Version classiqueVersion mobile

Bilingual discourse and cross-cultural fertilisation: Sanskrit and Tamil in medieval India

Whitney Cox
Vincenzo Vergiani

Section III. Written in stone? Shifting registers of inscriptional discourse

Words for Worship: Tamil and Sanskrit in medieval temple inscriptions

Leslie C. Orr

Texte intégral

I am grateful to the organizers, Whitney Cox and Vincenzo Vergiani, and to the other participants in the Cambridge workshop where the papers in this volume were first presented, and have benefited greatly from the suggestions I received and the lively discussions on that occasion. Many of the participants of the workshop are also regular conversation partners on matters epigraphical and textual whose expertise and insights have been of great value to me. As always, in my study of the Tamil inscriptions, I am indebted to the Epigraphy Office of the Archaeological Survey of India, in Mysore, and particularly to S. Swaminathan. Special thanks are due to Kannan M., whose invitation to contribute a paper on Tamil and Sanskrit in the inscriptions resulted in an earlier foray into some of the material presented here.

1South Indian inscriptions constitute a field within which there are many opportunities to explore the interactions of Tamil and Sanskrit, both because there is a corpus, stretching across many centuries, of numerous epigraphical documents to examine and because these documents deal with a wide range of matters — from poetic to pragmatic. Other contributors to this volume demonstrate how the inscriptions can be employed to answer questions about the languages and modalities of law and of eulogy. The field of inquiry in the present chapter is the milieu of the temple — where, in fact, we find so many of the inscriptions themselves — and I propose to provide a preliminary analysis of the modes employed, and of their Tamilness and/or Sanskritness, when medieval Tamil inscriptions talk about the objects, acts, and spaces of worship.

  • 1 See Davis and Orr (2007) for a comparative study of how participants in temple festivals are repres (...)
  • 2 My sense of the Āgamic vocabulary is derived from an examination of the Ajitāgama, Mahotsavavidhi, (...)

2One of the central themes of this volume is the creation of technical vocabularies and discursive strategies in Tamil which are parallel or equivalent to Sanskrit systems. In a consideration of words for worship in medieval Tamil inscriptions, the obvious parallel Sanskrit system is that of the Āgamas — particularly those later Āgamas, dating from perhaps the 12th century onward and showing the impress of South Indian culture, in which temple ritual, rather than individual spiritual pursuits, is increasingly central (Goodall 2006). A detailed comparison of the language of the inscriptions and that of the Āgamas is beyond the scope of this essay, although it would be an immensely valuable undertaking, which could suggest answers to questions about the influence of ritual and architectural practices on the production of the medieval Āgamic texts — or vice versa.1 Instead my focus is on the Tamil inscriptions themselves, but I examine the language of the inscriptions against the backdrop of the Sanskrit vocabulary for worship provided by the Āgamic corpus — a vocabulary that seems quite consistent across the sectarian boundaries between Śaiva, Pāñcarātra, and Vaikhānasa, and a vocabulary that would have been available to at least some of the people of the medieval Tamil temple milieu.2

1. Ways to write about worship

3In the Appendix are six texts which provide an idea of how words for worship appear in the context of the Tamil inscriptions. These six texts appear in inscriptions ranging from the 8th to the 20th centuries — and only the oldest and the most recent are given in their entirety. It is important to emphasize that these texts are not in any way “representative”, but rather suggestive and illustrative. They are not necessarily typical of the time and place where they were produced, and they are certainly atypical in terms of the identity of the donors whose gifts are described, half of whom are royal figures — a Pallava queen, a Chola queen, and a Pandya king — since the vast majority of temple patrons through most of medieval Tamil history have not been members of the ruling houses. My choice of texts is also misleading inasmuch as they are all engraved on the walls of temples dedicated to Śiva; it is the case, however, that the inscriptions of Vaiṣṇava and Jain temples are very similar to the Śaiva ones in terms of their form, content, and vocabulary.

  • 3 According to K.G. Krishnan (1990), linguistic mixing has been a feature of epigraphical practice in (...)

4The texts I have chosen are typical of Tamil stone inscriptions in their use of a mixed Tamil-Sanskrit language — an inscriptional maṇipravāla in which Sanskrit words are combined with Tamil ones.3 Sanskrit words — or even single syllables — will be written either in Grantha script (indicated by italics in the Appendix) or transliterated into Tamil script, or the Vaṭṭeḻuttu script used for Tamil, especially in inscriptions of the far south or of early medieval times. Some Grantha characters — for example ś, as in śrī or īśvara — are so ubiquitous in Tamil inscriptions that they must be considered Tamil rather than Grantha. All six of the inscriptions in the Appendix use Grantha letters, except for TEXT F — but even here, in the “pure Tamil” of the 20th century, we find the Tamilized form (tarmam) of the Sanskrit word dharma, meaning, in this context, an endowment. Twelve centuries earlier, in the inscription from Kanchipuram presented as TEXT A we see the word dharma in Grantha used in precisely this same sense, in the last line of the inscription; however, the same word as part of the name of the queen Dharmamahādevi is transcribed both in Grantha and in Tamil (taruma), suggesting complete interchangeability in script usage in this case.

5Whitney Cox (2010) has recently suggested that the choice of using one writing system rather than another (just as the choice of language) deserves more detailed study in terms of its cultural-political implications. This certainly bears consideration in the case of Tamil inscriptions — not only the uses of Grantha, but of Vaṭṭeḻuttu as well. Sylvain Brocquet (1997) argues that the Grantha script used in the stone inscriptions of the Pallava period would have had a powerful visual impact, enhancing the expressive qualities of the Sanskrit used in these records. But in the profoundly mixed language of later and less magnificently engraved inscriptions, this consideration is of little use to us as we seek an explanation for why Grantha was or was not employed. It may be that the use of Grantha in a Tamil inscription signals a self-conscious appropriation of Sanskrit — a display of learning or sense of connection with a trans-regional cultural paradigm on the part of the scribe (as is perhaps demonstrated in TEXT E, from 15th-century Tenkasi) — although if, as is likely, the ur-text were a palm leaf document, it is difficult to know who was meant to be impressed (other scribes?). Or it may be (as is suggested by TEXT A) that these writing forms were freely interchangeable and not loaded with a significance derived from their Sanskrit or Tamil derivation or appearance.

6Before turning to the examination of Tamil texts engraved in stone, like those in the Appendix, I would like to look briefly at two bilingual copper-plate charters that pre-date our inscriptions and that (unusually for copper-plates) record the support of temple worship. The fact that we have both Tamil and Sanskrit descriptions of the grant provide us with an opportunity to compare the two modes of expression literally side by side.

7The Pallankoyil copper-plate grant (PCM 1 = IP 17) — dated by some scholars to as early as the mid-6th century — may contain the earliest inscriptional reference to worship in Tamilnadu. Although unearthed in Tanjavur district, this inscription concerns an endowment of land made by the Pallava king Simhavarman to a Jain ascetic of Tiruparuttikkunram, or “Jina Kanchi” in present-day Chingleput district. The Tamil portion of the inscription says only that the giftwas made as paḷḷiccantam — i.e. a land grant to a Jain institution — while in the Sanskrit portion, the object of the giftis made clear: to provide for arrangements for worship (pūjā) for “the multitude of the greatest among jinas, jainas and munis” (Subramaniam 1958).

  • 4 Note that the word devakarman also appears eight centuries later in TEXT E.

8Another bilingual Pallava royal grant, an endowment for a temple and a brahmadeya near Kanchipuram, dating from the late 7th century, is the Kuram copper-plate inscription (SII 1.151 = IP 46). This inscription presents a similar picture in terms of the division of labor between Tamil and Sanskrit: the Tamil section of the copper plate is relatively terse and provides only a general idea of how worship is to be offered, while the Sanskrit portion provides much more detail. In the Tamil portion of the Kuram plates, it is said that provision is to be made for worship (expressed by the good Tamil word vaḻipāṭu, literally “proceeding on the [proper] way”), customary offerings (oḻukkavi, with Tamil oḻukkam having a meaning similar to that of vaḻi in the preceding expression, combined with the Sanskrit word havis in its Tamil form avi), and devakaruman, acts for god or worship duties, again based on Sanskrit.4 In the Sanskrit portion of the Kuram plates, we find that the specific offerings for the temple deity are listed — pūjya-snapana-kusuma-gandhadhūpa-dīpa-havir-upahāra-bali-śaṃkha-paṭaha-ādi: “worship, bathing, flowers, perfumes, incense, lamps, food offerings, bali, conches, drums, etc.” While this list seems as though it could have been extracted from an Āgamic text, it is noteworthy that the standard term in the Āgamas for food offerings — naivedya — is absent. Instead, as in the Tamil portion, the word havis appears, and its meaning of “sacrificial oblation” seems to have been modified so that both the Sanskrit word and its Tamil form are now appropriate for use within the context of temple worship to designate food offerings.

  • 5 The word havis / avi is used to designate food offerings in the context of Jain worship in the 9th (...)
  • 6 Uthaya Veluppillai demonstrates in her forthcoming doctoral thesis that the phrase amutu cey, which (...)

9Such usage in 7th-century Tamilnadu was evidently relatively new, since in Tēvāram, the word avi is used exclusively with reference to sacrificial oblations. There are several occurrences of avi in inscriptions of the 9th century, and it is noteworthy that it is used in connection with worship at Jain as well as Hindu temples.5 But this term disappears after the 9th century, to be replaced by amutu (e.g. in SII 12.58 of the 9th century), with its variant forms amirtu (SII 7.523, 9th century) and amitu (SII 14.16a, 9th century) — a term that is ubiquitous in medieval Tamil inscriptions, first appearing in the 8th century (e.g. in SII 8.331; cf. also the 10th-century TEXT B). Derived from Sanskrit amṛta, “nectar of immortality”, it takes on a new and very specific meaning in the temple milieu of medieval Tamilnadu. And as with avi, this meaning within the inscriptional context appears to be different from its use in Tamil literature of the same or earlier periods.6

  • 7 In the “Tamilization” of Sanskrit words, a final long a becomes ai. Thus pūjā becomes pūcai.

10Although the word pūjā appears in the Sanskrit sections of both the Pallankoyil and Kuram copper-plate grants, it is not taken up very quickly in the Tamil inscriptions. There are a few appearances of the Tamil word pūcai or pūcaṉai as early as the 9th century, but the word only gains currency in the 12th century.7 (It appears once in the Appendix, in the 15th-century TEXT E.) Instead, the more commonly occurring words for worship — for what priests do — are other terms borrowed from Sanskrit: arccaṉai, which we see in TEXT A and which is found in 8th-century inscriptions in the far south (e.g. SII 14.17) as well as the northern area of the Tamil country, and ārādhanā, together with the related verb ārādh, Tamilized as tiruvārātaṉai and ārāti, which also appear, beginning in the 8th century (SII 6.448 and 6.356), and which we find in TEXT B, from the 10th century.

2. Spaces and structures for worship

  • 8 In keeping with the orthography of the Tamil inscriptions, where long e and o are not distinguished (...)
  • 9 In 9th-century inscriptions, we also find the form viṣṇugraham — e.g. in SII 12.47 and SII 7.420.

11If worship itself is referred to in the Tamil inscriptions by Sanskrit-derived terms, the term for the venue where worship activities take place — the general word for “temple” — is almost invariably Tamil. The most frequently occurring terms are taḷi (as we see in TEXT A and in TEXT B — in the latter, we have tirukaṟṟaḷi, “sacred stone temple”) and koyil (used in TEXTS C and E).8 Both of these words begin to appear as early as the 7th century (taḷi in the Kuram plates, discussed above; koyil in EI 32.23, of the second half of the 7th century) and continue to have widespread currency in subsequent times. Meanwhile, there are also some Sanskrit-derived terms, notably gṛha — “house” — which is used in Tamil as it is in Sanskrit to refer to temples (also geha), although in Tamil inscriptions it has a relatively restricted use. In TEXT A, the name of the temple — dharmmamahādevīśvaragra[*ha]ttu — incorporates “graham” (if the epigraphist’s emendation is correct), and we see a similar usage in the Kuram plates of the late 7th century with the name for a Siva temple of parameccuragaram (see also the early 9th-century SII 4.132 for the use of gruham to mean temple). But the main place where the Tamilized graham/garam finds a place is in the nomenclature for Viṣṇu temples — viṇṇu + karam, “Viṣṇu’s house,” producing viṇṇakaram. This usage is found in the Nālāyira Tivyap Pirapantam in Tirumaṅkai āḻvār’s poem on the Vaikuntha Perumal temple in Kanchipuram, as well as in inscriptions of the early 9th century (SII 4.131) — which may be more or less contemporary with Tirumaṅkai’s composition — and, by the end of the 9th century (e.g. in SII 8.633), the word is shortened to viṇṇakar.9

  • 10 The term garbhagṛha is found more among the later inscriptions of the Tamil corpus. Tamil translite (...)
  • 11 It may be that the relevance of the word nāḍikā to the temple sanctum arises from its denotation in (...)

12The Tamil inscriptions rarely make use of gṛha in the compound garbhagṛha, and when we do find this latter term or its Tamil variants — as in inscriptions from the 12th and 13th centuries (ARE 1924, no. 121 and 1922, no. 554) and in the 15th-century TEXT E (where it is spelled in Grantha letters) — it appears to refer to the central shrine in its entirety (from upāna to stūpi, as TEXT E says), as a term interchangeable with vimāna, rather than to the interior chamber housing the central image.10 The word more commonly used to denote the temple’s inner sanctum, uṇṇāḻikai, appears in inscriptions as early as the 8th century (for example, in TEXT A). This is a truly hybrid word, composed of Tamil uḷ (“inner”) together with Sanskrit nāḍikā (cf. also Skt. nālika). It is difficult to understand, however, how the Sanskrit word — which has the meaning of tube or channel (or a measure of time of twenty-four minutes, which is also one of the meanings of Tamil nāḻikai) — can possibly be an appropriate borrowing here.11 Evidently there has been a major shift in meaning for the Sanskrit term in the Tamil milieu — a shift which has also taken place in medieval literature, at least in part, since there is at least one occurrence of the word uṇṇāḻikai with its inscriptional meaning in Tēvāram 6.51.3, where Śiva is described as uṇṇāḻikaiyār, the one in the inner sanctum. A variant of this composite word, found in inscriptions from the 9th century onward is akanāḻikai (e.g. the 9th-century TAS 1.6 and the 10th-century ARE 1933, no. 233), where the first element is akam, another Tamil word meaning “inner”. And in our TEXT E, we find yet another, less common usage — here there is the statement that the garbhagṛha and the ardhamaṇḍapa have been built as a central shrine (iṭaināḻikai).

  • 12 TKC says that tānam is used as a synonym for koyil — temple — beginning in the 10th or 11th century

13The Sanskrit words that seem to be most commonly used in the Āgamas to denote a temple site — not necessarily a building, but a place where worship is carried out — are sthānam and ālaya. The first of these words, in Sanskrit or in its Tamil form tāṉam, is encountered quite frequently in medieval inscriptions — again more with the sense of a sacred space in general than of a constructed one. Such usage of tāṉam starts as early as the 7th and 8th centuries (IP 42 and SII 14.27). In the earlier inscriptions, we often find the word used in the expression mūlattāṉam, which we may consider to have the meaning “root place” or “original foundation”; the earliest appearance of this expression is as part of the name of god — e.g. Tirumūlattāṉattuppaṭārar, “the Lord (Śiva) of the original foundation”.12 Ālaya, on the other hand, although common in the Āgamas, is scarcely ever found in Tamil inscriptions. There are a scattering of references in inscriptions of the 8th to 10th centuries, where we find the usage jinālaya for a site sacred to Jains (ARE 1934, no. 35/no. 22) — after which the term drops out of view altogether, until we come across an inscription of the 15th century at Tenkasi, which describes the temple complex as a whole, from garbhagṛha to gopuram, as an ālaiyam (SII 26.554).

14In the Āgamic literature, the temple as a built structure — as opposed to the temple as a site — may be referred to as ālaya, but more often we find the words vimāna (especially in Vaiṣṇava texts) and prāsāda or harmya (in Śaiva texts). The only one of these terms that occurs at all in the Tamil inscriptions is vimāna — usually in the form śrīvimāna. We find this term in TEXT B (of the 10th century) with reference to repairs made to the vimāna and in TEXT D (of the 12th century), which records the building of a vimāna. While appearing as early as the 9th century (e.g. SII 12.55), this word for the temple as a building occurs rather infrequently in the Tamil inscriptions. Instead, we seem to find Tamil or Tamilized words, either rather general terms like taḷi or koyil that have no specific structural referent (although they may often be considered to be equivalent to the vimāna) or terms that designate particular zones or constructions within the temple compound.

  • 13 Among the early references māḷikai is found in an 8th-century inscription (SII 2.73) and tirucuṟṟāl (...)
  • 14 In a 12th-century inscription from Tirunelveli district in the far south of Tamilnadu, we find tiru (...)
  • 15 For example, in SII 3.1092, it is said that the piṟākāram was rebuilt. Another inscription of the 1 (...)

15Among these is tirumuṟṟam, a term for the temple courtyard which is found in our TEXT B, that seems thoroughly Tamil (a second 10th-century reference for this word is EI 7.20H). The wall (or walls) enclosing the temple compound, and the circumambulatory passage within these walls, are most often called cuṟṟālai or māḷikai, or some variant or combination of these words. That these terms, or others denoting an enclosing wall, are rarely met with before the 12th century is perhaps more an indication of the history of the growth of the South Indian temple complex than it is linguistically significant.13 Cuṟṟu, a good Tamil word meaning to encircle or surround, seems highly appropriate within this context. It is paired with the Sanskrit-derived māḷikai from as early as the 10th century (SII 17.617). A bilingual stone inscription of the early 12th century at Chidambaram (SII 4.225), listing the gifts made by the lord Naralokavīraṉ, mentions that one of these is a tirucuṟṟumāḷikai — which appears to correspond either to sāla or to prākāra in the Sanskrit portion, both words meaning “wall”. With māḷikai we have a case where a Sanskrit word, mālaka — with the meanings of a “storeyed house” or “things arranged in a line” — is taken up in early Tamil literature, becoming māḷikai and being used to mean “mansion” or “upper storey”, but in the inscriptional context making a semantic turn to become the equivalent of prākāra, and to suggest something a bit grander and more elaborate than simply a wall, perhaps a gallery or multi-level construction.14 Māḷikai is combined not only with cuṟṟu but also with the Tamil naṭai — meaning walking, a route, or corridor — such that tirunaṭaimāḷikai, found in inscriptions from the 12th century onward, points toward the ritual use of the enclosed space as a circumambulatory or processional path. As for the word prākāra itself (Ta. piṟākāram), this appears occasionally in the Tamil inscriptions, side by side with the more Tamil usages, but in its earlier occurrences it seems to refer (like tirumuṟṟam) to the space enclosed, only in the 13th century (also) designating the enclosing wall.15

  • 16 Kōpuram appears fairly regularly in Tēvāram, but in that context refers to a tower more generally r (...)
  • 17 We also find olakka kūṭam, with the same meaning, for example in two inscriptions of the 11th centu (...)
  • 18 KMK, p. 27, which may date from the middle of the 12th century, contains the phrase upānāti stūpipa (...)
  • 19 An inscription of 1447 says that the ploughing ceremony was performed by the king about a year befo (...)

16The characteristic South Indian temple tower is almost always referred to in inscriptions as kōpuram — derived from Sanskrit gopuram — from the 9th century onwards (SII 3.91), and it is often spelled as a Sanskrit word even in early inscriptions as well as in later ones such as our 15th-century TEXT E.16 The Tamil tiruvācal as a term for the temple gateway seems to gain currency somewhat later, starting in the 12th century. Also embraced in inscriptional Tamil is the Sanskrit word maṇḍapa, which appears as a Sanskrit word in Grantha in the 8th century (SII 6.356), and in its Tamil form, maṇṭakam, even earlier (for example in the 7th-century Kuram plates — SII 1.151). Starting in the 10th century we find references to ardhamaṇḍapas and mahāmaṇḍapas, and in the 11th and 12th centuries to the tiruvolakkamaṇḍapa — the audience hall — as we see also in TEXT E.17 The 15th-century TEXT E provides us with still more technical architectural language when it is said that the garbhagṛha of the Viśvanātha temple at Tenkasi was built under the Pandya king’s sponsorship from upāna to stūpi, i.e. from the base to the crowning finial. There are perhaps a couple of examples of the use of this “upāna to stūpi” phrase in the 12th and 13th centuries, but it seems to be especially in fashion in the Tamil inscriptions of the 14th to 16th centuries, particularly those dated in the reign of the later Pandyas based in the western Tirunelveli district.18 The Tenkasi inscriptions that concern themselves with temple building, like TEXT E, have not only an unusual degree of architectural specificity, but also on occasion make reference to one of the details of the ritual of consecration, stating that the sponsor of the temple — the king — participated in the preliminary ploughing ceremony, using the Sanskrit word karṣaṇa or the Tamilized form kariḻcaṉai.19

  • 20 Although Ajitāgama has an entire chapter devoted to karṣaṇavidhi, the topic receives little attenti (...)
  • 21 In this connection, it is noteworthy that the term for temple consecration now in very common use i (...)
  • 22 This seems to be the term used in the earliest Tamil inscriptions that record temple foundations, f (...)
  • 23 This is an expression found in inscriptions of the 12th and especially 13th centuries (e.g. IPS 124 (...)

17The inscriptions’ specific mention of this ritual — rather than the other procedures associated with temple consecration that loom much larger in the Āgamic texts20 — provides a clear indication of the inscriptions’ orientation toward the concerns of donors and temple patrons, and not toward priestly responsibilities involved in the foundation of temples. But in fact the inscriptions rarely record the establishment of temples, nor do most temples in the Tamil country have foundation inscriptions. It may be worthwhile, however, to examine the language used in the cases where the building of a temple is recorded — as in TEXT D. Here we find the expression mahadevarai tirupratiṣṭhaiyum paṇṇi śrī vimānamum eḻuntaruḷivitta — the donor caused the vimāna to graciously appear and arranged for the consecration (tirupratiṣṭhai) of Mahādevar. Evidently the Tamil expression eḻuntaruḷivitta is applied to the establishment of the temple by the donor, and Sanskrit pratiṣṭhā (pratiṣṭhai in its Tamil form) to the consecration of the deity within. But in an inscription at the same temple as TEXT D, engraved eleven years earlier (SII 26.4 = IPS 132), we find that precisely the same act of establishing the vimāna (and the Śivaliṅga), by the same donor, is recorded using the term pratiṣṭhai, suggesting that the Tamil and Sanskrit expressions are interchangeable. This usage also suggests that pratiṣṭhā, when employed in inscriptions recording the foundation of temples, typically does not refer to the ceremonial consecration conducted by priests, but rather to the actions carried out by the donor.21 This suggestion is borne out by terms found elsewhere that record the establishment of temples — all of them Tamil: the donor eṭupitta (caused [it] to be built),22 ukantaruḷuvitta (caused [it] to graciously rise up),23 and, as we have already seen, eḻuntaruḷivitta (caused [it] to graciously appear). These last two expressions point not only toward the role of the donor in temple foundation but even more they emphasize the importance of the deity by whose grace the building has come into being.

18When it comes to the setting up of images, we again find the term eḻuntaruḷivitta — by far the most frequently encountered in this context — conveying this same notion. Although both eḻuntaruḷivitta and pratiṣṭhā, referring to the establishment of images, are found in Tamil inscriptions from the 9th century onward, the latter term is much less common, at least in the early medieval period (pratiṣṭhā seems to be over-represented among the inscriptions in the Appendix). Our TEXT C, of the 12th century, may indicate that the two terms were not entirely synonymous in the case of the setting up of images: in this inscription, eḻuntaruḷivitta is applied to the donor’s gift of the image of the goddess to the temple, while pratiṣṭhai is apparently paired with kalyāṇam, the ceremony of marriage, as a celebration being sponsored by the village assembly. This indication of a ritual dimension for the pratiṣṭhā of images provides us with a rare glimpse of the ceremonial consecration that was, no doubt, a part of the establishment of images, but that is scarcely ever attested by the inscriptions. It is significant, perhaps, that there is no Tamil term that reflects the ritual aspect of the setting up of images, and it is the Sanskrit term — of very frequent occurrence in the Āgamas — that is pressed into service in this context.

  • 24 Willis’ (2009) argument — based on his reading of Sanskrit Gupta-period inscriptions — that an imag (...)
  • 25 Tirumeṉi is also used in the inscriptions to refer to the person of the king, when, for example, a (...)
  • 26 In the bilingual stone inscription of Naralokavīraṉ at Chidambaram (SII 4.225), which I have alread (...)

19As for words for the objects of worship, the Sanskrit vocabulary of the Āgamas is very rich: we have pratimā, mūrti, bera, bimba, and more (see Smith 1975; Brunner 1990; Colas 1996). A few of these words appear in the Tamil inscriptions. For example, pratimā (Tamil paṭimam) appears as early as the 8th century (e.g. EI 4.14A) — and apparently especially in the context of Jain worship. And we find mūrti from the early 11th century onward (e.g. SII 2.30).24 More frequently met with than either of these Sanskrit terms is the Tamil tirumeṉi, appearing in inscriptions from the 9th century onward (e.g. SII 8.581), if not earlier.25 But the most common way of referring to images in Tamil inscriptions is without words at all. So, for example, none of the texts in the Appendix has a word for “image,” even though two of them (TEXTS C and D) refer to the establishment of an image. It is not a divine form, but the deity her or himself — nācciyār or mahādevar — who is the subject (eḻuntaruḷi) or object (mahādevarai) of consecration in the Tamil inscriptions.26

3. Making offerings and organizing worship

  • 27 Interestingly, both Sanskrit vitāna and Tamil vitāṉam mean not only “canopy” but can also denote a (...)
  • 28 In the Tamil inscriptions some of these more standard Sanskrit terms for bathing are actually found (...)

20We find in TEXT B quite a full account of the services and offerings made to the gods enshrined in the temple built by Queen Cempiyaṉ Mahādevi, in the late 10th century. In this excerpt from a larger inscription (SII 13.170), we see that provision was made for food offerings (tiruvamutu), the sacred bath (tirumañcanam), water offerings (taṇṇiramutu), anointing of the deity (tirumeypūccu), and the bearing of the canopy (vitāṉam), and further arrangements made to support three garland-makers indicate that garlands (tiruppaḷḷittāmam) also figured in the regular offerings and adornments required in the temple. The first item in the list of offerings is tiruvamutu, a Sanskrit-derived term, but one — as we have established above — that has taken on a particular meaning in the context of inscriptional Tamil, and of medieval temple life, which is quite distinct from both its Sanskrit root and its usage in early Tamil literature. With two of the other services mentioned — tirumeypūccu and vitāṉam — we have in the first instance a purely Tamil expression, “the anointing of the sacred body,” and in the second a Tamil word borrowed directly from Sanskrit.27 In the part of the inscription reproduced as TEXT B, we find two other terms: tirumañcaṉam (sacred bath) and tiruppaḷḷittāmam (garland). However Tamilized they are, both of these terms have Sanskrit words at their core — majjana and dāma — prefixed by tiru, and in the case of dāma, also by paḷḷi (which seems to mean “royal” in this context). Nonetheless, these Tamil words are not related to the Sanskrit words that are employed by the Āgamic texts to denote these temple services, where we find kusuma or puṣpa for flowers and snapana, snāna or abhiṣeka for ritual bathing.28 Tirumañcaṉam and tiruppaḷḷittāmam, even if they have Sanskrit roots, appear to have developed independently from the Āgamic discursive milieu.

  • 29 To nantal, meaning “perishing,” is added the negative suffix –ā producing the word nantā, meaning “ (...)
  • 30 The terms for the four cantis found in this inscription (SII 13.170) are a wonderful mixture of “pu (...)

21In other parts of the lengthy and fragmentary inscription which is excerpted in TEXT B, considerable attention is given to provisions for burning lamps, which is a type of gift very frequently recorded in temple inscriptions, particularly those of the 9th and 10th centuries. Three types of lamps are mentioned in fragments #3 and #6 of the inscription: perpetual lamps (tirunontā viḷakku), lamps for daily services (sandhi viḷakku), and garlands of lamps (dīpamālai). The last of these expressions is purely Sanskrit, but it is in fact quite unusual to encounter such terms as dīpa (Tamil tīpam) in the inscriptions, and other Sanskrit terms for lamps and lamp services used in the Āgamas, such as ārātrika (“night lamp”) and nīrājana (“waving of lamps”), are not found at all. The Tamil word viḷakku is, on the other hand, ubiquitous — and very often paired with the Tamil nontā (or variants nuntā or nantā).29 The word canti (or sandhi) also frequently appears in connection with the donation of the means to fuel lamps; in this case the lamps are meant to burn at the time of a specific daily service — a canti. From its Sanskrit base in the denotation of the three sandhyās — dawn, noon, and evening, at which time it is appropriate to offer worship — the word in the Tamil inscriptional context rapidly developed into a temple-specific technical term, denoting the occasions for daily worship whether these numbered three, five, or more. In the inscription recording Cempiyaṉ Mahādevi’s patronage of the temple at Kuttalam, arrangements are made for food offerings four times daily (fragments #4 and #5).30

  • 31 The term bali appears in the Sanskrit portion of the 7th-century Kuram plates (SII 1.151 = IP 46), (...)

22Among the various people providing service in the temple who are mentioned in this inscription — priests, “flywhisk women”, singers of hymns (tiruppatiyam), Brahmans reciting Vedic texts, and many others — drummers loom particularly large, as we see in TEXT B. The drummers (ukaccu/uvaccar) are implicated in a round of daily temple observances somewhat different from those associated with the cantis; they are required to perform at the time of the sacred bath (tirumañcanaṉam), and at the times of the food-offerings (tiruvamutu), bhūtabali, śrībali, the midnight service (arddhayāmam), and the service of waking the god (paḷḷiyeḻucci). The mention of the śrībali ceremony (and of the role of drummers in its daily celebration) is very common in the Tamil inscriptions, and the Sanskrit word śrībali, or its Tamil equivalent tiruppali, is found in inscriptions from the 8th century onward (e.g. IP 71 = SII 8.331).31 Here we have a case where there is a close correspondence between the Āgamic vocabulary and the inscriptional one, although (as with sandhyā) the Sanskrit word seems to have slipped away from its earlier sphere of application to become useful within the specific framework of temple ritual. The range of services and substances offered to the deity on a daily basis are referred to in the Tamil inscriptions by a rich assortment of terms in which we find various mixtures of Tamil and Sanskrit. But temple festivals are almost always referred to by Tamil terms tiruviḻā or tirunāḷ — as in TEXT B, although Sanskrit utsava makes an occasional appearance, as we see in the 15th-century TEXT E, as also in inscriptions as early as the 10th century. And Tamil words are used by preference for processions, with the expression eḻuntaruḷ, “graciously appear” — a term we have already encountered in the context of consecration of images — being virtually the only one employed (Orr 2004).

  • 32 There are two examples of the use of ubhayam in the 9th and 12th centuries, but most references to (...)

23In both TEXT A of the 8th century and TEXT F of the 20th century, we find the Sanskrit word dharma, and/or its Tamil forms tarumam and tarmam, used to mean “endowment”. The Sanskrit word nibandha is used in the 10th-century TEXT B (in fragment #10) and, as perhaps also with the word dharma, there seems to be a narrowing of the semantic field of Sanskrit nibandha (binding, attachment, root, restraint, grant of property) to result — in Tamil — in the sense of endowment or obligation. Variants of this word — nimantam and nivantam — are very common in the Tamil inscriptions from the 10th century onward, and demonstrate further semantic extensions, including the meanings of arrangement, temple service, agreement, and allotted duties. Within a few centuries, there is a proliferation of words in the Tamil inscriptions denoting various kinds of property transactions, including temple donations, and especially words referring to various rights in property or modes of recognition of those who were granted property. So starting in the 13th century we begin to see the Sanskrit term ubhaya and the Tamil term kaṭṭaḷai designating endowments.32 And words like koṟṟu (daily allotment of rice — which we see in TEXT A), kāṇi (rights in property), and paṅku (shares) or nāḷ (“days” — see texts D and E), which are common in earlier inscriptions, are gradually replaced, especially from the 15th century onward, by kāṇiyāṭci, mariyātai, carvamāṉiyam, and cutantiram — words that designate rights and privileges for temple personnel. The first of these terms is Tamil, formed by the addition of āṭci (“ownership”) to the word kāṇī that was in use earlier. The others are clearly derived from Sanskrit words — maryādā (custom, agreement, propriety of conduct), sarva plus mānya (to be honored), and svatantra (independence) — although there is a considerable shift in meaning from the “original” Sanskrit in the case of all of them.

4. Concluding discussion

24What do we make of all these words for worship? When compared to earlier or contemporary Tamil literature, our words for worship seem to constitute a vocabulary in large part specific to and developed within the inscriptional context. Even those words that appear in both contexts — avi, amutu, kōpuram — have different meanings and application in the inscriptions than they do, for example, in Tēvāram. And we see that from their very earliest appearance in the inscriptions of the 7th century, many of the Tamil words employed, like these, have Sanskrit roots. But very few of the Sanskrit base words are part of the vocabulary of temple worship found in the Āgamic literature: in these texts there is no amṛta, majjana, dāma, or nāḍikā. Further, the Sanskrit words that have been taken up into the Tamil of the inscriptions have often undergone considerable alteration in their meanings. Sometimes this is quite a radical shift, as in the case of amutu and nāḻikai. In other cases, we have the adaptation of a Sanskrit word and a restriction of meaning in its use in the Tamil inscriptions — as with gṛha and māḷikai. Or we find that there is an expansion in a word’s semantic field, as in the case of canti and nibandha.

25Sheldon Pollock has given us a great deal to think about with reference to language use — whether Sanskrit or vernacular — in terms both of choice and of change. To consider first the question of change over time, one would be hard-pressed to discern a trend in the Tamil inscriptions. If forced to choose between the two processes of vernacularization and Sanskritization, I would probably come down on the side of Sanskritization, since in the case of at least some vocabulary items, particularly those to do with temple architecture and with gifts of and rights to land or honors, we seem to see an increasing use of Sanskrit. At the same time, however, such Sanskrit terms as ubhaya and svatantra are used in the Tamil inscriptional context with scant regard for their Sanskrit meanings, taking their place within the particular vocabulary of temple affairs. And inasmuch as Sanskrit is imbedded in the “vernacular” language of the inscriptions from the very beginning, and since this language continues throughout the medieval period to be thoroughly mixed, it is not possible in this context to trace the kinds of processes over time about which Pollock has written — or to disprove him either.

26So let us look at the question of choice: why choose Sanskrit? Why use Sanskrit words — ārātinai and arccaṉai — for worship and Tamil ones — koyil and taḷi — for the place where worship takes place? Is it significant that in the two early bilingual copper-plate grants concerning temple worship, it is in the Sanskrit portion of the inscription that we find a more explicit indication of the fact and the character of worship? If there is some prestige or even sacrality associated with the use of Sanskrit in this context, from what precise sources does this derive? One thing seems clear: the Āgamic texts are not the basis for such Sanskrit usages or any cachet they may have had. The lack of fit between the inscriptional and textual corpora in the vocabulary used, the shared use of words for worship in Jain and Hindu inscriptional contexts, and the indications that terms and practices migrated from temples to texts rather than vice versa — all of these factors argue against this notion. Daud Ali (2011) has recently emphasized the necessity of considering how Sanskrit as a language of prestige and power relates to specific local contexts and political practices. He has found, in the complex distribution of Sanskrit and vernacular lexical items in Indonesian inscriptions, “a deep entanglement of Sanskritized ‘cosmopolitan’ elements with local usage”… “While Sanskrit words denoted some of the most quotidian elements of everyday administrative and material culture (belying any claim for any linguistic division of labour), some of the most exalted and highly symbolic terminology was expressed in indigenous vocabulary” (pp. 288–89). This is very much the case for the Tamil inscriptions I am considering, as well, and the logic of language choice is not at all transparent.

27One point of entry to this puzzle may be the preferential use of eḻuntaruḷivitta rather than pratiṣṭhai, and the virtual absence of words for “icon” in the Tamil inscriptions. The form of the inscriptional discourse seems to reflect a particular perspective — that of the devotee and donor, who assists in making present a particular manifestation of god, for whom the image is god, and on whose behalf the record of his or her gift is engraved on the temple wall. Although temple inscriptions are often voiced in the first person (as in TEXT A), it is unlikely that the text of the record was composed by the donor. So who is responsible for choosing whether to use the Tamil or Sanskrit word? Who is making the selection of words to use, in whose consciousness are the alternatives being considered? The authors of these texts are temple and village accountants — kaṇakku. Despite the fact that they were locally based, their modes of expression are remarkably consistent over time and space. Like the record-keepers in South India in later times — karaṇams and niyogi Brahmans — they were by no means unilingually Tamil. Like the latter-day kaṇakkuppiḷḷais whom Bhavani Raman has studied, they had specialized computational skills (we see a glimpse of this in TEXT B). Their expertise was not only required in drawing up documents of agreements, land transfers, and contracts but — I would argue — it was formative in the development of a particular type of language that had widespread currency in inscriptions and other types of records. Their knowledge and use of Sanskrit would have been quite different from that of the temple priests — who may indeed have been conversant with the vocabulary of the Āgamas. So we have in TEXTS D and E the perhaps ironic situation of the first-person voice being that of the Śivabrāhmaṇas, whereas the framer of their words did not share in their priestly expertise, and employed his own language.

28It is difficult to know what our medieval accountants thought about Sanskrit — or to know whether they even thought it was Sanskrit. It may well be the case that the Sanskrit words for worship and places of worship that they employed were completely naturalized in the Tamil of their times. But, consciously or not, in the hybrid expressions that they coined and the new applications of words that they made, by crafting a vocabulary specific to temple life and responsive to legal exigencies, they created a language which draws on the lexical heritage of both Tamil and Sanskrit in novel and sometimes surprising ways.



I. Primary Sources

I.1. Epigraphical Sources


ARE: Annual Reports on Indian Epigraphy

EI: Epigraphia Indica

IEP: Inscriptions of the Early Pāṇḍyas (from 300 B.C. to 984 A.D.)

IMK: Īrōṭu Māvaṭṭak Kalveṭṭukaḷ

IP: Inscriptions of the Pallavas

IPS: Inscriptions (Texts) of the Pudukkottai State

KMK: Kāñcipuram Māvaṭṭak Kalveṭṭukaḷ

PCM: Pallavar ceppēṭukaḷ muppatu/Thirty Pallava Copper-plates (prior to 1000 A.D.)

SII: South Indian Inscriptions

TAS: Travancore Archaeology Series

TKC: Tamiḻk kalveṭṭuc collakarāti (Glossary of Tamil Inscriptions)

Annual Reports on Indian Epigraphy. Delhi: Manager of Publications, 1905–1996. (Transcripts of the inscriptions abstracted in these annual reports have been graciously made available to me at the Office of the Chief Epigraphist, Archaeological Survey of India, Mysore.)

Epigraphia Indica. Calcutta-Delhi: Director General, Archaeological Survey of India, 1892 ff.

Inscriptions of the Early Pāṇḍyas (from 300 B.C. to 984 A.D.). Edited by K.G. Krishnan. New Delhi: Indian Council of Historical Research/Northern Book Centre, 2002.

Inscriptions of the Pallavas. Edited by T.V. Mahalingam. New Delhi: Indian Council of Historical Research, 1988.

Inscriptions (Texts) of the Pudukkottai State arranged according to Dynasties. Pudukkottai, 1929.

Īrōṭu Māvaṭṭak Kalveṭṭukaḷ. Edited by Cītārām Kurumūrtti. Chennai: Tamil Nadu State Department of Archaeology, 2007.

Kāñcipuram Māvaṭṭak Kalveṭṭukaḷ. Edited by Tī. Śrī. Śrītar. Chennai: Tamil Nadu State Department of Archaeology, 2006.

Pallavar ceppēṭukaḷ muppatu/Thirty Pallava Copper-plates (prior to 1000 A.D.). Edited by Ti. Nā. Cuppiramaṇiyaṉ. Chennai: Tamiḻ Varalāṟṟuk kaḻakam, 1966.

South Indian Inscriptions, vol.2–26. Delhi: Director-General, Archaeological Survey of India, 1891–1990.

Tamiḻk kalveṭṭuc collakarāti (Glossary of Tamil Inscriptions), 2 volumes. Edited by Y. Subbarayalu. Chennai: Cānti Cātaṉam, 2002–2003.

Travancore Archaeology Series, vol. 1. Edited by T.A. Gopinatha Rao. Madras: Methodist Publishing House, 1910–1913.

I.2. Texts

Ajitāgama. Edition critique par N.R. Bhatt, 3 volumes. Pondichéry: Institut Français d’Indologie, (Publications de l’Institut Français d’Indologie no 24), 1964–1991.

Mahotsavavidhi. Edited, translated and annotated by Richard H. Davis, in A Priest’s Guide for the Great Festival: Aghorasiva’s Mahotsavavidhi by Aghoraśivācārya. New York: Oxford University Press, 2009.

Nālāyira Tivyap Pirapantam with commentary by Dr. S. Jagathratchagan and English translation by Sri Rama Bharati, 3rd edition. Chennai: Āḻvārkaḷ āyvumaiyam, 2002.

The Pañcāvaraṇastava of Aghoraśivācārya. A Twelfth-Century South Indian Prescription for the Visualisation of Sadāśiva and His Retinue. An annotated critical edition by Dominic Goodall, Nibedita Rout, R. Sathyanarayanan, S.A.S. Sarma, T. Ganesan and S. Sambandhaśivācārya. Pondichéry: Institut Français de Pondichéry/École française d’Extrême-Orient, (Collection Indologie no 102), 2005.

Rauravāgama. Edition critique par N.R. Bhatt, 3 volumes. Pondichéry: Institut Français d’Indologie, (Publications de l’Institut Français d’Indologie no 18), 1961–1988.

Le Rauravāgama. Un traité de rituel et de doctrine śivaïtes, 2 volumes. Introduction, traduction et notes par B. Dagens et M.-L. Barazer-Billoret. Pondichéry: Institut Français de Pondichéry, (Publications du Département d’Indologie no 89), 2000.

Somaśambhupaddhati, 4 volumes. Texte, traduction, introduction et notes par Hélène Brunner-Lachaux. Pondichéry: Institut Français d’Indologie, (Publications de l’Institut Français d’Indologie no 25), 1963–1998.

Tēvāram: hymnes śivaïtes du pays tamoul, 3 volumes. Edition établie par T.V. Gopal Iyer. Pondichéry: Institut Français d’Indologie, (Publications de l’Institut Français d’Indologie no 68), 1984–1991.

Tēvāram. Translated by V.M. Subramanya Aiyar. In Jean-Luc Chevillard and S.A.S. Sarma (Editors), Digital Tēvāram/Kaṇiṉit Tēvāram [CD-ROM]. Pondichéry: Institut Français de Pondichéry/École française d’Extrême-Orient, (Collection Indologie no 103), 2007.

II. Secondary Sources

Ali, Daud (2011). “The Early Inscriptions of Indonesia and the Problem of the Sanskrit Cosmopolis”. In P.-Y. Manguin, A. Mani and G. Wade (Editors), Early Interactions between South and Southeast Asia: Reflections on Cross-Cultural Exchange, pp. 277–297. Singapore: ISEAS — New Delhi: Manohar (Nalanda-Sriwijaya Series no 2).

Barazer-Billoret, Marie-Luce (1999). La grande fête du temple (mahotsava) d’après les āgama śivaïtes (Unpublished doctoral dissertation). Université Paris III — Sorbonne Nouvelle, Paris, France.

Brocquet, Sylvain (1997). Les inscriptions sanskrites des Pallava: poésie, rituel, idéologie. (Unpublished doctoral dissertation). Université Paris II — Sorbonne Nouvelle, Paris, France.

Brunner, Hélène (1990). “L’image divine dans le culte agamique de Śiva: Rapport entre l’image mentale et le support concret du culte”. In A. Padoux (Editor), L’Image Divine: Culte et médiation dans l’hindouisme. Paris: CNRS.

Chevillard, Jean-Luc (2000). “Le Tēvāram au XXe siècle”. Bulletin de l’École française d’Extrême-Orient, 87, pp. 729–740.

Colas, Gérard (1986). Le temple selon Marīci. Pondichéry: Institut Français d’Indologie (Publications de l’Institut Français d’Indologie no 71).

Colas, Gérard (1996). Viṣṇu, ses images et ses feux: Les métamorphoses du dieu chez les vaikhānasa. Paris: Presses de l’École française d’Extrême-Orient.

Cox, Whitney (2010). “Scribe and Script in the Cālukya West Deccan”. Indian Economic and Social History Review, 47, pp. 1–28.

Dagens, Bruno (1984). Architecture in the Ajitāgama and the Rauravāgama. New Delhi: Sitaram Bhartia Institute.

Davis, Richard H. and Leslie C. Orr (2007). “People of the Festival”. In D. Goodall and A. Padoux (Editors), Mélanges tantriques à la mémoire d’Hélène Brunner/Tantric Studies in Memory of Hélène Brunner, pp. 73–97. Pondichéry: Institut Français de Pondichéry/École française d’Extrême-Orient (Collection Indologie no 106).

Francis, E. (2009). Le discours royal: Inscriptions et monuments pallava (IVème — IXème siècles, (Unpublished doctoral dissertation). Université Catholique de Louvain, Louvain, Belgium.

Goodall, D. (2006) Initiation et délivrance selon le Śaiva Siddhānta. In G. Colas and G. Tarabout (Editors), Rites hindous: transferts et transformations, pp. 93–116. Paris: Editions de l’EHESS.

Granoff, Phyllis (2005). “Images and their Ritual Use in Medieval India: Hesitations and Contradictions”. In P. Granoff and K. Shinohara (Editors), Images in Asian Religions: Texts and Contexts. Vancouver: University of British Columbia.

Krishnan, K.G. (1975). “Architectural Terms in South Indian Temple Inscriptions”. In P. Chandra (Editor), Studies in Indian Temple Architecture. New Delhi: American Institute of Indian Studies.

Krishnan, K.G. (1990). “Characteristic Features of Tamil Inscriptions”. In H.M. Nayak and B.R. Gopal (Editors), South India Studies: Dr. T.V. Mahalingam Commemoration Volume. Mysore: Geetha Book House.

Lefèvre, Vincent (2006). Commanditaires et artistes en Inde du Sud. Paris: Presses Sorbonne Nouvelle.

Lefèvre, Vincent (2010). “‘Portrait’ or ‘image’? Some terminological and literary perspectives on portraiture in ancient and mediaeval India”. Paper presented at the conference on “Portraiture in South Asia”, London, May 2010.

Orr, Leslie C. (2000). Donors, Devotees and Daughters of God: Temple Women in Medieval Tamilnadu. New York: Oxford University Press.

Orr, Leslie C. (2004). “Processions in the medieval South Indian temple: Sociology, sovereignty and soteriology”. In J.-L. Chevillard and E. Wilden (Editors), South Indian Horizons: Felicitation Volume for François Gros on the Occasion of his 70th Birthday, pp. 437–470. Pondichéry: Institut Français de Pondichéry/École française d’Extrême-Orient (Publications du Département d’Indologie no 94).

Orr, Leslie C. (2009). “Tamil and Sanskrit in the Medieval Epigraphical Context”. In Kannan M. and J. Clare (Editors), Passages: Relationships between Tamil and Sanskrit, pp. 97–114. Pondichéry: Institut Français de Pondichéry — Tamil Chair, Department of South and Southeast Asian Studies, University of California at Berkeley (IFP — Publications Hors Série no 11).

Pollock, Sheldon (2006). The Language of the Gods in the World of Men: Sanskrit, Culture, and Power in Premodern India. Berkeley: University of California.

Raman, Bhavani (2007). “Tamil Munshis and Kacceri Tamil under the Company’s Document Raj in Early-Nineteenth-Century Madras”. In T. Trautmann (Editor), The Madras School of Orientalism: Producing Knowledge in Colonial South India, pp. 209–232. New Delhi: Oxford University Press.

Sethuraman, N. (1994). “The Later Pandyas (1371–1750)”. Journal of the Epigraphical Society of India, 20, pp. 96–116.

Slaczka, Anna (2007). Temple Consecration Rituals in Ancient India: Text and Archaeology. Leiden: Brill.

Smith, H. Daniel (1975). A Descriptive Bibliography of the Printed Texts of the Pāñcarātrāgama. Baroda: Oriental Institute.

Subramaniam, T.N. (1958). “Paḷḷankōvil Jaina Copper-Plate Grant of Early Pallava Period”. Archaeological Society of South India: Transactions for the year 1958–59, pp. 41–83. Madras: Archaeological Society of South India, Madras Museum.

Takashima, Jun (2005). “Pratiṣṭhā in the Śaiva Āgamas”. In S. Einoo and J. Takashima (Editors), From Material to Deity: Indian Rituals of Consecration, pp. 115–142. New Delhi: Manohar.

Willis, Michael (2009). The Archaeology of Hindu Ritual: Temples and the Establishment of the Gods. Cambridge: Cambridge University Press.



In this appendix, italics are used to indicate characters appearing in the text of the inscription in Grantha script. The numbers in curved brackets provide the number of the line of the inscription as engraved in situ. Three dots indicate an ellipsis made by me in reproducing the text or translation here; more (or fewer) than three dots indicate a break or missing section in the text of the inscription itself. Square brackets in the Tamil text of the inscription indicate that the character within is a reading provided by the epigraphist in the case of a hard-to-decipher character, or, if there is an asterisk, the character is not in the original and has been supplied by the epigraphist. English words within square brackets are provided by me, as summaries of content in the text of the inscription or, in the translations, as proposed additions that render the inscription more intelligible.


SII 4.827 = ARE 1893, no. 14 = IP 80; cf. Francis 2009, pp. 470–72 / W, S and E walls of the Muktīśvara temple, Kanchipuram / 28th regnal year of Nandivarman (II?) — i.e. 758–759 CE (?)

{1} [svasti śrī] na[nti]va……….[ru]patteṭṭāvatu dha[rmmama] hādevīśvaragra[*ha]ttu māṇikkadevarkku devatāṉamāyiṉa nāyivaṭṭikuḷattūr ka[ṇ]ṇantaikuṭi[p] pāṭṭuḷpa[ṭa] nilam pattup paṭṭiyu māṇikkatevarkk uṇṇāḻikaippuṟamāka paṇittukkuṭutteṉ dharmamahādeviyeṉ itu arccaṉayi ceytuṇpārāṉār innilattāl ūr varintatellam

{2} [tatta]śivaṉum dharmmakaṇṇaṉukkum aṉantaśivaṉukkum iva[*r] kaḷ makkaḷ makkaḷe arccaṉai ceyituṇṇap paṇittuk kuṭutteṉ ittaḷik[ku] t taṭṭaḻi[vu]yivaraik koṇṭu koṭṭuvāṉ ā[mu]tarayaṉ nāyaṟṟukkaṇi i[*va]ṉukkum māṭalar makaṉ mātaśarmmanukkum iruvarkkumm utakapūrvvakam ceytu iraṉtu koṟṟu kuṭuttom

{3} a string of female names

{4} mantrācāryyar aivar viḷakkuntavacikaṇavar paṉṉiruvar tavacikaḷ uṇṇakkuṭukka taḷiyāḷvā[r ū]ṭṭārāka paṉṉiruvar kuttika[ḷai] meytu kuṭutteṉ mutalmuppatiruvarum ippaṉṉiru[var e]ṟṟi nāṟppattunālvar kampa[ṉa]ttu kuṭutteṉ tarumamādeviyeṉ ittaḷipuṟamāṉa nāyvaṭṭi kuḷattūrum ittaḷi parivāramum taḷiyum idharmam aḻiyāmai kāttāṉ aṭi eṉ muṭi mel atu

In the twenty-eighth regnal year of Nantiva[rmapallava]

I, Dharmamahādevi, gave 10 paṭṭi of land in Kaṇṇantaikuṭi [part of] Nāyivaṭṭikuḷattūr as a temple endowment (uṇṇāḻikaippuṟam) for Māṇikkadevar, this land being the devadāna for Māṇikkadevar of Dharmamahādevīśvaragṛham.

I gave this land of those possessing the hereditary right to perform arccaṉai, whose boundaries are within the village, to Tattaśivaṉ, Dharmakaṇṇaṉ, and Aṉantaśivaṉ, so that arccaṉai might be performed by them [and] their descendants.

To Mātaśarman the son of Māṭalar, and to the drummer (koṭṭuvāṉ) Āmutarayaṉ Nāyaṟṟukkaṇi who receives a living allowance (uyivarai = uyirvaḻkai? cf. cīvitam) to play the taṭṭaḻi (drum) for this temple (ittaḷi), to these two we gave with the pouring of water two [grants for] subsistence (koṟṟu).

I, Dharmamahādevi, gave [the means of support for a group called] Nāṟpattunālvar Kampaṉam (“The forty-four of Lord Ekampaṉ”?) [consisting of] the first thirty-two plus these twelve — the twelve brahmacārins (kuttikaḷ) I appointed (? meytu kuṭutteṉ) to allow to be fed (i.e. to cook for) the deity (taḷiyāḷvar), and those ascetics who are given food (the first thirty-two) [including] twelve ascetics of great purity, five mantrācāryas, and fifteen women (whose names are given).

Whoever protects Nāyivaṭṭikuḷattūr Kaṇṇantaikuṭi as a temple site (ittaḷippuṟam), this temple establishment (of personnel — ittaḷipparivāram), and this temple (itself) and maintains this charity (dharmam) — I place his feet on my head.

[Given that the temple bears the same name as the donor — probably a Pallava queen — it seems likely that she is also the founder of the temple.]


SII 13.170 = ARE 1926, no. 103 / E and N walls of the Uktavedeśvara temple, Kuttalam, Thanjavur district / 7th regnal year of Rājarāja (I) — i.e. 992 CE / Fourteen disconnected fragments — below are extracts from pieces #8 and #9 (which appear to be continuous) and from #10 and #11 (see also Orr 2000: 89–91)

#8 {7}… ā[*ḻ]vārait tiruvārātiṉai ceyyum brāhmaṇan oruvaṉukkuk kappaṭamuṭpaṭa nicati nellup patakku nāṉāḻiyāka orāṭṭaikku nellu eḻupattaiṅ{11}kalattukku nilaṉ oṉpati mākkāṇiyaraik kāṇiyum tirukkoyiluṭaiyār{13}kaḷ muṉvelinilaṅ koṇṭu ceytuvarummāṇ muṉṟiṉotun tiruvamuta[*ṭ] ṭavun tirumañcaṉanirum taṇṇiramutu cumantaṭṭavum tiruvāratiṉai ceyyu nampikku paricārakañceyyavum{17}tirumeypūccut tekkavum vitāṉam piṭikkavum uṭpaṭa veṇṭumāṉa iraṇṭiṉukku nicata nelluk kuruṇi nāṉāḻiyāka orā{20}ṭṭaināḷaikku nellu nāṟpattain kalattiṟ nilam kāley araimāvaraikkāṇi tiruppali{22}ttāmam paṟittut toṭuppār muvarkku nicata nellu kuruṇi nāṉāḻiyāka orāṭṭaikku nellu nāṟpattaiṅkalattuk{25}ku nilañ kāḷe araimāvaraik kāṇi tirumu[*ṟ] ṟan tiruvalakiṭṭut tirumeḻukkir muvarkku nicata nelluk kuruṇi nāṉāḻiyā{28}ka orāṭṭaikku nellu nāṟpattaiṅ kalattukku nilaṉ kāley araimāvaraikkāṇi{30}ukaccut talaippaṟai yoṉṟu mattaḷa muṉṟu kaṟaṭikai oṉṟum caṅkira{32}ṇṭum kāḷam iraṇṭum cekaṇṭikai yoṉṟum tāḷam oraṇaiyum kaimaṇi oraṉaiyumāka āka pa{34}nṉiraṇṭum tirumañcaṉamum tiru- #9{1}vamutum bhūtabaliyum śrībaliyum arddhayāmamum paḷḷiyeḻucciyum{5}koṭṭuvatāka… [amounts of paddy for drummers]

For one brahman to do worship (tiruvārātiṉai) of the deity [provision], including for cloth, of 1 patakku & 4 nāḻis of paddy daily, which is 75 kalams of paddy yearly, thus 9 mā & ½ kāṇi of land. For what is required — including provision for those who take charge of the land in the presence of the tirukkoyiluṭaiyārkaḷ (temple trustees?), together with [provision for] the three māṇis (brahmacārins), and for making food offerings (tiruvamutu), water for the sacred bath (tirumañcanam), abundant offerings of water (taṇṇiramutu), for the performance of auxiliary worship services, the anointing of the deity (tirumeypūccu), and the bearing of the canopy (vitāṉam) — 1 kuruṇi & 4 nāḻi of paddy twice daily… [amount yearly, and amount of land needed to produce]. For three garland-makers 1 kuruṇi & 4 nāḻi of paddy daily… [similarly]. For three people to sweep and smear the temple courtyard 1 kuruṇi & 4 nāḻi of paddy daily… [similarly]. For twelve drummers (ukaccu) — one to beat the chief drum, three the mattaḷam drum, one the kaṟaṭikai (rattle?), two to play the conch, two the trumpet, one the gong, two the cymbals, and two the hand-bells — at the time of the sacred bath, and at the times of the food-offerings, bhūtabali, śrībali, the midnight service, and the early morning service (paḷḷiyeḻucci) …

#10 {2–6}…āḻvār tirttamāṭiyaruḷum vaikāci vicākattiṟkut tiruviḻāviṟkkum veṇṭu palavicattiṟku … [amounts of paddy]

[provision] for the various expenses for the bathing of the deity and for the requirements for the festival on Vicākam in Vaikāci

#10 {21} nibandhañ ceykaveṉṟu ittirukkaṟṟaḷi eṭupittu ide{23}vark….. tta śrī uttamacoḻarai tiruvayiṟuvāytta{25}uṭaiya pirāṭṭiyar cempiyaṉ mahādeviyār irājarājakesaripanmakku yāṇṭu 7 vatil…

Queen Cempiyaṉ Mahādeviyār, mother of Śrī Uttamacoḻa, having built this stone temple and made the foregoing endowment, [gave this agreement to] this god, in the 7th year of Rajaraja…

#11 {17} śrīvimāṉamum maṇṭapamum uḷḷiṭṭaṉa putukka…

[provision] for renovation of the vimāna, the maṇḍapa, etc.


ARE 1918, no. 192 / W and S walls of the Pāṭalīśvara temple, Brahmadesam, South Arcot district / 14th regnal year of Rājarāja (II) — i.e. 1160 CE (I am grateful to the Epigraphy Office of the Archaeological Survey of India for graciously granting me permission to consult the transcript of this unpublished inscription).

svasti srī pumaruviya tirumātum… (meykkīrtti) parakesarivarmaṉ tribhuvanacakkiravattikaḷ rājarājadevarku yāṇṭu 14 rājarājavaḷanāṭṭup paḷairaiyūrnāṭṭu brahmadeya teriyūr śrī rājarājacaturve dimaṅkalattup peruṅkuripperumakkaḷom ivvāṭṭai makara nāyiṟṟu pūrva pakṣattu titiyaiyum putan kiḻamaiyum peṟṟa pūraṭṭāti nāḷ namakku muladeya mahādevar koyilil devakanmikaḷum srīmāheśvarakkaṇkāṇi ceyvārkaḷum kaṇṭu ikkoyilil.. ruṅkoḷar eḻuntaruḷuvitta tiruppaḷḷiyaṟai āḷuṭaiya nāciyā… tirupratiṣṭhai paḷḷit tirukkalyāṇam paḷḷukiṟa aṉṟu āḷuṭaiya nācciyaṟku mahāsabhaiyār strīdhanamāka nirpārttuk kuṭutta nilam paḷḷiraṇmovum nam māmpiṭākai kiḻar ceñciyil aṅkatevirttaruḷiṉa śrī kulottuṅkacoḻadevarku yāṇṭu 21…

In the 14th year of Rajaraja, etc…, we the village assembly (peruṅkuripperumakkaḷ) of the brahmadeya Teriyūr Śrīrājarājacaturvedimaṅkalam, in Paḷairaiyūrnāṭu within Rājarajavaḷanāṭu, in this year on this day [precise details given] [issued a document concerning] the deity’s property which was received by the devakanmikaḷ and the srīmāheśvarakkaṇkāṇi ceyvārkaḷ of the temple of Mahādevar:

In the 21st year of Kulottunga, the brahman assembly (mahāsabhaiyār) gave land (? in a hamlet of eastern Ceñci) as strīdhana to [the goddess] Āḷuṭaiya nācciyār to celebrate at that time the consecration (tirupratiṣṭhai) and marriage (tirukkalyāṇam) of this “bedchamber goddess” Āḷuṭaiya nācciyār which had been set up (lit, “caused to graciously appear”) in the temple by Iruṅkoḷar.

The inscription continues with details of the land granted.


SII 26.9 = ARE 1909, no. 9 = IPS 137 / N wall of the central shrine in Choleśvara temple, Ponnamaravati, Trichy district / 19th regnal year of Rājarāja (II) — i.e. 1164–65 CE / incomplete

{1} [date] rājarājappāṇṭināṭṭu rājendracoḻavaḷanāṭṭu puṟamalaināṭṭu poṉṉamarāpati rājentracoḷīsvaramuṭ[*ai] ya mahadevarai tirupratiṣṭhaiyum paṇṇi śrī vimānamum eḻunt’aruḷivitta vīmaṉ rājentracoḻaṉā{2}ṉa niṣadharājar muṉpu civappirāmaṇarkku nīrvārttu tānam paṇṇikkuṭutta nāḷkaḷ piṉpu nāṅkaḷ enkaḷil saṃharicca paricāvatu…

Vīmaṉ Rājendracoḻaṉ a.k.a. Niṣadharājar built (lit, “caused to graciously appear”) the temple and arranged for the consecration (tirupratiṣṭhai) of Mahādevar, the Lord of Rājendracoḷīśvaram in Poṉṉamarāpati in [places named]. Formerly, days [with worship rights] had been bestowed as gifts to Śivabrahmanas, with the pouring of water; we now subsequently gather together [to distribute these days] as follows [the inscription continues with the assignment of days to a number of named bhaṭṭar]


SII 5.762 = ARE 1895, no. 198 / S wall of mandapa in front of central shrine of the Kāśi Viśvanātha temple, Tenkasi, Tirunelveli district / Jaṭilavarman Kulaśekhara’s regnal year 2+42 — i.e. 1474 CE, according to Sethuraman’s (1994) dating of these later Pandya kings.

{1}kojaṭilavarmmarāna tribhuvanaccakravattikaḷ śrī kulaśekharadevaṟku yāṇṭu iraṇṭāvatiṉ etir nāṟpattu iraṇṭāvatu … [details of day] teṉ{2} arināṭṭu citranadi uttaratirattu dakṣiṇakāśiyil muṉnāḷ aṇṇāḻvi parākrama pāṇḍyadevar tiruvuḷḷampaṟṟi uṭaiyār viśvanāthaṉaiyum pratiṣṭhippittu tirukkoyil upānādi stūpi pariyantam garbhagṛham arddhamaṇḍapam iṭaināḻikai{3}mahāmaṇḍapam sopānam ivaiyum kuṟaivaṟat tiruppaṇiyuñ ceyvittu āvaraṇa gopudikaḷum ceyvittu pūjaikku nityanaimittikaṅkaḷukku veṇṭum tevatāṉamum viṭṭu utsavavaryantamāyuḷḷa karmmaṅkaḷum naṭatti aṉai{4}ttuk kottil veṇṭuvāraiyuṅ kaṟpittu naṭantu potukac ceyte nāmum iṉṉayaṉāṟkut tiruvolakkamaṇḍapamāka kulacekaraṉ maṇṭapamun tiruppaṇi ceyvittu muṉpu kaṟpitta per niṅkalāka eṭṭām paṅku devakarmmamākak kaṟpi{5}tta… [list of bhaṭṭar with gotras, and their shares]

In the 2nd opposite the 42nd year of Jaṭilavarman Kulaśekhara (on a specific day),

In accordance with the intention of [the king’s] elder brother Parākrama Paṇḍyadeva -- who had earlier (munnāḷ) caused Lord Viśvanāthaṉ to be consecrated (pratiṣṭhippittu) at Dakṣiṇakaśi on the northern bank of the Citranadi in Teṉarināṭu, and had made [for this god] a temple (tirukkoyil), [constructing] the garbhagṛha from upāna to stūpi and the ardhamaṇḍapa as a central shrine (iṭaināḻikai), and the mahāmaṇḍapa and the steps, and had work done to remove all defects, and made the surrounding wall, gopura, etc., [this Parākrama Paṇḍyadeva] who had released [land as] devadānam to provide for pūjai and for daily and occasional observances --

we all (the Śivabrāhmaṇas), for the ongoing proceeding [of these observances] have commanded those among the personnel who are required such that all services including festivals (read utsavaparyanta, “up to festivals”) be carried out; we have arranged for worship in the Kulacekaraṉ maṇḍapa which serves as an audience hall (tiruvolakkamaṇḍapa) for this god; and we have ordained for [serving the] purposes of the god (devakarmam) [the distribution of shares among the Śivabrāhmaṇas, including] the previously mandated eighth share which had not been assigned (pēr nīṅkalāka, “being excluded from the people”) [as follows]…


Unpublished, viewed on site. On a marble plaque mounted on the south wall of the mahamandapa of the Kṛpapurīśvara temple in Tiruvennainallur, South Arcot district.

ka. a. cīṉivāca ceṭṭiyar

manaivimār vaḷḷiyammai



po. cellammāḷiṉ

elakṭirikṭlaiṭ tarmam



The endowment (tarmam) for electric lights is by Po. Cellammāḷ, the daughter of Vaḷḷiyammai Viruttāmpāḷ and the wife of Ka. A. Cīṉivāca Ceṭṭiyar


1 See Davis and Orr (2007) for a comparative study of how participants in temple festivals are represented in the Mahotsavavidhi and the Tamil inscriptions. The nature of the relationship between Āgamic texts and temple practices, and the question of how over the course of time the Āgamas came to be treatises on temple ritual, have begun to be explored (e. g. in Brunner 1990, Colas 1996, Goodall 2006, Goodall et al. 2006, Granoff2005), but mysteries abound and debates remain unresolved. In terms of one aspect of this issue — the matter of temple architecture — we may contrast the view of Gérard Colas with that of Bruno Dagens. Colas maintains that the Vaikhānasa text Vimānārcanākalpa shows signs of few if any interpolations, and that it was a text actually used in the design of the 8th-century (?) Sundaravarada Perumāḷ temple in Uttaramerur (Colas 1986: 5–7, 57–59). Dagens, examining the sections on temple architecture of the Śaiva Ajitāgama and Rauravāgama, concludes that “it is almost certain that these treatises, which have never ceased to be read and employed, have several times been adjusted according to the prevailing fashion. In their present form they are the result of a long evolutionary process” (Dagens 1984: 12). Takashima expresses a similar view: “I think that Āgamic instruction concerning the structure of temples and organization of temple rituals do not preceed [sic] actual temples but follow the real development. So, the Ajitāgama, for example, must be later than the eleventh-century expansion of big Saiva temples” (Takashima 2005: 115). The view of Colas is in part informed by the fact that there is an inscription at the Uttaramerur temple, in Sanskrit verse, that praises the temple as one built by experts knowledgeable in vāstu and in accord with āgama (āgamavidhaiḥ) (SII 6.333). Elsewhere we find one or two similar inscriptions: an undated Tamil inscription from Idakal in Tirunelveli district in which the various constructions — including gopura, mahāmaṇḍapa, and enclosing wall — are said to have been made in accord with the ordinances of āgama (ākamattiṉ paṭi vāḻa) (ARE 1916, no. 501); and another Tamil inscription from the far south, engraved at Mannarkoyil and perhaps dating from the early 13th century, which claims that the second prākāra was built according to the śāstras (śāstrārtthattiṉ paṭi) (ARE 1916, no. 408). In my view, it is quite problematic to connect these (very few) inscriptional references to āgama, or śāstra — which may denote simply “tradition” or “authoritative knowledge” in a general way (see Chevillard 2000) — with the specific texts that we today call the Āgamas.

2 My sense of the Āgamic vocabulary is derived from an examination of the Ajitāgama, Mahotsavavidhi, Rauravāgama, and Somaśambhupaddhati, in addition to references and citations gleaned from the works cited in the preceding footnote, as also Barazer-Billoret (1999), Colas (1996), and Smith (1975); help in the review of these sources has been provided by research assistants Steven Engler, Nathalie Kalina, and Gisele Pritchard. In my survey of the inscriptional usages, I have relied on TKC, Krishnan (1975), and Lefèvre (2006), but even more on my own collection of references — a collection made possible by the labor of my research assistants Michelle Bakker, Ashleigh Delaye, Michael Gollner, Kirsten Hansen, Sacha Mathew, Tanisha Ramachandran, Astrid Schau-Larsen, and Shaun Turriff.

3 According to K.G. Krishnan (1990), linguistic mixing has been a feature of epigraphical practice in Tamilnadu from the beginning, with the Tamil Brahmi inscriptions of two thousand years ago employing Prakrit words within Tamil inscriptions, but he suggests that in the second half of the first millennium, the incorporation of Sanskrit into inscriptional Tamil came to be more pronounced. On this topic, also see Francis’ contribution to this volume.

4 Note that the word devakarman also appears eight centuries later in TEXT E.

5 The word havis / avi is used to designate food offerings in the context of Jain worship in the 9th century, for example in IEP 32 = SII 14.22 and IEP 54c = SII 14.118.

6 Uthaya Veluppillai demonstrates in her forthcoming doctoral thesis that the phrase amutu cey, which first appears in Tēvāram, is associated in that context with Śiva’s act of consuming poison at the time of the churning of the ocean — transforming the poison into amṛta. In the inscriptions, precisely the same expression refers to the consumption of food offerings by the deity of the temple. In the inscriptional context the god’s partaking of the food causes it to become amṛta / amutu. The meaning of amutu is not, however, restricted to substances already offered (i.e. prasādam — a word rarely found in the medieval Tamil inscriptions), and the term refers also to materials intended for the deity’s consumption.

7 In the “Tamilization” of Sanskrit words, a final long a becomes ai. Thus pūjā becomes pūcai.

8 In keeping with the orthography of the Tamil inscriptions, where long e and o are not distinguished from the short vowels, I do not mark these vowels in discussions of inscriptional usages; hence, koyil instead of kōyil.

9 In 9th-century inscriptions, we also find the form viṣṇugraham — e.g. in SII 12.47 and SII 7.420.

10 The term garbhagṛha is found more among the later inscriptions of the Tamil corpus. Tamil transliterations of the Sanskrit word — e.g. kerpakkiṟaka — appear in 17th-and 18th-century inscriptions from Erode district, for example at Sivagiri and Bhavani (IMK 1).

11 It may be that the relevance of the word nāḍikā to the temple sanctum arises from its denotation in Sanskrit of a vessel or container — perhaps something like a garbha? Or the meaning of nāḍikā as a bodily channel, connected with the circulation of prāṇa, may point toward the use of the term in the temple context as a designation for the divine life source enshrined within. Many thanks to Shaman Hatley and to Dominic Goodall for taking the time to puzzle over this word with me. Michael Willis (2009) tells us that the reason nāḍikā means both a tube and a measure of time is because of the use of hollow sections of bamboo in the construction of water clocks.

12 TKC says that tānam is used as a synonym for koyil — temple — beginning in the 10th or 11th century.

13 Among the early references māḷikai is found in an 8th-century inscription (SII 2.73) and tirucuṟṟālai in the 9th century (SII 3.91).

14 In a 12th-century inscription from Tirunelveli district in the far south of Tamilnadu, we find tirumāḷikai apparently referring to a maṇḍapa (ARE 1916, no. 482). Also very commonly met with is the Tamil word tirumaṭil, as a more straightforward term for wall, found in inscriptions beginning in the 8th century (e. g. SII 14.19). In TEXT E we have the unusual use of āvaraṇa to designate the temple’s enclosing wall; this usage is similar to what is seen in Pāñcarātra texts and quite unlike the meaning in Śaivāgamas, where āvaraṇa refers to the circle of attendant deities — visualized or present in image form as parivāra devatās (see Goodall et al. 2006).

15 For example, in SII 3.1092, it is said that the piṟākāram was rebuilt. Another inscription of the 13th century (IPS 255) suggests that piṟākāram might refer to the actual enclosing wall, and that tirumāḷikai designated the space (and also possibly built structures such as galleries) enclosed by the wall; it records that the gateway (tiruvācal) of the second pirkāram was demolished, the tirumāḷikai enlarged, and the gateway widened.

16 Kōpuram appears fairly regularly in Tēvāram, but in that context refers to a tower more generally rather than to a temple entrance (Chevillard 2000).

17 We also find olakka kūṭam, with the same meaning, for example in two inscriptions of the 11th century (SII 8.68; ARE 1908, no. 452). The Tamil word ōlakkam is perhaps derived from ōlai, palm leaf, the medium for documents of various kinds including orders and proclamations — which might have been read out in the ōlakkamaṇḍapa of a palace or temple. The tiruvolakkamaṇḍapa that we see in the inscriptions appears to be the exact equivalent of the āsthāna maṇḍapa referred to in the Āgamas — a term that appears in 18th-century Tamil inscriptions as well (ARE 1970–71, no. 533; SII 6.430).

18 KMK, p. 27, which may date from the middle of the 12th century, contains the phrase upānāti stūpipariyantam. The phrase is found in inscriptions recording the construction or renovation of temples by Pandya kings at Kuttalam, Ilanji, and Tenkasi; in other places in the southernmost region of Tamilnadu in the period of the 14th to 16th centuries, non-royal figures are named as sponsors of similar projects, with similar phrasing in the inscriptions.

19 An inscription of 1447 says that the ploughing ceremony was performed by the king about a year before the construction of the central shrine of the Kāśi Viśvanātha temple at Tenkasi (SII 26.537). About a hundred years later, in the mid-16th century, another Pandya king is said to have established the Kulaśekhara temple at Tenkasi “beginning with the ploughing ceremony and ending with the consecration” (karuṣaṇayāti pratiṣṭhāntam, SII 5.765). In the first decade of the 16th century, we have an inscription from Melacceval, near Tirunelveli, which records the performance of the karṣaṇa ceremony by a non-royal figure in the case of the foundation of a goddess shrine (ARE 1916, no. 602). The earliest reference to this ritual that I have found is in an inscription of the late 12th or early 13th century engraved on a temple in the western part of Tamilnadu, in Coimbatore district (SII 26.145). This records that a local king, of the so-called Kongu Chola dynasty, established a Śivaliṅga — evidently as the deity in the central shrine — and set up images of the god and goddess, carrying out this foundation “beginning with the ploughing ceremony and ending with the consecration.” Here we see the formulaic Sanskrit expression inserted in the midst of a Tamil inscription; it may also be significant that in this context there is no mention of the building of the temple itself — only the establishment of the gods — and that, outside of the formulaic phrase, this “establishment” is described using precisely the same mixed Tamil-Sanskrit vocabulary that is found in TEXT D, and that is discussed below, in which pratiṣṭhā does not seem to have a technical ritual meaning.

20 Although Ajitāgama has an entire chapter devoted to karṣaṇavidhi, the topic receives little attention elsewhere in the Śaiva literature or in Pāñcarātra or śilpaśāstra texts (see Dagens 1984, Smith 1975, Slaczka 2007).

21 In this connection, it is noteworthy that the term for temple consecration now in very common use in Tamilnadu — kumbhābhiṣekam — is not found in inscriptions before the 19th century, unless ARE 1911, no. 382 can be dated as early as the 16th century.

22 This seems to be the term used in the earliest Tamil inscriptions that record temple foundations, for example in the 9th-century SII 7.420 and SII 12.47, and in the 10th-century inscriptions recording the establishment of the Śiva temples at Kuttalam and at Vriddhachalam by the Chola queen Cempiyaṉ Mahādevi (TEXT B and SII 19.302).

23 This is an expression found in inscriptions of the 12th and especially 13th centuries (e.g. IPS 124; SII 12.257; SII 26.478).

24 Willis’ (2009) argument — based on his reading of Sanskrit Gupta-period inscriptions — that an image is referred to as pratimā before it undergoes pratiṣṭhā, and through this ritual is transformed into something called a mūrti, does not accord with the use of these terms in the inscriptions of medieval Tamilnadu, including those in Sanskrit, where the meanings of such terms appear to be considerably less precise. Lefèvre (2010) also suggests that in various inscriptional corpora and Sanskrit literature more generally there is considerable variation in the application of these terms (including their use for representations of humans), and overlap among them.

25 Tirumeṉi is also used in the inscriptions to refer to the person of the king, when, for example, a gift is made for the sake of the king’s health.

26 In the bilingual stone inscription of Naralokavīraṉ at Chidambaram (SII 4.225), which I have already briefly discussed, there are three references to mūrtis in the Sanskrit portion of the record, and no word for “image” at all in the Tamil portion.

27 Interestingly, both Sanskrit vitāna and Tamil vitāṉam mean not only “canopy” but can also denote a sacrificial oblation.

28 In the Tamil inscriptions some of these more standard Sanskrit terms for bathing are actually found earlier than tirumañcaṉam — which makes an appearance in the early 10th century — and continue to be employed apparently interchangeably with tirumañcaṉam after this word gains currency. Abhiṣekam appears as early as 796 (SII 4.135) and snapanam a century or so later (SII 7.421); the Sanskrit-derived tīrttam also appears in inscriptions of the 10th and 11th centuries, as we see in fragment #10 of TEXT B.

29 To nantal, meaning “perishing,” is added the negative suffix –ā producing the word nantā, meaning “not-perishing” or “perpetual.”

30 The terms for the four cantis found in this inscription (SII 13.170) are a wonderful mixture of “pure” and Sanskrit-derived Tamil: ciṟukālai (“early dawn”) combines the Tamil adjective ciṟu (little) with Sanskrit kāla (time, morning); uccampotai (“noon”) uses a Tamil word for time (pōtu) with a Sanskrit word (ucca, elevated) as the first element in the compound; similarly antiyampotai (“evening”), with Sanskrit antya (end); and arddhayāmam (“midnight”) is borrowed directly from Sanskrit.

31 The term bali appears in the Sanskrit portion of the 7th-century Kuram plates (SII 1.151 = IP 46), discussed above. Michael Willis (2009) has recently argued that the use of bali — together with caru and sattra — as terms appearing in Gupta period inscriptions in the context of temple worship (especially in the 6th century), is indicative of the derivation of pūjā from domestic Vedic ritual, and rituals of hospitality in particular. Whether or not this is the case, it is clear that in early medieval Tamilnadu bali offerings are treated as essential elements of temple worship. That both bhūtabali and śrībali are mentioned in TEXT B suggests that there is a distinction by the 10th century between the former ritual, which is the bali observance that might be regarded as being tied to the domestic Vedic realm, and the śrībali, which seems rather to belong to the temple milieu. The affinity between inscriptional and Āgamic descriptions of the śrībali ceremony — involving processions and special images (see Orr 2004) — suggests to me that the Āgamas were in this case being influenced by actual temple practice. Although bali offerings are not referred to in inscriptions outside the “Hindu” context, we do find caru as the term used for offerings to a Jain Tirthankara in 9th-century Kalugumalai (SII 5.308). We have already seen how the “Vedic” havis/avi finds its way into use in both the Jain and Hindu temple contexts by the 9th century.

32 There are two examples of the use of ubhayam in the 9th and 12th centuries, but most references to this term are quite a bit later. It is not clear how Sanskrit ubhaya, meaning “both,” came to have a completely different sense in inscriptional Tamil.


Associate Professor in the Department of Religion at Concordia University in Montreal, where she teaches in the areas of South Asian religions and women and religion. Her book Donors, Devotees and Daughters of God: Temple Women in Medieval Tamilnadu (NY: Oxford University Press) appeared in 2000, and she is also the author of a large number of journal articles and book chapters. Her research is concentrated in two areas: (1) the roles and activities of women in Hinduism, Buddhism, and Jainism; and (2) the organization of religious life in the history of South India up to the late medieval period, especially with respect to interactions among religions. Her current research focuses on temple building and temple renovation in Tamilnadu.

© Institut Français de Pondichéry, 2013

Conditions d’utilisation :


Volume papier
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search