1 Senegalese National Archives, Dakar (ANS) 9G86 (107). The archive documents reproduced later in the chapter are from this file.

2 Thesis in Social Anthropology and Ethnology, École des Hautes Études en Sciences Sociales, Paris, under the supervision of Pierre Bonte (CNRS, LAS, Collège de France), titled “Construction symbolique et sociale de la hiérarchie : noblesse, statut et état pour les Ahl Barikalla (Mauritanie).”

3 Research for the Pacoba program on the peopling of the Banc d’Arguin national Park and territorial dynamics partly published in B. Acloque, “Mémoire et enjeux territoriaux sur la côte mauritanienne : Le conflit d’Eyznaye et la convention de 1930 sur les puits et la pêche.” in Le littoral mauritanien à l’aube du XXI^e siècle: Peuplement, gouvernance de la nature, dynamiques sociales et culturelles. Edited by S. Boulay and B. Lecoquierre (Paris: Karthala, 2011), 125–149Google Scholar.

4 Mauritanian National Archives, Nouakchott (ARIM) E1–75. This was correspondence between the governor of Mauritania and that of French West Africa, as well as with lower-level authorities. The most important documents were transmitted to the French West African governor. Access to the archives was particularly difficult, and I would like to acknowledge the kindness of Heybetna Ould Sidi Heyba, assistant director of the National Archives in Nouakchott.

5 A dialect of Arabic spoken by the Bizhan, an ethnic group characterized as Moorish in colonial writing. They inhabit Western Sahara and Mauritania and extend into neighboring countries.

6 I use qabila (pl. qabail) rather than the ambiguous “tribe.” A qabila is a family group with a common ancestor from whom the group’s name is usually taken.

7 The largest religious qabila in the Sahel (the vast coastal region straddling Mauritania and Spanish Sahara).

8 To facilitate understanding and in the absence of phonetic spelling, I have adopted the spelling most commonly used by colonial administrators.

9 Known in the West from the tales of the Arabian Nights, genies (jnunn in Hassaniyya) make up a parallel world to the visible world. Their existence is attested by the Koran.

10 The Koran explicitly condemns recourse to sihr. The word is ambiguous and can designate any magical practice. On magical customs and their judgment by Islam, see C. Hames (ed.), Coran et talisman: Textes et pratiques magiques en milieu musulman (Paris, 2007)Google Scholar.

11 See, for example, L. Hurbon, “Sorcellerie et pouvoir en Haïti,” Archives des sciences sociales des religions 48:1 (1979), 43–52Google Scholar.

12 Testimony at the time as well as present-day testimony, all acknowledging constraint, is very contradictory on the subject.

13 Malicious gossip as the primary source on witchcraft is also true of Western Europe: “Unfortunately, we know less about what the sorcerers and witches themselves believed than what was believed about them. Possibly witches and sorcerers had more complicated emotions and systems of beliefs than those who believed in them. And this makes them more difficult to study.” J. Caro Baroja, The Witches and Their World, trans. O. N. V. Glendinning (Chicago, 1964), 243Google Scholar. It therefore seems imprudent to conclude, following E. Ann McDougall, that sell may be “evidence of slave resistance against an oppressive master society.” E. A. McDougall, “Slavery, Sorcery, and Colonial ‘Reality’ in Mauritania, c. 1910–60,” in C. Youé and T. Stapleton (eds.), Agency and Action in Colonial Africa: Essays for John E. Flint (New York, 2001), 75. I would incline toward the contrary view, that it is a manifestation of the fear of the masters of an overthrow of the social order, which would occur if slaves began to control the fate of those who dominated them.

14 The verb is sell, isell (tear off, remove, extract); in the south, more common is mass, imass (suck) for the same practice, and the person doing it is called massa.

15 This does not make it possible to make an absolute distinction between witchcraft and sorcery, in Evans-Pritchard’s terms, one unconscious and inherited, the other voluntary and acquired. E. E. Evans Pritchard, Witchcraft, Oracles, and Magic among the Azande (Oxford, 1937). In the case considered here, people seem to have believed rather in a possession or a gift. A piece of written testimony mentions that some slaves are “destined” for sell (letter of Ahmed Baba Ould Mohamed, presented later). In addition, Cheikh Ould Abd El Aziz is said to have attempted to eradicate the evil with a hajab (statement of Mohamed Ould Batta, Nouakchott, January 13, 2009, presented later).

16 The Haratin are a social group of dependents, reputedly of slave origin but free, traditionally employed in agriculture and small animal raising.

17 There is a persistent rumor that blacksmiths are the descendants of Jews.

18 This idea is developed by E. A. McDougall, “Slavery, Sorcery, and Colonial ‘Reality’ in Mauritania.” However, she wrongly identifies witchcraft in general with the particular case of sell.

19 E. Doutté, Magie et religion dans l’Afrique du nord (Algiers, 1909). He does mention the belief in witches flying on broomsticks, who suck blood, in Kheibar in Arabia, Magie et religion, 51.

20 This belief is shared farther north: “Blacks are generally thought of throughout the Maghreb as particularly adept at witchcraft,” Magie et religion, 34.

21 A. L. Ech Chenguiti, El Wasît, trans. Mourad Teffahi (Saint-Louis du Sénégal, 1953)Google Scholar. Ann McDougall’s article includes a summary of this passage in English. The author speaks only of sihr (witchcraft) in his description, never of sell (vampirism from a distance) (Personal communication from Abdel Wedoud Ould Cheikh, April 2009). It is probable that his audience of eastern Arabs (the Arabic original was published in Cairo) would not have understood the second expression. Trab el-Bizhan means the land of the Bizhan, an expression designating the territory occupied by Hassaniya-speakers, roughly from Goulimime in southern Morocco to Timbuktu in Mali and Saint-Louis du Sénégal.

22 Abbé P. D. Boilat, Esquisses sénégalaises (Paris, 1853), 315ff. Thanks to Alice Bellagamba for this reference.

23 See A. Zempleni, “La ‘maladie’ et ses ‘causes’,” L’Ethnographie, 96–97 (1985): 13–44.

24 Busquet does not distinguish between sell and sihr. Many passages from this document have been translated into English by E. A. McDougall, “Slavery, Sorcery, and Colonial ‘Reality.’”

25 ARIM E1–8 (?) (the classification number is not fully legible). Political report on the subdivision of Atar, July 1936. On the language of administrators about slavery, see B. Acloque, “Embarras de l’administration coloniale: La question de l’esclavage au début du XXe siècle en Mauritanie,” in M. Villasante-de Beauvais (ed.), Groupes serviles au Sahara (Paris, 2000), 97–119.

26 One case of the execution of a Haratin for that reason was reported less than ten years ago.

27 The cases of witchcraft in Adrar and Tagant known to the administration since 1900 and catalogued by Ann McDougall do not seem to clearly involve sell before 1934, with the possible exception of a case in 1922.

28 The circle was the lowest-level administrative region of the colony of Mauritania to which the listed qabail or groups were attached. The administrator in charge of a circle was usually called a commander, even if he was a civilian.

29 At the time, the region of Inchiri was attached to the circle of Adrar. The post of Akjoujt was not made a circle until 1931.

30 I have not been able to locate this document, which is mentioned in another letter.

31 The Ouled Bou Sba are a large qabila of traders, who mostly came to Mauritania in the early nineteenth century, that imposed its presence by arms. For this reason it was in a state of conflict with almost all the qabail of the region.

32 Although located in Sénégal, Saint-Louis remained the capital of Mauritania until the eve of independence.

33 Even though they date the events to early October 1928. Was the January date just a mistake? In a preliminary draft, Carde was already writing “in the second half of January.” I have found no satisfactory explanation for this difference in dates.

34 Aside from the file from the Dakar archives (ANS 9G86 (107)), which contains the correspondence from the governor general, I consulted Mauritanian archives, in which I located some documents (ARIM E1–75) containing the correspondence of the governor of Mauritania, but the condition of the files suggests that documents might have been lost or mislaid. Other information may be in the local archives of Nouadhibou (formerly Port-Étienne) and Atar.

35 According to the testimony of Mohamed Ould Batta later in the chapter, it is possible that the execution happened around August.

36 Military control there was long limited to the peninsula of Cap Blanc. An edited book on this port city on the border with the former Spanish Sahara is now in preparation: Nouadhibou: Fortunes et infortunes de la “capitale économique” mauritanienne (provisional title), B. Acloque (ed.).

37 On this border and the way it was perceived over time, see B. Acloque, “L’idée de frontière en milieu nomade,’ in M. Villasante Cervello (ed.), Colonisations et héritages actuels au Sahara et au Sahel (Paris, 2007), II, 351–382Google Scholar.

38 In several documents, the importance given to the support of the Ahel Abd El Aziz is evident. The administration seemed to have blind faith in their reliability. But it is hard to know how real this support was. The manipulative skills of these clever politicians, as we shall see, cast doubt on their apparently unshakeable loyalty.

39 Some remarks on this issue are in the notes to the documents presented later.

40 According to Mohamed Ould Batta (testimony later in the chapter), he was engaged in the getna (date harvest) in Adrar, which would place the events in or around August 1928.

41 See the testimony of Mohamed Ould Batta later in the chapter.

42 See the testimony of Ahmedou Ould Moni, Joffriyat (Agneitir), January 10, 2009, presented later in the chapter.

43 The administration’s investigation and the oral reprimand seem to have left no traces in memories.

44 See particularly the testimony of Ahmedou Ould Moni later in the chapter.

45 One of his letters sets out the jurisprudence concerning slaves’ witchcraft (presented later), the other warns against “slanderers.”

46 See the testimony of Ahmedou Ould Moni later in the chapter.

47 Who was then Sid Ahmed Ould Ahmed Ould Sid Ahmed Ould Aïda.

48 We know nothing of the statements made at the time by the members of this qabila. A delegation nonetheless did come to support the Ahel Barikallah (probably because of internal divisions), according to the governor of Mauritania.

49 Doubt remains as to the administration’s ignorance, particularly that of the governor of Mauritania, regarding the artificial construction of the Ahel Habiballah’s defense. The embarrassment of the colonial authorities over the question of slavery suggests that they may have been showing accommodating stance. See Acloque, “Embarras de l’administration coloniale.”

50 For clarity, names of places and persons have been uniformly presented in the form most common in the colonial period. This is also the case for the notes. Tifersas is located north of Akjoujt, more than 100 miles northeast of the site of the events.

51 Abbreviations have been replaced by complete words. Nomad Groups were mobile camel-mounted military units that marked the colonial presence in desert zones. Trarza is the region in southwestern Mauritania occupied by the French since 1902.

52 In the document, which is typewritten unless otherwise indicated. This copy made in Port-Étienne (established in 1906 on the Nouadhibou Peninsula) was one of the seven documents initially sent to the central administration of Mauritania in Saint-Louis du Sénégal, transmitted to the governor general of French West Africa in Dakar.

53 Agneitir is the coastal dune region near Cape Timiris, an extension of the dunes of Akchar. Inchiri is the rocky plain of the Akjoujt region in southeastern Akchar that in 1931 gave its name to a circle.

54 The Ouled El Lab were a small warlike qabila, separated from their Ouled Delim cousins whom they led in the late nineteenth century and close allies of the Ahel Barikallah.

55 The Adrar is the mountain range southeast of Sahel, the large northern coastal region of Trab el-Bizhan. It was made into a circle in 1909, following a difficult military conquest.

56 The Tijirit is the long valley located between the lines of dunes of Azefall and Akchar, running 300 miles from northeast to southwest.

57 The persons questioned from this qabila remain unknown.

58 Génin’s conclusions are the same as those reached by various administrators in their letters, despite blatant contradictions, particularly as to the unfolding of events, that came out in the depositions. See later in the chapter.

59 Bou Ghabra on current maps, a well in the extreme southwest of Akchar.

60 It is worth noting that the distinction between slaves according to sex adopts the local terminology that reserves ‘abd (slave) for a male slave and khadem (female servant) for a female slave. The identity of the slaves, their origin, and the possible practicing of witchcraft gave rise to no investigation by the administration. Slaves, even as victims of murder, were of little interest to administrators. In his letter 248 AP/d of March 13, 1929 (ANS 9G86 (107)), the director of political and administrative affairs of the AOF, however, wrote to the governor of Mauritania: “The victims were indeed ‘born servants’; the transcript of one of the interrogations mentions that they had changed masters about thirty years ago; they had thus lost any contact with their natural relatives, whose domicile is unknown and who besides could not receive indemnification in these circumstances without provoking new political difficulties.” If the change in masters thirty years earlier was indeed mentioned by the principal accused, it was not established (see later in the chapter). Moreover, nothing in the archive documents allows one to say that they were born slaves or that they had no existing family connections. As for the location of any possible relatives, nothing was done to establish it. Political questions undeniably trumped the exercise of justice.

61 Phrase underlined in the margin.

62 Phrase underlined in the margin. The pressure exerted by the Ouled Bou Sba, notoriously on bad terms with the Ahel Barikallah, was reiterated in all the colonial writings. And yet, a delegation of Ouled Bou Sba came to the administration to support the Ahel Barikallah (letter 8 BP from the lieutenant governor of Mauritania to the governor general of the AOF, February 18, 1929, ANS 9G86 (107)). Recent testimony attributes no role to them and does not even recall their presence (see later in the chapter).

63 From the point of view of Islamic law, the master alone has the authority to inflict punishments of mutilation or death on his slaves. The emphasis witnesses apparently place on this point is worth noting.

64 This was a frequent form of statement from the Ahel Barikallah as from most religious qabila, for whom reserve is the basis of diplomacy.

65 Génin erroneously characterizes the Ahel Habiballah as a tribe. They were in fact one of the six groups of the Ahel Barikallah recognized by the French administration. It was the only one to be listed by the circle of the Baie du Lévrier (Port-Étienne).

66 Among the Ahel Habiballah, the succession to the chieftainship of Mohamed Ould Abd El Aziz was still recent (1926). His brother Cheikh was allegedly excluded because the origin of his mother (Ouled Tidrarine) was not sufficiently noble in the eyes of the Ahel Barikallah and because of his judged-too-warlike conduct. His nephew Ahmed Baba Ould Mohamed Ould Abd El Aziz was quickly designated by the djemaa (assembly of notables), although his authority was never fully established, particularly over Cheikh.

67 Administrative documents repeatedly use the term “witch” (sorcier), whereas vampirism, although identified with witchcraft, was clearly differentiated for the Bizhan.

68 This sentence is underlined in the margin. The words “marabouts” and “warriors” adopt an indigenous distinction between the Zwaya noble groups, who do not bear arms and justify their position through religious authority, and the ‘Arab, who acquired their position by arms. The two groups, who have different codes of honor, consider intermarriages to be misalliances. Captain Bousquet in Port-Étienne and the lieutenant governor of Mauritania limited responsibility for the murders to the Ouled Bou Sba notables (warriors, at least as far as the colonial administration was concerned) and the Ahel Barikallah (marabouts).

69 Sidi M’Bareck Ould Ahmed Baba, chief of the Démouissat, principal warrior group of the Ouled Bou Sba, was well enough known to the administration for it to be unnecessary to use his full name.

70 Handwritten.

71 Many typos and spelling and punctuation mistakes, made because of the direct transcription of the interrogation, have been corrected for greater clarity. Cheikh Ould Abd El Aziz was the principal person accused of the three murders. He was initially summoned to Atar like the others involved. But because it was established that the events took place in the territory of the circle of the Baie du Lévrier (Telegram no. 924A, Mauritania to Atar Circle, December 1, 1928, ARIM E1/75), he was heard in Port-Étienne by Captain Bousquet. Learning of his questioning, Governor Choteau of Mauritania wrote in a telegram (no. 403 P, February 2, 1929, ARIM E1/75) to his subordinate to “treat him with consideration and to urgently inform [him] of [his] opinion about [the] political repercussions [of] this affair and [the] solution that it would be appropriate [to] come to.” In reply (no. 16 p, February 8, 1929, ARIM E1/75), Bousquet offers reassurances: Cheikh, “as well as those who accompanied him to Port-Étienne, have been treated with all the requisite consideration. All continued to enjoy complete freedom. In my opinion and for the reasons set out in my letter 14/P previously mentioned, this affair should be settled by making the Ahel Barikallah understand, without offending them, that in the future they will have to refrain from such acts.”

72 The next day at the same time, his nephew Mohammed Abdallahi Ould Mohamed Ould Abd El Aziz, who had already been heard by Génin, was questioned. I will point out the most important divergences between the two in the notes. On the other hand, his two sons, Sid Ahmed Ould Cheikh and Mohamed Nami Ould Cheikh, who were “deeply implicated, one [being] ill in his camp and the other in the north sent by the Ahel Barikallah to deal with a matter involving the return of camels,” as Bousquet wrote (no. 14 P, February 3, 1929, ANS 9G86(107)), were not questioned.

73 This is the only point at which there is a suggestion of a judicial proceeding against Cheikh Ould Abd El Aziz. The transmittal letter (no. 14 P, February 3, 1929, ANS 9G86 (107)) shows a reluctance to carry it out: “there was a preliminary investigation. Before continuing it, I thought I should explain the reasons for my reluctance to go any further.”

74 This well is more precisely located in southwestern Tijirit.

75 The father’s name of the slaves appears only three times in the whole documents: twice in this interrogation, once in the transmittal letter (no. 14 P, February 3, 1929, ANS 9G86(107)). The spelling is each time different: “Ould AB EL MOYLA,” “OULD AB EL SOYLA,” and “ould AB EL HOVLA.” The appellation “Abd el ...” is usual for Muslims. It means “slave of...” and it is followed by one of the ninety-nine names of God. It is not reserved for slaves. Cheikh father’s name “Abd el Aziz,” “slave of the Beloved,” or of his nephew Mohamed “Abdallahi,” “slave of God,” are two examples. It might be “Abd el Moula,” “slave of the Lord,” which is why I use the closest spelling. In today’s testimonies, neither name nor first names are known.

76 Thirty years earlier, the French were not present in Mauritania. Repression of the slave trade, particularly enforced after the law of December 12, 1905, was thus not applicable if the transaction has taken place before that date. The statement may have taken that fact into account. That possibility is bolstered by the fact that they were in the prime of life in 1928 and, according to Mohamed Ould Batta (see later in the chapter), the fact that they had been tortured for witchcraft before their purchase. It is hard to imagine young children being tortured for that reason. Moreover, some remember that the events took place shortly after their arrival (Ahmedou Ould Moni, presented later).

77 This is a mistake. The Ahel Barikallah had separated from the Ideïkoub in the seventeenth century. They were thus two independent, albeit related, qabail.

78 A djemaa is an assembly of notables of varied composition that meets to make common decisions working toward a consensus.

79 Mohamed Abdallahi Ould Mohamed Ould Abd El Aziz sets these events back “about ten years.”

80 The word is ambiguous. It may suggest the use of the torture technique called the question. It appears later on with the same ambiguity.

81 Cheikh’s father, Abd El Aziz Ould sheikh Mohamed El Mami, died at the very beginning of the twentieth century. There is certainly confusion here between him and Cheikh’s brother, Mohamed Ould Abd El Aziz, chief of the Habiballah group, who died in 1926, which agrees with the statement of Mohammed Abdallahi Ould Mohamed Ould Abd El Aziz.

82 According to Mohamed Abdallahi Ould Mohamed Ould Abd El Aziz, three or four years earlier, the three slaves “were again accused of stealing the blood of two Ouled Bou Sba of the family of the Ahel Sidi Béchir,” then of causing the death of Mohamedun Ould Ahmed (Ahel Meki) and of Minita (Aminata or mint (daughter of)?) El Hajj Omar (Ahel Maouloud) of the Ahel Barikallah and of Toybani Ould Mohamed El Moctar of the Ideïkoub.

83 This account of events is confirmed by Mohamed Abdallahi Ould Mohamed Ould Abd El Aziz, who concluded: “Hamadi dragged himself to Sid Ahmed, who told us that Hamadi had given [him] back his blood, but no one saw the blood. Hamadi died a few moments later from the blows he had gotten from all of us.”

84 According to Mohammed Abdallahi Ould Mohamed Ould Abd El Aziz, it was following another application of the gris-gris that Sid Ahmed accused Souélim.

85 The mistake here is obvious: Mohamed Ould Abd El Aziz was the brother of Cheikh Ould Abd El Aziz. He did have a daughter named Fatimatou. Their death was attributed to these three slaves by Mohamed Abdallahi Ould Mohamed Ould Abd El Aziz.

86 Another mistake: Mohamed Abdallahi Ould Mohamed Ould Abd El Aziz was the brother of Ahmed Baba Ould Mohamed Ould Abd El Aziz. Mohamed Abdallahi said he had been sent to his uncle to “tell him that he absolutely had to kill the three witches.” He also said he arrived three days before Cheikh returned from his trip. According to him, Sid Ahmed complained that his blood had been taken. To the demand to give back the blood, Zénabou “answered that she had already given her part back, that Hamadi had not given his back.” Hamadi had run away and only returned with Cheikh and Souélim three days later. The administration did not seem to have any doubts of the presence of both throughout the events. The delegation was probably sent after all of this in order to present a common version of events.

87 These were Cheikh’s brothers-in-law, uncles of Mohamed Nami and Sid Ahmed.

88 These were the only three people outside the Habiballah whose presence was confirmed by the statements Mohamed Abdallahi Ould Mohamed Ould Abd El Aziz.

89 These were probably the members of the delegation of Habiballah sent by the chief of the group Ahmed Baba mentioned earlier.

90 The two versions diverge seriously. According to Mohamed Abdallahi Ould Mohamed Ould Abd El Aziz, the death of the slaves was not decided on; it was the result of beating them to make them confess. Hamadi first, then at nightfall “Souélim did not want to confess to being a witch: he died from the blows we all struck him with.” When the others went out to pray, Mohamed Nami Ould Cheikh Ould Abd El Aziz fired a shot into the corpse of Souélim. It was not until two days later, after Sid Ahmed suffered a relapse, that the presentation of the gris-gris led him to designate Zénabou. “Like her brothers, she protested that she was not a witch; she was beaten until she was dead.” Again taking advantage of a prayer, this time it was Sid Ahmed who fired “a pistol shot into the corpse of Zénabou.” These differences were not picked up by the colonial administration.

91 This assertion did not convince the administration. Sid Ahmed and Mohamed Nami were considered co-perpetrators of the murders to the same degree as Cheikh.

92 This argument carried some weight. In his transmittal letter, Captain Bousquet wrote: “It is also appropriate to take into account the fact that they all believe they are covered by Muslim law and that they acted as their ancestors would have acted” (no. 14 P, February 3, 1929. ANS 9G86(107)).

93 In reality chief of the Ahel Habiballah group of the Ahel Barikallah.

94 Nephew.

95 This accusation carried conviction in the administration. Captain Bousquet, for example, wrote that if the “original” guilty parties were Cheikh and his two sons, that this was “an act of collective fanaticism”; “the principal instigators were the Ouled Bou Sba, in particular Sidi M’Bareck, whose threats overcame the reluctance of the Ahel Barikallah” (no. 14 P, February 3, 1929, ANS 9G86(107)).

96 Handwritten.

97 Translation of one of the two letters written by Ahmed Baba Ould Mohamed Ould Abd El Aziz, chief of the Ahel Habiballah group of the Ahel Barikallah. Two other letters, one from a “djemaa of the residents of Adrar,” the other from Mohamed Ould El Khalil of the Rgueibat, and all the originals in Arabic, are included in the file. They seem to have been provided by Cheikh Ould Abd El Aziz first to the commander of the Adrar circle, then to the commander of the circle of the Baie du Lévrier. They then went up the chain of the hierarchy and were finally translated in writing at the request of the AOF administration by Touradou Ben sheikh Kamara in February 1929. The entire document is handwritten.

98 Customary introductory formula.

99 Colonel Henri Gaden was governor of Mauritania from 1916 to 1926 (with the title “Commissaire du Gouvernement général de l’Afrique occidentale française” until 1920, and then “lieutenant gouverneur” when Mauritania became a separate colony). Alphonse Choteau held the office in 1928–29.

100 The Arabic text reads “witchcraft (sihr) called sell.”

101 These qabail were small nomadic religious groups that had good relations with the Ahel Barikallah.

102 Assertion not confirmed by any testimony.

103 Idiomatic formulas that the translator chose to put in parentheses.

104 Although the attachment of the Ahel Barikallah to their dependents and their reluctance to free them are still well known in Mauritania today, this “proverb” is unknown to me.

105 The document was probably written in late 1928. The last letter of Ahmed Baba Ould Mohamed Ould Abd El Aziz was certainly later, because it only intended to warn against “slanderers.”

106 A hajab is a talisman of Islamic inspiration or not made by a specialist to obtain a particular magic effect. See introduction.

107 Tiris is a vast plain with famous pastures in northern Azefall, mainly in the formerly Spanish zone.

108 Date harvest that took place in July and August in regions with oases, the occasion for festive gatherings. The events, therefore, took place two or three months before the colonial administration thought they had.

109 The custom among the Zwaya groups, including the Ahel Barikallah, was to be unarmed. Horses, as well as rifles, are warrior objects associated with the ‘Arab nobility.

110 The translation (into French) was made by Farida Mint Habib and Ahmed Mouloud Ould Eida, in the presence of Hélène Artaud. My interlocutor said he got this information from his father and from Hamadi Ould Saguti, who will be mentioned later.

111 This interlocutor is the only one who attributes ownership of the slaves to Mohamed Nami. He does not even mention his father, Cheikh.

112 The hill of slaves.

113 The verb used (sell, isell) is translated as “extract.” See introduction.

114 The verb used (hajer, ihajer) refers to the Prophet’s flight to Medina from the domination of the polytheists in Mecca, the origin of the Muslim calendar (Hegira). It is the word used in Mauritania to designate the flight from colonial domination to unoccupied territory. The Spanish did not travel through the territory granted by treaty until after 1934. See Acloque, “L’idée de frontière”.

115 Hartani (pl. Haratin) designates a freedman or an assimilated foreigner in a clientelist relationship with his master. He belongs to the category of the tlamid. It can also be a euphemism for slave.

116 Located in the Taziast (north of Azefall), 45 miles north of the site of the events, Erzmeilat was 60 miles south of the border with Spanish territory. It was also on the road to Port-Étienne.

117 The post of Port-Étienne was always known to the Bizhan by the name of the peninsula, Nouadhibou, now the name of the city.

118 Cheikh Ould Mouknass of the Ahel Laghzal represented the El-Graa qabila before the colonial administration in Port-Étienne. He was very influential there. The Ahel Laghzal were tlamid (religious disciples) of the Ahel Abd El Aziz, the family of Mohamed Nami Ould Cheikh.