Visit Us | Contact Us | Membership | Make a Gift | Calendar | Order Online | What's New

Print this Page

Home > Research > Publications > Annual Reports > 1997-98 Annual Report

The Aqaba Project


Donald Whitcomb

Archaeological research in the Aqaba region has revealed a succession of settlements, from the Chalcolithic to the modern period. The excavations of the Oriental Institute, beginning in 1985, produced a completely unexpected chapter in this story, the remains of the early Islamic city of Ayla in the heart of the modern city. This town was occupied from ca. ad 650 to the arrival of the Crusaders in ad 1116, a period of ca. 450 years. The archaeological evidence of this period provides important information on the sequence of cultural changes between Late Antiquity and the formation of Early Islam. Further, it provides documentation on the stages of development during the great fluorescence of the Abbasid and Fatimid periods antecedent to the transformations of the medieval world. The Islamic Aqaba Project has been engaged in the interpretation of walls and sherds in light of these historical contexts, while being concerned for the touristic development and explanation of this site in the modern city (as discussed in the 1996/97 Annual Report).

The interpretative understanding of archaeological evidence holds an importance at least equal to the process of discovery. This will be an obvious statement for members of the Oriental Institute, yet needs some repetition for many others, especially some governmental agencies. With this in mind, the Islamic Aqaba Project has shifted its emphasis to processing information already recovered and, perhaps more importantly, examining this data in the context of a new theory of urbanism, the explication of the beginnings of the Islamic city. The study of the early Islamic city may profitably take its focus from study of the so-called desert castles located in Bilad al-Sham (Syria-Palestine). These monuments, mostly dated to the late Umayyad period, have been intensively analyzed from art historical vantages, with multiple hypotheses on their functional rationale. More recently, however, the author has combined the results from Aqaba with evidence from Anjar in Lebanon and other sites to bring a more nuanced perspective to these sites.

The present thesis hypothesizes that these early Islamic settlements were all constructed as incipient urban entities. They contain structural elements typical of more recognizable cities, such as bath houses, gates, the palatial or better administrative structures, mosques, and residential elements. Thus monuments, from the perspective of internal archaeological context, are considered as aspects of urban planning. The Muslim conquest initiated a conscious attempt to recreate specific morphological features that constituted an urban pattern characteristic of western and southwestern Arabian culture. The institutional components of this South Arabian city were adapted to the religious, administrative, and commercial needs of the new Islamic polity, a transformation which set a trajectory for medieval cities throughout the Middle East (and perhaps even Europe of the early Middle Ages).

Thus an Arabian concept of urbanism lies at the foundation of the early Islamic city; the existence of a distinctive "Islamic city" from the beginnings of Islam begins to take form with specific archaeological characteristics. This hypothesis is derived from Aqaba and other urban plans and can be tested on other sites in Arabia and the Levant.

Spatial Patterns in the Early Islamic City

An initial breakthrough in the study of structural elements centered on the location of the bath house. In the early Islamic city, the bath house (hammam) was a primary urban element, one which adopts a Hellenistic technological apparatus. While one finds baths in extra-urban situations, there seems to be a pattern within cities. Analysis of a number of sites revealed a constant relationship in distance and direction to the "palace" (or administrative center), with the bath house located to the north (or northwest) at 50-60 m distance. Further, the bath is often located just east of the north gate, which appears to function as the principal entrance into the city.

The city of Ayla is oriented with corners to the cardinal directions; nevertheless, the northeast gate appears to be associated with the direction of Syria and hence a functional "north." As luck would have it, there is no evidence for the location of a bath house at Aqaba, though some reports during building of the Corniche road suggest that hypocausts might have been found east of the Syrian Gate. During the 1995 season of excavations, a large building, decorated with external pilasters, was found just northeast of the Central Pavilion. If this was the administrative structure or "palace" of Ayla, its location would be the predicted distance and direction from the putative bath and north gate (fig. 1).

In general, there were two administrative structures in the early Islamic city. One of these was the Dar al-Imara, the place of the amir, of the social leadership, and of the military. The second structure was the Balat (a term deriving from Palatium), related to the Diwan, to the administrative apparatus, to the financial offices, the bureaucracy. Both of these buildings had fixed locations relative to the mosque. The Dar al-Imara was located to the qibla (south) side of the mosque; and the Balat was located west of the palace. This topographic configuration would correspond to the axial relationship of the Ayla congregational mosque, excavated in 1993, and the palace structure mentioned above. The model would predict a Balat or financial offices to be located in the unexcavated area west of the palace.

Nothing is ever completely straightforward, even in the best of models. The mosque excavated in Aqaba is a secondary structure, belonging to the Abbasid period (after the mid-eighth century), which imitates the original congregational mosque. The Umayyad mosque, which was probably that founded by `Uthman ibn `Affan, is not beneath the later mosque and has not been found. Another pattern may be postulated: location of the palace on the qibla side of the mosque appears to have been an innovation of the caliph al-Mu'awiya (ca. 680). Before this time, the palace seems to have been either north or east of the mosque, possibly in imitation of the locational relationship between the house of the Prophet and the mosque (haram) in Medina. This pattern would suggest that the first mosque at Ayla will be found south of the palace building. The area has not been excavated but lies in the path of the wadi; if the wadi is a fault formed in the 748 earthquake, then there was an ample reason for abandoning its first location and building an imitation on solid ground to the north of the palace.


This very brief description of a model of the early Islamic city has an obvious function in providing an empirical basis for understanding Ayla or other archaeological sites. Perhaps of more interest for the Oriental Institute is the strong indication that, among other aspects, the early Islamic city was a ritual city. Rather like Persepolis, the ritual city is a functional framework that has underpinned urbanism since very early times. One may trace this phenomenon from the ancient Near East into its late antique manifestations in southwest Arabia. In this sense, the early Islamic city was an oriental phenomenon, one of cultural continuity which encompassed varied tribal and ethnic identities into the Islamic community (umma). This was an Arabian concept of urbanism that proved appropriate as the material referent for the theocratic state which began in the Hijaz during the early seventh century.

It may not be inappropriate at the Oriental Institute to express the generalizing pattern of this urbanism in an adaptation of Robert Braidwood's famous "dipchart," originally intended to display much earlier patterns of settlement (fig. 2). Like Jarmo, the centrality of the excavations at Aqaba may prove to be accidental with accumulation of much further archaeological study.

Revised: February 7, 2007

Home > Research > Publications > Annual Reports > 1997-98 Annual Report