“Few societies can so conspicuously lack those judicial, administrative, and political procedures which lie at the heart of the western conception of government. The traditional northern Somali political system has no chiefs to run it and no formal judiciary to control it. Men are divided amongst political units without any administrative hierarchy of officials and with no instituted positions of leadership to direct their affairs.” (Lewis 1961 p. vii)

Somalia was created in 1960 out of the colonies of British Somaliland (north) and Italian Somaliland (south). The exiting colonial powers set up a democratic central government, possibly not the best option for a society whose traditional institutions were decentralized and stateless. The democracy lasted for nine years and was followed by the dictatorship of Siad Barre. He was ousted in 1991, the central government disintegrated, and the Somalis were back with their traditional system.

With two differences. First, the experience of a past central government and the expectation of a future one encouraged some, especially those near the capital of Mogadishu, to engage in a power struggle aimed at putting themselves in the very profitable role of rulers instead of the very unprofitable role of ruled. Second, outside powers, acting through the U.N. in the belief that the country needed a central government, attempted to reestablish one by military force. The result has been an extended period of violence and chaos, especially in and near the capital, by groups acting outside of, and often in violation of, the traditional legal system.^¹

Elsewhere, especially in what had been British Somaliland and now calls itself simply the Republic of Somaliland,^² the traditional system of customary law reestablished itself. While Somaliland has a government—arguably created in part as a so far unsuccessful attempt to persuade foreign states of its legitimacy—it is a government based on traditional institutions, with an upper house of clan elders, and one that seems for the most part to defer to customary law privately enforced in the traditional manner—the same policy followed earlier by the British officials.

This chapter is based primarily on the work of I.M. Lewis, a British anthropologist who has been writing about Somalis since the 1950’s, especially on the detailed description of the institutions of the northern Somali pastoralists in his A Pastoral Democracy: A Study Of Pastoralism And Politics Among The Northern Somali Of The Horn Of Africa. My other source is The Law of the Somalis, written by Michael Van Notten, a Dutch legal scholar who married a Somali wife and lived as part of the Samaron clan of North-west Somaliland for twelve years, until his death in 2002. The book was edited and published by Spencer MacCallum, a social anthropologist with an interest in stateless societies. As Van Notten makes clear, his description applies in detail only to the area where he lived (Awdal), but traditional law elsewhere, while not identical, is generally similar.^³

Both of my sources are concerned primarily with the pastoralists of northern Somalia, and it is their legal system I will be describing. The institutions in the south, where camel herding is to a considerable extent replaced by agriculture, appear to be somewhat less based on kinship, more on geography.^⁴ Van Notten’s observations were from about 1990 to 2002; Lewis based his detailed account on observations made in the 1950’s, when northern Somalia was still under British control. The accounts are broadly similar, allowing for the difference in the time periods and the fact that Van Notten is dealing with a more restricted area.^⁵

The institutions through which the Somali enforce rights and settle disputes are based on two principles—kinship, primarily agnatic kinship (defined through the paternal line) and contract. Every Somali memorizes, as a child, his genealogy through the paternal line up many generations, an important piece of information since it defines his relationship to every other Somali. A clan, which may number in the hundreds of thousands, consists of individuals all of whom^⁶ are believed to be descendants of a common ancestor in the paternal line, possibly twenty or thirty generations up. The closer the linkage between two Somalis—the smaller the number of generations to a common ancestor—the more likely they are to be allies, ceteris paribus.^⁷ “As the Somali themselves put it, what a person’s address is in Europe, his genealogy is in Somaliland.”^⁸

The system of alliances through which the legal system works is fluid in two sense. To begin with, coalitions are created by agnatic kinship at multiple levels. If, to simplify considerably, there is a conflict between two individuals whose common great-great-grandfather in the paternal line had two sons, the group that becomes engaged on the side of each will be the descendants of the sons from whom he is descended. If a conflict arises involving a member of one of those groups against someone whose genealogy links with theirs higher up the genealogical tree, the two groups that were enemies in the first round may ally. In the case of a conflict between individuals in different clans, all of each clan is, in principle if not always in practice, allied in support of its member.^⁹

The coalitions are not defined entirely by kinship. The nearest thing to a well defined and stable unit below the clan level is what Lewis, following the practice of the British administrators, refers to as a “dia-paying group” (“dia” is the Arabic term for blood money, the Icelandic wergeld).^¹⁰ The dia-paying group is responsible for paying for offenses by its members, collecting for offenses against its members, and, in the latter case, using force or the threat of force to obtain payment.^¹¹ It also deals with conflicts between its members, under rules defined in part by contract.

The dia-paying group’s membership and internal rules are defined by explicit contract.^¹² In most cases it is made up of a number of jiffo-paying groups, each of which may consist of the descendants in the paternal line of a single common ancestor, or several such. Within the dia-paying group, the separate jiffo-paying groups have special responsibility for the offenses of their members defined by the group’s contract, usually paying all of the cost up to some limit, commonly a third of the payment for a killing. The division of costs above that point among the components of the larger group is also defined by contract, usually in proportion to either the number of males in each Jiffo-paying group or its wealth in livestock.

The separate jiffo-paying groups within a dia-paying group may or may not be linked by agnatic kinship. If smaller groups not so linked wish to combine for the purpose, they may take advantage of links through maternal kinship to do so, producing a uterine alliance, and similarly for coalitions at other levels of the system. Sometimes coalitions are formed with neither agnatic nor uterine links as justification; such are referred to as gaashaanbuur, meaning literally 'pile of shields,'^¹³ since they are made in order to acquire sufficient fighting strength.

By Lewis’ account, the size of a dia-paying group is limited at the low end, about 300 males, by the need for enough resources to pay blood-money when necessary without an undue burden on any individual and for sufficient military power to enforce claims for payment of blood-money. It is limited at the high end, about 3000 men, by the problems of internal conflict—too many able and ambitious would-be leaders, situations where one of the component groups feels it is being unfairly burdened by payments for offenses by members of other groups, and similar problems. Component groups can and do split off, to form their own dia-paying group or join with a different one. And if circumstances force an alliance among multiple dia-paying groups, they may temporarily constitute themselves as a single such group.

Somalis distinguish between general laws, applying to all members of the clan, and special laws, applying to members of a particular group, in particular a dia-paying group, judged within the group. Special laws are established by explicit contract,^¹⁴ general law seen as a set of broad principles, with the application up to the particular judge; thus there are, in principle, no situations on which the law is silent. Decisions are to be based on customary practice. A sufficient number of consistent decisions may result in a rule followed by judges coming to be treated as a law; in this sense it is a system of informal case law.^¹⁵ Within a clan, significant differences among judges tend to disappear over time, but such difference between the jurisprudence of different clans may persist.

Judges are not officials with a position and salary but arbitrators accepted by the disputants. A judge has no special rights,^¹⁶ such as the right to summon or cross-examine a witness. Nor is the judge viewed as an authoritative source of law. His job is to settle conflicts by applying the rules that people in the community normally observe. A judge who produces verdicts that meet general disapproval is unlikely to be asked to judge cases in the future. To quote Van Notten:

“A Somali judge is free to develop his own principles of law and his own doctrines. The test of whether such principles and doctrines are acceptable to the community comes as soon as he has given his verdict on a conflict. If a verdict deviates from what the community finds reasonable and just, there is little chance that its author will be asked again to sit as a judge.”^¹⁷

Neither politicians nor religious dignitaries are responsible for developing or interpreting the general law, and, as a rule, neither can function in the law as judge, witness, or enforcer. Folk wisdom includes the sayings “One can change one’s religion; one cannot change the law” and “Between religion and tradition, choose tradition.”

One notable exception to the separation of law and religion is that matters of marriage and inheritance are usually brought before a judge who applies Koranic law, almost all Somalis being Muslims.^¹⁸ A second exception is that a judge who needs to know the extent of injury of an assault victim will ask a religious leader to investigate the question and testify on it. The schedule of payments of blood-money for death or injury is based on that in Shari’a, modified by custom and contract, with the amount sometimes larger or smaller depending on the relationship between offender and victim. The contract for a dia-paying group may specify a lower charge for injuries of one member of the group by another, a larger sum may be paid for death or injury to a particularly well respected victim or occurring under especially ugly circumstances, and groups sometimes agree to increase the amount in order to reduce intergroup violence.

Somali political institutions at the clan level exist but are limited. Traditionally, most clans had a titular head, appointed for life, whose functions were mostly ceremonial. The position is apparently not essential, since in recent years several clans have failed to choose any replacement for one who died.^¹⁹

Each clan also has an assembly (Guurti), made up of the heads of the most important families. It has the power to declare war or peace with other clans, but not to conscript or hire soldiers. It can recommend changes in the traditional law but has no power to legislate them. Decision making in the Guurti, as in the kris of the Vlach Rom, is not by majority vote but by consensus,^²⁰ with the result that meetings can last for months. If a conclusion is reached that some members still oppose, they (and, presumably, their families) have no obligation to help implement it.

The mechanics of trial and law enforcement as described by Van Notten are straightforward. When a dispute arises between members of different dia-paying groups, the elders (oday) from each side^²¹ form a court with themselves as judges, ask the parties to state their cases, hear witnesses and state a verdict. Judges may, before agreeing to judge a dispute, require the parties to agree, sometimes in writing, to obey their verdict. If force is needed to make the losing party obey the verdict in an intra-clan dispute, the judges can recruit all able-bodied male villagers for the purpose; anyone who refuses is considered an associate of the defendant and owes a fine to the plaintiff’s family. In the case of an inter-clan dispute, however, enforcement is up to the clan of the victim.

What if no court is agreed on in time, or the oday of one party’s family refuses to participate (particularly likely if the parties are from different clans), or the court fails to give a verdict? At that point, the victim and his family are entitled to self-help, imposing restitution or compensation by force. If the other party believes that excessive force was used or excessive compensation extracted he is entitled to sue for compensation.

Thus the Somali system is ultimately a feud system, one in which law is enforced by the private application of force or the threat of force, but a feud system with institutions for avoiding violence via widely respected mechanisms to arbitrate disputes.^²² Part of what makes it successful, according to Van Notten, is that families are obligated to help defend their kin but not to help attack their opponents, with the result that armed conflicts are likely to lead to stalemate, and from there to arbitration.

Conviction for what we would regard as a civil or criminal offense usually results in a fine, traditionally stated in camels. The offender’s dia-paying group (like the Irish fine^²³) is a guarantor for payment of the fine. In inter-clan disputes, the clan is the guarantor. If an individual repeatedly violates the law and is unable to pay the resulting fines, his dia-paying group may publicly announce that it is no longer responsible for him. Since there is now not only nobody to guarantee his fines but nobody to defend him or threaten force in vindication of his rights, he is effectively an outlaw, and likely to leave the territory for that of another clan. In less extreme situations, the kin group, having had to pay for the offense of its member, may restrict him in ways designed to prevent a repeat offense, for example by forbidding him from going armed.

Both plaintiff and defendant have the right to appeal a verdict, with the number of appeals permitted depending on the rules of their clan. The appeals court must have more judges than the court that produced the original verdict, drawn from a wider group of families or clans; if one party refuses to go along with an appeal, the family of the other is entitled to self-help.

In the course of a trial, a disputed fact is admitted as evidence only on the testimony of three witnesses. Parties can call in experts and character witnesses to support their case. Each party gets to state his case, or have a representative do so for him, call witnesses and present evidence; witnesses are not normally subject to cross-examination. In some cases the judge of the defendant may simply agree, on behalf of his clan or kin group, to pay the required compensation. As in some other legal systems, oaths are sometimes required as part of the legal process. If a fact is disputed, or supported by fewer than three witnesses, parties may be required to swear one of several different oaths to the truth of their position. One such oath consists of the oath-giver swearing by his marriage; if it later turns out that his oath was false, the marriage is dissolved. If the plaintiff fails to establish his case, the defendant must still swear to his innocence before the case will be dismissed.

If the convicted defendant refuses to pay within the specified time, he is subject to penalties ranging from a fine in honey to having one of his animals slaughtered, cooked, and eaten by the villagers each day.

As with Arabic jinayat, penalties are computed in camels—in the Somali case, healthy she-camels aged three to six years. It may be paid in other livestock, at a fixed customary rate, by agreement between the parties, or in money. The payment goes to the family of the victim. There exists a standard schedule of fines for unintentional bodily injury. The family of the victim may, and often will, accept less, possibly as a good will gesture. For an intentional injury, fines are doubled.

For intentional murder, the penalty is a life for a life; if the murderer succeeds in fleeing abroad, a member of his family of equal status may be put to death in his stead, a rule that gives his family a strong incentive not to help him escape. In most cases the victim’s family can choose to accept blood-money instead, at a rate of 100 camels for a man and 50 for a woman, although if the murder was sufficiently outrageous the court may insist on execution of the murderer. If the killer and victim are of different clans, the victim’s family are less likely to accept blood-money; if the killer escapes, his family owes two lives instead of one.

The fines in Somali law are based on the fines specified in jinayat, on which the Somali system is based, although the rules differ in detail.^²⁴ Jinayat fines for intentional death or injury are specified as a number of female camels, equally divided between one, two, three, and four year old (the heavier diya^²⁵) and, for accidental homicide, as a number of camels from a group of which one fifth are male one-year olds and the rest female camels equally divided between one, two, three and four year old. The Somali rules do not, and the jinayat rules do, treat the loss of a body part of which the victim only has one as equivalent to the loss of life. Otherwise the patterns are similar but not identical.

Somali legal rules for bodily injury have one other interesting feature. If a man seriously wounds another, his family must take the victim into their household and nurse him back to health—the same requirement as in ancient Irish law.^²⁶

The rules for accidental damage to property, including animals, are the same as in modern tort law. The person responsible must make the victim whole by replacing the property or its value. The penalty for stealing an animal, however, is the return of two.

Somali customary law also covers subjects such as breach of contract and defamation. The legal rules with regard to property are complicated by the fact that not all sorts of property can be privately owned. Grazing land is treated as a commons, with a rule of first come, first served—once one herd is grazing in a particular pasture or drinking at a water hole, it has temporary ownership. When water is scarce, grazing land and water sources in a clan’s territory get treated as possessions of the clan, which others are supposed to use only with permission. The definition of clan territory is based on past practice, and imprecise. It sounds from Lewis’ description as though the combined effect of the imprecision between clans and the lack of well defined property rights to grazing within clans results in a good deal of conflict.

Most of the people in the area studied by Von Notten were nomadic pastoralists, but there are also partly or entirely sedentary populations that recognize individual ownership of small parcels of grazing land, with the restriction that they may only be sold within the clan and lent or rented only to clan members or outsiders who have some tie with the clan, such as marriage to a clan member. Agricultural land in areas where the locals practice agriculture is similarly subject to a restricted form of ownership. Some land also belongs to groups or an entire clan, and can be divided and allocated to individual members only with the assent of all male members of the group.

One odd feature of Somali customary law that is that a wealthy man is required, with detailed legal rules, to share his wealth with neighbors and relatives. Seen from one angle this can be viewed as a form of social insurance, from another as a disincentive to successful business dealings.

The state of Ethiopia dealt with as a clan: In 1992, a year after the Tigraens took control of the Ethiopian government, some federal soldiers wantonly killed a Somali merchant near the village of Sheddher. Their reason was that he had refused to give them some of his merchandise, which happened to be qhat. An hour later, the family of the victim killed two federal soldiers who happened to be passing through the village. The military commander of the Somali region thereupon ordered a punitive expedition and sent an entire platoon to Sheddher. On its arrival, the soldiers learned that the villagers could have killed several more federal soldiers that day but had not done so. They had acted according to their customary law, which stipulates that when someone of another clan murders a clansman, two members of that other clan will be killed. Shortly thereafter, a similar incident happened in the same territory, in the village of Lafaissa, where a federal soldier had sought refuge in a military camp after wantonly killing a Somali. The family of the victim went to a military camp in the nearby village of Herigel and killed two federal soldiers. The military commander in Harar chose to take no action against the clan and informed his soldiers that henceforth they had to respect the customary law. As a result, no more killings occurred in the territory. (Van Notten, pp. 181-182)

Ḥassan Ugaas are currently estimated to number about 1,500 men. They comprise four main segments (jilibs) which act as jiffo-paying groups.

A petition delivered to the District Commissioner and dated the 8th of March, 1950, states their ḥeer to be as follows:

1. When a man of the Hassan Ugaas is murdered by an external group twenty camels of his blood-wealth (100) will be taken by his 'next of kin' (i.e. his sons, brothers, father, and possibly uncles) and the remaining eighty camels shared amongst all the Ḥassan Ugaas.

2. If a man of the Hassan Ugaas is wounded by an outsider and his injuries are valued at thirty-three-and-a-third camels (a standard rate for non-fatal but quite serious injuries),^²⁸ 1 ten camels will be given to him and the remainder to his jiffo-group.

3. Homicide amongst members of the Hassan Ugaas is subject to compensation at the rate of thirty-three-and-a-third camels, payable only to the deceased's next of kin. If the culprit is unable to pay all or part, he will be assisted by his lineage.

4. In cases of assault within the Hassan Ugaas for which compensation up to the value of thirty-three-and-a-third camels is payable (i.e. according to the Shariah) only two-thirds will be paid.

5. Ḥaal of 150 shillings (East African) is payable to the person attacked^²⁹ when a man of the Ḥassan Ugaas joins another to fight with a third.

6. If one man of the Hassan Ugaas insults another at a Hassan Ugaas council (shir) he shall pay 150 Shs. to the offended party.

7. If a man of the Ḥassan Ugaas marries a girl already betrothed to another man of the group, or a widow whom it is the customary right of another to marry, he shall pay ḥaal of five camels to the aggrieved party.^³⁰

8. If the Ḥassan Ugaas kill a man of another group they will pay his blood-wealth in equal shares (amongst the four lineages) by 'peniscounting' (qoora tiris).^³¹

9. Compensation for serious wounds valued at thirty-three-and-athird camels or more, owing to a person of another group, will be paid collectively by all the Ḥassan Ugaas by 'penis counting'.

Lewis, I.M., Peoples of the Horn of Africa: Somali, Afar and Saho, International African Institute, London, 1955 (reprinted 1969).

A Pastoral Democracy: A Study Of Pastoralism And Politics Among The Northern Somali Of The Horn Of Africa, Published for the International African Institute by the Oxford University Press London New York Toronto, 1961.

Blood and Bone: The Call of Kinship in Somali Society, The Red Sea Press, Lawrenceville, 1994.

Understanding Somalia and Somaliland: Culture, History, Society, Columbia University Press, N.Y., 2008.

Schacht, Joseph, An Introduction to Islamic Law, Clarendon Press, Oxford, 1986.

Van Notten, Michael, The Law of the Somalis, (edited by Spencer Heath MacCallum), The Red Sea Press, Asmara, 2005.

1 A brief period of peace in the south was created by an alliance of Islamic courts (the ICU, “Islamic Courts Union”) backed by local clan militias. In late 2006 that period was ended when Ethiopian troops, acting with U.S. and U.N. support in alliance with the U.N created “Transitional Federal Government,” invaded Somalia and eventually took Mogadishu. Lewis 2008 pp. 87-91. For Lewis’ view of the situation in Somalia in recent years, see the URL in the references.

4 About the Sab, Southern Somali agriculturalist, Lewis writes: “Their conquest political structure if not that of a tribal state is certainly more formalized and more hierarchical than the political system of the northern nomads. … Sab tribes—for here there are territorially defined jural and political communities—do not appear to base their political relations upon kinship but upon what Maine called local contiguity.” (Lewis 1961 p. 13.)

5 There is, however, an important terminological difference between the two sources; Van Notten inteprets kinship terminology in absolute terms; Lewis argues that it actually describes relative position in the genealogical tree. He writes:

The four terms, qabiil, ³qolo, jilib ⁴and reer which are commonly used to refer to different levels of grouping, and which have been consistently misinterpreted by most writers on the Somali, can legitimately all be applied to the same genealogical and political unit in different contexts. Generally speaking, apart from its restricted use to designate the descendants of a living man, reer means lineage in an extended sense. Thus the Ḍulbahante clan is a lineage (reer) of the Daarood clan-family and smaller units within the Ḍulbahante are lineages (reers) of the Ḍulbahante clan, and so on down to the minimal unit which consists of a man and his sons. The Arabic loan-word, qabiil, is generally restricted in use to the larger units such as clans, sub-clans, and primary lineage-groups. But here again its use is entirely relative. What, in relation to a smaller unit is referred to as a qabiil, in relation to a larger unit is spoken of as qolo or jilib. If a man of the Ḍulbahante, referring to his relations with other members of his clan, describes his qabiil as Ḍulbahante, his qolo as Barkad (a primary lineage-group) and his jilib as a smaller Barkad lineage, he may in another situation refer to Daarood as his qabiil his qolo as Harti (see genealogy at end of text) and finally his jilib as Ḍulbahante. None of these terms can be assigned any precise or fixed significance except by reference to other units. All are relative terms within the hierarchy of segmentation based on agnatic connexion.

7 the more distant units are genealogically, the greater the likelihood of extended strife between the larger units of which they are segments. Thus strife between a segment of the Ḍulbahante and a segment of the Habar Tol Ja'lo is more likely to lead to fighting between the Ḍulbahante as a unit opposed to the Habar Tol Ja'lo as a unit, than is strife between one group of Faaraḥ Garaad and one segment of Maḥamuud Garaad to a general war between these two segments of the Ḍulbahante. ... The smaller a unit the greater is its cohesion, and the larger a unit the more likely is it to split into hostile groups following the lines of genealogical cleavage. (Lewis 1961 p. 152)

9 “In principle, within the clan dia-paying groups oppose dia-paying groups, primary lineage-groups primary lineage-groups, and within the clan-family clans oppose clans. But the simple model of agnatic segmentation with equipoised units at every level is distorted by the recognition of irregular growth and by the importance given to the uneven distribution of man-power and fighting potential. These inequalities are counteracted partly by uterine ties which act as a built-in compensating mechanism, and partly by alliances (gaashaanbuur) outside kinship.” (Lewis 1961 p. 159)

12 Under British administration, the contracts were in writing, with copies filed with the British officials; for an example see Appendix 1. Lewis suggests that prior to that they were normally oral contracts, but may sometimes have been in writing and deposited with the custodian of the shrine of a prominent saint. (Lewis 1961 p. 176).

14 “It is at least refreshing to confront the various hypothetical theories of social contract with a contract case where this institution is systematically employed, within a kinship ambience, in the formation and definition of political units, to an extent which, so far as I know, is unparalleled elsewhere.” (Preface to the 1982 edition of A Pastoral Democracy, p. xiv)

15 This decription is from Van Notten. Schacht, describing pre-Islamic bedouin law, writes: “If protracted negotiation between the parties led to no result, recourse was normally had to an arbitrator (ḥakam) … . … the parties were free to appoint … any person on whom they agreed, … . The decision of the ḥakam, which was final, was not an enforceable judgement (the execution had indeed to be guaranteed by the security), but rather a statement of right on a disputed point. It therefore became easily an authoritative statement of what the customeray law was, or ought to be; the function of the arbitrator merged into that of a lawmaker, … . Schacht pp. 7-8.

17 Judges are normally the elders of the parties’ jilibs. Presumably the point of the passage is that a judge whose decisions are generally disapproved of loses his position as elder. Von Notten mentions that an individual unhappy with the elder of his jilib may leave his extended family and start a new one, possibly joining up with other dissatisfied individuals from other jilibs, or can try to convince the other members of his family to replace their elder.

23 Von Notten is not entirely clear about the membership of juffo (his term for what Lewis calles the jiffo-paying group) and jilib (dia-paying group). At one point the former is defined as all the descendants in the paternal line of a common great-grandfather and the latter is a group of juffo’s whose members are related. At another point, the jilib is defined as “all the living descendants of a given paternal great-grandfather or further removed ancestor.” The fine in ancient Irish law, discussed in Chapter XI, is made up of all the descendants in the paternal line of a common great-grandfather. It sounds from Lewis’s description as though the juffo (his jiffo-paying group or “tertiary section”) is defined by paternal kinship, the jilib (dia-paying group or “secondary section”) by contract among the member juffos, usually but not always ones that have a common paternal ancestor. Describing the logic of kinship alliances, he writes: “The tertiary sections support each other against the secondary, the secondary against the primary, and all combine in the solidarity of the tribe when this is threatened or attacked from outside. Lewis 1955 p. 108.

26 This is the second striking similarity we have observed between Somali and ancient Irish law. The Celts, in the course of their history, occupied many different places, but it is hard to believe that they got as far as the horn of Africa, so the similarities are presumably examples of parallel legal evolution—or, just possibly, common descent from some legal system in the very distant past. For a third similarity, consider that a favorite sport of both cultures appears to have been cattle raiding—in the Somali case camels.