1When the Tuareg upheavals in northern Mali and in northern Niger began in the 1990s, they aimed at building up autonomous regions or even a state that would embrace all members of the « Tuareg nation » (temust or tumast). The ideology of a common « Tuareg nation » (or temust), however, had only slowly developed during several decades in Libyan and Algerian exile where Tuareg of various origins had been perceived by their hosts as members of a homogenous ethnic group. This ascription from the outside had been (partly) taken over by Tuareg and translated into the political project of temust (the « Tuareg nation »), represented as a unit of people sharing common descent, speaking the same language, and, above all, equals in dignity and morals. All quarrels between the Tuareg themselves, i.e. former conflicts between regional and tribal groups, or differences between various classes of the traditional hierarchical society, were to disappear and make room for a national identity in a future Tuareg state. This state should reach ideally even beyond the boundaries inherited from the post-colonial state (Klute, 1995, p. 55-56). Despite its discourse of national unity, however, the rebel movement soon split up into several factions which opposed one another politically at first, and later also militarily.

2I have shown elsewhere (Klute, 1995) that the divisions between the various factions do not only reflect diverging political orientations, but that they also refer to friendly or hostile relationships between tribes or confederations of tribes. These relationships though are not very old ; their present form goes back to the colonial conquest at the beginning of the 20^th century.

3This article deals mainly with the changing nature of relationships between various Tuareg rebel movements during the so-called « Tuareg rebellion ». Though I will concentrate on the relations between two Malian rebel movements during a fratricidal war that took place in northern Mali in 1994, I will refer to the whole area affected by the Tuareg upheavals in the 1990s, i.e. large parts of the southern Sahara and the northern Sahel, reaching from the Mediterranean shores, the Libyan, Algerian and Chadian Sahara, to the northern parts of Niger and Mali. In order to make my arguments more understandable, I will not only deal with events connected directly to the fratricidal war between two Tuareg rebel movements in 1994, but also with the period between 1980, when the first « revolutionary » organisation of exiled Tuareg was founded in Libya, and the year 1996 when the Tuareg upheaval in Mali ended.

4The following observations are based on field studies conducted between 1990 and 1998 in southern Algeria, northern Mali, northern Niger, and France. Besides « classical » methods of data-collection in anthropological fieldwork, such as observations and interviews, my reflections are also, and mainly, based on biographical studies. These studies helped to follow trajectories of people, including the varying relationships they had to one another and in particular to asset the changing nature of these relationships given in legitimatory discourses.

6As I will explain, friendship played an important role particularly among Tuareg migrants during the years of exile that preceded the military upheaval in 1990. Here, friendship fulfilled two functions : firstly, it served as an ideological tool that was supposed to strengthen the unity of an imagined utopian Tuareg nation. Secondly, relationships stemming from friendship had the potential to cut across tribal or kin identity and solidarity. Political orientations, indeed, followed the same transversal logic.

7Identities were negotiated and reshaped again in the course of and after the military upheavals, which took place over the 1990s. My main thesis is that the logic of descent eventually prevailed over any other logic of relation-building. This holds true in particular for the year 1994 when two Tuareg rebel movements in northern Mali fought a fratricidal war.

The Ideology of Friendship

10The history of modern migrations of the Tuareg dates back to the last years of colonial rule in the 1950s. Hundreds of Tuareg from Mali and Niger migrated to southern Algeria where they were employed in the oil-industry, or became mercenaries in the French colonial army. From the middle of the 1950s onwards, the French recruited more and more Tuareg in order to cope with the growing anti-colonial National Resistance Movement (Front de libération nationale, FLN) in Algeria.

11In the following years, the number of Tuareg migrants in Algeria increased. In 1963/64, groups of Tuareg in northern Mali revolted against the Malian government. As they were small in number, and poorly armed, the Malian army had no difficulty in suppressing the upheaval rather easily, despite the fact that the Tuareg had a much better knowledge of the terrain and that they were superior individual fighters. Hundreds of Tuareg, including women and children, fled the suppression campaigns of the Malian Army. A decade later, when a heavy drought decimated the herds of many Tuareg nomads in northern Mali and Niger, a new wave of thousands of refugees arrived in Algeria, and from there, continued also to Libya. A similar drought occurred in the 1980s ; again, thousands of Tuareg from Mali and Niger fled to the neighbouring countries in the north.

12Sentiments of common origin and common destiny developed in the milieu of Tuareg migrants in Algeria and Libya. This was partly due to the fact that their Algerian and Libyan hosts treated these migrant workers and their families quite badly as Tuareg from abroad, without differentiating between regional groups, tribal identities, or social strata. Tuareg migrants started to address one another as « friend » (amidi ; pl. imidiwan), and not as « cross-cousin », etc., as they used to do before. As their Algerian and Libyan hosts thought the (exiled) Tuareg to form a homogenous ethnic group, the exiled Tuareg, too, started to imagine themselves as an ethnic group, united by common traits.

13There is virtually no scientific work, based on fieldwork and direct observation, which covers this period of Tuareg exile in Algeria and Libya preceding the beginning of the so-called « Tuareg rebellion » in 1990. All publications concerning this period are based on data collected retrospectively. As « revolutionary » movements and violent wars are very dynamic processes indeed, statements of social actors can undergo dramatic changes in short periods of time. What was said yesterday can be omitted or forgotten today. This is particularly true if one takes into account that the Tuareg still live in a predominantly oral culture, where all statements are actually shaped by circumstance and, as in this case, by the balance of power prevailing at that moment.

14In order to overcome these methodological difficulties, I have used the migrants’ poetry as a historical source. I have chosen poetry as a historical source for two reasons : firstly, because poetic texts are, so to say, more resistant to changes than prosaic ones ; secondly, because the Tuareg migrant poets recorded their works on audio-cassettes, and in this way, prevented all later modifications. On the whole, I have collected some hundred poems (or songs) ; from this collection, I have selected a corpus of about fifty pieces according to the following criteria : audibility, possibility to date the creation, possibility to identify the author and circumstances of the piece’s creation. This corpus covers the period from 1978 to 1996. The careful interpretation of the migrants’ poetry, combined with other data, has actually contributed to the reconstruction and analysis of the Tuareg upheaval in Niger and Mali.

15With regard to emic appellations, the analysis of the migrants’ poetry reveals several interesting points :
(1) In my collection of several hundreds pieces of Tuareg migrant poetry, no tribal name and no appellation of a social stratum is ever mentioned, whereas geographical terms do appear.
I interpret this omission as a clear indication of the authors’ desire to overcome social or tribal differences within Tuareg society. Each and every Tuareg should be the same in a future, though still utopian, nation-state, regardless of his/her tribal or social belongings.
(2) The second interesting point is that in the corpus of the fifty or so poems I collected, and which I can date exactly, a clear distinction is made between the appellation of the Tuareg who migrated, and those who stayed « at home ». Whereas the Tuareg who stayed at home are addressed, with only few exceptions, in kinship terms, Tuareg migrants are not addressed in kinship terms ; instead, the poets mostly just call them « friends » (imidiwan).

16Again, neutral appellations such as « friends » have to be interpreted as ideological attempts at homogenising rather heterogeneous communities of Tuareg migrants in Algerian or Libyan exile. The fact that Tuareg who stayed at home are addressed, with much affection, in kinship terms is a first hint at the political project of the Tuareg migrants : to give up exile, to return home and to fight for political and cultural autonomy.

17On the other hand, life in exile seemed to have favoured friendships and the development of intimate, personal relationships that cut across habitual social or tribal divisions : appellations such as « friends » clearly had concrete and material foundations.

18The first revolutionary organisation of Tuareg migrants was founded in Al Khums (Libyan transcription) near Tripoli at the end of August 1980 with the semi-official support of Libya ; it was called Front populaire de libération du Sahara arabe central (FPLSAC : Popular Front for the Liberation of the Arabic Central Sahara). This organisation included Tuareg from Mali as well as from Niger. The same year, Ghaddafi opened the first military camp for Tuareg in order to give the Tuareg military training and to recruit them into his famous Islamic legion.

19Apparently, life in these camps very much favoured the building up of personal relationships. Friendship bonds were created between Tuareg from Mali and from Niger, between slaves and nobles, and even between different tribal confederacies which were long since considered sworn enemies at home in Mali or Niger. Personal relationships grew even stronger, when Tuareg mercenaries became comrades in arms. Hundreds of Tuareg migrants fought as Libyan mercenaries in various campaigns in Lebanon or Chad, sent there by the Libyan leader for his ambitious political goals.

20In retrospect, the local narrative in northern Mali underlined the coherence of the revolutionary movement that was based on friendship bonds only at that time, by saying that even close kin could become personal enemies in Libyan or Algerian exile, and, moreover, could start hating each other – the local expression of the phenomenon that migration and exile were experienced as transitional situations, changing even traditional representations of what personal relationships should be based upon (see also Meier, 2004).

21At that time migrants started to call all Tuareg Kal Lamiger – a poetic neoplasm composed of the Tuareg word Kal (« people of » ) and the terms « Mali » and « Niger ». This new composition symbolised the wish to overcome all the various divisions among the Tuareg – even state boundaries – in order to erect a common Tuareg nation-state. As late as 1992, when I did fieldwork in southern Algeria, it was impossible to ask for tribal names or the social status of persons because this was considered as detrimental to the national cause.

22The first division of the united resistance movement of exiled Tuareg took place in 1987. One day after the death of Seyni Kountché, Niger’s dictator and General-President, the Malian Tuareg, and only the Malian Tuareg, founded a new clandestine resistance movement in Libyan exile. At the same time, the Tuareg of Niger completely withdrew from the project that promoted a common nation-state of all Tuareg. The Niger Tuareg hoped for democratic change in their home country after Seyni Kountché’s death. They, therefore, voted against the plan of a military and necessarily violent upheaval and opted for a peaceful political struggle. After that, there was no longer a common military or political project of the Tuareg from Mali and Niger, despite the fact that contact between Malian and Niger rebel movements continued to exist, and despite the fact that some rebels fought on an individual basis, and occasionally, on the other side of the border.

From Friends to Enemies

23At the end of 1989 and the beginning of 1990, around 20,000 Tuareg migrants were sent back from neighbouring Algeria and Libya to Mali and Niger. Among them and their followers were a few hundred who had been mercenaries for a number of years in Libya’s Islamic Legion ; some had even fought in campaigns for the Libyan Leader, as I mentioned earlier. These people had organised clandestine associations of Tuareg migrants in exile. The migrants’ return to their home countries marked the beginning of the so-called « Tuareg rebellion ».

24The military upheaval in Mali had three phases. The first phase was marked by a successful struggle against the Malian army, which ended with the Treaties of Tamanrasset in January 1991.

26In this situation the Tuareg rebels benefited from the changed global political situation at the beginning of the 1990s and the serious crisis of the military regime in Mali, which was due to fall in March 1991. In order to rid itself of one of the conflicts, the Mali government signed a peace treaty under the mediation of Algeria in January 1991. In the « Treaties of Tamanrasset », as the peace agreement was known from then on, the signatories agreed to grant special status to the north of Mali, plus to economic concessions, which was practically equivalent to autonomy for the Tuareg in that part of the country.

27During the first phase of the military upheaval, the poets of the rebels continued to address those Tuareg who had not joined the armed rebellion yet, in kinship terms. Apparently, affective appellations, such as kinship terms, were supposed to convince those Tuareg who had stayed at home to join the armed struggle and to become members of the imagined, though still utopian, Tuareg nation-state. The rebels, on the other hand, were still called « friends », « my friends », etc., but more and more martial appellations were used too, such as « wild beasts », « fighters », or « martyrs ».

28Implementing the Tamanrasset Treaties proved almost hopeless. Furthermore, the successful fight against Mali had unleashed neotribal, social, and political dynamics among the Tuareg which would become the driving forces for the second phase of the rebellion : the splintering of the rebel movement. Under the mediation of France and Mauritania the peace agreements of the Pacte national in April 1992 ended the second phase of the rebellion.

29In spite of the agreement of Tamanrasset, some rebel groups continued their attacks in the spring of 1991. Soon afterwards, the united rebel movement split into first two, and later three, politico-military movements which acted independently both militarily and politically. Apart from the original movement MPLA (which now had changed its name to Mouvement pour la Libération de l’Azawad, MPA), the first division organised itself as Front Populaire pour la Libération de l’Azawad (FPLA), and the second one as Armée Révolutionnaire pour la Libération de l’Azawad (ARLA). In an earlier publication, I have shown that the lines of division between today’s rebel movements correspond to former divisions between Tuareg groups in what became to be later known as Mali (Klute, 1995). As already mentioned, however, these divisions are not very old. They go back to the time of colonial conquest at the beginning of the twentieth century. In face of the French colonial army, Tuareg groups essentially had three options : to resist (which most groups opted for), to collaborate, or to wait and see what the French conquerors were like and how they would act.

30What is of importance here, is the fact that the leaders of the three movements had known one another personally for a long time : they had been in Libyan exile, they had lived in the same military camps, they had received the same military training, and they had fought in the same military campaigns in Syria, Lebanon, and Chad. Equally important as their shared experience in exile, however, is the fact that they were of about the same age. This also holds true for most members of the three armed militias. Friends of the same age group had become political enemies and, as I will show later, military enemies as well.

31There was not only an internal differentiation within the rebel movement, but also a differentiation among the opponents of the rebels. In reaction to Tuareg raids against settlements along the Niger River, a militia of the « black » Songhay was mobilised against the « white » Malians in a Chasse aux Blancs. With the support of certain elements in the army, it carried out pogroms against Malian Tuareg and Moors. These pogroms resulted in many Tuareg and Moors fleeing en masse, first to Mauritania, but also to Burkina Faso, Niger or Algeria ; the number of war refugees reached 100,000.

32However, the political situation in Mali changed radically after the fall of the military regime of Moussa Traoré in March 1991. After many months of talks, mediators from France and Mauritania helped the warring parties conclude the Pacte national in April 1992. In this treaty the new transitory government of Mali was able, politically speaking, to get its way in many respects : the idea of federalism was dropped, together with the idea of an Azawad region, which was again replaced by the term « the North of Mali » (Nord du Mali). The Malian government also agreed to incorporate a great number of rebels into its armed forces and public service.

34Nevertheless, there was still hope of peace. In the second half of 1992, the French helped the Malian government to form the first military units from army and former rebel personnel which they called « mixed patrols ». The « mixed patrols » deployed in the north were to guarantee public safety. By the beginning of 1993, the regular Malian army had absorbed more than six hundred rebels. Official plans provided for a further 3,000 fighters. State administration was slowly gaining ground in the northern regions again. International donor organisations made significant financial means available for the reconstruction of the war-torn region and the promotion of peace. The newly elected democratic President Konaré convinced the northern population of his peaceful intentions. In the summer of 1993, all rebel movements signed the National Pact.

35This did not bring about peace, however. Instead, the war entered its third and bloodiest phase. Its major characteristics were a renewed and ongoing dissolution of administrative structures of the central government in the northern region, and ethnic strife between « white » and « black » Malians, particularly in 1994.

37Particularly during the phase of ethnic strife of the Tuareg rebellion, interethnic as well as intra-ethnic friendship became very problematic indeed. Personal interethnic relationships of various kinds, including interethnic friendship, of course did exist in pre-war times. In the context of the armed conflict, however, it became difficult to openly declare friendship with members of other ethnic groups. One informant put this as follows :

• 7 Iswadan ag Saghid, Kidal, November 1996, my translation.

« [...] And these pogroms have been blind. They did not save anybody. They were directed against an ethnic group, against the Tuareg, against the Moors. Good relationships were of no use. Marriage was of no use. Mixing was of no use. And what happened? Everybody was forced : either you are on this side or the other. Neutrality did not exist any more ; one could not be neutral. You had to choose sides, and some choices were inevitable. There were people who did not have any choice, because, if you are light-skinned, even if you wanted to choose the side of the government, you could not. [...] This has forced everybody to one side »7.

38The few cases of people I recorded, who refused « to choose sides », were not actually sanctioned by killing, as in many other contemporary small wars, but by avoidance, and, more importantly, through the disapproval of women. In fact, some Tuareg women actually left their husbands to follow rebel lovers, if they felt that their spouses adopted an in-between-position. Interethnic friends could not – at least in the Malian case – prevent violence or help in conflict resolutions, as the elder literature supposes (e.g. Mühlmann, 1940). Particularly during the phase of ethnic strife in 1994, factors of exclusion were so strong that it became dangerous to openly declare interethnic friendship (see also Horstmann and Schlee, 2001).

39Almost at the same time, at the beginning of 1994, the second conflict broke out. This was the fratricidal war between the two Tuareg rebel movements of MPA (former MPLA) and ARLA. Both movements had their military base in the region of Kidal in northern Mali, where both movements also recruited the majority of their fighters. The first movement had a very small social basis, though. At the beginning of the war, its members consisted almost exclusively of women and men from one of the leading and most respected tribal groups in the region, the Ifoghas tribes. The second movement, by contrast, was a heterogeneous coalition of all those who wanted to end the dominance of the Ifoghas as the region’s leading tribal group ; it recruited members among all non-Ifoghas groups, of free and vassal strata alike.

40At the very beginning of this fratricidal war, ARLA committed two strategic errors, which in retrospect decided the outcome of the war, though fights still continued for another ten months after. The first error was an ambush, at the end of February 1994, in which the second in command of the MPA movement was killed while in pursuit of bandits. Colonel Bilal Saloum was one of those officers who had been highly decorated by the Libyans for their service in the Islamic Legion in Libya and Chad. He was, in accordance with the agreement of the National Pact, a colonel in the Malian army and was a member of the Supervisory Commission of the Agreements of the National Pact. In the beginning of 1994, with the so-called « mixed brigades », he hunted down bandits, who included fighters of the ARLA movement. The latter, however, believed the sole purpose of these supposed « bandit hunts » was to crush their movement, so that the colonel’s brigade, hence the MPA, would be the only militia in the north of Mali. Nevertheless, the fact that the colonel was of slave origin was of crucial importance. What appeared to be a heavy loss was in fact a great opportunity, for it allowed the MPA to make up its numerical disadvantage and to enrol Tuareg of slave origins into its ranks.

41To convince former slaves, two main arguments were used : the first argument was protection and the second one friendship.

42ARLA was asked to hand over the murderers immediately and to compensate the family of the victim. Besides offering to protect the « weak » of society, the argument of friendship was also put forward. It was said that the leader of the MPA, (of Ifoghas and « noble » origin) and his second in command, Bilal Saloum (of slave origin) had been intimate and close friends. This friendship was of greater moral value than the republican, egalitarian and anti-noble discourse ARLA upheld. Pursuing the colonel’s murderer, the Ifoghas of the MPA argued, proved two things : the Ifoghas were ready to protect the « weak » of society against violence ; and, secondly, friendship was of greater value than social origin. Both arguments convinced the region’s former slaves to join the Ifoghas’ movement, the MPA, in great numbers.

44My biographical data of the local elite in the Kidal region shows that the stratum divide during the fratricidal war in 1994 was very clear : not a single person from the stratum of former slaves had joined ARLA, by now exclusively composed of so called vassal tribes (Imrad) ; instead, most (80 %) were members of MPA, the Ifoghas’ militia. The same holds true regarding the stratum of free and noble tribes. Of course, no one from the leading Ifoghas tribes had joined ARLA, now considered the vassal movement. Nearly all members of other tribes had left ARLA after the kidnapping of the Ifoghas tribes’ head, as I have mentioned above. Most had joined the MPA movement ; some had left the armed fight altogether. From the middle of 1994 onwards, only people of vassal origin were members of ARLA.

The Prevailing of the Descent Logic

46At the end of 1994, the situation at the end of both wars – against Mali and between both rebel movements – was as follows : the only militia in the Kidal region is the MPA, the militia of the Ifoghas. The Ifoghas have won the war ; they still have weapons, even if not officially. They have forbidden all fighters of the ARLA movement to enter the region for five years. They threaten to immediately shoot any armed person they come across. Moreover, stolen goods must be returned.

47A military victory, however, does not mean that the war wounds are healed and that sentiments of hate and vengeance are forgotten. The return of the Ifoghas’ uncontested power is not the result of mere opportunism on the part of the defeated vassals. The Ifoghas won their position by clever moves of reconciliation which aimed at the reintegration of the vassals into the local society. Some of these measures are of particular interest for our discussion : the first measure used was a marriage, and the second one was the rewriting of local history.

48News of the marriage caused quite a stir in the region. It was no less than the marriage of the military leader of the MPA movement, Iyad ag Rhali. The bride was the former wife of the man who used to be the groom’s best friend and comrade in arms, but, above all, she was from a vassal tribe, and all three, the bride and the two men, shared a common, long-standing past. They were together in Libyan exile where the two men were recruited by the Islamic Legion. If one was Ifoghas and the other one vassal, it did not prevent them from fighting together with the Legion in Lebanon and preparing the upheaval in Mali. At the beginning of 1990 the two men secretly returned to Niger ; they fought together in the very first Tuareg attack in Mali. Both men were friends and excellent and courageous fighters. So the eventual break between the friends was particularly dramatic at the beginning of the war between « brothers » : close friends had become arch enemies.

49Everyone in the region was sure about the political nature of this union. There were those who simply declared the bride to be the spoils of war – the vassals could not even keep their women. Others looked on the marriage as a sign of reconciliation. The union with the former wife of the vassals’ best fighter meant that the vassals belonged once again to the local community. On the other hand, the marriage also clearly demonstrated that friendship no longer served as a solid foundation upon which to build political relations or to achieve political objectives. Instead, friendship had been replaced by kinship bonds.

51The strategy was as follows: the Ifoghas delegated Mohammed Lamine ag Mohammed Fall to re-establish the affiliation of the vassals, through their genealogy, with the community of the region. Mohammed could be described as a propaganda minister and was just the man for the job. He is a fluent, multilingual speaker and has learned the art to persuade his audience during his exile in Libya. There, his job was to persuade the Tuareg in Libyan exile to join the Islamic Legion. In the tradition of an Arab scholar, he reconstructed the genealogy of each group and each family in the region, and recorded it in school notebooks. He believes, with a few exceptions, to be able to trace all families back to five forefathers who were related to one another. For him, this is a new historical outlook explaining the relations between all groups in the region. The most important aspect is : the existence of vassals is a myth. There are no vassals in the region. All Tuareg are interrelated according to this reconstruction of history. In a society in which kinship is one of the foremost principles, this conclusion is of crucial importance : the label « vassal » is not a primordial, but a secondary, less significant social distinction ; it has established itself in those groups who, over the years, have become weak both politically and militarily and who had been forced to submit to the supremacy of strong families.

52With his school notebooks and genealogical evidence in hand, Mohammed made a tour of the region in order to convince the vassals of their true origins. A summary of the message is as follows: the vassals may exercise no violence against the Ifoghas. The Ifoghas are related to the vassals, and people are not violent towards members of their own family.

Conclusion

54From the very beginning, friendship was an important issue for the Tuareg resistance movement. Alongside the terms of national, regional, tribal, or strata identities, friendship was also much discussed, negotiated and renegotiated during the time of exile of Malian and Niger Tuareg migrants. In order to overcome regional, tribal, or social divisions of « traditional » Tuareg society, members of clandestine Tuareg resistance movements avoided habitual appellations which were supposed to evoke tribal, social, or regional identities ; instead, they addressed one another as « friend », because friendship bonds were thought to strengthen the unity of an imagined, but still utopian Tuareg nation.

55After thousands of migrants were sent home at the end of 1989 and the beginning of 1990, some groups of Tuareg in northern Mali started a military upheaval which became to be known as the « Tuareg rebellion ». Though the rebels were surprisingly successful, the united movement soon split up into various factions, opposing one another politically, and later also militarily. In particular, during a phase of ethnic strife in 1994, when the Songhay militia Ganda Koy called for ethnic cleansing of all « white » Malians, northern Mali was forced into an ethnic divide. Though interethnic friendship had existed in pre-war-times, ethnic identity now became the dominant logic of political relation-building.

56The same year, a fratricidal war between the two Tuareg resistance movements of MPA and ARLA broke out in northern Mali. I have shown that these internal conflicts and intra-ethnic divisions replaced friendship bonds that had cut across « traditional » ties, by tribal and social identities.

57The winning side of the fratricidal war in northern Mali, the dominant tribes of Ifoghas Tuareg, undertook measures of reconciliation in order to reintegrate the defeated into the local community. What is of particular interest here is the fact that attempts at overcoming former (tribal or social) divisions were not based on friendship, as during Tuareg exile, but on kinship and consanguinity.

58Friendship is not only a relational term, friendship is a political term as well. In much the same way as kinship, consanguinity, or descent, it may be used to legitimise political relation-building, alliances or hostile relationships.

59In the context of violent conflicts, I would argue that friendship bonds which cut across identities of the respective warring parties are rather difficult to maintain. In the case of the Tuareg rebellion at least, social actors built relationships on what seemed more stable and more trustworthy than mere friendship.

60But still friendship remains important among Malian Tuareg, though in a somehow liminal way. In the very same year that the Tuareg rebellion ended, one of the rebel-poets composed a song which became a local hit. The song undermines, so to say, the kinship ideology of the ruling groups by praising again the values of an (idealised) friendship :

Better you stay in this world
But stay at its shores
For there is no centre
Which destruction will not come
The world has just three other days
So love your friend because of his love
For there is no centre
Which destruction will not come.