Using Children in Armed Conflict:

A Legitimate African Tradition?

Culture, Tradition and Human Rights

Age Grades and Age Sets

Military Action and Socio-Political Structures

Age Sets in the Zulu Kingdom

Age Sets Under Colonial Rule

The Definition of Childhood

The Consequences of Violating Childrenıs Rights

Criminalising the Recruitment of Child Soldiers

The Nature of the Problem

Existing Protections Under Humanitarian Law

International Human Rights Law

Problems of Enforcement

Conclusion: Taking Stock



"On the eve of the new millennium we are witnessing an abomination ‹ an abomination directed against children in the context of armed conflict."

Olara Otunno, Special Representative for Children and Armed Conflict, United Nations, speaking at the UN Security Council, July 1998.


We selected under-researched topics for our current series of monographs, trusting that they will help in defining mechanisms for stopping the practice of using children in and for war. We are grateful for eminent legal scholar, Professor Tom Bennettıs contribution, which we believe brings a fresh and critical look at the historical, anthropological and legal aspect of child warriors in the African tradition. In his other work on criminalising recruitment of children in armed conflict, he makes no bones about the work that still needs to be done in this area. There is a crying need to provide a deterrent to war crimes against children and to see international justice in action. To this end ACT is calling for the establishment of a specialised international tribunal on War Crimes Against Children as a matter of the gravest urgency.

ACT the Action Plan Project for Children in Armed Conflict forms part of the activities of the Institute for Security Studies (ISS) in South Africa. The mandate of the ISS is to enhance human security in Africa and ACT supports the quest to stop the use of children in armed conflict situations by undertaking applied research and making recommendations on this issue. The present practices and abuses pose a direct threat to the dignity of humankind, they contradict all principles and rights associated with international norms and make a mockery of the notion of basic human security. Unless the present situation is immediately redressed, generations of children will continue to be exposed to a culture of violence which neither offers alternatives for intellectual growth nor contributes to peace and nation-building processes. This ultimately robs children of their future and their humanity.

I wish to thank Virginia Gamba, head of the Arms Management Programme at the ISS for her professional help and Tom Bennett for his fine and scholarly understanding of the topic Elizabeth Bennett Head, ACT project on Children in Armed Conflict




Estimates of the number of children being exploited for military purposes are inevitably only approximations. Nonetheless, from studies conducted by the International NGOs Coalition in 24 countries worldwide, it appears that more than 300 000 children, both boys and girls, are being used as soldiers, saboteurs, spies, carriers, "wives" and general camp-followers.1 This practice is becoming pervasive, as the many civil wars in Africa - ranging from Angola to Uganda and Sierra Leone - have shown. What is more, the International NGO studies show that children may constitute a significant percentage of armed forces. Of the 60 000 combatants engaged in Liberiaıs civil war, for instance, about 10 percent were children.

Force is not necessarily used to engage these conscripts. By appealing to a sense of patriotism - as happened during the prolonged conflict between Ethiopia and Eritrea - children may be persuaded to enlist. Revenge for past grievances may be the motivation - as in the Hutu-Tutsi conflict in Rwanda - or children may simply attach themselves to an army as their only hope of surviving the social dislocations of war. However, in the great majority of cases, recruitment is coercive. RENAMOıs tactic of simply abducting batches of children from their home villages during Mozambiqueıs civil war and putting them under arms is typical.

Whether children enlisted of their own free will or were forcibly  conscripted, their involvement in armed conflict presents quite obvious dangers. Not only are young people ill-equipped to cope with the physical dangers they encounter, but their immaturity poses an additional threat to the safety of other combatants. Although less obvious, the long-term social consequences are possibly even more harmful. Children taken from their families and communities are deprived of the normal processes of socialisation and education, and, when peace returns, there is little hope of veterans being successfully reintegrated into society. Instead, the child brutalised in its formative years is primed to perpetuate a cycle of killing and lawlessness. As a result, entire generations have been written off as "lost".

1 Culture, Tradition and Human Rights While the harm done to children personally and the harm done to society generally is at first sight morally indefensible, the practice of conscripting child soldiers is being justified as an African cultural tradition. Before assessing the validity of this claim, we need a clear understanding of what is entailed by culture and tradition.

Culture is an amorphous concept denoting anything that contributes to the unique character of a social group, thereby distinguishing it from other groups. It follows that culture may include artefacts, language, laws, customs and moral codes, in fact, a peopleıs entire intellectual and material heritage.2 Tradition, which is the process of transmitting knowledge and beliefs to future generations, is part and parcel of culture.

Not only is tradition the means for keeping a culture alive but it is also the means whereby items of culture gain moral authority. Thus, the continuity or persistence of a practice over time is the principal way of testing its normative value.

A peopleıs sense of their culture presupposes their conviction that they differ from other groups (together with recognition of this difference by a wider society). The differences might not be great - Serbs and Croats in the former Yugoslavia and Protestants and Catholics in Northern Ireland are well-known examples of very similar cultures - but to the people concerned they are vital. Self-awareness suggests in turn that members of the group have gained a certain distance from their culture, in other words, that they have begun to reflect upon what they do and think.

The successful justification of a morally suspect social practice on the grounds of culture needs an audience sympathetic to cultural relativism.

(More specifically, to persuade a human rights activist that using children in armed combat is right or legitimate would require a high degree of tolerance for deviation from the clear prohibition laid down in the international code of childrenıs rights.) Earlier this century, cultural relativism was something of a dogma in western intellectual  circles. For instance, the executive board of American Anthropological Association objected to the proposed Universal Declaration of Human Rights in 1948 on the grounds that it would encode a set of Euro-American values. Hence, in its comments on the provisions of article 5, which dealt with the prohibition on torture or cruel, inhuman and degrading treatment, the board refused to condemn the standards of other cultures: "what is held to be a human right in one society may be regarded as anti-social by another people."3

By the 1960s, attitudes started to change, largely as a result of work done by theoretical Marxists. They began a critical reexamination of the assumptions on which cultural relativism was based.4 Probably the most important of these assumptions was the discrete unity of each cultural group, with the implication that culture was unchanging and impervious to outside influence. Critics showed that pure cultures, ones uncontaminated by external influences, were mythical.

Once it was accepted that the boundaries between social groups can never be absolute, it followed that cultures must always be fluid and that different groups will share common practices and institutions. (No clearer evidence was needed than the global spread of that icon of American culture, Coca Cola, even in the former Soviet bloc and the Islamic world, two areas highly resistant to western influences.) Conceding this point led inevitably to the charge that all cultures are derivative.

The fact that a culture is not truly indigenous, however, does not necessarily imply that it is somehow unauthentic. Rather, the objectivity entailed in defending a culture and the existence of shared cultural practices suggest that cultural traditions are created and that the process of their creation may be contested. Deciding which elements will determine a culture involves choice. What is considered culturally typical can clearly never comprise the totality of a groupıs ideas, material output or behaviour. Only a limited number of traits - selected from clothing, language, values, habits or whatever - are chosen as emblems of cultural uniqueness. Over time, as the group changes, these distinctive features are changed. Accordingly, maintenance of a culture entails a process of selecting and discarding.

Emblems may be appropriated from another culture, such as the adoption in England of German customs for celebrating Christmas. They may be taken from a peopleıs own past. The revival of traditional African modes of dress for the opening of parliament is a recent instance from South Africaıs new democracy. Or, as with the rituals created in the 19th century to mark the coronation of British monarchs, they may be "invented" to meet the needs of the moment.

When we appreciate that culture is a conscious construction, we must realise that it can be manipulated by dominant forces in a society to achieve particular ends. Earlier, relativists had taken it for granted that culture was the spontaneous outgrowth of community practice, the product of some sort of primordial democracy. Critical scholarship of the 1960s, however, showed that particular ideas and practices were usually generated by elite groups or even imposed by outsiders.5

Possibly the best example of the latter point was the heavy emphasis the South African government placed on the distinctiveness of African cultures in order to justify apartheid. The segregation of black and white and the   further division of the African population into 10 homelands would have seemed at best arbitrary, or even worse self-serving, if the government had not rationalised its policy by generating notionally separate cultural traditions. Hence, although diverse forces had drawn the people of South Africa together to share common places of work, worship, transport etc, they were forced apart to live in separate areas, to work under different conditions, to learn in different schools, and, eventually, to take different citizenships. This costly experiment was celebrated as the logical outcome of different cultures.6

To acknowledge that a culture might not be accepted by all members of a group presents the advocate of a cultural tradition with several awkward questions. The first is whether an alleged tradition was generally observed in a society. If it was not, then by what right can a particular faction argue for its preservation? Who, in other words, is entitled to define and maintain culture? These questions prompt another inquiry. If a practice has been imposed or has been conceived of for ulterior, political ends, then the ends (and the motives for attaining those ends) should be scrutinised. All of these questions lead us, ultimately, to ask how far fundamental rights and freedoms may be compromised in order to tolerate cultural diversity.

The international human rights movement has a well-nigh invincible authority, since it rests on a claim that human rights represent universal values. By implication, these values are culturally neutral (and arguments of cultural relativism are therefore redundant). There is no need to accept this universalist claim at face value, however, for it is clear that human rights originated in western systems of law and philosophy. If that is so, an African apologist could contest the precedence given to international human rights by pointing to the fact that the continent has its own culturally unique way of securing human dignity. Since human dignity is also the value underlying all human rights, this argument has an immediate plausibility.

The first response would be that African nations have been eager to accede to international human rights instruments, especially the United Nations Convention on the Rights of the Child (1989). This treaty, with its decidedly western slant, provided the inspiration for Africaıs own Charter on the Rights and Welfare of the Child (1991).7 The second response would be to acknowledge that, although Africans are entitled to preserve their cultural heritage, the welfare of children takes priority over cultural rights.

Most human rights instruments guarantee both individual rights and a right to culture.8 Contradictions between the two are bound to arise. In the event of a conflict, which right should prevail? At the outset, it must be noted that a "right to culture" is ambiguous. The right can be interpreted to mean either an individualıs entitlement to participate in a culture of choice9 or a groupıs entitlement to demand recognition of its culture (and thus preservation of its cultural identity).10 In the former case, an individual may assert the right against a cultural group (which would have the duty to admit new members). In the latter case, the group may assert its right against the state (which would be obliged to take cognisance of the groupıs cultural tradition).

A third interpretation of the right to culture would allow a group to insist that its members remain true to their cultural tradition whatever the predilection of particular individuals. This understanding has found no support in international human rights law, principally because it would entail denying the first interpretation: the individualıs free choice to decide whether to participate in a culture or to abandon it. In view of the fact that cultures might be imposed or "created" to achieve ulterior ends, such a construction of the right to culture is prudent. Where a culture works to the benefit of a specific group, there could be no ground for disturbing an individualıs claim to fundamental human rights. For reasons such as this, the human rights movement has traditionally preferred individual rights over group rights.11

The conflict between individual and group rights is especially acute where children are concerned. Whose interests should prevail: those of the group or those of the child? African tradition stressed the welfare of the extended family and the immediate political community. Individual interests were submerged in the common weal and the normative system stressed an individualıs duties rather than rights.12 Children had no especially favoured position.13

Conversely, the most important tenet of the human rights law on children is that a childıs interests are to be given paramount consideration. This principle is encoded in the UN Convention on the Rights of the Child14 and the African Charter on Childrenıs Rights.15 Admittedly, "the childıs best interests" principle is vague, and it is said to be an open invitation to apply whatever cultural norms on upbringing happen to be current.16 Yet, because the principle is worded in relative terms, it performs the vital function of ranking contradictory claims or rules. Hence, whenever the right to preserve a culture comes into conflict with a childıs interests, the latter prevail.17 Article 21 of the African Charter on Childrenıs Rights is a case in point: it prohibits cultural practices that might be prejudicial to a childıs health or life. Not only are the wellbeing and safety of children of overriding importance but group rights to culture are relatively weaker and subordinate to other human rights. The issues above indicate that to construct what might be called an "Africanist" argument in favour of using children for military purposes is   a daunting prospect. In the first place, clear evidence would be needed of a constant, uniform usage in Africa. If such a tradition did exist, was it perpetuated into modern times or was it recreated to serve new and contemporary purposes? Related to this question is whether those asserting the cultural tradition are the only ones to support it. If the evidence discloses no general social acceptance, there seems no good reason to permit the practice, given the compelling moral weight of modern human rights norms.

2 Age Grades and Age Sets

An argument claiming the authority of a cultural tradition would need to assert something more than the chance occurrence of a particular activity or its fleeting popularity. A tradition implies that the activity was institutionalised in the sense that it was persistent and widely approved.

Age grades and especially age sets are two institutions that would have the most direct bearing on the use of children in armed conflict, since both functioned to coopt youths to the service of society.

Everyone is subject to the physical process of ageing, and there is probably no society that does not use age for the broader purpose of defining status.

The term "age grades" denotes these social divisions.18 A grade implies a fixed position within a hierarchy of statuses, through which individuals progress as they grow older.

From attributes thought characteristic of particular age groups, society may allocate appropriate roles. For instance, because aggression and physical prowess are commonly associated with young men, this group is  expected to play a warrior role. Older people, on the other hand, are thought to be repositories of knowledge and moral virtue; these attributes dictate their roles as rulers, policymakers and arbitrators. Each status carries with it various rights, duties and privileges in relation to the other statuses in the system, and differences in status are reinforced by rules stipulating approved modes of dress, deportment, eating, drinking, speaking and address.

In all parts of the world, age grades provide a basis for social stratification. In Africa, however, and in certain areas of North America, Brazil and India, they also provided a primary means of political control.19

Because age grades cut across ties created by kinship and common residence - the usual principles of social organisation - they enabled senior rulers to maintain command over young and potentially unruly subjects in the  junior grade.

Three social ages - child, junior and elder - are perhaps universally accepted, but these divisions may be considerably varied and nuanced. The Karimojong of north-eastern Uganda, for instance, emphasised only a single division between seniors and juniors. The Nuer of southern Sudan, too, ranked the total body of men into only senior and junior groups according to whether or not they had been initiated. With the Arusha of northern Tanzania, on the other hand, the division between senior and junior grades was further elaborated into older and younger subdivisions.

The senior male grade is generally taken to be norm-setting, since it is an ideal to which everyone should aspire and towards which children should be educated. Nonetheless, the subordinate junior status is the one that usually attracts most social attention. Of the three age grades recognised by the Nupe in Nigeria, for example, the junior grade is considered the crux, for it is then that young men are "educated for citizenship". When Nupe men reach 30, most have married, and the social importance of the group starts  to decline. Thus, while elderly coevals continue to maintain a degree of comradeship, their grade ceases to have any particular social function.20

Members of a junior grade are at an ambiguous age. They can no longer be treated as children but they still lack the maturity of adults. The grade is therefore a bridging period, a time when irresponsible youths will be taught to become adults. As will become apparent below, it is the flexibility of this age category that opens the way to exploitative practices: although adolescents may be physically capable of performing certain tasks, for most intellectual and emotional purposes they are still children.

Movement from one age grade to the next is generally accompanied by a rite de passage, a ritual display serving to focus public attention on changes of status. A common rite de passage, occasioned by puberty, is initiation. This ceremony involves a period of seclusion, shared hardships and special training, and it often includes circumcision (or another form of bodily mutilation). When initiates emerge, they are deemed to have shed  attributes typical of childhood, such as dependency and confinement to the family domain, and to have gained some of the qualities of adulthood.

Certain African societies had an elaboration of the age grade system, whereby members of a grade formed, in anthropological terms, "age sets". Age sets are social groupings that have many characteristics of a corporate entity. At regular intervals - varying from six to 15 years - young people of similar age in a community are collected into distinct units. At the end of this period of recruitment, admission to that particular set closes and a new recruitment can begin. To indicate that initiates are no longer children, admission to a set normally entails a ceremony of initiation.

Members remain in the same set all their lives. They are bound together by a network of reciprocal rights and duties which is designed to promote intimate, lifelong relationships. Each age set progresses through the scale of age grades as a unit. Because everyone in a particular set changes social status at the same time, transitions from childhood to youth to old age do not occur imperceptibly and individually, but by single and clearly marked occasions.

The age set system is normally associated with various East African societies, where it was a vital element in socio-political organisation. The pastoral Masai of southern Kenya, for example, had two age grades - warriors and elders - each of which contained a number of age sets.21 When a group had been recruited as a set to the warrior grade, it was segregated into a separate establishment. The whole set was subject to the command of a captain, who was responsible for discipline, planning raids on neighbours and settling disputes. When men married they moved as a unit into the elder age grade. While each set took its own decisions as a whole, the elders generally had authority to decide supervening economic and political issues.

With the Nandi of western Kenya, every 15 years or so, a group of similarly aged young men were gathered together to form a highly integrated warrior set. As they did so, their predecessors moved on to the next age grade. The more senior status attracted the usual privileges of age but it had fewer responsibilities and so tended to degenerate in social importance.22 The Nyakyusa at the northern end of Lake Malawi marked the separateness of each age set in a manner similar to the Masai by detaching members from their families and housing them in their own distinct villages. For the remainder of the membersı lives these villages then functioned as independent units.23

In southern Africa, too, age sets were common (although they differed from the East African institution by functioning as an adjunct to the chieftaincy rather than the main basis for organising political power).24 Among the Sotho and Tswana, every four to seven years, both men and women were liable to be recruited into age sets at the time of their initiation. All those who had passed puberty since the last school would be summoned to attend.

(Siblings were deliberately kept apart; hence, a younger brother would have to wait until his older brother had been initiated.) Through the medium of songs, riddles, dramas and dances, initiates were taught traditional values (especially the respect and obedience due to seniors) and a sense of solidarity with their age mates.

While age sets were a distinctive feature of societies, such as the Masai, with pastoral economies, no reductive correlation is possible between the age set system and a particular type of economy. Agriculturalists, such as the Arusha and Chagga, who live near the Masai in Tanzania, also had age sets.25 Nor is the age set system the preserve of a particular type of political structure. The Masai were an acephalous (or so-called "stateless") society, in that no one person or group exercised full political powers, but age grades and sets were also a significant feature of the highly centralised Edo state of Benin.

By about 1440, under its ruler Oba Ewuare, Benin had become a major power in the area that stretches from modern Lagos to the Niger Delta.26 In every village under the cityıs control, men were graded as youths (irhoghae), warriors (ighele) or elders (edion). Unlike the Masai and other peoples of East Africa, no ceremonies marked the formation of age sets or the movement from one grade to the next. Whether a boy was a youth and whether a batch of youths should move into the warrior grade was decided by the elders as and when necessary.

The members of each grade had responsibilities appropriate to their age and position. Hence, youths from about 12 to 30 years of age were expected to perform lighter public duties, such as constructing buildings, maintaining roads and paths, repairing shrines or council houses and carrying the twice-yearly tribute of yams and palm-oil to the Oba in Benin. Warriors in the ighele grade, ie those who were over 30 but not yet elders, acted as a local police force and performed communal tasks requiring special skills.

For military purposes, the ighele was subdivided into a senior grade (which was used for fighting) and two junior grades (which were called up only in emergencies). Elders (edion), those who were about 60 years or older, formed village councils. They appointed leaders for the other grades, controlled access to village lands, collected dues from outsiders and organised communal works. Elders were exempted from the more arduous tasks,  especially military duties.

3 Military Action and Socio-Political Structures

From what has been said above, it is evident that, although age grades were omnipresent in sub-Saharan Africa, age sets were found in a more limited (albeit economically and politically diverse) range of societies. Whether a society was organised only on an age grade system or into age sets too, the duty to fight for the realm would have fallen on junior men. The argument that African tradition supports the use of children in armed conflict therefore begins on the apparently sound basis of a well-established social institution.

Nevertheless, whether a community would expect children to participate in military activities would depend, in the first instance, upon whether it faced a situation of attack or defence. To defend against an armed attack, children, together with all other available bodies, could be called upon to do whatever lay within their capabilities.27 Compared with initiating an attack, however, defence is a less considered activity. Thus, using children to ward off an imminent threat to life and limb can hardly be regarded as an institutionalised practice.

Inducting children into combat groups in preparation for an armed attack, on the other hand, would be a matter of deliberate policy and so could be deemed institutional. And it is significant in this respect, that the use of age sets is normally associated with strategies of attack rather than defence. What is more, because the age set system allows armed forces to be rapidly marshalled for battle, such sets are often a feature of aggressively expanding societies. It follows that, if the argument in favour of conscripting child soldiers rests largely on the age set tradition, children could be expected to have played a military role in polities embarking on a career of territorial conquest.

As it happens, only a small minority of the pre-colonial African polities fell into this category. These were typically of a centralised or "state" structure, as opposed to a decentralised (or acephalous) structure.28 To describe a polity as "acephalous" implies that no particular person or group wields absolute power, and, without the organising potential of a political superstructure, the society must depend for its cohesion on bonds of kinship. Hence, in a so-called "stateless" society, families are more or less autonomous.

Very few societies fully realised this theoretical model. Families of hunter-gatherers, such as the San in Botswana and the Mbuti in Congo, who recognised no superior political unit, came closest (and even here families cooperated for occasional economic or defensive purposes). Masai and Nuer polities were also apparently acephalous, because, although senior males exercised a certain degree of political authority, the people had no formal leaders with supra-familial powers. Conversely, in another type of acephalous structure found in many areas of Africa - examples would be the Tonga in Zambia, the Kikuyu in Kenya and the Ibo in Nigeria - groups of families were drawn together by kinship and collective economic activities into permanent village communities administered by councils of senior elders.

While military aggression and territorial aggrandisement are not usually associated with the stateless societies described above, the relative simplicity of their political structure did not necessarily militate against conflict. Acephalous societies could be just as violent and destructive as their centralised counterparts.29 Nonetheless, their capacity for warfare was restricted, mainly by the availability of manpower.30

For instance, historical evidence shows that conflicts among the pre-state societies of 17th- and 18th-century southern Africa were a matter of petty raids and counter-raids. Because manpower was needed primarily for hunting, herding and farming, a fighting force could not be kept in the field indefinitely. Warriors were mobilised for short periods when necessary.

Without the means to sustain extended campaigns, conflicts tended to be episodic and battles were short and limited to the respective groups. Common causes for dispute were quarrels over territorial boundaries, rights to land and water, stock thefts, interference with visiting subjects and refusals to hand over fugitives from justice. If the aggrieved party did not succeed in gaining satisfaction through diplomatic means, violence could well ensue. A formal challenge might be issued and a day fixed for battle.

Warriors in each section of the realm would be called up by their leaders.

When the entire national force was assembled, it would proceed to battle, often urged on by women and children. Alternatively, the tactic might be stealth. An enemy village might be surrounded at night and attacked at dawn. In the assault, huts would be burned, adult men would be killed and the victors would then retreat carrying with them livestock, women and children from the vanquished community. Retaliation might follow and hostile relations  could persist until one party sued for peace. Nevertheless, offensive operations could normally not be sustained for longer than a week or two.31

The nature of intergroup conflict changes when centralised or "state societies are involved. These polities are characterised by the concentration of power and authority in a single person or group, cleavages of rank and wealth (often cutting across kinship ties), systems of administration that bind together disparate communities, regular sources of revenue and, of course, organised military force.

By the 16th century, polities of this type had evolved in both western (Ghana, Mali, Songhai, Kanem-Bornu, Hausa and Benin) and central Africa (Kongo and Mwenemutapa). By the 19th century, the process of state formation had intensified, but several polities, such as the Tswana chiefdoms in Botswana, still fell short of fully centralised state structures. Particular leaders ("chiefs" in colonial parlance) might have acquired power, but their rule depended on cooperation with councils of elders drawn from the major families of the realm. Because the leaderıs authority was precariously poised on his powers of patronage and prestige and the competing powers of rival families, such societies were continually fracturing and regrouping.

In the process of creating states, the raids and counter-raids typical of the conflicts in acephalous societies were replaced by military force of a different and more destructive nature. The classic state presupposed a permanent power that would be capable of expanding territorially and permanently subjugating conquered groups. Earlier  methods of mobilising all able-bodied men for raids or reprisals had been viable because operations were sporadic; and a large mass of warriors could often compensate for what might have been lacking in discipline and technology.

If new peoples and territories were to be subdued, however, experienced standing armies were essential. Such forces were sometimes obtained by recruiting mercenaries, pressing slaves into service or even by engaging the support of a European power.32 A more radical solution was to establish anational army by instituting changes to the structure of society as a whole.

4 Age Sets in the Zulu Kingdom

A classic, and a particularly well-documented example of the latter process is the formation of the Zulu state in the south-eastern part of South Africa at the turn of the 19th century.33 Although this development had parallels elsewhere in the region - among the Sotho under Moshoeshoe, the Swazi under Sobhuza and Mswati, the Pedi under Thulare and the Ngwato under Kgama - the scale and speed of the phenomenon make it exceptional.

Various explanations have been put forward to account for the appearance of the Zulu kingdom, ranging from the demographic and the environmental to competition for control of trading links with Delagoa Bay. While causes may be disputed, there is general agreement about the means used to forge the new regime. Existing age sets were transformed into military regiments which provided a fighting force capable of welding together the remains of more than 100 different peoples under a Zulu aristocracy.34

Before the Zulu ascendancy, age grades and age sets were, by all accounts, ubiquitous in the region.35 Originally, these sets had been formed periodically when a local ruler called together young men of roughly the same age from the homesteads under his authority to undergo initiation and circumcision. Among peoples of the eastern seaboard, however, initiation was (and in the southern areas still is) a local affair that placed little emphasis on fighting or transcendent national values.

It is true that among the Sotho-Tswana peoples of the interior initiation was organised at a national level and that considerable stress was laid on acquiring national identity.36 In this case, members of the age sets were more readily turned to the service of the ruler - and all age sets had performed military duties. When combat was impending, men from each family within a neighbourhood would be mustered at a central homestead. This detachment would then be marched to the chiefıs homestead where it would be joined by others to constitute the national army. Such a force, however, was bound to be loosely structured. Because neighbours and kinsmen fought side by side, they preserved their parochial loyalties and in consequence a tendency for independent action.

According to popular tradition, Dingiswayo, leader of the Mthethwa, was responsible for transforming the traditional age sets (sg:ibutho pl:amabutho) into armies of conquest. Yet it is unlikely that the idea originated with him. Reliable evidence indicates that, in the 18th century, various other peoples in the area - the Mabhudu, Ndwandwe, Hlubi, Qwabe and Ngwane - already had military regiments of this nature. Oral traditions further claim that Mabhudu and Ndwandwe age sets were engaged in campaigns of conquest.

Even so, the Mthethwa seem to have used amabutho for the additional purpose of imposing their rule within the realm. As the Mthethwa had expanded, they had incorporated conquered groups. While the defeated were allowed to retain a measure of autonomy, they became liable, as subject peoples, to pay tribute. Amabutho were used to extract these payments. Hence, as an internal police force, amabutho were active in controlling subordinate peoples and establishing a centralised and stratified state.

Use of the amabutho to secure control internally and to spread Mthethwa dominion externally was probably responsible for generating a dynamic of raiding and conquest. The longer amabutho were kept in the field, the greater the need to acquire cattle as sustenance and as a reward for services. Because this demand could be met only by capturing more cattle, cattle-raiding became, from the late 18th century onwards, a structural necessity for maintenance of the Mthethwa state.

The demand for cattle intensified when the south-eastern region of Africa was devastated by drought in the early 19th century. At the same time, relations between the rival Mthethwa and Ndwandwe finally erupted into war.

The Mthethwa were defeated and Dingiswayo was killed. But the Zulu, a client group of the Mthethwa, under their legendary leader Shaka, survived. In 1819, in a decisive battle at Mhlathuze River, the Zulu overcame the Ndwandwe and expelled them from the region.

Shaka then immediately launched an ambitious campaign of territorial expansion. Because Zulu dominance relied even more heavily on controlling a fast-growing number of subjugated and potentially rebellious people, the age set system was extended. Young men of like age from all over the conquered territories were drafted into amabutho. By dispersing neighbours and kinfolk into different regiments, this system had the effect of undercutting local loyalties and family ties. Affiliation to the regiment and state now became paramount.

The use of age sets to provide a permanent national army coincided with the disappearance of circumcision. Previously, all the peoples in South Africa had regarded this custom as mandatory to distinguishing boys from adult men:any uncircumcised male would have been socially scorned, denied participation in council meetings and deemed unmarriageable.37 Enrolment into amabutho now became the principal rite de passage. Tradition  attributes the abolition of circumcision to Dingiswayo, who is said to have ordered all impending ceremonies be deferred until his conquests were complete, but the reason why the practice was so suddenly abandoned is in fact obscure.38

The social origins of amabutho lay in the large homesteads of the Nguni people, where children of similar age constantly associated with one another and played together.39 Those living in a group of neighbouring homesteads formed a natural group (sg:intanga pl:iintanga) and were treated as a unit by adults. From early childhood members of these units were imbued with a martial spirit. Even before puberty, neighbouring age sets were expected to fight one another, and, although the mock battles involved sticks instead of spears, they prepared children for much tougher contests in later life.

Fights between members of the more senior iintanga, spurred on by their neighbours and relatives, were a constant occurrence.

Ceremonies marking puberty divided the junior iintanga from the senior. The local ruler of a district (induna) would choose an older person (often someone belonging to the next age set in seniority) to be the leader of an intanga. That person retained his position for life. The intanga continued to take on members from a neighbourhood until enough had been received to form a regiment. It then became a fixed group admitting no more recruits.

When the king decided to constitute a new regiment, officials were instructed to send boys of between the ages of about 18 and 20 to the royal homestead for ukubuthwa (drafting or enrolment).40 Even before an enrolment occurred, boys might abscond to a regimental cantonment to await the event.

They might run away one or two years after undergoing puberty ceremonies (or whenever they plucked up the courage). Until the ukubuthwa, however, these boys were not active fighters. When the army went into attack, they drove the accompanying cattle and carried the food and beer.41 Otherwise, they were expected to use their time productively by learning the arts of battle and by attending to the more menial chores of herding cattle or bearing shields and mats for senior warriors.

While the duration of military training depended on the kingıs will, it generally used to last about six months. Each ibutho had its own name, insignia (such as headgear and shields) and regimental officers. For purposes of speedy mobilisation, troops were assigned to specially built royal cantonments (sg:ikhanda pl:amakhanda). Each ikhanda was under the control of a military commander appointed by the king.42

The arduous drill and the complex regulations governing life in the regiments were obviously designed to instil discipline and a military temperament. As with the iintanga of children, rivalry between groups was encouraged and they were frequently set to fight one another. Training was also intended to inculcate a sense of loyalty to the king and to induce warriors to identify him as their leader and source of their welfare.43

Although the regimentsı prime duty was to fight, they also had economic functions.44 One of the consequences of transforming the traditional age set system into an organ of state was to divert labour power from the menıs natal homesteads to public purposes. Thus warriors serving in amabutho were responsible for guarding the kingıs herds; they built their own quarters, collected firewood, carried water, hoed the kingıs fields and did any other public works that he ordered.

The Zulu state assumed command over the lives not only of men but of women. While female age sets had origins in the late 18th century, Shaka appropriated the institution, thereby ensuring state control over the processes traditionally dominated by women: reproduction and agricultural production. Previously, female initiation had been conducted at the homestead of a girlıs father, with the principal aim of ensuring fertility after marriage. In the heyday of Zulu power, however, a girlsı intanga was formed to match to every boysı intanga, and, when boys were drafted into regiments, girls were drafted into sister regiments.45

Female age sets had no active military duties; they served the state mainly as pools of labour. By regimenting the lives of women, the state could also control marriage and the formation of new families. Members of male and female regiments were permitted to marry only when the king decided to stand the regiments down. Long delays occurred in doing so, with the result that in Shakaıs reign women of up to 40 years of age still did not have husbands.46

5 Age Sets Under Colonial Rule

A question that should now be considered is whether militarised age sets, such as the Zulu amabutho described above, persisted into the post-colonial period, since the argument that African culture endorses the use of child soldiers would be greatly strengthened if it could be shown that traditional institutions had survived the imposition  of colonial rule. Of course, even if an institution were no longer a living practice, its existence as a past tradition could still be a legitimate basis for the Africanist argument, but the element of continuity is important in this context too.

When a particular group is casting about for some suitably convincing justification for its views, and when the basis of its claim is an obsolete institution, the argument in favour of contemporary human rights norms becomes all the more compelling, for these norms do not displace an existing cultural practice. To the contrary, the argument for using child soldiers then appears suspect, because it suggests that a dominant group is attempting to impose its will on society at large.

The first aim of the European colonisers was, of course, to suppress local resistance to their rule. Hence, defeat or disablement of the military power concentrated in the age set system would have been a self-evident priority.

As it happened, colonial authorities took no action to prohibit the formation of age sets. Instead, the pre-colonial institution disintegrated only when its functions disappeared and recruits fell away. In Zululand, for example, although the amakhanda were disbanded, the amabutho survived, albeit with only ceremonial functions.47

One of the immediate effects of colonisation was to put an end to internal conflicts among subject peoples. Another was to transfer responsibility for defence against external aggression to armies conscripted from the metropolitan states. Indigenous regiments therefore lost their military duties. Even so, colonial administrations were not above exploiting existing age sets to achieve their own imperial objectives.

On occasion, warriors from collaborating or conquered nations were turned against less tractable tribes elsewhere in the colonies.48 Colonial authorities could, of course, have conscripted members of age sets into their own armies. At the turn of the century, this option had a considerable body of support in Britain, since it was felt that traditionally warlike peoples, like the Zulu and Hausa, could be profitably harnessed to imperial defence. According to the opposing school of thought, however, which initially at least prevailed, Africans were unsuitable recruits. Not only were their loyalty and reliability suspect but the British believed that Africans should not be used to fight white menıs wars.49

As it turned out, South Africa was one of the few territories to resist the so-called "Gurkha syndrome". Under the 1912 Defence Act, men between the ages of 17 to 60 could be called up; but only white males were liable formilitary service. In other British territories, the massive loss of lives in the First World War led to voluntary recruitment and eventual conscription of Africans. By the middle of the Second World War, 400 000 were serving in the armed forces, most as labourers and non-combatants but a significant number as active soldiers.50

French policy towards the recruitment of indigenous armies was quite different.51 As early as 1912, France had introduced universal male conscription and, unlike Germany or Britain, it was prepared to use African troops in any part of its empire. 170 000 Africans - drawn primarily from the supposedly martial tribes in French West Africa, such as the Bambara, Malinke and Wolof - served in the French Army in the 1914-18 War.

Immediately thereafter, in 1919, conscription was intensified to make up for French losses on European battlefields. Any male between the ages of 20 and 28 was expected to serve for four years in the French army.

It can readily be imagined that service in colonial armies was far from popular with the African population. Conscription was widely resisted and where possible evaded. The colonial authorities therefore turned to traditional rulers for assistance, and, when chiefs were called upon to deliver able-bodied men to the government, they looked to members of the age sets.

Traditional authorities had similar, personal interests in preserving the age set system. In the 1940s in Bechuanaland, for instance, regimental members were regularly being called upon to contribute labour for community projects (and, of course, for the chiefsı own establishment). Moreover, to build up local revenues traditional leaders were sending men abroad to earn money as migrant labourers in neighbouring South Africa. As an institution, therefore, circumstances were favourable to persistence of age sets.52

Yet, ultimately, colonial and chiefly reasons for maintaining the age set tradition did not prove as strong as the forces undermining it.53

Christianity was one such force. The rituals involved in initiation and recruitment, for instance, were often regarded as morally offensive, and missionaries forbade participation by their converts. The force that proved to  be most destructive of the old order, however, was the colonial economy.

Throughout Africa, chronic shortages of manpower threatened the viability of colonial ventures into farming and mining. Once again, members of the junior age grades and age sets provided a ready source of labour. British authorities normally resorted to indirect fiscal devices, notably hut taxes, to force able-bodied men onto the labour market.54 The French and Portuguese forcibly conscripted men into labour corvées, and in Benin, too,  British officials simply appropriated the services of men in junior age grades to provide a free supply of carriers and workers on the roads.55

Young men therefore had to leave home early to find employment on farms, in the mines and in cities. Because potential recruits for the age sets were working as labourers they were either unavailable when new regiments  were formed or unable to afford the time for training. As a result, the system was disrupted.56

6 The Definition of Childhood

An issue central to the argument for conscripting children into military service is whether African traditions involved children. Many of those currently bearing arms in Africa are 10 years old and some are even younger.

Were people as young as this conscripted into the pre-colonial age sets? The answer to this question obviously depends upon what is meant by a "child".

Every society withholds some, if not all, powers and responsibilities from the young until they can behave in a mature and fully rational manner.

Partly to protect the child and partly to protect the wider society, children are placed under the control of a guardian. In literate societies adulthood is usually ascribed to individuals when they reach certain predetermined ages. The transition from childhood is therefore fixed at an age when it is presumed that individuals will be capable of conducting themselves as adults. In preliterate societies, on the other hand, where it is not possible to keep exact records, movement from one age category to another is related to physical processes, such as puberty, and to social events, such as initiation.

There would be a large measure of agreement between the fixed-age system of written codes and the more flexible African system on what should be regarded as a totally dependent infant. Both systems treat individuals under an approximate age of seven as socially powerless. Adolescence, however, is a period of ambiguity, when individuals are to some extent capable of responsible behaviour but are still in need of guidance and protection. In these circumstances, status as a child is flexible and thus open to manipulation to achieve ulterior ends.

Cultural stereotypes of ageing, together with the broader demands of the economy and society, are clearly the major determinants of social capacities. A particular type of economy may prove crucial in shortening or lengthening the duration of childhood. If people are living at a subsistence level, for instance, the capacities and responsibilities of adulthood begin early. Because an average life span is short and survival is a struggle, a long period of dependency as a child is a luxury that families cannot afford.57 Further variables dictating the termination of childhood can be found in other, more specific social demands, and, for different purposes, attainment of particular capacities may be staggered at different ages.58

Full adulthood should therefore be seen as the outcome of a gradual process.59 In Africa this process began at about the age of six or seven, when children took their first step from infancy. They would then be expected to contribute to the running of the household. Boys would usually herd livestock or keep birds away from the crops, while girls would attend to domestic chores. While this occasion was not necessarily marked by any special ceremony, some peoples, such as the Nguni, slit or pierced a childıs ear lobes, thereby "opening its ears". This minor rite de passage indicated that the child now had sufficient intellectual ability to respond to reasoned  nstruction.60

The next stage in a childıs social development was puberty. As a physical manifestation of maturity, puberty is almost universally a time for performing dramatic public rituals to signify enhanced status. Groups of boys reaching puberty (and usually girls too) would therefore be segregated from the community to undergo special training.61 In these initiation schools the children were taught about the duties owing to their families, their ancestors and to the wider community.

Initiation admitted youths into adulthood for some but by no means all social purposes. For example, boys who had been initiated were treated with greater respect than mere children, and they were no longer subject to the discipline of their mothers and other female relatives. Nonetheless, an initiate did not become completely independent of his guardian. He was expected to remain at home, as a ward of his family head, who would continue to manage his property and legal affairs.

Adulthood was achieved only with the accumulation of further criteria (such as enrolment into a military age set). Perhaps the most basic requirement designating a totally independent male - few, if any African societies permitted women this position - would be a manıs ability to act for himself and in his own interests. This capability presupposed access to and control of the material means of support. Hence, marriage and the establishment of a new homestead were two prime indications of an adult male, although even persons who met these requirements might not be completely separated from their natal families. A man might well remain subject to his fatherıs moral and spiritual authority until the latter died.

Anyone who had not attained the adult status of a male, whether a woman, a child, an unmarried son or even a married son who had not yet set up his own establishment, was to a greater or lesser extent denied full capacity. The family head would be responsible for such dependants: he was obliged to give them physical support, he was liable for their wrongdoing and he would be required to argue their cases in court. Conversely, dependants had to submit to his discipline and their income would be pooled as a family resource.

In this scheme of things, men would have become members of age sets and hence liable to perform military duties before they married. The reason for conscription at this stage of their lives had to do with their physical prowess and their lack of attachment to wives and children. (In the Zulu regimental system, for instance, marriage was deliberately delayed in order to keep men on active service.) This generalisation is as true of the traditional African age sets as it is of modern western systems of conscription.

Although the warriors of pre-colonial African armies would not have attained complete adult status, children were not recruited to regiments nor did they bear arms. At most, as we have seen in the case of the Zulu, children gave incidental support as non-combatants. While a degree of uncertainty might always exist about a recruitıs precise age, the evidence is clear that men were drafted into regiments three or four years after puberty, and it should also be noted that (contrary to current practices) girls were never used as combatants. Coopting only relatively mature men was less a matter of principle than of pragmatism: armies need men who can fight effectively.

Warrior status was therefore reserved for relatively mature, able-bodied men. On attitudes to recruitment in the post-colonial period, we have no clear evidence, due mainly to the disintegration of the age set system. We do know, however, that traditional rules for regulating attainment of adulthood have been transformed as the result of the socio-economic changes that came in the wake of colonialism. Throughout Africa, urbanisation has fragmented extended families. Breakdown of the family has, in turn, resulted in an ever-increasing number of illegitimate births, runaway children and single-parent families. Through wage labour, sons and daughters have gained financial independence of their guardians. Christianity and Islam have discouraged performance of initiation ceremonies, and everywhere tradition has succumbed to a modern secular culture.62

People in all parts of Africa are now aware that rewards are to be won only through a western style of education; and, for those attempting to escape from a life in subsistence agriculture, the period of childhood dependency may be prolonged. When a stable form of employment in commerce or the state bureaucracy can be found only with school qualifications, children remain as charges of their families until secondary or even tertiary education has been completed.

Further changes to the old order came about with the introduction of European laws to the colonies. The new legal regimes displaced, at least partially, indigenous rules on attaining adult capacity. For purposes of marriage, succession and land tenure, the colonial powers were prepared to tolerate traditional systems of law and custom, since they had little interest in how Africans regulated their domestic affairs. But the economic and political framework of the state was another matter. Here European laws were imposed. Hence, where public administration (and especially, of course, liability to do military service) was concerned, the new law was applicable.63

This legal regime has persisted even after African states won their independence. South Africa is a typical example. Section 28(3) of the Constitution now fixes 18 as the age for distinguishing children from adults for purposes of enjoying constitutional rights and the various public benefits of schooling and healthcare. The age of 18 was chosen, because it is now an international norm64 and because a fixed age removes the uncertainties of traditional African methods for defining child- and adulthood.

7 The Consequences of Violating Childrenıs Rights

Because the military traditions of Africa neither permitted nor encouraged the conscription of children (especially girls), whether the concept of "child" is defined in African or international terms, the argument in favour of children bearing arms would fail. With not even the justification of culture to support this practice, what consequences should attach to breach of an international human rights prohibition on the recruitment of children? States that ratified the UN Convention on the Rights of the Child or the African Charter undertake to protect children within their jurisdiction from abuse and exploitation. Special provisions in both Conventions oblige states to refrain from recruiting children and to ensure respect for rules of international humanitarian law which affect children during armed conflicts.65 States parties are also obliged to pass appropriate legislative and administrative measures to implement the rights listed in the treaties.66

The number of states prepared to  undertake these obligations was surprisingly high; but then a verbal assent to the morally unquestionable cause of children cost little. Compelling signatories to carry out their undertakings, however, is another matter. The major enforcement mechanism in the treaties is a reporting/inquiry procedure. Under the UN Convention parties must submit regular reports to an international committee on steps they have taken to fulfil their treaty obligations.67 Under the African instrument a like committee may initiate its own investigations about partiesı efforts to implement the Charter.68

These committees have no authority to receive complaints from other states, NGOs or, more significantly, from individuals. Nor do the committees have any power to penalise states that have violated childrenıs rights. They can do no more than offer an offending state their comments and recommendations for improvement. In the absence of any material sanction against offenders, child conscripts must rely on the legal infrastructure (which is often inadequate) and political goodwill of states to take action on their behalf.

A government could conceivably be sued in its own courts for failing to protect children, if a concerned NGO or childrenıs organisation were prepared to take up the issue. Nevertheless, the success of such an intervention would depend upon whether the stateıs constitution included a justiciable bill of rights or upon whether the international Conventions had been fully incorporated into its legal system. Because there are so many obstacles in the way of holding states responsible, a simpler option would be to impose criminal liability on the persons who conscript children. Establishing the necessary elements of the crime in particular cases might be troublesome - was the child coerced into bearing arms? did it voluntarily become attached to an army? how should distinctions be drawn between bearing arms and performing non-combatant duties? - but this approach has the merit of fixing responsibility on individual military commanders. What is more, if the crime is deemed an international offence, prosecution of the offender would not be left to the vagaries of the government in whose jurisdiction the offence was committed.

Any state is entitled to try and punish those who commit international crimes.

In the circumstances, criminalising the exploitation of children for military purposes seems the most sensible way of eliminating this extremely harmful practice. It is to be hoped, therefore, that a specific offence of this nature will be included in the charter of the proposed International Criminal Court and, of course, in the charter of the Africa Court.



1 These figures were published in the UN-sponsored Graça Machel Report, Impact of Armed Conflict on Children, UN Department of Information, A/51/306, New York, 1996.

2 For a definition, see D Kaplan & R A Manners, Culture Theory, Prentice Hall, Englewood Cliffs NJ, 1972, pp.2-4 and, for a general text on the topic, see E E Roosens, Creating Ethnicity: the process of ethnogenesis, Sage Publications, London, 1989.

3 Statement on human rights, American Anthropologist, Menashe, 49, 1947, pp.539ff.

4 See the papers in D Seddon (ed), Relations of Production: Marxist approaches to economic anthropology, Frank Cass, London, 1978 and Kaplan & Manners, op.cit., pp.5-8 and 37-8.

5 See Thornton in E Boonzaier & J Sharp (eds), South African Keywords: the uses and abuses of political concepts, David Philip, Cape Town, 1988, pp.17-24, for an account of how the African cultural tradition in South Africa was manipulated by the government to further apartheid.

6 See W F Lye & C Murray, Transformations on the Highveld: the Tswana and Southern Sotho, David Philip, Cape Town, 1980.

7 Although the latter attests to the value of African culture - the preamble, for instance, urges African states to take "into consideration the virtues of their cultural heritage, historical background and the values of African civilisation" - it clearly drew its inspiration from international norms.

8 Thus article 27 of the International Covenant on Civil and Political Rights provides that: "persons belonging to ... minorities shall not be denied the right, in community with the other members of their group, to enjoy their own culture ...."

9 Which is governed by article 17(2) of the African Charter on Human and Peopleıs Rights and article 30 of the UN Convention on the Rights of the Child.

10 Article 22 of the African Charter on Human and Peopleıs Rights.

11 See F Kaganas & C Murray, The contest between culture and gender equality under South Africaıs Interim Constitution, Journal of Law & Society, London, 21, 1994, p.409ff.

12 See Mojekwu in J L Nelson & V M Green (eds), International Human Rights: contemporary issues, Human Rights Publishing Group, Stanfordville NY, 1980, pp.86-7 and MıBaye in K I Vasak (ed), The International Dimensions of Human Rights, Greenwood Press, Westport Conn, 1982, pp.588-9. Hence Chapter 2 of the African Charter on Human and Peopleıs Rights lays down a series of obligations that individuals owe their families and society, and art 29(1) obliges each person "to preserve the harmonious development of the family and to work for the cohesion and respect of the family; to respect his parents at all times, to maintain them in case of need".

13 And article 31 of the African Charter on Childrenıs Rights obliges children to preserve African cultural values and family cohesion and to respect parents and elders.

14 Article 3(1).

15 Article 4.

16 See Alston and Parker in P Alston (ed), The Best Interest of the Child: reconciling culture and human rights, Clarendon Press, Oxford, 1994, pp.19-20 and 39, respectively.

17 See Alston, op.cit., pp.15-16.

18 Two standard-setting texts on what follows are S N Eisenstadt, From Generation to Generation: age groups and social structure, Free Press, Glencoe Ill, 1956 and F H Stewart, Fundamentals of Age-Group Systems, Academic Press, New York and London, 1977.

19 Thus S N Eisenstadt, Africa 24, 1954, London, p.105 claims that the existence of age sets implies the development of a society from a kinship base to a more centralised socio-political structure. Cf I Schapera, Government and Politics in Tribal Societies, C A Watts & Co, London, 1956, pp.216-7.

20 See S F Nadel, Black Byzantium: the kingdom of Nupe in Nigeria, OUP, 1942, pp.115ff.

21 See P Spencer, The Maasai of Matapato: a study of rituals of rebellion, Indiana University Press, Bloomington and Indianapolis, 1988, cf.5ff.

22 See G W B Huntingford, The Nandi of Kenya: tribal control in a pastoral society, Routledge & Kegan Paul, London, 1953, ch.3.

23 M Wilson, For Men and Elders, International African Institute, London, 1977.

24 D F Ellenberger (translated by J C MacGregor), History of the Basuto; ancient and modern, Caxton Publishing Co, London, 1912, pp.264 and 272.

25 P H Gulliver, Social Control in an African Society: a study of the Arusha, Routledge & Kegan Paul, London, 1963, ch.3 deals with the Arusha. Evidently, by 1800 Masai raids had induced the Chagga to introduce this system in order to integrate disparate families into closer political units. See J Iliffe, A Modern History of Tanganyika, CUP, 1979, p.59.

26 See generally R E Bradbury, Benin Studies, OUP, London, 1973, pp.170-5 and P A Igbafe, Benin under British Administration, Longman, London, 1979, pp.13-15.

27 E E Evans Pritchard, Political structure of Nandi-speaking Peoples, Africa, 10, London, 1940, pp.260-7 argues that defence is per force organised on a local basis.

28 This distinction, although by no means problem free, is frequently drawn in analyses of pre-colonial African polities. See M Fortes & E E Evans-Pritchard (eds), African Political Systems, OUP, London, 1940, Introduction.

29 L H Keeley, War before Civilisation, OUP, London, 1996, ch.2.

30 And due partly to the principle of proportionality, which is inherent in feuding relationships that may develop between family or village groups. See the classic study of the Nuer by E E Evans-Pritchard, The Nuer: a description of the modes of livelihood and political institutions of a Nilotic people, Clarendon Press, Oxford, 1940, pp.155-62.

31 See accounts given by M Hunter, Reaction to Conquest, 2ed, OUP, London, 1961, pp.400ff and J H Soga, The Ama-Xosa: life and customs, Lovedale Press, 1931, ch.4 of military actions in the southern eastern areas of South Africa.

32 In fact, by the middle of the 19th century, most standing armies in Africa were being supported by alien powers: R Hallett, Africa to 1875, Heinemann Educational Books, London, 1974, p.264.

33 See J Wright & C Hamilton in A Duminy & B Guest (eds), Natal and Zululand from earliest times to 1910, Univ of Natal Press, 1989, pp.62ff and J D Omer-Cooper, The Zulu Aftermath: a 19th-century revolution in Bantu Africa, Longman, London, 1966, pp.28-9

34 Offshoots of Mthethwa and Zulu exploits - the Swazi, Ndebele and Zwangendabaıs Ngoni - carried the military age set system far into the interior of Africa. For the Swazi see H Kuper, An African Aristocracy - rank among the Swazi, OUP, London, 1947, pp.119ff and The Swazi; a South African kingdom, Holt, Rinehart & Winston, New York, 1963, pp.52ff.

35 See generally W Hoernle in I Schapera (ed), The Bantu-speaking Tribes of South Africa, Maskew Miller, Cape Town, 1956, pp.84ff and Hunter in M Hunter & L Thompson, The Oxford History of South Africa, vol 1, OUP, London, 1969, pp.124-5 and 159-61.

36 See H O Mönnig, The Pedi, Van Schaik, Pretoria, 1967, pp.122-4 and I Schapera, Married Life in an African Tribe, Penguin, Harmondsworth, 1971, pp.229-31.

37 Circumcision was (and is) still practised by peoples neighbouring the Zulu, such as the Xhosa, Thembu and Mpondomise. The Pondo also stopped it, however, under Mqikelaıs reign. See Hunter op.cit., (n31) pp.165 and 396.

38 Omer-Cooper, op.cit., p.27 fn1. He argues that the period of seclusion needed for circumcision and initiation deprived the tribe of fighting strength and left initiates vulnerable in times of war. See too E Krige, The Social System of the Zulus, Longmans, Green & Co, London, 1936, pp.116-17.

39 Krige, op.cit., pp.36ff.

40 Described by Krige, op.cit., pp.106-7.

41 See Krige, op.cit., p.273.

42 Krige, op.cit., pp.262 describes the distinctive regalia for each regiment; at pp.404-7 she lists the regiments and at pp.264-6 the amakhanda.

43 The kingıs position in fact corresponded with the propaganda. He controlled the nationıs major rituals; he supplied the troops with arms and regalia; he fed them with cattle from the royal herds and grain and beer from the royal crops. A personıs only hope of advancement (and even of marriage) lay with the king. Omer-Cooper, op.cit., p.36.

44 Guy in S Marks & A Atmore (eds), Economy and Society in Pre-Industrial South Africa, Longman, London, 1980, pp.115-17.

45 Krige, op.cit., p.38.

46 Omer-Cooper, op.cit., pp.34-5.

47 Enrolments therefore continued to be called every seven or eight years. See, for instance, Krige, op.cit., pp.108ff.

48 As the Germans did in the Hehe wars: Temu in M H Y Kaniki (ed), Tanzania under Colonial Rule, Longman, London, 1979, p.114.

49 See D Killingray, The idea of a British imperial African army, 20, Journal of African History, London, 1979, p.421.

50 Although most of the combatants were put into service in Africa, some were sent as far afield as Burma and Ceylon. See D Killingray & R Rathbone (eds), Africa and the Second World War, Macmillan Press, London, 1986,  esp pp.15 and 68ff.

51 The main impetus in French policy came from the governor of Senegal, Faidherbe, who created an indigenous regiment in 1857, the Tirailleurs sénégalais. See generally M Echenberg, Colonial Conscripts: the Tirailleurs Sénégalais in French West Africa, 1875-1960, Heinemann, Portsmouth, James Currey, London, 1991 and A Clayton, France, Soldiers and Africa, Brasseyıs Defence Publishers, London, 1988.

52 Schapera in Schapera, op.cit., p.382. Thus, although initiation was becoming less relevant to social life in Bechuanaland, membership of a regiment was still considered "indispensable to full recognition as an adult". Schapera op.cit., (n36) Married Life p.233.

53 See examples given in S Roberts, Botswana I: Tswana Family Law (Restatement of African Law: 5), Sweet & Maxwell, London, 1972, p.24, Kuper op.cit., (n34) The Swazi p.55 and E J Krige & J D Krige, The Realm of a Rain Queen; a study of the pattern of Lovedu society, OUP, London, 1943, p.100.

54 See Laband & Thompson in Duminy & Guest, op.cit., p.222 for Zululand.

55 Requisitioning labour in this manner ended only when direct taxation was imposed and paid public servants took over these responsibilities. Igbafe, op.cit., p.331.

56 See, for example, Wilson, op.cit., (n23) p.186.

57 In medieval Europe, for instance, P Ariès P (trans R Baldick), Centuries of Childhood; a social history of family life, Random House, New York, 1964, p.128 argued that the concept of childhood did not exist at all.

58 See S B Burman & P Reynolds (eds), Growing up in a Divided Society; the contexts of childhood in South Africa, Ravan Press, Johannesburg, 1986, pp.8-11.

59 For what follows, see Van der Vliet in W D Hammond-Tooke (ed), The Bantu-speaking Peoples of Southern Africa, 2ed, Routledge & Kegan Paul, London, 1974, ch 7.

60 Krige, op.cit., 81-7; B A Marwick, The Swazi; an ethnographic account of he natives of the Swaziland Protectorate, CUP, 1940, pp.152-3; H A Junod, The Life of a South African Tribe, vol 2, Attinger, Neuchatel, 1912, p.95.

61 M Gluckman, Essays in the Ritual of Social Relations, Manchester UP, 1962, p.16. Not all societies have initiation rituals. The Shona in Zimbabwe, for instance, do not: J F Holleman, Shona Customary Law, with reference to kinship, marriage, the family and the estate, OUP, Cape Town, 1952, p.73.

62 It is remarkable to what extent some traditions have survived. Boys in South Africa, for instance, still undergo initiation, although in some rural areas the rite has been transformed into a period of wage labour abroad. See B A Pauw, The Second Generation; a study of the family among urbanised Bantu in East London, OUP, Cape Town, 1973, pp.88-9; McAllister in P Mayer (ed), Black Villagers in an Industrial Society, OUP, Cape Town, 1980, p.243.

63 T W Bennett, The Application of Customary Law in Southern Africa: the conflict of personal laws, Juta & Co, Cape Town, 1985, pp.39-40.

64 Under both article 1 of the UN Convention on the Rights of the Child and article 2 of the African Charter on the Rights and Welfare of the Child. While 18 is the age chosen for general purposes, article 38(3) of the UN Convention established 15 years as a minimum age for recruiting children into military service. The UN Commission on Human Rights, however, has drafted an optional protocol to the Convention that sets 18 as the minimum age for compulsory recruitment or participation in hostilities.

65 Article 22 of the African Charter and article 38 of the UN Convention.

66 Article 1 of the African Charter and articles 4 and 19 of the UN Convention.

67 Articles 43 and 44.

68 Article 45.




Go to Home Page         Go to Issue/ Theme Page          Go to Country Page