1Socio-linguistic practices documented in inscriptions from South and South-East Asia between the 4^th and 16^th centuries display a type of “functional diglossia” characteristic of legal discourse in states influenced by the transregional Dharmaśāstra tradition in Sanskrit.1 This diglossia can take two forms. Sanskrit itself may be used as an acrolect, either alone or in bilingual records, where it has primarily expressive or ceremonial functions. But the focus in this article is on the other form: the use of a highly Sanskritized, formal and formulaic register of the local vernaculars. Starting with some observations about the impact of Sanskrit legal discourse on Old Khmer and Old Javanese legal records, comparable inscriptions in Medieval Tamil are examined in detail, noting the influence of first Prakrit and then Sanskrit on legal idiom. It is concluded that the use of this Sanskritized register reflects not simply the prestige of Brahmanical high culture but also the perceived value and utility of an imported specialized conceptual system of law and administration. This study further suggests that the introduction of written legal documentation, simultaneous with the spread of Brahmanical legal ideas, could also provide a mechanism for the formal recognition of local customary norms as law, in keeping with the Dharmaśāstric principle that customary norms constitute Dharma.

1. The problem

2Anyone who has ever participated in courtroom proceedings, or watched a lawyer reformulate personal matters in the jargon of a legal document, is likely to have been struck by the gulf between “ordinary terms” and “legalese”. It is not just a matter of the specialized lexicon of the lawyer, but a whole way of thinking about personal relationships and social activities, and of making arguments about them, that is distinctively “legal”. The differences between these modes of thought and expression are multiple: there is the difference between vernacular speech and specialized technical language, but also between the multiplicity of localized, group-specific, and idiosyncratic understandings and that of a shared, standardized, translocal idiom, and between subjects of the law and the experts and authorities who apply it. The parties to a dispute or those charged with a crime “all have their own definitions of their relations, conflicts, and hardships — their own conceptions of justice or injustice — but these must conform to existing legal criteria” (Zelizer 2000: 833).

3The movement between these two extremes has often been conceived of as a translation (e.g. Zelizer 2000: 833). In Europe and the Americas today, the translation is one between registers of the same language, with the legal language both more formal and more reliant on a vocabulary of Latin loan-words drawn from the classical Roman law tradition. The oscillation between these registers, or “codes” (as linguistic anthropologists call them) can be called “code-switching”. The legal situations that I will discuss in what follows would then be examples of domain-specific code-switching, in which the change between codes or registers correlates closely with shifts of audience, purpose, and function that are closely linked in a well-bounded institutional domain. However, the term “code-switching” is so closely associated with the ad hoc and largely unconscious habits of conversation that it is not necessarily the best way of representing what happens in interactions between colloquial and legal discourses.

4In legal discourse, in fact, the formal register is a regular, wellbounded idiom that can at times be incomprehensible to those unfamiliar with it; this amounts to something more systematic than code-switching. Furthermore, one can find instances of a more extreme difference between registers, including the use of completely different languages, or the use of a mixed dialect for legal business. There, we are confronted with a sort of extended diglossia.2 In extended diglossia, the language of everyday life (L1) alternates with another, learned language (L2) deemed appropriate to certain uses and contexts. An intermediate situation is the one in which L2 is not a different language altogether, but an artificial version of L1 infused with many elements of a learned language in the form of loan-words and foreign phrases: perhaps we might call it L1a.3 Where L1a is primarily a written register, the situation has been called “medial diglossia” (Ramseier 1998: 264), and where specialized function determines usage, “functional diglossia” (Rash 1998: 50–52).

5The notion of “legal diglossia” has already been proposed, albeit briefly. In comparing the oldest versions of two 13^th-century Slavic legal texts, Matejka (1977: 195) proposed to explain the occurrence of “certain East Slavic features in the Church Slavonic Zakon [Sudnyj Ljudem] and certain Church Slavonisms in the East Slavic Pravda [Russkaja]” in terms of legal diglossia “defined as two co-occurrent varieties of a single system of literacy” used for legal expression. In this case, one codex was written in a local dialect infused with a few high-register elements, while the other was written in the high register with a few vernacular elements. Matejka noted that over time, the increasing prestige of Church Slavonic led to its displacing of East Slavic dialect in later versions of the Pravda. Around the same time, in a long essay, a sociologist of language suggested treating “legal English as a form of diglossia” (Danet 1980: 473–74), calling attention to the distinctive features of the high-register language in a diglossic situation, including its specialized functions, prestige, literary heritage, status as a learned language, and standardization — features identified in Ferguson’s “classical” model of diglossia.4

2. Code-switching, loan-words, and diglossia

6We find a striking ancient example in a Latin marriage-contract from 2^nd-century CE Egypt drafted by Greek-speakers.5 In Roman Egypt of the time, Latin was the official language of the army and of testaments made by Roman citizens, but other legal work was usually done in Greek. In this document, the legal terms were in correct Latin, but the terminology of ornaments and domestic items composing the dowry was full of Latinized Greek and Greek words transliterated directly into Roman letters. The majority of these words are unattested elsewhere in Latin, and were instances of code-switching — Greek words used in place of equivalent Latin terms which were either unknown or unfamiliar to the writers, whose Latin training was focused on formal legal composition rather than colloquial expression (Leiwo and Halla-Aho 2002: 576–580). However, although Leiwo and Halla-Aho speak of “code-mixing” — that is, “lexical and morphological interference” in one code from another — in texts shaped primarily by speech habits, they suggest that “where text genre or tradition had an effect on language choice” diglossia is at work (561–562).

7Walter Ong called this sort of situation “cultural diglossia” (1984). Although Ong saw the case of medieval Latin as historically unique, much of what he says about it could apply to the use of Sanskrit in medieval India and early historical South-East Asia. For instance, he proposes that “a textualized, chirographically controlled language such as Learned Latin aided greatly in establishing the distance between observer and observed, between the knower and the known, that science, and especially modern science, required” (Ong 1984: 9). This claim that the use of such a language promotes objectivity and provides “distance” applies well to the impersonality and abstraction of most Sanskrit discourse.

3. Bilingualism in the spread of Brahmanical law

8Robert Lingat once made just such an observation about the role of Sanskrit scholastic tradition in influencing regional legal practice:

Hindu law, rather than imposing its injunctions and superimposing itself on customary rules that it might have claimed to supplant, sought only to offer principles and methods, classes and distinctions, thanks to which a truly local law was able to take shape. Hindu law brought not so much prescriptions as norms, not ready-made institutions but frameworks for structuring institutions already in existence or in the process of formation.6

9Along with Śaivism, Vaiṣṇavism, and Buddhism, and the formal resources of Sanskritic literary culture, Indic legal models, expertise, and institutions were prestige imports into South-East Asia.7 Without rehearsing the details of this process, and the historiographical problems they pose, it may suffice to note that the process seems to have begun when local South-East Asian elites began to emulate the cultural norms and practices of high classical India, to which they were exposed by the Indic trade communities (initially from the Kadamba-and Cāḷukya-ruled west coast) that developed around ports in present Vietnam, Sumatra, Borneo, and Java.8 This emulation — likely motivated by a combination of material and intellectual aspirations — gave rise to royal states that presented themselves according to Indic models, and sought to participate in the reciprocal relations with kingdoms in India, with diplomacy (then as now) as the facilitator of both trade and cultural transmission. Like Indian kings, the kings of mainland South-East Asia and the archipelago embraced and patronized Indic cults, relying on priests and monks to confer on them divine blessings and the marks of sacral authenticity, and later sponsoring the construction of grand temples.

10As an integral part of these religious systems, and in particular as a corollary of adopting the Brahmanical idiom of kingship, these rulers sought to institute Dharmaśāstra, a system of brahmin-authored, textually based precepts, as the standard of justice. Just as with other Indic cultural traits, the legal elements underwent various sorts of adaptation as they were established. This process casts into relief the largely analogous processes that were always at work in India.

11The most basic of these is the elusive interplay between the scholastic, textually codified system of the Dharmaśāstra and the regional, local, and group-specific customary legal traditions which generally remained unwritten (lex non scripta). It is well known that there is a distinction between the norms prescribed as eternal or divinely revealed in the Śāstras, and rules that actually serve as the criteria by which transactions are made and disputes are settled in any given jurisdiction. It is also well known that the Dharmaśāstra recognizes this distinction, accounts for it in its procedural rules, and broadly sanctions the application of such local standards and statutes. What is not so well known is how the two overlapping sets of principles were applied in particular cases, since the recording of such cases is comparatively very rare, and those records which exist do not tell us everything we should like to know.

12Our documentary sources for actual procedures and decisions in India are thus just as fragmentary as the Śāstric sources are copious. What we do have are epigraphical records of various types, formulary compendia (such as the Lekhapaddhati), and, at the cusp of the colonial era, the Maratha legal records (1600–1818, Maharashtra and Tanjore).9 For South-East Asia, we likewise have inscriptions and some legal books, modeled initially on the Manusmṛti but composed in local languages and frequently departing markedly from the Indian Śāstric model. My contribution here will be to synthesize parts of this evidence — viz. to note certain trends in the epigraphy, with close attention to some examples from the Tamil country; to make some observations about discursive practice in such documents; and finally to propose some interpretations.10

3.1. Cambodia

13George Cœdès, citing Lingat with approval, produces as supporting evidence a terminological division of labor in an inscriptional legal record from 10 th-century Cambodia. He observes that the “external apparatus” (l’appareil extérieur) of a lawsuit was recorded using Indic vocabulary, while the “essentially local material facts, which constitute the very substance of the suit” (les faits matériels, essentiellement locaux, qui constituent la substance même du procès) are expressed in the Khmer language (Cœdès 1954: 66–67):

14The result is a mixed idiom that might be compared with the Sanskritized Tamil called maṇipravāḷam (“pearls-and-coral”), in which medieval Vaiṣṇava theology was written. The 10^th-century Khmer record in question records the details of a legal dispute over the status of the last of a series of women who had been handed over to the owner of a buffalo as surety (phtā) against the buffalo itself, which has been loaned on interest. The point in dispute was whether or not the holder of the surety had the right to require this woman to serve in a temple of Śivapattana. The suit (gati) was brought not by either direct party to the loan but by a couple presumably having some interest in the woman’s fate. The plaintiffs’ charge turned on the circumstance that this woman, Tai Kaṃpit, the niece of the first woman (Tai Kanhyaṅ) handed over as surety, was a replacement for her aunt who had absconded (and an intervening replacement, who had also run away). In each case, the women serving as surety had in turn been given by donative act (duk ta praśasta) to the Śivapattana temple. The plaintiffs objected to the employment of Tai Kaṃpit in the temple and had removed her from there, on the grounds that she was merely the replacement for her aunt, and that indeed the aunt herself had been handed over as a surety against a loan, not as a price of purchase. These circumstances, it was argued, implied that employment of the niece in the temple was irregular. Thus, the question hinged on the rights of the surety-taker over the surety.

15After the hearing, the court found the complaint without merit, penalized the plaintiffs with fifty slaps in the face — a penalty unheard of in India — and ordered Tai Kaṃpit to be returned to the owner of the buffalo for continued service at the temple. The decision was signed by the sabhācāre (a representative of the king, according to Cœdès), the four sabhāsats, and two other officers of the state. Thus, Cœdès found that Sanskritic expressions are used in the Khmer text to denote jurisprudential categories and the components of legal process, as well as allusions to Śāstric authority;11 and the local vernacular to frame circumstantial particulars, especially the commercial terminology.12 Moreover, it can be shown that this pattern is replicated in a number of other Khmer inscriptions dealing with legal matters.13

3.2. Java

16The situation in Java differs in that, besides legal inscriptions, a tradition of law-books has survived to the present day.14 The earliest inscriptions in Java (5^th–8^th c.) were in Sanskrit or in a coastal dialect of Old Malay, and were primarily religious or political in import. Although the precise circumstances are not known, the use of inscriptions to document financial transactions and related legal obligations begins with the monetization of the Javanese economy. We have here a shift in the use of writing (or at least the use of epigraphy) from an explicitly sacral or political function to broader legal and economic applications. The corresponding broadening of the range of society having recourse to such documents is, not surprisingly, confirmed by the fact that the legal inscriptions are almost always written in Old Javanese, which no doubt was understood by a larger portion of the population than understood Sanskrit.15

17These legal inscriptions are not limited to transactional records. There are several examples of jayapattra or jayasoṅ (records of successful suits), beginning with judgments recorded in inscriptions of the early 10^th century16 (and continuing in legal usage up to the 18^th century).17 These records refer to legal procedures that are explicitly part of Dharmaśāstra jurisprudence: one, the “Jayapattra Dieduksman”, notes that “his dispute was adjudicated18 by the Samĕgĕt Pinapan” (pinariccheda guṇadoṣa nira de samaggat pinapan), and then gives an account of the “cause of the dispute” (sambandha nikaṅ guṇadoṣa). The Sanskrit loan-word pariccheda, “decision [between right and wrong]” (here inflected with the Javanese passive infix-in-), used with the compound guṇadoṣa (“merit and fault”, “right and wrong”; hence, legal dispute), unmistakably echoes such phrases from Dharmaśāstra as guṇadoṣaparīkṣaṇam (“investigating what is right and wrong”, Mānavadharmaśāstra 1.117d) and guṇadoṣavicakṣaṇam (“distinguishing between right and wrong”, Mānavadharmaśāstra 9.169b).19 The dispute is “fully resolved” (sampun śuddhapariśuddha20 ikaṅ guṇadoṣa) in the presence of “witnesses” (sākṣī), who are listed, and the writer of the actual “written document” (likhitapātra) is identified — in all of this relying on Indic technical terms for these components of the legal process. The record concludes: “The purpose of this ‘victory-document’ is so that the matter may never be spoken of again” (kunaṅ sugyan tatān paṅujara ya muvaḥ dlāha niṅ dlāha ya donikeṅ jayapātra), a sentiment expressed in some Indian legal inscriptions as well. Discharge of a debt could also get recorded in a document, a śuddhapattra, which certifies that the debt “was cleared” (śuddha hutaṅ).21

18Neither of those 10^th-century records cites textual authority, but there are other indications that the Javanese had become familiar at least with the Indic notion of authoritative textual authority (i.e. “Āgama”) from the 8^th century at the latest: the earliest such allusion comes in the Caṅgal inscription of 732, in which King Sañjaya compares his father to Manu, the legendary Indian law-teacher. Three 12^th-century inscriptions22 record orders of the Balinese king Jayapaṅus that claim a basis in the “essence of Dharma in Manu’s teaching” (pöh niṅ Mānavaśāsanadharma) or “the essence of the Mānawa [Dharmaśāstra] and Kāmandaka [Nītiśāra]” (pöh niṅ Mānava-Kāmandaka), though without quoting directly from them.23

19The well-known Bendosari jayapattra (“Decree Jaya Song”, 14^th c.),24 like several other Majapahit-era inscriptions, refers to a composite tradition called Kuṭāra-Mānava. If anything, Śāstric idiom is more prominent than before: each “litigant” (vyavahāri) in the dispute (kavivakṣanya) is eager to argue his “case” (pakṣa, pakṣadharma) before the magistrate (pragvivaka = Skt. prāḍvivāka, plate 4 recto).25 The Dharmaśāstric principle of long-term possession (bhukti) as evidence of ownership of property (dravya) gets invoked (as in the Tamil “Exhibit E,” below): “People acknowledge me as owner of 33 lirih on the basis of possession… that is firm possession since the time of my ancestors” (ka bhukti deniṅ amadṛvyakĕn lirih 33… punika ta sthiti bhukti saṅkeṅ tuha-tuha; plate 5 recto 1, with Sanskrit elements in bold), “because it has been in our possession from time immemorial” (makahetu anadi kābhuktyanipun; 5 verso 1).26 Finally, the record describes the criteria used to decide the case:

5v5 …pinametakĕn śastradṛṣṭa, deśadṛṣṭa, udāharaṇa, guru kaka,
6r1 makataṅgvan rasāgama ri saṅ hyaṅ kuṭāramānavādi, maṅanukāra pravṛttyacāra sang pāṇḍita vyavahāraviccheda 6r2 ka riṅ puhun malama (plate 5 verso 5–6 recto 2)

…the norms of the Śāstra, the norms of the country, illustrative cases, and ancient teachers, relying on the essence of tradition found in the holy Kuṭāra, Mānava, and other books, imitating the character and conduct of the scholars who decided lawsuits in olden time.

20These parameters are not merely stated using Sanskrit legalisms: they reproduce Dharmaśāstric rules. Thus, Mānavadharmaśāstra 8.2 prescribes that the king render judgments “taking the norms of the Śāstra and the norms of the country as his grounds” (deśadṛṣṭaiś ca śāstradṛṣṭaiś ca hetubhiḥ), and in the Kauṭilīya Arthaśāstra, udāharaṇa (“illustrative case”) is mentioned alongside Dharmaśāstra and Arthaśāstra (1.5.14), and also alongside hetu (“good reason”) and dṛṣṭānta (exemplum, 2.10.9), as the basis for royal decisions.27 Imitation of the conduct (ācāra) of the learned and virtuous is well known as the third “root” of dharma.

21As in Khmer legalese, terms of art native to local institutions, or words for everyday affairs, social statuses, and more generic legal notions, are more likely to be vernacular:28 patuku-sava or pamidara (“blood-price”), patéh-patéh (“orders”). Likewise, the standards of taxation, ownership, and financial obligations are all defined by local standards and mainly in local terms.29 Moreover, in Old Javanese legal usage, many Sanskrit terms underwent semantic developments not found in Indian usage: viśeṣa (in Sanskrit, “distinction, excellence”) means “power, authority” in Old Javanese;30 similarly: guṇadoṣa as shorthand for “legal dispute”, kāyānurūpa (referring to honorific gifts “in accordance with rank”), and sukha-duḥkha added at the head of lists of offences punishable by a fine paid to the beneficiary of a grant — its meaning in this context is not clear, but it is not found in analogous lists of the “ten crimes” (daśāparādhāḥ) in the Sanskrit śāstras.31 Such lexical items were “naturalized” in Old Javanese legal discourse to the extent that certain typical idioms reflecting Old Javanese syntactic habits were developed as part of Old Javanese legalese. For instance, Sanskrit sīman (“border”), which comes to refer by extension to a piece of land with fixed boundaries, is used with the Old Javanese demonstrative particle para, which signifies multiple members of a category, in the reduplicative compound sīma-para-sīma (“various sorts of land parcels”). Similar sounding compounds, but employing Sanskrit prefixes, are also used: śuddha-pariśuddha (“completely cleared of debt”), bhāṇḍa-paribhāṇḍa (“all kinds of merchandise”).32 Thus even as Sanskrit itself fell out of use, it lived on in loans and fixed phrases in Old Javanese legal and religious usage.33 And although literary Old Javanese was highly Sanskritized throughout, some of the words and expressions found in legal contexts seem to be uncommon elsewhere,34 and thus may be regarded as distinctive of a functionally specialized register of the language.

3.3. Tamil Nadu

22This situation in South-East Asia in fact has a close parallel in India itself, which I will attempt to illustrate with some examples from the Tamil country.35 The Sanskrit legal and political terminology encountered in records from the Tamil country derive ultimately from literary discourse shaped by śāstric precepts and by ideals illustrated in the Sanskrit epics and Purāṇas — that is, from a thought-world far removed from vernacular South Indian civil society.

23The earliest such records that we have from South India are inscriptions in Prakrit from as early as the 4^th century, and Sanskrit from the 5^th, but these soon gave way to the use of Tamil for the technical portion of the record, and later even for the conventional ornate versified eulogy of the reigning king, the meykkīrtti.36 The legal content of these inscriptions was from the start composed in a register of the language that was as clearly marked offfrom the vernacular as English legalese is from everyday English — that is, often unintelligible to the uninitiated. The distinctive features of this Tamil legalese are: (a) extensive borrowing of Indo-Aryan terms for offices and functions for which no Tamil word existed; (b) the preferential use of Indo-Aryan-derived terms of art in place of Tamil synonyms that may simply have seemed too colloquial or too imprecise for a special sort of official business; (c) the influence of Indo-Aryan categories and śāstric idiomatic patterns even where Indo-Aryan-derived terminology is not present.37

4. Criteria for distinguishing differences of register in ancient South Indian texts

24It might well be objected that it is difficult to distinguish nuances of register in texts such as these in the absence of a large surviving corpus of definitely colloquial vernacular usage. More particularly, it might be argued that the writers of these texts did not recognize terms of Indo-Aryan origin as being loan-words or as different linguistically from near synonyms of Dravidian origin. This of course is a legitimate concern. But on the basis of analogies with similar situations in other languages, I propose certain suggestive (though not infallible) criteria:

• The presence of Indo-Aryan tatsamas — loans that retain their original form, and have not been subjected to the modification required to represent the word within Tamil phonological parameters.

25The presence of Indo-Aryan tatsamas is more likely to reflect the user’s awareness of using a technical term or item of elite register than may be presumed for tadbhavas, insofar as the latter are more likely to have been a part of the Tamil lexicon for a longer time, and thus less likely to be perceived as foreign or high-register lexemes on this basis. This seemingly unambiguous criterion actually has a wide grey penumbra. Indo-Aryan loans have been coming into Tamil since the earliest surviving Tamil texts, and have come in from different sources (Sanskrit, Prakrit, Marathi, etc.). Many of the most commonly occurring Indo-Aryan legal terms in Tamil were taken over from South Indian Prakrit inscriptional usage. A good example is āṇatti, “executor of a grant” (cf. Skt. ājñapti)38 and the corresponding finite verb āṇapayati, both of which were used in the Pallava-era Prakrit inscriptions from at least the 4^th century CE.39 The word appears in early inscriptional legal documents in Tamil (from ca. 7^th c.).40

26Nevertheless, it might be argued that even a tatsama like āṇatti or āvaṇam (“deed, document” < Pkt. āvaṇa; cf. Skt. āpaṇa)41 has been so far appropriated as no longer to be perceived as a loan-word (which is not to imply that it is not a technical term all the same). But the very diversity of Tamil phonological realizations of the Indo-Aryan word for “order” (Pkt. āṇā / Skt. ājñā) — āṇai, āññai, ākñai, and ākkiṉai all occur — might be taken as a signal of the word’s “learnedness” (even if part of the reason for the diversity may be the word’s multiple entries into the language).

• Orthography: the use of Grantha characters, a modified and expanded version of the Tamil alphabet that was designed expressly for writing Sanskrit.

27The presence of Grantha characters in a record may also provide some clues, at least to the perceptions of the actual writer. It is arguably a confirmation that the writer recognizes the Sanskritic origin or association of a term, although the non-use of Grantha does not necessarily entail the opposite conclusion. This is a complex factor. It is common for the same Indo-Aryan-derived word to be written with Grantha letters and in Tamil letters (which require phonological simplification of the word) just a few lines apart in the same document. Grantha and Tamil characters are very commonly mixed, and even when Grantha is used, it is not necessary used correctly or consistently. Moreover, it might be argued that professional scribes may not even have considered these different scripts, simply viewing the Grantha forms as supplementary characters that fill out the phonological range of the Tamil alphabet (though Grantha actually has distinct forms for many of the phonemes that Sanskrit shares with Tamil).

28There may be some such cases, but my sense is that in general, scribes, though inconsistent, recognized at least part of the Indo-Aryan-derived technical jargon as “foreign”, or at least in need of a special, learned orthography. It is notable that, despite the highly unstandardized state of Tamil epigraphic orthography, one never (or almost never?) finds Grantha characters used in non-Indo-Aryan-derived words. In fact, in my experience, Grantha is used only for Sanskrit loans; Prakrit tatsamas (e.g., āṇai, āṇatti, āvaṇam, veṭṭi,42 etc.) are written in Tamil letters, a circumstance that may be explained by two complementary factors: (1) Prakrit is phonologically simpler than Sanskrit, and has fewer sounds that are absent in Tamil — it “fits” the Tamil script somewhat better; (2) in legal discourse at least, Prakrit words were borrowed at an earlier stage, and thus may have seemed to “belong” to Tamil more, while Sanskrit elements, received later, would have stood out as more stilted and artificial.

• Other stylistic or formal factors, such as a close affinity with actual śāstric usage, or with Prakrit or Sanskrit precedents, which can reinforce our impression that a term or construction was chosen on the strength of that affinity, i.e. that it is a marked usage.

29We have a clear example of the influence of Prakrit discourse in the idiomatic habit of using the same verb in two forms, the second time with a causative stem, so as to cover both direct and indirect agency in following or breaking the law, and thus to avoid ambiguity and the use of loopholes in subsequent disputes. Consider these early 4^th-century Pallava examples:

pariharatha pariharāpetha ca43
exempt, and cause [others] to exempt [the brahmadeya from the usual obligations].
pariharitavaṃ parihāpetavva ca44
[the brahmadeya] should be exempted, and caused [by others] to be exempted.
jo amhasāsanaṃ atichitūna pīlā bādhā karejjā kārāpejjā vā45
he who, having violated our order, should cause, or cause others to cause, even a small impediment [to this order] …
atha koci… pilā bādhā kareyya kāravejjā46
now if anyone… should cause, [or] cause others to cause, even a small impediment [to this order] …

30Likewise, in a 5^th-century Pallava Sanskrit inscription:

tad imaṃ sarvvaparihā[raiḥ] pariharttavya[m] parihārayitevyañ ca47 This [granted land] should be exempted and caused to be exempted by others in accordance with all the [eighteen] exemptions.

brahmadeyīk[ṛ]ty[ā]smābhis samprattas tatas sarvvaparihārair a[yaṃ] grāmaḥ pariharttavyaḥ parihārayitavyaś ca48
I have created this as a brahmadeya. On that account this village should be exempted and caused to be exempted by others in accordance with all exemptions.

31This device appears later in Tamil inscriptions, sometimes in even more elaborate constructions. It appears to have been found especially suitable for use in referring to violation of a decision or order, and its consequences (as in two of the Prakrit examples above). Here are two examples from 11^th-century Cōḻa inscriptions:

inta samahāram kūṟu aḻivikkappaṇṇuvārum paṇṇuvippārum viṇṇappañ ceyvārum ceyvippārum tiruvāṇai49 […] innilamum maṉaiyum cuṭṭi tuṭarvārum tuṭarvippārum kuṭuppippārum tiruvāṉai50
Those who abolish this division [of land], and those who cause others to abolish this division, and those who petition for [such an abolishment], and those who cause others to petition [are violators of] the royal order (tiru-āṉai); those who usurp and those who cause others to usurp these lands and house sites [are violators of] the royal order.

ippaṭi allatu ceyvippārum ceyvārum tiruvāṉai enṟu sabhāvyavastaipaṇṇinom mahāsabhaiyom, ivviyavastap paṭiye cilālekai ceyka-v-eṉṟum tiruvāymoḻintaruḷit tirumukam pirasātañ ceytaruḷi vantamaiyilum, ippaṭi allatu ceyvippārum ceyvārum tiru[vā]ṉai eṉṟu cilālekai ceytu kuṭuttom peruṅkuṟiperumakkaḷom51

We the members of the great assembly have promulgated the “decision of the assembly” (sabhā-vyavastai), saying that those who cause others to act otherwise than in this way, and those who [themselves] act otherwise [are violators of] the royal order. And having gratefully received a royal order which was pleased to decree that an inscription should be made in accordance with this decision, we have made and presented a stone inscription (cilālekai = Pkt. silālehā or Skt. śilālekhā), saying that those who act otherwise than in this way, or who cause others to act otherwise [are violators of] the royal order.

32Another formal transfer is in the reference at the end to the verbal instruction of the king. In the earliest, Prakrit examples we find only a simple statement of who issued the order to inscribe the record, that is, who acted as āṇatti (ājñapti). This was sometimes, but not always, the ruler who made the grant or decree:

āṇatī sayatti dattā paṭṭikā52
The copper-plate record has been given by my own order.

sayam āṇataṃ53
Ordered by myself.

33But as early as the 5^th century, the formula (now in Sanskrit) was made more emphatic by the mention of the king’s “own mouth”:

bhaṭṭ[ā]rakā[ṇāṃ] svamukhājñāptyā likhita[m i]daṃ54
This was written by the command from the king’s own mouth.
prabhos svamukhājñāptyā neminā likhitaṃ55
Written by Nemi by the command from the king’s own mouth.

34It is this particular trope that inspired phrases that become typical of the Tamil records:

tiruvāy moḻintu
tiruvāy moḻint-aruḷa/aruḷi
tiruvāy moḻint-aruḷiṉār-eṉṟu
tiruvāy moḻint-aruḷiṉapaṭikku
tiruvāy moḻint-aruḷiṉamaiyil
the royal mouth having spoken
the royal mouth having had the grace to speak
since the king has graciously told by his sacred mouth56

tiruvāykkeḷvi
officer “who listens to the royal mouth”

35The notion even gets ritualized in the treatment of the royal order itself, which is referred to as the tirumukam (< Skt. śrīmukha, “holy mouth”) and placed on the heads of those receiving the order (a practice described already in one of the earliest Tamil records, the bilingual Pallankovil plates of Siṃhavarman III, ca. 546 CE).

5. The exhibits

36The following sample records, chronologically arranged, are intended to illustrate some of the phenomena that I have described. In the texts, I have set in bold type words of Indo-Aryan origin occurring in Tamil passages, including proper names and titles (interspersed stanzas wholly in Sanskrit are identified as such). The purpose of the bold type is to give an impression of the frequency and distribution of Indo-Aryan lexical items in epigraphs composed in Tamil. It is important to bear in mind my caveat that some of these items — early borrowings from Prakrit, in particular, were less likely later on to have been perceived as foreign in the way that many later Sanskrit borrowings were.

37Exhibit A: A bilingual grant to endow a temple (mid-9^th century)

38First, consider a bilingual copper-plate grant from the middle of the 9^th century. It begins with a lengthy panegyric in Sanskrit praising the donor, towards the end of which the grant itself is mentioned in broad terms. The terms of the grant are spelled out in detail in the Tamil portion:

39Vēlūrpāḷaiyam Plates, 6^th year of Nandivarman III = ca. 852–3 (ARE 1911, no. A.24; SII 2.5.98 [501–517] and 12.49; IP 121)

41–47 [Skt. panegyric of the donor ends with:]
tasmai devāya śarvvāya pūjāsatrādikarmaṇe /
so dāg57 grāman tirukkāṭṭuppaḷḷināmānam īśvaraḥ //
vijñaptim atrākṛta coḷavaṃśacūḷāmaṇir vviśrutavikramaśrīḥ /
dhīraḥ kumārāṃkuśanāmadheyas tyāgena rādheyasamaḥ kṛtajñaḥ //
atrājñaptir abhūn mantrī nampanāmā mahīpateḥ /
agradattānvayavyomaśarannīhāradīdhiti[ḥ] //
vāgmanaḥ[k]āyakarmmāṇi parārtthāny eva yasya saḥ /
māheśvaro manodhīraḥ praśasti[ṃ] kṛtavān imām //
47 [Tamil:] Puḻaṟkoṭṭattu Nāyaṟunāṭṭut Tirukkāṭṭuppaḷḷip pañcavaram āïrakkā-
48 ṭi. itu Kovicaiya Nantivarmmaṟku yāṇṭu āṟāvatu Coḻamahārājar viṇṇappattā-
49 ṉ iraiyūr uṭaiyāṉ Nampaṉ āṇattiyākat Tirukkāṭṭuppaḷḷic Caṉṉa-
50 kkuṟi Yajñabhaṭṭa[r] eṭuppitta Yajñeśvarattu mahādevarkku nāṭṭu nīṅkalā-
51 y uṭpuravāy tevardāṉamākap peṟṟataṟkup peṟṟa parihāram: nāṭāṭciyum
52 ūrāṭciyum puravupoṉṉum tirumukkāṇamum vaṭṭināḻiyum putāḻiyum taṭṭu-
53 kāyamum īḻampūṭciyum iṭaippūṭciyum maṉṟupāṭam(t) tarakum taṟikkūṟ[ai]-
54 yum kūlamum nallāvum nallerutum nallāṭum nāṭukāvalum ūṭupokkum
55 kallāṇakkāṇamum kucakkāṇamum pāṟaikkāṇam paṭṭiṉaceriyum maṟṟum ivvū-
56 rai[e]llai u[ḷ]ḷakappaṭṭatu kottoṭṭuṇṇappālat ellām ev[v]ak[ai] ppaṭṭa-
57 tum kokk[o]ḷḷappeṟāte iv-Yajñeśvarattu mahādevarēyi koḷḷape-
58 ṟṟataṟkup peṟṟa vyavasthai: cuṭṭoṭṭāl māṭamāḷikai eṭukkappeṟuvatā-
59 kavum, tamaṉakamum iruveliyum ceṅkaḻunīrum uḷḷiyum naṭappeṟuvatākavum, kāvute-
60 ṅkiṭappeṟuvatākavum, turavukiṇaṟiḻittappeṟuvatākavum, peruñcekkiṭappeṟu-
61 vatākavum, ivvūr ellai u[ḷ]ḷiṭṭa teṅkum paṉaiyum ivarkaḷ maṉamiṉṟi īḻavar eṟap-
62 peṟātārākavum, ivvakaippaṭṭa vyavasthaiyiṉūṭu Yajñeśvarattu mahādevark-
63 kut tevatāṉamāy sarvvaparihāramāka paraṭatti ceṉṟatu / [Skt:]
64–68 sukṛtam idam ajasraṃ rakṣateti kṣitīśās sakalanṛpatiketus so [’]yam āgāmino vaḥ.
haracaraṇasaro[jo] ttaṃsacih(ve) namūrddhnā.
mukuḷitakara[pa]tmo vandate nandivarmmā |
sarvvān etāṃ bhāvinaḥ pā[rtthi] vendrān bhūyo bhūyaḥ prārthayaty eṣa rāmaḥ.
sāmānyo [’]yan dharmmasetu[r*] nnṛpāṇāṃ kāle kāle p[ā]lanīyo bhavadbhiḥ /
karakauśalakṛtayaśasā ciṟṟayaputreṇa patrasaṃgho [’]yam.
69 likhitaṃ perayanāmnā sthapatikulavyomacandreṇa.
[Tamil:] Kaccip-
pēṭṭaim Maṉaiccērik-Kāṣṭakārimakaṉ Pērayaṉ eḻuttu /

[in Sanskrit:]… To that god Śarva, the king granted the village called
Tirukkāṭṭuppaḷḷi for the rites of worship, etc.
The heroic crest-jewel of the Cōḷa race named Kumārāṅkuśa, wise, right-acting, Karṇa’s equal in generosity, made the request (vijñāpti) for this [grant].
The executor (ājñapti) here was the king’s minister named Namba, the autumnal moon in the firmament of the Agradatta family.
The Maheśvara-devotee Manodhīra, who in speech, thought, and body acts only for the sake of others, composed this praśasti.

40[in Tamil:] In the sixth year of Kōvijaya-Nandivarman, at the request (viṇṇappattāṉ) of Chōḷa-Mahārāja and with Iraiyūr-uḍaiyāṉ Nambaṉ as the executor (āṇattiyāka), this [village] of Tirukāṭṭuppaḷḷi in Nāyaṟunāṭu, in Puḻāṟkoṭṭam, whose pañcavāram (land tax)58 is 1000 kāṭi (i.e., measures of paddy), was excluded from the district (nāṭṭu-nīṅkal) [and] has been assigned as a giftto god of taxable land (uṭpuravu59 devadāna) in favor of Mahādeva of [the temple of] Yajñeśvara, and the [following] exemptions (parihāra) were secured: tax collected by the nāṭu, tax collected by the ūr, tax to be paid in gold, tirumukkāṇam, sales tax per measure (vaṭṭināḻi), putāḻi (?), tax on goldsmithing (taṭṭukāyam), tax on toddy tappers (īṛampūṭci), tax on shepherds (iṭaippūṭci), court fees (maṉṟupāṭu), brokerage fees, tax on looms, tax on grains and pulses (kūlam), good cows, good bulls, and good sheep; district watch-fee, tax on marginal plots (?) (ūṭupōkku);60 taxes on marriages, potters, quarries, and fishermen (paṭṭinacēri); and all other [income] of any kind which the king could receive and enjoy within the boundary of this village, shall not be collected by the king but by this Mahādeva of Yajñeśvara alone. The [following] vyavasthā is [also] granted: brick multi-storied houses may be built; southern wood (tamakaṉam), cuscuss grass (iruvēli), red water lily (ceṅkaḻuṉīr), and flowering bulbs (or garlic) (uḷḷi) may be cultivated; gardens can be planted; tanks and wells may be sunk; big oil-presses may be made; and toddy-tappers may not climb the coconut and palm trees within the village without consent. With the vyavasthā thus defined, [the village] was conferred on the assembly (paraṭatti) as a dēvadāna, with all exemptions, to Mahādeva of Yajñeśvara.

[in Sanskrit:] “Protect this good deed in perpetuity,” Nandivarman, banner of all kings, his head marked by the lotus chaplet of Śiva’s feet, the lotuses of his hands closed [in prayer], commends to you [who are] the future kings.
This Rāma prays again to all these great kings to come: “This, the common boundary-wall of dharma should be protected by you from age to age!”

This set of plates was written by Pēraya, son of Ciṟṟaya, famed for his feats of manual dexterity, moon in the sky of the family of carpenters.
[in Tamil:] Pērayaṉ, son of the carpenters of Maṉaiccēri in Kaccippēṭu, wrote this.

41In this example, we have extended diglossia in the full sense — the alternation between two distinct languages, with Sanskrit (L2) for ceremonial and literary purposes, and Tamil (L1) for practical matters. But as I argue here, the Tamil portion itself displays a diglossia of registers. The phrases that frame out the legal structure of the record — that make it conform to a pan-Indic Brahmanical model — are more formulaic and have recourse to Indo-Aryan legal terms.

42For the “business portion” of the record, there is a shift into Tamil, but a highly formulaic Tamil infused with technical terms that are loans from Sanskrit. These loans likewise denote legal concepts that belong specifically to Brahmanical jurisprudence, and thus for which no exact Tamil equivalent was available: parihāra (exemption), vyavasthā (decision or decree), and devadāna (endowment to a deity) represent the central legal “facts” created by the record. The two legal agents of the acts are also designated by their Prakrit/Tamil titles viṇappattāṉ and āṇatti, which are so central to the legal task that they are mentioned in the Sanskrit introduction, using the cognate Sanskrit terms vijñapti and ājñapti. Most of the other legal terms, denoting various taxes,61 administrative divisions, and governing bodies, are in (un-Indo-Aryanized) Tamil. Though also rather formulaic, these passages have a somewhat more descriptive or discursive quality that was not as far from non-legal Tamil as the parts laden with foreign jargon.

Exhibit B: Contract regarding a giftto a temple (mid-10^th century)

43My second example is a late Rāṣṭrakūṭa-era temple inscription from Bahur (966) that, rather than being a simple deed of gift, serves as a contract in writing between a socially prominent donor and a group of temple servants called the puṇyagaṇa. The members of the latter group commit themselves to providing ceremonial meals to the donor in perpetuity using food extracted from lands purchased for the village council using the donor’s funds (dravyam). Like other records of this period and region, it stipulates a penalty for defaulting on the agreement: a large fine to be paid to the dharmāsanam, the royal court of justice.

Tirumūlanātar Temple, Bahur, year 27 of Kaṉṉaradeva = 966 (PI 9; ARE 1902, no. 183; SII 7.810)

1 svasti śri śrikaṉṉara deva[r*]kki yāṇṭu [2]7 āvatu vākūr śrimulast[ā]
2 ṉattu puṇṇyakaṇattup perumakkaḷ[o]m ivvūrāḷum perumakkaḷuḷ ampa-
3 ṭukkai kovintapaṭṭaccomāciyār makaṉ mati[cū] tyakramavi[t*]taṉukku oṭṭi iṭṭuk kuṭutta
4 paricā[va*]tu i [va*] ṉukku iśrikoyi[li]le vaiccu ira[ṇ]ṭu kalam uttamakgraṃ iraṇṭukaṟiyu[m]
5 neyyum aṭṭuvirum kāyamilaiyoṭum cantr[ā]tittaval ūṭṭuvata[r*]kku veṇṭu[m*] dravya
6 m ivaṉpakkal koṇṭu idravyamey kuṭut[tu] ūrār pakkal cantrātittaval i[ṟ]ai
7 yili poṉ kuṭuttu koṇṭu tiru[t*]tiṉa pūmiy ivvur kaṭamperi kiḻ teṉkaḻaṉi ka-
8 ḻuveṟimeṭṭu umāpiṭāri eṉṉum ceṟuvu mikitik kuṟaimai uḷḷaṭaṅka patiṉeḻu ceṟuvi[l]
9 bh[o]gattilum ivaṉuk[ku] iraṇṭu kalam u[t]tamākram ūṭṭavom āṉom niṉṟa [po]-
10 kamum uttamākram ūṭṭuvom ākavum / itu[kku] virotam cetārait dhanmāsanamu-
11 talākat tā[n] veṇṭu kovukke vevveṟṟu vakai [i]rupat[tu*] [ai]ñ kaḻañcu poṉ maṉ[ṟa]p(pe)
12 peṟuvār ā-
13 kavum / maṉ-
14 ṟupāṭu iṟu-
15 ttum ippa-
16 ricey iv-
17 [vu]ttama agra-
18 m ūṭṭuvera-
19 māṉom pu(ṇ)-
20 ṇya(ya) kkaṇap-
21 perumakkaḷo-
22 [m]

[1–8] Blessings! In the year 27 of King Kaṉṉaratēvar, we the honorable members of the Puṇyakaṇam (temple council) of the Mūlastāṉam temple in Vākūr provided upon mutual consent (a document) to Matisūtyakkramavittaṉ son of Ampaṭukkai Kōvintapaṭṭaccōmāciyār, a member of the leading citizens of this village, to the effect that:

we shall feed him with two pots of uttamāgram (choice consecrated food), one from each crop, from the land consisting of about seventeen divisions, called Teṉkaḻaṉi Kaluveṟimeṭṭu Umāpiṭāri, in the irrigation area of Kaṭampēri of this village, which was bought on payment of gold from the village committee as tax-exempt for as long as the moon and sun endure, after he (M.) provided the means (dravyam) to pay for provision of two pots of uttamāgram cooked at the temple, along with two side dishes, ghee, areca nut, and betel leaf;
[9–10] we shall feed him with this uttamāgram even if the crop fails;
[10–13] from those who contravene this [contract], let the fine collector (maṉṟappeṟuvār) collect 25 kaḻañcus of gold in various forms as fine [paid] to the Court of Justice (dharmāsanam).
[13–22] In that case, after having paid the fine, we the honorable members of the temple administration shall feed him this uttamāgram.

Exhibits C and D: Certification and Reconfirmation (late 11^th and late 12^th century)

44One common context in which Śāstra-derived lingo occurs is where the writer seems to be concerned to emphasize the “official status” or authoritative validity of the document being prepared — that is, where appeal is being made to a principle developed primarily in Sanskritic jurisprudence. An inscription from the late 11^th century was inscribed in the Vaṭamūleśvara temple as a public attestation to the authenticity of a deed of sale, as stated in the final section of the record:

Kīḻappaḻuvūr, near Tiruccirappalli, 20^th year of the reign of Kulottuṅga I = 1090–91(SII 3.71)
22 … ikkācu onṟum āvaṇakkaḷiye kaiccellaṟakkoṇṭu [vi]ṟṟu vilaiyā-
23 vaṇa[m] ceyi[tu] kuṭutt[o] m brahmate[yam] ciṟupaḻuvūr sabhaiyom innilam oru[m] āvukkum ituve vilaiyolai āvatākavum itu[ve] poru[m] āvaṟu-
24 [ti]ppo[ruḷci]la[v]olai[y]āvatākavu[m] itul a[l*]latu [v]eṟu porumāvaṟutipporuḷcila[vo]lai kāṭṭakkaṭavar allātārākavum ippaṭi icaiñ-
25 cu ikkācu onṟu[m koṇ] ṭu innilam [o]rumāvum vilaikkaṟa viṟṟu poruḷaṟakkoṇṭom ciṟupaḻuvūr sabhaiyom ivarkaḷ paṇikka ip[pi]ramāṇam eḻutine-
26 n madhyastan pa[ḻuvūru]ṭaiyān ā[yi]rattirunūṟṟuvan muṭikoṇṭānen ivai enneḻuttu …

[22–24] Having received this one kācu in full into the hand … we, the assembly of the brahmadeya of Ciṟupaḻuvūr, sold (the land), and made and gave a deed of sale [vilaiyāvaṇam]. For this one-twentieth (vēli) of land this alone shall be the record of sale [vilaiyōlai],62 and this alone shall be the record of the final payment of the money, and they (viz., the purchasers) shall not be bound to produce another record of the final payment of the money besides this
[24–25] Having thus agreed, having received this one kācu, and having sold this one-twentieth [vēli] of land at the full price, we, the assembly of Ciṟupaḻuvūr, have received the money in full.
[25–26] By their order, I, the Madhyastha Pa[ḻuvūr-U]ṭaiyāṉ Āyirattirunuṟṟuvaṉ Muṭikoṇṭāṉ, wrote this [authoritative] document (piramāṇam); this is my writing [i.e. attestation]… [Other signatures follow.]63

45In this example, we find Sanskrit terms to denote certain official statuses and institutions: the brahmadeyam, the sabhai, and the position of madhyastha. The seeming synonymy of the terms vilaiyāvaṇam and vilaiyōlai, “document”, might indicate that the element āvaṇam no longer sounds more “official” than the term ōlai (literally denoting the physical object, the “palm leaf”); where a more explicit description of the content of the document is desired, a non-Indo-Aryan expression is used: porumāvaṟutipporuḷcilavōlai (“record of the final payment of the money”).

46But where legal force of these documents according to the principles of Dharmaśāstra needs to be attested, the inscription refers to itself, using the term piramāṇam, identical with Skt. pramāṇam, “authoritative proof or record”.64 It functions rather like the seal of a notary public, which transforms an ordinary written record into something that would “stand up in court”. Another example is an inscription from Tirunedungalam, near Trichy, which was intended to constitute a vilaipramāṇa icaivutiṭṭu, “deed of agreement which is proof of sale” (SII 23.718, p. 510, Tirunedungalam, Lalgudi taluk, Tiruchirappalli district, from the reign of Sundarapāṇḍyadeva I).

47Similarly, one also finds instances of a renewal or reconfirmation of the authority of a document, as in a record from the Kailāsanātha Temple at Rishiyūr, Tanjore district, from the 33^rd year of the reign of Kulottuṅga III (1211, SII 23.476, pp. 352–54):

[7] … ivar aruḷicceyya ippuṉapramāṇa icaivutiṭṭeḻutiṉeṉ ippiḻicūṟ karaṇattān āvaṇamuṭaiyān tiruvātavuṭaiyān civāsudevan eṉ eḻuttu …

By his gracious command [i.e. of the king, Arayaṉ
Tiruvaṉaikkāvuṭaiyāṉ Malaiyappiyarāyaṉ of Perumūr], I,
Civāsudevan son of Tiruvātavuṭaiyāṉ of Āvaṉam, accountant of Piḻicūṟ, write this “reauthorizing” deed of agreement (ippuṉapramāṇa icaivutiṭṭu). [This is] my signature.

48Another case is noted by R. Tirumalai (1983) in a Nelliyappar temple inscription from Tinnevelly (ARE 1894, no. 122 = SII 5.411, pp. 138–40, reign of Maṟavarman Kulaśekhara Pāṇḍya). That inscription constituted eka-pramāṇa, a single document to replace multiple earlier deeds. The sabhā of Kulaśekhara-Caturvedimaṅgalam thereby created a new agrahāra called Anavaradadāna-Caturvedimaṅgalam, incorporating prior grants and decrees made by King Maṟavarman for the benefit of the individual brahmins who would hold shares in the new settlement, and replacing earlier svāmibhoga rights (share-ownership as recompense for loss of cultivation rights on land converted into a tank-bed.)

Exhibit E: Appeal to the jurisprudential principle of bhukti (mid-13^th century)

49This example is interesting in that it invokes juristic categories from the Dharmaśāstra, albeit without citing any particular sources. It describes a land dispute between some brahmins (who claim to have been deprived unjustly of some lands) and some Veḷḷāḷas (a high-status agricultural caste) who are accused of usurping the land:

Saumyanātha Temple, Nandalūru, Rajampet taluk, Cuddapah district, year 13 of Rājendra Coḷa III = 1257–1258 (SII 23.580, section III-E, pp. 400– 405)
1 maturāntakap pot-
2 tappiccoḻan olai. peruṅkaṇṭūr akarattu pāka-
3 yaṉum karaṇattāṉum kaṇṭu taṅkaḷūrukkup piṭā-
4 kai āna koḷūr cilakālam tāṅkaḷ aṉupavikka
5 ppeṟāte iṉumpuṭolil veḷḷāḷar balattā-
6 le aṉubhavitte itu nāṉum nāṭavarum kūṭa ārā[y]ntu
7 ivveḷḷāḷarai ivvūr brāhmaṇakṣetram alla v e[ṉ]ṟu niṅ-
8 kaḷ maḻuppiṭittal brāhmaṇarai maḻuppiṭippittu ko-
9 ḷḷukutal ceyyuṅkoḷ enna iraṇṭuppikkum māṭṭā-
10 maiyāl ikkoḷūr pūrvakālattil brāhmaṇarutey ā-
11 kaiyāl piṅkaḷasaṃvatsara[t]til cittiraiviṣuvile annaji
12 yar tirukkāladevamahārājāvukku dharmmam āka taṅkaḷukku nirvā[r]t-
13 tuk[ku] ṭuttu ((i))nta sāsanamum kuṭuttom. innāḷ mutal
14 aṉubhavippate. manmasiddhe(s) sāmānyo [’]yan dharmasetu-
15 n nṛpāṇām kāle kāle pālaniyo bhavadbhiḥ sarvān e-
16 tān bhāvinaḥ pārtthivendrān bhūyo bhūyo ācate rā-
17 mabhadraḥ. dānapālanayor maddhye dānāt śreyo [’] nupālanam
18 dānāt svargam avāpnoti pālanād acyutam padam. svadattām pa((ra))-
19 dattām vā yo hareta vasundharām [ṣa] ṣṭiṃ varṣasahasrāṇi vi-
20 ṣṭāyāj jāyate kṛmiḥ . ippaṭi[k]ku aṟivom periyanāṭṭu vi-
21 ṣayattāro[m] yā[ṇ]ṭu patumuṉṟu ivaiccāl vanta vari iru-
22 nūṟṟu orupattu oṉṟu

[1-14] [This is] the palm-leaf [order] (ōlai) of Maturāntaka, the Cōḻa of Pottapi[nāḍu]. Pākayaṉ of the brahmin settlement of Peruṅkaṇṭūr and the accountant having seen that the Veḷḷāḷas in Iṉumpuṭol have by force, for some time, enjoyed possession of Koḷūr, which is a suburb of their own village, I [= the king] together with the men the nāṭu have investigated, saying to the Veḷḷāḷas, “Grasp the hot axe and say ‘This village is not brahmins’ land,’ and you yourselves get the brahmins to grasp the hot axe”. Since this is not acceptable to both parties, [and] since this (village of) Koḷḷūr belonged to brahmins in earlier times, we have given it with the pouring of water to Aṉṉajiyar [maṭha-priest] Śrī Kāladeva Mahārāja as a pious donation (dharmam) on the equinox of Cittirai in the Piṅgala year, we have given this inscribed edict (śāsanam). From this day forward, [you] may enjoy possession.

[14–20: Sanskrit stanza]
This common (i.e. generally applicable) “boundary-wall of dharma” of Manmasiddhi is to be protected by kings in age after age;65 Rāmabhadra requests all those great kings yet to come [to protect it] longer still.
Between giving and protecting, protecting is better than giving: from giving, one wins heaven; from protecting, [one wins] the Immortal Place.
He who appropriates land that has been given, whether by himself or by others, is born as a worm in dung for sixty thousand years.

[20–22] Thus we who belong to the “country of the greater region” know [i.e. affirm] in the year 13. The tax coming through this [grant] is 211.

50This inscription is interesting on more than one count. First, it alludes to the very ancient ordeal of grasping a red-hot axe head. This is one of the “classical” judicial ordeals described in the Śāstras, though that need not mean that this inscription is evidence for the application of śāstric norms in legal practice. The very contrary may be the case: that such ordeals were an element of customary practice that ended up getting recognized in Śāstra. It is notable, anyway, that the language used to describe the ordeal is not in any way Sanskritic. It is found also in other Tamil inscriptions, such as one from Nandalūru Saumyanātha temple in Cuddapah district (SII 23.577), where members of two families vow to perform the ordeal in order to have hereditary privileges restored. In that case, by contrast, the record is signed by brahmin officials of the king with quite śāstric Sanskrit titles: the dharmādhikārin (“dharma officer”) and the divyādhikārin (“ordeals officer”).

51The other interesting point here is the reference to “enjoyment of possession” as a legal proof of rightful ownership. In Sanskrit works, this is usually referred to as bhukti, one of the three “human” pramāṇas (or proofs) alongside documentation and witnesses. The Vēlvikuṭi Grants in fact cite nīṭu bhukti, “longtime possession”, as a proof.66 Here, instead, four centuries later, we find the maṇipravāḷam verb aṉupavittal derived from the Sanskrit noun anubhava, which in legal contexts means “enjoying possession”, used both to confer (or reaffirm) the brahmins’ right to the land. The Veḷḷāḷas’ recent possession of the land is marked as illegitimate by the use of two Indo-Aryan-derived legalisms of a śāstric character: they had possession of the land only for a short period (cilakālam) and only through the use of force (balattāl; balam, “force”, here inflected with the Tamil ablative suffix). Indo-Aryan kāla, though highly naturalized into Tamil by this time, is probably marked in this context as a legal idiom, alluding to the Dharmaśāstric principle that “possession acquires [legal force] with the passage of time: this is a settled point in the Śāstras” (kālātiharaṇād bhuktir [scil.: balavatī] iti śāstreṣu niścayaḥ, Nāradasmṛti 1.66), though possession over more than one generation is understood; in the case of property acquired rather than inherited, “even if he has used it for a long time, his legal possession of it is not recognized” without other proof (cirakālopabhoge ’pi bhuktis tasyaiva neṣyate, Kātyāyanasmṛti 324). The legal invalidity of forcible possession is clearly expressed in Mānavadharmaśāstra 8.168, in the same terms (bhukta, with bala in the ablative).67

Exhibit F: A highly Sanskritized record of a ‘statutory reform’ (15^th century)

52This much-discussed 15^th-century example is unusual: it is an inscription containing a statutory law adopted by a council of brahmins and intended to apply to all brahmins in South India.68 Its aim was to criminalize the payment of brideprice and to affirm the Sanskritic scripturally based model of kanyādāna as the only legal form of marriage. In effect, then, this inscription aims to abrogate a customary standard widespread in the region, and to substitute an arrangement conforming to Brahmanical Dharma.

Viriñcipuram Temple, r. of Vīrapratāpadevarāya of Vijayanagara, Śaka 1347 (exp.) = 1425 (SII 1.56)69

1 śubham astu
2 svasti śrīmaṉmahā-irājādirājaparameśvarāṉa śrī[vī] rapratāpadevarāyamahārāja pri
3 thivirājyam paṇṇi aruḷāṉiṉṟa śakābdam 1347 eḻiṉ mel cellāṉi[ṉ] ṟa viśvāvasu-
4 varuṣam paṅkuṉi m. 3 ṣaṣṭiyu[m] budhaṉ kiḻamaiyum peṟṟa aniḻattu [i.e., anuṣattu] nāḷ paṭaiviṭṭu irājyattu
5 aśeṣavidyamahājanaṅkaḷum arkkapuṣkaraṇi gopināthasannadhiyi[l]e
6 dharmmasthāpanasamayapatram paṇṇi kuṭuttapaṭi iṟṟai nāḷ mutalāka inta-
7 ppaṭaivīṭṭu rājyattubrāhmaṇaril kaṉṉa[ṭi]kar tamiḻir teluṅkar ilāḷar mutalā-
8 ṉa aśeṣagotrattu aśeṣasūtrattil aśeṣaśākh[ai*]yil avakkaḷum70 vivāham paṇ-
9 ṇumiṭattu kanyādānamāka vivāham paṇṇakkaṭavar ākavum [/] kanyādānam paṇṇāmal
10 poṉ vāṅkip peṇ kuṭuttāl poṉ kuṭuttu vivāham paṉṉiṉāl irājadaṇḍattukkum uṭpaṭṭu
11 brāhmaṇyattukkum puṟampākak kaṭavāreṉ[ṟu] paṇṇi[na] dharmmasthāpanasamayapatram [//] ippaṭikku aśeṣavidyama-
12 hājanaṅkaḷ eḻuttu [//] …

Let there be prosperity! Hail! On the day of (the nakṣatra) Anusham,71 which corresponds to Wednesday, the sixth lunar day, the third (solar day) of the month of Phālgunī of the Viśvāvasu year, which was current after the Śaka year 1347 (had passed), while the illustrious mahārājādhirāja-parameśvara, the illustrious Vīrapratāpa-Devarāya-Mahārāja was pleased to rule the earth, the great men (brahmins of the assembly) of all branches of sacred studies of the kingdom (rājyam) of Paḍaivīḍu drew up, in the presence of (the god) Gopinātha (of) Arkapuṣkariṇī, a document (which contains) an agreement fixing the sacred law. According to (this document), if the Brāhmaṇas of this kingdom (rājyam) of Paḍaivīḍu, viz., Kaṉṉaḍigas, Tamiḻs, Teluṅgas, Ilāḷas, etc., of all gotras, sūtras and śākhās conclude a marriage, they shall, from this day forward, do it by kanyādāna. Those who do not adopt kanyādāna, i.e. both those who give a girl away after having received gold, and those who conclude a marriage after having given gold, shall be liable to punishment by the king and shall be excluded from the community of Brāhmaṇas. These are the contents of the document which was drawn up. The following are the signatures of the great men of all branches of sacred studies … (trans. Hultzsch, with some italics romanized)

53This inscription is noteworthy for a couple of reasons. First, it is common to hear that India has no legislation in the sense of statutes passed by a legislative body to govern a general class of actions (rather than applying only to a particular case at hand) and promulgated publicly. However, this is clearly such a statute. The inscription describes itself with a long Sanskrit compound, dharmmasthāpanasamayapatram, as a “document of an agreement establishing a law”. The signatories to the agreement (and authors of the record) refer to themselves collectively as aśeṣavidyamahājanaṅkaḷ, “great men learned in every science” (a Sanskrit title with Tamil plural suffix), which conventionally denotes the members of the local brahmin council. This ceremonial title explicitly asserts their comprehensive expertise in all areas of learning, which is a rationale for their legitimate authority in this matter.

54The jurisdiction of the law is also clearly indicated: it shall apply to all brahmins of whatever community throughout the kingdom of Paḍaivīḍu. In fact, it is not anachronistic to speak of this law as a “statutory reform” of marriage practices among brahmins in the kingdom. The practice being rejected is one that obtained among some groups in South India. Although the Sanskrit word vivāha, the broadest Brahmanical label for marriage, is applied to this rejected arrangement, it is otherwise described using “plain language,” in contrast to the statutory standard which is designated by the Śāstric term kanyādāna, “gift of a virgin”. Royal titles, calendrical units, temple institutions, and brahmin lineage terms are also in Sanskrit.

55Finally, two penalties are prescribed for violation of this law: a punishment imposed by the state, and a social disability imposed by the caste elders. The first is denoted by the Sanskrit technical term, rājadaṇḍa “the king’s punishment”, a concept central to the Brahmanical ideology of kingship, which was invariably the vehicle for the spread of Brahmanical law. The second penalty, exclusion from caste, is a customary practice which, though recognized and provided with remedies in śāstric jurisprudence, is readily expressed in vernacular terms — in this case, a Tamil expression for excluding (puṟampākak kaṭavār).72

56More than half of the text consists of Sanskrit terms, many inflected with Tamil endings, and linked syntactically by a skeletal Tamil sentence structure. Here, we have a full-blown epigraphic maṇipravāḷam. Official titles, calendrical data, temple institutions, brahmin lineage markers, and legal terms are all Indo-Aryan-derived, and in fact Sanskrit rather than Prakrit. A notable exception are (most of) the terms used to describe brideprice, the customary marriage practice being outlawed here: poṉ vāṅkip peṇ kuṭuttāl poṉ kuṭuttu vivāham paṉṉiṉāl, “both those who give a woman after receiving gold, and those who perform vivāha after giving gold”.

6. Conclusions

57This study was undertaken in the spirit of a scouting expedition; more definitive results will require a more extensive survey of Tamil inscriptions (and of South-East Asian epigraphy as well), with more attention to shifts in usage over time and in relation to political, economic, and religious developments. In the meantime, I offer some tentative conclusions.

58The use of the Indo-Aryan terms is not just for show; they are used because the legal categories and concepts being appealed to are most unambiguously expressible in those terms. This of course entails the larger question, why are Dharmaśāstric (or in any case pan-Indian) legal categories and concepts required? The short answer has to do with the desire of regional South Asian and South-East Asian elites, and rulers in particular, to adopt the political structures, social ideals, and legal apparatus of the Brahmanical courtly culture — what Sheldon Pollock calls the Sanskrit cosmopolis. In South India, the transmission of this cosmopolitanism occurred in tandem with the widespread penetration of brahmin communities that were in close interaction with other social groups even as they remained part of a pan-Indian intellectual network. This may have assisted the diffusion of Brahmanical ideas. Furthermore, the existence of an established Tamil literary tradition provided a reservoir of non-Indo-Aryan-derived expressions that could be used to express concepts otherwise denoted by Indo-Aryan technical terms.

59By contrast, in South-East Asia, brahmin institutions seem to have been less integrated into the social fabric, more a phenomenon of the court and the royal temple. Sanskritic features passed into the local language the way French vocabulary made its way into English and Russian: by emulation of courtly language habits. Throughout India, vernaculars developed literary expression of their own, which even came to replace Sanskrit in the literary portions of inscriptions. But despite a legacy of a thousand years of correct, polished Sanskrit composition, Old Khmer inscriptions yield not a single surviving example of Khmer poetry; Javanese, by contrast, produced a large literature, initially inspired by Indic models, including “Dharmaśāstras” composed in the local language (something that remained rare in India). So although the transplanting of Brahmanical culture occurred simultaneously in South and South-East Asia, the process unfolded differently in different places, in spite of some similarities in the circumstances of introduction.

60In all cases, Sanskritic legal terminology was adopted, but also adapted to the exigencies of use in a new cultural and linguistic matrix. These adaptations included shifts in the meaning of particular terms,73 the accommodation of regional or otherwise non-Brahmanical practices and norms alongside those inherited from Śāstric law, and the application of Śāstric legal structures, procedures, and forms of documentation to local needs and purposes.

61All this does not prove that Sanskrit permeated legal discourse in these environments because it had been made the language of state; indeed, the fact that a regional vernacular was the vehicle of most of the documentary matter in inscriptions (as opposed to expressive or benedictory material) indicates the opposite: the practical matters of administration and law were conducted primarily or exclusively in the vernacular. The Indo-Aryan elements however served important functions. Prakrit, although it was not carried beyond the subcontinent, first played the role of language of state, imported from the North into South India; subsequently, Sanskrit was learnt and used in Brahmanically influenced religious, political, and literary circles, and Sanskrit panegyrics were dramatic “performances” that publicly established the king’s or the noble family’s belonging to a “global” elite that was conversant in aesthetically refined discourse in this learned language.

62At the same time, Tamil, like the regional vernaculars of South-East Asia, was “massively invaded by Sanskrit at the lexical level”,74 while Sanskrit absorbed a far smaller number of Tamil loans in return, and none at all from Khmer, Cham, Malay, or Javanese. The result of this lopsided relationship was the emergence of particular sophisticated versions of the vernacular: “Classical” Tamil (especially its maṇipravāḷam forms) and literary Javanese (Khmer being the exception here), but also the legalese versions of Old Tamil, Old Khmer, and Old Javanese as they are preserved in inscriptions. Their emergence, under the influence of Brahmanical legal and religious categories and concepts, created a situation of medial or functional diglossia: everyday usage confronted with a specialized idiom relying on technical jargon largely drawn from a foreign, learned language. No doubt, over time, many of those foreign, learned elements got “naturalized”; to the degree that this occurred, the distance between the L1 and L1a decreased. Certainly, many Indo-Aryan-derived lexical items, including proper names, survived even in Islamicized Indonesia to the present day.

63But during the centuries when Brahmanically influenced legal systems were in full force in Tamil Nadu, Cambodia, and Indonesia (at least in certain social spheres), a specifically “legal” diglossia seems to have operated, whereby indigenous legal norms were rearticulated under the influence of a prestigious, transregionally recognized (viz. “global”), document-producing legal regime that was simultaneously adapted to local purposes. The Tamil examples selected here illustrate a range of purposes to which this hybrid system was applied, and in particular the way in which Indo-Aryan lexical items got introduced in the first place to denote technical features of the imported legal framework (official legal functions, śāstrically defined obligations and immunities, or the institutional elements of Brahmanical religion), while the vernacular supplied much of the local legal or administrative terms (e.g., divisions of land and financial arrangements).75

64This linguistic phenomenon is the outward sign of a larger process, the spread of a tradition of formalized law in which writing served two purposes: the production of codes “for academic purposes or propaganda,” and the production of documents with legal force.76 Although the codes seem to have been used more in the training of legal experts than directly in administration and litigation, they did provide a conceptual and linguistic framework for the idiom used in the “courtroom” and in legal records. In Tamil Nadu, as in Cambodia and Java, this involved the adoption of a foreign scholastically influenced set of categories and its adaptation to local norms.

65In some ways, Javanese inscriptions hew closer to Śāstric models: jayapattras and śuddhapattras are more common there than in Indian epigraphy, which may be the result of self-consciously adopting the model “as a package” rather than gradually and piecemeal. On the other hand, the introduction of written documentation, which was concomitant with the introduction of foreign legal concepts in both Tamil Nadu and South-East Asia, provided a mechanism by which local customary norms were “recognized” as legal rules (or deliberately rejected, as in Exhibit F) within a broader, transregional framework. In Java (as in Burma and Thailand) — though not (it seems) in Cambodia and rarely in India77 — this further led to the creation of Dharmaśāstra-like codes in local languages. It may be that there was not as much of a “market” for such vernacular codes in India because the audience of the codes were mainly brahmins, who were moreover diffused throughout the countryside over many centuries. Sanskrit-literate brahmins could thus purvey their conceptual wares in the local language. In South-East Asia, by contrast, the imported Indic legal learning had a smaller “footprint” in the social matrix, its cultural content would have looked more alien to local eyes, and the social groups supporting it more cut off from brahmin intellectual life in India. This may have made it seem more imperative to recast the genre to reflect local ācāra; it may even suggest a closer relationship to legal practice.

66Much more study will be needed to test these hypotheses, and to flesh out the relationship between terminology and practice in each of these regions.78