7

Śāktism, polity and society in medieval Malabar

Rich Freeman

Introduction

This paper is an attempt to contextualize a series of interesting historical and anthropological problems around textual remnants of Śākta cults in Kerala, focusing on the particular region of northern Kerala, or Malabar, proper. The goal is to provide what can only be a brief overview of a series of interconnections between Sanskrit and vernacular texts, and forums and practices of worship which survive in the region today, or which have a still-legible historical purchase among present communities of the region. After some three decades of fieldwork and study in the region, I have collected a good deal of material, both ethnographic and textual, which I hope to draw together into a larger monographic reflection on issues and questions introduced here in a necessarily summary fashion.

Many of the major historical questions I confront here have been stimulated and informed by the work of Alexis Sanderson. From my own interdisciplinary perspective as a textually trained anthropologist of the Kerala region I perceive two major, connected projects in Sanderson’s oeuvrethat I have found invaluable. The first has been his painstaking years of searching out, characterizing and ordering much of the largely unpublished medieval Śaiva material (including its allied Śākta strains), and his analysis of this literature in terms of actual practices and doctrines which can be usefully situated in regard to their potential historical, social, and political significance (e.g. 1988, 2007a) The second phase of his endeavors, as I see it, builds on this first, partly by way of comparative work on parallel or derived developments in Vaiṣṇava, Buddhist, and Jain circles, but more directly through a wide-ranging engagement with the inscriptional record, and serves to formulate general historical theses concerning the success and reach of what he calls the ‘Śaiva Age’ (e.g. 2009). This brings his work as a rigorous text-critical Indologist into convergence with more conventional historians of South India, particularly those concerned with the intersection of Indian polity and religious institutions (e.g. Stein 1994; Heitzman 1997; Champakalakshmi 2011), but, as he also suggests, and as I hope to demonstrate here, with anthropology. What has been most exciting for me is that many of the matters I have puzzled over and come to tentative findings on during my years of anthropological and literary research in the region of northern Kerala, have been refocused and often confirmed in unexpected ways by Sanderson’s way of characterizing an issue, or by his more specific textual findings.

In what follows, I will proceed by providing a sketch of some key texts and socio-historical problems in the Śāktism of Kerala, while trying to connect these with larger trends in Indian Tantric and Hindu developments. In many cases I will draw schematically on earlier published work and my dissertation, but in a way that reframes this as a kind of overview of what I hope to consolidate into a monograph on the Tantric tradition of Kerala.

Śāktic and Tantric Kerala

The Puranic syntheses in Kerala temples and performative cults have tended to elide the various influences of more discriminate lines of Śākta cults (e.g. the Trika, Kubjikā, and Krama, which we know were textually represented in Kerala from their manuscript evidence), into an amalgam which few have tried to probe, beyond the vague invocations of ‘Mother Goddess’ or ‘Bhagavati’ worship, as being generically Śākta. It turns out, however, that there are far more interesting anthropological and historical implications for the whole Śākta/Bhairava milieu than were suspected, once we begin to engage actual Sanskrit and vernacular texts, and work to situate them in their regional context, on the one hand, and place these latter in terms of the larger dynamics that Sanderson has charted out in trans-regional scope, on the other. I will begin by reviewing a few of the principal texts and social structural correlations I find relevant that pertain to all of Kerala, including Malabar, and then narrow my focus to more particular Malabar material, which I will also try to fit back into the larger patterns.

As with the attribution that many Kerala temples are nebulously ‘Śākta’, one of the striking descriptive usages of wider application is that Kerala Hindu worship is avowedly ‘Tantric’. The expositors and managers of this tradition, the modern Nampūtiri Brahmans, however, are quick to point out that this is merged in a ‘Vedic’ cult, meaning that, aside from drawing on Vedic as well as Tantric mantras, any association with Kaula practices of blood sacrifice, non-vegetarian or alcoholic offerings, and ritualized sexual practices or imagery is entirely absent (Unni 2006: 8). The Tantric attribution carries a particular social valency, however, in that many of these prominent Brahmans are/were also, in local usage, Tantris,¹ meaning that they exclusively control the hereditary rights and knowledge to perform the installation, renovations, annual festivals, and purification of all the major temples of Kerala which were divided up hereditarily under their jurisdiction. They furthermore appointed, and still ritually authorize and oversee, all priests who are engaged in the mere daily rites of worship, who are known in Kerala as Śānti(-kārar), and are generally drawn from socially and ritually subordinate Brahman families. I believe this distinction was historically significant in both social and ritual terms to which I will return.

Whatever the social constitution of Nampūtiri Brahmans historically (and I suspect there was a far greater degree of historical diversity in their origins and community than is usually acknowledged), they have presented themselves as the single, more or less homogeneous and uniquely authoritative Brahmanical tradition of Kerala since the later medieval period. This regional ritual consolidation was reflected textually in the fifteenth-century handbook of temple installation and rituals, the Tantrasamuccaya, which is the standard, authoritative manual regulating temple ritual from one end of the state to the other. Its importance was demonstrated and authority shored up through two layers of Sanskrit commentary, within the next generation and apparently from within the same family (Ramaswamy Sastri et al. 1945: iii–iv; Narayana Pillai 1951: 3–4), followed by a Malayalam commentary in the eighteenth century, which was abstracted into a free-standing and complete handbook and translation. It is the latter, the Kuḻikkāṭṭu Pacca, which was updated and published in modern times to circulate throughout Kerala as the contemporary standard which closely adheres to the Tantrasamuccaya.

Textual interventions

The Tantrasamuccaya attempted, much as in the pañcāyatana of the Smārta tradition (Bühnemann 1988: 49–51), to reduce the pantheon of the earlier Kerala temple cult to a mere seven gods: Viṣṇu, Śiva, Śaṅkaranārāyaṇa, Durgā, Subrahmaṇya, Gaṇapati, and Śāstṛ. Ritually, the work looks very much like an Āgamic handbook of the Tamil country, except for its pointed melding of Vaiṣṇava and Śaiva worship, by fronting Viṣṇu as the paradigmatic deity in a clearly Śaivite ritual frame, and by installing the constitutionally composite Śaiva-Vaiṣṇava deities of Śaṅkaranārāyaṇa and Śāstṛ. That this work represents an attempt at a drastic reduction of the ritual life that preceded it seems apparent both from the textual production that preceded and followed it.

The most prominent work in circulation before it was the compendious and encyclopedic Tantric digest, the Īśānaśivagurudevapaddhati. This latter work is evidently composite, with older parts in a dominantly Śaiva Saiddhāntika stream that probably go back to the twelfth century, yet which nevertheless incorporate significant Vaiṣṇava and Śākta cults of worship. While the prevalence and concentration of the manuscripts would make one suspect it is of Kerala provenance, as well as featuring cultural items particular to that region, the medieval overlap and entanglement between what we think of as Kerala and Tamil Nadu today (Freeman 1998), as well as the mobility of the lineages propagating such texts, means it could just as likely represent an originally Tamil tradition that was later indigenized and redacted from several sources in Kerala. Particularly significant in this regard is the large Mantrapāda, in which, as Goudriaan has pointed out (1977: 158–160), there is broad overlap in subject and language with the undeniably Kerala work, the Tantrasārasaṅgraha, perhaps of the fifteenth century or earlier. The latter is a digest of mantravādafor toxicology, demonology, and kāmya worship with a pronounced Śākta inclination, but this merely serves to intensify the already Śākta elements of Īśānaśiva’s paddhati. In the absence of a systematic examination of these relations, and with the paddhati inadequately edited or researched, derivation from or influence on the shared material could have gone in either direction, with the possibility that both could have drawn on a tradition of applied Śākta rites associated with a common source called Nārāyaṇīya which figures in both textual corpora, as Goudriaan has noted (1977: 160).

So in any case, if Īśāna’s compendium reflects the dominant stream of the earlier Tantric tradition of Kerala, the Tantrasamuccaya certainly represents a narrowing of that tradition, particularly with regards to the Śākta deities and components. The only female deity prominently featured in the Tantrasamuccaya, Durgā, is assimilated to the Vaiṣṇava structure of mantric worship, and shorn of the more sanguinary ritual or iconographic associations one would expect of the Puranic war-goddess.

That the systemizing and narrowing effect of the Tantrasamuccaya on the pantheon invited a compensatory move is evident in a subsequent textual production, the Śeṣasamuccaya, composed in the next generation, again within the same family (Narayana Pillai 1951: 3–4). As an explicit supplement to the Tantrasamuccaya, it took up a remainder (śeṣa) of deities who required a commensurate textual treatment to accord with Kerala practice. Aside from four male gods who were present in the Kerala temple campus,² these additional deities comprised: (1) the potentially Śākta pantheon of Sarasvatī, Śrī, Pārvatī, Jyeṣṭhā, and Bhadrakālī; then (2) a generic pantheon of the Seven Mothers (with accompanying Vīrabhadra and Gaṇapati), and Kṣetrapāla; but then (3) another, discrete complex, characterized as the demon-slaying Bhradrakālī named Rurujit, joined by Śiva, Kṣetrapāla, the Mothers, and allied deities, and the Lokapālas in their own right, headed by Indra.³

Sarasvatī, as the embodiment of speech, and therefore mantra itself, was a principal vehicle of the Śākta dispensation, especially as the syllabary (lipis) installed in the worshiper’s body, as we shall encounter later. Śrī, though originating as Viṣṇu’s spouse, readily took on attributes of a Śākta entity in South India, as can be seen from the Lakṣmītantra. Pārvatī, as Śiva’s principal spouse, was readily assimilated to the embodiment of Śaktī herself. Jyēṣṭhā as the elder sister of the Goddess, is popularly associated with negative feminine power and sorcery, and Bhadrakālī is the fierce, potentially violent goddess, created by Śiva for the destruction of demons, who reinstitutes all those sanguinary and sexual associations the Vaiṣṇavized Durgā seemed to earlier expunge. Indeed it is Bhadrakālī (in various forms and modes, including Cāmuṇḍā) who becomes the more prominent liturgical representative of the independent, martial offspring of Śiva, representing those ‘Bhagavatis’ for which Kerala is famous.

As for the ubiquity of the Mothers in Kerala temples, even those dedicated primarily or exclusively to male deities, normally feature the Seven Mothers, in a row, flanked by Vīrabhadra and Gaṇapati on their ends, rendered in a low-lying modular plinth (pīṭha) with all these deities represented by identical little surmounted liṅgas. These Mothers are, of course, the seven consorts or Śaktis of the principal gods, as their named feminine counterparts. Set outside and to the south of the central shrine, as part of the inner circle of offering stones, they are given balis as a regular part of worship, everywhere (Figure 7.1).

Figure 7.1  Nampūtiri Brahman novitiate priests offering bali as part of regular worship to the standardized, modular representation of the Seven Mothers, with accompanying Vīrabhadra and Gaṇapati on the southern side of the main shrine

Kṣetrapāla, the fierce guardian of the temple, is conceptually part of their entourage, but he has his own offering stone, in the northeast corner of the temple’s outer courtyard, within the main walls.

What seems initially odd in terms of the format of the Śeṣasamuccaya as given above, is the last sequence of partly repeating deities, comprising the special demon-slaying Bhadrakālī, Rurujit, along with Śiva, and Kṣetrapāla, and various forms of the latter gods. It turns out these constitute a special paradigm and configuration of temple layout and worship, of which the formerly discussed little modular group is merely a reduced and schematized rendering adapted to more ordinary temples. For while the first two groups of gods are all treated in the first six chapters of the Śeṣasamuccaya, the seventh through the ninth chapters enjoin a coherently discrete ritual program (vidhāna) dedicated to Rurujit, the form of Bhadrakālī (or Cāmuṇḍā) who is independently installed in her own shine, configured with another separate one for Śiva, an accompanying Kṣetrapāla, and a fully iconically differentiated and free-standing set of the Seven Mothers in a hall (śāla) flanked by Vīrabhadra and Gaṇapati. These chapters deal sequentially with this special kind of temple in three separate paṭalas devoted to installation (pratiṣṭhā), to the liturgy and rites of worship, and to the festivals and renovations. The importance of this discrete section is that it clearly represents a sustained and relatively coherent unit of Śākta intrusion into the Kerala pantheon. As we will see, it reaches back into earlier strata of Śākta goddess texts, on the one hand, but has also left its unique historical impress in a chain of related, royal temples throughout the Malabar region, on the other. Before taking up these matters, however, and by way of background, I want to briefly review the social configuration of castes, ritual personnel and their forums of worship, and their relation to the traditional chiefs of the region. This will again be a necessarily cursory sketch of these relations as they bear directly on the matter of Śākta cults, since this should give us a better grasp of the context in which these textual remnants were developed and deployed.

Ethnological reflections

While Nampūtiri Brahmans present themselves as a consolidated and coherent caste from the outside, they were internally differentiated into a complex gradation of ritual and socio-economic status-groups with many practical boundaries to intermarriage.⁴ What was most prominent about them from an inter-caste, anthropological perspective was their famous symbiotic connubial relations of sambandham with castes under them. This institution of sambandham dictated that only the eldest Nampūtiri male in any sibling cohort could legitimately marry a Nampūtiri woman and reproduce the next generation. All the younger siblings had sambandham sexual liaisons with lower-caste women of clean, Śūdra grade (temple servants, royals, chiefs, and militia), whose matrilineally reckoned offspring were legitimate inheritors in their own castes, but had no claim against their Brahman sires or their estates. The women of these matrilineal castes, the politically and socio-economically dominant castes of Kerala, could freely form multiple liaisons with any men of their own or higher castes and thus legitimately reproduce the social identity of their own castes through their matriline. These latter castes fell into the two larger caste-groupings generally known as Ampalavāsi, the temple servant castes, and the Nāyar, which included most royals, political dominants, and soldiery of the land.

As with any Brahmans, the Nampūtiris original claims to ritual suzerainty came from their custodianship of the vedic knowledge and rites, and they have been made famous by the fact that a restricted circle of families retained this knowledge in an unbroken line down to the present (Staal 1984). In terms of the modern narratives of Kerala historians, however, Nampūtiris seem to have been endowed with a mysterious power which supposedly allowed them to engineer the ‘feudal’ order of sambandham that at once preserved their land holdings and gave them license to sexually predate on the hapless (and implausibly gullible) dominant castes under them. Credit goes to Louis Dumont, however, who, in his survey of Nampūtiris in their local and pan-Indian context, registered what they gave up in this process as well. Since only legitimately married Brahmans might keep the vedic fires, perform the vedic rites, and secure their place in heaven through legitimate ritual heirs who carry these on, Dumont noted that the majority of Nampūtiri males would seem to have been individually faced with a kind of religious disenfranchisement in this arrangement (Dumont 1983: 112–113, 123–125).⁵ Faced with such impediments to the normal notions of vedic entitlement and ritual status, I would suggest that the alternate tantric order of installing deities in temples and performing pūjās would seem to offer an obvious socio-religious compensatory validation. In fact, in reproductively staffing both temples and royal courts with their consorts and offspring, these institutions became a kind of surrogate, religiously extended family, a kind of tantric clan which compensated for their vedic debility, yet continued to draw on fraternal and ancestral ties to a vedic patrimony. From the perspective of the martial and royal castes themselves, the predominantly feminized cults of their militant, chiefly goddesses tapped the status of Brahmanical (vedic) religion, but underwrote this masculine formal authority with the claims to their own greater Śākta powers of sexual generation, psycho-physical energies, and bellicose political capacity. While I have charted out something of the complex imbrication of these religious ideologies through the vernacular works of their hybrid, Maṇipravāḷam literature which I cannot review here (Freeman 2003a, 2003b, 2011), I hope the reader can nevertheless appreciate my basic point on the social structuring of the caste-configured religious order. At a basic structural level, I want to suggest that the systematic correspondences between the caste- and gender-skewed reproductive order of upper-caste society, and the avowedly tantric religion which permeated the whole, with the Brahmans as Tantris and their martial and royal caste consorts as predominantly Tāntrika/Śākta initiates and devotees, cannot have been mere happenstance. The working out of this historical scenario requires a more complex analysis of the caste-histories and structures involved, which I must defer here. A further complication comes from these structures having been heavily overwritten, both through a later, largely pre-modern, Smārta reformism, and then the colonially inspired caste reformism of modern, regional historians entangled with their own caste-conditioned cultural politics. The religious ideology that features in most of these historical narratives is a vaguely formulated bhakti movement which draws selectively on its medieval Tamil Śaivasiddhānta and Śrīvaiṣṇava counterparts with little sense of the pan-Indic and varied Tantric historical developments which significantly informed the doctrines and practices of temple culture in South India. The critical interventions of Sanderson and other Sanskrit scholars like Dominic Goodall (1998, 2004), showing the lines of historical derivation from pan-Indian sources into south India have yet to really impact Tamil scholarship on these matters, let alone the extension of related historical influences to Kerala. This paper attempts to inaugurate something of the latter sort with regard to a fairly clear line of Śākta development in Kerala.

A major inroad into these problems at the most general level of the history of tantric practices has come from understanding that the ‘classical’ South Indian temple and priesthoods emerged socially and doctrinally by stages in the transformation from originally private cults of initiates undertaking individual worship for personal spiritual goals, into the consecration of permanent images installed in temples for public worship.⁶ In this way, whatever the earlier nature of public worship and shrines may have been in the South, these were successively colonized by Saiddhāntika and other tantric schools to generate the medieval temple culture we think of as classically ‘Southern’. As Sanderson has demonstrated, the important class of early pratiṣṭhātantras reveals the critical intermediate stage in which Śaiva officiants had expanded their role to regularizing the installation (pratiṣṭhā) rites of permanent images for worship in temples, but left the daily rites of worship and interaction with the public itself to other functionaries (2009: 274ff.). It is this historical stratification which I am convinced is sedimented into the Kerala system, between the different classes of the Nampūtiri formation, whereby Tantris, as learned ācāryas (in the language of the Tantrasamuccaya), have higher rights to install, ritually regulate, and perform annual rites of festivals, renewal and renovations in temples, but delegate daily pūjās to the subordinate Śāntikārar, whom they seem to have historically initiated for their role.⁷ The subordinate status of the latter is seen in the contrastive pairing preserved in this terminology, as well, for śānti rites of ‘pacification’ clearly suggest a regime of propitiating lower, quasi-demonic godlings of the temple, compared to the foundational and fortifying exercise of tantra, as the science of loftier image-divinization in the Kerala context.⁸

In returning to the more specifically Śākta complex of the Śeṣasamuccaya introduced earlier, we can see this same historical dynamic of accommodation more clearly evidenced in the social disjunctures of officiants brought together around the cult of Rurujit. For it turns out that the worship that was historically connected to this textually preserved system (as either its progenitor or legatee)⁹ is still flourishing in Malabar. I was intrigued in my original work around the royal shrine of the erstwhile chief of Nīlēśvaram, at Mannampuṟattu Kāvu, where I came to witness the prominent festival of spirit-possessed Teyyam dancers, to find an avowedly Śākta priestly caste I had never heard of in daily charge of the temple. Traditionally, this caste of Piṭārar (called Mūssatu, further to the south), worshipped the royal goddess inside the temple with the blood-sacrifice of cocks, and offerings of liquor, and the flesh as naivedya, as part of the regular service on Tuesdays, Fridays, and at other times of special request.¹⁰ In other respects, these Piṭārar are just like Brahmans, being invested with the sacred thread, studying their own śāstras for worship in Sanskrit, and even having their own temple servant castes for drumming and ritual assistance, with whom they engaged in sambandham relations, traditionally, exactly like Nampūtiris. The Nīlēśvaram Rāja, however, also supports another principal temple, the major Śiva temple of the town, as a regular Nampūtiri-staffed establishment conforming to the normal Kerala pattern of worship that we find in the Tantrasamuccaya. The goddess is considered the consort of the royal Śiva, receiving a garland from his temple daily, and during the Pūram spring festival, their portable images (tiṭampu, or utasava-mūrti), still process from their respective temples to meet at the royal tank where they bathe together. It was reported to me that married couples also bathe together with them, as a kind of spring rite for auspicious marital life and fertility.

On the occasion of the Teyyam festival at Mannampuṟattu Kāvu, many of the principal castes of Nīlēśvaram send their own shrine officials to the goddess temple, where they gather in the outer compound, many of them undergoing ritualized spirit possession, to participate in the twinned Bhagavatis elaborately costumed, possessed Teyyam worship, along with the Teyyam of Kṣetrapālan. The major factions of the martial, Nāyar castes, and their ritual officiants, are prominent in the managerial membership of this, their former king’s, royal shrine, as are the toddy-tappers, the Tiyyas, who daily supply liquor for the goddess worship. Rather anomalously, as it seemed to me then, the installation, renewal, and annual purification and certain festival rites, are nevertheless performed under a Brahman Tantri, and on those occasions, the Piṭārar have to vacate their office and turn ritual jurisdiction briefly over to him. While I was unaware then of the full historical context or textual import, even in my dissertation I wrote of this temple and its priests as clearly representing a separate, Śākta order of wide social significance, more oriented towards power than towards the purity of the Brahmanical order and its Śaiva temple (Freeman 1991). The brief episodes of Brahman custody also seemed then, as now, designed somewhat anxiously to demonstrate the formally superior claims of Brahmans over the demonstrably more powerful and popular Śākta complex. In light of earlier observations, what we clearly have here is a similar division between the Tantri and the Śānti offices, but where these latter are in fact of different caste, socially differentiated and ritually specialized for Śākta rites.

Subsequent research, travel, and textual study have now made it clear that the royal shrine in Nīlēśvaram, and the whole chain of related royal shrines of various local kings and chiefs southward to Calicut, and beyond, undeniably represent the temple-type, cult, and liturgy of the Rurujit-vidhāna sections of the Śeṣasamuccaya.¹¹ The layout of the royal temple of the Kōlattiri Rāja that I visited at Māṭāyi Kāvu corresponds in its constituent shrines, deities, and their orientation to the description in the opening of the seventh paṭala of the Śeṣasamuccaya on pratiṣṭhā (Figure 7.2), as does the temple at the site of his royal martial gymnasium, Kaḷarivātukkal, 17 kilometers to its south.

Figure 7.2  Rough sketch plan giving major shrines and their relative orientation in the Kōlattiri’s royal temple of Māṭāyi Kāvu. This accords with the Rurujit-vidhāna incorporated into the Śeṣasamuccaya

I also later confirmed with one of the chief Piṭārar of Nīlēśvaram, by going through the eighth paṭala on pūjā with him, that his pantheon and liturgy corresponds in most of its essentials to the Rurujit chapter on pūjā of the Śeṣasamuccaya. Of course, the seventh chapter on pratiṣṭhā, and the ninth on festivals, renovations, etc., would only pertain to the Brahmans who supplant the Piṭārar on these occasions, though it is doubtful that any of them follow the Rurujit-vidhāna in this context, as I explain below.

Socially, the Malabar chiefly patrons of these temples intermarried and intermittently warred on each other, centered on two principal kingdoms: that of the Kōlattiri Rāja, whose original capital and nearby main temple were some 30 kilometers south of Nīlēśvaram (near the famous coastal mountain of Ēḻimala); and that of the Sāmudiri or ‘Zamorin’ of Calicut, another 65 kilometers south of there. All nine of these major chiefly temples, from Nīlēśvaram to Calicut, north to south, were and are staffed with the same clan of intermarrying Piṭārar or Mūssatu priests and their Ampalavāsi assistants. Heading further south from Calicut, there were an additional five related temples, reaching all the way down into central Travancore, and including the famous royal Cēra temple of Kodungallur Bhagavati, roughly midway between Trichur and Cochin. While these temples are reported to have had the same Śākta installation, liturgy, and priesthoods historically, except for Kodungallur, they have reportedly all reverted to generic Nampūtiri rites and priestly control in the modern period, so I will confine myself mostly to the Malabar complex.

Alexis Sanderson’s main focus in his political treatment of the ‘Śaiva Age’ is naturally with prominent kings, those who left a noteworthy inscriptional record, had associations with prominent Śaiva Rājagurus, and undertook major temple and palace building projects (2009). But he further surmises, rightly, as I believe this Malabar material shows, that similar socio-political processes diffused through these cults more widely into the smaller chiefdoms and with a comparable reach deeper into society. Indeed, the higher political echelons would have necessarily relied on the latter infrastructural relations for their support. Relatedly, from the Brahmanical perspective, Sanderson has also shown with regard to the Paippalādin Atharvavedins of Orissa how the priestly formation adapted a Trika- and Kālīkula-inspired Śākta cult into their ritual repertoire, under the attraction of royal patronage (2007b: 236ff.). In that context, he has also noted the presence of a parallel cult of Bhadrakālī in South India, based on the Brahmayāmala tradition of Śāktism, derived at some remove from the older north Indian text of the same name (ibid.: 277–278 n.141). It is in fact this southern Brahmayāmala which seems to have been the ultimate source for the cult of Rurujit that is incorporated into the Śeṣasamuccaya and informs the whole royal cult of Malabar Śākta worship. Finally, Sanderson has also suggested in the same article that the mantric digests of the Prapañcasāra and Śāradātilaka were composed in the same Paippalādin milieu of Orissa (ibid.: 230–233). The former of these two textual traditions (but not the latter) turns out to be directly relevant to the Kerala material as well, and in certain regards highlights the whole Nampūtiri imbrication with Śākta cults, as we shall see.

Accordingly, in the remainder of this paper, I will try to broadly sketch out a number of intersecting social and ritual dimensions of the worship of Śākta and related deities in Malabar. This will entail surveying how Malabar chiefs articulated puranic goddesses and their cohort with local histories and Śākta cults and officiants, how Brahmans positioned themselves in negotiating between their institutional interest in Āgamic temple culture and the challenge from tantric Śāktism, how tantric yogis made their appearance and their accommodation with Brahmans and the royal and caste order, and finally, how one bit of material culture around the charged issue of blood-offerings demonstrates an accommodation that is diagnostic of the complex layering of both doctrine and caste around the issue of ritual status and purity.

Polity and Puranic Śākta goddesses

As we have seen, the Kōlattiri Rāja’s tutelary shrine of Māṭāyi Kāvu provides an exemplary architectural realization of the Rurujit-vidhāna in the Śeṣasamuccaya. While the name Rurujit, however, clearly implies that the goddess is the conqueror of the demon named Ruru,¹² this demon is virtually unknown in Kerala. While the worship of this goddess, who is variously known within the liturgy by names and epithets like Caṇḍā Kāpālinī, Aghorī, and Bhairavī that reach back into the Brahmayāmala texts, including the old northern one (cf. Hately 2007: 8–9), she is known in local performance and in her local Puranic persona, including most prominently in the southern Brahmayāmala, as the killer of the demon Dāruka.¹³ This is the form of Bhadrakāḷi, of pan-Kerala scope in Sanskrit and Malayalam texts, who hearkens back to the royal Cēra tutelary in her temple at Kodungallur. At Māṭāyi Kāvu, however, we find a common pattern of localization that I have repeatedly encountered: the demon is assimilated to a former nemesis of the local kingdom, and the Goddess, after killing this enemy, is pacified and installed as the royal protectress. Though this demon-enemy has the generic named Dāruka, at Māṭāyi Kāvu the site of his old fortress and the scenes of the battle are all localized in folklore and programs of ritual performance. For at Māṭāyi Kāvu, as at Nīlēśvaram, the royal goddess is annually incarnated in the temple’s outer compound in the elaborately costumed and spirit-possessed vehicle of a Teyyam dancer, who sings his lineages as orally preserved liturgies of the goddess, after which he interacts with worshippers ritually and verbally as her living presence. In this way, the king’s royal goddess reaches out into the common classes, and she has spread in this manner (as did scores of other Teyyam deities from various sites) through the whole region of the Kōlattiri’s domains.

While the Dāruka narrative occurs in brief form in at least some versions of the Liṅgapurāṇa (1906 [1857]: 138–139), we need a thorough study of its textual lineage in order to trace either its origins from, or adoption into Kerala. The slayer of Dāruka is further the puranic goddess whose myth is encapsulated in the inauguration of the southern Brahmayāmala tradition, where her creation by Śiva trumps the powers of the Seven Mothers as Ekavīrā, the ‘Solitary Heroine’.¹⁴ I have found a long Sanskrit version of this myth in what most likely comes from northern Kerala, purporting to be part of the Mārkaṇḍeyapurāṇa, and positioned to follow in that text just after the famous Devīmāhātmya (Tirumumpu 1975). Although it is in grammatically sound Sanskrit, it includes several instances of the word kūḷi for ‘demon’, which only occurs in Malayalam and medieval Tamil. While the work’s frame story is narrated to a king of Kashmir when he was hunting in the Daṇḍāraṇyaka forest, its action around the demon Dāruka devolves from Gokarṇa, on the west coast, the northernmost outpost of what was Brahmanically reckoned as Kerala.¹⁵ It therefore seems highly likely to have been composed near the region where it is extant and published, i.e. the far north of Malabar, bordering on the Tuḷu territory.

Cultically speaking, the text shows very clearly the impetus to incorporate Śākta esoterica into the puranic frame of the war goddess. The Kālī of this text is herself a yoginī, and a kāpālinī who carries the skull bowl with which she eventually drinks the blood of Dāruka as it froths from his chest when she tears him open. She has also associations with foresters and the wilds. When she plucks out Dāruka’s heart, roasts it in the fire of her eye, and eats it, she is compared to a hunter’s wife, digging up a forest tuber.¹⁶ And yet, her ferocious powers are clearly derived from the application of textually warranted Śākta traditions. For a run of sixteen of verses in her praise encode into their first syllables the fifteen-syllable kādi-version of the Śrīvidyā mantra, and add a final śrī(ṁ) beginning the sixteenth verse, for the ṣoḍaśākṣarī, the sixteen-syllable version of that vidyā.¹⁷ The goddess is then further explicitly praised as being Kubjikā, and again as Śrīvidyā and Tripurasundarī.

In social terms, there is a recurring narrative motif in most Kerala versions of the goddess myth which I have argued reflects centrally the tension between the martial castes and the Brahmanical temple order that we have seen as the central fact of kinship underlying the reproduction of these castes and their integration into the Tantric order of worship. In these myths the goddess, created by the Śiva of the Brahmanical temple order, is so crazed with blood-lust after she has killed Dāruka that she advances to attack the temple or Kailāsa itself. In order to pacify her, Siva turns two of his offspring or servants from within the temple (in the case of this Dārukavadha text, Gaṇapati and Nandin) into wailing infants and orders them to lie in her oncoming path. Their cries rouse her maternal instincts, milk flows to her breasts, and as she takes them up and proceeds to suckle them, her rage is gradually quelled in the course of nursing them. In addition, the Brahman ṛṣis are urged to sing her praises, and temple servants to produce music and dance for her, and this completes her domestication which gains her admittance into the outer precincts (gopura) of the temple. We could hardly ask for a clearer set of parallels to the ambivalent relations of martial servitude, connubium, and hybridly popular rituals by which, from the Brahman perspective, the dangerously violent and impure, yet necessary, matrilineal military castes were incorporated into the temple cult under the person of their goddess.

The royal cults of actual Malabar history reflect both this puranic thematization, and the hybridly Śākta cultic content of the Sanskritic Rurujit-vidhāna, as seen in the practices and liturgies at the vernacular level. In one of the Teyyam liturgies of the royal goddess, it is Kṣetrapāla and Vīrabhadra who are transformed into children to nurse at the Devī s breast, for these are the actual deities installed inside the temple, likewise manifested in Teyyam performances in the outside compound. In another such song, we see a more surprising instance of the osmosis by which esoterica of the temple cult passes into the liturgical domain of the possession deities. One of the diagnostic features of the Rurujit-vidhāna, coming from the Mātṛsadbhāva texts on which it is based, is the presence of ten differently named Kṣetrapālas as guardians of the directions (the four cardinals, their intermediates, the nadir and the zenith).¹⁸ While in the folk-imagination the guardian godlings of temples shade readily into demonic collectives, in recollection of the earlier narrative of Śiva creating children to pacify the goddess, there is a further carry-over from the liturgical presence of ten Kṣetrapālas:

As Kāḷi, creating havoc, approached Kailāsa in rage, in his fear,

The lord of Kailāsa deftly took his gaping, howling demons,

And in the form of ten infants there quelled her rage.¹⁹

The protean nature of these martial, guardian, quasi-demonic beings allows them to mediate across caste and political boundaries, forming various sub-pantheons that can alternately merge or individuate, constellating around the main dyad of Śiva and the Goddess (though even the Goddess tends to bifurcate, as we shall see). The very formation of the sub-kingdom of Nīlēśvaram, from an allied attack on the region by the Kōlattiri to its south, and the Zamorin, further to his south, exemplifies this. In the mythology of these events, an alliance of the gods Kṣetrapāla, Vīrbhadra, Vēṭṭakkarumakan, and Śāstāvu (Śāstṛ) ventured north and conquered the various demon-kings of that region. In historical fact, it was the lineages of Nāyar warriors represented by these tutelaries who conquered Nīlēśvaram, settled there, and established temples to these deities, often on the sites of earlier temples, and in association with goddesses of those sites. As these Nāyar transitioned from soldiers to chieftains, their deities could similarly morph from being lowly temple guardians, commensurate with their common status in their original territory, to higher, fully enshrined deities. Thus we have two major temples to Kṣetrapāla, and one to Vīrabhadra under three of the major Nāyar lineage-formations near Nīlēśvaram, each in a complex alliance with the earlier goddesses. At Nīlēśvaram itself, the center of the new kingdom where the Rāja established himself, the Dāruka-template is worked out in an illustrative local variant. Being afflicted by an evil demon-king named Mannan (meaning simply ‘king’ in old south Dravidian), his subjects eventually petitioned the Kōlattiri’s goddess at Māṭāyi Kāvu to liberate them. She indeed came north, killed Mannan at the site of his old palace-temple, and established herself there in the temple of Mannampuṟattu Kāvu.²⁰ Since she was the goddess of Māṭāyi Kāvu, the Kōlattiri naturally sent up a lineage of Piṭārar priests from his own tutelary temple to worship her according to the proper Śākta rites of her original shrine. There are actually two goddesses enshrined at Mannampuṟattu Kāvu, the principal, eight-armed Caṇḍayogeśvarī, who is worshipped more or less according to the Rurujit-vidhāna, and on the same pīṭha, a silver Bhadrakālī, who is reportedly worshipped according to her liturgy in the main part of the Śeṣasamuccaya. Correspondingly, there are also two Teyyams to the different goddesses that are also performed for the Teyyāṭṭam festival, in the temple’s outer compound, one of whom represents the Kōlattiri’s tutelary who conquered Mannan, while the other represents the earlier chiefly goddess of Mannan himself. In addition, Kṣetrapāla himself is incarnated as a Teyyam, as the warrior-guardian son of his Goddess-mother. The spirit-possessed Teyyams have offerings of huge kalaśams of toddy and cock-sacrifices offered to them outside the temple, while traditionally, the Cāmuṇḍā inside had blood-sacrifice and liquor offerings, as well. So as a simultaneous orchestration what we see, both at Mannampuṟattu Kāvu, and the Kōlattiri’s royal shrines (and the others through northern Malabar), is an inner temple-cult of Śākta, Sanskritic rites, and an outer cult of possessed folk-worship coordinating the martial and lower castes all under royal patronage, and framed in local, historical versions of Kerala-wide Puranic charters of conquest and divinely sanctioned rule. This was the way Tantrism and Śāktism functioned socio-politically in medieval Kerala, and this accords very well, it seems to me, with the general principles that Sanderson has laid out for the broader Śaiva-Śākta pattern. While I have written at length of the Brahmanical, royal, martial, and lower-caste articulations of Teyyam deities throughout these realms (Freeman 1991), what I now realize is that the temple-internal, royal cult which ordered these realms was this Rurujit-vidhāna Śākta cult, which even the external, Teyyam manifestation of it made clear was for royal purposes. The royal Teyyam liturgy itself praises the Goddess as the ‘essence of the ṣaḍādhāra’ in Tantra-yogic terms, as the ‘syllabary of the fifty-one’ of Sarasvatī, from which mantras derive, as the ‘nine-syllabled’ mantra of the Brahmayāmala, and the ‘sixteen syllabled’ mantra of the Tripurasundarī cult. This goddess was generated ‘to destroy the lineages of enemies’, in ‘great affection for the king’ as ‘the Kerala hero who would not falter’, and she was the personification of ‘the dīkṣā that was enjoined to protect the entire lineage of the Kōlattiri king’ (Viṣṇu Nampūtiri 1993: 26, 39, 42, 49).

Lineage of the Śākta officiants

In turning to the principal Śākta officiants themselves, we confront a historical puzzle. While we do not know the origins of the Piṭārar, or the circumstances of their entering the cult of Malabar’s royal shrines, there are nevertheless clues to their tantric milieu. The name Piṭārar is generally thought to be a vernacular form of bhaṭṭāraka, which is phonologically quite plausible, and contextually fits very well as an extremely common title of Śaiva and Śākta officiants in medieval India. Indeed the earliest references to Piṭārar from twelfth-century Kerala refer to them as looking after temple matters (śrīkāryam) with their students (śiṣyar), as involved with Yogis (cōkikaḷ) in the context of temple grants, and as imparting religious teaching (upadeśa) (Vēṇugōpālan 2009: 880).²¹ Local tradition attributes the authorship of the Sanskrit Śākta hymn (stuti), Laghubhaṭṭāraka, to a Piṭārar of Mannampuṟattu Kāvu, after whom it was supposedly named (C. N. Nampūtiri 2004: 8–9). While this claim carries little weight as to authorship of this particular work, it lends cultural support to the association of Piṭārar with Śākta worship, for this stuti to the Bālā form of Tripurasundarī has been quite popular in Kerala.²² The samayavidyā of Tripurasundarī is indeed incorporated into the Rurujit-vidhāna of the Śeṣasamuccaya, and in the Calicut goddess temple of the Zamorin, a śrīcakra is actually installed and worshiped there.

Furthermore, it is clear that the common name for such Śākta goddesses in the Tamil and Kerala countries, Piṭāri, is cognate with this complex of Śaiva/Śākta appellations. Sanderson has discussed a thirteenth-century Cōḻa inscription from Kōlār, in Karnataka, as exemplifying a cult identical to, or closely allied with the southern Brahmayāmala tradition (2007b: 277 n.140). The modular pantheon of the Mothers there would indeed fit perfectly into the Malabar array, and the separate, principal goddess is Cāmuṇḍeśvari, who is also referred to as Piṭāriyār (with an honorific ending), which the editors of the inscriptions note is derived from Bhaṭṭārikā, a regular Sanskrit title for the Goddess. Evidence further convinces me that the occasional appellation for such goddesses in Kerala, Piṣāri, is merely the attempt to (re-)Sanskritize the medieval Dravidian vernacularized form, Piṭāri. A temple that is one of this Malabar Śākta group, in the territory of the Kuṟumpranāṭu chief, some 30 kilometers north of Calicut, is indeed called Piṣāri Kāvu, after the goddess there, and I believe such variants were once common for the Śākta goddess throughout Kerala. In central Kerala, I studied several goddesses around the great temple of Peruvanam named Piṣārikkal who were all said to derive from the Śākta goddess, Kāmākṣi, at Kāñcipuram, by translocation through spirit possession of persons or objects. While the -kal element in her name is simply an old locative, I believe the projected cognate of Piṭāri was actually Bhaṭṭāri (as though a feminine formed from bhaṭṭāra),²³ rather than Bhaṭṭārikā, and I indeed found that the principal village goddess of Peruvanam, now called simply ‘Mother of Ūrakam’ (Ūrakattamma, after that village), originally bore the name Bhaṭāri according to an inscription of perhaps the eleventh century (Srinivasan 1986: 60). So the point is that Piṭārar, as a contemporary caste-title for a Śākta officiant no doubt derived from Bhaṭṭāra(ka), as a title for Śaiva officiants, but was also closely associated with the feminine Bhaṭṭāri(kā), as a term for the Śākta or assimilated folk-goddesses whom they served, convergent with the pan-South Indian Piṭāri and her cult.

Further evidence of originally Śākta-yoga associations, is found in a major caste of Kerala temple-servants called Piṣāraṭi. They are a highly Sanskrit-learned caste, and the ones I studied with in central Kerala were also noted for their expertise in indigenous modes of temple performance-arts and music. Most interesting as to their origins is the claim that they derived from Brahmans who undertook renunciation, but then lapsed back into householder life, being subsequently demoted to the status of non-Brahman temple servants. They were indeed matrilineal, having had sambandham with Nampūtiris, but lending credence to the legend of their previous renuniciant status, they were traditionally not cremated, but buried, sitting in ‘samādhi’ in a grave filled with salt. (This is exactly the treatment given to the bodies of sannyāsis and the caste of Yogis I shall discuss shortly.) It seems fairly certain that the name Piṣāraṭi is again derived from the Piṭārar/Piṭāri complex,²⁴ and in northern Kerala they are closely associated with the Śākta temples. At the aforementioned Piṣāri Kāvu, a Piṣāraṭi officiant must be present for the announcement of the date of main festival, in addition to the regular Mūssatu (the southern equivalent of the Piṭārar) who serve as the pūjāris there.

So an intriguing question becomes what the historical relation may have been between the Piṭārar and the Piṣāraṭi, and whether in fact they derive from the same or analogous communities of Śaiva/Śākta yogis, gurus, or officiants who took different social trajectories. While the Piṭārar have largely assimilated to a Brahmanical model (though in consuming alcohol and meat they are rejected as such by Nampūtiris), their elevation into the royal cult by Malabar kings may have allowed a path different from that of the Piṣāraṭi, who were more or less co-opted into a second-tier status in the Nampūtiri-dominated temples.²⁵ It is notable, however, that the Piṣāraṭis have remained some of the finest Sanskrit scholars among the non-Brahmans, and in my experience studying with them, they seemed to think this learning distinguished them from other Ampalavāsis and indeed from Nampūtiris.

Brahman mantravāda, yogis, and popular religion

In looking to other social strata for the Tantra-Śākta influences that might have shaped Malabar, we must turn back to the Teyyam cults, and consider not just the relations to royalty and the puranic legitimation of polities of conquest, but to the Śākta identities of these ‘folk’-deities, and their wider relations with society. Certain families of Brahmans actually patronized Teyyam worship at the domestic level of their manor-shrines, and especially those Brahmans associated with mantravāda. These Teyyams are generally grouped in a pantheon of five, of slightly variable constituency, but always so as to include Bhairavan as their head, along with Kuṭṭiccāttan, and Poṭṭan. This pentad of Teyyams are known as Mantramūrtis, and are associated particularly with the practice of mantravāda. While I have written at length about these deities elsewhere, I will just say of Kuṭṭiccāttan that he is the apotheosis of a virulently Śākta sociopath (Freeman 2006), and of Poṭṭan that he is an untouchable tantric gnostic who is Sanskritized into the Caṇḍāla who accosts and humbles Śaṅkārācārya as the narrative basis for the well-known Vedāntic stuti, the Maṇīṣāpañcaka(Freeman 2009). Bhairavan brings together two personas: one is the Tamil version of Bhairava who, as a Bhairava-yogi, cannibalizes his devotee’s son in the famous Ciṟuttoṇṭar story of the twelfth-century Periyapurāṇam; the other tells of Bhairava’s similar destruction of a Śaiva-yogi pontiff’s lineage in a famous Mangalore maṭha, linked to the migration of house-holding lineages of Yogis into Malabar (Freeman 2009). The last is particularly important in its likely historical influence, to which I will return, but for the moment, I want to just point out that these Mantramūrtis are the most violently anti-Brahmanical of the Teyyams, in an explicitly tantric mode, and that the Brahmans who master or worship them are thereby considered as powerful mantravādins. While I found such mantravādins therefore to be not highly regarded by some of the more orthodox, temple-centered Nampūtiris, it is these families who are the Tantris of the chiefly and martial temples where Teyyams are performed (along with another lower division of Brahmans called Emprāntiri). These mantravādins were known as the Sampradāyis in the vernacular folk-liturgies and literature, and this would seem to have continuities with the old northern Brahmayāmala tradition, as well, where the usage suggests similar attainment and transmission of mantrasiddhi (cf. Hately 2007: 415–417 n.114). As we shall see, this mantravādi-sampradāyam was also transmitted across the caste-barriers.

It is significant to note in this context that mantravāda occupies a different position between the presumably original private cults of Tantric initiates, and the public worship in temples. For it is a kind of private attainment (mantrasiddhi, enabled through initiation), applied to the very limited public of clients in a restricted, domestic setting. In fact, it could be argued in terms of earlier reflections on royalty that it was perhaps when that client was a king, that we begin to get the genesis of an expansion through his patronage and following into a temple cult. This touches on the development that Sanderson has treated in regard to Tantric gurus becoming Rājagurus, and an issue that is central, I believe, to the Rurujit-vidhāna emerging from adaptations of the older Brahmayāmala tradition. This hinges on the competition between established Brahmans and later sectarian mendicants for royal patronage, and this brings us back to the Teyyam figure of Bhairavan (Figure 7.3).

Figure 7.3  The Teyyam of Bhairavan, performed by Malayan dancers, modeled on the costumed ritual-begging of the Cōyi (Yogi) caste who settled in Malabar

As I have stated earlier, and worked out in some detail, there were certainly one or more historical movements of Bhairava Yogis into Malabar. From their maṭha in Mangalore, associated with the Kadri temple complex that dates back to the eleventh century, they had built up a substantial holding of lands and a kind of pontificate that was reported on in its decline by Pietro della Valle in the early seventeenth century.²⁶ By this time, their establishment had been sacked by their rival Vīraśaivas under Nāyaka rule, and if there had not already been intermittent migrations of sedentized, householder Yogis (called Cōyi) into Malabar, there was certainly an impetus to migrate after they lost their lands and management of the Kadri temple. It is they whom I have demonstrated brought in performative episodes of ritualized begging in full Bhairava-yogi regalia, and it is this costuming that was incorporated into the Teyyam of Bhairavan, with its narrative liturgies I have touched upon earlier. These liturgies relate both to the earlier Bhairava cult in Tamil Nadu (which I have found spread from that source not only into Kerala but into the Karnataka and Andhra countries with Vīraśaivism), as well as to the local memories of the wreckage of their pontiff’s lineage-control of Kadri. What is of great historical interest, in both these streams, is the recollection of origins or affiliations of these yogis with Ciṟṟampalam and Ponnampalam, that is to say, with the shrines of the great Tamil temple-complex at Cidambaram. An esoteric Bhairava cult in association with the Naṭarājan of Cidambaram is attested, inscriptionally, iconographically, and in texts, as is the Śākta goddess there as Piṭāri, both in an independent temple as Tillai Kāḷi, and earlier in the Ciṟṟampalam shrine itself (Soudararajan 2004: 49–50). On the Malabar side, the Yogis are reported as resorting to the aforesaid shrines as their central refuge, and dancing there in possessed ecstasy. They are also celebrated as having known multiple languages and as having traveled widely, both in the liturgies of Teyyam, and in a number of folk-rites in the spring Pūram festivals I will allude to shortly. So we clearly have evidence, (though over what time period, and in how many successive movements we cannot tell), of Bhairava Yogis most immediately from Mangalore, but with ties into Tamil Nadu, of learned and relatively cosmopolitan profile, settling in Malabar. They either brought their women with them when they entered the region, or they found wives among the local women, and transformed themselves into the settled caste known locally as Cōyis.

They either were, or have become matrilineal, and in certain lineages individuals are still schooled in a Bhairava-Śākta cult and initiated into the office of Gurukkaḷ, (recalling the name of the Śaiva officiant-caste of the Tamil country). These Gurukkaḷ, who govern the community and their shrines, worship Bhairava and various fierce, local goddesses, with blood-sacrifice, meat-naivedya, and alcohol, as staunch and avowed Śāktas. There is a commensurate lore of these Gurukkaḷ, best exemplified by the Kurikkaḷ (Gurukkaḷ) Teyyam (Viṣṇu Nampūtiri 1981: 54–57). This Teyyam’s liturgy tells of a local Cōyi who rose to eminence through learning and mantric-power, and was admitted to all the various kings’ palaces where he was honored. Eventually the Kōlattiri king was afflicted by a debilitating spirit possession that none could cure, except the Gurukkaḷ, through various of his Śākta blood-rites. He was rewarded by the king with titles and gifts, but then on his way home was murdered by those in the realm who could not tolerate one of Gurukkaḷ status rising to such prominence. On his death his gods turned him into a Teyyam, and he is still worshipped today by rites of costumed possession.

What this liturgy shows is that these Gurukkaḷ were both honored in the courts, and were in competition for royal patronage with others. Some of these others were local Brahmans, and one of the well-known events of Malabar history is that a Kōlattiri Rāja (perhaps of the fourteenth century or later) came into conflict with the Brahmans of his own realm and so brought in others (called Emprāntiri) from the Tuḷu country as rivals and replacements.²⁷ The earlier treated stratification among the temple-based Nampūtiris between Tantris and Śāntis, the resort of Brahmans to mantravādawith the more active embrace of the lower-caste Teyyam worship, and the division between these older Brahmans and the imported Emprāntiri, all testify to the internal and external pressures among Brahmans to compete in the religious fora of the chiefly socio-political order. Śaiva and Tantric mendicants must certainly have been a force in this competition.

The strategy of Malabar kings and chiefs, as in Nīlēśvaram, to patronize both a Brahmanical Śiva temple, and a Śākta goddess temple, and to bring them together in various complimentary rites, but also to have a Brahman Tantri in ultimate charge over his Śākta counterparts, testifies to the entangled hierarchies that could emerge. The Piṭārar are elevated in their own Śākta domain, and treated as Brahman priests by all the Nāyar and other middling castes of their worshipping constituency for most of the ritual calendar, yet are not ‘real’ Brahmans in the highest temple order. The same is the case for their lineage-mates throughout the chain of Malabar temples.

A similar situation came to eventually obtain, but lower down the caste-ladder, for Cōyis and their Gurukkaḷ. These, I was surprised to learn, perform the pratiṣḥtā, and function analogously to the Tantris in the Teyyam and other shrines, except that they do so for major castes in the avarṇacategory, those traditionally falling below the Nāyar as putative ‘Untouchables’. Recognizing this role of the Cōyi Gurukkaḷ, however, now explains some peculiar features of the Pūram song-festivals which are central to the collective spring worship among the avarṇa castes. For these communities, whom one would expect to have been largely illiterate, have amongst them Sankskritically learned masters, called Paṇikkar, who train singing and dancing troupes to perform various genres in the Pūram rites directly related to our concerns. These include: celebrations of Śiva and Śakti, specifically linked to their dance in the shrines of Cidambaram; depictions of Bhairava Yogis, including deportment, dress, and material from the Tantric lore of mantra and yogic psycho-physiology; various versions of histories of the Yogi’s wanderings, and their settlement as farmers in Malabar; and various tellings of the Cidambaramāhātmya (narratively related to, yet independent of, Tamil and Sanskrit versions of those texts).²⁸ It seems to me likely that this tutelage came from the close intimacy (and perhaps some original connubium) between the Gurukkaḷ and prominent castes like the Tiyyas and Maṇiyāṇis among whom the Pūram rites flourish.

The presence of the Gurukkaḷ (as literal gurus) among these castes would also explain the many Teyyams originating among these lower castes, whose protagonists attain their apotheosis through being celebrated as great yogis, Śākta initiates and mantravādins. The Teyyams also testify to relations of inter-caste tutelage from the Brahman Sampradāyi-mantrvādins, which shows there were attractions, among these echelons of Brahmans, to compete with Yogis even for avarṇa clientele. It is also likely, despite the later wave of Smārta reformism that has made such caste relations seem unthinkable, that the force of a Śākta ideology at one time licensed such intimacies as it did among non-Saiddhāntika Tantrics elsewhere in India. Even the Teyyams of Pulayas, the traditionally lowest caste of agrestic laborers, include their own Toṇṭaccan or Guru whose liturgy recounts his studying under a female Yoginī to learn her special art of shape-shifting into animal form.

Textual dynamics in social context

If overtly Śākta teaching and practice were eventually pushed socially downward by the hegemony of Brahmanical orthopraxy, the Piṭārar and the royal cult of the goddesses, as well as Brahman mantravādin practice, carried much of the doctrine and discursive content upward. It was at the level of orthodox Brahmanical practice and textual production that this tension was most clearly registered, and I want to return to these issues in the dynamic we started with in terms of speculatively placing the Rurujit-vidhāna in the milieu of the Śeṣasamuccaya and Tantrasamuccaya. At a similarly Brahmanical textual level in Orissa, Sanderson has treated a comparable dynamic among the Paippalādin Atharvavedins, where they both incorporated Śākta pantheons and variants of their rites into their worship, on the one hand, but also expunged them doctrinally and ritually of unacceptable elements, on the other. Relatedly, he suggests that the great mantric digests, the Prapañcasāra and Śāradātilaka were products of this milieu, and likely also derived from the Orissan region (2007b: 336ff., 230–233).

In reviewing the general trajectory of the Tantrasamuccaya, I would argue in a similar vein that it represents the attempt of its author to first reduce the much vaster and Śāktified content of the Īśāna compendium to a controlled program of Smārta-style worship, but that the inadequacy of this initial attempt to keep up with Śākta pressure on various fronts, led to the additional accommodations of the supplementary Śeṣasamuccaya. This force is seen at the vernacular level by the fact that while the Tantrasamuccaya lacks any treatment of Bhadrakālī, the vernacular renderings of the Tantrasamuccaya all add her on, as though she should have been included with the seven deities of the original’s pantheon, and which cannot do without her. Within the Śeṣasamuccaya itself, this force is even clearer in the modular addition of the Rurujit-vidhāna to the end of the work, with its partial redundancy in pantheon, but the enhanced Śāktification stemming from its origins in the blood-cult of the southern Brahmayāmala. This last legacy is particularly interesting in social terms because, again as Sanderson has noted, the daily priests of this Dāruka-slaying Kālī-cult are not Brahmans, but Pāraśavas, defined as the offspring of Brahman men and Śūdra women who should officiate for Kālī (2007b: 277–278 n.142). This taxon would of course apply to what all the Ampalavāsi temple-servant castes in Kerala theoretically were, under their sambandhamrelations with Nampūtiris. We are thus left ultimately wondering about several scenarios for the origins of the actual Śākta survivors of this order in the persons of the Piṭārar. Were they originally Ampalavāsis of this Pāraśava-sort, who were upgraded into their quasi-autonomy as quasi-Brahmans in the cult of the Malabar kings? Or were they Śaiva mendicant Bhaṭṭārakas, Rājagurus who were originally autonomous Śākta officiants, later demoted and somewhat subordinated under the Smārta orthodoxy of Nampūtiris? It seems likely, looking comparatively to the varying status-trajectories of the Piṭārar, Piṣāraṭi, and Cōyi that we have evidence for different social trajectories for different Śākta constituencies in different times and places. But what is clear, especially given the now documented Śākta influence at all levels of the caste-system and socio-political formations, is that there was a major socio-religious and historical impact represented by the various textual remains available to us. This then invites a marshaling of these sources with a sharper set of historical agenda in mind applied to their text-critical study.

Construing the Śeṣasamuccaya and its Rurujit-vidhāna chapters in these terms (but with much of the fine-grained textual study still to be done), I would suggest that as with the Paippalādin materials, Nampūtiris were responding to larger Śākta and Tantric movements and trying to gain ritual control of them, to bring them within the orbit of their social authority. It is an interesting fact, however, that no one seems to know of any temples where this Rurujit-vidhāna, expurgated though it is, has been a part of any Nampūtiri Tantric practice. We have only the remnants of these temple layouts and greatly redacted pūjās of the Piṭārar in Malabar, and the correlated practice of Tantris, during their brief annual roles in these temples, which seems to have lapsed into the generic Nampūtiri Bhadrakālī-template of the earlier parts of the Śeṣasamuccaya. While it is possible that the inclusion of the Rurujit-vidhāna section was never intended for Nampūtiri practice (but then why expurgate it, and why include it in the whole project of the Tantrasamuccaya and Śeṣasamuccaya?), it is also possible that the two chapters on pratiṣṭhā, and festivals and renovations were intended for Nampūtiri Tantris, while the elaborate pūjā-paṭala was for Tantris only during their special festivals, with the ordinary pūjās being left to local practice and unspecified. Or perhaps they also intended the text as a project of expurgating even the ordinary pūjās of others in accordance with their own predilections. In any case, while the project was a historical failure, in terms of not being followed by Nampūtiris today, it shows that at the time, there was some urgency felt for these Brahmans to gain control of this cult and incorporate it into their ritual program.

To close these textual considerations, tentative and provisional as they are, I would like to suggest that the Prapañcasāra (but not the Śāradātilaka) stands in a similar relation to the Tantra-Śeṣasamuccaya tradition of Kerala as either of the former texts (or both of them) perhaps did to the traditions of the Orissan Brahmans. It is first of all noteworthy that the Tantra-Śeṣasamuccaya tradition does not seem to draw on the Śāradātilaka, while it does, quite significantly, on the Prapañcasāra. This suggests that we may need to drive more of a historical wedge between the Prapañcasāra and Śāradātilaka, and query whether the former is as tightly bound to an Orissan milieu as the latter.²⁹ More immediately important to our present concerns, however, is the similar force the Prapañcasāra has as a template for expurgated Śāktism, and its redeployment from its centrally mantravādin context, to one of pratiṣṭhā in the Kerala tradition. Hence we find, significantly I think, that the dhyāna-ślokas in the Tantrasamuccaya for Śiva and (by tradition) for Sarasvatī are taken directly from the Prapañcasāra.³⁰ It is interesting that when Nampūtiris had a substantial Śaiva Siddhānta corpus to draw upon, they took their Śiva from here, along with their (Lipi-)Sarasvatī in the context of the extensive lipi-nyāsa that is generic (and Śāktically so) for the pūjā of all the deities in the Tantrasamuccaya. Tellingly, when we come to the Śeṣasamuccaya, this derivation expands significantly, for Sūrya, Sarasvatī, Śrī, and Pārvatī are all taken directly from the Prapañcasāra.³¹ Of the two dhyāna-ślokas from within the Rurujit-vidhāna for Bhadrakālī/Cāmuṇḍā, the first for her four-armed form is found in our grantha manuscript of the Mātṛsadbhāva (no. R5126: 117), while both this and the śloka for her eight-armed form are found together from another manuscript of a free-standing text titled Rurujidvidhānapūjāpaddhati (no. R3365: 17), the origin for which is indeed the palace of Kōlattiri Rāja (ibid.: 102). The provenance of this latter text makes one wonder, finally, whether the Rurujit-vidhāna chapters found their way into the Śeṣasamuccaya from such Malabar sources, or whether the Kōlattiri’s scholars extracted this material for use of the Piṭārar, or their prototypes. It is perhaps significant that the Tantri of the Zamorin’s temple in this Malabar complex belongs to the Cēnnās family, bearing the same name as that of the authors of the Tantra- and Śeṣasamuccayas. No doubt some of the questions I have raised in a provisional and speculative fashion might be answered with more thorough-going combing of these textual corpora (and field investigations), which I hope to continue. The pantheons, mantric arrays, rituals, and any accompanying cultic and socio-political information must all be garnered from the different versions of the southern Brahmayāmala, the Mātṛsadbhāva, the Rurujidvidhāna, and the Śeṣasamuccaya and carefully compared in hopes of filling out the history of this system.

Artifice and expurgation in blood-rites

I wish to end this discussion with an example of material culture in the present which bears on these texts and redirects us to considering an important part of their socio-cultural dynamics. In my fieldwork, I encountered the ubiquity of what is called gurusi, or an artificial blood that is made of turmeric and building lime mixed in water. From my study of early Tamil, I realized that the word was a phonological Sanskritization of the old Dravidian kuruti, which is simply ‘blood’, but especially used in the Caṅkam literature of human blood shed in battle and sacrificial blood. The latter usage is still retained for actual blood-sacrifice among the non-Brahmanized castes in Kerala, while the Nampūtiris use the artificial word and substance, gurusi, for propitiating Śākta goddesses to whom they offer huge cauldrons of the stuff (Figure 7.4).

Figure 7.4  Nampūtiri of a prominent Vādhyar family offering gurusi-tarpaṇam from a large cauldron of artificial blood at his feet over a woven maṇḍalam to their family goddess of Raktēśvari in a domestic rite

I have also seen these Brahmans combine the gurusi constituents with rice for certain balis in festivals and mantravāda in explicit replacement of meat. It has been evident from my fieldwork that there is a kind of mimicry at work here, in which Nampūtiris strive anxiously to capture the force of Śākta blood-rites, while avoiding the attendant impurity of slaughtering animals.

I was puzzled, however, to find non-Brahmans also using gurusi, even in real sacrifice, by dipping the severed necks of their victims into a cauldron of the substance and swishing it around, before emptying the vessel in oblation. Though I suspected they might be emulating Brahmans in this, when I asked why they would use a substitute for blood, when they had the real thing with all its attendant powers (and unavoidable impurity), they revealed that a completely different logic was operative: they wished to offer huge pots of real blood, but couldn’t afford to sacrifice so many animals or such large ones, so they used the bloodied gurusi to express their good intentions to the gods, despite their paltry resources! These obversely twinned logics of substitution in sacrificial offerings have directed me back to the texts, Sanskrit and vernacular, and I wish to close with some interesting connections among them.

A number of Teyyam worshipers reported to me that particular enshrined goddesses could no longer be performed as Teyyams, since they would require human sacrifice, which could not be provided now, as in the past. Certainly it was evident from the liturgies and performances of even less demanding deities that the blood-sacrifices they accept are representative of the ideal of human sacrifice. Most telling for me was the so-called Kōḻi-stōtram, which reveals the history of the sacrificial chicken (kōḻi), to have originated as Dāruka, who fled the goddess by taking various forms as he was successively slain, earlier as a warrior, and finally as a chicken. The modern avarṇa officiant recites this history to his goddess before slaying the fowl, and summarizes its ritual portent with the lines:

Previously, I devotedly hacked a man and gave you his blood. Then when a man could not be happily gotten, I slaughtered an elephant for you. Getting neither man nor elephant, I cut a goat. Now getting not a single elephant, nor man, nor goat, by cutting a chicken I am giving you blood today.³²

I was immediately reminded of these lines when I read the following on the substitutional chain of victims for pacifying the Mothers from the southern Brahmayāmala:

The generation of the Mothers was in order to destroy Asuras;

One should effect the welfare of all beings by duly pacifying these Mothers.

The sacrificial victim is said to be of three sorts: highest, middling, and lowest.

Killing an elephant, a horse or a man is designated the highest; while a buffalo or camel is middling; and a goat or pig the lowest.

Again, a deer is the highest, while a chicken is the lowest.³³

Clearly this earlier Śākta logic is closer to the sentiments of my Teyyam informants today than it is to Brahmans.

In turning to Īśānaśiva’s paddhati, we find further information to illuminate these issues, a transitional situation involving Brahmans and Pāraśavas that perhaps anticipates the Nampūtiri/Piṭārar split. The paṭala on the Mothers there (Kriyāpāda 59), presenting substantially the same cult of our concern, ends with a pacifying śānti-rite giving a social ranking to the order of sacrificial victims (Gaṇapati Śāstrī 1922: vol. 77: 579). While we have developed away from overt consideration of offering a human victim, it is nevertheless designated mahāpaśu, which is the code-word for human sacrifice in the Śākta-Bhairava milieu which this text emulates, recalling the ‘man-beast’ (narappacu) whose flesh the Tamil Bhairava-yogi demands in the Periyapurāṇam story that is ancestral to his Kerala Teyyam. The mahāpaśu in our paddhati, however, is redefined as being the deer of a Brahman, the elephant of a Kṣatriya, the buffalo of a Vaiśya, the goat of ‘others’, or, if these are unavailable, a buffalo that belongs to everyone. This mahāpaśu is bound to the sacrificial stake, slaughtered, its blood offered to the Mother or Mothers, and its blood and flesh mixed with rice and distributively offered as balis. Though the chicken does not appear here (as probably unworthy of this collective mahāśānti rite), we can find here the logic for the earlier Brahmayāmala‘s high ranking of deer in its coda-line, due to its association with Brahmans, and the general logic, throughout, for the intersecting ranks of size and social association of the other animals, with the human being latent in the mahāpaśu nomenclature, and the chicken as residually the lowest.

I think we can also find a further social differentiation among the officiants implicit in the sequence and performance of the rites in this paṭala. The pratiṣṭhā and initial worship I believe are done by Brahmans, which is all that would make sense of the stipulation that offerings are made in substitution (pratinidhi, pratirūpa) explicitly for goat’s meat and other flesh, with the itemization of the vegetarian substitutes (ibid.: 572). Indeed, I believe the claim of this paṭala to being a ‘purified’ (śuddha-paṣkataḥ … viśuddham) rendering of the Svacchanda- and Mātṛ-tantras is not intended eulogistically, but carries this ritual significance of Brahmanical expurgation (ibid.: 567, 580). Even into the first part of the sanguinary śānti-rite, a Brahman seems to be superintending the festival installations, with substitute-meat stipulated for his performing community-wide balis that sound very much like the modern Nampūtiri grāma-bali as a part of major festivals (ibid.: 578). The context shifts, however, with the deity’s setting out in full, royal grandeur, accompanied by warriors bearing the swords of Kālī (khaḍga), for there we are told of the general body of the community festively engaged in the nightly slaughter of animals (ibid.: 578–579). It shifts again, however, with the appearance of the Pāraśava, who appears with mudrās and dance to take up his place in the festival maṇḍapa (ibid.: 579). It is just after this that we get the rendition of the mahāpaśu rites, detailed above, and it seems quite likely that it is the Pāraśava who superintends over or performs these bloody balis to the Mothers, just as we have seen was his mandate in the Brahmayāmala, and still farther back in South Indian history.³⁴

But what of the antecedents of the modern gurusi? Again Īśānaśiva’s text provides a context where this comes clearly into play. While I had come across the reference to raktodaka, ‘blood-water’, and suspected this might be gurusi, it is indeed confirmed in the section of this text dealing with the exorcism of bhūtas. And there its Kerala recipe is explicitly given as combining lime (cūrṇa) and turmeric (niśā) in water, to be offered as inducements and balis to feed and placate possessing demons (Gaṇapati Śāstrī 1921: vol. II: 303–304).³⁵ The context here is the earlier differentiated mantravāda (a kind of permanent way station between sādhana and pratiṣṭhā), into which the Nampūtiris were drawn, and which they also extended in their officiation over the various enshrined goddesses who themselves first entered their purview as quasi-demonic beings. In the earlier text of the Kurikkaḷ Teyyam, the Bhairava Yogi there cures the Kōlattiri Rāja of his demons with actual blood-sacrifice, as well as gurusi, and he would have been the Brahman mantravādin’s natural competitor in such ventures. The gurusi-tarpaṇam I have seen as private worship by Nampūtiris of their own lineage goddesses, plays out like an enactment of this exorcism from Īśānaśiva’s text. While the implications and correlations are too detailed for the present exposition, I am convinced that on both the domestic front, as well as in public fora, the Śaiva-Śākta mendicants and lineages that entered South India posed considerable challenges to the region’s Brahmans. The latter met these challenges in the texts and practices we find today, which are testament to their struggles, and artifacts of their multifaceted response. The ultimate triumph of an (admittedly transformed) Smārta orthodoxy further explains the survival of much of the original impetus and content of Śāktism in the folk-religion of Malabar.

The Śeṣasamuccaya and the cult of Dāruka-Rurujit with which I started, records the effort to claim this Śākta territory for Brahmans, expurgating what was threatening as they developed their new, and now authoritative canon. To come full circle I may thereby close with an interesting apparent oversight made by those authors. In the listing of the dravyas to be offered to the very Śākta powers of the oldest Brahmayāmala, Raktā, Karālī, Caṇḍākṣī, and Mahocchuṣmā, the compiler of the Rurujit-vidhāna section of the Śeṣasamuccaya seems, accidentally I imagine, to actually have enjoined the offering of meat.³⁶ His commentator, probably his son or nephew, immediately made good the lapse by explaining that in the place of meat, we are to understand the use of a substitute.³⁷ In the modern Malayalam commentary, a Tantri Guru specifies that the substitute should be the steamed rice-cakes known in Malayalam as aṭa (K. Divākaran Nampūtirippāṭu 1977: 242). Representing the modern Nampūtiri consensus on such matters, this substitute can be seen in Staal and Gardner’s film on the Agnicayana (1976), where the Tantric expurgation has carried over even to their traditional practice of the Veda. For lined up in their banana-leaf packets, with little leashes tied round them, are these same little rice-cakes, awaiting their ‘sacrifice’ in lieu of the textually mandated goats.

Notes

1  I follow Malayalam spellings of Sanskrit words in a Kerala context, so that I write Tantri, rather than Tantrin, or Tantrī following Western or Indian Indological conventions. The latter seem artificial in these contexts, and there is little scope for confusion.

2  These are Brahma, Arka (Sūrya), Vaiśravaṇa, and Kṛṣṇa. While the first three are merely installed as aniconic pīṭhas in temple compounds, the last, according to the Vimarśinī, is qualified as indicating gośālāsthaḥ kṛṣṇaḥ. (Narayana Pillai 1951: 2nd part, p. 1). This would seem to refer specifically to the rather late intrusion of an ancillary, modular shrine to what is still locally called Gośāla-Kṛṣṇa, found within the outer walls of larger Kerala temples, such as at Peruvanam and Tṛprayār in central Kerala.

3  I again follow the same passage of the Vimarśinī, which explains the syntactic aggregation and grouping of the deities here, and in fact accords with their distribution and topical handling in the overall organization of the text and their ritual complexes.

4  One can get a sense of the complex internal status grades (ābhijātyam) of the Nampūtiri formation from Padmanabha Menon (1984: 9–48), with the caveat that his is a textually formalized projection of the system reliant on such works as the Śāṅkarasmṛti and Keralotpatti.

5  I disagree, however, with Dumont’s ultimate analysis, which basically reads Brahman aspirations for power as distorting what should have been their ritual values. Dumont was, as are most historians and anthropologists, unaware of the social historical significance of tantric and Śākta movements in South India.

6  Early forays into this nexus as posing a major problem of historical development were raised by Brunner (1975, 1990), but it is in the introductions to Goodall’s aforementioned edited works (1998, 2004) and allied articles that he really brings text-historical clarity to these issues for the South.

7  The subject of tantric initiation in Kerala has recently been briefly surveyed by S. A. S. Sarma (2010), who argues that tantric initiation had devolved by the time of the Tantrasamuccaya from its earlier salvific purpose into a mere mantropadeśa or mantradīkṣā for purposes of temple installation and worship. I have confirmed in my fieldwork cases of institutional Tantris still administering mantropadeśa to students to ratify them for daily pūjās, though this may be the recent revival of a tradition that had essentially lapsed with modernity. While Sarma’s brief observations seem to posit a soteriological rationale for this shift, a historically anthropological perspective would need to consider the social organizational and political implications of this change.

8  Cf. the brief discussion of the mahāśānti for the Mātṛ deities from Īśānaśiva’s paddhati, below.

9  As I discuss below, despite the obvious basic identity relation between the Rurujit-vidhāna texts and contemporary practice, the social and ritual disjunctures that do exist between the two could suggest historical scenarios by which either one is earlier.

10  I say they performed such sacrifices ‘traditionally’ because current law would render such practices illegal, despite the fact that the majority of such Hindu law-makers in Kerala are themselves non-vegetarian and may freely drink alcohol. When harassed with such meddlesome statutes there are various stop-gaps possible, such as presenting live chickens in the temple, but then taking them home for slaughter and preparation of the meat which is later offered and consumed as naivedya of the goddess.

11  I refer to these sections and the cultic complex as the Rurujit-vidhāna, following modern, local usage. There is, in fact, a discrete pūjāpaddhati ms. titled Rurujidvidhāna to which the Śeṣasamuccaya is closely tied, mentioned briefly, below.

12  The Vimarśinī on Śeṣasamuccaya 1.1 reads, rurujiditi rurunāmno daityasya hantrī bhadrakāḷī. (Narayana Pillai 1951: part 2, p. 1).

13  There are two different texts under this name. I have seen copies of only one of the manuscripts (IFP T. no. 552), while the other (Trivandrum, possibly no. T. 982), a different text, I know only from extracts which Alexis Sanderson has kindly shared with me.

14  A brief rendition of the creation of Ekavīrā or Bhadrakālī and her killing of Dāruka, opens the Brahmayāmala (IFP ms. T. no. 552, p. 1), while the goddess finds mention in the Trivandrum ms. as Dārukāsuranāśinī, according to the corrected reading supplied by Sanderson (from his extracts, p. 4), in the context of her identification as Ekavīrā or Kālarātrī.

15  gōkarṇṇaṃ nāma pāścātyasyārṇṇavasya taṭe śubhe

śaṅkarasya mahākṣetraṃ śaṅkaraṃ sarvadehināṃ

(I.32, Tirumumpu 1975: 16)

16  hṛdayāmbujam utkhāya kandam vyādhavadhūr iva

pakvā‘kṣipāvake devī bhakṣayāmāsa tanmuhuḥ

(8.41, ibid.: 130)

17  The verses in the sixth chapter, from 39–54 begin, respectively, with the syllables ka, e, ī, la, hrī(ṁ), ha, sa, ka, ha, la, hrī(ṁ), sa, ka, la, hrī(ṁ), and the sixteenth adds śrī(ṁ) (pp. 88–92).

18  Matṛsadbhāva, Trivandrum ms. 792, p.93; Rurujidvidhānapūjāpaddhati, ms. pp. 1–2. Śeṣasamuccaya, mantras following 8.4. (Narayana Pillai 1951: 93–94).

19  iṭar ceyta kāḷi matamōṭu kaila-y-aṇṇayuṃ aran bhayattāl

iḻalōṭu bhūtakāḷi karāḷi bālakaḷ kailanāthan

daśavuruvāya kuḻavikaḷ koṇṭu matam aviṭe taḷatti

(Viṣṇu Nampūtiri 1993)

20  Literally this means the grove (kāvu) outside (-puṟattu) of Mannan (or the king), or of the mannam, the assembly ground. There is indeed a large sacred grove around three sides of the temple, and a small shrine outside the west (facing) side of the temple, which houses the goddess supposed to be original to Mannan on his palace-site.

21  I have confirmed these references in the Malayalam Lexicon, under the older variant, Piḻārar, which include a twelfth-century reference from the Sucīndram temple, and a fourteenth-century Maṇipravāḷam work, Uṇṇiyāṭicaritam. The latter reference to upadeśam is confirmed by the description of Piṭārar as the chiefs of Bhaṭṭar (Paṭṭar), who imparted śāstric learning in temples.

22  A Kerala Sanskrit commentary was composed on it in the fourteenth century by Rāghavānanda (Gaṇapati Sāstrī 1917), and two in Malayalam in the nineteenth century (C. N. Nampūtiri 2004).

23  This is analogically confirmed, in meaning, variance of form and derivation from the old Tamil piṭār, meaning ‘arrogance, greatness’, and paṭārar, meaning ‘God … venerable persons, as priests’, and their derivation from Sanskrit bhaṭṭāra. (Madras University (Committee) 1982: vol. IV: 2435; vol. V: 2652).

24  The element -aṭi in their name is a common honorific suffix, referring to the foot (aṭi) or sacred foot (tiruvaṭi) of a deity. A popular etymology traces their name from Viṣahāri (‘poison-remover’) which is also applied to the goddesses, Piṣāri(kkal), mentioned above. While linguistically implausible, this nevertheless supports their co-association in folkloric memory.

25  Most perplexing in the history of the Piṣāraṭi is their designation ‘Vaiṣṇava’ as a kind of alias for their caste-name in Sanskritic literary works, but whether this represents a period prior to their purported renunciant phase, or after their re-assimilation to the caste-order needs further investigation.

26  For these details, I draw freely on my earlier work (Freeman 2006), though two corrections are necessary there. One concerns the early dating of Tirumantiram, in which I would now follow Goodall (1998: xxxvii–ix). The other concerns the location of Candragiri/pura, the original sites of which I have now more firmly identified further north in the Koṅkan; I hypothesize, however, that the earlier identified site in South Kānara was linked with these, in the spread of Paścimāmnāya movements down the West Coast. I have recently discovered textual evidence for these movements among my informants in Malabar.

27  Well known in various Malayalam sources, an account of these events found their way into the Sanskrit work Udayavarmacarita (Sāmbaśiva Śāstrī 1938).

28  While I cannot dwell on these fascinating genres here, representative selections can be found in Viṣṇu Nampūtiri (1998, selections in 2005).

29  There are a number of historical problems I have, from the Kerala perspective, on understanding how and why the Prapañcasāra, if an originally Orissan cultural production, would end up so intimately embraced on the far West Coast (and I don’t think the attribution of it to Śaṅkara alone would carry much clout). While Sanderson’s case for the Orissan influence is compelling (though more so in the case of the Śāradātilaka), my own inclinations are to look to the Cōḻas and perhaps Cōḍa-Gaṅgas as complicating these matters, but I cannot pursue these arguments here at present.

30  While the dhyāna-śloka for Śiva is given (7.66), there is none for Sarasvatī in the published Tantrasamuccaya’s treatment of lipinyāsa. Since the ṛṣi, chandas, and devatā are given for her there, however (5.20), the Nampūtiri vernacular traditions all supply this dhyāna, as found in Prapañcasāra (7.3). Śiva’s dhyāna in the latter is at 27.41.

31  The correlations between Śeṣasamuccaya and Prapañcasāra are, respectively: for Sūrya (3.99 = 15.40); for Sarasvatī (3.102 = 8.41); for Śrī (3.103 = 12.4); and for Pārvatī (3.104 = 10.10), with some slight changes.

32  mumpināl āḷe veṭṭi ampināl cōra tannēn. impamōṭ' āḷe-k-kiṭṭāññappōḻ iṅṅ' āna veṭṭi. āḷeyuṃ āneyuṃ kiṭṭāññ' āṭṭine aṟuttu. pinne aneyum āḷeyum āṭum ēkavuṃ kiṭṭāññippōḷ kōḻiye aṟuttuṃ koṇṭu cōraye-t-tarunnēn innu (Cantēra 1973).

33  asurāṇāṃ vināśārthaṃ mātṛṇām udbhavaṃ kramāt

mātṛṇām śāntikaṃ kṛtvā sarvabhū{ta}hitaṃ bhavet

uttamāmadhyamadhama trividhaṃ paśur ucyate

hastīhayanaram hatvā uttamantu vinirdiśet

madhyaṃ māhiṣam evoṣṭraṃ tv adhamaṃ tvajasūkaraiḥ

hariṇam uttamañcaiva kukkuṭañcādhamaṃ bhavet

(No. T522: 91–92)

34  It is quite significant that the early text that enjoins the employment of Pāraśavas for Bhadrakālī‘s worship (Caland 1982: 226–227), as Sanderson has pointed out, is unambiguously South Indian; its very syntax suggests its origins in Tamil Nadu (ibid.: xiv–xv).

35  I believe it may be historically significant that there are many synonyms deployed for the Sanskrit term (rakta-udaka, -uda, -ambhas, -toya, -jala), vs. the unitary term, and substance in Malayalam.

36  mūlaṃ madhye raktā hṛdayakarāḷyau śikhā ca caṇḍākṣī kavacaṃ ca mahocchuṣmā śiraś ca pūrvādidikṣu devyāṃ syuḥ samidājyānnasiddhārthamāṁsāny atha tilā yavāḥ dravyāṇi vrīhayaś cājyaṃ sarvatrāṣṭaśataṃ hutiḥ (Narayana Pillai 1951: 88).

37  māṁsasthāne pratinidher eva grāhyaḥ (Vimarśinī on 7.92–94, p. 65).

References

Manuscripts

Brahmayāmala, Institut français de Pondichéry, no. T522. Devanāgarī script.

Brahmayāmala, Trivandrum University Manuscript Library, no. 11170(?) or 1982(?); from extracts supplied by Alexis Sanderson.

Mātṛsadbhāva, Institut français de Pondichéry, no. 792; Copied from Trivandrum University Manuscript Library, no. 1642. Devanāgarī script.

Mātṛsadbhāva, Institut français de Pondichéry, Government Oriental Manuscript Library, Madras, no. R5126/SR5623. Grantha script.

Rurujidvidhānapūjāpaddhati, Institut français de Pondichéry, Copied from Government Oriental Manuscript Library, Madras, no. R3365/SR1865. Grantha script.

Books and articles

Brunner, H. 1975. ‘Le sadhaka, personage oublié du sivaisme du Sud’. Journal Asiatique, 263: 411–443.

Brunner, H. 1990. “Ātmārthapūjā Versus Parārthapūjā in the Śaiva Tradition”. In The Sanskrit Tradition and Tantrism, ed. Teun Goudriaan, pp. 4–24. Panels of the VIIth World Sanskrit Conference, Gen. J. Bronkhorst. Leiden: E. J. Brill.

Bühnemann, G. 1988. Pūjā: A Study in Smārta Ritual. Vienna: Institut für Indologie der Universität Wien, Sammlung De Nobili.

Caland, W., trans. 1982 (rpt). Vaikhanasasmartasutram: The Domestic Rules and Sacred Laws of the Vaikhanasa School Belonging to the Black Yajurveda. New Delhi: Ramand Vidya Bhawan.

Cantēra. C.M.S. 1973. Kaṇṇakiyum Cīrmmakkāvum. Kottayam: The Author; distributed by National Book Stall.

Cantēra. C.M.S. 1982. Uttara Kēraḷattile Śāktēyakkāvukaḷ. Cannanore: A. V. Sanjeevan.

Champakalakshmi, R. 2011. Religion, Tradition, and Ideology: Pre-colonial South India. New Delhi: Oxford University Press.

Divākaran Namūtirippāṭu, K., ed. 1977. Śēṣasamuccayaṃ (Commentary). Covvannur: Tantravidyāpīṭham.

Divākaran Nampūtirippāṭu, K., ed. 1986. Kuḻikāṭṭu Pacca (Tantra Grantham), Compiled by D. Subrāya Tantri. 3rd edition. Kunnamkuḷam, Kerala: Pañcāṅgam Press.

Dumont, L. 1983. “Nayar Marriages as Indian Facts”. In Affinity as a Value: Marriage Alliance in South India, with Comparative Essays on Australia, ed. Louis Dumont, pp. 105–144. Chicago, IL: University of Chicago Press.

Freeman, R. 1991. ‘Purity and Violence: Sacred Power in the Teyyam Worship of Malabar’. Ph.D. dissertation, Dept. of Anthropology, University of Pennsylvania. Ann Arbor, MI: University Microfilms Incorporated.

Freeman, J. R. 1998. ‘Rubies and Coral: A Lapidary Language for Kerala’. Journal of Asian Studies, 57(1): 38–65.

Freeman, J. R. 2003a. ‘Hindu Literature in Malayalam’. In A Companion to Hinduism. Gavin Flood, ed. Oxford: Basil Blackwell.

Freeman, J. R. 2003b. “Genre and Society: The Literary Culture of Pre-Modern Kerala”. In Literary Cultures in History: Reconstructions from South Asia, ed. Sheldon Pollock. Berkeley, CA: University of California Press.

Freeman, J. R. 2006. “Shifting Forms of the Wandering Yogi: The Teyyam of Bhairavan”. In Masked Ritual and Performance in South India, eds, David Shulman and Deborah Thiagarajan, pp. 147–183. Ann Arbor, MI: Centers for South and Southeast Asian Studies, University of Michigan Press.

Freeman, J. R. 2009. “Untouchable Bodies of Knowledge in the Spirit Possession of Malabar.” In The Body in India: Ritual, Transgression, Performativity, eds, Axel Michaels and Christoph Wulf, pp. 135–164. Paragrana 18(1). Reprinted as Images of the Body in India. New Delhi: Routledge 2010.

Freeman, J. R. 2011. ‘Artful Selves in a Libidinal Economy: The Cult of Courtesans and Inter-caste Connubium in Kerala’. History and Anthropology Colloquium, February 14, Duke University.

Gaṇapati Sāstrī, T., ed. 1917. The Laghustuti of Śrī Laghubhaṭṭāraka with the Commentary of Śrī Rāghavānanda. Trivandrum Sanskrit Series, no. 60. Trivandrum: Government Press.

Gaṇapati Sāstrī, T., ed. 1920–22, 1925. Īśānaśivagurudeva Paddhati of Īśānaśiva Gurudeva. Trivandrum Sanskrit Series, vols. 69, 72, 77, 83. Trivandrum: Government Press. Rpt. 1990 with Introduction by N. P. Unni. Delhi: Bharatiya Vidya Prakasan.

Goodall, D., ed. and trans. 1998. Kiraṇavṛttih/Bhaṭṭarāmakaṇṭhaviracitā = Bhaṭṭa Rāmakaṇṭha’s commentary on the Kiraṇatantra. Pondicherry: Institut française de Pondichéry, École française d’Extrême-Orient.

Goodall, D. ed. and trans. 2004. Parākhyatantra, A Scripture of the Śaiva Siddhānta. Pondicherry: Institut français de Pondichéry, École française d’Extrême-Orient.

Goudriaan, T. 1977. ‘Khaḍga-Rāvaṇa and his Worship in Balinese and Indian Tantric Sources’. Wiener Zeitschrift für die Kunde Südasiens, 21: 143–169.

Hatley, S. 2007. ‘The Brahmayāmalatantra and Early Śaiva Cult of Yoginīs’. Ph.D. Dissertation, Department of Religious Studies, University of Pennsylvania. Ann Arbor, MI: University Microfilms Incorporated, # 3292099.

Heitzman, J. 1997. Gifts of Power: Lordship in an Early Indian. Delhi and New York: Oxford University Press.

Liṅgamahāpurāṇa (with the Ṭīkā Śivatoṣiṇī) 1906 [1857], Mumbai: Veṅkateśvara Steam Press.

Madras University (Committee). 1982. The Tamil Lexicon. Vols. IV and V. Madras: University of Madras.

Nampūtiri, C.N., ed. 2004. Bālāviṁśati athavā Tripurāstōtraṃ Viṁśati (Vyākyānadvayasamētam). Kodungallur: Devi Book Stall.

Narayana Pillai, P.K., ed. 1951. Śeṣasamuccaya. Trivandrum Sanskrit Series, no. 166 Trivandrum: Government Press.

Padmanabha Menon, K.P. 1984. History of Kerala Written in the Form of Notes on Visscher’s Letters from Malabar, Volume 3. T. K. Krishna Menon, ed. New Delhi: Asian Educational Services.

Ramaswami Sastri, V. et al., eds 1945–62. Tantrasamuccaya of Nārāyaṇa with the Commentaries Vimarśini of Śaṅkara and Vivaraṇa of Nāṛāyaṇaśiṣya. Trivandrum Sanskrit Series, nos. 151, 169, and 200. Trivandrum: Government Press.

Sāmbaśiva Śāstrī, K., ed. 1938. The Udayavarmacarita. Trivandrum Sanskrit Series, no. 133. Trivandrum: Government Press.

Sanderson, A. 1988. “Śaivism and the Tantric Traditions”. In The World’s Religions, eds, S. Sutherland, L. Houlden, P. Clarke and F. Hardy, pp. 660–704. London: Routledge and Kegan Paul.

Sanderson, A. 2007a. “The Śaiva Exegesis of Kashmir”. In Mélanges Tantriques à la Mémoire d’Hélène Brunner, eds, Dominic Goodall and André Padoux, pp. 231–442; pp. 551–582. Pondicherry: Institut français d’Indologie/École française d’Extrême-Orient.

Sanderson, A. 2007b. “Atharvavedins in Tantric Territory: The Āngirasakalpa Texts of the Oriya Paippalādins and their Connection with the Trika and the Kālīkula, with Critical Editions of the Parājapavidhi, the Parāmantravidhi, and the *Bhadrakālī-mantravidhiprakarana”. In The Atharvaveda and its Paippalāda Śākhā: Historical and Philological Papers on a Vedic Tradition, eds, Arlo Griffiths and Annette Schmiedchen, pp. 195–311. Aachen: Shaker Verlag.

Sanderson, A. 2009. “The Śaiva Age: The Rise and Dominance of Śaivism During the Early Medieval Period”. In Genesis and Development of Tantrism, ed. Shingo Einoo, pp. 41–350. Tokyo: Institute of Oriental Culture, University of Tokyo.

Sarma, S.A.S. 2010. ‘Why Have the Later Ritual Manuals of Kerala forgotten “Initiation”?’ Acta Orientalia: Acadiae Scientarium Hungaricae, 63(4): 443–454.

Soudararajan, J. 2004. Naṭarāja in South Indian Art. Delhi: Sharada Publishing House.

Srinivasan, P.R. 1986. Annual Report on Indian Epigraphy for 1975–76. New Delhi: Archaeological Survey of India.

Staal, F. and R. Gardner, eds 1976. Altar of Fire [film]. Berkeley, CA: University of California Extension Media Center.

Staal, Frits. 1984. Agni: The Vedic Ritual of the Fire Altar. Delhi: Motilal Banarsidass.

Stein, B. 1994. Peasant, State, and Society in Medieval South India. Delhi: Oxford University Press.

Tirumumpu, T.S. 1975. Śrī Bhadrakāḷī Māhātmyam athavā Dārukavadham (With Text and Malayalam Translation). Trikkarpur, Kerala: The Author.

Unni, N.P. 2006. Tantra literature of Kerala. Delhi: New Bharatiya Book Co.

Venugopalan, P.V., ed. 2009. Malayalam Lexicon. Volume VIII. Thiruvananthapuram: University of Kerala.

Viṣṇu Nampūtiri, M.V., ed. 1981. Tōṟṟampāṭṭukaḷ. Kottayam, Kerala: National Book Stall.

Viṣṇu Nampūtiri, M.V., ed. 1993. Tiruvarkāṭṭu Bhagavatittōṟṟam (Tōṟṟampāṭṭu). Kottayam, Kerala: National Book Stall.

Viṣṇu Nampūtiri, M.V., ed. 1998. Pūrakkaḷi. Kottayam: Current Books.

Viṣṇu Nampūtiri, M.V., ed. 2005. Nammuṭe Paṇṭattepāṭṭukaḷ. Kottayam: DC Books. Distributors, Current Books.