Pahari Harihara or Harahari: Nuanced Iconography, in Indian Painting. Themes, Histories, Interpretations | Essays in Honour of B. N. Goswamy. Edited by Mahesh Sharma and Padma Kaimal,

286 Pahari Harihara or Harahari

Pahari Harihara or HarahariNuanced Iconography

Mahesh Sharma

Early western Himalayan states were largely influenced by the political and religious developments in Kashmir. As a result, the rise of Saivism in Kashmir witnessed a corresponding development in the hills; later on, the rise of Vaisnavism in Kashmir was also reflected in the Chamba temples and epigraphs. The shift is evident in contemporary inscriptions, particularly in Chamba, where the rulers called themselves “Paramavaisnava” or “Paramamahesvara”, as the case might be (Sharma, Mahesh 2009: ch. 1). That the iconography of images housed in temples might also be so influenced, particularly by texts like the Visnudharmottara Purana (VDP hereafter), which is of Kashmiri provenance, is not surprising. This is evident in early images, such as Tryambakesana Rudra, or Narasimha, and later, Vaikuntha.

Shifts in sectarian ideologies and accompanying symbols, indicative of the changing sectarian affiliations of the state, were sometimes accompanied by discord and sporadic violence. Rajatarangini, for instance, castigates the state for plundering the “treasures bestowed to the temples by former kings” and recounts “divine images” (devotpatananayaka) were desecrated (Sharma, Mahesh 2002: 49–66; Eaton: 104–14). Moreover, sectarian identities were further sharpened by the rise of monastic-organisations, which competed for state patronage, occasionally leading to hostilities, as noticed by Lorenzen (1978), Kolf (1990) and Pinch (2006). The sectarian antagonism, however, was not unmitigated. The local community and the state sought solutions, particularly in the domain of symbolic syncretism. One such experiment was the reconciliation of the Vaisnava and Saiva ethos in the symbolic amalgam known as Hari-Hara. An example is provided by the 1223 CE Davangere inscription, where apparently a conscious effort had gone into fusing the two antithetical symbols of Hari-Visnu and Hara-Siva into one anthropo-morphic form, namely Harihara. In order to rise above sectarian differences, this symbol was projected as the guiding principal of the universe, as a Supreme God, all other being but associations or projections of this: “Some saying there is no god on earth [but] Hari, and some saying there is no god on earth but Hara, to remove their doubts Hari-Hara murti-idol was revealed in a single form; which form of glory may it protect us.” The inscription goes on to declare that the Hari-Hara Ksetra (sacred area) was greater on earth than Setu (Ramesvaram), Varanasi or Kuruksetra—all major centres of pilgrimage in the “Hindu” tradition (Mysore Inscriptions: 20).

Mahesh Sharma 287

If the inscriptions took pains to create sectarian harmony, the medieval religious literature was equally concerned with promoting reconciliation. The point in case is the Ramacaritamanasa of Tulsidasa—the most popular of all religious texts, particularly in north India. Sensing the need to resolve sectarian strife, the author uses a syncretic strategy: the text is a transmission of the grand story of Rama narrated by Siva to Parvati. Moreover, even as the text is primarily Vaisnava, extolling the exploits of the Rama incarnation of Visnu in popular language and idiom, Tulsidasa cautions the readers and sectarian followers that (Lankakanda: 1.6b–8):

…/None is dear to me as is Siva.The one who hates Siva to become my follower/ such a person cannot achieve me even in dreams.The one who worships me at the cost of Sankara/ he is surely short-sighted and ill-witted.

Similarly, the Lankakanda (1.1) begins by the proclamation of the author that he adores Rama, the supreme deity, who is worshiped even by Siva; and continues, that he strives to glorify Sankara, who is the lord of Kasi (2.1), and the allayer of sins in the Kali age (2.3). It is, therefore, not without significance that the associated pilgrimage centre reflecting the syncretic ethos is Ramesvaram (i.e. the place of Rama and Siva). The significance of the fused symbol is best echoed in Balakanda (104.1a): “There is no order without Hari-Hara/….” Hari-Hara, however, never became a prominent deity in north India. While there are some examples of Harihara images in temple niches, there are no prominent temples dedicated to this deity except in Osian (Rajasthan) and Davangere (Karnataka). It is interesting that Harihara gained prominence in southeast Asia, particularly in the Pre-Angkorian Khmer civilisation and the Devaraja kings (Lavy 2003: 21–39); and to an extent in south India, primarily as Ayyapan—the son of Hara and Mohini or Visnu (therefore, Harihara). In fact, most known exegetical texts to Harihara are in Telugu, and are of later origin. While there are some early medieval to pre-fifteenth century sculptures in south Asia, there are some late (seventeenth century onwards) painted versions also: from Travancore, Kalighat, and significantly, from the western Himalayas. In this paper we shall discuss some Pahari Harihara paintings from Mankot, Chamba and Mandi to point out how differing political and religious affiliations and considerations subtly influenced the painted variations.

IOne of the earliest examples of Harihara in the western Himalayas is the stunning sculpted image located in the northwestern corner of the 1240 CE Baijnath Vaidyanath Temple in Kangra District of Himachal Pradesh. Hari-Hara is seated in pralambapadasana on a throne. He holds a sceptre in his lower hands, resting on his thighs—perhaps indicative of the divine king. The meaning of the sceptre is ambiguous. Depending upon which half of the fused deity holding the object is emphasised, it may represent either the staff of Lakulisa or the mace of Visnu. Most likely, a combined implement is intended. The fusion is brought out dexterously. The rod held in the hand becomes a small

Fig. 25.1 Baijnath Harihara, 1240 CE. Northwestern corner, Baijnath Temple

288 Pahari Harihara or Harahari

Mahesh Sharma 289

mace to the left-proper; while it stays straightened to the right. Ornamented, the magnificence of Visnu is represented by the dangling kaustubha jewel; while, the eternal transcen- dence of time is represented by the third eye of Siva. The crown similarly is vertically split—the matted locks, or jatamukuta, to the figure’s proper right, and Visnu’s crown or kiritamakuta to the left. In the raised right and left hands are the trident and disc respectively. On each flank of the throne are the spouses, seated on Nandi and Garuda—in anjalimudra—carrying fruit and flower. However, the image is distinct from other Harihara examples. It is a three-headed Vaikuntha-Sadasiva (fig. 25.1). To the left side, one can see the Varaha head facing the disc of Visnu; while on the right side is Bhairava, facing the trident.

The decision to split the figure into right and left halves may have been inspired by the iconography used for androgynous icons like Ardhanarisvara and Visnu-Kamalaja. Even the texts explicate that the Haryarddhamurti (Half of Hari) or Harihara was fashioned on the principle of anydrogynous images (Gopinatha Rao: II.1: 332–37). The text, it seems, is aptly followed by the architect of the Baijnath temple. Opposite the Harihara, to the southwest corner is the eight-handed Visnu-Kamalaja, riding atop Garuda (fig. 25.2). The four-handed Visnu is to the normal right, holding from top to bottom: the disc, conch, mace and lotus-bud; while Laksmi holds a shield, water-pot, lotus flower and a pestle. The comparison between the above two images would suggest that the goddess half was substituted by Visnu, to create a syncretic composite anthropomorphic form of Hari-Hara. The juxtaposition of the two examples would suggest that Visnu is considered the creative principle or the prakriti tattva while Siva is the inert purusa principle. It is Hara, in the timeless eternal state of inertia, the formless and expressionless. The kriya sakti or energizer ‘I’, the dormant embodiment of which makes Hara the Har-i, embodies the creative principle of form, vision and movement. The conjoined image may represent the concept of purusaprakriti, where prakriti embodies the purusa principle and expresses in the worldly efflorescence. It is in this sense that the Vaman Purana says that within Hari is Hara. Theoretically, it is also an attempt to harmonise the radicalism of Tantrantra and Vaikhanasa schools, popular by the turn of the first millennium CE in Kashmir.

It seems that this left-right division was not a comfortable one from the Saiva–Vaisnava sectarian positions. The left side or vama-bhaga was considered the weaker, emotive and creative; it was associated with the feminine. The right or daksina-bhaga was the ritually pure and dominant side. The left–right division is also central to divine ranking, whereby the left is considered subservient to, or lower in hierarchy to the right. The division based on two equal halves therefore belies the tacit sectarian compromise that created a composite Hari-Hara form. The politics of appropriating the right side of the body is a pointer to this discomfiture. The early eleventh century Tantrasara of Abhinavagupta in fact calls the left side as the “Vaisnavadha” or the Vaisnava half. Abhinavagupta, however, cautions that sarvajnana or complete knowledge (comprehension of reality) can only be accomplished by cultivating the paramatattvam or the dominant elements of the other half (Tantrasara 22 vs 4). Both the above sculptural examples, coming from a Saiva temple, are therefore purposeful. That Visnu-Kamalaja is subtly used to establish the principle of vama-bhaga or prakriti tattva, and thus to explicate the Hari-Hara image, displays an astute understanding of iconographical correspondences. The same principle was utilized in formulating the genesis of Ayyapa, the son of Hara and Mohini, who is a feminine form of Visnu. It establishes the dominance of Siva as the cosmic principle, even if the embodiment of this is possible only through the creative force of Hari. That this was a deliberate, if subtle, manipulation is evident from the various names and iconographic switches in the texts as well as images. Thus, the Haryarddha of Uttarakamikagama (Gopinatha Rao: II. 2: 168) is Hari-Sankara of Agni Purana and Vamana Purana; Harihara is the Hara-Hari of Tantrasara and Siva-Narayana of Matsya Purana (Chandra, L. 1988: I: 274). There are other combinations. For instance, VDP describes the left half of four-armed Harihara as Hrisikesa, which is dark, holding cakra and abja or lotus, riding on Garuda; the right half being white-coloured Sadasiva riding a vrisabha or bull, holding trident and in varadamudra (Rao II.2: 171). The iconographic variations across texts and images range from two to four arms; and one, three or four heads. Implements held by the composite god are not limited to the standard trident, skull-cup, hourglass-drum or damaru, staff; or conch, mace, lotus and disc; but may sometimes include parasu (axe), rosary, shield, kamandalu (waterpot), serpent and antelope (Chandra, L. 1988: I: ch. 10: 274–81). facing page Fig. 25.2 Visnu-Kamalaja, 1240 CE. Southwest corner,

Baijnath Temple

290 Pahari Harihara or Harahari

IIThe iconographical puzzle becomes even more interesting when painted versions are compared. We may first refer to the c. 1720–40 opaque watercolour painting of Harihara on paper (fig. 25.3). This painting—perhaps from Mankot and in the collection of Doris Wiener Gallery, New York—strictly follows the iconographic prescriptions of VDP. Harihara stands atop the cosmic ocean, depicted by five horizontal strokes at the bottom. Strictly following the text, Harihara stands straight, without any bends in the legs (Gopinatha Rao: II.1: 334), and is divided into two equal halves: Hari to the normal left and Hara to the right; Hari is dark blue, the colour of the rain-bearing cloud; while Hara is white. Hari wears a yellow ankle-length dhoti and sports a pitavastra; Hara is draped in a leopard-skin (krittivasan, tiger-skin), the head and paws of which extend to the deity’s feet. Hari wears a pearl necklace, which turns into rudraksa-necklace on the Hara half. Similarly, the pendant kaustubha of Hari is balanced by the hooded serpent of Hara. The head is also vertically divided: the diademed crown of Hari sporting three glowing flowers; and the matted hair of Hara that is neatly tied in a bun (jatamukata).The single earring similarly is divided: the ornate half towards Hari and the plain half on Hara’s side. The face of Hari and Hara are both shown in profile, creating two neat divisions, identified by the typical tilaka marks: the vertical pundarika denoting Hari and three horizontal lines indicating Hara. The tilaka marks are carried over to arms and navel, as if the image has been worshipped and invoked appropriately (nyasa). Finally, it is a four-armed image, with arms separated from the elbows, rather than the shoulders. While Hari holds a yellow coloured mace and lotus; Hara is shown with his usual emblems, the hourglass-drum or damaru (sketched in its bare outlines) and a long white-coloured trident. The length of the trident rather counterbalances the tips of pitavastra and mace—aesthetically balancing the right and left sides. The colour balance of the image rests on the play of dark-blue and white. We may note that while the colour dark-blue absorbs all hues into its body; the white reflects. The dark-blue takes away the property of all colours; the white reflects the quality of colours. Symbolically, Hari (Vasudeva) is dark-blue—embodying the ether, the formless pervasive spatial substance and is, therefore, symbolic of creative energy. Hara is white, the colour of cohesion and the quality of satva, the truth. The dark-blue is symbolic of creative element, the prakrit; the white is the face of annihilation, the destructive energy associated with Hara. The two—sitting at the opposite end of the colour-spectrum—combine to create a cycle of life and death, the principle of creation and

above Fig. 25.3 Standing Harihara from Mankot. Opaque watercolour on paper, 1720–40. Courtesy: Sotheby’s 2012 (Lot 232)

facing page Fig. 25.4 Seated Harihara on Lotus from Chamba. Gouache on paper, c. 1750. Courtesy: Chandigarh Museum, Acc. no. 5022–B (1)

Mahesh Sharma 291

292 Pahari Harihara or Harahari

destruction, the attributes of the Supreme God. The dark colour is the symbol of prakriti, while white is symbolic of vikriti. As has been clarified by VDP, vikriti is embodied by prakriti (III. ch. 47. vs 1–18); hence Hara is embodied by Hari. While this embodiment comes across in most painted images of Harihara or Harahari (except that the colour blue becomes sky-coloured), the left–right switch in some is significant. In the sitting Harahari from Chandigarh Museum, painted perhaps in Chamba in the 1750s, the dark blue-coloured Hari in profile is to the normal right, while the skin (off-white) coloured Hara in three-quarters is to the left (fig. 25.4). Vertically split into two halves, Harahari is seated on a double-petaled lotus, perhaps in padmasana. The lotus is placed on a gem-studded octagonal golden throne, the four legs of which are all visible. Hari is draped in a tawny or brownish-yellow dhoti, which mingles with the dull cinnabar

dhoti of Hara. Hara is also draped in the tiger-skin around his waist and thighs. Hara is wearing the mundamala which is joined with Hari’s vanamala. The kaustubha and other necklaces in the right Hari half are counterbalanced by the hooded serpent to the left. The dark red urdhava-pundra tilaka is prominent on the dark-blue forehead of the right-hand figure; Hari also wears a diademed tiara with five-plumed mayurapankha or peacock’s feathers. In contrast, Hara is depicted in three-quarter view, perhaps to showcase the third-eye of Tryambakama, highlighted further by tripundra smears of vermilion and a barely visible crescent moon. The matted locks of Hara are neatly tied into a knot. The respective tilakas are also etched on the neck, arms and abdomen of their bodies. The Chamba Harahari is four-handed, carrying disc and conch in the proper right hands and trident with damaru attached to its rod and skull-cup or kapala in the proper left hands. The left-right reversal in this image may be usefully compared to similar switches in paintings from Mandi. One such painting, published by Glynn (1983: 60, fig. 31), playfully swaps the attributes in a manner that suggests some ambivalence about how best to portray Harihara/Harahari. In this painting, Siva-Natesvara and Krisna exchange their ayaudhas, cognizance objects (fig. 25.5). Siva takes Krisna’s conch in his right hand and staff (as of gopa, cowherd) in his left and wears Krisna’s yellow dhoti; while Krisna takes the speared-trident and damaru and is draped in leopard skin. Furthermore, Krisna is shown sporting a serpent; while Siva wears the kaustubha and vanamala. Yet, the dancing Krisna is blue-hued and wears the mayura mukuta and is marked with the urdhava-pundra tilaka. One can see a tuft of hair below his mukuta. The white faced Siva, on the other hand, is painted in three-quarters and his matted locks are tied neatly in a bun. On his forehead rests the crescent moon and third-eye, along with three horizontal lines of his tilaka. Both the deities dance, playfully looking at each other, while they are watched approvingly by the four-faced bearded Brahma, sage Agastya, and others. Likewise, Garuda and Nandi have switched positions, their gaze resting on their masters. There are reverberations of painted Nata-Narayana from the Kangra Ragamala in this dance. In the Nata-Narayana, standing on a lotus, the normal left half exhibits the iconographic features of Visnu and the right, those of Siva (Randhawa 1958: 81, fig. 72). The Mandi painting presents profound philosophical ideas in lighthearted form. Krisna and Natesa look playfully at each other as they dance; the suggestion is that the two are identical. Their positions can therefore be switched. Left or

Fig. 25.5 Dancing Krisna and Natesvara in Switched Roles. Unknown private collection. Courtesy: Artibus Asie. Black and White reproduction. Published by Glynn (1983: 60, fig. 31)

Mahesh Sharma 293

right sides in dance become meaningless, as they keep changing in a flow, merging into each other, as it were. Despite this indication of identity at a deeper level, however, the painter subtly conveys sectarian dominance by using the scale of hierarchy. Siva is tall and elegant (even though he acts as Krisna); while Krisna is scrawny and diminutive. A similar subtle assertion of dominance is perceptible in the final two examples of c. 1710–30, largely identical paintings of Sadasiva-Narayana from Mandi. The first (Kramrisch 1981: 168, pl. 8), from the Howard Hodgkin collection—now displayed in the Ashmolean Museum, University of Oxford—is an image of cross-legged, seated ten-armed and five-headed Sadasiva, commissioned by Raja Sidh Sen of Mandi (r. 1684–1727). Draped in a dhoti of leopard-skin— the paws of which hang near the feet of Sadasiva, giving an illusion of four feet, while its tail becomes the waistband—the hirsute potbellied torso of Sadasiva is bare. Sadasiva is in natural skin colour, and his matted locks hang over his puffed-up chest. The figure wears a serpent knotted head to tail, a rudraksamala with a golden pendant, and a garland of skulls (mundamala), each skull tied to the next by the hair on its crown. Around the head, toes and arms is wrapped the multi-stringed long vanamala. On his earlobes hang the black kundalas or earrings of animal hide (fig. 25.6). The five heads—Sadyojata, Vamadeva, Aghora, Tatpurusa and Isana—symbolising the earth, water, heat (fire), wind and sky, respectively, are borne by a corpulent body with ten arms. All the heads have three eyes, except that of Aghora (who is to the rear and is therefore not visible). This invisible head of Aghora is symbolically represented by the flowered mound at the rear top of the crown. The fifth head, of Isana, is on the top, looking skywards. The three eyes are representative of sun, moon and fire. While the two normal half-closed eyes, as of a mahayogi, regulate the cosmic order, the third eye, bearing fire, causes annihilation at the end of the eon. Just below the third eye, a crescent moon is cleverly placed along with the smear of horizontal tilaka on all the visible heads. The VDP (III. ch. 48. 1–20) provides a corresponding common name to each of these five heads. The ego-head or the central head is that of Mahadeva. The head not seen is that of Bhairava, who is supposed to be of terrific form (hence hidden). Significantly, the head to the left, of Rudra, is known as Umavaktra or the one in the direction of the goddess, and is symbolic of water; and the head to the right is Nandivaktra, i.e. facing its vehicle Nandi, the bull. The top one is “the purifying” face of Sadasiva.

The most interesting aspect of this painting, from the point of view of our discussion is the ayudhas or the cognizance signs associated with each of the ten hands. These are divided into a left–right group of five each. The emblems carried by the proper right-sided hands, from top to bottom are: trident; damaru; transparent skull-cup full of fresh blood (indicated by the tuft of hair hanging below the kapal); the unsheathed sword or khadaga; and hooded sarpa or serpent. The top two and the bottom wrists are wrapped in strings of rudraksa seeds; the third wrist is wrapped in a red rudraksa (?) string; while the fourth wrist band is of red and black rudraksas. Interestingly, the proper left sided emblems defy the known Sadasiva iconography. From top to bottom, these are: mace; conch or sankha; cakra or disc; lotus; and shield or dhala

Fig. 25.6 Harihara Sadashiva from Mandi, c. 1720–30. Howard Hodgkin Collection. Acc. no. LI118.4. Also published by Kramrisch (1984: 168, pl. 8). Courtesy: Ashmolean Museum of Art and Archaeology, University of Oxford

294 Pahari Harihara or Harahari

along with white rosary. The wristbands are of white and yellow basil or tulsi beads. The association of tulsi, basil along with the emblems in the top four hands, is unmistakably with Visnu-Narayana. The painter is subtly portraying the image of Narayana-Sadasiva. By locating the Narayana towards Umavaktra or the goddess’ head, the painter is also making a subtle statement of hieratical dominance. This is established by placing the ayudhas associated with the goddess, the rosary and shield in the lowermost left hand. The iconographical intention is more clearly evident in the c. 1730–40 variation of this painting, now in Victoria and Albert Museum (fig. 25.7). In this gouache on paper—carrying an inscription in English “GA P 44”, referring to the Gayer-Anderson collection (painting no. 44) that it belongs to—the five-faced ten-armed Sadasiva, draped in the leopard-skin dhoti, is seated cross-legged atop a stepped stone-pedestal, just like an idol placed in a temple (Goswamy and Fischer 1992: 2007, pl. 84). This imagery is perhaps a visual representation of the pancanna-linga, housed in the Pancavaktra temple in Mandi, although the iconography has been changed to match the Sadasiva painting discussed above. Seated on an elephant hide, the tusked head and neck of which is casually slung across his left shoulder as though it were the upper garment of the unseen Bhairava, (VDP: III. ch. 59 vs 1–7), this Sadasiva is of greyish-blue hue, perhaps because he is smeared in ash. The hairiness of the torso is accentuated by shading and stippling. The figure wears a mundamala and knotted serpent. Interestingly, he also wears a yogi’s nada (horn shaped object for sound) around his neck, with jatas covering his upper body. All three of the main faces are depicted in three-quarter view, and bear the crescent moon and horizontal tilaka on the forehead. The matted locks are heaped to form a crown in which lies the Sadasiva face, looking towards the sky. The subtle variations are in the earrings. Mahesvara, or the middle head, sports none. The Vamadeva-head, the one towards proper right, sports the kundal of animal hide. A serpent of invisible Aghora head is partly visible. The head of the goddess half—the left side—wears a pearled earring. The ten arms are divided into two groups of five each. The proper right hands hold a fine trident encased in a black rod in the middle; a damaru, drum; blood filled skull-cup with a tuft of hair visible; an unsheathed sword; and hooded dark-blue snake. There are gold rings on the fingers and the horns of Nandi, the bull, who is towards the right side of the pedestal are also encased in gold. The proper left half

Fig. 25.7 Harihara Sadashiva from Mandi, c. 1730–40. Given by Col. T. G. Gayer-Anderson and Maj. R. G. Gayer-Anderson, Pasha. Courtesy: Victoria and Albert Museum. IS.239-1952. Also published by Goswamy and Fischer (1992: 207, pl. 84)

Mahesh Sharma 295

similarly has Vaisnava emblems in the four upper hands—the mace; gold-encased conch; the disc or cakra; and resplendent lotus. The lowermost hand holds a rosary and a dhal, shield. These attributes are emblematic of the goddess, reinforced by a growling tiger to the left side of the stone pedestal. Clearly the painter is making an emphatic statement. The proper left side is the domain of Visnu, as it is also the domain of goddess. That the Mandi images display a clever, and clearly intentional, manipulation of iconographical elements is evident when we compare these Sadasiva paintings with other contemporary western Himalayan examples. For instance, the Kangra-style painting of Sadasiva from Mandi, now in the Bharat Kala Bhavan, Varanasi, portrays the deity with his standard cognizances: bow, lotus, shield, mriga, trisula, sword, varadamudra, abhayamudra, parasu and vajra (Panthey: 39). Another example of Sadasiva seated on an elephant skin from Basohli, depicts him holding the usual parasu, damaru, trisula, shield, bow, khadaga, varadahasta, arrow, skull-cup and aksamala (Ibid). Obviously, the comparisons are pointers to the intention, perhaps not of the painters, but of their patron (in this case, the ruler). As Goswamy (1975) points out, the painters had little access to and understanding of complex iconographies. They were instructed by the Brahamana ideologues associated with the court about which elements to depict. These Brahmanas wrote instructions in the sidelines or margins of the paintings to remind the painters what were exact requirements. It is apparent from the comparison of Sadasiva paintings in different styles in the western Himalayan region that both the Brahamanas and painters were however aware of exact iconographies. The switch therefore was intentional, though subtle.

IIIAt one time, ordinary people might have seen these fused images as pointers to the subsumption of all deities in the person of one supreme god (as do the devout Saivites of Mandi and Benares today). With the benefit of the images and texts we have surveyed, we know that the historical situation was more complicated. There is a saying (narrated by Rai Krishnadas of Benaras) that at the time of death, Visnu whispers the name of Siva in the ear of the dying so that he may attain the domains of Siva in perpetuity. Also, the dying utters the name of Visnu to curry favour with Siva (Panthey: 65). Interestingly, the Saiva ascetics are often heard uttering the name of Visnu, again in the hope of bonding with Siva. Such a belief has also been documented in a Mandi painting. In this painting, among the worshipers

of Siva and Parvati is Krisna, who is seated in the lap of a devout bearded Saivite. Krisna is apparently whispering the name of Siva to this Saivite, who is perhaps touching the feet of Krisna in veneration (Panthey: 68). Such a belief is reinforced even in the medieval literature. Again I draw your attention to the Uttarakanda of Ramacaritamanasa. There is a story of a devout Saivite who hated the very sight of Visnu and therefore could not understand the love of his Saivite guru for Hari (104–06). His guru finally explained to the distraught disciple that only by worshipping Hari could one attain Hara. Hari is the beloved of Hara; just as Hara of Hari (Uttarakanda: 106.1c–2b):

The sole reward, my son, of worshipping Lord Siva is uninterrupted devotion to Sri Rama’s feet. Siva himself, as well as Brahma, adores Sri Rama.

Yet, we also need a cogent explanation for the painted variations, which were perhaps dictated by the dominant religious affiliations of each of these hill states. One way of looking at the question follows Lavy (2003: 21–39), who argued that the popularity of Harihara in pre-Angkor Cambodia was due to the conjoined deity’s scope as a model for a multiple concept of power, which could be “deployed” for the religious and political unification of the state. This theory applies equally well to the western Himalayan states, where such a symbol unified the differing ideological orientations of the state and the people, particularly after a sixteenth century marked by Mughal inroads into the hills and a simultaneous Vaisnavaisation of polity (Mandi, however, continued with its Saivite association). The emergence of a hybrid form unifying two contesting sectarian positions in one supreme symbol makes sense against this backdrop, particularly if we concede that the subtle iconographic nuances, as discussed above, were the dominant principle of the hybrid imagery. The subtle switching perhaps provided differing sectarian orientation to a similar looking object. Thus, a Harihara or Harahari reversal embodies and may explain the ambiguous religious political environment of that period. The nuances embedded in the form and the multiple connotations and meaning thereof may point to an elusive peace achieved by the two warring sectarian factions.

n