Wutun grammatical sketch
- Version: 20141217a (AD)
Wutun (ISO 639-3: wuh), an OV language, is an aberrant variant of Northwest Mandarin Chinese (hereafter NWC) with extensive Amdo Tibetan lexical and grammatical features. This language is spoken by a population of ca. 4000 (Janhunen et al.= J2008:19) in northeastern Tibet in what is now Rebgong རེབ་གོང་ (Ch. Tongren同仁) county of China's Qinghai province. Speakers live in three villages of the Rongwo river valley (Ch. Longwu 隆务河, a tributary of the Yellow River): Upper Wutun上五屯, Lower Wutun下五屯, and Jiacama 加倉玛villages. The Wutun population is Tibetan Buddhist and have been officially assigned to the Tibetan or Monguor (Tu土族) Nationalities, the latter group including other local populations that speak Mongolic, not Sinitic languages.
This sketch is a preliminary analysis of about four hours of Wutun discourse recorded by Shawu Dondrup (Xiawu Dongzhou夏吾东周) in Dwyer's DOBES project (2002-2006), evaluating these together with the few published materials on the language (Chen 1982, Chen 1986, J2008, Sandman 2012, Xi 1983). We compare our materials primarily with J2008, whose recordings and preliminary transcriptions were also done by Wutun native speaker (and Janhunen co-author) Shawu Dondrup. (The two projects overlapped.) Despite both projects having the same native speaker-researcher creating the primary transcriptions and providing one of several opinions on sentence glosses, the sound system as reflected in the DOBES (current) project's data is more NWC-like (and simpler), while that of the J2008 project analyzes the data as much more Amdo Tibetan-like (with many more sound distinctions). We hope that this very preliminary analysis will stimulate further work with Wutun materials; our materials are publicly available on the Interactive Inner Asia website (https://ittc.iaia.ku.edu), and the text and line numbers are referenced here. (The Wutun forms herein are primarily transcribed in IPA; examples from others and corresponding forms in Tibetan are transcribed in the source orthography. [ ] enclose phonetic forms; < > enclose transliterations of orthographic forms.)
1. Sound system
Wutun has an inventory that is similar to NWC, except that it does not have tones. Strong Amdo Tibetan influence has particularly affected the consonant inventory and co-occurrence restrictions.
1.1 Vowel inventory
Wutun has a system of six vowels: Vowel inventory:
i u
e (ə) o
a
Since Wutun is an unwritten language without orthographic standardization, we observe a high degree of both conditioned and free variation in our corpus:
/i/ [ɿ~iː~ʲi ] [piːli] negative copula (不是） [ʲi tatsɿ] 'all' [ʲilitə] 'ancient'
/a/ [a~ɑ~ja] [ra] 'and' [taqtaq] 'just' [pəʨan] 'clothe' [xɑxua] 'paint'
/o/ [o ~ ɔ] [toŋja] 'carefully' [ʷotsətsɿ] 'at that time' [ʨhopa] 'water offerings'
/u/ [u ~ ə ~ ʷu] [ku] 'that' [bə] 'not' (negation) [ʷuli] 'there'
/e/ [e~ẽ~ɛ~ej] [nẽ] 'read' [məntse] 'this way' [kɛnʤa] 'shirt' NWC and Wutun /e/ corresponds to a diphthong, usually /ai/ in Modern Standard Chinese (hereafter MSC) , and reflects the pattern of NWC and Wutun having monophthongs and diphthongs where MSC has diphthongs and triphthongs, respectively.
The phonemic status of ə (ɨ ɯ) as such is unclear; in Chinese words it corresponds with MSC /e/ [ɤ], but contrasts with Wutun /e/ : [be] 'white' (MSC 白）(Jan 08) [le] 'come' （MSC来) (Jan 08) (in mani is [li]) so we should put ə as a phoneme above or just leave the example here? [ə] [kə] indefinite single particle (MSC 个) [ʨə] 'this' (MSC 这) Diphthongs There are only three diphthongs in Wutun, all off-glides, and primarily occurring in Chinese words (but cf. 'Tibet').
/uo/ [ʂuo] 'say' (MSC 说) [xuo] 好 'good' (MSC 好)
/ui/ [ʨui] 'tuck into' (MSC 揣) [tʰoŋsui] 'peer, person of the same age' (MSC 同岁)
/ua/ [ɕoxua] 'laugh at' (MSC 笑话) [ʨʰuan] 穿 'wear' (MSC 穿) [suan] 'Tibet' (adx)
Due to the off-glide of palatal stops to a non-high vowel and the variable linguistics background of the transcriptionists, the sequences ɕa ʨa ʨʰa etc. are in our corpus sometimes erroneously rendered ɕia ʨia ʨʰia, e.g. [ɕianla] 'display,' [ʝiu] 'have, be', [ɕiosi] 'astonishing.'
1.2 Consonant inventory
Our corpus shows the following consonant inventory. Allophones are in parentheses.
Table 1. Consonant inventory
labial alveolar retroflex alveo-palatal palatal velar uvular glottal plosives pʰ p (b) tʰ t (d)
kʰ k (g) (q)
nasals m n
ɲ
ŋ
fricatives f s z ʂ ɕ ç ɣ x (χ) h affricates
ʦʰ ʦ tʂʰ tʂ ʨʰ ʨ (ʥ)
lateral fricatives
(ɬ)
lateral approx. l
r
approx.
y w
Some sounds occurs only in loanwords, such as: /f/, which occurs in loanwords from Chinese as in: [fenɕi] 'research' (MSC 分析) [ɬ] and [ʥ] (contrasts with ʨ), which occur in Tibetan loanwords as in: [ɬaɣaŋ] 'temple' (cf. adx ɬa ɣaŋ Hua 1993) [ʥaŋkup] 'headdress' (adx hʥan 'jewelry, ornaments' Hua 1993)
The phoneme /r/ mimics its use in MSC: initially, it is realized as a retroflex fricative [ʐ], whereas syllable-finally it is either a retroflex approximate [ɻ] or a syllabic approximate [ɚ]. The latter is a common Mandarin syllabic extension on nouns, as in 'sun' below. /r/ [ʐ ~ɚ ~ ɻ?] [ʨartur] (adx) 'hair dress' (WT skra, Rebgong hcça 'hair' Hua 2002) tur unknown [ritʰur] [ ritʰɚ] 'sun' (MSC热头) [ ʉɻʂʅ] 'twenty' (MSC 二十)
1.3 Phonological contrasts
NWC and Amdo Tibetan contrast obstruents in two different ways: the first one presents a contrast in aspiration (/p/ : /pʰ/), while the second one both in aspiration and voicing /b/ : /p/ : /pʰ/ (and/bʰ/). Examples showing obstruents in initial and medial position, with words of Sinitic and Tibetan origin.
[b]V [bu] 'cloth' (MSC 布） <bujian> [pəʨan]
[d]V <dawa> [dawa] 'communication', 'news', 'message' (likely cognate: hda 'signal' Hua 1993) [doso] 'coal' (WT rdo sol, Rebgong hdo sho Hua 2002)
[g]V <guore> [g̬uri] 'mgoras boots' (WT mgoras) [p]: <bian>[pian] 'change' (MSC 变）
[pʰ] <pizi> [pʰiʦɿ] 'leather' (MSC 皮子)
V[p]V <aba> [apa] 'dad' (NWC a pha; MSC 阿爸)
V[p]V <qiaoba> [ʨhopa] 'water offerings' (WT mchod pa, ʨʰol pa Hua 1993)
V[pʰ]V <xepa> [xepʰa] 'be afraid' (MSC 害怕）
C[p]V <gerenba> [kərəmpa] 'ceremony conductor' [mampa] 'doctor' (WT sman pa, Rebgong hman ba Hua 2002)
So the basic contrast is in aspiration as in Chinese (voiceless aspirated : voiceless unaspirated), but nonetheless, in our view some Tibetan loan words are realized as voiced, even in initial position. Upon closer examination, these are lexical items with pre-initials. As we see below, J2008 analyze these loans as preserving the pre-initial in the Wutun copy:
[g] [guan] 'monastery' (WT dgon pa, Rebgong hgon ba Hua 2002) IAIA orth IAIA-IPA Jan orth (JJ:85) Jan IPA guan guan gguanba
ʱguanba
Not all Tibetan loan words with voiced pre-initials follow this pattern, however, as our corpus shows: <buri> [puri] 'mgoras boots'.
Our corpus shows a smaller consonant inventory compared to that of J2008 (2008): -Our initial voiced obstruents, e.g. [b], [d] and [g] are represented in J2008 as pre-initials [mb], [nd], [ng]; -Our palatal affricates [ʨ] and [ʨʰ] are represented in J2008 as: [cç] , [ʨ] and [cçʰ], [ʨʰ], respectively, distinguishing front and back obstruent variants co-occuring with; thus [cç] and [cçʰ] with [i], and [ʨ] and [ʨʰ] occur with [e], [ə]; -Our inventory sounds varying from J2008's consonant inventory include [ç] (J. et al. [ɧ]); [χ] (J. et al. [x]); [ʂ] (J. et al. [ʂʰ]), and [ɲ] (J. et al. ). -In Tibetan words only, J2008 further claim that the Amdo Tibetan s : sʰ (xxx '?' vs. sʰa 'meat' AD) distinction can be present in Wutun, and cites the example [sʰɛ̃]三 'three', which in our corpus is unambiguously [sɛ̃] and usually transcribed broadly as [sen].
1.4 Structure of the syllable
The minimal syllable is V, as in a 啊in a=ge 'who, which'. The maximal syllable is CCV(V)C as in [nʣam~(d)zam]. Cf. nʣamlaŋ ~ nʦʰa laŋ] 'world' (Chen 1982, WT nʣam gling); [nʣam hlaŋ ~ nʣam bə hlaŋ Hua 1993).
1.4.1 Initial consonants and consonant clusters (pre-initials)
For a single consonant initial, any consonant phoneme in the inventory can occur. (The phones [ɣ] and [q] do not occur initially, [q] being an allophone of /ɣ/, and [ɣ] being disallowed initially in NWC. In a so-called consonant cluster, the two consonants C1 and C2 are always homorganic, or close to it. C1 are always pre-initials. In our corpus,C1C2 clusters are restricted to ng, rzh, and rj: (all but the first from Jiacama): [rʦɿwa] 'successful' (the only consonant cluster in mani) <nguanna> 'temple' ( [ʱgʷɛ̃] Jan 08) [often transc. as <guan> in Mani <rzhewa> 'guest' (Rebgong ndʐu wa Hua 2002) <rzha> 'enemy' (Rebgong hɟja ja Hua 2002) <rjiacuo> 'ocean' ([ʱɟʝɑ̃ʦʰo] Jan 2008, Rebgong tsho 'lake' Hua 2002) <rjitoh> 'change' ([nʥirla] Chen 1986; Rebgong nɟjər Hua 2002) The initial n is likely a laryngeal pre-initial, the initial r likley nasal and laryngeal pre-initials. (Some of them as [nʥirla] and hɟja are not superscripted, but they precede consonant that can be found with pre-initials.) All of these examples are in words of Tibetan origin, which have original pre-initials. Our corpus shows three Tibetan preinitials, ɦ, N, and h, preserved as initial clusters in Wutun, though several of these alternate freely with single consonants (e.g. [ʱg~g]). Thus, from Written Tibetan to Wutun: {add Wylie in leftmost column - Wylie for retroflex and palatals=? AD} Wylie Wutun (broad) narrow IPA rg ɦg ʱg rzh? Ndʐ ndʐ rj? ɦʥ (??Nʥ) ʱɟ, nʥ gl ʰl hl
J2008 (2008: 40-41) also list three pre-initials: N (=m, n, ɳ, ŋ), ɦ, h, but no less than 21 pre-initial-initial combinations not attested in our corpus. This difference could be due to two plausible explanations: either Wutun is losing or never had as many consonant clusters as Rebgong Tibetan (the local variety of Amdo Tibetan), or the close phonemic and Tibetan-focused transcription of the former project allowed for more phonetic detail to be expressed. The actual explanation likely lies between the two options.
1.4.2 Monopthongs and Diphthongs
Any vowel can occur as a monopthong in a syllable; as exemplified above, only the following off-glide diphthongs occur: /uo/ /ui/ /ua/.
1.4.3 Finals
For the final consonant, nasals and non nasal consonants (stops, fricatives, sonorants, approximants, and lateral approximants) only are permissable. Stop finals are only found in words of Tibetan origin, and not allowed in NWC. The sonorant r is also common as a final, and less frequently, l (the latter unlike NWC).
1.4.3.1 Nasal finals Nasal finals are present both in Chinese and Tibetan. While Modern Chinese present just two nasal finals [n] and [ŋ], Written Tibetan and some Amdo-Tibetan dialects (such as the Amdo dialect of Labrang) show the nasal [m]. Some examples from mani text: -[dsamlaŋ] 'world' (WT nʣam hlaŋ, 'world' Hua 1993) -[ʨʰuan] 'wear' (MSC 穿) The final [m] is present in Amdo-Tibetan loanwords such as: [rogam] 'coffin' (WT ro hgam, 'coffin' Hua 1993) [k̬orlam] 'route' (WT skor lam, tʂə lam Hua 1993) [tirəm]~[tiran] 'ceremony' (adx?)
1.4.3.2 Consonantal finals Non-nasal finals were introduced to Wutun from Amdo Tibetan dialecʦ, which preserve a series of Tibetan finals. Our corpus has the following non-nasal finals: [b] [p], [g] [k], [ɣ] [q], [l] and [r]. -[p] ([b] ) is typically realized as [p ~ ˀ]: [geiʦʰab]~[geʦʰup] 'monkly deeds', (adx hge hʦa 'deeds' Hua 1993) [tʂəkap]~[tʂəkaˀ] 'seventh day after death' (adx [ʨʰə] 'seven', hkap 'time' Hua 1993) [təmʦap] 'offering bowl' (adx təm 'bunch', 'wrap'; tʰep 'bowl' Hua 1993) [ʨaɣtʰap]~[ʨaqtʰap] 'stove' (adx hʨak tʰap Hua 1993)
-[g] [k] [wuxog] '?' 'few'? 'many'?
-[ɣ] [q] [alaq] 'incarnate Buddha, living Buddha' (cf. adx a lak Hua 1993) [mi=tə=tərraɣ] (EXIST.NEG=NZR=AZR.ctfac); [rəngaq] 'attire' (cf. adx rən 'treasure' Hua 1993) [ʦʰoq] 'ceremony' (cf. Rebgong tshok 'assemble' Hua 2002) [taqtaq] 'just' (tak tak Hua 1993)
-[l] [ ʨʰuan pol] 'wear out' (MSC 穿破)
-[r] [ritʰur] 'sun' (MSC热头) [taça wotʰar] 'Thar Lo' (adx?) [çawotʰar] 'Shotar' [çər] 'drink' (WT bʑes/adx mtɕhod,Rebgong ʑe Hua 2002)
In our corpus finals seem to occur with Amdo-Tibetan loanwords. Exceptions are the words [ʨhuan pol] 'wear out' and [ritʰur], which has a Sinitic origin. Concerning the latter, the final -r can represent in this case the retroflex r present in Northern Chinese dialects. Moreover, it is possible to remark the presence of non nasal finals that, as J2008 et al. claim (2008 : 46), have been segmentally lost, such as r and b.
2. Syntax
Wutun is an OV language. It shows features from NWC, as the constituent order and the negation particles, as well as features from Tibetan, as the marking of the direct and indirect experience in copula constructions. For a question of clarity and to make the text more easily readable we chose to write Wutun elements presented in the sections dedicated to syntax and morphology in italics.
2.1 Constituent order
As other languages spoken in the Amdo-Tibetan area (such as Qinghai Baonan and SE Monguor), and unlike MSC, Wutun is a verb final language, with a rigid OV structure. The first constituent of Wutun utterances is constituted by a noun phrase, which can be formed by the subject or by a topic. Wutun can be indeed considered a topic prominent language, that organizes its structure according to topic and comment, rather than subject and predicate (J2008 101-02). The focus on the topic is often showed by the use of the focus marker hi=la. The two examples below aim at showing the constituent order and the existence of both topic prominent and subject-predicate utterances.
(1) ŋa=nitə xua jilitə=kʰanla=ma hi=la jitatsɿ pian=ku=ma jiu=tə rə we=GEN language ancient=than=and COP=COND a.lot change=PERF=SEQ EXIST=NZR COP.fac 'As regard to the language spoken before, many more changes are have occurred'. (wuh20020125a_mani.19) The utterance shows a topic prominent structure, with the topic constituted by the sequence ŋa=nitə xua jilitə=kʰanla=ma hi=la 'As regard to the language spoken before', and the comment by the sequence jitatsɿ pian=ku=ma jiu=tə rə 'Many more changes are have occurred'. The topic prominent sequence is followed by the topic marker hi=la, formed by the equative copula hai and the conditional clitic =la.
(2) wotsətsɿ ɲirəŋ ɲəçə=təra=ja ku çə luotʰi ʨhuan jiu=mu that.time women women=PL=PRT.emph that black boot wear EXIST=PRT.inter 'At that time, women also wore black shoes with long legs'. (wuh20020125a_mani.46) In (2) the sentence shows a clear SOV structure. The subject is represented by ɲirəŋ ɲəçətəra 'women', and the predicate, at the end of the utterance, by the verb ʨhuan jiu 'wore'. The direct object, ku çə luotʰi '(that) black shoes', is placed between the subject and the predicate. We can also remark that the adjective precede the noun, construction that is present in NWC and Mongolic languages, but not in Tibetan, where the adjective follows the noun. In Wutun, the adjective typically precedes the noun, but it can be also found in predicative position, as in the following example: (3) hoŋwo xuo=li ɲɿ health good=INDIR you 'You are in good physical condition'. (wuh20020125a_mani.127) In (3) the predicate is constituted by the adjective xuo (MSC 好). We can also remark that the constituent order can vary when it comes to the oral speech. Here, we can observe that the personal pronoun ɲɿ, is placed after the predicate xuo=li. This particular construction is certainly due to a conversational need to focus on a particular element in the utterance. (ck with AD)
2.2 Finiteness
Finite clauses in Wutun are marked by aspectual markers, such as the perfective =li and =ku, often followed by the indirect marker =li and the direct marker =jiu (from MSC 有, cf. MSC you). Non finite clauses can be subordinated to the finite clause, with verbal markers, e.g. =sa (conditional) or =ta (consequential). (cf. section 3.5, verbal morphology).
2.3 Negation
Wutun shows two different verbal negative particles: the particle pu~pə~pi and mi, both of Sinitic origin. The former, pu, is an imperfective negation. The latter mi is an existential negative that indicates a perfective negation. Both particles cliticized to the verb they are followed to.
2.3.1 The negative particle pu (MSC 不）
(4) ta ʨarkara pə=ʂuo=li=mu ʨarkara ʂuo ta ɕiosi=kə=li ʨarkara ʂuo jiu nowadays funny NEG=say=IMPF=PRT.inter funny call and astonishing=CL=PRT funny call EXIST 'Nowadays, people do not use it. It's laughable to use it. It's called "funny"'. (wuh20020125a_mani.38) In (4) the morpheme pu negates an action or a state with reference to the presence. Unlike MSC and NWC, where the negation pu precedes the verb, in Wutun pu can follow the verb that it negates, if this verb is followed by a nominalizer, as in: ck with AD if we want to keep this example, this construction occurs just once in the corpus.
(5) jitatsɿ ʨʰuan=ti pi=li everyone wear=NZR NEG=IMPF 'No one would wear (it).' (wuh20020125a_mani.26)
As for MSC, in constructions with verbal complements the negative particle pu is placed between the verb and the verbal complement, as in : (6) ɲian=pə=ʂi=li ʨʰopa ʨə ka:lama chant=NEG=RES=IMPF water.offerings so.that difficult 'Because there are many water offerings to make so it is impossible to be on time'. (wuh20020125a_mani.213)
2.3.2 The negative particle mi（没）
Unlike the negative particle pu, the existential negative mi negates an action with reference to the past, as shown in: (7) ʂe=li mi ku=li=tə hi=li house=LOC EXIST.NEG pass=PERF=NZR COP=INDIR 'Didn't he pass away at home?' (wuh20020125a_mani.302)
2.4 Copula utterances
Wutun shows different types of copula utterances. This variety seems to be related to the presence of both Sinitic and Tibetan copulas, and to the marking in Wutun of the direct and indirect perspective. The types of copula present in our corpus are: re, hi=li, hi=jiu, and pe=ra. In Wutun is present, even if rarely employed in our corpus, a zero copula construction.
2.4.1 Factive copula re~ri
The factive copula re~ri is of Tibetan origin (WT red) and it expresses certainty. (8) jilitə xua ŋoma mi:=li ri ancient language really EXIST.NEG=IMPF COP.fac 'There is a very little of our own old language'. (wuh20020125a_mani.6)
In our corpus the factive copula is also present in interrogative questions, followed by the interrogative particle =mu, as in:
(9) ʨarkara ʂuo=la jatsʰan ʂuo=li=tə ri=mu maŋʤa ʂuo=li=tə ri funny say=CZR interesting say=PERF=NZR COP.fac=PRT.int pathetic say=PERF=NZR COP.fac 'Does "skya ga ra" mean something like funny or pathetic?' (wuh20020125a_mani.39)
2.4.2 Indirect copula hi=li
The copula hi=li is formed by the equative copula hai and by the indirect marker =li. It expresses second-hand information and distance from what has been said. (10) ɕoɕa lian xuŋ hi:=li xija lian xuŋ ha#ha#ha trousers two inch COP=INDIR size two inch ha#ha#ha 'The size of a pair of trousers was about two inches – ha ha, two inches'. (wuh20020125a_mani.82) The indirectivity is clear in (10), as the Wutun speakers are referring to an information concerning the past, that apparently they did not experienced directly.
(11) ʤaʦʰaŋma=tə hi=li awɯ Rgya.tsa.ma.village=GEN COP=INDIR child 'We are people of the Jiacama village.' (wuh20020125a_mani.168) In (11) the use of the indirectivity is more complex. The speakers are talking about themselves and about being part of the Rgya Tsa Ma village community, statement that would require directivity. On the contrary, the use the indirectivity is motivated by the fact that they are expressing a general statement (ck with AD!)
2.4.3 Directi copula hijiu~hijia~jija
The copula construction hi=jiu is formed by the equative copula hi and the direct suffix =jiu. It is used in first person subject/topic utterances.
(12) ŋun=ʨikə ɲiantʂɿ#kuiɲian ʂuo=la=mu ʨarkara ʂuo jiu hi:=ja I=PAUC time#past.year say=CZR=PRT.int funny say EXIST COP=DIR 'We used to call it "funny"'. (wuh20020125a_mani.37)
In interrogative clauses, the directivity is revolved, as it is related to the sphere of the second person the question is asked to (J2008 : 98). This is shown in: (13) ka ɲiɲi ɕoŋ=kə=hə ta wɯʦəʦɿ li#ʤa=la li=li=tə hi=jia young old.lady village=CL=FOC and then Li#house=ABL come=PERF=NZR COP=DIR 'Is the younger old lady from the family of Li when she was young?' (wuh20020125a_mani.151)
2.4.4 Negative copula pə ra
The copula pə ra is a negative copula, which is formed by the Sinitic negative particle pu~pə and by the Tibetan factive copula re~ra.
(14) jəmtumpa ɲian=ti=tə pə ra Yundunba recite=PROG=NZR COP.NEG COP.fac 'Isn’t the Yundumba Scripture?' (wuh20020125a_mani.267) In (14), the negative copula pə ra is used to form an interrogative-negative question.
2.4.5 Zero copula
As said in the introduction to the copula constructions, the zero copula construction is not common in our corpus. Indeed, it is found just once. (15) ʨaŋ tʂəkap today seventh.day.after.death 'Yes, today's the seventh day from one's death'. (wuh20020125a_mani.287)
2.5 Direct and indirect utterances
The verb ʂuo 'say' follows the reported speech.
(16) oh ku=tə kui mani=li ʨə=ti=li ʂuo=la oh that=GEN spirit temple=LOC go=NZR=PERF say=IMPF 'Oh, is it said that his spirit goes to the village shrine?' (wuh20020125a_mani.359)
3. Morphology
Wutun morphology is formed by Sinitic, Tibetan, and Mongolic elements. The language shows a large amount of morphological markers (prefixes and postpositions) that can be considered as clitics or affixes. Their status of clitic is related to the isolating nature of some suffixes, mostly of them of Sinitic origin, (e.g. the nominalizer -de), which the bounding to the stem is not as consistent as that one of a suffix. Moreover, as claimed by J2008 (2008: 53), on the segmental level the only morphological dependence from the stem is found in pronouns. Nonetheless, a few morphological markers, such as the benefactive or the nominalizer, can be semantically more strongly bounded to the stem, but not on the phonological level. Another indicator of boundness are orthographies. Xiawu Dongzhi, the transcriber of our corpus, does not have background in linguistics and did not received any special training in segmentation. Therefore, his segmentation choices can be seen as possibly indicative of boundness. Looking at the orthography, markers as the genitive are written separately; other case markers, such as the locative and the ablative, can be attached or separated from the noun. On the contrary, verbal markers such as aspect markers, perspective markers, or verbal complements are written together. Therefore, the verbal morphology is more closely bound. Considering this subtle different degrees of boundedness and the isolating nature of Wutun morphological markers, we chose to conform the segmentation of our corpus indicating them as clitics, using the sign =.
3.1 Nouns
Nouns can be followed by suffixes indicating number and cases. As in MSC, the mark of the plural is an optional category: unmarked nous can be either singular or plural. As in Tibetan and MSC, the grammatical gender is not marked in nouns (ck with AD, Tibetan marks gender only in pronouns).
3.1.1 Plural
The singular marker is =kə (from the MSC general classifier =ke 个, which is Wutun has the role of an indefinite single particle). The plural is marked with the clitics =təra and =ʨəkə. The latter is usually attached to singular pronouns. Wutun has also a dual marker, the postposed clitic =liaŋkə 'two together'.
3.1.2 Cases
Wutun presents three cases: genitive, locative and ablative, which are express ed through the use of postpositions. The first two, genitive and locative, are of Sinitic origin, while the the third one, the ablative, is of Tibetan origin. Wutun shows also two focus markers.
Nominative: unmarked
Accusative: unmarked
Genitive: =tə (MSC 的） (17) ʂe=tə doɣxo=kə house=GEN owner=CL 'owner of the house' (wuh20020125a_mani.360)
Locative: =li (MSC 里) (18) ɕiŋ=li heart=LOC 'in the heart' (wuh20020125a_mani.22）
Ablative: =la (WT mkr-la) (19) li#ʤa=la Li#house=ABL 'from Li’s house' (wuh20020125a_mani.151)
Focus markers: =xa and =mu
The focus markers =xa and =mu cover different functions: they act as case markers as well as topic markers. In the next examples the focus marker =xa follows a pronoun and it seems to play to role of an oblique case, in which the element marked by the focus marker is the recipient of the action. (20) ʨʰəta ɲiɲi ŋa=xa kui#ri kəra tu=ti=li ŋu ŋu=tə=ku ɕin=man ʨʰuan=ʂi=ti wo Khyida daughter we=FOC past#day and! talk!=NZR=PERF 1s 1s=GEN=CL find=RES wear=RES=NZR PRT! 'Khyida's daughter was also talking to us the day before yesterday. I am going to find them and wear them'. (wuh20020125a_mani.73)
(21) ɕuxua=ti=tə rə ŋa=mu=ʨikə=xa laugh.at=PROG=NZR COP.fac we=ASSOC=PAUC=FOC 'People would laugh at us'. (wuh20020125a_mani.8）
In (20) and (21) we could identify the function of =xa as a dative. However, from a semantic point of view, the elements marked by =xa are not only the recipient of the action, but also the patients, constituting what Dede (2007) calls an 'anti-ergative'. This recalls the function of the Tibetan dative la, which influences the use of the particle =xa in a variant of NWC, the Huangshui Chinese spoken in Qinghai (ibid.). Another evidence of this influence is shown in (22), where =xa marks the patient, or the object, of the action: (22) tʰaŋtətʰŋtə jitia hə=ʂi=ma poʤa=li xan ɕoɕa=xa jitʰiu ʤu=ʂi=ma ̥ in.a.rush! together eat=RES=SEQ edge=LOC also trousers=FOC together! tuck=RES=SEQ çilotʰi=tə=jitiu ʤu=ʂi=ma boots=GEN=together! tuck=RES=SEQ 'After eating in a rush, the edges of the trousers are tucked in the boots and then head out towards the upper village'. (wuh20020125a_mani.97） Here, =xa is marking the preverbal patient of the nominal phrase, acting as the Chinese 把. Again, we can identify in =xa 'anti-ergative' marker. It is important to remark that, looking at the different layers of the language, the particle =xa borrowed from Amdo-Tibetan the semantic role of anti-ergativity. This function was not borrowed with its phonetic Amdo-Tibetan form la, but transferred in the NWC particle［xa] 哈 (or 下)??. (We were discussing this in the last meeting, Dede (2007) writes this particle as 哈, but I think we were talking about 下..)
The marker =mu is defined by Sandman (2012) as a morphological topic marking, which function is similar to that one of the Baonan topic marker ma. The same topic marker is present in Salar. In the next two examples, =mu act as a focus markers, highlighting in the utterance the element that deserves more attention, that is more central in the message:
(23) mani=li=mu hi=tə=ʂə ʨuntə ji hok jiu=li temple=LOC=FOC COP=NZR=RES! like.this one set EXIST=INDIR 'There is a set of bowls like this in the temple'. (wuh20020125a_mani.188）
(24) çi lothi=kə jiu ɕi:=mu xa=ʂə=la ta k̬orlam=li xuo=li ʂuo=tʂi=ja black boots=CL EXIST snow=FOC come.down=RES=CZR and route=LOC good=INDIR say=DUR=PRT.emph ka:lamala məntse ʂuo=li ʂuo=tʂi=ja children this.way say=PERF say=DUR=PRT.emph 'There is a pair of those black boots that are really helpful to go on circumambulation in the snow, but kids still make such comments'. (wuh20020125a_mani.64）
However, in (25) we can notice the anti-ergative function of =xa in the topic marker =mu.
(25) ʨʰopa=mu ɕianla=ʂə ta titsho ʂɿ#tian=mo to=ku=li water.offerings=FOC display=RES and time ten#hour=FOC arrive=COMPL=PERF 'After displaying the water offerings, it's already ten o'clock'. (wuh20020125a_mani.206) Here, the marker =mu follows the object marking, the patient of the action, acting as an accusative case. In the same utterance, its variant?? =mo, underlines a definite element in the sentence, acting as a topic marker.
We can remark here a functional overlapping. The particle =mu, which has as principal feature the function of topic marker, which we related to the topic prominent nature of MSC and NWC, can act as well as an accusative case marking, taking with it the Tibetan semantic feature of the particle =xa.
3.2 Pronouns
The table below present the pronouns found in our corpus, which are of Sinitic origin.
Table 2. Pronouns PN1s ŋo ~ ŋɵ ~ ŋu ŋa (oblique) PN2s ɲi PN3s ku PN1p ŋa PN2p ɲi=ʨəkə (paucative) ɲi=mu (associative) PN3p ku=ʨəkə
As shown in the table Wutun pronouns do not show gender distinction. As indicated by the second person plural, plural forms can be formed through the use of the paucative =ʨəkə, which find its origin in the MSC form 几个 'a few', and the associative marker =mu, probably derive from the MSC plural 们 (J2008: 65). Interestingly for a Sinitic language, Wutun pronouns can take/indicate? cases with morphological variation. Our corpus shows the example of the dative ŋa, which differentiates itself from the unmarked form through the alternation u : a.
Chen 1982 and 1986 attested in his works two Mongolic forms pronouns of Mongolic origin are absent in our corpus: aʨʰaŋ for the third person singular (Chen 1982), aʨʰaŋɢula (Chen 1982) and aʨʰaŋla (Chen 86) for the third persons plural. These forms are not attested in our corpus, they have been probably lost because of an intense contact with NWC, or they probably belonged to the speech of non-Wutun speakers.
3.2.1 Demonstrative pronouns
The demonstrative pronouns have a Sinitic and a Mongolic form. As regard to the proximal demonstrative pronouns, our corpus shows the Sinitic form tʂə~tʂɿ 'this' (in the mani text is almost always ʨə, but I don't think is the right IPA), which is usually followed by the indefinite single particle =kə, and the Mongolic one eɯ=la 'these' (ela in Xi 1983). The Sinitic deictic form, made with the locative case =li is ʨʰə=li 'here'. As regard to the distal demonstrative, our corpus shows the Sinitic form -ku (which is also the singular third pronoun), and the a Mongolic-Sinitic hybrid form, the distal deictic form wuli~ wɯli 'there' (wuh20020125a.71), formed by the Mongolic distal demonstrative and the Sinitic locative =li (MSC里).
3.2.2 Interrogative pronouns
Interrogative pronouns are formed with the interrogative stem of Tibetan origin a- (WT xe) and the Sinitic one ma- (from 什么?). a=li 'where', formed by the interrogative stem a and the locative suffix =li ma=kə 'which', formed by the interrogative stem ma and the indefinite single particle =kə amən=tə=kə 'how many', formed by the interrogative stem a, the Sinitic plural ending =mən, the genitive =tə, and the indefinite single particle =kə. a=kə 'who', formed by the interrogative stem a and the indefinite single particle =kə. Interrogative pronouns can also have the meaning of indefinite pronouns, such as ma=ge, which means 'what, which' but also 'whatever'. The last meaning is shown in the following: (26) ta məntsa jaɣla jiu xə ma:=kə huŋ=ʂə jiu also this.way beautiful EXIST and whatever=CL decorate=RES EXIST 'It was made beautifully'. (wuh20020125a.71)
3.3 Postpositions
Postpositions in Wutun are locative, instrumental, temporal and comitative.
Locative ʂə ~ ʂɑ̃ ~ ʂʅ 'on' (MSC 上) li 'in', 'inside' (MSC 里) taŋ 'above' (WT sgaŋ/steŋ/thog, Rebgong thok, Hua 2002) joˠ 'down' (Jan et al. 08) (WT yog)
Instrumental liaŋkə 'with', 'using', 'by means of'
Temporal ŋanʨʰe 'before' (Jan08) ʂe 'when, during'
Comitative liaŋkə 'together with'
3.4 Numerals
Our corpus shows that most numerals are of Sinitic origin (as was found also by J2008: 71), with the opportunistic use of some Tibetan forms.
Table 3. Primary numerals in Wutun one
ji
two ɤr ~ ək (Jan 08) three san four sɿ five wu (or better write ʷu?) six liu seven ʨʰi eight pa nine ʨiu ten
ʂɿ (or better write ʂʅ?)
As show in the table Mongolic and Tibetan simple numeral are absent in our corpus. We can suppose a common origin for san 'three', which can have as cognate the WT gsum, Rebgong_H02 [hsəm]. On the contrary, Chen 1982 and Chen 1986 show the predominance of Mongolic numerals, e.g. tʰolon ~tʰolo 'seven' (Chen 1982), tʰaːwun 'five' (Chen 1986), χabraŋ 'ten' (ibid.). As we claim for the pronouns, the presence of Mongolic basic vocabulary in these previous works and the large amount of Sinitic forms in our corpus can be related to the presence of non-Wutun speakers in the area or to a situation of multilingualism in the area in which the switch from a language to another could have been a common language practice. (for AD: in the lexicon we found the Mongolic form ɢuar ~guar 'two' (wuh20020125), but I didn't find it in mani, and all the numerals in the mani text and in the Jiacama are of Sinitic origin, so it could be an exception but it could be also a mistake in the lexicon.)
A similar scenario is showed for the decade and above. Table 4. Numerals - the decade and above fourteen ʂɿsɿ (or ʂʅsɿ?) twenty ərʂʅ twenty-two liaŋ=kə ŋanʂɿ eighty-one paʂɿji hundred pi
As for the decade and above, we see that the numeral present in our texts are of Sinitic origin, with the exception of a hybrid Chinese-Tibetan word: liaŋ=kə ŋanʂɿ 'twenty-two' (wuh20020125a I.225, I.254, I.258), from Chinese 两个 'two=CL' + adx ɳi ɕu ~ Rb. ɳə xhə 'twenty' (Hua 2002). As for primary numerals, Chen 1986 shows Mongolic forms, e.g. tʰaːwuraŋ ~ tʰeːraŋ 'fifty' and ɲiːmaraŋ tʰolon 'eighty-seven'. Interestingly, in J2008 08, where most of the numerals are as in our corpus of Sinitic origin, we find two Amdo-Tibetan forms: ʰtʰɵŋ ~toŋ 'thousand' and ʂʰə ~ tʂʰi 'ten thousand'. We also remark that Wutun's use of counting liaŋ 'two' is idiosyncratic; ordinarily ərʂʅliaŋ=kə 'twenty-two=CL' (wuh20020125a.256, I.259, I.275). Nevertheless, it also preserves he Chinese distinction between an enumerating ər 'two' and a counting liaŋ 'two (of)': ərʂʅ 'twenty', ku liaŋ=kə 'the two of them'. Apart from this hybrid form, Tibetan numerals are absent, as well as the Mongolic ones. Ordinals are formed with the clitic =wa, after the number (from WT ba) e.g. <yiwa cang> 'the first chant' (wuh20040708.142). J2008 (2008 : 73) cites also the Sinitic form composed by the prefix di= (from MSC 第).
3.5 Verbal morphology
Wutun has a complex system of verbal morphology, showing a large amount of verbal markers and verbal type elements. Its verbal morphology does not have a temporal-aspectual nature, but rather a completive-non completive one. This feature is typical of MSC, which is the source of most of the verbal markers here presented. Therefore, even though some verbal markers can show a temporal distinction, we opted for a classification based on completion. MSC features can be recognized in the verbal complements and auxiliaries. On the contrary, an Amdo-Tibetan feature is represented by the marking of the direct-indirect experience.
3.5.1 Completion
Our corpus shows two finite markers, =li and =ku, and three non-finite markers =la~li, the durative =tʂə and the progressive =ti.
3.5.1.1 Completive markers
Perfective =li (MSC 了)?
(27) ʤaʦʰaŋma=la ʨə=ta ʦʰɯ=li=tə ri rgya.tsa.ma.village=ABL here=AV come=PERF=NZR COP.fac 'I came here from Jiacama village'. (wuh20020125a_mani.156)
In (27) above, the perfective =li occurs in finite clauses and marks the anteriority of the action, that is here clearly happened in the past: ʦʰɯ=li=tə ri 'I came'. The perfective =li can also be followed by the form ʦɿ=li as shown below: (28) jə=li ʨə tʰi=ʂi=kʰu ʨi=kə=li ʨi=kə jo=li ʦɿ=li EXIST=INDIR this happen=RES=COMPL here=CL=LOC this=CL must=PERF EXEC=INDIR 'Something like this has (had?) to happen initially in this area'. (wuh20020125a_mani.34)
As shown in (28), the form ʦɿ=li, derived from the Sinitic verb 做, acts as an executive marker and has also the function to develop a base for the objective marker =li (J2008 et al. 2008 : 95).
Completive =ku (MSC 过）
As in MSC, the completive marker =ku is the equivalent of the verb 过 'to pass', and it marks the completion of an action. The completion of an action can be situated in the past, but not necessarily, as shown in the following examples.
(29) ɲɿ=tə ku hi=jiu=ja ŋu=tə=ku ŋu ʨʰuan pol=ku=li you=GEN that COP=DIR=PRT.emph I=GEN=CL I clothe wear.out=COMPL=PERF 'What you just have said is truly the case. I have worn them out by myself'. (wuh20020125a_mani.107) Here, =ku occurs in a finite clause and is followed by the perfective ＝li. The two markers indicate the completion of an action in the past.
As said before, the completive guo can also mark the completion of an action not situated in the past, but in the near future as in: (30) ʨaɣtʰap=ʂə kʰei=ku=li stove=SUPER clean=COMPL=INDIR 'They will be cleaned on the stove later'. (wuh20020125a_mani.147)
Anteriority is also expressed by the sequence verb + existential jiu, as in: (31) ŋun=ʨikə ɲiantʂɿ#kuiɲian ʂuo=la=mu ʨarkara ʂuo jiu hi:=ja I=PAUC time#past.year say=CZR=PRT.int funny say EXIST COP=DIR 'We used to call it "funny"'. (wuh20020125a_mani.37)
3.5.1.2 Non-completive markers
Imperfective =la~li (WT *.pa ~.ba)
(32) tʂha kun=ti=li tiaŋʦawo li ku=pə=ʨi=la thaŋ rənʨʰan#tʂoma tea infuse=NZR=INDIR Ting'tdzin come pass=NEG=away=IMPF perhaps Rinchen#Sgrolma ʨən=kə tʂa=ti jiu injection=CL make=PROG EXIST 'If that's the case, the old man Ting'tdzin is not coming since Rinchen Sgrolma is receiving injections.' (wuh20020125a_mani.241) In (32), =la marks the non-completion of the activity in the time. The imperfective =la is frequently found in finite clauses. It is seems to play a different role from its homonymous marker =la, which indicate non-anteriority in non-finite clauses, as show in:
(33) jilitə xua ʂuo=la ɕuxua=ti=tə ri old language say=CZR laugh.at=PROG=NZR COP.fac 'We get laughed at when we speak the old language'. (wuh20020125a_mani.7)
Durative =tʂə (MSC 着)
(34) ʨaŋmənʦa mani ɲian=la=ra tʂəmə=kə ʤula=ku me=kə ʂuo=tʂə=ja nowadays mani recite=CZR=also thus=CL change=PERF and=CL say=DUR=PRT.emph 'These days, (maybe is better to add 'it's been said that?') changes also have occurred in the way mani is chanted'. (wuh20020125a_mani.4) The durative =tʂə is frequently present in finite clauses, it expresses an action or a state that is still going on at the moment and continuing to the future, but also a habitual fact or action.
Progressive =ti (35) ŋa=ni xua xuo=ti=tə rə=mu we=GEN language good=PROG=NZR COP.fac=PRT.inter 'Is our language a neat one?' (wuh20020125a_mani.14)
(36) ajirən ʨʰuan=ti=tə ku guri old.women wear=PROG=NZR that mgoleb.shoes 'What the old women wear is called mgoleb shoes'. (wuh20020125a_mani.59)
As shown in (35) and (36) the progressive =ti occurs together with the nominalizer =tə. (but according to JJ =ti is a fusion of 的+有). Compared to the durative =tʂə, =ti seems to indicate in particular an habitual, established fact or action.
3.5.2 Mood
The indicative tense is unmarked.
The conditional is expresses through the verbal marker -la (from adx case mkr -la), as in: (37) wɯtsɿtsɿ lətsu=la ʂuo=la ləwa ʂuo=la ɲi tsʰɯ ta ɲi təku tʰin=ku=ti=tə ri at.that.time lyrics=ABL say=COND singer say=COND you sing so.that you each.other! listen=PERF=PROG=NZR COP.fac 'If one were a singer at that time, singers would be able to learn lyrics from each other'. (wuh20020125a_mani.37)
The Wutun imperative can be formed just using a bare form as in: (38) ku=ta=la li mantə tuo=ku=li=tə pə ra there=AV=ABL come that.way much=COMPL=PERF=NZR NEG COP.dir 'Please come that way, is it too much?' (wuh20020125a_mani.131)
Sometimes, the imperative can be formed by a bare form followed by a resultative, marking the completion of the action asked by the speaker, as in: (39) ɲɿ ʤe ku=ta tsʰo=ʂə ʤiumi you so.that there=AV seat=RES 'Jigasmed 'Please sit there, Jigasmed!' (wuh20020125a_mani.202)
3.5.3 Verbal complements
In this section we present a series of verbal markers that we call verbal complements. These complements constitute originally single verbs that, as verbal complements, semantically modify the verb they are attached to, indicating the completion, the inchoaction? or the direction of the action expressed by the main verb, as shown in:
(40) ʨiwa=li=ra wɯ=li ŋa=nitə ku xɑxua=tə ku aʤa ta ku ja=li hi jḁ ku ʨʰuanti home=LOC=and there=LOC we=GEN that paint=NZR that monk and that courtyard=LOC and courtyard that clothes ʨʰuanti jitatsɿ kʰaŋ=man clothes all observe=RES 'The monk artist of my family observed the different clothing styles in the yard of the monastery'. (wuh20020125a_mani.29) Here, the verbal complement =man, whose meaning as a independent verb is 'fill, satisfy' indicates the completion of the action.
(41) ɲɿ=ʨikə go=la tʂʰa çə rḁ ɲɿ=ʨikə tun=ti=la momo ri=ʨʰei=tʰala you=PAUC pass=RES tea drink COP.fac you=PAUC freeze=NZR=IMPF bread hot=RES=TERM 'Are you all cold? Please come and have some tea till the bread is heated'. (wuh20020125a_mani.140)
In (41) above, the verbal complement =la, whose meaning as an independent verb is is 'come' (MSC 来), indicates here the movement toward the speaker.
This type of construction is typical of Standard and NWC. Therefore, the verbal complements here show are all of Sinitic origin. We show here below the verbal complements present in our corpus:
ʂə 'do successfully, complete' (MSC 上） la 'come, to be close'?? (MSC 来) man 'finish, become full' （MSC 满） ʨʰi 'to start' (MSC 起）
3.5.4 Auxiliaries
In this section we include a series of modal auxiliaries present in our corpus. Compared to verbal morphology analyzed above, the modal auxiliaries act as verbal type elements. The modal auxiliaries shown here below follow the main verb indicating ability, possibility, and necessity. They do not occur often in our corpus: as we will see in (44) modality can be expressed through the use of other grammatical constructions. The examples here below are preceded by negative particles, that cliticized to the following element.
Ability/possibility: ʨʰe 'be able, can' (42) atia=kə mi=la ɲian ku pə=ʨʰe=li monk=CL EXIST.NEG=COND recite that NEG=be.able=INDIR 'A religious service cannot be performed without the assistance of a monk from the family'. (wuh20020125a_mani.281)
Necessity: to~ti 'need to' (MSC 的＋要) (43) ʤaŋla=ti jiu=tʂə=mu ʤaŋla=tə jiu ji mo=to learn=NZR EXIST=DUR=PRT.inter learn=NZR EXIST and! EXIST.NEG=must 'There was no need to learn, no need to learn at all'. (wuh20020125a_mani.105)
In the following example we show how the necessity can be formed in Wutun without the use of modal auxiliary, but with the nominalizer =de followed by the existential jiu (from MSC 有). 44) ku naŋʨʰan=li tʂə=ti jiu jiu pa that Rongbo.monastery=LOC go=NZR EXIST EXIST PRT 'They would have to go to the Rongwo monastery'. (wuh20020125a_mani.397)
Necessity: jo 'must' (45) jə=li ʨə tʰi=ʂi=kʰu ʨi=kə=li ʨi=kə jo=li tsɿ=li EXIST=INDIR this happen=RES=COMPL here=CL=LOC this=CL must=PERF EXEC=INDIR 'Something like this must have happened in this area'.
3.5.5 Non-finite serial markers
Wutun as a series of non-finite serial markers that cover a coordinative, temporal, concessive, conditional and counterfactual function. The general serial marker is =ma of Tibetan origin (WT glu dbyangs), which indicates coordination between two actions (it is translated with 'and'), as show in: (46) ta ʨaŋmənʦa ma:=ke hi:=ja ʤurla=ku=ma pian=ku=ma awo and nowadays whatever=CL COP=PRT.emph!DIR change=PERF=SEQ change=PERF=SEQ boy 'Boy, so many changes have occurred these days'. (wuh20020125a_mani. 92)
A second serial marker is the marker =ta, which is often precede by =ma, a verbal marker of unclear origin, that conveys a resultative meaning. The form =ma=ta introduces resultative-consequential and hypothetical clauses, as shown respectively in (47) and (48). (47) kʰu rogam kʰan=ma li:=ma=ta that coffin see=SEQ come=RES=CONSEQ 'Since he has seen the coffin,...' (wuh20020125a_mani.309)
(48) lə tsʰɯ=ma=ta han hi=la ra ləwa hi:=li song sing=RES=CONSEQ still COP=INDIR COP.fac singer COP=INDIR 'She is still a great singer if she sings'. (wuh20020125a_mani.106)
3.5.6 Perspective
The marker of the perspective refers to the epistemic estimation of the speaker in regard to what is expressing in the utterances. Therefore, the term perspective include notions as that of the inferentiality, evidentiality, indirectivity, etc. As we have introduced for the copula utterances, Wutun marks the direct (egophoric) and indirect experience. These are features presents in the Bodic languages, and not in the other two language families which influences Wutun: the Sinitic one and the Mongolic one. This feature has been borrowed from Wutun, as well as from other languages spoken in the Amdo-Tibet area, such as the Mongolic SE Monguor and Baonan.
3.5.6.1 Indirectivity
Indirectivity is marked with the suffix =li. It usually refers to second-hand information and expresses distance from what has been said. (49) ɲiɲi=ʨikə ma:=kə ʨaŋ tʂʰa kun=ti jiu ʂuo=li lady=PAUC whatever=CL today tea infuse=PROG EXIST say=INDIR 'I heard that ladies were going to make some tea today'. (wuh20020125a_mani.240) Here, the speaker is giving an information that she heard but is not sure about, as it does not concern her personal sphere. The indirect marker =li has been described as a verbal marker (J2008) but, as shown in Sandman (2012), it can be found also after nouns, as in: (50) Ane tiema=ge=li ana ni kan=da EX bike=CL=INDIR mother PN2s look=IMP 'Oh it’s a bike! Mother, look!' (Sandman 2012: 383; morphological glossing added here.) Here, =li has in particular the function of a sensorial inferential marker.
Also in our corpus is possible to find the marker =li not preceded by a verb or by a verbal marker, as in: (51) ta ʨarkara pə=ʂuo=li=mu ʨarkara ʂuo ta ɕiosi=kə=li ʨarkara ʂuo jiu nowadays funny NEG=say=INDIR=PRT.inter funny call and astonishing=CL=INDIR funny call EXIST 'Nowadays, people do not use it. It's laughable to use it. It's called "funny"'. (wuh20020125a_mani.38) Here, the indirect marker =li is found after an adjective and a the indefinite single particle =ke. Probably, the speaker is expressing a general statement, and its use is a morphological device to express distance. In the next example, the marker =li follows a demonstrative: (52) ku ku=li ʂoupa ʨartur sɯlur jitatsɿ kuan=li ʨə=la ku jitatsɿ=te: tə=te:=ma ʨi=kə ti=tə ri that that=INDIR hair.decoration hair.dress sort all monastery=LOC go=CZR that all=GEN wear=NZR=SEQ this=CL wear=NZR COP.fac 'It is not the case that they were allowed to visit the monastery wearing all sorts of head and hair dresses'. (wuh20020125a_mani.95)
Looking at the (50), (51), and (52), we can suppose the extension of the use of =li to to non-verbal elements, such as nouns, adjectives,and demonstratives, with the function of expressing sensorial inferentiality, and epistemologic distance from what the speaker is expressing.
3.5.6.2 Directivity
Directivity is marked with the direct marker =jiu, and appears with first person subject/topic utterance, or utterances in which the subject/topic belongs to the sphere of a first person speaker (J2008). In our corpus, the direct marker is often present in copula utterances, and, because of its limited occurrence (exclusively with first person pronouns, that often are not marked with a direct marker), it is not found very often in our corpus. Therefore, we present here below just an example of directivity that is found with a verb, and we refer for others examples to the copula utterances section. (53) ʨʰopa amən=tə=kə ɕanla=ti=jiu taŋdsən awo offering.bowl how.many=GEN=CL offer=NZR=DIR Ting'dzin uncle 'Uncle Ting'dzin, how many offering bowls are you putting on the display?' (wuh20020125a_mani.95) In (53) we find an example of directivity reversed, as we have already seen in 12). The utterance is indeed a question and the directivity is directed to the sphere of the second person respondent.
4. Conclusion
This grammatical sketch aimed at presenting the linguistic features of Wutun, looking at the contact with language families present in the Amdo-Tibetan area, such as Tibetan, Mongolic, and Turkic. Looking at the different layers of the language, we remarked the Sinitic principal features of Wutun, and its similarity with MSC and its closest neighbor language, NWC. We have noticed similarities with NWC in the phonological system, in the constituent order and its topic-prominent nature, in pronouns and numerals, in some case markers (e.g. the genitive and the locative), in the completive/non-completive nature of the verbal morphology as well as in the form and function of verbal markers such as verbal complements and modal auxiliaries. In this NWC-like structure, we have remarked the presence of Amdo-Tibetan in different layers of the language. At the phonological level, we observed the presence of pre-initials and stop finals, not allowed in NWC. At the morphological level, we found the presence of cases such as the ablative and the anti-ergative. The latter is an emblematic example of semantic borrowing, in which an Amdo-Tibetan function has been transferred to a NWC form. Another important feature is the morphological marking of the direct and indirect perspective in copula constructions and verbal forms, which seems to be extended also to other parts of speech, e.g. nouns, adjectives, and demonstratives, acting as a sensorial-inferential marker. As for the other contact languages, Mongolic and Turkic, their influence is mostly found in the lexicon, but still limited. Even though not analyzed directly, we have seen that in the various examples shown that the Wutun basic lexicon is mostly of Sinitic origin. This is shown in pronouns, numerals, basic verbs and adjectives. Looking at the vocabulary employed in our corpus, we can affirm that Amdo-Tibetan is the source for the lexicon concerning religion and its rituals, because of the Wutun's faith in Buddhism. All these features make Wutun another example of the contact situation in the Qinghai-Gansu Sprachbund. In this particular case, the cultural proximity with Tibetans and the common element of Buddhism played an important role in the evolution of the language. In an area where also Mongolic and Turkic languages are spoken, Wutun kept its NWC typological structure incorporating Amdo-Tibetan phonological, morphological and lexical elements.
Works cited
- Chen Naixiong 陈乃雄. 1982. 五屯话初探. [A preliminary investigation into the Wutun vernacular]. 民族语文 Nationality Languages: 10-18.
- Chen Naixiong 陈乃雄. 1986. 关于五屯话 [On Wutun language/An Outline of the Wutun Linguistic Structure]. Journal of Asian and African Studies 31: 33-52.
- Dede, Keith. 2007. The Origin of the Anti-ergative [xa] in Huangshui Chinese. Language and Linguistics 8.4: 863-881.
- Hua Kan. 1993. ###
- Hua Kan. 2002. ###
- Janhunen, Juha, Marja Peltomaa, Erika Sandman and Xiawu Dongzhou. 2008. Wutun. Munich: Lincom Europa. (= J2008)
- Sandman, Erika. 2012. Bonan Grammatical Features in Wutun Mandarin. Per Urales ad Orientem. Iter polyphonicum multilingue. Suomalais-Ugrilaisen Seuran Toimituksia=Mémoires de la Société Finno-Ougrienne 264. Helsinki.
- Xi Yuanlin 席元麟. 1983. 同仁土族考察报告。 四寨子（五屯）的民族历史，语言和艺术。第二部分：语言 [A report on an investigation of Tongren Monguors: the history, language, and art of the four villages (Wutun). Part II: Language]. Qinghai Nationalities College Tongren Monguor Research group ms.: 1-52.
Abbreviations:
suffix boundary = clitic boundary
ASSOC associative
AV adverb (was „ADV“!; ADV >> AV in texts!!!)
adx Amdo Tibetan (modern)
CL classifier
COMPL complement
CONSEQ consequential
COP.fac factive copula
CZR complementizer
DIR directive
emph emphatic
EXEC executive
EXIST existential copula
GEN genitive
IMPF imperfective
INDIR indirect
inter interrogative
LOC locative
NEG negative
NWC Northwest Mandarin Chinese
NZR nominalizer
PAUC paucal
PERF perfective
PN pronoun (1, 2, 3 s/p singular/plural)
PROG progressive
PRT particle
RES resultative
SEQ sequential
TERM terminative
WT Written Tibetan
wuh Wutun