1 K. 598, Stanza 14. The K number refers to a Khmer inscription listed in Online Appendix A. Where there is a letter after a K number, it refers to an inscription face. Key dates and events are set out as a timeline in Online Appendix B. All dates are CE. Where the date of an event differs from that of the inscription (e.g., land bought 50 years before a text was written), the earlier date is cited and used in the analysis. Undated texts are commonly ascribed to a century, but in this study, we have often refined the range, e.g., where a reigning king or his posthumous name is mentioned. The numerical analyses often require texts to be assigned to a period such as a quarter of a century. Where the date is uncertain, its estimated range is allocated proportionately, e.g., if a text spans 100 years, a quarter of a text is assigned to each 25-year period. The study uses over 500 inscriptions written in Old Khmer and Sanskrit, dated from the beginning of the Angkorian period up to the reign of Jayavarman VII (1181). Where inscriptions consist of two or more distinct parts, having different authors, dates or subjects, they are taken as discrete texts.

2 Lustig, E.J., ‘Money doesn't make the world go round: Angkor's non-monetisation’, in Economic development, integration, and morality in Asia and the Americas, ed. Wood, Donald C. (Bingley: Emerald, 2009), pp. 184, 186Google Scholar.

3 Sahai, Sachchidanand, ‘Central administration in ancient Cambodia’, South East Asian Review 3, 1 (1978): 18Google Scholar; Wolters, Oliver W., History, culture, and region in Southeast Asian perspectives (Singapore: ISEAS, 1982), p. 19Google Scholar.

4 Vickery, Michael, Society, economics, and politics in pre-Angkor Cambodia: The 7th–8th centuries (Tokyo: Centre for East Asian Cultural Studies for UNESCO, Toyo Bunko, 1998), pp. 299, 309Google Scholar.

5 Ibid., p. 271.

6 Chakravarti, Adhir, The Sdok Kak Thom Inscription, Part II: Text, translation and commentary (Calcutta: Sankrit College, 1980), pp. 54–7Google Scholar; Daguan, Zhou, Customs of Cambodia, trans. Uk, Solang and Uk, Beling (Phnom Penh: DatASIA, 2016[1297]), p. 32Google Scholar; Sahai, Central administration, pp. 18, 26–30.

7 Mabbett, Ian W., ‘Kingship in Angkor’, Journal of the Siam Society 66, 2 (1978): 30–34Google Scholar.

8 Vāp: the title for a middle-ranking male, possibly stemming from a Mon-Khmer word for father. Loñ: the title for a middle-ranking male and counterpart of teṅ. It may be related to the Pre-Angkorian kloñ (chief). Teṅ: a title for women, often the counterpart of loñ, is derived from the Pre-Angkorian tāṅ, of similar meaning. Chloñ: a male title found in only 15 inscriptions. For vāp, loñ and teṅ, see also Vickery, Society, economics, and politics in pre-Angkor Cambodia, pp. 406–8, and Lewitz, Saveros, ‘Note on words for male and female in Old Khmer and Modern Khmer’, in Austroasiatic Studies, ed. Jenner, Philip (Honolulu: University of Hawai‘i Press, 1976), pp. 763, 768Google Scholar. For chloñ, see Vickery, ibid., p. 365.

9 The Pre-Angkorian mratāṅ/ mratāñ carried over into the Angkorian period and increased in frequency as mrateṅ/ mrateñ (khloñ) under Sūryavarman I (1001–1049). Vraḥ kamratāṅ ‘añ (VKA), which Vickery points out, pertained originally to gods and was then adopted by kings in the Pre-Angkorian period (Society, economics, and politics in pre-Angkor Cambodia, pp. 207–8, 406), was given increasingly to high officials in the Angkorian era. In Vickery's A history of Cambodia: Summary of lectures given at the Faculty of Archaeology, Royal University of Fine Arts, 2001–2002, Vidya Series (Phnom Penh: Pre-Angkor Studies Society, 2002), p. 99Google Scholar, he also points to a similar process under Udayādityavarman II (1050–1066), where dhūli jeṅ vraḥ kamrateṅ ‘añ, given to rulers from the 8th century, came to be held by some high officials.

10 Chloñ, comprising 2.5 per cent of officials, will not be analysed separately. They are mentioned up to 1049 as landholders, founders and in elevated temple roles. The title appears to have ceased being used, rather than its status having declined as did that of the loñ.

11 Vickery, A history of Cambodia, p. 82.

12 Vickery, Society, economics, and politics in pre-Angkor Cambodia, p. 406.

13 In the late 13th century, the visiting Chinese emissary Zhou Daguan (Customs, p. 118) reported that every village had a Buddhist temple or pagoda.

14 Wheatley, Paul, ‘Satyānṛta in Suvarṇadvīpa: From reciprocity to redistribution in ancient Southeast Asia’, in Ancient civilization and trade, ed. Sabloff, Jeremy A. and Lamberg-Karlovsky, C.C. (Albuquerque: University of New Mexico Press, 1975), p. 252Google Scholar.

15 Some texts are not obviously associated with foundations, but given that the inscriptions were found mainly in temple precincts, and that much of the land in question was given to foundations, they are likely to be foundation-related.

16 Jacques, Claude, ‘Khmer epigraphy’, Museum International 54, 1–2 (2002): 3CrossRefGoogle Scholar.

17 Aymonier, Étienne, Le Cambodge: Le groupe d'Angkor et l'histoire, vol. 3 (Paris: Ernest Leroux, 1904), pp. 540, 545–6Google Scholar; Ricklefs, Merle C., ‘Land and law in the epigraphy of tenth-century Cambodia’, Journal of the Asiatic Society 26, 3 (1967): 419Google Scholar (citing Cœdès, pers. comm.).

18 Mabbett, Ian W., ‘Varṇas in Angkor and the Indian caste system’, Journal of Asian Studies 36, 3 (1977): 431CrossRefGoogle Scholar; Vickery, Society, economics, and politics in pre-Angkor Cambodia, p. 406. For example, K. 1152A (962) concerns the appointment of a vāp as mūla of pages in the royal chamber of diversions and a sruk (village) he receives as a royal benefice. K. 831 (968) records the endowment of a foundation by three vāp in 924. A loñ in K. 215 (949) builds an āśrama on land he has bought and endows it with ricefields and slaves.

19 In K. 989: A20 (1008), loñ are said to be descended from Jayavarman IV; in K. 91: B1 (1080–1107), from Indravarman II (877–89); in K. 956/2: 9–10 (910–25), loñ and teṅ are descended from Jayavarman II.

20 See K. 165N (952); K. 158 (1003); K. 693 (1003).

21 See K. 956/2 (post 922); K. 91 (1080–1107); K. 989 (1008). Some ancestors in K. 91 and K. 989 are also titled chloñ.

22 In K. 1152A (979) a mrateñ is said to be the nephew of a vāp; in K. 1229: A47 (979) a vāp is the son of a mratāñ khloñ, also titled kaṃsteṅ ‘añ; in K. 958: N30–33 (947) a teṅ and 2 vāp were the children of a teṅ. Vāp are also seen in families with chloñ in K. 572: 11 (878–977), K. 238: B12 [949], and K. 1152: A12 [962]).

23 Vickery, Society, economics, and politics in pre-Angkor Cambodia, p. 270. See also Barth, A. and Bergaigne, M.A., Inscriptions sanscrites de Cāmpa et du Cambodge, 4 vols. (Paris: Notices et extraits des manuscrits de la Bibliotheque nationale, 1885–1893), vol. 27, pt. 1, pp. 124–5Google Scholar; Cœdès, George, The Indianized states of Southeast Asia, trans. Cowing, Susan Brown (Honolulu: University of Hawai‘i Press, 1968[1964]), p. 291Google Scholar.

24 The Sanskrit term varṇa has meanings including ‘class’, ‘group’ and ‘caste’. These were elite corporations appointed and given property by kings, reportedly from the time of Jayavarman II (K. 989: 9 [1008]) and may once have been associated with ceremony, the palace or certain localities. The corpus of Khmer inscriptions names at least 19 varṇa. The Sanskrit term varga has similar connotations of ‘division’ or ‘group’. It is first used in this sense in a Pre-Angkorian text, K. 1241: 7 (776). Varga appear to have been of lower status than varṇa, though the terms are sometimes interchangeable. In K. 1229: C4-21 (977), many land vendors of the Dhruvapura varga, mostly vāp, are sorted according to their courtly ‘function’ (varṇa). In K. 235: D14, Vijayapattana is a varṇa, but in K. 233: A5 (968–1001), it is a varga. See also Mabbett, ‘Varṇas in Angkor and the Indian caste system’, and Chakravarti, Adhir, ‘Caste system in ancient Cambodia’, Journal of Ancient Indian History 6, 1–2 (1972–3): 143–58Google Scholar.

25 K. 255: 16 (978); K. 128: 7 (1204).

26 The incomplete K. 61 (912) suggests that six women titled teṅ were given or offered themselves to serve a god. In 1029, King Sūryavarman gifted a chloñ ‘So and a teṅ Hyaṅ to the temple at Phnom Chisor (K. 31: 2–6). In 1096, Jayavarman VI (1080–1107) made an offering of a single teṅ to a god (K. 814: C3–6).

27 K. 814C (1096); K. 852 (1107); K. 32 (1116); K.383 (1121); K. 194 (1119); K. 34 (1113–49); K. 1036 (1113–1149); K. 366 (1139); K. 200 (1145); K. 850 (1155?).

28 With si and tai: K. 366: A22–b21; K. 383; with gho: K. 200: from A12 (1145). K. 938 (uncertain date).

29 K. 852: 6–10. However, in a contemporaneous text, K. 258, several teṅ tvan are vendors of land.

30 K. 258 (post 1107, but land purchases are ca. 1094); K. 249 (1109); K. 397E (1112); K. 523BD (1118); K. 254BCd (1129).

31 K. 207: 24–7 (1042) illustrates this. The differential status was first noted by Étienne Aymonier, Le groupe, p. 545.

32 However, a few women with the more commonly male titles of steṅ/ steñ and kaṃsteṅ, may have been counted as males.

33 These include all ācārya,’nak sañjak, most individuals transmitting royal orders, chaplains, treasury officials and functionaries in the royal courts.

34 Not necessarily in the army. See K. 91: B17 (1080–1107) and K. 397: E2 (1109).

35 Vickery, Michael, ‘The reign of Sūryavarman I and royal factionalism at Angkor’, Journal of Southeast Asian Studies 16, 2 (1985): 232CrossRefGoogle Scholar.

36 Lustig, Eileen J., ‘Using inscription data to investigate power in Angkor's empire’, Aséanie 27 (June 2011): 48–51Google Scholar.

37 Michael Vickery, in ‘The reign of Sūryavarman I’ (pp. 232–6, 243) highlighted texts with spurious claims, such as proximity of ancestors to an early ruler, albeit written mainly by rival families more powerful than those of the vāp and the loñ. One might query whether other aspects of these texts might also be open to scrutiny. We would argue that while some details may be inaccurate, the fact that the same kinds of stories are being repeated in so many of the texts strongly suggests that the concepts were not only feasible, but probable in the society. Thus, land buyers and sellers, disputants, witnesses, court officials and village elders, most probably held titles and fulfilled roles as are seen in the texts.

38 Vickery, Society, economics, and politics in pre-Angkor Cambodia, pp. 279–81, 295.

39 See for example, K. 449: XVI–VII (1069). Sedov, Leonid, in ‘Angkor: Society and state’, in The early state, ed. Claessen, Henri J.M. and Skalnik, Peter (The Hague: Mouton, 1978), pp. 114–7Google Scholar, has hypothesised that Jayavarman II began to integrate the population into a unified political system by incorporating ruling elites and clans (kula) into varṇa and temple communities. As the territorial units of the clans became state administrative divisions, some of these were subdivided into branches and reorganised.

40 Ricklefs, Land and law, p. 419.

41 However, there are far fewer contemporaneous than historical epigraphic reports of land gained under karuṇāprasāda, raising the question of their veracity.

42 Land without inheritors: K. 208: 45–47 (1050–66); K. 219: 6–13 (1050). Unoccupied land: K. 566: B1–3 (978–1077); K. 598: B6–9 (1006); K. 697: B3–16 (878–977).

43 K. 598B (1006).

44 K. 444: B27; K. 868: A34; K. 175: S12 (974).

45 Mabbett, Ian W., ‘Some remarks on the present state of knowledge about slavery in Angkor’, in Slavery, bondage and dependency in Southeast Asia, ed. Reid, Anthony (St. Lucia: University of Queensland Press, 1983), pp. 45–6Google Scholar.

46 K. 420: 15–9 (1001–49).

47 Cœdès, Georges, Inscriptions du Cambodge: Collection de textes et documents sur l'Indochine, ed. de Boccard, E., 8 vols. (Paris: École française d'Extrême-Orient, 1937–66) (1952), vol. IV, p. 164Google Scholar.

48 K. 842 (968).

49 K. 353: S17–24 (944–1001).

50 K. 598: B31–2 (1006).

51 See for example, K. 212A (1027); K. 217 (1026); K. 618 (1026?), offered and then ‘returned’ to the founder as karuṇāprasāda; K. 850 (1107–13); K. 1152 (962).

52 Hall, Kenneth R., Maritime trade and state development in early Southeast Asia (Honolulu: University of Hawai‘i Press, 1985), pp. 158–9Google Scholar.

53 For example, K. 1141: B21–4 (972); K. 348: 30–2 (954).

54 For example, K. 572: 8–12 (1011); K. 245: 29–30 (1001–49).

55 K. 205 (936).

56 K. 216 (1005).

57 The jurisdictions of the officials are not always clear. Some had up to six administrative duties affecting landowners and religious establishments. These included transmitting royal ordinances, arranging for land to be measured and marked out, implementing court rulings, witnessing transactions and ascertaining facts for court cases.

58 Sahai, Sachchidanand, ‘Fiscal administration in ancient Cambodia’, South East Asian Review 1, 2 (1977b): 133–4Google Scholar. Inscription K. 957: 16–8 (941) is illustrative: ‘[The slaves and livestock] are not subject to the authority of the viṣaya chief, rice chief, oil chief, [or] the head of the gāp jnval. The land under the authority of K.A. jagat Liṅgapura is not to be subject to taxes.’ (Translated from Cœdès, Inscriptions, vol. VII, p. 139.)

59 In K. 391: W37–8 (1082), granted by Jayavarman VI.

60 Where several tracts were bought by a single official, this buyer is counted for each transaction.

61 K. 258. This text is also unusual in detailing the sale of many tracts of land, with many vendors involved, some in the sale of more than one tract.

62 mratāñ (chloñ/ khloñ); steṅ (‘añ); (vraḥ) kaṃsteṅ (‘añ); kamrateṅ (‘añ).

63 The latest certain date for land purchased by a vāp is 974 (K. 343: S1), although a vāp still inherited land in 1011 (K. 569: 9). The latest mention of land purchased by a loñ is ca. 978 (K. 933: B4–6), but another inherited land as late as 1094 (K. 260: S12).

64 The latest sale by a vāp is in 1042 (K. 207) and by a loñ is in 1084 (K. 258). Sales by khloñ are in the period 1067–96 (K. 258).

65 K. 239: S24 (966).

66 K. 1198: B5–6 (1014).

67 Chakravarti, Caste system, p. 147, but see Mabbett, ‘Varṇas in Angkor and the Indian caste system’, p. 434. Something akin to this is seen in the Philippines, Cambodia, Nepal and India, today, where group members, mostly households, may hold ‘permanent or temporary rights to particular resource niches within the common property’. See Kirsten Ewers Andersen, ‘Communal tenure and the governance of common property resources in Asia: Lessons from experiences in selected countries’ (Rome: Food and Agriculture Organisation of the United Nations, 2011), http://www.fao.org/docrep/014/am658e/am658e00.pdf (last accessed 18 Sept. 2017), p. 8.

68 K. 258 (post 1107).

69 K. 207: 7–29 (1042).

70 K. 221: S7–11 (1011).

71 Dominique Soutif has translated K. 1218 (1007–8) in his ‘Organisation religieuse et profane du temple khmer du VIIe au XIIIe siècle’ (PhD diss., Université Sorbonne Nouvelle, Paris 3, 2009), p. 593. This is an inscription on a bronze vase of about 9 kg, inscribed with its weight, 3 tula 16 kattitki. The Sanskrit kattiki is equivalent to 1 liṅ, making the value of the liṅ 5.9 g.

72 K. 348 (954); K. 165 (957); K. 181 (962); K. 674 (966); K. 885 (968); K. 425 (968); K. 353S (878–977); K. 257S (979); K. 262S (983) [2 cases]; K. 263B (984); K. 344 (985); K. 1116 (992); K. 257N (994); K. 233 (968–1001); K. 158 (1003); K. 598 (1006); K. 720 (1006); K. 1198B (1014); K. 843B (1025) [2 cases]; K. 373 (11th century); K. 588 (11th century); K. 1238 (1036) [case in 1003]; K. 1074 (1078–1277).

73 K. 1238: A39 (1036); K. 885: 6, 10 (968), in which a loñ twice withdrew his objections to the placement of boundary markers by a VKA.

74 The total number of disputes (25) is small, putting these curves close to the limits of statistical validity. Nevertheless, the fact that the percentage of disputes increases and decreases sharply, while the number of texts is reasonably constant, implies significant social and economic pressures at that time.

75 Vickery assigns the land disputes more generally to the first part of the 11th century in ‘The reign of Sūryavarman I’, p. 232.

76 There are at least two accounts of disputes in which witnesses had been paid by the ultimate winner to corroborate evidence. See K. 257: N11–5 (994); K. 1198: B38 (1009–15).

77 We might infer too that one benefit for loñ who had been promoted was that they could purchase land with fewer restrictions.

78 Zhou Daguan noted in his Customs, p. 29, that there were strict rules about the type of floral design permitted to be worn, based on the status of the wearer.

79 K. 374: 12–23 (1042).

80 We see another example of a premium required for an exchange in the Ta Prohm inscription, K. 273: LII (1186), where it is specified that if unhulled rice is supplied, the quantity needs to be four times that required for hulled rice.

81 Jenner, Philip N., A dictionary of Angkorian Khmer, ed. Cooper, Doug (Canberra: Pacific Linguistics, Research School of Pacific and Asian Studies, Australian National University, 2009)Google Scholar.

82 Once money was introduced into Khmer society, around the 16th century, the need for such conversions would have fallen away. The idea of paying double could have remained.

83 Monier-Williams, , Cologne digital Sanskrit lexicon (Cologne: Institute of Indology and Tamil Studies, University of Cologne, 2005)Google Scholar, http://webapps.uni-koeln.de/tamil/; http://www.sanskrit-lexicon.uni-koeln.de/monier/indexcaller.php (last accessed 25 Nov. 2014).

84 K. 1198: B28 (1014).

85 K. 353N (1046).

86 K. 233: A15 (968–1001); K. 165: N39 (952); K. 105: 19 (968); K. 257: N4 (994); K. 153: 8, 11 (1001); K. 1198: B28 (1014); K. 353: N6 (1046); K. 420: 5, 7, 12, 19 (1001–1049): K. 1238: A11 (1036).

87 In K. 105:19 (968) the pul was to acquire some buffaloes to exchange for laterite to build a sanctuary and in K. 257: N4 (994) it was to purchase goods to buy a mandira.

88 Vickery, in his chapter ‘Some remarks on early state formation in Cambodia’, in Southeast Asia in the 9th to 14th Centuries, ed. Hall, D.G. and Milner, A.C (Singapore: ISEAS, 1986), pp. 102–9Google Scholar, draws upon anthropological studies to propose a ‘conical clan’ structure in which people and gods are ranked in a hierarchy according to their proximity in descent from a common ancestor, in this case Indravarman II, rather than under the system of primogeniture, where the throne passes from father to son. This line of descent from Indravarman is mentioned in K. 253: B III (1005); K. 125: A10 (1007); K. 380: W17 (1037); and somewhat ambiguously in K. 136AB (1066–80).

89 K. 420: 32–4 (1001–49) records the gift to a foundation of settlements confiscated by the king from two individuals who had risen up against him.

90 This may have been the rationale for his Oath of Allegiance, K. 292 (1011).

91 Briggs, Lawrence Palmer, The ancient Khmer empire (Bangkok: White Lotus, 1999[1951]), pp. 134, 140, 145Google Scholar. These roles, religious in origin, had become largely secular in India. See Vickery, ‘The reign of Sūryavarman I’, pp. 229–30.

92 Briggs, Lawrence Palmer, ‘The genealogy and successors or Śivāchārya: Suppression of the great sacerdotal families by Sūryavarman I’, Bulletin de l’École française d'Extrême-Orient 46, I (1946): 184Google Scholar.

93 Genealogies commencing with Jayavarman II: K. 989 (1008); K. 235 (1052); K. 449 (1069), genealogy from Jayavarman II to Harṣavarman III); K. 91 (1080–1107), genealogy likely from Jayavarman II to Jayavarman VI. Land or careers starting with Jayavarman II: K. 253B; K. 278 (1007); K. 92 (1028); K. 1238 (1036); K. 661 (1050–1066); K. 289 (1066); K. 275; K. 1036 (1113–1149); K. 834.

94 Mabbett, ‘Some remarks on the present state of knowledge about slavery in Angkor’, p. 53, finds no direct Angkorian evidence for debt slavery in temples. Debt servitude existed in 19th century Cambodia. See Aymonier, Étienne, Le Cambodge: Le royaume actuel, vol. 1 (Paris: Ernest Leroux, 1900), pp. 98–9Google Scholar.

95 E. Lustig had previously suggested in ‘Using inscription data to investigate power in Angkor's Empire’ (p. 57), that this process ceased some time after Sūryavarman I (1002–1050).

96 For example, in Bagan, the temples and clergy became extremely wealthy and powerful, largely through fiscal immunities. Periodically, their lands and property were confiscated by the rulers in an exercise described as ‘purification’ of the saṅgha. See Thwin, Michael Aung, Pagan: The origins of modern Burma (Honolulu: University of Hawai‘i Press, 1985), pp. 169–209Google Scholar. In Java, the sīma grants to religious communities were curtailed in the 11th and 12th centuries and replaced by titles and other status symbols, as rulers attempted to gain more control over their income. See Jan Wisseman Christie, ‘Javanese markets and the Asian sea trade boom of the tenth to the thirteenth centuries A.D.’, Journal of the Economic and Social History of the Orient 40, 4 (1998): 354Google Scholar.

97 K. 258ABC (1107–13); K. 850: 3 (1150–6?); K. 397: E3–4 (1108); and K. 383: B15 (1121).