फेरि कोसीको पिरलो

फेरि कोसीको पिरलो
अघिल्लो वर्ष भदौ २ गते कोसी नदीले पूर्वी तटबन्ध भत्काई आफ्नो दिशा बदलेपछि नेपाल र भारतका लाखौं बासिन्दा घरवारविहीन भए । प्रलयकारी बाढीले सडक, सिँचाइ, खानेपानी, कृषिलगायत भौतिक पूर्वाधारलाई ठूलो क्षति पुर्‍यायो । त्यतिबेला तटबन्ध भत्कनुको मुख्य प्राविधिक कारण थियो- ँस्पर’ हरूको नियमित मर्मत-सम्भार हुन नसक्नु । नदीको बेगलाई कम गरी तटबन्धतर्फ सीधा धार रोक्ने संरचना स्पर हो ।
भदौमा फुटेको कोसीको भेललाई तीन महिनाअघि पुरानै स्थितिमा ल्याउन भएका प्रयत्न झन्डै असफल भएको थियो, जबकि वषर्ायाम सुरु नहुँदै आएको बाढीले अस्थायी बाँध भत्काएर ठूलो विपत्ति निम्तिने संकेत आइसकेको थियो । किनभने अनुमानभन्दा निकै छिटो पानी परेर कोसीमा बहाव बढ्न गई निर्माणाधीन स्परलाई पुनः धक्का पुर्‍याएको थियो । गत अपि्रलमै स्पर र तटबन्ध पुनर्निर्माण गरी कोसीलाई दुरुस्त राख्ने भारत सरकारको प्रतिबद्धता पनि असफल भयो । उसले तीनपटक म्याद थप गरेर कोसीको अन्तिम कार्य आगामी जुन १५ मा पूरा गर्ने नेपाल सरकारलाई जानकारी दिएको छ । भारतले यसरी जानकारी दिएको केही दिनभित्रै निर्माणाधीन केही संरचनालाई बाढीले बगायो । यसले गर्दा कोसीको विपत्ति दोहोरिने आशंका र भयमा स्थानीय बासिन्दा सुरक्षित स्थलतर्फ शरण लिनसमेत जानुपर्‍यो ।
नेपाल र भारतबीच थुप्रै संवेदनशील र जटिल सम्बन्धमध्ये कोसी सम्झौता प्रमुख हो । यही सम्झौताले कोसीका सम्पूर्ण संरचना तथा अवयवको सञ्चालन, नियन्त्रणमात्र नभई मर्मत-सम्भारको जिम्मा पनि भारतलाई दिएको छ । गत भदौमा बाढीको विपत्ति भोगिसकेपछि पनि भारतले पूर्वनिर्धारित समयमा मर्मत तथा पुनर्निर्माण कार्य गर्न सकेन । उसले त्यतिबेला हिउँदभित्रै युद्धस्तरमा कोसी पुनर्निर्माण गर्ने घोषणा गरेको थियो । कोसीजस्तो संवेदनशील मामिलामा समेत भारत संवेदनशील नभएको हो कि भन्ने आशंकालाई उसका गतिविधिले मलजल गरेका छन् । यसपालि गत अपि्रलभित्रै पूरा हुनुपर्ने मर्मत कार्य भारतीय ठेकेदार र सरकारबीचको विवादले गर्दा ढिलो भयो । एक दिन पनि नबिराई चौबीसै घन्टा काम गर्नुपर्नेमा सरकार र ठेकेदार कम्पनीबीच निकै दिन जुहारी चलेको थियो । यसबाहेक नेपालका कोसी पीडितको आन्दोलन, दिनहुँजसो हुने बन्द, हडताल, चक्काजामले पनि निर्धारित समयभित्र काम पूरा नहुनमा मद्दत गरेको देखिन्छ । भारतको विहार प्रान्तका जलस्र्रोतमन्त्रीले कोसीको भ्रमणताका नेपाल सरकारबाट सुरक्षा नपाएको गुनासोसमेत गरेका थिए । कोसी बाढीले आफ्नै उठीवास भइरहेका बेला पीडितहरूले विभिन्न माग राखी मर्मत कार्यमा समेत अवरोध पुर्‍याए । यी र यस्ता कृत्यप्रति नेपाल र भारत दुवैतर्फका सरकारी संयन्त्र संवेदनशील देखिएनन् ।
कोसीको समस्या कुनै एक वर्ष वा यामको होइन । आगामी दिनमा गत सालको जस्तो घटना दोहोरिन नदिन सरकारले सतर्कता अपनाउनुपर्छ । ँभारत सरकारको जिम्मा हो’ भनेर नेपाल चुप लागेर बस्ने हो भने यस्ता घटना सामान्य बन्न थाल्नेछन् । कोसीका सूचना आदानप्रदान गरी छिटोभन्दा छिटो समस्या समाधान गर्न दुई देशबीच सम्पर्क अधिकारीको व्यवस्था छ । तर कोसीसम्बन्धी जानकारी वा ज्ञान भएकालाई नियुक्ति गर्नुभन्दा आफ्ना कार्यकर्ता भर्ती गर्ने चलनले दुई सरकारबीच प्रभावकारी संवाद हुनसकेको छैन ।
कोसीको घटना दोहोरिन नदिन नियमितरूपमा स्पर तथा तटबन्धहरूको मर्मत सम्भार गर्नुपर्छ । यसका लागि नेपाल-भारत संयुक्त कोसी तथा गण्डक समितिको बैठक नियमित बसेर प्राविधिक टुंगो लगाउने प्रचलनलाई निरन्तरता दिन आवश्यक छ । आगो लागिसकेपछि कुवा खन्ने नभई आगो लाग्नै नदिने व्यवस्था हुनुपर्छ । मर्मत सम्भारको जिम्मा भारतको भएकाले उसले पनि बीच-बीचमा आइपरेका समस्या तथा अड्चनबारे नेपाललाई समयमै अवगत गराउन जरुरी छ । कोसी क्षेत्रमा दिनहुँजसो हुने बन्द-हडतालले पनि यसलाई प्रभावित पार्दै आएको छ । स्थानीय प्रशासनले कोसी संरचनाको मर्मतमा संलग्न प्राविधिकलाई जसरी भए पनि सुरक्षा दिनैपर्छ । यति गर्न सके मात्र पनि विगतको विपत्ति दोहोरिने छैन ।

थुलुंग राइ केशव कुमार (jasper)

Himalayan Linguistics: Archive No।११

Thulung RaiAimée LahaussoisLACITO/CNRS, Villejuif, France1. IntroductionThulung Rai is a Kiranti language spoken by about a thousand people in Eastern Nepal.The dialect represented here is that of Mukli. Mukli is near the confluence of the Solu Khola andDudh Khosi rivers, and the village is spread out over a hillside at an average elevation of 1500 m.The language is in a situation of what I estimate to be ‘intense’ contact (using Thomasonand Kaufman’s standards, 1988: 83) with Nepali. Lexical borrowing is widespread, and there isalso a good deal of structural borrowing. This is due to the fact that most speakers are bilingualwith Nepali. There do not appear to be any fluent speakers younger than 20.N. Allen spent several years in the village of Mukli in the early 1970’s and produced aSketch of Thulung Grammar (1975). This is a particularly valuable document in light of thechanges which have come about in the last thirty years. Ebert has also incorporated Thulung datagathered from Allen’s texts into her comparative analyses of Kiranti languages (such as 1994).2. PhonologyInitial consonants (in native words)unvoiced unvoiced voiced voiced nasalaspirated aspiratedglottal çvelar k kh g «affricate ts tsh dz dzhdental t th d dh nretroflex £labial p ph b bh mcontinuants h, s, l, rsemi-vowels j, wFinal consonantsk ¢ p « n m l rVowelsi y e a ¹ o » uVowel lengthAllen marks length distinctions in vowels, but I do not consider vowel length tobe distinctive in the modern language. While some speakers sometimes producedminimal pairs distinguishing vowel length in some words, these length distinctionsLahaussois: Thulung Rai2were not found to be consistent. This may be the influence of Nepali, which does nothave long vowels.Diphthongsai, au, eu, ¹i, ¹u, »i, uiFalling diphthongs are considered to be glide plus vowel combinations, as they can occurword-initially (while rising diphthongs cannot).In the syllable cannon below, diphthongs are represented by the single V, but never occurwith a preceding glide.ToneAllen reports finding certain pairs of words for which “the most obvious differencewas that in... one member was pronounced faster and in a more fortis manner.”(1975: 32) He describes these words “as having tense tone” and the other member of thepair as having “lax tone.” He goes on to point out that “the great majority of Thulungwords are not members of such minimal pairs and sounded equally natural... whetherpronounced tense or lax.” (ibid.) I found no examples of tonal contrast, even in the pairslisted in Allen. Presumably the traces of tonal contrast disappeared under pressurefrom Nepali.Syllable cannon(Ci) (liquid) (glide) V (Cf)3. Typological summarySalient typological features of Thulung include the following:• Basic S O V order.• Nouns can be inflected for dual or plural number.• Morphological ergative split, with ergative marking for second plural, third persons andother NPs.• Verbs are inflected for person and number, with up to two arguments marked.4. Basic sentences4.1. IntransitivesA basic intransitive sentence has an unmarked subject and a verb agreeing with it in personand number.1. a-wotsy pakha l¹-mri1POSS-husband outside go-3sP:PST‘My husband went outside.’There are several types of predication with the copula, which also agrees with the subject.Equative predication2. m» mytsy a-pap-ku «ops» buthat man 1POSS-father-GEN friend be:3s‘That man is my father’s friend.’Himalayan Linguistics: Archive No.13Locational predication3. gumi neb-ra bu-mi3sP house-LOC be-3sP‘She is at home.’Existential predication4. dher prots»sama bumany Rai caste be:3s‘There are many Rai castes.’Possessive predication5. go-nu« £okpudzahanbu1s-COM big family be:3s‘I have a big family.’4.2 TransitivesThulung has ergative marking in transitive sentences, but with a split conditioned by personand number: agents are nominative, with no overt marking, if they are from the class 1s, 1de,1di, 1pe, 1pi, 2s, 2sP, 2d, and they are marked with ergative -ka if they are 2p, 3s, 3sP, 3d, 3p andany other NP.There is also primary object marking: -lai marks the animate object of monotransitives(always when human, sometimes when non-human) and the recipient of ditransitives.Monotransitive6. go mag djo-uto1s mug drop-1s/3s:PST‘I dropped the mug.’7. go mam-lai ts»m dwak-pu1s mother-DAT much like-1s/3s'I like my mother a lot.'8. i-lwak-ka i-mam-lai khl»i2POSS-y.sibling-ERG 2POSS-mother-DAThelp:3s/3s'Your younger sibling helps your mother.’Ditransitive9. mam-ka u-ts»-tsi-lai po-mu-¢hok gwak-tymother-ERG 3POSS-children-DU-DAT eat-INF-stuff give-3s/3s:PST'The mother gave her two children food.'10. go a-mam-lai ts» gwak-tomi1s 1POSS-mother-DATchild give-1s/3sP:PST'I gave the child to my mother.'Lahaussois: Thulung Rai44.3 Experiencer constructionsExperiencer constructions are those for which the experiencer is expressed with dativemarking and the verb has 3s (agent) agreement.11. go-lai kwara ly-ra1s-DAT thirst feel-3s:PSTI am thirsty12. go-lai tsoktsho b¹k-ta1s-DAT anger arise-3s:PST‘I am angry.’ (‘Anger arose in me.’)13. go-lai bira «im ly-ra1s-DAT leech fear feel-3s:PST‘I am afraid of leeches.’5. The Noun Phrase5.1. Word and affix orderThulung has the following basic word order in the noun phrase:DEM NUM Classifier Adjective poss-N-caseN-GENNouns can be marked for number (5.4), for case (5.8), and pragmatic status (5.9).Modifiers can also function as nouns, and can be any of the following categories:demonstrativenumeral (+ classifier)adjectiverelative clausepossessive pronoun, possessor-GENplace-LOC-NOM5.2. PronounsThulung has a fairly complete set of pronouns, with dual forms, an inclusive/exclusivedistinction, and polite forms (marked 2sP and 3sP for 2 singular polite and 3 singular polite respectively.)1s go, 1de gutsuku,1di gutsi,1pe guku,1pi gui2s gana, 2sP gani, 2d gatsi, 2p ganimim3s gu, 3sP gumi, 3d gutsi, 3p gumimimAllen noted that polite forms were sometimes used, whereas this now seems to be a fairlywell-established pattern. The polite forms are based on what were, in Allen’s time, plural forms,and the plurals have been reinforced by the plural morpheme.5.3. DemonstrativesThulung has proximal, o/oram, and distal, m»/m»ram (also meram), demonstratives.The demonstratives can also be used for third person referents, optionally with dual and pluralforms formed with the corresponding suffixes (-tsip and –mim respectively).Himalayan Linguistics: Archive No.155.4. Numerals, number5.4.1. Numerals and classifiersNative numerals go up to ‘three’ for most speakers, and are ko, n¹, su (‘1’, ‘2’, ‘3’ respectively).The etymon for ‘four’ (blu) appears in some temporal expressions (bluha££a ‘in fouryears’, blunem ‘four days ago’, blu ‘in four years’, bluna ‘four years ago’) but is not used withthe remaining classifiers.The only classifier seen with any frequency is the generic classifier -le, and this appears incombination with the three remaining native numbers. The classifier is not used when what isbeing counted are periods of time, such as year, week, day, in which case numerals occur alone.The exception to this is ko-lem, n¹-lem, su-lem, used for ‘one day’, ‘two days’ and ‘three days’respectively. Another somewhat productive classifier is –phe used for counting generally roundobjects, such as coins, bananas, bread.5.4.2. NumberThe plural marker, -mim, is used to pluralize nouns, and its use with pronouns appears tobe a new phenomenon, calqued on the Nepali. The distribution of the pluralizer -mim is as follows(and presumably the dual patterns paralelly, although I have fewer data on its distribution):it is an optional plural marker, but there is a scale in the frequency of appearance with certainnoun classes. It appears most frequently with kin terms, where its use is fairly consistent, althoughexamples are found where it is ommitted. It is optional but generally used with non-kinhumans, and considerably less frequent with non-human animates, such as herd animals. It tendsnot to appear very frequently with inanimate nouns (but this depends on the speaker). The samedistribution applies to the dual marker -tsip.5.5. AdjectivesAdjectives in Thulung are defined by requiring no genitive or nominalizing material toappear attributively. There is a small class of lexical adjectives, which are given in the followinglist, while the biggest class of adjectives are deverbal, seen at the end of this section.khrekhreja ‘bumpy, rough’pl¹pl¹ja ‘smooth’£okpu ‘big’jakke ‘small’«ats» ‘old’malomtsʼ ‘young’happa ‘much’The color terms are based on a reduplicative pattern, which seems to be uniquely restrictedto colors.lalam ‘red’gigim ‘green’kekem‘black’çoçom ‘yellow’bubum ‘white’nunum ‘blue’Lahaussois: Thulung Rai6The following adjectives are deverbal, being non-past participial forms, for which not allinput verbs are synchronically recognizable.jepa ‘high’ <jemsimu ‘to stand’dhypa ‘long’te¢pa ‘smart’tsapa ‘strong’ <tsammu ‘be able’dz»pa ‘good’tsh¹kpa ‘cold’dzalpa ‘hot’tsisi¢pa ‘wet’sa¢pa ‘dry’ <samu, ‘to dry’khepa ‘bitter’ <khemu, ‘to be bitter’br¹pa ‘tasty’ <br¹mu, ‘to have taste’lempa ‘sweet’ <lemmu, ‘to lick’dukpa‘spicy’dzyrpa ‘sour’5.6. Possession5.6.1. Pronominal possessionEach person has a generic possessive prefix.1 This prefix is unspecified for number, andcan be used for singular, dual and plural referents. There is also a possessive pronoun for eachperson, which can be used prenominally or as an independent nominal. They appear to be combinationsof the prefix and a nominalizing element. Like the prefixes, the forms given in the tablebelow are unspecificied for number.1 person a- ama2 person i- ima3 person u- umaThese prefixes and possessive pronouns can also be specified for number with the additionof a suffix: atsi-/ atsima (1de), aki-/ akima (1pe), itsi-/ itsima (1di), iki-/ ikima (1pi), itsi-/itsima (2d), ini-/ inima (2p or 2sP), utsi-/ utsima (3d), uni-/ unima (3p or 3sP).The prefixes are most often seen with first person prefixes and pronouns, but occasionallywith second and third persons as well.5.6.2. Genitive markingDependent-GEN Head is the structure of the possessed noun phrase.The genitive marker is interchangeably –ku or –kam.14. je-ku r¹«clothes-GEN color‘the colour of the clothes’1 I believe these to be prefixes, but have found one example where it behaved like a clitic:a-£okpu tsʼ ‘my big child’. Quite possibly £okpu tsʼ is interpreted by this speaker to be a single lexical element,corresponding to English ‘eldest’.Himalayan Linguistics: Archive No.1715. gai-kam gunucow-GEN inside‘the inside of the cow’When the head noun is a body part or kin term, the head can optionally be marked withthe possessive pronoun in addition to the genitive marker on the dependent.16. bwa-kam u-bwipig-GEN 3POSS-head‘the pig’s head’17. po-ku u-s»chicken-GEN3POSS-meat‘chicken meat’There is also a special subset of possessed NPs: possessed time words are marked differently,and take -m or -mim (-mim postconsonantally).18. nemtha-m dzamevening-NOM rice‘the evening meal’19. anep-mim dintoday-NOM2day‘today’s days (these days)’5.7. Relative clausesThere are two means of relativizing, one construction being based on a finite verb followedby a nominalizer, and the other on a participial form.5.7.1. -m/-mim relativizationThis type is based on a finite verb. DeLancey (2002) has noted that “relativization in TBis a subspecies of clausal nominalization” and this applies to Thulung, where the relative markersare in fact nominalizers. The distribution of these nominalizers in relativization is as follows: -mis used for general relativization, in all cases (past and non-past) except for with mono-syllabicverbs; -mim is only used for the non-past, and always for mono-syllabic verbs.20. [go khok-to-m/*-mim] dzam br¹pa bai-ra1s cook-1s/3s:PST-NOM/NOM2 rice good be-3s:PST'The food I cooked was good.'21. [go dwak-pu-m/-mim] kitap gani-lai gwa-gwa1s like-1s/3s-NOM/NOM2 book 2sP-DAT give-givedwa.like:3s/3s'I want to give you a book I like.'Lahaussois: Thulung Rai822. [gui si-i-mim /*-m] din1pi die-1pi-NOM2/NOM day'the day we die'The original distribution (according to Allen) appears to be based on tense, with –mim fornon-past and –m for past, but the distribution is now partly phonologically conditioned.5.7.2. -pa/-ma relativizationThis type is based on a participle, either non-past (-pa) or past (-ma). The main semanticdifference is that in a participial relative, the participants need not be specified. (In -m/-mim relativization,the verb agreement shows who the participants are.)23. [nem bane-pa] a-lwakhouse make-Npst.PRT 1POSS-y.sibling‘my brother who builds houses’24. khok-ma dzamcook-Pst.PRT rice‘cooked rice (by anyone)’5.8. Case markersErgative –ka (exs 8, 9)Instrumental –ka (see also causal, 7.4.)25. ganimim-ka m» mi-h¹npa betho-ka mi-dzhak-ni2p-ERG that NEG-sharp knife-INSTR NEG-cut-2p/3s‘You should not cut wood with that blunt knife.’Dative –lai (exs 7, 9, 11)Genitive -ku, -kam (see 5.6.2.)Locative -ra, but also -nu (same level), -la (high), -jy (low)Locative –ra (see also purposive, 7.3.)26. go iskul-ra angredzi si-pu1s school-LOC English teach-1s/3s‘I teach English at the school.’Level locative -nu27. Deusa-nu-m Darim Popnar ra-ma dadzjuDeusa-levLOC-NOM Darim Popnar call-Pst.PRT o.brotherbhai noktsho get-tsi.y.brother shaman come.up-3d:PST‘A pair of brother magicians from across at Deusa, called Darim and Popnar, came up.’High locative –la28. Lukla ra-ma ¢au-laHimalayan Linguistics: Archive No.19Lukla call-Pst.PRT place-hiLOC‘In the place up from here called Lukla’Low locative -jy29. gumi-ka bloku-jy-m ku khe-saka pem-thal-miri3p-ERG river-loLOC-NOM water bring-AC drink-3p-HAB-3p/3s:PST‘They brought the water from down in the river and were drinking it.’Comitative –num30. go pap-num bu-«u1s father-COM be-1s‘I live with my father.’Ablative -lam, -la«ka31. ba-lam/la«ka rok-na-mwhere-ABL come-2s:PST-NOM‘Where have you come from?’5.9. Discourse markers5.9.1. Topic ne, Focus reThe topic marker signals given information, in contrast to the focus marker which marksnew information. They appear after whatever case marker may be required by the grammaticalcontext.The contrast between the two is best seen in examples where they both appear.32. ¹sin£a ne mytsy re bu-mi tshahere TOP man FOC be-3p be‘There’s a man here!’ (indicating surprise, as noone was present before)33. go ne bia re be-uto1s TOP marriage FOC do-1s/3s:PST‘I got married.’ (said to someone who hasn’t seen me since)The topic and focus markers need not appear in the same sentence, and there are numerousexamples of their occurring independently of one another.5.9.2. Contrast tsahiThe contrast marker is borrowed from Nepali. It is used to set up a contrast between severalentities, in which case all are marked with tsahi, or between a marked entity and a referencepoint obvious from context, as in the following.34. khlea-ka tsahi m» b¹¢¹l-gunu u-bwi phik-y.dog-ERG CONTR that bottle-inside 3POSS-head stick.in-3s/3s:PST‘The dog stuck his head inside the bottle.’The entity with which khlea is contrasted is the other participants in the story, who are carryingon other activities.Lahaussois: Thulung Rai105.9.3. Emphasis marker -«aThe emphasis marker can be affixed to adverbs and adjectives as well as noun phrases,and generally emphasizes the marked element. For NPs the result is similar to the focus marker,highlighting the new information.35. oram je-ku r¹« go-«a tshen-to-mthis clothes-GEN colour 1s-EMPH choose-1s/3s:PST-NOM‘I am the one who chose the colour of these clothes.’6. The Verb Phrase6.1. Word and affix orderAdverb negation-verb-directional-aspectivizer-agreement/non-finite-hearsay6.2. AdverbsAdverbs are very limited in number, and express manner (wakha ‘slowly’, dala ‘fast’) andextent (ts»m ‘much’, kits» ‘little’).There is another means of expressing manner on a verb: a limited number of verbs cantake a particular prefix on the root, to express that the action is carried out suddenly. These adverbialprefixes are not productive, and are limited to the particular verb they occur with. Theymay have their origins in some kind of onomatopeic expression.Examples:bu«-swa-mu ‘to flee rapidly, suddenly’, cf swa-mu ‘to flee (vi)’la-than-mu ‘to pull out quickly’, cf than-mu ‘to pull out (vt)’puli-tshar-mu ‘to make something fall quickly’, cf tshar-mu ‘to make fall (vt)’tsha-gro-mu ‘to throw immediately’, cf gro-mu ‘to throw (vt)’ts»-krwa-mu ‘to push in violently, suddenly’, cf krwa-mu ‘to push into (vt)’leb-bo-mu ‘to throw someone down violently during a fight’tho«-kon-mu ‘to drench all of a sudden’, cf kon-mu ‘to make wet (vt)’6.3. NegationNegation is a prefix to the verb. It has the form mi-.In order to negate past forms of verbs, extra material is needed: the irrrealis marker is suffixedto the finite non-past verb form.36. mi-lwa-ku-wa vs. lwat-tokoNEG-see-1pe/3s-IRR see-1pe/3s:PSTWe did not see it. We saw it.37. mi-l¹-«-wa vs l¹-«roNEG-go-1s-IRR go-1s:PSTI did not go. I went.6.4. VerbThe verb has a monosyllabic root.A multisyllabic stem can occur when the root is augmented by an aspectivizer (this is acategory of suffixes which appear to be derived from full verbs and which cover, in additional toHimalayan Linguistics: Archive No.111aspectual shadings, valence increasing and decreasing morphology—see 6.7.). The root can alsobe augmented into a bisyllabic stem by suffixing a directional. This is a set of suffixes whichcombine with the verb romu ‘to come’, specifying the vertical dimension involved. In some casesthese are compounds, and the suffix identifiably comes form a motion verb, in other cases, theorigin of the suffix is lost.The compounds formed are limited to the following list:rojomu ‘to arrive here from up above’ (jomu ‘to come down’)rogemu ‘to arrive here from down below’ (gemu ‘to come up’)rothimu ‘to arrive there (from any direction)’rophamu ‘to arrive here (from any direction)’robimu ‘to arrive here (from any direction)’ (bimu ‘to come’)6.5.Agreement6.5.1. Agreement paradigmsVerbs have a suffixal agreement slot, and the agreement suffixes can encode up to twoparticipants. The suffixes are portmanteau morphemes that also encode tense. For intransitives,the only participant is encoded, for monotransitives, the agent and patient, and for ditransitives,the agent and the recipient.There are however cases in which it looks like it is the primary object is unmarked on theverb, but this appears to be an omission of the distinctive number marker rather than the markingof a single argument. The following example shows that even within the same sentence, verbswith the same arguments can show a shift in marking. (The relevant suffixes are marked in bold:the action clearly refers to a 3s agent and 3d patient, yet some verbs have 3s patient suffixes.)38. mesimma pet-tsi-m patshi th¹k-tytsi, l¹-ry-mathen eat-3d/3s:PST-NOM after hide-3s/3d:PST take-3s/3s:PST-ASth¹k-ty-ma phoka-ka «jak-tytsi-çehide-3s/3s:PST-AS ash-INSTR cover-3s/3d:PST-HS‘Then after they ate, she hid them, she took [them] and hid [them] and covered them inashes.’Similarly, the verbal marking in the following example makes it seem like it is the patient,and not the recipient, which is used for verb agreement, but it appears to be the same type of phenomenonas in 39.39. mam-ka u-ts»-tsi-lai po-mu-¢hok gwak-tymother-ERG 3POSS-children-DU-DAT eat-INF-stuff give-3s/3s:PST'The mother gave her two children food.'The language has different agreement paradigms for past and non-past.Lahaussois: Thulung Rai12Table 1: Intransitive personal endings, non-past and pastNonpastPast1s -«u -«oro1de -tsuku -tsoko1di -tsi -tsi1pe -ku -toko1pi -£i -£i2s -na -na2d -tsi -tsi2sP,2p-ni -ni3s -£a3d -tsi -tsi3sP,3p-mi -miriTable 2: Non-past intransitive personal endingsPat►Ag▼1s 1de 1di 1pe 1pi 2s 2d 2sP,2p3s 3d 3sP,3p1s -ni -nitsi -nini -u -utsi -umi1de -tsuku1di -tsi1pe -ku1pi -i -itsi -i2s -«i -«itsi /-tsiki-ki -na2d -«itsi -tsiki -tsiki -tsi2sP,2p-«ini -kini -kini -ni3s -«i -«itsi/-tsiki-tsiki -ki -sa -na -na -nimi -y -ytsi -ymi3d -«itsi -«itsi/-tsiki-sa -sa/-kini-sami -natsi -natsi -nimi/-nitsi-ytsi -ytsi -ytsi3sP,3p-«imi -tsiki -sami -sami/-kimi-sami -nami -nitsi/-natsimi-nimi -mi -ytsi/-mi-miHimalayan Linguistics: Archive No.113Table 3: Past transitive personal endingsPat►Ag▼1s 1de 1di 1pe 1pi 2s 2d 2sP,2p3s 3d 3sP,3p1s -ni -nitsi -nini -to -totsi -tomi1de -tsoko1di -tsi1pe -toko1pi -li2s -«iri -tsiki -na -natsi -nami2d -«iritsi -tsiki -tsi2sP,2p-«irini -tiki/-tsiki-tiki-ni -nitsi -nimi3s -«iri -tsiki -tsiki/-sa££atsi-tiki -sa££a -na -natsi -ly -lytsi -lymi3d -«iritsi -tsiki -sa££atsi-tiki -sa££a -natsi -natsi -lytsi -lytsi -lymi3sP,3p-«irimi -tsikimi -sa££ami-tikimi -sa££ami -nami -natsimi-nimi-miri -miri -miriThus the verb jal-mu ‘to hit’ has a verb stem jal-. For a 1s agent and 2s patient, the formis jal-ni (the past and non-past are identical for this combination.)One thing that becomes apparent upon looking at the transitive suffixes is that the third person isvirtually unmarked: a paradigm with a 3s patient is practically the same as an intransitive paradigm.6.5.2. Verb stem alternationsThulung has a system of alternating verb stems, with three main verb classes.Lahaussois: Thulung Rai14Class I Class II Class IIINon-alternating verbsThese are verbs with roots endingin -l,-r, and also some of theverbs whose roots end in -m andin -n (although these are divided,with some of the -m final rootsbeing subclass IIb, and some ofthe -n final being subclass IIc.)Examples of some of the verbs inClass I are mal-mu ‘to search’, kurmu‘to carry’, plym-mu ‘to soak’,mun-mu ‘to establish’.Alternating stemsVerbs in class II divide into threesubclasses depending on the alternationthey show. The tablebelow lists the two stem endingsfor each subclass, with stem 1appearing in imperative formsand stem 2 in infinitive forms.12Subclass IIa-kOSubclass IIb-pmSubclass IIc-£O/-nAlternations between stem 1 andstem 2 occur according to a welldefinedpattern, depending onthe person and number of theagent. The distribution is set outin table 4.Alternating stemsVerbs in class III divide up untilthree subclasses as well, but inthis case there are three alternativestems per verb type. Stem 1appears for certain person combinationsas well as in the imperative,stem 2 appears for otherperson combinations and for theinfinitive, and stem 3 appears formost past forms (Stem 3 is stem 2with an additional past-tensemorpheme -t, which assimilatesto -n before -n)123Subclass IIIa-sO-tSubclass IIIb-iO-tSubclass IIIc-«O-tAlternations occur but in muchmore reduced circumstances thanfor Class II verbs, and the distributionof the three stems is laidout in table 5 below.For verb classes with alternating stems, it seems clear from comparative evidence thatthese alternations were originally phonologically based, with significantly different verbal suffixesfor non-past and past forms. The phonological conditioning is no longer relevant in determiningthe alternations, and is now set out according to person/number combinations.Himalayan Linguistics: Archive No.115Table 4: Class II alternation patterntransitive intransitiveNonpastPastNonpastPas1s 1 1 2 t21de2 1 2 11di 2 1 2 11pe2 1 2 11pi 1 1 1 12s 2 1 2 12d 2 1 2 12p 2 1 2 13s 1 1 2 13d 2 1 2 13p 2 2~1 2 2Table 5: Class III alternation patterntransitive intransitiveNonpastPastNonpastPas1s 2 2 2 t21de2 3 2 31di 2 3 2 31pe2 3 2 31pi 1 1 1 12s 2 3 2 32d 2 3 2 32p 2 3 2 33s 1 1 2 33d 2 3 2 33p 2 2 2 2Lahaussois: Thulung Rai16Table 6: Examples of verb paradigms illustrating Class II verbs (shading represents stem1)Subclass IIa Subclass IIb Subclass IIcrja-mu ‘to write’ rem-mu ‘to look’ sen-mu ‘to kill’Non-past Past Non-past Past Non-past Past1s/3s rjak-pu rjak-to rep-u rep-to se¢-pu set-to1de/3s rja-tsuku rjak-tsoko remtsukureptsokosentsukusettsoko1di/3s rja-tsi rjak-tsi rem-tsi rep-tsi sen-tsi set-tsi1pe/3s rja-ku rjak-toko rem-ku rep-ku sen-ku set-toko1pi/3s rjak-i rjak-ti rep-i rep-£i ser-i se£-£i2s/3s rja-na rjak-na rem-na rep-na sen-na se£-na2d/3s rja-tsi rjak-tsi rem-tsi rep-tsi sen-tsi set-tsi2p/3s rja-ni rjak-ni rem-ni rep-ni sen-ni se£-ni3s/3s rjak-y rjak-ty rep-y rep-£y ser-y se£-£y3d/3s rja-tsi rjak-tsi rem-tsi rep-tsi sen-tsi set-tsi3p/3s rja-mi rja-mri rem-mi rep-miri sen-mi se¢-miriTable 7: Examples of verb paradigms illustrating Class III verbs (light shading is Stem 3,dark shading is stem 1, and no shading is stem 2)Subclass IIIa Subclass IIIb Subclass IIIclwa-mu ‘to see’ tsa-mu ‘to burn’ Du-mu ‘to drink’lwamu, ‘tosee’Non-past Past NonpastPast NonpastPast1s/3s la-u la-uto tsa-u tsa-uto £u-u £u-uto1de/3s lwatsukulwattsokotsa-tsuku tsat-tsoko £utsuku£uttsoko1di/3s lwa-tsi lwat-tsi tsa-tsi tsat-tsi £u-tsi £ut-tsi1pe/3s lwa-ku lwat-toko tsa-ku tsat-toko £u-ku £ut-toko1pi/3s lwas-i lwas-ti tsa-i tsai-ri £u«-i £u«-ri2s/3s lwa-na lwan-na tsa-na tsan-na £u-na £un-na2d/3s lwa-tsi lwat-tsi tsa-tsi tsat-tsi £u-tsi £ut-tsi2p/3s lwa-ni lwan-ni tsa-ni tsan-ni £u-ni £un-ni3s/3s lwas-y lwas-ty tsa-y tsai-ry ;;u«-y ;;u«-ry3d/3s lwa-tsi lwat-tsi tsa-tsi tsat-tsi £u-tsi £ut-tsi3p/3s lwa-mi lwa-mri tsa-mi tsa-mri £u-mi £u-mri6.6. Mood6.6.1.irrealisThe irrealis is either -wa or -ja, with the following conditioning: -wa after back vowels, -ja after front vowels. It is also seen assimilating to a preceding bilabial nasal resulting in -ba.The irrealis appears in the following contexts:Himalayan Linguistics: Archive No.1171) in some conditional clauses: either in both clauses (eg. 40), or in the protasis marked with -mala (eg 41).40. m» nem £i-s¹¢-pu-wa-malathat day leave-DEF-1s/3s-IRR-CONDdzhjal-la«ka b¹tse-«a mi-dyp-sa-wawindow-ABL survive-EMPH NEG-become-2IMP-IRR‘If I had left her alone that day, she would not have survived the window (iewhen she climbed out and fell)’41. mi-tsap-sy-ja-mala kits» phul kam-mu basiNEG-able-3s-IRR-COND little flour add-INF OBL‘If one is not able [to guess the right amount], one must add a little flour.’2) in the past tense forms of negative verbs (as mentioned in 6.3.)6.6.2.imperativeThe imperative is formed with the suffixes -a for 2s, -tsi for 2d, -ni for 2p.jal-mu ‘to hit’ 2s jal-a 2d jal-tsi 2p jal-niVerbs which have alternating stems use stem 1 with the imperative suffixes.ro-mu ‘to come’ (Subclass IIa) 2s rok-a 2d rok-tsi 2p rok-nibre-mu ‘to buy’ (Subclass IIc) 2s brer-a 2d bret-tsi 2p bre£-niThe exception to this is for verbs from subclass IIb which use a suffix -ra for the 2s imperativeform.rem-mu ‘to see’ (Subclass IIb) 2s rep-ra 2p rep-niFor verbs from class III, stem 1 is only used for the 2s imperative form, and stem 2 for thedual and plural forms.£u-mu ‘to drink’ (Subclass IIIc) 2s £u«-a 2d £u-tsi 2p £u-niThe imperative conjugation can also encode a second participant.The verb in the following example is rem-ben-mu (look-CAU-INF)i-k¹l rem-be¢-«i / rem-be¢-«itsi / rem-be¢-«ini‘Show me your face.’The participants encoded are, respectively, 2s, 2d, 2p as agent and 1s as recipient.6.7. Aspect6.7.1. “aspectivizers”These are suffixed onto the verb root to convey aspectual shadings. (There is a piece ofagreement morphology which appears between the aspectivizer and the verb stem for certain persons;this point needs further study.)Lahaussois: Thulung Rai18Habitual -thal42. go athal iskul l¹«-thal-«uI nowadays school go-1s-HAB-1s‘I go to school regularly these days.’Stative -ta (‘to keep on doing X’)43. g¹n-ta-si-musit-STA-DET-INF (the detransitivizer is also in this example)‘to keep sitting’Definitive -so (perfective + permanence of action)44. gu-ka surti £u-mu £i-so£-£y3s-ERG tobacco drink-INF stop-DEF-3s/3s:PST‘He gave up smoking for good.’Ponent -dz»l (perfective + actions carried out ahead of time)45. go beno-lai ghas ph¹l-dz»l-to-m bu1s ox-DAT grass cut-PON-1s/3s:PST-NOM be:3s‘I have cut grass for the ox.’Resultative -le (perfective + focus on end-point of action)46. go mi-g¹k-thi-«a a-bep si-m-le-mri1s NEG-be.born-NEXP-EMPH 1POSS-grandfather die-3p-RES-3p:PST‘My grandfather was dead before I was born.’6.7.2. perfect construction : nominalized finite verb followed by an inflected copula47. bante l¹-mri-m bu-miwhere go-3p:PST-NOM be-3p‘Where have they gone?’6.7.3. progressive construction: verb root + -sa;;a followed by an inflected copula48. anebdika pare-pa-ka ¢au-«a mi-lwa-sa«a bu-minowadays study-Npst.PRT-ERG place-EMPH NEG-find-PROG be-3p‘Nowadays people who study are not finding jobs.’6.8. Non-finite forms6.8.1. “Infinitive”The infinitive suffix is –mu.The very notion of infinitive is challenged by the fact that these forms can sometimesshow agreement with the patient/recipient, in the context of an expression of obligation or negativeobligation.For patients/recipients which have number marking (ie duals, plurals, and polite referents),the infinitive suffix can be augmented to make reference to number. In other words, patients/recipients corresponding to the following person.number combinations can bring about alternativeinfinitive marking as follows:Himalayan Linguistics: Archive No.1191pi, 2sP, 2p, 3sP, 3p: infinitive form in –mmi (as an alternative to the form in –mu)1de, 1di, 2d, 3d : infinitive form in –mutsi (as an alternative to the form in –mu)The context on these infinitives forms if somewhat limited though. While they can beused in obligation constructions, they cannot be used in complement constructions that normallytake infinitives.49. go a-pap-lai lwa-mu la-u1s 1POSS-father-DAT see-INF see-1s‘I get to see my father.’50. *go a-pap-lai lwa-mmi la-u1s 1POSS-father-DAT see-PL.INF see-1s51. gumi-lai dzam £e-mmi basi3p-DAT rice feed-PL.INF OBL‘(I) must feed them rice.’52. *go m»-mim-lai dzam £e-mmi khap-u1s that-PLU-DAT rice feed-PL.INF be.about-1s‘I am about to feed them rice.’53. go m»l-lai dzam £e-mu khap-uI that-DAT rice feed-INF be.about-1s‘I am about to feed him rice.’6.8.2. ParticiplesThere are two participles in Thulung, the past and the non-past. The past participle isformed by suffixing -ma to the verb root, whereas the non-past is with suffix -pa. These participlesare used in relativization (see 5.7.2.)6.8.3. ConverbsThulung has two converbs: one is for expressing simultaneous events, and the other for asequential relationship between clauses. In both cases, the converb is built by suffixing the relevantconverbal affix to the verb root: -to for the simultaneous converb, suffixed to stem 1 (forverbs from classes II and III); -saka for the anterior converb, suffixed to stem 2 (for verbs fromclasses II and III). The converbs are seen in 7.5. and 7.6.6.9. EvidentialityThulung has a hearsay evidential marker, used with great frequency in narratives to relatean event not personally witnessed. This particle is -çe.54. meram khram-lo m¹ni l¹-mi-çehe cry:3s-SS good.man go-3p-HS‘Apparently, when he cries, good people die.’Sentence nominalization can occur before or after this marker.Lahaussois: Thulung Rai2055. “dape-£ola re ¹«-«u” rak-ta-çe-ma ¹ms-ta-çe-mrack-above FOC sleep-1s say-3s:PST-HS-AS sleep-3s:PST-HS-NOM‘“I’ll sleep up on the rack” he said and slept.’56. meram tsahi kits» u-kantshi.aula-rathat CONTR little 3POSS-baby.finger-LOC£ar-ry-m-ka dh¹milo luk-ta-m-çemeet-3s/3s:PST-NOM-INSTR cloudy.liquid exit-3s:PST-NOM-HS‘Because he had been injured in the baby finger, cloudy liquid came out.’7. Complex sentencesComplex sentences are those in which multiple clauses are combined. These clauses canbe based on either a finite or a non-finite verb, the input depending on the construction. In thecategory of non-finite verbs, the possibilities are bare stems, infinitives, converbs, whereas forfinite verbs, the possibilities are finite verbs or nominalized finite verbs. It is interesting that regardlessof the type of complex sentence, the subordinate clause is frequently nominal in form:either finite and nominalized, or non-finite and infinitival (which results in a nominal element).7.1. Complement clausesComplement clauses are of two types: those which are finite, and those which a nonfinite.These divide along semantic lines, with complements to verbs of cognition and sensation,such as to hear, to see belonging to the finite types; complements to modals, on the other hand,and non-finite, and based on an infinitive verb form, the bare stem (or the reduplicated barestem), or a verbal noun (-si suffixed to the verb stem).7.1.1. Verbs of utterrance and cognition:‘to tell’57. a-lwak-ka mysy-s» py-ry-m a-mam-ka1POSS-y.sibling-ERG buffalo-meat eat-3s:PST-NOM 1POSS-mother-ERGbastaka s»-mriyesterday tell-3sP:PST‘My mother said that yesterday my brother ate buffalo meat.’‘to hear’58. sokmu-ra gupsy «ur-mim ¹sin£a th»-siforest-LOC tiger roar:3s-NOM here hear-1pi‘Here we hear the tiger roar in the forest.’59. ¹ni sintha s¹lla by-ry-m u-£okpu ts»-kaand night counsel do-3s/3s:PST-NOM 3POSS-big child-ERGth»s-ty-mhear-3s/3s:PST-NOM‘And the big child heard as they made a plan at night.’Himalayan Linguistics: Archive No.121‘to say’60. ma«-ka ne hopmam mi-dz»pa lamdimother-ERG TOP like.this NEG-good roadkhrekhreja be-pa lamdi l¹k-tsi rak-ta-m bu-mi.bumpy make-Npst.PRT road go-2D say-3s-NOM be-3sP‘Mother said “take the bad, bumpy road like this.”’7.1.2. Modals:‘to want to, to like to’V-V dwamu, V-mu dwamu(the case of the subject can be either nominative or dative--when it is dative, the verb agrees witha neutral 3 person)61. go-lai phaplu l¹-l¹ dwa21s-DAT Phaplu go-go want:3s‘I want to go to Phaplu.’62. go/go-lai phaplu l¹-mu dwak-pu/dwa1s-DAT Phaplu go-INFwant:1s/want:3s‘to need to, have to’V-mu tsahemu/tsahe bomu, V-mu basi63. go homlo ljasi po-mu tsahe-u1s now banana eat-INF need-1s/3s‘I have to eat bananas now.’64. go dika mukli l¹-mu basi1s tomorrow Mukli go-INF OBL‘I need to go to Mukli tomorrow.’‘to be able to’V-mu tsammu, V-si then-mu65. ko«mi tsokpu-mim hun-mu tsam-mi, ko«mi mi-tsam-misome bird-PLU fly-INF can-3p some NEG-can-3p‘Some birds can fly, others cannot.’66. lamdi-si then-muwalk-VN know-INF‘to know how to walk’7.2. CausativeThere are several ways to express the causative in Thulung. The most common is to usethe suffix –be, which attaches to the verb root and is followed by agreement morphology.2 Reduplication of the verb root before the modal dwa-mu ‘to want, to like’ is seen quite frequently, but by no meansnecessary. If the verb complement is not reduplicated, then it must be in infinitive form, as in examples 7 and 8.Lahaussois: Thulung Rai22An alternative means of expressing causativity is seen below. The verb b;;nemu ‘tomake, to prepare’ (which is a borrowing from Nepali banaaunu) is used in combination with anon-past participial form.67. go a-lwak-lai ri-pa b¹ne-u1s 1POSS-y.sibling-DAT laugh-Npst.PRT make-1s/3s‘I make my brother laugh.’68. go a-lwak-lai mysy-s» pe-pa b¹ne-u1s 1POSS-y.sibling-DAT buffalo-meat eat-Npst.PRT make-1s/3s‘I make my brother eat buffalo meat.’7.3. Purpose clausesA purpose clause is built from the bare stem of the verb, with locative marker -ra suffixed.The cognitive connection seems to conceptualize the purpose as being a goal, which has alocative connotation.69. grenem theb-£a l¹s-ta-m bunettle pick-LOC go-3s:PST-NOM be:3s‘She went to pick nettles.’7.4. Causal clausesA nominalized finite verb is the base for a causal clause, with the ergative/instrumentalmarker as a suffix.70. go basi dzam pe-uto-m-ka homlo n»pa bu-«u1s leftover rice eat-1s/3s:PST-NOM-INSTR now sick be-1s‘Because I ate leftover rice, I am sick now.’7.5. Temporal clauses7.5.1. Simultaneous relationshipA simultaneous relationship between two clauses can be expressed with three differentconstructions: an infinitive verb followed by the loan word bela ‘time’ and a temporal marker -ka; with a converbal form in -to; with a sequencer -lo suffixed onto a finite verb. The converbalclause is subordinate to the main clause, as a converb is by nature a non-finite verb form. Thesequencer on the other hand is suffixed to a finite verb, and it serves to coordinate the clauses in acertain temporal relationship. Equivalent examples are given of both, as well as of another meansof expressing another simultaneous temporal relationship, with the expression bela-ka followingan infinitive verb form.71. m» lo b¹ne-mu bela-ka m» deuta rok-tathat frog prepare-INF time-TEMP that god come-3s:PST72. m» lo b¹net-to m» deuta rok-tathat frog prepare-SC that god come-3s:PST73. m» lo b¹ne by-ry-lo m» deuta rok-tathat frog prepare do-3s/3s:PST-SS that god come-3s:PSTHimalayan Linguistics: Archive No.123‘While he was preparing the frog, the god arrived.’7.5.2. Sequential relationshipLike a simultaneous relationship, a sequential relationship can be expressed in three ways:a finite nominalized verb followed by the loan word patshi ‘after’; a converbal form -saka, on theroot; a sequencer construction, with -ma suffixing to the finite verb. The relevant suffixes are theanterior converb –saka and the anterior sequencer –ma (the label anterior serving to express thenotion that the marked clause contains the material coming first temporally).74. mari mu ho¢-miri-m patshi happa mwasy tshabe¢-mirimuch fire light-3p/3s:PST-NOM after much soot spread-3p/3s:PST75. mari mu ho-saka happa mwasy tshabe¢-mirimuch fire light-AC much soot spread-3p/3s:PST76. mari mu ho¢-miri-ma mepmam happa mwasy tshabe¢-miri.much fire light-3p/3s:PST-AS like.that much soot spread-3p/3s:PST‘After lighting a huge fire, they spread the soot all over.’7.6. ConditionalsConditional clauses are marked with -mala (sometimes just -la), and either one of theclauses can be non-past or past, or irrealis.The following sentence has non-past verbs in both clauses.77. go mukli mi-bi-«u-mala ama-mam-ka dykha bo-mi1s Mukli NEG-come-1s-COND 1POSS-mother-ERG difficulty do-3sP‘If I don’t come to Mukli, my mother will struggle.’The same sentence, refering to an unrealized past situation, is in the irrealis mode3.78. go mukli mi-bi-«-wa-m-mala1s Mukli NEG-come-1s-IRR-NOM-CONDama-mam-ka dykha be-m-ba1POSS-mother-ERG difficulty do-3p-IRR“If I hadn’t come to Mukli, my mother would have struggled.”The past can also be used in both clauses to refer to a hypothetical situation.79. ama-wa-ka £okpu mytsy-num bia be-mri-mala1POSS-o.sibling-ERG big man-COM marriage do-3sP/3s:PST-COND£okpu neb-ra dym-mibig house-LOC become-3sP‘If my big sister marries an important person, she will live in a big house.’3 A prototypical irrealis mode makes no assertion that a specific event or state of affairs has actually happened.(Payne 1998: 244) This calrifies why the irrealis would occur with an unrealized event.Lahaussois: Thulung Rai248. ComparisonComparative constructions: -ram, probably a locative (-ra) followed by a relativizer (-m),is suffixed to the object of comparison. (The Nepali bhanda is also used)80. ama-lwak go-ram jepa bu1POSS-brother 1s-than tall be:3s‘My brother is taller than I am.’Superlatives also show this alternation between native and borrowed material, in additionto which the very form of the superlative construction is the same across the two languages.81. ama-del khotle-ram/-bhanda dz»pa bu1POSS-village all-than/-than beautiful be:3s‘My village is the most beautiful.’Most clear as far as comparatives and superlatives go is that Nepali is the source of an alternativecomparative, with bhanda. Also possible, but potentially an areal pattern instead, is the fact thatNepali is the source of the superlative construction using ‘all’ followed by the comparative.AbbreviationsABL ablativeAC anterior converbAS anterior sequencerCAU causativizerCONDconditionalCONTR contrastiveDAT dativeDEF definitiveDET detransitivizerDU dualEMPHemphasisERG ergativeFOC focusGEN genitiveHAB habitualhiLOChigh locativeHS hearsayINF infinitiveINSTRinstrumentalIRR irrealislevLOC level locativeLOC locativeloLOClow locativeNEG negativeNEXP negative experientialNOM nominalizer -mNOM2nominalizer -mimNpst.PRT non-pastparticipleOBL obligationPL.INF plural infinitivePLU pluralPON ponentPOSS possessivepronounPROG progressivePst.PRT past participleRES resultativeSC simultaneousconverbCOM comitativeSS simultaneoussequencerSTA stativeTEMP temporalTOP topicVN verbal-nounPersonal agreement markers show the person and number of the two argumentsseparated by a / in the order A/O.Reference to person/number combinations is as follows: 1, 2, 3 are for first, secondand third persons respectively; s is singular, d is dual, p is plural, sP is singular polite,and e and i stand for exclusive and inclusive respectively.Himalayan Linguistics: Archive No.125All lexical items (not grammatical) borrowed from Nepali (and in a few cases,English) appear in italics in the examples. Within the text of my discussion, italics serveto distinguish Thulung words from the English.BibliographyAllen, Nicholas J. 1975. Sketch of Thulung grammar. Ithaca: Cornell University [CornellUniversity East Asia Papers No. 6].DeLancey, Scott. 2002. “Relativization in Bodic.” Berkeley, CA. Proceedings of the 28thannual meeting of the Berkeley Linguistics Society.Ebert, Karen. 1994. The structure of Kiranti languages. Zurich: Universitaet Zuerich[Arbeiten des Seminars fuer Allgemeine Sprachwissenschaft 13].Thomason, Sarah, and Terrence Kaufman. 1988. Language Contact, Creolization, andGenetic Linguistics. Berkeley: University of California Press.aimee@vjf.cnrs.fr