Summary report

In Madinat al-Zahraʼ (Córdoba, Spain), the foundation, expansion, and destruction of a 10th-century AD caliph's residence can be studied in exemplary fashion. The focus of a 6-year project is the question of the material form of the caliph's role in society and the contact zones between ruler and urban population. In cooperation with the Conjunto de Madinat al-Zahra, the forecourt of the caliph's palace, the so-called Plaza de Armas, has been investigated since 2016. The goal of the fifth and final field campaign was to investigate the south side of the plaza. Here, the plaza borders on a residential and commercial area of the city, with a height difference of around 16 m between the plaza and the urban area. In a 13 m long and up to 4 m deep trench the southeast corner of the square was discovered. Remains of a massive terrace wall were found, which may have been more than 6 m thick. It was also possible to investigate part of the building complex adjoining the plaza to the east. In this area there was a paved courtyard surrounded by rows of rooms, including a latrine, as well as the section of a paved passage.

Report of work at Madīnat al-Zahrā’

The aim of the fifth and final excavation season at the Plaza de Armas of Madinat al-Zahraʼ was the investigation of the southeast corner of the plaza (Fig. 1). The area where the corner one stood is in an area that proved difficult for excavation, with a pronounced slope toward the south. During a season of four weeks in October 2022 a 3 m wide and 13 m long trench was excavated, down to a depth of up to 4 m (Fig. 2). Remains of the corner itself was indeed found, at a level of 187.4 m, 6 m below the level of the plaza. Because of the great depth of the trench and the reduced space, information on the adjacent areas is limited, however. The results regard for the most part the building located to the east of the plaza, less so the plaza itself. Basic question on the morphology of the southern limit of the plaza remain unanswered. Additional geophysical investigations planned for the fall of 2023 may help to address some of these questions.


Fig. 1: The Plaza de Armas of Madinat al-Zahra’. Areas investigated in 2017 to 2022 are indicated in red. Drawing: Felix Arnold, DAI/Conjunto de Madinat al-Zahra.


Fig. 2: Area investigated in 2022 at the southeast corner of the plaza. Foto: María Latova/DAI.

The enclosure wall of the Eastern Building Complex

Within the excavation trench studied this season, the remains of the southwest corner of the Eastern Building was found, as well as a section of the western wall of the complex (Fig. 3). The wall is extremely thick, up to 1.7 m. As in sections studied previously, the wall was built of regular ashlar blocks of limestone, about 100 cm long, 25-30 cm thick and 40 cm high. Instead of lime mortar, reddish-brown soil was placed in the horizontal joints of the masonry. The western face of the wall is stepped, reducing the width of the wall by as much as 17 cm in each course. At the level of the plaza, the thickness of the wall may well have been only 1 m, as is the case further north.


Fig. 3: Remains of the southeast corner of the plaza, looking north. Foto: María Latova/DAI.

Along the western wall, a pronounced joint is visible in the masonry, indicating a break in the fabric. The steps north and south of the joint do not align, and diverge in orientation (Fig. 4). The northern section was built first, with very regular steps, and with carefully finished outer faces. The steps of the southern section was built with less care, with steps fanning out toward the south. The joint between the two sections was constructed in an interlocking manner, with stretchers jumping back and forth, to create a bond to the added section. The two sections therefore appear to have been the result of a single plan, but executed at two moments of the construction process, by two different crews of masons.


Fig. 4: Stepped foundation layers of the enclosure wall of the Eastern Building. Foto: María Latova/DAI.

At the northern end of the trench a header reaches beyond the outer face of the wall. The stretcher adjoining it to the north was placed at an irregular angle, indicating that it was part of a core masonry, not the outer face of the wall. At this position, a buttress must therefore have existed, to reinforce the enclosure wall. The size of the buttress is not clear, but must have been more than 1.7 m in width.

Some 7 m south of the buttress, a corner is preserved on the inner side of the wall, indicating the location of the southern wall of the enclosure. On the outer, western face of the wall, the steps also end at this point, indicating the existence of a buttress at the corner of the enclosure wall. The southern face of this buttress was encountered 3.4 m further south. Since the outer faces are also likely to have been stepped, the size of the buttress probably was reduced further up, to about 2.3 m.

Of the southern wall of the enclosure only some remains were found in the eastern profile of the trench. The width of the wall is likely to have been less than that of the buttress. The layout of the pavement to be discussed below suggests that the inner face of the wall was also stepped, reducing the width of the wall further, possibly to only 1 m like the western wall.

The remains of the corner excavated this season provides information on the size of the Eastern Building Complex (Fig. 5). The width of the enclosure from north to south may now be reconstructed to have been 116.4 m on the inside, and some 118.4 m on the outside, 121.5 m including the buttresses (250 cubits à 47.3 m = 118.25 m). The gate found in 2019 was placed almost exactly in the center of the enclosure wall, only some 67 cm north of the exact center. The geomagnetic survey of 2017 indicates that the building was about 155 m long.


Fig. 5: Original layout of the Eastern Building, with excavated areas marked in red. Drawing: Felix Arnold/DAI.

Results from previous seasons suggest that the space inside the enclosure wall had a pronounced slope in the beginning, falling from about 197.5 m in the north to 182.5 m above sea level in the south. The original ground level was not encountered in the trench of this season, but must lie below 183.5 m. Already during the construction process, soil was placed inside the corner, raising the level up to a level of 186.1 m. The fill is comprised of volcanic rock, alternating with layers of limestone chip at the level of every course of the adjoining stone masonry. Outside the enclosure wall, the ground sloped further down toward the south, reaching a level of 177 m.

Inside the enclosure, the isolated building investigated in 2018 may have been the only structure to be erected in the first building phase. It was located on a high platform, built of walls constructed without the use of lime mortar. Its function still awaits further study.

Structures inside the Eastern Building Complex

In a second building phase, additional structures were built inside the enclosure. These are all built using lime mortar, as is the case also of the structures found in 2021 within the northwest corner of the enclosure. Within the area investigated this season, a wall oriented east-west was found to abut the inner side of the western enclosure wall (Fig. 6). The wall is about 1 m at the top, but was supplied with a thicker foundation, of more than 1.6 m, with a sequence of steps on either side.


Fig. 6: Profile along the eastern side of the trench excavated in 2022, showing walls and associated layers. Drawing: Felix Arnold/DAI.

The ground level to the south of this interior wall was raised at this time, up to a level of 187.75 m. A section of a pavement was still found, directly on the present-day surface (Fig. 7). Rows of limestone ashlar blocks, three courses deep, served to consolidate the ground. Parts of three such rows, running east to west, are preserved. A fourth row is likely to have been located further south. If it was placed at an equal distance, the paved space measured about 3.25 m in width. The space in between the rows of ashlar blocks was filled with volcanic rocks originating from the subsoil of the area, placed on a fill of earth.


Fig. 7: Pavement inside the Eastern Building, made of volcanic rocks and rows of limestone blocks. Foto: María Latova/DAI.

The area adjoining the wall to the north was also leveled at this time. Remains of a similar pavement, at a level of 187.6 m, may be observed further east and north, indicating that the entire paved area was about 57.6 m wide from north to south and 72 m long from east to west. To the north it was limited by a high terrace wall, on which the building studied in 2018 now stood. To the east, it reached up to a wall or row of rooms, of which some remains are still visible on the ground. This wall apparently subdivided the Eastern Building Complex into two more or less equal halves. The eastern half was also paved at this time. The function of these two large open areas is not clear. The kind of pavement composed of volcanic rocks placed between rows of ashlar blocks is usually associated with horses, the area thus possibly having been used as forecourts to the plaza, or as associated stables.

Along the side of the trench excavated this season the remains of further structures were found that had been built along the western side of the courtyard. The preserved remains may be interpreted as parts of a latrine. Still in situ is the drainage system, composed of a slab sloping toward the west, a shaft, and a 29 cm wide canal running below and sloping toward the east (Fig. 8). The continuation of this canal could not be investigated this season. The top of the latrine may have been formed of violet limestone slabs, of which several were still found in the surrounding area. The level of the latrine may have been about 188.15 m, the level of the floor adjoining it to the east about 187.8 m.


Fig. 8: Section of the latrine inside the Eastern Building. Drawing: Felix Arnold/DAI.

The latrine is likely to have formed the southern end of a 5 m wide row of rooms built against the western enclosure wall. The gate studied in 2019 would have formed part of this row of rooms, though on a much higher level, at 193 m. Along the eastern, outer side of these rooms a ramp led up to the higher terrace, of which the upper end was found in 2018.


The information gained this season on the plaza to the west is not conclusive. According to the work done in 2019 and 2021, the level of the plaza may be expected at about 193 m, considerably above the remains encountered this year. To create a plaza at this level, a massive retaining wall would have been needed to the south, where the ground slopes down to a level of about 177 m. Indications of for this wall – or rather a sequence of walls – is still visible on the ground today, and some remains have been cleared at the western end, which were documented in 2017. How these walls joint the southwest corner of the Eastern Building is not clear. To the west of the enclosure wall massive levels of debris were encountered, up to 4 m deep. A detailed study of these layers suggest that they are neither part of the original fill of the terrace nor a destruction layer, but a fill of a trench dug by stone robbers. Apparently, a wall was dismantled that had run parallel to the enclosure wall of the Eastern Building and was attached to its outer face. Such a secondary wall had been found both in 2019 and in 2021 further north. No remains were encountered this season, the wall having been robbed to a much lower level. With it all indications of adjoining walls, including the remains of the eastern portico, are gone, at least in the area studied this season.

The only indication of a massive retaining wall is a section of masonry found to the south of the corner of the Eastern Building. The masonry was clearly attached in a later phase, using lime mortar. To southern limit of this masonry could not be reached this season. The masonry may be the eastern end of the latest phase of reinforcement of the terrace wall of the plaza, or a reinforcement of the buttress of the Eastern Building.

In either case, the masonry indicates the existence of massive walls delimiting the plaza to the south. The wall is likely to have been more than 6 m high and 16 m high. On top of the wall probably stood the outer wall of the plaza, quite possibly another portico, creating a southern façade more than 25 m high and170 m long. This wall would have separated the plaza from the urban area adjoining the plaza to the south, including the mosque and its surrounding buildings. Weather an access existed at this point from the city up to the level of the plaza remains unknown, but considering the significant difference in level of 16 m is rather unlikely, at least in the later phases of construction. In any case, the wall would have been a very impressive, truly monumental structure, creating a clear separation between town and palatial area, and an unambiguous statement of the power of the caliphate. At least at this moment of its history, the plaza would clearly have formed part of the palace, with highly restricted access from the outside.

Architectural fragments

Within the debris the top end of a column shaft was found. The shaft was made of granite, the first time a shaft of this material has been found at Madinat al-Zahra. An indentation in the center of the top side indicates the use of a dowel to connect the capital to an impost or architrave. Such dowels are not found in the architecture at Madinat al-Zahra, but were commonly used in the Roman and Late Antique period. The column shaft therefore appears to be of an earlier date, and to have been reused at Madinat al-Zahra as a spolia. Roman spolia have been found at the site, but so far only Roman sarcophagi (reused as water basins) and other sculptures. The reuse of columns is mentioned in historic texts, however. Both Maslama and Ibn Baškuwāl report a shipment of 10 columns of green and pink marble from Carthage, Tunis and Sfax in North Africa, 40 columns and two fountains from Constantinople, and another 19 columns from the Frankish empire.

Among the debris excavated this season were also two fragments of a window grill (Fig. 9). The two fragments display a rich vegetal ornamentation on one side, which was pierced and the perforations painted in red. A window grill of this type is known only once more in Madinat al-Zahra, from the House of the Water Basin. In this case, the window was placed in a semicircular frame and probably had been placed above the main entrance to one of the halls of the house. The origin of the newly found fragments is not known.


Fig. 9: Two fragments of a decorated window grill. Fotos: María Latova/DAI.



Felix Arnold, La Plaza de Armas, in: Vaquerizo, Disiderio, Rosón Lorente, Francisco [Hrsg.] Arqueología de Madinat Qurtuba: reflexiones, novedades, historias (Cordoba 2022), pp. 269–274

Felix Arnold, „Madīnat az‐Zahrāʾ, Spanien. Die Arbeiten des Jahres 2022”, e‐Forschungsberichte 2023.1, in print


Feix Arnold and Dirce Marzoli, “Die Madrider Abteilung des Deutschen Archäologischen Instituts.” Lecture in the residence of the German ambassador to Spain (Madrid), 17.05.2022

Felix Arnold, “Córdoba”. Lecture in the series “DAInsight. 50 Jahre Welterbekonvention” (Berlin), 22.09.2022, []

Felix Arnold, “Erecting Walls, Segregating Society in 10th Century Córdoba” (Hamburg), 16.11.2022

Guided tours:

During the campaign of 2022, guided tours were organized for the general public every Saturday and Sunday.

Media reports:

Benjamin Leonard, “A Caliph’s Shining City,” in: Archaeology, March/April 2023, pp. 47-53.

Final publication

Preliminary reports are published regularly online in German in eBerichte des Deutschen Archäologischen Instituts (last issue in 2023) and in Spanish in the Anuario Arqueológico de Andalucía.

The final results of the project will be published at the end of the project (2024) within the series of monographs of the German Archaeological Institute (Madrider Beiträge). A Spanish version will be published by the Junta de Andalucía.