You are here:Home/Resources/ Middle East Garden Traditions/ Introductory Essays/ Early Islamic Gardens in Syria, Jordan, and Iraq

Early Islamic Gardens in Syria, Jordan, and Iraq

Antonio Almagro and D. Fairchild Ruggles

This introduction has been edited for style. Read and download the original essay (PDF).


In Syria, Jordan, Lebanon, and Iraq, between fifty and one hundred residential establishments dedicated to agricultural production and animal husbandry, as well as pleasure, were built in the countryside away from urban centers in the seventh and eighth centuries.Oleg Grabar, The Formation of Islamic Art (New Haven: Yale University Press, 1973), 31. Although situated in remote areas, these were permanent settlements that were strategically located to take advantage of whatever water sources existed as well as to play an intermediary role between the nomadic tribes of the desert and the market centers where agricultural and animal products were traded. The first Islamic gardens emerged in this relationship between architecture, water resources, and the landscape.

Amman citadel Aerial view of the Amman Citadel.

Individual members of the Umayyad family generally built their residences in isolated locations that today seem like harsh, arid environments. But they were places of cultural refinement, graced by small gardens that in almost all cases provided the spatial connection between baths and audience halls, an architectural model in which the pleasure provided by both elements was an integral aspect of the expression of power and dignity. The grouping of elements in this way occurred in residences of a more or less private character and also soon became a feature in urban palaces, as demonstrated in the case of Amman. The Amman Citadel contains a unique palatine complex, built intramuros and of enormous size, which survives from the Umayyad era. Its baths were located next to the grand entry vestibule that led from the mosque and a large open plaza into the administrative zone of the palace. A large cistern stood east of the baths, providing water for the baths as well as the walled area that sloped to the south. A. Almagro,  P. Jiménez, and J. Navarro, El palacio omeya de Amman, vol. 3, Investigación arqueológica y restauración (Granada: Escuela de Estudios Árabes, CSIC, 2000).

Qusayr ‘Amra (711–715) was probably the earliest of these residences, known today for its relatively well preserved bathhouse adorned with murals on its interior. Near the bathhouse, the remaining fabric suggests the existence of a small zone irrigated by a well and a noria, distinct and independent from the well and noria that provided water for the bathhouse. Martín Almagro Basch et al., Qusayr Amra: Residencia y baños omeyas en el desierto de Jordania (Madrid: Ministerio de Asuntos Exteriores, Dirección General de Relaciones Culturales y Junta para la Protección de Monumentos y Bienes Culturales en el Exterior, 1975). Until only a few years ago, one could see the remains of a small wall that enclosed the garden. The residence itself, today in ruins, stood hardly 600 meters from the bath but at a slight elevation that kept it separate from the bath and cultivated area.

Hammam al-Sarah The garden area of Hammam al-Sarah.

In Hammam al-Sarah (between 800 and 825), a building with a structure very similar to Qusayr ‘Amra and scarcely 40 kilometers away, the garden enclosure is perfectly visible. Its existence is confirmed, furthermore, by the presence of a fountain 60 meters from the building, in the center of the cultivated area and on axis with the portal of the bathhouse’s audience hall. The fountain was fed from the same source as the bathhouse and its channel system provided water for irrigating the garden. This bath must have formed part of an extensive agricultural estate whose residential center was at Qasr al-Hallabat (between 800 and 825), two kilometers to the west. The estate included other cultivated areas that were enclosed by walls. This union of an audience hall/bathhouse with garden, independent of the main residence, makes clear the intimate connection between two forms of personal and social pleasure that are always found in Islamic culture.

An even more evolved example on these same lines can be seen in the palace Khirbat al-Mafjar (ca. 739–743), built by the Umayyad caliph Hisham in the Jordan Valley near Jericho. There the bath gains monumental dimensions that rival those of the main residence. The mosque stands between them, and as a mediating element for all three structures, there is a large garden with a free-standing pavilion (or kiosk) of sophisticated design.

Al-Rusafa (Sergiopolis) in Syria had an enclosure that can absolutely be identified as a pleasure garden. Thilo Ulbert, “Ein umaiyadischer Pavillion in Resafa-Rusafat Hišam,” Damaszener Mitteilungen 7 (1993): 213–31 and D. Fairchild Ruggles, “Il giardini con pianta a croce nel Mediterraneo islamico." In Il giardino islamico: Architettura, natura, paesaggio, ed. Attilio Petruccioli (Milan: Electra, 1993), 143–54. Sergiopolis was a Byzantine settlement of handsome, white gypsum walls located 180 miles east of Aleppo in a desert landscape. It was taken over in the Umayyad period and became the site for the caliph Hisham’s palace estate between 724 and 743. The city known as Rusafa flourished thereafter as a minor settlement, with a mixed population of Muslims and Christians, until the Mongol invasions in the mid-thirteenth century, after which it was abandoned. Outside and to the south of the Byzantine stone city, a team of archaeologists excavated an extensive residential area of mudbrick walls traversed by a seasonal stream. They discovered an irregularly shaped enclosure with a central pavilion that was slightly elevated, open on each of its four sides toward the surrounding garden and probably domed. The fact that handsome stucco fragments dating to the Umayyad period were also recovered at the site indicated that this pavilion was part of a palace complex, one of two built on the outskirts of Rusafa. On three of the four sides, paths with steps leading up to the pavilion were discovered, and the archaeologists speculated that there was formerly a fourth path that had since been obliterated. This is the earliest known Islamic example of a pavilion occupying the intersecting point of the axes in a quadripartite space, and as such it is of considerable importance for garden history. The four-part layout was to become a classic and ubiquitous (although by no means exclusive) garden type that would be repeated with variations throughout the Islamic world. When the Umayyad prince who was to become ‘Abd al-Rahman I, fled from Syria to al-Andalus in the mid-eighth century, he built himself a garden estate outside of Córdoba, which he named al-Rusafa in memory of his grandfather’s desert palace where he had spent part of his youth.D. Fairchild Ruggles, Gardens, Landscape, and Vision in the Palaces of Islamic Spain (University Park, Penn.: Pennsylvania State University Press, 2000), 42–45. Historical accounts mention that the plantings of Rusafa’s gardens in Córdoba included many exotic varieties; however, they do not describe actual layout.

In the ninth century, the new city of Samarra with its successive expansions offers examples of palaces with extensive gardens that were well integrated in the design of the entire complex. The Abbasid caliph al-Mu‘tasim (833–842) founded  Samarra in the second quarter of the ninth century to serve as his palatine seat, and immediately set about developing the area, encouraging the planting of trees and the cultivation of a variety of delicious and sweetly scented plants in kitchen gardens. According to Ya‘qubi, he commanded a workforce of “gardeners knowledgeable in sowing, planting, and tending date palms.” Descriptions of the caliph’s palace are detailed, perhaps because of their extravagance. On the occasion of the reception of the Byzantine ambassador in 917, al-Khatib al-Baghdadi reported a garden with orange trees, 400 tall date palm trees, and a metal-rimmed pool. In other gardens, leafy trees and birds made of gold and silver, and fountains spouting perfume, astounded the senses.

Dar al-Khilafa Plan of Dar al-Khilafa.

Al-Mu‘tasim built the Dar al-Khilafa (also known as the Jawsaq al-Khaqani) in 836, which was followed by new palaces built by his sons and followers. These palaces had large interior courtyards that, with basins and water channels, surely contained gardens. They received their water from canals tapping the Tigris River, along the banks of which Samarra’ sprawled for 50 kilometers. The Dar al-Khilafa had a square courtyard with a fountain in its center aligned along the palace’s central east–west axis. Beyond lay an even larger courtyard with water channels and two fountains. Both of these may have been gardened.K. A. C. Creswell, A Short Account of Early Muslim Architecture, rev. ed. (Aldershot: Scholar Press, 1989); Khalid K. Hammudi, “Qasr al-khalifa al-Mu`tasim fi Samarra’,” Sumer 38 (1982): 168–205; and Alastair Northedge, “An Interpretation of the Palace of the Caliph at Samarra,” Ars Orientalis 23 (1993): 143–70. At Balkuwara Palace (850s), the partially enclosed courtyard at the southwestern end had a water basin in its center and a prospect of the wide riverbed and its opposite bank. Although there is no direct evidence of planting in this courtyard, it is hard to imagine it otherwise. At the palace’s northeastern end, there were three interior courtyards that the archaeologist Ernst Herzfeld showed as quadripartite with cross-axial walkways.K. A. C. Creswell, Early Muslim Architecture, 2 vols. 2nd ed. (Oxford: Clarendon Press, 1969), 2:265–70. However, Alastair Northedge, on the basis of both archaeological evidence and Herzfeld’s own notes, has corrected the plan to show the garden courtyards were divided bilaterally.Alastair Northedge, “Creswell, Herzfeld, and Samarra,” Muqarnas 8 (1991): 74–93. It is quite likely that Herzfeld, upon discovering courtyards that may have been gardened, imposed a teleological interpretation of the space as a chahārbāgh (a term that his generation of scholars used to refer to an Islamic garden divided into four roughly equal rectangles). While we have seen that the cross-axial quadripartite plan existed in Iraq at least as early as the mid-eighth century, archaeologists in the first half of the twentieth century did not know this. Thus, when Herzfeld cast the Samarra garden as a cross-axial type, he was extrapolating, not from Rusafa, but from much later Islamic models where the quadripartite form could easily be observed.

At Samarra and elsewhere in the first centuries of Islam, it can be hard to match the archaeological evidence to the almost effusive historical descriptions of gardened palaces and fertile landscapes, and yet both are valuable sources for the formative period of Islamic gardens. One of the reasons that so little is known about the early period of Islamic garden-making is that, whereas the walls and architectural structures survive at least in crumbling condition, the gardens have completely vanished. An archaeologist can identify the presence of a semi-ruined structure and explore it to determine its original form and function, but it takes considerable imagination and often sheer luck to discover the site of an early garden. Usually, it is only the trace of a hydraulic system that attests to the former presence of cultivated plants.