^{page 111 note 2} ‘Gurung’ and ‘Gurungkura’ are the traditional spellings recognized by the War Office Handbook on Gurkhas and the LSI; the Nepali words would be transcribed guru, gurukura in the transcription made use of by T. W. Clark, Introduction to Nepali (in preparation), which has been used for all Nepali words quoted here.

^{page 111 note 3} Hodgson, B. H., Essays on the Languages, Literature and Religion of Nepal and Tibet, London, 1874.Google Scholar

^{page 111 note 4} e.g. Money, G. W. P., Gurkhali Manual, 3rd edition, Bombay, 1942.Google Scholar

^{page 111 note 5} Part I, 182-8, with word-lists on pp. 254-270.

^{page 111 note 6} LSI, III, I, 183

^{page 111 note 7} For the terms ‘system’ and ‘sub-system’ see Part II of this paper and footnote 3, p. 116.

^{page 112 note 1} i.e. in terms of a ‘prosodic’ approach to tone rather than a ‘tonemic’ one. See here particularly A. E. Sharp, ‘A Tonal Analysis of the Disyllabic Noun in the Machame Dialect of Chaga’, BSOAS., xvi, pp. 167 ff.

^{page 112 note 2} Brief reading conventions are given here, but it is to be understood that the transcription is based on phonological rather than broad-phonetic criteria, and values assigned here can be no more than approximate, i and e are front unrounded vowels, respectively close and mid. yW represents a front rounded vowel, half-close. & is an open unrounded front-central vowel, while o and U are mid and close partially-rounded back vowels respectively. ey represents a front-closing diphthong whose starting-point is an unrounded central vowel. Nasality is represented by ~ above the vowel-letter (or above the first where more than one are used), k t p, g d b, t d are plosives, of which t and d are dental, t and dretroflexed and (]_ retroflexed; C and j are prepalatal affricates more closely represented as [Cc] and [IZ]. Of these k t p t and c are not aspirated; aspirated units are represented by kh, etc. tr represents a post-alveolar affricate, described in Part II of this article, n n n and m are the velar, palatal, dental, and bilabial nasals respectively, and are voiced, y and w are palatal and labiovelar semivowels respectively; the labiopalatal semivowel is written, like the front-rounded vowel, as yw, which in practice causes no ambiguity, r is a single-tap alveolar fricative [f], and 1 is clear; both these are voiced, except in the clusters kl and kly where the 1 represents a voiceless lateral fricative. S and ∫ are respectively alveolar and alveolo-palatal tongue-tip-down fricatives, with or without voice. The ‘glottal fricative’ h may be voiced as a tone-2 initial. (q represents the glottal stop and is only associated with tone 1. gh jh dh dh and bh are used only in transcribing Nepali loanwords, as are n and ew. These are read in Gurungkura as g j d b n and O respectively. A more detailed phonetic transcription has been given where appropriate; this is printed within square brackets and its symbols are those of the I.P.A. Alphabet. It has been found necessary to recognize three pitch-levels, high, mid, and low. These have been represented graphically by the following symbols:— A. High level – B. Mid level – C. Low level D. Fall \ E. High fall \ F. Low fall \ G. Fall-rise √ H. Rise-fall ^ I. Rise / each of which is associated with one syllable of the phonetic transcription. (This does not necessarily correspond with the syllabication of the phonological spelling; cf. the close transcription of humu QU in Example 12.) Word-division has been represented by space. Where a hyphen is used (e.g. Example 13) it has no significance beyond the help it gives to the eye in separating the components of a compound word.

^{page 113 note 1} This does not apply to the 1 written in the clusters kl and kly; cf. f.n. 2, p. 112 above.

^{page 113 note 2} The ‘ translation-meanings ‘ appended are for convenience of identification only; they do not necessarily imply congruence in terms of meaning at either the grammatical or lexical levels.

^{page 113 note 3} Following Sprigg, R. K., ‘Verbal Phrases in Lhasa Tibetan’, BSOAS., xvi, 134–156, 320-350 and 566-591, ‘tone-1 words’ and ‘tone-2 words’ have been abbreviated as 1W and 2W respectively, and fS and sS have been used as convenient abbreviations for fast and slow speeds of utterance respectively.Google Scholar

^{page 114 note 1} A tonal category to which a monosyllable in isolation may have been assigned is irrelevant when such ‘monosyllable’ may be identified with the second morpheme in a disyllabic word. E.g. the features of the initial of the syllable in isolation, as described above, are not coincident with those which could be stated for the initial of that syllable in fine compositi. The exponents of intraverbal junction, of which an account is outside the scope of this paper, have no relevance as exponents of tonal categories. It may, however, be necessary to write the tone-mark before the second morpheme in order to ensure consistency in lexical entry, e.g. Cu'ni ‘twelve’ contra cuni ‘seventeen’; these are both 1W, and this is symbolized by the absence of tone-mark initially. The morpheme ni is tonally neutral, but partakes of the features of tone 2 in a monosyllable and, in this case, tone 1 in a disyllable.

^{page 114 note 2} An instance of ‘creaky’ voice has, however, been recorded in one 2W in fS: -creak- naji 'pogi [naizl bri:] § § \ _ ‘I carried it’, beside sS [brBi:]; here, however, a velar would be stated in other members of the grammatical paradigm, e.g. / [brBo] (fS), [brgo:] (sS) ‘carry it’, and this feature need neither be considered as ‘glottalization’ nor specially signified in the transcription.

^{page 115 note 1} For the term ‘colligation’ see H. F. Simon, ‘Two Substantival Complexes in Standard Chinese’, BSOAS., xv, 327-355, and J. R. Firth, footnote contributed to Sprigg, op. cit., p. 136.

^{page 115 note 2} For convenience of identification Nepali translation-equivalents of these counting-series are given here:— Example 5. ek, dui, tin ddf … … de∫ Example 7. yewta, duita, tinawla ‥ de∫ewta

^{page 115 note 3} For the term ‘contonation’ and its relation to intonation, see J. R. Firth, footnote contributed to Sharp, op. eit., p. 164.

^{page 115 note 4} No translation-meanings can be given in English to distinguish Examples 10 and 11 from Example 9, and Nepali translations are therefore given here:— Example 9. sat/dui dkh, sat/dui jana manche. Example 10. sat/duila akh, sat/dui jena manche. Example 11. satewta/duita akh, satewtay/duitsy pna manche.

^{page 116 note 1} Being here exponents of ‘emphasis’.

^{page 116 note 2} cf. a similar observation made by J. Carnochan in ‘A Study in the Phonology of an Igbo Speaker’, BSOAS., i n, 424.

^{page 116 note 3} It must be emphasized that the examples given above are presented in limited intonationcontexts in illustration of the principal points made in this part of the article, and are not to be considered as exhaustive in terms of either intonation or contonation. From an investigation of the recorded utterances of the informant it would appear that intonation requires to be stated, provisionally, in terms of at least three systems; it is possible that a study of Gurungkura in the field may require a statement with a larger number of terms to account for a wider range of emotional contexts than have been recorded from a single speaker. Furthermore, while conclusions drawn from illustrations such as those in Example 6 may be valid for all grammatical categories, it does not follow that generalizations made on the basis of the characteristics of words in isolation are necessarily valid for all categories in all contexts, nor that such generalizations are necessarily valid for words outside the main system, which have been excluded from this statement.

^{page 117 note 1} op. cit,, 183

^{page 117 note 2} In ‘Sounds and Prosodies’, TPS., 1948, 127.

^{page 117 note 3} In ‘The Phonology of Loanwords in some South-East Asian Languages’, TPS., 1951, 133.

^{page 117 note 4} -n-, which must be set up for Kat, hman_4u Nepali, was not required for a statement of the informant's Nepali speech, which might be described as the Western koine strongly influenced by his Army environment.

^{page 117 note 5} cf. W. S. Allen, ‘Some Prosodic Aspects of Retroflexion and Aspiration in Sanskrit’, BSOAS., xni, 939-946, and ‘Retroflexion in Sanskrit: Prosodic Technique and its Relevance to Comparative Statement’, BSOAS., xvi, 556-565. I am indebted to Dr. Allen for allowing me access to this second article before its publication and for checking my statement against his own conclusions.

^{page 117 note 6} In the second case the LSI occasionally writes t; this is probably a slip.

^{page 118 note 1} R. Shafer, in ‘Classification of Some Languages of the Himalayas’ [sic], Journal of the Bihar ] Research Society, xxxvi, 3-4, 192 ff., refers (p. 200) to ‘stop consonants + -r-' having become j ‘supradentals’ in modern spoken Central Tibetan dialects. If by this extraordinary term he; means alveolar or retroflexed plosives, his assertion is not in agreement with the facts. The articulations in question, although tr, dr may be written in Tibetan for the retroflexed plosives in words borrowed from Indian languages (e.g. Tib. drag for Hindi (j_ak), are in fact not plosives but affricates. See here Sprigg, op. cit., f.n. to p. 142.

^{page 118 note 2} cf. Eugenie J. A. Henderson, ‘Notes on the Syllable Structure of Lushai’, B80AS., xn, 722.

^{page 118 note 3} The two members of this term are commutable only among the initial alternances in Gurungkura; this fact provides further reason for dissociating them from the ‘retroflexed ‘ term.

^{page 119 note 1} It would appear that R. Shafer, loc. cit., has been misled by Hodgson's and other specimens into making this comparison without regard to the tonal discrepancy between Gurungkura and Tibetan. As most of Shafer's material has been culled from Hodgson, the LSI and other equally untrustworthy sources, it is difficult to accept his conclusions without reservation.

^{page 119 note 2} This may be inferred from a consideration of the lolchor terms in Gurungkura, used in counting years in the 12-year cycle, vis-a-vis their Tibetan counterparts. Of the twelve terms, cu`lo`lo to`lo `ywi`lo `mu`pryw`lo sa`pra`lo ce`lo khi`lo and pho`lo (cf. Tibetan byi-ba (-lo) glay stag yos hbrug sbrul Ha lug spre(l) bya khyi and phag), at least `mu`pru and sa` pryw must be set outside the main system on account of their patterns in isolation (-^ and the series is also distinguished by the ‘animal’ words being used only in colligation, and collocation, with Mo ‘year’ (other words are in common general use to represent the animals which correspond to the ‘ animals ‘ of the cycle). They are thus to be suspected as loanwords, and are sufficiently far removed from the spoken Central Tibetan terms as to make it very improbable that they are modern importations by the selected youths the Gurungs send for study in Tibetan monasteries, many to Lhasa itself, to return as priests.