The entourage and highest-ranking members of a ruler’s court were known as his khāṣṣa. The literal meaning of khāṣṣ is private and particular. Its negative double is ‘ām which means general and common. The khāṣṣa constituted the core of the ruling elite and played an active role in politics, performing crucial functions such as determining heirs-apparent and taking the oath of allegiance. They were pillars of the dawla. Members of the khāṣṣa of Ḥafṣid emirs were not necessarily the same as his chief officials—although that was generally the case. They belonged to a group that was privy to the ruler’s secrets such as the location of his treasure, the terms of negotiations and agreements with enemies and allies, and his military plans. The composition of the khāṣṣa of a particular emir was subject to change. It depended on solidarity between members and the loyalty they demonstrated their leader. It is important to note in this regard that although most of those who were part of a ruler’s khāṣṣa were members of the upper classes of society, not everyone who belonged to the social elite was automatically a member of the khāṣṣa.
Whereas the term khāṣṣa described a specific group of individuals with ties to the ruler and each other, ‘āmma (also kāffa) was a generic term that lumped together a number of very distinct groups, from notable scholars and wealthy merchants, to craftsmen and petty officials. Members of the ‘āmma were subjects of the ruler who were not part of his khāṣṣa, and thus not part of the dawla. In fact, the term does not correspond to any actual sociopolitical group and has very little descriptive or analytical value. It simply distinguished those who were part of the ruler’s circle from everyone else. Revolts against the ruler that were not led by recognized political factions of elites were automatically attributed to the ‘āmma as a whole, even if it was clearly a mere portion of the ‘āmma that was involved. Interestingly, Bedouin “tribes” belonged neither with the ‘āmma nor the khāṣṣa, even when they were allied or tributary. They stood as separate and distinct entities. The urban emir may have secured the allegiance of individual Bedouin leaders, but the latter did not become members of the khāṣṣa, and their fellow tribesmen were not part of the ‘āmma.
The political imaginary suggested by the concepts of khāṣṣa and ‘āmma poses the question of the bias of the sources in a way that goes beyond the political activities of individual authors. Tying political conceptions to the politics of regionalization allows us to delineate the contours of political discourse in Ifrīqiyā and thus move beyond the issue of the objectivity, reliability, and representativeness of individual sources and statements.
The Learned and the Evidence
The category “intellectual” is a modern one and using it to describe educated men and women in medieval Ifrīqiyā can be anachronistic.42 One of the problems associated with using the concept is that historians might project onto the past the modern idea that intellectual activity is socially differentiated or autonomous. Yet, the idea that it is easy to situate intellectual activity, and thus intellectuals, in modern societies is itself problematic. Clearly, identifying some individuals as intellectuals and excluding others is itself a political act that, some might say, serves to legitimize the prevailing social order.
In medieval Ifrīqiyā, the notions of knowledge or science (‘ilm) and of the learned man who practiced it ‘ālim (pl. ‘ulamā’) had similar effects. They delineated the form and content of acceptable thought and served to legitimize a social order, especially by establishing a hierarchy of knowledge. The ulama’s writings not only constitute most of the historical archive, they present the historian with a particular ordering of it.
Moreover, medieval ulama did not always agree on who could be considered an ‘ālim or the criteria that identified one as such. When the Chief Judge of Tunis, Ibn ‘Arafa (d. 1401), heard that the historian Ibn Khaldūn held a position as judge in Cairo, he scoffed. For him, Ibn Khaldūn was not a ‘ālim because his knowledge of Mālikī jurisprudence was too poor to qualify him as one. The modern historian’s decision to identify some individuals as ulama risks taking sides and reproducing the judgments of medieval authors. One must somehow find a way to acknowledge Ibn Khaldūn’s involvement at the Ḥafṣid court, his authorship of an important historical narrative, and the opposition he inspired in some of Ifrīqiyā’s most prominent learned men.
As if this were not enough, the activities of the ulama, and the established forms of knowledge that they maintained, systematically marginalized the ideas of many in their societies—so many, in fact, that historians have a hard time recovering the contingency of the social hierarchy that they represented in ways that made them seem part of the natural order. For the ulama, an individual whose ideas rallied hundreds and opposed the power of elites by organizing armed resistance was not a thinking individual. Their habit of denigrating this type of person may seem perfectly justified in retrospect since often they were barely literate. But for modern historians, the actions of these illiterati are quite relevant. Without them, the particular institutional arrangements that prevailed in Ifrīqiyā may appear as having descended from the heavens.
We have a choice: we can use the nomenclature of the medieval ulama and thus reproduce their biases, or we can use modern terms such as “intellectual.” Though the latter path equally risks bias, I have chosen it. In this book, I use the term intellectual to apply universally to all those who articulated political ideas, and not merely to those designated as ulama in the sources. Doing so has the benefit of recalling the political partiality of the historical evidence and its bias against certain social groups. In the absence of neutral terms, medieval or modern, it might have been helpful to use a completely different concept or to invent one to serve my purposes. My sense is that doing so would not be necessarily more fruitful. Instead, “misusing” the notion of intellectual in the particular sense I propose maintains an awareness of the conceptual work involved when historians take into account medieval and modern ideologies simultaneously.
The Mediterranean
The Mediterranean encapsulates a set of connected historical problems related to the categories that frame research and the character of the archive. Following Peregrine Horden and Nicholas Purcell’s distinction between “history in the Mediterranean” and “history of the Mediterranean,” this study of Ifrīqiyā is primarily a study of history in the Mediterranean.43 Its argument about the multiplicity of economic arrangements and the integration of some, but not all of them, agrees with Horden and Purcell’s view of the prevalence of microregions in the Mediterranean.
Map 3. The western Mediterranean.
Whether their work is a history in or of the Mediterranean, however, Mediterraneanists have tended to reinforce the idea that North Africa was a single region by framing their research in terms of relations between a European city, kingdom, or region, and all of North Africa or the Maghrib.44 This routine act of lumping and generalizing has weakened the degree of specificity and nuance that some Mediterraneanists want to enshrine. By challenging the habit of assuming that Ifrīqiyā, and by extension North Africa or the Maghrib, should be understood as regions, this book hopes to contribute to the development of new habits based on the ongoing rethinking of historical knowledge about the Mediterranean.
Another common scholarly shorthand that requires some attention is the “Islamic Mediterranean” which, as far as one is able to tell, stands for the “lands of Islam” that have a Mediterranean coast.45 Although it is rarely meant as a serious analytical category, this epithet lends credence to the practice of understanding history through the prism of religion. Merely stating that some societies were Islamic does little to illuminate why understanding them as such improves our interpretation of their histories. Unfortunately, this practice is as common as it is problematic. It is thus necessary to resist the sirens that call for a greater inclusion of the “Islamic Mediterranean” into the Mediterranean field—at least under this rubric.
Al-Andalus