¹ According to ‘Abd al-Rahmān ibn Muḥammad ibn Muhammad ibn Khaldūn, Kitāb al-‘ibar wa dīwān al-mubtada wa’ l-khabar, partial edition of section concerning North Africa, Algiers, 1847, n, 194, Znāga (or Zanāga or Aznāga) is a more correct form of the name than Sanhāja, often used in Arabic literature, which, he said, was an Arabic deformation of the name, which could not be exactly transcribed in Arabic. Znāga, Zanāga, and Aznāga are also closer to what the writers of the accounts of the Portuguese explorations of the western Sahara in the ninth/fifteenth century, who were probably not familiar with Arabic literature and whose information was obtained from local informants, called them. Alvise de Ca’ da Mosto, Relation des voyages ὰ la Côle. occidentale d’Afrique, translated from the Italian into French by Charles Soheffer, Paris, 1895; Valentim Fernandes, Description de la côte d’Afrique de Ceuta au Sénégal, Portuguese text published and translated into French by Pierre de Cenival and Théodore Monod, Paris, 1938; Diogo Gomes, De la première découverte de la Guinée, Latin text and translated into French by Theodore Monod and G. Duval, Bissau, 1959; Duarte Pacheco Pereira, Esmeraldo de situ orbis, Portuguese text published and translated into French by Raymond Mauny, Bissau, 1956; and Gomes Eannes de Zurara, Chronique de Ghiinée, translated from Portuguese by Lèon Burdon, Dakar, 1960, called them Azanaghi, Azenegues, Cenegii, and Azaneguys. The few thousand remaining Znāga speakers, who are all Arabic-speaking as well, today call themselves Iznāgen (sing. Iẓnug). Znāga belongs to the group of languages which has been designated as Berber, but since the origin of the Znāga speakers is not known, nor is it known if they were of a single origin or physical type, the term will be used here in a linguistic and cultural sense only. Znāga has come to mean ‘ tributaries ’ in some parts of the western Sahara, since most of the tributaries were of Znāga-speaking origin, so in those areas it can be pejorative, while in other areas it refers to a specific) warrior tribe, the Idaw ‘Ish, whose chiefs claim descent from the Almoravid chiefs.

In early outside Arabic literature the western Sahara was called simply the Sahara. In later outside Arabic literature it was called Shinqīṭ, after Shingī ṭi, one of the famous towns of the area. The western Saharans themselves today call their land Trab al-Bīẓān, the ‘ Land of the Whites ’ (see n. 3). It is bounded by the Atlantic on the west, the Senegal river and the Niger bend on the south, an imaginary line from the Niger bend to Tuwāt on the east, and another imaginary line in the north from Tuwāt to the Sāgiya al-Hamrā’, on the Atlantic. The Bīẓāsn are not the sole inhabitants, nor are they entirely within the above boundaries. Present political boundaries are of recent origin and do not reflect the ethnic situation. Before the colonial period there were no important urban centres in the area nor does there seem to have ever been a unified state.

² The only major Arabic-speaking group to move into the western Sahara was the Ḥassān, who came from the north in the eighth/fourteenth century, and western Saharan dialectal Arabic, which derives from their language, is called Ḥassāniyya, lughat al-Ḥassāniyya, the ‘ Ḥassāniyya language ’, kalām Hassān, the ‘ words of the Ḥassān ’, and kalām al-BīḤ ān, the ‘ words of the Whites’, on whom see n. 3. With only a few exceptions, transcription in this article generally follows the standard Arabic spelling of Arabic and Znāga personal names, without taking into account dialectal variants due to lack of certainty about how some of the names were pronounced at the time the texts being discussed were written and due to local variations. Place and group names are spelled approximately as they are pronounced today in Ḥassāniyya, and names of persons of the present century are spelled as they are pronounced or as they appear in their writings: tj is used to represent a Znāga sound, sometimes occurring in Ḥassāniyya also, written as jīm, sometimes with three dots under it, and pronounced somewhere between j and d; e is used to indicate an intermediate vowel sound between a and i which occurs in some Znāga and Ḥassāniyya words.

³ Bīẓ ān, from Biḍān, the regular literal Arabic word for ‘ Whites ’, and a Ḥassāniyya singular, Bīẓ ānī, derived from it, are the terms the Arabic-speaking western Saharans now use to describe themselves. The Znāga equivalent is Gudatjen, possibly related to the Gudāla or Judāla, who according to early outside Arabic accounts were a Znāga group who lived along the Atlantic coast of the western Sahara. It is not known when Bīẓān and Bīẓāni first came into use in the western Sahara, although it was probably not before the beginning of the spread of Arabic as a spoken language there, in the eighth/fourteenth century or later. It was probably originally used by peoples of North African descent to describe themselves in distinction to the black peoples to the south of the Sahara, but it has come, paradoxically, to include blacks who have become part of the society of the Bīẓān, such as slaves of the Bāẓān and their freed slaves and descendants; and persons of mixed origin belonging to Bīẓān society. The Bīẓān are now characterized by the Ḥassāniyya dialect of Arabic, Znāga, or Azayr (a dialect of Soninke which has now nearly died out), belonging to one of the tribes and classes of Bīẓāani society, and the observance of certain distinctive customs and manners which are different from those of their neighbours and other Arab Muslim peoples. Tribe, to the Bī ẓān, usually means a group which feels a sense of cohesion because of a belief in descent from a common ancestor, whose name the tribe usually bears. The class structure of Bīẓānī society is complex and there are many variations in the organization and nuances in terminology. The relationship of segmentation to stratification is also complicated, and a satisfactory ‘ model’ of Bīẓāni society has yet to be constructed. It is said that the classes are endogamous, hierarchical, and occupationally specialized. Generally the classes are: religious, warrior, tributary, freed slaves and their descendants, slaves, musicians, smiths, fishermen, and hunters.

⁴ Unpublished Arabic text. A partial French translation, which is not completely satisfactory, was published by Hamet, Ismael in an article entitled ‘Les Kounta’, Revue du Monde M usulnum, xv, 9, 1911, 302–18. The summary in this article is based on a copy of the Arabic text given me by Dr. ‘A bd al-‘Azīz ‘Abd Allāh Batran, fols 61–72.Google Scholar

⁵ Al-Qayrawān, or Kairouan, is in north-eastern Tunisia.

⁶ ‘Uqba ibn Nān’ was of the Fihriyīn branch of the Quraysh and was one of the early Arab governors of Ifrīqiya (first/seventh century).

⁷ Walāta is at the southern edge of the Sahara, to the west of the Niger bend.

⁸ Zāb is a region in north-eastern Algeria around Biskra.

⁹ Ῑday qūb is also spelt Ῑdaygūb, Id Ya ‘ qūb, and Ῑda Ya ‘ūb.

¹⁰ Mustaghānim, or Mostaganem, is a town in north-western Algeria, on the Mediterranean coast near Oran.

¹¹ Amzāb, or M’zab, is an oasis in the north-central Algerian Sahara.

¹² Tuwāt, or Touat, is a series of oases in the central Algerian Sahara.

¹³ al-Zarkashī, Muhammad ibn Ibrahīm, Chromiques des Almohades el Hafçides, translated by Fagnan, E., Constantine, 1895, 183–210.Google Scholar

¹⁴ What the Abdūkal, Lamtūna, and Murābitin might have been in the ninth/fifteenth century is not certain. The Abdūkal were probably a tribe or a confederation of tribes which were not necessarily closely related. According to Mukhtār wuld Ḥāmidun, the leading Znāga-speaking Bīzanī scholar (to whom I am indebted for most of the information in this article on Znāga names and words), the name Abdūkal is derived from the Znāga verbal root d k l ‘to assemble ’, and Abdūkal means ‘ those who have joined together ’. They may have been part of the Lamtūna, but it would seem from the Ghallāwiyyatht they did not include all who now claim Lamtuna origin, such as the Tajakūnet, who were clearly a different group. The Lamtūna were the leading element in the fifth/eleventh-century Almoravid movement, but it is not known what the original nature of the group was. The name may have been applied to all who joined them or were conquered by them and became their tributaries, so the name as it is used in the Ghallāwiyya could be considered to mean the Znāga of the western Sahara in general. Small tribes called Abdūkal and Lamtīna exist today, but it is probable that if they do descend from the earlier groups bearing these names, which there is no particular reason to doubt, they only include the descendants of a small part of the earlier members, the rest having adopted other names. In the western Sahara, Murā biṭīn and Murābiṭun now mean the Almoravids, which is derived from a Spanish deformation of the Arabic name. There is no evidence that their successors continued to rule the western Sahara up to the ninth/fifteenth century, as the Ghallāwiyya says, nor that the name was still applied to their descendants at that time, but in the Ghallāwiyya, the name may be considered more or less equivalent to Lamtūna, that is, the Znāga, as distinct from the Hassān. Although Murābiṭ, the singular form, is now used in the western Sahara to designate a member of the religious class, the plural is used only to mean the Almoravids. That it did not designate only the religious at the time the Ghallāwiyya was written is demonstrated by the assertion that when they were defeated a part of them became tributaries, while their zwāya remained as before.

¹⁵ Sabta, or Ceuta, is a city in northern Morocco.

¹⁶ What their status may have been is not certain.

¹⁷ A lengthy criticism of the current version of Kunta origins is contained in ‘Abd al-‘Aziz ‘Abd Allāh Batrans’s Sī dī al-Mukhtār al-Kuntī and the recrudescence of Islam in the western Sahara and the middle Niger, Ph.D. thesis, University of Birmingham, 1971. I have only summarized parts of Dr. Batran’s analysis of the current view as background for discussion of earlier material.Google Scholar

¹⁸ E. Lévi-Provenèal, ‘Arabica occidentalia. I: 1. ‘Un nouveau rècit de la conquête de I’Afrique du Nord par les Arabes’, Arabica, i, 1, 1954, 17–43. The authenticity of this account has not been confirmed.Google Scholar

¹⁹ Encyclopaedia of Islam, iii, 2, 974, and ii, 2, 1157.

²⁰ Batran, op. cit., 47–8, says t h a t he failed to discover their names elsewhere.

²¹ Information furnished by Mukhtār wuld Ḥāmidun (M.W.H.).

²² Abū ’I-‘Abbās al-Sabtī, one of the seven patron saints of Marrākush, was born in 524/1129–30, and died in 601/1204–5, according to t he Arabic sources cited by Castries, Henri de, ’Les sept patrons de Marrakech’, Hespéris, iv, 3, 1924, 268, 271. The name of the city of Sabta is said to be derived from the Latin name Septem Fratres ‘ the Seven Brothers ’, the seven patrons of Marrākush are called in Arabic Sab ‘atu Rijāl.Google Scholar

²³ Leriche, Albert, ‘Une tribu maure: les ‘Arūsyin’, Conferência International dos Africanistas Ocidentais, Vol. v, 2^a Parte, Bissau, 1947, 73–4.Google Scholar

²⁴ Unpublished Arabic text. Information from the work in this article is from the copy in Butīlimit.

²⁵ They are part of a collection of loose pages belonging to Muhammad ‘Abd Allāh ibn Ahḥmad Abū ‘1-A’rāf al-Mfūū ‘Alī al-Taknī al-Wād Nūnī al-Sūsī al-Tinbuktī, of Tinbuktu. They were photographed by Dr. H. T. Norris, of the School of Oriental and African Studies, who has kindly made his film available. A grant from the University of London Central Research Fund made possible a trip to the western Sahara and discussions of the text with Bīẓānī scholars, especially Mukhtar wuld Hamidun and Mohammed el Chennafi.

²⁶ Al-Ghallāwiyya, fol. 69.

²⁷ Kitāb al-ṭarā’if.

²⁸ ibid.

²⁹ ibid.

³⁰ ibid.

³¹ Tadhkirat al-nisyān, translated by Houdas, O. (Publications de I’ȁcole des Langues Orientales Vivantes, iv^e Sèr., Vols. xix - xx), Paris, 1899–1901, Arabic text, 142, French translation, 231.Google Scholar

³² Chronicle of Aḥmad ibn Twayr al-Janna (thirteenth/nineteenth century). From the notes of M.W.Ḥ.

³³ Tadhkirat al-nisyān mentions a person named Sayyid Saddīq al-Kuntawī, who Bīẓanī scholars feel was probably al-Ṣaddīq, son of Sīdī Muḥammad.

³⁴ Kitāb al-ṭarā’ if.

³⁵ ibid.

³⁶ The top edges of both folios are badly frayed and a re illegible.

³⁷ Sīd Aḥmad is a widely-used name among the Bīẓān, although it is usually written by them in its standard Arabic form, Sīdi Ahmad. Sīd/Sīdī is often used by the Bīzān as an integral part of the proper name. Al-Bakkāy means ‘ the Weeper ’. It is said that following an accidental omission in his devotions, tears flowed from Sīdī Aḥmad’s eyes for the rest of his life, and that his eyes were still moist when his body was exhumed years after his death to be moved to another burial place.

³⁸ Sīdī Muhammad al-Kuntī’s nisba was probably derived from the name of his mother’s father’s father, Kunta. Cf. n. 114, below.

³⁹ Allu could be a Znāga deformation of ‘Ali. ‘Ayn in Arabic words and names is sometimes dropped in Znāga, and lām, tj, and yā’ are often interchangeable. Znāga variants of ‘Ali in use at present are Āju, Atji, and Ayū. Alla is a currently used Tuareg man's name in the area of Tinbuktu, but the meaning is not known (M.W.Ḥ.). Cf. n. 75, n. 82, n. 102, n. 111, and n. 123, below.

⁴⁰ This name is preceded by Yabni on t he page numbered 8 only. There were several well-known western Saharan figures who had t he kunya of Yabnī, or more correctly Yabnī’ he builds but they all had names other than al-Ḥasan, so none can be identified with the one in this genealogy. Cf. n. 44 and n. 124.

⁴¹ Ayshif is a Znāga variant of Yūsuf. According to Tajak ānet tradition, Ayshif al-Jakānī was the founder of the Ῑdayshif branch of the tribe. The lineage given in this text agrees with Tajakānet tradition. (M.W.Ḥ.). Cf. n. 131 and n. 140.

⁴² Malyam is a Znāga form of Maryam (M.W.Ḥ.).

⁴³ Yaruzga is a Znāga name derived from the Arabic root r z q (M.W.Ḥ.).

⁴⁴ See n. 40 and n. 124.

⁴⁵ This name could be rendered by ‘Akd …’ or ‘Agd …’, a Znāga term of filiation meaning ‘the son of … ’ or ‘the son …’, followed by a personal name which may have been omitted where the blank occurs in the text; or ‘Ak Dagh …’ or ‘Ag Dagh …‘, ’a son of the sons of …’, that is, ‘a person of the tribe …’, ‘Dagh’ and the name of the tribe having been omitted. Ak and Ag are both used in Znāga, and the compiler or copyist of the text used kāf to represent both k and g. It would seem that dāl could have been mistaken for rā’ e.g. as in ‘ k r Agh Zaynbu ’, (see n. 70), which would seem best rendered by Ag Dagh Zaynbu ‘ a son of the sons of Zaynbu ’, that is, a member of the tribe called Idagh Zaynbu in Hassāniyya. There is also a proper name, Akur, presumably of Znāga origin, which occurs in the lineage of a seventeenth-century chief of the Ijemmān, a tribe of the Ḥawẓ. He was al-Tālib Saddīq, son of al-TṬlib al-Ḥasan, son of Atjfagha Maḥ ḥam, son of Ṣālih, son of Yarag, son of Ahmad, son of Akur, son of Fanwa, son of ‘Abd Allāh Ajum. Al-Ṭālib Ṣaddīq died in 1073/1663, according to Muhammad ‘Abd Allāh ibn Abī Bakr al-Ṣaddīq ibn ‘Abd Allāh ibn Muhammad ibn al-Ṭālib ‘Alī al-Bannānī al-Bartilī al-Walātī (d. 1219/1805), Fath al-shakūr fiī ma ‘rifat ‘ulam ā’ al-Takr ū r, unpublished Arabic text belonging to Mukhtār wuld Hāmidun. Both Akur of the Ijemmān lineage and A k r of this text had sons named Muhammad. Yarag, the grandson of Akur of the Ijemmān lineage, and Yabni, the grandson of A k r in the Kunta text could have been the same person. Atjfagha Maḥ ḥam, the great-great-grandson of Akur in the Ijemmān lineage, might also be identified with Muhammad, the great-grandson of A k r in the Kunta text, as Maḥ ḥam is a form of Muḥammad. From Sīdī Ahmad al-Bakkāy back to A k r is eight generations, or about 200 years, on the arbitrary basis of 25 years to a generation. If Sīdī Aḥmad al-Bakkāy died 920/1614–15, as the Ghallāwiyya says, A k r would have died in about 715/1315. On the same basis, the seven generations between al-Ṭālib Ṣaddīq and Akur in the Ijemmān lineage would span 175 years; if al- ṭ;ālib Ṣaddīq died in 1073/1663, as Fath al-shakūr says, Akur would have died in about 895/1490, a discrepancy of 175 years. According to Ṣālih ibn ‘Abd al-Wahhāb al-NāṢ ṣirī (d. c. 1270/1854), al-Ḥaswa ’ l-baysāniyya fī ’l-ansāb al-Hassāniyya, unpublished Arabic text, original belonging to grandson of author who lives in the Ḥawẓ, the Ijemmān say that Ajum was the paternal uncle of Abyayrī, who, they say, lived in the tenth/sixteenth century, while Ajum, his paternal uncle, would have lived a century earlier according to calculations based on the Ijemmān lineage, or even earlier according to calculations based on the lineage of Sīdī Aḥmad al-Bakkāy, if the Akur of the Ijemmān lineage and A k r in the lineage of Sīdī Aḥmad al-Bakkāy are the same. This would strengthen the theory, mentioned in al-Ḥaswa, that the Ijemmān are not cousins of the Awlād Abyayrī, but descended from the Izammāten Tuareg. This in turn could be strengthened by the proximity of A k r in the genealogy of Sīdī Ahmad al-Bakkāy to the Tamagnīt or Tamgūna (see n. 53), who claim Massūfa origin, as the Tuareg are also held by some to have been Massūfa.

Another possibility is that A k r was the beginning of the name Akrayl (see n. 132).

⁴⁶ Amghār is a Znāga word meaning ‘chief’, which is now used as a proper name (M.W.H.).

⁴⁷ Page numbered 8 gives only one Muhammad in this lineage.

⁴⁸ Tagnīwā, one possible reading, could be a feminine form of a Znāga word for ‘ black ’ (cf. janāwa and ganāwa) equivalent to the Ḥassāniyya kawrī (pl. kuwār), referring to the black peoples south of the Sahara who do not speak Arabic or Znāga. These names, and similar ones such as al-Bambārī, are often used by the Bīẓān as a good omen, as such peoples are reputed longer-lived than the Bīẓān, or so that any evil or bad luck would fall on those whom the name correctly describes and not on others so named. Another possible reading of the name could be Taknīwā, derived from takna, the Znāga word for ‘ co-wife ’ in a polygamous marriage, which is the name of a tribal confederation between the Wad Nun and the Sāgiya al-Ḥamrā’. This interpretation would be supported by the probable proximity of the persons in the lineage of Sīdī Aḥmad al-Bakkāy to the country of the Takna, although it is not known when the Takna were formed; the first known mention of them in a written document is in a tenth/sixteenth-century work (Monteil, Vincent, Notes sur lea Tekna, Paris, 1948, 16).Google Scholar

⁴⁹ Reading uncertain. M.W.Ḥ. suggests Ῑhīr, a masculine personal name used today in the north-western Sahara and by the Tuareg, the origin and meaning of which are unknown.

⁵⁰ The vocalization of the first name was suggested by M.W.Ḥ. The second is vocalized in the text. The origins and meanings of both are unknown. (Cf. n. 58.)

⁵¹ Origin, meaning, and vocalization unknown, although presumably Znāga.

⁵² Alim is a Znāga deformation of the Arabic ‘Ālim (M.W.H.). (Cf. n. 113, below.)

⁵³ This is probably t h e Znāga name which is now pronounced Tamagnīdh by Znāga speakers (written Tamagni’dh by Muhammad Wālid ibn al-Muṣṭafā ibn Khālunā al-Daymānī (d. 1212/1797), Kitāb al-ansāb, published and translated by Hamet, Ismael in Chroniques de la Mauritanie Sénégalaise, Paris, 1911, Arabic text, 88), and which is now pronounced Tamgūna in Ḥassāniyya. The variations in spelling follow usual patterns: kāf is used to represent g; td’ is used in place of the final dhāl of the Znāga form (cf. the Znāga tribal name Tashidbīdh which became Tāshidbīt in Ḥassāniyya); and the vowels of Znāga names and words written in Arabic script vary, probably reflecting a lack of exact equivalents in some cases, variations in pronunciation in different areas, by different groups a n d a t different times, and modification by Hassāniyya speakers in accordance with the principles of euphony in their own language. The Tamagnīdh or Tamgūna are said to have been of Massūfa origin a n d are now one of t h e branches of t h e I d aw al-Hajj tribe. The name means the same as t h e Arabic mustaqīm ‘ right, straightforward’ (M.W.H.).Google Scholar

⁵⁴ The origin and meaning of this name are unknown. Vocalization suggested by M.W.Ḥ.

⁵⁵ The name may be incomplete as it is followed by a blank.

⁵⁶ Vocalized in both copies. The origin and the meaning of this name are unknown, although its form seems to be Znāga. In (masc.) and tin (fern.) mean ‘ of…’ or ‘the possessor of…’ in Znāga. Personal and place-names often take these forms. In this case it is a person, as it is preceded by bin ‘son of’, but the context does not indicate whether it was a man or a woman. Sometimes Tuareg men's names are feminine in form, and it may have once been the case among the Znāga, although it is not now so.

⁵⁷ Ashsha is a Hassāniyya form of the Znāga name Ashshan, which is a deformation of the Arabic name ‘Ā’isha (M.W.H.).

⁵⁸ The origin and meaning of this name are unknown. Cf. n. 50.

⁵⁹ This phrase is of unknown origin and meaning. Part or all of it may be a proper name, partially or entirely Znāga. The only thing which is identifiable is an, which means ‘ of’ in Znāga. The phrase is followed by ṣaḥ ḥ, an Arabic word meaning ‘ correct’, assuring the reader that it was accurately copied, suggesting that the phrase was unknown to the copyist.

⁶⁰ It is likely that a name is missing here. The fragment ‘…j r ’ suggests Ῑdayjjer, the name of one of the legendary founders of Shingīti and the name of a tribe claiming descent from him and who still live at Shingīti and are now tributaries of the Ῑdaw ‘Ali. Cf. the Tinbuktu-Koy, Muḥammad an-Adḍa, al-Ṣanhājī (Znāgī) of the tribe of Ajjar, a native of Shingīti, mentioned in ‘Abd al-Rahmān ibn ‘Abd Allāh al-Ṣa’dī, Tārīkh al-Sudān, translated by O. Houdas, edited by E. Benoist (Publications de 1’ÉScole des Langues Orientales Vivantes, iv^e Sér., Vols. xii - xiii ), Paris, 1898–1900, Arabic text, 22, French translation, 38.

⁶¹ This is probably the same as given in n. 43.

⁶² Q ṭ ḥ is unknown. Hawa appears only on the page numbered 8. Perhaps Faṭim or Faẓma (Hawa) was intended. It might be a deformation of the Znāga word kuẓ ẓih ‘ the last (or) the youngest child’. Hawa and Ahawa are Znāga forms of the Arabic Hawwā’ ‘ Eve’, which is pronounced Hāwa in Hassāniyya (M.W.H.).

⁶³ The origin and meaning of this name are unknown. Yajmar means ‘ strong’ or ‘ capable’ in Znāga. The suffix ’ āk could mean ‘ nevertheless’. Ak is also a Znāga masculine proper name. Cf. Bādhal Agdh Ak al-Majlisī, Bādhal, son of Ak of t he tribe of Madlish, mentioned in Muhammad Wālid ibn al-Muṣṭtafā ibn Khālunā al-Daymānī, op. cit., Arabic text, 91. Ak, whose name means ‘ master of knowledge’, came, according to the traditions of t he Madlish, to t h e Sahara from Morocco with Abū Bakr ibn ‘Umar, the Almoravid chief (M.W.H.).

⁶⁴ Origin, meaning, and gender unknown. Transcription hypothetical. Cf. T r j t, discussed in n. 104, and Targa, from which Tuareg may have been derived. Targa is also a common toponym in north-western Africa, and in some Berber dialects or languages it means a canal. (Cf. n. 84 and n. 97.) Another possibility is the Znāga word tureg ‘ she has a child’ or ‘ she has given birth to a child’ (M.W.H.).

⁶⁵ This name could also be vocalized Fāṭma and Fāṭim. It is preceded by a blank space, so its position on the tree is hypothetical.

⁶⁶ The final alīf this name could be rather the first letter of the next word, which would thus be ibn rather than bin. Amza or Amzā resembles the Znāga word amthih ‘ small’ (M.W.H.).

⁶⁷ Origin and meaning unknown; vocalization hypothetical. Aburaj means ‘ lazy’ in Znāga. Aktūlīl could be a deformation of Ak Tū’tjitj ‘ son of a naked woman’, or Tūlīl or Tū’tjitj could have been the name of a tribe (M.W.H.).

⁶⁸ This name is followed by an illegible word in both copies which might be a b 1. Ahamut is a Znāga form of Aḥmad. The second word might be a deformation of awbwaytj, Znāga for ‘slave’. Cf. n. 108. (M.W.H.)

⁶⁹ See n. 57.

⁷⁰ Possibly an alīf at the beginning of t he name was omitted and the third letter was a dāl which was confused with rā’, which is written in a similar way. The name could thus be read Ak, or Ag, Dagh Zaynbu, a Znāga form of filiation meaning ‘ a son of the sons of Zaynbu’, that is, a member of t he tribe called Ῑdagh Zaynbu in Hassāniyya. It is possible that this form, the tribe, and the Znāga language had disappeared in t he north-western Sahara by t h e time the text was compiled or copied, and that the original text was mis-read. See n. 130 on the Ῑdagh Zaynbu. Another possible interpretation is Aknr Agh Zaynbu ‘ Akur, son of Zaynbu ’.

⁷¹ See n. 57.

⁷² This name could also be vocalized Hunṭ or Hint. It is a Znāga, form of Ahmad (M.W.H.).

⁷³ This name appears only on the page numbered 8. It s origin and meaning are unknown, although i t is presumably Znāga (M.W.H.).

⁷⁴ The most common pronunciation of this name is Hunz, but it may also be pronounced Hanẓ or Hinẓ. It is a Znāga form of Ahmad (M.W.H.).

⁷⁵ This could be a form of the name of the tribe which is now called Dhawāyu (by the Tandagha and Ikumlaylan tribes) or Dhawātji (by the Awlād Daymān and Idabalḥasan tribes) in Znāga and Ῑdaw ‘Alī in Hassāniyya (M.W.H.). The Ῑdaw ‘Alī claim that their ancestor came from Tābalbāla to the Adrār at the time of the Almoravids, and that they founded the town of Shingīṭi in the seventh/thirteenth century. See Norris, H. T., ‘The history of Shinqīt according to the Ῑdaw ‘All tradition’, Bull, de l’IFAN, sèr. B, xxiv, 3–4, 1962, 393–413. Cf. n. 39, n. 82, n. 102, n. 111, and n. 123.Google Scholar

⁷⁶ A Znāga form of Fāṭima (M.W.H.).

⁷⁷ See n. 43.

⁷⁸ M.W.Ḥ. feels that this may be al-Ḥājj Ya ‘qūb Yabnī, one of the founders of Wadān and ancestor of the Ῑdayqūb branch of the ūdaw al-Ḥājj, according to the traditions of that tribe. The descendants mentioned here, however, are not known from Idaw al-Hajj tradition, and the Ya’qūb in this genealogy would have lived several centuries after al-Hajj Ya’qūb Yabnī, who, according to Ḥdaw al-Ḥajj tradition, lived in the sixth/twelfth century, although it may indicate that he was not among the founders of Wadān or that it was not founded in the sixth/twelfth century.

⁷⁹ See n. 62. This name is followed by a blank space on the page numbered 7, but by ‘daughter of’ on the page numbered 8.

⁸⁰ This is a variant of Agīlāl, a Ḥassāniyya deformation of the Znāga Agūtjitj or Agūyayh ‘ he who has his tail cut off’, meaning ‘ without children’. This was one of the names of good omen given in the hope that what was described would not become a reality (cf. n. 48). A major branch of the Tajakānet is named Lagwālil, after their reputed ancestor A’mar Agilāl, and a branch of the Ῑday Shilli tribe is also named. The name was once common, and the person mentioned in the text cannot be identified (M.W.H.).

⁸¹ A Znāga woman’s name (M.W.H.).

⁸² Āmir is an Arabic name. On Allu, see n. 39, n. 75, n. 102, n. 111, and n. 123. A w d ṭ is of unknown origin and meaning, and of uncertain vocalization (M.W.H.).

⁸³ See n. 42. This name is followed by a blank space, so the position of the following names is hypothetical.

⁸⁴ The kāf bears a fatha in the second version on the page numbered 8. It might be a second name of Malyam, although there is a blank separating them. It would not be her mother's name, as the next phrase says that her mother was named Fāṭima. It might be a man's name (of. Tarja, n. 64), who might have been her father, as the names of the fathers of most of the other women in this branch are given. Another reading might be T z g, as the dot on the zāy could have been omitted or erased, and the copyist used kāf elsewhere to represent g (cf. n. 110 and n. 127). This reading would suggest the tribal name of Tayzegga, which reading is supported by the tradition that many of the Tayzegga are of Bāfūr origin, who are associated with T r k or T z g in this text, and it is also said that the Tayzegga originated in the town of Az؛gi, which, according to some traditions was inhabited by Bāf؛r before the arrival of the Almoravids in the fifth/eleventh century. If T z g is a correct interpretation, it could also be a variant of Az؛gi. Ab؛ ‘1-’Abbās Aḥmad ibn ‘AlĪ ibn ‘Abd al-Qādir al-Ḥusaynī Taqī al-Din al-Maqrīzī (d. 845/1442). Al-bayān wa ‘l-I‘rāb ’ammā bi arḍi Miṣr min al-A‘rāb, Cairo, 1961, 57, mentioned a tribe of the Sanhāja called Tizkīk who made shields of lamṭ hide, who were perhaps the Tayzegga. Another possibility is al-T r k, an ancestor of the Īdnān, who are now considered Tuareg, mentioned by Muḥammad Wālid ibn al-Muṣṭafā ibn Khālunā al-Daymānī, op. cit., Arabic text, 91. Still another possibility is the name of a tribe of uncertain origin called al-Tūwārīk, formerly affiliated with the Īdaw Hājj, now with the Brākna, a branch of the Hassān, who are today very few in number (M.W.H.). Ibn Khaldūn, op. cit., Beirut edition, vi, 408,-mentions T r i k h in connexion with the Lamt؛na.

⁸⁵ According to various traditions in the Adrār, the Bāfūr were blacks, Jews, Portuguese, or giants who lived in the Adrār before the arrival of the army of Abū Bakr ibn ‘Umar, the Almoravid chief, but it would seem likely that they were simply an old Znāga tribe which may have been of considerable importance, but which was defeated, possibly disgraced, and absorbed by other tribes, as most of the Gudāla or Judāla may have been. The presence of their name in this text and the names of persons called al-Bāfūrī in the twelfth/eighteenth-century accounts of the history of the Tashumsha confederation, would seem to confirm that the Bāfūr were in no way different from other Znāga tribes. The original form of the name of the region of the Adrār may have been Adrār an-Bāfūr ‘the mountain of the Bāfūr’, as suggested in the accounts of ninth/fifteenth-century Portuguese explorations.

⁸⁶ A Znāga man's name still in use. A person by this name is said to have been the ancestor of the Dayllan branch of the Angādis (see n. 106 and n. 110), and it occurs in the name of a well Tindaylla ‘of the Dayllan’, in the Gibla (M.W.H.).

⁸⁷ This name may indicate that the person was from the Adrār or of the tribe of the Awlād Tīdrārīn, who are said to have originated in the Adrār, taken their name from the region, and been originally associated with the Īdaw al-Hājj. They now live mainly between the Adrār and the ocean (M.W.H.).

⁸⁸ Incomplete. Unidentified.

⁸⁹ The page numbered 7 gives a kasra (i) for the vowel following the initial tā’ but M.W.Ḥ. says that it would seem to be an error and the correct form is probably Taruzga, a feminine form of Yaruzga (see n. 43), but it would seem to have been a man according to what seems the best interpretation of the text. Another interpretation could be that Taruzga was the mother of … b sh, and that al-Zaharā was the mother of Taruzga. It could also be that names feminine in form were once used for men among Znāga speakers, which is the case among the Tuareg today, but not by Znāga speakers.

⁹⁰ This name suggests Dashaq, a Znāga deformation of al-faqīh, which is the name of a branch of the tribe of Ahl Buradda, who are a branch of the Awlād ‘Allūsh, a branch of the Hassān. It also suggests the Znāga name Shugan, called Ishugānan in Hassāniyya, a branch of the Tandagha, a zwāya tribe, part of which is still Znāga-speaking (M.W.H.).

⁹¹ Origin, meaning, and gender unknown. Presumably Znāga (M.W.H.).

⁹² See n. 57.

⁹³ This name is not now in use as a personal name. It resem bles the Znāga tribal name Dheg Ajmallan, or Idagjmalla in Hassāniyya (M.W.Ḥ.). In Muḥammad Wālid ibn al-MuṢṭafā ā ibn Khālunā al-Daymānī, op. cit., Arabic text, 91, it is written Ijmalla.

⁹⁴ This could have referred to someone from the tribe of Īdagh Zaynbu. See n. 70, n. 129, and n. 130.

⁹⁵ See n. 57.

⁹⁶ Tagunda now means ‘sheep-raising nomads’ in Znāga, but the name seems to be incomplete (M.W.Ḥ.).

⁹⁷ Not clear in text. See n. 64 and n. 84.

⁹⁸ Origin, meaning, and vocalization unknown. Probably Znāga. The context and form would suggest that the name is feminine (M.W.Ḥ.).

⁹⁹ Unknown to M.W.Ḥ. It is a Tuareg woman's name. See Foucauld, Charles de, Dictionnaire abrégé touareg-franç de noms propres (dialecte de l’Ahaggar), published by André Basset, Paris, 1940, 291.Google Scholar

¹⁰⁰ Origin, meaning, vocalization, and gender unknown. Because of blank spaces in the text at this point, the position of the following name is hypothetical.

¹⁰¹ Unidentifiable.

¹⁰² See n. 39, n. 75, n. 82, n. 111, and n. 123.

¹⁰³ Also pronounced Hanṭ and Hinṭ A Znāga form of Aḥmad (M.W.Ḥ.).

¹⁰⁴ This is probably the same name as Turjūt or Tūrjūt mentioned by Ibn ‘Idhārī in al-Bayān al- Mughrīb fī akhbār mulūk al-Andalus wa ‘l-Maghrib, published by Miranda, Ambrosio Huici, ‘Un fragmento inéditio de Ibn ‘Idārī sobre los Almorávides’, Hespéris Tamuda, 11, 1, 1961, 63;Google Scholar Turqūt mentioned by ’l-Ḥasan, Abū ‘Ali ibn Muḥammad ibn Aḥmad ibn ‘Umar ibn Abī Zar’, Kitāb al-anīs al-muṭrib bi-Rawḍ al-Qirī ās fī akhbār mulūk al-Maghrib wa tārīkh madīnat Fās, Arabio edition, Rabat, 1946, 11, 45; and Turkūt mentioned in Maqrīzī, op. cit., 57; who according to these three authors of the eighth/fourteenth and ninth/fifteenth centuries was an ancestor of the Almoravid chiefs. The variations in spelling are of minor importance: Ibn ‘Idhāri used two different spellings in the same passage. Jīm, kāf with or without three dots, and qāf with one, two, or three dots are often used interchangeably in the transcription of Znāga names by Arabic authors, and the vowels used in Arabic transcription of Znāga names also vary. The name seems to appear only in the lineage of the Almoravid chiefs and of Sīdī Aḥmad al-Bakkāy, and its meaning is not known. According to Ibn ‘Idhārī and Maqrīzī, Turjut was the great-grandfather of Abū Bakr ibn ‘Umar and Yūsuf ibn Tāshfin, but according to Ibn Abī Zar’ only of Yūsuf ibn Tāshfīn. According to Ibn ‘Idhārī and Ibn Abi Zar’, Turjut was a man, but according to Maqrīzī Turjut was a woman. The fragment of the text of Sīdī Aḥmad al-Bakkāy does not indicate the sex of T r j t, since the name is the last in the lineage. The form would seem to be feminine, and such names are currently used among the Tuareg for both men and women, but it is not known if this usage ever existed in the western Sahara. Possibly Hunṭ (Aḥmad), son of T r j t in the lineage of Sidi Aḥmad al-Bakkāy could be identified with Ḥamid or Muḥammad, sons of Turjut in Ibn ‘Idh āri. From T r j t to Sīdī Aḥmad al-Bakkāy is 12 generations, or about 300 years, if generations averaged about 25 years, which may be about right as eight of the twelve persons were women, three men, and one undetermined (T r j t). If Sīdī Aḥmad al-Bakkāy died in about 920/1514–15, as the Ghallāwiyya says, his ancestor T r j t would have lived at the beginning of the seventh/thirteenth century, while Turjut, ancestor of the Almoravid chiefs, would probably have lived in the fourth/tenth or fifth/eleventh century, although names could have been omitted from or added to either or both of the lineages. The early position it occupies in the two lineages, and the fact that it does not seem to occur elsewhere, suggest that it may be the same person, and it could have been a woman.Google Scholar

¹⁰⁵ This name is only partially vocalized in the text. Origin, meaning, and gender unknown. Presumably Znāga.

¹⁰⁶ This is probably a group which is now part of the Īday Shilli tribe and which is now called Īdagh Maḥḥam. Hum and Maḥḥam are both deformationas of Muḥammad (M.W.Ḥ.).

¹⁰⁷ This is probably the Antāda branch of the Idagh Hum or Idagh Maḥḥam, which still exists today (M.W.Ḥ.). See n. 106, above.

¹⁰⁸ This is a Znāga deformation of Aḥmad Fāḍil (M.W.Ḥ.). Cf. Humud Faḍaj in Muḥammad Wālid ibn al-MuṢṭafā ibn Khālunā al-Daymānī, op. cit., Arabic text, 77.

¹⁰⁹ The blank space may have contained the name of a branch of the Angādis, which follows, which the copyist may not have been able to read. A d could be read as Ada, a Znāga form of Aḥmad, or it could be read as Id, a Ḥassāniyya deformation of the Znāga prefix Dhaw, equivalent to the Arabio Awlād (M.W.Ḥ.).

¹¹⁰ According to western Saharan tradition, the Angādis were one of the major Znāga warrior tribes of the central western Sahara until the tenth/sixteenth or eleventh/seventeenth century, when they moved to the south-west where they declined in power and numbers; there are fewer than 10 persons surviving today. The origin and meaning of their name is unknown, and no memory of their exact origin or of the names and genealogies of their chiefs has been conserved, although it is said that they were of the Lamtūna. The tribe was divided into nine branches: Ῑdaw Yidda, Ahl Aḥmad an Adanan, Ῑdaw Zaghar, Ῑd Ag Shatjah, Ῑdayndik, Ῑdaw Yidder, Ῑdayndukān, Daygartan, and Dayllan. A part of the now important warrior tribe called Ῑdaw ‘Ῑsh, also called the Znāga, may be the Ῑdaw Yidder under a new name, as Yidder and ‘Ῑsh have similar meanings, being derived from the verb ‘to live’, in Znāga and Arabic (M.W.Ḥ.).

¹¹¹ Seen. 39, n. 75, n. 82, n. 102, and n. 123 on Allu. A w y a n is of unknown origin, meaning, and gender, and the transcription is not certain. The context does not indicate whether the name is masculine or feminine, although Āmir is masculine and Allu usually precedes masculine names.

¹¹² See n. 62.

¹¹³ See n. 52.

¹¹⁴ Origin and meaning unknown. Presumably Znāga. It is always vocalized Kunta. The context would make it masculine. It is no longer in use as a personal name, and this would seem to have been the only person known in the western Sahara to have had this name (M.W.H.).

¹¹⁵ Origin and meaning unknown. It is written Zam in the Challāwiyya; Zam is a currently used man's name in the Gibla among Ḥaraṭīn (M.W.H.).

¹¹⁶ Origin, meaning, and vocalization unknown. Presumably Znaga (M.W.H.).

¹¹⁷ This could be a deformation of the Znāga phrase lanaffi’dh an-babba ‘useful to his father ’ (M.W.H.).

¹¹⁸ Origin, meaning, and vocalization unknown. It could be an incorrect rendering of Aghashshant, a diminutive of ashshan ‘known’ (M.W.H.).

¹¹⁹ See n. 14.

¹²⁰ This is probably a Ḥassāniyya deformation of the Znāga Yidda an-Hund ‘the grandfather of Ahmad’ (M.W.H.).

¹²¹ Origin and meaning unknown. It is presumably Znāga, and the context indicates that it is masculine.

¹²² This name, which is more commonly written Ῑday Shilli, is a Ḥassaniyya form of the Znāga name, Dhaw Yashshitj, meaning ‘the descendants of the living one’, thus being similar in meaning to Dhaw Yidder and Ῑdaw ‘Ῑsh (cf. n. 110). It is not known whether Yashshitj was a person or if the name was only one of good omen (cf. n. 48). According to tradition, the Ῑday Shilli were of Ḥimyar and Lamtūna origin, and they were among the forces of the Almoravids, although they are not mentioned by name in any of the known Almoravid chronicles and no memory has been preserved of the names and genealogies of their chiefs from that period, nor exactly how they might have been connected. They were considered a major warrior tribe of the central western Sahara until their defeat by the Ḥassān in the eleventh/seventeenth century, when they were reduced to tributary status. They also apparently included a zwāya branch, the Ῑdaylba, a daughter of the founder of which was the wife of Sīdī ‘Umar al-Shaykh, son of Sīdī Ahmad al-Bakkāy (M.W.H.).

¹²³ See n. 39, n. 75, n. 82, n. 102, and n. Ill on Allu. The origin, meaning, gender, and vocalization of A w l m is not known. The name would be masculine in this context, and Allu usually precedes masculine names.

¹²⁴ Reading uncertain. Neither possible reading suggests any known name.

¹²⁵ A name currently used in the Tagānit according to Kunta from that region.

¹²⁶ This name would mean ‘the son of Akkādi’ or ‘the son Akkādi’. Ak is a Znāga form of filiation. Akkādi would be a proper name; its origin and meaning are not known.

¹²⁷ See n. 110.

¹²⁸ See n. 120.

¹²⁹ The third letter could be read as a rā’ or as a dāl. See n. 70 and n. 94.

¹³⁰ Now usually written Ῑdagh Zaynbu. See n. 70 and n. 94. According to present western Saharan tradition, the Ῑdagh Zaynbu are shurafā’ descended from Zaynab, the daughter of Fātima, the daughter of the Prophet. It is said that four men, Sīdī Muhammad al-Kuntī; Muhammad ‘Abd Allāh, ancestor of the Madlish; his brother, ‘Abd al-Rahmān al-Rakkāz; and the ancestor of the Ῑdagh Zaynbu came to the Sahara with Abū Bakr ibn ‘Umar, the Almoravid chief (Muḥammad ibn Ḥabat al-Ghallāwī, unpublished Arabic text in the collection of the Ḥabat family of Shingīṭi). According to Kunta tradition, Sīdī Muḥammad al-Kuntī would have lived some 400 years later, so, if the tradition of the Kunta were correct, which is more likely, he would not have been among them. Whatever the origins of the Ῑdagh Zaynbu were, one could deduce from the text of Sīdī Ahmad al-Bakkāy that they existed by the end of the seventh/thirteenth or the beginning of the eighth/fourteenth century. According to western Saharan tradition, their original territory was to the north of the Adrār, between the Adrār and the Zammūr, but they are now in the extreme south-western part of the Sahara. They are now a small zwāya tribe linked to the Tajakānet, which confirms their northern origins and early relationships.

¹³¹ See n. 41. Ῑd al-Fagha is a Ḥassāniyya deformation of a Znāga equivalent of the Arabic Awlād al-Faqīh. It would seem that this was a branch of the Ῑdayshif, although it is not known today (M.W.H.). Cf. n. 41 and n. 140.

¹³² Origin a nd meaning unknown; presumably Znāga. This name occupies the same place in Tajakānet tradition.

¹³³ This name occupies the same place in Tajakānet tradition.

¹³⁴ Origin and meaning uncertain. Presumably Znāga. This person, who is also called Jākan, and Jākir in Arabic and Tjegen in Znāga, is considered by the Tajakānet as their ancestor. According to Shaykh Muḥammad ibn al-Mukhtār ibn Sa‘id al-Yadālī (d. 1166/1753), Kitāb shiyam al-zawāya, published and translated into French by Hamet, Ismael in Chroniques de la Mauritanie Sénégalaise, Paris, 1911, Arabic text, 73, ‘The Ῑ;daw al-Ḥājj, the Tajakānet, and the Tāfullālt are the sons of Sharuwāl, son of Awān, son of ‘Alī, son of Amsam, son of Yahkadkar, son of Nabtān, son of Lamtūn, and their ancestor Amsam is the ancestor of the Masūma, and the origin of all of them is the Masūma’.Google Scholar

¹³⁵ This person occupies the same place in Tajakānet tradition and is considered the ancestor of the Ramāẓin, one of the principal branches of the Tajakānet.

¹³⁶ Mussā’ is a Znāga deformation of Mūsā, and Mussa’ān means ‘my Mūsā’ (M.W.H.). According to Tajakānet tradition, he is known as Mūsānnī and is considered the ancestor of the Awlād Mūssānī branch of the tribe.

¹³⁷ According to Kunta tradition, Sīdī Muḥammad al-Kuntī al-Kabīr is buried at a place named Faểk, between the Adrār and the ocean. Kunta tradition says also that Sīdī Aḥmad al-Bakkāy was young when his father, Sīdī Muḥammad al-Kuntī al-Kabīr died.

¹³⁸ Origin, meaning, and vocalization unknown. The name might be related to the name Ward which appears in the same lineage in the Kitāb al-ṭarā’if and to Wārith, which appears in similar lineages for the Ῑdaw al-Ḥājj Ῑdayqūb.

¹³⁹ M.W.Ḥ. believes that this is a Znāga deformation of al-Ḥasan. When Arabic words are borrowed by Znāga, lām often becomes tj and sometimes yā’ or i. Ḥā’ usually becomes hā’. Various Znāga forms of al-Ḥasan are Yahas, Atjhas, and Adhas. The tribe called Ῑd Ab al-Ḥasan in Ḥassāniyya is called Dabwayhas and Dabutjahas in Znāga.

¹⁴⁰ If the fragment of Sīdī Aḥmad al-Bakkāy is authentic, it would contain the first known use of the nisba al-Jakāni, but since the tribal name Tajakānet is contained only in a marginal note, which may not have been part of the original, the earliest known mention of the Tajakānet in more or less this form would be somewhat later, so the exact relationship between the two is not absolutely certain. The Tajakānet say that they are descended from a person named Jakān, Jākan, or Jākir (see n. 134), and that the name of the group, Tajakānet, is derived from his name, the singular form being al-Jakānī. Present-day Znāga speakers say that the ancestor of the tribe was named Tjegen, and they call the tribe Dheg Tjegent ‘the descendants or the people originating from Tjegen’, the singular of which would be Agda Tjegen, and it is assumed that Tajakānet is a deformation of Dheg Tjegent. The Tajakānet were presumably Znāga-speaking at the time of their formation—the last of the Znāga-speaking Tajakānet may have died around 1318/1900—but there is no evidence as to what terminology they may have originally used, and there is no known written text in Znāga supporting the antiquity of current Znaga usage. Furthermore, the authenticity of the now accepted genealogical explanation of the origin of the Tajakānet is far from certain: neither Jakān, Jākan, Jākir, nor Tjegen may have been the original form of the name, and it may not have been a person; perhaps the original collective name was neither Dheg Tjegent nor Tajakānet, and possibly it did not mean what it is now assumed to mean.

¹⁴¹ Nevertheless, no attempt was made to link Jakān with the Lamtüna and Ḥimyar, as is now done. It is now claimed that the Tajakānet, the Ῑday Shilli, the Angādis, and the Abdūkal were Lamtūna, descended from Ḥimyar; that the Tamgūna were Massūfa, cousins of the Lamtūna; and that the Ῑdagh Zaynbu and the Ῑdaw ‘Alī (Awlād Ayyaw?) were shurafā’, although no such claims are made in the fragment of Sīdī Aḥmad al-Bakkāy, and it is possible that the present claims arose comparatively recently, as may have been the case with the Kunta's claim to having descended from Quraysh. In the case of the latter two, the Ῑdagh Zaynbu and the Ῑdaw ‘Alī (Awlād Ayyaw?), an accidental similarity of names with those of early Islamic figures may have suggested the connexions.