ABSTRACT

Traditional Somali politics was based on two dialectically related principles: kinship and contract (tol iyo xeer). Kinship was founded on the segmentary lineage system under which people traced their descent to common male ancestors. Agnates functioned as corporate political groups because they were blood relatives. But the blood relation was not sufficient to establish a political system. Agnates functioned as corporate political groups because they negotiated a social contract that defined the spheres of their collective unity.

Text:

Traditional Somali politics was governed by two dialectically related principles: kinship and social contract (Lewis 1961). Although kinship has received a great deal of attention in the literature produced about Somalia since the collapse of the state in 1990 (Clarke 1995) the social contract has not. Yet it is hardly possible to understand kinship fully without taking into account the social contract, which is the legal charter, so to speak, of the kinship system. The kinship system is based on blood relation, but the ties that bind blood relatives are grounded on social contract--on a public system of rules publicly negotiated. Without an understanding of the public rules of kinship, it is impossible to fully grasp the form and content of Somali politics. Nothing illustrates the lack of full comprehension of Somali politics better than the abject failure of the various international conferences held to resolve the Somali political crisis. The aim of the conferences was to resolve the crisis by distributing political office to the clans according to strength, naively assuming that the Somali political system was based only on clanship. Clanship is the first principle of Somali politics. But the second principle--the social contract--is equally important, because it establishes the rules of clanship. Clanship, in itself, is 'imaginary' as Ibn Khaldun pointed out. This article attempts to bring back into the analysis of Somali politics its traditional governing principles.

The article argues that traditional Somali politics was governed by two dialectically related principles: tol iyo xeer (kinship and contract) (Lewis 1961: 3, 126, 193). Kinship was founded on the segmentary lineage system under which people traced their descent to common maie ancestors. The use of the word 'tribe' is inappropriate in discussing Somali kinship because in the anthropological literature from the seventeenth century onwards 'tribe' meant small, centralized or acephalous groups, each with unique characteristics such as territorial boundaries, distinct origin and language (Lewis 1961: 2). A more appropriate nomenclature is the division of Somalis into corporate political groups: clan-families, clans, sub-clans, primary lineages, and dia-paying groups (Lewis 1961: 4). The clan-family-such as the Darood, Hawiya, Isaaq, or Dir--was the upper limit of clanship, while the dia-paying group which Somalis call jilib--which extends from five to seven ancestors--was the lower limit of the system. The dia-paying group was the most important corporate political group. It consisted of people who shared the most important responsibilities such as the payment of blood money, common defence, and common help (Lewis 1961: 6).

Although the jilib was the foundation of the whole system and was the point at which 'a man most frequently acts', a central feature of the Somali political system was the 'relativity of its political divisions' (Lewis 1961: 6-7). Corporate political activity, in other words, existed and developed at other levels of the segmentary system because every ancestor was a point of 'potential division, and of unity' (Lewis 1961: 7). The segmentary lineage system had no 'stable hierarchy' of political or administrative offices (Lewis 1961:196). At all levels of the corporate political groups, the leaders were the elders, which technically meant all adult males, since all adult males were 'classed as elders (oday, pl. odayaal) with the right to speak at the council (shir) which deliberates matters of common concern' (Lewis 1961: 196). At the dia-paying moot as in the clan moot, 'All men are councillors, and all men are politicians' (Lewis 1961: 198). Although all men were equal inside and outside the moot, not all men had the same level of influence and power. The difference in influence and power was conferred by the society according to specific standards of virtue. The foundation of the kinship system was the social contract (xeer). Agnates were corporate political groups not just because they shared a common ancestor but because of explicit treaties between them that defined the terms of their collective unity. These treaties had 'contractual elements having close affinities with those political theories which saw the origins of political union in an egalitarian social contract' (Lewis 1961: 3). The treaties defined 'people's relations at the level of segmentation represented by the dia-paying group. So that in this group, which is in effect the basic political and jural unit of pastoral society, contract and clanship meet to give it its specific character' (Lewis 1961: 6). They galvanized 'the diffuse and manifold bonds of kinship at any point and through any ancestor, giving rise to opposed political units' (Lewis 1961: 3).

Politics in traditional Somali society had this two-fold character: agnation held by the explicit ties of the social contract (Lewis 1961:127). This article deals with that issue. It is divided into two sections: the first section examines agnation; the second how the ties created by the social contract make possible the functioning of the relations of agnation. Agnation is not in itself sufficient to create an effective political system.

Traditional Somali society was fiercely republican. As a Buganda police escort disapprovingly told Drake-Brockman: 'Somalis, Bwana, they no good; each his own Sultan' (Drake-Brockman 1912: 102). Although 'each his own Sultan', Somalis had leaders as well as political and judicial institutions. The leaders had no power of coercion even when some carried impressive titles such as Sheikh, Ugaas, Garaad or Sultan. They had no more power than an elder (Lewis 1961: 206-7, 241). As Ibn Khaldun pointed out, there is a radical difference between leadership and kingship. Leadership (riyasa) was based on the voluntary loyalty of people to certain men because they admired certain qualities about them; such leaders, however, had no power of coercion (qahr) to exact loyalty. They had the symbols of power--respect, followers, ability to wage war and conclude peace, etc.--but no titles. Kingship, in contrast, is based on titles acquired through conquest (taqalub) and coercion (qahr) (Ibn Khaldun 1999: 244). Somali society had leadership but no kingship. And their leadership system had a clear public system of rules based on kinship, virtue and contract. Though all men were politicians by right, men had difference in status determined by wealth, poetic gift, political ability, talent for oratory, knowledge, piety, wisdom, hospitality, courage, tactical acumen in war and personal strength (Lewis 1961: 196, 198). These virtues, so to speak, gave a person prestige and status in the society; except wealth, none could be inherited. Moreover, the society had institutions with a clear public system of rules. The most important institutions were the shir (Political Council) and xeerbeegti (Law Council). In the shir political issues were settled, and in the xeerbeegti legal disputes were mediated.

The political council was the key political institution in the country. The council could be held anywhere: the shade of any large tree or any other place of convenience would suffice. Although there may be a 'great deal of argument and wrangling, all those present are expected to behave courteously and breaches of good manners may be punished' (Lewis 1961: 198-9). The political council may be held by the dia-paying group or the clan, and it may meet to discuss various issues: it may recognize the promulgation of a new dia-paying group, adapt a peace treaty, plan war against another group, appoint a war leader for such war, or discuss any other issue of relevance to the group (Lewis 1961: 199).

The second important institution was the law council. The key function of the xeerbeegti was, like the shir, to resolve conflict, albeit legal conflict. The council dealt with legal disputes that concerned dia-paying groups or lineages or clans or individuals, including issues such as murder, insult, injury, divorce, inheritance, theft, robbery and affray. In the traditional legal system justice was done and was seen to be done, as it were, behind 'a veil of ignorance' (Rawls 1973: 11). Ideally no one was given advantage in the system, even when the disputants were not equal because of natural talent or wealth or personality or lineage affiliation. The law council was supposed to listen to both sides (Lewis 1961: 229-31). The muduci (plaintiff) and mudaacaly (defendant) had the burden of proving their case. Each could either speak for himself or herself or appoint a lawyer (afhayeen) and, if the case was complex, the lawyer would also have an assistant lawyer (hojiye) whose task was to remind the lawyer about the key issues in the case--if he forgot any. In complex cases the law council also appointed a recorder--ladhaye or doodqaade--to record the proceedings orally for the benefit of the council, reminding them of the main points. The council based its judgement on the evidence submitted to it; hence in reaching a judgement it depended heavily on the presentation of the plaintiff, the rebuttal of the defendant, and the testimony of witnesses. That was why rhetoric--the ability to persuade--was important in Somali society and politics. The council consisted of four members to make sure that, in case a split occurred, a simple majority would not impose its view. (1)

The Somali oral tradition is full of stories about the sort of men that were admired in the society, as illustrated by 'Baroordiiq' (Lament) a poem composed by Cali Jama Haabiil (Haabiil 1993-4; Andrzejewski 1993: 24-6). The poem is a eulogy for the poet's friend Xirsi Cilmi Goonle. He describes Xirsi in successive stanzas as a man of excellence in war, judicial disputes, hospitality, political council, and service to the community. These were the virtues (moral and intellectual) as Somalis admired, understood and practised them. For Haabiil, as for Aristotle, the virtues were the result of 'purposive disposition' (Aristotle 1953: 1107a)--of ethos (Aristotle 1953:1103a15)--since they are within 'our power' (Aristotle 1953:1113b5). The aim of the virtues is the doing of 'good' (Aristotle 1953: 1094a) and the reward for doing 'good' is honour (Aristotle 1953: 1124a). Xirsi Cilmi Goonle is honoured by the poet and by the society because of his virtues. Though Xirsi is not poor--since he is capable of liberality in hosting guests--wealth does not distinguish him. Wealth, qua wealth, does not confer honour; wealth confers honour in the way it is used publicly. What distinguishes him, then, is his courage, ability to be just and persuasive in legal disputes, contribution to political discussions in the political council, liberality, sophistication (he is able to entertain his guests without resorting to vulgarity), and patience with the young and the old. He is a statesman because he leads free and equal men, men who are neither dependants, nor slaves, nor subjects (Aristotle 1962: 1255b16).

He is honoured, thus, because he combines, to use Lewis's words, the 'ideals of Somali belligerence, independence, daring and panache' (Lewis 1961: 198) as well as sophistication. Somalis call that kind of person hal-karaan (a capable man; a man capable of at the least one of the virtues), or, in an ironic way, belaayo (Lewis 1961: 198) (disaster or catastrophe to his enemies). Such persons were accepted as leaders not just because they were virtuous members of society but also because they 'hosted' the interests of the community (Ibn Khaldun 1999: 334-5). This is what politics is all about. This is a point about which Ibn Khaldun and Aristotle- whom Ibn Khaldun admired as 'the first teacher' of philosophy--agreed: politics is about the 'good' (khayr) (Ibn Khaldun 1999: 251; Aristotle 1953: 1094a).

Colonial administrators did not understand traditional Somali politics. They believed that Aqils (chiefs) ran traditional Somali society, and that they ran it along 'tribal' lines. The chief task of the administrators was to identify the 'tribes' and the Aqils that rule them, and then to use the Aqils as intermediaries between the administration and the 'tribes'. Indirect rule depended on the successful completion of that task. The administration failed for three reasons. First, the administration failed to win over the confidence of men of influence and authority among the leaders of the corporate groups. Second, the Aqils that already existed and which the government attempted to transform into chiefs in the tradition of the warrant chiefs were not the traditional rulers of the country and they were without influence and credibility among the rural people; and those who had such influence were reluctant to identify themselves too closely with the government. Third, there were no 'tribes' in the country and no tradition of 'tribal' hierarchy and cohesion at a level that the administration could mobilize (Mohamed 2002). This puzzled British officers. They assumed that since the society was 'tribal', the Aqils must have full authority to govern and mobilize the clan families. Commissioner Cordeaux blamed the failure on 'detribalization', which he assumed was triggered by the existence of a 'higher authority' than the Aqils that had reduced their standing and power (Cordeaux 1906).

Cordeaux was determined to reverse the process and restore their 'influence and authority' (Cordeaux 1906). He decided to create a 'strong and elastic form of organization' that would yet again tie together the 'tribes' 'by the common tie of mutual interest' (Cordeaux 1905). That organization would have both administrative and military aspects: administratively, it would consist of a civil authority run by the Aqils; and militarily of 'tribal militia' that would revive 'the "esprit de corps" of the tribes' (Cordeaux 1905). This double process would create a 'political body' (Homby 1905) that would make it possible for the Somalis to govern themselves for the benefit of the colonial power.

The trouble was that rather than building up society along 'tribal' lines, the creation of Aqils and other paraphernalia of indirect rule actually 'weakened the stability of dia-paying groups' because it created 'artificial positions of leadership' (Lewis 1961: 203) such as Aqils and Sultans who had no traditional roots although they have since become 'traditional' (Hudson 1952). The only clan with a long tradition of the institution of Sultan was the Gadabursi; the Isaaq never had it in the pre-colonial era. The Gadabursi gave their sultan the title of Ugaas. Even so, the position was never stable. It lapsed at some time in the nineteenth century as a result of rivalry between different claimants who could not achieve a following among the population; and when the Gadabursi tried to revive it in the colonial era they could not (Lewis 1961: 204). Whether they were called Sultan or Ugaas or Garaad, men who held these offices never had any institutional authority and their influence was no greater than that of any influential elder. In the colonial period the number of Sultans--like the number of Aqils--increased, probably at the behest of the administration. Many clans and lineages--such as the Isa, the Habr Awal (Isaaq), Arap (Isaaq), Idagalla (Isaaq), Habr Jeclo (Isaaq), Dhulbahanta (Darood), and Warsangeli (Darood)--elected their own Sultans. By multiplying political offices such as Aqils and Sultans, indirect rule created 'artificial political positions' that sowed jealousies and rivalries among the elders and thus weakened the stability of kinship groups. This process was further intensified in the late colonial period when the colonial bureaucracy expanded dramatically as a result of post-war imperial development policies.

During the late colonial period, imperial policy makers believed that the acceleration of development initiatives 'would make colonies simultaneously more productive and more ideologically stable in the tumult of the postwar years', and so 'sent waves of experts to Africa to refashion' all aspects of colonial life. Hitherto the district commissioner, the governor, the native commissioner and the army officer were the dominant figures in the colonies. But in the late colonial period the Whiteman-as-Expert made his appearance- the expert in agriculture, irrigation, livestock husbandry, soil and forestry improvement and protection, public health, formal and informal education, institutional development, and political reform. Imperialism in late colonial Africa 'was the imperialism of knowledge' (Cooper 1997). Political reform was central to the new imperial project of modernization: its aim was quite simply to establish 'local authorities based on traditional institutions' (Hudson 1952). The idea was the same as that articulated by Lord Lugard in the 1920s: indirect rule was supposed to create a community of interests between the Somali elite and colonial rule. Unlike the early colonial period, when the emphasis was on establishing chiefs, in the late colonial period the emphasis was on establishing modern institutions such as the Protectorate Advisory Council (1946), Local Government Councils (1950-4), Legislative Council (1957), and Executive Council (1959). The institutions were supposed to be modern, yet they were to be built on a 'traditional' basis.

In 1941, the 'question of Indirect administration through tribal or other Somali Leaders occupied the attention' of the government (Chief Secretary 1948). The most urgent task of the government was the displacement of the Aqils. They were unpopular with administrators and with the rural population. The rural population viewed them as government agents with no 'inherent authority' (Hudson 1952). Colonel Jameson, the Chief of Civil Staff in the East African Command, dismissed them as a 'by-word for bribery, corruption, and intrigue'; they depended, he said, 'for their livelihood on promoting personal and tribal discord, for the settlement of which they extract large fees' (Jameson 1943). He quoted an Aqil who wrote a letter to a client in Aden in which he urged the client to send more money, 'because, I must let you know, justice is a matter of commerce' (Jameson 1943). G. T. Fisher, the governor of the country from 1941 to 1948, was also sceptical about the usefulness of the Aqils. They were generally corrupt and unpopular, and the district commissioners in 1943 'reported unfavourably on them' (Fisher 1943). The 'existence of such a body, that is known to be corrupt, as part of the machinery of Government cannot fail to bring the Administration into disrepute, and so weaken constituted authority' (Fisher 1943).

He thought that the Jawarbadars--unpaid representatives of some sections of the lineages--were more useful and effective 'because they owe nothing to the government, and realize that they depend for their livelihood on the straight-forward dealing with the tribes' (Fisher 1943). But the Jawarbadars were not enough. They were too few, too traditional, and too independent. What was needed was a modern elite 'invested with natural powers of leadership' and 'the necessary education and training to carry out an administrative policy acceptable to World opinion and modern requirements' (Chief Secretary 1948). The only group that could approach such high standards were the civil servants and traders 'with rudimentary education or a local reputation for possessing such knowledge' (Chief Secretary 1948). Although this 'new governing class' (Chief Secretary 1948) was to be the fulcrum of indirect rule, the Aqils and Sultans were not to be excluded, since indirect rule would establish modern institutions along 'traditional' lines.

The key modern institutions established by the government for political reform were the Protectorate Advisory Council, Local Government Councils, and the Legislative Council. These had different functions. The Protectorate Advisory Council had no executive role; its role was merely advisory. It held meetings several times a year which were rotated between towns. The institution was supposed to be national and representative of all sectors of society and all 'tribes', and so those who took part in its activities were civil servants, traders, Aqils, Sultans and other prominent figures in the society. Unlike the Protectorate Advisory Council, the local government councils had executive powers, albeit very limited. They were to be dual in form: councils for the countryside and councils for the urban areas. In the countryside, the local government councils would replace the Aqils. The new local authorities there would, like Lord Lugard chiefs, have executive powers such as the power to tax, to hold court, and to run a police force. In the towns, the local government authorities would run the municipal affairs of the towns. The Information Office hailed the policy with some exaggeration as 'the most prominent landmark in the history of Somaliland' (Information Office 1951) since it was supposed to prepare Somalis in 'suitable stages' for 'legislative functions in the same way as Town Councils or County Councils in England have served as excellent training schools for many men who later became distinguished members of Parliament' (Information Office 1951).

Although the policy was successful in the towns, it was a total failure in the rural areas. It met with widespread resistance for three reasons: the public objected to the idea of giving powers that the Aqils never traditionally had; the public rejected the attempt to make the Aqils into chiefs of lineages rather than dia-paying groups--in essence expanding their role; the public suspected that the government was attempting to expand its powers and entrench its hold over the country. Various rumours justified the resistance: 'It is widely thought by the country people that under the new Ordinance nomadic pastoralists would not be allowed to move in search of fresh grazing without the permission of the Government'; that government control would be so complete that 'in the future all pregnant women will have to register themselves with the new authorities and it has even been said that Government officials will keep registers of sexual intercourse!!'; 'that the Government is making this reorganization merely to put into power their own men who will enable some new form of taxation to be enforced and this will inevitably lead to slavery and degradation' (Reece 1951). R. S. Hudson witnessed a dia-paying group meeting held specifically to protest against the policy (Hudson 1952). The resistance to the policy forced the government to cancel the attempt to create chiefs and reappoint the Aqils in early 1951 as local authorities for rural areas.

In the towns, the government established local government councils in Berbera and Hargeysa in early 1951 by appointing its 'own men' (from a pool of traders, civil servants, and Aqils). In 1953, the local government councils were extended to Burao, Erigavo, Borama and Las Anod. Since the aim of the policy was to establish local authorities based on traditional institutions, the government not only incorporated Aqils into, but maintained 'tribal' balance in the councils (Lewis 1958). The same principle was followed in the establishment of the Legislative Council. In February 1955, an Order in Council signed by Her Majesty was published in the country, which provided for the transformation of the Protectorate Advisory Council into the Legislative Council, and a 1955 White Paper announced the explicit arrangements made for the creation of this body (Colonial Office Reports 1954 and 1955). The Legislative Council was opened in May 1957.

It consisted of a majority of official members and a minority of unofficial Somali members (War Somali Sidihii 1957). Although it was the governor's choice to appoint whom he wanted, he made a show of consulting the Protectorate Advisory Council in the selection process, since the Legislative Council was supposed to replace the former body. The governor selected 24 candidates for the Legislative Council from the Protectorate Advisory Council, and asked the advice of Protectorate Council members to submit six candidates out of the 24 for appointment to the Legislative Council. An important criterion for the nomination to the council was 'tribal' balance (Lewis 1958: 152-3). The Council was convened on 21 May. Sir Theodore Pike stated in his inaugural address to the Council that this legislative body had been established in order to introduce Somaliland 'into the great and growing tradition of parliamentary democracy' (Lewis 1958).

The modern political elite of the country emerged out of these three institutions. The problem that the administration faced was two-fold: how to incorporate the elite into these institutions and how to make the institutions representative. The first problem was easy: the government appointed them to the institutions; but the second problem was difficult, for it concerned the criterion for making the institutions legitimate and democratic. This was a problem that the administration could not solve. When the local government councils were established under the 1950 ordinance, the key question that confronted the government was: should the councils represent the population geographically or 'tribally'? In other words, since in each district various 'tribes' lived together and since the 'tribes' were also dispersed geographically over large areas, how was the representation to be made? This question was the recurring motif in all the discussions about making the institutions representative. Gerald Reece stood on one side of the debate, the Colonial Office on the other. The Colonial Office feared that the administration was creating a system of direct rule rather than a system of indirect rule.

For instance, Wallis, a senior research officer in the Colonial Office, argued that the local government ordinance did not 'associate the people' closely 'with the management of their own affairs' because it was based on the Kenya system (1937 Native Authorities Ordinance) and would thus lead to 'straightforward headman arrangement'. There is a difference 'between the headman/agent concept belonging to direct rule and tribal/local authority concept belonging to indirect rule'. The building of the authorities should be from the bottom up: 'Only when the local people have proved in what way and in what combinations they can really do what is required of them will the time have come for powers to be vested directly in them' (Wallis 1950). Lanham agreed. The 'draft ordinance is a half-baked affair which will perhaps produce the outward form of local government but which is not likely to produce the inward spirit to make it work'. The government, therefore, should reconsider the issue (Lanham 1950). The Colonial Office was convinced that Reece's idea would lead to direct rule, which would be 'very expensive and will soon lead to a final breakdown of tribal organization' (Jerrom 1950).

Reece insisted that 'if it were possible to reorganize these people entirely along a tribal basis nothing would please us better, nor be more suitable and simple'. But it was not that simple. The problem was that the 'tribes and the sections are permanently intermixed and intermarried'. In the Borama District, 'it would be impracticable to separate the Issa and the Gadabursi, who are intermarried and at all times intermixed geographically'. It was the same in the Hargeysa and the Burao districts, where it 'would be quite unworkable to have separate councils for the Habr Yunis, Habr Toljaala, Habr Awal and Eidagalla tribes'. In the Erigavo District, too, there were no pure 'tribes' that were isolated from others, except the Wersangli. It was possible to 'have a Warsangeli tribal council', but you could not have pure Habr Jeclo or Habr Younis councils. Besides, the object should not be separation but rather the intermixing of the 'tribes' so that they could 'sit together and agree sensibly about sharing water and grazing and combining in schemes for grazing control and soil conservation'.

'We want to retain', he added, 'all that is good in the existing tribal life and tribal organization ... but it is utterly impossible now to sort them all out and compel them to live again as separate tribes and sections--even if this were really advisable from the point of view of peace and progress.' The idea, furthermore, of giving the tribal authorities judicial powers--though that is exactly what the ordinance did--would be a 'retrograde step, and it was inconsistent with the idea of subordinate courts which Fisher started, and with which you (Colonial Office) agreed, and of which the Protectorate Council are in favour'. One of the reasons Fisher introduced the subordinate courts was the 'corruption of the old Akil courts which were composed of tribal elders'. To revert to that corrupt system would be anything but progressive. Lugard and Hailey suggested that 'judicial powers give tribal authorities enhanced prestige'. But 'except amongst very primitive people, the idea nowadays is surely to separate the executive from the judiciary?'
"It seems to me that in London there is a tendency to think that because most of these people live (and will always live) as pastoral nomads, they are still very simple and primitive and unchanged. That is not so. Conditions have altered a good deal in recent years and even in the wildest places one now finds some seamen from Cardiff and ex-soldiers from Burma. Communications are much more rapid and the standard of knowledge of World affairs is often surprising". (Reece 1950)

Though 'in many respects the people are very backward, all the old administrative systems which we applied successfully to very primitive and unspoilt African tribes are not now always suitable'. New ideas and systems of government were necessary. Of course many of the 'petty affairs' of the nomads would be 'settled out in the bush by tribal elders in accordance with their old systems'. Nonetheless, there was a need for ideas that would take into account modern changes: in short, the organization of the councils should be on a geographic rather than a purely 'tribal' basis (Reece 1950).

J. Griffiths made the response of the colonial office to Reece. Griffiths recognized the 'difficulties in the way of organization of the Somali people on a purely tribal basis'. There is obviously a necessity in forming 'groups composed according to the family ties of the tribes, geographical position and other identity of interests'. The problem was that such groups 'would not be tribal authorities in the accepted sense of this designation, and that they would not bring to their duties a tribal tradition of authority giving them traditional powers to which the local government powers which you would propose to give them by legislation would be supplementary'. But since such traditions were unavailable for use, greater emphasis must be placed on the 'younger and more progressive elements in the community'. The administration must solicit the 'advice and assistance of suitable persons who would be unpaid and who would, together with the paid local authorities, form a group council responsible for the matters prescribed in the draft legislation'. This would in effect
"make use at the lowest level of local government for the services of the public spirited persons of knowledge and experience whose advice would, I suggest, be of considerable benefit to the local authorities in the performance of their duties and should assist them in the goodwill of the community towards their work". (Griffiths 1950)

Reece agreed that the 'young and the sophisticated Somalis ... must share the future Government of this country with the nomadic elders' (Reece 1950). But how to create such a collaborative framework between traditional elders and 'sophisticated' politicians? And anyway, who were the 'nomadic elders'? As one document on 'native administration' queried a bit breathlessly: would it 'not be possible, outside the towns, to search diligently and find the traditional head or heads and their traditional advisors'? (Native Administration 1951). There was utter confusion about the existence in the country of 'some mysterious form of tribal organization' (Reece 1953b). When the government decided to make the Legislative Council representative in 1958, that 'mystery' became a difficult problem to unravel. The government appointed two commissions to investigate the issue of representation: the commission to investigate constitutional advance was appointed in February 1958 (Executive Council 1957) and the commission on representational reform in April 1959.

The 1958 commission to investigate constitutional advance consisted of N. P. Carrick-Allan, the Attorney General, who was the chairman of the commission; two official representatives, E. M. Wilson and N. L. Thomas; and two Somali representatives, Michael Mariano and Mohamoud Haji Ahmed Ali. The remit of the commission was as follows: how many representatives should be elected? Who should vote? Should women vote? What should be the voting age? Should the vote be restricted to persons born in Somaliland? What of those of local nationality but born elsewhere? Should there be a residence requirement? Should there be any other qualifications to vote? How is the voter to be registered in a pastoral society? Should voting be direct or indirect (electoral college)? How would constituencies be determined? The last question was perhaps the most difficult: was the constituency to be based on 'tribe' or territory? 'The inescapable fact', the Attorney General wrote, 'is that at present tribal loyalties amongst the Somalis are very strong and this is likely to be the case for many years to come. That being so voting will be primarily on a tribal basis'. He added it is 'better to face the facts squarely and to base the constituencies on tribes. In other words tribes A and B elect one member of the Council.

C, D and E elect another, and so on.' But how would 'tribal' representation work geographically, since the 'tribes' were dispersed over wide areas and since they were intermingled (as Gerald Reece insisted in 1950), and since representation must be based geographically either on a town or a district or both? The Attorney General was confused. He believed that a distinction could be made between geographically and 'tribally' based elections. Yet he felt that the distinction was hard to make in practice. Unable to resolve the issue, he concluded that elections could be 'organized on a tribal basis although apparently on a geographical' ground. Tribes are geographically based, he reasoned, so if the elections were organized 'tribally' they would also lead to geographical representation (Carrick-Allan 1958). The commission published its report in June 1958.

It recommended the election of twelve Somali members to the Legislative Council on the basis of universal male (not female) suffrage. The vote was to take place through secret ballot in the towns, and public acclamation in the rural areas (Lewis 2002: 153). The seats were to be distributed to the six districts: Hargeysa, Borama, Burao, Erigavo, Las Anod, and Berbera; the district representation--which was geographical--also satisfied 'tribal' representation since the 'tribes' were based in districts. In March 1959, elections were held for the Legislative Council in the towns and rural areas. The Legislative Council consisted of thirteen Somali elected members; three nominated to the Council by the governor; fourteen official members; and three ex-officio members. Despite the elections, colonial officers were still in the majority. The February 1960 election would give elected members the majority, although in the Executive Council colonial officers would retain their majority until the day of independence.

The 1959 commission on representational reform was chaired by A. L. Scawin. The other members were Jirdeh Hussein, Mohamed Haji Ibrahim Egal, Hassan Jama and Ibrahim Adan. Its remit was to enquire into the expansion of the number of seats in the Legislative Council, the method of selection of unofficial members, the method of election of elected members, the division of the electoral districts, the qualification and disqualification of electors and candidates, and the place of 'traditional tribal authorities in relation to the Legislative and Executive organs of Government and how these authorities may best be associated with the future working of the Constitution' (Scawin 1959).

The four areas of importance were the form that the election should take, the number of seats to be contested, the distribution of the seats, and the role of the 'traditional authorities' in the politics of the country. The commission recommended that the elections be divided into urban and rural elections. In the rural areas, the 'tribes' were to choose their representatives; in the urban areas, the urbanites were to choose their representatives. In the urban area voters were to be registered and use a secret ballot. In the rural areas, voters were to use a secret ballot but were not to be registered since there was insufficient time to take a census of the population between May 1959 (the date the commission finished its report) and February 1960 (the date the commission recommended elections be held). The commission recommended that 33 seats be contested. These were to be distributed between towns and rural areas, with the following breakdown in the towns: Hargeysa Township, 3 seats; Burao Township, 2 seats; Berbera Township, 2 seats; Borama Township, 1 seat; Erigavo Township, 1 seat; Las Anod Township, 1 seat. Of the rural districts, Hargeysa, Burao, Borama, Erigavo, and Las Anod each had 4 seats, while Berbera, because it was smaller, had three. The distribution of the seats to the rural areas and the towns was supposed to guarantee 'balanced representation' geographically and 'tribally'.

To some extent the number of seats (33) recommended by the commission was arbitrary. Many people wanted the Somaliland Legislative Council to have the same number of seats as the Somalia Legislative Council, which had 60 up to 1959 and 90 thereafter. But this was considered unnecessary. The commission recommended the smallest possible number of legislative seats for the country- 33 seats. Like the 1958 commission, the 1959 commission recommended that only men should vote and run for office. The commission also recommended that 'traditional authorities' play a key role in the politics of the country. But the commission's formulation was ambiguous: who were such authorities--the Aqils and Sultans, or urban politicians who were anxious to find a reliable constituency? The commission did not say. Unlike the 1958 commission, the majority of the members of the 1959 commission were Somalis; some of them were political heavyweights like Jirdeh Hussein, one of the richest men in the country, and Mohamed Ibrahim Egal, the secretary-general of the Somali National League in 1959 and the first Prime Minister of Somaliland in 1960. They shadow-boxed around the issue of 'tribal' representation because they were 'nationalists' and thus presumably anxious to de-emphasize what they considered to be the weakness of Somali nationalism, even though they ruthlessly used the 'bonds of kinship in attaining their political aims' (Lewis 2002:168).

It was British administrators rather than the Somali elite who discussed the issue frankly, but they were not any wiser for their frankness. As the above discussion illustrates, they were confused about the 'tribes'. This is surprising since the British have been writing about Somali 'tribes' since the nineteenth century. Lieutenant Cruttenden was the first European to talk about the Darood and the Isaaq (Cruttenden 1849); Burton devoted a chapter on the 'tribes' in his famous book; after them every explorer in the nineteenth century and every administrator during the colonial era had something to say about 'tribes'. A central feature of these discourses was the genealogical tree. The earliest contributions to the reproduction of the Somali genealogical tree were made by Captain S. B. Miles and Captain H. M. Abud (Miles 1870; Abud 1893). Captain Abud wrote his report as a 'ready reference book to guide' the 'newly appointed' officers to the coast 'in their work'. He concentrated on the genealogical tree which he 'obtained from the elders of the tribes' and which he 'checked and rechecked over and over again' to ascertain 'that the main points' were accurate. Although he focused on the genealogical tree he did not want to produce a 'perfect genealogical tree' since that 'would necessitate the entry of the name of every male Somali living or dead, a task which would be supererogatory indeed'. He talked about the mythology of Somali origins, and then discussed the division of the kinship groups into 'tribe, sub-tribe, clan, and family' (Abud 1893).

The discussion was brief, but the genealogical tree was detailed. In 1945, the administration produced a detailed report on the topic (Fisher 1945). After a brief background discussion of the country, a history of the connection of the Somalis and the British, and the claims of the Somalis to 'racial individualism that would seem to be at least as good as our own', the report plunged into a straightforward classification of the 'tribes', followed by a detailed genealogical tree. Then, in 1950, John Hunt summarized all the knowledge he could muster about the genealogies of the 'tribes' in his exhaustive survey of the country. Yet all these discourses provided no useful insight into the workings of the Somali political system. The question that consistently perplexed the administration was: how did Somalis make political decisions? Who were their leaders? How did the genealogical tree work? In 1952, Reece invited R. S. Hudson, an amateur anthropologist who worked and lived in Tanganyika, to visit the country. Hudson's remit was difficult: to explain the mystery of how Somalis made political decisions (Hudson 1952). Hudson, however, could not explain the mystery. He could not explain how Somalis made political decisions or how the various kinship groups cohered. This was a common failing among those who tried to understand the problem. And so they almost always fell back on the colonial archive--and especially on the genealogical tree, which since Abud's report had been refined and completed.

But in itself, the genealogical tree is useless (Ibn Khaldun 1999: 226); it is like 'abandoned roads on an old map' (Bourdieu 1977: 38). There is a radical difference between genealogy (nasab) and group feeling (casabiya). They have different rules in politics. Politics is, by necessity, a corporate activity because the human being is, by necessity, a social being (Ibn Khaldun 1999: 69) who has two fundamental needs. First, he is not independent in acquiring his needs in food and other necessities (Ibn Khaldun 1999: 641,678); second, he needs protection from other human beings for self-preservation (Ibn Khaldun 1999: 69, 540, 641). For Ibn Khaldun, who was the first philosopher and historian to deal with the social contract, man enters into political society because of the inconveniences of the state of nature, which are considerable since man is by nature evil. Man has a dual nature: he has a 'human' nature and an 'animal' nature. The first is the source of his 'good' qualities, the second of his evil qualities. The 'animal' nature makes the human being by nature evil; if he sees what he likes he is likely to take it by any means he can (Ibn Khaldun 1999: 223).

This is why the first requirement for the survival of the human being is security. To achieve security he needs cooperation with others. The most basic cooperation takes place between silat al-raham (very close agnates--the dia-paying group): people tied by blood, because the relationship between blood relatives is natural to human beings (Ibn Khaldun 1999: 225). Close agnates are connected by nasab (genealogy; in Somali, abtirsiimo--literally the counting of fathers). When the nasab is 'very close' then al-itihad (alliance) and al-iltiham (closeness or togetherness) is the 'clear result' (Ibn Khaldun 1999: 225) which makes possible al-munasara (to come to aid of one other; in Somali isu hiilin) (Ibn Khaldun 1999: 226). This is the foundation of casabiya (group feeling; in Somali, tolnimo). (2) The casabiya is especially necessary in war. Since man is by nature evil, war is a 'natural thing among human beings' (Ibn Khaldun 1999: 479). The causes of war are revenge, hatred and struggle over power and resources; the most common cause, however, is revenge (Ibn Khaldun 1999: 479). What makes one group victorious over others in war is casabiya, because it makes possible 'common defence, resistance, protection, and the making of common demands' (Ibn Khaldun 1999: 249). The group that has such common feeling wins as a group. Two things, then, make the foundations of a political society possible: silat al-raham based on nasab (abtirsimo) and group feeling (casabiya; tolnimo). The two (genealogy and group feeling) are different things. The first relates the connection between individuals genealogically; the second is the actual feeling that ties them together and which makes it possible for them to defend themselves (and impossible to do so if that feeling is absent). The two are distinct. Colonial anthropology confused them badly.

Genealogy (nasab) is natural to people, yet it is wahmi (imaginary) (Ibn Khaldun 1999: 326-7). In itself it does not create the togetherness (al-iltiham) of close agnates. What creates such togetherness is 'social intercourse, common defence, long association and familiarity, and companionship in childhood and in all other aspects of life' (Ibn Khaldun 1999: 327). This is the basis of group feeling, which consists of two kinds: the group feeling of the smallest group (close agnates), and the group feeling of the larger group such as the lineage or the clan or the nation (casabiya kubra jamia: big group feeling that unifies) (Ibn Khaldun 1999: 522). Practical kinship rather than 'official' kinship is the basis of group feeling, whether small or large (Bourdieu 1977). That is the key to understanding Somali kinship, not the genealogical tree. The 'simple model of agnatic segmentation with equipoise units at every level' is not very useful since the lineage system is an 'ongoing structure, continually developing by segmentation over the generations as the population expanded' (Lewis 1961: 159). In 'this historical process' the importance and role of ancestors as 'reference points of contemporary segmentation in the day-to-day relations between lineages' constantly changes. An ancestor who once represented a 'primary lineage-group, may, in time, come to represent a clan or even eventually clan-family', or his descendants may have declined in number, or may have even completely disappeared. Growth in the system is not uniform: 'some lineages expand with historical segmentation, while others contract and even disappear. This historical process provides the framework of contemporary political lineage cleavage. Groups unite and divide by reference to points of historical cleavage' (Lewis 1961 : 158).

Agnates counteracted the inequality in the system by using various systems of alliances. Some of these alliances were based on uterine connection, and some on pure invention. Although affinal ties were not part of lineage morphology, yet they were important in lineage segmentation (Lewis 1961: 141). When a section of a primary lineage was smaller than another, Somalis formed baho (uterine alliance) to counteract the strength in number of the other group within the same lineage (Lewis 1961: 156); hence, 'Real or putative uterine alliances are ... as much a feature of the Somali lineage system as is agnatic segmentation' (Lewis 1961: 156). There were other alliances such as the gaanshaanbuur (piling of the shields) which were 'struck across agnation without this [uterine] justification, and to a corresponding degree without honour' (Lewis 1961: 156). Somalis also formed other relations based on affiliation such as relations formed by marriage xidid (literally root) across clans and lineage cleavages. Burton, for instance, noted that the Habr Younis (Isaaq) 'refuse maidens of the same or even of consanguineous family'. The men were encouraged to marry from other clans as a 'political device' to create 'connections' with the other clans; this was a 'common' practice for all Somalis. One of the uses of the cultivation of such connections, for instance, was to 'entitle the stranger to immunity from the blood-feud' (Burton 1966: 95; Lewis 1961: 180).

There were other relations of affiliation at the individual level based, for instance, on religious orders (tariiqa). The 'tariiqa affiliation is a matter of personal choice' and so 'cuts across clan and lineage boundaries' (Lewis 1961: 220). Relations were also built on the basis of business connections, or on the basis of vicinage (Lewis 1961: 125). The winning of friends and allies, wrote B. W. Andrzejewski, was always an 'unceasing preoccupation of every' individual and corporate group (Andrzejewski 1967); these friendships and alliances were formed in various ways. Although agnation was the 'given element in Somali politics' (Lewis 1961: 136), it was 'not the sole idiom in which political relations are cast' (Lewis 1961: 192). Myriad alliances were formed in everyday life that transcended agnation. The Somali system was famous for the relativity of its 'political divisions' (Lewis 1961: 7) as well as the 'relativity of affiliation' (Lewis 1961: 136). That relativity of political divisions and affiliation stretched over the group feeling from the smallest (dia-paying group) to the largest (clan, nation). The genealogical table does not tell anything about the real politics of the people. But as a mythology it is very useful, for it naturalizes the historical (Barthes 1957: 202)--it empties the historical (content) from human relations and reduces it to an abstract table (form).

The problem for colonial anthropology was that it had difficulty in understanding the society: on the one hand, it was confronted by too many clans, lineages, and dia-paying groups that were connected and yet independent, that functioned politically and yet appeared to be anarchic; on the other hand, it had somehow to make sense of the politics of the country. Between the 'excessively real' (Barthes 1957) and the need for practical ways of governing the country, colonial anthropology could find no other alternative but to rely too heavily on the genealogical table. But the table, because it ignored the historical, had no practical value. How could one appoint leadership from the outside and by proclamation for such a flexible political system? How could one fix it as a chart if it is not fixed and if it is always already in flux? How could one replace a system based on give-and-take, on constant haggling, on consent, to one based on fiat and expect it to work and function the same way? How could one change a system in which the people are sovereign to one in which sovereignty is claimed by a central authority? The genealogical tree in the traditional Somali political system has no meaning in itself; indeed, beyond a certain point Somalis simply do not care for it (Lewis 1961: 130). It is social intercourse--the politics of everyday life--that is the key to Somali kinship. Colonial anthropology was unable to comprehend that point. Moreover, it was unable to understand that the 'togetherness' of the group, whether it was the dia-paying group or the clan, was based not just on genealogical connection and social intercourse, but also on social contract. Without understanding the form and content of the social contract one could hardly make sense of the Somali kinship system.

II

The foundation of the kinship system was the social contract (xeer). Agnates were formed as corporate political groups not just because they shared a common ancestor but because of explicit treaties that defined the terms of their collective unity and regulated the behaviour of the individual by a system of delicts. As already pointed out, the human being is by necessity a social being because he has two fundamental needs. First, he is not independent in acquiring his needs; second, he needs protection from other human beings. These two fundamental needs force human beings to form alliances for protection. The basic form of such alliance is the close agnates. But this is not enough since all human beings organize themselves as agnates. Neither is the acquisition of arms. Arms are useful in protecting human beings from wild animals, but against other human beings arms are not very effective since all groups arm themselves. Against 'injustice armed' (Aristotle 1962: 1253a29) human beings arm themselves in defence. Hence human beings need something else beside basic alliances and arms to protect themselves.

They need wazic (constraint, restraint, check) (Ibn Khaldun 1999: 71, 72, 73, 223, 333, 670). This is the key word that Ibn Khaldun uses repeatedly: human beings need wazic, he says again and again, to prevent destruction, conflict, war, the shedding of blood, and wanton killing (Ibn Khaldun 1999: 333). The formulation and preservation of wazic (social contract) is the 'proper role of politics' (Ibn Khaldun 1999: 670). For the wazic (social contract) to be effective, the 'culture of politics' ought to be based on ethics (Ibn Khaldun 1999: 252, 253). But ethics is not enough. The culture of politics ought to be based also on laws that distinguish between what is 'just' (haq) and what is 'unjust' (badil) (Ibn Khaldun 1999: 61). For Ibn Khaldun, as for John Rawls, the idea of justice is the charter--the foundation--of society (Rawls 1973: 10). The sharing of a common understanding of justice 'makes a polis' (Rawls 1973: 214). The greatest danger to the social contract is one-man rule (infrad bil almr) which engenders oppression (dulm), corruption and economic ruin (Ibn Khaldun 1999: 517). One-man rule impoverishes the society as it increases the wealth of the ruler and that of his family, close relatives and dependants- ministers as well as policemen (Ibn Khaldun 1999: 502). The social contract is most effective in preventing mayhem, oppression and corruption--the lowest level (dia-paying group) or the highest level (casabiya kubra jamia: big group feeling that unifies) (Ibn Khaldun 1999: 522)--when the relations between the leaders and the people are defined by 'closeness' (that is, equality in status), 'ease' (that is, familiarity) (Ibn Khaldun 1999:514), companionship, mutual protection and common aims (khulq fil rufq bil ricaya wal qasd) (Ibn Khaldun 1999: 522).

Sir Henry Maine traced the evolution of law from one based on custom to one based on contract (Maine 1861). Somali law was based on contract (Lewis 1961: 299). A contract is by definition an
"agreement between two or more persons intended to create a legal obligation between them and to be legally enforceable. The law of contract is accordingly the body of principles relative to those kinds of agreements which are intended to, and do, create binding and enforceable bilateral obligations, conferring rights in personam on each party and which are therefore legally contracts". (Walker 1980: 284)

Likewise, the Somali social contract regulated relations between groups and within groups by explicit contracts (Lewis 1961: 6). These contracts were recorded orally although Cruttenden pointed to a treaty recorded on paper and deposited in the grave of a venerated Sheikh in Berbera (Cruttenden 1849; Lewis 1961: 176). Many of the contracts collected by I. M. Lewis during the late colonial period were written. Take, for instance, the contract signed by the three dia-paying groups that constituted the Hinjiinle (Dhulbahante): the Yusuf Adan, Farah Adan and Ahmed Adan. The contract consisted of five articles. The first and second articles specified how blood money of males and females was to be shared by the Hinjiinle; the third article stated the procedures to be followed in the event of homicide that occurs within the group; the fourth article delineated the procedures to be followed in the event of homicide that occurs outside the group; and the fifth stated the procedures to be followed in the event of the killing of a non-Hinjiinle woman by a Hinjiinle (Lewis 1961: 178, 177-80). Like agnation, the social contract was 'not a static principle, but dynamic within the framework of the lineage system' (Lewis 1961: 195). If a new situation arose, for instance, the Hinjiinle would meet and sign another xeer about the issue. This was the case not just within a dia-paying group or lineage but also between lineages and clans. The making of legislation, in other words, was in the hands of the corporate groups. The alliances discussed in the previous section, such as the baho and the gaashaanbuur, were formed by contract. The contract may be renegotiated or rescinded according to the requirements and needs of the corporate groups (Lewis 1961: 170).

The social contract, moreover, defined individual responsibility (and thus restrained it) by a system of delicts. The key delicts were homicide (dil), injury (qoon) and insult (dhaliil): for dil the compensation was mag, for qoon it was qoomaal, and for dhaliil it was xaal (Lewis 1961: 162). The mag was the same for male and female irrespective of age, rank or wealth: for the male it was set at 100 camels although between the Gadabursi and Eisa it was set at 10 she camels, 10 cows, 100 sheep and goats, and one 'nubile girl fitted out for marriage' (Lewis 1961: 164); for a female it was set at 50 camels. The payment of xaal was regulated by a code. Delicts for the payment of xaal included striking with a shoe or a whip, slapping with a hand (which was divided into three kinds: dhirbaxo (striking with the palm of the hand), faragorgor (striking with the back of the hand) and tanatoomo (striking with the hand by a blow from the shoulder)), entering a house without permission, failure to marry a girl after betrothal, uninvited conversation with a man's wife. The compensation for these delicts varied. The payment for striking a person with a whip was one pony (which was valued at 5 camels or 45 sheep, since one camel was valued at nine sheep). If the blow with the whip caused injury (qoon), the injury would be assessed and compensation paid above the compensation for the xaal. The compensation for qoon, like that for xaal, was regulated by code. The compensation for the loss of an eye or a limb was half the blood money, that is, fifty camels for male and twenty-five camels for female injuries. Other injuries were divided into two: (1) those that caused swelling (barar) and (2) those that caused an open wound that led to bleeding (nabar).

The latter is divided into 'several classes', the most important of which are: a penetrating wound to the throat or abdomen; a wound that severely destroys tissue without cutting it off--that, for instance, caused limping (naafo); and a wound that severs a finger or a toe. Injuries to the skull were graded according to code and four degrees were recognized: (1) a wound that exposed the skull (lafcadaatay); (2) a wound that caused a depression; (3) a wound that caused fracture; and (4) a wound that exposed the brain. The compensation for exposing the scalp was five camels; for causing a depression in the skull ten camels; for fracture of the skull fifteen camels; and for exposing the brain thirty-three and one-third camels. The teeth were valued: each was worth five camels. So were fingers and toes: they were valued at ten camels each, while every joint of the fingers and toes was equivalent to three and one-third camels (Drake-Brockman 1912; Wright 1941). There were also codes for marriage, divorce, inheritance, theft, robbery, affray, etcetera. This public system of roles 'constrained' human behaviour and protected 'property', so to speak. As such it made possible the creation of a political society; agnation in itself is not sufficient to create such system.

This traditional public system of rules was subverted during the colonial period in two ways: the administration imposed on the society another system of rules to resolve legal disputes, and the elite waged a vicious campaign against the social contract for political reasons. The administration, as already pointed out, had two different ambitions in the early and late colonial periods. In the early colonial period, it wanted to create chiefs; for such an institution customary law was presumed to be of great importance. In the late colonial period, it wanted to establish modern institutions led by a modern elite, although it still wanted to create chiefs in the rural areas. Once the attempt to build chiefs in the rural areas failed, the government decided to focus on the development of a modern court system. The administration created two courts: the Qadi (Islamic) court and the district courts, supervised by the Appellate court. The Qadi courts dealt mainly with marriage, divorce, inheritance, etcetera, while the district court dealt with delicts such as murder, injury, theft, robbery, property disputes, etcetera. The district court was the key legal institution in the country through which the administration replaced the 'traditional unofficial settlements' (xeer) with a new system (Reece 1953a).

The new public system of rules was based on the India Code of Civil Procedure and Indian Penal Law. This process had already begun in the early colonial period when the government (especially in the 1930s) increasingly tried murder cases even when the cases were settled in the rural areas according to Somali law. Despite the introduction of a totally new and alien system of rules, lawyers trained in the Western legal tradition were not allowed to practise in the country; even petitions to protest against the injustices of the legal system were prohibited. According to Rawls, a society is well ordered only when it is 'regulated by a public conception of justice' (Rawls 1973: 4). This conception must be a shared conception since it is the 'fundamental charter of a well-ordered human association' (Rawls 1973: 5). Under colonial rule, however, the old shared conception of justice was replaced with a new conception that was never 'shared', never transparent, and hence never public. The new rules were written in an alien language that Somalis could neither read nor write nor speak; and those who were trained in the new rules were not allowed (at the least in the early colonial period) to practise law, and in the late colonial period, even though lawyers were allowed to practise, not a single Somali was trained formally in the Western legal tradition.

Meanwhile, the Somali elite waged a vicious campaign against the traditional contract. The elite justified their war against Somali xeer in two ways: they said that it was a backward system, and that it weakened the nationalist movement. As I. M. Lewis put it, the elite 'attribute[d] their difficulties in creating a stable national patriotism to the force of heer, rather than to the force of clanship (tol)' (Lewis 1961: 195). This distinction was their alibi for their political use of clanship. They had two contradictory ambitions: on the one hand, they wanted to mobilize supporters through the clan; on the other they wanted unchecked power. As already pointed out, under the public system of rules of kinship all members of the clan, lineage or dia-paying group participated in the decisions of the group: the formulation and following of rules for the kinship group were social practices in which the whole group participated. The elite had no interest in such a democratic system. They wanted the kinship relation but not the burden of the rules and ethics of kinship. And so they waged a fierce ideological campaign against the rules that governed kinship even as they ruthlessly used it for political ends. What they wanted was kinship politics without any rules. The logical end of such politics was the total collapse of the state in 1990 because Somali politics became a politics without rules, which in reality is not politics at all.

The elite waged a fierce campaign against xeer in all public forums such as the Legislative Council, Protectorate Advisory Council, community centres, Radio Hargeysa, and the cafes. They insisted that there was a contradiction between the requirements of modernity and Somali law. This contradiction was especially flagrant with respect to murder. Colonial law--which they viewed as modern--recognized only individual responsibility, while traditional law recognized 'tribal' responsibility. This was of course a naive view of colonial law. Colonial law recognized and insisted upon collective responsibility. The Collective Punishment Ordinance, after all, remained on the statute books of the protectorate until the day of independence. The courts prosecuted people as individuals when they committed minor crimes, and sometimes when they committed murder. But more often than not, punishment was apportioned to the whole 'tribe'.

Haji Khalif Hassan and Yusuf Meygag Samatar debated the issue in 1953 on Radio Hargeysa. The title of their topic was 'The Payment of Dia and Wound Compensation' (War Somali Sidihii 1953). It was also debated at length in the Protectorate Advisory Council in 1955. Omer Horreh (Erigavo) argued that the people of the Erigavo town were 'unanimous in their wish that the collective payment of dia (blood money) and wound compensation as governed by Somali custom should be abolished'. Somali law is 'bad for public peace because it encourages murder and other acts of violence'. Under Somali law any crime committed by an individual would be shared by all 'members of the tribe and for his act he [the criminal] would have to pay a fine which might be as little as 8 annas'. We must place 'responsibility on the wrongdoer and not on the tribe as a whole'. If Somali law was abolished 'there would be fewer murders' (War Somali Sidhii 1955). The motion was endorsed by 24 members of the Council and rejected by 19 members (War Somali Sidihii 1955). But since the Council was an advisory body its decisions were not binding.

The Legislative Council also discussed the issue in 1958. Ahmed Haji Dualeh, an unofficial member of the Council, proposed the abolition of Somali xeer under which 'every Somali is compelled to take part in the payment of blood money' ('Debates in Council' 1958). The official members of the Council abstained from voting, but were in general sympathetic to the motion, which was passed by the unofficial members. (The motion, though, was non-binding since only the unofficial members of the Council passed it.) Another (unofficial) member of the Council, Michael Mariano, proposed a motion with radical pretensions: the enactment of a law to enable anyone who wished 'to break away from the injurious custom of paying or receiving blood money in respect of dia or wound compensation to do so'. The motion was defeated because the official (majority) members rejected it. The Commissioner of Native Affairs, Mr. P. Carrel, characterized the motion as 'premature' ('Debates in Council' 1958). In 1959, Ahmed Haji Dualeh, an unofficial member of the Executive Council, proposed a motion to abolish Somali 'tribal Her', presumably because Somali law contradicted the requirements of modern law, which emphasized individual responsibility. After a long discussion, the Council defeated the motion and decided to continue the effective implementation of the Collective Punishment Ordinance (Executive Council 1959).

According to Ibn Khaldun, the etymology of the word siyasa--the word Somalis use for politics--is sas (3) (foundation). Politics is about building the foundation of society for a common 'good'. The social contract was of central importance in Somali politics because it constituted the public system of rules of politics. In the colonial period, and especially in the post-colonial period, the social contract was increasingly ignored both in the practices of politicians and in academic studies. The Somali elite was embarrassed by the traditional Somali contract, which they assumed to be the symbol of the backwardness of Somali politics even as they eagerly used kinship for achieving their political ambitions. They were encouraged in their selective policies by their conquest, so to speak, of the colonial administration, which gave them a power centre independent of the traditional political structures of the people. This gave them the ability to pick and choose from the old system: to use kinship relations for political purposes and to marginalize the social contract. In short, since the colonial period and especially since independence, Somali politics has been hobbling about with one very weak leg (kinship) as the other leg (social contract) withered under the assault of colonial rule and post-colonial hypocrisy.

REFERENCES

Abud, H. M. (1893) PRO, FO 78/4602, 'Genealogical trees of the Somal, Dolbahanta and Habr Gerhajis, Aidagalleh, October'.
Andrzejewski, B. W. (1967) 'Poetry in Somali society', New Society 1 (25) (21 March).
--(1993). An Anthology of Somali Poetry. Bloomington IN : Indiana University Press.
Aristotle (1953) The Nicomachean Ethics. London: Penguin Classics.
--(1962). The Politics. London: Penguin Classics.
Barthes, R. (1957) Mythologies. Paris: Editions du Seuil.
Bourdieu, P. (1977) Outline of a Theory of Practice. Cambridge: Cambridge University Press.
Burton, R. (1966) First Footsteps in East Africa. London: Routledge and Kegan Paul.
Carrick-Allan, N. P. (1958) PRO, CO 830/16, Acting Attorney General and Chair, 'Commission of Inquiry into the Unofficial Representation in the Legislative Council, 8 February'.
Chief Secretary (1948) PRO, CO 537/3618/15363, 'Notes on British Somaliland, June'.
Clarke, W. S. (1995) The Humanitarian Intervention in Somalia: a bibliography. Washington: Centre for Strategic Studies.
Colonial Office (1954, 1955) Colonial Office Annual Reports on the Somaliland Protectorate, 1954 and 1955. London: HMS.
Cooper, F. (1997) International Development and the Social Sciences. Berkeley: University of California Press.
Cordeaux, H. E. S. (1905) PRO, CO 879/87, Acting Commissioner to Mr Lyttelton, 23 August.
--(1906) PRO, CO 537/44, 'Memorandum by H. E. S. Cordeaux on Somaliland, 30 June'.
Cruttenden, Lt (1849) 'Memoirs of the western or Edoor tribes, inhabiting the Somali coast of northeast Africa', Journal of the Royal Geographical Society 19.
'Debates in Council' (1958) Africa Digest 5 (4) (January-February).
Drake-Brockman, R. E. (1912) British Somaliland. London: Hurst and Blackett, Ltd.
Executive Council (1957) PRO, CO 830/11, 'Minutes of the 18th Meeting of the Executive Council Held at Government House on Wednesday, 27 November at 9 a.m.'.
--(1959) PRO, CO 830/13, 'Abolition of Heer'.
Fisher, G. T. (1943) PRO, WO 32/10862, 'Annual Report on the Administration of British Somaliland'.
Fisher, G. T. (1945) PRO, CO 1015/672, 'British Somaliland and its Tribes, January'.
Griffiths, J. (1950) PRO, CO 535/152/2, J. Griffiths to Gerald Reece, 30 March.
Haabiil, A. J. (1993-4) 'Baroordiiq', Hal-Abuur: Journal of Somali Literature and Culture 1 (1-2).
Hornby, M. L. (1905) PRO, CO 879/87, Lt Colonel M. L. Hornby, Commanding the Troops, Somaliland Protectorate, to His Majesty's Commissioner and Commander-in-Chief, Somaliland Protectorate, 16 August.
Hudson, R. S. (1952) PRO, CO 1015/560, 'Some preliminary notes on problems relating to the constitutional development of the Somaliland Protectorate, March'.
Ibn Khaldun. (1999) al-Muqaddimah. Beirut: Dar al-Kitab al-Lubnani.
Information Office. (1951) PRO, CO 1015/366, 'New notes on Somaliland Protectorate, July-September'.
Jameson, F. R. W. (1943) PRO, WO 230/94, 'Tour impressions 5 September'.
Jerrom (1950) PRO, CO 535/152/2, 'Minutes'.
Lanham (1950) PRO, CO 535/152/2, 'Minutes'.
Lewis, I. M. (1958) 'Modem political movements in Somaliland--II', Africa 28 (4).
--(1961) Pastoral Democracy. London: Oxford University Press.
--(2002) A Modern History of the Somali. Oxford: James Currey.
Maine, H. (1861)Ancient Law. London.
Miles, S. B. (1870) PRO, FO 78/3186, 'A rough genealogical sketch of the Somalees from Tagurah to east of Berbera, 22 April'.
Mohamed, J. (2002) '"The evils of locust bait": popular nationalism during the 1945 anti-locust control rebellion in colonial Somaliland', Past and Present 174.
'Native Administration in Somaliland' (1951) PRO, CO 535/152/3.
Rawls, J. (1973) Theory of Justice. Oxford: Oxford University Press.
Reece, G. (1950) PRO, CO 535/152/2, Gerald Reece to C. E. Lambert, Esq., 5 March.
--(1951) PRO, CO 535/152/2, Gerald Reece to C. E. Lambert, Esquire, 29 January.
--(1953a) PRO, CO 1015/363, Reece to Secretary of State for the Colonies, 9 May.
--(1953b) PRO, CO 1015/560, Gerald Reece to J. E. Marnham, 6 August.
Rosenthal, F. (1967) The Muqaddimah. Princeton: Princeton University Press.
Scawin, A. L. (1959) Oxford University, Rhodes House Library, 751.13s.2/1959(1), 'Somaliland Protectorate: Report of the Commission on Representational Reform, April-May'.
Walker, D. (1980) The Oxford Companion to Law. Oxford: Oxford University Press.
Wallis (1950) PRO, CO 535/152/2, 'Minutes'.
War Somali Sidihii (1953) PRO, CO 1015/868.
War Somali Sidihii (1955) PRO, CO 1015/868.
War Somali Sidihii (1957) PRO, CO 1015/868.
Wright, A. C. A. (1941) Oxford University, Rhodes House Library, Mss.Afr.S.1104, 'The interaction of various systems of law in British Somaliland with particular relation to the payment of blood money and the infliction of the death sentence'.
(1) The administration added a fourth member to the council in order to expedite the judicial process (Wright 1941).
(2) The etymology of the word is cusba, which is used in the Quran many times: it means jamaaca (group); casabiya is the feeling that ties the cusba (group) or 'group feeling', as Franz Rosenthal translates the word (Rosenthal 1967: xi).
(3) Somalis pronounce the word sas as 'sees'; they use the word literally for the foundation laid for a house.

JAMA MOHAMED writes about the social history of Somaliland. He lives in Toronto, Canada.