D. Genequand, An Early Islamic Mosque in Palmyra, 2008

An Early Islamic Mosque in Palmyra

Denis Genequand

This article presents the results of preliminary investigations on a mosque that was recently

rediscovered in the centre of Palmyra. The general organization of the mosque, the way in which it

reuses a pre-existing Roman building, its situation south of the Tetrapylon and the data at hand on

the later periods in Palmyra all point to an Umayyad date for the structure. This new element of

early Islamic Palmyra will be described. This in turn will lead to some considerations on its context,

in particular in relation with the Umayyad suq already known in Palmyra and more generally on

caliphal involvements in the large scale programmes of urban development or urban creation.

Keywords: Early Islam, Mosque, Palmyra, Umayyad Caliphate and Urban Pattern

Introduction

The city of Palmyra in the Syrian steppe benefited

from the ideal position of a large oasis situated

between the Euphrates and the cities of western

Syria (Busra, Damascus, Homs/Emesa) and the

Mediterranean coast. As a caravan city Palmyra

was at the height of its fame under the Principate

during the 1st to the 3rd centuries AD. The attempt at

usurpation by Zenobia and her son Wahballat,

followed by the victorious military campaign of

Aurelianus in AD 272, marked the end of its great

prosperity and some sort of decline, but certainly not

its disappearance. If, after that, Palmyra no longer

had its former power, it nevertheless stayed an

important city. It was first and foremost a garrison

city, where the Legio I Illyricorum was based after the

provincial reorganization under the Tetrarchy

(Notitia Dignitatum, Or. XXXII, 30). It then became

a bishopric and the city sent a bishop to the council of

Nicaea in AD 325 and again in AD 351 to the council

of Chalcedon (Devreesse 1945, 206). By the 6th

century, the Episcopal group situated in the middle of

the city was pretty large and included several

churches and chapels (Gawlikowski 2005); other

churches distributed around the city bear witness to

a settlement that was still thriving. Its relative

importance is also demonstrated by the fact that in

AD 527 the emperor Justinian ordered the restoration

of the city, especially its rampart, churches and water

system (reported by Ioannes Malalas, Theophanes

and Procopius, cf. Kowalski 1997, 50–52; on the

ramparts: Seyrig 1950; Gawlikowski 1974).

During the early Islamic period Palmyra is less well

documented, but there are some textual sources and

some good archaeological evidence that needs to be

better exploited to demonstrate its continuing impor-

tance. Also, in the extensive region around Palmyra,

there are a certain number of caliphal palaces,

aristocratic mansions or residences and new urban

settlements that point indirectly to this importance.

These ‘desert castles’ have been the subject of study of

a Syrian-Swiss project since 2002 (Genequand 2004a;

2004b; 2005). During the 2006 season of this project,

a large mosque was rediscovered in Palmyra which,

as will be demonstrated below, is likely to be the

Umayyad congregational mosque. As would be

expected with such a structure, the rediscovered

mosque is also shedding new light on the early

Islamic period in the city. Only a very short time was

devoted in the field to the Palmyra mosque in 2006,

but a larger Syrian-Swiss project is currently being

organized for its detailed investigation in the coming

years. This short article is intended as a preliminary

presentation of the newly rediscovered mosque, of its

context and of its interpretation.

Early Islamic Palmyra: the Textual Sources

Since before the coming of Islam, Palmyra/Tadmur

and its region were part of the domain of Kalb, a

Yamani tribe. Palmyra and its region were conquered

in the mid-630s, perhaps even as early as H 13/AD 634

by the general Khalid b. al-Walid (al-Baladhuri,

Futuh al-Buldan, 111). After this, Palmyrena stayed in

the hands of Kalb and was incorporated in the jund of

Denis Genequand, Council for British Research in the Levant, 1 Pl.Mussard, CH-1290 Versoix, Switzerland, e-mail: [email protected]

� Council for British Research in the Levant 2008Published by ManeyDOI 10.1179/175638008x284143 Levant 2008 VOL 40 NO 1 3

Homs. According to Abu Mikhnaf, cited by al-

Tabari (Ta’rikh, II, 482), in H 64/AD 683–684 most of

the Banu Umayya were driven out of Medina, Mecca

and the Hijaz by ‘Abdallah b. al-Zubayr and went to

Palmyra. It was there that Marwan b. al-Hakam

received the oath of allegiance by the Banu Umayya

and by the people of Palmyra, before moving with an

army against al-Dahhak b. Qays at the battle of Marj

Rahit and then becoming caliph in H 65/AD 684. This

was an important political event, which marked a

shift of power from the Sufyanid to the Marwanid

branch of the Umayyad family. If it really happened

in Palmyra, it shows the political weight the city and

surrounding area had at that time.

In H 126/AD 744 the city was apparently still well

fortified, as during the rebellion of Yazid b. al-Walid

someone proposed to al-Walid b. Yazid to take

refuge there (al-Tabari, Ta’rikh, II, 1796). The latter

finally decided to go to al-Bakhra’, 21 km south of

Palmyra, where he was killed by the soldiers of ‘Abd

al-‘Aziz b. al-Hajjaj b. ‘Abd al-Malik and buried

there (al-Tabari, Ta’rikh, II, 1795–1807; Genequand

2004b). The city, and particularly its ramparts,

subsequently suffered an amount of damage difficult

to estimate after having supported Sulayman b.

Hisham against the caliph Marwan b. Muhammad

in H 127/AD 744–745 (al-Tabari, Ta’rikh, II, 1892,

1896, 1912; Ibn al-Faqih, Kitab al-Buldan, 110;

Yaqut, Mu‘jam al-Buldan, II, 17) and again in H

132/AD 750 during the fights related to the fall of the

Umayyad dynasty (al-Tabari, Ta’rikh, III, 53–54;

al-Isfahani, Aghani, XVIII, 150).

A bishopric was maintained in Palmyra during the

Umayyad and Abbasid periods and bishops were still

ordained in, or shortly after, H 177/AD 793 (bishop

Symeon, formerly in the monastery of Mar Jacob in

Cyrrhus) and H 202/AD 818 (bishop Iohannan/John

III, formerly at the monastery of Mar Hanania)

(Michael the Syrian, vol. III, 451, 453).

By the end of the 10th century AD, the geographer

al-Muqaddasi described Palmyra as a qasaba, a word

implying a rather small town or settlement encom-

passed with some sort of enclosure (Ahsan al-

Taqasim, 156).

Early Islamic Palmyra: the Archaeology

There have been archaeological excavations in

Palmyra since the 1920s, most of them devoted to

the impressive remains of the Roman period (Fig. 1).

Until recently, with some exceptions like Seyrig’s note

on the ramparts (Seyrig 1950) or Huart’s and

Sauvaget’s studies of a handful of medieval Arabic

inscriptions (Huart 1929; Sauvaget 1931), remains

belonging to the late Antiquity and the Islamic period

were not studied, nor even mentioned in many cases.

Since the 1960s, there have more often been mentions

of later modifications or hints about the fate of some

monuments after the Principate.

Amongst the few elements that give a more precise

idea of early Islamic Palmyra, one should note several

houses excavated by the Polish mission during the

1960s in the western part of the city (area of the

Camp of Diocletian), where there is ample evidence

of an occupation that was still important during the

so-called ‘Arab’ period (Michalowski 1960, 69–81;

1962, 54–77; 1963, 41–60, 188–89 (coins); 1966, 27–

29). Nevertheless, the simplistic equation [well built 5

Byzantine / poorly built 5 late Byzantine / badly built

5 Arab] makes the precise date and evolution of the

area difficult to follow. In one of these houses, a

hoard of coins was found, dating back to the mid-7th

century and composed of Byzantine coins minted

between the reigns of Phocas (AD 602–610) and

Constans II (AD 641–668) (Michalowski 1962, 222–

36). Amongst the few sherds published with a

photograph (Michalowski 1960, fig. 87; 1962,

fig. 70), some seem to belong to the early polychrome

glazed ceramic and can be dated broadly to the first

half of the 9th century AD, providing a rough date for

the abandonment of the area.

Similarly, Syrian excavations in the Annexe of the

Agora (Basilica) brought to light ‘Byzantine’ houses

and ‘Arab-Islamic’ or ‘Ayyubid’ houses and pottery

workshops (Bounni and Saliby 1968, 101). These

were not fully published, only a rough plan of the

houses and two photographs with pottery. On these

photographs, one rounded lamp fits well with a

second half of the 8th to first half of the 9th century

AD date (Bounni and Saliby 1968, pl. X.3) and there is

no illustrated material that is later than this. One may

wonder about the precise date of these structures,

especially the ‘Arab-Islamic’ houses and workshops

for which an early Islamic date seems plausible. With

the documentation presently at hand and before an

examination of the preserved material, there is

nothing sustaining an Ayyubid date, moreover there

is no other place showing signs of an Ayyubid

occupation in this part of the city. One may also

recall that in Baysan or Jarash, Umayyad and early

Abbasid pottery workshops were located in several

places in the centre of the ancient cities, for example

in Jarash’s North Theatre/Bouleuterion facing the

former Agora (Schaefer and Falkner 1986).

Recent Polish excavations in and around the large

Episcopal group of Palmyra also provide ample

evidence for an occupation of that area until the

Genequand An Early Islamic Mosque in Palmyra

4 Levant 2008 VOL 40 NO 1

beginning of the 9th century AD. Three churches were

excavated (churches I, II and III) (Gawlikowski 1992;

1994; 1998; 1999; 2002; 2005). Church I had a

complicated history as a monument built in the 4th

century AD and abandoned by the beginning of the

7th century AD. Church II was built or at least largely

modified in the Umayyad period, while Church III

was probably built in the 5th century AD. Churches II

and III were in use until the early 9th century AD.

They both show the vitality that the Christian

community of Palmyra exhibited during the

Umayyad and early Abbasid periods. During the

excavations in the atrium of Church III, two hoards

of Sasanian silver dirhams were found, containing

respectively 757 and 18 coins (Gawlikowski 2002,

266–69). After a preliminary study, it appears that a

majority of coins were minted between AD 534 and

AD 626, from the reign of Khusraw I to the reign of

Khusraw II. But the hoard also includes pre-reform

early Islamic coins minted under Ziyad b. Abu

Sufyan (dated H 53/AD 673) and ‘Abdallah b. al-

Zubayr (dated H 63/AD 682 and H 72/AD 691). In all

likelihood, the hoard was collected during the

Sasanian occupation of Syria and buried together

with later coins during the Umayyad period after H

72/AD 691. In a neighbouring block, a very large

house was also excavated (Gawlikowski 1992; 1994).

Built in the second half of the 2nd century AD, the

house witnessed modifications to its original plan and

stayed in use until the late 8th to early 9th century AD.

Between 1977 and 1984, the Museum of Palmyra

undertook the excavation of the main Colonnaded

Street to the west of the Tetrapylon, the eastern part

having been cleared earlier during the late 1950s and

Figure 1 General plan of Palmyra (drawing, Thibaud Fournet)


Levant 2008 VOL 40 NO 1 5

early 1960s by the Directorate General of Antiquities

and Museums. While excavating the western part of

the Colonnaded Street, the remains of a suq or

market were discovered (al-As‘ad and Stepniowski

1989). It consists of a row of at least forty-seven

shops built in the middle of the street. Shops are

aligned for over 170 m and in groups of three to

fifteen; each of the six groups represents a different

construction phase. All of them were accessible from

a passage left on their northern side, in between their

facade and the sidewalk of the Colonnaded Street.

The whole structure is built in stone and can be

considered as a substantial monumental construc-

tion. It is not known whether the suq also extended to

the east of the Tetrapylon and was removed when this

part of the Colonnaded Street was cleared, or

whether it stopped there.

This suq was first thought to be of medieval date

and of not much importance. It was thanks to the

Polish mission that two of the shops were excavated

and some soundings done elsewhere. This allowed the

precise dating of the suq (al-As‘ad and Stepniowski

1989, 210–11, 222). Three different main layers were

pointed out: a pre-suq layer dated to the Umayyad

period, a layer of occupation and use of the suq also

dated to the Umayyad period, and a secondary use

and abandonment layer dated to the Abbasid period.

The proposed date of construction of the suq is the

very late 7th or first half of the 8th century AD.

As the first structure of its kind to have been

discovered archaeologically in Bilad al-Sham, the

Umayyad suq in Palmyra was not given its real

importance when published. Since then, another

Umayyad suq was excavated in Baysan/Scythopolis

(Khamis 2001). This later structure was two-storeyed

and very different from the one in Palmyra, but it is

clearly identified by a dedication inscription of the

caliph Hisham b. ‘Abd al-Malik dated H 120/AD 738.

Another commercial structure identified as part of a

market was discovered in Rusafa/Sergiopolis, a

temporary capital city of the same caliph (Ulbert

1992). It is L-shaped and extends for over 100 m

along the main street leading from the palatial area

outside the rampart to the congregational mosque on

the northern side of the Saint Sergius sanctuary. The

markets in Palmyra, Baysan and Rusafa testify to the

importance accorded by the Umayyad caliphs to

commercial and economic development. As large

structures that can only have been commissioned by a

strong political power — this is especially clear with

the size and situation of the Palmyra suq — they also

show how much Umayyad caliphs were involved in

urban development in Bilad al-Sham.

Apart from Palmyra itself, the whole area called

Palmyrena shows good evidence of its importance

during the Umayyad period. There are a good

number of caliphal palaces, aristocratic mansions or

residences and new urban settlements dotted all

around Palmyra (Genequand 2004a). The most

famous are Qasr al-Hayr al-Gharbi (Schlumberger

1986) and Qasr al-Hayr al-Sharqi (Grabar et al. 1978;

Genequand 2005), both built by the caliph Hisham b.

‘Abd al-Malik respectively in H 109/AD 727 and H 110/

AD 728–729; amongst the others, one should mention

Nu‘man b. Bashir’s qasr in al-Bakhra’ (Genequand

2004b) or Qudaym, another large aristocratic

mansion of an unknown owner (Genequand 2004a,

11–12).

The Mosque

Situation

The mosque is situated on the main east-west

Colonnaded Street, to the south of the Tetrapylon,

directly to the north-west of the Theatre and some-

what north-east of the Agora (Fig. 2). It is the centre

of the Roman city. The mosque is built on top of and

reusing parts of a large Roman structure that was

tentatively interpreted as the temple of the imperial

cult (Caesareum). Depending on the original extent of

the Umayyad suq, the mosque is situated either at its

eastern extremity or in the middle of it.

The mosque and the underlying Roman building

were excavated in 1962 by a team from the

Directorate General of Antiquities and Museums of

Syria led by Adnan Bounni (Bounni 1963). The

excavation was apparently done without any record-

ing of the stratigraphy. Antoni Ostraz, when planning

the eastern part of the Colonnaded Street, produced

an architectural plan, but it includes only the Roman

remains (Ostraz 1969) (Fig. 3). It is possible now to

do more work on the standing remains, but it is

difficult to know how much of the two structures

were removed during the excavations, particularly for

the later one.

Roman Building

The pre-existing Roman building measures 42.50 633 m externally (Figs 3 and 5). It is oriented north-

east to south-west, but for convenience will be

described as if oriented north to south. It is entered

by a monumental tripartite door situated in the

middle of its northern facade and flanked by a series

of shops opening into the Colonnaded Street. After

the door, a corridor leads to a large paved courtyard

with a peristyle. Five out of six columns of the

southern row are well preserved and still have some



pieces of architraves on top (Fig. 4). They are fluted

columns and capitals belonging to the Corinthian

order. These columns open into a hypostyle room,

where only fragments of bases or negatives indicate

the positions of columns. Their exact insertion in the

structure needs to be better defined as it is not

completely clear whether they all belong to the

original monument. Another point that needs to be

better understood is the presence on Ostraz’s plan

(1969) of two side rooms at both ends of the

hypostyle room. They are no longer visible (founda-

tions preserved at a lower level?), but the general plan

of the structure would make more sense if they had

really existed. The street facade, the peristyle and the

courtyard pavement were built with finely dressed or

carved ashlars, shafts, bases or capitals of hard

yellow limestone, while most of the side and internal

walls, which would have been covered with a coat of

plaster, were made out of regular ashlars of brown

limestone.

The southern wall of the building is interrupted by

a different piece of masonry that represents what

seem to be the remains of an exedra. It has a diameter

of about 3.60 m and is paved with flagstones made of

yellow limestone (Fig. 10). It is not clear whether this

belongs to the Roman period or if it belongs to later

rebuilding activities; this point will be discussed later.

The eastern part of the building is composed of

another row of shops opening into a smaller

colonnaded street leading to the Agora and bordered

by the semicircular plaza around the Theatre. The

axis of the street oriented towards the Agora involved

an oblique facade for the shops and means that the

southernmost ones are especially small.

The function of the building is not clear. Its central

position and the apparent absence of any other

rooms mean that it is unlikely to have been a house.

The interpretation that it was the temple of the

imperial cult (Caesareum) is based solely on the find

of an inscription mentioning an autocrator Caesar

Figure 2 Plan of the central sector of Palmyra, including the Agora, the Theatre, the Tetrapylon, part of the Colonnaded

Street and the Roman building/mosque. The Umayyad suq was extending to the west of the Tetrapylon (draw-

ing, Thibaud Fournet)



Figure 3 Tetrapylon, Colonnaded Street and Roman building planned in the 1960s by Antoni Ostraz (after Ostraz 1969)

Figure 4 The courtyard and columns of the Roman building seen from the south-east; the Tetrapylon is in the back-

ground (photo, Denis Genequand)



divus (Bounni 1995, 17), which was clearly not in its

original place and therefore not really indicative of

the building’s function. Temples of the imperial cult

do not necessarily follow defined plans and a

hypostyle hall with an exedra could be a solution.

But recent research on the imperial cult in Palmyra

tends to show that the temple is more likely to have

been located elsewhere. Further investigations on the

building will therefore be necessary in order to really

understand its function.

Architecture of the Mosque

For the description of the mosque, one should bear in

mind that it is based on the standing and visible

remains and that some of the later walls may have

been removed during the excavations. The mosque is

built on top of the Roman structure (Fig. 5). It reuses

the eastern Roman wall without modification, while

the western wall is rebuilt on the same alignment with

reused architectural elements (columns shafts, archi-

traves, etc.) and ashlars. The qibla wall is built anew

with the correct orientation to the south of the

Roman building. Because of this orientation towards

Mecca, it is set obliquely, departing from the eastern

Roman wall and linked to the western one by another

oblique wall. The qibla wall and the linking wall are

also built with reused architectural elements and

ashlars. The northern wall is more difficult to assess

for the moment. A quick look at it gives the

impression that the Roman wall was reused appar-

ently without modifications, like the eastern wall,

conserving its facade with shops in front of the

Tetrapylon. But, in the original Roman doorway,

massive and isolated monolithic jambs of another

door set obliquely, approximately on the same

orientation as the qibla wall, may indicate that it

was rebuilt or at least modified (Fig. 7). The jambs

rest directly on the Roman threshold.

The internal arrangement of the structure seems

quite clear given the standing pre-existing remains.

The peristyle becomes the courtyard. The still

standing southern row of columns of the peristyle

marks the limit between the courtyard and the prayer

hall. It was perhaps closed by wooden elements

placed between the columns or even by stone

masonries. The prayer hall corresponds more or less

to the former hypostyle room and its ceiling was

probably partly supported by the same row of

disappeared columns (Fig. 6). As previously stated,

it is nevertheless not impossible that the rather

irregular remains of this inner colonnade belong in

parts to the mosque phase. In the western half of the

prayer hall, there was certainly another row of

columns, the width being otherwise too large for its

ceiling. This second row of columns was perhaps in

line with the southern wall of the former Roman

Figure 5 Plan of the currently standing remains of the mosque and former Roman building (survey by Marie-Cecile

Bosert and Denis Genequand; drawing, Marie-Cecile Bosert and Marion Berti)



building. Because of the obliqueness of the qibla wall,

there is definitely no need for this second row in the

eastern part. Of course, the two possible side rooms

of the Roman structure no longer existed. Before

further work is done in the field and with the elements

presently at hand, it therefore seems possible to

reconstruct a prayer hall with three bays, respectively

two bays to the east, and nine aisles. As for the

courtyard, it cannot be known from standing remains

whether there were side porticoes or not, but given

the general plan, it is more likely to have been the

case.

The mosque as a whole is a rather large structure

with an outside width of 41 m (south) to 42.50 m

(north) and a length of 26 m (east) to 42.50 m (west).

The inside dimensions of the courtyard are at least 41

6 15.45 m (not including the former shops, the

preservation of which is unknown at that time) and

the prayer hall measures internally 39.50 (south)/41

(north) 6 18 (west)/1.40 (east) m. The oblique

position of the qibla wall explains these rather odd

dimensions.

Of course, the qibla wall is the best element for

identifying the structure and, indeed, it is amongst

Figure 6 View of the prayer hall of the mosque (photo, Denis Genequand)

Figure 7 Standing door jambs inside the original Roman doorways (photo, Denis Genequand)



the best preserved elements. It contains a central

semicircular niche, a mihrab indicating the direction

of Mecca, backed by a trapezoidal salient projecting

outside (Figs 8 and 9). This mihrab is rather small,

having a width of 0.90 m for a depth of 1.30 m, and

bordered by two vertical rectangular monolithic

piers. Inside it, the lime mortar floor is well preserved

and gives an indication of the level of the original

floor of the mosque (i.e. all the remains visible now in

the prayer hall were below the floor). There is a small

niche 2 m to the east of the mihrab. Only one course

high and situated just above the floor level, it is not

sure whether it was intended to be a small lateral

mihrab, or if it is only due to the reuse of former

architectural fragments and was supposed to be

covered with plaster. Further investigations are

needed.

The eastern half of the qibla wall presents several

phases, modifications and additions that will not be

described in detail here, some more fieldwork also

being necessary. Suffice it to say now that a

rectangular room was added to the south of the qibla

wall (Fig. 5), the walls of which are preserved much

higher than for the rest of the mosque and are also

made of reused ashlars and architectural elements. It

can only be hypothesized that it was the base for a

structure higher than the mosque, perhaps a tower.

Figure 8 Central part of the qibla wall with the mihrab and a lateral niche (photo, Denis Genequand)

Figure 9 The mihrab (photo, Denis Genequand)



Fortunately, it has not been completely excavated

and will in the future provide some stratified levels

and valuable information about its architecture,

phasing and insertion into the mosque.

The last point that needs to be made in relation to

the qibla wall is the possibility — I say the possibility,

because it is only an hypothesis that needs to be

verified — of a first qibla being on the same alignment

as the southern wall of the Roman building. It was

noted above that the southern Roman wall is

interrupted by some masonry apparently related to

some sort of exedra (Fig. 10). The exact phasing of

the exedra, whether original or much later, is not yet

known. When planning the building shortly after its

excavation, when it was presumably cleaned for

detailed observations, Ostraz (1969) considered it as

a later addition and left a blank space on the plan. It

could therefore be hypothesized that it belongs to a

first mosque that fully reused the rest of the plan of

the Roman building. In this case, it would have been

a rather large mihrab, like those of Jarash (Walmsley

and Damgaard 2005) or ‘Amman Citadel (Almagro

and Jimenez 2000). Such a possibility, with two

different phases of mosque, would explain better

certain oddities in the plan, especially in the

south-west corner, where secondary masonries are

not all bounded together. After a first mosque quickly

set in the centre of Palmyra, its Muslim community

would have felt the need of a qibla properly oriented

towards Mecca and modified the monument.

Finally, there are several elements related to the

mosque in front of the northern facade, including a

cistern built on top of former canalizations. It could

have been used for ritual ablutions before entering

the mosque or to bring water inside the courtyard for

the same purpose.

Dating of the Mosque

The mosque was not excavated with modern archae-

ological methods and nothing is known about the

stratigraphy or other dating elements from the

excavations. The only way for this to be done now

would be by excavating south of the qibla wall,

especially in the rectangular room abutted to it.

Nevertheless, before further investigations in the

field, the context and setting of the mosque strongly

suggest a likely date.

First, it is set in the centre of the former Roman

city in clear relation to the Umayyad suq, a well-

dated structure. Secondly, when the mosque was

Figure 10 Remains of the exedra in the southern wall of the Roman building; it is possible that it was a first mihrab

(photo, Denis Genequand)



built, the Roman topography of the city was still

predominant, the streets still in use, parts of the

Roman monuments still standing and the floor levels

not much elevated (as shown by the reuse of the

original Roman paved courtyard, by the reuse of

the Roman threshold of the door and by the floor in

the mihrab only slightly higher); all this indicates the

absence of a long period of abandonment and

involves a date still close in time to late Antiquity.

Thirdly, from what is known about the evolution and

stratigraphy of excavated areas in Palmyra, the centre

of the city does not present layers or phases that are

dated later than the first half of the 9th century AD

(see above), the exceptions being the temple of Bel,

where the occupation seems to concentrate from the

9th–10th century AD onwards and which was

refortified in H 527/AD 1132–1133 (Sauvaget 1931,

143), and the late 12th to early 13th century AD

fortress dominating the city (Bylinski 1999). By the

middle of the 9th century AD, most of the Roman city

was abandoned.

From what precedes, the most likely date for the

mosque is the early Islamic period and more precisely

the Umayyad period. On the one hand, a later date of

the Ayyubid or Mamluk period seems impossible

given the absence of other contemporary remains in

the central part of the city. And even if there were

some, the large size of the mosque would make it an

unlikely candidate for a small occupation. On the

other hand, the Umayyad period is a very good

candidate, having a lot of evidence for a still thriving

occupation and, not least, evidence of large-scale

modifications of urbanism with the nearby suq. For

the moment and before more data are obtained about

the stratigraphy, it is difficult and unreliable to

propose a more precise date within the Umayyad

period. Let us only say that the first half of the 8th

century is the more likely given the general trends

established by archaeologists for Umayyad mosque

building activities in Bilad al-Sham.

Discussion

Early Islamic congregational mosques were certainly

numerous in Bilad al-Sham, as reflected by textual

sources, by mosques with Umayyad origins that are

still in use and by archaeology. Nevertheless,

mosques that are still in use are often difficult to

consider with an archaeological perspective because

the exact extent of what is really Umayyad has, in

most cases, never been properly assessed.

Archaeology has contributed quite a lot to the

knowledge of Umayyad mosques, but most of

them are small structures related to princely and

aristocratic residences, leaving only a handful of

Umayyad congregational mosques that are well

known and can provide good comparative evidence.

In respect to the theme dealt with here, i.e. a con-

gregational mosque in a former antique and late

antique city, the relevant structures are ‘Amman

Downtown (Northedge 1992, 63–69), ‘Amman Citadel

(Almagro and Jimenez 2000), Jarash (Walmsley and

Damgaard 2005), ‘Anjar (Hillenbrand 1999), Qasr

al-Hayr al-Sharqi (Grabar et al. 1978, 46–51) and

Rusafa (Sack 1996).

The plan consisting of a hypostyle hall and a

courtyard is found everywhere for a congregational

mosque. Of course the rather irregular plan of the

Palmyra prayer hall is entirely due to the partial reuse

of a more ancient building combined with the

necessity of having a qibla wall properly oriented

towards Mecca. It would not have parallels else-

where. As hypothesized before, the prayer hall seems

to have been divided into three bays to the west,

respectively two to the east, and nine aisles. All the

other mosques cited before have three bays and it

seems to be the custom for urban congregational

mosques. Nine aisles are found in Jarash, Rusafa and

‘Anjar, while the two mosques in ‘Amman and Qasr

al-Hayr al-Sharqi have only seven aisles. It is not yet

known if there were porticoes in the courtyard in

Palmyra, but one can expect that they existed as in

every other congregational mosque, or that they were

at least planned, but not necessarily built, as in

Rusafa. For the moment, the closest architectural

parallels are ‘Amman Downtown, Jarash, Qasr al-

Hayr al-Sharqi and Rusafa. The hypothesis of two

different phases for the qibla wall needs more

research and, if confirmed, would prove to be very

interesting. It would be the first archaeological

attestation of such an evolution for an Umayyad

mosque in Bilad al-Sham.

The overall dimensions of the mosque (41 642.50 m, considering the width of the qibla wall and

the longest side) correspond roughly to the dimen-

sions of the Jarash mosque (38.90 6 44.50 m) or of

the Qasr al-Hayr al-Sharqi mosque (38 6 48.50 m).

That makes the structure one of the smallest of the

Umayyad congregational mosques built in Bilad al-

Sham. But the dimensions and proportions of the

Palmyra mosque fits well with a trend that prevailed

during the reign of Hisham b. ‘Abd al-Malik (H 105–

125/AD 724–743), as recently demonstrated by

Walmsley and Damgaard (2005, 375).

The situation of the Palmyra mosque in the centre

of the ancient city, precisely at the main crossroads

and at the end of a market street, is also strongly



reminiscent of other antique urban centres modified

by the Umayyads or new Umayyad urban settle-

ments. Of course, the closest parallel is offered by

Jarash, with its congregational mosque situated to the

south-west of the South Tetrakonia, at the intersec-

tion of the Cardo and South Decumanus. As in

Palmyra, shops bordered the eastern side of the

Jarash mosque and a larger Umayyad market area or

suq was identified around the plaza of the South

Tetrakonia and perhaps along the western part of the

South Decumanus (Walmsley 2003, 112–13). In

Rusafa, an L-shaped market or suq street led to the

congregational mosque situated north of the Saint

Sergius basilica. In ‘Anjar and at the ‘Amman citadel,

two new urban settlements, a large colonnaded street,

respectively a colonnaded plaza, bordered by many

shops are the main accesses to the centrally placed

congregational mosques.

All the previously cited examples of urban centres,

whether new (‘Amman Citadel, ‘Anjar and Qasr al-

Hayr al-Sharqi) or antique (‘Amman Downtown,

Baysan, Jarash and Rusafa, to which should be

added Damascus, Jerusalem, Homs, Busra or

Aleppo) testify to the involvement of the Umayyad

Marwanid caliphs in large-scale urban development

or refurbishment. Only a strong political power was

able to impose this kind of modification to the urban

core of large cities or to create new ones, and was able

to make the necessary investments. Mosques were

linked to the affirmation of a new religion and a new

political power. Markets were linked to economic

development. Nevertheless, one may wonder if these

latter should be interpreted as proof of a still thriving

and continuous economic life in Bilad al-Sham or as

an attempt by the new rulers to revitalize a declining

economy. This is, of course, a very large debate that

cannot be properly addressed within the parameters

of this paper.

Palmyra, with its Umayyad mosque and suq,

become fully integrated in this urban development

phenomenon. The city and its territory — the latter

including several important caliphal and aristocratic

residences or new urban settlements — are strong

indications of the vitality of the region. Palmyra was

still a significant regional centre. Nevertheless, unlike

other urban centres, Palmyra did not seem to have

had an important administrative role. Rusafa was

temporarily Hisham b. ‘Abd al-Malik’s capital.

‘Amman was the seat of the governor of al-Balqa’,

a sub-governorship of the jund Dimashq, and also had

a garrison and a mint (Northedge 1992, 49). Jarash

and Baysan were administrative centres, capitals of

districts (kura) of the jund al-Urdunn, and both had a

mint (Walmsley 2003, 111–12). Therefore, there

should be another reason for the important develop-

ment and vitality of Palmyra and Palmyrena under

the Umayyads and it can be hypothesized that it was

their political role. As mentioned at the beginning of

this contribution, they were the core area of settle-

ment of Kalb, one of the leading tribal groups in

terms of support to the Umayyad caliphate. This

support, shown for example at the time of the coming

of the Marwanid branch, was certainly as important

to the caliphs as the governorship of regional

subdivisions of the ajnad and would justify the

development of central Palmyra and the different

investments in Palmyrena. Both were the scenes

where the caliphs were exercising their political

power.

Conclusion

The rediscovery of a large mosque in central Palmyra,

which can very likely be attributed to the Umayyad

period, increases the knowledge of the ancient city

during the early Islamic period and allows its

importance to be reassessed in comparison with other

Umayyad urban centres. Yet, if the general organiza-

tion of the mosque and of the pre-existing building is

quite clear, there remains much to do to gain a

detailed understanding of them. This is also the case

for a more precise date of the mosque. Given the

architecture of the structure and the present state of

research about Palmyra, the Umayyad period seems

to be the most satisfactory date, but it is not definitely

proven. Therefore, more fieldwork needs to be done

on both the Roman building and the early Islamic

mosque. This will be done by a new Syrian-Swiss

project currently being organized by the author. It

will include a complete architectural analysis of both

structures and a reassessment of their stratigraphy by

a few soundings in sectors that were not excavated

before. This project should make possible a better

knowledge of the function and evolution of the

Roman building. It should enable the date of the

mosque to be assured and provide an understanding

as to precisely how it was inserted into the previous

building. It should also afford the detailed study of

the architecture of one of the few early Islamic

congregational mosques known archaeologically in

Bilad al-Sham. Finally, it will shed new light on the

transition from late Antiquity to early Islam in

Palmyra.

Acknowledgements

The preliminary investigations on the Palmyra

mosque were done in April 2006 by Marie-Cecile



Bosert (surveyor, Paris), Hugo Amoroso (archaeolo-

gist, Lausanne) and Denis Genequand, with the help

of Walid al-As‘ad (director of the Palmyra office of

the DGAMS), during the 2006 season of the Syrian-

Swiss archaeological mission at Qasr al-Hayr al-

Sharqi. I wish to thank the Directorate General of

Antiquities and Museums of Syria, its General

Director, Dr Bassam Jamous, and the Director for

Excavations and Archaeological Studies, Dr Michel

al-Maqdissi, for welcoming us, for help and support,

and for allowing us to work on Umayyad settlements

in Syria. It is also the occasion to thank again the

Swiss-Liechtenstein Foundation for Archaeological

Research Abroad (SLFA, Zurich) for their support

and for funding the project at Qasr al-Hayr al-Sharqi.

I am grateful to Isabelle Ruben (Amman) for

correcting the English text and to Marion Berti

(SCA, Geneva) for producing the final version of

Figure 5. I also wish to thank Jean-Baptiste Yon and

Thibaud Fournet (CNRS, Lyon) for answering some

questions and inquiries and for providing the plans

for Figures 1 and 2, plans that were first published in

their Inscriptions de Palmyre. Promenade epigraphique

dans la ville antique de Palmyre (IFAPO: Beyrouth-

Damas-Amman, 2001), and Michel Gawlikowski for

giving some indications on the date of abandonment

of the area he is studying in central Palmyra.

Bibliography

Sourcesal-Baladhuri, Ahmad b. Yahya’ b. Jabir, Kitab Futuh al-Buldan. Ed.

M.J. De Goeje (1866). Brill: Leiden.

Ibn al-Faqih al-Hamdani, Ahmad b. Muhammad, Mukhtasar Kitab al-

Buldan. Ed. M.J. De Goeje (1885). Brill: Leiden (BGA V).

al-Isfahani, Abu ’l-Faraj ‘Ali b. al-Husayn, Kitab al-Aghani. Dar al-

Tiba‘a: Bulak, 1285 (1868).

Michael the Syrian, Chronique. Ed. and trans. J.-B. Chabot, 4 vols

(1899–1910). Leroux : Paris.

al-Muqaddasi, Shams al-Din Muhammad b. Ahmad, Ahsan al-Taqasim

fi Ma‘rifat al-Aqalim. Ed. M.J. De Goeje (1906). Brill: Leiden

(BGA III).

Notitia Dignitatum. Ed. O. Seeck (1876). Weidmann: Berlin.

al-Tabari, Abu Ja‘far Muhammad b. Jarir, Ta’rikh al-Rusul wa-l-

Muluk. Ed. M.J. De Goeje et al. (1879–1901). Brill: Leiden.

Yaqut, ibn ‘Abd Allah al-Hamawi, Mu‘jam al-Buldan. Ed. F.

Wustenfeld (1866–73). Brockhaus: Leipzig.

WorksAl-As‘ad, K. and Stepniowski, F.M. (1989) The Umayyad Suq in

Palmyra. Damaszener Mitteilungen 4, 205–23.

Almagro, A. and Jimenez, P. (2000) The Umayyad Mosque of the

Citadel of ‘Amman. ADAJ 44, 459–75.

Bounni, A. (1963) Hawla al-Mashru‘ al-Tadmurı al-Istithua’ı. AAAS 13,

115–24.

—— (1995) Vierzig Jahre syrische Ausgrabungen in Palmyra. Pp. 12–20

in A. Schmidt-Colinet (ed.), Palmyra. Kulturbegegnung im

Grenzbereich. Philipp von Zabern: Mainz am Rhein.

Bounni, A. and Saliby, N. (1968) Fouilles de l’annexe de l’agora a

Palmyre. Rapport preliminaire. AAAS 18, 93–102.

Bylinski, J. (1999) Qal‘at Shirkuh at Palmyra: a Medieval Fortress

Reinterpreted. Bulletin d’etudes orientales 51, 151–208.

Devreesse, R. (1945) Le patriarcat d’Antioche depuis la paix de l’Eglise

jusqu’a la conquete arabe. J. Gabalda: Paris.

Gawlikowski, M. (1974) Les defenses de Palmyre. Syria 51, 231–42.

—— (1992) Palmyra 1991. Polish Archaeology in the Mediterranean 3,

68–76.

—— (1994) Palmyra 1993. Polish Archaeology in the Mediterranean 5,

133–43.

—— (1998) Palmyra, Excavations 1997. Polish Archaeology in the

Mediterranean 9, 197–211.

—— (1999) Palmyra, Excavations 1998. Polish Archaeology in the


—— (2002) Palmyra, Season 2001. Polish Archaeology in the


—— (2005) Palmyre chretienne et sa cathedrale. Dossiers d’archeologie

309, 70–71.

Genequand, D. (2004a) Chateaux omeyyades de Palmyrene. Annales

islamologiques 38, 3–44.

—— (2004b) Al-Bakhra’ (Avatha), from the Tetrarchic Fort to the

Umayyad Castle. Levant 36, 225–42.

—— (2005) From ‘desert castle’ to medieval town: Qasr al-Hayr al-

Sharqi (Syria). Antiquity 79/304, 350–61.

Grabar, O., Holod, R., Knustad, J. and Trousdale, W. (1978) City in

the Desert. Qasr al-Hayr East. Center for Middle Eastern Studies

of Harvard University: Cambridge, Mass.

Hillenbrand, R. (1999) ‘Anjar and Early Islamic Urbanism. Pp. 59–98

in G.P. Brogiolo and B. Ward-Perkins (eds), The Idea and Ideal of

Town between Late Antiquity and the Early Middle Ages. Brill:

Leiden-Boston-Koln.

Huart, Cl. (1929) Inscriptions arabes de Palmyre. REI 3, 238–44.

Khamis, E. (2001) Two Wall Mosaic Inscriptions from the Umayyad

Market Place in Bet Shean/Baysan. BSOAS 64/2, 159–76.

Kowalski, S.P. (1997) Late Roman Palmyra in Litterature and

Epigraphy. Studia Palmyrenskie 10, 39–62.

Michalowski, K. (1960) Palmyre. Fouilles Polonaises 1959. PWN:

Warszawa.

—— (1962) Palmyre. Fouilles Polonaises 1960. PWN: Warszawa.

—— (1963) Palmyre. Fouilles Polonaises 1961. PWN: Warszawa.

—— (1966) Palmyre. Fouilles Polonaises 1963 et 1964. PWN:

Warszawa.

Northedge, A. (1992) Studies on Roman and Islamic ‘Amman. Volume 1:

History, Site and Architecture. Oxford University Press: Oxford.

Ostraz, A. (1969) Note sur le plan de la partie mediane de la rue

principale de Palmyre. AAAS 19, 109–20.

Sack, D. (1996) Resafa IV. Die Grosse Moschee von Resafa—Rusafat

Hisham. Philipp von Zabern: Mainz am Rhein.

Sauvaget, J. (1931) Inscriptions arabes du temple de Bel a Palmyre.

Syria 12, 143–53.

Schaefer, J. and Falkner, R.K. (1986) An Umayyad Potters’ Complex

in the North Theatre, Jerash. Pp. 411–59 in F. Zayadine (ed.),

Jerash Archaeological Project 1981–1983. Department of

Antiquities: ‘Amman.

Schlumberger, D. (1986) Qasr el-Heir el-Gharbi. Geuthner: Paris.

Seyrig, H. (1950) Antiquites Syriennes. Syria 27, 239–42.

Ulbert, T. (1992) Beobachtungen im Westhofbereich der Grossen

Basilika von Resafa. Damaszener Mitteilungen 6, 403–16.

Walmsley, A. (2003) The Friday Mosque of Early Islamic Jarash in

Jordan. Journal of the David Collection 1, 110–31.

Walmsley, A. and Damgaard, K. (2005) The Umayyad Congregational

Mosque of Jarash in Jordan and its Relationship to Early

Mosques. Antiquity 79/304, 362–78.



D. Genequand, An Early Islamic Mosque in Palmyra, 2008

Documents