and conflation that sustained colonial confabulations about the race, culture, civilization, traditions, and religion of the natives.3 These, rather than the activities of the colonizers, were used to explain the wretchedness of the North Africans.4
One of the consequences of the reification of North Africa was its universal appearance in scholarship as an acceptable category that supported scholarly generalizations. The situation continues today. One need only look for the prevalence of references to “Roman” or “ancient” North Africa which, more than the anachronistic proclivities of some historians, demonstrate the adaptation and accommodation of colonial knowledge to post- or neocolonial conditions.5 What is at stake here is not the historians’ use of a colonial category, but rather the fact that in following this convention they risk designing their research in ways that incorporate ideas they might not necessarily agree with.
The refitting of the terms of colonial knowledge, which has been ongoing since the 1960s, has toned down what scholars now consider derogatory statements about the natives. The way to convey respect for the natives has been to replace North Africa with a native category, the Maghrib, though doing so merely maintains the pretense of the colonial body of knowledge.6 As it happens, the Maghrib too is an ideological construct produced and reproduced in the process of supporting the domination of a number of pre-modern elites.7 In other words, like modern North Africa, the medieval term Maghrib was a construct that served amalgamation and conflation, not elucidation and clarification. However, unlike North Africa, it has been in use for centuries, incorporated into a number of discourses, not always for the same reasons or with the same consequences. Consequently, using the Maghrib, or one of its subregions, as a container within which to fit history does more than support colonial discourse. It gives credence to the dominant discourses of the successive medieval dynasties that utilized the category to prop up their rule. Neither option is particularly attractive.
This book proposes an alternative perspective focusing on Ifrīqiyā. It examines how intellectuals associated with the Ḥafṣid dynasty (1229–1574) represented the dynasty’s rule over specific groups and areas as its rule “in” or “of” Ifrīqiyā. It analyzes the ways their writings have served as the empirical basis upon which modern historians have transformed Ifrīqiyā into the easternmost region of the Maghrib. The book argues that the perspective of these texts gains significance in relation to political struggles that led to the elimination of the political autonomy of urban emirates in Ifrīqiyā and the imposition of rule from Tunis at the end of the fourteenth century. Intellectuals of the era described the ruler (emir) of Tunis as ruling over all of Ifrīqiyā because he had eliminated what they saw as political fragmentation—not because he actually controlled the entire “territory” of Ifrīqiyā or all those who inhabited it. In order to establish this, the book will bring into focus the evolution of the city of Bijāya, the main city under Ḥafṣid rule, beside Tunis. It will show that Bijāya was not always under the rule of Tunis or the Ḥafṣids, that it was not dependent on its integration into an Ifrīqiyā-wide economy, and that the range of activities of its elite demonstrate the Tunis-centric bias of the sources.
Arguing that contemporary writers partook of an ideology that became increasingly dominant after the end of the fourteenth century requires a careful examination of the evidence. It involves using the descriptions of sociohistorical processes found in the texts to demonstrate that claims about the extent of the emir’s influence were politically motivated embellishments. The conceptual tools with which scholars have understood such processes make the articulation of this argument a complex task. While the notion of region figures prominently because of the specific claims of Ḥafṣid intellectuals, it is not the only issue this study tackles. Scholarly ideas about regions and regionalization are part of a broader discourse sustained by a nexus of related conceptions and approaches. A discussion of some of these will shed light on the scope of the challenge ahead.
How Ifrīqiyā Was Made
Maghrib specialists have not analyzed the evolution of the category Ifrīqiyā over time in relation to sociohistorical and intellectual changes. Instead, they have tried to define it, or ascertain the most reliable description of it, by cataloguing references in Arabic texts written over seven or eight centuries, from al-Andalus to Baghdad. This approach collapses chronology and ignores the conditions within which various authors produced those references. Unsurprisingly, it has also yielded conflicting bits of information. This is what the historian Mohamed Talbi discovered: “The details given by the various Arabo-Muslim geographers and historians do not always agree,” he wrote. Apparently, these authors did not precisely understand “the exact frontiers of Ifrīḳiya.” Talbi’s challenge was therefore to explain their confusion.
It may be said that the geographical Ifrīḳiya consisted essentially of the ancient (Numidia) Proconsularis and Byzacena, which formed the nucleus of it, to which were later added Tripolitania, the Numidia of the Aurès, and even a part of Sitifian Numidia. Upon this geographical concept was superimposed an administrative concept. Because of this, Ifrīḳiya tended to be confused, in the writings of the chroniclers, with the territory that, in the Middle Ages, was ruled in turn from Qayrawān, from Mahdiyya or from Tunis—a territory which expanded or contracted according to the vicissitudes of history.8
In order to eliminate this confusion, Talbi proposed that late antique Roman categories be used instead of medieval ones.9 This unsatisfactory position rests on the unexplained and not obvious distinction between a late antique “geographic” conception and a medieval “administrative” one, the former assumed to be more accurate than the latter. In addition to fostering anachronism, this interpretation implies that geographical notions develop independently of politics, an idea that deserves, if not demonstration, at least explanation.
Moreover, if in the medieval period Ifrīqiyā was a territory that expanded and contracted, then medieval authors were not necessarily confused when they disagreed about its frontiers. For instance, the “divergence” over whether the city of Bijāya belonged to western Ifrīqiyā or was the capital of the central Maghrib was not necessarily due to confusion. The city only became important in the second half of the eleventh century when it became the capital of the Ḥammādids (1015–1152). As such, it became the capital of the central Maghrib, a region associated with that dynasty. This explains why the great geographer al-Idrīsī (d. 1166) considered Bijāya to belong to the central Maghrib.10 Later, when the Ḥafṣids claimed Bijāya, it was reasonable for the Mamlūk official and author Ibn Faḍl Allāh al-‘Umarī (d. 1349) to believe it belonged to Ifrīqiyā.11 But when the Bijāyan judge al-Ghubrīnī (d. 1304) wrote of “our central Maghrib” including Bijāya in it, he was expressing thoughts supportive of the autonomy of the Ḥafṣid emirate of Bijāya from Tunis and its standing as the capital of the central Maghrib.12 Deciding where the borders of Ifrīqiyā were had political implications and learned contemporaries seem to have understood that.
Recently, Dominique Valérian has suggested that, “in spite of the vagueness of the sources, [historians] must preserve the political and administrative sense that [Ifrīqiyā] had in the early years of the Islamic Maghrib and that it retained in many later texts,” namely, that “Bijāya actually belonged to Ifrīqiyā, which experienced a real political unity in the Ḥafṣid period.”13 Valérian did not explain why he believed that historians should preserve earlier notions or why the fact that later authors reproduced earlier notions made those notions more valid. Nevertheless, he gave weight to political considerations in deciding the frontiers of Ifrīqiyā. Although he also noted the importance of narrative practices in producing what he described as a blurry and vague conception of Ifrīqiyā, he argued that the political unification imposed by the Ḥafṣids warranted that Bijāya should be included in it. The problem is that Ifrīqiyā was not always politically unified and was certainly not always under the control of the Ḥafṣids. Why the Ḥafṣid period should become the standard rather than the Ḥammādid period is also not self-evident.
In spite of these difficulties, Valérian’s original insight connecting region and politics is critical because it enables the notion of “Ifrīqiyā” to become the object of historical analysis rather than to be taken for granted. Instead of using