¹ Cœdès, G., Inscriptions du Cambodge, VI, Paris, 1954.Google Scholar

² The distinction between phonological postulates and the transliterated characters (for which M. Cœdès' system is followed) is maintained throughout by using bold type for the former and italics for the latter.

¹ Borrowing the useful term, of Henderson, E. J. A., ‘The main features of Cambodian pronunciation’, BSOAS, XIV, 1, 1952, 149–74.CrossRefGoogle Scholar

² The hyphen is here used to help in demonstrating the infix. Elsewhere, in transliterated words it is used exactly as in Cœdès, Inscriptions du Cambodge; in the phonology it is used to indicate disyllables.

³ The word rnoc is in the Angkorian dialect often spelt rṇṇoc, but this is an alternative for the spelling with dental n.

⁴ The fact that the modern pronunciation of this character is unvoiced does not seem relevant here; the important point is that modern b and v represent distinct initial consonants.

¹ The same system of transliteration is used for Modern Khmer as for Old Khmer with the additions of the phonetic symbols, ε, , and uə used to represent the vowels, and .

² The numeral occurs on many other pre-Angkorian inscriptions but is usually not written in word-form. In Angkorian Khmer the word-form is vyar.

³ The digraph hv is also used to represent a final consonant in Old Khmer which is, in Modern Khmer, v; e.g. OK jāhv, MK jāv, ‘to barter’.

⁴ The modern script is used here to facilitate printing.

⁵ As in Modern Khmer.

¹ For the phonological interpretation of both taet and ta-et as t?et, see p. 361.

² ṛ and ḷ do occur occasionally in Pre-Angkorian and Angkorian Khmer, e.g. tṛ (MK trī ‘fish’) but the examples are too few to need discussion here.

¹ As in the words yi, yuṅ, yok ‘to take’, yoṅ, yon, yol.

² Discussed on pp. 360–4 below.

³ Modern Khmer kryel ‘crane’, in which ye represents a diphthong.

⁴ No precise phonetic knowledge of the realization of this or any other vowel-unit is claimed; the use of the length-mark here is due to the need to distinguish phonologically between īə (pre-Angkorian and Angkorian) and iə (Angkorian only), ūə (pre-Angkorian and Angkorian) and uə (Angkorian only) are also distinguished in this way.

⁵ Representing both b and v.

¹ Contrast OK consonantal v > MK v, e.g. pre-Angkorian svā, lvā, MK svā ‘monkey’, lvā ‘fig’.

² See pp. 360–4 below for discussion of permissible patterns in initial sequences.

³ See p. 356, n. 4.

¹ ai and au are held to represent a + final y and final v respectively and should therefore be mentioned here.

¹ There is an interesting distinction here between the behaviour of surd and sonant, a distinction which the writing suggests but which no longer operates in Modern Khmer. Thus while the junction of a surd and a liquid or nasal is often marked by aspiration, that between a sonant and a liquid or nasal is often marked by the anaptyctic vowel.

² Many female slave names are preceded by ya, a Mon female slave title (I am indebted to Professor G. H. Luce for this suggestion); e.g. ku ya māy. Orthographically one cannot tell whether two words, ya māy, or one word, yamāy, is intended, but phonologically ya māy is suggested.

¹ op. cit., III, p. 75, n. 3.

² N is used with the implication ‘any one of the four nasal consonants, ṅ, ñ. n, m’.

¹ Modern Khmer babil supports the phonological transcription b for v.

¹ By definition, so to speak, v must be construed as b here (since v is an initial of restricted use). Modern Khmer bhṅā corroborates the b in bmṅā

² Modern Khmer drabaṅestablishes the b here.

¹ There seems to have been a prefix kN too; cf. je, je (capacity measurement) and kañje, kñje, basket used for measuring. A very large proportion of words have a 3-place initial of which the first two consonants are kN.

² Numerals follow tloṅ but precede tanloṅ, e.g. tloṅ 2 tariloṅ.

¹ dialect, Angkorian, pvasGoogle Scholar, pūəs, pamnvas, pmnūəs.