Introduction^{Footnote 1}

The historiography of pre-nineteenth century Aïr and Ader relies heavily on the published translations of copies of Arabic texts, known as the Agadez Chronicles, that local chiefs made available to colonial officers at the beginning of the twentieth century.^{Footnote 2} Basing their insights on these sources, historians of the region accepted a particular interpretation of the establishment of the Sultanate of Agadez that has two main strands. According to the first of these, the representatives of the main Tuareg confederations living in this region toward the end of the fifteenth century sought to find a sultan who would arbitrate amongst them. Having identified such a ruler, after some failed attempts at finding an appropriate location for him, they installed their sultan in Agadez and imposed him upon the whole of Aïr. In the second strand of this interpretation, a group of named “supporting tribes”^{Footnote 3} – Lissawan, Iberkoreen, Imiskikian, and Inoussoufan – became the sultan’s closest supporters in Aïr and in the second half of the seventeenth century played a critical role in the conquest of Ader by Agabba, son of Muḥammad al-Mubārak the Sultan of Agadez. Some of these tribes, the Lissawan in particular, were consequently given land and political privileges in Ader.^{Footnote 4}

This article suggests that this interpretation should be re-examined critically because it relies mainly on specific versions of the Agadez Chronicles (text LVII published by H.R. Palmer in 1928^{Footnote 5} and text B published by Yves Urvoy in 1934)^{Footnote 6} and Y Tarichi, which have not been subjected to careful critical analysis. It shows that these published versions of the Chronicles come from exactly the same source: they are copies created in 1907 and circulated by the Sultan of Agadez, Ibrahim Ed Dasouqy. The Y Tarichi was made available, also in 1907, to Captain Laforgue when he visited Agadez.

At the beginning of the twentieth century, when France was struggling to crush the resistance of the two powerful Tuareg confederations of the Kel Gress and Iwellemmeden Kel Denneg, the Y Tarichi and some versions of the Agadez Chronicles legitimated the consolidation of the Lissawan’s administrative authority over a large portion of Ader and the establishment of a line of Lissawan canton chiefs. These texts confirmed the authority of the Lissawan, Aïr Tuareg, and Agadez Sultanate by magnifying their role in the establishment of the main political institutions of Aïr and Ader in earlier centuries. By the same token, they eclipsed the role of other actors – particularly the Iwellemmeden Kel Denneg – who had mounted a strenuous resistance against the French and against those Tuareg who supported the French with varying degrees of conviction.

Urvoy mentions casually in his introduction to the Agadez Chronicles that the manuscripts he translated had been “copied and, when appropriate, updated in 1907 during the reign of Sultan Ibrahim-ed-Dasouqy.”^{Footnote 7} It is suggested here that these documents had been altered by persons operating in the circles of the Sultan of Agadez Ibrahim Ed Dasouqy and the Ader-based Lissawan chief Amattaza Ennour, who believed they would benefit from circulating a revised text that aggrandized the role of their ancestors in the political affairs of the Sultanate of Agadez from the fifteenth to the seventeenth century.^{Footnote 8} The second group of so-called “supporting tribes” (Lissawan, Iberkoreen, Imiskikian, and Inoussoufan) appears only in copies of the chronicles “copied and updated” in 1907. The names of the second group of tribes supporting the sultan do not figure in other versions of the Chronicles or in independent sources on the history of this region, and it is likely that they were included in the versions updated in 1907 for reasons of political expediency. To be sure, in our present state of knowledge this must remain a hypothesis. The data examined in this article do not allow us to reconstruct with any degree of certainty either the genealogy of the manuscripts of the Chronicles of Agadez, or the historical dynamics described in these manuscripts. However, while we wait for a critical edition of various versions of the Arabic manuscripts of the Agadez Chronicles to appear, this article suggests why such updates were made, by whom, and with what consequences for political relations in the colonial period.

By comparing different versions of the manuscripts known today as the Chronicles of Agadez and Y Tarichi, and reconstructing the conditions of their circulation, this article exposes the political projects of some Lissawan elites, the sultans of Agadez and Sokoto, the Iwellemmeden Kel Denneg, and French colonial administrators who served in the Aïr and Ader regions in the first decade of the twentieth century. It follows Paulo Fernando de Moraes Farias’s warning not to read chronicles as “bundles of raw information,” but rather to consider critically the intellectual and political projects of those who commissioned, wrote, copied, and circulated these texts.^{Footnote 9} These individuals were not only involved in struggles over material resources such as land, people, and livestock. They were also fighting over the interpretation of a contested past. Following military defeat by colonizing powers, the former political rulers of this region became unwillingly accountable to new European authorities. Amongst various strategies, they used historical sources as political resources and circulated revised versions of manuscripts that supported their claims and undermined those of their enemies.

The aim of this article is twofold. First, it aims to reopen the debate on the historical evolution of Central Saharan political structures, and particularly on the development of the peculiar institution of the Agadez Sultanate and its hierarchical relations with surrounding satellite kingdoms and/or emirates. Second, it contributes to the comparative literature on the colonial construction of knowledge about colonized societies.^{Footnote 10} Africanist research shows that the colonial encounter gave rise to epistemological struggles over the interpretation of Africa’s past. The most contested aspects of this past were those most likely to influence the distribution of political power in the new colonial context: for example, the boundaries of ethnicities and territorial polities,^{Footnote 11} the interpretation of customary law,^{Footnote 12} and the production of an official discourse about pre-colonial ruling hierarchies.^{Footnote 13} Pioneering studies promoted the idea of a colonial “invention” of African traditions.^{Footnote 14} These interpretations were criticized by researchers who showed that Africans had contributed actively to shaping (new) understandings of tradition by feeding colonial administrators information that would be advantageous to them.^{Footnote 15} Some of this information consisted of chronicles, manuscripts, and other forms of written or oral data that the colonial administration considered sources for historical reconstruction.^{Footnote 16} This article looks at transactions and negotiations that involved this type of material.

Far from rendering sources unusable, critical inquiries in their production and circulation can shed light on two interrelated orders of questions: questions on the sources’ topics and/or contents; and questions on the strategies of their authors, copyists, compilers, and publicizers. Reconstructing processes of textual inclusions, exclusions, and rewriting of sources can yield valuable information on the intellectual and political visions of those making textual revisions.

It goes without saying that this approach is not relevant only to the moment of the colonial encounter. Critical studies of African sources have revealed many occasions in pre-colonial times when oral performers, writers, chroniclers, and copyists intervened in the presentation of various types of texts with the aim of unfolding their own idiosyncratic projects. Thus, for example, de Moraes Farias has shown that Al-Sa’dī’s Ta’rīkh al-Sūdān was written to provide an account of history that would reconcile the visions of descendants of the Moroccan invaders and those of the Sahelian communities of the Niger Bend.^{Footnote 17} Murray Last, M.G. Smith, John Hunwick, and Paul Lovejoy have investigated possible authors, compilers, and editors of various versions of the Kano Chronicles in order to explain the intellectual and political functions of these texts.^{Footnote 18} Most recently, Nobili and Mathee have studied multiple versions of manuscripts of the Tārīkh al-fattāsh, reassessed a century of critical studies of these sources, and provided a new interpretation of the political motivations behind successive interventions in these texts.^{Footnote 19} This article turns to the Chronicles of Agadez in the hope of stimulating new inquiries into the history of this region and its deep historiography.^{Footnote 20}

Following this introduction, the article’s second section discusses early sources on Aïr and Ader and introduces the main institutions under discussion here. The third examines the main published versions of the Agadez Chronicles and Y Tarichi, and compares their contents. The fourth reconstructs the production and circulation of different versions of the Agadez Chronicles and Y Tarichi in the first decade of the twentieth century, focusing particularly on the agency of the sultans of Agadez and Sokoto. The fifth section discusses the significance of revisions in the context of Lissawan–Iwellemmeden Kel Denneg–French relations at the beginning of the twentieth century. Finally, the conclusion advances some reflections on the implications of this reassessment for the historiography of the region.

Agadez and Ader Outside the Agadez Chronicles and Y Tarichi

This section introduces institutions that figure in the Agadez Chronicles and Y Tarichi as they had been observed by early travelers and authors who wrote about the central Sahara and its people. It discusses sources that provide information on political relations in Aïr and Ader from the fourteenth to the seventeenth century – roughly the period covered in greater detail in the Agadez Chronicles and Y Tarichi. Of course, it cannot be excluded that the sources examined in this section simply omitted to refer to events and names mentioned in the Agadez Chronicles and Y Tarichi. However, the comparison is useful because these early sources are precisely where one would expect to find these events and names mentioned, had they been perceived as important in their own time. Some early sources provide snapshots of the political institutions that preceded the installation of a sultanate in Aïr, while others represent the first recorded evidence on the young sultanate. The primary function of the various versions of the Agadez Chronicles, when they were composed, was to explain the origin of the sultanate. By contrast, the texts discussed in this section mention Aïr in writings that were not focused specifically on Aïr’s political institutions.

Writing in the 1330s, al-’Umarī mentioned “three independent white Muslim kings who are Berbers; the Sultan of Ahīr (Aïr), the Sultan of DMWShH, and the Sultan of Tādmakka. (…) Each of them is an independent sovereign; and no one of them rules another, but the greatest is the king of Ahīr.”^{Footnote 21} When in 1353 ibn Baṭṭūṭah reached “Kahir” travelling in a caravan of six hundred female slaves from Takedda, he described this area as “the country of the Sultan of al-Karkari,” but did not specify the name of the place where this unidentified sultan resided.^{Footnote 22} It is noteworthy that al-’Umarī’s description of the Sultan of Aïr as white, Muslim, and Berber contrasts with later representations of Agadez sultans emphasizing blackness and non-Berber origin. Agadez appears again as a seat of power in the second half of the fifteenth century, when the Egyptian scholar al-Suyūṭī wrote a treatise for the kings and sultans in the land of Takrur, in which he addressed himself specifically to King Muḥammad b. Sattafan, “Lord of Agadez” and his brothers, and to Ibrahim Sura, King of Katsina.^{Footnote 23}

The sobriquet “bin Sattafan” implies a dark complexion. Yet “dark complexion” can be interpreted in various ways depending on the cultural colour code that identified people and places – and the moral implications of such a code in folk theories of race and identity. Baz Lecocq’s analysis of Tuareg theories of race suggests that “sattafan” referred to a shiny greenish or bluish black that in some regions was associated with the Tuareg nobility – and not, as has been suggested, with blacksmiths, slaves, and Sudanic Africans.^{Footnote 24} This implies that al-Suyūṭī’s reference to the King of Agadez as King Muḥammad b. Sattafan in the second half of the fifteenth century may have implied Berber noble descent and not – as commonly believed – Sudanic, or possibly slave, ancestry. I return to the significance of this in the following section.

The Taʾrīkh al-fattāsh and Taʾrīkh al-Sūdān mention an expedition of the Askyia Muḥammad in 1500–1501 against “Tildza” in “Ayar.” While Houdas and Delafosse do not advance any interpretation on this point,^{Footnote 25} Henri Lhote identifies “Tildza” with Tadeliza in Aïr.^{Footnote 26} Hunwick does not discuss Lhote’s hypothesis but, stressing that the translation of this sentence presents difficulties, explains that “Tilza Tanat was the sobriquet of Muḥammad b. ʻAbd-ar-Raḥmān, the tenth sultan of Agadez,” and renders al-Sa’dī’s relevant passage in the Taʾrīkh al-Sūdān as follows: “[Askiya Muḥammad] in 907/1501–1502 campaigned in Aïr and drove Tilza from his sultanate,” imposing a degree of Songhay rule over this region.^{Footnote 27}

Leo Africanus, or al-Ḥasan ibn Muḥammad al-Wazzān al-Zayyātī al-Fāsī, claims to have visited Agadez in 1513.^{Footnote 28} He portrays Agadez as a walled city with well-built houses that hosted a flourishing trade with Kano and Gobir,^{Footnote 29} and mentions that the sultan was a “native of Libya.”^{Footnote 30} Leo writes that in Agadez the sultan’s power depended on the support of the desert people, who also manned his army. This made him a weak and unstable regent. The inhabitants of the southern part of the Agadez kingdom are characterized as pastoralist nomads, who transport their tents of straw mats on oxen in their travels. According to Leo Africanus, Agadez was “built by the moderns,” and paid a tribute of 150,000 ducats to the king of “Tombutto.” Hunwick gives 1413 as the date when Agadez was built,^{Footnote 31} based on Luis del Mármol Carvajal’s indication that Agadez was founded one hundred and sixty years before his writing in 1573.^{Footnote 32} Hamani discusses alternative dates and concludes that it was founded between 1413 and 1424.^{Footnote 33} It is noteworthy that by specifying that the sultan was a “native of Libya” Leo Africanus was suggesting that the sultan was a local – that is, one who came from the desert region (or the third division of Africa as described at the beginning of his Descrittione) and not from the fourth division or “land of the blacks” (la terra di negri).^{Footnote 34} This again suggests that the tradition of the black descent of the sultan, or of the servile origin of the sultan as the son of a concubine, must have been introduced after the sixteenth century.

Regrettably, information on the historical development and expansion of the Agadez Sultanate does not figure in the writings of the rare Europeans known to be acquainted with the Central Sahara in the seventeenth and eighteenth centuries. Those we know of were not explorers, but traders, missionaries, and, possibly, slaves. Amongst them, Vincenzo Matteo, a trader from Ragusa (Dubrovnik) cited by Lorenzo d’Anania as one of his key informants, appears to have visited Kano in the first half of the sixteenth century.^{Footnote 35} Yet d’Anania’s “Universale Fabrica” contains scanty information on Agadez and follows Leo Africanus’s description of the city. Uncertainty surrounds the odyssey of Pieter Farde, a Franciscan brother supposedly captured by pirates in 1686 in the Mediterranean and sold to a man who brought Brother Farde to Agadez as a slave. The sultan during Farde’s captivity would have been Muḥammad al-Mubārak, who, together with his son Agabba, figures in the Chronicles and Y Tarichi. Farde’s letters, of doubtful authenticity, mention the presence in Agadez of more than forty Christian slaves, mostly from Italy and France. But they say nothing of regional political arrangements and do not mention the names of the sultan or local dignitaries.^{Footnote 36} In 1711 the Franciscan missionaries Father Carlo Maria di Genova and Father Severino da Silesia may have visited Agadez, whose sultan at the time would have been Muḥammad Agabba b. Muḥammad al-Mubārak, supposedly the leader of Agadez’s expansion into Ader. The two Franciscan missionaries may have reached Agadez during a trip aimed at finding Christian communities in the Sudan. From Agadez they continued their journey to Katsina, where they are reported to have died without leaving records of observations made during their journey.^{Footnote 37}

Records deriving from early travel reports on the central Sahara and Sahel do not mention Ader. Perhaps the earliest mention of Ader is to be found in Manuscript J of Urvoy’s Chroniques corpus, the biography of Abou-Bakr, son of Attaher-Tachi, born in 1657, which does not mention the “supporting tribes,” but indicates that during the author’s lifetime Ader was conquered by the sons of the Sultan of Agadez; that on its soil wars were fought between Tuareg and Gobirawa; and that Agadez had diplomatic and perhaps commercial relations with Ader, as the author’s profession led him to “travel in Ader.”^{Footnote 38} This text is the autobiography of an otherwise unknown person, and was not written to legitimate claims to political leadership. It confirms information provided by Muḥammad Bello and in “Notes on Some Asben Records” on the southwards expansion of Agadez and its occupation of Ader following wars with Gobir in the late seventeenth and early eighteenth century.

Muḥammad Bello’s Infāq al-maysūr fī taʾrīkh bilād al-Takrūr became available in Europe before colonial invasion.^{Footnote 39} It is possible – as suggested by Augi – that the Chronicles of Agadez and the Infāq al-maysūr were inspired by a common (lost) source,^{Footnote 40} and that they are not “independent” as Norris suggested.^{Footnote 41} However, the Infāq al-maysūr could not have been manipulated with the intent of supporting the particular configuration of interests that took shape in the initial years of colonial occupation. Like Leo Africanus, Sultan Muḥammad Bello comments on the instability of the peculiar political arrangements of Aïr, where the appointed sultan’s power depended on the support of the main Tuareg warrior tribes amongst whom he arbitrated. Aïr was initially

Rennell of Rodd interpreted Bello’s reference to a sultan from the “family of Ansatfén” as a “reference to the selection of a ruler from a slave family (…) for Ansatfén, i.e. n’Sattafan, means ‘the black ones’ from the word sattaf = ‘black’.”^{Footnote 43} As mentioned above, this interpretation may require revising, if Lecocq’s translation of “sattafan” as a hue of shiny black that implies Tuareg nobility is accepted. The second group of “supporting tribes” mentioned in some versions of the Agadez Chronicles – Lissawan, Iberkoreen, Imiskikian, and Inoussoufan – does not appear in Muḥammad Bello’s text written in the early nineteenth century,^{Footnote 44} or in nineteenth-century European sources that explicitly refer to this region, namely, those of Ulrich Seetzen, Hugh Clapperton, and Heinrich Barth.

Most significantly, the second group of “supporting tribes” does not appear in Barth’s section focusing specifically on the history of the sultanate and supporting tribes.^{Footnote 45} There are two possible explanations for this silence: either the power of these four tribes had been eclipsed temporarily in the nineteenth century, and possibly earlier, due to the growth of influence of other groups, such as the Iwellemmeden and Kel Gress; or these four tribes, under these names, had not played a central role in connection to the early sultanate, as more recent representatives of these groups claimed in their exchanges with colonial officers at the beginning of the twentieth century, when, it would seem probable, their names were written retrospectively into Aïr and Ader’s main historical chronicles.

It is important to emphasize “under these names:” the ethnonyms of the second group of “supporting tribes” found in text LVII published by H.R. Palmer 1928 and text B published by Yves Urvoy in 1934 appear to be relatively modern. This does not mean, of course, that the groups they refer to are also “modern” in any sense of this term, nor that their ancestors had not occupied positions of power in the Saharan region in earlier periods. It is particularly odd to find the ethnonym “Lissawan,” in various spellings, in a document that claims to be a copy of a sixteenth-century source, because to my knowledge a Lissawan identity developed late.^{Footnote 46} Nevertheless, one of the chiefly tribes of the Lissawan is that of the Illemteyen, which is generally acknowledged to be a form of the name of the old Lemta mentioned by Ibn Ḥawqal in the tenth century and Ibn Khaldun in the fourteenth century.^{Footnote 47} The ethnonym “Inoussoufan” is a recent variant of the older “Massufa.”^{Footnote 48} The Iberkoreyen’s history is linked to the site of In Tedoq.^{Footnote 49} And I could not find references to an ancient form of “Imiskikian,” a group which in the twentieth century was closely associated with the Lissawan chiefdom in the Keita region.

We are, therefore, confronted with two distinct questions: (1) whether the ancestors of groups known (in the twentieth century) as Lissawan, Iberkoreen, Imiskikian, and Inoussoufan had played a role in the establishment of the Aïr Sultanate; and (2) when these specific ethnonyms first appeared in the Agadez chronicles, and why. The history of ethnonyms and the history of people are separate phenomena.^{Footnote 50} The date from which an ethnonym is attested reflects the existence of a shared identity for persons grouped under such label – it does not mean that those persons, with their particular ways of living, had not existed before under one or more different labels. A “Lissawan” identity does not seem to have been present in fifteenth- or sixteenth-century Aïr (although the appearance of new evidence could change this perception). Yet it is possible that in these centuries people known as Illemteyen (Lemta), some of whose descendants would eventually become known as Lissawan, played a role in the establishment of the Agadez Sultanate and Ader Emirate or Kingdom. I suggest that at present we should question the assertion in Palmer’s text LVII and Urvoy’s text B that the Lissawan, Iberkoreen, Imiskikian, and Inoussoufan had played a role in the establishment of the Aïr Sultanate; and that this second set of ethnonyms was probably included in copies of the chronicles made in the twentieth century.

To sum up, before Muḥammad Bello’s writing we find no mention of specific, named “supporting tribes.” Limited evidence found in the writings of earlier authors suggests that up until the mid-fifteenth century, Aïr-based chiefs were supported by various groups, and their residence may have shifted in response to changing alliances and conditions of pastures. By the early sixteenth century, when Leo Africanus visited or learned about the region, it seems that the Sultanate of Aïr had become a more centralized (but not more stable) political institution, with its capital in Agadez. While asserting that the sultan was a “native of Libya” (and not of “the land of the blacks”), Leo did not provide a racialized description. Norris suggested that the discursive emphasis on a “black sultan” could have developed from close ties with Mamluk Egypt.^{Footnote 51} While the exact meaning of sattafan as “black” needs revisiting in light of Baz Lecocq’s discussion of colour coding, contacts with Egypt are mentioned in the section of the Chronicles that addresses legal questions and trade. Hamani interprets references incorporating the word sattafan in the texts of al-Suyūṭī and Muḥammad Bello as suggesting a western origin for the first sultans:

These two hypotheses are not mutually exclusive. Indeed Norris, too, had highlighted the existence of the village of Sattafan in the Adrar-n-Ifoghas in Mali, suggesting that this was the origin of the first sultans up until the reign of Sultan Ahmat Tilza in the late sixteenth century.^{Footnote 53} It is possible that some of the earliest sultans had originated from a village or group known as “Sattafan” in the Adrar-n-Ifoghas and that at some point sattafan may have acquired new racial/moral connotations due to contacts with Mamluk Egypt, first, and then with European notions of race in colonial and more recent times. Now let us turn to a close examination of the most detailed sources for this history.

The Agadez Chronicles and Y Tarichi

The Agadez Chronicles are available in the form of published translations of copies of manuscripts made in the twentieth century and handed over to colonial administrators. Between 1910 and 1936 four publications appeared that contained translations of Arabic manuscripts, published respectively by Herbert Palmer,^{Footnote 54} R. Tardivet,^{Footnote 55} and Yves Urvoy.^{Footnote 56} Palmer’s two publications and Tardivet’s article are variants of two of the ten manuscripts published by Urvoy (Manuscript A and part of Manuscript B). Urvoy’s Manuscripts C-J had not been published previously. The Y Tarichi was annexed to Colonel Peignol’s “Monographie du Cercle de Tahoua,” unpublished and dated 1907. Djibo Hamani published it as an appendix to his Au carrefour du Soudan et de la Berberie in 1989. ^{Footnote 57} Hamani provides the classification references for the Arabic manuscripts that he identified as corresponding to the published translations of the Agadez Chronicles held in the archives of the Institut de Recherches en Sciences Humaines (IRSH) of Niamey, alongside information about each of the manuscripts.^{Footnote 58} In Volume II of their bio-bibliographical reference work on the Arabic literature of Africa, Hunwick and O’Fahey include bio-bibliographical information on Aïr and Ader. Hunwick (who compiled Volume II) follows Hamani and adds information on copies held outside Niger (mainly in Nigeria) and on published translations.^{Footnote 59}

The published translations include precious information on how the translated manuscripts were obtained. Coupled to a comparison of differences across these published versions, this information sheds light on the perceptions of the past of those who passed on copies of manuscripts to French administrators. This knowledge constitutes a fundamental “interpretative key” for analysing the different versions of manuscripts that reached us.

Palmer published two sets of texts, which he believed to be derived from the same source. The first one, entitled “Notes on Some Absen Records” and published in 1910, is his translation “of some notes in Arabic compiled by a Hausa Mallam [whose] information was derived from Tuareg sources – probably in part from MSS.”^{Footnote 60} It is not clear in what form Palmer accessed these “records,” whether he translated written texts handed over to him, or transcribed and then translated an oral account. Palmer says that he “lays no claim to be translating an ancient MS. The ‘Records’ rank as ‘oral testimony.’ This, however does not imply that they are not reliable, for the Sudanese memory is usually more accurate and reliable than Sudanese copyists.”^{Footnote 61} This suggests that Palmer accessed some Arabic notes written on paper by a Hausa mallam who drew upon an orally transmitted repertoire of historical knowledge and probably some Arabic manuscripts (possibly obtained from Tuareg sources) in his possession.

Palmer divided the Mallam’s account in three sections numbered I, II, and III. Section I corresponds roughly to Urvoy’s MS B, Palmer’s text LVII in Sudanese Memoirs (1928), and the initial paragraphs of Tardivet’s translation. Section III is a chronological list of sultans of Agadez, and corresponds roughly to Urvoy’s MS A, Palmer’s text LVIII (1928), and the bulk of Tardivet’s translation. Section II is very short (one paragraph) and corresponds to the sub-section of Urvoy’s MS B entitled “Origine des Sandals.”

In 1928 Palmer published a large compendium of translations of Arabic manuscripts intermixed with historical information, under the title Sudanese Memoirs. The third volume of this work includes an English translation of the Chronicles of the Sultanate of Aïr, based on an Arabic manuscript, Palmer noted, that had been “sent by the ruling family of Agadez to the present Sultan of Sokoto, and by him given to the Resident, Mr. G.W. Webster, MBE.”^{Footnote 62} He arranged this material in two separate texts numbered LVII and LVIII, which correspond roughly to Urvoy’s MS B and MS A, respectively.

R. Tardivet provided another list of the sultans of Aïr, published in 1928. He consulted a set of Arabic manuscripts that belonged to “mon ami, le Sultan de l’Aïr, Oumarou”^{Footnote 63} (son of Sultan Ibrahim Ed Dasouqy) and had one manuscript translated by the district interpreter, a man from Agadez called Adam, and re-translated by the colonial interpreter Kouradou Kamara.^{Footnote 64} This text mixes information similar to that provided in Urvoy’s MSS A and B and in Palmer’s two publications.

The longest and most detailed publication is Yves Urvoy’s article “Les Chroniques d’Agadès,” published in 1934. Urvoy translated ten manuscripts, designated A to J and preceded by an introduction in which he summarized the history of the region, and provided some information on the manuscripts and his translation thereof. Text A is a “list of the sultans of Agadez” from Yunus (1405) to Oumarou (1923).^{Footnote 65} Text B describes the respective origins of the Sultanate of Aïr, the Sandal Tuareg, and the Sultanate of Ader. Text C contains an account of the origin of the Sultanate of Bornu. Text D is a list of the sultans of Gobir. Text E provides a concise chronology of events in the Aïr region between 1683 and 1771. Text F continues this chronological history from 1788 to 1890. Text G is a short list of the female apical ancestresses of the different tribes of the Kel Ewey Tuareg. Text H is another chronological list of events occurring in the Aïr region from 1683 to 1829, continued in Manuscript I from 1816 to 1869. Finally, Text J is the “Diary of Abubakar, son of Attaher-Tachi,” which presents in an autobiographical style a mix of political, environmental, and personal events that occurred in the lifetime of the author between his birth in 1657 and 1699. I could not identify any data on the history of this manuscript.^{Footnote 66}

H.T. Norris provided an English translation of “the diary of Abu Bakr b. Al-Tahir Tashi.” Norris consulted a copy in the possession of the Sultan of Agadez, which, he noted, did “not seem to be identical with the copy previously translated [i.e. Urvoy’s copy], here and there incorrect dates are given. Yet, there is no mistaking its originality. An exciting period in the history of Agadez is disclosed.”^{Footnote 67} Regrettably, however, in spite of – or perhaps due to – his enthusiasm, Norris does not tell us why he chose to characterize this text as “unmistakably original,” nor on which basis some “dates” (which dates?) in one or the other copy (which one?) should be deemed “incorrect.” According to Abdullahi Augi, “the chronicle pertaining to the memoirs of Muhammad Attahit [sic] Tashi (b. 1657), which is called manuscript J in Urvoy’s translation, was certainly compiled by the author, i.e. in the seventeenth century.” Like Norris, however, Augi omits to tell us whence he derives such certainty.^{Footnote 68}

In contrast to the better-known Agadez Chronicles, the Y Tarichi is cited exclusively in the specialist historiography of Ader. It is a text copied and part-written in 1907 (possibly on the basis of older manuscripts), obtained by Captain Laforgue in his second trip to Agadez in September–October 1907. Captain Laforgue transmitted it to Lieutenant Peignol, who was stationed in Ader near the Lissawan canton chief.^{Footnote 69} The Y Tarichi illustrates the role played by the “supporting tribes,” in particular the Lissawan, in the seventeenth-century politics of Aïr and Ader. According to Lieutenant André Maurice Peignol, who included a typed French translation of the Y Tarichi as an appendix to his Monograph of the Cercle of Tahoua, Laforgue:

In his 1947 study of the Lissawan, the administrator Souchet further explains that the Y Tarichi consists of copies of an Arabic text composed in Agadez by Malam Ahmadou Ben Detchoukou, cleric at the service of the Sultan Ibrahim Ed Dasouqy, and copied by Malam Djibrina, cleric at the service of the Lissawan chief Ennour.^{Footnote 71} Peignol’s monograph and the Y Tarichi in its appendix are both typewritten and now kept in the National Archives of Niger in Niamey. The Y Tarichi describes events leading to the installation of a sultan in Agadez and the creation of a cadet “sultanate” in Ader. Some of the information provided in the Y Tarichi on the tribes supporting the sultan coincides with sections of Manuscript B of Urvoy’s “Chroniques” (which Urvoy considers “very ancient”) and sections of Palmer’s text LVII in Sudanese Memoirs. The Y Tarichi expands substantially the information provided in the section entitled “Origins of the Ader Sultanate” in Urvoy’s MS B. But, unlike the latter, it does not tell Ader’s history from the perspective of Birnin Ader and Illela, where the emir of Ader was installed up until colonial rule. It narrates Ader’s pre-colonial history following Agabba’s conquest from a Lissawan-centred perspective that emphasizes relations between the Lissawan, Kel Gress, Iwellemmeden, and Sokoto.

The table below summarizes schematically the themes covered by each of these five sources, showing areas of broad thematic overlap. Because Urvoy’s article is the longest and most detailed, I have placed it first to show thematic correspondences with the other texts. I only included in this table those themes in Urvoy’s texts A and B that are the most relevant to this article’s focus.

The thematic overlap between various versions of the Agadez Chronicles and Y Tarichi is not precise. Although the same theme may be covered in multiple texts, the names or numbers of tribes mentioned or the details of particular events often differ across versions. The question addressed here is not whether it might be possible to establish which version is closest to a hypothetical “original” or “true” source, for there is no such thing. The question is what explains the differences between versions.

Let us examine four main orders of difference that can be grouped as follows: (1) acknowledgement of sources; (2) number and names of Tuareg groups; (3) identity of the first sultan; (4) the Sultanate of Ader.

Production and Circulation of the Manuscripts: the Positions of the Sultans of Agadez and Sokoto

To make sense of the differences across versions of these texts and attempt to establish which versions are most reliable for the purpose of historical reconstruction, it is necessary to consider how the copies of the manuscripts that were eventually published had been circulated in the first decade of the twentieth century.

Urvoy obtained his version of the “Chroniques d’Agadès” from Sultan Oumarou of Agadez, who according to him provided a “very clean copy” made by clerics at his court. We know that the documents held by Sultan Oumarou had been copied and updated in 1907 under Sultan Ibrahim’s power:

Urvoy goes on to say that the section entitled “Origin of the Sultanate of Ader” was redacted (rédigée) in 1907.^{Footnote 101} In this same year Lieutenant Peignol obtained the Y Tarichi on the history of the Lissawan, the Tuareg group which in 1900 was placed at the head of the newly created district of Tamaske (later renamed Keita) in north-eastern Ader. As has been mentioned, the two texts on the Sultanate of Aïr (numbered LVII and LVIII, respectively) published in Palmer’s Sudanese Memoirs “are translated from two manuscripts sent by the ruling family of Agadez to the present Sultan of Sokoto, and by him given to the Resident, Mr. G.W. Webster, MBE.”^{Footnote 102} The Sultan of Agadez who transmitted the documents to the Sultan of Sokoto was almost certainly Sultan Ibrahim. Palmer’s text LVII, coming from the same source as the texts translated in Urvoy’s “Chroniques d’Agadès,” refers to both sets of tribes and cites the same names as those found in Urvoy’s copy.

Urvoy did not attempt to clarify the possibly common derivation of his Manuscripts A and B and texts LVII and LVIII published by Palmer in Sudanese Memoirs. He stated simply that Manuscript A is “surely authentic and at all times multiple copies of it must have existed. The sultans had and have the habit to share it easily in order to demonstrate the antiquity of their lineage. In particular, some copies of it existed amongst the southern Hausa sultans. One, which was in Sokoto, has already been translated in English by M. Palmer.”^{Footnote 103} A few paragraphs below this, Urvoy reiterated that “Manuscript A and the first part of Manuscript B are already known. A translation was published in 1909 in the Journal of African Society based on a copy found in Katsina [this is Palmer’s “Notes on Some Asben Records”]. A second, better one [was published] by Sir H.R. Palmer based on copies held by the Emir of Sokoto and published in the author’s Sudanese Memoirs.”^{Footnote 104} Urvoy then compared Palmer’s 1928 text and his Manuscript A and highlighted a few minor differences (mainly in the dates), explaining where he had adopted variants of names from Palmer’s version and where he had followed his copy of Manuscript A.^{Footnote 105}

Urvoy seems to have imagined that the Sokoto version of the manuscript had circulated in Sokoto for an indeterminate number of years, possibly decades, and that it was one in a set of copies that the Agadez sultans recurrently circulated like pamphlets for propagandistic purposes. But Palmer’s information, while very general, suggests a more focused process, through which this particular manuscript had been “sent by the ruling family of Agadez to the present Sultan of Sokoto, and by him given to the Resident, Mr. G.W. Webster, M.B.E.”^{Footnote 106} Palmer’s word “sent” suggests that this manuscript had been dispatched from Agadez by a sultan residing there. However, in 1907 Sultan Ibrahim Ed Dasouqy was in fact residing in Sokoto, having been driven out of his sultanate by Mohamed El Baqary, who was reclaiming a position usurped just a few years earlier by Ibrahim Ed Dasouqy. Following these turbulent events and having lost the sultanate to his old foe, Ibrahim Ed Dasouqy sought refuge in Sokoto. This is where the French administration found him in August 1907 and brought him back to Agadez to reign as Sultan.^{Footnote 107} Shortly afterwards, Ibrahim Ed Dasouqy was deposed by the French and replaced with the son of Mohamed El Baqary, Abd-Er-Rahim Tagama – the Sultan Tagama who participated in the anti-colonial revolt of 1916–1917.^{Footnote 108}

After the brutal suppression of the revolt, Ibrahim Ed Dasouqy was reinstated as Sultan. In 1918 Sultan Ibrahim publicly summoned the people of Agadez and forbade them to provide information to Europeans. He “systematically forbade all natives to visit the French Post without passing through him and receiving orders on what they should say or do.”^{Footnote 109} The Sultan was arrested on 10 August 1919 for “disloyal behaviour.” He was provisionally liberated from prison but deposed in November 1919 for “incapacity” and replaced with his son Oumarou, who reigned until 1960.

It is possible that in 1907 Ibrahim Ed Dasouqy could have handed over a copy of some of the Agadez manuscripts (A and B) to the Sultan of Sokoto, Muhammadu Attahiru II, in Sokoto: these manuscripts were later made available to the Resident Mr. Webster and were eventually published by Palmer in Sudanese Memoirs. Also in 1907 Ibrahim Ed Dasouqy left Sokoto and returned to Agadez, where copies of the manuscripts were made available to the French in that same year. These manuscripts would be published by Urvoy as “Agadez Chronicles.” The second set of tribes appears in both these versions of the Agadez Chronicles (and not in other versions). The second set of tribes also appears, with the same names, in the Y Tarichi and the “Origin of the Sultanate of Ader” (in Urvoy’s Manuscript B), both created and circulated in 1907. In short, the second set of tribes only appears in the “1907 versions” of the Aïr and Ader chronicles, which were – as mentioned above – “copied and, when appropriate, updated in 1907 during the reign of Sultan Ibrahim-ed-Dasouqy.”^{Footnote 110}

Palmer’s mention of “the present Sultan of Sokoto” in his 1928 publication is confusing. Sokoto’s Sultan in 1928 would have been Muhammadu Tambari (1924–1931).^{Footnote 111} Sultan Muhammadu Tambari was deposed by the British in 1931 and forced to abdicate. He absconded from house arrest on 6 February 1931 and moved to the French Colony of Niger, where he remained until 1932.^{Footnote 112} The Sultan of Agadez at the time of Muhammadu Tambari’s stay in Niger was Oumarou, son of Ibrahim Ed Dasouqy. Hence, the copy of Manuscript A translated by Palmer could either have been given to Sultan Muhammadu Attahiru II by Sultan Ibrahim Ed Dasouqy during his sojourn in Sokoto in 1907, or else have been sent to Sokoto’s Sultan Muhammadu Tambari by Agadez’s Sultan Oumarou, son of Ibrahim, before 1928. Muhammadu Tambari’s stay in Niger in 1931–1932 suggests close relations with Niger’s Islamic elites of the time: the Sultan of Agadez and, further south, the Lissawan chiefs in Ader and the paramount chiefs of the Itesan and Kel Gress Tuareg in the Gobir area.

Sultan Muhammadu Attahiru II had witnessed the British occupation and faced a Mahdist revolt in 1906. He collaborated with the British in the revolt’s suppression.^{Footnote 113} Both he and his “guest” – Sultan Ibrahim Ed Dasouqy – were well aware of the military strength of the European invaders and had developed strategies to appear collaborative while possibly harboring resentment toward the European administration.^{Footnote 114} It is not surprising that the Sultans of Aïr and Sokoto harbored ambivalent feelings toward Christian invaders, feelings that would not have changed one generation later. Their relationship with colonial authorities was complex, as they knew they owed their positions (and precarious safety) primarily to European support. In northern Nigeria, both Sultan Muhammadu Attahiru II and Sultan Muhammadu Tambari had been appointed by the British administrators, in disregard of the Fulani electorate’s views.^{Footnote 115} In Agadez, Sultan Ibrahim Ed Dasouqy and Sultan Oumarou faced similar predicaments. Like most native chiefs who had witnessed the fate of those who dared to resist European rule, they were anxious to convince colonial authorities of the legitimacy of their power and the sincerity of their loyalty to colonial government. Whatever their personal feelings toward the French invaders, they seem to have convinced the French administrators that it was in France’s interest to support them. Lieutenant Camille-Charles Jean, who was in Aïr in 1903–1904, argued in his study of the Tuareg of Aïr, completed in May 1907, that “the role of the Sultan of Aïr should be enhanced by us, his relations extended so that he could act as our agent and feel obliged to us.”^{Footnote 116} It is surprising how little consideration post-colonial historians have given to these circumstances, which are the context in which the collection of manuscripts took place.

The text published by Palmer as “Notes on Some Asben Records” does not coincide with the “view from Agadez,” which was actively publicized by Sultan Ibrahim of Agadez in 1907 through the distribution of copies of the Chronicles to French and (via Sokoto) English colonial administrators. In his introduction to “Notes on Some Asben Records,” written before the version provided by Sultan Ibrahim became available, Palmer noted that “the jealousy with which such information [historical information in manuscripts on Tuareg history] is released by the Tuareg, and to some extent by the Sudanese, is well known. With regard to Asben, a recent French writer, Lieutenant Jean, states that he was informed that all records had been burnt some years ago,”^{Footnote 117} that is, before the 1907 copies were provided through official channels.

Had manuscripts been destroyed to keep them from falling into the hands of European occupants? Or was the “fire” an initial excuse to avoid having to comply with French demands and to buy time during which extant Arabic manuscripts could be copied, and appropriate alterations introduced in the copies? As for the Y Tarichi, Urvoy tells us that the original manuscript written by “Mallam-ben-Detchoukou, marabout du sultan d’Agades” had “disappeared” (disparu) in a Tuareg raid against Keita of 1917.^{Footnote 118} No attention was paid to the curious circumstances in which manuscripts were first denied, and then circulated. Nor to the fact that the internal consistency that ran through the manuscripts circulated from Agadez was not supported by other available sources on the history of Agadez and Ader.

Colonial and post-colonial scholars lent unwarranted support to the versions of the Agadez Chronicles published by Urvoy in 1934 and the corresponding texts published by Palmer in Sudanese Memoirs, both of which mention two – not one – groups of Tuareg tribes; and both of which were congruent with the political arrangements that existed in the first decade of the 1900s. Yet such congruence should have triggered caution rather than confidence. The set of names contained in Urvoy’s Chronicles is likely to be the most unstable element of the traditions pertaining to Aïr and Ader’s pre-nineteenth-century history. Colonial researchers expressed awareness of the potential bias of these texts. They knew that these texts had first been denied, then made available by rulers anxious to prove their chiefly credentials. But they chose to ignore these facts. Doubts were raised only to be cast aside, and sources declared problematic continued to be used as foundations for doubtful reconstructions.

The Inclusion of the Second Set of Tribes in the 1907 Versions: the Position of the Lissawan

The absence of the second set of tribes from sources that became available before the French occupation of the Central Sudan suggests that these names might have been added to Palmer’s text LVII and Urvoy’s text B shortly before they reached European hands, together with other documents gathered in the same period and illustrating the history of the Lissawan and their allies and dependants (the Y Tarichi). At the arrival of the French the Lissawan were given the role of traditional chiefs of the Canton of Tamaske. Later the Tamaske area was rendered independent from the area of Keita, and the latter section of the original “Group of Tamaske” was renamed “Keita” and remained under Lissawan control to this day. The Agadez Chronicles legitimize Lissawan power in this region by linking it to the history of the foundation of the Sultanate of Agadez.

In the first years of the twentieth century, this version of history appealed to French colonialists, Agadez, and the Lissawan. I have not been able, yet, to establish what may have been the relationship between Ibrahim Ed Dasouqy and Amattaza Ennour in 1907. But it is clear that both chiefs sought to gain French confidence in order to avoid, for them and their people, colonial repression and violence.^{Footnote 119} The Chronicles strengthened the credentials of the Agadez Sultan and of the Lissawan, who did not pose a threat to French interests, and omitted reference to the Tuareg confederations of the Iwellemmeden Kel Denneg, who had ruled the region of Tamaske and Keita at the time of the French occupation and had mounted a strenuous resistance against France.

Amattaza Ennour allied himself with the French in 1900. As a result of this alliance, his power was considerably strengthened. The Lissawan were the first local chiefly group to sign a peace treaty with the French. As shown in the quotation below, this earned them the resentment of the chiefs of the Iwellemmeden Kel Denneg, first and foremost of Makhammad, paramount chief (amenokal) of the Kel Denneg, who refused to collaborate with France and conspired against French power until 1917:

The French administration was in a highly precarious position in Ader in the first two decades of the twentieth century. Up until the bloody repression of 1917 they were unable to constrain the activities of the Kel Denneg resistance, which pillaged and raided continuously villages that used to fall in their area of power before French occupation. In this context, the French saw the Lissawan’s collaboration as “the most important [submission] of all, that of the Amattaza Ennour, the sole hereditary chief of the region who, thanks to his authority and energy, was going to be a serious support to us.”^{Footnote 121}

On 2 November 1901 the Convention of Tamaske, also known as “the Peace of Tamaske,” was promulgated publicly.^{Footnote 122} Lieutenant Colonel Péroz, who commanded the French Third Military Territory, explained at a public meeting the norms regulating relations between Ader’s Tuareg and Hausa groups and the French government. Most chiefs attended the meeting, except for Makhammad and the Iwellemeden. The privileged position of Amattaza was forcefully asserted. At the beginning of this speech, Péroz saluted “Amattaza, chief of the Lissawan: I gave him and all his children command of the country and the Asna villages of Tamaske, because he is the only Tuareg chief who came to us and understood that we brought the justice and peace of God.”^{Footnote 123} Péroz went on to greet the other chiefs, including Ouarzagane Tambari of the Kel Gress and the Asna. Then he spoke to Makhammad, who was not there to hear his warnings, and reminded him of the necessity of his submission and the requirement that he pay taxes to the French authorities and obey their orders.

In his early correspondence with French administrators, Makhammad attempted to undermine Amattaza’s position and influence:

By 1907 Makhammad had died, and his succession was disputed between Ismaghil and Rhezi. In January 1907 the chiefs of various Iwellemmeden sections were summoned to the office of the Commandant de Cercle in front of the Amattaza of Tamaske^{Footnote 126} for the public announcement of Decision No. 122, which was an attempt to settle once and for all tensions within the Kel Denneg, and support the feeble authority of Ismaghil, the French-designated chief. Until 1908 the Lissawan Amattaza ruled over a large region with the town of Tamaske as its main centre. But in 1908 this region, known as “Group of Tamaske,” was divided into three separate cantons: Tamaske, Laba, and Keita. Only the canton of Keita remained under Lissawan power, while Tamaske and Laba selected their own chiefs, who were seen as representatives of the Hausaphone Asna.^{Footnote 127}

In 1907 Amattaza Ennour must have felt vulnerable, faced with the defection of Hausaphone groups that initially had supported him and were now seeking to achieve independence from him, and under the ongoing threat of Kel Denneg resentment that concretized tragically ten years later, during the 1917 revolt. On 14 February 1917 Keita was sacked and the Lissawan canton chief Afedandan was assassinated “by the men of Kaocen, Manit and Rhezi,”^{Footnote 128} some of whom were the last representatives of Kel Denneg resistance in Ader (including the same Rhezi mentioned in the paragraph above). These events prove unequivocally that Lissawan anxieties were well-founded: anxiety would have been intense in 1907, just before the partition of the canton that had been given to them in 1900. And this would have pushed the Lissawan, alongside other chiefs, to “update” the historical records so as to strengthen their position in the eyes of the French.

The Chronicles that were circulated in 1907 endorsed both the Lissawan’s claims to power and French support to the Lissawan. No one raised questions. As we have seen, French researchers knew – or suspected – that the texts of the Agadez Chronicles and Y Tarichi had been circulated with the aim of advancing a specific agenda. Yet, if the French administration had good political reasons not to question the version of history contained in the documents they received from the chiefs they supported, it is perhaps more surprising that recent scholars have not been more inquisitive.

It seems likely that the main function of the Agadez Chronicles, whenever these texts were composed orally and put into writing, was to explain how political transformation from a less to a more centralized structure of rule came about. Subsequent revisions, such as the introduction of the specific names cited in the copies updated in 1907 and given to colonial administrators in the first decade of the twentieth century, accommodated the political concerns of the times.