This book is based on seventeen months' fieldwork on Jabal Razih in the far north of the Republic of Yemen, where I spent from March to October 1977, October 1979 to May 1980, and January and February 1993. The "ethnographic present" therefore refers to that period. Throughout my fieldwork I lived in the tiny "town" or madinah of al-Nazir, the main settlement of a tribe in southern Razih. During my first two stays I maintained a semi-independent household; and on my third I lived en famille in the house next door. The madinah was an ideal fieldwork base. It is socially and occupationally heterogeneous. Men prominent in tribal and government affairs were among my neighbors. And visitors came from far and wide to its lively weekly market. I did most of my research in and on the tribe of al-Nazir, one of the ten small tribes of Razih, partly because I wanted to understand one community well, and partly because travel was difficult in the mountainous terrain. There were no motor tracks during my first fieldwork in 1977, and although the construction of a trans-Razih track and feeder tracks had begun by 1979, travel remained slow and most places could still only be reached on foot. Although I traveled widely in Razih and visited the neighboring tribal region of 'Uqarib, therefore, my deepest firsthand knowledge is about the tribe of al-Nazir, and I describe the tribal system of Razih mainly from that perspective.
No-one in Razih spoke English, so all my fieldwork was conducted in Arabic. The local dialect, or language, is extremely unusual, and was always a difficulty, but some male informants could switch registers to a form of Arabic I could understand more easily. My linguistic struggles were greatly helped by Bonnie Glover Stalls, a specialist in Arabic dialects, who accompanied me to al-Nazir for two months in 1979 to study spoken Razih, and has continued to provide helpful advice. I was also aided in the field by Cynthia Myntti, who came for a month in 1977 to do a health study, and by Ian Dunn and Michael Dunn, who visited for a month in 1977 to make a film. Ian also kindly helped me map the madinah and tribal territories during a second short stay in 1980. Apart from these visits, I was unaccompanied during my fieldwork. Being a lone foreign woman was an advantage. I could cross the gender divide as no man could do, and women could visit me without prejudicing their reputations. I explained that I was studying "customs and traditions" and "history," and this was accepted.
This research was conducted under the auspices of the British Museum, where I was curator for Middle Eastern ethnography at the former Museum of Mankind. I initially intended to study crafts (Weir 1975), and I made a collection of artifacts for the Museum. I was also interested in the local economy, on which I have previously published (Weir 1985a, 1985b, and 1987). Then at the start of my second visit, in 1979, a dispute broke out between al-Nazir and a neighboring tribe. This breakthrough event, which I describe in Chapter Eight, opened a fascinating window on tribal law and politics, which thereafter became my primary focus. Hitherto I had understood the tribal system largely through my informants' abstract and idealized descriptions; now I saw tribal politics being practiced, and tribal law being implemented, and glimpsed the richness and complexity of this system. This dispute also first alerted me to the immense importance of documents in Razihi culture. I discovered that hoards of scrolled papers were preserved in peoples' homes, and realized that this was an anthropological treasure trove. By a stroke of luck two Naziris owned photocopiers which ran off portable generators, and during my second two visits I was able to copy more than three hundred and fifty papers spanning nearly four centuries. A short catalogue of most of these is provided in Appendix 2, and they are referenced in the text with the prefix D. I regret that it has been impossible to translate or reproduce many of these fascinating documents in this work because of lack of space, but I plan to do so elsewhere, and the reader is meanwhile referred to Weir 1996.
Most of the documents I copied are pacts and treaties among and between tribal groups, between tribes and state authorities, and among the ruling elite themselves. Whenever problems arose, men gathered (as they still gather) to resolve them, and recorded their decisions, judgments, and agreements in handwritten papers which they preserved for future reference. These documents are an invaluable primary source, for they catch local people "inscribing" their own system for their own purposes at different periods. They provide a wealth of information about tribal political and legal matters. Because they are usually dated, and invariably name the participants of the meetings at which they were written, they also make it possible to track individuals and groups, and their relationships, back through time; however, they only rarely and briefly provide information about events such as would facilitate a narrative history. Overall, the documents testify to the predominantly contractual basis of political relations in Razih, and to remarkable continuities in structures and practices, both of which are major themes of this work.
Throughout my fieldwork I gathered social information indiscriminately while participating in everyday life and special occasions. I also collected data by more formal methods: photography, tape recordings, censuses of selected settlements, and surveys of occupations, land-ownership, and the market of al-Nazir. With local help I also did preliminary translations of documents—a difficult task which has continued over subsequent years, and delayed the completion of this work. This variegated material has allowed me to present my themes and arguments with an eclectic mixture of formal description and analysis, quotations from informants and documents, and narratives of events and cases based on documentary evidence and informants' accounts. Stories had undoubtedly been adjusted in their countless retelling to fit current concerns; but Razihis are politically sophisticated, and their accounts reveal a refreshing absence of mystification about how their system operates, and are filled with information and insights. Although my authorial overview of the tribal system of Razih undoubtedly differs from that which any local person would produce, I hope that Razihis who one day read this book will recognize the picture I paint as true in essentials to their own understandings.
This book describes the politico-legal system of small tribes of farmers and traders which have existed on Jabal Razih, in much the same form, for at least four centuries, and considers their historical relationship with a continuous succession of religious rulers and the present republican state. Throughout the book I have addressed fundamental questions of governance. What are the key political groups, and how are they conceptualized? What accounts for their size and positions? How are power and authority distributed and exercised, and curbed or resisted? How are disputes settled and order restored, how effectively, and in whose interests? And how are the institutions, principles, rules, and procedures for maintaining law and order sustained and reproduced, or changed? These issues are especially compelling when a political system exhibits, as does that of Razih, remarkable structural and cultural continuities, and an apparently abiding concern for containing and resolving conflicts and minimizing violence.
The entities usually called "tribes" found throughout rural North Africa and the Middle East are diverse polities. Anthropological attempts to formulate a detailed, universally applicable definition have therefore invariably fallen foul of exceptions. Some anthropologists of Yemen have therefore employed the term tribe without defining it, or have opted for a less loaded term, such as community. We need to categorize and label, however, in order to discuss, differentiate, and compare sociological phenomena, and the term tribe is still a useful portmanteau term, I believe, for territorial polities whose members share a common allegiance, which exist in a matrix of similar polities with which they have relations, and which have always been potentially or actually formally subordinate to some kind of "state," also, of course, a problematic term. These minimal criteria at least distinguish "tribes" from other sub-state entities, such as "ethnic groups" or "peoples," which are not necessarily political organizations; from political parties which are not territorial; and from state-administrative units which do not interact politically, and whose members owe allegiance only to the state. At the same time, my general formulation deliberately invites description of tribal features which vary through space and time, and which should always be empirically determined, if possible, for different regions and periods.
Among important variables are economies. Not all tribes are nomadic or transhumant, as some have assumed or implied (Gellner 1981:24, 89; Asad 1986; Eickelman 2002:46). Many in Afghanistan, Iraq, Iran, Morocco, and Algeria, and most in Yemen, comprise populations of settled farmers. Tribes also vary in their criteria of membership and idioms of identity. Many are conceptualized as large descent groups, but others place greater emphasis on idioms of place and leadership. If we wish to understand political action, therefore, we need to avoid thinking of tribes as invariably comprising "large kin groups, organized and regulated according to ties of blood or family lineage" (Khoury and Kostiner 1991:4), and to be alert for other organizing principles.
Much of Yemen is divided into mostly sedentary tribes, and has been since antiquity; the only major region which seems never to have been tribally organized surrounds the towns of Ibb and Ta`izz in "Lower Yemen." Anthropological studies show that while the tribes of Yemen share a similar political culture, especially those of the northern highlands ("Upper Yemen"), they also vary regionally in size, forms of identity, and modes of organization. Unfortunately, as Gingrich (1993) notes, this been obscured by homogenizing generalizations about "the tribes of Yemen." Such approaches not only mask diversity, but also fail to address the environmental and historical factors underlying it. Dresch even explicitly denies the relevance of ecology for understanding Yemeni tribes, the structures of which he portrays as sets of cognitive categories on a linguistic model. But tribal groupings are more than systems of definition or classification which can be analytically divorced from their varied contexts as he implies.
Yemen is a land of dramatically differentiated topography, climate, and ecology. Some tribes have territories in arid regions of marginal agriculture and widely scattered populations; others occupy fertile, well-watered regions, produce surpluses and cash crops, and are densely populated. Some tribes occupy easily traversed plains, others steep, inaccessible mountains. Tribes also vary in their proximity to, or remoteness from, major towns and trade routes, and the peripatetic centers of past states. All such factors have, I contend, affected the sizes of tribes, how they are organized, and their inclination and ability to solve their problems peaceably or violently in different historical circumstances. The tribal relationship with local or colonial states was also contingent on environmental factors. Some tribes were too poor or remote to attract rulers, and could evade or resist their control; others were strategically or economically attractive to states, and were forced or induced to submit. The kind and degree of state control, and the policies and methods of different rulers, also affected tribes. It is therefore essential to examine each tribal system on its own terms, and within its particular geographical, economic, and historical context, as I have attempted to do in this book. This is not, I should stress, to argue for some kind of economic or political determinism. On the contrary, I see tribes as polities created, maintained, or changed by people acting, individually or collectively, in their own perceived interests, and striving to achieve concrete goals. With this in mind, I have throughout emphasized the instrumental and administrative aspects of tribal politics and state-tribe relations, and tried to show how people construct and operate their system, resist or comply with its dictates, and compete for power, prestige, and the rewards of office.
The dominant anthropological model of tribal political organization in the Middle East and North Africa when I embarked on fieldwork was "segmentary theory." This attempts to explain the maintenance of order in so-called "acephalous" societies which lack (or are assumed to lack) permanent governmental institutions. The ideal-type "tribe" of this theory comprises a population which usually claims patrilineal descent from a common eponymous ancestor, and is subdivided into a hierarchy of "nested" lineages or "segments" named after subsequent ancestors. Although anthropologists use various labels to differentiate these tribal "segments" (typically, in ascending order, family, lineage, clan, tribe, and confederation), the segmentary model insists that these structures are essentially homologous—that each comprises more or less egalitarian kin groups which replicate in all but size those of which they are part, or into which they are divided. Above all, and this is the crux of the theory, no segment has specialized or permanent political functions—there is no "crucial level of social organization" (Gellner 1981:117; 1991:109).
The fundamental concept of segmentarism as a theory of politico-legal action is that of "balanced opposition." In the absence of effective leaders, order and the balance of power are maintained by collective action: equivalent groups at different levels of the system mobilize in response to threats, then dissolve when they abate. This happens, of course, in many societies, but in segmentary societies it is this mechanism alone which operates. The action groups typically take is violent revenge or "feud," and conflicts are resolved by temporary mediators with little power—a role for which Gellner notably argued holy men ("saints") are specially qualified, in Muslim societies, because of their religious prestige and political neutrality. But peace is temporary, and dormant feuds are repeatedly resuscitated. The segmentary model has been challenged by several anthropologists, and many think it now defunct. It retains explanatory power, however, for societies lacking specialized order-maintaining institutions, has a degree of fit with what some anthropologists have observed, especially among nomadic tribes, and remains useful as an ideal type for comparative purposes.
As we shall see, the tribal system of Razih shares several characteristics with segmentary systems: it comprises nested groups (though they are not homologous); political relations are expressed in kinship idioms; collective responsibility is fundamental; groups can take revenge; and mediation is centrally important, including by religious specialists. But Razih has a crucial level of organization, its long-lived tribes, and other stable governmental institutions: dynastic leadership; administrative and judicial structures; written laws, and specialized personnel and procedures for their enforcement; and durable alliances based on contracts and treaties reflecting interests, not genealogical connections. All these features are incompatible with the segmentary model.
My research therefore led me to adopt an alternative, geo-political approach to understanding the tribes of Razih, which I eventually conceived as tiny sovereign domains, each governed and represented by leaders with constitutional authority and powers of office. The concepts of tribal governance and sovereignty have a long history in the anthropology of different countries and continents (see Dole 1968; Vincent 1990:42-46), but have been under-used in studies of North African and Middle Eastern tribes. This might be partly because key polities are often small, unmarked by dedicated buildings, and have governmental practices which can seem "rudimentary" and "informal" from the (sometimes condescending) perspective of members of modern states (see Vincent 1990:46; Gledhill 2000:11). It might therefore be necessary to focus closely on individual communities and families in order to find "government," as Mundy (1995) has notably demonstrated for another area of Yemen. Tribal governance has also been insufficiently recognized because, as Munson (1989) and Hugh Roberts (2002, 2003) have argued, some anthropologists have been blinkered by segmentary theory, with its anarchic and agonistic vision of tribal societies. They have therefore tended to see tribesmen more as warriors to be mustered in battle than as citizens subject to the same jurisdictions. Like the citizens of states, however, and as the evidence from Razih shows, they can be either according to circumstances.
Razih lies due west of the northern plateau town of Sa`dah, and due east of the Red Sea port of Jizan. It is a fertile and populous region with a productive economy based on agriculture and trade; it bestrides an important trade route across the northern mountains; and in the west it commands the coastal plain (the Tihamah). For fiscal and strategic reasons, therefore, it has probably attracted some kind of supra-tribal or "state" control since antiquity—perhaps even since Sabaean times in the mid-first millennium BC.
Razih has experienced great cultural continuity in state governance. Since the birth of Islam in the seventh century, it has known nothing but Muslim regimes, and from the foundation of the Zaydi-Shi`ite state in Yemen in the late ninth century until Razih joined the Republic after the 1960s Civil War, it was usually under some kind of Zaydi dawlah—a polysemic term which can be translated as "state," "regime," "ruler," or "government," according to context. The only major interruption to Zaydi rule was seven decades of Ottoman occupation from the mid-sixteenth to the early seventeenth century. For over three centuries thereafter, Razih was ruled by members of the Qasimi dynasty, whose founder initiated the Zaydi insurrections which eventually ousted the Ottomans from Yemen. For periods it was under the main Zaydi-Qasimi rulers (sing. imam) based in various highland seats including Sanaa. At other times it was part of dissident imamates or dawlahs based in Sa`dah. And for long periods it was under locally based dawlahs. It will be noted that, in common with the rest of North Yemen, Razih has never been colonized by a European power.
Zaydi dawlahs were very different from the Weberian-western model of states. None had sole jurisdiction, nor a monopoly of the legitimate use of force, within the domains they claimed and aspired to rule. All were also weak in human and material resources, and had tiny, highly personalized administrations which mainly aimed to administer shari`ah law, collect taxes, and preserve or expand their own hegemony; and the tribes which constituted most of their domains had substantial politico-jural autonomy and were well armed. These Zaydi domains also fluctuated in size as tribes and territories were won or lost, and it was not until the twentieth century that the Zaydi imamate gained fixed, if contested, borders.
In Razih these weak and fissiparous dawlahs were superimposed on relatively stable tribes, which functioned as the prime and constant units of state governance. Razihis referred to this composite polity by the collocation "state-and-tribe" (dawlah-wa-qabilah). As I will show in Part III, by their policy of indirect rule, imams and other religious overlords reinforced the tribal system of Razih, both ideologically and instrumentally, by accepting it as the principal and legitimate form of local governance, and by co-opting and exploiting its structures and practices for their own fiscal, legal, and military purposes. At the same time, they caused modifications in tribal structures, and triggered changes and disturbances in the tribal system by strengthening or weakening particular leaders, tilting local balances (or imbalances) of power, and creating or exacerbating conflicts of interest.
I touch on aspects of the state-tribe relationship in Razih throughout the first two parts of the book, which focus mainly on tribal politics and law, but have deferred detailed consideration of the state-tribe relationship in a historical context until last, after the tribal system has been described. I have taken the expulsion of the Ottomans from Yemen as my starting point for this more chronological treatment, partly because it was a major juncture in local and national history, and partly because the earliest Razihi documents I found date back to that period.