None of the royal palaces and pavilions of Isfahan built prior to the 17th century are extant. In contrast, of all the monuments of Isfahan, Safavid palaces represent the most coherent group of buildings to have survived from a single period (Galdieri, 1974). Seventeenth-century palaces in Isfahan, like other principal imperial monuments of the Safavids, served in their architectural and spatial particularities the ceremonial and ritual needs of a reconfigured household. This was a time when the Safavid household increasingly abandoned its peripatetic ways of rule, necessitated in the age of confederacy, for a sedentary demonstration of authority and legitimacy. Whereas in the 16th century, the Safavids entrusted the princes, the family’s bloodline, with the powerful qezelbāš governors of key provinces of the realm, in the Isfahan phase of imperial rule, the princes were kept at the harem in the capital under the supervision of their concubine-mothers and the tutelage of the eunuch ḡolāms (Babayan, 2004 pp. 30-31).
European visitors to Safavid Persia, for example, found themselves increasingly bound by Isfahan, where they were able to gain a royal audience or conduct their business with the court and government bureaucracy without having to follow the itinerant monarchs (Babaie, 1994a, pp. 264-68). Pietro Della Valle (q.v.), who arrived in Persia in 1617, did move from city to city and camp to camp to meet with Shah ʿAbbās I the Great. In contrast, those following him, Jean-Babtiste Tavernier (six journeys to Persia between 1632 and 1668) and Adam Olearius (the secretary of the Holstein envoy in Isfahan in 1637-38), for example, could not have done any of their “official” business outside Isfahan.
Isfahan, more than any other seat of power in the history of Islamic Persia, assumed unique centrality for the performance of kingship during its time as the Safavid capital. A representation of this shift in emphases is the massive urban campaigns carried out during the first half of the 17th century. Its critical imperial role is also clearly reflected in the sheer number of its palaces and the diversity of their architectural articulation of a range of functions. The ʿĀli Qāpu (q.v.), the Tālār-e Ṭawila, the Čehel Sotun (q.v.), and the Hašt Behešt constitute the principal palaces within the Dawlat-ḵāna, the royal precinct of Isfahan. Others, such as Hazār Jarib (reign of Shah ʿAbbās the Great), Āʾina-ḵāna (reign of Shah Ṣafi I), and Faraḥābād (reign of Shah Solṭān-Ḥosayn), served as retreats away from the urban hustle and bustle but still within reach of the city. Instead of moving to Māzandarān summer palaces, for example, which was the habit especially of Shah ʿAbbās the Great, the later Safavids spent the hot summer months at the Hazār Jarib or the Faraḥābād, both garden-palace retreats located south of the city. Centralization had fixed the geography of authority unto the capital city of Isfahan and the convergence of the production and consumption of politics and culture required the architectural accommodation of its operations. The royal precinct and its palaces were at the heart of this imperial enterprise.
Notwithstanding the distinctive Islamic practice of dividing the royal precinct into the private (andarun, q.v) and public (birun) zones, any discussion of palaces in the Islamic world will also have to take note of the different typology that applies to the Western terminology of the architectural experience. The conglomerate that makes up the Dawlat-ḵāna in Isfahan is not a contiguous series of open and closed spaces compounded into a single, albeit massive, building like the Louvre in Paris. Neither is it a barricaded clustering of rooms, halls, and loggias around successive courtyards that is exemplified by the Topkapı in Istanbul. Instead, the Dawlat-ḵāna of Isfahan is composed of a number of large ceremonial palaces and clusters of office buildings, workshops, and residential apartments that are held together through the intermediary of gardens and within boundaries that are marked by “soft” walls.
At the time of the initial constructions, begun in 1590-91, the Safavid royal precinct was the site of the old, extra-urban Bāḡ-e Naqš-e Jahān. As the building campaign progressed, by the official transfer of the capital in 1598, the Dawlat-ḵāna had been flanked by the two urban foci of the Meydān-e Naqš-e Jahān and the Čahārbāḡ Avenue and had become the imperial seat of Safavid dominion (McChesney, pp. 119-20). The boundaries of the Dawlat-ḵāna, unlike in either earlier or contemporary examples, were articulated by the dual function of urban features: the two-story shops and coffeehouses that formed the Meydān periphery and served also as a wall of separation between the royal precinct and the public space; and the verdant gardens along the Čahārbāḡ that lined the western side of the Dawlat-ḵāna.
Access to the precinct was regulated through gates that were strategically positioned to enunciate their multifarious imperial functions: the ʿĀli Qāpu Palace gateway served as the seat of the judiciary as well as for the ceremonial events of the court; the harem gate, just south of the ʿĀli Qāpu, accentuated the inseparable ties in the administration of the centralized empire between the Safavid household and the slave-based political structure. Another gate, known as the Čahār Ḥawż (Four ponds), was located north of the ʿĀli Qāpu and accessed the Daftar-ḵāna or Imperial Chancellery. The Kitchen Gate (Darb-e matbaḵ), located south of the harem gate, brought to the edge of the public sphere the symbol of royal generosity, as it was through this gate that food would be distributed to the poor and dispatched to the guests of the shah. Together, these gates communicated through their architectural mediation of functions the relationship between the Safavid household, the locus of power and authority, and the denizens of Isfahan and, by extension, of the entire Safavid dominion.
ʿĀli Qāpu. The first amongst the urban Safavid palaces in Isfahan was the ʿĀli Qāpu. Its conceptualization coincided with that of the Meydān project (Galdieri, 1979, pp. 1-40; Babaie, 1994a, pp. 99-135; see Plate I ). During the initial campaign of 1590-91, the preparation of the Meydān as the new urban center was accompanied by the construction of the Qayṣariya Bazaar and a two-story building that served as the gateway into the royal precinct of the Bāḡ-e Naqš-e Jahān (Galdieri, 1979, pp. 3-15; Babaie, 1994a, pp. 115-24). This initial, modest gateway was to be transformed in the 1602 campaign, during which the first “block” was raised by the addition of two stories. This phase of construction invested the gateway with the monumentality of a palace. It also redefined the functional parameters of this most accessible royal seat of the Dawlat-ḵāna complex.
The first block of two stories, as archeological investigations and extant evidence demonstrate, was an integral part of the Meydān construction. It was designed on a cross-in-square (or nine-fold) plan in which the spaces or rooms formed by the arms of the cross and the intervening spaces in the four corners rotated around a central area. This type of building, known as a hašt behešt (eight paradises, for its eight spaces around the central opening; q.v.), was associated with such royal pavilions as the 16th-century so-called Čehel Sotun of Safavid Qazvin and the 15th-century Timurid pavilions in the royal gardens of Samarqand and Herat (known only through descriptions). The initial block of the ʿĀli Qāpu gateway then represented an instance of continuity in architectural traditions of the past.
In contrast, the second block, the additional two stories of 1602, marked a radical departure in the conception of a royal seat, a palace (Plate II ). Here the third floor is principally given over to a rectangular audience hall, the narrow side of which faces, through shallow porches (ayvān, q.v.), the Meydān and the royal precinct sides. The audience hall rises to the full height of the two stories with its vaulted roof held aloft through an innovative forked-arch system that allowed for fenestration on the upper zone of the side walls (Galdieri, 1979, pp. 16; Ferrante, pp. 142-43; Babaie, 1994a, pp. 101-2). An ingenious double-story mezzanine occupies the sides of the audience hall unit on the fourth floor. This little-known architectural feat is hidden into the floor that was reserved for the women of the harem, who were provided a glimpse of the ceremonial gatherings below through windows that filled the in-between squinch (triangular) surfaces of the forked-arches. The addition of the final fifth floor (comp. by around 1615), of a private cluster of intensely decorated rooms, concluded the first phase of the transition from a gateway into a ceremonial palace during the reign of Shah ʿAbbās the Great (1587-1629).
With its original functions as the seat from which emanated imperial justice, the enhanced functionality of the ʿĀli Qāpu as a ceremonial palace rendered this the uncommon example of the marking of the threshold between the public and the royal arenas of a capital city in the early modern age. Whereas the ceremonial seat of authority and its performance of kingship in the Ottoman Topkapı and the Mughal Red Forts were withdrawn deep inside the fortified royal precincts, the Dawlat-ḵāna of Isfahan placed its most ceremonial palace at the threshold between the royal and the public. The placement of the royal chapel-mosque of Shaikh Loṭf-Allāh (see below) directly across from the ʿĀli Qāpu Palace gateway further illustrated for the denizens of Isfahan and of the imperial dominion of the Safavids the significance of ritual accessibility and humility that was necessitated by the discourse of legitimacy in a Twelver Shiʿite construction of kingship.
Unlike the assumption of aloofness in varying degrees of intensity that befitted a Sunnite, caliphal model of kingship (adopted by the Ottoman sultans and Mughal emperors), the Safavid imperial posture was predicated on the imperative of humility at the threshold of Imam ʿAli. In fact, Shah ʿAbbās the Great, who identified himself as “the dog (or dust or servant) of the threshold of ʿAli” (kalb-e āstān-e ʿAli, kalb-e āstān-e welāyat; see, Shah ʿAbbās’s poem in Jalāl-al-Din Monajjem, pp. 163, 237; Falsafi, II, the plate facing p. 24) instituted the practice of dismounting before the ʿĀli Qāpu, for this gate was sanctified with the installment of a door from the Shrine of ʿAli in Najaf, and was thus made holy for the shahs ruled on behalf of the Mahdi (Messiah), the Hidden Twelfth Imam.
Tālār-e Ṭawila and Āʾina-ḵāna. Like Shah ʿAbbās, his successor Shah Ṣafi I (r. 1629-42) spent a considerable length of time, especially during the first years of his reign, outside of Isfahan, engaged mostly in the renewed conflicts with the Ottomans on the western frontiers. Shah Ṣafi I’s coronation in 1629 took place at the ʿĀli Qāpu, but the waning of the Palace-Gateway’s functionality in his reign indicates a shift in Safavid ceremonials (Babaie, 1994a, pp. 234-39). Two palaces are attributable to his reign on the basis of their mention, for the first time, in Persian chronicles of his reign: the Tālār-e Ṭawila (the Hall of the Stables) in the Dawlat-ḵāna and the Āʾina-ḵāna (the Hall of Mirrors) on the southern bank of the Zāyandarud River. Neither one has survived, but descriptions and some engravings help us reconstruct their basic architectural configuration and functions (Babaie, 1994a, pp. 245-53; Honarfar, 1965, p. 752; Mirzā Ḥosayn, p. 23; Floor, pp. 149-63).
According to Adam Olearius, the secretary of the Holstein Embassy, and other European visitors, the Tālār-e Ṭawila was a large elongated building located in a garden a little way to the southwest of the ʿĀli Qāpu. Set at the end of a long pool, the Tālār-e Ṭawila was raised on a platform and was subdivided along its longitudinal axis into three sections, beginning, from the outermost, with a pillared hall with a fountain and concluding at its innermost with an ayvān porch for the throne. Wooden pillars held up the roof in all three zones while low wooden banisters separated the sections thus demarcating the degrees of remove and significance (Olearius, 1669, p. 202; Tavernier, 1678, p. 178).
The architectural configuration of this palace is indeed extraordinary, for it is the antithesis of the earlier hašt behešt model. Not only is the plan laid out on the longitudinal axis instead of the centralized cross-in-square, it is also here that the traditional ratio between solid and void has been completely reversed in favor of considerable openness. Pillared halls were not a novelty, but the particularity of the pillared halls at the Tālār-e Ṭawila ushers in a new chapter in the Persian palace designs. Unlike the horizontal stretch of pillared halls of Achaemenid palaces or even the contemporary Mughal Divān-e ʿĀmm, the tālār in Tālār-e Ṭawila consists of a sequence of pillared spaces that are completely open on three sides, and that are placed one behind the other on the longitudinal axis (Babaie, 1994a, pp. 253-61; Koch, 1994, pp. 143-65). The privileging of the tālār in the Tālār-e Ṭa-wila may in fact be an indication of a consciousness about its architectural novelty.
European descriptions of feasts and audiences at the Tālār-e Ṭawila indicate that this type of palace was especially suited for large gatherings (some describe 500 guests) in a space that could be hierarchically zoned but that also allowed for visibility and access, at least in their potential. The Tālār-e Ṭawila, in fact, became the most frequently utilized building for feasts and audiences during the reign of Shah Ṣafi I and continued to be one of the principal ceremonial palaces in the Dawlat-ḵāna. Equally open and ceremonial was the Āʾina-ḵāna, the other palace built during the reign of Shah Ṣafi I, which is also based on a tālār design. The Āʾina-ḵāna was amongst the nearby suburban palaces developed along the southern bank of the Zāyandarud River, and as such the tālār there utilized its unobstructed view of the river in the same way that the Tālār-e Ṭawila communicated with the long pool in its garden setting. The Āʾina-ḵāna, better known thanks to 19th-century engravings by Pascal Coste, may also be dated to the 1630s on the evidence that it was utilized for ceremonial events and was also praised for its beauty in a poem by Moẓaffar Torka (Coste, Pl. XXXV; Babaie, 1994a, pp. 250-51; Honarfar, 1965, pp. 578-80).
ʿĀli Qāpu Tālār. The dominant transparency of the Tālār-e Ṭawila and the Āʾina-ḵāna, as well as their spatial and visual intermingling with the outside (water and garden elements in these cases) mark a new turning point in palace architecture in the Iranian world. Soon after Shah ʿAbbās II ascended the throne in 1642, his grand vizier, Mirzā Moḥammad Sāru Taqi, the most likely mastermind behind the idea of the tālār type palace, was entrusted with the construction of a tālār in front of the existing five-story tower of ʿĀli Qāpu (Babaie, 2004, pp. 106-7; see Plate III ). The addition consisted of a massive two-story unit that provided the gateway with a vaulted passage on the ground floor and the audience hall on the third floor with a tālār.
This tālār radically altered the functionality of the original palace. As the Safavid court ceremonials had increasingly tended towards enlarged and elaborate feasts and gatherings in the hundreds, the ʿĀli Qāpu’s audience hall and its shallow ayvān facing the Meydān had seized to be functional. With the tālār addition, the ʿĀli Qāpu resumed, as contemporary evidence testifies (e.g., Waḥid, pp. 198, 222), its role as one of the premium ceremonial palaces in the Dawlat-ḵāna. Here, however, the tālār at the ʿĀli Qāpu afforded the public an even greater visual access to the performance of kingship.
Čehel Sotun. Following the architectural intervention and the functional updating of the ʿĀli Qāpu, Shah ʿAbbās II (r. 1642-66) sponsored the building of the Čehel Sotun Palace within the grounds of the Dawlat-ḵāna (Honarfar, 1965, pp. 557-74; Mehrābādi, pp. 339-61; Ferrante; see Plate IV ). A chronogram by the poet Sāʾeb and other evidence date the building to 1647 while the completion of its mural decorations may have taken until the early 1650s (see Honarfar, 1972b and Babaie, 1994b). The Čehel Sotun marks the full architectural realization of the experiments that began with the Tālār-e Ṭawila. It furthermore signals the institutionalization by this mid-century date of the practice of feasting as an imperial ritual that enunciated the particularity of the Safavid kingship that necessitated formal spaces for its performance.
Jean Chardin (q.v.), the French traveler, exemplified the feasting rituals of the Safavids by describing just such an event at the Čehel Sotun, in which the mass of guests were wined, dined, entertained and were engaged in political conversations in the company of the shah (Chardin, V, pp. 468-79). This manner of feasting the multitudes of guests, where the shah himself resided as its host, was the Safavid answer to the symbolic need for accessibility as the deputy of the Imams. As such, it stood in sharp contrast to the aloofness of the Ottoman sultans and Mughal emperors, for example, whose remoteness in the manner of the ʿAbbāsid caliphs prohibited any public display of such earthly and mundane acts as eating, much less sharing food with others of profane status.
The Čehel Sotun is also laid down on its longitudinal axis. Set within a huge garden, its elongated plan is compellingly emphasized by the landscaping of the grounds in which a long pool in front and a smaller one in the back appear as if extensions of the structure itself. A massive tālār of twenty wooden pillars (forty with the reflection in the pool and hence the name, Kāḵ-e Čehel Sotun [Forty-Columned Palace]) forms the front of the building (Plate V ). This tālār, enlivened by a fountain-pool at its center, leads to ever decreasing spaces in size, which are also increasingly more contained by walls and vaults. At the culmination of this processional and hierarchical spacing, which had, for performative purposes, begun at the far end of the long pool, sat the throne ayvān. Here, the decorative scheme of painted, gilded and mirrored columns and walls intensified into an explosive, prismatic mirror-work in the vault of the ayvān and a pair of mural panels of scenes of courtly pastime on the side walls.
Exceptionally in the Čehel Sotun, a large triple-vaulted audience hall lies behind the tālār-ayvān reception area on the west (Babaie, 1994a, pp. 138-39). The hall is a rectangle turned perpendicular onto the main axis of the building. It opened on the sides (north and south) to two verandas (both with a single row of wooden columns and with mural decoration), while at its corners were placed four smaller rooms of intimate decoration and scale, most likely suited for private use. The architectural articulation of the hall and its extensive mural program enunciate its purely ceremonial function. Mural paintings range in size and subject matter: the smaller ones on the lower walls are of scenes of courtly leisure; the four large rectangular panels on the upper walls are of historicized themes in which Safavid monarchs are portrayed as victorious in battle as well as in politics.
The decorative scheme then visualized the centrality of the Safavids as the arbiters in matters of sovereignty within the competitive politics of the Persianate world at the same time that the architectural articulation of the Čehel Sotun facilitated the performance of royal feasts, which had become increasingly the stage for the enunciation of Safavid kingship.
Hašt Behešt. One of the most famous Safavid palaces in Isfahan is the Hašt Behešt, a relatively well-preserved building dated to 1080/1669 (Honarfar, 1965, pp. 622-26; Mehrābādi, pp. 328-39; Naṣrābādi, pp. 487-90; Mirzā Ḥosayn, pp. 25-26; Ferrante, 1968c; see Plate VI ). As had been the pattern in the Safavid constructions of palaces in Isfahan, Shah Solaymān (r. 1666-94) also embarked on the construction of a new palace soon after he ascended the throne. His palace, however, was constructed not in the birun or public zone of the Dawlat-ḵāna, but in the vast gardens of Bolbol “Nightingale,” located to the south of the harem. As such, the Hašt Behešt manifested, both in its location and its architectural and decorative strategies, the priorities of a more private function than the feasting rituals and ceremonial events held at the Tālār-e Ṭawila or the Čehel Sotun.
As its name indicates, this was a building based on a hašt behešt (q.v.) plan. Raised on a platform, the structure reads as a square with its corners cut away to resemble a modified octagon. The corners are accentuated by the massive, two-story rise of pylons that are perforated with stacked pointed-arch niches. Each of the four sides of the building, on the other hand, is given to spacious porches roofed over with the placement of a few tall and thin wooden columns. The large and tall central space is a fully articulated octagon, rising well above the two-story height through an elegantly tapering and mirrored internal lantern (Plate VII ). The sensuality of the volumetric alternating of solids and voids within the structure of the Hašt Behešt itself is contrasted with the rich surface articulation through murals, perforated woodwork, prismatic mirrors and extensive tile work of figural subject matter (Luschey-Schmeisser, pp. 7-28). Descriptions of events at this building, of leisurely outings of the harem women and of the private gatherings, clearly synchronize with the architectural and decorative particularities of the Hašt Behešt (Chardin, pp. 39-43).
Suburban palaces. With the Tālār-e Ṭawila, the ʿĀli Qāpu, and the Čehel Sotun, the Dawlat-ḵāna had been equipped by the middle of the 17th century with ceremonial palaces singularly suited to feasting that was accommodated by the tālār spaces. These palaces were complemented by the suburban retreats that successive Safavid monarchs built in order to facilitate the need for retreat while still centered at the capital city. The Hezār Jarib (or ʿAbbāsābād) of Shah ʿAbbās the Great, built in the 1590s on the model of gently sloped terraced gardens, mediated the transition from the Čahārbāḡ Avenue and the Ṣoffa Mountain, on the foot of which it was laid out (Honarfar, 1975, pp. 76 ff.; Alemi p. 75; Mirzā Ḥosayn, p. 27).
Palaces along the Zāyandarud River, especially the Āʾina-ḵāna and the so-called Haft Dast ensemble of Shah Ṣafi I and Shah ʿAbbās II lined the southern bank of the river between the two bridges of Allāh-verdi Khan and Ḵᵛāju (see below; Haft Dast was destroyed in the late Qājar period, see Honarfar, 1965, pp. 678, 680). Together with garden retreats and parks on the northern bank, this complex made up the riverfront ensemble of Saʿādatābād, all of which was destroyed during the Qajar period (19th century). Equally large and complex was the Faraḥābād palace-city built by Shah Solṭān-Ḥosayn in the early years of the 18th century (Honarfar, 1965, pp. 722-25; Mehrābādi, pp. 80-81). This, however, was not simply a retreat for the last Safavid king but served as his principal place of residence during a reign that was mired by the delegation of power to the eunuch ḡolāms of the harem and the powerful clergy. When Maḥmud the Afghan advanced on Isfahan in 1722, it was from Faraḥābād that he led the siege of the capital city and it was to Faraḥābād that Shah Solṭān-Ḥosayn came to deliver his ancestral crown and the Safavid Empire (Krusinski, II, pp. 36-37, 98-100).
(Sussan Babaie with Robert Haug)
Originally Published: December 15, 2007
Last Updated: April 5, 2012
This article is available in print.
Vol. XIV, Fasc. 1, pp. 14-20