Umm el-Jimal

Continuity and Change in the Urban Character of the S. Hauran from the 5th - 9th c. AD: The Archaeological Evidence at Umm el-Jimal

TABLE OF CONTENTS

• Summary
• Introduction: Umm el-Jimal before the fifth century
• The Byzantine Town, 5th-7th C.
• Islamic Umm el-Jimal
• Continuity
• Change
• Conclusion: Some concluding statements and thoughts
• Bibliography

Summary

Pre-Islamic Umm el-Jimal was less affected by Roman grid-planning than many settlements in Jordan and hence never looked like cities such as Jerash. Umm el-Jimal grew rapidly in the 5^th c., reaching its optimum size in the 6th, and received a serious refurbishing in the 7th. This remodeling involved the thorough removal of pre-Islamic occupational layers, which significantly transformed the internal material culture without, however, changing the general architectural frame of the town. This late 6th / early 7th remodeling phase can be seen as a clear example of regional early Islamic (Umayyad/Abbasid) culture, because it remains mostly unaltered by the post-9th c. phases. Moreover it may be viewed as the proximate context of the contemporary "desert castles."

Back to Top

Introduction: Umm el-Jimal before the fifth century

The least understood phase of UJ's archaeological history is its earliest, the 1st-4th c. A. D. The most extensive remains of this period are at el-Herri, the rough-strewn area of ruins to the SE of the surviving Byzantine/Islamic site. There, founding levels go back to the 1st c. A. D., contemporary with the Nabataean hegemony at Bostra. These are succeeded by a 2nd c. phase, contemporary with the coming of the Romans, a 3rd/4th c. abandonment, and 4th c. use as a dumping site. The building remains examined so far indicate simple structures built of mostly undressed basalt cobbles and boulders, mostly domestic in character (de Vries 1986; 1993: 443-444, 450-451; Momani and Horstmanshof 1995).

Contemporary with that apparently local settlement, more formal Roman construction took place beginning in the 2nd c. (e.g. the gate dedicated to M. Aurelius and Commodus) and continued with the construction of the so-called Praetorium and the large reservoir. Though the three just mentioned survived, other structures did not. Evidence of relatively monumental imperial style structures appears only in a number of decorative fragments built as spoils into later phase architecture, taken from buildings being dismantled in proximity of the Commodus Gate and the Praetorium.

Thus the monumental character of the Roman imperial era was largely lost in the subsequent extensive building phases of the 4th and 5th - 8th centuries. It is safe to say that at Umm el-Jimal classical monumental Roman and Nabataean models, having disappeared by the 4th c., played virtually no role in the architectural shaping of the Umayyad /Abbasid phase of the settlement. A strong exception to this is the Praetorium, which received a 7th/8th c. refurbishing (Robin Brown in de Vries 1998: 161-193). Though of complex design, this building's exterior was decoratively austere, with the only classical "touch" the four late antique ionic-capitalled columns of the atrium.

Even the earlier, Tetrarchic, castellum of the early 4th c., though a deliberate Roman imperial construction on the model of legionary camps (de Vries 1986: 1993: 435-436), included spoiled decorative architectural elements from the earlier building phase, re-tooled as construction blocks without regard to decorative value. Locally, the precedent for the late antique austere style was the functional, undecorated Barracks (de Vries 1982: 111-112; Parker 1986: 26-29, in de Vries 1998: 131-142), with its foreboding, blank exterior walls, single entry and strictly utilitarian interior barracks rooms surrounding the large interior courtyard. This, rather than the more elaborate Praetorium, served as a model for Umm el-Jimal's late antique houses, the "rustic villas" of the 5th to 8th century. What survived this metamorphosis from Roman to late antique was the tradition of careful and structurally sound design, for which the basis must be the accumulated experience of imported Nabataean, Roman and local construction traditions visible in late antique structures throughout the region.

Back to Top

I. The Byzantine town, 5th-7th C.

Thus, the construction surge of the 5th and 6th c. which gave Umm el-Jimal its characteristic shape benefited from Roman construction techniques, but was already far removed from Roman monumental design and gridded town planning (de Vries 1985, 1988, 1998: 91-127). The rapid growth of the 5th-7th c. proceeded according to a non-gridded plan, in which clusters (or insulae) were laid out with rectangular courtyard houses at the core. Such houses, which may have their heritage in both the traditional East Mediterranean house of earlier periods and the Roman rustic villa, were plainly constructed, with austere exteriors and simple but warmly secure interiors.

The clusters consisted of several such houses attached to each other, with those on the perimeter sometimes of irregular, trapezoidal plan necessitated by the available space to the property line. Dividing lines between such clusters form the streets and allies of irregular width which twist and turn, and sometimes open onto small open areas between houses reminiscent of the hosh of early modern Jordanian villages.

The open space left between the several groups of these clusters were mostly partitioned by field walls, defining animal corrals, and possibly, garden plots. The net effect is that shared or public spaces were limited to the area east of the Commodus Gate (with the Klaudianos Church at the northern and the Cathedral at the southern end). Such spaces would not have served the public function of the open streets and squares of the Roman cities, but could have been the camping locale for caravans, had Umm el-Jimal played a role in regional and interregional trade and transport in the 6th/7th c. (de Vries 1998: 236-239).

This town was also devoid of public buildings in the classical sense. The Praetorium was by now privatized, that is, incorporated into a newly constructed domestic complex (R. Brown in de Vries 1998: 161-193). The same happened to the so-called Nabataean Temple (really constructed in the 4th c., on the eve of Umm el-Jimal’s adoption of Christianity, S. T. Parker and L. de Veaux in de Vries 1998: 149-160). The Barracks (or "later castellum") appears to have lost its military role in a remodeling that included the addition of the tower with its famous Christian inscriptions. (Whether it continued in a military role, became monastic, or served some other purpose is not clear Parker 1986: 29, in de Vries 1998: 140. But if the open spaces were used for caravans, this building is the only available candidate for the role of hostel or caravanserai.)

A major investment in public construction was the repair of the town wall and gate system (S. T. Parker in de Vries 1998: 143-147), which, along with the new lookout-tower of the Barracks, provided local protection at the time that the imperial army had relinquished its policing role to the mobile army of the Arab phylarch.

The fifteen churches were the only other public structures (Schick 1995: 469-472; maps in de Vries 1981: 58-59, 1982: 102-103 or 1998: 14, 49). However, most of these (eight) were built into private complexes, some of which, like that of the Double Church, were domestic (de Vries 1988). Four of these had their entries from the enclosed courtyard of those complexes, so that they appear to be entirely private, from an architectural planing point of view. The other four, as well as the remaining seven (that is, eleven in total) had entrances accessible from the street. Most of these, however, are not centrally located, and give the feel of accessibility to the residents of the housing clusters into which they are built. Of the three free-standing churches, the Numerianos Church, the West Church and the Cathedral, the Numerianos Church was built by private donors, and had its entrance accessible from the attached cloister. The West Church, so prominent in the visual landscape today, stands in a large walled courtyard, accessible through an impressive but lockable gate, which pierces the earlier town wall just south of the Commodus Gate. This leaves the Cathedral, with its central location in main open area of the town and with street access to its triple-doored narthex, as the only clear candidate for a truly public church that would have been readily accessible to visitors and strangers as well as residents.

Such was the shape of Umm el-Jimal on the eve Islam: a sturdily built and elaborately laid out community with both building design and town plan very different from the imperially sponsored Roman cities built in the 2nd-3rd centuries. The classical stress on public spaces had been replaced by an equivalent stress on private space. Evidence for participation in public life remained in the maintenance of the perimeter wall, the shared use of intra muros open spaces, and the openness to the general populace of at least some of the churches.

Back to Top

II. Islamic Umm el-Jimal

A. Continuity

Architecturally, the basic shell of the Byzantine town just described became the "container" for the Umayyad/Abbasid settlement (de Vries 1982: 70-71, 1993: 448-449, 1995: 430-431, 1998: 231-232). Comparison of the town map as it was c. A.D. 550 to that of ca. A.D. 700, would reveal essentially the same plan. Though other aspects of material culture shifted (pottery, for example -- better dealt with by Dr. C. J. Lenzen) the architectural aspect of society remained virtually the same. There is no indication of a drastic interruption, nor of a population turnover. In the main, the same population appears to have continued, and the major change is a gradual decrease in its size, down to the end of occupation of the site ca. A. D. 900. Religiously, also, society remained mostly Christian, though there is also some evidence for the introduction of Islam (see below). Thus, the architectural shape of Islamic Umm el-Jimal, and by extension, of similar communities in the southern Hauran may be termed as continuous with that of the basically un-Roman 6th-7th c. Byzantine town.

B. Change

Evidence of deterioration in the 7th c. comes from the presence of Umayyad occupation levels on top of the collapse debris of rooms immediately west of the Barracks (later castellum) corner tower. Though no repair was done here, such deterioration elsewhere may have led to major refurbishing of much of the town late in the 7th and early in the 8th century. In both the Praetorium complex and House XVIII (that of the double windows) excellent plaster-on-cobble floors in all rooms furnish evidence for such thorough refurbishing (R. Brown in de Vries 1998: 178-184, 195-202). In at least one case, House 119 (east of the Barracks) an entirely new house was constructed on Byzantine foundations (de Vries 1993: 428-431), its stables including characteristically Umayyad manger rows (like those in the burned stable at Pella). The water system was maintained and extended to the reservoir at the SE corner of the town. The NE Church may have been constructed de novo, with the apse area left unfinished, but the chancel area paved with characteristic plaster-cobble flooring, painted with red, white and black "tile squares" -- an alternative to the earlier use of mosaic flooring (e.g. in the Numerianos Church). Other churches (e.g. the north half of the Double Church) were remodelled as churches in the Umayyad period.

At least two churches appear to have been converted to mosques, the Numerianos and the West Churches. Such conversion consisted of the construction of a blocking wall across the apse, using spoiled masonry made available by the nearby construction activities (e.g. at House 119), and, possibly, the creation of a qibla by removing some masonry blocks from the interior face of the south wall. Though Christian symbols are prevalent throughout the churches and houses, no Islamic symbols have been found at the site. In the Numerianos Church, however, it appears that the conversion included the careful removal of Byzantine occupation debris from the latest pre-Islamic plaster floor by heaving it over the blocking wall into now-abandoned apse.

This specific instance points to a very interesting feature of the Umayyad refurbishing, which may be called the "cleanup phenomenon." Before the construction of the solid Umayyad floors in the Praetorium and House XVIII all earlier occupation debris was carefully removed down to preoccupation levels, so that the Umayyad floors were actually installed at levels lower that the earlier ones. In case of House 119 the Byzantine structure was dismantled and all occupation debris removed, so that all that was left was a single cobble and cement Byzantine pavement, which was so hard that it would have required the ancient equivalent of a jack-hammer to remove it. At House 119, this scrape-down produced a two-meter thick dump fan of numerous alternating ash an soil lenses, in which the latest pottery was 6th/7th c. Byzantine from top to bottom (de Vries 1995: 428-431). Similar dumps remain as distinctive mounds spaced throughout the site to indicate that such cleaning took place everywhere. Two others we tested, the thick mound NE of the Roman Reservoir and the huge fan outside the West Gate, revealed the same single period stratigraphy (J. Sauer and B. de Vries in de Vries: 1998: 129-130).

Such a radical cleanup may be indicative of a break with past culture, but it is probably motivated by hygienic necessity. Ancient communities that lasted for centuries tended to "drown" in their own accumulated debris. At Umm el-Jimal an earlier instance is the general cleanup of the 1st-4th c. accumulated debris and its removal to the neighboring ruined remains of el-Herri in the 4th c. (Momani and Horstmanshof 1995). Thus, in the Umayyad case, the cleanup involved the removal of debris accumulated from dense occupation stretching over at least three centuries. That the debris was stored by the formation of dumps on the site itself, may indicate a less serious demand for space in the 7th/8th c. than may have been true in the 4th. One might like to think that such radical cleansing was a specifically targeted health measure, but there is no literary evidence for the association of the bubonic pandemic, which ravaged Umayyad society so severely, with community hygiene (de Vries 1998: 240; see Conrad 1981). Rather, the residents may have been faced with a practical choice: Either abandon the refuse-laden site for a new location (as was done in late antique Egypt) or clean it up in order to take advantage of the already existing facilities. The existing buildings and the available water system made it worth staying, to clean up and remodel.

Back to Top

Conclusion: Some concluding statements and thoughts

1. The transition from Byzantine to Umayyad Umm el-Jimal appears to be less radical than the transition from Roman to Byzantine Umm el-Jimal.

2. The non-Hippodamic character of Umm el-Jimal's town plan was determined in the Byzantine era, and inherited by the Islamic. The relatively austere lack of decoration and monumentality is also continuous from the Byzantine to the Islamic.

3. The changes from Byzantine to early Islamic culture at Umm el-Jimal include significant non-religious and non-political factors, among them gradual population decline, and the major cleanup and refurbishing to preserve site viability. However, both the sponsorship of the extensive remodeling and the conversion of churches to mosques may be directly associated with the activities of the Islamic Imperial center at Damascus.

4. The 7th/8th c. flourishing of Umm el-Jimal is in the context of the construction and early use of the desert castles. The refurbished Praetorium, though not a bath, would not have felt unlike Hammam es-Sarakh or Qusayr Amra. The frescoes of Amra could derive from the same artistic milieu responsible for the church frescoes at Umm el-Jimal and contemporary communities like it. The artisans doing the remodelling at Umm el-Jimal could have belonged to the same pool of builders and artists responsible for the qusur.

5. Umayyad/Abbasid Umm el-Jimal represents a stratum of early Islamic culture, that by virtue of the town's 8th c. collapse and 9th c. abandonment, remained unaltered by subsequent occupation and building phases (except for the rather easily distinguishable late Ottoman (Druze) occupation of the early 20th c.). Its study can therefore give good insight into the urban-architectural character of early Islam in Bilad esh-Sham, outside the major centers (Bostra, Damascus, Jerusalem) where subsequent occupation added the deep layers of Medieval Islamic culture. Thus, early Islamic Umm el-Jimal's houses still stand to inspire Jordan's lively architectural creativity in the late 20th century.

Back to Top

Bibliography

1993 Egypt in Late Antiquity. Princeton: Princeton University Press.

Momani, Ahmad and Horstmanshof, Michael

1995 Umm el-Jimal Area R 1994 Field Season. Annual of the Department of Antiquities of Jordan 39, 469-476.

1995 The Christian Communities of Palestine from Byzantine to Islamic Rule (Princeton: The Darwin
Press, Inc.)

Back to Top