Version classiqueVersion mobile

The rules of Barat. Tribal documents from Yemen

 | 
Paul Dresch

End matter

Contrasts and Principles

Texte intégral

  • 1 See e.g. Fā’iz Mūsā al-Ḥarbī al-Tanẓīmāt al-qānūnīyah al-qaḍā’īyah laday qabā’il al-ḥijāz (2 vols., (...)
  • 2 The three traditions from northwest Europe have little to do with one another, and none, I assume, (...)

1Those of us intrigued by Yemen become obsessed with purely Yemeni sources, and reading even of North Arabian badu seems a poor distraction. There is a great deal to be done on existing material from South Arabia (the documents examined here are merely provocation); there is much to do also with evidence from al-Ḥijāz.1 It may, however, be worth standing further back. Shared history is not the only prism through which to view customary law, and neither is generalized comparison of a simple kind. Although partly the product of translation, where English falls without thought into set rhythms, the resemblances between Yemeni ʿurf, or “custom”, and for instance early Irish, Icelandic, or Anglo-Saxon law-codes are sometimes striking, and one wonders if we are flirting with elementary forms of self‑definition.2 Let us start by clearing away some obvious differences, then explore what if anything underpins the feeling of familiarity.

Rank and order

  • 3 A convenient presentation of the main texts is D. Whitelock (ed.) English Historical Documents c. 5 (...)
  • 4 The wergeld is compensation for a man’s life, as in murder. The mundbryce is the value of his peace (...)

2Going back to the Anglo-Saxon law-texts one is struck first of all by the interest they show in rank.3 The earliest Kentish laws, those of Aethelbert (c. AD 600), thus distinguish two or three grades of freemen, and three half-free; the West Saxon law of Ine (c. AD 690) is broadly similar, and it is hard to avoid the impression that men in the Saxon world were considered naturally unequal. Nor was southern England alone in this. The so-called “Law of the North People”, a good deal later, lists ten different grades of wergeld (blood-money), from kings and archbishops down to Welshmen who hold no land. The Celtic ideals of early Ireland are in many respects different –so far as kings and their relation to moral unity go, Ireland obeys a separate logic– but the honour‑price or face‑price (lóg n‑enech) varies with rank, and such abilities as that to form contracts or that to swear oaths vary with the honour‑price as they do with Anglo-Saxon wergeld and perhaps mundbryce.4 The enthusiasm of later clerks and commentators for elaboration in the Irish case only strengthens one’s impression that legal inequality was a feature of common sense.

  • 5 The Icelandic material is here quite complex. The wergeld ring-lists, which nowhere appear in e.g. (...)
  • 6 For the phrase “aggravated crimes”, Murray Sons of Ishmael pp. 206, 209 and passim. In Sinai and th (...)
  • 7 This is not to say such law never recognizes political authority. The Bin Zinbāʿ text makes referen (...)

3This seems not to have been so in the (tribal) Arab world. As people at Baraṭ insist, tribesmen are all of one marṭabah, one rank or level; indeed, the blood-money is the same even for non‑tribesmen such as Jews and market-folk. Although there is surely rank in evidence –as with tribesmen, the manūʿ, and “weak” people– it only takes legal form in the ability to give protection, and tribesmen themselves are nominally equals. In this respect the legal common sense of South Arabia resembles more that of early Iceland, where formal compensation differentiated only free persons and slaves.5 All these systems recognize “aggravating circumstances”;6 the relation of circumstance to status can itself be intricate. But the lack in our Baraṭ documents of distinctive provisions for chiefs –that is, for nuqabā' and shaykhs– is fairly typical of Yemen,7 and separate documents that accord some of these figures or their families hijrah‑status, like that of protected qāḍīs (part three, above), stress that la-hum ʿalay-hum, what is due to them is [also] due from them. Their status is singular but not unequivocally superior.

  • 8 The same theme looks to occur in the famous pièces justificatives that make up part four of A. Hano (...)

4Anglo-Saxon laws show also an extraordinary interest in relations between men established through women, and in what can properly be called a traffic of women among men. This too is informed by rank. “If anyone lies with a maiden belonging to the king, he is to pay 50 shillings compensation. If it is a grinding slave, he is to pay 25 shillings compensation; [if a slave of] the third [class], 12 shillings”. Purloining a ceorl’s serving‑woman means payment of six shillings but illicit sex with his slave-woman of the third class requires only thirty sceattas. A welcome degree of clarity is provided by Alfred’s prescription, c. AD 890, that a male slave who rapes a slave‑woman be castrated (the freeman who does the same owes money). Throughout the record, however, compensation for offences against females, whether or not free‑born, varies both with the woman’s status and with that of the men to whom she legally attaches.8

  • 9 Such provisions occur in the Rossi and Serjeant texts. Some are dimly visible, though very hard to (...)
  • 10 In Arab practice the link of a woman to her paternal kin is never wholly lost: M. Meeker, Meaning a (...)
  • 11 Even setting aside the polemical tendencies of Imams, there is little doubt that sexual morality in (...)

5Besides the obvious wrong of rape, we find mention in Arabian material of forced miscarriage and also of “transactional” concerns with fornication and adultery, marriage and divorce.9 These have their echoes in European law-codes. A distinction is occasionally implied between women who are veiled or secluded (ahl al-ḥajb) and others, the former, one assumes, being those of respectable families. But in customary law a woman’s blood-money is simply half that of a man, and amends are due (sometimes discrete payments, sometimes multiples of damage) which vary not with rank but with circumstance, very much as with the world of men: for a man to kill an honest woman, meanwhile, is inherently disgrace, or ʿayb.10 None of this is to say that wrongs against women all met the same response. An offence against women of a prominent and respected family might well draw collective outrage, and lesser families receive tepid commiseration; non-tribal families might fail to gain even that. Nor do rules about compensation tell us anything of sexual morality more broadly.11 Again, however, by comparison with much of northern Europe this seems a “flat” moral landscape in which men are opposed symmetrically to men, and women each fall within male-defined spaces.

  • 12 Kelly Irish Law pp. 140-1. The amount owed was the protector’s honour-price or face-price. The Angl (...)

6What makes the Anglo-Saxon world, and the otherwise very different worlds of Ireland or even (with due caution) of Norse tradition, seem familiar to a Yemenist is first merely vengeance and compensation. Where wrongs that a state claims to deal with as “criminal” offences, thus monopolizing legitimate force in the way that Weber spoke of, are dealt with by exchange between the parties at odds, all the latter cases look falsely similar. But the parties themselves are conceived on occasion in similar ways. Often men have bodily value (so much for their life or death, so much for a limb) but also a value that attaches to their protection of other men, and the right to protect defines a freeman or tribesman as against the dependant, the slave, or foreigner. Sometimes this is accorded a separate price, sometimes the value of the man’s own life: in a slightly different mode, it is marked by recompense for an “aggravated wrong”. But men of full legal status are commonly treated as minor sovereignties. In old Irish law, as in many others, injuring someone under protection was an offence against the protector himself and specifically required amends;12 a formal “power to protect” was recognized in early England, and parallels may be found elsewhere in Europe.

  • 13 As we suggested earlier (part three, above), subordinate status, as for instance within a household (...)

7Before we ever come to such plainly collective issues as markets, courts or meetings, we find certain features marked off from the social. Not infrequently there are echoes of the Baraṭ provisions (text B section 26: part six, above) on houses, and thus physical trespass; more importantly, domestic hierarchy is treated as preceding custom or legislation.13 The sets within which the elements lie vary not just in size but in nature, and the elements themselves are therefore of different value. Male‑centred spaces, which households exemplify in concrete form, are nonetheless often marked in law. We saw earlier (part three) how a rhetoric of close-range marriage may give “symbolic” clarity to the opposition of men or sets of men and to mutual refuge. More widely, however, one finds what looks to be a partial equation between the value of male-defined spaces and that of spaces or events between them. A mosque or a church for instance, or a pagan temple, may be set aside from the usual run of interaction. So too may markets and meetings. Breach of the peace there is commonly due amends for an “aggravated” wrong as is breach of protection, both being distinguished from mere exchange of violence.

  • 14 Obviously we are here on ground mapped out by Louis Dumont. I do not want to pursue formal argument (...)
  • 15 No evolutionary logic underpins the sequence. If an explanation is wanted of hierarchical developme (...)

8Here we encounter difference that in so far as it attaches to persons may equally be read in terms of circumstance or rank. An arbitrator’s status, such that amends are due for wronging him, depends in Yemen on temporary circumstance; the status of a sayyid (descended from the Prophet) was often more akin to rank, and was permanent. Rank suggests a vision of the whole.14 But in many cases, where the peace of God is only one of several, it is hard to know how we should best describe formal prominence. For example, are bishops in the Anglo‑Saxon texts “set aside”, as one might translate the Arabic muhajjar, and because of this in Aethelbert’s laws due compensation eleven-fold (bi-l-muḥaddash, as one says in Yemen) or are they politically superior? One can become the other, as happened perhaps with kings.15

  • 16 For this and the points following, Whitelock (ed.) English Historical Documents pp. 391, 395, 399‑4 (...)
  • 17 P. Wormald Legal Culture in the Early Medieval West (1999) p. 194, cf. op. cit. p.193. The seeds of (...)

9In the laws of Aethelbert (c. AD 600) the king is due fifty shillings for violence in his presence.16 In addition to this, injuries done at a meeting convened by him are paid two-fold; so are injuries done in someone’s house if the king is drinking there. But in Hlothheare and Eadric (c. AD 685) anyone who draws a weapon when drinking with others, “yet no injury is done”, owes the householder a shilling and owes the king twelve shillings although the latter is nowhere at hand and is personally unconnected with the matter. A few years later “we reach the point that the huge sum of 120 shillings (also payable for a fight in church) is applied to a punch-up in a labourer’s home, or even in open countryside...”, which payment, again, is due the king. The code of Ine (c. AD 690) says also that someone fighting in the king’s house now forfeits his property. The king can decide whether even to spare his life. The value of the king at home is perhaps less important analytically than the reach of his peace, which encompasses, where it does not replace, the peace of other men: “it is surely not too bold to detect a trend here”.17

  • 18 18 Wormald Legal Culture p. 61. For the quote following, ibid. p. 341. Cf. idem. The Making of Engl (...)

10What exactly the trend consisted in must be left to experts. But the effects are evident in language: “mund, once the value of the king’s protection, was now the premium on his commands; bót, once the redress of wrong to an injured kin, was now a fine for damaging society as a whole”.18 Other men’s mund existed still, as evidenced in redress for trepass, and well after Alfred’s laws (c. AD 890) compensation for violence against others surely varies in practice with people’s status. But all men now live more or less in the king’s peace. “Meanwhile, [c. AD 950] wergeld was payable for violating the king’s laws, for incest or precipitate second marriage (!)...” To disentangle our political and legal language from that involution and from others elsewhere in Europe, is far from easy. Difference, rank, and equality –the very stuff of equity and justice– are often hard to conceptualize apart from an imagined whole identified with public order.

Decentred society

  • 19 Claims to ancient precedent are made at most dates: English kings ruled as they did, in theory, bec (...)
  • 20 The involution of the federalist American system is a striking case. Rarely can such care have been (...)

11A long early history in Europe and a far later historiography attaching to what might be called the bourgeois state (that form of state which reached its apogee in the nineteenth century and seems now to be evaporating) both turn on ideas of a body politic, that is, of a bounded entity in which power and morality coincide. Claims to continuity are difficult to unpick. Suffice it to say for the present purpose that legal distinctions between, for instance, private and public domains are established within a political whole, not preserved from an earlier state of society before the whole was formed.19 Rights claimed as the moral basis for several current polities, whether phrased as ancient freedoms or generically as les droits de l'homme, are in practice thus treated as “liberties” granted by the body politic, and in the name of this imagined unity such liberties can be suspended.20

  • 21 Kelly Irish Law p. 141. For a useful sketch of law’s history before the Elizabethan period see R. S (...)
  • 22 Whether under British or French imperialism, it was not that all law was unified (often quite the o (...)

12What now seems the “pre-modern” world is characterized at many times and locations by a plethora of different spaces to which different rights attach: churches or mosques, markets, particular roads, and the territory of settlements. None outranks and validates the others. Rights may be validated, and seemingly preserved, by larger sovereignties, as in the English conception of immemorial practice: such validation may underpin the phrase “customary law”. But state authority shows often a marked intolerance of rights to refuge. Kelly thus cites an English treaty of 1583 that forbids an Irish magnate giving legal protection (commairce, or “comrick”) to other lords.21 The history of France in North Africa (1830 onwards) provides ample cases. The attempt to form a Berber “civil law”, for instance by Hanoteau and Letourneux in the 1860s, had it been pursued might have led to judgement distinct from others, under state tutelage, yet surely without local refuge. There is nothing specific to the nineteenth century, meanwhile, about monopolizing validation.22

  • 23 This is hardly the place to pursue the history of European ideas, but “civil society” (again now a (...)
  • 24 Hundreds” probably pre-exist Canute by at least two generations. But typically (I would guess, dia (...)

13The state in the North Atlantic world, no matter how organized internally, has long been a jealous state. It tolerates at home no rivals. Its domestic debates in recent centuries throw often an historical shadow, as if the state succeeded lesser groupings of broadly similar kind and formed or imposed peace among them.23 The only alternative, one might think, is a pre-social (individualist) condition of unstructured hordes. Actually this is not so. Rather than “units” of society existing naturally they were often enough, at all periods, constituted by state-like powers: famously, Canute (c. AD 1020) required all freemen above the age of twelve to be part of a “hundred” if they were to have legal standing, which included having each a blood‑price and being oath-worthy. What bounded groups existed before this is hard to judge.24 Indeed, the very breadth of modern literature on “kindreds” suggests how dubious the question is, yet plainly there was custom, sociability, and even law.

  • 25 The mere distinction between locals and others is asymmetrical. But incomers are commonly depicted (...)

14What we loosely refer to as social “worlds”, such as that of Yemeni tribalism, are made up not of groups or of individuals but of shared assumptions. In early Europe, in Africa, in the tribal Middle East, or indeed among the Inuit of the frozen north, foreigners had no status save as they attached to locals, any more than they did in Fustel de Coulanges’s ancient city. That is not to say locals were defined in the way of those cities as a body politic. Asymmetrical rhetorics of first-comer and second-comer, host and guest, or indeed of mothers’ brothers and sisters’ sons –centred merely on persons– allow visits, expulsion, or incorporation of strangers;25 moral equals may themselves be defined by no more than, for instance, mutual recognition of ability to speak the truth. For an incoming king like Canute to attach oath-worthiness to membership in bounded groups is a compelling means to assert control.

  • 26 Cf. J. Pitt-Rivers, Honour and social status in J.G. Peristiany (ed.) Honour and Shame (1965) p. 33 (...)

15Oath itself can do many things. It can underline the seriousness of a statement or constitute a promise of future action; it can specify (retrospectively) “intent”. Often, though by no means always, it appeals to a source of value beyond the social such as God or spirits, but its crucial importance is surely that it excludes a lie, and “to lie is to deny the truth to someone who has the right to be told it”.26 Who that is depends upon wider circumstance. Guarding a truth, and the corresponding capacity to tell it, may be ordered symmetrically through opposition, or hierarchically such that some can demand unilaterally the truth of others: appeal to the sacred fits either case. But the capacity to swear a common oath (to establish truth) in the former case is often limited to arms-bearing men. In Yemen individual oath that another has met an obligation is usually limited to the same type of people, yet outsiders of tribal status, too, are oath‑worthy. Despite variability in the detail of local practice, the nature of obligations and the need for oath are explained as “the customs of the tribes” (aʿrāf al­-qabā'il) supposedly common to much of at least Arabia.

  • 27 A famous example is Ibn Ḥubaysh of Sufyān and his men raiding Lower Yemen at the turn of the 18th c (...)

16The moral field here is indefinitely wide, not bounded. This does not mean that it encompassed everyone: the tribes of Upper Yemen repeatedly acted in Lower Yemen and in other places as if connected in no way at all to people they raided there.27 It did not, however, and still does not, have obvious limits. Any people recognized as “tribes”, no matter how little is known of them in practice or how scant interaction, can be recognized as equals. Earlier (part three) we stressed the importance of refuge with neighbouring tribesmen (arbāʿ al-qubul) as a means of pressuring one’s fellows to right a wrong, and the reciprocal importance of accepting refugees. A concern with reputation extends the logic to comparative strangers (aʿshār). Some part of the network of relatedness, meanwhile, may be actualized for a time by a pact of “brotherhood”; on the other hand, a unit in such moral schemes may sub-divide and the parts decline common action. To study a tribe in isolation would be perverse, for the values of tribalism are, by definition, not localized. What roughly we shall call morality and politics are therefore not coincident.

17The complex geography of Baraṭ (part two, above) should not distract us from how important territory is, and the borders between many tribes in Yemen are extremely clear. Inviolability of territory (ḥurmat al-waṭan) is a principle all recognize; pacts refer almost always to shared borders (ḥudūd). But such lines take their value from the opposition of tribes to tribes, or sections to sections, and in no way bound a “corporation”. There is no rule determining which elements of a tribe should speak with one voice (determining for instance how many guarantors should sign for a section); and, although specific rights of course differ sharply, no distinction of kind is recognized between men on one side of a border and those on the other. The same laws (shurūʿ) are appealed to within tribes and between them.

  • 28 All these rival descriptions might be used of local societies with their own logic. The Irish túath(...)
  • 29 Meeker ‘Meaning and society’ (part one) pp. 254-5, 259, discussing specifically “segmentary” sharaf(...)
  • 30 For a famous North African instance, E.L. Peters The Bedouin of Cyrenaica: studies in personal and (...)

18There are several ways locality may be established and lines thus drawn among people who potentially each recognize the others. In much of Europe at different dates, lordship and kingship (the former here a general term, the latter often quite ambiguous) are at least as conspicuous as territory or kin‑links;28 even godar or chiefs in Iceland, though to start with their power was minimal, seemed the obvious answer to who belonged where and with whom for such purposes as oath. Moral unity in Arab tribalism, however, is expressed as descent from a common ancestor who is long dead, or thought indeed never to have lived. Identity forms no focus of authority, and “nested” identity no political hierarchy.29 The power of shaykhs, when it emerges, tends to misfit the map of tribal divisions,30 which themselves are not “groups” in the usual sense but better spoken of usually as abstract “sets”. Expectations of moral and juridical unity can be misleading.

  • 31 A. Bernard and L. Milliot, Les qânouns kabyles dans l’ouvrage de Hanoteau et Letourneux, Revue des (...)
  • 32 The simple question of boundedness, very obviously problematic in the case of Arab tribalism, recur (...)

19How bounded the village communities in Berber-speaking parts of French North Africa were before French administration is open to doubt, but certainly the idea was there; the principle almost certainly did not extend to tribes, however.31 Across the Arabic-speaking world even villages are seldom conspicuous as moral units. The value of tribes, for their part, is predicated on what supposedly they share, on what Gibbon in a different context refers to as “a general resemblance of religion, language, and manners”. The laws of the tribes (shurūʿ al-qabā'il) are thus like a ius gentium. In this respect tribalism in Arabia is very different from the small societies that have often drawn anthropology’s attention where morality might begin and end at boundaries.32 No self-evident “groups” can be isolated, yet persons have value only through relatedness.

  • 33 Where states are conceived of as legal persons, the logic permeates discussion of international rel (...)
  • 34 Mutual recognition, of course, may be denied, cf. Harrison ‘Symbolic construction of aggression’. N (...)

20Much Western (and now global) argument proceeds on individualist lines as if persons pre-existed society. Only contract would link them morally,33 or otherwise a state align them. Most other traditions, by contrast, (Arab tribalism is an obvious case) have recognized people as connected of their nature and seen groups, whether transient or permanent, as deriving their value from relations in a broader world. They may do so as conquerors, which tends to paradox in that no equals are admitted to recognize one’s worth; they may do so as elective subordinates, hugging their values to themselves (in short, Nietzsche’s Rome and Judaea). But a common mode is to accept mutual emulation and repute over wide distances, and to admit –or to claim, when oppressed– a common humanity.34 For all their experience of this phenomenon, analysts find it awkward to rationalize as sociology.

  • 35 For an excellent discussion of these issues see R. Parkin The Dark Side of Humanity: the work of Ro (...)
  • 36 What religion was (the veneration of idols and so forth) he does not say”: Dresch ‘Mutual deceptio (...)

21Anthropology, like jurisprudence, seeks often a whole where no body politic is evident. Provoked by individualism in the Europe around them, the students of Durkheim for instance sought a body moral. The model for this communion, briefly put, was of a “church”, and the ill-defined qualities associated with matters sacred were pursued as the stuff of society, leaving endless confusion as to what the profane might be.35 A classic case is Bichr Farès’s justly famous L'Honneur chez les Arabes avant 1’Is1am (1932) where the language of ʿirḍ and sharaf –that is, of honour– is equated with “religion”.36 But insistence on the social was surely right. The current alternative to visions of the whole in English-language material is a mindless recourse to “agency” or to “practice”, which reduces the sacred to obfuscation and makes all values null or equal. It becomes very hard to say why, for instance, oath should matter or what the meaning is of aggravated wrongs.

  • 37 We shall touch briefly, later, on Western distinctions between public and private. The associations (...)

22The values of what roughly we call honour are identified in the case at hand with a relation of males to females, with what earlier we referred to as domestic hierarchy. That theme recurs, it might seem, in a vast range of ethnographic cases. But in Yemen, as elsewhere, it does not reduce to mere sexual difference (the sacred of the left and of the right as Bourdieu would have it). The same language of reputation and protection applies in Arabia to land, to territory, to guests or protégés, indeed to a range of instances, some permanent and others temporary, that give persons value. These hierarchical relations of responsibility are what insert men (most obviously, but women also) in the world of symmetrical opposition. Without them neither persons nor sets of persons need be anything but empirical phenomena and would lack obvious moral, not to say legal, worth.37 One can readily see why Farès discussing honour was drawn to Durkheimian parallels with religion and the sacred.

  • 38 The arbitrariness of venerated things and places requires thinking through. The law-stone of Icelan (...)
  • 39 E.g. Mauss Essai sur le don pp. 180-81. Between the idea of congrégations dont la permanence peut e (...)
  • 40 Dermis et al. Laws of Early Iceland part one, pp. 99, 111,221.

23The opposition of sacred and profane is no more (and no less, be it said) than that between extra-ordinary and ordinary, though formalized and marked off as such, and points of “ultimate” reference are thus arbitrary.38 The principle at issue is apparent in children’s games: one has only to trace a circle on the ground and say anyone within the circle is “it” for a compelling logic to emerge which may or may not be rationalized with explanation. No “religious” cosmology validates the ritual status of parade-grounds or common rooms. Those transient collective events to which Mauss pointed in non-state systems,39 meanwhile, have recurrent value and very often appear in law. Breach of the (temporary) peace at meetings in Yemen is an obvious example. Even in Iceland –one of the “loosest” polities in the European record in that households were thoroughly autonomous– people’s personal compensation-value was doubled at the spring and the autumn council-assemblies.40

  • 41 Calling a meeting “sacred” will seem to some readers frivolous by comparison with oath. Yet the com (...)
  • 42 An older anthropology spoke of “jural communities”. Actually, anyone not obviously dependent, or de (...)

24Pitt-Rivers’s account of the sacred as an inner space to which (mysterious, and therefore perhaps themselves sacred) strangers may be admitted needs supplementing. Hospitality, protection, and mutual honour indeed depend, as Pitt-Rivers argues, on avoiding simultaneity through “an alternation of roles”. But simultaneity can mean more than conflict. Where the moral spaces identified with arms-bearing men are brought together at meetings or markets they generate a separate form or instance of the sacred, the breach of which is due amends. No one group’s or person’s peace encompasses the others, although the honour of guarantors may be at stake; no political whole specifies right conduct. Yet, rather as with oath, the values which insert one in the world (which, in a symptomatically trivial modern image, “put one in the game”) are committed and displayed.41 To deny those values by violence after formal exchange of greetings is to deny the commonality which makes all of those present recognizable to their fellows as moral persons. This, like protection, is marked in law. Injury at a meeting might thus incur multiple amends as may injuring one’s guest.42

Murder, recompense, and loss of right

  • 43 This was once widespread in the (tribal) Middle East. See e.g. Jaussen Pays de Moab p. 224, Murray (...)
  • 44 Often one has no means of knowing whether compensation-values correspond with such ideas. To what e (...)
  • 45 See e.g. S. Hutchinson Nuer Dilemmas (1996), which deals admirably with a morality of vengeance and (...)

25The very idea of paying for a life with money deserves note. An anthropologist might expect instead forms of settlement that restore the dead person’s place in society and thus restore in some measure the integrity of a line or group to which that person belonged. The provision of a woman in marriage, whose male offspring in patrilineal schemes could perhaps be named for a murdered man, denotes an obvious moral recuperation.43 So too may provision of farm-land to a household. Still more obviously, and regardless of any spiritual identification with people (or none that is made explicit), livestock provide the promise of fecundity and continuity which death removes.44 In many societies, indeed, one finds some notion of “life” or of shared substance that makes sense, potentially, of restitution in blood or in beasts alike and allows one to pursue analytical coherence.45

  • 46 The term capacities needs adding here because loss of sanity, the senses, continence, or reproducti (...)

26Yet even the classical Islamic valuing of full blood-money as one hundred camels was surely a dead metaphor. In custom one can almost treat the value of a life as the sum of the body’s parts or capacities,46 and those recent instances where valuation of blood was in livestock appear less “survivals” than simply cases among others. Values are (potentially) all convertible, as they also were in early Europe. The sums due in compensation are rarely paid wholly to the victim’s close relatives, and in the case of death do not serve obviously to maintain, say, widows and orphans; on the other hand, even when paid as large amounts in full they seem not to figure in imagery of collective subsistence. Typically they are subject to dispute, they appear precise (therefore just) once agreed on, and often they are afterwards all but impossible to track. The insistence that damage of all sorts has precise cash-value is meanwhile made the more intriguing by a probable lack of actual coin.

  • 47 To the person outside the system it always seems there “should be” a logic of e.g. kinship or relig (...)
  • 48 See e.g. H.R. Loyn, Kinship in Anglo-Saxon England, Anglo-Saxon England vol. 3 (1974), p. 197. A fa (...)

27In the introductory essay we touched on divisions of compensation. The arbitrariness of political response was also stressed: the groups which form to pursue or compose a dispute are not identical with the sets defined by shared descent or territory, producing an endless deferral of “collective” logic.47 This theme is apparent elsewhere. Many authors have noted the theoreticism of Northern scribes, for instance, who worked out divisions of amends among degrees of kin to a point where the sums prescribed were tiny and such kin might be difficult to find;48 on the other hand, cognatic cousinship must have meant most people could count a great many relatives or be counted as relative by others, not all of which connections coincided with acquaintance. Payments defined a group in practice as much as the opposite was true in legal theory. The agnatic kin, meanwhile, came in several degrees. In Irish law, for instance, the fine was everyone connected by six generations of male descent, but the “real” kin (more properly what the laws call “certain” kin) were descendants of a great-grandfather, and for many purposes this less negotiable set was probably the one that mattered. The same is likely true of Anglo-Saxon kindreds.

  • 49 As always one is not comparing quite like with like. In certain respects the khamsah, so called, in (...)
  • 50 See e.g. Jaussen Pays de Moab p. 160, Murray Sons of Ishmael pp. 205-6. In all the badu systems whi (...)

28It is true in extreme form of Arab cases.49 Stewart’s instance from Sinai of basic vengeance groups (perhaps a dozen men) being thought of as contractual appears rare, although not unique, but assertions that larger units share “one blood” are usually of this form, written out as pacts. Meanwhile, the claim by wronged outsiders that a section or tribe, who share an “ancestor”, of their nature share responsibility is met with slaughter-beasts, pledged rifles, or sworn denial from those implicated. In North Arabia the “camel of sleep” (baʿīr al‑nawm) excuses kin two generations from the killer, leaving few men who must move or confront vengeance;50 among settled folk in South Arabia they may be trapped at home. In either case, the extreme possibility exists of a miscreant being declared a non-person by his own brothers. While this core of joint responsibility thus recedes indefinitely, the same offence and arguments over payment may involve whole tribes.

  • 51 A generalized idea of damage and wrongs as having value in objects recurs elsewhere (Kroeber’s acco (...)

29In many legal traditions besides those of Arab tribes the exact and general equivalence which is promised by coin or by precious metals co-exists with real indeterminacy as to who is involved and how, and insult does not reduce to injury. Gathering and distributing payment, which in effect determines who is involved in which respect, is properly no business of outsiders; the payment itself, from outsiders or to them, is by contrast everyone’s concern. Honour is satisfied by agreement to a sum, potentially applicable to all, that all can recognize.51 This pretence to uniform valuation, whether or not in societies that differentiate men by rank, is apparent often in a practical distinction: while personalized objects are used in pledges (the obvious Yemeni case is the dagger as surety), compensation and amends, even if paid partly or largely in household goods, demand always cash-valuation. Insults considered breach of law may be dealt with, by the same generalizing logic, as cash‑multiples.

Non-state law

30Justice (ʿadl) means exact reparation, but no standing authority exists to enforce payment. As in many systems of decentred law so in tribal Yemen, personal surety is conspicuous (part six, above) and the stake the guarantor must pledge is not the value of a debt but his own honour. At markets guarantors must pay multiple amends on a culprit’s behalf (tajbīr, in text 1, above), but even here default results first in “blackening”, or loss of the guarantor’s standing. In an ordinary dispute the debts themselves are pursued and specified, or accepted in judgement at the end, by those at odds. For them only shameful action incurs definitively loss of honour, in which case their antagonists must invoke not contingency but a breach of law; and across the whole scheme of ʿurf, as we outlined in the introductory essay, runs a legal distinction between compensation for damage (naqṣ) and amends for shame.

  • 52 Stewart ‘Contract’ p. 246. The idea of tajbīr introduces complications here, but it does not, I thi (...)

31Stewart’s question in the case of Sinai whether loss of honour is strictly loss of right52 receives here a legal answer. In the Baraṭ texts instances of ʿayb, or shame, including default on guaranty, mean loss of sinnah (texts A and B, above). More is therefore at stake than mere reputation, on which people will anyway disagree. The negative complement of ability to provide surety or an oath is formal denial of the right to exchange refuge. Again early Europe and its fringes provide parallels in the form of outlawry.

  • 53 Miller Bloodtaking p. 276 of a society where money-compensation was in practice common, or at least (...)

32The most striking case is perhaps that of Iceland, where the enormously full and complex laws of Grágás (written out in the thirteenth century, but plainly containing much that is older) have little to say, by comparison with what seems “procedure”, about amends: strictly speaking, “monetary damages for injury or killing were not available at law, as they were at arbitration”.53

  • 54 The outlaw himself named his three places at the compensation court, held 14 days after the ruling (...)

33For many wrongs one threatened formal suit for outlawry, then removed the case to arbitration and potential settlement. Full outlawry left a man no call of any kind on others: if he escaped from Iceland, despite a ban on people giving passage to full outlaws, he could be killed with impunity abroad as he could among the barren places of home. Lesser outlawry was more like temporary banishment. Though the outlaw lost his property, his exile and loss of right were for only three years, and he had three summers in which to arrange passage abroad, during which time, still in Iceland, he had three named homes.54 Within bow-shot of each, and once a month within bow-shot of the roads between them, he was safe much as other men, as he also was beyond Iceland.

  • 55 Miller Bloodtaking p. 175. For the founding myth of escape from royal domination, ibid. pp. 13‑14. (...)

34The divisions of space and time here are typical of many pre‑modern systems. Within such worlds bounded groups (in effect corporations) may offer refuge to outsiders or deny rights to their own members, perhaps expelling them formally; in other cases immunity may be denied by far looser congeries of persons. Often in both types of case justice is fragmented, with one locality’s exile being the other’s refugee. Iceland, indeed, is unusual as a European instance that defined itself in part by the absence of kings yet “the collectivity with well-defined boundaries, from which expulsion was the ultimate sanction, was the society itself”.55 In some legal dispensations a whole community might be bound to hunt down the outlawed man, in others (perhaps the more common case) he was left to fortune or to those he had initially wronged. But loss of right ensues where amends are either not admitted or are wilfully refused.

  • 56 The tale from Ḥabshūsh of Nihm and Khawlān, which we mentioned earlier, in part three, is a case in (...)
  • 57 A distinction is conventionally drawn between substantive and procedural or adjectival law. Sometim (...)

35In Arabia a presumption of refuge appears to have been usual. One could demand that another set of people expel one’s opponent if he had committed disgrace, or ʿayb, and then fled to them: one could not demand they hunt him down. Inviolability of territory was recognized, and those with whom the culprit had taken refuge would be appealed to publicly on grounds of justice, perhaps summonsed legally, to withdraw protection.56 A man whom all considered “blackened”, or whom none would stand up for, disappeared from the tribal world. Pacts, or qawā’id, meanwhile defined a space in which rights are agreed specifically, not least rights of refuge and when “covering” (ḥajb) should be recognized by whom.57 The divisions of space within such pacts are comparable in effect to those of truce ʿarrah wa-lā jarrah, whereby one stays off the other person’s or people’s land. In settlement of disputes it is not unusual to find a provision that one or other party not attend a specific market; more rarely by far, those at odds may be allotted alternate weeks. Social space, unlike abstract value, is differentiated of its nature, and the idea of peace takes forms rather different from those of public order in the rhetoric of states.

36Such phrases as “loss of immunity” (very apt to Iceland, where immunity might translate helgi) need treating with caution in Arabia. Even where protection is withdrawn of the kind that allows one escort, or to stay with others as qaṭīr, the dispute is still between only the original parties. If someone else killed the miscreant for some other reason, a whole new set of blood-debts and procedures would arise as if the victim had no different history than other men. If a tribe with whom the miscreant took refuge withdrew their protection then his proper antagonists could kill him. But so long as any recognition at all were left the man, each exchange of violence would usually implicate someone, or some set of people, who would have to decide for themselves on the justice of what had happened. In short, there is no body politic. Instead there is a string of practical decisions, with or without legal process, that redefine the possibilities of violence but do not impose general order.

  • 58 See e.g. Wormald Legal Culture p. 63. Gluckman, of course, was a friend of the great J. Wallace‑Had (...)

37To ask that law of itself provide political order is to ask a great deal, but the habit of states rhetorically is often to equate the two; and attempts are made frequently to show that non-states achieve somehow the same equation. Law is identified with its application, and two different meanings of order are conflated, that of categories or definitions and that of statistical (administrative) regularity. Mechanisms of social control are sought. Sophisticated historical work on northern Europe thus continues to cite with surprising tolerance less sophisticated anthropology drawn from southern Africa, such as Gluckman’s “peace in the feud”.58 One can seek all too easily (in Yemen as elsewhere) types of regularity, or predictability, that those within the system simply do not expect or seek for they do not think in terms of bodies politic.

  • 59 This is not to say poverty causes violence: only relative poverty, compared against one’s neighbour (...)

38Nineteenth-century accounts of tribal Yemen (Ḥabshūsh is the best instance) depict persuasively the insecurity in which people lived and the arbitrariness of endemic violence. Yet mentions of organized warfare over farm-land or grazing, for instance, are probably less numerous in the histories than those of small-scale bitterness and of pillage from strangers in tribal territory, all these being dwarfed by vicious dispute conducted in other terms or in no terms at all. Given the poverty of the region and the lack of an established logic allotting wealth from elsewhere, it would be surprising to find much else.59 “Order”, meanwhile, was promised by strong Imams. The strong Imamate of the twentieth century was remembered less favourably in its aftermath, however, than the “time of corruption” it succeeded, and since then the tribes have passed through periods of sustained violence and periods of calm. Rarely does one find oneself drawn to a logic of collective feud.

  • 60 al-Ḥasan Al-Hamdānī Ṣifat jazīrat al-ʿarab (D. H. Müller ed. 1968) pp. 194-5. For the war between S (...)
  • 61 A striking recent case was a dispute between two sections of Banī Ṣuraym, which went on for the bes (...)
  • 62 Settlement of violent dispute is important. But it often draws one to what Hart called the “bad man (...)

39Drama on a large scale is surely possible. Al-Hamdānī (writing c. AD 930) decribes the death of three hundred men in a war between Duhmah and Wā’ilah that resulted from a “weak” person under Wā’ilah’s protection being killed by somebody from Dahm;60 at almost the other end of the historical record, in the late nineteenth century, a shaykh from Sufyān carried off women from Ḥāshid (an appalling affront in local terms), who responded by laying waste part of Sufyanī’s territory. But it is scarcely the threat of such drama that informs the course of daily life. The tribal world is one of endless procrastination, avoidance, and locally explicable “exceptions to the rule”, making patterns of dispute untidy. People should and do step in where possible, imposing temporary truces (most confrontations between large sets of men, in fact, are smothered by mediation). But where something like feud ensues, on whatever scale, neighbours stand aside.61 Although there is a generalized idea that peace results from justice, there is no collective means to secure such peace, and those intractably at odds are in practice left (with regret, perhaps, yet in local terms justly) to pursue their quarrel. Settlement (ṣulḥ) depends on the willingness of those at odds, and conjuring such willingness from mutual anger may prove impossible. Yet “custom” is valued as promising just settlement.62

  • 63 The scrolls referred to can be very long, sometimes covering at least a century, the original issue (...)
  • 64 The supposed coherence of “rational” law was on this score contrasted by Weber with “qadi‑justice”, (...)

40Disputes are seldom resolved definitively but nearly always are somehow contained or managed. It is partly for this reason, whether in tribal areas or in other parts of Yemen, that custom has often been preferred to sharīʿah judgement. The scrolls of repeated judgements and repetition of the underlying case may be equally long in the two modes of law; but in ʿurf, or custom, the process unfolds among people one knows, and provisions to contain violence are as integrally part of the process as are sums decreed in compensation.63 Each is specific to the individual case. In the absence of jurisprudence, the inherent slippage between rule and application can seem to outsiders wide;64 and the structure of guaranty depends (rightly, in the local view) on expediency.

  • 65 Hart’s celebrated “secondary” rules of recognition, change and adjudication, and his suggestion tha (...)

41Rhetorics of state-law in Europe would surely denounce both features. Law and its administration need not be identified so closely, however; nor should we be drawn unthinkingly to the lawyer’s common identification of law with jurisprudence. Both equations are typical of those specific bodies politic which define themselves juristically.65 But custom, or ʿurf, in Yemen does not work with “propositions” as in much of Europe; nor is there a system of formal “precedent” as in the English system. Those knowledgeable in law and skilled in arbitration, meanwhile, are not a specialist class of the kind to whom Weber drew attention in Europe’s history but generally are tribal shaykhs, and those who turn to them for arbitration or advice are simply tribesmen. Mismatch between reality and justice is extruded, without need for further thought, to the realm of “politics” and of personal character.

  • 66 For the quote following, F.L. Cheyette, Suum cuique tribuere, French Historical Studies vol. 6/3 (1 (...)

42Reconciliation is what rules judgement (al-ṣulḥ sayyid al‑aḥkām). No conflict of principle is assumed between law and fair settlement, ex aequo et bono as lawyers say; no abstract “rationality” requires maintaining. Cheyette depicts the logic for early mediaeval France, where (as in Yemen later) payments were made back and forth in settlement:66

The practice of giving everyone something was ... so prevalent that it is impossible to reconstruct any objective rules of decision on the basis of arbitral judgements in lower Languedoc, at least before the mid-thirteenth century ... When there were conflicting claims to property, nothing would be settled until everyone was satisfied.

  • 67 The French distinction was geographical, some regions claiming more immediate reference than others (...)

43Studies of non-state law thus evaporate often into study of “dispute resolution”. The existence of law itself may be identified with the claims of practitioners in courts and chanceries; or the absence of a lawyer’s “objective rules” may be attributed to droit coutumier as opposed to droit écrit.67 The former is known just as much as the latter through written evidence, and often its value is defined in advance by myths of state. But much of that myth is extremely recent, forming part of the discourse that gives us the notion of “early law”. Such generalizing schemes work poorly.

  • 68 Cheyette ‘Invention of the state’ p. 161, cf. Wormald Legal Culture p. 23, and English Law p. 95, w (...)
  • 69 The form of pacts is shown by the texts translated earlier. An idea of more generalized law is give (...)

44Cheyette points elsewhere to a style of composition, rather different from that of professionalized law, which accounts for much of the initial feeling of resemblance between, say, Arabia and early Europe. He does so by quoting not “custom” but codes from early English kings: “If someone does this, then the following...”; “If it should be, and someone does that...”.68 While later European law tends often to abstraction, the rules here are immediate and concrete. The condition of using such a style is that the terms make sense independently, and no totalizing system is assumed in which categories must themselves be defined, through endless self-reference, by the law or philosophy which uses them: in other words, no jurisprudential enterprise claims exhaustive truth. The law of the tribes in Yemen takes the concrete form. Definitions of persons, for instance, are treated as given, and rules concern instead what counts among persons, where and when, as right action or responsibility. That is what one might expect, perhaps, of pacts (qawāʿid). But it is also the style of law (sharʿ, pl. shurūʿ) supposedly not limited by time or place, or voluntary commitment.69 The law is not seen in lawyers’ terms.

  • 70 For these practical distinctions and the forms of rhetoric see e.g. Dresch ‘Episodes in a dispute’, (...)

45Arabic ḥukm might, abstractly, be translated as arbitration, judgement, or even political dominion. A ḥākim (or someone muḥakkam) in ʿurf is an arbitrator, agreed to by those at odds. But the ḥukm he issues is a ruling. Usually he specifies certain amounts due, and often he refers some to ʿurf al-jihah, the supposed “custom” of the plaintiff’s area or people: an element of self-judgement is thus approved. Yet all, if justly done, accords with “the laws of the tribes” (shurūʿ al-qabā'il), which supposedly apply to all.70 The lawyer’s question of how “custom” becomes “law” has a quite different meaning here than it would in a jurisprudential state. While all things are arguable, what is actionable is most obviously default on specific responsibilities. Certain instances, such as escort, hospitality and guaranty, are thus the focus of what was once called ḥukm al-manʿ (roughly, law or judgement which centres on ideas of protection, and therefore honour), while custom itself (that is, simply ʿurf) includes potentially the whole of life.

  • 71 Cf. Cheyette ‘Suum cuique’ p. 287. The English term is notoriously all-inclusive. In French what on (...)

46To argue over definitions is not to think. What we recognize as “law” may be an appeal to custom, the legislation of a sovereign body, or simply the world inhabited by lawyers.71 Yet rules are invoked widely, and resemblances among ethnographies are numerous. The circumstances defy reduction to instrumental order. Early kings in Europe were thus associated by name with law-codes, and references to such law being used in practice are to start with significantly absent: what mattered was promulgating a written text. Elsewhere, as with Ireland, we find bodies of specialists identified with formal law. Iceland provides a further type of case, with its vast amount of (later) manuscript that appears to regulate almost everything:

  • 72 Miller Bloodtaking p. 228. For the earlier points about kingly law, Wormald Legal Culture chapter 1 (...)

this mass of rules was mercifully unaccompanied by any state enforcement mechanism. There was no bureaucracy to oversee their administration, no permanent presence, other than the inscrutable Law Rock, that had a vested interest in their being honoured...72

  • 73 Kant, of course, built an approach specifically to ethics around the categorical imperative that on (...)
  • 74 Concept of Law p. 155, discussing specifically the idea of justice. For all that rules are not the (...)

47Whole societies on occasion, some with ideas of communal sovereignty and some with none at all, seem fascinated with legal procedure. More generally, however, the world is seen as governed morally by rules. This is hardly to say more than that life can or should make sense. Where such rules are invoked as law it is not primarily because they constrain individuals or offer such individuals a bargain with their fellows (both may be true of course) but that they express in intelligible form what justly cannot be otherwise.73 Not all are of equal weight, nor all of equal generality (some, indeed, are trivial or highly localized); all, however, might fit Hart’s simple precept, “Treat like cases alike and different cases differently”.74

48Like state-centred law, customary law in Yemen specifies rights and obligations, and often (by no means always) the sanctions or penalties for wrong. Jurisprudence, which custom does not have, produces in other systems –not all of them state-centred– a coherent account of how the parts fit together, or on rational grounds do not: it produces, indeed, an hierarchical vision of the whole. But whether law is enforced, or enforceable, is rarely fundamental. Acceptance that there simply is law (we need not all know, of course, what it says in detail) is far more basic, and not surprisingly its significance depends on wider self‑definition. Some societies thus attributed law to kings, others to mythic legislators; still others, rather oddly but with huge results, have claimed to discover or invent law as they go along. The tribes of Yemen are one of that large set of people who treat law as having always been there.

  • 75 The law of nations, as ius gentium, was for Rome first a lowest common denominator, inferior to civ (...)
  • 76 L. Oppenheim International Law vol. 1 (1958) p. 38. Higher rules can as well be called basic rules, (...)

49Custom, or ʿurf, among the tribes is in no sense an agreement among bodies politic. Nonetheless, its nature is less that of a complement or precursor to the internal law of states than of a parallel to what used to be thought the law of nations,75 where again the “whole” to which anthropology so often appeals is felt by lawyers: “it is only by reference to a higher legal rule in relation to which they are all equal, that the equality and independence of a number of sovereign States can be conceived”.76 But that rule, if its existence be admitted, is merely imminent in what is done. Something of the kind is true of many systems. Indeed, where no jurists exist to formulate doctrine (that is, to formulate the world in the terms of their profession) “law” may be as little bounded as the moral field it expresses and in part defines.

The order of (national) states

  • 77 S. Merry, Legal pluralism, Law and Society Review vol. 22 (1988). In “modern” states, however, such (...)
  • 78 L. al-Zwaini, Mediating between custom and code: Dâr al-Salâm, an NGO for tribal arbitration in Ṣan (...)

50Within what students of international law call “municipal” jurisdiction, more is going on than government prescribes or legislates. Most social formations exhibit in this sense legal pluralism.77 In Yemen rhetorical contrasts among sharīʿah (Islamic law), ʿurf (or appeal to “custom”), and qānūn (governmental legislation) obscure a more complex practice where ʿurf itself varies among activities and geographic areas, and sharīʿah remains conceptually unlimited: an empiricist might find as many “domains” of law as there are villages, activities, or city neighbourhoods. The current world, however, is that of the jealous state. State law in Yemen, for instance, thus recognizes and in its own terms “validates” the fact of non‑governmental arbitration.78 But the political whole in whose name such legitimacy is granted is itself not institutionalized uniformally, any more than elsewhere, and people, still, are related of their nature. Let us take a case from the 1990s.

51Bal Ḥārith, just north of Ṣanʿā’, were afraid the government were about to sequester part of their land to extend the airport. They went to Dhū Muḥammad, with whom they have no immediate connection of politics or genealogy, and slaughtered beasts to demand help, just as they might in dispute with some other tribe: talaḥḥaqū fawq dhū muḥammad ʿalā l-dawlah (or min al-dawlah). In other words they formed a connection (luḥqah). Dhū Muḥammad duly approached the Governor of Ṣanʿā’, who gave them rifles in the usual way to promise good faith and asked a shaykh of Dhū Muḥammad to arbitrate. Bal Ḥārith’s concerns were taken up by Dhū Muḥammad and compensation agreed in the event that Bal Ḥārith land be needed.

  • 79 Tribalism” is my abbreviation of the arguments, not a local term (although reported from western Y (...)

52Whatever the chances of a happy outcome to the process, this is ʿurf or “custom”, as much as are rules defining vengeance. There is simply an accepted way of doing things. Yet this is not what state enthusiasts, or tribesmen when denouncing the state’s shortcomings, mean by “tribalism”.79 One set blame disorder on tribalism, the other blame disorder on tribalism’s demise: writers on politics translate this as geographical division. The distinction between tribal territory and “the land of the state” in our eighteenth-century material is long gone, however, and was anyway contingent on the Qāsimī Imamate’s perceived injustice. Nationalist states in the current world present different questions. If we wish to abstract from the present two concepts of law, then a difference of principle suggests itself: states of their nature are monist, so to speak, and tribalism pluralist.

  • 80 No-one who has read Mauss will be surprised at the aligning of dispute and trade.
    Quite widely in th
    (...)

53Settlement of a dispute in custom means not subordination to a common good but separation of those at odds. Ideally, as the phrase runs, “there is nothing between them”. Appeals to more general relatedness through all-encompassing (agnatic) descent are appeals not least to a proper autonomy, and the nominal equality apparent in law, which ignores so many practical inequalities, expresses this unthought axiom. Tribes are of different size (thus strength), men are of different wealth and influence. All identities save that of the person are contextual or provisional. But the comparison implicit in disputes over compensation, which threatens always to rank those at odds, is never generalized, for each of the parties at odds, apart from the encounters of dispute and trade,80 subsists in their own moral space.

  • 81 Changes in what some call advanced democratic states have provoked much literature on “governmental (...)

54Such themes recur elsewhere in systems of morality that could hardly be more different. In the classically economic space of nineteenth- and early twentieth-century Europe, for instance, domestic finance was a private sphere: overt comparison of wealth and wage-rates meant a ranking that compromised the space for self-respect implied in decency. Nor was “privacy” a matter only of morality. The “rule of law” expressed precisely the ideal that moral persons, and associations of persons, were not subject arbitrarily to collective valuation by the state which encompassed all.81 What many worlds exhibit, in fact, is a systematic non-comparability of persons, or of sets or groups, in all but defined spheres.

  • 82 As the constitutional (administrative) state took increasingly solid form in Britain, at the turn o (...)

55This is a principle to which anthropologists, and perhaps historians, have given less thought than it deserves, accepting too often as exhaustive the distinction between hierarchy (where value derives from a vision of the whole, promulgated often by a specialist class, of whom the legal profession are for Weber a type case) and egalitarianism, where individualism threatens always disorder or a violently imposed conformity. All hierarchy need do –if the term be appropriate– is to allot the spaces, be they physical, as with territory, or abstract, such as domains of civil life. The bourgeois state, with its invention of “private” space as against public rights and duties, thus counterfeited within a moral whole what many societies had the luxury of treating as a natural order. Turn the prism another way, indeed, and tribalism and the bourgeois state appear in the same moral set: with tyranny as their shared opposite.82

  • 83 To appeal to lack of evidence is to raise the possibility merely of weak ethnography.
    But no account
    (...)

56The case might be made that in tribalism there is no whole, and that the language of hierarchy is superfluous beyond the domestic realm, or beyond the more general realm of protection which domestic examples typify.83 It is hard to construct a “cosmology” to which tribal divisions would correspond as part to whole; the “sacred” appears not as validating or encompassing society but often as imminent in social events and circumstance. Yet there is no doubt that in law (amply evident in the settlement of disputes) men are opposed to men, and tribes to tribes, as nominally equals. This is so without recourse explicitly to a “higher legal rule”. What is evident in actual rules is a certain priority of procedural over substantive law, and a concern to separate, not subordinate, those at odds.

  • 84 A value-judgement would be premature and in any case irrelevant to understanding.
    But here, analytic
    (...)
  • 85 One would not contest the status of international law with experts, and the principle that it appli (...)

57The vulnerability of agreements to disagree is notorious. The complement to the endless recurrence of tribal pacts in Yemen, for instance, is an endless recurrence of “strong leaders”, such as many of the Qāsimī Imams in the eighteenth century who promised order at sword-point. In Europe elected dictatorship recurs (from Rome if a classical past is sought) through Napoleon III of France to Italy’s and Germany’s disasters of the century just past. The rule of law offers little safeguard morally where law is conceived of as the expression of collective will. In our own days the bourgeois distinction between public and private spheres collapses in the face of populism, and concerns dealt with previously as embedded in “contexts” of responsibility are reflected in an abstract vision of community.84 All forms of autonomy become suspect. At the same time, and among the same constituencies as favour brisk action to right perceived wrongs at home, the pluralism that characterized international law is displaced by a demand for effective pan‑national sovereignty.85

58The detail of short-term politics aside, the conceptual stakes are high and room for bitterness extensive. The provisional nature of social truths becomes hard to argue, and a space between morality and politics becomes difficult for moral activists to conceive of as other than condoning wickedness; yet many traditions, widely separated in time and space, have placed refuge over sanction, as if on the assumption that claims to absolute truth are to be mistrusted. The conceit was maintained that what is right is not defined by groups, even by bodies politic, but only exemplified by them in varying degrees. Behind or beyond what Europe called “positive law” thus lay values that for all but state functionaries were more basic; and the prominence in both American and Continental jurisprudence of what amounts to natural law exemplifies how the argument continues. After all, law could be “posited” by tyrants, and where kings are replaced by “the people” the threat to morality is all the greater, the claustrophobic constraints of “the state” all the less escapable.

  • 86 For the quotations following, E. Gibbon The History of the Decline and Fall of the Roman Empire (18 (...)

59The logic of the modern state is total, and expanding it to global dimensions, as even those who cite Kant seem sometimes to envisage, is the stuff of nightmare, for community and morality are different things, the distinction being sometimes allowed by mere contingency of history. The nineteenth century in Europe gave us evolutionary and historical arguments over where “law” had come from; the eighteenth century had posed the (misleading) terms of individual against society. Such terms present always the dilemma whose horns are commonly chaos and sovereign power. But enlightenment thinkers worked within a social order as little theorized as that of Yemeni tribes with their notions of a common, uninstitutionalized morality. Edward Gibbon in the third chapter of his Decline and Fall of the Roman Empire (1776) depicts the contingency with his usual charm:86

  • 87 Judgements of this kind are never mere reports on empirical similarity or difference.
    In passing on
    (...)

The division of Europe into a number of independent states, connected, however, with each other by the general resemblance of religion, language, and manners, is productive of the most beneficial consequences to the liberty of mankind.87 A modern tyrant, who should find no resistance either in his own breast or in his people, would soon experience a gentle restraint from the example of his equals, the dread of present censure, the advice of his allies and the apprehension of his enemies. The object of his displeasure, escaping from the narrow limits of his dominions, would easily obtain, in a happier climate, a secure refuge, a new fortune adequate to his merit, the freedom of complaint and perhaps the means of revenge.

60Readers more concerned with order than with liberty might place this as a mere concomitant of supposedly endemic civil war. One might equally complain that while Voltaire could skip from court to court, most of the population could not, and even in Gibbon’s time there were those who saw justice as better served by unity on the Roman model.

But the empire of the Romans filled the world, and when that empire fell into the hands of a single person, the world became a safe and dreary prison for his enemies. The slave of imperial despotism, whether he was condemned to drag his gilded chain in Rome and the senate, or to wear out a life of exile on the barren rock of Seriphus, or the frozen banks of the Danube, expected his fate in silent despair. To resist was fatal, and it was impossible to fly... Beyond the frontiers, his anxious view could discover nothing except the ocean, inhospitable deserts, hostile tribes of barbarians, of fierce manners and unknown language, or dependent kings, who would gladly purchase the emperor’s protection by the sacrifice of an obnoxious fugitive. ‘Wherever you are,’ said Cicero to the exiled Marcellus, ‘remember that you are equally within the power of the conqueror’.

61In the modern age the erstwhile barbarians live under dependent kings, and neither oceans nor deserts can offer refuge: the frontier, in a sense, is closed. To constrain the emperor by municipal law, or its global equivalent, scarcely resolves the problem. No matter how just the polity, the language of the jurisprudential state invites always the equation of one moral regime with humanity itself, and of moral action with the exercise of power. Other legal traditions such as customary judgement (ḥukm ʿurfī) offer means at least to retain distinctions of understanding among justice, morality, and political order. Natural law, for instance, need not be seen as municipal law write large, a code ideally to be imposed on all. It need only be the recognition of shared humanity in the same degree as are the laws of the tribes (aʿrāf al-qabā'il). The “procedural” features such as escort and guaranty to which ḥukm al-manʿ always gave such prominence provided means ideally for settlement by ʿurf al-jihah, the supposed custom of the plaintiff’s area or people, and what that was, precisely, was not for the plaintiff’s antagonist to define.

Notes

1 See e.g. Fā’iz Mūsā al-Ḥarbī al-Tanẓīmāt al-qānūnīyah al-qaḍā’īyah laday qabā’il al-ḥijāz (2 vols., 2000), Ashhar al-quḍāh wa-katabat al-wathā’iq wādī farʿ (2004).

2 The three traditions from northwest Europe have little to do with one another, and none, I assume, shares much history with Arabia: such simple parallels as paying for the violation of a man’s protection are therefore worth noting. Obviously one could reach out to unwritten rules of dispute and settlement in Africa, New Guinea and elsewhere, but literate traditions by themselves provide food for thought.

3 A convenient presentation of the main texts is D. Whitelock (ed.) English Historical Documents c. 500-1042 (1979). For Ireland and the face-price, below, F. Kelly A Guide to Early Irish Law (1988) pp. 8-9, 126.

4 The wergeld is compensation for a man’s life, as in murder. The mundbryce is the value of his peace or protection, and mund itself means a general power to protect. The first manuscript use of mundbryce seems to come in the 10th century, but the idea is plain in e.g. Hlothhaere and Eadric’s laws.

5 The Icelandic material is here quite complex. The wergeld ring-lists, which nowhere appear in e.g. saga accounts, might allow one to differentiate free farmers from freedmen and the latter from slaves. In practice the amount of compensation paid, or extorted, varies hugely, and “the likelihood of commuting blood into money varied inversely with the popularity of the corpse” (Miller Bloodtaking p. 203, cf. ibid. p. 278). But the law seems to envisage formal compensation as properly the same for all free Icelanders: ibid. p. 37, A. Dermis et al. Laws of Early Iceland part one (1980) p. 8.

6 For the phrase “aggravated crimes”, Murray Sons of Ishmael pp. 206, 209 and passim. In Sinai and the Western desert as in Yemen the usual penalty was demanded “four-fold”. Al-ʿAlīmī drawing on more generalized legal terms to describe Yemeni custom, distinguishes ẓurūf takhfīf and ẓurūf mushaddidah (Qaḍā’ qabalī pp. 72-3).

7 This is not to say such law never recognizes political authority. The Bin Zinbāʿ text makes reference to duties owed to one who “orders the good and prohibits the reprehensible”, that is, a person claiming to uphold Islam: Abū Ghānim Bunyah qabalīyah pp. 378, 383. Others of Rossi’s texts (‘Diritto’ pp. 21, 24) mention sultans.

8 The same theme looks to occur in the famous pièces justificatives that make up part four of A. Hanoteau and A. Letourneux La Kabylie et les coutumes kabyles (3 vols., 1873). Closer inspection shows, however, that most of these deal with conceptually fairly equal transactions among families, concerning marriage, adultery, divorce, and inheritance. What is striking about the Berber qânouns by comparison with Arabian material, meanwhile, is the sheer amount they contain concerning women.

9 Such provisions occur in the Rossi and Serjeant texts. Some are dimly visible, though very hard to read, in Abū Ghānim Bunyah qabalīyah pp. 373, 380, 383, 384, cf. Rossi ‘Diritto’ p. 24. For ahl al‑ḥajb, below, Bunyah qabalīyah p. 364.

10 In Arab practice the link of a woman to her paternal kin is never wholly lost: M. Meeker, Meaning and society in the Near East, International Journal of Middle East Studies vol. 7 (1976). But even close‑range marriage among paternal relatives, though the case is hard to frame in customary-law terms, seems not to mean women drop out of moral space in the way that happens with modern prisons or with ancient slavery. Obviously one should not search for parallels with bourgeois ethics, but neither does any parallel come to mind with the struggle the church often had in early Europe to promulgate the idea that men harming women was inherently a wrong.
It is common, nonetheless, for women in Yemen seeking to escape local constraints, or these days state initiatives, to appeal specifically to Islamic law. See A. Wurth, Mobilizing Islam and custom against statutory reform
in B. Dupret (ed.) Le Shaykh et le procureur (2005). Ibn al-Mujāwir (Tārīkh al‑mustabṣir pp. 103, 154-5) suggests long ago some intriguingly parallel worlds, where for instance only women’s testimony would be admissable in disputes among women.

11 Even setting aside the polemical tendencies of Imams, there is little doubt that sexual morality in early Yemen was very different from present ideals: Serjeant ‘Tribal affinities’ pp. 19-20. What the practical expectations were in different areas later is hard to guess.

12 Kelly Irish Law pp. 140-1. The amount owed was the protector’s honour-price or face-price. The Anglo-Saxon rates are set out in Whitelock (ed.) English Historical Documents.

13 As we suggested earlier (part three, above), subordinate status, as for instance within a household, although even this only illustrates the more general logic of protection, may “feminize” men symbolically if they cannot reciprocate. This is not the same as substantive identity. Loose talk of “patriarchy”, as if all the world were like legal Rome, is equally unhelfpul, for rather few societies grant men powers of life and death over free-born women though many have done so over slaves. Aristotle stressed (Politics, Book 1, chapters 2 and 12) that wives are of a different nature to slaves though both were subordinates in households: indeed it is here, in family affairs, that political virtue supposedly has its origin. The failure to grasp the distinction between slaves and free-born wives he claimed typified certain barbarians, much as Western propagandists now claim is true of the Muslim world. Muslim suspicions of European barbarism concern often the same domain.

14 Obviously we are here on ground mapped out by Louis Dumont. I do not want to pursue formal arguments about hierarchy, for one ends up with logic-chopping; and when I say “opposition” I mean always a symmetrical relation. In practice we need only concern ourselves with how values are ordered, and with whether and how an order is made explicit.
Although Meeker (‘Meaning and society’ part two p. 421) is properly cautious, claims to Islamic righteousness complement the provisional nature of claims made in terms of e.g. territory or kinship. It would be useful to know far more of the practical ideologies informing such less than sectarian Islamic states as that of the Rasūlids.

15 No evolutionary logic underpins the sequence. If an explanation is wanted of hierarchical developments across early Europe, it lies not with the qualities of kin-groups but with monopolies of writing and with relations of the church to worldly power, cf. F.L. Cheyette, The invention of the state in B. Lackner and K. Philp (eds.) Essays on Medieval Civilization (1978).

16 For this and the points following, Whitelock (ed.) English Historical Documents pp. 391, 395, 399‑400. Not only are injuries in the presence of the king paid two-fold (below), so are those at church or at meetings. The principle is amply familiar from Arabia.

17 P. Wormald Legal Culture in the Early Medieval West (1999) p. 194, cf. op. cit. p.193. The seeds of the idea are evident from the start in that Aethelbert’s code prescribes payment to the king of his protection price for the killing of any freeman: Whitelock (ed.) English Historical Documents p. 391. The important point for our purpose is not whether kings at a given date actually controlled people, but whether the moral claim to hierarchy made sense.

18 18 Wormald Legal Culture p. 61. For the quote following, ibid. p. 341. Cf. idem. The Making of English Law (1999) p. 105 and, for a later stage, p. 312.

19 Claims to ancient precedent are made at most dates: English kings ruled as they did, in theory, because “the law doth permit”. But that is not to say precedent unrolled of itself into later constitutional thought and ideas of “civil society”. The English constitutional case starts on this score from a newly formed system of absolute monarchy, the Tudor state, not hugely different from that of inter-lacustrine Bantu kingdoms. For an interesting, if sometimes too elaborate, examination of “modern” politics arising within such a moral whole see S.L. Collins From Divine Cosmos to Sovereign State (1989).

20 The involution of the federalist American system is a striking case. Rarely can such care have been taken in establishing a polity both to anticipate the threat of populism and to limit the powers of government, yet moral alarms phrased in terms of the unmediated whole repeatedly over-ride provisions in the founding document. Winning back space from absolutist claims requires often legal additions at federal level to precisely the institutions that acted ultra vires: the idea of rolling back federal jurisdiction to its original bounds, thus preventing such questions arising, has only a quaint charm.

21 Kelly Irish Law p. 141. For a useful sketch of law’s history before the Elizabethan period see R. Sharpe, Dispute settlement in medieval Ireland in W. Davies and P. Fouracre (eds.) Dispute Settlement in Early Medieval Europe (1986).

22 Whether under British or French imperialism, it was not that all law was unified (often quite the opposite occurred) but that autonomous conduct of proceedings was forbidden: “native courts” or the like were subordinated to the state, if not wholly state inventions. To look for signs of imperial power in “the letter of the law” is often to look in the wrong place. Validation of courts and denial of rights to refuge are usually of more importance.

23 This is hardly the place to pursue the history of European ideas, but “civil society” (again now a current phrase) is often written of as if it precedes the state both logically and historically. Rather the opposite is true, in fact. People are social of their nature, but to speak of “society” in this sense is to speak of what does not exist without the state. The idea of preventing disorder through political hierarchy, meanwhile, runs deep. From the legal historians of the 19th century (indeed, from Milton) to recent historians and anthropologists, one need only note how “anarchy”, or having no leader, is used as a synonym for chaos.

24 Hundreds” probably pre-exist Canute by at least two generations. But typically (I would guess, diagnostically) we know little about what they were: it is only as monarchy co-­opts ways of doing things that the latter come into view and take solid form. The point about “kindreds”, below, is a separate one but surely obeys the same logic.

25 The mere distinction between locals and others is asymmetrical. But incomers are commonly depicted as, for a time at least, inferior to or controlled by those already present, and in many cases it seems to be recognition of this role as patron that makes absorption of outsiders unproblematic. When the status of patron is removed, as it is by modern states, foreign residents become easily an impersonal, undifferentiated threat.

26 Cf. J. Pitt-Rivers, Honour and social status in J.G. Peristiany (ed.) Honour and Shame (1965) p. 33, see also p. 34. For the related issue of intention, ibid. pp. 26-8, 32. Early anthropology showed a fascination with oaths invoking deities, and it is well to remind ourselves that oaths may equally well be sworn on personal value: one’s beard, one’s mother, one’s mother’s grave, and so on.

27 A famous example is Ibn Ḥubaysh of Sufyān and his men raiding Lower Yemen at the turn of the 18th century, and women’s ear-rings, some with ears attached, appearing for sale in the Ṣanʿā’ markets. See Muḥammad Zabārah Nashr al-ʿarf vol. 2 (1958) p. 670.

28 All these rival descriptions might be used of local societies with their own logic. The Irish túath or “tribe” is a good example. It supposedly needed not only a king, but a churchman and a poet; the physically umblemished king was associated with prosperity. The fact that a king took no fine from disputing parties does not mean that justice was therefore, by default, “purely kin-based” (Wormald Legal Culture pp. 195-6).

29 Meeker ‘Meaning and society’ (part one) pp. 254-5, 259, discussing specifically “segmentary” sharaf. In practice different levels of identity are not all stressed equally. We saw earlier how Ṣubārah, half of Sufyān, could agree its own pacts; so can smaller units. But the word tribe (qabīlah) in Upper Yemen is used mostly of such units as Dhū Muḥammad, Sufyān or Banī Ṣuraym, not of their components. I cannot provide a sound explanation why.

30 For a famous North African instance, E.L. Peters The Bedouin of Cyrenaica: studies in personal and corporate power (1990) pp. 112-37. The misfit is as prominent, though for different reasons, in Upper Yemen. A good idea of the span of shaykhly agreements, from which readers can draw hints of influence, is given by the documents in Abū Ghānim al­-Qabīlah wa-l-dawlah fī l-yaman (1990) pp. 471‑639.

31 A. Bernard and L. Milliot, Les qânouns kabyles dans l’ouvrage de Hanoteau et Letourneux, Revue des Études Islamiques vol. 7 (1933), pp. 22-3. The diagnostic question is whether “fines” are payable to a group as such or “amends” to named guarantors.

32 The simple question of boundedness, very obviously problematic in the case of Arab tribalism, recurs in a valuable discussion of Lowland New Guinea: S. Harrison, The symbolic construction of aggression and war in a Sepik River society, Man n.s. vol. 24/4 (1989). “Society” is often something to be carved from the social by suppressing or denying relations of common humanity with those outside it, but humanity itself, and thus the unbounded social, extends indefinitely far.

33 Where states are conceived of as legal persons, the logic permeates discussion of international relations, as in Kant’s famous essay on “Perpetual Peace” (constitution.org/kant): “the harmony of politics with morals is possible only in a federative alliance...” But, whether in Europe or New Guinea, an assumption of common humanity keeps recurring; and the ability to give one’s word (surely the logical requirement of Kant’s “federation”) was assumed in European relations with many non‑European powers until after Kant’s time.

34 Mutual recognition, of course, may be denied, cf. Harrison ‘Symbolic construction of aggression’. Nor need one go to New Guinea for cases. Aggressive states in Europe’s recent history or the conquest kingdoms of Africa provide striking examples of moral exceptionalism. “No triumph of peace is quite as great as the supreme triumphs of war” said Theodore Roosevelt, chief of a polity with no natural enemies that almost from its founding showed a popular taste for military adventure and an earnest moral singularity.

35 For an excellent discussion of these issues see R. Parkin The Dark Side of Humanity: the work of Robert Hertz and its legacy (1996) pp. 27-30, 34-5 and passim. Presumably the question begins with generalizing William Robertson Smith’s view of Greek and Roman cults.

36 What religion was (the veneration of idols and so forth) he does not say”: Dresch ‘Mutual deception’ p. 129. Farès, to be fair, is cautious: “Nous ne prétendons point que les éléments de l’honneur sont, de par leur nature, des croyances et des pratiques religieuses. Ça serait s’abuser outre mesure. Cependant, ils en remplissent la fonction: ne font-ils point communier dans les mêmes représentations et fraterniser dans les mêmes actes la totalité des Arabes?” Plainly “religion” here is not at all the same as Islamic dīn.

37 We shall touch briefly, later, on Western distinctions between public and private. The associations of such distinctions more generally with sexual asymmetry mean the distinctions are deeply (one might also think rightly) suspect to those concerned with the moral autonomy of women. But when sexual hierarchy is removed, we still face a fundamental problem: without a domain of their own to control, legal subjects, be they female or male, become morally mere objects to be ordered by administration.

38 The arbitrariness of venerated things and places requires thinking through. The law-stone of Iceland (Miller Bloodtaking p. 228) and the tomb of Jubar b. Khawlān (Ḥabshūsh Yémen p. 70, Ru’yah p. 38) are related cases. North Africa, despite an easy analytical recourse in the literature to barakah, provides ample cases where one need only white-wash a rock to produce a “saint”.

39 E.g. Mauss Essai sur le don pp. 180-81. Between the idea of congrégations dont la permanence peut excéder une saison de concentration sociale and that of l’instant fugitif où ... les hommes prennent conscience sentimentale d’eux-mêmes lies room for much investigation.

40 Dermis et al. Laws of Early Iceland part one, pp. 99, 111,221.

41 Calling a meeting “sacred” will seem to some readers frivolous by comparison with oath. Yet the common use of the term in anthropology is arbitrary, and very often it serves to round out an holistic analysis by appeal to external value. In tribal Arabia the social appears not to be validated by anything beyond itself except presumed antiquity.

42 An older anthropology spoke of “jural communities”. Actually, anyone not obviously dependent, or defined in advance as such, who is willing to recognize the penalty can thus recognize the law and become part of the moral field of those willing to live and let live as mutually recognizable persons. Powerful outsiders who refuse the formalities appear savages, intent on something akin to rape.

43 This was once widespread in the (tribal) Middle East. See e.g. Jaussen Pays de Moab p. 224, Murray Sons of Ishmael p. 203, Peters Cyrenaica p. 77. Dislike of the practice by the women involved was parallel, symptomatically, to their dislike of exchange marriage.

44 Often one has no means of knowing whether compensation-values correspond with such ideas. To what extent, for instance, did the Irish at different periods think in terms of silver, and to what extent in terms of actual heifers, milch cows or female slaves, which provide the terms for units of compensation? Cf. Kelly Irish Law pp. 112-16.

45 See e.g. S. Hutchinson Nuer Dilemmas (1996), which deals admirably with a morality of vengeance and compensation encountering cash transactions. A set of very different cases needs thought, where before any notion of “economy” was present lives were valued in money.

46 The term capacities needs adding here because loss of sanity, the senses, continence, or reproductive ability is commonly valued at the full blood-money. See e.g Document 7 in part three, above. Nonetheless, a limb is half a life and vice-versa. Islamic law, unlike custom, introduces a distinction by denying a general lex talionis: for murder one has a choice between killing and payment, but wounds should be settled by paying arsh.

47 To the person outside the system it always seems there “should be” a logic of e.g. kinship or religion. It can also appear that way to people within the system (hence “family ties are not what they were”, “in those days there was really Islam”). But for them a complementary view suggests equally that all other values must be denied to uphold those of honour which make one what one is: hence extravagant tales of e.g. killing one’s kin who wrong a protégé.

48 See e.g. H.R. Loyn, Kinship in Anglo-Saxon England, Anglo-Saxon England vol. 3 (1974), p. 197. A famous case of this sort is the elaborate scheme of baugatal, paying blood-­money, in saga Iceland, for which there appears to be no evidence but jurists’ musings (Miller Bloodtaking pp. 144-5), not to mention the likelihood of a population too small to furnish the requisite degrees of cousinship.

49 As always one is not comparing quite like with like. In certain respects the khamsah, so called, in North Arabia suggests ambiguities that may not have arisen with e.g. the Irish fine. The older literature on descent and marriage illustrates a very deep-rooted pattern in the Arab world: where close-range marriage is culturally marked, as Murphy and Kasden said 45 years ago, men who share agnatic descent, “do not occupy a status in Maine’s sense, but nor are they related just by contract”. Relatedness is unavoidable and inherently ambiguous.

50 See e.g. Jaussen Pays de Moab p. 160, Murray Sons of Ishmael pp. 205-6. In all the badu systems which the literature describes those immediately liable in a killing move elsewhere. Typically there is a brief period in which unthinking retaliation is admitted, and then the party at risk are either under others’ protection or are simply absent.

51 A generalized idea of damage and wrongs as having value in objects recurs elsewhere (Kroeber’s account of the Yurok in California used once to be well known). Money introduces other questions. Across southern Europe, as for instance in Visigothic documents, one finds land-values expressed in gold, where presumably almost no-one had any; in Ireland, by contrast, only kings could pledge a gold object. But the idea of cash-compensation haunts many cases where authors have wished to speak of honour.

52 Stewart ‘Contract’ p. 246. The idea of tajbīr introduces complications here, but it does not, I think, invalidate Stewart’s point that a guarantor’s liability involves primarily his honour: vengeance, as much as payment, would discharge his duty to the initial victim.
Indeed, quasi-contractual links only illustrate a more general priority of status over sanctions. The implications for law of this stress on honour (specifically on
ʿirḍ) as defining liability and duty will be footnoted later.

53 Miller Bloodtaking p. 276 of a society where money-compensation was in practice common, or at least was much argued over. Obviously a great deal might end up at arbitration rather than at a formal council or thing, giving “law” in mediaeval Iceland a very limited meaning. The law itself agreed at some stage that arbitration was binding. But this, and the provision that arbitrators who reneged on a commitment to give judgement should themselves be outlawed, appears less than conspicuously in Grágás: A. Dermis et al (eds.) Laws of Early Iceland part two (2000) pp. 204-6, 355-7.

54 The outlaw himself named his three places at the compensation court, held 14 days after the ruling of banishment: Dermis et al. Laws of Early Iceland part one p. 93. Lesser outlawry and full outlawry do not derive from one Icelandic term, but there seems every reason to see them as locally two degrees of the same practice.

55 Miller Bloodtaking p. 175. For the founding myth of escape from royal domination, ibid. pp. 13‑14. For Anglo-Saxon provisions, Whitelock (ed.) English Historical Documents; for Ireland, Kelly Irish Law pp. 222-4. An interesting feature of the latter cases, is that kings might retain in their households men whose rights at home were forfeit.

56 The tale from Ḥabshūsh of Nihm and Khawlān, which we mentioned earlier, in part three, is a case in point. One has to go to those who have provided refuge and invoke the principle that “no-one but a shameful person can accept a man who committed ʿayb”. Disagreement in such a case threatens serious insult of one tribe or section by the other.

57 A distinction is conventionally drawn between substantive and procedural or adjectival law. Sometimes one finds the latter described, in a way suggesting how the basic idea is centred on a body politic, as “delegatory”. Even Hart’s more subtle distinction between primary and secondary rules (H.L.A. Hart The Concept of Law, 1972, pp. 78-9, 91-2) fits uncomfortably with customary law here. But in so far so as primary rules are identified with sanctions (cf. ibid. p. 27) one would have to say that in ʿurf they come second to rules defining who is vulnerable to whom. As we shall see later, various “types” of rule need not form a self-defining whole.

58 See e.g. Wormald Legal Culture p. 63. Gluckman, of course, was a friend of the great J. Wallace‑Hadrill. But it is odd that his Judicial Process Among the Barotse (1955), which actually contains much of interest, has been forgotten and his Custom and Conflict in Africa (1956) remembered. The reductio ad absurdum comes at the end of Custom and Conflict with a chapter on “the bond in the colour-bar”, as if apartheid South Africa had somehow been sustained by a balanced tension and interlocking ties between black and white. What sustained the old South Africa less fortuitously, as Gluckman himself says, was the South African police and army.

59 This is not to say poverty causes violence: only relative poverty, compared against one’s neighbours and lacking a rationale, does that. Nor are the chronicle sources unproblematic. But the typical cases of local violence seem to be pillage of the unprotected and bitter dispute over external funds. One suspects that part of that bitterness would be removed by recognition of the parties as simply having worth in the world of states and powers.
Certain areas of the country, such as the Jawf, were and are notorious for endemic, fragmented violence, where the usual forms of containment are said to fail routinely. Unfortunately, state sensitivities make research difficult. The question remains, however, as to why the area south-east of al-Ḥazm towards Ma’rib has so much harder a time than, say, Baraṭ or Wādī Madhāb where customary law is similar.

60 al-Ḥasan Al-Hamdānī Ṣifat jazīrat al-ʿarab (D. H. Müller ed. 1968) pp. 194-5. For the war between Sufyān and Ḥāshid, below, Glaser ‘Arḥab’ pp. 1, 15, Dresch Tribes pp. 111-12.

61 A striking recent case was a dispute between two sections of Banī Ṣuraym, which went on for the best part of a generation so bitterly as to prevent children attending school. But not even the two whole “ninths” of the tribe became closely involved: a handful of villages on each side were left to snipe across their own land and ambush each other elsewhere. On an even smaller scale, at Baraṭ two neighbouring households were at odds for years. One side of the house was in effect a free-fire zone, and even strangers were ill advised to expose themselves; on the other, neighbours met in the shade as if nothing unusual were going on.

62 Settlement of violent dispute is important. But it often draws one to what Hart called the “bad man” approach, and thus to sanctions preventing wickedness. “Why should not law be equally if not more concerned with the ‘puzzled man’ or ‘ignorant man’ who is willing to do what is required, if only he can be told what it is?” (Hart Concept of Law p. 39). The conventional problem cases in text-books, of international law and so-called “primitive law”, both make more sense if this is recognized. In Yemen people often want ʿurf defined, even at formal arbitration, for quite other reasons than avoiding a blood‑bath.

63 The scrolls referred to can be very long, sometimes covering at least a century, the original issue being never settled. In “custom” that surprises no-one. Meanwhile, most people would say that sharīʿah is, by definition, superior to all forms of law. Its administration is not. Unhappy outcomes (or the lack of any outcome at all) are usually blamed on judges, but the problem is often more that the process involves antagonists with a style of argument they do not command and with people they do not know.

64 The supposed coherence of “rational” law was on this score contrasted by Weber with “qadi‑justice”, an image used later by Justice Frankfurter. It is well to remember that Weber used the phrase of English law. The Court of Chancery did not follow even its own previous decisions until at least the 17th century, “binding precedent” had to wait for proper law-reports in the 19th century, and equity in England was not formally aligned with common law until the 1870s.

65 Hart’s celebrated “secondary” rules of recognition, change and adjudication, and his suggestion that their union with primary rules is the essence of law (Concept of Law p. 151), place us firmly within this sort of world. The reference to Weber, below, is of course to the long eighth chapter of Economy and Society (Roth and Wittich eds., 1968, vol. 2). In retrospect it seems odd, in fact, that Hart left Weber so thoroughly aside.

66 For the quote following, F.L. Cheyette, Suum cuique tribuere, French Historical Studies vol. 6/3 (1970), p. 293. For useful background on this kind of discussion, W. Davies and P. Fouracre (eds.) The Settlement of Disputes in Early Medieval Europe (1986).

67 The French distinction was geographical, some regions claiming more immediate reference than others to Roman law, and the pays du droit coutumier sometimes revelling in local difference. As Maine remarked, this encouraged an interest in relations between civil and natural law, not least on the part of Montesquieu. But twisting the original meaning of the terms here seems justified.

68 Cheyette ‘Invention of the state’ p. 161, cf. Wormald Legal Culture p. 23, and English Law p. 95, where the prevalence of simple conditionals in Aethelbert is noted. Cheyette’s argument that a particular type of abstraction links Western law to theology and philosophy, and thus to conceptions of the state, is persuasive. A very different style of coherence was sometimes sought for Islamic fiqh: see e.g. Wael Hallāq, The development of logical structure in Sunni legal theory, Der Islam vol. 64/1 (1987).

69 The form of pacts is shown by the texts translated earlier. An idea of more generalized law is given by the Bin Zinbāʿ text in Abū Ghānim’s Bunyah qabalīyah. But the fragments I have of Serjeant’s Kitāb al-ādāb wa-l-lawāzim (they are only fragments, unfortunately) suggest extraordinary detail on duties and obligations, and links with Islamic reasoning. Even to guess at the style of the work overall would be premature.

70 For these practical distinctions and the forms of rhetoric see e.g. Dresch ‘Episodes in a dispute’, where “the laws of Ḥāshid and Bakīl” are invoked by a plaintiff. An attempt to specify what exactly those are would be effort misdirected.

71 Cf. Cheyette ‘Suum cuique’ p. 287. The English term is notoriously all-inclusive. In French what one practices or appeals to may be la loi yet a student of law is said to faire son droit. In English, “law” covers both and much else besides. All the major languages of Europe, despite this, seem in practice to work with the same indeterminacy of reference. The suggestion that “law” can be simply the world as seen by lawyers is not intended frivolously. See e.g. B.Z. Tamanaha A General Jurisprudence of Law and Society (2001).

72 Miller Bloodtaking p. 228. For the earlier points about kingly law, Wormald Legal Culture chapter 1, English Law chapter 2. For Ireland, Kelly Irish Law. Stewart’s Sinai material provides a non‑written case of what has to be admitted as in a full sense “law”.

73 Kant, of course, built an approach specifically to ethics around the categorical imperative that one only act on maxims one can will as universal. In the secondary literature on law (not in Kant, I think) this degenerates into instrumental arguments: only will rules that will not contradict your interests if the rules are generalized to others. Far more to the point for an anthropologist is the original idea of universalizing statements.

74 Concept of Law p. 155, discussing specifically the idea of justice. For all that rules are not the whole of social life (perhaps not even a large part), talk of “law”, or even “legal anthropology”, without a focus on rules is simply talk of blinkered sociology.

75 The law of nations, as ius gentium, was for Rome first a lowest common denominator, inferior to civil law. Mediaeval theorists, in a quite different mood, seem almost to equate it with natural law. Late in the renaissance it resurfaces as ius inter gentes, governing how states interact, what the Romans might have called ius fetiale and what the moderns called droit des gens. What I am thinking of here is the last of these in the period, say, 1645-1945, from Grotius to Nuremberg.

76 L. Oppenheim International Law vol. 1 (1958) p. 38. Higher rules can as well be called basic rules, and one sometimes find experts in international law writing of ius cogens, law that all acknowledge or which “gathers” all, something like natural law. But what this is appears difficult to pin down: we might all agree there are human rights, for instance, but what exactly they are is only specified by treaty. This is very much the position with tribal ʿurf.

77 S. Merry, Legal pluralism, Law and Society Review vol. 22 (1988). In “modern” states, however, such pluralism exists within an ultimate claim to legitimacy, which has real-world effects. For example, in Southern California a great many people seem to be driving cars without insurance (often without a license), and a whole folk-jurisprudence governs how to get one’s fender fixed. In following that accepted system one presumably places oneself in a vulnerable position as regards California law, by e.g. “not reporting an accident”. The state at any point can define the private pursuit of equity as wrong.

78 L. al-Zwaini, Mediating between custom and code: Dâr al-Salâm, an NGO for tribal arbitration in Ṣanʿā’ in B. Dupret (ed.) Le Shaykh et le procureur (2005), pp. 327-9. The organization al-Zwaini examines in her paper hopes to codify custom (ibid. p. 314), thus inviting the self-contradictions involved in codifying sharīʿah. But they are at least left to pursue their project. In other cases, implicated more deeply in the problems of e.g. vengeance, attempts to formalize arbitration meet suspicions of political subversion.

79 Tribalism” is my abbreviation of the arguments, not a local term (although reported from western Yemen and from Khawlān al-Ṭiyāl, qabyalah is not common usage in most places north of Ṣanʿā’). In arguments of this kind, what people of all sorts now often say, symptomatically of modern conditions, is “the tribal system” (al-niẓām al-qabalī).

80 No-one who has read Mauss will be surprised at the aligning of dispute and trade.
Quite widely in the Middle East, however, we find exchange being treated as inherently problematic. It used to be that direct exchange at market had to be mediated by non-tribal “weak” people. See e.g. Dresch ‘Mutual deception’ pp. 126-7.

81 Changes in what some call advanced democratic states have provoked much literature on “governmentality”. Public-private distinctions are judged outmoded. Yet the 18th century ambition expressed in John Adams’s splendid line “a government of laws, and not of men” encounters a paradox: regimes of total law, whether mediated by commercial institutions or governmental administration, leave no private space for moral judgement. Right action often comes to exist in a domain akin to that of reparations from uninsured drivers in the Southern Californian case.

82 As the constitutional (administrative) state took increasingly solid form in Britain, at the turn of the 19th century, such authors as Elphinstone and Burckhardt, with Tacitus and the Roman republic in mind, found political virtue in the tribal worlds of Afghanistan and Arabia. The tyrannical opposite they found in respectively lowland India and the Turkish empire. A generation before, in Kant for instance, the tribes of Arabia were imagined, with equal freedom, as examples of quasi-savagery. It may be that their history again demands thought from those whose experience is wholly of state order.

83 To appeal to lack of evidence is to raise the possibility merely of weak ethnography.
But no account of Arabian tribalism, over a span now of some century and a half, suggests a fixed scale of values such as is commonly reported elsewhere. Does “honour” encompass “religion”, for example, or vice-versa?

84 A value-judgement would be premature and in any case irrelevant to understanding.
But here, analytically, is where we find our clues for the guess made earlier that the specifically “bourgeois” state is at its end. The non-public space par excellence, the family, no longer has, for better or worse, any obvious legitimacy: intervention there is no longer considered an exceptional necessity but a moral imperative.

85 One would not contest the status of international law with experts, and the principle that it applies to individuals, not only to states, was enforced at the end of World War Two. Nonetheless political appeals to such law as over-riding national sovereignty have blossomed, not coincidentally, since the collapse of a balance between blocs of states circa 1990. In popular discourse almost any act of mass brutality committed elsewhere is now called a war crime, and certain governments are willing to use such language.

86 For the quotations following, E. Gibbon The History of the Decline and Fall of the Roman Empire (1895) vol. 1, p. 109.

87 Judgements of this kind are never mere reports on empirical similarity or difference.
In passing one should note Gibbon’s view of language, so very much at odds with recent analyses of ethnicity and nationalism. Gibbon’s French was excellent, but he is not under the illusion that everyone who mattered spoke French or English. What he might perhaps have claimed is that certain assumptions about the world were broadly shared.

© 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