Download PDF | Brian Ulrich - Arabs in the Early Islamic Empire_ Exploring Al-Azd Tribal Identity-Edinburgh University Press (2019).
268 Pages
Introduction
Explanation and Historiography of the Problem
Classical Arabic writers frequently portrayed the Arabs as a tribally
defined people. During the 1200s, Ibn Manzur’s Lisan al-ʿArab quoted
numerous prior authorities as saying that an ʿarabī was anyone descended
from the ʿarab.
1
Peter Webb, however, notes that in al-Khalil b. Ahmad
al-Farahidi’s dictionary Kitab al-ʿAyn, which was finalised by a colleague
after his death in 791, ‘Arab’ (ʿarab) was defined on the basis of purity of
speech without reference to language, and was thus the opposite of ʿajam,
characterised by impure speech. Genealogical definitions of Arabness rose
to the forefront during the ninth century, though one suspects they drew
upon earlier antecedents.2
It was in the early ninth century that Hisham b.
Muhammad Ibn al-Kalbi (d. 819) wrote his works on genealogy, which
drew on earlier material from a culture in which descent was a common
way of characterising identity.
Particularly in Nasab Maʿadd wa alYaman al-Kabir, Ibn al-Kalbi presents what came to be the two branches
of the Arabs, often called ‘northerners’ (Qays or Nizar) and ‘southerners’
(al-Yaman).3
Al-Khalil b. Ahmad would have had reasons for finding congenial the
linguistic definitions of ‘ʿarab’ circulating in his day. Born in Oman, he
belonged to a tribal grouping called al-Azd.4
In Ibn al-Kalbi’s influential
genealogical work, al-Azd was a name applied to Diraʾ b. al-Ghawth,
one of the descendants of Qahtan, progenitor of al-Yaman.5
As Webb
demonstrates, ‘Arabness’ was a fluid and contested idea during the early
caliphate. On the basis of pre-Islamic Arabic poetry, he argues that no one
self-identified as an Arab in late pre-Islamic Arabia, and that the major
ethnic grouping for the poets was actually Maʿadd, later turned into a
key ancestor of the northerners.6
Robert Hoyland has argued that Webb
underestimated both the epigraphic and poetic evidence for a potentially
language-based Arab identity prior to Islam, saying ‘he operates with the
notion that either we have a coherent all-embracing Arab identity or no identity, whereas a much more nuanced approach is needed.’7
Regardless
of the degree to which Arab identity existed prior to Islam, however, it
certainly gained prominence in the wake of the conquests which brought
those of Arabian heritage to political and cultural power, and what it
entailed became a subject of contestation and debate, with the southerners
generally and often the Azd specifically denied the status of an Arabness
constructed on a Maʿadd base.8
The central argument of this book is that al-Azd identity existed
from pre-Islamic times into the Abbasid period, though the meaning of
that identity shifted in different contexts across time and space. This
persistence is seen not only through the use of tribal ties as political and
economic assets under certain conditions, but also through the composition and preservation of narratives which reveal a concern with tribal
reputation and heritage. As a contribution to the literature on tribes and
states in Middle Eastern history, this work problematises concepts of
both tribe and state in arguing that the incorporation of tribally organised
populations into the caliphate should be viewed as a long-term process
of incorporating and redeploying both material and cultural resources.
As a work of early Islamic history, it argues that the continuity of
development between late antiquity and early Islam that scholars have
observed in other areas is also shown in the behaviour of tribespeople
and the relationship between tribes and states in the region.9
This means,
in part, that tribes continued to be a useful form of social organisation
in ways more similar to those of pre-Islamic Arabia than has sometimes
been supposed.
This study also has implications for several larger issues.
The first is
of concern to anthropologists as well as historians, and involves tribes
and states as overlapping forms of social relations. The argument herein
is that while Bedouin perceived tribe in terms of a network of assets,
options and obligations, tribal groups often depended upon key leaders
who served as guarantors of those assets and options. As the configurations of power took on more statelike qualities, state policies shaped the
elite world inhabited by those leaders, leading to shifts in the nature of
such leadership and the function of tribal identity. Another issue involves
ethnicity in the early Islamic Middle East and the role played by tribal
ties in forming a more general sense of Arab identity, as described above.
Not only was Arabian descent an important part of the status claims of
urban notables, but also tribal ties were a means of establishing solidarity with substantial rural populations which could serve as military and
economic resources.
Historiographic Background
Much of the early work on tribes in early Islamic history focused on
the origins of the division between the ‘northern’ and ‘southern’ Arabs
mentioned above. Some scholars, such as W. Robertson Smith, claimed
these represented ‘two great races’ into which Arabs had been divided
‘from time immemorial’.10 However, in his crucial study of Ibn al-Kalbi’s
genealogical work, Werner Caskel found that there was no evidence for
this division prior to Islam, though one could perhaps discern the conditions for its development in the cultural differences between the settled
agricultural civilisation of south Arabia and the pastoralists and oasis
dwellers further north.11 M. A. Shaban argues that Qays and al-Yaman
were actually political parties, which he goes so far as to compare to the
Whigs and Tories of seventeenth-century England.12 Partially following
Shaban, Fukuzo Amabe sees the two groups as tribal confederations with
different duties, and attributes the enmity between the two to their favouring different policies based on those duties. However, Amabe himself
notes numerous exceptions where groups of the Qays and al-Yaman performed the duties generally associated with the other group.13 Shaban’s
idea of political parties was definitively refuted by Patricia Crone.14 There
has, meanwhile, been some movement toward arguing that the larger confederations of al-Yaman, Mudar, and Rabiʿa, the latter two subdivisions of
the ‘northerners’, may in fact have origins in the sixth century, with their
consolidation during the seventh and eighth representing a continuation of
pre-Islamic developments.15
More recently, scholars have focused attention on the social, economic
and political factors which contributed to changes in the tribal structure.
Fred Donner has examined the effect on the Arabian tribes of the formation of the early Islamic polity, arguing that the state provided means of
tying individuals to itself through the payment of military stipends and
shares from state-owned lands, while simultaneously using tribal ties as
mechanisms of classification and enforcement.16 Crone’s Slaves on Horses
has been another influential interpretation. In this work, Crone ‘presents
an explanation of how and why slave soldiers came to be a central feature
of the Muslim polity.’17 One major component in her explanation is that
the early caliphal state destroyed the Arab tribal structure, replacing it
with newly created military administrative units in the garrison towns
that simply bore tribal names and were ruled indirectly through a class of
ashrāf (notables) from the leading families among the tribes.18 A further
development took place during the Marwanid period, in which these moieties in turn became purely military units, and the conflicts between them an example of military factionalism.19 Although she highlights differences
between Central Asian and Arab conquests, on the issue of tribal military
units Crone may have been influenced by Chinggis Khan’s reorganisation
of the Central Asian tribes into an ordered military structure, the elements
of which bore tribal names.20
Crone’s factionalism model has much to recommend it as a means of
understanding Marwanid politics. It is, however, inadequate as a means
of understanding relations between the Arab tribes and the early Islamic
state apart from the military realm, and her account of its evolution needs
revisiting in light of more recent understandings of pre-Islamic Arabia and
the evolution of Arab identity. For Crone, the Arabs ‘enjoyed an ethnic
and cultural homogeneity quite without parallel in Central Asia or Europe’
and ‘had lived in freedom from ethnic and social disturbance from very
ancient times.’21 A characteristic of the Arabian Peninsula was thus ‘tribal
immutability’.22 None of this can now be accepted. Arab identity prior to
Islam, if it existed at all, definitely did not extend to the entire peninsula
and was probably of little importance, while states had long been involved
with the politics of the peninsula and its tribes.
Crone’s model of the interaction of tribes and states in the region is
also flawed. Her definition of ‘tribe’ is ‘a descent group which constitutes
a political community’ and, specifically, ‘that descent group within which
control of pasture land is vested, within which particular rules regarding
blood-money and other aspects of behavior apply, which is endowed with
a chief, and within which most of social life is conducted.’23 She sees these
tribes as a completely alternative method of social organisation to states,
which as superior socio-political structures undermine and ultimately
destroy them.24 As Richard Tapper has pointed out, the last assertion
seems to run contrary to the history of the Middle East in which tribes and
states coexisted for long periods, and tribes sometimes gained ascendancy
over and even conquered states.25 It does, however, apparently lie behind
her view of the Arab tribes in the early caliphate.
Most recently, Eva Orthmann has examined the Arab tribes from the
last years of the Umayyad period into the 800s and the rise of Turkish soldiery during the reign of al-Maʾmun. In a book that ranges widely through
source criticism, case studies of events, Arabic terminology, tribal structure, and more, Orthmann concludes that contrary to Crone, Arab tribes
continued to play a role in political and economic life in cities, towns and
the steppe throughout the period under study. The Abbasid caliphs faced
many tribal revolts, and tried various strategies for managing the tribes,
including co-option, the appointment of leaders, and dividing and conquering.26 She concludes that rivalry between Qays and al-Yaman did not motivate actions, but was instead one of the tools which groups in certain
times and places used to promote their interests.27
Orthmann’s view that membership in a broad tribal confederation could
serve as a situational asset will be mentioned again below. Another implication of her work focuses attention on studies of smaller tribal groups
which held more regular significance. M. J. Kister’s doctoral dissertation,
‘Tamim in the Period of the Jahiliyya: A Study in Tribal Traditions’, uses
poetry, akhbār reports and genealogical literature to compile a valuable
picture of how the Tamim represented their pre-Islamic past, while also
calling attention to that tribe’s important links with both Mecca and the
Sasanians.28 Michael Lecker published a prosopographical study of the
Banu Sulaym, a tribe near Mecca during the life of Muhammad, partly to
establish that ‘dedicating a monograph to a single tribe is a feasible and
useful project’. 29 Ella Landau-Tasseron drew upon her earlier work concerning the Wars of the Ridda in publishing a brief study of Asad during
that period, concluding that the term ‘Asad’ in the sources usually represented only part of the tribe as outlined by Arab genealogists.30 Donner
has come to similar conclusions regarding the Bakr b. Waʾil.31 ʿAbd Allah
Ibrahim al-ʿAskar has written an examination of B. Hanifa’s transition
into the Umayyad period, and Mohammad Rihan has studied the history
of the ʿAmila of Lebanon under the Umayyads.32 Rasheed Hosein has
written about Thaqif, and particularly their relationship with Quraysh.33
Most recently, Georg Leube has taken a prosopographical approach to
the study of Kinda during the first three Muslim generations.34 As a contribution to the history of Arab and Arabian tribes, this work seeks to go
beyond the others by taking a broad chronological scope and offering a
new model for understanding the ongoing dynamic interaction between
structures identified as those of tribe and state.
Conceptual Approaches: Genealogy and Identity
The historiography above has noted definitions of ‘tribe’, and in fact
scholars have produced multiple definitions for that term, with implications for any group such as al-Azd to which it is applied. Often ‘tribe’ has
been used to connote primitiveness and denigrate certain societies and
their history, leading scholars of most regions to abandon it altogether. It
remains current, however, in Middle Eastern Studies. ‘Nation’ and ‘ethnicity’, which have replaced ‘tribe’ among those studying the Americas
and Africa, must in Middle Eastern usage go with linguistic groups such
as the Arabs and Kurds, and so ‘tribe’ translates the Arabic qabīla and
sometimes ʿashīra, much as it does the largest component groups of ancient Israel, for which it was used in English long before colonialism.
However, an implication of this varied use is that scholars must avoid
lumping together all societies labeled ‘tribal’, as social structures vary
widely between the Middle East and, say, Central Asia, where the political
economy is quite different. In addition, especially in the Arab world, the
term should not be associated with nomadism.35
For the authors of our primary sources, al-Azd was defined through
genealogy, particularly that of Ibn al-Kalbi, in which, as noted above, the
Azd are all those descended from Diraʾ b. al-Ghawth, who was sometimes
known as ‘al-Azd’ because he displayed the courage of a lion (al-asad).
Ibn al-Kalbi’s works became standard; the work of his most prominent
successor, Abu Muhammad ʿAli b. Ahmad Ibn Hazm (d. 1064), was
largely built upon them.36 However, the Ibn al-Kalbi corpus, and by extension other extant genealogical texts from the ninth century, should be seen
in the context of the articulation of an Arab identity embracing all cultural
groups with real or claimed roots in the Arabian Peninsula.37 Webb has
ably charted the way genealogical texts are one source illustrating debates
over inclusion and primacy within Arabness.38 Furthermore, there is clear
evidence of alternative schemes specific to al-Azd in the Kitab al-Tijan fi
Muluk Himyar, which in its extant form is by Ibn al-Kalbi’s contemporary
ʿAbd al-Malik b. Hisham (d. 833). Not only is there a Hamdan b. al-Azd
who became the ancestor of some al-Azd in Iraq (specifically at Babylon),
but many of the Azd of the Sarat Mountains are tied to a Hubayr b. alHubur b. al-Azd.39 Despite the prominence of being sons of the al-Azd
eponym himself, neither figure features in Ibn al-Kalbi’s work.
In place of viewing al-Azd as a descent group traceable through genealogical literature, this study adopts a definition of tribe from William and
Fidelity Lancaster: ‘a definition of identity against other similar units of
self-definition acting in a moral arena’. These tribes are not corporate
entities, and in fact may be widely dispersed.40 They have reputation, and
tribespeople partake of that reputation in assuming and being accepted
into the identity, as well as gaining access to any economic resources
belonging to the tribe either through communal rights to land use or government stipends. Tribes are also formed by individuals and sub-groups
who conceive of themselves as freely in association with each other.41
This definition allows not only for the groups’ decentralised organisation
and dispersal throughout the Muslim lands, but also for the consistency in
conceptual associations, both personal, as when individuals use the al-Azd
label to move around the caliphate, and with the perception of common
heritage seen in the Oman-related insults directed at al-Azd in different
times and places.
For those who adhere to them, tribal identities and the relationships
they involve provide assets, options and obligations.42 These pertain not
only to the material realm, but to the ideological one as well. The ‘reputation’ described by the Lancasters is similar to Michael Meeker’s concept
of ‘segmentary sharaf’. With this term, Meeker referred to the deeds of
ancestors which brought honour to their clan, deeds the power of which
derived in significant ways from their recognition by other clans.43 The
concept is strikingly similar to the modern Jordanian tribal histories
studied by Andrew Shryock, who writes, ‘Ancestral ties to men of renown,
or the lack thereof, were offered as proof of a tribesman’s social standing’,
though in the context of his research different tribal histories were kept
apart to avoid social division.44 Here, too, we see the potential theoretical
significance of tribes using higher identity levels as assets, such as with
Orthmann’s conclusion about Qays and al-Yaman.
The Lancasters’ definition of tribe provides a framework which resonates with our sources, but as a corrective to views which emphasise an
egalitarian ideal, this work occasionally references the idea of ‘aristocratic
orders’ postulated by David Sneath. These aristocratic orders were composed of individual houses which over a period of generations mobilised
genealogy to claim social and legal privilege and access to resources.
Thus problematising the distinction between classically conceived states
and stateless societies, and in our terminology tribal societies, Sneath goes
on to suggest that aristocratic orders could provide ‘a common mode and
language of power that could allow rapid incorporation’ into larger state
structures.45 Sneath’s broader view of ‘headless states’ has been criticised,
however, and there is no evidence it is relevant to Arabia. In addition,
kinship probably plays more of a role than he allows, and the dynamics
between kinship and leadership need to be handled sensitively in different
historical contexts.46
Although Sneath addresses Arab societies only briefly, we have an
indication that there could be aristocratic orders in the term ‘sharīf’,
which for Ibn Manzur was a noble status one could only get through one’s
ancestors.47 Christian Robin shows that even in ancient times, Yemen had
a highly stratified society dominated by noble lineages.48 J. C. Wilkinson
has identified tribal formation around loyalty to local shaykhs as a recurring pattern in Oman’s history.49 Ibn Khaldun’s view of nomadic tribes,
informed by North Africa and the Middle East, also emphasised the role
of tribal leaders more than scholars have generally noted.50 Aziz al-Azmeh
has highlighted the ways in which pre-Islamic Arabia was clearly hierarchical: ‘There existed a very decided reality to hierarchy and social differentiation, to the prerogatives of noble lineages and, within these lineages, to their chiefs . . . Pre-Islamic poetry is replete with a vocabulary of hierarchy, dominance and superordination, as it celebrated feats of arms.’51
This study will frequently note the status of leaders among al-Azd and
their significance for their followers, as well as the way that significance
factored into the incorporation of the tribes into the caliphate.
Amidst this corrective, however, we should bear in mind that shaykhs
in much of Arabia had far less control over their followers than did those
in Central Asia, perhaps as a consequence of the fact there were fewer
resources for them to extract and use to develop coercive potential.
There is much in the sources to suggest that among their own followers,
leaders acted in ways parallel to the modern ethnographic observations
of Lancaster that shaykhs maintained their influence through reputation,
persuasion and mediation.52 The key point, however, is that egalitarianism
was an idea, not a reality. Al-Azmeh explains: ‘Individual honour, ferocious as it may have been, being dependent upon the collectivity, was,
by some perennial covenant, deposited in the chief, the sayyid.’53 Again,
tribespeople perceived tribal identity in terms of assets, options and obligations. The power of the tribal leader, however, guaranteed the resources
which gave rise to the assets, and he led during the reputational conflicts
which could be a significant obligation.
In order to understand the relationship of tribe and state in the early
caliphate, this work is also influenced by scholarship on comparative
empires. Even though the Middle East was home to the first societies
scholars describe as empires, the caliphate has seldom merited more than
a passing mention in the scholarship dedicated to the theme.54 The most
in-depth discussion is that of S. E. Finer, who draws a distinction between
an ‘Empire Mark I’, characterised by the domination of one identity group
over others, and an ‘Empire Mark II’, in which the evolution of a common
imperial culture served as the common marker of a porous ruling class.
Finer specifically cites the difference between the Umayyad and Abbasid
caliphates as an example of this distinction, along with the development
of China between the Han and Song dynasties.55 As with all pre-modern
states, the caliphate ruled through cooperation with local notables, the
nature of which varied across time and space but could include ʿulamāʾ,
tribal leaders and wealthy landowners.56 The present work will highlight
ways in which the shifting types of intermediary between state and society
had an impact on the signification of al-Azd identity, while the incorporation of that identity into the common imperial culture shows how elite
Azdis helped to shape that culture while finding a place within it.
Searching for al-Azd Identity
The present work, then, focuses on the development and meaning of
al-Azd identity and its role in the development of the early Islamic polity.
It remains to understand the methods of working with the source material,
which consists almost exclusively of Arabic texts containing information about the times and places under investigation. Apart from papyri
and inscriptions, our earliest Arabic sources date from the ninth century,
much later than the period for which most critical points of this study will
be developed. While it is true that these writers primarily compiled and
edited for their own purposes accounts from earlier generations of scholars, with isnads showing the purported chains of transmission, the main
bulk of those accounts dates from only the eighth century. Such source
material behind our extant works included written works from earlier
authors, notes for or from lectures, and perhaps ‘sourcebooks’, meaning
here distinct but related materials which were combined organisationally
in ways Sarah Bowen Savant compares to packets of reading for modern
university courses.57 Part of the historical investigation, then, shall involve
understanding the texts themselves and the ways in which we can use them
to understand periods before their composition. The approach favoured in
the present work is what Fred Donner calls ‘tradition criticism’, distinguished by an emphasis on the means and context of the transmission of
information within the texts.58
Our approach to the texts recognises that they contain what James
Fentress and Chris Wickham call ‘social memory’, which in their description ‘identifies a group, giving it a sense of its past and defining its aspirations for the future’.59 Social memory, because of its role for the group, is
reflected upon and contested. As the authors say, ‘This means that we must
situate groups in relation to their own traditions, asking how they interpret
their own “ghosts”, and how they use them as a source of knowledge.’60
Many of the characteristics of social memory comport well with the observations of scholars analysing the Arabic source material. For example,
social memory is not only articulated in words, but also in images and
rituals. These images, however, are both simplified to aid in transmission
and conventionalised so as to achieve meaning for the entire group.61
This, of course, recalls the frequent use of topoi in ancient and medieval
literary sources, which may also have been linked to memorisation techniques. Noth and Conrad have identified a number of topoi in early Islamic
history which one must understand in order to properly assess the source
material.62
The Arabs eventually remembered themselves as a genealogical family, and the genealogical works produced from the ninth century on often
served to record not only what were believed to be the relations among
tribes, but also the deeds of generations past which could serve as the
heritage of the Arab people. Writers of genealogies often wrote works
refuting those they called shuʿūbīs, a term which should probably be read
as simply connoting anti-Arab bigotry.63 One recurring source throughout
this work is a genealogical text called either al-Ansab or Ansab al-ʿArab.
Although attributed to Salama b. Muslim al-ʿAwtabi, who authored some
Ibadi religious works, it is probably the work of an anonymous Omani
author from the mid-tenth century, albeit with later emendations, and so
referenced herein as ps.-al-ʿAwtabi.64 As Hinds notes, this work is probably more profitably regarded as an Omani heritage project than that of a
particular tribe, but the significance of al-Azd in Oman means that they
dominate the second volume.65
Scholars have debated the degree to which information in these genealogies goes back to pre-Islamic times.66 The present work generally follows
the views of Zoltan Szombathy, who sought to trace the method by which
scholars developing the genealogies worked. A key part of Szombathy’s
argument is a distinction between general claims of descent or relationship
and the systematic tracing of lines of ancestors and kinship connections,
or genealogy.67 The concept of nasab among Bedouin before and during
the early days of Islam was simply that of relationship.68 The pre-Islamic
nassāba was someone who knew the heroes, legends and dishonorable
relations of the tribes with which he was concerned, knowledge that served
both to create group heritage and to attack the honour of rivals.69 Bedouin
in the desert had no reason to be interested in the nasab of tribes far afield,
though elites in areas close to societies with more developed literate cultures have left inscriptions showing that they, at least, maintained longer
and more comprehensive genealogies.70 In addition, pre-Islamic Arabs
seem not to have had the belief that all tribes were composed of descendants of a common ancestor, as opposed to what became the general if
not universal assumption during the Abbasid period. Southern Arabia in
particular had tribes whose names derived from toponyms.71 Robertson
Smith noted that as late as Umayyad times, some tribes were represented
in poetry as having female eponyms.72 Asad Ahmed and Webb have both
seen a decline in the significance of matrilineal lines during the eighth
century.73 Because of these shifts, the present work will, particularly in
dealing with earlier periods, privilege as most likely older accounts which
do not conform to the patterns expected in the Abbasid period.
Szombathy describes the macro-genealogy as composed of ‘patches’
and ‘filling stuff’. A ‘patch’ was the name of one or more real or legendary figures drawn from the source material. Thus, the prophet Muhammad and
his immediate family represented ‘patch number one’. The ‘filling stuff’
was scholarly conjecture, such as names added to a genealogy to push a
figure further into the past or relationships between patches that seemed to
the genealogists to belong together.74 In addition, scholars felt compelled
to provide all manner of details, both to secure their scholarly reputations and out of a sense that not knowing the answer to a question was an
intolerable failure.75 The ‘patches’ for the systematised genealogy came
from tribal tales and poems, accounts of local relationships among tribes,
hadith, and the official registers for the tribally organised stipend system
in the garrison towns. Although some of this information was collected
among Bedouin in the desert, far more stemmed from Bedouin or their
immediate descendants in the cities of Iraq, especially the market west
of Basra. Poetry from pre- and early Islamic times also had information
which the developers of Arabic genealogy integrated into their scheme.76
However, although adopting Szombathy’s model for the formation
of the genealogies, as well as his terminology, the present work resists
drawing too firm a distinction between pre-Islamic Bedouin and Arabs in
the early caliphate, for the former were also impacted by elites and state
structures. Aziz al-Azmeh believes overarching genealogies were articulated under the Jafnids and Nasrids drawing upon ‘ancient ethnological
lore prevalent in the steppe and desert’.77 Ibn al-Kalbi, for example, drew
in his works on material from al-Hira, where history was produced during
the reign of al-Nuʿman III.78 In other words, the ‘patches’ might often
be larger than Szombathy appears to allow, and as with the critique of
Crone’s view of Arab tribal identity discussed in the introduction, a view
of pristine desert tribes is misleading.
Because genealogies can still serve as a useful reference for orientation, this work will note how different al-Azd groups mentioned fit into
them. The thrust of reconstruction, however, will rely on other material.
One important source is poetry, which survives even from the preIslamic period.79 Although there was some forgery of poetry, most is now
accepted as authentic, but unfortunately the bulk of the pre-Islamic Arabic
poetry comes from northern Arabia and the Hijaz, and there is little from
the pre-Islamic Azd.80 One possible reason for this is a lack of the Bedouin
themes that appealed to Abbasid-era collectors influenced by what Webb
has called the ‘Bedouinisation of memory’, an issue which will also figure
into Chapter 1.81 The same may account for why al-Azd is so little referenced in the adab literature, material delineating the attributes necessary
to become a cultured member of Abbasid urban society,82
Many sections of this work will rely heavily on prosopography, defined by Lawrence Stone as, ‘the investigation of the . . . characteristics of a
group of actors in history by means of a collective study of their lives.’83
Crone, springing off her deep scepticism about the reliability of the
sources’ narratives, argued that ‘early Islamic history has to be almost
exclusively prosopographical’.84 As noted above, Lecker and Leube use
prosopography in studies of the Banu Sulaym and Kinda respectively.85
Our extant sources are, to put it mildly, filled with lots and lots of names,
and these usually have nisbas, such as al-Azdi, from which we can see
their tribal identity. The imperfections of this technique are obvious, for
only a small percentage of individuals appear in written texts, and there
is a significant bias towards groups involved in events which caught the
attention of later chroniclers, such as the life of Muhammad and the conquests. Nonetheless, from these names, we can discern useful patterns of
geographic concentrations and what levels of identity existed and were
most important in certain times and places.
Particularly for the period after the conquests, the histories produced
by medieval Islamic society also become useful sources of information
on general developments. Meaning can also be teased from the specific
narratives they contain. As noted above, our extant sources from the ninth
century and later drew on earlier material, and through the isnads we have
chains of sources whom we can identify from biographical dictionaries.
Scott Savran notes how narratives are a ‘bridge linking collective memory
and identity construction, to the extent that narrative serves as the vehicle
by which a group imagines its history.’86 In addition to origin stories, the
didactic structure of narratives allows us to consider issues of relevance
to those to whom they can be traced, such as a group for which segmentary sharaf is being claimed and defended.87 This is not to say that we
can ‘reconstruct lost texts’, as we are at the mercy of decisions made in
later recensions and different ways the authorities themselves may have
presented material at different times.88 However, the material of different authorities often has clear patterns and themes which vary by time
and place. In this way, the book also contributes to ongoing scholarship
into historical memory in early Islamic history and our understanding of
the sources on which our knowledge is based, including the work of individual transmitters of historical traditions.
Al-Azd
Al-Azd is a particularly useful tribe for the purposes of this study. First, it
includes populations from both Oman and the Sarat Mountains in western
Arabia, allowing us to highlight the differences between those two areas. It played a significant role in key events of early caliphate, and was well
represented in the conquests of the east. One of its leading families, that of
al-Muhallab b. Abi Sufra and his descendants, were pivotal players on the
Umayyad political scene at a point when Arab and tribal identities were
very much in flux, and thus al-Azd became a key element of the broader
al-Yaman grouping. Finally, the availability of sources by and about
people continuing to use the al-Azdi nisba well into Abbasid times allows
us to consider what the identity contained in the nisba meant for them and
how it fits the broader society of which they were a part.
In Ibn al-Kalbi’s system, al-Azd included not only the groups known
as the Azd Sarat, Azd ʿUman and Azd Shanuʾa (or Shanuʿa), but also
the Ghassanids and the Ansar of Medina. However, there is good reason
to omit the latter two groups from this study. The Ansar, for example,
are treated separately from al-Azd in virtually every context except for
genealogical works and accounts of the ‘Scattering of al-Azd’, which
will feature in Chapter 1. An exception is two lines of poetry attributed to
both Hassan b. Thabit and Saʿd b. al-Husayn al-Khazraji, both poets from
Muhammad’s time, in which both the Ansar and Ghassan are said to be
al-Azd, and Ghassan is described as a watering place.90 However, there is
much poetry falsely attributed to Hassan b. Thabit, and the fact these lines
are so unmoored from authorship suggests they may have been written
much later.91 In his mid-eighth-century Tafsir, Muqatil b. Sulayman
al-Azdi lists the Azd as migrating to Bahrayn and Oman, with the Ansar
and Ghassan listed as separate groups moving to Medina and al-Sham.92
Ghassan was also separate from al-Azd in pre-Islamic inscriptions.93
Despite the historically tenuous relationship, however, Ibn al-Kalbi’s
concern with tribes of historical or religious significance means that in
Nasab Maʿadd wa al-Yaman al-Kabir, roughly half of the al-Azd material involves the Ansar or the Ghassanids. Nothing at all is said of many
of the influential clans found in the Taʾrikh al-Mawsil of Abu Zakariyya
al-Azdi (d. 945), and the Azd ʿUman, supposedly the most numerous, are
squelched together in just four pages.94
That leaves us with the three groups which have ‘Azd’ in their name
above. Szombathy saw the identification of the Azd Sarat and Azd ʿUman
as a single tribe as comparable to a fallacy described by hadith scholars
in which persons or groups with the same name were conflated.95 Caskel
also suggests that the two groups seem forced together to create a single
tribe where none had previously existed. At the same time, given the
degree to which they are mixed together genealogically, he does wonder
if there may in fact have been some sort of pre-Islamic unity.96 Caskel also
suggests that the sheer height of the al-Azd eponym in the genealogical scheme testifies to its artificiality.97 In discussing the important al-Azd
subgroup of Daws, Caskel notes that the Tarif and Sulaym b. Fahm still
lived in the Sarat mountains when Ibn al-Kalbi undertook his genealogical work, that the Munhib had lived there previously, and then argued that
the large Malik b. Fahm grouping from Oman had been added to the
line in Basra.98 This, however, depends on his incorrect claim that the
Azd in Basra prior to the arrival of Malik b. Fahm groups in the city had
mostly been Daws. He also argued that further answers to the joining of
the al-Azd branches probably lay in Kufa.99 The belief in an early Daws
primacy among al-Azd comes from Julius Wellhausen; however it is supported only by Wellhausen’s errant belief that the Huddan were Daws.100
In fact virtually all of the Azd in Basra during the early Islamic period
were from Oman, and all those of Kufa were from western Arabia. All
of the Azd of Basra who have entries in Ibn Saʿd for his first five generations are from Omani tribes. In Khalifa b. Khayyat’s ninth-century Kitab
al-Tabaqat, the only western Azd in Basra is ʿUqba b. Wassaj al-Bursani,
a descendant of the Ghitrif.101 There is nothing in the al-Azd settlement
in Basra to account for the way Oman’s al-Azd fit into the genealogical
tables.
Focusing solely on the Azd ʿUman and Azd Sarat also ignores the Azd
Shanuʾa, the division of the tribe which is most intermingled between
the two sides of Arabia. Wilkinson claims that the difference between
the Azd ʿUman and Azd Shanuʾa was based on different organisational
structures in Oman prior to Islam combined with a special association of
the Azd ʿUman with the Muhallabids.102 He identifies Azd Shanuʾa with
the Omani tribes of ʿUthman b. Nasr, the most famous of which were the
Maʿwala, the Huddan and the Yahmad.103 However, there were also tribes
identified as Azd Shanuʾa found entirely in western Arabia. According
to Abu al-Faraj al-Isfahani (d. 967), the Azd Shanuʾa were the Azd who
settled in the Sarat Mountains after the flood which followed the breaking
of the Maʾrib dam.104 Ps.-al-ʿAwtabi included the Daws as an Azd Shanuʾa
group, as well as the ʿUthman b. Nasr.105 Both Abu ʿUbayd al-Bakri and
Yaqut (d. 1229) included an account naming as the Azd Shanuʾa the tribes
of Ghamid, Bariq and Daws, tribes which drove the Khathʿam out of much
of the Sarat Mountains running between Mecca and Yemen.106 A stray reference to the B. Lihb as Azd Shanuʾa may relate to their association with
the Ghamid.107 Yaqut also traced the Azd Shanuʾa to a Shanuʾa region of
Yemen 42 farsakhs from Sanʿaʾ.108 The early Umayyad poet Suraqa b.
Mirdas al-Bariqi identified the Azd Shanuʾa as his own people (qawm).109
Also, during the Wars of the Ridda, Humayda b. al-Nuʿman al-Bariqi and
his followers fell back on a place called Shanuʾa.110 This region is today identified with the region of the B. Rizam near Abha in Saudi Arabia’s
ʿAsir province.111 Ibn Hisham’s Kitab al-Tijan also links the Azd Shanuʿa
with a place called Shanʾ (sic), and places ancestors of the Ghamid and
Bariq prominently among them.112
The Azd Shanuʾa were thus not a strictly genealogically defined grouping, though they did include the dominant al-Azd subgroups during the
first half of the eighth century.113 Chapter 1 will make the case that they
had a common highland valley lifestyle in both Oman and western Arabia,
and that their rise to prominence and geographic spread occurred in conjunction with the defeat of the Kinda during the sixth century. However,
they were not the only divided grouping. Ibn al-Kalbi also placed descendants of ʿAmr b. al-Azd in both Oman and the Hijaz, though mostly the
former.114 Caskel and Strenziok’s register, however, mentions that some
were no longer al-Azd, such as the al-Saʿiq b. ʿAmr and Saʿd b. ʿAmr
becoming part of ʿAbd al-Qays.115 Caskel and Strenziok also believed
that ʿImran b. ʿAmr Muzayqiya, an important figure in one branch of the
Azd ʿUman, was originally ʿImran b. ʿAmr b. al-Azd.116 Abu Muhammad
al-Hamdani (d. 945) noted groups of al-Asd b. ʿImran in the Sarat
Mountains.117
Preview of the Chapters
Chapter 1 deals with al-Azd origins in pre-Islamic Arabia. The first
section of this chapter reviews the evidence of pre-Islamic inscriptions
and archaeology, arguing for the existence of one or two al-Azd tribal
kingdoms during the third century ce which are dimly reflected in later
historical memory and which represent the earliest known basis of the
al-Azd identity. It also examines the accounts of al-Azd origins from the
early Islamic written tradition, concluding that while they contain some
pre-Islamic elements, they represent the issues of the early caliphate, and
are thus a product of that period in Iraq. Information on the Azd in both
Oman and the Sarat Mountains is then reviewed in detail, which allows
for conclusions about tribal groups and leadership on the eve of Islam and
how they were impacted by the development of the Islamic state. This
chapter also introduces a distinction between a directly controlled western
Arabia and an eastern Arabia that was at best loosely allied with the early
Islamic state, a distinction which will extend further into discussions of the
expansion of the Arabs into the greater Middle East.
Chapter 2 focuses on the conquests and the establishment and society
of the garrison towns in lower Iraq. It begins by arguing that whereas
the conquests of Syria and those which resulted in the foundation of Kufa were centrally directed from the Hijaz, Basra resulted from a more
independent tribal movement which only slowly saw increasing levels
of administration on behalf of the caliphate. It establishes that military
factionalism played an early role in tribal political alignments which
explains the alignment of the tribes in the Battle of the Camel during the
First Civil War, including an alliance of Oman’s al-Azd with Thaqif and
the Umayyads that persisted through the end of the seventh century. It
also examines the shifts in identity in Iraq as larger identity groupings
such as ‘al-Azd’, which seem to have played little practical role in Arabia,
became significant due to their use as administrative units under the early
Umayyads, though it highlights ways in which this process had far more
continuity with pre-Islamic Arabia than is sometimes supposed.
Chapter 3 deals with the Muhallabids, the most important al-Azd
family during the eighth century. A key theme is the changing nature of
elite status during the period, changes which impacted the types of figures
seen as assets and options by Azdis. The Muhallabids were of non-Arab
origin, but after rising to prominence on the eastern frontier both presented
themselves and were presented by later al-Azd sources as a long-standing
shaykhly family. They also represent a transition from independent tribal
generalship in the east to central control of the conquests, with al-Hajjaj’s
rivalry with the Muhallabids marking the end of the al-Azd-Umayyad
alliance and potentially the rise of opposition between ‘northern’ and
‘southern’ Arabs in the caliphate. Throughout, the chapter shall highlight
ways in which the Muhallabids served as ‘sites of memory’ the textual
commemoration of which allows us to see the shifting nature of tribal
identity in later times.
Chapter 4 focuses on the conquests during the first half of the eighth
century and the role of the Azd in the society of the eastern frontier. One
key point developed is that whereas some scholars have posited that the
commercial interests of Omani Azdis were a driving force behind the
conquest of Sind under al-Hajjaj, a careful look at the sources reveals that
Omanis were already present in the region at that time, and that al-Hajjaj
probably attempted to centralize control over movements which began
long before him. This also suggests a pattern, seen also in Yazid b. alMuhallab’s conquests south of the Caspian Sea, according to which factions avoided conquering regions from which they were already benefiting
commercially, regions which were then conquered by others. The last part
of this chapter examines al-Azd settlement in Khurasan and significance
of tribal identification in the province
Finally, Chapter 5 will explore the Azd in the al-Jazira region of northern Iraq under the early Abbasids. This area had a significant tribal population, and much of the conflict in the area took the form of tribal ʿaṣabiyya.
The Abbasid period, however, saw new types of intermediary between
state and tribal population in the form of the learned classes and locally
important leaders who were co-opted with estates and appointments to
official offices. These city-based elites often relied on rural reserves of
manpower to actualise local authority, especially when the central government was weakened by the civil war of the early 800s. The sources on this
period show evidence of a living genealogical idiom, but also new ways
of deploying the pre-Islamic heritage and identities which had been reified
in writing since the formation of the garrison towns’ administrative units.
Link
Press Here
