Encountering the Goddess in the Indian Himalaya: On the Contribution of Ethnographic Film to the Study of Religion

This paper examines the benefits of ethnographic film for the study of religion. It argues that the exploration of gaps between colloquial descriptions of divinities and their practical manifestation in ritual is instructive of the way religious categories are conceptualized. The argument is developed through an analysis of selected scenes from the documentary AVATARA, a meditation on goddess worship (Śaktism) among the Khas ethnic majority of the Hindu Himalaya (Himachal Pradesh, India). Centering on embodiments of the goddess in spirit possession séances, it points to a fundamental difference between the popular depiction of the deity as a virgin-child (kanyā) who visits followers in their dreams and her actual manifestation as a menacing mother (mātā) during ritual activities. These ostensibly incongruent images are ultimately bridged by the anthropologically informed edition of the material caught on camera, illustrating the added advantage of documentary filmmaking for approximating religious experiences.


Introduction
This paper examines the benefits of ethnographic film for the study of Indic religions. Specifically, it explores multiple manifestations of the pan-Indian cult of the goddess (Śaktism) in a West Himalayan hinterland as depicted in the documentary AVATARA (Harel 2020). The film investigates popular notions and ritual practices associated with the goddess among the Khas ethnic majority of the hilly tracts in Himachal Pradesh, India. The shooting of the film took place in Karsog tehsīl, a sub-region in the erstwhile princely state of Suket that is contiguous with a larger area known as "Serāj" (from Hindi, "Svarāj", "independent") on account of its residents' egalitarian traditions. It is currently located in the southeastern parts of Mandi District and in the southern portion of the Kullu District.
In addressing a region that was peripheral to the regulatory Rajput kingdoms that surrounded it and that is therefore poorly documented in historical sources, this article draws upon a substantial body of anthropological literature about Khas society and its traditions (Berreman [1963(Berreman [ ] 1999Berti 2001Berti , 2009Halperin 2019;Luchesi 2018;Moran 2018Moran , 2019aSax 1991Sax , 2003Sax , 2009Sutherland 1998Sutherland , 2003. The abundant information that these studies hold about the Khas and their highly political form of civil religion provides the context for the investigation of popular belief in the goddess (śakti). The pervasive hold of the goddess on her followers' psyche and social life highlights the peculiarities of this regional iteration of what is in essence a pan-Indian tradition: while the goddess is commonly described as a virgin child (kanyā) who visits people in their dreams, her physical embodiment in mediums (gur) and family elders through highly institutionalized forms of spirit possession invariably assumes the form of an authoritarian mother (mātā).
The split between the goddess as benevolent child and the goddess as punishing mother is consistent with the division between the domestic and public sphere in Indic religions (Simmons et al. 2018, Introduction). As Frank Korom (2020) notes, this distinction is part of a broader pattern of religious behavior that is both universally relatable and specificallyŚākta. In a thoughtful review of recent works about the goddess, Korom (2020, p. 402) underlines the importance of the movement between emic and etic perspectives as a fruitful method for studying the goddess. This approach also guides the enquiries below, where historical context, cultic links, and ethnographic data are used to contextualize the individual statements about the nature of the goddess and the dialogues and ritual practices caught on camera. A careful application of this method should, in principle, lead to an approximation of Khas perceptions of the goddess as both a universal entity that is familiar to scholars of South Asian religions and as a specific deity that assists her followers in a highly personal manner (by exorcising evil spirits, resolving disputes, lending money, etc.).
The specificity of the dual functions of the Himalayan goddess of the Khas becomes apparent when compared with the contribution of goddess cults to the regulation of social behaviors in other parts of South Asia. For example, the stress on containing female sexuality through marriage in the sophisticated urban setting of Tamil Nadu manifests in a complex ritual calendar in the Meenakshi Temple of Madurai, where the goddess's wedding constitutes the apex of yearly religious activities (Harman 1989). Slightly north of the Tamil heartland, in the Telugu south, the female attendants of the goddess have come under the gradual authority of high-caste male ritual specialists (Flueckiger 2015). Similar cases of appropriation recur across the Hindi heartland in various settings (Rao 2003).
In nearly all of these cases, the power of the goddess is attributed to her chastity. In the Tamil South, this belief is epitomized in the concept of "kar pu, a word that points at a combination of (what are considered) 'good' feminine virtues" that are associated with married goddesses and that are encouraged in women (Peres 2021, p. 356). Despite catering to a patriarchal system of values (loyalty to one's husband, self-restraint, sexual abstinence, etc.), these virtues are also believed to generate extraordinary powers. As Ofer Peres notes, the concept of kar pu mimics the Sanskritic notion of tapas insofar as it "represents both the [ascetic] practice and the power that one accumulates through the practice". In this respect, the South Indian idealization (and adoration) of female chastity as a source of womanly power (śakti) agrees with North Indian (and Himalayan) Rajput traditions, in which wives who join their husbands' funeral pyre are deified as "satīmātās" (Harlan 1992). 1 By way of contrast, in the predominately pastoral setting of the West Himalayan Khas, the goddess is fiercely independent in virtually all her forms. Local traditions and ethnographic data are remarkably scant regarding stories of the goddess's "domestication". Instead, the sites deemed most potent for ritual activity are those that are believed to permeate with divine female energy (śakti), from earthen pits that are simply called "place" (zāga) into which blood offerings are let (Bhatt et al. 2014, pp. 81-82;Negi 2012, pp. 20-27) to the Sanskritized pilgrimage sites of Tantric provenance that are known asśakti pīt . has or "seats of power" (Erndl 1993). While these beliefs are no guarantee for female emancipation insofar as Khas women continue to perform the majority of chores in house and field, they do hint at a consensual reading of the goddess as the ultimate animator of existence and therefore as the most powerful entity in the hills.
After this introduction, the first part of this essay reviews the background to the shooting of AVATARA (Harel 2020), the formulation of the documentary's central research questions, and the historical and cultural characteristics of the area in which it was filmed. Since the research questions that guided this project had a profound impact on the practical and analytical approaches that were developed in the course of filming and editing, the following sections shift to a first-person narrative so as to clarify the contingencies and considerations behind the analysis of religious phenomena in this study. An overview of the geography, history, and central religious sites in Karsog and its environs follows. The religious institutions in the valley are dated on the basis of material evidence to establish a continuity in religious activity in the region and the primacy of the goddess within this setting. The second part of the essay presents the main case studies that are explored in the film: an exorcism ritual that is performed by the goddess Chindi (Can . d .ī ka) and an intimate spirit possession séance in the hilltop temple of Shikari (Śikārī), the goddess of hunters. The myths and stories that are told in connection with each of the goddesses are reviewed, and an exposition of their physical manifestations and modes of worship as they were caught on film is presented and analyzed. The final section takes stock of the different iterations of the goddess in Karsog to reflect on the viewer-researcher's ability to comprehendŚaktism in twenty-first-century Himachal Pradesh through local categories.

Background: In Search of a Question
The shooting of AVATARA (Harel 2020) came about by chance or, more precisely, by way of serendipity. We had originally planned to document a mysterious "human sacrifice" ritual that we had read about in an unpublished ethnography of the Khas that is held at the British library, London (Emerson n.d., ca. 1930). Since the ritual only takes place at irregular intervals in a handful of villages, it was crucial that we reach the correct location at the right moment. Having made enquiries with local contacts, we set out for the village of Hurang in Mandi (Himachal Pradesh, India) in the summer of 2016 to document the rite but found out that the ritual had been postponed to the following year soon after our arrival on site. 2 We thus found ourselves "stranded" in the middle of a beautiful alpine state with all the equipment necessary to shoot a documentary, but with no clear idea of what it was that we were after.
The transition from excited anticipation to confused disappointment marked the beginning of a joint creative process that culminated in the shooting of AVATARA. No longer "predator-filmmakers" on the hunt for a specific ritual, we began to ask larger questions about the beliefs and practices of Himalayan societies at the fringes of the Indian subcontinent. Our conversations were shaped, to a large degree, by the very different backgrounds of the team that had assembled for the task: a scholar with over twenty years of experience in researching the history and culture of Himachal Pradesh, a veteran filmmaker who specializes in scientific films, and a co-producer doubling as soundman who possesses an extensive knowledge of cinema history.
Still committed to the documentation of religious rites, our discussions gradually narrowed down to the specifics of religious beliefs among the Khas. Since the team members were not overly familiar with each other, the questions that we formulated seemed rather basic, almost naïve. For example, we wanted to know whether the peasants and shepherds of the hills had a unified set of beliefs? If so, how did they practice them? Who were their gods? How did they manifest? While the substantial body of ethnographic literature about the region that had developed over the past two centuries provided ample data to answer these questions, our conversations soon revealed that what primarily interested us was the nature of the religious experience or, more precisely, the sentiments (bhāv) that underlie religious experiences in this particular strand of mountain culture.
Having defined our research interests, it remained to be seen whether it would be possible to relate them in cinematic language. The answer to this question would have a profound impact on the type of film that we were making. Was this to be an ethnographic film in the traditional sense of a descriptive document that seeks to capture, convey, and in some sense "explain" a social reality that is unfamiliar to its viewers? Or was this, rather, a documentary that uses the artistic power of cinema to convey extra-mundane phenomena in a coherent, intelligible fashion? In scholarly parlance, this distinction harkened to a fundamental divide in the study of religion: were we embarking on the documentation of Durkheimian "social facts" or were we rather pursuing the phenomenological manifestation of religion through subjective experiences of the divine? In other words, would we be documenting religious discourses and (ritual) actions that establish a sense of solidarity among their participants or were we on a somewhat more precarious search for encounters between individual believers and the ineffable?
Since these dual aspects of religion were not mutually exclusive and given that we had yet to locate a site for shooting our film, we decided that for the time being, at least, we would pursue them both. 3 Filming the social manifestations of religious behaviors was, theoretically at least, easy to achieve insofar as Khas society is, similar to most South Asian societies, rife with ritual traditions. However, capturing the experiential dimension of religion presented a challenge of a different order altogether. For one, the mediation of religious experience through language rendered most interviews "useless" insofar as they could only produce a discourse about the ineffable and not the religious phenomenon itself. 4 Most of our interviewees were averse to discussing their subjective experiences of the goddess in front of the camera and even when they did acquiesce, their comments tended to be too cursory and "dry" (descriptive)-that is, lacking in "feeling" (bhāv)-to engage viewers who were unfamiliar with the particulars of Khas culture. This was a far cry from the overpowering awesomeness of divine encounters that we had hoped for. We clearly had to tweak our approach.
Our search for a universal phenomenon that would be relatable to audiences beyond South Asia ultimately focused on the numerous stories of godly visitations. Amply reported in the anthropological literature and commonly popping up in conversations, the appearance of gods and goddesses, primarily in dreams, is a prominent feature of South Asian culture. I had personally heard numerous such stories during my stays in the hills, as apparently did the author of the aforementioned ethnography of the Khas a century before, who explained that It is a common belief in the hills that the soul or life of man consist of two parts. In the more Hinduised tracts the two elements are known as chitr and guptmemory and speech. 5 Gupt does not ordinarily leave the body until death, but chitr is of a wandering disposition, given to nocturnal adventures. It is he who leaves the body during sleep, and dreams are but the record of his adventures. With his return to the body, a dream ceases, although it may happen that he does not return at all. (Emerson n.d.,fo. 867) Although we would hear a near verbatim explanation of the splitting soul during our interviews with locals, the prospect of filming peoples' dreams seemed rather slim. At the same time, we felt that enquiring about the appearance of the gods in dream states would make a good conversation starter and help steer our interviews in a fruitful direction. Having agreed on a research topic and the main themes to be explored, we were ready to decide on which area in the hills would best serve our purposes for filming.

From Valley to Hilltop: Geography, History, and Culture in a Himalayan Hinterland
The administrative sub-unit (teshīl) of Karsog, today part of Mandi District but geographically and culturally contiguous with the egalitarian region of Seraj, provided an ideal setting for our project (Figure 1). Off the beaten track of most of the commercial and tourist traffic that traverses the alpine areas of Himachal Pradesh, Karsog fit our purposes perfectly in the sense that it was sufficiently populous to represent a "typical" Khas region with diverse communities but was also considered more conservative than the better visited parts of the hills. 6 It was also familiar to the author from an earlier visit in 2013.
Geographically, the valley of Karsog is locked between two ridges that diverge from the northerly summit of the "Hunter Goddess" (Shikari Devi) at its head and run all the way down to the mighty Sutlej River due south where they terminate. The only town of significance in the valley is the moderate settlement of Karsog. Situated at the center of the fertile valley, Karsog serves as the gateway for those arriving by road from other parts of the hills. The view of the valley is exceptionally serene. The absence of spectacular snowcapped mountains that feature in other parts of the hills is more than compensated for by the quilt-like canvas of lush greens (pre-harvest) and yellowish-brown paddies (postharvest) that expands in a soft decline between the ridges in the shape of an uncannily neat, symmetrical triangle. 7 Interspersed with various huts and settlements, the view gives the (deceptive) impression of a sealed off, self-contained area that is cut off from the broader world that surrounds it ( Figure 2). Geographically, the valley of Karsog is locked between two ridges that diverge from the northerly summit of the "Hunter Goddess" (Shikari Devi) at its head and run all the way down to the mighty Sutlej River due south where they terminate. The only town of significance in the valley is the moderate settlement of Karsog. Situated at the center of the fertile valley, Karsog serves as the gateway for those arriving by road from other parts of the hills. The view of the valley is exceptionally serene. The absence of spectacular snowcapped mountains that feature in other parts of the hills is more than compensated for by the quilt-like canvas of lush greens (pre-harvest) and yellowish-brown paddies (post-harvest) that expands in a soft decline between the ridges in the shape of an uncannily neat, symmetrical triangle. 7 Interspersed with various huts and settlements, the view gives the (deceptive) impression of a sealed off, self-contained area that is cut off from the broader world that surrounds it ( Figure 2).
Historically, Karsog was part of the Rajput state of Suket until Indian independence. The Sens of Suket claim origins in Bengal, their progenitors having migrated into the hills at some point during the eighth century CE (Hutchison and Vogel 1933, Vol 1, pp. 341-3). Despite their lofty pedigree, the Sen rulers (rājās) do not seem to have been overly involved in the affairs of the valley. The kings of Suket thus only appear in oral traditions as rivals of the territorial gods and goddesses (devī-devtā) that dominate social and political life in the valley. In this respect, the teshīl provided an ideal setting for examining grassroots religion insofar as it was only tangentially affected by the Rajput kingdoms that overshadow Himachali history today (Moran 2019b). The peripheral status of Karsog in relation to the Rajput kingdoms that surrounded it has important consequences for the type of sources that are available for learning about the region. If the Rajput Houses of (from the north, clockwise) Kullu, Bashahr, Shangri (an offshoot of Kullu), Suket, and Mandi (led by a junior branch of the Suketi Sens) that surround the teshīl have bequeathed dynastic rolls (vaṃśāvalīs) from which to construct historical narratives, Karsog and Seraj remained in the hands of the political substratum of "lords" and "barons" (rāṇā-ṭhākur) that permeates the hills. Since the latter rarely left any written accounts or inscriptions, our sources for reconstructing their pasts are primarily found in oral traditions and specimens of material culture. While many of these traditions were committed to print by Himachali scholars and devotees in recent decades, their popular transmission remains primarily oral in nature. The repetition of stories and rituals that appear in local publications in our encounters in Karsog tended to confirm the notion Historically, Karsog was part of the Rajput state of Suket until Indian independence. The Sens of Suket claim origins in Bengal, their progenitors having migrated into the hills at some point during the eighth century CE (Hutchison and Vogel 1933, vol. 1, pp. 341-43). Despite their lofty pedigree, the Sen rulers (rājās) do not seem to have been overly involved in the affairs of the valley. The kings of Suket thus only appear in oral traditions as rivals of the territorial gods and goddesses (devī-devtā) that dominate social and political life in the valley. In this respect, the teshīl provided an ideal setting for examining grassroots religion insofar as it was only tangentially affected by the Rajput kingdoms that overshadow Himachali history today (Moran 2019b).
The peripheral status of Karsog in relation to the Rajput kingdoms that surrounded it has important consequences for the type of sources that are available for learning about the region. If the Rajput Houses of (from the north, clockwise) Kullu, Bashahr, Shangri (an offshoot of Kullu), Suket, and Mandi (led by a junior branch of the Suketi Sens) that surround the teshīl have bequeathed dynastic rolls (vam .śā valīs) from which to construct historical narratives, Karsog and Seraj remained in the hands of the political substratum of "lords" and "barons" (rān .ā -t . hākur) that permeates the hills. Since the latter rarely left any written accounts or inscriptions, our sources for reconstructing their pasts are primarily found in oral traditions and specimens of material culture. While many of these traditions were committed to print by Himachali scholars and devotees in recent decades, their popular transmission remains primarily oral in nature. The repetition of stories and rituals that appear in local publications in our encounters in Karsog tended to confirm the notion that these locally significant narratives are still accepted as "historical truths" today.
The structural similarity between the political history of the rān .ā s and t . hākurs of Seraj and that of the Rajput kingdoms is instructive of broader cultural patterns among the Khas. In both cases, the relations between the communities and their leaders were fraught with feuds of a limited nature over the control of lands, prestige, and revenue (Hutchison and Vogel 1933, vol. 1, preface). However, if the rān .ā s and t . hākurs who were subsumed under the apparatuses of Rajput states tend to be incorporated in the dynastic histories of their masters, the local leadership of Karsog was only nominally incorporated in such histories. Rather, it was the host of territorial gods and goddesses that comprised the ultimate source of authority in Karsog, and it is their stories, myths, and rituals that provide the most credible source for comprehending the region's past, religion, and culture.

Village Gods, Regional Goddesses, and the Sacred Geography of Karsog
Society in Karsog follows the general model of social organization among the West Himalayan Khas, which is commonly referred to as "government by deity" (devtā kā rāj). Broadly conceived, this "system" of indigenous governance centers on village gods (devīdevtā) that function as the "kings" of "little kingdoms", whose sovereignty is enacted through complex ritual protocols involving the entirety of the community (Sax 2003;Sutherland 1998). The devtā cult is thus a civil religion that regulates social life and political relations between community members and between similarly structured communities across the region.
The pronouncedly political and territorial character of the devtā regime (Berti 2009) is complemented with a broad spectrum of beliefs and religious sentiments among its followers. In this respect, Karsog recalls many other parts of South Asia, in which popular creeds are underlined by a pervasive belief in the goddess (Śakti) as the ultimate manifestation of reality. While this belief does not preclude the admittance ofŚaiva, Vais . n . ava, Sikh, and other creeds, our enquiries indicated that the notion that the world is an illusory (māyā) construct that is sustained by divine female energy (śakti) was an assumption shared by most, if not all, of the area's inhabitants.
Widely attested to in scholarly literature (Erndl 1993) and unequivocally confirmed by enquiries on site, it quickly became apparent that it would be impossible to shoot a film about religion in Karsog without mentioning the goddess. The existence of the Goddess as supreme reality is complemented with specific manifestations in particular geographical sites, where she is worshipped under various names and in different guises. Recording stories that are associated with such sites and documenting the rituals that are performed there became primary filming goals. These goals were complemented with interviews that aimed to delineate the goddess's place in her followers' psyche.
The story that opens the film, for example, is a narrative from Chamba State that was told to us by a teacher from the valley floor. According to this tradition, the goddess appeared to the king in a dream warning of impending floods, droughts, and calamities that would wreak havoc if she were not propitiated. The goddess demanded the king's only son in sacrifice but, after some haggling, agreed to take the latter's wife (bahū) instead and thus prevented a catastrophe. 8 That these traditions are centuries-old may be deduced from the story of raja Madan Singh of Suket (c. 13th century CE), who abandoned the ancient capital of Pangni (due west of Karsog) after the goddess appeared to him in a dream to complain that his palace had been constructed over her home (Hutchison and Vogel 1933, vol. 1, p. 350). The material remains in temples from Karsog proper provide even firmer proof of the area's antiquity.
The temple of Mamleshwar Mahadev, alias "Gaurī-Śaṅkar", lies adjacent to the main road and is home to a local manifestation ofŚiva. The inner sanctum houses a metal cast image of the deity's family ("Śiv-Parivār"), which consists ofŚiva, Pārvatī, and their sons Kārtikey and Gan . eś. The temple is also famous for housing oddities such as a gigantic grain of wheat, a legacy from the time when the Pān . d . avas had allegedly sojourned in the valley during the Dwapar (dvāpara) Age. According to Om Chandra Handa, this architecturally unique temple replaced an earlier stone temple or temples that housed stone lingams (Handa 2001, pp. 272-75). The lingams, along with various stone sculptures of different styles and rock types, are now worshipped in the temple courtyard.
While theseŚaiva relics are clearly ancient, the dating of the site is ultimately facilitated by a Vais . n . ava image. Situated to the right of the inner sanctum, a near life-size polycephalic image of Vis . n . u with Varāha and Narasim . ha faces known as Vaikun . t . ha bears strong stylistic affinities with images from medieval Kashmir ( Figure 3). These resemblances suggest the black stone image was already established in the temple during the eighth century, that is, a century after the Bengali Sens gained a foothold in Suket. 9 That a work of this level of sophistication was executed in a sub-region of the state soon after the arrival of the new dynasty-and its evident links with coeval works in Kashmir-suggests a vibrant religious culture was already well in place in Karsog prior to the parvenu Sens' establishment in the hills. Following the road for another 7-8 km through the paddies leads to the terminus at Kau, the home of Kāmākśā ("Kāmākhyā") Devi. A terrifying manifestation of Kālī, this goddess is visited by pilgrims throughout the year and is particularly popular during the Navaratras. Numerous blood sacrifices (balidān) used to be offered here before the outlawing of animal sacrifice by the state government in 2014. 10 A third site of importance Following the road for another 7-8 km through the paddies leads to the terminus at Kau, the home of Kāmākśā ("Kāmākhyā") Devi. A terrifying manifestation of Kālī, this goddess is visited by pilgrims throughout the year and is particularly popular during the Navaratras. Numerous blood sacrifices (balidān) used to be offered here before the outlawing of animal sacrifice by the state government in 2014. 10 A third site of importance mirrors Kamaksha in the northern part of the valley. Known as "Shankar Dehra", the site is some 7-8 km north of the main town in the last settlement of the valley at the foot of the forested slopes that climb up to Shikari Devi. A rudimentary cement structure serves as a temple (dehrā) toŚiva (Śaṅkar) housing a stone that is identified as a lingam.
The temples of Mamleshwar Mahadev and Kāmākśā Devi are not only the most sizeable religious structures in the valley, but also the ones that most clearly depict a connection between its denizens and the Khas society of the higher reaches. They are thus constructed in the Pahari style of interlaced wooden beams and stones that is common to the higher hills, have extensive courtyards, and have separate wooden storehouses or "treasuries" (bhan . d .ā r) attached to them as per the custom of territorial deities elsewhere. These links are also evinced in the temples of the ridges above the valley. As in the Shimla Hills to the south (Moran 2019a), the residents of the mountains that surround Karsog identify as descendants of the warring sides of the Mahābhārata. The eastern ridge claims Kaurava descent, while those on the west purport to be the progeny of Pan . d . u. Each of these "factions" is subject to a prominent territorial deity of the ancient serpent (nāgā) type that are headquartered in impressive bastion-like fortresses of timber and stone. 11 Despite their pronounced significance to social and political life, the nāgā sovereigns did not register as exceptionally important at the level of personal belief. It was, rather, the goddess in her various manifestations that recurred in peasant narratives and who, somewhat fittingly, occupied the mountains' highest reaches. After a few days of filming and interviews in the valley, we decided to relocate to the western ridge above the valley, halfway towards the headquarters of Mahu Nag. Having shifted from fields to orchards, our move set the tone for a cinematic journey from the valley floor to Shikari Devi, the "Goddess of Hunters", who resides on the tallest peak between the Beas and Sutlej Rivers in Seraj (3359 m). As we soon found out, there was at least one additional goddess of significance between the valley and the mountaintop.

Encounters with the Goddess
In the second part of this essay, the relationships between the goddess and her followers are analyzed through selected scenes from AVATARA (Harel 2020). The background and "behind the scenes" context of each of the goddesses that was encountered is provided first. These are followed by the presentation of the scenes in which they feature and an analysis of their contribution to the anthropology of religion among the Himalayan Khas. Since this section provides "spoilers" to the cinematic experience of viewing the documentary, it is recommended to read it only after having watched the film.

Chindi (Can . d .ī ka) Devi
We set camp in a damp building with walls in dire need of plastering at the edge of an apple orchard that stretched above us on a hill. Apple packing for wholesale markets in the plains seemed to have replaced rice planting as the agricultural activity of choice, and a quick stroll along the road soon brought the familiar fields of Karsog into sight. Our dwellings were some two kilometers away from the village of Chindi, which spread west from the temple of the goddess Cin . d .ī (alias "Can . d .ī ka") on the high point of the ridge. 12 Although there were no written sources to work from to construct the goddess's history during our visit, becoming regulars in the chai shop opposite the temple helped overcome this lacuna in knowledge.
During our two weeks in Chindi, morning hours were mostly spent conducting interviews with locals and travelers at the chai shop, from whom we gathered abundant information about the site. We learned that the goddess was a form of Durgā, the undis-puted ruler of the area and emphatically juxtaposed with Kāmākśā of the lower valley, whom locals equated with Kālī. This identity was evinced in her strict "vegetarianism", that is, non-acceptance of blood sacrifices. We were also told that she was a "mountain girl" (pahār .ī lar . kī) who had originally reached Chindi via the Jalori Pass that leads into Kullu. The owner of the chai shop, an elderly woman from Lahaul who had eloped many years ago with her now-deceased husband to resettle in Chindi, was particularly proud to inform us that the goddess had never married, no doubt reverberating her own experience as a young escapee from the northern regions.
Like the answers that we had received in the valley below, the patrons and proprietors of the chai shop claimed that the goddess visited people in their dreams. Her form was almost always that of a virgin (kanyā kā rūp), who was exceedingly beautiful (khūbsūrat), wore a red dress, and was bedecked with golden jewelry. However, in certain cases, she might appear dark and horrid with flies hovering around her, a sure indication of impending death. Regardless of form, the meetings with the goddess were universally described as elusive, fleeting moments, "like the wind blowing through the bazaar (man . d .ī kī havā)". While this input was promising insofar as it was consistent with descriptions of the goddess and celestial encounters in other parts of South Asia, our hopes of capturing on film the goddess that inhabited peoples' minds remained rather slim. We decided to enquire further with the temple officiates.
In keeping with the structure of the "devtā system" (Sax 2003;Sutherland 2003), the sovereign goddess of Chindi ruled through an entourage of "god people" (devtā log) that included a chief administrator (kārdār), an official spokesperson or medium (gur), and several other functionaries from the community. The charismatic medium is usually the most important of these officials in the sense that he embodies the deity in ritual activities and can be approached for help and guidance by members of the community. As we later found out, local tradition holds that the mediums of Chindi were descended from an ancestor who had been visited by the goddess many years ago (Harnot 1991, pp. 102-3).
The story goes that the ancestor, a farmer from the village of Churag (halfway between Chindi and Mahunag), had heard a knock on the door one night, and when he opened it, he found a beautiful, lonesome girl (kanyā) asking for assistance. He offered her food, drink, and a bed for the night, and although the child never touched her food, she did drink amply of the milk (an indication of her followers' association with Vais . n . avism). After they went to bed, the goddess, impressed with her host's kindness, appeared to him in a dream. She introduced herself as "Durgā Bhagvatī" and offered to stay and protect him and his neighbors provided they build her a temple (Harnot 1991, p. 102).
The farmer started from his sleep and immediately went looking for the girl, but her bed was empty. He looked outside and saw nothing at first, but then heard a gentle tap of footsteps. He followed the sound all the way along the ridge until he reached the site where the temple currently stands, where the tapping suddenly stopped. Looking down towards the ground, he found a stone statue of the goddess that is now enshrined in the temple (the image was covered in cloth during our visit and so could not be made out in detail). The farmer told his neighbors about the encounter, had a temple built in her honor, and became her official spokesperson, a role that has passed in his lineage ever since (Harnot 1991, p. 103).
Somewhat oddly, the current medium ignored this story altogether in his account of the goddess's history. A householder in his mid-30s, Jyoti Sharma claimed to have been chosen by the goddess when he was a child. He reports having had severe problems concentrating at school as the goddess would continually call to him and occupy his mind. One night, when he was twelve, he wandered off from his bed and woke up in the goddess's temple, after which point his identification with her was made complete. Although Sharma admitted that his role passed in the family and that he was thus a "dynastic medium" (khāndānī gur), he seemed to deem his personal connection with the goddess more important to his position than the family credentials that may have helped in establishing it.
The bond between Sharma and the goddess was particularly palpable in his skillful response to our queries. Captivating with his enthusiasm, his narrative of the goddess's history was delivered with exceptional gusto and charisma. 13 Thus, although we had arrived unannounced, Sharma quickly accommodated our needs by seating himself in the courtyard outside the temple to recount the goddess's tale (Figure 4). A circle of listenersprimarily men and young boys-quickly grew around us, listening in and contributing with tales of their own. The story that Sharma presented was far richer in details and decidedly different from the one recorded in print.  The oral account of the goddess's origins, which was never (to the best of my knowledge) committed to print, carried significant weight as an "official" narrative of Chindi's history that is told by a leading member of her entourage. Sharma's discourse whetted the appetite of his listeners, who joined his performance with additional stories of encounters with the goddess. We learned a historical tale about a devotee who had been imprisoned by the raja of Suket and saved by the goddess, who manifested in a swarm of bees that carried him to safety. More recently, a middle-aged villager recounted how the goddess had answered his prayers just as he was about to undergo brain surgery in a Chandigarh hospital. The surgeons had cut his cranium open to remove a massive tumor only to discover that the tumor had miraculously disappeared, stitched the patient back, and sent him home (Thakur n.d. Dipu Thakur, personal communication, July 2016).
The consensus between the medium and the goddess's followers as regards Ciṇḍī's superhuman powers notwithstanding, their accounts of her manifestation differed strongly. For her followers, the goddess was a benign figure who answered prayers and assisted in challenges. Sharma did not openly oppose such salvific narratives, but he did place an emphasis on her tempestuous, puerile nature as a child who needs cajoling lest her temper burst and liquidate her surroundings. When we asked whether it would be possible to see the goddess at work, our hosts beamed with delight. We were invited to visit the temple some ten days hence, on the first day of the month (saṃkrānti), when the goddess would manifest her healing powers for all to see by exorcising demons from the afflicted.
Encounter 1: The Goddess as Exorcist Saṃkrānti arrived. We made our way through a thick fog towards the temple to the beat of the ḍhol drum that announced the happening. 14 The temple veranda and courtyard were crammed with families and a general sense of commotion emanated from inside.
Stepping over the threshold, we found families, children, and couples sitting patiently on According to Sharma, the goddess had migrated from Kullu via the Jalori Pass in an arduous journey that culminated in her installation on the ridge. The tale was consistent with bits and ends that we had gathered from locals in the chai shop but was delivered with an authority that far exceeded the scholarly discourses we had encountered thus far. In Sharma's narrative, the goddess had wondrous encounters with supernatural entities and people of diverse backgrounds across the region. She travels with a familiar in the form of a cat who transforms into a tiger, turns a blacksmith who fails to build her a sickle (drāntī) into a stone, and uses her magical powers to create lakes and natural water tanks by cutting through the earth in the forest. She also subdues local demons who become her protectors. In all of these cases, the goddess comes across as a childish figure of exceptional power (śakti) who must be appropriated lest she turns wrathful.
The oral account of the goddess's origins, which was never (to the best of my knowledge) committed to print, carried significant weight as an "official" narrative of Chindi's history that is told by a leading member of her entourage. Sharma's discourse whetted the appetite of his listeners, who joined his performance with additional stories of encounters with the goddess. We learned a historical tale about a devotee who had been imprisoned by the raja of Suket and saved by the goddess, who manifested in a swarm of bees that carried him to safety. More recently, a middle-aged villager recounted how the goddess had answered his prayers just as he was about to undergo brain surgery in a Chandigarh hospital. The surgeons had cut his cranium open to remove a massive tumor only to discover that the tumor had miraculously disappeared, stitched the patient back, and sent him home (Thakur n.d. Dipu Thakur, personal communication, July 2016).
The consensus between the medium and the goddess's followers as regards Cin . d .ī 's superhuman powers notwithstanding, their accounts of her manifestation differed strongly.
For her followers, the goddess was a benign figure who answered prayers and assisted in challenges. Sharma did not openly oppose such salvific narratives, but he did place an emphasis on her tempestuous, puerile nature as a child who needs cajoling lest her temper burst and liquidate her surroundings. When we asked whether it would be possible to see the goddess at work, our hosts beamed with delight. We were invited to visit the temple some ten days hence, on the first day of the month (sam . krānti), when the goddess would manifest her healing powers for all to see by exorcising demons from the afflicted.
Encounter 1: The Goddess as Exorcist Sam . krānti arrived. We made our way through a thick fog towards the temple to the beat of the d . hol drum that announced the happening. 14 The temple veranda and courtyard were crammed with families and a general sense of commotion emanated from inside.
Stepping over the threshold, we found families, children, and couples sitting patiently on the floor at various distances from the inner sanctum, where Jyoti Sharma and a group of priests sat adjacent to the central alcove with the primary image (mūrti) of the goddess ( Figure 5). We were free to film so long as we remained outside the inner hub of ritual activity.
since he first started serving the goddess (keeping hair uncut is one of the various restrictions that mediums are expected to follow). His expression was stern and attentive. Facing him was a young woman accompanied by a man who appeared to be her husband and the latter's mother. The young woman was sobbing and mumbling unintelligibly. Even as this drama was unfolding, a strong heaving sound was heard from the opposite wall of the temple, where another loose-haired medium sat on a pedestal. This was Amrit Pāl, the goddess's servant and protector. The young woman was carried over to face the pedestal before the second medium, escorted and supported by her kin. The medium was tasked with consolidating the goddess's dictums by placing his sword on the back of the client and tying rice from the temple treasury in their hair for protection. This type of exchange between the goddess and her aids, on the one hand, and the in-married women and their husband's families, on the other hand, continued throughout the day in varying degrees of intensity. In some instances, the patient would burst into violent shrieks, signifying that they were utterly inconsolable, and in others, they would be more explicit about what haunted them, which almost always had to do with quarrels in the family, most often over land. Once a patient had been treated and transferred to the goddess's deputy for the rites that solidify the deity's ruling, the medium would double tap an iron ring that was attached to the inner sanctum's door, signaling that the next patient come forth. The medium's long hair, which is usually kept covered under a woolen cap, was exposed and reached well beyond his shoulders, evincing the many years that had passed since he first started serving the goddess (keeping hair uncut is one of the various restrictions that mediums are expected to follow). His expression was stern and attentive. Facing him was a young woman accompanied by a man who appeared to be her husband and the latter's mother. The young woman was sobbing and mumbling unintelligibly. Even as this drama was unfolding, a strong heaving sound was heard from the opposite wall of the temple, where another loose-haired medium sat on a pedestal. This was Amrit Pāl, the goddess's servant and protector. The young woman was carried over to face the pedestal before the second medium, escorted and supported by her kin. The medium was tasked with consolidating the goddess's dictums by placing his sword on the back of the client and tying rice from the temple treasury in their hair for protection.
This type of exchange between the goddess and her aids, on the one hand, and the in-married women and their husband's families, on the other hand, continued throughout the day in varying degrees of intensity. In some instances, the patient would burst into violent shrieks, signifying that they were utterly inconsolable, and in others, they would be more explicit about what haunted them, which almost always had to do with quarrels in the family, most often over land. Once a patient had been treated and transferred to the goddess's deputy for the rites that solidify the deity's ruling, the medium would double tap an iron ring that was attached to the inner sanctum's door, signaling that the next patient come forth.
Although none of the women we saw professed to be possessed by an alien entity, it was commonly understood that they had come under the control of demons (asur, rākśas) or ghosts of various sorts (bhūt-pret) and that it was in the goddess's power to remove these pestering entities. That Jyoti Sharma was embodying the goddess during these sessions was made explicit when a bus driver showed up (in uniform, apparently on a break from his route) to discuss a grievance unrelated to spirit possession.
The camera caught the séance when the sides were already in dialogue. The driver stood with his hands folded in namaskar halfway between the inner sanctum and the temple door. A short wooden barrier separated the inner court where exorcisms were taking place from the outer ring of participants, onlookers, and patients in waiting. The driver was closely listening to the goddess's words, and when he spoke, it was in the normal manner with none of the signs of possession that we had observed before. [Driver bows his head.] As with the earlier exchanges, Sharma was enacting the goddess in her capacity as a regulator of the community's social life. However, contrary to the exorcisms, the dialogue between the parties had little to do with the "expulsion" of "unnatural" forces affecting psy-chological discontent. Rather, the goddess and her follower were negotiating the conditions of a loan made out to the latter, most likely from the temple's considerable earnings.
In this respect, the exorcisms that were slated to be the main ritual activity of the day were temporarily interrupted for another central function of the temple establishment, namely as a banking institution that services the community (Singh 1989;Thakur and Bhatt 2017). As the dialogue between the bus driver and the goddess indicates, devtā temples can use their income to compete over clients in a bid to increase their authority. The threat that "all your family history is known here" and the demand that the client "not visit other temples" (a bus driver is exceptionally mobile and can easily improve the conditions of his loans by checking for competing offers in other temples) are thus intended to place the goddess in a position of social and political dominance.
The rituals at Chindi were revealing of the way the goddess is conceptualized by the West Himalayan Khas and of broader affinities to do with marginalized social groups writ large. Although the goddess was popularly conceived of as a virgin-child in interviews and public discourse, her actual presence when possessing her medium cast her as a "mother". While this split may serve to substantiate the deity's authority during ritual exchanges, it also highlights the protective powers that it purports to extend to its followers.
That the medium is a Brahmin male and his clients in-married daughters (dhiyānī) from other parts of the hills placed our enquiries on firm anthropological ground. The exorcisms in the temple conform with a wider pattern of "ecstatic religion" that Ioan Lewis (1971) had famously documented in the Sudan. According to Lewis, the institutionalized nature of séance possessions in which men officiate as exorcists plays an important part in smoothening ruptures in the social fabric. Specifically, they provide socially disadvantaged in-married women with an opportunity to vocalize their otherwise suppressed criticism of conflicts and difficulties in their new surroundings through ritualized behavior that is deemed socially acceptable. As the hapless "victims" of malicious entities, the goddess's patients are given a public platform on which to express their grievances without risking a backlash in the domestic sphere afterwards (Lewis 1971; for other modes of "venting" in Garhwal, see Sax 2009).
The encounters with the goddess at Chindi, where the medium and his team alternately approximated community "social workers" and moneylenders, underline the centrality of village gods to Khas social life. Embodied by the medium who had provided us with the narrative of Chindi's arrival to the ridge some two weeks earlier, we saw how the benign goddess that is popularly conceptualized as a virgin-child transformed into an authoritative motherly figure in situations that revolve around ritual healing and economic dealings. The meetings at Chindi had thus achieved at least two of our central goals: the documentation of religious practices along "traditional" ethnographic lines and a nuanced appreciation of the way the goddess manifests in the minds and actions of her followers.

Shikari (Śikārī) Devi
After a few days of exploring the Chindi ridge and its temples, we began our ascent to the mountaintop temple of Shikari Devi (Śikārī Devī). As her appellation implies, the goddess used to be frequented by hunters who patronized the thick forests that surround the temple in search of game, and it is said that it is impossible to succeed in a hunt without receiving her blessing (Harnot 1991, pp. 108-9). While the conversion of the forest around the site into a wildlife sanctuary put an end to hunting, the goddess's name and reputation remain unchanged. Today, it is primarily frequented by pilgrims from the hills and, to a lesser degree, by tourists and sightseers from outside the state.
The temple of Shikari differs from the majority of Pahari temples in several ways. As the highest peak in the region, the site was most likely the object of a mountain cult prior to its current rebranding. The pervasive association of mountain peaks with goddesses aside (Emerson n.d.,Chapter 12), the site also seems to have been the focus of bloody sacrifices that are witnessed in stone slabs known as "bars . elas" (on account of their being in ritual use for only one year) that commemorate satī and that are found in the back of the temple. The standard image is of a large mounted or standing warrior who is surrounded by female figures of smaller sizes (changes in size indicating rank), apparently the women of the harem who had followed the deceased to the funeral pyre, often along with their maidservants and several high-ranking officers in the departed ruler's service (Bindra 1982).
The pilgrims who visit the site climb a long set of stairs from a collection of d . hābās and tents that serve as pilgrim's houses (dharmśālās) before reaching the base of the goddess's temple. Devoid of distinct architectural features, the "temple" is in fact nothing more than a cement platform with an elevated inner sanctum that is conspicuously lacking a roof. The snow that falls over the mountain in abundance during the winter is said to never accumulate within the sacred compound itself, indicating its sanctity.
The temple proper consists of a smaller concrete dais that is surmounted on the base platform. This inner section is divided into a large open area into which visitors enter and a narrow stretch at its end which houses a series of sacred images (mūrtis) and which is separated from the "hall" by a cement railing. The images are almost entirely made of stone in different, somewhat crude styles. Instead of one central image, the temple houses a collection from the Hindu pantheon with a relatively prominent Nāginī flanked by Kālī, Durgā, Bhairav,Śani, and so forth ( Figure 6). The pilgrim's exchange with the deities is mediated by a priest (pujārī) who occupies the space between the cement railing and the images. As is customary in Hindu temples, visitors present gifts (flowers, cash, sweets, etc.) to the deities via the priest, who reciprocates with similar items containing the deity's blessing (prasād).
Religions 2021, 12, x FOR PEER REVIEW 15 of 23 mediated by a priest (pujārī) who occupies the space between the cement railing and the images. As is customary in Hindu temples, visitors present gifts (flowers, cash, sweets, etc.) to the deities via the priest, who reciprocates with similar items containing the deity's blessing (prasād). The history of human sacrifices and the goddess's explicit association with hunting suggested that Shikari was perhaps not quite as "vegetarian" an entity as Chindi. There was, however, little to indicate this upon arrival. Having climbed up the ridge from Chindi for the better part of the day, our first meeting at the ḍhābā was with a Brahmin pandit from Mandi Town. The pandit had been engaged by four young professionals to perform the "yoginī pūjā", a ritual service aimed at securing good brides for his clients. The dedication of the ritual to yoginīs is consistent with the broad perception of mountain tops as abodes of "Kālī" and her attendant spirits, who are alternately called yoginīs, joginīs, ḍākinīs, or ḍāyns (Emerson n.d., Negi 2012). The terms and powers associated with these entities suggest a medieval provenance in Tantrism (Shaw 2006), but the site itself is likely to have been held sacred long before that.
Our conversation with the pandit revealed that he was a priest by profession and well versed in Sanskrit. He also professed a personal connection with the goddess, whom he claimed to frequently meet and converse with in his dreams. The goddess, he explained, appears as a child in a red dress and sits next to him to discuss this or that topic for hours before retiring. Encouraged by the auspicious presence of the goddess, we concluded to follow the pandit and his clients the next day.
We found them on the platform outside the temple early in the morning with bundles The history of human sacrifices and the goddess's explicit association with hunting suggested that Shikari was perhaps not quite as "vegetarian" an entity as Chindi. There was, however, little to indicate this upon arrival. Having climbed up the ridge from Chindi for the better part of the day, our first meeting at the d . hābā was with a Brahmin pandit from Mandi Town. The pandit had been engaged by four young professionals to perform the "yoginī pūjā", a ritual service aimed at securing good brides for his clients. The dedication of the ritual to yoginīs is consistent with the broad perception of mountain tops as abodes of "Kālī" and her attendant spirits, who are alternately called yoginīs, joginīs, d .ā kinīs, or d .ā yns (Emerson n.d.;Negi 2012). The terms and powers associated with these entities suggest a medieval provenance in Tantrism (Shaw 2006), but the site itself is likely to have been held sacred long before that.
Our conversation with the pandit revealed that he was a priest by profession and well versed in Sanskrit. He also professed a personal connection with the goddess, whom he claimed to frequently meet and converse with in his dreams. The goddess, he explained, appears as a child in a red dress and sits next to him to discuss this or that topic for hours before retiring. Encouraged by the auspicious presence of the goddess, we concluded to follow the pandit and his clients the next day.
We found them on the platform outside the temple early in the morning with bundles of ritual paraphernalia ready for use. The pandit oversaw the construction of an earthen altar, into which the clients' offerings were to be poured, and arranged bekhal branches into a meticulous quadrangle pile for the fire. He gave further instructions for preparing a mix of vegetable offerings (coconut, rice, ghee, dates, etc.) in an aluminum bowl while building an elaborate yantra from flour and rice into which the deities and planets necessary for the rite were to be summoned. His slightly confused urbanite clientele did its best to follow the instructions as they assumed positions facing the altar with individual trays of offerings in front of them.
Encounter 2: The Goddess as Counselor While the pandit was busy chanting mantras and gesturing his clients to cast ritual offerings into the fire, a second group of pilgrims made its way up the mountain. 15 Unlike the urbanites who had travelled to the temple from Mandi by car, the newcomers treaded a narrow footpath from the mountainous interior of Jhanjheli, due east of Shikari Devi. They were eight: an elderly married couple, their son and his wife, and their two daughters and baby boy, who had just had his first head shaving ceremony (mund .ā ). The elderly woman's brother accompanied the pilgrims with a goat that he led on a leash made of string. Taking care not to disturb them (we only found out about the relationship between the party members later), we observed them settling in at the edge of the platform.
Although our presence was acknowledged, it did not seem to impact the conduct of those in the temple to any significant degree. While the pandit, who had already met us the night before, was hardly surprised to find us on site, the indifference that the newly arrived pilgrims displayed towards our presence was both welcome and puzzling. It soon transpired that they had arrived on a rather pressing business and were thus occupied with concerns far greater than the unexplained presence of the camera-wielding foreigners in their midst. This unfussy attitude is also connected to the nature of Shikari Devi as a destination for pilgrims (yātrīs) from various backgrounds. In this respect, the hilltop temple provided a more open-ended setting for our enquiries than the starkly community-oriented temple at Chindi had, wherein engaging with the temple authorities was a prerequisite for filming.
A few minutes later, the pilgrims mounted the steps towards the shrine, leaving the pandit and his clients to pursue their ritual on the outer platform. The heads of the family entered first. The husband was hissing and panting as if trying to cool down, and his wife who followed him bore a countenance of grave concern. The uncle and the goat, visibly reluctant to climb the stairs, entered next and were followed by the rest of the family. Once inside the compound, the head of the family made a surprising move and entered the inner sanctum, assuming a position next to the priest. His wife faced him in supplication.
The hissing grew stronger as the husband started nodding his head in disapproval, wincing as if undergoing some kind of biting pain. While we would only comprehend the details of their conversation much later with the aid of translators from Seraj (the dialogue was in a strong local dialect of Western Pahari), it was obvious from the physical gestures and spatial setting that the husband was in fact speaking on behalf of the goddess or some other divine entity. The dialogue that ensued soon clarified the reason for the group's visit.
Wife/Supplicant: Mother, bless and protect my family.
Husband/Śikārī: Why did you come here?
Wife/Supplicant: Look at the baby, it needs help.
Husband/Śikārī: You made a mistake. Why did you use a machine to cut its hair? We won't allow that, neither will Kamru Nag. 16 Wife/Supplicant: How are we supposed to know these things? It's the Dark Age (Kālī Yug), we don't know anything.
Husband/Śikārī (hissing strongly): It is very bad. Why did you do this?! Wife/Supplicant: We apologized already. Now, fix it! The family had thus come to amend for the ritual transgression of shaving the newborn's head with a machine instead of the blade that is traditionally prescribed. The conversation between the deity and the wife developed further. The goddess demanded two vast sacrificial ceremonies (yajña) be made in honor of Kamru Nag over the next couple of years. The supplicant responded that she was not sure they'd have enough crops to finance such lavish ceremonies, but her pleas fell on deaf ears. The husband/goddess was adamant.
This discussion was accompanied by an exchange of substances between the goddess and her client. The goddess kept blowing into grains of rice and handed them over to the supplicant, ordering her to keep them "in the usual place" for the family's protection. 17 The wife counted the grains and grimaced negatively whenever the number drawn proved inauspiciously even. She kept demanding more grains until gaining an auspicious seven, which slightly relaxed her state.
With the matter for which the family had gone on pilgrimage resolved, one would have expected a return to normalcy. Just then, however, something remarkably interesting happened. The goddess turned to the mistress of the house with a stern warning: A new space seemed to have opened for addressing more pressing topics than the ritual shaving of the baby's head. Judging by his reactions, the person who was "drinking in secret" was none of other than the couple's son, who had shied away from most of the dealings till then and who was now actively avoiding eye contact with the party (Figure 7). Sensing that the debate between the husband/goddess and his/her wife/supplicant was heating, the temple priest (pujārī) broke his silence in a bid to assuage tensions over the drunkard son: "Don't be harsh. He is family (apne haiṅ). Just pray to the goddess to teach him manners (shudh-buddh)." A new space seemed to have opened for addressing more pressing topics than the ritual shaving of the baby's head. Judging by his reactions, the person who was "drinking in secret" was none of other than the couple's son, who had shied away from most of the dealings till then and who was now actively avoiding eye contact with the party ( Figure  7). Sensing that the debate between the husband/goddess and his/her wife/supplicant was heating, the temple priest (pujārī) broke his silence in a bid to assuage tensions over the drunkard son: "Don't be harsh. He is family (apne haiṅ). Just pray to the goddess to teach him manners (shudh-buddh)." The angry spirits somehow subsided and the husband/goddess asked the wife/supplicant to purify the goat with grains of rice. The wife complied and cast a couple of grains on the animal. The entire family turned its gaze towards the goat in expectation that it would shiver, a sign that it accepts to be sacrificed. It never moved. The angry spirits somehow subsided and the husband/goddess asked the wife/ supplicant to purify the goat with grains of rice. The wife complied and cast a couple of grains on the animal. The entire family turned its gaze towards the goat in expectation that it would shiver, a sign that it accepts to be sacrificed. It never moved.
Having concluded their dealings, the entourage exited the temple to resume the same path on which it had arrived. Once clear of the temple grounds, the troupe sat for a picnic next to a fountain. The wife's brother decapitated the goat with a single blow and sadly mused over the lack of food in its system ("it's a shame it died hungry") as he dissected it into parts. The meat was packed in a bundle to be cooked later and the liver placed over an open fire. Maggi noodles and flat bread chapatis were distributed for all (including the unsolicited filming crew) along with salted pieces of liver.
The head of the family, no longer hissing and heaving, kissed and fondled the baby. The wife chewed on the liver and more food was passed around. The prodigal son picked the severed goat head from the pool of internal organs at the site of the killing. Relieved of no longer having to face his drinking habits in public, he joined the family in packing bags and preparing to depart. They left us soon afterwards.
The rituals at Shikari Devi shed further light on the beliefs and practices linked with the goddess in Seraj. As a pilgrimage site that is far removed from settled society, the hilltop shrine caters to different clients in various ways. The rituals that we witnessed were consequently far less institutionalized than those encountered at Chindi, where the goddess's role as sovereign mandates periodic "social work" in the form of exorcism sessions and "banking services" to villagers far and wide.
Surrounded by dense forests and impressive wildlife, the hilltop shrine provides a dramatic setting on which worshippers may project their inner worlds and fantasies. The lack of a permanent community also facilitates a more individualistic approach to the deity, from the Sanskritic ritual and vegetarian offerings that were administered by the Brahmin pandit of Mandi to the seemingly "spontaneous" instance of spirit possession and blood sacrifice by the highland villagers of Jhanjheli. The dialogues between the Pahari pilgrims evince the often reported (but rarely filmed) prevalence of spirit possession in family circles as a means for resolving emotionally charged conflicts.
Our documentation illustrates how the "official" business of ritual duties (the mund .ā "shaving rite") acts as a gateway for tackling domestic difficulties (the drunkard son). As the conversation between the husband/goddess and the spouse/supplicant makes clear, the couple had already discussed their son's drinking habits prior to the visit to no effect. The disapproval of a third, superhuman party (the goddess) is intended to breach the impasse in domestic relations by endowing the speaker with extraordinary authority, although cases reported during fieldwork indicate that this is not always successful. 18 The open-air setting of the temple and the overt discussion of family sensitivities also help ventilate tensions that could risk escalating in more quotidian circumstances. The seeming spontaneity in which the family's most intimate issues are raised thus enables the priest to intervene and diffuse the rising tensions. Additionally, while the "offender" may have been shamed in the process, the affair is ultimately contained in a way that preserves the family's unity. 19 That these matters unfold in the context of a private family pilgrimage and not in an institutionalized ritual setting such as the exorcisms at Chindi is significant, for it underlines the pervasiveness of spirit possession as a method for contacting the goddess in intimate circles beyond those of the community.
Finally, similar to the exorcisms at Chindi, the goddess's universal acknowledgement as a virgin-child seems to be limited to the realm of dreams. In practice, when her followers seek to communicate with her, the goddess is clearly addressed as a mother. The split between the benign, virginal purity of the goddess as child and the wrathful, bossy mother whose maternal care is somewhat hidden behind the overbearing menace of an omnipotent matriarch thus remains central to the conceptualization of the goddess in and outside of settled society. This is consistent with the alleged rarity of the child-goddess's presence, whose encounters are brisk and fleeting like "the wind in the bazaar", in the dream-worlds of her believers' individual psyches. The goddess as mother, on the other hand, is ubiquitous, manifesting in village mediums and in family circles through the readily familiar practice of possession.

Epilogue: The Goddess Manifests
A strange silence set in once the family had left the grounds. 20 We lingered next to the fountain, and the smoky remnants of the fire made their way skywards in the cold forest damp. Then, quite unexpectedly, a soft jingle of bells came from the trees above us. Before long an inquisitive goat popped its head through the bushes and was staring at us intently. More goats joined and within less than a minute we had become surrounded by a herd of snow-white fur flowing down from the clearing above the ridge.
Fairly perplexed, we remained silent and observed. The goats crossed the picnic ground sniffing grassy patches in search of food, indifferent towards the blood-soaked spot where one of their own had just found its death. Retracing our gaze through this wondrously eerie scene towards the hilltop from where the goats had arrived, we suddenly realized that we were not alone.
A girl-child had appeared on the rock above us. No more than five or six years old, her hair disheveled and a set of tarnished crimson rags covering her body, the little shepherdess had followed her flock down the hill and positioned herself on a prominence above us ( Figure 8). Our gazes met. Startled by the unexpectedness of the encounter, neither of us dared speak. wondrously eerie scene towards the hilltop from where the goats had arrived, we suddenly realized that we were not alone. A girl-child had appeared on the rock above us. No more than five or six years old, her hair disheveled and a set of tarnished crimson rags covering her body, the little shepherdess had followed her flock down the hill and positioned herself on a prominence above us (Figure 8). Our gazes met. Startled by the unexpectedness of the encounter, neither of us dared speak. We gestured for her to warm up by the fire and she obliged us with a smile and a twinkle of the eye. Another shepherd girl, most likely the red-clad child's older sister, appeared a few minutes later. She unloaded a cloth bundle that she carried on her back, and they quickly changed into a more "respectable" set of clothes, covering their hair with handkerchiefs. Flashing a quick happy smile in farewell, the sisters had no sooner appeared as they were gone, following their flock into the thick of the forest below.
We kept silent for a long while after the encounter. Everything that we were told about the goddess seemed to be true. Here was a little girl dressed in red, who had chosen to appear before us for a fleeting moment. She communicated with us easily through signs and gestures even though she never uttered a word. 21 The power of her sight (darśan) and the meeting of her eyes were intensely present long after she had gone. The meeting was a complete, immersive, impenetrable, and, for lack of a better word, auspicious experience.
Was this mere serendipity, or were there higher forces at work? Was the child even aware that we had immediately recognized her as the goddess? As chance creatures on her path, were we not better described as part of her divine play (līlā) than as researcherfilmmakers? Was the fact that this was a Muslim Gurjar girl of any importance? Perhaps We gestured for her to warm up by the fire and she obliged us with a smile and a twinkle of the eye. Another shepherd girl, most likely the red-clad child's older sister, appeared a few minutes later. She unloaded a cloth bundle that she carried on her back, and they quickly changed into a more "respectable" set of clothes, covering their hair with handkerchiefs. Flashing a quick happy smile in farewell, the sisters had no sooner appeared as they were gone, following their flock into the thick of the forest below.
We kept silent for a long while after the encounter. Everything that we were told about the goddess seemed to be true. Here was a little girl dressed in red, who had chosen to appear before us for a fleeting moment. She communicated with us easily through signs and gestures even though she never uttered a word. 21 The power of her sight (darśan) and the meeting of her eyes were intensely present long after she had gone. The meeting was a complete, immersive, impenetrable, and, for lack of a better word, auspicious experience.
Was this mere serendipity, or were there higher forces at work? Was the child even aware that we had immediately recognized her as the goddess? As chance creatures on her path, were we not better described as part of her divine play (līlā) than as researcherfilmmakers? Was the fact that this was a Muslim Gurjar girl of any importance? Perhaps a message intended to teach us about the transcendent nature of religious experience? Our journey pressed on towards more temples and new encounters, but the encounter on the mountain had brought our pursuit of the goddess to an end: an act of grace to conclude an open-ended search for the divine. Our mission was completed.
This article investigated the means through which ethnographic film can contribute to the study of religion. Pursuing an approximation of religious experience among the West Himalayan Khas, it explored the discourses and ritual practices surrounding a regional variant of the pan-Indian cult of the goddess. As elsewhere in South Asia, the ethnographic material caught on camera evinced the pervasive hold ofŚaktism on Khas society, where it crystallizes around the polar manifestation of virgin-child (kanyā) and powerful-mother (mātā). Thus, if colloquial discussions of the goddess in interviews and conversations between locals (e.g., peasants, shop owners, religious officiates) indicated a decisive preference for the child-goddess among the believers who commonly encountered her in their dreams, the rituals caught on camera revealed that it was the goddess as a mother whom they more commonly encountered in public. Mediated through the regionally normative phenomenon of spirit possession, those speaking for the goddess embodied the mother with an authority that, somewhat paradoxically, stems from the purity of her virginal manifestation (cf., the discussion of kar pu, tapas, andśakti in the introduction, above).
By capturing the coexistence of the seemingly contrasting manifestations of the goddess on camera, the film renders the gap between the discursive and ritual expressions ofŚaktism intelligible without sacrificing its complexity. The patrons of the chai shop at Chindi, for example, were unanimous in describing the goddess as a child, but when the time came for exorcising demons, it was the goddess as mother whom they turned to. Similarly, the goddess of hunters was likened to a fairy-like joginī child, but when pilgrims sought her help, she was instinctively referred to as "mother" (mātā). The serendipitous conclusion of the film in an encounter with a Gurjar shepherd-girl breached this impasse by transitioning from the visual representation of ritual to the approximation of religious experience. While the manipulation of sight and sound that is integral to filmmaking plays an important part in conveying this culminating "leap" in documentary content, the meaningfulness of its message is equally reliant on the presentation of the ethnographic data that precede it in a culturally informed narrative; the shock value of "meeting" the shepherd-girl would be largely moot without the persistent descriptions of the goddess's form by film participants in the buildup to the final encounter.
In this respect, AVATARA (Harel 2020) demonstrates the capacity of ethnographic film for breaching and connecting disciplinary boundaries. Specifically, it brings the ethnographic fieldwork that underlies Social and Cultural Anthropology into conversation with the scholarly analysis of religious experience that informs Religious Studies. By insisting on a faithful rendition of colloquial discourse and ritual actions relating to goddess worship, the "appearance" of the girl after the sacrifice creates a conclusion that is both ethnographically accurate (the meeting is presented as it actually happened) and internally consistent with the conceptualization of the goddess by those who participated in the film (a virgin-child whose spontaneous appearance in dream states is often contingent on sacrifice). What the documentary analyzed here ultimately demonstrates is how a sensitive editing of ethnographic data that is attuned to local sentiments can both refine the scholarly understanding of social phenomena (by conveying the child/mother divide in goddess worship in a coherent, unitary manner) and instantiate emic conceptualizations of the divine (by bringing the goddess "to life" in the final scene).  Acknowledgments: The author wishes to thank the residents and participants of Karsog tehsil who for sharing their knowledge in the course of making the documentary AVATARA; Chetu Thakur and Mahesh Rajput for translating dialogues from the local dialect; Nadav Harel for documentation and discussions on religion and ritual; and Csaba Kiss for reading through an early draft of this paper.

Conflicts of Interest:
The author declares no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

1
For a detailed discussion of satī in the Rajput setting of North and Western India, see Moran (2019b, pp. 105-23). On satī in the area explored in this paper, see Section 3.2, below. 2 A couple of months later, we ultimately did find an alternative performance of the ritual that we were able to shoot (Harel and Moran 2018) and then analyze in a separate research article (Moran 2018). 3 For a useful introduction and discussion of the societal and phenomenological approaches towards religion, see (Pickering 1975;Proudfoot 1985), respectively. 4 For a lucid defense of religious experience as a viable analytical category that furthers the study of religion, see (Bush 2012). 5 It is unclear how Emerson arrived at these terminologically confusing definitions of "memory" and "speech", which are better translated as "colorful image" (chitr) and "hidden/secret one" (gupt). 6 The pace of cultural change in the area is visibly slower than in areas with better connectivity (e.g., Kangra, Kullu). Women in Outer Seraj, for example, were expected to cover their heads on festive occasions until fairly recently; Brahmi women wore red scarves (dhātu), while others donned black (Bhatnagar 1998, p. 143). 7 Rice and maize are harvested in the kharīf season (September-October). Wheat is the main rabī crop (April-May). 8 The princess was buried alive in a wall, echoing a universal folk theme (Dundes 1996). For an analysis of the Chamba tradition, see Sharma (2016). 9 My thanks to Alexis Sanderson for references on this topic (email communication, December 2013). According to Sanderson, this is the farthest east that a specimen of the Kashmiri type has been reported, indicating the medieval floruit of Kashmiri artisanry and culture had extended well into the eastern hinterlands of present-day Himachal Pradesh. For similar specimens from Kashmir, see (Siudmak 2013, pp. 380-98, especially plate 178 on p. 388, which dates to the beginning of the third quarter of the eighth century). At over a meter high, the stone image in Mamel recalls the near life-size bronze Vaikun . t . ha in the Laks . mīnārāyan . a temple at Chamba, which has been assigned to the ninth century (Pal 1975, plates 84a-c). 10 In 1987, some 30 buffaloes were offered in sacrifice in the course of the Navaratra festival (Harnot 1991, p. 101). 11 The eastern ridge is the domain of Dahmunī Nāg, while the western ridge that leads into Suket is the preserve of Mahu Nāg, alias "Rājā Karn . ". For more on this class of deities, see Handa (2004). 12 The cult of Cin . d .ī /Can . d .ī ka is especially powerful in central Kinnaur (Emerson n.d.,, but it is unclear whether the two goddesses share any affinities beyond their name. 13 On charisma as a central feature of Khas mediums, see Berti (2001). 14 For the film footage corresponding to this section, see https://drive.google.com/file/d/1Hj18Av1XD3-VUGhyjvdmGsFkYpw rDVSK/view?usp=sharing. Accessed 18 November 2021. 15 For the film footage corresponding to this section, see https://drive.google.com/file/d/1l9niHSkM_CQvGm4wfcDmrzB6FoD V_D8u/view?usp=sharing. Accessed 18 November 2021. 16 The head of the family was, in fact, speaking on behalf of two deities:Śikārī Devī and Kāmrū Nāg. The latter is a powerful weather god, whose temple is located at a few hours' march west of Shikari. 17 The transition from the powerful goddess who blesses the family's rice grains to the familiar husband who is intimately aware of "the usual place" at home where they should be stored is telling of the continuum of mental states entailed in spirit possession in South Asia. For a salient example from South India, see (Handelman 1995). 18 A friend from Kullu, whose name shall be kept anonymous for the sake of privacy, often recounts how his father used to become "possessed" and threaten his wife every time she asked him to stop drinking. He had consequently lost all faith in the authenticity of those claiming to speak for the gods. 19 For a powerful example of spirit possession surfacing issues to do with incest, see (Lecomte-Tilouine 2009). 20 For the film footage corresponding to this section, see https://drive.google.com/file/d/1SeXAdDdEZp0n1Wcfo5UWgFS3z9V 2Q-CP/view?usp=sharing. Accessed 18 November 2021. 21 A shepherd from the higher reaches explained to us that the "joginī", as he called Shikari Devi, can communicate with every person in whatever manner that is possible (language, signs, etc.). On the sign language (chommā) of yoginīs, see (Hatley 2007).