Stela found during excavation in Oc Eo archaeological site (An Giang Province, Vietnam) (K. 1426), 7th century CE

The stela appears to be of sandstone and measures 50 × 28 × 5 cm. Its upper silhouette has the shape of a downturned curly brace. The stela has no tenon at its base, but the bottom 15 cm were left blank, as though this part was originally intended to be inserted into the ground or into a stone mortise. When discovered, it lay on the ground, somehow broken into two fragments on one face and in four fragments on another. The pre-modern breakage has not been repaired yet. The stela is now kept in the storage room of Óc Eo Museum.At the time we finish this article, none of us has yet had the chance to observe the stone directly. We rely on information shared by Nguyễn Khánh Trung Kiên 2019 and Đỗ Trường Giang 2023

The stela is inscribed on its two broad faces, face A in Sanskrit verse and face B in Old Khmer prose. Both faces are decorated with a blossoming lotus above one or two strings of pearls two on A, only one on B at the top of fourteen lines of text and one string of pearls below the text on each face. The Khmer text runs continuously in one column whereas the Sanskrit one is arranged in two columns, even-numbered verse quarters pādas in one and odd-numbered ones in the other. Both texts are written in an early form of Khmer script typical of the second half of the seventh century, without long descenders on the characters ka and ra. Since no other king going by the name Jayavarman is known to have ruled in the seventh century, it is clear that the king Jayavarman mentioned in the Sanskrit text must have been Jayavarman I 657–681 CE. Besides its paleographic aspect, the inscription shares two main features of its content with other inscriptions of the same king: the opening expression ājñā vraḥ kamratāṅ añ ni “order of My Lord” with addition of the particle ni where the inscriptions belonging to the reigns before and after use the phrase ājñā vraḥ kamratāṅ añ without said particle and the details of the threat of punishment at the end of the Khmer portion.For discussion of the characteristics of the inscriptions of King Jayavarman I, see Vickery (26, 165–168), and for his dates see Goodall and Revire (27121).

The character ra consists of a double stroke and does not extend farther down than the body of other characters. While the subscript of the character ṇa in the word pūrṇṇamī in the Khmer text and pūrṇṇodita- in the Sanskrit one still preserves the ‘‘x’’ shape characteristic of the fifth and sixth centuries, the character ṇa above it has the two sides of the ‘x’ split and connected by a stroke at the bottom. The decoration of the stela and the paleographic characteristics of the script present similarities with several other inscriptions of Jayavarman I. The stela from Phum Chrei K. 563, Kompong Speu, 7th century CE, for example, also has a decorative lotus at the top and a string of pearls below the text. The character ka does not have a long descendant loop, except the ones in the lines 9 and 12 of the Sanskrit text. The dominance of the ka without the descendant loop can also be found in the doorjamb from Tuol Kuk Preah K. 493, Prey Veng, 657 CE; by contrast, the stela from Preah Kuhea Luong K. 44, Kampot, 674 CE presents more characters ka with the descendant loop than the other type, whereas the two types are equally represented in K. 563. The shapes of the vowel-markers markers i and ī. The first one is written as a round small circle and the latter is slightly larger with a horizontal stroke in the middle. In some cases, the vowel marker ī appears like a spiral turning to the left, which is comparable to that in the inscription K. 493.

As Diwakar Acharya pointed out to us personal communication of 28.vi.2023, vineya could refer also to the king’s subjects for cf. vinayādhāna in Raghuvaṁśa 1.24. This means that a second punning interpretation is at least suggested, even if it is not perhaps meant to expand fully in the mind of the reader, in which the king is the subject of the stanza, his name, Jayavarman, being already adumbrated or suggested by the opening word: He is victorious, that Victor jinaḥ, a veritable sun for destroying the darknesses of all the world, whose rise spreads out to awaken the lotuses that are his subjects who are to be disciplined. For a handful of comparable instances of pre-Angkorian Śaiva inscriptions that open with a verse that is manifestly in praise of Śiva but that allow themselves also to be interpreted as referring to the ruling king, see Goodall and Revire (270). This is a trope we find elsewhere, for instance in the 5th-century Guḍnāpur pillar inscription from Karnataka India, commemorating the creation and endowment of a temple of Kāmadeva. Its first verse, as Cecil and Gomes suggest (1616), could describe both Kāma and the ruling king Ravivarman.

The use of kaṇṭaka literally “thorn” to refer metaphorically to enemies or trouble-makers within the realm is an old cliché see for example Raghuvaṁśa 14.73. One is thus meant to understand kaṇṭakabhayam simultaneously as “fear of thorns” and “danger of trouble-makers”. Although the “burning of enemies” might in such a Buddhist context allude to the destruction of Māra, it seems more likely to us that this stanza speaks about the king. The correlative yasya shows that this stanza is to go with the next two stanzas which contain yas and then tena, each time denoting the king. The vocabulary tejas, prajāḥ, arivaṁśe also suggests that this stanza is only about the king.

One could imagine a layer of meaning in which he “was straight as the rod of a pair of scales because of the fact that he was engaged in cultivating a meditative state of awareness that all things are of equal value”. Such a yogic awareness is spoken of as sāmarasya in Śaiva traditions, or as śaktisamarasa e.g. in 5:36 of the Uttarasūtra of the Niśvāsatattvasaṃhitā and in 4:61 of the Nayasūtra, both edited in Goodall et al. ().

As Diwakar Acharya observed to us personal communication of 28.vi.2023, the combination of a genitive and a dative candanavihārasya arthāya would be more normal for expressing this sense. Nonetheless, the first dative could be the result of attraction to the case of the second. Or we can interpret as we have, with two parallel purposive datives: “for wealth” and, at the same time, “for the vihāra”. He further suggested that arthāya might mean simply “for the purpose / sake of”. So “he gave this charter for the sake of the Śrī Candana Monastery”. Note that it is not said that the king made any donation himself. Pre-Angkorian inscriptions that record expenditure by the ruling king seem to be rare Goodall (27–36. Here no benefactor is mentioned, which perhaps suggests that the king’s edict relates to the disbursement of wealth that is regularly donated in small amounts by many individuals.

We could have taken the expression mādhave mādhave distributively to mean in “every month of Vaiśākha”, but since the element niyata- already expresses fixedness and therefore calendrical regularity, there is no need for the repetition of mādhave to express this. Indeed this repetition seems elegant precisely because the word is used in two different senses. Admittedly one could argue that niyata- is instead used to qualify the festival as “restrained”, in other words a ceremony in which the participants were expected to restrain themselves from sensual enjoyment and fix their minds on the Buddha and his teachings. But this might seem an anachronistically post-protestant interpretation, since Hindu and Buddhist festival processions often allow a joyful inversion of norms. This is furthermore arguably suggested by the opening mādhave, “In Springtime !...”, a time for joyous festivals all over the world from time immemorial. We have silently assumed that pratimā-gamanam is intended as a metrically expedient alternative to some more standard expression for a procession, such as pratimāyātrā, although it is just conceivable that it refers to “the approaching of the Buddha image for veneration” pratimā-āgamanam. Still, this is less likely, since the image, once installed, was presumably always there to be approached, not just on some calendrical festival. Festivals archetypically involve processions in South India invariably so, to the point that utsava is often used to mean “procession”, e.g. in the expression utsavamūrti. For other allusions to religious processions in the Khmer epigraphical corpus, see K. 290 9th c., st. LXVI–LXVII, Buddhist; K. 853 9th c., st. XIV, Śaiva; K. 356 10th c., st. XI, a processional image, described as utsavayāyin, of a golden Viṣṇu mounted upon a silver Garuḍa; K. 276 11th c., lines 4–6, Śaiva; K. 277 11th c., lines 33–34, Śaiva; K. 1222 12th c., Southern face, Śaiva. See also the passages referred to by Bhattacharya in his brief discussion of the subject (89–91, to which we have alluded above. There are numerous other references of course to utsavas, such as a Śaiva one in K. 55, st. 14, which also belongs to the seventh century, but without explicit references to processions or processional images.

As for dāsa and karmakara which latter are presumably servants with a different legal status from the dāsas, perhaps because they are not as completely owned, the Śaiva author Brahmaśambhu draws this distinction in his Naimittikakriyānusandhāna of 938 CE. According to this text f. 73v, when the religious master ācārya transfers his responsibilities to his successor, he should say: idaṁ sthānam iyaṁ vr̥ttir iyaṁ pustakasaṁhatiḥ amī vai bharaṇīyās tu dāsāḥ karmakarāś ca naḥ etat sarvaṁ mayā tubhyaṁ dattam adya tvayāpi ca pālanīyaṁ tathā samyag yathāsmābhiḥ prapālitam “Such is the foundation; such is the revenue; such is the library; and these are the slaves and workers that are to be supported. All this I have given to you this day, and you should properly maintain it as I have done.”

The Governor mentioned here is presumably one of the successors to Vidyāviśeṣa, the governor of Tamandarapura who authored K. 604 and K. 1235 in 627 CE for editions and translations of which, see Goodall (). The fact that this figure is assigned the administrative burden of ensuring the correct use of these resources seems indeed to confirm that we may speak of him as a “governor” as argued by Goodall, 2019 passim. It was clear already that the various other mentions of Tamandarapura in the epigraphic corpus K. 9, K. 604 and K. 1235, when all taken together, suggested the location somewhere in South Vietnam. The discovery of K. 1426 at Oc Eo might indicate that Tamandarapura was in fact the ancient name of Óc Eo. As for the nature of the compound name Tamandarapura, which, apart from the element -pura, does not seem to be Sanskrit, see the summary of past discussions given by Goodall (29), quoting Vickery (182–183) and correspondence with Griffiths, and see Chhom and Griffiths forthcoming, p. 102, where we conclude: While pura means ‘city’ in Sanskrit, the element tamandara is hard to explain in Sanskrit or Khmer and may perhaps reflect an indigenous language of the Austronesian family: in Malay, for instance, taman darat could mean ‘flatland garden’, and it is likely that a very similar expression also existed in the ancient Cham language. So the name may be an indication that the ethnolinguistic profile of the ancient Mekong Delta, and the polity of Fu-nan, included other people beside Khmers.

Regarding the term saṁ paribhoga, the reproductions at our disposal make it imaginable that the dot we interpret as anusvāra is actually accidental, which would allow reading saparibhoga. Previous scholars have sometimes hesitated between saṁ paribhoga and saparibhoga see e.g. the note in Cœdès 1936 on the inscription K. 6. But it is probably undesirable to read saparibhoga, as the expression is attested multiple times with or without p- prefix as psaṁ paribhoga, often with clear presence of anusvāra see further occurrences in the inscriptions K. 51, K. 163, K. 426, K. 561, K. 582, K. 600, K. 818, K. 904, K. 926 and K. 1275.

Taṁrāṅ is a rare toponym. In a variant spelling of presumably trivial significance, it appears, as Taṁraṅ in the inscription of Ban Hin Khon Nakhon Ratchasima, Thailand, 7th c.. Is Taṁrāṅ the Khmer form of the Sanskrit Tamandarapura? To give an example of the sometimes rather substantial differences between presumably primary vernacular names and their guise in Sanskrit context, we may first refer to the epigraphy of Campā, where Panrāṅ is the vernacular toponym to which Pāṇḍaraṅga or Pāṇduraṅga corresponds in other contexts Griffiths and Southworth (285–291). We are also reminded of the possible correspondence between the Sanskrit name Kāmaraṅga for Arakan in present-day Burma and the Old Burmese toponyms Kaṁ Raṁ or Kamḥ Yaṁ (307). If Taṁraṁ does indeed form a pair with Tamandara, then the Malay or Cham explanation of the latter, proposed just above, may need to be reconsidered, or else it must be assumed that the vernacular Khmer term came into existence only after the Sanskritization of an originally Austronesian name had taken place.

Regarding ther term kloñ ñan, it might be read kloñña naṁ “superintendent of temple naṁ = vnaṁ” but this would yield a unique example of the word vnaṁ followed by the preposition ai “at”. Nevertheless, there are several examples in the pre-Angkorian inscriptions where vnaṁ stands in front of a name.

The word cnaṁ is clearly the Old Khmer word meaning “year”, as found for instance in the inscription K. 90 Kampong Cham, 6th c., lines N7–10 nivandha ge ta gui utsava ta pon hvat ta gui cnaṁ ta moyya “Provision for the people at the four festival occasions in one year”. The word cracar, by contrast, is not recorded in Pou’s Old Khmer dictionary () while in Jenner’s dictionary of pre-Angkorian Khmer (144), based on a single attestation in the inscription K. 1004 Kandal, 691 CE, it is analyzed as a derivation from the verb car “to write note, jot down in order”. However, in unpublished notes on that inscription, to which we have access, Jenner translates the word cracar tṅai “daily” literally cracar “every” and tṅai “day”. If our intuition is correct that the context in which the word cracar appears in the inscription K. 1426 is comparable to the one where cracar tṅai appears in K. 1004, our translation of cracar cnaṁ as “every year” stands to reason. We would then propose a different morphological analysis from Jenner’s: cracar could be derived from the otherwise unknown Old Khmer antecedent of the word cuor word that means “line, row” in modern Khmer. With the prefix of intensification cra-, the resulting meaning could then be “one after another in a line or a row”, somewhat like the words “règle” and “régulier” in French which can refer both to straight lines and regularitFor a discussion of the prefixes c- and cra- in Khmer, see Jenne and Pou (xxxvii–xxxviii).

The word cmap is formed with the infix -m- whose original function seems to have been to create agent nouns, so c-m-ap “seizer” from cap “to seize” (307). However, it needs to be recognised that the inherited affixation of Khmer had already become fossilized to a certain degree by the stage of Old Khmer.Cf. Jacob (608): “The fossilization may be described as follows. Although the function of many Mod.K. words containing an O.K. affix could in a given context be recognized by means of the affix, there are also many words of similar construction in the use of which the original function does not operate any more. In addition to this, other, newer affixes have confused the picture. The infix m occurring between two consonants, for example, can be nominalizing or causative, at least since the Mid.K. period.” In the context where cmap occurs in K. 1426, it is likely that the word functions as equivalent to the verb cap although the alternative interpretation as an agent noun cannot be entirely dismissed. The syntactic context is similar to that where c-m-er occurs, interchanging freely, it seems, with unaffixed cer “to transgress”, in the formulaic phrase of King Jayavarman I’s inscriptions: ge cer/cmer ājñā vraḥ kamratāṅ añ ge daṇḍa. This can be translated by “those who transgress the order of My Lord, are punished” if cmer is understood as a verb or “those who are transgressors of the order of My Lord, are punished” if it is taken as an agent noun.

Like the word cracar, cralak is formed with the prefix of intensification cra-, here attached to the verb -lak “to cut a notch, groove, make an incision; to hollow out; to trim, clip; to cut into; to scratch, score, groove, mark; to incise, engrave; to gouge, chisel”. The resulting meaning “to cut intensely” underlies our interpretation “to cause damage” in the context of the inscription (417). To date, the word cralak appears in three inscriptions, i.e. K. 502 Chanthaburi, 7th c.; this inscription is fragmentary, K. 1267 Kompong Speu, 7th c. and K. 1426. The context in which the term cralak appears in K. 1267 seems to confirm our interpretation. The sentence ge ta cralak gui paṅ tiṅ ti mās kaddi moy can thus be translated as “Persons who cause damage to the donation of the king shall pay a fine of one kaṭṭi of gold” (194).

Analogously to the word cmap above, the noun k-n-ar is derived from the verb kar “to obstruct” with another affix, -n-, whose original function was likewise to form agent nouns, but contextually more likely to have been intended as a verb.

The expression qnak naṁ pitai “naṁ pitai people” cannot be entirely understood: while the meaning of qnak “person” is certain, that of the word pitai remains obscure and the word naṁ might be a noun meaning “cake” or a verb meaning “to lead, bring, take”. According to Jenner (310), pitai is the name of an unidentified cake naṁ which was presumably offered to the dead. However, if one considers naṁ to be a verb, then pitai must be its object and it may no longer be assumed to be any kind of cake. The expression qnak naṁ pitai figures also in another inscription of King Jayavarman I, i.e. the stela of Preah Kuhea Luong K. 44. The text alludes to the presence of some naṁ pitai people within the precincts of a sanctuary. Similarly, these people in K. 1426 probably had the privilege of entering the precincts of the Lord.

The restitution of the world dap is based on both the meaning and the space on the stone. Given that the prohibitive word dap is followed by the word pre or tel pre in three previous phrases, the same wording is expected in the lost passage here. Nevertheless, the available space is too large for the word pre and too small for tel pre. The meaning of dap suits the context better than that of pre “to use, to order”.