III. LINEAR EXPANSION:
Phimai & Kandariya Mahadeva
No Khmer temple displays four-way symmetry throughout nor is any built with perfectly concentric terraces or tiers crowned by a central tower, all of whose parts can be arranged inside a single square mandala. Instead, these buildings characteristically reveal an east-west asymmetry with a clear directional “liturgical axis” leading from the profane peripheral world to a sacred center, the shrine and its shikhara. Most Hindu and Buddhist temples are east-facing, the direction of the rising sun, associated with birth, while the west is usually associated with death. Even Angkorian “temple mountains” or pyramids, which in theory should be square, circular or a combination of the two, are generally off-set towards the west and preceded by extended “processional paths” from their eastern gopuras or gateways, past a cluster of buildings whose purposes often remain obscure, to the garbagriha, sanctum sanctorum or sanctuary. The most practical explanation for this anomaly is the need of Hindu liturgy to prepare worshippers to behold the god in his earthly home and receive his darshan, to see and be seen in his auspicious gaze. Although Hindu worship or puja is normally individual rather than congregational, this still requires space for ritual purification, chanting, invocations, sermons, perhaps sacred dancing, an assembly hall or sabhamandapa where devotees can gather before the small, dark, cave-like garbagriha to glimpse the deity, as well as an ambulatory or corridor for pradakshina, ritual, solar (clockwise) circumambulation, (though there is no evidence that this was practiced at Angkor.)
Major ceremonial centers or pilgrimage sites such as Borobudur and Angkor Wat, seem to have resolved the need for an elongated liturgical path by bending their ritual axes around the edges of concentric terraces with galleries lined with shrines and didactic bas relief, in effect, turning a straight line into an inward-turning, right angle spiral which might require days-long liturgies, as well as, a background in esoteric theology. Such expensive and time-consuming “processional paths” would be impractical for smaller, local temples intended for daily rituals performed by resident brahmins, the Buddhist sangha and the laity. Architects generally solved the need for an eastward extension from the cella or garbagriha by the simple expedient of stringing one or more mandapas or halls along the “liturgical axis.” Indeed, one way to trace the evolution of the Indian temple is through its sthapakas’ and sthapatis’ efforts to unify this chain of functionally dissimilar structures into a coherent whole, a challenge usually thought to have received its classic solution at the Kandariya Mahadeva Temple (c.1030) at Khajuraho (see figure I.)
The architectural demands of darshan find a counterpart in Christian Holy Communion, another transubstantiation of the profane into the sacred and a symbol into its signified. Indian and European architects thus faced a surprisingly similar conundrum: how to accommodate an extended, ritual axis culminating at the point where the god becomes physically manifest and, at the same time, place this intersection of the celestial and terrestrial at the center of their shrines? In India, this problem translated into having the garbagriha occupy the central padas of a square mandala, while, in the West, it meant placing the altar in the middle of a centrally-planned church - a Greek Cross, circle or octagon. A comparison of the strategies each pursued to reconcile these competing demands may serve to highlight the trajectories of these two traditions, as well as their unresolved tensions. The early Christian church adopted the rectangular, aisled Roman basilica or law court, with an imperial legate officiating from an apse or exedra at its far end, despite such readily available centrally-planned models as the Pantheon and Temple of Vesta. Liturgical practicality, inevitably, trumped architectural theory, so that Christian churches can appear as concatenations of discordant parts: a narthax, long nave with side chapels and transept, followed by a choir, apse, ladychapel, as well as, baptistry, sacristry, bell tower and spire. This asymmetrical, ”Latin cross" plan only became church doctrine with the Counter Reformation, codified in the encyclicals issued by the Council of Trent (1545-1563, which removed the rood screen or iconostasis, so the congregation in the nave would have an unimpeded view of the hebdomodal thaumaturgy on the elevated altar in front of them, officiated by clergy invested by an uninterrupted, "apostolic succession" from St. Peter and Christ. These prescriptions deformed some of the church's most important monuments, for example, compromising Bramante and Michelangelo's centrally-planned design for the new St. Peter's by tacking-on Maderno's three, perfunctory bays, as an afterthought. Palladio (1508-1580) too was forced to abandon his favored design for a centrally-planned votive church, Il Redentore, an offering for ending the plague of 1576. Here, however, the demands of orthodoxy provoked an even more original response: a dramatic sequence of three, distinct spaces as the climax of an unprecedentedly extended "processional path" - stretching from the monk's choir behind the altar, across the transept, down the nave, out the doors, onto the fondamenta, then over a pontoon bridge of boats from Giudecca to the Zattere. Ironically, as a result of the church's diktats, it was the Vicentine architect's infidel contemporary, the Christian convert to Islam, Mimar Sinan (c.1489-1588,) who achieved the most imposing expression of the Renaissance's ideal of a building with four-way symmetry, his Sehzade or Prince's Mosque in Istanbul (1547-1548.)
Indian and Khmer sthapakas, like their European counterparts, could not fit the necessary liturgical axis within a square mandala or grid; instead, they restricted its four-way symmetry to the garbagriha and shikhara or vimana above it, then added smaller structures to the east, often derived from this shrine or proportional to it, which might therefore be regarded as its “aedicules” – a process which, in deference to Hardy’s more rigorous analysis, might be named “aedicular addition” or “linear expansion,” a kind of architectural polysyndeton. In India, this solution can be traced from the 5m square Vishnu shrine at Deogarh (c.500,) through the seven-column deep porch and mandapa of the Durga temple at Aihole (c.650-700,) to the twenty-three column, hypostyle hall before the Brihadisvara Temple, Tanjavur (1003-1110.) This process often begins, like a paramasayika mandala, with a square sanctuary or cella which then projects four similar, “porch aedicules,” forming a Greek or equal-armed cross whose eastern arm is then elongated by the addition of a vestibule or antarala connected to a mandapa and ardhamandapa or entrance porch.