Version classiqueVersion mobile

The rules of Barat. Tribal documents from Yemen

 | 
Paul Dresch

Part 2. Customary law in context

Legal summary: substantive and procedural

Entrées d'index

Mots clés :

femmes

Keywords :

women

Texte intégral

  • 1 The only recent case on which I have much detail concerned a dispute between fifths of Dhū Muḥammad (...)

1The first thing, says text B (section 2), is wounding a woman. A culprit who shows by collective oath this was done in error pays the blood-compensation twice (dam wa‑dhamm), plus a camel and a length of cloth referred to as the price of tashwīf. If he fails to show the wounding or killing was unintentional, he pays four-fold and his right (sinnah, to give and receive protection) is valid or again recognized. This latter clause is surprising. The payment sounds too little. For cash-payment to be taken at all for killing a woman these days it would have to form part of a more complex settlement, whereby honour was somehow satisfied, and even then would usually be described as eleven-fold if it could not be defined as non-intentional (bi­-ghayr al‑ʿamd).1 We have no way of telling whether provisions in the Code for a womanʼs wounding or death being compensated four-fold were a theoretical provision seldom acted on or whether Baraṭ two hundred years ago was in some sense a rougher place than it is now.

  • 2 Circumcision used to be done between the ages of about seven and 14. It was also done this late in (...)
  • 3 Obviously this is a difficult subject to pin down (a gentleman does not enquire about another manʼs (...)
  • 4 Text A section 4 says also that amends are due, as well as compensation, if one uncircumcised man w (...)

2Text B section 3 (text A section 4) describes offences against uncircumcised males, where the detail of expectation has changed considerably. Circumcision is nowadays done in the first weeks of life. In those days it came much later,2 and those particularly in danger of being killed (the only son of a family owing blood to determined neighbours would be the obvious case) had circumcision postponed indefinitely. Although many at Baraṭ now reject the possibility as un-Islamic, it may have been that men could even marry without being circumcised;3 most insist also that uncircumcised males could not have born arms and their position must have been “like that of women”, but text A section 4 suggests perhaps they had sinnah, so that too must be in doubt.4 What we do know (or at least this is widely remembered) is that uncircumcised males were once distinguished by a ḥawṭ, an iron ring or collar around the neck, and their blood-money was that of other men. In the world of the documents, a man or boy who is not circumcised, whatever his age, cannot honourably be killed or wounded.

  • 5 Cf. al‑Sudumī Naẓarīyah pp. 119, 120, 122, Ḥabshūsh Ruʼyah p. 55. Even when blood-money is paid in (...)

3We are told that if he is killed, then the payment due, in addition to the standard blood-money, is a hundred sheep and a camel and a length of cloth (the camel and cloth here are the same as for tashwīf). If the killing or wounding can be shown by oath to be unintentional, then three quarters of the fine are dropped and the culprit again has his right (sinnah) to escort and protection. But, despite this, “the blood remains to be judged”. In other words, the dead personʼs kin may continue to seek revenge: the blood-debt itself and a threat of vengeance can persist indefinitely.5 Here is a point where rules, so to speak, are governed by other rules. If the culprit has paid the recognized amends, and thus has sinnah, one must visualize a kind of shell-­game whereby he moves from the protection of one group or person to another, perhaps seeking escort for specific journeys or refuge for specified periods, and is therefore often safe (in theory) from those who still want his blood.

  • 6 We mentioned in part two, above, provisions for escorting people to “the land of the state” when th (...)

4In the following section (text B section 4, text A section 5), which deals with mutilation, such as wounding a man in the face for instance (the term used again is shawfah or tashwīf), even that possibility is omitted and perhaps denied. If one secures the oath that oneʼs act was not intentional, one pays amends; but if the oath is not secured, and one is judged to have maimed a man deliberately, then one has no right, no companion, and no “link” to others (mā la-h sinnah wa-lā ṣāḥib wa-lā ʿulqah). The culprit has no right to escort or protection. In the end there are only two real options: either the wrongdoer stops constantly at home, besieged in his own house, or he leaves the tribeʼs territory and ceases to exist in local terms.6

  • 7 What was interesting about this case was its specificity. How realistic the sonʼs fears were I cann (...)
  • 8 Obviously if I know where he is then so do all Dhū Muḥammad, including his antagonists. But leaving (...)

5Some provisions in the present documents seem archaic at Baraṭ these days, but the principles informing such provisions remain familiar in most places north and east of Ṣanʿāʼ. It does not often happen that a man is trapped in his house indefinitely; it is not unknown, however. The son of a particular friend in Sufyān, for instance, having committed a squalid act of adultery close to home, spent more than three years like this, never leaving the house in daylight and after dark outdoors, when moving among the sorghum for a breath of air, never striking a match or lighting a cigarette for fear that someone was out on the hill-side with a rifle.7 At the time of writing, a friend from Baraṭ has chosen the other option and is living at the far end of Yemen, having fled the tribe. He killed a woman in the course of a gun-battle, and although he certainly did not intend to do so, he is not willing even to secure the oath and pay the resulting fine, so living at Baraṭ would mean a constant guard against assassination and for several years he has lived near Taʿizz.8 In the end most disputes are contained satisfactorily, if not resolved. But it can take some while.

  • 9 There is a whole series of Vth form verbs like this that usually take the preposition from”: e.g. (...)
  • 10 al‑Sudumī Naẓarīyah pp. 150-3. The account he gives of guaranty is accurate and useful, overlapping (...)
  • 11 For the verb istaqāma see al‑Sayāghī (ed.) Ṣafaḥāt p. 71, where someone at anʿāʼ in the 1850s stan (...)

6To concentrate on extremes is to misrepresent how such moral and legal systems work, and what appear, in for instance fieldwork, as “exceptions” to a rule of collective liability are actually examples of something else. On the middle ground, between exile and unconstrained violence, a person will yataḥajjab, that is, place himself under or be under anotherʼs care.9 The root hajaba suggests “covering”, and the meaning of this cluster of terms in ʿurf is that someone takes the culprit under his protection and makes himself responsible (yaltazim) for managing the debt the culprit has incurred. If a man from group A injurs someone from group B then not only those two individuals but the two whole groups are potentially at odds: someone else from A will often step in and yaḥjib (yaḥjub) ʿalā the original culprit, perhaps giving a rifle (bunduq al‑ḥujbah or bunduq al‑wifāʼ to the claimants. The immediate dispute is displaced to a relation between the claimant and a representative, so to speak, of the person against whom the claim is made, and the representative (if the rifle is taken) is not to be treated as an enemy. Muḥammad al‑Sudumī quotes in parallel the formula that man ḥajaba aw taʿayyaba ḥurima qitāl-hu, meaning roughly that a person who agrees to see through the process of payment cannot be offered violence.10 This, almost certainly, is what the qawāʿid (text B section 13, cf. A section 13 and B section 21) mean by the phrase al‑jār yastaqīm muṣān, “the neighbour stands in [for the debtor] as protected”.11

  • 12 This is what Stewart (ʻContractʼ p. 179), discussing the badu of Sinai, calls a “three party contra (...)
  • 13 Such arrangements are specific. If a killing is being paid for, then a separate set of guarantors w (...)

7Acting in this way entails responsibilities, and intervening to “cover” oneʼs kinsman or neighbour is voluntary. If kinsmen and neighbours refuse any part in the dispute, the culprit is on his own; if they take on the person involved, they also take responsibility for the debts and claims. Numerous distinctions are available in the language of ʿurf, but the texts at hand describe ḥujbah consistently in general terms which recur in a range of settings (e.g. text B sections 5, 13, 37, 40, 41). The man from section or family A who gives the rifle, and acts in effect as a guarantor, becomes the lazīm: his mulazzim is the initial victim in B, who can invoke the obligation (malzam) the lazīm has taken on and can claim from him ʿayb, or amends, if it is not discharged.12 Text B section 20 says payment for blood is to be made in three instalments, as it seems to have been also in, for instance, al‑ʿUṣaymāt until recently and apparently still is with certain Jawf tribes such as ʿAbīdah. It is done through a lazīm, thus displacing the original antagonism between creditor and debtor until full compensation has been paid.13

  • 14 Most of the offences that are reckoned ʿayb are listed in the early parts of text B: wronging someo (...)

8The same could apparently be done, which is less predictable, with amends for a disgraceful action (ʿayb or ʿatb) if the creditor agreed accepting these. Text B section 37 again specifies payment over three years, but says any claim for blood (nuqṭat dam) underlying the amends is a separate matter. Creditor and debtor are still not at peace, and al‑dhamm qabla l-dam, only once amends for the insult (ʿayb or dhamm) are paid can settlement begin for compensation (naqṣ or dam). What cannot honourably be done in a case of ʿayb is to prevent the claimant seeking vengeance if he so wishes from the man who insulted or disgraced him, for mā ʿalā al‑dhamm iʿtirāḍ, “there is no intervening [no protecting the culprit] in cases of shame or insult” (text B section 23).14 Nonetheless the qawāʿid set limits on action within the pact. The claimant cannot summon his whole group for a raid (ghazū) on the person who shamed or insulted him but can take only twenty men (text B section 21, cf. text A section 15). All this depends, of course, on the culpritʼs neighbours and relatives recognizing what he did was shameful and therefore standing aside. (If they do not, which means they consider their fellowʼs initial action was justified, then one is facing a major dispute between tribes or sections). In most circumstances, violence itself is formalized among the pactʼs members.

  • 15 It goes almost without saying that in the absence of any sovereign authority, events can be reinter (...)

9The claimant for an ʿayb, like a man seeking vengeance within the pact, is said to get only three “rifles”, three shots on each occasion, at the man who wronged him (text B sections 20, 21). If he exceeds this, amends are due. The interpretation of the same idea is nowadays liberal, and at present there is almost a routinized process: on the first day one fires over the house, on the second at the top of the house (but avoiding the kitchen area, or tajwāb, where women might be), and on the third day into the walls half-way down. After that one starts on the windows, on the manʼs car if he has parked it by the house, on whatever material objects are in reach, while the man himself is placed in effect under personal blockade. He should by then have decided to settle the claim, or one of his relatives or neighbours should yajib ʿalay-h, take responsibility and arrange payment.15 If no-one does, or if the claimant refuses cash-amends, then the dispute drags on until blood is taken, and perhaps beyond.

  • 16 To be at all confident we have the meaning here it is as well to compare carefully all the passages (...)

10If protection can be withdrawn by immediate neighbours in case of ʿayb, it can also be extended by others in cases of naqṣ or simple loss. Provisions for internal disputes among the pact’s signatories are complex, and the texts here are difficult to read. Text A (section 14) and text B (sections 19, 20) deal in closely parallel terms with compensation (naqīṣah) for wounds or killing. Both use verbs to do with refuge or covering (yaḥjib or yataḥajjab, cf. ḥujbah) and refer to obligations (malāzim), but each is unclear at first reading what the obligations are and the texts seem at first to contradict each other on whether a man has or does not have sinnah, the right to take and give protection. Each takes a great deal as understood.16 A recent document from Ṣubārah in Sufyān, dated Shawwāl 1384 (March 1965), provides a useful parallel:

فَمَن قتل على مَحْض النقاء فله أول دَرَك يُدَرِّك به وبقبائله ومَن حضر مِن صبُارة فله درك والدرك هو خمسة عشر يوماً كذلك الرُفقة تشل بين صُبارة وهو كذلك خمسة عشر يوماً هذا اذا لم قد به درك... وكذلك للقاتل سِنَّة والسِنَّة ثلاثة أشهر واليه يتنقّل لسِنَّته فوق لِحَام صبُارة

Whoever kills for an honourable reason (ʿalā maḥḍ al‑naqā’, to clear his name) has the [right to the] first intervention (darak) offered him and his tribesmen.
Whoever of
ubārah is present has the right to intervene, and that intervention is [good for] fifteen days ... Likewise the killer has right (sinnah, to protection), and the right is three months: he must move to [take] his right from the sections of ubārah.

  • 17 laysa la-h bayna ṣubārah ʿulqah tashtal sawā'a rafaq al‑mashall wa-huwa khamsat ʿashar yawman wa-il (...)

11The first intervention (darak), from anyone present of Ṣubārah, can be replaced by rufqah, a more general relation of care and protection, also good for fifteen days but announced at market. A man who commits ʿayb has no rights to protection, whether rufqah or darak, only to escort from Ṣubārahʼs territory.17 Killing for an honourable reason, however, leads to three monthsʼ more general right (sinnah), and the verb yatanaqqal (to move) suggests clearly that to claim this the killer must leave his home.

  • 18 Again one notes that solidarity is not automatic. In a case like this the manʼs own fellows within (...)

12Among the signatories of the present documents, and particularly among Dhū Muḥammad, there seems to have been a system of refuge or protection such that, if one man wounded or killed another in circumstances that involved no ʿayb, each fifth but his own would shelter the culprit for a month in turn, or a month and ten days by some accounts.18 This status of qiṭār or refuge was personal. While it lasted, those who had suffered the loss (ahl al‑manqūṣ) could not attack the culprit himself but could seek vengeance from his fellows. If they sought vengeance and, rather than balancing the account, lost another man, then the initial culpritʼs right of residence was forfeit. According to text B section 19, when the period of residence and shelter with the other fifths expired he retained right of escort but not of refuge.

  • 19 The provision applies only to compensation, not to amends, i.e. payment for ʿayb or shame. It shoul (...)
  • 20 The guarantors, as well as the person whose protection has been violated, are then due amends. How (...)

13Whether or not he moves to the shelter of another section, the rule is that, provided no ʿayb or disgrace has been committed, a man whose opponent refuses payment and seeks revenge has sinnah (right of protection, such as escort) with his fellows; on the other hand, he who refuses payment when those wronged within the pact have said they will accept it has no such right (text B sections 20 and 21, text A section 14). This general morality does not start or end at tribal boundaries, however, but applies to everyone in the region and beyond simply as “tribal custom” (aʿrāf al‑qabā'il). An outsider who owes compensation to someone in the pact thus has no place in their territory or at market (text B sections 7, 25). But the same outsider can take shelter (rabāʿ) with another of the pactʼs members so long as he pays or agrees to pay this second man what he owes (text B section 7): the second man thus “covers him” (yaḥjib ʿalay-h, takes responsibility), and in due course will see his own fellow in the pact is somehow recompensed.19 Neither tribes nor sections are solid, homogeneous blocs. Debtors and creditors, fellow-tribesmen and foreigners, are inter­mixed geographically and socially. What the pact does is claim a moral coherence among the signatories in that everyone agrees what his fellowsʼ rights are and how these should be recognized. A deliberate breach of those rights by one of the pact against another shames or insults all the pactʼs guarantors.20

  • 21 This is not say everyone knows why someone is present in their territory. In the course of a seriou (...)
  • 22 Glaser Maʼrib p. 134, al‑Sudumī Naẓarīyah pp. 103, 109-10, 136, 140. Ḥabshūsh (Ruʼyah p. 99) mentio (...)

14A man protected by others because living in their territory is referred to as qaṭīr (text B sections 6 and 7, cf. sections 19, 23, 30), whether taking refuge or simply living there for other reasons (in much of the area, as we mentioned in part two, people can graze within the borders of other tribes): the two meanings of qaṭīr are distinguished only, though clearly enough, by context.21 Most people at Baraṭ these days also decline to draw distinctions between qaṭīr and rabīʿ. Glaser, in the Jawf of the nineteenth century, while agreeing these terms overlapped, suggested the qaṭīr was someone “delegated” by his tribe, whose residence elsewhere expressed an existing relation, while the rabīʿ invoked a general right of hospitality; more recently, perhaps with Khawlān al‑Ṭiyāl in mind, al‑Sudumī explains the rabīʿ as someone seeking protection and the qaṭīr as someone who moves for economic reasons.22 Text B section 6 implies that at Baraṭ two centuries ago rabīʿ was a general term, encompassing both the sayyir (escort) and qaṭīr, but if so, this is no longer common usage, and in fact the word rabāʿ itself is little heard.

  • 23 The Yemeni qaīr is the North Arabian qaṣīr. What philological relation there is between the words, (...)

15The qaṭīr (or rabīʿ), we should also note, can be either the person protecting or the one protected. Usually the one protected is the one who moves, whether to escape threats of vengeance or simply to graze flocks, but the spatial relation can be reversed. Although not common, this remains intelligible practice. In the late 1970s, for example, people from al‑Maʿāṭirah found themselves at odds with Āl Aḥmad b. Kawl of Dhū Muḥammad, and a minor dispute over grazing sheep (in fact, over maḥjar of the kind described in part two) led to several killings. Al‑Maʿāṭirah invited in people from the other fifths of Dhū Muḥammad. So long as such a person lived as qaṭīr alongside the particular families of al‑Maʿāṭirah who feared attack, the latter were reckoned under his protection and that of his agnatic kin.23

  • 24 Turning back to the Dāris genealogy in part one, we see that eponyms like this are only five or six (...)
  • 25 No great flurry of rifles and meetings preceded this arrangement: it seemed simply common‑sense. No (...)

16In brief, the definitional sets expressed through a language of shared male descent do not correspond to the groups who accept responsibility in given case, or who are held responsible, and neither the sets nor the groups correspond to the pattern of people on the ground. Nor, as we explained earlier, do sections or tribes as wholes clash unavoidably in warfare. Let us take a contemporary example that probably corresponds to older practice. Al‑Qaḥūm are a small fraction (what text A section 9 of the qawāʿid refers to as a badanah) within Āl Dumaynah of Dhū Muḥammad, and number about 150 people. Their genealogy (or place in the local classification of fractions, at least) is Qaḥūm b. Aḥmad b. Ṣāliḥ.24 Not long ago someone from al‑Qaḥūm killed someone from al‑Maʿāṭirah, and at the time of writing there is blood between them: members of al‑Qaḥūm are thus in danger of vengeance from the fraction of al‑Maʿāṭirah involved. But their closest relatives within Āl Dumaynah, Āl Qāsim b. Ṣāliḥ (the fraction that includes the shaykhly family of Bayt Dāris), have no quarrel with al‑Maʿāṭirah and where need be can escort members of al‑Qaḥūm anywhere at Baraṭ, There is no question of al‑Maʿāṭirah versus all of Āl Dumaynah, let alone all of Dhū Muḥammad.25

  • 26 Small numbers of people may invoke the tribeʼs name in doing this for themselves, but others in the (...)

17In other circumstances (very much as with populist nationalism) a small-scale dispute can erupt without thought into large-scale conflict. To constrain the spread of conflict, meetings are called; so they are to widen a conflict and gain support from whole sections or tribes, or indeed from distant, unrelated allies. Such meetings are held at market on market-day or at meeting-places called mawṭāh or miṣrākh, Animals may be proffered and rifles pledged, as described earlier. Everything is cross-cut by specific relations, concretized in the system of guaranty, and irregularly members of tribes come together to guarantee their own moral solidarity as they do the safety of borders or the market,26 actualizing what is meant to be the case anyway. The preambles to texts A and B (section 1 of each) thus speak of “renewing” together the partiesʼ brotherhood, their guarantors, and the customs of fathers and forefathers.

  • 27 For simple brotherhood formulae see Dresch Tribes pp. 85,330-1, ʻSufyānʼ p. 52, al‑Sudumī Naẓarīyah (...)
  • 28 Cf. Dresch ʻSufyānʼ p. 57 for a similar provision in case of simple blood-debts and thus of revenge (...)

18The qawāʿid al‑malāzim, as texts A and B are usually called, are also a qāʿidat ṣaḥb or qāʿidat ikhāʼ, a pact of brotherhood. The usual brief formulae of such documents (in effect, that the signatories undertake to be a single unit in vengeance)27 are spread out through texts A and B, but concentrated particularly in text A, which deals solely with Dhū Muḥammad. If someone rejects (yashidd) his fellow, for example, and refuses the claim for contributions to collective debts or costs, then not only must his section pay in his place but he himself loses his fellowship (ṣaḥb) in the tribe and thus his right of protection (sinnah) until he pays his share (text A section 9). As we saw earlier, certain elements such as the market, borders, women and protected people, are marked as collective matters: defence of these, in significant contrast to the usual rule, means suspending internal blood-debts “even if [a man] has disgraced/insulted his [own] tribesmen” (text B section 11, cf. text A section 15).28 Specific commitments can ideally be of equal weight (text A section 17):

Whoever fails his fellow (man fāwata āib-ah) in their obligations (malāzim), or in a payment (ghurm) that gathers them as some obligation gathers Dhū Muammad, ... no longer [himself] has any obligation he can call on or to which he is subject (lā baqā [sic] la-h malzam wa-lā ʿalay-h malzam).

19The reference in the first line here is to undertakings that are concluded by individual tribesmen but recognized as just by all.

  • 29 For the equivalence of such relations see text B section 13 (text A section 13). How far the obliga (...)
  • 30 Apart from the problem of foreigners having no ultimate guarantor to back them up, one should not t (...)
  • 31 Cf. Ibn al‑Mujāwir Tārīkh al‑mustabṣir pp. 253-4. The details he gives are uncertain, and not quite (...)

20A malzam (pl. malāzim) is an “obligation”. Some are permanent, such as those to immediate kin or dependants in the same household. Others are voluntary “links” (ʿulaq), such as escort or guaranty or that deriving from commensality.29 The law of “the bread and the salt” has entered English literature, and doubtless French, and acquired a dubiously “oriental” air, but among tribes in eastern Yemen as late as the mid-1980s it was real, and even the foreign visitor was taught the rules by concerned acquaintances in Arḥab, Sufyān, and Dhū Ḥusayn before being allowed to wander off by himself in the Lower Jawf.30 Text B (sections 13, 25) mentions the convention briefly; text A section 17 spells out the details –so much time for a cup of coffee, so much for eating dates, so much for a meal of bread and meat.31 Any offence during that period between people who shared food was actionable in the same way as a breach of formal guaranty.

  • 32 In al‑ʿUṣaymāt, far to the west of Bara, a little west of Ḥūth, the word malzam can mean relatives (...)
  • 33 Joint payment in effect defines tribal unity, to the point where the men of a tribe may be called s (...)

21The ability to take on such obligations would distinguish the tribesmen of Dhū Muḥammad and their fellows from, for instance, the manūʿ, who live at Baraṭ but cannot give escort or protection. It defines tribal status. But recognition of “rights” and “links” as a common concern also defines one set of tribesmen against another, as in text B section 23 where the signatories to the pact decide not to support or take on obligations made by Dhū Ḥusayn.32 Commitments may be limited or divided out even within a pact, as they are between Dhū Muḥammad and al‑Maʿāṭirah in text B section 36. What makes a pact like this, however, is the shared recognition of which links and obligations are actionable by whom, which expenses (ghurm or gharāmah) are shared,33 and which commitments, as for instance to protect the market, are held in common: a particular “obligation” is the concern of all in so far as its breach now compromises shared reputation. In fact malzam overlaps here with ʿulqah and more generally with sinnah, the word we have translated everywhere as “right”. Text A section 14 thus gives maḍat malāzim (pl. of malzam) for a set of rights or obligations running out, and text B section 19 in exactly the same sentence gives maḍat sinan (pl. of sinnah). The terms sinnah, ʿulqah, and malzam appear usually as denoting three aspects of the same thing.

  • 34 Al‑ʿAlīmī: Qaḍāʼ qabalī p. 118, cf. al‑Sudumī Naẓarīyah p. 127. Aside from our qawāʿid the only cle (...)

22This brings us to a key phrase which recurs throughout text B, ahl al‑malāzim. Given there are three major parties to text B and that no ancestral name defines them neatly against others (part one, above), a general phrase of some kind is needed, and as al‑ʿAlīmī suggests, it clearly means the tribes or persons “committed” (multazimīn) to the pactʼs provisions.34 The singular presumably would be malzam, an “obligation” or “undertaking”, perhaps even a section or clause (standard Arabic malzamah) in which such an undertaking was agreed. A reasonable gloss of ahl al‑malāzim would thus be “the confederates” or “adherents to the pact”, and throughout text B it is glossed here as simply “the allied people”. In a more limited sense, tribesmen from Baraṭ say, it means the documentsʼ signatories, the guarantors. In most passages it plainly includes those for whom they sign, and some malāzim are of their nature joint. But in many places (e.g. text B, sections 10, 20, 21, 23, 24) malzam is an individual status which can be recognized or suspended by these signatories and the people for whom they give guaranty, a concrete instance of sinnah or general “right”.

  • 35 One does not want to be drawn into false oppositions between e.g. status and contract. Suffice it t (...)
  • 36 The balance of power I had thought of as a peculiarly British view of the world, drawn in part from (...)
  • 37 An obvious indication of this is the way that agreements concerning one tribe or one set of section (...)

23The qawāʿid al‑malāzim (texts A and B) set out what types of obligation are recognized and upheld by whom. They do not imply unconditional solidarity. Tribal identity is constitutive, if that word be allowed, and says who is who; it is not predictive in a simple way of who next does what, or of who confronts whom on which occasion. The assumptions apparent in the documents are consistent with what one sees and hears now, or has done in recent decades. Personal obligations are formed or invoked, and upheld by others, in pursuit supposedly of what is just; and a call to arms (nakfah), of the kind described at the start of this chapter, is answered or ignored on the same grounds. He who suffers a loss (al‑manqūṣ) is spoken of as oppressed (maẓlūm), which all should recognize. Restitution or absolution, in money or blood, restores in some measure the victim's standing; but arranging restitution, or its promise, as we have seen, involves a vast range of possible interventions by neighbours, kin, or virtual strangers.35 One has to get away from ideas of solidary groups, of inherent propensities to violence, and thus of a balance of power,36 for the balance at issue is a moral balance expressed through tribal identities, each of which has its reputation (sumʿah) as against other tribes and sections.37

  • 38 Whole sections and tribes can change identity by brotherhood (see Dresch Tribes pp. 329-32); indivi (...)
  • 39 This is not so as a formal provision of ʿurf, but the tribe to which the claimant goes will accept (...)

24Text B section 38 thus prescribes what to do if a member of one of Dhū Muḥammad's fifths or of al‑Maʿāṭirah (in other words a member of the pact) fails to get satisfaction from his fellows and goes to some other tribe to claim brotherhood (ṣaḥb, ukhūwah or ikhāʼ).38 His fellows, the members of the pact, have to go after him within fifteen days. They give rifles as surety to the people who have taken him in, and if his claim is found to be justified they must pay him his due to retrieve him: judgement in effect rests with those who take in the claimant.39 The same rules apply, symmetrically, if a person from another tribe claims shelter with members of the pact, and “these are weighty laws” (shurūʿ ghaṣṣābah). If someone from Dhū Muḥammad or al‑Maʿāṭirah goes to another “fifth”, however, claiming oppression by his specific fellows and seeking justice through shelter with others of the pact, two shaykhs are appointed, one from each fifth, to judge the matter. If those he left are judged at fault and now fail to give him recompense, their guarantor, who signed the pact for them, is “vilified” (shutama) as one would be for reneging on a link of escort.

  • 40 For a fuller account see Dresch ʻPlacing the blameʼ, the title of which plays off the term malām, b (...)
  • 41 I am indebted to Geert van Gelder for suggesting, at last, a derivation of jidhn. Al‑Zabīdī’s Tāj a (...)

25The details vary from tribe to tribe, and I have not heard of anything so elaborate south or west of Sufyān, but the general idea remains widespread in the further parts of northern and eastern Yemen. The process is known as bayḍā' wa­-khiyār, roughly “whiteness and choice”.40 If one has a claim against a guarantor, an escort, a protector of any kind who defaults on a commitment (see text B sections 12, 15, 23, 38, 41), then, to take the most concrete form of the practice, one announces the claim publicly and puts up a white flag and black flag (bayḍā' and sawdā'). If the person summonsed does not discharge his obligation, or give rifles that guarantee doing so, then the white flag comes down and the black one is left in place. The final period of “choice” (khiyār) begins. Dhū Muammad, unlike for instance Āl Sālim or al‑ʿAmālisah, seem rarely to place actual flags now and simply make announcements at market by way of the dawshān, though the same terminology is used: bayḍā’, sawdā’, khiyār. The first phase is usually reckoned at eight days, and the second at eight days, giving a total of two weeks (“fifteen days” as they are counted locally). If the defaulter has still not responded (istaqāma, irtadda) then one places a jidhn or malām, known also as shutūmah, from shatama to “vilify”. The jidhn, at simplest, is a blackened stick which stands for the defaulter.41 It is reckoned an extreme insult, and deprives the people vilified of all rights to escort, refuge, or access to the claimant’s market. It therefore matters where the jidhn is planted and by whom.

26In the 1983 note from Sufyān to Banī Ṣuraym that we quoted earlier we find a distinction between two different cases within the tribe.

اذا تخلى عنه رفيقه... فقد صقط عرضه... وله شاسة سوداء في السوق وشتم في بادية. جواب ثاني أما الكفيل المدعى بما كفل به يدعى في السوق ولم يقوم على وجهه بالوفاء فله سوداء وبيضاء في السوق وله خيار ومهله خمسة عشر نهار يبيض وجهه بالوفاء والالزم له الجذن في السوق

  • 42 The passage goes on in terms we shall encounter later: wa-yastarafaq ʿalā l-jidhn lā yunṭa attā y (...)

27The escort or companion (rafīq) who fails in his duty is due a black flag in the market and, according to Ṣubārah’s custom, is vilified in the surrounding countryside (shutama fī l-bādīyah, in other words a jidhn is placed there, presumably at the claimant’s village); on the other hand, the bondsman or guarantor (kafīl) who fails to act after being formally summonsed has fifteen days’ “choice” and then the jidhn, like the black and white flags, is placed at the market in al‑Ḥarf itself.42 In our documents from Baraṭ, supposedly two hundred years earlier, both the flags and the jidhn are placed in the market if the claim is against outsiders (text 1 section 11, text B section 23): so can the flags be if a claim is made within the pact. It is forbidden, however, for one of the members of this pact ever to place a jidhn against another member in their own market, on pain of an eleven-fold fine (text A section 12, text B section 12).

  • 43 For a parallel, Jaussen Pays de Moab p. 153, where within a ṣuḥbah livestock must simply be returne (...)

28Around this concern for the protected market and for common borders forms a shared agreement that certain actions are inadmissible. For example, if one of those governed by the pact steals livestock from another, he must pay four times the value of the animal (text B section 17): if he fails to do so, then his rights as a tribesman are suspended.43 If he cuts the produce from a fruiting palm-tree, he must pay four times the damage plus a pair of sheep (text B section 31). Certain rules about land-use, not least the seasonal status of maḥājir and rights concerning wells (text B sections 30, 43), were touched on in part two. Assignment of responsibility in killing is mentioned (text B sections 27, 33, 34, 35), and much space accorded to arbitration (text B sections 15, 39, 40, 42). Substantive law is, in a sense, outweighed by procedural.

  • 44 For a brief discussion of traditional legal history on this score, Miller Bloodtaking pp. 248-9, 36 (...)
  • 45 Cf. Millerʼs comments on Iceland (Bloodtaking pp. 21, 227-9), where the possession of law, without (...)

29There is scant sign in the texts, any more than in recent ethnography, of formalism: plaints, for example, do not depend on a set form of words.44 But the bulk of the pact concerns, in effect, when a man has and does not have right, or sinnah, and no distinction can usefully be drawn between law and self-help, or even self-help and litigation. The “custom” invoked or specified in these texts is procedural in the sense that it allows or does not allow those at odds with others to claim help from their fellows. One may “have no fellow” (lā la-h ṣāḥib), or be backed by the tribe, or more usually and more importantly, be allowed one's sinnah as the process of dispute and settlement unfolds. This is nonetheless law. Indeed, in discussions of ʿurf one is often reminded of the Norse proverb quoted famously by Njál: “With laws shall our land be built, but with disorder [ólog, unlaw] laid waste”. It might be that, in some measure, possessing such a pact as text B gave a feeling of identity that reached beyond the nested obligations of shared descent and residence.45

  • 46 Steffen et al. Final Report p. II/208, total figures for Nāḥiyat al‑ʿInān and Nāḥiyat Rajūzah. One (...)
  • 47 Of the 62 delegates nominated, 55 had taken an active part; at the end there were 39 signatories to (...)

30The common assertion by Baraṭis that this is a “constitution” is well founded. The agreement presents itself as between free but related persons, signed for by guarantors who are very numerous. The total population (male and female, of whatever age, tribal and non-tribal families alike) in the whole area was of the order of 15,000 souls in the late 1970s.46 In the 1770s it can scarcely have been half that, and the relevant part, the tribal families of Dhū Muḥammad, may have been far less. The relation between Dhū Muḥammad in Upper Yemen and in Lower Yemen at the time is unclear. Nor is an exact count of names possible (a mere stroke of the pen or a line on a photocopy can divide a name in two by accident or run two names into one), but there are something of the order of 150 signatories named on text B. The ratio of representation, if we can coin a phrase, is therefore extremely high, about one signatory for every ten adult men. (Very roughly speaking, the famous Constitutional Convention at Philadelphia in 1787, at around the time of our Baraṭ texts, would have needed not 55 delegates but approximately 10,000).47 The sheer number of independent parties to a pact like text B suggests the difficulty of ever enforcing it as a set of rules; it also suggests a general acceptance of the categories and definitions which the pact supposedly “renews”.

  • 48 Haykel Revival and Reform pp. 64-5, Dresch Tribes p. 233 n. 15, al‑ʿAmrī Miʼat ʿām min tārīkh al‑ya (...)

31Importantly, these texts, and those that al‑Akwaʿ gives, provide a glimpse of moral structures at Baraṭ, which chronicles and learned biographies ignore or suggest can hardly have existed. The nature of that literature was touched on in parts one and two. We might simply remind ourselves of how conspicuous Baraṭ was at the level of events. If the dates of texts A and B are correct (that is, AD 1763 and AD 1796), then they straddle the al‑ʿAnsīsʼ call from Baraṭ for rejection of Qāsimī doctrine (1768), a massive revolt by the Baraṭ tribes against the Imam in Ṣanʿāʼ (1771), and a rising by Ḥusayn b. Ḥasan al‑ʿAnsī in 1793, through all of which one finds little in the town-centred chronicles but denunciation of the “disbeliever tribes” as in effect savages.48 Plainly there was more to the land of the tribes (bilād al‑qabā'il) at the period than this. Yet text B saw print in Arabic twenty years ago and remains uncited, and al‑Akwaʿ ten years later on saw no reason to correlate it with his own texts concerning Baraṭ, As so often, the published editions of standard histories obscure our view of who Yemenis beyond learned circles may have thought they were.

  • 49 Sayyid Muṣṭāfā Sālimʼs Wathā'iq yamanīyah (1982) marked a quiet revolution in Yemeni studies. For p (...)

32The habit that Yemeni authors have formed of including In their work photocopy documents is a huge step forward,49 but there is surely a great deal out there still to be sorted, assembled, and understood. There may well be other codes like qawāʿid al‑malāzim.

33There are certainly many market-guaranties and pacts of brotherhood; dispute documents, like inheritance papers, are numerous; clues to ʿurf at different times and places are often apparent in documentary fragments, and these are often thrown away as the house-cleaning attendant on a change of generation now unfolds in changed conditions. Yemen is simply littered with writing. The “land of the tribes” is no exception. But while epigraphists concerned with the pre-Islamic age correlate their smallest finds to build up a picture of a distant society, analysts of Yemen's long Islamic history seem content to leave for termites the paperwork of everyday life beyond a few major cities. If the present effort, with all its faults, suggests only that such material is worth exploring, then it serves a purpose.

Notes

1 The only recent case on which I have much detail concerned a dispute between fifths of Dhū Muḥammad. Āl Amad b. Kawl lost a woman killed and blamed Dhū Zayd, who said she was killed by a stray shot from Āl Amad b. Kawl themselves. An arbitrator found against Āl Amad b. Kawl and judged the shooting deliberate. Though the judgement has still not been implemented at the time of writing (and may perhaps not be) it was certainly bi-l­-muḥaddash, meaning 11 times the womanʼs blood‑price, which is half that of a man.

2 Circumcision used to be done between the ages of about seven and 14. It was also done this late in many of Yemenʼs towns. But accusations that circumcision was never done recur in for instance Saudi denunciations, pre-World War Two, of tribal “ignorance” in al‑Ḥijāz: Dickson, quoted Serjeant ʻSunnah, Qurʼānʼ p. 46.

3 Obviously this is a difficult subject to pin down (a gentleman does not enquire about another manʼs genitals), but I am sure I knew people 25 years ago who were married yet not circumcised. I suspect that the system has evaporated –if it has evaporated– only recently. In the case of which I am surest the man had been shot in the leg when a young boy.

4 Text A section 4 says also that amends are due, as well as compensation, if one uncircumcised man wounds another. But, unlike the case where a full tribesman offends against such a person, no oath of intent is specified, a point Muḥsin al‑Ghuraybī promptly explained as due to the fact the antagonists are “equal”.

5 Cf. al‑Sudumī Naẓarīyah pp. 119, 120, 122, Ḥabshūsh Ruʼyah p. 55. Even when blood-money is paid in an ordinary killing the debt of blood in a sense remains. Guaranty for the individual parties at odds may be made permanent, for after money is taken, and the public relation between sections or tribes restored, the wish for revenge might itself endure.
This theme recurs elsewhere. See e.g. Bourdieu ʻSentiment of honourʼ pp. 205, 216. Even in the very different world of southern Sudan, where payments are not only complex formally but embedded in elaborate connections of cattle, blood and life, there remains the ideal of eventual vengeance: S. Hutchinson
Nuer Dilemmas: coping with money, war, and the state (1996) p. 62.

6 We mentioned in part two, above, provisions for escorting people to “the land of the state” when their other relations expired or were forfeit (text A sections 15, 22, text B section 21). At Bara the short-term arrangement that allows this is called rafaq al‑mashadd.

7 What was interesting about this case was its specificity. How realistic the sonʼs fears were I cannot say, though I remember from many occasions the shutters kept closed at night in the heat of summer; the father, who was my special friend, did not feel at risk and wandered around the area in daylight without a rifle, which in those days was unusual. Similar specificity is found with less personal wrongs than adultery once rifles or slaughter-beasts have been given which make clear the attitude of the culpritʼs kin.

8 Obviously if I know where he is then so do all Dhū Muḥammad, including his antagonists. But leaving tribal territory or staying in oneʼs house both illustrate a principle that is badly understated in most accounts of Middle Eastern tribalism: “out of sight, out of mind”. A standard form of truce at Bara is thus mā fī-h lā ʿarrah wa-lā jarrah, the main condition of which is that each party simply stays off the otherʼs land and away from his house or houses.

9 There is a whole series of Vth form verbs like this that usually take the preposition from”: e.g. yataraffaq or yatasayyar min-nī, he is escorted by me, under my protection. Some, confusingly, seem also to serve as transitive verbs.

10 al‑Sudumī Naẓarīyah pp. 150-3. The account he gives of guaranty is accurate and useful, overlapping with that of “brotherhood” as obliging men of a section or tribe ideally to act for each other in this way (ibid. 130-1). “Representative”, in the line before, is my own brief gloss: it is not how tribesmen speak of the relation.

11 For the verb istaqāma see al‑Sayāghī (ed.) Ṣafaḥāt p. 71, where someone at anʿāʼ in the 1850s stands forward to pay blood-money on behalf of his fellows and thus contain a dispute. In fact it can mean taking on or accepting any obligation. In text B section 23, concerned with enforcing obligations on men from other tribes, it corresponds to “doing the right thing”, tabayyaḍa, in text A section 18.

12 This is what Stewart (ʻContractʼ p. 179), discussing the badu of Sinai, calls a “three party contract”. His account is complex, not least because of close comparison with European legal thought, but the essential points apply equally to Yemen.

13 Such arrangements are specific. If a killing is being paid for, then a separate set of guarantors would be appointed to maintain the peace after payment was completed. If a truce formed part of settlement, then again separate guarantors would be chosen. The man who “covers” his fellow (yaḥjib ʿalay-h) is responsible only for arranging the agreed payments. Someone who arranges payment in this way as a lazīm, with the agreement of debtor and creditor, is formally entitled to part of the payment, by many reckonings a third.

14 Most of the offences that are reckoned ʿayb are listed in the early parts of text B: wronging someone escorted or granted refuge; violence at a meeting; and breach of guaranty, for instance. Note that if payment of amends for ʿayb is arranged through a lazīm, then the lazīm is inviolable, as he would be in payment of compensation; but the culprit, for whom he pays, is not. He would not, for instance, be due anything if shot at (cf. text B section 14).

15 It goes almost without saying that in the absence of any sovereign authority, events can be reinterpreted as they unfold. This is what happened in a claim for tashwīf, many years ago now, when someone firing at another’s house ignited a pile of brushwood (Dresch Tribes p. 56). It turns out that the neighbours of the man who owned the wood had agreed he was liable for amends, and his antagonist was proceeding in the way just described. When the wood ignited, the neighbours changed their minds.

16 To be at all confident we have the meaning here it is as well to compare carefully all the passages where sinnah is recognized or suspended after blood is shed. The provisions in text A section 14 are spread across text B sections 19, 20, 21.
It should be noted clearly that, despite the currency of such terms elsewhere, friends from Bara
have not been comfortable with reading taajjaba or yataajjab in the present documents. In text A section 14 and text B section 19 many have felt more comfortable reading najaat (to end or run out) than taajjaba at the start of the section or have insisted taajjaba itself must somehow mean to run out. Though this works for isolated lines, it does not work for the passages as wholes.

17 laysa la-h bayna ṣubārah ʿulqah tashtal sawā'a rafaq al‑mashall wa-huwa khamsat ʿashar yawman wa-illā fa-lā la-h akh wa-lā ṣāḥib wa-lā ʿulqah wa-lā wazā wa-man azā'a-h fa-kān muʿayyib al‑jamīʿ min ṣubārah. The wording is almost identical to that of the much older Bara documents. Both at Bara and in Sufyān people have said rafaq al‑mashall here is the same as rafaq al‑mashadd, but I have not heard the first phrase used spontaneously and it may be a scribeʼs error.

18 Again one notes that solidarity is not automatic. In a case like this the manʼs own fellows within the fifth have not washed their hands of him, but nor have they agreed to support him actively and thus provoke a fight with the whole of his antagonistʼs fifth. If the initial dispute involves men from two different fifths, then there are three fifths left to offer the man refuge in turn, giving a total of three or four monthsʼ cover. During that time a truce or some process of arbitration might presumably start and the problem thus be contained by other means.

19 The provision applies only to compensation, not to amends, i.e. payment for ʿayb or shame. It should also be noted that where the text says pay (more literally, as we shall see, “cough up”) what is meant is agree to pay. Even nowadays, when wealth in cash is far more available that it would have been previously, debts are often managed gradually, though they are not, I think, deliberately drawn out beyond three years to maintain guaranty.

20 The guarantors, as well as the person whose protection has been violated, are then due amends. How these might have been divided, if they were paid at all, is unclear. Perhaps more important is that wronging the guarantors would clearly have lost the man his sinnah.

21 This is not say everyone knows why someone is present in their territory. In the course of a serious dispute between Dhū Muḥammad and Dhū usayn I met a Muḥammadī living in Ḥusaynī territory; in a later dispute with Wāʼilah I encountered a Wāʼilī living with Dhū Muḥammad. People simply accepted that each was somebodyʼs qaīr. The first seemed, as it were, an economic migrant; the second had moved to his in-laws because of trouble at home.

22 Glaser Maʼrib p. 134, al‑Sudumī Naẓarīyah pp. 103, 109-10, 136, 140. Ḥabshūsh (Ruʼyah p. 99) mentions rubʿān (pl. of rabīʿ) as meaning share-croppers, perhaps of less than full tribal status; the same meaning attaches to quṭarāʼ (pl. of qaṭīr) in al‑ʿUṣaymāt, where a refugee or someone moving for economic reasons would usually be referred to as labīth.

23 The Yemeni qaīr is the North Arabian qaṣīr. What philological relation there is between the words, if any, is unclear. But for a close North Arabian parallel to the incomer protecting his hosts near Bara see Jaussen Pays de Moab p. 218.

24 Turning back to the Dāris genealogy in part one, we see that eponyms like this are only five or six generations removed from the present. People will often fill in the gaps quite confidently, and consistently, with personal names. For detail at this level within Āl Dumaynah see al‑Ḥajrī Buldān p. 111.

25 No great flurry of rifles and meetings preceded this arrangement: it seemed simply common‑sense. Nor does the involvement of an important family of shaykhs make any difference. This kind of pattern is actually far more common than large-scale confrontations between whole sections or tribes even in the event of murder.

26 Small numbers of people may invoke the tribeʼs name in doing this for themselves, but others in the tribe may reject the provisions or simply ignore them. If one turns back to the agreement on borders given as Document 4 in part two, above, one sees that not only was one fifth of Dhū Muḥammad missing but the signatories for the others are rather few. Quite possibly the practical effect of their agreement was unimpressive.

27 For simple brotherhood formulae see Dresch Tribes pp. 85,330-1, ʻSufyānʼ p. 52, al‑Sudumī Naẓarīyah pp. 95-6. Where no redefinition of tribal membership is involved, usually pacts of this kind seem to follow on a crisis, but what was happening at Bara at the time when texts A and B were first signed is a mystery.

28 Cf. Dresch ʻSufyānʼ p. 57 for a similar provision in case of simple blood-debts and thus of revenge or compensation (naqṣ). But so far as I am aware, this is the only exception in the Bara scheme to the rule that someone owing ʿayb loses his general right or sinnah.

29 For the equivalence of such relations see text B section 13 (text A section 13). How far the obligation extended is unclear, as too is the reason commensality receives special mention at a number of points, e.g. text A section 8, text B section 25.

30 Apart from the problem of foreigners having no ultimate guarantor to back them up, one should not take too romantic a view of this: Ḥabshūshʼs experiences with e.g. escorts in the 19th century Jawf were not uniformly happy. But other researchers twenty years ago will remember how easy it was, provided one were plainly harmless, to go anywhere one wished in tribal territory and feel entirely safe. With the spread of state power, free movement has become far more difficult.

31 Cf. Ibn al‑Mujāwir Tārīkh al‑mustabṣir pp. 253-4. The details he gives are uncertain, and not quite identical with those here, save that water gave one dayʼs link as coffee does in text A. Nonetheless the resemblance over so large a span of time is striking.

32 In al‑ʿUṣaymāt, far to the west of Bara, a little west of Ḥūth, the word malzam can mean relatives by marriage, as it can further south in Khārif, which conceivably it could do here. Though the usage is not known at Bara these days, and I consider this most unlikely, it is not impossible that even in text B sections 19-21 we have two words that look the same: malāzim meaning “relatives/fellow tribesmen” who give the man support and protection (section 19), and malāzim meaning “rights of protection” that have run out (section 21).

33 Joint payment in effect defines tribal unity, to the point where the men of a tribe may be called simply al‑gharrāmah (those who pay ghurm): al‑Sudumī Naẓarīyah pp. 50, 73, 75, 130 and passim. In text B Dhū Muḥammad and al‑Maʿāirah share responsibility for e.g. the market; but it is not clear whether ghurm more broadly is shared as against outsiders.

34 Al‑ʿAlīmī: Qaḍāʼ qabalī p. 118, cf. al‑Sudumī Naẓarīyah p. 127. Aside from our qawāʿid the only clear reference I have to a similar phrase is in Ṣafaḥāt majḥūlah (al‑Sayāghī ed. 1978) p. 68 n. 171, where the editor mentions ahl al‑malzam as supporters of a leader who takes responsibility for their costs and losses. In passing one should also note an association of these words in early sources with ideas of fighting or confrontation, e.g. Nubdhah mushīrah pp. 106, 120.

35 One does not want to be drawn into false oppositions between e.g. status and contract. Suffice it to say that right action (that which “whitens” a man) is external to the individual and can be sought, accepted, or rejected. The bias of the material at hand is obvious, but it is worth browsing the glossary at the end of the present work and noting how many terms relate to taking on, discharging, or rejecting claims.

36 The balance of power I had thought of as a peculiarly British view of the world, drawn in part from Hobbes and Harrington, in part from a position at the edge of Europe. But discussion with French, German, and American colleagues keeps reverting to this same idea of tribes as solidary masses of people who would press on each other as wholes. Nothing could be more misleading.

37 An obvious indication of this is the way that agreements concerning one tribe or one set of sections are often formally witnessed by signatories from elsewhere. For examples see al‑Akwaʿ Madkhal pp. 33, 43-4.

38 Whole sections and tribes can change identity by brotherhood (see Dresch Tribes pp. 329-32); individual families can do so as naqāʼil without leaving memories of profound dispute. Seeking brotherhood because one is oppressed or abandoned by oneʼs own tribe is nonetheless a powerful statement.

39 This is not so as a formal provision of ʿurf, but the tribe to which the claimant goes will accept him as long as they see his claims as justified, and the tribe he left are thus disgraced until the other tribeʼs acceptance is withdrawn. (The story about Hādī Ḥasan Abū Asbaʿ of Dhū Muḥammad taking refuge with Dhū usayn that we gave earlier is an illustration). Conversely a tribe who accept an outsider may expel him if convinced he has acted shamefully. Banī Jubar of Khawlān thus expelled al‑Jarādī, who had gone to them for refuge, when convinced by Bin Miʿṣār of Nihm that al‑Jarādī had wronged a Jew: Ḥabshūsh Ru'yah pp. 37-9, Yémen pp. 69-71.

40 For a fuller account see Dresch ʻPlacing the blameʼ, the title of which plays off the term malām, below. Also Glaser Ma'rib pp. 76-7, 103, 180-1, where the form lā'imah is given for the final part of the process. When I first encountered the procedure it seemed something of an ethnographic rarity, but I am struck by how references to bayḍā' wa-khiyār run through all three of the present documents. Whether common or rare empirically, it seems to have been conceptually important.

41 I am indebted to Geert van Gelder for suggesting, at last, a derivation of jidhn. Al‑Zabīdī’s Tāj al‑ʿarūs lists jidhn as a variant of jidhl, a tree-trunk or wooden post, not least one set up for mangey camels to scratch against (ʿūd yunṣabu li-l-ibil al‑jarbāʿ). As to the truth, allāhu aʿlam.

42 The passage goes on in terms we shall encounter later: wa-yastarafaq ʿalā l-jidhn lā yunṭa attā yashhar wa-yaẓhar wa-man la-h aqq lā yasqaṭ aqq-ah. The jidhn can only be “struck down” when the culprit is seen explicitly to accept his obligation.

43 For a parallel, Jaussen Pays de Moab p. 153, where within a ṣuḥbah livestock must simply be returned or paid for, but within a bin ʿammah theft is compensated four-fold. Nonetheless, the provision in text B that failure to pay means ones loses ṣaḥb is striking.

44 For a brief discussion of traditional legal history on this score, Miller Bloodtaking pp. 248-9, 363 n. 40. Even the great Maitland was inclined to think rigid formulae distinctive of early law, though there is little evidence for it in the early English record: see P. Wormald, Legal Culture in the Early Medieval West (1999) pp. 290, 294, 309 and passim. The assumption was that empty ritual would be later displaced by a care for “objective” evidence. As Wormald suggests, and as one might pursue through much recent literature on early Europe, this rested particularly on a profound misunderstanding of collective oath.

45 Cf. Millerʼs comments on Iceland (Bloodtaking pp. 21, 227-9), where the possession of law, without anything resembling kingship, did much to define what Iceland was. For the quote from Njál, above, ibid. p. 229.

46 Steffen et al. Final Report p. II/208, total figures for Nāḥiyat al‑ʿInān and Nāḥiyat Rajūzah. One could substitute Nāḥiyat al‑Marāshī for the latter and get about the same total figure: ibid. p. II/194.

47 Of the 62 delegates nominated, 55 had taken an active part; at the end there were 39 signatories to the famous draft, and the best part of three yearsʼ work was needed to gain ratification from all the states. Much of the effort required to produce the US constitution comes before and after the conference: the popular assemblies that ratified it at state-level give us a far higher ratio of representation, and the agreement reached gave the federal government almost as limited a remit as that of al‑ʿInānʼs market guarantors.
Nonetheless the very idea of a delegate or deputy that allowed the process was more inclusive and binding than that in the tribal system. The signatories of text B are merely guarantors
(ḍumanāʼ) of and for their men. To mimic Philadelphia in the Yemeni system one would need to assemble a vast and unmanageable meeting, and constantly reconvene it.

48 Haykel Revival and Reform pp. 64-5, Dresch Tribes p. 233 n. 15, al‑ʿAmrī Miʼat ʿām min tārīkh al‑yaman pp. 100, 107-8, 113, 115, 159 and passim, Zabārah Nayl al‑waṭar part one pp. 262, 299, part two p. 294.

49 Sayyid Muṣṭāfā Sālimʼs Wathā'iq yamanīyah (1982) marked a quiet revolution in Yemeni studies. For photocopy material relevant to the tribes see e.g. Abū Ghānim Bunyah qabalīyah, al‑Qabīlah wa‑l‑dawlah, al‑YāziMin al‑thawrah al‑bikr. The new convention of giving photocopy texts extends to e.g. al‑Akwaʿ Hijar al‑ʿilm wa-maʿāqil-hu (5 vols., 1995).

© 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