ARCHITECTURE v. Islamic, pre-Safavid



v. Islamic, pre-Safavid

The beginnings of an Islamic architecture in Iran are still almost impossible to identify properly. Remaining monuments are few, most of them are very uncertainly dated, and literary information is scanty or difficult to interpret. Only numerous and thorough archeological enterprises will alleviate this unfortunate state of affairs. In the meantime, three important assumptions underlie many of the interpretations given to the development of Islamic architecture in Iran. Though unproven, they all contain some element of likely truth; at the very least they are logical. One is that under Islam, many architectural techniques and purposes remained as they had been before in the two clearly identified centers of culture: the Sasanian empire in the west and especially the southwest and the Sogdian city states in the northeast. The second is that Islam brought one new architectural function, the mosque, and a new form, the hypostyle hall, occasionally used for mosques; there is a so far inconclusive debate on whether most mosques tended to use the new hypostyle form or whether traditional pre Islamic sanctuaries (especially the čahārṭāq or single domed building open on four sides) were commonly transferred to the new faith. Finally, it is assumed that a number of pre Islamic monumental and official forms like the ayvān and the dome on squinches and decorative techniques like stucco were immediately adopted for Islamic secular buildings and through Islam spread all over the new Muslim world. Whether correct or not, these assumptions are only of limited help in understanding the first three centuries of Islamic architecture in Iran. It is only from the 4th/10th century onward that enough monuments and literary or epigraphical monuments remain to lead to more secure conclusions and more fruitful hypotheses.

Chronology. Iran did not develop unified, easily defined stylistic and technical tendencies until Shah ʿAbbās’s time. Two factors present throughout the first six or seven centuries of its known architectural development modify the drawing of general conclusions. One factor was regional distinctions: what may be true of Khorasan at one time is not necessarily valid for Fārs or Azerbaijan. There were notable instances of artists and motifs moving from one area to another, as with the architects and decorators from Shiraz who worked in Khorasan in the middle of the 9th/15th century, or with a certain type of brick decoration, or with the moqarnas, which seems to have migrated from Khorasan westward in the 5th/11th century. The second modifying factor was sociopolitical. Especially from the middle of the 5th/11th century, after the rule of the Great Saljuqs, there seems to have been two levels of architectural patronage: an “imperial” level of major princes and dynasties and a “local” one of urban notables. The phenomenon is most clearly definable in the 8th/14th century, when an Il khanid imperial tradition operated simultaneously with several local ones in the areas of Isfahan and Yazd, but it was present both earlier and later.

Keeping these two factors in mind, the following chronological ordering of Iranian Islamic architecture can be proposed on the basis of remaining monuments: Fourth/tenth—fifth/eleventh centuries. A period best known in Khorasan, Transoxania, and Afghanistan, its major monuments are the Samanid mausoleums of Bukhara and Tīm, the small mosque in Balḵ, the Samanid houses and palaces at Afrāsīāb, and the palaces of the Ghaznavids in Termeḏ, Ḡaznī, and Laškarī Bāzār. It was the time of the development of baked brick construction and decoration, of remarkable growth of stucco decoration, and of the first changes introduced into the squinches of more traditional architecture. It may be legitimate to call this period Samanid Ghaznavid, although eventually the architecture of the two dynasties should be definable separately, and Ghaznavid inspired architecture did continue well into the 6th/12th century, especially in Afghanistan. It is a period poorly known in western Iran except for major recent discoveries of Buyid monuments in Isfahan, which may have formed an entirely separate school.

Fifth/eleventh—early seventh/thirteenth centuries. This is the first period known in most provinces of Iran. Its most brilliant group of monuments is in Jibal (Isfahan, Zavāra, Ardestān, Barsīān). There are many local variants, major subgroups occurring in Azerbaijan in the later 6th/12th and early 7th/13th centuries (Marāḡa and Naḵjevān), and in Khorasan in the 6th/12th century (Rabāṭ Šaraf, Saraḵs, Bukhara, Marv, Dāya Ḵātūn). Most of the remaining architecture consists of mosques, mausoleums, minarets, and caravanserais. It was the time when the four ayvāns plan became truly ubiquitous, when dome construction and moqarnas acquired particularly logical architectonic characteristics, when complex geometry ruled over decoration and composition, when formal portals and tall minarets spread all over Iran, and especially when a most successful balance was reached between construction and decoration. Except in special cases like the mausoleum of Sanjar in Marv or the north dome of Isfahan’s Great Mosque, where imperial taste and patronage are clear, most of the monuments of this period have not yet been analyzed to separate imperial and local features. It is appropriate to call this era the Saljuq period, provided all its features are not necessarily attributed to the dynasty.

Eighth/fourteenth century. Because of the ravages of the Mongol invasion the 7th/13th century has left few traces, but with the consolidation of Il khanid rule at the end of the century one of the most fruitful periods of Iranian architecture began. Though apparently dominated by the great imperial monuments of Tabrīz and Solṭānīa, recent scholarship has shown that the era is probably more significant as the period in which the cities and even villages of western, southwestern, and central Iran (Isfahan, Naṭanz, Yazd, Dašt-e Lenjān, Kermān) acquired their quasi permanent architectural setting. The situation elsewhere in Iran is more difficult to assess properly, as only individual and somewhat isolated examples (Varāmīn, Darband, Bākū, Marand) have been identified. It is still too early to be able to define the stylistic characteristics of this century; it is possibly by regions that 8th/14th century style will eventually be most meaningful. Some innovations, however, are evident: particular concern for light in cupolas, alleviation of walls in selected areas, experimentation with vaults, color in decoration, many new techniques of ornament.

Ninth/fifteenth century. This Timurid century is easier to define. Its main creative centers were the dynastic ones in northeastern Iran (Herat, Samarkand, Mašhad, Turkestan, Anau, Tayābād, Kargerd), but its masterpieces are found in Tabrīz, Qom, Yazd, and Kermān as well. While the functions and purposes of architecture did not change much in this era, its techniques, styles, and especially mood did. Technically there occurred a new series of experiments with vault construction—almost the last such in Iran—leading to an almost total separation between exterior and interior effects of domes and to the logically brilliant, if at times overly complex, arrangement of ribs and of curved surfaces. Timurid architectural style concerned itself with the elaboration of interior proportions, with the composition of surface decoration, and especially with the uses of faience mosaics and multi colored tiles. As a result its mood is one of luxurious effects which are perhaps more secular than religious, more externally brilliant than directly inspiring. Yet it possesses a unique quality of almost magic attraction. Its colors, subtle compositions, and structurally compact monuments serve as focal points in the cities or villages where they are found.

Characteristic forms and purposes. The most common monuments of Iranian Islamic architecture are primarily religious: large congregational mosques, small private mosques, madrasas, ḵānaqās, mausoleums for holy men—often with elaborate attending institutions. Most of these functions are typical of all Islamic architectural traditions, although the development of large sanctuaries around the burial places of imams is a more specifically Iranian phenomenon. Also peculiarly Iranian is the existence of dynastic and personal mausoleums. From the Samanid one in Bukhara to Sanjar’s at Marv, Oljāytū’s in Solṭānīa, and Timur’s in Samarkand, major monuments were built to glorify a specific prince. As is apparent from the inscriptions on some of these mausoleums, they tended to reflect more than royal vainglory and often served as semi religious cultic centers.

Secular functions of Iranian architecture have not been preserved as well as religious ones. A large number of caravanserais remain, mostly from the 5th/11th and 6th/12th centuries. Palaces and fine houses are known primarily in the East, and much more for the 4th/10th to 6th/12th centuries than for later times, except for fragments of Timurid buildings. Baths, private houses, bazaars are very poorly known outside the excavations at Sīrāf, although later and better preserved examples (from Kāšān or Marv) may be taken as representative. The same is probably true of formal gardens. Largely fragmentary city walls, citadels, and military architecture remain almost everywhere, and a 4th/10th century bridge is still standing near Isfahan. For any further study of the typology of functions an essential source consists in waqf documents, the most notable examples being a 4th/10th century waqf for Samarkand, a 6th/12th century one inscribed on the walls of Qazvīn’s congregational mosque, the waqf nāma of Rašīd al dīn, and several waqfs dealing with Yazd.

It is probable that there was a distinctively Iranian density of architectural functions in some areas or at some times, but so far this can not be identified with certainty. Iranian originality is clearer when one turns to architectural forms, for regardless of the function of a monument, the same small number of formal units appears constantly at all times, in almost any setting, and in almost every area. Five of these are particularly notable: the dome, the ayvān, the court, the tower, and the wall.

The dome, majestic at Solṭānīa and Samarkand or infinitely varied in Isfahan, is one of the true glories of Islamic architecture in Iran. It can stand alone or be set in series and cover large spaces. It can be artificially heightened as in the mausoleum of Tamerlane or almost flat as in the small bays of the congregational mosque of Isfahan. Most significantly, it was the source of the most imaginative and exciting technical developments of Iranian architecture: the endless variations introduced into the pre Islamic squinch ending up with the moqarnas or intersecting ribs of late Timurid and especially Safavid monuments. A further peculiarity of the Iranian dome is that it was not used simply as a complete entity. It was also segmented: its parts (halves, thirds, quarters) became separate compositional units of their own.

The ayvān and the court are less original to Islamic Iran, since both have a long pre Islamic history in Iran and Iraq. Two notable things happened to them in Islamic times. One is that they both became widespread morphological elements throughout Iran. (There is still much uncertainty about the development of the ayvān in early Islamic times.) The second is that the court and from one to four ayvāns were transformed into the compositional and modular axes of most major buildings. The ways in which this actually happened have not yet been investigated adequately, but from the simple 6th/12th century mosque at Zavāra all the way to the spectacular early 9th/15th century one in Samarkand, these are the elements around which the monument is organized. In some of the earlier masterpieces (before the growth of axial portals) the court with its ayvāns formed the inner facade of most monuments.

There is still much debate about minarets, best understood simply as towers, the tall cylinders (at times polygons) which rise above so much of the Iranian landscape. Some did serve for the call to prayer, others were watchtowers or commemorative monuments, some were even tombs. Although some of them are squat and ungainly, most are tall and thin. They can be undecorated or, as in Bukhara or Mašhad, carry a most impressive ornament. They can stand alone, in pairs framing portals, or in groups serving to emphasize the main forms of as widely different monuments as the Solṭānīa mausoleum or the mosque of Bībī Ḵānom in Samarkand. The origin of the form is still obscure, as are the reasons why it acquired such a uniquely wide use.

The last morphological unit of Iranian architecture to emphasize is the wall. As such it is obviously not unique to Iran, but it acquired certain peculiar characteristics in Iran. There are several reasons for this: the limitations of brick as the main medium of construction; the lack of true column tradition; the absence of well developed means to light interiors. As a result the wall became one of the most elaborate elements in a monument. Grandiose and grandiloquent in the mosque of ʿAlī Shah in Tabrīz, it is pure design in the Gonbad-e Qābūs. In the mausoleum of the Samanids its construction and decoration became one, while in the Ḵārraqān mausoleums it supports an elaborate decoration. In Isfahan’s north dome it reflects the superstructure, but in Ḵārgerd or Oloḡ Beg’s madrasa in Samarkand it is articulated by extraordinary color designs. This variety of treatment applies not only to the wall’s surface, but also to its shape, for one of the peculiarities of Iranian architectural supports is that almost all of them (some early exceptions notwithstanding) tend to become segments of walls.

Further investigations will no doubt bring out additional morphological elements in Iranian architecture, but the important point is that there did not always exist a clear and consistent correlation between forms and functions. At times converging, they had very different histories.

Decoration and esthetic values. The techniques of architectural decoration found in Iranian Islamic architecture are quite varied: stucco, brick, terra cotta, faience mosaic, tiles, carved wood, in any combination. The history and stylistic definition of these techniques still remain to be written. This is also true of the designs and motifs used on them: writing, vegetal designs, geometry, much more rarely figural themes. At this stage of our knowledge we can only outline two broad questions about Iranian architectural decoration.

The first of these is whether there was an iconography of architectural decoration or whether much of it should be considered simply as ornament. The hypothesis is that most of it was not mere ornament, if one excepts obvious devices emphasizing for instance detail of construction. At times, as in palaces (usually only known through texts), there was a concrete iconography of decoration—illustrations of Persian epics, for instance, or, in mosques, precisely chosen Koranic references. Much more frequently decoration served to create a spiritual or sensuous mood, without concrete meanings necessarily attributed to any one motif. There is the mood of geometric patterning, which derives directly from Islamic concerns for mathematical theory and which is also an expression of a certain rational conception of divine and natural reality. Then there is the mood of luxuriance, brilliant flowers and pseudo gardens, festivals of colors, in which the beholder’s eye is lost; it is possible that this mood is a reflection of a paradisiac vision of another world, serving to suggest that the monument of architecture (especially religious) is an intimation of an esoteric reality. In secular buildings it may rather be wealth or pleasure which is thus suggested, but information for the period before 1500 is still too fragmentary. Much research is still needed before any interpretations are fully anchored.

The second question is perhaps more technical but quite important for history. It deals with schools of decorators and regional or chronological styles. Decorators moved a great deal throughout Iran, but we still do not know very well whether they carried with them original motifs and styles or whether they adapted themselves to whatever may have been required by patrons.

The last question about Iranian Islamic architecture is whether it possessed a unified esthetic outlook. Some writers have argued that a particularly Iranian sense of architectural and decorative values has permeated all Iranian architectural creation. Others have felt that the pre Safavid Islamic centuries of Iran were a time of interchange with surrounding areas, and that overall period styles best characterize these centuries. It is perhaps not possible to settle the debate, as both propositions may well be true. Brick architecture, the ayvān, color decoration, sensuous forms, geometry, are all very Iranian features, some of which go back to pre Islamic times, but their Islamic expression was perhaps much more culturally than ethnically definable.


Bibliography: See in print ed., EIr. II/4, London, 1986, pp. 342-45.

(O. Grabar)

Originally Published: December 15, 1986

Last Updated: August 11, 2011

This article is available in print.
Vol. II, Fasc. 4, pp. 339-345