Madīnat al-Zahrāʾ was the capital of the 10th century Umayyad caliphate in al-Andalus and is today one of the most important sites of Islamic culture in the west. In 2018 Madīnat al-Zahrāʾ was inscribed in the UNESCO World Heritage list. A five-year project is currently dedicated to the investigation of the Plaza de Armas, the main public square of the caliphal city. The aim of the field season 2018/2019 was the excavation of a hitherto unknown building on the eastern side of the plaza. The building was a pavilion-like structure erected on a 3.5 m high platform. The ground plan was dominated by a T-shaped hall, reminiscent of similar halls found at Samarra and Fustat. The only examples of this type known so far to the west of Egypt have been found in Fatimid palaces at Aǧdābīyā (Libya) and Ṣabra al-Manṣūriya (Tunisia). A geomagnetic survey revealed that the newly discovered building formed part of a larger compound, which may at first have served an administrative purpose. In a later phase the building appears to have been remodeled to be used as a stable for horses, possibly by visitors that dismounted in the adjoining Plaza de Armas. The building was abandoned in the 11th century and subsequently deteriorated gradually. A coin of John II of Castile (r. 1406–1454) found in a trench offers the first concrete proof from Madīnat al-Zahrāʾ that the walls of the site were dismantled for the construction of the nearby monastery of San Jerónimo de Valparíso. The excavation of this season has revealed not only a building of unique typology, but has added greatly to our understanding of the function of the Plaza de Armas and its overall development.


Report of the 2018-19 season of work at Madīnat al-Zahrā’

Madīnat al-Zahrā’ was founded in 325 AH/936 AD by ʿAbd al-Raḥmān III as the capital of the western Umayyad caliphate. The city with its palaces, gardens, mosques and workshops quickly became the cultural center of the West, rivaling cities like Bagdad and Cairo in the East. Because of its importance and state of preservation, the archaeological site was inscribed this year – in July 2018 – in the UNESCO World Heritage list. The site has since received increased media attention, and a significant rise in tourism.

At Madīnat al-Zahrā’ archaeological work has been conducted for more than a century. R. Velázquez Bosco (from 1911 to 1923), F. Hernández Jiménez (from 1923 to 1975) and others brought to light the central area of the caliphal palace, including the famous Salón Rico. Comprehensive restauration work was carried out thereafter by R. Manzano Martos from 1975 to 1985. A. Vallejo Triano, director of the site from 1985 to 2013, lead a thorough reinvestigation and restauration of several building complexes, oversaw the construction of a well-equipped site museum and published a monumental book on the site.

Unlike the central palace area, little is known so far of the city and the subsidiary palaces of Madīnat al-Zahrā’. The aim of the current project is the study of the so called Plaza de Armas, a large public square to the east of the zone excavated to date (Fig. 1). The plaza served as the main intersection between the palace and the city. To the west of the square lies the palace, to the south the congregational mosque of the city, to the east a so-far unknown building. The investigation of the plaza and its surrounding buildings offers new insights on how the caliph interacted with society and how this interaction developed over time.

mini Arnold Fig 1 199

Fig. 1: Madīnat al-Zahrā’ (LIDAR SCAN of the Instituto Geográfico Nacional 2014). The location of the Plaza de Armas is indicated in red.

Work at Madīnat al-Zahrā’ is currently conducted within the framework of an agreement signed between the Junta de Andalucía and the German Archaeological Institute in 2015. In May 2017 a five-year project (“Proyecto General de Investigación”) directed by F. Arnold and A. Montejo Córdoba (director of Madīnat al-Zahrā’), was approved by the Junta de Andalucía, with the aim of investigating the Plaza de Armas. The aim of the first season of field work (in June and July 2017) was the investigation of the portico on the western side of the plaza. The second season – which is the subject of this report – was carried out in June and July 2018, again in collaboration with A. Canto García of the Universidad Autónoma de Madrid. Participants were F. Bastian, A. Braveman, K. Czarnitzki, J. Garzón, M. Hofmann, W. Jablonska, H. Lehmann, I. O. Roibu, R. Stolle, A. Ugolini, Y. Yosuoka and A. Zamorano Arenas. In addition, geophysical work was conducted in October 2018 by T. Herbich and R. Ryndziewicz.

Over the course of the past year the project has increasingly become the platform for a number of other undertakings. The aim of these projects is to place the investigation of the site on an international footing and thus intensify the study of Madīnat al-Zahrā’ in all its aspects. Regular meetings of the project members foster the exchange of information and ideas. The first meeting took place in March 2018. A second meeting is planned for July 11th-12th, 2019. Currently studies of glass objects (N. Schibille and A. Zamorano, funded by the ERC), metal objects (A. Ugolini) and architectural elements (Y. Yasuoka, funded by the Japanese Academy of Science) are under way. A project aimed at a reinvestigation of the so called Salón Basilical has been approved by the German Research Foundation (Technical University of Berlin). Independently, C. Duckworth from the University of Newcastle has been conducting a geophysical survey of the site, initially aimed at identifying workshop areas in the city. Taken together, these projects greatly increase our understanding of Madīnat al-Zahrā’ and its development.


Excavation work conducted in 2019

The aim of the second campaign of field work at Madīnat al-Zahrā’ was the investigation of a building on the eastern side of the Plaza de Armas (Fig. 2). The area has never been the object of archaeological work before. The topographic survey conducted by F. Hernández in 1923 revealed the eastern limit of the plaza and indicated the location of a singular building to the east. Since then, scholars including R. Castejón and A. Vallejo Triano interpreted this building as a multi-naved hall similar in design to the Salón Basilical and the Salón Rico. A clear understanding of the nature and meaning of this building is essential for the interpretation of the Plaza de Armas and its function.

mini Arnold Fig 2 199

Fig. 2: The Plaza de Armas and adjoining buildings (drawing F. Arnold).

A geomagnetic survey conducted in 2017 had revealed a monumental portico along the eastern limit of the Plaza de Armas, similar to the portico on its western side. Beyond this portico the ground is divided into two terraces, an upper terrace to the north located more or less on the same level as the Plaza de Armas, and a lower terrace to the south. On the upper terrace a distinctive, tell-like hill is visible in the topography, which constitutes the remains of the supposed multi-naved building (Building 63).

In June 2018 the surface of this hill was cleared of vegetation, revealing the remains of several walls built of ashlar masonry (Fig. 3). Over the course of three weeks, five sections were excavated with the aim of gaining information on the structure and function of Building 63 (Fig. 4). In section 1, at the southern foot of the hill, the remains of a portico were found buried beneath a thick layer of debris (Fig. 5). The portico appears to have been constructed against a platform, which formed the foundation for a pavilion-like building (Fig. 6). In order to clarify the ground plan of this building, sections 3 and 4 were dug on top of the platform. In addition, section 2 was excavated to the south of section 1, revealing the southern of the terrace on which Building 63 stood and a ramp connecting this terrace to the adjoining lower terrace. In sections 1–5, five phases of construction, use and destruction could be differentiated.

mini Arnold Fig 3 199

Fig. 3: View from Building 63 across the Plaza de Armas, with the western portico in the background (photo F. Arnold).

mini Arnold Fig 4 199

 Fig. 4: Ground plan of Building 63 with the location of Sections 1-5 (drawing F. Arnold).

mini Arnold Fig 5 199

Fig. 5: Remains of the portico of Building 63 covered by debris (photo M. Pijuán).

mini Arnold Fig 6 199

Fig. 6: Southeast corner of the platform of Building 63 (photo M. Pijuán).

Phase 1

In the first phase a 25.7 m wide, 19.7 m deep and 3.5 m high platform was erected. On top of the platform stood a pavilion-like building, which occupied the entire area of the platform. The remains excavated so far suggest that a 23.7 m wide and 3.1 m deep hall (E3) was located along the southern outer wall of the building, stretching along its entire façade (Fig. 4). Openings in the façade would have provided an impressive view across the city of Madīnat al-Zahrā’ and the surrounding landscape. In the central axis of the building a second, 4.8 m wide and 13.2 m deep, iwan-like hall (E4) opened onto the front hall (E3), thus forming a T-shaped ground plan. The two halls were connected by a 3.4 m wide opening, which appears to have been spanned by a wide arch. A step constructed of marble slabs led from the front hall (E3) to the 10 cm higher back hall (E4). The organization of the side rooms located in the northwestern and northwestern corners of the pavilion (E5 and E6) still needs to be clarified by further excavation. Also unclear is how the pavilion was accessed from the outside. A ramp or staircase may have been located in the back.

T-shaped ground plans of this kind are well known from the Near East, particularly from Abbasid Sāmārra and al-Fusṭāṭ. The only examples known so far to the west of Egypt are found in the Fatimid palaces at Aǧdābīyā (Libya) and Ṣabra al-Manṣūriya (Tunisia). At Madīnat al-Zahrā’ the influence of T-shaped ground plans has been noted in the Salón Rico and the house of Ǧaʿfar. The new building appears to be a much more explicit copy of the Near Eastern prototypes, however.

The results of the geomagnetic survey suggest that the pavilion stood inside a large, tripartite compound, which at first lay outside the confines of the palace of the caliph (see below, Fig. 9). The orientation of the compound diverges slightly from that of the palace area. The compound may at first have served an administrative purpose, the pavilion functioning as the reception hall of a high official. The compound appears to predate the construction of the city wall and thus to the 940’s, soon after the foundation of Madīnat al-Zahrā’.


Phase 2

In a second phase, a portico was added to the southern side of the platform (Fig. 4, E2). The 25.6 m long and 5 m wide portico appears to have comprised seven pillars. It is the first portico found at Madīnat al-Zahrā’ which is open to three sides. So far an L-shaped pillar at the southeastern corner has been excavated, as well as two square pillars in the center (Fig. 5). The pillars were furnished with a polished lime plaster, and a thicker dado at the base. The portico may have supported a kind of balcony, which would have allowed visitors to exit the hall of the pavilion and enjoy the view across the landscape (Fig, 7a).

mini Arnold Fig 7 199

Fig. 7: Reconstruction of Building 63 in Phase 2 (top) and Phase 3 (bottom) (drawings F. Arnold).

The area surrounding the pavilion (E1) was now paved with limestone slabs. To the south, the ground was reinforced by a terrace wall. Along the inner side of this wall the remains of an underground canal were found which may have served to drain water from the neighboring Plaza de Armas. The building activity appears to be contemporary with the construction of the city wall and the two porticos of the Plaza de Armas, all dated to the 950’s CE.


Phase 3

Building 63 was later put to a different use. Both in the portico (E2) and the upper hall (E4) benches with were constructed along the walls (Fig. 4 and 6). Basins on top of the benches and holes along the edges suggest that animals were now kept here, the basins serving as troughs for feeding the animals. At the same time, the floor of the portico (E2) and the surrounding area (E1) was raised and paved will cobble stones. Elsewhere at the site such pavements are common in areas frequented by horses, for example in the streets and ramps leading up to the palace area. In addition a wall was added to the southern edge of the terrace to close the area off. A ramp now connected the terrace to the adjoining lower terrace (Fig. 7b). All these elements suggest that the area was now used by animals, presumably horses.

The location of Building 63 on a raised platform and the high terraces did not make the area very suitable for keeping animals. The change in use indicates on the one hand that the original function of the building – probably in the context of the state administration – was no longer needed or had been shifted to another area. On the other hand the change is indicative for the heavy use of horses in the neighboring Plaza de Armas. According to contemporary sources visitors usually arrived at the palace on horseback and dismounted near its entrance. Building 63 may have served as a short-term stable for these horses. In addition, the plaza may have been used as a training ground for military horses. Such a function is attested for the area outside the palace gate in the city of Córdoba.

The date for the change in function is not yet clear. The role of horses in the military and in palace protocol appears to have increased in the time of al-Ḥakam II (r. 961–976). The changes might have been effected later, however, either while al-Hišam lived at Madīnat al-Zahrā’, or when the site was used as the head quarter of the army of Sulaymān al-Mustaʿīn in 1010 CE.


Phase 4

Unlike in the portico on the western side of the Plaza de Armas, no indication of destruction by fire was observed in Building 63. Instead, a fine layer of sediments suggests that the building stood empty for some time and deteriorated gradually. First the roof collapsed, creating a layer of roof tiles (Fig. 8). In the T-shaped hall only few roof tiles have been found. Instead a large quantity of fragments from a brick pavement was discovered which may originate either from a second story or from a roof terrace. Subsequently the walls deteriorated and collapsed, creating an up to 4 m thick layer of debris.

mini Arnold Fig 8 199

Fig. 8: Profile through the destruction debris covering the walls of Building 63 (drawing F. Arnold and Y. Yasuoka).


Phase 5

At a later time the stones of the building were reused for construction projects elsewhere. In many cases trenches were observed during the excavation of the debris which originated from the demolishing of the walls, sometimes down to their foundations. The trenches were later filled with destruction debris, usually composed of broken blocks.

In one of the trenches a coin with the name of John II of Castile was found, providing the first concrete evidence for the dating of the demolition of the walls. During the reign of John II (r. 1406–1454) the monastery of San Jerónimo de Valparíso was under construction. The monastery is located only 500 m to the northeast of the excavation area and in large parts built of stones originating from buildings of the 10th century. It is therefore more than likely that Building 63 was demolished for this purpose.

The ruined remains of Building 63 were subsequently covered by a layer of earth originating from the decomposition of the growing vegetation. On the surface a coin dating to the reign of Philip IV of Spain was found, minted in Segovia in 1622.


Geomagnetic survey

In October 2018 T. Herbich of the Polish Academy of Sciences was able to extend the geomagnetic survey begun in 2017 to encompass the entire Plaza de Armas and the area to the east. Because of the irregular topography, heavy vegetation and disturbances by modern constructions like metal fences etc. the conditions of the area are not favorable to such a survey. The results are nevertheless of great significance for an understanding of the urbanistic structure of the area (Fig. 9).

mini Arnold Fig 9 199

Fig. 9: Results of the geomagnetic survey of 2017 and 2018 in the area of the Plaza de Armas (T. Herbich).

On the geomagnetic image the continuation of the city wall can be traced, as well as the highway running alongside its outer face. No indications for towers are visible, suggesting that the city wall was constructed in this section without such towers. Some walls can be identified outside the city wall, suggesting that buildings existed also outside the limits of the city wall. The most interesting result inside the city walls regards the orientation of the buildings, which differs from the orientation deduced from topographic surveys. The geomagnetic image suggests that the Building 63 excavated this season stood inside a larger compound, which was divided into three sections. Each section comprised an upper terrace occupied by buildings and a lower terrace, possibly an open courtyard (or garden). Building 63 occupies the upper terrace of the westernmost section of the compound.


Documentation and study of archaeological materials

During the excavation some 564 fragments of pottery were recovered, the majority in the debris of Phase 4 (Fig. 10). For the most part the pottery can be dated to the second half of the 10th century and include 174 fragments of glazed pottery (15 pieces of “verde y manganeso”). A few sherds may derive from the 15th century, from the time when the building was being dismantled.

mini Arnold Fig 10 199

Fig. 10: Study of roof tiles in the courtyard of the museum of Madīnat al-Zahrā’ (photo A. Ugolini).

Interesting are the finds of metal objects, which were studied by A. Ugolini. Most frequent are iron nails, some up to 25 cm long, originating from the roof construction and the doors of the building. Several applications of doors were identified, including a half-moon shaped example. An iron trowel was found in the debris, which may originate from the construction of Building 63 in the 10th century rather than its demolition. Among the debris of the 15th century two horseshoes were recovered.

In the debris (Phase 4) some animal bones were found, which C. Liesau (Universidad Autónoma de Madrid) identified as deriving from cattle, sheep and/or goats. Interesting is the shell of a scallop (“shell of St. James”), which was found in a trough of Phase 4. It may have served as food or as a tool.

Botanical remains gained from earth samples by flotation were studied by M. O. Rodríguez (University of Jaén). Most frequent are the remains of black pine and Aleppo pine (Pinus nigra and Pinus halepensis), which were both used as construction material, especially for the roof of buildings. Remains of olive, plum and almond trees might point to cultivation.  Remains of ash, evergreen oak, mastic, phyillyrea, and willow trees probably derive from vegetation subsequent to the abandonment of the site.

Dr. Felix Arnold
German Archaeological Institute, Madrid, Spain