¹ Figure 13 gives a listing of all Sukhothai inscriptions as of 1990 with their presumed dates, location, and citations. For a list of the abbreviations used in this article, see below p. 561. Other conventions are as follows:

// indicates phonemic transcription; [ ] indicates phonetic transcription; {} indicates morphophonemic transcription; in line citations from inscriptions # is used to indicate sentence boundaries as they are marked in the inscriptions; ** indicates forms neither documented nor attested nor grammatically acceptable.

The transliteration used here does not note the presence or absence of virama on a final consonant. However, double finals are transliterated as -CC. Short -a-, when marked by mai han akat, is transliterated as ă. Sd.6 [#9] uses subscripts when writing post-initials; these have then been transliterated as CC-. Ambiguity may arise in the transliteration of rhotacized clusters, whether complex initials in ETh. are to be interpreted phonologically as CərC- or CrəC-; the presence of OM loans such as /kərya/ ‘gear, regalia, etc. ‘or serdha/ ‘to have faith (v.), faith (n.)’ in ETh. would suggest the former. The vowel support in the context C_(C) is transliterated here as å. Subscript consonants in ETh. are transliterated (-) CC-. The vertical stroke ‘is transliterated here either as’ (minor syllable) or as’ (tone mark). - is transliterated as.

² In Yoneo, Ishii, Aewrisvongse, Nidhi, Osamu, Akagi, Wichienkhiew, Aroonrat, Endo, Noriko, A glossarial index of the Sukhothai inscriptions [hereafter cited as GISI], Bangkok: Amarin, 1989), 198Google Scholar, and in Griswold, A. B. and Nagara, Prasert ṇa ‘King Lödaiya of Sukhodaya and his contemporaries (Epigraphic and historical studies, 10)’, JSS, LX, 1, 1972, 21–152Google Scholar (7 plates).

³ This reading follows Griswold and Prasert's edition of the inscription. Transliteration [C.B.]: [60] # kū mī sardhā hnakk [61] hnā kū ciṅ dhåt tan ku ‘oy d ā n hay jihitt khāt wā cakk sadāmsā sanānay lankā dīp cakk faṅṅ gām bra? [62] pen cau duk ‘an lē # …

⁴ Griswold/Prasert (1972,131).

⁵ Griswold/Prasert (1972: p. 131, n. 164) state ‘[sadāṁ] is equivalent to [kraḥdāṁ]’

⁶ The following errors in GISI should be corrected: at Kb.l.A.55, Nw.2.A.15, Sd.lO.B.10, Sd.l 1.A.14 no occurrences are found; Sd.2.B.24 is based on an earlier reading. At Nw.2.B.9 dām is preceded by a lacuna, and may thus be read as [ka]dāṁ or [ka'jdāṁ. Sd.2.B.56 has no occurrence of gawar. For cakk 106 (Sd.18)-1–17 read 106 [Sd.18]-l-7. GISI readings are based variously on Cœdès, CI and ISP; in one instance Cœdès's reading of Sd.ll.A. 14 as kådām kāmben dăṅ sī tān has been amended in ISP (p. 213) to kå kāmbeṅ dăn sī tān. A gloss for kå is not provided. Not having examined the inscription I have omitted this form from figs. 1 and 2.

⁷ In epigraphic Thai the diacritic ‘indicates not only tone B in non-stopped syllables but also minor syllable vocalism, possible /#/or/o/</o:/, and may be related to stress-assignment. This convention is prevalent throughout Sukhothai epigraphy. Exceptional in this respect is LB.38 (#62), of Wat Phra Yuen (1369). where the symbol’ marks unstressed initial syllables, including Indo-Aryan loans; initial vowels a-,ă -,ā - i,-ī- and medial -au- (of the first syllable) are reduced here to the initial vowel support ‘a- marked by the symbol’. Griswold and Prasert have called the mark’ on one occasion ‘separation marker’ (Griswold, A. B. and Nagara, Prasert na ‘The inscription of Wat Pra Yün [#62, Lb.38] (Epigraphic and historical studies No. 13)’, JSS LXII, 1, 1974, 123–41Google Scholar, no doubt thinking of clitics: ‘In some cases it [sc.‘] is a tone marker; in others it appears to be a separation marker, as it sometimes is in Sukhodayan and Ayudhyan inscriptions (p. 125)’). But this is inaccurate insofar as the first syllable in some disyllabic words may also be marked by’. Prosodic markings in Thai epigraphy are still little understood, although my own preliminary analysis suggests that these are context-sensitive, that is, tone marking interacts with stress; in Sd.2, for instance, the word for ‘cloth, fabric, textile’, mod./phâ:/ has the tone marker - at A.89 in 4 cases out of 5. In the introduction to Sd. 2. Griswold/Prasert state ‘diacritical marks’ and + are used in this one [sc. inscription] in a very haphazard manner, frequently being omitted where we should think them necessary, sometimes being added where they are not needed; and the same word is often written indifferently with or without them’ (1972: 84). However, it has gone unnoticed that in the first case where ETh. Phā is unmarked, at Sd.2.A.89, it is part of a nominal compound reīan phā ‘tent’ whereas in the following passage phā marked by * is a nominal head followed by an attributive verb phā’ ten² hleian phā² tām phā² (kh)yaw phā² khā(w). Adducing forms attested in other Sukhothai inscriptions it is found that this variation is systematic, and the notations therefore consistent: forms marked by are found in Ay.l.C.2, Sd,18.B.17, 20, unmarked forms in Sd.6.(A)30, Sd.ll.A. 16, Sd.10.B.13x, and Sd.18.B.39; Sd.12.D.10, 11 has junctural forms ph-, followed there by attributive verbs. Unmarked phā in Sd.6 and Sd. 11 are part of blendforms consisting of the nominal head pha and an IA loan, Sd.6 (A) 30 phā sănghāthi, Sd.11.A.16 phā peṅcatī. In Sd.18.B.39 one finds unmarked phā clause-initially: # phā nap brah råy kau1 sip # ‘cloths for saluting the Buddha, a hundred and ninety’ (Griswold, A. B. and Nagara, Prasert na ‘The inscription of Vat Jān Lom (A.D. 1384): Epigraphic and historical studies, No. 8’, Journal of the Siam Society, LIX, 1, 1971, 189–208; here p. 206)Google Scholar; lack of stress may account for its unmarked occurrence in Sd.10.B.13. The same rules apply to ETh. ṭan² ‘tree, trunk, beginning’; Sd.2.B.7 ṭay ṭan brah srīmahābaudhi’ below the sacred Bo tree’ [unmarked—unstressed] and Sd.B.l 1 pluk ṭåk māy ṭan² hñay’… to plant flowers and large trees’ [marked—stressed].

⁸ For details see my ‘Old Mon s-’, in J.H.C.S., Davidson, (ed.), Austroasiatic languages: essays in honour of H. L. Shorto (London: School of Oriental and African Studies, 1990), 241–9;Google Scholar‘Addenda, Old Mon s-’, Mon-Khmer Studies, XVIII-XIX, 1991, 250–3Google Scholar, and ‘Notes on Mon epig-raphy II’, JSS, LXXIX, 2, 1991 [1992], 61–79Google Scholar where also illustrations of the Jakata plaques, referred to in the Shorto Festschrift, appear.

⁹ For ka, ka’, kå the following corrections to GISI should be made: Sd.2.A.99 old reading has (ka'dāṁ), new reading has ka’, in Sd.2.B. 15 read kadāṁ. In Sd.2.B.27 it occurs only once, not twice. In Nw.2.B.23 and 37.21 there are no occurrences. Sd.ll.A.14 old reading kådāṁ, new reading kå kāṁbeṅ.

¹⁰ Shorto's reconstruction of -d- in this word as an imploded stop d- rests on the basis of its modern reflex SM /kānom/ (< OM/kəndom/), which would otherwise have resulted in a form like SM **kəlom, with OM medioclusters /-nd-/ being simplified to MM -I -, and its MM variant Frequent kanḍṁ. See DMI, q.v., and his ‘The register distinctions in Mon-Khmer languages’, Wissenschaftliche Zeitschrift der Karl-Marx Universität [Leipzig], Gesellschafts- und sprachwissenschaftliche Reihe, xvi, 1–2, 1967, 245–8Google Scholar. Notice that /kənom/ itself is an irregular form; the expected SM reflex of the OM form ought to be **həncm. If OM -d- in this instance were an imploded stop f then one would expect OM *kərdom to correspond to SM /hədom/ and not the actual SM Variation of the type -nd ∼ -nḍ- is attested elsewhere in the OM corpus, such as pindoṅ ∼ pinḍoṅ ‘to send, escort’, dindiṅ ∼ dinḍiṅ ‘to carry [litter]’; however, kindok ∼ kinḍok ‘spirit, ghost’ deviates from this pattern as its modern reflex /kəlok/ suggests the OM mediocluster OM -/nd-/, and not OM /-nd-/. To reconcile the variant vocalism of ETh. and its modern reflex -am with Shorto's reconstruction of orthographic -am rhymes as OM *-om is problematic, unless we assume the borrowing to antedate the OM shift *-am > *-am. An EOM reconstruction of Classical OM *om as *-am is not to be excluded.

¹¹ busac kyāk ‘to dedicate [Buddha] image’, kīr kuṁbār ‘to dig tank’, til ‘ārām ‘to plant grove’; pa pun dān ‘to perform acts of merit and charity’. Note that by the eleventh-twelfth century in central Burma kandam puñ (Lp. 1.B.3) was replaced by pa puñ

¹² Shorto does not derive OM sinraṅṅ, sanraṅ, sunraṅ ‘act, work, on behalf or in name of others,…‘ from OM sraṅ, presumably because of the latter's modern reflex’ to cast, coin ‘and because of occurrences such as OM pa sinraṅ (Ku.218) ‘to work for others’; however, cases such as sanraṅ pun ‘works of merit’ (I.A.49) raise doubts, sinraṅ, sunraṅ, sanraṅ may be the nominalized form -n- of sraṅ

¹³ SM /saη/ (< MM cruiṅ, LMM sruiṅ) applied to perishable materials such as wood

¹⁴ The vocabulary relating to the manufacturing of artefacts and of the setting up sīmā is quite precise and differentiated in epigraphic Mon; thus one finds at Thaton (Kawgun, eleventh century, DMI 6) pūc kyāk tmo’ ‘carved a Buddha [image] from stone’, in a variety of twelfth-century votive tablets lāt kyek ‘to mould …’ compared to the general expression kindam kyek thar moy ‘made a Buddha image’ (Myinkaba Kubyaukgyi 13). In MM sīmā may be kanarh (< OM kandaṁ) ‘to set up’, or saṁmut, samut, samit ‘to demarcate’; boundary markers nimit may be tuiw ‘to set up’ or daḅut; daḅut is also applied to relics dhāt. In SM /kanom/ collocates with /caik/ ‘to make a Buddha image’ and with /pra?koh/ ‘treatise, book’, ‘to copy [manuscript] as an act of merit’.

¹⁵ A reconstructed paradigm for *kdaṁ (base) would include *sikdaṁ (hypothetical s-), kindaṁ (frequentative -N-), skindaṁ (hypothetical s-, frequentative -N-), *kudaṁ (causative -ə-), *skudaṁ (hypothetical s-, causative -ə-), *kardarn (nominal -r-), *kandam (nominal -n-). Note orthographically contrastive pairs, which can be reconstructed internally based on other attested forms in the OM corpus, kindaṁ/*kandaṁ, *kdaṁ *[k^ədam]/*kudaṁ, -u- representing the vocalization of a former labial infix (Shorto in conversation June 1990).

¹⁶ Note that the spelling of other Mon loanwords in ETh. such as kwayan /kwian/’ cart’, also follows Khmer orthographic rules

¹⁷ Similarly, ETh. ṯāṁnakk (Sd.2.B.14), ṯāṁnak (Sd.lO.A.13, Sd.25.19, 20) corresponds to modern Khmer ṯaṁnāk’ ‘shelter, etc’.

¹⁸ Mod. Khm. khdam ‘hut, shelter’ is likely to be a borrowing from Mon.

¹⁹ This observation is based on the entire Sukhothai corpus and the north-eastern Thai corpus. In the epigraphic corpus from northeastern Thailand only căk and cah are attested: see Thawat Punnothok, Isan inscriptions, Bangkok, Ramkhamhaeng University/Toyota Foundation, I, 1986 (text), ii, 1988 (plates) [in Thai]. Inscriptions from northern Thailand have not yet been comprehensively examined.

²⁰ Lao /si?/, Luang Prabang dialect /?i?/, corresponds to Phuan, Neua and Tai Dam /ci?/, but the isoglosses suggest that the /si?/ form is confined to southern-most areas, an MK contact zone (information courtesy J. R. Chamberlain), implying to me, if not an MK borrowing, then at least paronymic attraction /ci?/ > /?si?/. Chamberlain suggests that, in fact, the form /ci?/ derives historically from /cak/, its reflexes having been retained in Thai as discussed here. Nho has a particle /na?/ for the ‘hypothetical’, with unknown origins, and Mène, a Northern Tai language spoken in a Southwestern Tai area, has /kham/ [< *g-]. In Southern Thai one finds /ci?/, /ca?/ and /?i-/ in environmentally conditioned variation (Diller, A. V. N. in conversation, October 1991, and A Southern Thai dictionary, Songkhla: Srinakharinwirote University, 1982Google Scholar [in Thai], for which Diller acted as consultant). The absence of cognates of OM [s-] elsewhere in Mon–Khmer, except for Praok /sag-/ (H. L. Shorto in conversation, June 1990)—likely to be a fused prefix consisting of *s- and *-N- —, Lawa /si/, both Northern MK languages spoken at the periphery of a former Mon area, and a possible cognate in Vietnamese se does not necessarily imply that its origin be sought in Lao /si?/.

²¹ ETh. caḥ is, in fact, a strengthened form of the weak forms ca and ca’. In fig. 5a forms spelt with mai han akat -ă- and those with reduplicated final consonant have not been listed separately.

²² See Bauer ‘Old Mon s-’, ‘OM s- Addenda’, and ‘Notes on Mon epigraphy II’, art. cit., for further details, such as exact references and citations from Mon epigraphy. Prior to the publication of DM1 Shorto referred to the ‘hypothetical’ also as the ‘preparative’. In DMI the ‘hypothetical’ is opposed to the ‘actual’.

²³ One might even argue that ETh. cakk, etc. share certain characteristics with inflectional affixes, such as high selectivity of its host (operative verbs), restricted functional range, phonological liaison (variety of weak forms), and strict ordering.

²⁴ Variant orthographies are excluded from the list, as are variations in the marking of the hypothetical; however, the position of the marker, that is, the locus, be it the OM/MM prefix s- or the ETh. grammaticalized verb cakk, etc., is not subject to variation. ETh. spellings have been regularized. V refers to ‘verb’, NEG to ‘negative marker’ (< OM *verb).

²⁵ Locations where gray, etc. are not marked by the ‘hypothetical’ are as follows: Ay. 1.A. 10, A.18x; Kd.30.A.17. In Ay.l.A.10 it modifies a noun, and in Ay.l.A.18 it is negated:... pa’ gray’ bīn... pa’ gray’ teit... ‘(he does) not desire (it)... (he does) not get angry …’. Notice, however,change of word order in modern Thai.

²⁶ Variation does occur in the OM/MM corpus where, as Shorto points out in DMI (mic, q.v.), the ‘hypothetical’ form is occasionally replaced by the ‘actual’.

²⁷ Translations by Blagden and Shorto, except for the last example.

²⁸ This construction occurs in the earliest recensions of the ‘Law of the Three Seals’; cf. Yoneo Ishii, Mamoru Shibayama, Aroonrut Wichienkhiew, The computer concordance of the Law of the Three Seals (Bangkok: Amarin, 1990), 5 vols.

²⁹ For details see Bauer, ‘Old Mon J-’, art. cit., especially figs. 8 and 9.

³⁰ For each occurrence of the clause-head sources, locations and dates are given. Whenever beia, etc. is followed by cakk, strong and weak forms of the latter are specified; ‘[ ]’ indicate lacunae. ‘Class’ indicates the word class of the element following the clause-head. ‘Dem.’ and ‘pt.’ refer to ‘demonstrative’ and ‘particle’ respectively, ‘N’ to ‘nouns’, ‘V’ to ‘verbs’.

³¹ Griswold/Prasert (1972: p. 142, n. 30), comment on beia in Nw.2 (#11): ‘[/phui:a/], which usually means “because” in Sukhodayan inscriptions, is found in No.2 at n/6 [Sd/2 (#2) B.6] with a sense closer to its modern meaning, “in order to”, which is evidently what it means here.’ Notice that broh borrowed from Khmer, occurs in the Sukhothai corpus only twice (Sd.l 1.B.18; Sd.28.B.27).

³² Further studies of interference in Thai should take into account cases of phonological interference, such as variant vocalism, as in NN. 1.A. 19 prejñābala, Skt. prajna and Lb.38.A.16, Skt. prajña, P. paña, ETh. preñā; ibid., Skt. krtajña, P. kataññutā, ETh. karteñatā. Griswold/Prasert (‘The inscription of Vat Jān Lom [1384 A.D.]: ‘Epigraphic and historical studies, No. 8’, JSS, LIX, 1, 1971, 189–208, here p. 203, n. 20) point out that in Sd. 18.28–29 ‘[recanā] may be Pali racanā, “arrangement”, “composition”, or else a mistake for Pali rajana, “coloring” (in either case, with A altered to E before a palatal sound, as it sometimes is in the Sukhodayan inscriptions, e.g. Sejanālai, for Sajjanālaya).’ Shorto states in DMI, p. xviii, that OM o before palatals may be written either a or e, and before y alwas e or, rarely eai ‘The writing of e before palatals presumably results from an attempt to note a terminal glide.’

³³ Such combinations are attested elsewhere in South-East Asian epigraphy, and are discussed in my ‘Notes on Mon epigraphy’, JSS, LXXIX, 1, 1991, 31–83, as for instance in Mg.2 where we find an 5- inflected Khmer-Mon blendform, or in KhK.16 which uses doubled clitics ‘awo wo’.

³⁴ Such a view can be further supported by structural borrowing, the existence of lexical caiques, and phonological interference. Further examples from Sd.2 are provided in the appendix. The difficulties concerning the interpretation of pronominal reference in Sd.2 may be seen in this perspective; further analysis is needed.

³⁵ Griswold/Prasert as in ‘The epigraphy of Mahādharmarājā I of Sukhodaya: epigraphic and historical studies No. 11, part 1’, JSS, LXI, 1, 1973, 79–111 (here p. 95, n. 20).

³⁶ Shorto glosses ma, ma’ in DMI as follows (p. 281): ‘attributive particle’, (i) ‘connecting relative cl. to n. (a) denoting agent.., (b) connecting cl. to n. denoting goal of action, (c) connecting cl. to n. denoting locus &c. of action; then never governed by pn.’ (ii) ‘occasionally without antecedent n., forming nom. phrases’ and (iii) ‘ma+v. is sometimes used interchangeable with the attr. form.’

³⁷ Shorto links this attributive marker with the OM infix -m-: ‘An attributive is derived from simple verb forms by means of the infix -m-, -um-, -uṁ- (-uṁ- in roots with two-place initial, -umfollowing a labial and -m- otherwise in simple-initial roots). It is equivalent to the construction with the clause-subordinating particle ma..., and is used chiefly when the verb has no object or other extension; thus especially when it translates an English adjective.... Like the construction with ma, attributives sometimes occur without accompanying noun, and then themselves function as agentive nouns’ (pp. xxiii-xxiv). The remark preceding the quotation in DMI ‘alone among affixes it [sc. s-] combines with secondary-system loanwords’ (DMI, p. xxiii) needs correction in this context: dmuk ‘to be poor, miserable’, in I.D.47 is the attributive form -m-,derived from OM duk, dok, dukkha, P. dukkha ‘misery’. It could then be argued that, byanalogy, OM/SM/pa?/ is derived from the causative prefix p- and the clitic OM/na?/ ~ /na/, from the infix -n- ‘instrumental’. According to Shorto the ‘agentive’ function of -m- is a derived one.

³⁸ Glosses are provided in fig. 11b.

³⁹ As shown in the glosses from the Ananda in fig. ll a.

⁴⁰ Figure lla, Sd.32. For an inventory, synopsis and index./glossary see my ‘The Wat Sri Chum Jātaka glosses reconsidered’, JSS, LXXX. 1, 1992, [in press]. Further structural parallels between Mon and Thai glosses are provided there, as for instance in the Old Mon glosses from W. Taungbi, #11 and #14 and Sd.32.11 and Sd.32.12 respectively.

⁴¹ Again, it is Sd.2 which shows close structural parallels with OM, Sd.2.B.34 ‘ann hñay’ being a caique of OM j-m-nok ~ ma jnok ‘large, big (non-predicative)’, unless ‘ann is to be interpreted here as a noun. Further parallels include instances such as ETh. ‘ann wā ‘regarding, as for’ in Kb.l.A.35, 42, 44, Nn.l.A.8, and Sd.26.A.36, to be compared with MM ma gah in MM a marker of quotation; but notice difference in word order which may be accounted for by Burmese influence on MM syntax. OM/MM gah, SM/kh/ is the lexical equivalent of wā, mod. Thai /wâ:/. No OM occurrence is attested.

⁴² EB inscr. V (dated 1098, from Myatheindan pagoda, Ayetthema hill, or ‘Phayre Museum inscription’) lines 1–42 are identical, apart from different spellings, with inscriptions EB inscr. Ill (Myagan tank, Pagan) and EB inscr. IV (Alanpagan). Following Blagden, citations noted V* in Figures 12a and 12b derive from EB inscr. Ill, with spellings properly adjusted for EB inscr. V; see Epigraphia Birmanica [EB], Rangoon, Superintendent, Government Printing, 1920 21960, vol. 1.2 [for inscription m pp. 131–43, iv p. 143, v pp. 143–7]. In EB inscr. V ma occurs 46 times (total number of words 426,207 lemmata). The other relative clause marker in OM mun occurs only once; three verbs are inflected for the attributive, the plural marker guṁluṅ (6 occurrences), puṁrey ‘to be noble’ (1), suṁmgihh ‘to be rich, wealthy’ (1) and tmuy ‘to be finished’ (1).

⁴³ For the former see ‘The Wat Sri Chum Jātaka glosses reconsidered’, art. cit.

⁴⁴ For details on OM wo’ see JSS, LXXIX, 1, art. cit., 36–43. In Old Khmer the demonstrative, when occurring sentence-initially, must be linked by gi, gī. In the ETh. corpus such a construction is attested only once in Sd.37.2 ‘ăn ni gi, making comparison difficult for lack of comparable data.

⁴⁵ The evidence from ETh. would suggest that by the fourteenth century OM medioclusters And-/ and /-nd-/ had not yet been simplified to /-1-/ and /-n?/-/ ∼ /-ṅ-/ respectively; this mediocluster reduction is attested by EMM, in the late fifteenth century.

⁴⁶ The latest Mon inscription in the upper Chao Phraya Basin is Nw.7, a terracotta stupa inscribed in Mon from Muang district, Nakhorn Sawan; for related finds—terracotta stupas inscribed in Pali—see JSS, LXXIX, 2, 1991, 61–79.

⁴⁷ Nw.l (K 966), a dated inscription in Pali and Khmer, from a district just north of the town of Nakhorn Sawan; cf. G. Cœdès ‘Nouvelles donnèes èpigraphiques sur l'histoire de l'lndochine centrale’, Journal Asialique, CCXLVI, 2, 1958, 125–42 (here pp. 132–9). As far as the identification of speaker population from language use is concerned, it is worth recalling Cœdès's remark in this context that ‘une inscription en vernaculaire est faite pour être lue et comprise par la population locale qui ne parle pas forcément la même langue que la Cour’ (art. cit., p. 137)

⁴⁸ cf. K992 which has eight lines of Sanskrit (A) and three lines in Khmer (B): [1] – [2] tai kes [t]ai – [3] krapī dnayaṁ vyar –. Cœdès, was unable to obtain any documentation; the inscription was stored at the National Museum, Bangkok, and said to have originated from Sukhothai (Cœdès, G., Inscriptions du Cambodge (Paris: Ecole Française d'Extrême-Orient, vol. vii, 1964)Google Scholar.

⁴⁹ Another five inscriptions in Old Mon have been recovered from sites in Chiangmai province which have not been documented in DMI; see JSS LXXIX, 1, inventory pp. 35–6: (1) 23/2523 and (2) 24/2523, two votive images from Saraphi district, now at the National Museum, Lopburi, together with (3) an unregistered tablet probably from the same site, now at the Los Angeles County Musem of Art, reproduced in Arts of Asia, xv, 6, 1985, p. 115 and JSS LXXIX. 2, 1991, pp. 62–63, (4) Jm.45 from Wieng Mano, San Pa Tong, and (5) Bp. 422/2524 from Wieng Tho. All of these inscriptions antedate those from Lamphun sites, incorporated into DMI.

⁵⁰ For a discussion of the Wat Sri Chum inscription, especially its date and authorship, see, apart from Griswold/Prasert, 1972, op. cit., Vickery, M. ‘A guide through some recent Sukhothai historiography’, JSS, LXVI, 2, 1978, 182–246Google Scholar (for inscription #2 pp. 209–16); id. ‘From Lamphun to Inscription No. 2’, Siam Society Newsletter [SSN], ii, 1, 1987, 2–6Google Scholar; id. ‘Some new evidence for the cultural history of central Thailand’ SNN, ii, 3, 1986, 4–6Google Scholar; Gosling, E. M. B. ‘The history of Sukhothai as a ceremonial center: a study of early Siamese architecture and society’, Ph.D. thesis, University of Michigan, 1983, xix, 498Google Scholar pp., 265 ill., 2 vols. (now publishedin a revised form as Sukhothai: Its history, culture, and art, Oxford, 1991Google Scholar; ch. ii published in anearlier version in JSS, LXIX, 1–2, 1981, 13–42, and partly as ‘Inscriptions and art history: the case of inscription II’, in: Bickner, R. J., Hudak, T. J., Patcharin, Peyasantiwong (ed.), Papers from a conference on Thai studies in honor of William J. Gedney (Michigan Papers on South and Southeast Asia, 25, Ann Arbor: University of Michigan, Center for South and Southeast Asian Studies, 1986, 149–57)Google Scholar; id. ‘On Michael Vickery's “From Lamphun to Inscription 2” ’, SSN, IV, 1, 1988, 5–7Google Scholar.