AYVĀN (or ṬĀQ)-E KESRĀ (the Palace of Ḵosrow), the most famous of all Sasanian monuments and a landmark in the history of architecture, now only an imposing brick ruin. While actually belonging to the city of Asbānbar on the east bank of the Tigris below Baghdad, the Ayvān-e Kesrā (also called Ayvān-e Madāʾen) is usually associated with the name of Ctesiphon, which lies immediately to the north of Asbānbar and is another of the cities which made up the multiple complex of Madāʾen (Figure 1). Situated near the modern settlement of Salmān-e Pāk, the Ayvān-e Kesrā is the legendary throne hall of the Sasanian kings of kings.
Pertinent studies of the monuments include the invaluable photographic record that was published in Dieulafoy (V, sec. 6) before the floods of 1888 destroyed one third of the standing ruin (Plate XXXIII); the investigations of Herzfeld in 1907-08 from which he produced a plan based on the traces of walls above ground and the shapes of the surrounding ruins (Sarre and Herzfeld, Archäologische Reise, p. 61, fig. 167); the excavations of the German expedition of 1928-29 which confirmed some of Herzfeld’s projections, as well as unearthing other features in the area (Reuther, Ausgrabungen, pp. 17ff.); and the probes, conducted in 1964 by the Italian Mission in Iraq, which were designed to prepare the ground for restoration (Bruno, “The Preservation,” pp. 89ff.). In 1972 the Directorate General of Antiquities of Iraq completed the restoration of the south wing, and in 1975 work began on the reconstruction of the collapsed north wing (Madhloom, “Restorations,” pp. 119ff.).
The standing monument consists of a large ayvān 43.50 m deep by 25.50 m wide, penetrating a blind facade that stretches 46 m in either direction from the center line of the ayvān and stood originally 35 m above ground level to the height of its cornice (Plate II, Plate III). The articulation of the blind facade is formed by a series of six stories of brickwork, consisting of columns, entablatures, and arched niches. The ayvān is roofed by a parabolic vault, with the side walls tapering from 7 m to 4 m and brought forward by slight corbelling below the impost in order to reduce the enormous width to be spanned. The true part of the vault, for which slanting lays of brick on edge were employed (permitting the construction of a vault without centering), is confined to the upper third of the structure; this arch tapers from 1.80 m at the point of the spring line to 1.30 m at the crown. The great arch dominates the layout, but the area behind the facade can be shown to be taken up by a pair of large rectangular and square chambers on either side (if one can assume symmetry of plan) which are separated from the axial ayvān by a vaulted corridor system (Figure 2). Access to the corridor is gained through a doorway which penetrates the facade. Behind the ayvān, and connected to it by a narrow door, is an arrangement of rooms which are remarkable for their comparatively small dimensions. Through a cross passage and the central chamber one can enter the large hall at the rear of the complex. It is a later addition and measures 26 m wide by 38 m deep. The great arch and its facade have aroused comments that generally have acknowledged its impressive qualities. Ebn al-Faqīh (p. 255) considered it to be one of the marvels of the world. Ṭabarī (III, p. 320) reports that the caliph al-Manṣūr was advised by his minister Ḵāled b. Barmak not to attempt its demolition. This advice reflects both the building’s prodigious dimensions and its reputation as a monument. In spite, however, of the past and present notoriety surrounding the Ayvān-e Kesrā, including the observations of Arab geographers, the accounts of travelers, and the attention of archeologists, it is still only largely circumstantial evidence which permits investigators to propose a specific date for the building’s construction. That such a problem is a crucial issue stems from the fact that any interpretation of the building’s significance in the history of architecture depends entirely upon knowing when it was built.
The problem of identifying the building’s sponsor began at least as early as the time of Ḥamza Eṣfahānī (mid-4th/10th century A.D.) who, as cited by Yāqūt (I, pp. 425-26), quotes a passage translated by Ebn al-Moqaffaʿ from a Pahlavi chronicle, the Ḵᵛaḏāy-nāmāg, which states that the Ayvān at Madāʾen was built by Šāpūr I (r. 241-72 A.D.). This date was preferred by Herzfeld (Archäologische Reise, p. 76). Ḥamza himself favored a different view, based on information received from a Zoroastrian priest, Omayd b. Ašūhast, who claimed that the Ayvān was the work of Ḵosrow II Parvēz (r. 590-628). In other sources it is Ḵosrow I Anōšīravān (r. 531-79) who is claimed as the founder. This attribution has received the most support in recent times for stylistic reasons as well. Yāqūt said that it was the work of several kings (loc. cit.). The problem is compounded by the fact that late medieval writers confuse the Ayvān-e Kesrā with the “White Palace.” The confusion may stem from the fact that, according to Ṭabarī (I, p. 2441), when the victorious Moslem general Saʿd b. Abī Waqqāṣ occupied Madāʾen in 16/637 “he resided in the White Palace and took his prayers in the Ayvān.” By the time of Yāqūt the White Palace had long since been demolished for the re-use of its building materials, and there may have been a tendency to think of the two structures as belonging to same complex, though it is clear from the early writers (e.g., Yaʿqūbī, Boldān, p. 321, and the Taʾrīḵ Baḡdād I, p. 128) that the White Palace was in the ʿAtīqa section of Madāʾen, while the Ayvān was in Asbānbar. The picture is further clouded by the way that the personal name Kesrā was used in post-Sasanian times as the generic title for all Sasanian kings of kings. The problem is only slightly alleviated by the theory advanced by O. Grabar (see AYVĀN) that the latter term was used in the medieval period to refer to a throne palace and had no specific connotations of architectural form. Thus, any reference to an ayvān in Madāʾen should presumably automatically relate to the Ayvān-e Kesrā.
Reuther arrived at a sixth century date, i.e., that of Ḵosrow I, on the grounds of the facade’s style (Survey of Persian Art, pp. 515-16). He argued that three of the decorative principles employed in the facade’s decorations are unknown in the eastern Mediterranean area before the sixth century. These include the combination of large and small elements of the same type in an arcade where stories of markedly different scale are linked by a “colossal order” of engaged columns; the rejection of vertical alignment for the elements of superimposed stories; and the use of a decorative arcade motif on the archivolt of the arch. Naturally, this argument presupposes that all these developments had to have taken place first in Syria. It relied heavily upon the accounts of Byzantine historians such as Theophylaktos Simokatta (Histories 5.6, ed. C. de Boor, Leipzig, 1887, cf. Šāh-nāma, ed. Borūḵīm, IX, pp. 2886ff.), who went to great lengths to emphasize the contribution made by Greek, i.e., Byzantine, architects. “Ctesiphon is the biggest of the royal residences in Persia. It is said that the emperor Justinian sent Greek stone, as well as architects who were experts in building and experienced in vaulting, to Ḵosrow the son of Kavād (Ḵosrow I). And they built the royal palace not far from Ctesiphon in the Byzantine manner.”
While one can not dispute that marble, as well as artisans and even architects, were in all probability brought to the royal court of Ḵosrow I for his foundation of New-Antioch after his capture of the mother city in 540 A.D., it is entirely to be disputed whether these individuals were involved in the conceptual planning of the Ayvān-e Kesrā monument. It would be natural for any Byzantine writer like Theophylaktos, with his obvious bias, to have made the smallest contribution by a western architect (even redecoration with marble slabs) seem like control over the entire project. It would be wrong, too, to deny that the origins of the individual elements of the decorative facade are to be found in the vocabulary of the Hellenistic and Roman architecture of the West. But these principles of decoration had long been adopted and even developed by the Parthian architects of Mesopotamia, as demonstrated by the facade of the Parthian palace at Ashur. In fact, one could argue that the types of liberties which eastern architects were already accustomed to taking with the strict Classical concepts, in the second century A.D., naturally permitted the decorative arrangement of the Ayvān-e Kesrā to have been conceived long before the era of Ḵosrow I.
Herzfeld and others relied heavily upon this argument, being inclined to accept the Ḵᵛaḏāy-nāmāg source which attributed the building to Šāpūr I. Gullini has acknowledged the obvious similarity between the Ayvān-e Kesrā and the Ashur palace in terms of the use of a blind facade to decorate a building, but he argues that the Sasanian example lacks “the uniform rhythmic vertical cadence” which distinguishes the Parthian version. Following basically the same stylistic argument as Reuther he is inclined to favor a sixth century date (Bruno et al., Mesopotamia 1, p. 92).
Reuther’s interpretation also depended upon evidence unearthed by the 1928 excavations. But it should be emphasized that this evidence is circumstantial in the extreme. Thwarted by the lack of material by which they could study the monument in detail, and with considerations of safety in mind, the German excavators turned their attention to the Tell al-Ḏabāʿī mounds that lay immediately to the south of the south wing of the Ayvān-e Kesrā. Extensive fragments of architectural decorations were found on the surface, mostly of colored marble, mosaic glass cubes, and ornamental stucco. The stucco was dated to the sixth century on stylistic grounds. By an unacceptable line of argument the hypothesized friezes of human and animal figures were interpreted as being appropriate remnants of the White Palace, with the Ayvān-e Kesrā (in which, according to Ṭabarī, I, p. 2441, Saʿd took his prayers) located conveniently close by. The similarities of pavement level between the two areas were seen by the excavators as additional supportive evidence for dating the Ayvān-e Kesrā to the sixth century.
A possible solution to the dilemma might be found if it could be determined that the great arch actually dated from the third century, while the decorations (in particular the marble slabs from Antioch) merely represented later refurbishings. The later addition of the enclosed chamber at the rear of the complex suggested by the different level of foundations, might even fit nearly into this arrangement—the greater mystique associated with the imperial throne and the closed draperies of the reign of Ḵosrow I would have been well served by the addition of this more secluded chamber behind the public audience hall.
Recent probes, however, by the Italian Mission around the footings of Ayvān-e Kesrā—designed specifically as studies to plan the restoration of the monument—have demonstrated fairly strongly that there is no archeological evidence for such suggested refurbishings. The Mission reports that the only major changes visible from the probes appear to have occurred after the tearing-out of the precious wall facings (presumably after the Moslem conquest). It would be surprising if all traces of earlier phases had been completely removed, even if the hypothetical sixth-century refacing had involved the complete removal of the earlier pavements. Virtually no finds were recovered from the fill associated with the footings which could in any way provide a satisfactory date, although the universally problematic greenish-blue glazed pottery was found in the upper layers and assigned “a definite early Islamic” date (Mesopotamia I, p. 103). The recent restoration projects on the Ayvān-e Kesrā itself seem to imply that no further refinement of this date will be possible using conventional stratigraphic excavation techniques, unless some of the small side and rear chambers can be safely excavated, with the chance that more diagnostic dating material can be found in them. Excavations in the general area will advance considerably our knowledge of the history of Madāʾen, but barring a laboratory breakthrough such as in the realm of thermoluminescence for the dating of its bricks, the date and therefore significance of the great arch is likely to remain somewhat elusive.
Ayvān-e Kesrā has been mentioned and often described in Arabic and Persian sources and it is the subject of a moving qaṣīda by the poet Ḵāqānī who visited its ruins in mid-6th/12th century.
A. Bruno, with contributions by G. Gullini and M. Cavallero, “The Preservation and Restoration of Taq-i Kisra,” Mesopotamia 1, 1966, pp. 89-108, pls. xvii-xxv, figs. 35-59.
Camb. Hist. Iran III, pp. 1062-64.
Christensen, Iran Sass., pp. 386, 389ff., 504ff.
M. Dieulafoy, L’art antique de la Perse, Paris, 1884, V, pp. 71-74.
J. M. Fiey, “Topography of al-Mada’in,” Sumer 23, 1967, pp. 3-38.
E. Kühnel, Die Ausgrabungen der zweiten Ktesiphon-Expedition, with English summary by S. Dimand, Berlin, 1933, pp. 1-35.
O. Kurz, “The Date of the Taq-i Kisra,” JRAS, 1941, pp. 37-41.
Lacoste, “L’arc de Ctésiphon ou Taq Kesra (Mesopotamie),” Sumer 10, 1954, pp. 3-22.
T. Madhloom, “Mada’in (Ctesiphon), 1970-71,” Sumer 27, 1971, Arabic section, pp. 129-46.
Idem, “Al-Mada’in,” Sumer 31, 1975, Arabic section, pp. 165-70.
Idem, “Restorations in al-Mada’in 1975-77,” Sumer 34, 1978, Arabic section, pp. 119-29.
O. Reuther, Die Ausgrabungen der deutschen Ktesiphon-Expedition im Winter 1918-29, Berlin, 1930.
Idem, “The German Excavations at Ctesiphon,” Antiquity 3, 1929, pp. 424-51.
Idem, “Sasanian Architecture,” in Survey of Persian Art, pp. 493-578.
A. Saleh, “Al-Mada’in and its Surrounding Area in Arabic Literary Sources,” Mesopotamia 3-4, 1968-69, pp. 417-39.
F. Sarre and E. Herzfeld, Archäologische Reise im Euphrat- und Tigris-Gebiet II, Berlin, 1920.
J. H. Schmidt, “L’expéditon de Ctesiphon en 1931-1932,” Syria 15, 1934, pp. 1-23.
(E. J. Keall)
Originally Published: December 15, 1987
Last Updated: August 18, 2011
This article is available in print.
Vol. III, Fasc. 2, pp. 155-159