¹ On the Aborigines of India. Essay the First; on the Kocch, Bódo and Dhimal Tribes, in Three Parts Calcutta: Baptist Mission Press 1847Google Scholar

² Miscellaneous Essays Relating to Indian Subjects London, 1880, I, pp. 403, 406Google Scholar

³ pe or pe-da lu-da, the narratives, could be glossed “principle” or “example” or, crudely, “principle word example word”. For alternative terms discussed in relation to Gurung material: Pignède, B.Les Gurungs: une population himalayenne du Népal, Paris, 1966, pp. 323–24;Google ScholarHöfer, A.Tamang Ritual Texts I, Wiesbaden, 1981, p. 69:Google ScholarMacdonald, A. W.Essays on the Ethnology of Nepal and South Asia, Kathmandu, 1975, pp. 130,Google Scholar 147 fn. 21, 150–151 fn. 35; Snellgrove, D. L.The Nine Ways of Bon Boulder, Colorado, 1967, pp. 20,Google Scholar 256 fn. 9; Stein, R. A.Tibetan Civilization London, 1972, p. 198Google Scholar s.v. gtam-dpe; “Du récit au rituel dans les manuscrits tibétains de Touen-houang” Etudes tibétaines dédiées à la mémoire de Marcelle Lalou, Paris, 1971, p. 504;Google ScholarTucci, G.The Religion of Tibet London, 1980, pp. 210, 232,Google Scholar 273 fn. 21; Allen, N. J. Studies in the Myths and Oral Traditions of the Thulung Rai of East Nepal, unpublished Oxford D.Phil, thesis, 1976, pp. 137 fn. 1, 256–60:Google ScholarGurung, Ṭekan SīBauddha Dharma: Tamu Guru Kyẽrlõ Butwal, 1980, II, p. 23Google Scholar s.v. pyẽtãlbūtã; Strickland, S. S. Beliefs, Practices, and Legends, unpublished Cambridge Ph.D. thesis, 1982,Google Scholar chapter 5.

The orthography adopted in this paper is based on the broad reading conventions used by Burton-Page “Two Studies in Gurungkura” BSOAS 1955, 17, pp. 111–19,CrossRefGoogle Scholar 112 fn. 2, and is therefore provisional. Insufficient material on the phonology of Gurung dialects prevents a systematic comparison with Glover's Gurung Phonemic Summary, Kathmandu, 1969.Google Scholar At the level of the initial syllable of the “phonological word”, Gurung is reported to show two phonemically recognised pitches, a level/contour contrast, and a breathy/clear contrast on forms in the high tone class, at least for the villages Ghandrung (Burton-Page “Two Studies”, p. 114 fn.l and p. 116 fn. 3), and Ghacok (Glover “Gurung Tone and Higher Levels”, Kathmandu, 1970, p. 65Google Scholar). But the pitch and melody of suffixes are conditioned by the lexical item to which they are added, an observation applied to all theTamang-Gurung-Thakali languages by Mazaudon (“Consonantal Mutation and the Tonal Split in the Tamang Sub-Family of Nepal” Kailash 1978, pp. 161–163Google Scholar). Although this is the case, little attempt has been made systematically to approach the relationship between the tone of the monosyllabic morpheme and prosodic features of larger phonological units. Until this is attempted, it would seem premature to represent tonal characteristics of phrases from dictated pe and to assume a direct relationship to the melody of performances, a subject requiring extended linguistic and musicological treatment in its own right. For these reasons, neither tone marks nor musical transcriptions nor the breathy/clear contrast on terms in the higher tone class, are given here. For clarity of presentation, reduplicated forms, and root terms and their suffixes have been hyphenated. “Filler syllables”, usually [-ye(–)], are represented but are not hyphenated. Paired limbs of binomials and polynomials are underlined.

⁴ “Classification of the Sino-Tibetan Languages” Word 1955, pp. 11, 100–01Google Scholar

⁵ “Two Studies”, p. 119Google Scholar fn. 1. Mazaudon notes that languages spoken in Nyi-shang (Manang) and the NarValley should also be included within this Branch (“Consonantal Mutation”. p. 158Google Scholar).

⁶ “Cognate Counts via the Swadesh List in Some Tibeto-Burman Languages of Nepal” Tone Systems of Tibeto-Burman Languages of Nepal, III, pp. 23–107,Google Scholar edited by F. K. Lehman, Illinois, 1970, pp. 24–5. In Glover's scheme, East Himalayish is raised to form a subdivision paralleling a Bodic subdivision within Bodic.

⁷ Benedict, P. K.Sino-Tibetan: A Conspectus Cambridge, 1972, pp. 5–7CrossRefGoogle Scholar

⁸ Hale, E. A.Research on Tibeto-Burman Languages The Hague. 1982, pp. 13–5;Google Scholar and Matisoff, J. A.Variational Semantics in Tibeto-Burman, Philadelphia. 1978. pp. 4–5, 91—106ff.Google Scholar

⁹ Hale, E. A., op. cit, pp. 55–6Google Scholar

¹⁰ Hooker, J. D.Himalayan Journals. Notes of a Naturalist in Bengal, the Sikkim and Nepal Himalayas, the Khasia Mountains, &c. Revised edition, London, 1855, I. p. 261;CrossRefGoogle Scholar but also p. 160. where “Geroongs” are noted as “shepherds”. For comparison, Malla has proposed an argument relating the term Newar to T-B roots nghet‘cattle’ and pa ‘man’ i.e. “herdsman” (“Linguistic Archaeology of the Nepal Valley: A Preliminary Report”, Kailash VIII (1–2), 1981, p. 19).Google Scholar

¹¹ See Hamilton, FrancisAn Account of the Kingdom of Nepal, and of the territories annexed to this dominion by the House of Gorkha, Edinburgh, 1819, p. 243Google Scholar s.v. Seshant; Hodgson, , Essays, II, p. 30Google Scholar fn. s.v. sén Snellgrove, D. L. “Expériences népalaises” Objets et Mondes Vl, 1966, 2, p. 108Google Scholar s.v. seṃ; Höfer, Tamang Ritual Texts, pp. 6–7Google Scholar s.v. seṁ or sẽ; Allen, N. J. “Fourfold Classification of Society in the Himalayas” Himalayan Anthropology edited by Fisher, J. D., The Hague, 1978, p. 11Google Scholar fn. 2; Macdonald, A. W.Essays, 1975, p. 129Google Scholar & fn. 9; Jackson, D. P. “Notes on the History of Se-rib, and Nearby Places in the Upper Kali Gandaki Valley”, Kailash, VI, 3, pp. 195–228Google Scholarpassim. Se is the only term for “Gurung” to occur in the pe recorded, tamu being the name in colloquial use. The se are divided into two groups of clans: the sõ-bu se “Three Se” and the ku-bu se “Nine Se”, which correspond to the contemporary kro-məy and pwẽməy or (in Nepali) cãr jāt “Four People” and sora jāt “Sixteen People”. It is artificial to discuss the name Se apart from Tibetan traditions considered by Stein, R. A. in Les Tribus anciennes des marches sino-tibétaines: légendes, classifications et histoire (Paris, 1961, pp. 3 ff.Google Scholar), and from the place occupied by “Guruṅg” in the Nepalese caste order (Höfer, A., The Caste Hierarchy and the State in Nepal, a Study of the Muluki Ain of 1854, Innsbruck, 1979);Google Scholar but these two subjects lie outside the scope of this study.

¹² Allen, N. J. “Tibet and the Thulung Rai: Towards a Comparative Mythology of Bodic” Tibetan Studies in Honour of Hugh Richardson edited by Aris, M. and Kyi, Aung San Suu, New Delhi, 1980, pp. 1–2.Google Scholar

¹³ Glover's lexicostatistical approach, using Swadesh 100–word lists, suggests that Gurung Branch and Tibetan languages of Bodish share about 23% cognates, diverging around 2.900 BC; while the Bodic Subdivision shares with East Himalayish some 13% cognates, diverging about 5.000 BC (“Cognate Counts”, p. 25Google Scholar). More eloquently, Hodgson's judgement was that,

“It must suffice at present to observe that the legends of the dominant races indicate a transit of the Himálaya from thirty-five to forty-five generations back — say 1,000 to 1,300 years, and that I prefer the remoter period because the transit was certainly made before the Tibetans had adopted from India the religion and literature of Buddhism, in the seventh and eighth centuries of our era. This fact is as clearly impressed upon the crude dialects and cruder tenets of the sub-Himálayans as their northern origin is upon their peculiar forms and features, provided these points be investigated with the requisite care; for superficial attention is apt to rest solely upon the Lamaism recently as imperfectly imported among them, and upon the merely exceptional traits of their mixed and varying physiognomy.” (Essays on the Languages, Literature, and Religion of Nepal and Tibet, London, 1874, II, p. 31Google Scholar).

¹⁴ In one secret language, pharsyo kywi; spoken by girls in the now old-fashioned village “dormitory” (rodi), the form of modification of words is as follows: 1. for any monosyllabic term CV, read mVCVrka; 2. for any disyllabic term C¹V¹C²V², read mV¹C²V²C¹ V¹rka; 3. for any polysyllabic term, the initial monosyllable alone is relocated and receives the suffix -rka. The term pharsyo was glossed a-mildi-wa meaning “discordant, unharmonious” (from Nepali milnu, and therefore a loan word). None of these secret speech forms, nor even the Tsō kywi, was studied by Glover. Although he states that he “did not locate a recognized story-teller” (Sememic and Grammatical Structures in Gurung (Nepal), kathmandu, 1974, p. 185Google Scholar), he nonetheless includes in the Dictionary a table of shamanic or priestly ceremonies which, had they been studied linguistically, would have provided him with valuable informaiton (Glover, W. W., Glover, J. R.. Gurung, D. B.Gurung Nepali-English Dictionary, Canberra 1977, pp. 311–13Google Scholar).

¹⁵ Macdonald, A. W., Essays on the Ethnology of Nepal and South Asia, Kathmandu, 1975, pp. 145–46Google Scholar fn. 13, citing Roerich, (“Osnovnye problemy tibetskogo yazykoznanija” in Sovetskoe Vostokovedenie, Moscow, 1958 (4), p. 104Google Scholar) and Das, (Journey to Lhasa and Central Tibet, London 1904, pp. 4–5,Google Scholar fn.) who link this province to several Nepalese populations, particularly the Limbu. Indeed, it is sometimes said that Tsõ language is partially Tibetan in vocabulary. The place named widzu tsõ-ru is the “highest” spot to which the shaman travels in his search for lost souls, and is also the home of a great mythical shaman whose name (cyõ poju tida poju) carries the sense of a northern origin; cf. Höfer, Tamang Ritual Texts, p. 19,Google Scholar s.v. Uiseme.

¹⁶ “Gurung Dialects” Papers in Southeast Asian Linguistics No. 7 edited by Traill, R. L. et al. , Canberra, 1980, pp. 34, 48.Google Scholar

¹⁷ Ibid., pp. 58–65.

¹⁸ Glover, W. W., Glover, J. R., and Gurung, D. B.Gurung-Nepali-English Dictionary, Canberra, 1977;Google Scholar but Gurung dialects need attention beyond the level of Swadesh lists.

¹⁹ “Gurung Dialects”, p. 35Google Scholar Table 3, and pp. 58–64 columns 2, 6, 57b, and 80.

²⁰ Cf. Matisoff, , Variational Semantics, pp. 19–21, 113–40, 147 ff.Google Scholar; Hale, , Research, p. 58.Google Scholar

²¹ “Gurung Dialects”, pp. 35–6.Google Scholar

²² Ibid., p. 37.

²³ Ibid., p. 63 column 65.

²⁴ Gurung-Nepali-English Dictionary, 89b; and Matisoff, Variational Semantics, p. 36,Google Scholar on primary and secondary alternation of nuclear vowels.

²⁵ Burton-Page, 1955, p. 119 fn. 2, suggests that most of these animal terms are “old borrowings”.

²⁶ Thus, discussing onomatopoeic forms, Emeneau, (Language and Linguistic Area: Essays by Murray B. Emeneau edited by Dil, A. S., Oxford, 1980, p. 75Google Scholar) has attempted to differentiate between evolutionary and diffusionary aspects of language. Thus, he argues that,

“when for the same phoneme of the proto-language, in the same phonemic context, two different phonemes are found in language A corresponding to one phoneme in language B (or two phonemes whose distribution is explainable by contextual conditioning), then that phoneme of A which is the more dissimilar to the phoneme of B is the straight-line development, i.e. the continuant, from the proto-language and the other is found in borrowings from B. In general, this applies to languages which are in such a situation that contact of some sort allows or allowed borrowing at a postulated period. It is notable that relative numbers of examples do not enter in consideration at all.”

But in T-B monosyllabic and tonal languages this is easier said than done as Hale points out (Research, pp. 55–6Google Scholar) and as recent work on tonogenesis would suggest (Mazaudon, “Consonantal Mutation”, and “Tibeto-Burman Tonogenetics”, Linguistics of the Tibeto-Burman Area 1977, III, 2, pp. 1–123Google Scholar).

²⁷ Diffloth, G. “Expressives in Semai” Austroasiatic Studies, Canberra, 1976, pp. 263–64Google Scholar fn. 2.

²⁸ Emeneau, op. cit., pp. 7–8, 114;Google ScholarNacaskul, K. “Types of Elaboration in Some Southeast Asian Languages” Austroasiatic Studies, Canberra, 1976, pp. 873–89;Google Scholar and Matisoff, Variational Semantics, pp. 58–72.Google Scholar J. Hoffmann and A. van Emelen's exposition of parallelism in Munda verse, and particularly its use of jingle words and compound forms, is a useful source of comparison with the present analysis (Encyclopaedia Mundarica Patna, 1930–1941, Vol.IV(D-D), pp. 1116–54,Google Scholar esp. pp. 1136–45). Compare also G. M. Hopkins' comment cited by Jakobson in “Grammatical Parallelism and its Russian Facet”, Language 1966, pp. 42, 313.Google Scholar Vitebsky's analysis of Sora poetics (“Sora Poetics: the creation of metaphor in Sora ritual through grammatically parallel verse forms” Proceedings of the Second International Congress of Austro-Asiatic Linguistics,Mysore, 1978,Google Scholar in press) first drew the author's attention to Jakobson's survey and prompted this study.

²⁹ On Gurung verbs “to be”, see Glover, “Three Gurung equivalents of English Be” Journal of Tribhuvan University, Special Linguistic Number 1969, pp. 36–56.Google Scholar

³⁰ Language and Linguistic Area, pp. 264–65;Google Scholar Matisoff comments on primary and secondary vowel alternation phenomena from some Tibeto-Burman languages and suggests the possibility of sound symbolism in this (Variational Semantics, pp. 36–9Google Scholar). There are slight traces of medial -w- with the mid-back vowel o in the syllables [-kõ-dze] or more narrowly [-kwõ-dze], exhibiting an alternation the reverse of that noted by Matisoff in modern Lahu (ibid, pp. 38–9).

³¹ Malkiel, Y. “Studies in Irreversible Binomials” Lingua 1959, VII, pp. 140–41.Google Scholar

³² See Benedict Conspectus, article 256: Tsangla mu-gu, Thabor and Bunan khu, Vayu ku-lu, Bahing ku-ni, Limbu me-ku, T-B *kuw. There is also the form kedi-keba “bamboo” (T-B *g-pa/g-pwa ibid., 210, & article 44 for T spa/sba), though it seems to have no colloquial source in contemporary Siklis speech (which gives mo). Hale notes that Sino-Tibetan languages have shown a tendency to develop polysyllabic words, occasionally through suffixation but more usually prefixation and compounding. He states: “These two processes have worked together in various ways so that in any given language we can find large assortments of variant morphemes which have descended from much smaller sets of protoforms. Variants which show both phonological and semantic resemblance can be viewed as members of a single word family” (Research, p. 57Google Scholar). Cf. also Matisoff, , Variational Semantics, pp. 13–14, 19–21.Google Scholar The complex vocabulary of ritual language forms suggests that some such polysyllabic forms might be innovated within this style of speech, without implying that this would explain any general tendency within the language as a whole.

³³ Cf. Benedict's *s-kyur/*su.r “sour”. Conspectus, article 42.

³⁴ “Studies”, p. 113.Google Scholar

³⁵ Ibid., p. 116.

³⁶ Ibid., pp. 124–25.

³⁷ Hale, K. “A Note on a Walbiri Tradition of Antonymy”, Semantics edited by Steinberg, D. D. and Jakobovits, L. A., Cambridge: University Press, 1971, p. 476.Google Scholar

³⁸ As apparently for the Thulung in Allen's, N. J. “Sewala Puja Bintila Puja: Notes on Thulung Ritual Language” Kailash VI, p. 4, p. 251.Google Scholar See also Vitebsky op. cit. for the Sora; and for the Munda, Hoffmann, & Emelen's, Encyclopaedia Vol. IV, pp. 1140–45.Google Scholar Compare Matisoff's general comments in Variational Semantics, pp. 58–9,Google Scholar though the comparative study of these forms poses interesting problems (ibid, pp. 130–33).

³⁹ “Grammatical Parallelism”, p. 423.Google Scholar

⁴⁰ The verb pya-shi, ~pya-wa “inserting”, was in a dictated version; on tape it was heard to be khya-shi, ~*khya-wa presumably with the same meaning.

⁴¹ “Studies”, pp. 126–29.Google Scholar

⁴² E.g. by Nacaskul, “Types of Elaboration”, p. 884;Google Scholar and Hightower, J. R. “Some Characteristics of Parallel Prose” Studia Serica Bernhard Karlgren edited by Egerod, S. and Glahn, E., Copenhagen. 1959, pp. 60–91.Google Scholar

⁴³ Lowth, RobertDe sacra poesi Hebraeorum, Oxford, 1753;Google Scholar revised edition translated by Gregory, G. as Lectures on the sacred poetry of the Hebrews, London, 1787;Google Scholar see Lectures III—VI.

⁴⁴ Compare Höfer's interesting discussions of Tamang shamanic language forms in “Is the bombo an Ecstatic? Some ritual techniques of Tamang shamanism” Contributions to the Anthropology of Nepal, edited by von Fürer-Haimendorf, C., Warminster, 1974, pp. 168–82;Google Scholar and his “Tamang Ritual Texts. Notes on the Interpretation of an Oral Tradition of Nepal”, JRAS, 1985, 1, p. 27.Google Scholar

⁴⁵ In the performance from which this example was transcribed, there was an erroneous phrase by the leading shaman's pupil who gave ki ngye-ngye “you are” instead of the form given here to which he was immediately corrected by the master.

⁴⁶ The character Kara-ba-rẽ was equated by Padam Sing with Mahādeu. The pe of Kara-ba-rē, named kara da, narrates his marriage to his two sisters. One sister bears him nine sons, the other seven daughters. The nine sons try to kill their father by cutting away a rope ladder from which he gathers wild honey on a cliff. Thinking their father has fallen to his death, the nine sons compete at archery for the prize of their mother's sister, Kara-ba-re's other wife. Kara-ba-rẽ is rescued from the cliff by a bird of prey (tikrēwati,~kwre “bird of prey”), and defeats the nine sons by striking the target of their competition with his own arrow. The phrase in example 5 comes at a point when Kara-ba-rē, having failed to persuade his nine sons to descend the ladder on the honey-yielding cliff, declares his power in these terms, and himself determines to climb down. A transcribed text, translation and commentary are given in Strickland, , Beliefs, pp. 152–93;Google ScholarStrickland, “Honey hunting by the Gurungs of Nepal”, Bee World, 1982, 63 (4), pp. 157–58CrossRefGoogle Scholar contains a summary of the tale and illustrations of current honey collecting practice.

⁴⁷ Höfer, AndràsTamang Ritual Texts, p. 40;Google ScholarAllen, “Sewala Puja Bintila Puja”, p. 251.Google Scholar

⁴⁸ Poucha, P. “Le vers tibétain” Archiv orientàlnì 1954, XXII, p. 585.Google Scholar

⁴⁹ Malkiel has suggested that pairs of words may be “ordered in accordance with a hierarchy of values inherent in the structure of a given society” (“Studies”, p. 145Google Scholar); while in an interesting account of Australian Walbiri aborigines' “antonymous” speech, K. Hale has suggested that many paired oppositions can be “fully understood only in reference to other aspects of Walbiri culture” (“Note”, p. 481Google Scholar). The latter phrasing could be questioned on the grounds that the idea of “full understanding” is unspecific, indeed inoperable; while the “sociological” idea of “inherent” values expressed by Malkiel appears somewhat schematic. It could be argued that collateral information can help to elucidate presumptions of native interpreters, but the process of selecting this material seems intuitive and arbitrary. Thus, this account does not claim to be exclusive.

⁵⁰ “Two Aspects of Language and Two Types of Aphasic Disturbances”, Fundamentals of Language, The Hague and Paris, 1956, II, p. 91.Google Scholar

⁵¹ The single performance tape-recorded was much abbreviated and partially obscured by incidental noise. The complex terms translated “proper” and “improper” here are kyawe tēngē and rime khakyura.

⁵² It will be noted that the ordering of the cardinal points E-S-W-N overrides the evaluations given by Padam Sing here, since the West precedes North in this list but not in value. But he did not think that this was important. Pignède cites a comparable enumeration from Mohoriya (Les Gurungs, p. 350 s.v. rimai khagyura prahwai Google Scholar) of which the manuscript translation is given and discussed elsewhere (Strickland, , Beliefs, p. 82 f.Google Scholar).

⁵³ Paired spirit names have been noted in ethnological accounts of other parts of the Himalayas and further afield; and the formal character of the names testifies to their metaphysical status. For the Thulung see Allen, N. J. “The Ritual Journey” Contributions to the Anthropology of Nepal edited by von Fürer-Haimendorf, C., Warminster, 1974. pp. 8–9;Google Scholar and “Approaches to Illness in the Nepalese Hills” Social Anthropology and Medicine edited by Loudon, J. B., London, 1976, pp. 532–33;Google Scholar for the Limbu see Sagant, P. “Tâmpuṅmâ, divinité limbu de la forêt” Objets et Mondes 1969, IX, 1, p. 114Google Scholar and fn. 12; for the Sherpa see Ortner, S. ‘The White-Black Ones’, Himalayan Anthropology edited by Fisher, J. L.. The Hague. 1978. p. 279;Google Scholar for the Nagas of Assam see von Fürer-Haimendorf, C.Return to the Naked Nagas, London, 1976, p. 224;Google Scholar for Hodgson's Bodo and Dhimal see his Aborigines, pp. 166–67.Google Scholar Comparison between Gurung myths from Siklis and Mohoriya suggests that the names can be interpreted to refer to single or dual numbers of beings: the Bird of Prey in example 3 becomes a male and female couple in a text collected by Pignéde, though it is not clear in which direction change would most likely have occurred.

⁵⁴ It is for this reason that, since Kara-ba-rē claims the ability to travel throughout the cosmos, as noted above, he might also be interpreted to make a claim to immortality.

⁵⁵ Compare the cosmic schemes given for Thakali by Gauchan, S. and Vinding, M. “The History of the Thakali according to the Thakali Tradition”. Kailash, 1977, V, 2, p. 163Google Scholar and fn. 46; for Tibetan material. Stein, Tibetan Civilization, pp. 203–04Google Scholar; and Haarh, E.The Yarluṅ Dynasty, Stockholm. 1969, pp. 135–36;Google Scholar for the Lepchas, Gorer, G.Himalayan Village, London, 1938. p. 223;Google Scholar and de Beauvoir Stocks, C. “Folk-lore and Customs of the Lap-chas” JASB, 1925, XXI, pp. 336–37.Google Scholar

For Thulung, Allen gives a brief account of an ancestral migration narrative as follows (“Tibet and the Thulung Rai” pp. 3–4Google Scholar):

“The Place of Origin is associated with a Primal Lake located to the South, perhaps at Bara Chatra in the Terai. but in some versions it is subterranean. The exit from it is barred by a ‘door’ which is opened by the sacrifice of a human or a bird. The four brothers who come forth are ancestors of the Rai subtribes, theThulung being the youngest. The brothers disperse the Thulung residing temporarily in various places including the Central Valley and Tarangan in Khumbu. They are finally led to the fertile site of their first permanent village (Mukli) by a wild boar: Ramli. the Founding Ancestor, ties a packet of ash to its tail and is thereby enabled to follow its trail.”

It should be added that, though Thulung seem to associate “down” with “South”, they do not attach great significance to the cardinal directions and differ in this respect from the Gurung: Allen, N. J. “The Vertical Dimension in Thulung Classification” Jnl Anthropological Society of Oxford 1972, 3, pp. 82–3.Google Scholar

⁵⁶ Thus, compare also Macdonald's report of an interesting Gurung myth in which the spirit or double of a shepherdess, becoming the aid of jhākri healers, takes on the form of a vulture when in Lhasa, a falcon in Nepal, a fish in the plains, and a wild duck elsewhere (Essays pp. 123–24Google Scholar).

⁵⁷ Stein, R. A.Tibetan Civilisation, London: Faber, 1972, pp. 43–4.Google Scholar

⁵⁸ Stein, R. A. “L'habitat, le monde et le corps humain en Extrême-Orient et en Haute-Asie” Journal Asiatique 1957, pp. 245, 47–8.Google Scholar

⁵⁹ Macfarlane, A. D. J.Resources and Population: A Study of the Gurungs of Nepal, Cambridge, 1976, p. 25.Google Scholar

⁶⁰ Höfer notes for gyṁgar dẹn (Tamang Ritual Texts, p. 46Google Scholar):

“Both meanings, ‘India’ (Tib. rgya-gar) and ‘womb’ were given. Since gyȧgar repeatedly occurs with ṭhuṅba [‘to be born’, ‘to originate’], one is tempted to see in this expression a reminiscence of the Tib. euphemism for ‘India’, namely ‘phags-’khruṅs. ‘the birth (place) of the holy ones’ ([S. C.] Das [A Tibetan-English Dictionary Delhi.] 1970 [(reprint).] p. 304).Google Scholar”

⁶¹ It can be argued that the scheme works also at the level of domestic use of space, though probably for historical reasons there are discrepancies between practice in different parts of Gurung country.

⁶² The personality of the deceased appears to be closely associated with the visual image of the person, with which the pla or “soul” is intimately linked. The pla is also connected with health and prosperity. Both tõ deities and mõ spirits are sometimes classed as tsẽ. with the qualification tsa-wa “consuming, eating” used to suggest danger or conflict; compare Tamang cen in Höfer, , Tamang Ritual Texts, p. 15;Google Scholar perhaps also Tibetan btsan in Tucci, G.Religion of Tibet, p. 164;Google Scholar and Pignéde, Les Gurungs, pp. 345, 350 s.v. chaẽ and caẽ.Google Scholar

⁶³ Mazaudon, “Tibeto-Burman Tonogenetics”. pp. 88–9;Google ScholarHale, E. A.Research, pp. 51–2.Google Scholar

⁶⁴ Emeneau, Language and Linguistic Area, pp. 3–4. 114–19;Google Scholar a somewhat general survey is published by Goral, D. R. in “Numeral Classifier Systems: A Southeast Asian Cross-Linguistic Analysis” Linguistics of the Tibeto-Burman Area 1978. IV, 1, pp. 1–72:Google Scholar the examples are drawn from Vietnamese, Burmese, Thai, Cambodian, Indonesian, Chinese, Lao and Lahu.

⁶⁵ Nacaskul, “Types of Elaboration”, p. 873;Google Scholar for some comparative material, see Omar, Asmah Haji “A Comparison of Malay and the Sarawak-Type Languages”, Sarawak Museum Jnl XXXII (53), 1983, pp. 277–80.Google Scholar