High above the mighty Indus, on hills commonly called the Salt Range, stand important remains of forts with citadels and temples (Fig. 1). Built from the 6th to the 11th centuries AD, these structures lie in what was ancient India’s far northwest (Fig. 3), now in the Panjab and North West Frontier provinces of Pakistan. Largely ignored by scholars in this century, and orphaned from the main stream of architectural scholarship since the partition of South Asia in 1947, these remains form an important link in the history of South Asian architecture. Remarkably, this region preserves an almost continuous record of temples that can define the evolution of a distinctive school of Gandhâra Nâgara architecture. An integrated archaeological study of these sites, undertaken by the author with colleagues in Peshawar, has begun to reveal new aspects of this important period of South Asia’s antiquity. What follows is a preliminary report and stylistic analysis of the region’s temples.
Archaeologically, the area is best known for the massive numbers of Buddhist sculptural and structural remains associated with the region of Gandhara from the 1st century BC to the 5th century AD. These Gandharan remains already show a local visual vocabulary in which architectural traditions from India, Central Asia, and the classical world appear side by side. This mélange of traditions is evident on many Gandhâran Buddhist narrative steles, as well as monuments such as the famous shrine of the double-headed eagle and the Dharmarajika stupa at Taxila (Fig. 2).
The Chinese pilgrim, Hsüan Tsang, visiting Gandhara in the 7th century AD, noted hundreds of Hindu structures along with many Buddhist sites then in decline (Batters 1904-05). If there is a Gandhtran legacy in the Hindu temple architecture of subsequent centuries, it takes two paths: one, a unique tradition of temples with pyramidal roofs built in Kashmir from before the reign of Lalitaditya in the 9th century AD (Fig. 5), the other an independent tradition in Gandhara itself. Our project focuses on the consequences of this second tradition.
We find perhaps the earliest example of the Kashmir tradition in two small 6th-century (or earlier) temples at Liduv (Meister et al. 1986 [hereafter ETA]: 361-63) and of temples related to the second tradition in several 6th-century masonry sub-shrines at the Hindu pilgrimage site of Katas in the Salt Range (Fig. 6). The square Lãduv shrine has a circular interior space and had a hemispherical dome under a peaked roof, for which a Gandhiran prototype—a masonry structure at Guniar in Swat—is sometimes cited (Kak 1933:55-56; EITA:362). The whole was once covered by a pyramidal roof, as indicated by the frame surrounding its doorway. Gandhiran antecedents for this type can be seen in the “classical” niche pediments represented on the 1st-century-BC shrine of the double-headed eagle at Taxila, or the split pyramidal pediments in Gandhara sculpture and on stupas such as that shown in Figure 2. This distinctive gabled pent roof became the signature for Lalitaditya’s powerful Kashmir dynasty in the 9th century. Bell-preserved examples, from the 6th to 10th centuries, survive on temples at Narastan, Pandrethan (Fig. 5), and Payar.
The type of temple found at Batas, while sharing with Laduv the formula of a simple square plan, plain masonry walls, and cantoned corner pilasters, formed its superstructure by quite different means. The Katãs sub-shrines elevation can be reconstructed as a series of cornices with tiny intermediate rows of pillars and a crowning ribbed stone (malaka) (Fig. 6). This early type of simply storied structure has parallels in coastal western India at Sarnath (Saurashtra) and elsewhere across northern India and the Deccan in the 6th century AD (see Meister 4966; ETA).
With its representation of many multiple stories, the Katäs sub-shrine can be considered a type of proto-Nagara tower. However, local experimentation with the full Nägara formula—the typical curved temple form of northern India—had already begun at Kafirkot (“foreigners’ fortress” in local parlance) west of the Indus in the North West Frontier Province (see Figs. 7, 8). The two earliest temples in this fort can most closely be related to early Garulaka or Maitraka dynasty temples in Saurashtra at sites like Bhanasarã and Dhank, from the 6th and early 7th centuries AD, and Saindhava dynasty temples from the same region in the 6th century (see plates). Even the name of the little understood Saindhava dynasty seems to indicate a link with the Indus (Sindhu is an ancient name for the river).
Stylistic Sources for the Salt Range Temples
Scholars have tended to date this whole ‘group of temples now in Pakistan to “post Islamic contacts,” that is, after the 7th to 6th century AD, because of their use of mortar, rubble-fill between masonry walls, arches, and squinched interior domes (Archaeological Survey of India Annual Report 1920-21:6-7). They have also tended to locate them as a branch of Kashmiri architecture, because of one aberrant temple (Fig. 4). Both Percy Brown (1942) and James Harlem (1986), for example, in their volumes on Indian architecture, place the Salt Range temples in chapters on the Kashmir tradition.
Nineteenth and early 20th century scholars, including Aurel Stein (1937), Alexander Cunningham (1872-73), and Ananda Coomaraswamy (1927), focused their attention on the 10th-century temple at Malot in the Salt Range (Fig. 4) and its formal links to the architecture of Kashmir, thus setting the direction for later scholarship. The temple at Malot does indeed mimic pent-roofed temples in Kashmir at a time of marital alliance between the Utpalas of Kashmir and the Hindu Shhi kings of Hund in Gandhara (Rehman 1979). It differs from the Kashmir temples, however, in placing the curvilinear Niagara shrine models on its walls (see box on Shrine Models). These shrine models mimic local Gandhâra-Nagara temples at other 10th-century Hindu Shhi sites, such as a pair of temples in a second important fortress, Bigot (south Kfirkot), near Dera Ismail Khan (Figs. 1, 17).
The Kashmiri form found at Malot, however, is an exception. Better sources for this Indus group of temples can be found in the Gandhâra substrata and in the ferment of Nagara formation in other areas of north and western India (Meister 1981) than in Kashmir. Whether in the domed Buddhist compounds at Takht-ibahi or the 5th-century moldings facing the Dharma-rajika stupa at Taxila (Fig. 2), Gandhara antecedents are close at hand. Certainly the basic molding sequence of Gandhara-Nagara temples begins as early as Taxied. The typical slender pseudo-Corinthian pilasters at Kfirkot (Fig. 10)—as well as true arches—can be seen also on the 10th-century Buddhist stupas at Guldhara in Afghanistan (Harle 1986:73). The characteristic sloping batter of niches and doorways (and sometimes walls) on these temples has clear antecedents in Gandhäran conventions. Much of the architectural ornament in these temples is familiar to the Gandhara region and even the use of interior squinches and masonry domes is not new.
What is new to the region is the Nãgara modality of superstructure as it had developed in north India for the first time in the 5th and 6th centuries AD (Meister 1986, 1989). The shrine model on the wall of temple D at Bilot (Fig. A in box on Shrine Models) bears close resemblance to the much better known proto-Nagara shrine model represented on the early 6th-century doorway to the “Gupta” temple at Deogarh in central India, for example, or one on a brick stupa base at Nalanda in eastern India (Meister 1986:46-47).
A Walk Through the Salt Range Temples
To frame this local and continuous craft tradition of the Salt Range and upper Indus, let me briefly review the remains in chronological order. At north Kfirkot (Fig. 7), temples 13 and A represent the earliest experiments in this region with the developing Nãgara formula (Fig. 8). At Bilot (south Kfirkot) the much larger temple D awkwardly formulates a Nãgara tower on a square base, much like the pre-Nãgara temple at Bilesvara in Saurashtra in the 7th century (EI TA 1986: 181-84), and incorporates a model of a proto-Nagara shrine on its walls (Fig. A in box on Shrine Models). Late in the 7th century, temple C at Ktfirkot (Fig. 10) and temple A at Bilot (Fig. 11), both with damaged Niagara towers, project one central offset on each wall and modulate ornamental elements of their superstructures in a more integrated way compared to Bilot temple D (Fig. 9). These temples display a new confidence in and knowledge of Nagara formulas. Temple C tentatively introduces for the first time a version of north India’s common vase-and-foliage capital for its corner pilasters, while retaining the local neo-Corinthian type for the central offset.
Two striking temples, located on hills east of the Indus opposite Kalabagh at Mari-Indus—which I would date in the 8th century—continue and refine this local Niagara tradition, but still with only a single central offset on their walls (Fig. 12). Temple A places thin pilasters on the corners of each offset, while temple B pairs pilasters for the first time on its corner buttresses (Fig. 13). Temples in this sequence in turn seem to provide a central shrine model on each wall to represent a slightly earlier local experiment with the formula for a Niagara temple (see box on Shrine Models). Each also seems to carry forward some architectural element, as in the trefoil arched niches at Bilot (Fig. 9), the trefoil doorway at Mari-Indus, and the five-cusped entry to the smaller 9th-century temple at Amb (Fig. 14).
The first temple in this tradition that can have its date confirmed by any evidence other than style and decorative context is the elegant fired brick structure at Kallar (Fig. 15). Its walls of five offsets (a central one with two on each side), and its developed ornamentation with vase-and-foliage pilasters and other distinctive details, place it parallel to temples in central and western India from late in the 8th and early in the 9th century D. This date is supported by a single coin found near the foundations struck in the reign of the first Hindu Shahi. ruler, Kalar, whose dynasty has recently been dated by an inscription as beginning in An 821 (Lehman 1993:31). Only further archaeological explorations, however, and perhaps carbon-14 dating of wood beams used to support the interior domes of some of these temples, can fix more firmly the dates and historical frame suggested here.
Early in the 8th century, perhaps, sub-shrines were added above the eastern corners of the platform supporting temple D at Blot. These echo but reorient two domed cells sunk into the front corners of the temple’s platform (Fig. 16). The small temple D at Kafirkot, built near the north gateway to that fort late in the 9th century, mimics some distinctive details of these sub-shrines.
A Distinctive New Turn
In the 10th century, larger temples were built under the patronage of the Hindu Shah’ kings in the spectacular fortress at Amb (Fig. 18), at Bilot (Figs. 1, 17), and at Nandana (Fig. 19) on the eastern escarpment of the Salt Range. Like earlier ones of the region, these still were Latina temples (that is, they had single curvilinear spires), but within their walls were stairways leading to an upper story where an interior ambulatory corridor surrounded an upper chamber embedded within the tower (Fig. 20a). In this respect they are unlike all other Niagara temples in India.
This remarkable regional experiment with multiple levels, folded within a Latina tower (Figs. 17, 19-20), came to an end early in the 10th century. At that time the great fortress at Nandand on the eastern flank of the Salt Range fell to Muhammad of Ghazni, who sought to control the significant routes across the Panjab leading toward Multan and Delhi. The Hindu Shahi kings then took refuge with their cousins in Kashmir. In this sequence of Salt Range temples, only the last one, built at Nandana, suggests corner turrets on its single-spired tower (Figs. 19, 20a). These turrets remind us, however, of the multi-spired Niagara shrine models represented on the walls of the 10th-century Kashmir-related temple at Malt (Fig. 4), even as they reflect a multi-spired convention that became common in central and western India by the 9th/10th century (Bin).
Across northern India, this multi-spired (sekhari) temple type sets a new standard in the 11th century at such famous sites as Khajuraho in Madhya Pradesh, but its origins lie in experiments carried out in western India (Gujarat and Rajasthan) in the century before—experiments marked and reflected in these late Shahi temples in the Panjab.
That these forts and temples survive along the Indus must be a reminder to us of how untouched many of India’s traditions are; of how severely partition has truncated our understanding of South Asia’s multiple civilizations, both Islamic and Hindu; and of our task as scholars to mend that historical wound, even as we have begun to reproblematize colonial scholarship and its assumptions.
I end this preliminary report with a footnote to demonstrate the mighty weight of finding a new monument in the field. At the site of Mani, in addition to the two 9th-century temples already discussed (Fig. 12), there are also two mounds higher up the hill to the west, badly ravaged by treasure hunters, that past reports have labeled primarily as places of residence (Cunningham 1679; Mumtaz and Siddiq-a-Akbar 1989). These in fact are ruins of two large temples placed on high platforms. One, Temple C, still preserves remains of an inner sanctum and an enclosing ambulatory wall. On the north side, this wall preserves a central niche with a distinctive “Kashmiri-style” pent roof (Fig. E in box on Shrine Models), but the shattered remains of the temple’s superstructure suggest instead a complex multi-spired tower with curvilinear Latina spirelets. This temple seems, in fact, to have been almost a reverse response to the unique local experiment with Kashmiri style found at Malt (Fig. 4), and an answer to it. Let scholars beware.
Shrine Models as Signatures of Architectural Experimentation
The architects of these temples in the Salt Range and along the Indus knew that they were working within a variety of options. Architecture could engage their creativity, and through their creative actions, temples could evoke in multiple ways. They seem consciously to have left a record of their architectural experiments by placing shrine models as niches on the walls of many temples. These often seem to represent slightly earlier local experiments with the formula for a Nagara temple, focusing on the nature of the temple’s superstructure. Temple D at Bilot, for example, uses a protoNagara model (Fig. A). Temple B at Mtri, on the other hand, uses curvilinear Nagara models with ornamentation placed across single cornice layers (Fig. B). In this respect the models at Mari resemble the superstructure actually built in the 7th century for Bilot’s temple D rather than either superstructure built at Mari in the 8th century for temples A and B (Figs. 12, 13).
On the 10th-century temple at Malt, the central shrine models have developed curvilinear Nagara towers flanked by extra turrets (Fig. D). Maori’s remark able temple C, on the other hand, had central niches marked by a split pent-roof pediment framing a trefoil arch (Fig. E.) that suggests the gabled pent roof that once actually crowned the temple at Malot (Fig. 4). The trefoil-arch pattern can be seen at Bilot, Mari, Alb, and Malot in association with either pent-roofed or curvilinear formulas (Figs. A, C-E).
Marking temple walls with images of past architecture provides an historical locus for architects working within a system of meaning which sees each niche as an expansion of the temple as a whole (Meister 1993). The rhetoric of architectural representation in South Asia more often relates to an ahistorical rather than historical reality, yet from timed to time the two overlap (Dhaky 1977). In Gandhara sculpture, for example, the variety of recognizable Buddha types seems sometimes to point to specific places of pilgrimage. So also in the Salt Range, architectural experimentation gave contemporary expression to how the minds of its architects worked as well as providing a model of God’s creation.