Muḥammad and Āl Sulaymān are not quite those of shared concern to either of them and, say, Dhū Ḥusayn, No mention is made, therefore, of where borders lie in the Upper Jawf.14 To the south, however, the borders are where Dhū Muḥammad meet Sufyān at the mountain ridge of Asḥar, adjoining Wādī Khabash (a tributary of Wādī Madhāb); to the west the border is al‑Jabal al‑Maflūq,

  • 15 Ghāwiyat al‑Rabāʿ, mis-pointed, is mentioned by al‑Baṣrāwī (Mashriq al‑yaman p. 17) as a border-lin (...)

12To the east of Maḍīq Khizām, says the document unhelpfully, the borders with the northern tribes are “known” (maʿrūf). Further east (and well to the north, in fact) is an area ḥayth naqīlat al‑qiṭār jāyizah bayna-hum wa-bayn al‑aʿshār, which, though Naqīlat al‑Qiṭār may also function as a place-name, presumably means the area where rights of refuge or co-residence are shared among them and other tribes. This is Ghāwīyat al‑Rabāʿ, where the signatories seem to share responsibility for the border of Dahm as a whole (al‑jāmiʿah la-hum ḥadd dahm.15 The relation is not clear, however, between Dahm’s shared rights or responsibilities and rights shared with other tribes: some in Dhū Muḥammad claim other tribes may seek refuge there with Dahm, others that refuge may be sought there with any of Dahm, Yām, or Wā’ilah. Also in the east are borders at Umm al‑Shidād (pronounced roughly Umm Ishdād), which lies east of Yatmah, itself a long way east of Baraṭ at the end of Wādī Silbah, and al‑Haḍbah, which is north of that.

13The pact says simply that all recognize their joint borders, “correcting what came before in the way of border-pacts between them”. Whatever happens within the borders is their joint concern, and the signatories will pay and be paid together, acting as “one hand and one arm”. But arrangements for payment are worth brief note. Generally badu compute ghurm (joint expenses and collective debts) on the basis of simple numbers; settled people often do so by tribal divisions, regardless of their exact size, as for instance the fifths of Dhū Muḥammad are meant each to pay the same although within a fifth counts of men or of households are made for each sub-section. Here it says expenses for the settled person (ghurm al‑hajarī) are to be assessed simply by the number of grown men (ʿalā l-rajul al‑kāmil) on what is usually the badu model. Both are mentioned, however, reminding us how far even Dhū Muḥammad extends. Members of the tribe may routinely be involved with grazing in places as far apart as al‑Jabal al‑Maflūq and al‑Haḍbah or Umm al‑Shidād, a straight-line distance of perhaps 120–130 km.

  • 16 The major shaykhs at Baraṭ have all for a long time been settled, but the minor sections to which t (...)

14The eighteenth-century qawāʿid (text A section 6, text B section 5) seem to give four days for news to reach members of the pact in the general area around Baraṭ, “whether nomad or villager”. But an announcement that a truce was over appears to have been effective immediately if made in the market at al‑ʿInān on market-day (text B section 20), perhaps on the assumption that news from there would reach even the badu at least as rapidly as their enemies might reach them. All the “fifths” contain badu as well as settled persons; often quite small sub-sections within a fifth also contain both.16 Although settlements in predominantly badu areas of course are small and scattered, the badu are present at Baraṭ itself as well in Wādī al‑Nīl or around al‑Marāshī and in regions to the north-east, making the signatories to these pacts a rather dispersed and variegated set of people whose nominal centre is the weekly market. I shall probably not have the chance of long-term fieldwork at Baraṭ, so let me try to put down what little I know in the hope that others will find out more.

Notes

1 The inhabitants seldom present a unified front, however, and Baraṭ falls repeatedly to deception and negotiation. A famous case in the early 1890s saw Fayḍī Pāshā slip into Baraṭ from Sufyān, retrieve his prisoners, loot al‑ʿInān and march out again, while 2,500 tribesmen assembled at Rajūzah to discuss what to do about it. See Muḥammad Zabārah, Ā’immat al‑yaman bi-l-qarn al‑rābiʿ ʿashar (1956) vol. 1/2, pp. 82-3, 97-100.

2 al‑Ḥasan al‑Hamdānī Ṣifat jazīrat al‑ʿarab (1968) pp. 194-5. Between the two parts of this flattering description of Baraṭ comes a famous description of its people, from Duhmah of Shākir bin Bakīl: “the bravest [or strongest] of Hamdān, protectors of women, and defenders of the weak [or of the neighbour]. They are called the Quraysh of Hamdān”. Cf. Rossi ’Dirittoʼ p. 9. They were not, however, called Dhū Ghaylān in those days.

3 H. Steffen et al. Final Report of the Airphoto Interpretation Project (1978) p. II/198.
At a very rough guess, about 20% of Dhū Muḥammad near Baraṭ these days might be nomadic; but in drought years many now move to the major cities.

4 A good deal of detail is given in Muḥammad al‑Ḥajrī Buldān al‑yaman pp. 109-11. See also Ibrāhīm al‑Maqḥafī Muʿjam al‑buldān wa-l-qabā’il al‑yamanīyah (2002), though this is less reliable and the proof-reading has left many names mis-pointed.

5 The location of al‑Faraḥ (cf. Document 1, above) is surprisingly difficult to pin down.
Many have said it is almost as far off as Shawābah, towards Wādī Dhībīn; some have said it is in Banī Nawfʼs territory. In fact it seems to be a well or cistern to the south-east of Kharāb al‑Marāshī, about a day’s journey from al‑ʿInān, beween Madhāb and al‑Ḥumaydāt.

6 Dresch Tribes pp. 343-4. When we come to mention areas further east again, the inadequacy of the standard maps becomes pressing. They do not all show the same settlements, or even show reliably the same physical features, and one is left flipping back and forth among different maps to guess where places stand in relation to each other. A great deal of basic survey work needs doing. But for the present purpose see the 1:500,000 map sections of which are reproduced by Steffen et al., and the 1985 1:500,000 map produced by the Yemen Arab Republic Survey Authority.

7 Presumably this would come as news to the tribes of Wādī Dawāsir or the Gulf hinterland. But the territory in question is not exclusive, cf. J.C. Wilkinson, Traditional concepts of territory in south-east Arabia, Geographical Journal vol. 149/3 (1983) p. 308.

8 Dresch Tribes pp. 338-43. Much further north and east, at al‑Jāhimayn, was an area where Dahm, Wā’ilah, and Yām all grazed animals in winter: Muḥammad al‑Baṣrāwī Mashriq al‑yaman al‑saʿīd (1974) pp. 16-17. For systems in northern Arabia, C.R Raswan, Tribal areas and migration lines of the North Arabian Bedouins, The Geographical Review vol. 20/3 (1930).

9 ʿUlah is a very ancient name, mentioned in al‑Hamdānī, and formed part of Madhḥaj not of Shākir. It recurs in southern Yemen. Whether in Wā’ilah’s case it is thought of as an ancestral name I do not know, but the phrase rijāl ʿulah is not uncommon.

10 Sundry other names are used. Ḥadrānī, for instance, seems to be used of Wā’ilah and Āl Salīm together. On occasion in the 1980s one heard it used as a “summons” to collective action. Ḥadrān, however, is not genealogical but apparently is the name of a wādī and mountain where the borders of the two tribes meet.

11 I am grateful to Aḥmad ʿAbduh Thawābah for a copy of the document, and to many for discussing it. Such pacts are made and remade constantly, see e.g. Document 3 in part one, above, dated 1231/1816. An agreement dated 1301/1884, which I was not able to copy, bound Dhū Muḥammad, Āl Sulaymān, and al‑ʿAmālisah to defend a section of border against Wā’ilah. The pact (ḥilf) between the former two tribes was renewed in 1977 and again in 1983, within a few months of Document 4, but directed against Dhū Ḥusayn.

12 The last of these is a tiny grouping which seems to have minor settlements near Ṣaʿdah and in the lower reaches of al‑Marāshī. The split distribution sounded so unlikely that I asked several Baraṭ shaykhs about al‑Kharashah specifically. There was fairly general agreement about where they live, but little about their place in the broad scheme of genealogy.

13 Relations with Wā’ilah seem seldom to have been good. In the mid-1970s, for reasons connected with national politics, Dhū Muḥammad were always at odds with them. Fighting recurred near Wādī al‑ʿAqīq, adjoining Wādī Amlaḥ, where Dhū Muḥammad do not, I think, share a border with Wā’ilah but Āl Salīm do, and thus so do Dahm as a whole.

14 Connections south and east would be particularly interesting to know more about. Āl al‑ʿUkām’s hijrah at al‑Marqab, for instance, although in al‑Marānāt, seems to involve shared borders of some kind among Dhū Muḥammad, Dhū Ḥusayn, Banī Nawf, and Sufyān.

15 Ghāwiyat al‑Rabāʿ, mis-pointed, is mentioned by al‑Baṣrāwī (Mashriq al‑yaman p. 17) as a border-line (ḥadd fāriq) between Yām, to the north, and Dahm to the south. Most people at Baraṭ describe it as actually quite a wide area. I have not had the chance to visit anywhere much north of Silbah, and my own information is thus hearsay.

16 The major shaykhs at Baraṭ have all for a long time been settled, but the minor sections to which their families belong often include nomads. Lesser families of shaykhs are very numerous; nor is the position of shaykh confined to a particular line. I doubt overall that there was much political or economic stratification between nomadic and settled members of the tribe. What mattered far more, as we shall see below, was access to funds from land elsewhere in Yemen.

Table des illustrations

Titre Document 4: Agreement on borders involving Dhū Muḥammad (1377/1958)
URL http://books.openedition.org/cefas/docannexe/image/824/img-1.jpg
Fichier image/jpeg, 501k

© 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