Penelope Chetwode took artist John Nankivell on one of her tours of discovery 200 miles north of Delhi
Originally published in AR February 1973, this piece was republished online in April 2016
The drawings reproduced here are of temples in the Mandi and Kulu districts of what used to be known as the Punjab Hills but are now part of the State of Himachal Pradesh. In Himalayan terms hills are still hills up to a height of about 12000ft. Alter that, I think it is legitimate to call them mountains.
Broadly speaking, Indian hill architecture is to be found all over the western Siwalik ranges of the sub-Himalaya between 3000 to 10000 ft, where the religion is Hinduism with a strong ‘hill’ flavour. It is characterized by vertical walls built of alternate courses of wood and dry stone surmounted by sloping roofs which help to throw off the monsoon rains. Mountain architecture is to be found in the several provinces of Indian Tibet such as Ladakh, Spiti, Lahul and Upper Bushahr, and in Chinese Tibet in all of which regions the religion is that form of northern Buddhism usually called Lamaism. It is characterised by walls having a marked batter and built of sundried bricks-sometimes of stone but not timber-bonded-and surmounted by flat roofs, for they are outside the influence of the western monsoon.
Img 1 jpg
A.H. Francke in his ‘Antiquities of Indian Tibet’, records that on the frontier between the Tibetan and Kanawari peoples (Dt Rampur Bushahr) he found some interesting intermediate forms of architecture: ’There are little houses built of rubble masonry with flat roofs, for instance the temple at Rogi, and houses bu ilt of sunburnt bricks with location map a slanting roof, for instance the temple at Rarang’. Likewise along these frontiers of Indian Tibet, Hinduism and Buddhism intermingle.
The chief natural glory of the hills is the magnificent Himalayan cedar ‘cedrus deodara’ - which has the great virtue of being insect proof like teak. It grows roughly between 6000 and 10.000ft, and its lifespan is up to a thousand years. According to one of the earliest writers on hill architecture, William Simpson, cedar wood is made to do duty for lime, and literally binds the stone courses together. He also points out that the wood is always laid in a horizontal position, and never perpendicularly or diagonally, as in English timber framing.
‘Under certain circumstances freehand drawing is more effective than the camera in recording architectural detail’
‘In the well-built houses and temples’, he continues, ‘the wood is very carefully arranged, the beams being perhaps a foot or so in depth, extending the whole length of the wall, a beam on the outside and another on the inside, the space between being filled up with stone. The wall at right angles has its beams laid on the two just mentioned; on these again rest the next set of beams of the first mentioned wall, and thus they go on alternately…From this it will be understood that this mass of woodwork is capable of holding together itself, without the stones, which are filled in between to make a solid wall.’
The first man to classify the principal styles of hill architecture was Captain A.F.P. Harcourt in his pioneer work ‘The Himalayan districts of Kooloo, Spiti and Lahool’. In the following captions I have kept to his classifications but reversed his types 3 and 4 because I think the tower and chalet styles belong essentially together. I have also added a fifth type, which we have provisionally named the Sutlej Valley style, because we found it only in Outer Saraj which lies along the right bank of the river Sutlej. On a former trek I also found it on the left bank of this river, and William Simpson’s drawings illustrating his lecture consist entirely of temples in this style, as he travelled extensively in the Sutlej Valley.
Img 2 jpg
With regard to dating, the stone shikhara (type 1) temples of the classical type (of which few remain standing in the area under review) were built roughly between the late Gupta period, seventh to eighth centuries AD - at which time they were introduced from the plains of lndia - down to the end of the Gurjara - Pratihara period in the 11th century. A revival of this style was inaugurated in the 17th century by Raja Jaggat Singh of Kulu, but such temples are dull compared with their richly carved prototypes such as the relatively well-known shrine of Vishveshvara Mahadeva at Bajaura in the Kulu Valley.
‘I soon became painfully aware that the wide overhanging eaves of types 2 to 5 hid the decorative carving beneath them’
As to the indigenous styles peculiar to the hills, these are largely undateable, apart from those built within living memory and still being built today. These comprise types 2 to 5: type 2 being the timber bonded tower or rectangular building used for both sacred and secular purposes; type 3 being the chalet, sometimes built entirely of cedar wood, sometimes having timber bonded cella walls.
Type 4, the so-called ‘Pagoda’ style, is to me the most interesting of all and its origin forms a leading question in art history. It was Fergusson who first noticed the similarity between Himalayan and early Scandinavian architecture, and the mast churches of Norway certainly form a striking parallel with pagoda type temples all over the far east. The art historian who has dealt most exhaustively with the origin of the pagoda style is the late A. H. Longhurst of the Archaeological Survey of India, who traces it to the plains of India in early Buddhist times where he maintains it developed from the superimposed ceremonial umbrellas.
Img 13 jpg
He also thinks that what I call the Sutlej Valley style - type 5 - consisting of a fusion of types 3 and 4, is earlier than the free-standing pagoda. As to the applied arts, wood carving varying greatly in quality decorates these shrines: chiefly round the classical style door frames consisting of several receding jambs and superimposed lintels; round the curious little windows divided into vertical sections by one or more squat pillars; and on the open galleries between the storeys of the pagoda temples. We found several quite distinct styles in the area, ranging from primitive folk patterns in shallow relief, through much more sophisticated figure carving in high relief, up to classical style carving on the sanctuary door frames of two temples in Nirmand. Mention must also be made of the effective decorative device of hanging fringes of wooden pendants along the beams immediately under the eaves and sometimes beneath the verandas of nearly all the hill temples of the Western Himalaya. These are attached by means of hooks and eyes so that they swing and rattle in the wind. Longhurst believes this feature is derived from the tassels which often fringed the ancient royal umbrellas which wer11, and sometimes still are, carried over the images in both Buddhist and Hindu religious processions. Another commentator thinks it may have been suggested by a row of icicles: ‘The dripping over the eaves,’ he asserts, ‘of the thawed snow collected on the lower part of the roof, followed by frost, forms just such a fringe of icicles.’
I have been keenly interested in ‘hill’ architecture for the past eight years and in my efforts to photograph it I soon became painfully aware that the wide overhanging eaves of types 2 to 5 hid the decorative carving beneath them. For this reason I decided that an artist was the answer. In 1968 John Nankivell, then teaching art in Wantage, showed my husband John Betjeman some of his excellent drawings of Victorian architecture, and two years later I invited him to come out to India to draw some of the hill temples. So it came about that we trekked together for some three and a half months during the spring and summer of 1971, in Mandi and Kulu, and John Nankivell was able to do between 40 and 50 drawings, nine of which are reproduced here. In the following pages we do not show any type 1 stone shikhara temples because of their general familiarity. Instead we show examples of types 2 to 5, all in the indigenous ‘hill’ style, and striking examples, of ‘architecture without architects’.
A bhandar in this region denotes a temple store in which are kept the band instruments of a devata (village god), his banners, khuda (a sort of sedan chair carried on long poles), and the several metal masks which are attached to it, and the scarves and other gay materials with which it is draped at festivals. Grain is usually stored in the second storey which used to belong to the god but· since he had his lands removed soon after independance it belongs to the pujari (village priest, not necessrily a Brahmin) who lives in the top storey in a single large room with a hearth in the centre. To let the smoke escape the great tiles are often pushed apart. This building is in the indigenous timber-bonded style, the walls being covered with a light mud wash. On the top storey the cedar heams project on all four sides to carry the gallery which runs right round the building. The huge stone tiles are not slates but pieces of quartzitic rock which glint brilliantly in sunlight. There are locally worked quarries near most villages. The great ridge beam is a deodar trunk. Type 2 style is used not only for temple stores, but occasionally for temples as well, and more often for farm houses, in which case the ground floor is not used for the god’s possessions but by the farmer’s cattle.
Img 4 jpg
CHAINI: Temple of Murlidhar; Tower temple of Shring Rishi
A sacred hamlet three miles above Banjar, the principal village of Inner Saraj. In the foreground is the five-storey temple of 111urlidhar (Krishna the flute player), the actual shrine containing the image being in the courtyard at the back. The tall tower (about 150ft) contains a shrine of Shring Rishi in one of the upper storeys (we were not allowed inside), and a little way away stands a smaller tower which acts as the temple store. All these buildings are in the type 2 style, but it is interesting to note that in the lower storeys of both the tower and the rectangular temple the horizontal timbers are laid at wide intervals and presumably serve to distribute the pressure on the stone walls below them. In the upper storeys on the othe>· hand, the usual interlacing framework is the rule. The great tower obviously served for defensive as well as for sacred purposes, the only access to the upper storeys being by notched tree trunks which could be pulled up when all the villagers were safely inside. the temple of Murlidhar at Chaini, with the tower temple of Shring Rishi beyond We found these towers (but none so tall as that at Chaini) mostly in Inner Saraj, espercially in the wild Sainj Valley.
Img 3 jpg
NITHAR DEHRA BHANDAR
Nithar is the name of a district in the Sutlej V alley with jive or siz hamlets in it, three of which have sacred buildings of outstanding merit. We consider this jour-storeyed bhandar to be one of the most beautiful buildings we saw on our trek. The ground floor is used as a store house for the little type 5 temple of Buhra Mahadev which stands on the opposite Bide of the courtyard. Other buildings in this complez are used as part of the village school, and a pillarboz is attached to the bhandar itself. The pillars on the ground floor Open veranda consist of deodar trunks squared off, with panels of confronting birds ( a common motif in the Western Himalaya) below the bracket capitals on all four sides. The two-storeyed and partly enclosed projecting verandas they support have obviously been renewed in recent years and are probably made of pine, as it is extremely difficult to get a permit to fell cedars today. The pent roof above the upper veranda, and below the principal roof, is an interestwg feature of the building, and between them the great bhunda rope is draped, once used in the Himalayan rope-sliding ceremony. We went on to the first and second storeys by a series of stepladders and found there was no grain stored in them but a series of living rooms interspersed with shrines. Several families have their quarters here. We did not penetrate to the attic storey.
Img 8 jpg
CHUNG: Temple of Devi Chungarsa
The beautifully situated village of Chung (also known as Chong) with its tiny ruined fortress on a rock above the temple, is on the old route up the Parbatti valley which was used by all travellers before the modern bus road was built some 2000ft below it. Two rather bad coloured lithographs of it illustrate Calvert’s book Vazeeri Rupi; the Silver Country of the Vazeers in Kulu (London 1873), and the ‘quaint carving’ on the temple is referred to in two other books. But no art historian, so far as I am aware, has so jar recorded it, yet it is one of the most interesting of the many different chalet type temples we saw. It is raised on a dry stone plinth with a steep flight of steps leading up to the door. The temple measures 24ft 2in across the front and 26ft 3in along the sides. It has a veranda along the facade only, enclosed by a screen which is pierced by four curious jagged-shaped windows (not regular Moghul cusped arches as are found in many enclosed verandas on both sacred and secular buildings in the hills). In between and below these windows are folk style decorative carvings in low relief of hookah smokers, confronting peacocks drinking out of a tall vase, and various floral and abstract patterns.
The magnificent screen, the only one of its kind which we have so far come across, divides the interior of the temple in half and acts as an iconostasis with a double door in the centre. The carving on it is far superior to that on the exterior of the building and probably date from the 18th century as the dress in most of the figures is Moghul which became very fashionable in court circles in the hills from the late 17th century on, when the local families of miniature painters started to be influenced by Moghul painting. The arch over the door is formed of elephant trunks as in Gaja Lakshmi images, and in the spandrels above them are the goddess’ lions, as Chungarsa Devi is simply a local name for Durga who always rides a lion into her battles with the demons. Reef knots figure repeatedly in the decorative scheme and may be a reminder of the Celtic origin of some of the tribes who settled in the Westem Himalaya and in Rajasthan (where this motif is also found) after the collapse of the Gupta empire in the sixth century AD. A wooden fringe a foot deep hangs from the projecting beam along the top of the screen, but the artist has deliberately left this out in order to show the carving behind it.
Img 5 jpg
Img 7 jpg
KHOKHAN: Temple of Brahma
This jour-tiered pagoda temple stands in the large village of Khokhan only one and a half miles west of Bhuntar in the lower Kulu valley. Considering its proximity to the main road it is curious that it has not attracted the attention of previous writers on Himalayan architecture. Harcourt records only jour free standing pagoda type temples in the area of which Khokhan is not one. Vogel and Longhurst stick to this number. We actually found 13: nine old ones and four modern, the latest, at Hurla, having been built in 1970. It is a well proportioned three-tiered pagoda painted bright blue and white.
The Khokhan temple must be somewhere between 70 and 80ft high, rather irregularly proportioned with an extra wide bottom tier. An interesting feature of the structure is that the top roof is rectangular instead of circular as in all the other examples we recorded. A heavy beam bearing three wooden finials runs along the top ridge as in type 2 and 3 buildings. Carved galleries separate the three upper storeys with elaborately curved struts springing from the backs of birds to support the successive pent roofs. Another feature is the stylised horses’ heads at the bottom of each corner of the second and third roofs which each have a projecting stone tile to protect them from the weather. The decorative carving on the lower storey is in low relief in fascinating primitive designs in what we named the Picasso style.
Img 11 jpg
PARASHAR: Temple of Parashar Rishi
This curiously attenuated pagoda type temple is situated in the Mandi district in a basin of green hills beside a small lake with a floating island on it, at a height of some 9000ft. To the south, valleys of bright red rhododendrons run down to the Beas river some 20 miles away, to the north there is a magnificent view of the 20 000ft snow peaks of Lahul and Spiti, about 50 miles away. It is obviously ancient in origin and the carving on the body of the temple, of intertwined snakes, bird, pot and foliage panels, gods and goddesses, heraldic beasts, and a variety of abstract patterns, is of such excellence that I would be inclined to put it earlier than that on the famous temple of Hidimba at Manali, which is dated 1553, and on which the decorative carving on the facade nowhere near equals the quality of that at Parashar. The sanctuary door frame is the most elaborate I have so far seen in the hills and as beautiful as any I have seen in the classical style of the plains: a supreme work of art. It is composed of seven doorjambs, each one receding a little to the actual door, and seven superimposed lintels. The outer doorjambs and the upper lintel are decorated with fat scaly intertwined serpents in one continuous design.
The successive roofs are covered with slates from the Mandi slate quarries. In the temple of Parashar Rishi is a panel 16in high of a four-armed female deity in a long pleated skirt wearing a mala (garland). She is standing on what I take to be a local version of a makara, a hybrid monster, usually resembling a crocodile, and very common in classical Indian art. She appears to be holding its tongue in one of her right hands and its tail in one of her left. Her two remaining arms hold up a sickle and a purna-kumbha (water pot). The latter is an attribute of the river goddesses Ganga and Yamuna and the fact that she is standing on a makara indicates that she is Ganga. She is not, however, in the usual position of a river goddess on a temple which is on either side of the cella door frame. At Parashar she is on the outside left wall of the shrine and it is the first time I have seen her depicted with four arms. For this reason she may represent Durga standing on the buffalo demon. Hindu iconography gets very mixed-up in the hills.
Img 10 jpg
Img 9 jpg
NITHAR DHANAH: Temple of Dhaneswari Devi (Durga)
This small village is one of several Brahmin pockets in Outer Saraj where the inhabitants are exclusively Brahmin and the indigenous population (known as Kholis) are relegated to a suburb a little way outside. It is situated in the Sutlej Valley some 2000 feet above the river and the exquisite little Durga temple is in what I call - at least provisionally - the Sutlej Valley style, being a fusion of types three and four, the chalet and the pagoda: the former acting as the mandapa (hall of a classical plains type temple), and the latter as the Shikhara (the tower above the sanctuary). The single circular pagoda roof is raised on a gallery decorated with the familiar confronting bird motif. Magnificent birds are also carved below the capitals on the square pillars of the veranda which runs right round the rectangular temple. Here they are not confronting but look round on the sides of the pillars to form one head on the corners. The walls of the temple are timber-bonded with three foot decorative cedar wood panels let into them at intervals of such beauty that they really (Trade up out of folk into fine art. The subjects are the major gods of the Hindu pantheon: Siva and Parvati on Nandi, Vishnu and Lakshmi on Garuda, and several versions of Durga killing the buffalo demon.
Img 12 jpg