Full text. By Jean Robert Opgenort, 2011. Appeared in Libju-Bhumju 45.
Introduction
Wambule is the name of a Kiranti language[1] which is spoken by more than 5,000 people living around the confluence of the Sunkosi and Dudhkosi rivers in eastern Nepal. The Wambule area comprizes the southernmost part of Okhaldhunga district, the westernmost part of Khotang district, the northernmost part of Udaypur district, and the northeasternmost part of Sindhuli district.
This paper is part of an ongoing investigation into the reconstruction of Proto-Kiranti (PK), the hypothetical ancestor language of the Kiranti family, and the internal classification of the various Kiranti languages (Opgenort 2004a, 2005, forthcoming). The goal of this paper is to examine a number of salient phonological developments which occurred in Wambule and which characterize it as a distinctive member of the Kiranti language family.
Data
I compared the Wambule data which I personally collected in Nepal from 1996 to 2003 with data from other languages currently available in my Comparative and Etymological Kiranti Database (CEKiD). This database contains lexical and morphological material on many Kiranti languages as well as hundreds of reconstructed Proto-Kiranti etyma.
The language data used here are taken from the following publications: Bahing by Hodgson (1857); Bantawa by Winter (2003) and Doornenbal (2009); Chamling by Hodgson (1857) and Ebert (1997, 2003); Chepang by Caughley (2000); Dumi by Hodgson (1857) and Van Driem (1993); Hayu by Hodgson (1857) and Michailovsky (1981); Jero by Opgenort (2005); Khaling by Toba and Toba (1975) and Toba (1979); Kham by Watters and Watters (1973); Kulung by Hodgson (1857) and Tolsma (1999); Limbu by Van Driem (1987); Sunwar by Bieri and Schulze (1971, 1973a, 1973b) and Borchers (2007); Thulung by Allen (1975); Tilung by Toba (2004) and Rāpacā (VS 2065); Wambule by Opgenort (2004b); and Yamphu by Rutgers (1998).
In the present discussion, Western Kiranti (WK) comprizes the following languages: Hayu, Sunwar, Bahing, Jero, Wambule, Khaling and Dumi. Midwestern Kiranti (MK) includes Thulung and Tilung. Central Kiranti (CK) is represented by Chamling, Bantawa and Kulung, and Eastern Kiranti (EK) by Yamphu and Limbu.
Phonological developments
The Kiranti languages can be divided into various subgroups on the basis of phonological developments regarding initial obstruents and sonorants, and innovations in the rhotic sphere and consonant clusters (van Driem 1990; Michailovsky 1994; Opgenort 2004a, 2005, forthcoming). I reconstruct the following Proto-Kiranti initial consonants. The consonants discussed in this paper are in bold typeface.[2]
|
*p- |
*t- |
*c- |
*k- |
|
*ʔp- |
*ʔt- |
*ʔc- |
*ʔk- |
|
*ph- |
*th- |
*ch- |
*kh- |
|
*b- |
*d- |
*j- |
*g- |
|
|
|
|
*kw |
|
*m |
*n |
|
*ŋ |
|
(*ʔm) |
(*ʔn) |
|
|
|
|
*r |
|
*ŕ |
|
|
*l |
|
|
|
|
(*ʔl) |
|
|
|
*w |
|
*y |
*h |
|
|
*s |
|
|
The obstruents include three positions of articulation (bilabial, alveolar, velar) and one series of affricates (alveolar). I propose a four-way contrast in manner of articulation for obstruents: voiceless, preglottalized, voiceless aspirated and voiced. The two most salient series of obstruents that may be used for Kiranti subgrouping are the preglottalized proto-series *ʔp-, *ʔt-, *ʔc- and *ʔk- and the voiced proto-series *b-, *d-, *j- and *g-, since these series have undergone different developments in the various Kiranti languages. By contrast, the plain voiceless proto-series *p-, *t-, *c- and *k- and the voiceless aspirated proto-series *ph-, *th-, *ch- and *kh- are not used for subgrouping because the plain voiceless proto-series have remained essentially unchanged in the modern Kiranti languages (i.e. *C > C) and the reconstruction of a voiceless aspirated series is somewhat doubtful. Still, the historical development of *ch- may also be significant for the classification of Wambule within Kiranti.
Other developments that can be used for the classification of Wambule are that of labiovelar *kw- and that of the preglottalized sonorant series *ʔl-, *ʔm- and *ʔn-.
Retention of voiced *b-, *d-, *j- and *g-
The first phonological development, which sets Central and Eastern Kiranti apart from Western and Midwestern Kiranti, is the devoicing of the voiced proto-series *b-, *d-, *j- and *g-. The Wambule reflexes of the voiced proto-series have voiced obstruents. The retention of the voiced proto-series is typical for Western and Midwestern Kiranti.[3] Here are some examples:
-
PK *bit ‘cow’
WK/MK: Sunwar bii, Bahing bing ‘bull’, Jero biya, Wambule biya, Khaling bay, Dumi bhiʔi (b- expected), Thulung beno, Tilung bī
CK/EK: Chamling pyupa, Bantawa pit-ma, Kulung pi, Yamphu bik (p- expected), Limbu pit
-
PK *del ‘village’
WK/MK: Bahing dyal, Jero dɛl, Wambule dyal, Khaling del, Dumi deːl, Thulung del, Del
CK/EK: Bantawa ten, Kulung tel, Limbu tɛn ‘place’
-
PK *ja- ‘eat’
WK/MK: Hayu dza-, Sunwar ‘dza-tsa, Bahing jáwo, Jero jacap, Wambule jacam, Khaling jö-nä, Dumi dzunɨ, Thulung jam ‘food’, Tilung jukhmā
CK/EK: Chamling ca-ma, Bantawa ca-(a), Kulung cama, Yamphu caˑma, Limbu cama
-
PK *gup ‘tiger’
WK/MK: Sunwar gupsu, Bahing gupsa, Jero gupso, Wambule gupso, Thulung gupsiu
CK/EK: Bantawa ki-wa, Yamphu kiˑba, Limbu keba
Some other Wambule words that reflect Proto-Kiranti *b- are bakcam ‘to be, sit’, bubu ‘white’ and bukcam ‘to get up, stand up’. Proto-Kiranti *d- is reflected in e.g. dukcam ‘to shake’, dumcam ‘to become’ and dwakcam ‘to fall down’. Proto-Kiranti *j- is reflected in e.g. japcam ‘to buy’, jikcam ‘to break’ and jwakcam ‘to know’. Proto-Kirant *g- is reflected in e.g. gagbo ‘crow’, gucam ‘to pick up’ and gwakcam ‘to give’.
Deglottalization of preglottalized *ʔp-, *ʔt-, *ʔc- and *ʔk-
The second phonological development, which sets Central Kiranti apart from Eastern Kiranti, and Midwestern Kiranti apart from Western Kiranti, is the deglottalization of the preglottalized series *ʔp-, *ʔt-, *ʔc- and *ʔk-. The Wambule reflexes of the preglottalized proto-series have voiceless obstruents. The innovation of *ʔC- > C- is typical for Western Kiranti. By contrast, in Midwestern and Central Kiranti, the preglottalization of *ʔp- and *ʔt- was accompanied by voicing. Here are some examples:
-
PK *ʔpak ‘pig’
WK: Hayu póg, Sunwar poo, Bahing po, Jero pa, Wambule pa, Khaling ‘po, Dumi poʔo
MK: Thulung boa, Tilung bo
CK: Chamling bose, Bantawa bak, Kulung boː
EK: Yamphu akma, Limbu phak
-
PK *ʔtuŋ- ‘drink’
WK: Hayu tũˑta ‘drunk’, Sunwar tuu-tsa, Bahing túgno, Jero tuːcap, Wambule tuːcam, Khaling tu-nä, Dumi tɨŋnɨ
MK: Thulung ɖu(ŋ)-, Tilung duṅma
CK: Chamling dungma, Bantawa duŋ-ma, Kulung duːma
EK: Yamphu uŋma, Limbu thuŋma
-
PK *ʔkaŋ- ‘look’
WK: Sunwar koo-tsa ‘see’, Bahing kwó-gno ‘see’, Jero kicap, Wambule kwacam, Khaling ‘ko-nä ‘know how to do something’
-
MK: Tilung keimā ‘see’
CK: Chamling khanga-, khõ-, Bantawa khaŋ-(u), Kulung khoːma ‘see’
EK: Yamphu khaŋma
-
PK *ʔcik ‘bird’
WK: Sunwar ‘tsiikbi, Bahing chikba, Jero cikmu, Wambule cwagbo
MK: Thulung cəkpu
CK: Bantawa choŋ-ga, Kulung chowa
EK: Yamphu soŋ(w)a
Some other Wambule words that reflect Proto-Kiranti *ʔp- are paccam ‘to cause to do’, pwaccam ‘to tie, bind’ and picam ‘to come across a horizontal plane’. Proto-Kiranti *ʔt- is reflected in e.g. teicam ‘to make drink’, tupcam ‘to beat, strike’ and twapcam ‘to beat, play’. Proto-Kiranti *ʔc- is reflected in e.g. cicimo ‘mouse’, and perhaps in carja ‘millet’ and cumcam ‘to catch’ (the latter two may also reflect *c-). Proto-Kirant *ʔk- is reflected in e.g. kuksyal ‘cloud’, kwakcam ‘to bite, peck’ and kwamcam ‘to cover’.
Spirantization of Proto-Kiranti *ch-
A additional sound change in the obstruent series that occurred in some Western Kiranti languages, including Wambule, Jero and Bahing, is the development of Proto-Kiranti *ch- to s-, but the evidence for this development is still rather circumstantial. The change *ch- > s- also occurred in Eastern Kiranti, but this is most likely an independent development.
-
PK *cham ‘song’ (compound with -laŋ ‘language, word, speech’)
WK: Bahing swálong, Jero saːʔlɛŋ, Wambule saːlaŋ
MK: no data
CK: Bantawa cham, Kulung cham
EK: Yamphu semluma, Limbu sam-lo
Glottalization of labiovelar *kw-
Another significant sound change that occurred in Wambule is reflected in words with bilabial implosive stops in Bahing (transcribed as ʔb-), Sunwar (transcribed as bw-), and Wambule (transcribed as ɓ-), which might be traced back to Proto-Western Kiranti *ʔw- (or *ʔb-) corresponding to prior labiovelars because the preglottalized Western Kiranti segments correspond to velars in closely related and more distantly related Tibeto-Burman languages, in particular Kham and Chepang.
-
PK *kwar- ‘wound’
Bahing ʔbar, Wambule ɓari
WK: Hayu buʔma, Sunwar gaar, Jero mari, Khaling ‘kwaar, Dumi kar
MK: Thulung par ‘sore, skin lesion of any kind’
CK: Bantawa khen, Kulung kher
EK: Yamphu huwa, Limbu kaːn
Kham ‘khxtera, Chepang khay
-
PK *kwa- ‘to eat, bite’
Sunwar ‘bwa-, Bahing ʔba-, Wambule ɓacam
WK: Jero macap, Khaling bət-
MK: Thulung p(e)-, Tilung khemā
CK: no data
EK: no data
Kham kxya ‘eat (chewy or hard things)’
Some other Wambule words that reflect Proto-Kiranti *kw- are ɓala ‘shadow, spirit’, ɓambu ‘cheek’, ɓasyam ‘shoulder’ and ɓo ‘chicken’. In Wambule, *kw- > *ʔw- > *ʔm- > ɓ-, whereas in Jero *kw- > *ʔw- > *ʔm- > m-.
Innovation (or retention) of a preglottalized liquid
Some Wambule and Jero dialects share the presence of preglottalized ʔl-, which may either reflect Proto-Kiranti *ʔl- or represent a typical ‘Chaurasia’ (i.e. the theoretical language unit combining Wambule and Jero) innovation for which no conditioning factors can be found, except prefixes that are no longer traceable.
-
PK *ʔlaŋ ‘foot, leg’
Jero lɔsu (possibly *ʔlɔsu)[4],Wambule ʔlɔsu, lwasu
WK: Hayu le, Sunwar -li in khoyli, Bahing -li in kho’li, Khaling -l in syál, Dumi -lɨ in phoʔlɨ
MK: Thulung -l in khel, Tilung philuk,
CK: Chamling -lu in phílú, Bantawa laŋ, Kulung lɔŋ
EK: Yamphu laŋ, Limbu laŋ
Some other Wambule words that reflect Proto-Kiranti/Proto-Chaurasia *ʔl- are (ʔ)la ‘arm, hand’, (ʔ)lam ‘road, path’, (ʔ)lima ‘lie’ and (ʔ)liːcam ‘to sell’.
Innovation (or retention) of preglottalized nasals
The most important phonological difference between Wambule and the other Kiranti languages is the presence of implosives ɓ- and ɗ-, corresponding to nasal m- and n- in other Kiranti languages. The glottalization of m- and n- in Wambule may either reflect Proto-Kiranti *ʔm- and *ʔn- or represent a typical Wambule/Chaurasia innovation for which no conditioning factors can be found, except prefixes that are no longer traceable. In Jero, Proto-Kiranti/Proto-Chaurasia *ʔm- and *ʔn- appears to have been deglottalized.
-
PK *ʔmik ‘eye’
Wambule ɓisi
WK: Hayu mé-k, Sunwar miiktsi, Bahing michi, Jero misi, Khaling mas, Dumi miksi
MK: Thulung miksi, Tilung mikcī
CK: Chamling micu, Bantawa mük, Kulung muksi
EK: Yamphu mik, Limbu mik
-
PK *ʔniŋ ‘name’
Wambule ɗi
WK: Sunwar ne, Bahing ning, Jero ni, Khaling nang, Dumi nɨ
MK: Thulung nəŋ, Tilung nung
CK: Chamling nung, Bantawa nüŋ, Kulung niŋ
EK: Yamphu niŋ, Limbu miŋ
Some other Wambule words that reflect Proto-Kiranti/Proto-Chaurasia *ʔm- are ɓeiso ‘buffalo’, ɓiːco ‘female’ and ɓulum ‘tail’. Proto-Kiranti/Proto-Chaurasia *ʔn- is reflected in e.g. ɗucam ‘to be good’, ɗusum ‘nose’ and ɗwabu ‘ear’.
Retention of consonant clusters
Most Western Kiranti languages, including Wambule and Jero but excluding Dumi, have /Cr-/ and /Cl-/ clusters which consist of an initial bilabial or velar plosive plus *l or *r. Central and Eastern Kiranti languages such as Bantawa, Kulung, Yamphu and Limbu lack these clusters. The lack of clusters is a non-Western Kiranti innovation.
-
PK *blo ‘arrow’
WK: Hayu blo, Sunwar -bra in libra, Bahing blá, Jero blu, Wambule blo
MK: Thulung blə ‘metal arrow head’
CK: Bantawa bhe (be expected), Kulung bei
EK: Yamphu -la in thula, Limbu -la in toːŋ-la ‘arrow-stick’
Some other Wambule words that exemplify the retention of consonant clusters are brikcam ‘to break, burst’, glwamcam ‘to lie down’, gramji ‘hate’, kraccam ‘to bite, gnaw’, plyacam ‘to leave, quit’ and prwakcam ‘to run’.
Conclusion
Wambule shows traces of various phonological changes that occurred in the Kiranti language family.
-
Firstly, Wambule exemplifies a number of typical Western Kiranti features, such as the retention of the voiced proto-series *b-, *d-, *j- and *g- and that of /Cr-/ and /Cl-/ consonant clusters, innovations in (i.e. deglottalization of) the preglottalized series *ʔp-, *ʔt-, *ʔc- and *ʔk- and a possible spirantization of Proto-Kiranti *ch-.
-
Secondly, like Bahing and Sunwar, Wambule has a bilabial implosive stop that can be traced back to a prior labiovelar *kw-.
-
Then, like Jero, but unlike all the other Kiranti languages, Wambule has a preglottalized lateral ʔl-, which may be a retention of Proto-Kiranti *ʔl- or represent an innovation.
-
Finally, unlike all the other Kiranti languages, Wambule has implosive stops ɓ- and ɗ-, which may be a reflect of Proto-Kiranti *ʔm- and *ʔn- or represent a language-specific innovation.
The sum of these phonological developments, especially the intricate changes in the glottalic range, characterizes Wambule as a unique member of the Kiranti language family.
[1] The Kiranti languages form a major family of Tibeto-Burman languages spoken in Nepal.
[2] In this paper, I will not discuss the development of front or apical *r- to y- and that of back or dorsal *ŕ- to y-. In Opgenort (forthcoming), I claim that innovations in the rhotic sphere seem to be rather irregular in Western and Central Kiranti, and cannot yet be successfully used for grouping non-Eastern Kiranti languages. By contrast, Eastern Kiranti is characterized by systematic innovations in the rhotic sphere, where both *r- and *ŕ- merge with *y-. It may well be the case that in Eastern Kiranti *r- initially merged with *ŕ, and that the resulting dorsal rhotic *ŕ- eventually merged with *y-. In Western Kiranti, by contrast, *r- did not merge with *ŕ and *y-, but was kept distinct. However, here the phoneme *ŕ seems to have been reanalysed: in some languages *ŕ- was absorbed by *r- or by *y-, but in Kulung *ŕ became a voiced velar stop, except in the case of the word for the trade good ‘salt’, which is perhaps a loan word. The split of *ŕ into *r- and *y- remains largely unaccounted for.
[3] Note that shared retention is generally a weaker argument for subgrouping than a shared innovation.
[4] In written Jero, the phoneme /ʔl/ is found in *ʔlā ‘hand’, which corresponds to Wambule *ʔla.