Khirbat al-Mafjar and Its Place in the Archaeological Heritage of Palestine

Abstract

Khirbat al-Mafjar is a well-known Umayyad palace complex located near Jericho. The monument is one of the most important cultural symbols of early Islamic archaeology in Palestine. This article discusses the site’s initial association with the caliphate of Walid ibn Yazid in light of new excavations. Recent discoveries suggest the existence of an early Islamic agricultural estate and the potential for a new understanding of early Islamic qusur or proto-urban settlements.

The archaeological heritage of Palestine for the Islamic periods, especially the formative early Islamic period, must focus on the monument of Khirbat al-Mafjar, known as Qasr Hisham, near the city of Jericho (Fig. 1). The extraordinary beauty of its vast hall of mosaic carpets, stone and plaster sculptures—including human and animal figures—has long captured the attention of Islamic art historians and archaeologists. However, aside from specialists, there has been little consideration for its role in the formation of Islamic civilization and its part in the cultural history of Palestine.

Recovery of the Monument

The first person to carefully investigate the site was Dimitri C. Baramki, who in the early 1930s as a young Palestinian inspector for the British Mandate Department of Antiquities, traveled from his home in Jerusalem throughout Palestine (Fig. 2). His attention was drawn to a series of large mounds north of the town of Jericho, where carved stones were being removed and taken for use in the town and local cemeteries.

Baramki began excavating in 1933 on the southern mound, revealing the fine-carved stones of a doorway and surrounding rooms along with stucco decorations (1936). His initial assumption of finding a Byzantine building soon changed as he recognized it was a quite rare early Islamic monument (Fig. 3). Baramki immediately realized the importance of these discoveries and wrote clear reports describing the archaeology of the site through the 1930s, with only a few comparative resources at that time.¹ When the excavations ended, he wrote a clear and concise booklet on Qasr Hisham ([1947] 1956) after his dissertation on these monuments (1953).

Formation of a Narrative

This archaeological analysis of Khirbat al-Mafjar stands in contrast to the writings of Robert W. Hamilton, then the young director of the British Mandate’s Department of Archaeology in Jerusalem. As early as 1945 he took an active hand in describing the stone and plasterwork of Khirbat al-Mafjar and began an architectural study for reconstruction of the spacious bath hall (Fig. 4).² These remains became the springboard for his personal fascination with early Islamic stories surrounding the Caliph Walid II (743–745 CE). This narrative, rather than the archaeology, would form his interpretation of the site in 1959 as an Arabian mansion (see below), so much so that ten years later, Hamilton applied early Islamic stories from the Kitab al-Aghani to Khirbat al-Mafjar (1969: 65–67). After criticism from Ettinghausen (1972), he defended his description of the bath hall as a “frivolity hall” with a weak linguistic argument, stating that the derivation of mafjar (usually indicating ‘flowing water’) as ‘the place of fujur (debauchery)’ described the site where these “ancient pleasures [are] dimly remembered” (1978: 138). Finally Hamilton published a novel, Walid and His Friends (1988), which used the buildings and embellishments of Mafjar as the setting for poetry and an “U mayyad tragedy.”³ It is important to note that the monuments provide no direct evidence for this in terpretation.⁴ As Walmsley has summarized recently, “ . . . the widespread representation of human and animal figures . . . in an Islamic building invited explanation as to its inspiration and purpose, on which debate still continues” (2007: 19) (Fig. 5).

Fig. 1.

View of Khirbat al-Mafjar from the southeast. (Photo by D. Whitcomb.)

The narrative proposed by Hamilton carries certain assumptions. The first is the chronology of the buildings, confined to the second quarter of the eighth century (i.e., Hisham’s reign, 723–743 CE). In 1942, Baramki had already clearly demonstrated the continued occupation of Mafjar into the Abbasid and later Ayyubid periods (eighth–twelfth centuries CE) through analysis of the ceramics, which showed continued use of the palace, if not other buildings (Baramki 1944a; Whitcomb 1988). The patronage of Walid II was never accepted by Creswell (1969: 574–76) and was explicitly rejected by Baramki (1953: 47–49), although criticisms have been less frequent, especially with the admittedly enjoyable book on Walid. Art historical interpretations still vary as evidenced by Bisheh (2000), who dismisses one fanciful interpretation of the fruit-and-knife motif, and Soucek (1993), who provides an alternative (and attractive) interpretation of the bath sculptures as depictions of the legends of Solomon.

Fig. 2.

Dimitri Baramki with workmen at an excavation in the 1930s. (Courtesy of Constantine Baramki and the Department of Antiquities and Cultural Heritage, Palestinian Authority.)

Fig. 3.

General plan of the palatial complex of Khirbat al-Mafjar. (Drawing by D. Whitcomb and the Department of Antiquities and Cultural Heritage, Palestinian Authority.)

Fig. 4.

View of the entrance into the bath from the east. (Photo by J. Yasin.)

Fig. 5.

Colorized reconstruction of the façade of the bath. (After Hamilton 1959: fig. 52.)

Archaeology of Khirbat al-Mafjar

Scholars have usually taken the monumental volume, Khirbat al Mafjar: An Arabian Mansion in the Jordan Valley, as a final report on this extraordinary archaeological site (Hamilton 1959). This was published without the direct involvement of Baramki, who had written his dissertation on Mafjar in London in 1953. A careful examination of this monograph reveals that it is not a complete archaeological report and one must still use Baramki’s four preliminary reports published in the Quarterly of the Department of Antiquities in Palestine (1936, 1937, 1939, 1944b). Hamilton’s book describes the architecture of the palace, bath, mosque, and pavilion with a focus on embellishments in carved stone and stucco, fresco paintings,⁵ and the magnificent mosaic carpets. He presented for the first time the mosaic called the Tree of Life (Fig. 6), found in the Diwan or private reception chamber, which remains the most famous of Umayyad mosaics.

Many categories of artifacts are missing, such as the coins and inscriptions, pottery, glass, and small objects. Some of these crucial artifacts are found in Baramki’s preliminary reports, but other categories of evidence that one might expect from major excavations remain unavailable.

As a more specific instance, archaeological information on the bath hall and its surroundings, excavated from 1945 until 1948, were prepared as reports but never published.⁶ This archaeological data was apparently not of interest to Hamilton and only appears as an appendix to Baramki’s dissertation in 1953 (and had been of limited access).⁷ Initial analysis shows a continued occupation around the destroyed bath into the Abbasid period.

Fig. 6.

Famous mosaic known as the Tree of Life. (Courtesy of E. Cirelli.)

One of the Qusur

Baramki’s interpretation of the monument was as an example of the so-called desert castles, an architectural phenomenon that has a long and varied history in early Islamic archaeology, although its location on the edge of the fertile fields of Jericho stretches this descriptor. There is a growing consensus for the use of qusur(sing. qasr) for these monuments—the word having a general definition of a ‘palace’ with a root meaning of an ‘enclosure’ (Conrad 1981). An understanding of such sites is essentially an archaeological problem, as these places rarely have an ancient name or referent in literature or historical documentation. Discovery of the qusur has been accidental and continuous; Baramki had only two or three comparanda in the early 1930s while Genequand (2010) can count some 38 foundations in his recent thesis (many his own discoveries). A foundation in the Umayyad period is usually accepted and occupation continues into the early Abbasid period at least;⁸ there are pre-Islamic antecedents and medieval re-occupations for many examples.

The elusive key to understanding Qasr Hisham lies in the nature of the Umayyad qusur and the varying explanations of this phenomenon. These buildings have the following attributes or elements:

1. 1. A qasr has sets of rooms or bayts arranged around a courtyard. The structure is two storied with a single entry, and the form is not dissimilar to that of later caravanserais. A number are embellished with decorative features, especially stuccos.

2. 2. A bath or hammam often lies to the north at a short distance. The building is a combination of a hall with pools and then a sequence of bathing rooms of increasing heat.

3. 3. A mosque may be included. The location seems to vary, either between the qasr and hammam or southeast of the palace entrance.

4. 4. Residential structures may be found as isolated buildings or grouped into larger compounds; they tend to be located to the west or north of formal elements.

5. 5. A large enclosure or hayr is delineated by a buttressed wall, stretching eastward from the site for about three kilometers. There are varied interpretations of such hayrs that include agricultural/orchard lands or an animal preserve or park.

Some of the qusur have all these characteristic elements but many have one or only a few. The position taken here is that inclusion of these elements may have been cumulative, leading toward an “urban” entity (see below). Many are obviously incomplete and patterns based on building location require further exploration. There is also a need for a comparative study to advance possible interpretations. For now, lively discussions among archaeologists must suffice.

Secondary Excavations at Khirbat al-Mafjar

As Mafjar passed to the authority of the Jordanian government, it witnessed a strange twist of fate. After the collapse of the British Mandate, the director of antiquities of Jordan became Dr. Awni Dajani (Fig. 7),⁹ whose father owned vast lands around Jericho including the fields in which these ruins lay buried. As a young man, Dajani must have witnessed and may have participated in Baramki’s excavations. He renewed the excavations in 1957, directly after submitting his doctoral dissertation to University College London.

Fig. 7.

Awni Dajani at his excavations in the Northern Area in the 1960s. (Courtesy of the Archives of the Rockefeller Museum.)

His excavations continued through the 1960s; regrettably, the only published reports are two short notes (Dajani 1958, n. d.).¹⁰ Dajani seems to have concentrated his efforts on the untouched Northern Area, about 70 meters north of the bath. This was the third or northernmost mound described by Bliss in 1894, whose plan includes walls exposed at that time from serious looting for building stones (1894: 178). The only plan available of these excavations was that made by the Palestinian Authority after they took control of the site in 1993 (Fig. 8), showing an amalgam of walls of at least three different architectural phases obscured by some thirty years of neglect.

Renewed Excavations

Current excavations began in 2006 under the direction of Hamdan Taha of the Palestinian Department of Antiquities and Cultural Heritage. The Jericho Mafjar Project, a collaboration between the Palestinian Department of Antiquities and Cultural Heritage and the University of Chicago, was created with the expressed goal of an integrated archaeological re-assessment of the original excavations and architectural studies by Baramki and Hamilton with new archaeological research. The initial assessment is based on excavations by Taha (2011) and studies by Whitcomb (1988, 1995).

There are two directions this archaeological re-assessment of Mafjar may take. The first is a careful stratification of the long history of the building complexes, an approach systematically pursued by Baramki (1944a, 1953). Based on Baramki’s evidence from the palace (1944a) as modified by Whitcomb (1988), an archaeological sequence of four periods may be proposed:

1. 1. 700–750 CE: Construction and destruction debris mixed with painted wares.

2. 2. 750–800 CE: Further occupation, suggesting less extensive damage from the earthquake in the mid-eighth century. Ceramics seem to be transitional types, similar to the Mahesh phase at Ayla (Aqaba) (Whitcomb 2001: 509).

3. 3. 800–950 CE: Major reoccupation of the site in the Abbasid period; continuities and introduction of cream wares (popularly known as Mafjar Ware), incised, molded, and glazed ceramics.

4. 4. 1100–1300 CE: Medieval reoccupation in the Ayyubid-Mamluk period; apparently limited use until the final destruction of the palace.

Areas selected for new investigations are focused on the transitional area northeast of the bath (following Taha 2011) and the neglected Northern Area. In general the chronological framework suggested above is confirmed, as three building phases have been found in every area (an initial Umayyad founding, Umayyad occupation, and an extensive Abbasid reconstruction). The final reoccupation in the Ayyubid period is evident only in the palace, though this is a preliminary assessment.

The second direction is an appreciation of the setting, the evolving context of the estate as an incipient urban place—an early Islamic city—and its relationship with Ariha (Jericho), the continuing Christian center of the entire oasis. Indeed, numerous sites around Jericho and salvage projects within the modern city continue to amplify our understanding of the late Byzantine and early Islamic periods. The archaeology of these two “cities” may provide a model for understanding various relationships during this crucial period in the Levant and the Middle East.

What We Have Uncovered

The results of the initial two seasons may be summarized very briefly and are available on the website.¹¹ Our first trench might be considered “beginner’s luck” since the North Gate, identical to the South Gate excavated by Baramki, was found (Fig. 9). The second area extended the excavations by Dr. Taha in 2006 (see Taha 2011) to the north of the bath. Both excavations confirmed the stratigraphic sequence and indicated a new architectural complex. The 2011 and 2012 trenches discovered a broad, double stairway leading down to a subterranean room from both the north and south. A doorway appears to lead into a large room for fuel storage for the bath. This elaborate, rather formal plan may indicate a more significant structure. Both the North Gate and stairway stand as transitional elements and may indicate the importance of the Northern Area to the site.

Fig. 8.

Plan of Khirbat al-Mafjar with the 1993 plan and the new excavation areas. (Drawing by D. Whitcomb and the Department of Antiquities and Cultural Heritage, Palestinian Authority.)

Fig. 9.

The North Gate discovered in 2011. (Drawing by D. Whitcomb and the Department of Antiquities and Cultural Heritage, Palestinian Authority.)

The Northern Area was excavated by Dajani but was left untouched since that time. The 2012 season of excavations cleaned the area of mounds of Jordanian backdirt, pruned extensive trees, and has successfully brought the Northern Area into archaeological prominence as an integral part of Qasr Hisham. Area 6 revealed a large grape press with white mosaic floors, representing the intensive agricultural activities of an estate in the Umayyad period (Fig. 8). In Area 5, a series of rooms and the courtyard of a large house indicate continued occupation during the early Abbasid period (ninth century) with some spolia from the palace (Fig. 8). The next season will investigate the suggested horse stables behind this house, perhaps an indication of the importance of horse breeding for the estate. Furthermore, there are clear indications of an early palace or residential settlement, perhaps antecedent to the palace complex of Qasr Hisham.¹²

Another day’a or estate

One may consider the Northern Area to have been an agricultural estate during the Umayyad and Abbasid periods, the economic foundation for support of the palace complex. Not only is the size of occupation at Qasr Hisham now twice as large, but it also has two different components: the elite palace, mosque, and bath; and a complex agricultural estate. More evidence remains to be discovered to support this interpretation of Qasr Hisham as both palace and estate; it is hoped that this pattern may become important as a model of early Islamic settlement.

A preliminary hypothesis suggested by this day’a (pl. diya’) is that this might be a sixth element characteristic of the qusur. Virtually all of the early leaders of Islam invested in the establishment of estates, first in the Hijaz and then elsewhere (Millward 1964). The intentionality of these foundations leads to a proto-urban hypothesis, as explicated in Genequand (2010) and associated with the idea of the misr (a system of early Muslim settlements) (Whitcomb 1995, 2001).¹³ The very idea of early Islamic foundations represents an important advance in conceptualizing settlement and economic enterprise after the Muslim conquest.¹⁴ The textual sources clearly indicate that estates realized immense profits; the settlement of residential populations in the context of agricultural and commercial facilities suggests an intentionality toward an urban status. Subsequently, and regrettably most often, the cessation of external capital and unsustainable production led to reduction, abandonment, and historical oblivion for most of these settlements. This would seem the fate of Qasr Hisham, or whatever this settlement was then named.

Khirbat al-Mafjar or Qasr Hisham?

The cultural significance of these monuments and the site in general has a curious double personality for Palestinians and others, whether archaeological specialists or casual visitors. The dichotomy is encapsulated in its two names: Khirbat al-Mafjar seems to designate the archaeology, the physical remains and their serious, but evolving interpretations. Qasr Hisham seems to evoke a cultural pride in the artistic accomplishments of the Umayyad dynasty and, by extension, the formation of Islamic civilization during this earliest period. Both of these “personas” may be seen in presentations and are equally, as with most archaeological manifestations, capable of popular distortions and inaccuracies. This is not, perhaps, the debate to enthuse the casual visitor or other observers, which might explain the uncritical acceptance of Hamilton’s narrative. No one would wish to deny that the Umayyads may have had some good parties, but this is hardly an adequate explanation for a great synthesis of structure and high decoration.

Recognition of the importance of Qasr Hisham as a cultural icon for Palestine has existed and grows with the development of the Palestinian Department of Antiquities and Cultural Heritage. The site of Khirbat al-Mafjar must take second place to the renown of Tell es-Sultan, the biblical Jericho (Fig. 10).

Fig. 10.

Map of the region featuring Khirbat al-Mafjar and Jericho. (Based on Hamilton 1959: Fig. 1.)

This site has received intense attention from a project from the University of Rome for research and conservation (Taha 2005, 2010; see Cirelli and Zagari 2000 for an assessment of the Islamic periods). One might prefer the name Qasr Hisham, which carries a cultural implication and may be viewed in context of the pervasive, and contemporary, Byzantine character of Ariha or Jericho in this period. As mentioned above, the archaeology of Jericho as an important town needs to be considered as well as a landscape of other sites such as Tulul Abu al-‘Alayiq and Khirbat al-Nitla.

The concern for Qasr Hisham as an icon and historical treasure has expanded and demands new efforts for preservation and reconstruction as well as education.¹⁵ The latter is a particular focus where new research should replace the old stories and produce a more significant appreciation of this aspect of the cultural heritage of Jericho for Palestinians and the wider world.