+ All Categories
Home > Documents > D. Genequand, The New Urban Settlement at Qasr al-Hayr al-Sharqi: Components and Development in the...

D. Genequand, The New Urban Settlement at Qasr al-Hayr al-Sharqi: Components and Development in the...

Date post: 07-Feb-2023
Category:
Upload: aventicum
View: 1 times
Download: 0 times
Share this document with a friend
25
1 INTRODUCTION Amongst the so-called Umayyad castles or ‘desert castles’, a general name that in fact involves very different realities, Qasr al-Hayr al-Sharqi is one of the major monuments or group of monuments. One of the reasons for this fame is the compara- tively well preserved state of two of its structures and their impressive aspects. Another reason to see Qasr al-Hayr al-Sharqi as one of the major Umayyad sites in Bilad al-Sham, is the variety of its components and surface area. The components comprise much more than the usual square castle (qaßr), a mosque, some hydraulic features and sometimes a bath. The site’s surface area extends over more than ten square kilometres. In fact, despite being situated in a typical ‘desert castle’ environment, Qasr al-Hayr al-Sharqi corresponds to a different kind of settlement: the new urban settlement, as described some ten years ago by Alastair Northedge 1 . This paper will present and discuss the nature and interpretation of the site in the light of the mass of archaeological evidence and the very limited historical evidence 2 . 2 THE SETTING OF THE SITE Qasr al-Hayr al-Sharqi is situated some 110 km to the east-north-east of Palmyra, near the crossroads of the ancient routes linking the caravan city to al-Rusafa and al-Raqqa on the upper course of the Syrian Euphrates on one hand, and to Dayr al-Zawr, the Jazira and lower Mesopotamia on the other. The site is on the edge of a vast plain extend- ing beyond the southern slopes of the Jabal Bishri, which is the far eastern range of the Palmyrene chain. The region is an arid steppe with a mean annual rainfall of less than 140 mm. The position of the site is just downstream of the entrance into the plain of one of the major wadis of the south- western piedmont area of Jabal Bishri, the Wadi al-Suq, which has an immense catchment area. In all previous studies of the site, three main components of the Umayyad settlement have been recognised (Fig. 1). Following Oleg Grabar’s work, they are respectively called the ‘small enclo- sure’, the ‘large enclosure’ and the ‘outer enclo- sure’ 3 . They are each very different in form, as well as in function. Built in stone and baked-brick, or partially in mud-brick over a stone base, all three components are easily recognizable, even without excavations. They correspond respectively to a palace, a large monument with residential and reli- gious functions and an irrigated and cultivated area. They are, however, not the only construc- tions related to the Umayyad settlement. Indeed, several areas of the site contain mud-brick build- ings, some of which were probably of importance, but now reduced by erosion to mere contoured mounds. First of all, there are houses and perhaps other kinds of buildings extending across very large areas covering over 30 hectares to the north and, to a lesser extant, to the east and south of the small and large enclosures. These areas were called the ‘northern settlement’ and the ‘eastern settle- ment’ by Grabar, but were not otherwise investi- gated. Some 2.6 km to the south of the palace and large enclosure, there are two other monuments belonging to the same kind of enclosed structures; they are now called the ‘southern castles’. Finally, another ‘outer enclosure’ was recently recognised to the west of the first one. 3 HISTORY OF RESEARCH After a few general surveys, Albert Gabriel first properly studied Qasr al-Hayr al-Sharqi during 1 Northedge 1994. 2 I want to thank Isabelle Ruben and Alastair Northedge for reading and commenting on a preliminary version of this paper. 3 Grabar et al. 1978. The New Urban Settlement at Qasr al-Hayr al-Sharqi: Components and Development in the Early Islamic Period Denis Genequand
Transcript

1 INTRODUCTION

Amongst the so-called Umayyad castles or ‘desertcastles’, a general name that in fact involves verydifferent realities, Qasr al-Hayr al-Sharqi is one ofthe major monuments or group of monuments.One of the reasons for this fame is the compara-tively well preserved state of two of its structuresand their impressive aspects. Another reason to seeQasr al-Hayr al-Sharqi as one of the majorUmayyad sites in Bilad al-Sham, is the variety ofits components and surface area. The componentscomprise much more than the usual square castle(qaßr), a mosque, some hydraulic features andsometimes a bath. The site’s surface area extendsover more than ten square kilometres. In fact,despite being situated in a typical ‘desert castle’environment, Qasr al-Hayr al-Sharqi correspondsto a different kind of settlement: the new urbansettlement, as described some ten years ago byAlastair Northedge1. This paper will present anddiscuss the nature and interpretation of the site inthe light of the mass of archaeological evidenceand the very limited historical evidence2.

2 THE SETTING OF THE SITE

Qasr al-Hayr al-Sharqi is situated some 110 km tothe east-north-east of Palmyra, near the crossroadsof the ancient routes linking the caravan city to al-Rusafa and al-Raqqa on the upper course of theSyrian Euphrates on one hand, and to Dayr al-Zawr, the Jazira and lower Mesopotamia on theother. The site is on the edge of a vast plain extend-ing beyond the southern slopes of the Jabal Bishri,which is the far eastern range of the Palmyrenechain. The region is an arid steppe with a meanannual rainfall of less than 140 mm. The positionof the site is just downstream of the entrance intothe plain of one of the major wadis of the south-western piedmont area of Jabal Bishri, the Wadi al-Suq, which has an immense catchment area.

In all previous studies of the site, three maincomponents of the Umayyad settlement have beenrecognised (Fig. 1). Following Oleg Grabar’swork, they are respectively called the ‘small enclo-sure’, the ‘large enclosure’ and the ‘outer enclo-sure’3. They are each very different in form, as wellas in function. Built in stone and baked-brick, orpartially in mud-brick over a stone base, all threecomponents are easily recognizable, even withoutexcavations. They correspond respectively to apalace, a large monument with residential and reli-gious functions and an irrigated and cultivatedarea. They are, however, not the only construc-tions related to the Umayyad settlement. Indeed,several areas of the site contain mud-brick build-ings, some of which were probably of importance,but now reduced by erosion to mere contouredmounds. First of all, there are houses and perhapsother kinds of buildings extending across verylarge areas covering over 30 hectares to the northand, to a lesser extant, to the east and south of thesmall and large enclosures. These areas were calledthe ‘northern settlement’ and the ‘eastern settle-ment’ by Grabar, but were not otherwise investi-gated. Some 2.6 km to the south of the palace andlarge enclosure, there are two other monumentsbelonging to the same kind of enclosed structures;they are now called the ‘southern castles’. Finally,another ‘outer enclosure’ was recently recognisedto the west of the first one.

3 HISTORY OF RESEARCH

After a few general surveys, Albert Gabriel firstproperly studied Qasr al-Hayr al-Sharqi during

1 Northedge 1994.2 I want to thank Isabelle Ruben and Alastair Northedge for

reading and commenting on a preliminary version of thispaper.

3 Grabar et al. 1978.

The New Urban Settlement at Qasr al-Hayr al-Sharqi:Components and Development in the Early Islamic Period

Denis Genequand

the 1920s4. It was an architectural survey and noarchaeological excavation was done at this time.Gabriel made a sketch plan of the whole site andundertook more detailed investigations on the twomonuments usually known as the small enclosureand large enclosure. His general idea was that thesmall enclosure was a late Roman structure datedto the 5th–6th century AD. He interpreted the largeenclosure as an Umayyad construction. After him,Henry Seyrig, director of the Department ofAntiquities, carried out some investigations on thehydraulic structures5. K.A.C. Creswell completedthe documentation on the architecture6.

During the 1960s, an American mission led byGrabar studied and partially excavated the site.They concentrated their fieldwork on the smalland large enclosures and on a bath that were thenall securely attributed to the Umayyad period. Ona more general level, they also concentrated on theinterpretation of the early Islamic settlement. Thiswork resulted in an important and stimulatingmonograph7.

During the late 1990s, work was resumed byKhaled al-Asfiad, formerly director of the Palmyraoffice of the DGAM. Large areas of both the smalland large enclosures were excavated and heavilyrestored. In 2002, this became a collaborative Syri-an-Swiss project investigating Umayyad settle-ments in the Syrian steppe. The project at Qasr al-Hayr al-Sharqi is now co-directed by Walid al-Asfiad, director of the Palmyra office of theDGAM, and the present author. I will take theopportunity to thank again the Directorate Gener-al of Antiquities and Museums of Syria for wel-coming us and for their help and support, especial-ly Dr. Bassam Jamous, director general, and Dr.Michel al-Maqdissi, director of the ExcavationsService. I also want to thank the Swiss-Liechten-stein Foundation for Archaeological ResearchAbroad (SLFA, Zürich) for funding the project, aswell as researchers and participants to the project8.The fieldwork is still in progress and some of theelements and interpretations that I will present inthe following pages may change in the future.

4 THE PALACE OR SMALLENCLOSURE

The palace or small enclosure is a square building,with sides 70 m long (Fig. 2 and 3)9. It had twostoreys and is still preserved up to 11 m in places.Its external wall is reinforced at its corners by fourtower-buttresses in the shape of a three-quartercircle and by eight tower-buttresses in the shape ofa half-circle along the sides. Inside the building,rooms are organised on two storeys around a cen-tral courtyard with a portico. The plans of both

storeys are nearly identical and represent the tradi-tional layout of Umayyad castles in their groupingof rooms as apartments (bayt/buyüt). In otherwords, apartments consist of a large central roomopening onto four or five smaller ones. The func-tion of the building is controversial. Grabar saw itas a caravanserai10. But it is much more likely thatit was a palace or the caliphal residence, as suggest-ed by Jean Sauvaget and Creswell and proposedagain by Northedge11.

Recent fieldwork in the palace included a newarchitectural analysis of the standing remains andstratigraphical soundings. One of the main resultsof the architectural analysis shows that the monu-ment never witnessed partial destruction andreconstruction during the early Islamic period, aspostulated by Grabar12, but rather bears traces oftwo distinct phases, reflecting a major interrup-tion, during a single construction operation13.Stratigraphical data point to a clear shift in thekind of occupation and status of the palace duringthe 3rd century H/9th century AD, when vernacu-lar architecture was built inside it14. They alsopoint to a continuous occupation until the begin-ning of the 4th/10th century, when the monumentwas abandoned.

5 THE LARGE ENCLOSURE

The large enclosure forms a square structure withsides measuring 167 m in length (Fig. 2)15. It hasseveral elements in common with the small enclo-sure. The most obvious of these is its enclosurewall, which is also reinforced by four cornertower-buttresses in the shape of a three-quartercircle and twenty-four semi-circular tower-but-

4 Gabriel 1927.5 Seyrig 1931; Seyrig 1934.6 Creswell 1969, 522–544.7 Grabar et al. 1978.8 Researchers involved in the project are: Cyril Achard (pot-

tery, Paris), Hugo Amoroso (archaeology, Lausanne),Musab Besso (archaeozoology, DGAM/Damascus), Mar-cia Haldemann (archaeology, Lausanne), Daniel J. Hull(geophysic, University of York), Marlu Kühn (archaeo-botany, IPNA/Basel), Margaret O’Hea (glass, Universityof Adelaïde), Sophie Reynard (topographical survey, Paris),Christian de Reynier (architecture, SPMS/Neuchâtel) andJacqueline Studer (archaeozoology, MHN/Geneva).

9 Grabar et al. 1978, 15–39.10 Grabar et al. 1978, 32.11 Sauvaget 1939, 7; Creswell 1969, 538; Northedge 1994,

235–236.12 Grabar et al. 1978, 34–35.13 Genequand et al. 2006, 183–188.14 Genequand 2003b, 88–90; Genequand 2004a, 75–76. In the

following text, all the dates will be given as respectivelyH/AD.

15 Grabar et al. 1978, 40–89.

Denis Genequand262

tresses along the sides. Four gates, each oneflanked by two of these tower-buttresses, are situ-ated at the centre of each side of the enclosure withstreets leading in from them. A huge courtyardsurrounded by a portico occupies the centre of theenclosure. Twelve units of similar dimensions, butwith different plans and functions, are organisedaround this courtyard: a mosque, an industrial unit(oil presses), eight dwelling units and two units,supposedly without constructions, being openyards. An inscription, found on a stone in a pier of themosque of the large enclosure, ascribes the con-struction of a madına (literally a ‘town’) to thecaliph Hisham b. fiAbd al-Malik (H 105–125/AD724–743) in H 110/AD 728–916. It was consideredto be in a secondary position and to have been thefoundation inscription of the large enclosure. Itgreatly helped in forming the interpretation of thisenclosure not as an urban centre in the traditionalsense of the word but as a “small, private, and aris-tocratic settlement with a minimal number offunctions”17.

Recently, the mosque in the south-eastern cor-ner was entirely cleared by the Palmyra office ofthe DGAM, and subsequently restored. Few newelements were added to the general layout of themosque and the plan remains basically the oneproposed by Grabar following his soundings18. Itis important to note that the plan of the mosque istypical of Umayyad urban congregationalmosques. One interesting new detail is the pres-ence of a lateral secondary mi˛rb. It does notbelong to the original construction and was addedat a later stage. Its addition requested the cuttingof the finely dressed stones of the qibla wall. With-out stratigraphical elements and due to the restora-tion of the qibla wall, it is quite difficult to pro-pose a precise dating for its addition. I think,nevertheless, that it still belongs to the early Islam-ic period, i.e. the 2nd/8th or 3rd/9th centuries, andnot to the medieval reoccupation of the site, dur-ing which the Umayyad mosque was used fordwellings and during which a new congregationalmosque was built between the palace and largeenclosure.

Amongst other elements brought to light bythe clearance of the mosque are a few details on thearchitectural decoration and subdivision of the dif-ferent naves. All the architectonic elements are ofRoman date. They were brought from Palmyra, asclearly demonstrated by the Egyptian granitecolumns, the typical yellow limestone used for thebases and capitals and the discovery of the secondpart of a Greek inscription being a dedication tothe emperors Marcus Aurelius and Lucius Verus.

In 2003, a sounding was made in the largeenclosure to define the stratigraphical sequence as

much as possible19. The place chosen was thesouth-western corner, where the deposits seemedthe deepest. It was one of the areas supposed tohave been left free of construction, thought to becourtyards belonging to the neighbouring residen-tial units. Apart from stratigraphical data pertain-ing mainly to the medieval occupation, the sound-ing provided a few elements showing, with someconfidence, that the area was built over in the orig-inal plan. The rather limited extent of the excava-tion did not provide any further detail about theplan of the unit. Nevertheless, it may explain someof the structures and walls uncovered by theAmerican mission and the difficulty with whichthey explained them. These are the few long roomsbrought to light in 1971 in unit VIII. Because of aslightly different style of construction, i.e. twocourses of stone for the base of the walls instead ofone, they were interpreted as later constructions20.Their regularity within the general layout of theenclosure, as well as the data provided by therecent sounding, strongly suggest that the south-western corner of the large enclosure simply com-prised another built unit with a plan similar to theothers. A hypothesis, which needs to be confirmedby further fieldwork, may be that the large roomsgrouped in an apartment (a Syrian bayt) in theother units were on the west, the two long, narrowrooms also found in each of the other units beingon the east. The sounding also suggested that theseoriginal constructions were violently destroyed atsome point during the late 2nd/8th to early 3rd/9th

century (deliberate destruction or earthquake?).They were followed by some cruder structuresbuilt on the top of them with reused materials andassociated with 3rd/9th to early 4th/10th centurymaterial.

Architectural analysis of the standing parts ofthe large enclosure’s rampart showed that, like thepalace, it was built in two distinct phases within asingle operation21.

Grabar’s excavations pointed out building orrebuilding activities – particularly stucco decora-tions – during the second half of the 2nd/8th centu-ry in the unit situated west of the mosque (unitVI)22. This rich stucco decoration, unparalleled inother residential units and dated stylistically to thesecond half of the 2nd/8th century, is applied tosome doors and windows, but it is not evidencethat the structure was completed in the Abbasid

16 Grabar et al. 1978, 12. 191 with older references.17 Grabar et al. 1978, 79–81.18 Grabar et al. 1978, 46–51.19 Genequand 2004a, 71–75. 79.20 Grabar et al. 1978, 64.21 Genequand et al. 2006, 183–188.22 Grabar et al. 1978, 67–68. 148–149.

The New Urban Settlement at Qasr al-Hayr al-Sharqi 263

period. However, it is a good indicator that theunit kept a higher status or had a more importantowner after the fall of the Umayyad dynasty.Proper stratigraphical and architectural datawould be welcome to assess the detailed evolutionand history of this unit. The same unit VI, whichdiffers slightly in plan from the others and has acloser connection with the mosque, was interpretedby Grabar as having an official or administrativefunction and being a dr al-imra (house of thegovernment)23. One must keep in mind that noneof these units was completely excavated and that itis all a matter of discussing hypothetical recon-structions. In fact, its plan is not so different fromthe others and, while not completely rejecting thepossibility of a dr al-imra, it seems that it ismuch more likely that all the units with more orless similar plans were intended to have the sameresidential function as elite housing. It wouldmake more sense to look for the official or admin-istrative function in the neighbouring palace.

As for the small enclosure, this structure showssigns of a continuous occupation until the begin-ning of the 4th/10th century, after which it wasabandoned. But there are indications of a shift inthe kind of occupation and status by the 3rd/9th

century and vernacular architecture seems to takeplace everywhere in what had formerly been elitehousing.

6 THE BATH

A bath is situated immediately to the north of thepalace and large enclosure24. It includes a largeroom with a basilical plan, which has no true utili-ty for the functioning of the bath. It could havebeen conceived as a reception hall, an element thatis missing in both the palace and the large enclo-sure. A reception hall linked to a bath is anarrangement which is found in other Umayyadcastles or palatial complexes such of QusayrfiAmra, Hammam al-Sarakh or, in a more magnifi-cent manner, Khirbat al-Mafjar25.

7 THE NORTHERN SETTLEMENT

As indicated by its name, the northern settlementextends to the north of the palace and large enclo-sure. It is a large area covered by mud-brick struc-tures now reduced to a series of mounds withsmooth contour lines. Together with the lessextensive eastern settlement and a few other isolatedbuildings to the south of the palace, it covers atotal surface area of over 30 hectares. None ofthese structures or areas was properly studied dur-ing the 1960s, except for a surface collection of

sherds in a transect of the northern settlement anda few very small soundings26. These led to the con-clusion that the area was occupied during the earlyIslamic period.

Given the seeming importance of these mud-brick structures for understanding the early Islam-ic period at Qasr al-Hayr al-Sharqi, the northernsettlement has been the subject of detailed studysince 2004. The fieldwork involves soundings,excavations and geophysical survey27. The prelimi-nary hypothesis was that the northern settlementcorresponded to the vernacular housing facilitiesof the site during the Umayyad and early Abbasidperiods. Indeed, after three seasons of work there,it seems to be fully confirmed.

The geophysical survey (magnetometry andresistivity) has covered a certain number of housesand will continue to cover the largest part of thearea. Three medium to large-sized mounds, eachexpected to be the remains of a single house, werealso chosen for soundings and excavations, whichshould continue over the next few years andextend to other structures.

7.1 FIELD A

In Field A, excavations revealed a large house ofabout 38 x 30 m organised around two irregularcourtyards (Fig. 4 and 5). This house had threemain phases of construction. The original struc-ture was a square house (22.50 x 21 m) built en-tirely of mud-brick. It comprised three wings,totalling ten square rooms around a courtyard,and a smaller protruding room sheltering a subter-ranean cistern for storing rainwater. In the secondphase, a new wing with five large rooms, anothercourtyard and smaller structures or rooms wereadded to the east of the house. These were all builtin mud-brick over a stone base. The third phasesaw the adjunction of two more rooms flanking avaulted entrance corridor to the east of the house,as well as modifications to the previous structures.This last phase was characterized by mud-brickmasonries, of which the inside facing of the base ofwalls was reinforced by stone orthostats. At itsfullest development, the house comprised at leasttwenty rooms of different sizes and functions.

Two architectural particularities need to beemphasized: all the rooms were square and all the

23 Grabar et al. 1978, 70–71.24 Grabar et al. 1978, 90–97.25 Almagro et al. 1978; Hamilton 1959.26 Grabar et al. 1978, 105–106.27 Genequand 2005b, 152–162; Genequand et al. 2006,

163–175; Genequand et al. 2007, 126–130. 135–144.

Denis Genequand264

dividing walls or walls common to two roomswere either larger than the others or doubled.These two points clearly indicate that earthendomes covered all the rooms. This is a traditionalbuilding technique of the steppe areas of centraland northern Syria up to the present. Walls andfloors were coated with lime plaster.

Fireplaces, many bread ovens (tannür and†bün) and water basins were the most frequentlyfound installations inside the house. The cisternfrom the first phase consisted of a 4 m deep shaftcut in the bedrock and covered by a massive andraised coping. The second phase included anotherrock-cut cistern fed by a small canal leading from anearby wadi. One of the rooms flanking theentrance corridor was a bathroom and includes araised washbasin. Some industrial installations,possibly parts of a wine or oil press, are related tothe first and second phases.

The ceramics retrieved from this house indicatean occupation in the 2nd/8th and the first half of the3rd/9th centuries. Of course, there is not muchceramic that is numerically significant belongingto the very first occupation of the house and itsconstruction date is perhaps more difficult toassess. Nevertheless, some early types, like thenorth Syrian painted amphorae, point to theUmayyad period, to the first half of the 2nd/8th

century. By contrast, there is a much clearer viewof the assemblage of the last occupation and aban-donment levels. The latter assemblage includes thefirst series of polychrome glazed ceramic (yellow-glazed-family) and some types belonging to theso-called ‘Samarra√ horizon’ (white glaze with blueor green decoration), but also relief ware, ‘brittleware’ cooking pots with plain thickened rim andtriangular ledge handles and new types ofamphorae that no longer belong to the late Romantypes (Fig. 6). Consensus has not been reached onthe dating of the introduction of the polychromeglazed ceramic and the different types of the‘Samarra√ horizon’: it is the period of residence ofthe Abbasid caliph Harun al-Rashid in al-Raqqafor some (180–192/796–808)28; slightly later,roughly around the beginning of the period ofSamarra’ for the others (ca. 215–225/830–840)29.This means that the house was abandoned by thevery beginning of the 3rd/9th century at the earliestand by the mid 3rd/9th century at the latest. Anoccupation during the whole first third of the3rd/9th century therefore seems plausible.

7.2 FIELD B

Soundings and geophysical survey in Field Brevealed another house also organised around acentral courtyard and entirely built in mud-brick

(Fig. 7). It comprised four wings and measuresabout 35 x 30 m. The house developed through atleast two main architectural phases and probablyhad up to twelve rooms. The gateway givingentrance into the courtyard is situated in the east-ern wing and took the form of a protruding porch(Fig. 8). The only installations brought to lightwere fire places and bread ovens (tannür). Theceramic evidence shows that this house was in usefor a rather short period in the 2nd/8th century andthat it was abandoned long before the house inField A. Polychrome glazed ceramic is completelyabsent from the assemblage of the last occupationand abandonment layers. On the other hand, thisassemblage still contains material in the lateAntique tradition, like the Palestinian Late RomanAmphora (Carthage LRA) 5/6 and Umayyad‘brittle ware’ types (cooking pots with verticalneck and triangular rim, bowl with doubleincurved rim). An abandonment by the mid2nd/8th century or slightly later seems likely.

7.3 FIELD C

In Field C, a large sounding was excavated in athird house comprising three wings organised inan U-shape plan around a courtyard (Fig. 9 and10). The house measures about 30 x 20 m and wasbuilt entirely of mud-brick. Like the others, itdeveloped in several phases with the addition offurther rooms in spaces that originally belonged tothe courtyard. The sounding covered part of thesouthern wing and part of the courtyard. One ofthe fully excavated rooms in the centre of the winghad large storage facilities, likely to have been usedas silos for grain or straw. In the courtyard, therewere two latrines and a kitchen delimited by thin-ner walls. Latrines are vertical drains covered by asmall corbelled dome with a central rectangularopening. The neighbouring kitchen includes twodouble fire places and a bread oven (tannür). Likein Field B, ceramic evidence points to the use ofthe house during the 2nd/8th century and an aban-donment at some point during the second half ofthe same century.

Excavations and surveys in the northern settle-ment brought to light a very important componentof Umayyad and early Abbasid Qasr al-Hayr al-Sharqi, namely the vernacular housing of peoplewho lived there and who were not necessarily partof the elite. The architecture is simple and well

28 Watson 1999, 86.29 Northedge 2001.

The New Urban Settlement at Qasr al-Hayr al-Sharqi 265

adapted to the environment. With its many archi-tectural phases and some small industrial installa-tions, it also reflects the facts that the northernsettlement was much more than temporaryhousing for the builders of the palace and the otherlarge structures of the site, but was in fact athriving settlement for a period during the 2nd/8th

and early 3rd/9th centuries.The material culture from the houses includes

luxury goods, such as a wide range of early poly-chrome glazed ceramic, a large quantity of glass(also lustre painted and scratched) and importedamphorae (LRA 5/6 from Palestine). This is agood indicator of the level of prosperity enjoyedby the inhabitants of the northern settlement andnot only by those of the palace and aristocratichousing of the large enclosure.

Finally, excavations and studies in the northernsettlement indicate that Umayyad and earlyAbbasid Qasr al-Hayr al-Sharqi was a much largersettlement than previously thought and had animportant core of permanently settled people.

8 THE SOUTHERN CASTLES

Amongst the new elements revealed by the recentwork at Qasr al-Hayr al-Sharqi, the so-calledsouthern castles occupy a prominent position.They are situated 2.6 km south of the palace andlarge enclosure, next to one of the gates of theouter enclosures (Fig. 1). They consist of twosquare enclosures with a central courtyard and arebuilt of mud-brick. They were noticed by theAmerican mission and interpreted as mere servicebuildings or later caravanserais30. A substantialpart of their mud-brick walls is still visible on thesurface. They were first surveyed to get prelimi-nary plans31 and excavations were then conductedin both structures32. The study was completed bya geophysical survey of the southernmost struc-ture33.

The two structures have very similar plans andface each other in the same manner as the palaceand the large enclosure (Fig. 11). Both are square,about 65 m long on each side, with four cornertower-buttresses and eight intermediary onesalong the sides and flanking the gateways. Insideboth structures rooms are organised around thecentral courtyard, but very differently one fromthe other.

In the northern structure, each of the fourwings has two rows of little square rooms (4 x 4 m)(Fig. 12). The gateway and a rectangular entrancecorridor interrupt one of the wings. The structurecounts eighty-six rooms. With the notable excep-tion of rooms in the four corners, which involve amore elaborated regrouping and internal circula-

tion, all the other rooms appear to be connectedonly in pairs, in line from the courtyard. In thecorners, some of the rooms have lateral doors toallow access to the ones abutting the enclosurewall. This results in the grouping together of sixrooms. There is no trace of a portico in the court-yard (no bases or postholes).

The building technique is close to the one usedfor the other Umayyad monuments in Qasr al-Hayr al-Sharqi, i.e. two or three courses of welldressed blocks of local limestone forming both thefoundation and the base of the walls. The rest ofthe elevation is built entirely in mud-brick (bricksof 40/42 x 40/42 x 8/9 cm). An interesting archi-tectural element of the northern structure is thateach square room was covered by an earthen domeand not by a flat roof. Where the walls are suffi-ciently well preserved, squinches insuring thetransition from the square to the round plan werefound (Fig. 13). This roofing system, here verywell preserved, is similar to that found in the hous-es of the northern settlement. It was an as yetunknown architectural feature of early IslamicSyria.

The results of the excavations point toward adomestic function and domestic related activitiesfor the last occupation before the abandonment ofthe structure. The trench in the rooms to the eastof the entrance displayed, amongst other elements,the presence of a slightly raised bench in the firstroom, small storage installations in the second(pits for jars and small silos), while there was abread oven beside the door in the courtyard. Thisis quite typical of domestic structures. Anothertrench contained a simpler oven in the courtyardand evidence of use of the rooms (damage andrefurbishing of the floors). It may indicate thatthey were only used for storage. The entrance cor-ridor had lateral benches.

Apart from architectural data, the excavation inthe northern structure provided some valuablechronological data. If it seems quite clear thatthese two monuments may have been built in rela-tion to the construction of the palace, the large andthe outer enclosures at the beginning of the secondquarter of the 2nd/8th century, they were only usedfor a short period before being abandoned at somepoint in the second half of the 2nd/8th century. Thepottery associated with the abandonment layersstill displays the characteristics of the Umayyad toearly Abbasid phase and contains small Palestinian

30 Grabar et al. 1978, 103.31 Genequand 2003a, 47–50.32 Genequand 2004a, 78–84; Genequand 2005b, 146–152.33 Genequand et al. 2007, 138–139.

Denis Genequand266

amphorae derived from the LRA 5/6 andUmayyad types of ‘brittle ware’. Polychromeglazed ceramic is completely absent.

The southern structure presents the same gene-ral layout with four wings around a central court-yard, but is very different in its internal organisa-tion (Fig. 11). Three of the four wings consist ofonly one very long room, while the fourth one issubdivided in three rooms, including the entrancecorridor from the gateway (Fig. 14). The westernand eastern wings are slightly larger than the othersand have a stone foundation separating them intwo equal parts in length (Fig. 15). It is very likelythat this foundation was intended as a base for pil-lars supporting a flat truss for the roofing. Whererecognised, the long rooms are accessible from thecourtyard by two wide doors (2.10 m). Excava-tions only brought to light benches in the entrancecorridor and in the room adjacent to it, as well asash layers and fire places elsewhere. Some installa-tions were also related to a secondary occupation,which involved at least partial closing of the gate-way and the addition of another small room insidethe entrance corridor.

The pottery retrieved from the southern struc-ture was scanty and points to an abandonment atthe same time as the northern structure in the sec-ond half of the 2nd/8th century.

When work began in the southern castles, theworking hypothesis was that they were probablytwo more aristocratic residences34. This was basedon comparisons with other Umayyad sites wherethere are several other castle-like structures besidethe main palace or residence, such as Umm al-Walid, Khan al-Zabib and Jabal Says35. After exca-vations and with a much clearer understanding ofthe plans of both structures, such an interpretationis, of course, no longer relevant.

It is hard to find a good late Antique or earlyIslamic parallel for the northern structure, but itsplan may help to define its function. The eighty sixsquare rooms, all similar in size and shape, areunlikely to have been conceived and initially usedfor a domestic purpose, even if archaeology pointsto such a use in some rooms for the phase justbefore abandonment. There are no subdivisions inreal apartments and there were no facilities associ-ated with daily life, such as latrines, bathrooms orwater supply, other than a few installations devot-ed to the cooking of food. The most likely inter-pretation seems to be to consider these rooms asstorage space, easily accessible from the centralcourtyard. This would explain why most roomshad no installations at all inside and showed fewsigns of occupation apart from damage to the plas-tering of the floors and walls. Therefore, the wholestructure was probably a storehouse.

The southern structure, with its very longrooms occupying a whole wing, has a very goodarchitectural parallel at Qasr al-Hayr al-Gharbi, acontemporary palatial complex also built byHisham b. fiAbd al-Malik in H 109/AD 72736.About 1 km to the north of the palace, there isanother square enclosure known as the ‘khn’,which was the building bearing the inscription ofHisham b. fiAbd al-Malik on its lintel. Despitebeing slightly smaller (55 m per side), the ‘khn’displays exactly the same plan in terms of internalorganisation (but includes a portico and has notower-buttresses). It also has two elements abuttingits main façade: a drinking trough and a smallmosque. It is quite clear that the southern structureunder discussion and the ‘khn’ at Qasr al-Hayr al-Gharbi had exactly the same role. Given their longsrooms with wide doors and the presence of adrinking trough at Qasr al-Hayr al-Gharbi, it islikely that they were stables for animals. Someother structures with rather similar internal planswere also documented elsewhere. Examples areKhirbat al-Khan in the Syrian Jazira37, Qasr Khab-baz and Qasr Amij along the road to Hit in Iraq38,but none of these has yet been properly investigat-ed and dated. They are now usually interpreted assmall road stations dating to the Islamic period,although some of them may be pre-Islamic. InSamarra√, very long rooms are interpreted as sta-bles, although differently organised39.

The southern castles, then, would grouptogether in the same place a storehouse and stablesand could thus be seen as structures to do with thelocal economy. They were certainly closely con-nected to the activities pertaining to the outerenclosure, as shown by their proximity to one ofits gates. In this case, it is possible to see the south-ern castles as storage facilities for agricultural pro-duction and stables for the animals used either onthe site for helping in agricultural activities or forthe transportation of the products elsewhere afterharvesting or after transformation. Given the gen-eral context of these structures in Qasr al-Hayr al-Sharqi, I do not think they should be interpret-ed as caravanserais in the medieval sense of theterm – i.e. overnight relays on the main traderoads, for which there is not much evidence duringthe early Islamic period – but really as buildingsplanned and built to play a role within the agricul-tural production system of the site.

34 Genequand 2003a, 49–50; Genequand 2005a, 353.35 Bujard – Genequand 2001; Brisch 1963; Brisch 1965.36 Schlumberger 1986.37 Poidebard 1934, Pl. 150.38 Gregory – Kennedy 1985, 185–199.39 Northedge 2005, 213 Fig. 95.

The New Urban Settlement at Qasr al-Hayr al-Sharqi 267

9 WATER SYSTEM ANDWATER MILL

Qasr al-Hayr al-Sharqi is situated in an area withless than 140 mm mean annual rainfall and there isno spring on the site. Therefore, it had to rely on acomplex water system based on distant perennialsprings and rainwater for its water supply. Themajor element of the water system is an aqueductbringing water from two distant undergroundperennial springs situated in al-Kawm and Ummal-Tlal, some 25 km to the north-west (Fig. 16)40.These are sulphur springs, under pressure, fed byartesian water tables; at their source, they are char-acterized by important mounds of gypseousdeposits. From there, two qants (undergroundchannels dug in the sediments and with regularlyspaced vertical shafts for digging and maintenance)brought the water, following the courses of theWadi al-Murr and Wadi al-Fatayya until they jointogether at about one-third of their way and con-tinue then as only one qant in the Wadi al-Suq. Ina few places and for rather short distances – cross-ing of wadis or depressions – an aqueduct built instone masonry replaces the sediment-cut channel.Two kilometres before Qasr al-Hayr al-Sharqi, anunderground aqueduct entirely built in stone andbarrel vaulted replaces the qant. The stone-builtaqueduct has irregularly spaced openings formaintenance41.

The aqueduct arrives at Qasr al-Hayr al-Sharqifrom the north-west, from the mouth of Wadi al-Suq. It then changes direction, bending south tocross the northern settlement. The latter being sit-uated on a small eminence above the wadi, it wasnecessary to rely again on the qant technologyfor a few hundred metres, until it reaches a largeopen-air reservoir (birka)42. From the reservoirseveral channels brought the spring water to thepalace, the bath and the large enclosure. Theredoes not appear to be a branch from the aqueductinto the northern settlement.

In the northern settlement, according to thedata presently at hand, the water supply of thehouses relied on rainwater and floodwater. Rain-water was collected on the roofs and stored in sub-terranean cisterns, while floodwater was broughtinto one of the houses from the Wadi al-Suqthrough a small channel cut in the bedrock andprobably originating from a diversion on the bankof the wadi. Floodwater was also used for irriga-tion and agriculture in the outer enclosures,according to a system described below.

From what precedes, it is interesting to note thatwater was used very differently in Qasr al-Hayr al-Sharqi depending on its origin. Spring water, thebest quality but also the most expensive to procurethrough a 25 km long aqueduct, was only destined

to the palace, the related bath and the large enclo-sure. The northern settlement had to rely on rain-water, easy to procure but irregularly distributed,and to a lesser extant on floodwater from Wadi al-Suq. Irrigation for agriculture, of course, reliedon rainwater but also heavily on floodwater.

Just north of the site, at the mouth of Wadi al-Suq, a water mill was excavated43. The mill wasinstalled directly on the aqueduct and was com-posed of a lower chamber for the wheel and anupper milling chamber (Fig. 17). Water used forthe rotation of the wheel arrived through a secon-dary aqueduct, which apparently branched off themain one a few hundred metres upstream andremained at a constant altitude. This was intendedto create an oblique waterfall high enough to pro-vide waterpower to rotate a small horizontalwheel. The latter was positioned in a recessed areaalong the channel of the aqueduct and bothstreams of water joined together just downstream.A system of direct transmission moved the millitself, which rested on the floor of the upper room.Only fragments of several flat to slightly conicalmillstones were found. These are not very thickand have a diameter of about 1 m, a shape and sizetypical of water mills.

The mill was not part of the plan of the originalaqueduct, but was added at a later stage, probablyquite soon after completion. Afterwards, a fewmore rooms and installations were added aroundit. The mill and surrounding structures remainedin use until the beginning of the 3rd/9th century.Parts of the mill and aqueducts were destroyedand modified during the medieval period, whenthe main aqueduct was cleaned and restored.

Schlumberger excavated a water mill contem-poraneous with this one at Qasr al-Hayr al-Ghar-bi44. It was technically comparable (oblique waterfall and horizontal wheel) and had a very similarposition on a secondary aqueduct along the mainone. Another presumably Umayyad water millwith rather similar characteristics was also discov-ered recently in Mafian, in southern Jordan45.

10 OUTER ENCLOSURES

Two of the most striking features of the site are theouter enclosures, which encircle two areas cover-

40 Genequand et al. 2006, 179–182.41 Grabar et al. 1978, 106–107.42 Grabar et al. 1978, 107.43 Genequand et al. 2006, 175–179; Genequand et al. 2007,

131–135.44 Schlumberger 1986, 4.45 Genequand 2003c, 28.

Denis Genequand268

ing respectively over 7 and 2.2 square kilometres.The largest one was partially studied by the Amer-ican mission46. It comprises a 15 km long wall,built in mud-brick over a stone base and orna-mented by semi-circular buttresses regularlyspaced on both sides (Fig. 1 and 18). At its north-ern and southern extremities, there are two sluicessystems. The first of these can be described as adam allowing the control, diversion, concentrationand storage of floodwater from the Wadi al-Suq.Its main purpose was to divert the floodwater intotwo canals or aqueducts departing from itsextremities and conducting water in the two outerenclosures to be used in a controlled manner. Thesecond of these permitted the disposal of the sur-plus water in case of heavy flood, after it haspassed throughout the whole enclosure (Fig. 19).Five gates, each one with a different plan, allowedentry into this outer enclosure which has beeninterpreted as having an agricultural function,probably mainly connected to animal husbandry47.Recent findings, in particular new irrigationdevices, as well as parallels at other sites such asQasr al-Hayr al-Gharbi or Mafian48, point towardsit being something much more orientated towardscultivation.

Remains of a second similarly enclosed and irri-gated enclosure extending to the south-west of thesite were studied in 200449. Walls departing fromroughly south of the large enclosure can be fol-lowed for 2.5 km. In a few places, buttresses orother structures are visible against the walls. Anaqueduct coming from the sluices system in theWadi al-Suq enters the enclosure and has many sec-ondary offshoots, followed by earthen channels.This indicates that the enclosure was also intendedto be irrigated and therefore was meant to be culti-vated. Both walls and the canal stop abruptly afterrespectively 2 and 2.5 km and it seem that thewhole structure was never completed.

The fact that both outer enclosures were devot-ed to cultivation, led to the question of what kindof cultivation was concerned: agricultural produc-tion or pleasure gardens? There are several ele-ments to the answer. First, structures like the olivepresses in the large enclosure or the water mill inWadi al-Suq can be interpreted as proof that someplants – olive trees and cereals – were cultivated onthe site. The second element adding to the proof ofagricultural production is given by archaeobotany.Samples were systematically collected from well-stratified contexts in the excavations in the north-ern settlement50. The preliminary results attestwithout doubt to the cultivation on the site of bar-ley and naked wheat and probably also of lentil,millet, vine, olive tree and fig tree. Other species,such as sesame, dates, coriander, walnut and hazel

nut, which are also attested but require a lot ofwater or better soils, were probably importedfrom the Euphrates valley or better rain-fed areasof western Syria. The last elements of the answerare the environmental conditions. Whether orient-ed towards agricultural production or pleasuregardens, the limited amount of water and the poorquality of the soils would only allow species withparticular drought resistance to be cultivated andthese are precisely those attested by archaeologyand archaeobotany. Even if the builders of Qasr al-Hayr al-Sharqi wanted only to have some sortof green paradise around the monuments, thechoice of plants was limited. Traditional agricul-tural species were much easier to grow efficientlyin arid conditions that delicate exotic species.

11 EXCURSUS: THE MEDIEVALSETTLEMENT

Finally, mention should be made here that the site,after a period of abandonment from the beginningof the 4th/10th century onwards, was reoccupiedaround the middle or second half of the 6th/12th

century, in the Ayyubid period. This reoccupationgrew enough to form a small town covering thewhole surface of both, the palace and the largeenclosure as well as the space between them withnew houses51. At the beginning of the 7th/13th cen-tury, a large mosque, with a square minaret and acemetery in its courtyard, was built in the newlywalled area between the enclosures (Fig. 3)52. It isperhaps worth stating again that the tower inbetween the palace and the large enclosure istherefore not an early Islamic structure at all,despite still often being quoted in the literature asone of the possible early minarets. To allow thecreation of this new town during the medievalperiod, the aqueduct coming from al-Kawm wascleaned, restored and put into use once again.Extensive remains of gardens or enclosed cultivat-ed areas related to the medieval settlement stillexist all around the palace and large enclosure.

12 DISCUSSION

This short survey of the different components ofQasr al-Hayr al-Sharqi shows that they were more

46 Grabar et al. 1978, 98–103.47 Grabar et al. 1978, 103–104.48 Schlumberger 1986; Genequand 2003c.49 Genequand 2005b, 145.50 Genequand et al. 2006, 188–200.51 Grabar et al. 1978, 36–39. 81–84; Genequand 2005a,

355–357.52 Genequand 2003b, 72–87; Genequand 2004–2005.

The New Urban Settlement at Qasr al-Hayr al-Sharqi 269

numerous and important than previously thoughtand add much to the understanding of how the sitefunctioned during the Umayyad and earlyAbbasid period. The early date of abandonment ofthe southern castles and fields B and C in thenorthern settlement points to some changes afteror possibly as a consequence of the fall of theUmayyad dynasty in H 133/AD 750. It must behypothesised that most of the peripheral struc-tures of Qasr al-Hayr al-Sharqi were abandoned inthe course of the second half of the 2nd/8th century,with some exceptions like the house in field A,where occupation continued into the first half ofthe 3rd/9th century. Only the palace and the largeenclosure were continuously settled – with a shiftin status – until the beginning of the 4th/10th cen-tury. A possible explanation is the progressiveabandonment of an isolated site by the elite andthe double phenomenon involving a contraction ofthe rest of the population and its progressiveregrouping inside the walled structures.

12.1 TEXTUAL SOURCES

Despite the huge archaeological evidence, Qasr al-Hayr al-Sharqi almost completely lacks anytextual sources for the Umayyad period and moregenerally for the early Islamic period. It must beremembered that even its ancient name is not cer-tain. It was most probably fiUrd/fiAriza, a namethat derives from the classical Oresa, name of theclosest Antique and late Antique settlement (mod-ern al-Tayyiba, 15 km to the west)53. But it is stillpossible that, at the time of its foundation, Qasr al-Hayr al-Sharqi had another name.

The only putative textual source referring toQasr al-Hayr al-Sharqi during the Umayyad peri-od is the mention of a site called al-Zaytuna, whichbelonged to Hisham b. fiAbd al-Malik and wherehe was staying until he moved to al-Rusafa whenhe became caliph54. The toponym al-Zaytuna isfrequent and many different opinions have beenexpressed on its location. As pointed out byGrabar, this identification cannot be completelyexcluded55. But it should be noted that the sourcesdo not provide any more details about al-Zaytuna,other than that is was some sort of small ruralestate related to olive growing and including ahouse (manzil). Anyway, if this identification wereshown to be correct – which is far from being thecase – it is clear that the site did not contain all itscomponents at that time, as the foundationinscription is dated to H 110/AD 728–9.

All other sources are from the 4th/10th centuryonwards, with the possible exceptions of a fewpassing mentions without details in 3rd/9th centurygeographical works56. Therefore, early Islamic

Qasr al-Hayr al-Sharqi is not a site for which tex-tual sources are of primordial value for interpret-ing the archaeological remains.

12.2 EPIGRAPHY

Despite the poorness of textual sources, Qasr al-Hayr al-Sharqi is one of the very few Umayyadsites, along with Qasr al-Hayr al-Gharbi and Qasral-Muwaqqar, where an inscription has beenfound57. It is moreover a rather rich inscription, asit provides the name of the builder, the date ofconstruction, the kind of the structure, the nameof the person responsible for the construction andthe origin of the workers involved in the construc-tion (Fig. 20)58:

“In the name of God, the Compassionate, theMerciful. There is no God but God, One, withoutAssociate. Muhammad is the Prophet of God. Theservant of God Hisham, Commander of the Faith-ful, has ordered the construction (making) of thismadına, and this was done by the people of Homsunder the direction of Sulayman ibn fiUbayd. Inthe year one hundred and ten [AH] (728–9 AD).”

The conditions of discovery of the inscriptionand a much later publication pose some problemsof interpretation. They do not concern the read-ing, which, I think, is very clear, but deal firstlywith its original position; secondly with the inter-pretation of its content due to its original position;and thirdly raise some methodological problemspertaining to the identification often madebetween archaeological facts and written evidence.

The inscription was found in 1808 by Jean-Baptiste Rousseau, a consul of France in Aleppo,at the top of one of the pillars of the mosque in thesouth-eastern corner of the large enclosure59. Itwas one of the western pillars separating the court-yard from the prayer hall, a pillar that was less wellpreserved than the others. It was the central west-ern pillar also connected to the courtyard’s porti-co60. It is not clear which direction the inscriptionwas facing. The inscription was removed byRousseau and brought to Aleppo, where it waslost after having been drawn. The inscription wasfinally published almost a century later by Charles

53 Grabar 1970; Grabar et al. 1978, 10–11.54 Al-Tabari 1879–1901, II, 1466–7; al-Baladhuri 1866,

179–180; Sack 1996, 49–51. 134. 155.55 Grabar et al. 1978, 13.56 Grabar 1970; Grabar et al. 1978, 11.57 Schlumberger 1986, 1; Hamilton 1946.58 Clermont-Ganneau 1900, 285–292; Combe – Sauvaget –

Wiet 1931, n° 28; Grabar et al. 1978, 191.59 Rousseau 1899, 146–154.60 Rousseau 1899, 151; Gabriel 1927, 321.

Denis Genequand270

Clermont-Ganneau, on the basis of Rousseau’sdrawing61.

There are two possibilities for the originalposition of the inscription. It may be that it wasfound by Rousseau in its original position or thatit was in a reused position. The latter solution wassuggested, amongst others, by Gabriel but finallyrejected62. Nevertheless, arguing that it wasGabriel’s conclusion, Grabar then adopted thereuse solution and it was apparently widely fol-lowed. Grabar proposed that it was perhaps origi-nally placed over one of the gateways of the largeenclosure and moved later, possibly during themedieval reoccupation of the site63. The interpre-tation proposed by Grabar was that it was thefoundation inscription of the large enclosure only,a madına being not a town (literal translation ofmadına) but some sort of enclosed “small, privateand aristocratic settlement with a minimal numberof functions”64. But if the inscription was reused,there is no reason to think that it was necessarilyplaced over a gateway, nor even that it originatedfrom the large enclosure. It may have come fromelsewhere on the site, i.e. not related only to thelarge enclosure. It is worth noting that none of theupper parts (tympanums) of the gateways of thelarge enclosure shows a place for an inscription,but rather for large stucco panels, or perhaps tiledecorations. Moreover, there is actually no evi-dence that parts of the mosque arcades wererebuilt during the medieval period or earlier.

On the other hand, there is in fact no evidenceat all that the inscription was in a reused position.It is hard to imagine that it was first included inanother place or monument, and then moved, afteronly a few years or decades, into the mosque. Iteven seems impossible after the fall of theUmayyad dynasty. In conclusion, it is much morelikely that it was found in its original position.

Secondly, the mosque, perhaps along with thebath, was apparently one of the very few struc-tures or monuments of Qasr al-Hayr al-Sharqithat had a public status and was accessible to any-body. The fact that the mosque was not onlyaccessible through two doors from inside the largeenclosure, but also from outside, through a thirdone giving a direct access to the courtyard, empha-sizes this public aspect. It was not a privatemosque for the community living in the largeenclosure, but a congregational mosque for theMuslim community of the whole site. Its architec-ture and plan relate without doubt to the series ofthe Umayyad urban congregational mosques, suchas the mosques of Jarash65, fiAmman downtownand citadel66 or al-Rusafa67, and not to the smallmosques usually found beside Umayyad resi-dences or palaces. This is a very clear indicator ofthe site’s status. Then, the presence of the founda-

tion inscription of the madına in the only placedevoted to the community of the whole site, alongwith the presence of a real urban congregationalmosque, give some weight to the idea that themadına was not only a single monument in Qasral-Hayr al-Sharqi, but represented the whole sitewith all its components. I will come back to thispoint shortly.

The third problem raised by the interpretationof the inscription is a methodological one. Often,there is a tendency to try to equate a supposedlywell defined category of monuments to a singleword providing both a definition and an ideal rep-resentation of the structure. In Qasr al-Hayr al-Sharqi, it has been thought that the termmadına could only have served to describe a con-crete and perfectly delineated structure. The largeenclosure, which elsewhere and without inscrip-tion would have been defined as a large qaßr or˛ißn, was a perfect candidate. The madına becamea single, fortified-like monument with some of thefunctions of a town. Meanwhile, the madına beingidentified, this led to the neglecting of otherarchaeological remains, considered as of secondaryimportance.

12.3 WHAT IS A MADˆNA? THEEVIDENCE FROM QASR AL-HAYRAL-SHARQI

All this led to ask the question once again of whata madına might have been seen through the evi-dence from Qasr al-Hayr al-Sharqi?

There are two ways of answering. One is tolook in the written sources to find how the wordwas used in the early Islamic period and then tryto apply the most suitable definition to one orsome of the structures of Qasr al-Hayr al-Sharqi.The other, which I deliberately choose, is to try todefine the word madına through the archaeologi-cal evidence provided by Qasr al-Hayr al-Sharqi. Iproposed before, due to the original position ofthe foundation inscription in the congregationalmosque, that the madına was in fact the whole siteand not only one of its components. This would fitmuch better with the idea of an urban or urban-like settlement, in the sense that it greatly increasesthe number and the variety of its components.This is especially true since it is known that most

61 Clermont-Ganneau 1900, 285–293.62 Gabriel 1927, 231–232. 327.63 Grabar et al. 1978, 12–13.64 Grabar et al. 1978, 79–81.65 Walmsley – Damgaard 2005.66 Northedge 1992, 63–69; Almagro – Jiménez 2000.67 Sack 1996.

The New Urban Settlement at Qasr al-Hayr al-Sharqi 271

of the structures of the northern and eastern settle-ments were large Umayyad houses. Moreover,these houses reflect different status, which issomething much closer to the reality of socialorganisation. The large enclosure alone wouldonly reflect a rich and very egalitarian society,something difficult to accept with our presentknowledge of early Islamic society. It would alsoreflect a very limited residential settlement.

One should expect to find in a new urbansettlement, whatever its form or its more precisedefinition, elements showing at least, political, reli-gious, economic and residential functions. At Qasral-Hayr al-Sharqi, the first is shown by a possibledr al-imra in the large enclosure, but more con-fidently by the palace or caliphal residence. Thereligious function is clear with the congregationalmosque. An economic function is amply shownby large-scale agriculture in the outer enclosures(agriculture is also demonstrated by botanicalstudies), by the southern castles and by industrialstructures such as the olive presses and water mill.The residential function is then to be found in thearistocratic housing of the large enclosure, butmostly in the vernacular houses in the northernand eastern settlements. It may be summarized asfollow, having in mind the differences of status:the different units of the large enclosure werethose of the leading figures of the settlement, nota-bles or ashrf; while tribe members and mawliwere living in normal houses in the northern andeastern settlements. This hierarchy of accommoda-tion reflects a similar situation in the cities.

12.4 TOWARDS AN INTERPRETATIONOF EARLY ISLAMIC QASR AL-HAYRAL-SHARQI

It is not known who was living in Qasr al-Hayr al-Sharqi. The site was built on the order ofHisham b. fiAbd al-Malik and he probably stayedthere occasionally in the palace, but there is noother indication of who was more permanentlythere. It may have been only mawli of Hisham,or, given the hierarchy indicated by the differenttypes of housing, an amır or some ashrf of hisentourage, with their tribal followers and mawli.They may have been sedentary peoples comingfrom elsewhere, but they may also have beennomadic peoples that the government wanted tosettle.

There are, however, other Umayyad andAbbasid new urban settlements that provide paral-lels to help to explain Qasr al-Hayr al-Sharqi. Twoof them, fiAnjar in the Biqafi valley and Madinat al-Far in the Syrian Jazira, were founded byUmayyad generals of the Byzantine wars and were

likely to accommodate parts of the Syrian army68.The fiAmman citadel was probably founded by thegovernor of the Balqa√ and was housing his admin-istration and the garrison69. Al-Ramla, of whichthe plan is unknown, was founded by Sulayman b.fiAbd al-Malik when he was governor of Pales-tine70; like the fiAmman citadel, it was probablyoriginally intended for the provincial administra-tion, for a garrison and for some economic struc-tures, but also in this case and according to al-Bala-dhuri for other inhabitants (al-ns). In theAbbasid period, the Round City of al-Mansur inBaghdad is another good example of a caliphalnew urban settlement, even if much larger than theUmayyad forerunners71. According to al-Yafiqubi,it comprised the palace of al-Mansur, the resi-dences of his children, the congregational mosque,the buildings of the administration and residentialstructures for the servants of the palace andthe military commanders probably with theirtroops72.

I propose, therefore, to interpret Qasr al-Hayral-Sharqi as another attempt to found a new urbansettlement announcing the Round City of al-Mansur on a reduced scale. It was a new urbansettlement in the sense of a caliphal or aristocraticcity intended to be independent and self-sufficienteven though limited in size. It was perhaps alsointended that eventually other people would comeand the population would increase, allowing thesettlement to grow and become an organic city likeother early Islamic cities in Bilad al-Sham. Such asituation is paralleled more successfully with thefoundation of al-Ramla. The latter was saidto have been founded as a madına (madınat al-Ramla), and then made a mißr (maßßara-ha) bySulayman b. fiAbd al-Malik73. The misr in this casemeans a much larger settlement. Al-Ramla in factdeveloped from a probably reduced Umayyadcore in a large and thriving city in the 3rd/9th and4th/10th centuries74. This development was mainlydue to a favourable economic situation linked toindustry and trade.

Nevertheless, there is still one unknown ele-ment at Qasr al-Hayr al-Sharqi and it is who wasliving there? Generals or governors founded theother Umayyad new urban settlements and theirinhabitants are therefore easily identifiable, as foral-Mansur’s Round City. Qasr al-Hayr al-Sharqi isthe only Umayyad new urban settlement to have

68 Northedge 1994, 233–235. 239; Hillenbrand, 1999; Haase2006.

69 Northedge 1992, 165.70 Al-Baladhuri, 1866, 143.71 Northedge 2005, 248–250.72 Al-Yafiqubi 1892, 240.73 Al-Baladhuri, 1866, 143; Northedge 2000, 70.74 Petersen 2005, 95–102.

Denis Genequand272

been founded by a caliph and there are two possi-ble hypothetical answers. The first is that Hishamb. fiAbd al-Malik intended to build Qasr al-Hayral-Sharqi as a secondary or temporary capital forhimself, the caliphal administration and some mili-tary. Indeed, he spent most of his reign in the areaof north-eastern Palmyrena, especially in al-Ru-safa, but it is not known otherwise that he hadanother secondary capital than al-Rusafa/RusafatHisham. The second is that he intended Qasr al-Hayr al-Sharqi for someone of his entourage,for an amır or a sharıf and his clan and followersor perhaps for some of his children.

Due to the changes in the economic and politi-cal situation, but perhaps also to a greater extant tothe important factor of a situation with geographicand environmental conditions generally adverse tosedentary occupation, the madına of Qasr al-Hayral-Sharqi never really grew before losing its roleand being progressively abandoned. Qasr al-Hayral-Sharqi represents the failure of an enterprisethat could have been very successful in differentconditions. It also shows that the Umayyadcaliphs were investing huge amounts of money forlarge enterprises and building programs that prob-ably had a direct political effect, but were not nec-essarily viable in the long-term. The easy accessthey had to financial resources from the whole

caliphate meant that they were not really con-cerned by these long-term considerations.

13 CONCLUSION

The resumption of the work at Qasr al-Hayr al-Sharqi has proved, after five field seasons, to bevery successful. Qasr al-Hayr al-Sharqi remainsone of the major Umayyad Syrian sites and it pro-vides a very good opportunity to study the modal-ities of creation, development and abandonment ofan early Islamic new urban settlement. It alsoallows the study of different kind of strategies forsubsistence in an arid and generally adverse envi-ronment.

As tentatively shown by the present paper,Qasr al-Hayr al-Sharqi presents a huge amount ofnew archaeological data that will need to be inves-tigated in more detail in the future, but it also pre-sents elements allowing more general reconsidera-tion of interpretations of early Islamic archaeologyand history. Whether these new interpretationswill prove to be right or not and accepted or not isanother problem, but they will at least allow forthe continuation of the debate on early Islamicsettlements in Bilad al-Sham on different and re-newed grounds.

The New Urban Settlement at Qasr al-Hayr al-Sharqi 273

BIBLIOGRAPHY

ALMAGRO, M. – CABALLERO, L. – ZOZAYA, J. –ALMAGRO, A. 1975

Qusayr fiAmra. Residencia y banos omeyas enel desierto de Jordania. Granada.

ALMAGRO, A. – JIMÉNEZ, P. 2000The Umayyad Mosque of the Citadel ofAmman, AAJ 44, 459–475.

AL-BALADHURI, A. B. Y. 1866Kitb futü˛ al-buldn. Edited by M. J. DeGoeje. Leiden.

BRISCH, K. 1963Das omayyadische Schloss in Usais, MDIK 19,141–187.

BRISCH, K. 1965Das omayyadische Schloss in Usais (II), MDIK20, 138–177.

BUJARD, J. – GENEQUAND, D. 2001Umm al-Walid et Khan az-Zabib, deux étab-lissements omeyyades en limite du desert jor-danien, in: B. Geyer (ed.), Conquête de lasteppe et appropriation des terres sur lesmarges arides du Croissant fertile, Lyon (TMO36), 189–218.

CLERMONT-GANNEAU, C. 1900Une inscription du calife Hicham (an 110 del’hégire), in: C. Clermont-Ganneau, Recueild’archéologie orientale, III, 285–293.

CRESWELL, K.A.C. 1969Early Muslim Architecture. Oxford.

COMBE, E. – SAUVAGET, J. – WIET, G. 1931Répertoire chronologique d’épigraphie arabe.Vol. I. Cairo.

GABRIEL, A. 1927Kasr el-Heir, Syria 8, 302–329.

GENEQUAND, D. 2003aProjet «Implantations umayyades de Syrie etde Jordanie». Rapport de la campagne deprospection (juin–juillet 2002), Schweizerisch-Liechtensteinische Stiftung für ArchäologischeForschungen im Ausland (SLSA/FSLA/SLFA)– Jahresbericht 2002, 31–68.

GENEQUAND, D. 2003bRapport préliminaire de la campagne de fouille2002 à Qasr al-Hayr al-Sharqi (Syrie), Schwei-zerisch-Liechtensteinische Stiftung für Archäo-logische Forschungen im Ausland (SLSA/FSLA/SLFA) – Jahresbericht 2002, 69–96.

GENEQUAND, D. 2003cMa‘an, an Early Islamic Settlement in SouthernJordan: Preliminary Report on a Survey in2002, AAJ 47, 25–35.

GENEQUAND, D. 2004aRapport préliminaire de la campagne de fouille2003 à Qasr al-Hayr al-Sharqi et al-Bakhra√(Syrie), Schweizerisch-Liechtensteinische Stif-tung für Archäologische Forschungen im Aus-

land (SLSA/FSLA/SLFA) – Jahresbericht 2003,69–98.

GENEQUAND, D. 2004bChâteaux omeyyades de Palmyrène, Annalesislamologiques 38, 3–44.

GENEQUAND, D. 2005aFrom ‘desert castle’ to medieval town: Qasr al-Hayr al-Sharqi (Syria), Antiquity 79/304,350–361.

GENEQUAND, D. 2005bRapport préliminaire de la campagne de fouille2004 à Qasr al-Hayr al-Sharqi (Syrie),Schweizerisch-Liechtensteinische Stiftung fürArchäologische Forschungen im Ausland(SLSA/ FSLA/SLFA) – Jahresbericht 2004,143–166.

GENEQUAND, D. 2004–2005 (2008)Nouvelles recherches à Qasr al-Hayr al-Sharqi:la mosquée ayyoubide et la nécropole, AAS 47–48, 271–293.

GENEQUAND, D. – KÜHN, M. – DE REYNIER, C.2006

Rapport préliminaire des travaux de la missionarchéologique syro-suisse à Qasr al-Hayr al-Sharqi (Syrie) en 2005, Schweizerisch-Liech-tensteinische Stiftung für ArchäologischeForschungen im Ausland (SLSA/FSLA/SLFA)– Jahresbericht 2005, 161–203.

GENEQUAND, D. – HULL, D. – STUDER, J. 2007Rapport préliminaire des travaux de la missionarchéologique syro-suisse à Qasr al-Hayr al-Sharqi en 2006, Schweizerisch-Liechten-steinische Stiftung für Archäologische For-schungen im Ausland (SLSA/FSLA/SLFA) –Jahresbericht 2006, 123–156.

GRABAR, O. 1970Le nom ancien de Qasr al-Hayr al-Sharqi,Revue des Etudes Islamiques 38, 251–266.

GRABAR, O. – HOLOD, R. – KNUSTAD, J. – TROUS-DALE, W. 1978

City in the Desert: Qasr al-Hayr East. Cam-bridge (MA).

GREGORY, S. – KENNEDY, D. 1985Sir Aurel Stein’s Limes Report. Oxford.

HAASE, C.-P. 2006The Excavations at Madinat al-Far/HisnMaslama on the Balikh Road, in: H. Kennedy(ed.), Muslim Military Architecture in GreaterSyria. From the Coming of Islam to theOttoman Period, Leiden-Boston, 54–60.

HAMILTON, R. W. 1946An Eighth-Century Water-Gauge at al-Mu-waqqar, QDAP 12, 70–72.

HAMILTON, R.W. 1959Khirbat al-Mafjar: An Arabian Mansion in theJordan Valley. Oxford.

Denis Genequand274

HILLENBRAND, R. 1999fiAnjar and Early Islamic Urbanism, in: G. Bro-giolo – B. Ward-Perkins (eds.), The Idea andIdeal of Town between Late Antiquity and theEarly Middle Ages, Leiden, 59–98.

NORTHEDGE, A. 1992Studies on Roman and Islamic fiAmman. Vol. 1:History, Site and Architecture. Oxford.

NORTHEDGE, A. 1994Archaeology and New Urban Settlement inEarly Islamic Syria and Iraq, in: G.R.D. King –A. Cameron (eds.), The Byzantine and EarlyIslamic Near East. II, Land Use and SettlementPatterns, Princeton, 231–265.

NORTHEDGE, A. 2000Entre Amman et Samarra: l’archéologie et lesélites au début de l’Islam (VIIe–IXe siècle),Unpublished HDR thesis, Université de Paris I.

NORTHEDGE, A. 2001Thoughts on the Introduction of PolychromeGlazed Pottery in the Middle East, in: E. Ville-neuve – P.M. Watson (eds.), La céramiquebyzantine et proto-islamique en Syrie-Jordanie(IVe–VIIIe siècles apr. J.-C.), Beyrouth,207–214.

NORTHEDGE, A. 2005The Historical Topography of Samarra. Samar-ra Studies I. London.

PETERSEN, A. 2005The Towns of Palestine under Muslim RuleAD 600–1600. Oxford.

POIDEBARD, A. 1934La trace de Rome dans le désert de Syrie. Paris.

ROUSSEAU, J.-B. 1899Voyage de Baghdad à Alep, 1808. Paris.

SACK, D. 1996Resafa IV. Die Grosse Moschee vonResafa–Rußfat Hishm. Mainz.

SAUVAGET J. 1939Remarques sur les monuments omeyyades,Journal Asiatique 231, 1–59.

SCHLUMBERGER, D. 1986Qasr el-Heir el-Gharbi. Paris.

SEYRIG, H. 1931Antiquités syriennes I. Les jardins de Kasr el-Heir, Syria 12, 316–318.

SEYRIG, H. 1934Antiquités syriennes XVI. Retour aux jardinsde Kasr el-Heir, Syria 15, 24–32.

AL-TABARI, M. B. J. 1879–1901Ta√rıkh al-rusul wa-l-mulük. Edited by M. J.De Goeje, 15 vols. Leiden.

WATSON, O. 1999Report on the Glazed Ceramic, in: P.A. Miglus,Die frühislamische Keramik von Tall Aswad,Raqqa I, Mainz, 81–87.

WALMSLEY, A. – DAMGAARD, K. 2005The Umayyad Congregational Mosque ofJarash in Jordan and its Relationship to EarlyMosques, Antiquity 79/304, 362–378.

AL-YAfiQUBI, A. B. A.Y. 1892Kitb al-Buldn. Edited by M. J. De Goeje(BGA 7). Leiden.

The New Urban Settlement at Qasr al-Hayr al-Sharqi 275

Fig. 1 General plan of the site of Qasr al-Hayr al-Sharqi (survey and drawing: Sophie Reynard, Denis Genequand &Christian de Reynier): 1 palace, 2 large enclosure, 3 northern settlement, 4 outer enclosures, 5 southern castles,

6 water mill and aqueduct in Wadi al-Suq.

Denis Genequand276

Fig. 2 Plans of the palace (right) and large enclosure (left) (after Grabar et al. 1978).

Fig. 3 Western façade and gate of the palace, with the remains of the medieval mosque in the foreground(photo: Denis Genequand).

The New Urban Settlement at Qasr al-Hayr al-Sharqi 277

Fig. 4 Northern settlement, plan of the house in field A (drawing: Marion Berti).

Fig. 5 Northern settlement, aerial view of the house in field A (photo: Denis Genequand).

Denis Genequand278

Fig. 6 Abbasid pottery of the last phase of occupation in field A/northern settlement. 6808-1: basin; greenish ware(Munsell 5Y 7/3, «pale yellow»), medium fine, medium hard, mica, black, white and brown grits; yellowish slip. 6811-1:bottle; orange-red ware (Munsell 2.5YR 6/6, «light red»), fine, hard, few black and white grits; yellowish slip. 6811-6:cooking pot; “brittle ware”, red fabric (Munsell 10R 4/8, «red»), fine, hard, some bigger white grits; dark brown to blacksurface, slightly metallescent. 6817-1: amphora; red ware (Munsell 2.5YR 6/6, «light red»), fine, hard, some white andblack grits; yellowish slip. 6852-1: juglet; greenish ware (Munsell 5Y 7/3, «pale yellow»), fine, medium hard, some mica,black, white and brown grits; yellowish slip. 6852-2: pot; greenish ware (Munsell 5Y 7/3, «pale yellow»), fine, mediumhard, some mica, black, white and brown grits; yellowish slip. 6811-9: bowl, ‘Samarra√ horizon’; green splashes on anopaque white glaze (interior and exterior); yellow fabric (Munsell 5Y 8/4, «pale yellow»), very fine, hard. 6811-20/10:bowl, “yellow-glaze-family”; interior covered by a thin white slip, green stripes on an opaque yellow glaze; exteriorwith only lines of the white slip; red fabric (Munsell 10R 6/8, «light red»), fine, hard, some mica, white and grey grits.6856-2: bottom of pot (?), “yellow-glaze-family”; interior covered by a thin white slip, splashes of yellow glaze; exteri-or covered by a thin white slip, manganese painting and yellow and green glaze painting; red fabric (Munsell 10R 5/8,«red»), fine, hard, white grits. 6811-32: steatite vessel with incised decoration. 6811-35: alabaster cup

with red painted and incised decoration (drawing: Marcia Haldemann, Iona McRae and Marion Berti).

The New Urban Settlement at Qasr al-Hayr al-Sharqi 279

Fig. 7 Northern settlement, aerial view of excavations in field B (photo: Denis Genequand).

Fig. 8 Northern settlement, detail of the protruding porch of the house in field B (photo: Denis Genequand).

Denis Genequand280

Fig. 9 Northern settlement, plan of sounding in field C (drawing: Christian de Reynier).

Fig. 10 Northern settlement, aerial view of the sounding in field C (photo: Denis Genequand).

The New Urban Settlement at Qasr al-Hayr al-Sharqi 281

Fig. 11 Plan of the southern castles (survey and drawing: Sophie Reynard, Marion Berti, Christian de Reynier &Denis Genequand).

Fig. 12 Southern castles, excavations in the corner of the northern building; some parts of the excavations werealready backfilled (photo: Denis Genequand).

Denis Genequand282

Fig. 13 Southern castles, squinch in the northern build-ing (photo: Denis Genequand).

Fig. 14 Southern castles, view of the west side of thegateway and entrance corridor of the southern building

(photo: Denis Genequand).

Fig. 15 Southern castles, sounding in the eastern wing of the southern building (photo: Denis Genequand).

The New Urban Settlement at Qasr al-Hayr al-Sharqi 283

Fig. 16 Map of the aqueducts bringing water from al-Kawm and Umm al-Tlal to Qasr al-Hayr al-Sharqi.

Fig. 17 View of the water mill with remains of the sec-ondary aqueduct and oblique water fall to the left; the tri-angular covering of the aqueduct in the background be-longs to the medieval period (photo: Denis Genequand).

Fig. 18 Remains of the wall of the outer enclosure (photo: Denis Genequand).

Denis Genequand284

Fig. 19 Southern sluices of the outer enclosure (photo: Denis Genequand).

Fig. 20 Foundation inscription of the madına (after Clermont-Ganneau 1900).

The New Urban Settlement at Qasr al-Hayr al-Sharqi 285


Recommended