Fisher G. Between Empires. Arabs, Romans, and Sasanians in Late Antiquity
Подождите немного. Документ загружается.
courts of an urban house’
118
) marks al-H
˙
ayyat as different. Further,
the method of dating to the establishment of the province of Arabia in
106 is not unusual, but the recognition of al-Mundhir in the dating
formula is a remarkable acceptance and acknowledgement of the
Jafnid leader’s prominence in the area. It is also not an isolated
case. To the north, approximately 20 kilometres east of Damascus,
an inscription was found in the vicinity of a structure at al-Burj and
close to the site of D
˙
umayr, originally thought by Brünnow and
Domaszewski to be a second-century Roman camp but which was
recently shown by Lenoir to date to the late antique period.
119
The
inscription, published by Waddington states that a tower was built by
Mundhir. He is described as ÆæŒ[Ø] and also takes the name
Flavius in the inscription.
120
Here, again, there is evidence for the
presence of the Jafnids in a settled area, close to Damascus, and the
inscription is also a rare attestation of direct Jafnid involvement in
building activity.
121
What can we make of this clear evidence for the presence of the
Jafnids and the people linked with them in both the H
˙
aurān and the
area around Damascus? In the first instance, it seems to lend some
support to the Muslim tradition, which places the Jafnids throughout
this region. A further indication of their presence here comes from
John of Ephesus, who tells of how Bostra became a convenient and
local target for disgruntled supporters of al-Mundhir after his arrest
in 581.
122
This, again, suggests a certain number of Jafnid supporters
in the region, roughly contemporary with the completion of the house
at al-H
˙
ayyat. It is certainly logical to expect the leadership and at least
some of the people who presumably made up the group to be found
in the same region, and we should not imagine that the Jafnids
could command any degree of local recognition without the consent
and support of the population on whom they were dependent for
manpower and political leverage. However, in the absence of archae-
ological evidence from Jabiya, or the identification of other sites
linked to the Jafnids in the H
˙
aurān, the speci fics for the presence
118
Foss, ‘Syria in transition’, 251.
119
M. Lenoir, ‘Dumayr, faux camp romain, vraie résidence palatiale’, Syria,76
(1996), 227–36, correcting Brünnow and Domaszewski, DPA, iii. 200; Genequand,
‘Some thoughts’, 79.
120
Wadd. 2562c.
121
Cf. Genequand, ‘Some thoughts’, 79.
122
Joh. Eph. HE 177 (3.3.42).
102 Empires, Clients, and Politics
and distribution of the wider group of people connected to the Jafnids
remains unclear.
It is also tempting from these two inscriptions to draw links
between the Jafnid presence and the expansion of settlement and
the economy in the region. It is obvious that the Jafnids cannot
have been the only factor, but their military and administrative role
as phylarchs, the health of the local sixth-century economy, and the
relative calm experienced by the H
˙
aurān during this period suggest
that they may have played some part.
123
Using archaeological data
showing the well-attested abandonment of Roman fortification sys-
tems throughout Jordan and southern Syria in the sixth century, as
well as data from regional archaeological surveys, J. Johns argued that
the ‘withdrawal’ of the Byzantine administration from the area, and
its administrative transfer to the Arab allies of the Empire, may have
relieved the local tax burden, allowing the community to retain its
agricultural surpluses and increasing the availability of disposable
wealth.
124
Johns’ argument is concerned with showing the prob-
lematic nature of traditional reasoning which finds a linear relation-
ship between the strength of the state, rural prosperity, and
population increase. His argument is thus designed to show that any
‘withdrawal’ of the state from southern Syria in fact boosted prosperity
and population, not the reverse.
125
The archaeological evidence for the
withdrawal of troops from the region is strong, and it is well under-
stood that phylarchs played, in general, some kind of administrative
role.
126
The place of the Jafnids in this is less clear, however, and it is
also hard to imagine that the government in Constantinople would be
willing to forgo tax revenue. The question remains open.
At the same time, however, the relative peace in the region (espe-
cially compared to the districts around Apamea and Antioch, further
north, which suffered repeatedly and more harshly from Sasanian
expeditions and natural disasters) favoured economic growth and
prosperity. The H
˙
aurān is also conspicuous for its lack of forti fica-
tions, specifically in the absence of fortified farmsteads or towers
123
Liebeschuetz, ‘Nomads, phylarchs and settlement’, 144.
124
J. Johns, ‘The longue durée: state and settlement strategies in southern Trans-
jordan across the Islamic centuries’ in E. L. Rogan and T. Tell (eds.), Village, Steppe
and State: The Social Origins of Modern Jordan (London, 1994), 1–31 at 7–10.
125
Ibid. 7–8, 30. Cf. Whittow, ‘Rome and the Jafnids’, 223–4.
126
For a recent survey of the evidence for troop withdrawal and the abandonment
of fortifications, see Fisher, ‘A new perspective on Rome’s desert frontier’.
Empires, Clients, and Politics 103
which are far more common further north.
127
Whether this peace can
be associated with the presence of the Jafnids will also remain open to
debate, and it may actually have more to do with the level of threat in
the south versus the north where natural invasion routes followed the
Euphrates, but not the road across the desert via Palmyra. It is,
however (as Liebeschuetz indeed suggests), reasonable to see some
kind of link between regional security and the extensions of settle-
ment, as well as the lengthy and predominantly stable relationship
between the Jafnids and the Empire.
In my opinion, the most likely link between prosperity and the
Jafnids is a financial one.
128
The presence of the al-Burj inscription
and reports of the tower (long-since demolished) clearly show that
the Jafnids were spending money in the region.
129
Construction and
the installation of inscriptions required funds, presumably derived
either from subsidy or from other sources of income. Ancient writers
describe the lucrative results of raiding, and it is likely that some of
this money found its way into local prestige projects as well as private
coffers and the broader economy.
130
It is possible that the Jafnids also
collected some form of local taxation to their benefit; as we saw above,
the Nas
rids were reported to have received the right to do so from the
Sasanians, and the tradition of tax from pasture land appears as a
factor in the ‘strata’ argument between al-H
˙
ārith and al-Mundhir the
Nas
rid in 537.
131
It is possible, then, that involvement with the local
economy in the H
˙
aurān or its environs may have revolved around the
influx of cash from stipendiary or tax sources which then found
material expression.
132
The presence of such wealth was also likely
to have not only stimulated the local economy, but also to have
provided increased prosperity in a cyclical fashion to those who
participated in it. The tower at al-Burj, for example, was therefore
not simply a monument to the activity of an elite member of society,
127
Foss, ‘Syria in transition’, 252; Tate, Campagnes,48–51.
128
Cf. Whittow, ‘Rome and the Jafnids’, 224.
129
Liebeschuetz, ‘Nomads, phylarchs and settlement’, 145.
130
e.g. Malalas, Chron. 447, on 20,000 slaves taken as booty in the Samaritan
revolt, and then sold in ‘Persia and India’ (c.528); cf. Josh. Styl. Chron. 52, on 18,500
prisoners taken by the Nas
rid al-Nu mān in a raiding expedition in c.502.
131
Proc. BP 2.1.8.
132
Liebeschuetz, ‘Nomads, phylarchs and settlement’, 143; Whittow, ‘Rome and
the Jafnids’, 224.
104 Empires, Clients, and Politics
but also represented the public disposition of money and the hiring of
workmen and artisans. At al-H
˙
ayyat, the house was not built by al-
Mundhir, but it is particularly interesting that Seos insists on the fact
that he and his son Olbanos paid for the construction themselves.
133
The affiliation of Seos is unknown, but this conspicuous, public dis-
play of local wealth is what we might expect to see from wealthy
members of society under the prosperous economic conditions of the
time, and the presence of al-Mundhir’s name on the inscription again
connects him to precisely that.
The construction of the al-Burj tower is a rare example of a so-
called Ghassānid building, in actuality a Jafnid building, but its
inscription, ostentatiously demonstrating the name of its founder,
shows it to be the type of project that was the traditional preserve of
local wealthy individuals. Al-Mundhir was engaging in normal elite
activity, suggesting that he was (or wished to give the impression
that he was) very much part of the local community landscape. This
activity also built on earlier acclamations, such as that from Ne-
māra. In its essence, the Nem āra inscription was a celebration of the
high-profile position of Imruʾ l-Qays, and the heart of this message
is preserved elsewhere, not only at al-Burj and al-H
˙
ayyat but also in
the sixth-century inscription commemorating Abū-Karib, from
Sammāʾ (Ch. 2 n. 71), the mosaics at Nitl (Ch. 2 n. 65), the
acclamation to al-H
˙
ārith at Qas
ral-H
˙
ayr (Ch. 2 n. 97) and the
al-Mundhir building at Res
āfa (Ch. 2 n. 74). Nemāra was ‘inside’
the Roman periphery, but it was a self-declared monument whose
genesis laid claim less to local involvement and more to the wide-
ranging abilities of Imruʾ l-Qays away from the settled lands of the
Roman Empire. Two hundred years later, the inscriptions of
Sammāʾ,Nitl,H
˙
ayyat, and al-Burj, in particular, contain an addi-
tional element: a greater level of integration, specifically in terms of
their imitation of the ways in which Roman elites were typically
celebrated on inscriptions.
In this context, it is worth mentioning here three additional and well-
known inscriptions. At Anasartha, nearChalcis in the north of Syria (see
Map 6), inscriptions on two martyria dating to the fifth century raise the
question of non-Jafnid Arab involvement in building in areas of sig-
nificant Roman settlement. The first is a martyrion to St Thomas, dated
133
Genequand, ‘Some thoughts’, 79.
Empires, Clients, and Politics 105
to 425/6 and dedicated by a certain Mabia/Mavia.
134
Fanciful notions of
a link between this individual and the famous queen of the same name
have been suggested on numerous occasions.
135
The placement of the
Sebaste
Caesarea
Melitene
Germanicia
Nisus
Samosata
Amida
Mardine
Edessa
Carrhae
Zeugma
Europos
Hieropolis
Bathnae
Beroea
Chalcis
Anasartha
Androna
Seriane
Palmyra
0 200km
Larissa
Antioch
Cyrrhus
Doliche
Seleucia
Epiphania
Salaminias
Raphanaea
Emesa
Tripolis
Heliopolis
Damascus
Berytus
Tartus
Barbalissos
Callinicum
Sura
Resafa
Zenobia
MEDITERRANEAN
SEA
RIVER EUPHRATES
Map 6. Northern Syria.
134
The text is published by R. Mouterde, Le limes de Chalcis: organisation de la
steppe en haute Syrie romaine: documents aériens et épigraphiques, 2 vols (Paris, 1945),
i. 194–5.
135
Mouterde, Limes, raised the possibility of a descendant of Mavia; Shahid,
Fourth Century, 222–4, believes it to be the queen herself. See the discussion in
D. Feissel, ‘Les martyria d’Anasartha’, in V. Déroche (ed.), Mélanges Gilbert Dagron,
T&M 14 (Paris, 2002), 201–20, at 205–9.
106 Empires, Clients, and Politics
martyrion outside the walls of the city raises parallels with other extra-
mural constructions, such as the al-Mundhir building at Res
āfa, and
their potential place in regulating the relationship between nomads and
settlers.
136
However, there is no indication that Mavia was a phylarch
and any argument in favour of ‘nomads’ rests insecurely on the fact that
the name is Arabic in origin; all that can be safely said is that a person
with an Arabic name was the dedicator of this martyrion.
137
The second martyrion is more interesting. Built during the same
period, it celebrates a clarissimus ([ºÆ]æÆ) Silvanus as its
dedicator, enigmatically described as having power K ¯æE.
138
The text, which presents numerous significant difficulties of inter-
pretation, contains a variety of Homeric ‘echoes’ and K ¯æE is
taken by Feissel, through analogy to the Odyssey, as a reference to the
Arabs.
139
The martyrion is also apparently dedicated to Silvanus’
recently deceased daughter, described as a young wife of a phylarch.
From this, Silvanus has been variously described as a Roman officer
or even a phylarch himself, but there is no evidence to support such a
claim.
140
What the inscription does show, in the link between the
family of a clarissimus and a phylarch, is a different level of the
connections between phylarchs and the local elite of the sort which
finds a parallel at al-H
˙
ayyat.
Finally, the third inscription, from a martyrion of St John erected at
H
˙
arrān, is dedicated in its inscription in Greek and Arabic by an
unknown phylarch named Sharah
˙
īl(Ææź), who is commemo-
rated as the one who paid (ŒØ) for it (Ææź ƺı
çºÆæå[] ŒØ e Ææ[æØ] F ±ªı øı).
141
Like the
other inscriptions with an Arabic component referred to here, this
example will be discussed in greater detail in the following chapter.
Here I wish simply to point out that, in contrast to the example above
136
Mouterde, Limes, 195.
137
Cf. Liebeschuetz, ‘ Nomads, phylarchs and settlement’, 144.
138
IGLS 297.
139
Feissel, ‘Les martyria d’Anasartha’, 213–14.
140
Mouterde, Limes, 193; Jalabert and Mouterde in IGLS ii. 168–70, consider him
to be a dux Arabiae; Feissel, ‘Les martyria d’Anasartha’, 213: ‘Titulaire d’une dignité
romaine, Silvanos n’était cependant pas un fonctionnaire impérial: son autorité sur les
Arabes (appelés en style homérique ...), qui plus est une autorité perpétuelle, ne peut
être que celle d’un chef indigène, autrement dit un phylarque arabe.’
141
Wadd. 2464. See too Sartre, Trois études, 177; Genequand, ‘Some thoughts’, 80;
E. Littmann, ‘Osservazioni sulle iscrizioni di Harran e di Zebed’, Revista degli studi
orientali, 4 (1911), 193–98, and the discussion of this inscription in Ch. 4.
Empires, Clients, and Politics 107
where the provenance of the phylarch was unknown, both the name
and language on the H
˙
arrān inscription clearly favour identification
with an Arab phylarch, demonstrating unambiguously an engage-
ment in an elite Roman activity by an Arab close to the main
recognised sphere of Jafnid activity.
Of these three inscriptions, the latter two provide the best evidence for
additional links between ‘Roman’ elites, phylarchs, and commemorative
or acclamative inscriptions, and suggest that the presence of the Jafnids
on inscriptions in the H
˙
aurān should be placed within the context of a
wider pattern of integration which had been going on for some time.
Either as benefactors of inscriptions or appearing as third parties, late
antique Arab potentates, characterised in the majority by the Jafnids,
were in essence taking the role of late antique Roman elites. With the
construction of a tower at al-Burj, celebrated by a dedicatory Greek
building inscription, al-Mundhir was firmly linking himself, as he had at
Res
āfa, with the practices of a specific and highly visible sector of society.
SETTLEMENT AND SEDENTARISATION?
A key question is the extent to which the Jafnids were connected to
settlement patterns in the regions where they were active. Of parti-
cular interest is their possible connection to the processes of settle-
ment for any nomads and semi-nomads who may have fallen under
their control in the steppe and desert regions of Syria and Jordan.
Sedentarisation is an ancient and ongoing process, but it is dif ficult to
detect in ancient sources; the close relationship between nomads,
semi-nomads, and settlers obscures its specifics, especially since the
literary evidence prefers to emphasise extremes by highlighting raids,
brigandage, and other minor disruptions.
142
Nevertheless, the on-
going debate continues to demonstrate that settlers (those who
chose to spend the majority of their time in a single domicile in either
urban or rural settings) and nomads were closely connected, even
during episodes of animosity and even while they both occupied
‘settled’ and ‘nomadic’ regions.
143
It is also now well understood
142
E. B. Banning, ‘De Bello Paceque:AReplytoParker’, BASOR, 265 (1987), 52–4, at 53.
143
For which (with specific reference to this period) see Donner, ‘The role of
nomads’,73–4; also, the lengthy debate in the Bulletin of the American Schools of
108 Empires, Clients, and Politics
that nomads did not follow pastoralism exclusively, but had distinct
ties to economic activities which are normally associated with settled
populations. A small number of ancient graffiti from the desert, for
example, discuss opportunistic crop-sowing carried out by nomads,
and this activity is also known from other sources.
144
Despite such promising information, attempts to demonstrate the
integration of settled and nomadic groups in the H
˙
aurān have been
largely inconclusive. For example, Macdonald has convincingly shown
that Villeneuve’s argument for a connection between the increase in the
settled population of the H
˙
aurān in the early Roman period, and the
sedentarisation of nomads, was based on tenuous evidence, since it
depended on accepting the very contestable premise that the evidence
for tribal names in the region reflected the presence of nomads.
145
Macdonald has, in fact, shown that the opposite case might be true,
arguing convincingly that the authors of the ‘Safaitic’ graffiti in question
were neither sedentary nor lived in the settled areas south of Damascus,
as has sometimes been claimed, and, in fact, probably had much more to
Oriental Research,esp.E.B.Banning,‘Peasants, pastoralists, and the pax romana’,
BASOR, 261 (1986), 25–50, and the reply from S. T. Parker, ‘Peasants, pastoralists and
pax romana: a different view’, BASOR, 265 (1987), 35–51, and Banning, ‘De Bello
Paceque’ (n. 114); also P. Mayerson, ‘Saracens and Romans, 71–9; D. Graf, ‘Rome and
the Saracens: reassessing the nomadic menace’, in T. Fahd (ed.), L’Arabie préislamique
et son environnement historique et culturel. Actes du Colloque de Strasbourg, 24–27 juin
1987 (Leiden, 1989), 341–400; and B. Shaw, ‘Fear and loathing: the nomadic menace in
Roman Africa’,inC.M.Wells(ed.),Roman Africa: The Vanier Lectures, 1980 (Ottawa,
1982), 29–50; G. Tate, ‘Le problème de la défense et du peuplement de la steppe et du
désert, dans le nord de la Syrie, entre la chute de Palmyre et la règne de Justinien’, Ann.
Arch. Syr., 42 (1996), 331–5, at 334; Khazanov, Nomads and the Outside World,1–4.
144
M. C. A. Macdonald, ‘Nomads and the H
˙
awrān in the late Hellenistic and
Roman periods: a reassessment of the epigraphic evidence’, Syria, 70 (1993), 303–413,
at 316–17; also, A. M. Khazanov, ‘Nomads in the history of the sedentary world’,in
A. M. Khazanov and A. Wink (eds.), Nomads in the Sedentary World (Richmond,
Surrey, 2001), 1–24, at 1–2; S. A. Rosen and G. Avni, ‘The edge of empire: the
archaeology of pastoral nomads in the southern Negev highlands in Late Antiquity’,
BA, 56 (1993), 189–99, at 196–7.
145
F. Villeneuve, ‘Citadins, villageois, nomades: le cas de la Provincia Arabia (II
e
–
IV
e
siècles ap. J.-C.)’, Dialogues d’histoire ancienne, 15/1 (1989), 119–140, at 134–5;
Macdonald, ‘Nomads and the H
˙
awrān’,315–16; cf. M. Sartre, ‘Tribus et clans dans le
Hawran antique’, Syria, 59 (1982), 77–97, at 88–9. Cf. too also problematically
H. I. Macadam, ‘Epigraphy and village life in southern Syria during the Roman and
early Byzantine periods’, Berytus, 31 (1983), 103–15, esp. 111: ‘these [areas where tribal
inscriptions are found] are the upland areas on the fringes of the desert, and the areas
which would be a natural choice of Bedouin in the transitory stage from nomadism to
urbanization.’
Empires, Clients, and Politics 109
do with the desert than the fertile agricultural lands of the H
˙
aurāneven
while they interacted with the population of the region.
146
Challenges remain in identifying processes of nomadic settlement
in Late Antiquity, as they do for the earlier periods discussed by
Macdonald, and it is very unclear to what degree any of the Arab
groups who were probably at the fringes of the Roman Empire
‘settled’. Shahid is a vocal proponent of the position that the Jafnids
and Ghassān completely sedentarised in Late Antiquity, but there is
absolutely no evidence for such a claim, and it is also necessary to
point out that sedentarisation can be a long-term process, and neither
irreversible, nor fixed, nor quick.
147
The simple opposite—that the
Jafnids and their ‘group’ remained wholly nomads—is also unlikely.
In fact, the most reasonable demographic description is that any
group of people at the edges of Rome’s desert frontier would have
included a mix of settled, nomadic, and semi-nomadic strata, with the
Jafnids working to extend their power and influence over the pastor-
alists of the steppe and desert regions.
148
There was of course clear
146
Macdonald, ‘Nomads and the H
˙
awrān’, 311, esp. 322: ‘the content and dis-
tribution of the Safaitic inscriptions point inescapably to the conclusion that their
authors were nomads. I can find no evidence to support the view that they lived
among the sedentaries on Jabal H
˙
awrān, even for part of the year, indeed all the
information these texts provide suggests the opposite.’ This is a strong rejection of
both J. T. Milik, ‘La tribu des Bani Amrat en Jordanie de l’époque grecque et romaine’,
ADAJ, 24 (1980), 41–54, and M. Sartre, L’Orient romain: provinces et sociétés provin-
ciales Méditerranée orientale d’Auguste aux Sévères (31 avant J.-C.–235 après J.-C.)
(Paris, 1991), esp. 333; see as well id., ‘Transhumance, économie et société de
montagne en Syrie du sud’, Ann. Arch. Syr., 41 (1997), 75–86.
147
Cf. once more Shahid, Sixth Century, ii/1, esp. 1–20, who considers Ghassānto
be completely sedentarised following (1) ‘the time [that] they crossed the Byzantine
frontier’. Shahid writes (1): ‘when the Arab poets of pre-Islamic times and later
authors identified the Ghassānid lifestyle as sedentary and not nomadic, they gave a
truthful description of it.’ Also Sixth Century ii/1, 136–40, where he criticises Fow-
den’s Barbarian Plain for suggesting that some of Ghassān may have been nomads.
For processes of sedentarisation, see, generally, Khazanov, Nomads and the Outside
World, 199–201; also, P. C. Salzman, ‘Introduction: processes of sedentarisation as
adaptation and response’ in When Nomads Settle,1–20, at 11–14; in the same volume,
R. Bulliet, ‘Sedentarisation of nomads in the seventh century: the Arabs of Basra and
Kufa’,35–47; W. G. Dalton, ‘Some considerations in the sedentarisation of nomads:
the Libyan case’, in Salzman and Galaty (eds.), Nomads in a Changing World, 139–64,
esp. 144–61. These articles provide valuable perspectives on sedentarisation as a long-
term process.
148
Cf. Genequand, ‘Some thoughts’, 78: ‘It is important to remember that all the
territory under Jafnid control was not only settled by people ...whether they were
sedentary, semi-nomads or nomads, but also by other tribes ...these independent
tribes were simply members of the confederation under a general Jafnid leadership.’
110 Empires, Clients, and Politics
motivation for the Jafnids to ensure that whatever manpower existed
in the desert came under their control, and not under that of the
Nas
rids. One particular problem in Shahid’s position (‘there is no
doubt about their sedentary nature’
149
) is that he does not mention
the possibility that different sections of the population of the peoples
led by the Jafnids would have had very different motivations for either
sedentarising or choosing an alternative path. In this scenario, the
most likely section to settle would be the Jafnids themselves, and
those most closely connected to them, as a result of their close
relationship with the Empire. Certainly the epigraphic evidence
from al-H
˙
ayyat and al-Burj leads naturally to an argument for the
sedentarisation of the elite, derived from several logical propositions.
First, that nomads do not tend to build structures of relative perma-
nence, favouring more ephemeral structures better suited to their
specific and seasonal, temporary needs; secondly, that an acknowl-
edgement of a nomadic leader in the H
˙
aurān makes less sense than to
posit one who identified more clearly with a settled lifestyle; thirdly,
that the use of Greek inscriptions suggests a desire to be associated
with the local settled population; and fourthly, that the location of the
two inscriptions favours an argument for activity within a settled area.
Of course, there is no proof of Jafnid settlement; all that can be said
is that the incentive to settle would be felt more keenly by the elites,
who stood to gain from closer association with the villages, cities, and
towns of the Empire. Furthermore, if it can be said that the Jafnid elite
were the most likely to settle, the situation for the balance of the
people under their control is very murky indeed. The indications
from John of Ephesus that they were present in some number in
the general region of the H
˙
aurān does not provide any clues as to
their mode of life, and details of the large number of settlements
connected to the Jafnids by later authors such as Hamza are very
much open to debate. In a different article, Villeneuve once again
suggested that population increase in the H
˙
aurān may be linked to the
settlement of nomads in Late Antiquity, as elements of Arab groups
moving north from Arabia chose to remain in the area.
150
The
objection to his argument based on the existence of tribal names
149
Shahid, Sixth Century, ii/1, 140.
150
Villeneuve, ‘L’économie rurale’, 117, who asks: ‘comment les villageois ont-ils
pu s’accommoder de cette poussée des nomades à l’époque byzantine?’; see too
Macdonald, ‘Nomads and the H
˙
awrān’, 315.
Empires, Clients, and Politics 111