AYVĀN (palace, veranda, balcony, portico), a Persian word used also in Arabic (īwān, līwān) and Turkish. In classical Persian or Arabic texts, ayvān refers most of the time to a palatial function, either a whole palace or the most important and formal part of a palace. By extension, it can mean the most official or impressive part of any building. It has been suggested that the word derives from Old Persian apadāna (W. B. Henning, “Bráhman,” TPS, 1944, p. 109 n. 1 = Acta Iranica 6, p. 195; W. Eilers, in Camb. Hist. Iran III, p. 495), but this derivation is no longer securely established. The most celebrated literary use of the term for a standing secular monument occurs with respect to the remains of the Sasanian palace, Ayvān-e Kesrā, in Ctesiphon, where it is synonymous with ṭāq, the latter term referring to a form rather than to a function. The other examples of the use of the term in texts can rarely be associated with a specific form. In descriptions of ʿAbbasid palaces the ayvān was the main reception and audience hall of a larger establishment with only hypothetical formal equivalents. However, a four-storied ayvān, presumably a discrete building, was erected by the Muzaffarid Shah Yaḥyā (r. 789-95/1387-93) in Yazd (Aḥmad b. Ḥosayn Kāteb, Tārīḵ-ejadīd-eYazd, ed. Ī. Afšār, Tehran, 1345 Š./1966, p. 86). In the Šāh-nāma, the word is consistently and almost exclusively used for palaces or for audience halls. At some still undetermined time, it is possible that the word ayvān acquired the more technically narrow meaning of the architectural form to be discussed below. Thus in an inscription dated in 768/1366-67, the eastern hall of the Great Mosque in Isfahan is described as “this high ayvān” (Honarfar, Eṣfahān, p. 137). Whether the reference is to a form or to a place of particular distinction is not clear, as ʿAlī-Šīr Navāʾī, for instance mentions an ayvān with many columns, which certainly does not correspond to the vault of Isfahan. The matter will only be resolved after a careful survey of literary sources in proper chronological order.

The second common meaning of the word was developed by western art historians and archeologists, possibly under the impact of the monument at Ctesiphon. In this sense, the ayvān is a single large vaulted hall walled on three sides and opening directly to the outside on the fourth. Seen strictly as a unit of architectural composition, the ayvān is obviously one of the most consistent features of Iranian architecture since Parthian times. From Iran it was allegedly exported both eastward and westward, as in many buildings of thirteenth- and fourteenth-century Anatolia and Syria, in the madrasa of Sultan Ḥasan in Cairo, or in monuments of Islamic India. Within the Iranian world it is found in palaces, houses, mosques, madrasas, sanctuaries, and caravanserais. In mosques it is usually called ṣoffa. In pre-Islamic and early Islamic monuments, a single ayvān appears frequently associated with a domed hall, but its most conspicuous and celebrated use is in the combination of four ayvāns around a court. More than any other architectural element, this combination became the modular axis around which decorative and architectonic compositions were organized. Initially, as with the Great Mosque of Isfahan, these compositions were centered exclusively on the inner courtyard, defining the most characteristic Iranian architectural esthetic of the interior facade. Later, as in Timurid or Safavid masterpieces like the Ḵargerd madrasa, the Bībī Ḵānom mosque in Samarqand, or the Masjed-e Šāh in Isfahan, the ayvān also appears on the exterior of the monument, as a forecast of its interior forms. The exact history of these formal developments is still to be investigated as are the cultural or other reasons for whatever changes occurred in the use of ayvāns. But, regardless of the results of future scholarship on these issues, the ubiquitous importance of the ayvān is obvious enough.

Within this general definition and shortened survey of the ayvān in Iran several problems have attracted the attention of scholarship and require elaboration. Three of them are particularly important: 1. origins of the form; 2. the problem of the four ayvāns; 3. practical, symbolic, and esthetic properties of the ayvān.

1. Origins. The origins of the ayvān are virtually unknown.

The earliest known examples are found in the Parthian and Parthian-inspired monuments of Iraq. The most celebrated ones are the well-preserved single ones from Hatra (first century B.C.-second century A.D.) but recent investigations by E. S. Keall among others have shown that by the first century A.D. it was a common form in temples, palaces, and residences and that already at that time it was used singly, as two similar units facing each other, or even (Ashuz) as four units around a courtyard. Earlier research had tended to see in the ayvān an indigenous Mesopotamian development, possibly when technical innovations in vaulting made it possible and desirable in lands without wood to imitate the more formal buildings of the Iranian plateau. Some even saw its origins in the vernacular reed constructions of Mesopotamia, but this kind of explanation is no longer accepted. More recently it has been thought that this ayvān grew out of the impact on the Near East of Mediterranean architectural forms and that it is merely an adaptation to Mesopotamian or Iranian constructional techniques of the tablinum at court. Historically and culturally this is a reasonable hypothesis, since Parthian times are precisely the period when Hellenistic motifs were incorporated into the traditions of Iran or Mesopotamia. Its assumption, however, is that it was in Iran and/or Iraq that monumental vaulting first developed and this assumption has been questioned recently by historians of Mediterranean vaulting. It is curious, also, that the otherwise spectacular monuments from Parthian Nisa in Central Asia show extraordinary domes but no ayvāns.

To decide between these hypotheses is well-nigh impossible at this time, and one reason is the absence of adequate information from Iran itself and especially from its northeastern and Central Asian provinces. These are the provinces where the ayvān was to become particularly important in later times, but their Parthian and even later pre-Islamic monuments show, until now, only one identifiable ayvān, in the presumably Parthian palace at Kūh-e Ḵᵛāja in Iranian Sīstān. But the exploration of these vast regions has barely begun.

The form of the ayvān appears at the beginning of the Christian era and is almost immediately used for different functions and in varying arrangements, especially in Iraq. Whether or not it was a form imported from farther east can not be demonstrated.

2. The problem of the four ayvāns. A special problem in Islamic architecture is posed by the plan of a court with four ayvāns. The reason is that the spread of this particular plan has been related by many scholars to the growth of a major new Islamic function, the madrasa. The classical reasoning (Godard) is that the madrasa appeared first in northeastern Iran, that its functions were initially accomplished in private houses, that the original Khorasan house was one with four ayvāns, and that this house type became monumentalized in the Neẓāmīyas (one of which is alleged to have been traceable in Ḵargerd fifty years ago) and subsequently transferred into the standard plan for all possible functions of monumental architecture.

In this simplified form the scheme is not acceptable. The examples of four-ayvān private houses in eastern Iran are neither very numerous nor consistent enough to be used as models for later architecture; in addition, houses with exactly the same plan are known in ninth-century Iraq as well. What remains of the Ḵargerd madrasa is hardly sufficient for an acceptable reconstruction; and the assumption that four ayvāns were a convenient arrangement for the purposes of religious teaching is a fallacious extrapolation based on the ecumenical meaning of a few later madrasas. Most early madrasas were devoted to a single, or at most, to two religions.

Yet, even if the scheme as such is wrong, two of its features have some historical validity. The increasing popularity of the four-ayvān plan in religious architecture, as it appears for instance in a group of twelfth-century mosques in western Iran (Zavāra, Ardestān, Bersīān, Isfahan), indeed corresponds to the time of the spread of the institution of the madrasa, and inscriptions at Ardestān and literary sources at Isfahan (Honarfar, passim) indicate some sort of new relationship between the traditional Muslim place of prayer and specialized teaching functions, especially in Iran. The second feature is the northeastern Iranian background of the plan. The discovery at Ajina Tepe of a Buddhist monastery with four ayvāns as well as the number of formal variations in the ayvāns (including four around a court) found in presumably early Islamic houses in Sīstān suggest that Khorasan in its broadest sense may well have been the area where the monumental use of four ayvāns really originated or that it was revived there in Islamic times. Already it was used there for palaces (Laškarī Bāzār and Termeḏ) as well as for religious monuments.

3. Practical, symbolic, and esthetic properties of the ayvān. The ayvān was rarely used alone (partial exceptions in the sanctuary of Pīr-e Bakrān at Lenjān near Isfahan or the mosque of Nīrīz in Fārs) and was not therefore a discrete, individually meaningful, form. In Sasanian times it appears most commonly in combination with a dome and this ayvān-dome combination remained a most consistent feature of Iranian architecture. Alternately, it is the dominant feature of a court, in later times flanked by minarets. It is with two minarets that the ayvān is frequently used as a gateway.

Since we are not able to equate the Persian and Arabic word ayvān with the form called ayvān by art historians, the official meanings and associations of the former can not automatically be transferred to the latter. Yet it is likely that ayvāns in their art-historical sense did have specific practical and occasionally symbolic functions. In Sasanian times they were almost certainly the place of formal royal appearances, as the imperial crown was hung from the apex of the vault and a curtain stretched in front of the ayvān was opened to reveal the royal presence to the people assembled in the court. It is impossible to reconstruct the proper architectural setting for Islamic royal practices, except in fairly early times, when, in Baghdad or Marv, Sasanian practices and forms were probably maintained. But the exact use of all four ayvāns at a palace like Laškarī Bāzār is difficult to imagine. Honorific meanings can also be assumed for the ayvān in such examples as the sanctuaries of Lenjān and of Gāzorgāh, although for pious and commemorative architecture the ayvān never acquired the importance of the dome.

It is difficult to define the practical attributes of ayvāns. The single ayvān has the advantage of creating a single large space and its communal usefulness is obvious. But in a large building with several ayvāns, the unit is on the contrary a means to break up large spaces into small parts, each one with its individual focus. Whether or not the development of multi-ayvān mosques or caravanserais is a reflection of internal social or other divisions is still a moot question. Yet it is clear that the form itself would be used in very flexible fashion and its adoption in Islamic times corresponds to the whole culture’s concern for forms which could be used in many different ways and which would not become straitjackets. Finally it should be noted that, just as the ayvān would not have been possible without the development of vaulting, so also it became during the centuries one of the best vehicles for Iranian construction techniques.

Two themes predominate in a definition of the ayvān’s esthetic value. One is that it becomes a screen between interior and exterior worlds, controlling, as it does, the compositional rhythms of all walls and often their decoration as well. It is the place where inscriptions proclaim whatever secular or pious function a building had and the glory of its founders. Yet the ayvān is also a passageway, as the shadow of its interior intimates other parts of buildings. Hence it is not an accident that the ayvān became a gateway and that its decoration occasionally resembles that of a magnified meḥrāb. Its value as a frame made it a characteristic form of architecture in miniatures.



There is no real study of the ayvān; the following works may be consulted: O. Grabar, “Īwān,” EI2. Honarfar, Eṣfahān, pp. 86ff., 137.

N. V. D’yakonova and O. I. Smirnova, “K voprosu ob istolkovanii pendzhikentskoĭ rospisi,” Sbornik v chest’ I. A. Orbeli, Leningrad, 1960.

F. Oelmann, “Hilani und Liwanhaus,” Bonner Jahrbücher 127, 1922.

G. Gullini, Architettura iranica, Turin, 1964, pp. 326ff.

L. Golombek, The Timurid Shrine at Gazur Gah, Toronto, 1969.

A. Godard, “La mosquée,” Ars Islamica 6, 1951.

B. A. Litvinskiĭ and T. I. Zeĭmal, Adzhina-tepe, Moscow, 1971.

E. J. Keall, The Significance of Late Parthian Nippur, doctoral thesis, University of Michigan, Ann Arbor, 1970.

H. Behrens and M. Klinkott, “Das Ivan-Hofhaus in Afghanisch-Sistan,” AMI, N.S. 6, 1973.

N. Ardalan and L. Bakhtiar, The Sense of Unity, Chicago, 1973.

Camb. Hist. Iran III, pp. 489, 495, 1058-80 (passim), 1147, 1149.

Search terms:

 ایوان iwan  eyvan eywan
iwaan eyvaan eywaan  


(O. Grabar)

Originally Published: December 15, 1987

Last Updated: August 18, 2011

This article is available in print.
Vol. III, Fasc. 2, pp. 153-155