Project B11:
Semantic roles, case relations, and cross-clausal reference in Tibetan

SFB 441 main page
Subprojects of the SFB
B11 project page
  B11 publications
  People in the field
  Language archeology
  Valency dictionary
  Clause types (overview)
  Dialect diversity: Kenhat
  Expressions of comparison
  Map
  Technical notes

Fieldwork in Ladakh

  • Map Dialect regions in Ladakh and Baltistan
Harvest:

A Valency Dictionary of Ladakhi Verbs – work in progress


As fieldwork in 2004 and 2005 has shown, not only the lexicon, but also the grammar and thus the sentence patterns vary considerably between the two main Ladakhi dialect groups. We thus started to collect the Kenhat data more systematically (2005 - 2008: Gya-Sasoma, Gya valley). The following overviews shows how our data base developped over the past four years and how it differs from presently available lexical ressources:

Valency Dictionary of Ladakhi Verbs

surveyed
dialects

main entries

without sub-entries

with sub-entries

sub-entries total

examples

2005

2006

2007

2008

2005

2006

2007

2008

2005

2006

2007

2008

2005

2006

2007

2008

2005

2006

2007

2008

total

895

905

924

925

654

641

634

596

241

264

290

329

664

786

886

1046

7443

9283

11193

12980

DOM (Shamskat)

783

794

791

780

539

(539)

511

466

235

256

280

314

589

704

738

821

5241

5683

6079

6252

GYA  (Kenhat)

261

354

572

814

121

152

299

502

140

203

273

312

349

535

745

908

1076

1930

3102

4492

SAS (S)

116

(116)

(116)

(116)

52

(52)

(52)

(52)

64

(64)

(64)

(64)

86

(86)

95

(95)

429

(429)

(429)

(429)

LEH (K)

1

101

102

(102)

62

62

(62)

2

34

40

(40)

2

56

69

(69)

4

434

523

(523)

CEM (K)

88

(88)

(88)

(88)

56

(56)

(56)

(56)

32

(32)

(32)

(32)

46

(46)

(46)

(46)

285

(285)

(285)

(285)

TIR (S)

62

(62)

(62)

(62)

32

(32)

(32)

(32)

30

(30)

(30)

(30)

47

(47)

(47)

(47)

144

156

(156)

(156)

ARA (S)

51

78

98

(98)

28

39

48

(48)

23

39

50

(50)

29

51

70

(70)

126

208

343

(343)

TEA (S)

40

9

31

44

220

WAK (S)

29

(29)

(29)

(29)

15

(15)

(15)

(15)

14

(14)

(14)

(14)

21

(21)

(21)

(21)

49

(49)

(49)

(49)

SKI (S)

15

(15)

4

(4)

11

(11)

16

(16)

59

(59)

NYO (K)

13

(13)

10

(10)

3

(3)

4

(4)

22

(22)

KRD (S)

5

(5)

(5)

(5)

1

(1)

(1)

(1)

4

(4)

(4)

(4)

9

(9)

(9)

(9)

20

(20)

(20)

(20)

LEH2 (K)

6

(6)

(6)

(6)

1

(1)

(1)

(1)

5

(5)

(5)

(5)

5

(5)

(5)

(5)

8

(8)

(8)

(8)

other

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

57

72

92

117


Comparison with other lexical sources

 

main entries

full examples

pattern variation

collocations

 varieties

BRGY  CT

ca. 1200

ca. 30 - 40%

few cases

no specifical focus

1

CDTD

1339

almost none

almost none

high number, incomplete sets

Hackett 2003 CT

694

ca. 25 - 30%

almost none

small number

1

Haller 2004 Amdo

566

ca. 90%

almost none

almost none

1

LhV 2005  Lhasa

750/1110

ca. 60 %

few cases

as separate entries

1

 

 

 

 

 

 

VDLV

925

> 100 %

special focus
 

special focus
 

2 dialect groups,
2 complete sets
 

case relevant readings

1641

> 100 %

readings total

2203

 

 

 

 

CT verbs tested

1398

 ca. 50 % shared in each variety, ca. 45 % shared by both varieties

 shared among both dialects: ca. 85%

 
 ca. 5% in each dialect not found as verbs in CT

Domkhar

780

Gya

814

non-attested CT verbs

563

adjectivals

36

verbs not in CT

38

unclear relation

171


Abbreviations

1. Dialects
ARA: Aranu, Nubra (Shamskat); CEM: Cemre (Kenhat); DOM: Domkhar (Shamskat); GYA: Gya-Sasoma (Kenhat);
KRD: Kardong, Nubra (Shamskat); LEH: Leh (Kenhat); LEH2: Leh, second generation (Kenhat); NYO: Nyoma (Kenhat); SAS: Saspol (Shamskat); SKI: Skindiang (Shamskat); TEA: Tea (Shamskat); TIR: Tirit, Nubra (Shamskat); WAK: Wakka (Shamskat)

2. Bibliography

BRGY:
Zhang, Yisun [Kraŋ Dbyisun] et. al. (eds.) 1993. BodRgya tshig­mdzod chenmo [The large Tibetan Chinese Dictionary]. Vol. 1–2. Pecin: Mirigs dpeskrunkhaŋ [Beijing: Nationalities Publishing House].

CDTD:  Bielmeier, Roland. In preparation. Comparative dictionary of Tibetan dialects. Vol. I: Verbs. [Preprint 2008]

Hackett, P.G. (2003): A Tibetan verb lexicon. Verbs, classes, and syntactic frames. Ithaca, Boulder: Snow Lion.

Haller, F. (2004): Dialekt und Erzählungen von Themchen. Sprach­wissen­schaft­liche Beschreibung eines Nomadendialektes aus Nord-Amdo. Bonn: VGH Wissen­schafts­verlag.

LhV: Bailey, G. and Ch. E. Walker (2004): Lhasa Verbs. A Practical Introduction. Lhasa: Tibetan Academy of Social Science.

Consultants



top
B11 project page
B11 publications
SFB main page




"Language change and the fossilization of the Old Tibetan b- prefix in Ladakhi and Balti."
Clause types (an overview)
Kenhat     

kuʃunaŋ trakuʃu (apples and peaches)on the comparison of comparative expressions in structurally differing languages





Clause types in Ladakhi (overview)

In the meanwhile we have enlarged our collection of exotic sentence patterns. As one will see in the following overview, almost all combinations are possible.We have tried to keep all patterns in an obvious order, but that also means that the pattern numbers had to be reassigned (and that the numbers given in Zeisler 2007b are no longer valid). We will use colours for case, bold face and italics for multiple use of the same cases in order to highlight the combinatory complexity and to make the unexpected patterns (such as three times absolutive or three times locative marking) more evident. Note that the siglum "~Loc" includes also postpositions, which are structurally nouns followed by case markers. In the right most column we give a non-exhaustive specification of the main type of verb associated with the pattern, or the specific verbs going with this pattern. In the latter case, the patterns can be expected to be very infrequent, but, for the time being, we are not able to specify the frequency of the patterns in the verbal lexicon.

We expect that many of the more 'exotic' patterns will not be found in Classical Tibetan (CT), but exactly which one, we do not yet know. As one can see from the few patterns we were able to include, the classical sentence patterns are not restricted to the main patterns.

Case assignement and pattern variation in West Tibetan (WT) do to a certain extent reflect the semantic transitivity hierarchy (but there are significant differences between the dialects, perhaps also between individual speakers of the same dialect). In the Kenhat dialects we can observe quite nicely how lowering transitivity in the case of reflexive and reciprocal actions lead to an almost regular drop of the ergative marker. Another factor that plays a crucial role, especially for those verbs that take an intermediate position on the transitivity hierarchy, is distance or closeness in terms of space, time, and emotion. Events that are perceived as close tend to receive less overt marking than those that are perceived as distant. "Emotional distance" includes all kinds of personal involvement: surprise, embarrassment, compassion or being highly affected.

Clause types – overview

‡ preferred order, change of position possible: “x‡   ‡y” means that x and y may exchange their position (thus y, x), “(x   y)‡  ‡z” means that the group x, y and z may exchange their position without changing the order within the group (hence z, x, y)

A. Main patterns

0-place predicates

00     

WT: /rgyal/ ‘o.k.’, /rden/ ‘true’, introductory /yot/ ‘once upon the time’

1-place predicates

01     

Abs

change, motion

2-place predicates

02     

Abs

Abs

predication, WT transformation, low transitive verbs

03     

03a 

Abs

~Loc

affection, oriented motions, position, change into

03b             

~Loc=top

Abs

existence (at a given place)

04     

Abs

Abl

get out [–ctr], move away [±ctr]

05     

Abs

Com

contact, separation [± ctr]; WT: [–ctr] & cause, media, instrument

06     

Aes

Abs

possession, WT: experience, affection

07     

Erg

~Loc

directional activity

08     

Erg

Abs

non-directional activity, transformation, high transitive verbs

3-place predicates

09     

09a 

Erg

~Loc

Abs

transfer (to R)

09b             

Erg

Abs

~Loc

deposit, transformation into (sgyur, byed)

09c 

~Loc=top

Erg

Abs

WT: drag along (topicalisation of LCT argument necessary)

10     

10a 

Erg

Abl

Abs

take away-type I

10b             

Erg

Abs

Abl

take away-type II

11     

11a 

Erg

Abs

Com

join, mix see, separate, exchange (theme oriented), fill with

11b             

Erg

Com

Abs

exchange (recipient oriented)

B. Marginal patterns

1-place predicates

12     

Aes/~Loc

WT: emphasised emotions

13     

Erg/Instr

some animal sounds; WT: non-focusing use of sense organ, ploughing, harvesting, work fast

2-place predicates

14     

14a 

Abs‡

‡Instr/Gen


WT: [–ctr] events & cause, media, instrument;
CT: fear

14b             

Abs

Gen

WT: fill with [–ctr] (Shamskat = 14a borrowed from Kenhat);

15     

15a 

~Loc/Aes

Gen

WT: fill into with [–ctr] (Shamskat); CT tshugs

15b             

Aes

Instr/Gen

WT: be harmful [–anim]

15c 

Aes Instr

CT: be damaged [–anim] (thugs)

16     

Aes Com

WT: be content, satiated (/tshims/)

17     

Aes/~Loc

~Loc

WT: /eloa cha/ 'be forgetful (about sth)' and other mental states; experiencer of 13

18     

Aes

Abl

WT: get scolded/beaten (/khoa gerganehane phok/)

19     

Erg

Com

WT: press, collocation: divorce

20     

Erg

Instr/Gen

CT: collocation: khus debs, phus debs, promise (possConstr); WT: non-focused use of sense organ

21     

Erg

Abl

CT: directional activity (partitive): drink from; begin with; WT: non-focused  use of sense organ

3-place predicates

22     

Abs

Abs

Abs

WT: reflexive transformation (/co/), ‘it’s my beer’

23     

(Abs

Abs‡)‡

‡~Loc

WT: reflexive transformation (/zgyur/) 

24     

Abs

Com

Abs

WT: collocation: mix with; come into a discussion

25     

Abs

Abl

Abs

WT: motion from, protection 

26     

26a 

Abs

Abl

~Loc

WT: motion from x to y

26b             

Abl

~Loc(‡

‡)Abs

WT: gapping from x to y

26c 

~Loc

Abl(‡

‡)Abs

WT: exceptive exist (03b + REL)

27     

Abs

~Loc

~Loc

WT: labour force exchange

28     

28a 

Aes/~Loc

Abs

~Loc

WT: get sth stuck, be left behind, have enough, obtain into; perception through sense organ; experiencer of 03a

28b             

Aes/~Loc

~Loc

Abs

WT: believe; experiencer of 03b

28c 

Aes

Aes

Abs

WT: experiencer of 06

29     

Aes

~Loc

~Loc

WT: be expert (in sth), be in harmony (with), /cikcigika eloa cha/

30     

Aes

Abl‡

‡Abs

WT: obtain from; perception through sense organ

31     

Aes

Abs

Com

WT: sense organ; experiencer of 05

32     

32a 

Aes

Abs

Instr/Gen

WT: have enough for; experiencer of 15a;  

32b             

Aes

Instr/Gen

Abs

WT: sense perception through sense organ

32c 

~Loc

Instr

Abs

CT: x-la y-kyis (medium) luspa med 'be  complete'

33     

33a 

Aes

Abs

Abs

WT: experience as; /ŋo šes/; get more; exper. of 02

33b             

~Loc

Abs

Abs

WT: become more (on a certain place 03b+ RST)

34     

Erg

Abs

Abs

WT: transformation (/co/), estimation, repetition (timeArg); collocation: decide; consider (Kenhat)

35     

Erg

~Loc

~Loc

remember, warn, praise, talk, ask, teach so about

36     

36a 

Erg

Abs‡

‡Instr/Gen

WT: fill with [+ctr], lower price by, be enough by Ken/Sham; OT/CT: collocation: promise

36b             

Erg

Abs

Gen

WT: fill with [+ctr] (Shamskat)

37     

37a 

Erg

~Loc

Instr

CT: collocation: promise

37b             

Erg

~Loc

Gen

WT: fill (into), cover (upon) with [+ctr] (Shamskat)

38     

Erg

~Loc‡

‡Abl

WT: protect, focusing use of (sense) organ, directional, chase away

39     

Erg

Com‡

‡~Loc

WT: fill with; touch with organ (/ńuk/) (Kenhat)

4-place (including collocations)

40     

Abs

Abl

~Loc

Abs

WT: collocation: crawl & SRC/GOAL

41     

Abs

Abl

~Loc

~Loc

WT: go, come from X to Y for work

42     

Abs

Com

~Loc

Abs

WT: colloc.: get into a discussion (Kenhat)

43     

Abs

~Loc

~Loc

Abs

WT: collocation: believe [+ctr]

44     

Abs

~Loc

Abs

Abs

WT: collocation: believe [+ctr]

45     

45a 

Aes

Abs

~Loc

Abs

WT: experiencer of 21

45b             

Aes

~Loc

Abs

Abs

WT: /rden šes/ 'believe'

45c 

Aes

Aes

Abs

Abs

WT: experiencer of 29a (/ŋo šes/);

46     

46a 

Aes

~Loc

~Loc

Abs

WT: believe so with respect to sth

46b             

Aes

Aes

~Loc

Abs

WT: experiencer of 27b

47     

Erg

(Abl

~Loc‡)‡

‡Abs

WT: transfer from to; exchange / barter

48     

Erg

Abs

Abl

Abs

WT: colloc.: turn back animals (Kenhat)

49     

Erg

~Loc

‡Abs‡

~Loc

WT: remind, sell for I, collocation: divide, promise, believe (+ctr)

50     

Erg

(Abs

Abs‡)‡

‡~Loc

WT: transfer/put as, sell for II, colloc. invite, /rden šes/ (+ctr)

51     

Erg

Abs Com Abs

WT: establish social relationship between 2 persons

52     

Erg

Com

Abs

Com

WT: exchange / barter

53     

Erg

Com‡

‡(Abs‡

‡~Loc)

WT: exchange / barter

54     

Erg

~Loc

~Loc

~Loc

WT: write to sb about sth in a letter

55     

Erg

~Loc

Abs
Instr/Gen

CT: promise; WT: fix the price for with, lower the price by

56     

Erg

~Loc

~Loc

Instr/Gen

Not yet attested! OT/CT promise (CONTind)

5-place (collocations)

57     

Erg

~Loc

~Loc

Abs

Abs

WT: lower the price & beneficiary

58     

58a 

Erg

Abl

Abs

~Loc

~Loc

WT: take a bride

58b             

Erg

~Loc

(Abl

~Loc)
Abs

WT: lower the price

59     

Erg

Abs

Abl

~Loc Abs

WT: turn back animals

60     

Erg

~Loc

~Loc

Abs
Instr/Gen

WT: lower the price

 

Erg (coll)

~Loc

Com

Abs

WT: exchange, barter, mix

C. Special contexts:

impersonal constructions (often possessor constructions with first argument)

1-place

61
Abl

WT; have pain (zer gzer); CT starting with

2-place

62     

Abl

Abs

hon speech CONTdir

63     

Abl

~Loc

hon speech CONTind

3-place

64     

Abl

~Loc

Abs

honorific speech CT CONTdir

65     

Abl

~Loc

~Loc

Not yet attested! honorific speech CT CONTind

4-place

special negation patterns

66     

Aes

Abl

have (nothing) but (CT, TVP)

(30)

Aes

Abl

Abs

 

special numerical patterns

01

Abs

the two are one (LAD)

03b

~Loc

Abs

at X are  2, 3, 4 Y

13

~Loc

at X (are 2, 3, 4 (number adjektivals!), TVP v74)

D. Experiencer amplification (only WT)

33a

Aes

Abs

Abs

experiencer of 02

28a

Aes

Abs

~Loc

experiencer of 03a

30

Aes

Abs

Abl

experiencer of 04

31

Aes

Abs

Com

experiencer of 05

67     

Aes
Abs

Gen/Instr

experiencer of 15

68

Aes
Loc

Gen/Instr

experiencer of 16

28c

Aes

Aes

Abs

experiencer of 06

69     

Aes

Erg

~Loc

Not yet attested! experiencer of 07

70     

Aes
Erg
Abs

experiencer of 08

71     

Aes
(Abl
~Loc)
‡Abs

experiencer of 26a

46b
Aes
Aes
~Loc
Abs

experiencer of 28b (with /rden šes/)

72     

Aes
Erg
~Loc
Abs

experiencer of 09a

Consultants



top
B11 project page
B11 publications
SFB main page




"Language change and the fossilization of the Old Tibetan b- prefix in Ladakhi and Balti."
A Valency Dictionary of Ladakhi Verbs
Kenhat  

kuʃunaŋ trakuʃu (apples and peaches)on the comparison of comparative expressions in structurally differing languages







Kenhat – a group of Upper Ladakh dialects



(11th Himalayan Languages Symposium, hosted on 6-9 December 2005
by the Department of Linguistics, Faculty of Arts, Chulalongkorn University, Bangkok, Thailand.
The full paper is still to be published.)

The various dialects of Ladakh (and Baltistan) have been classified along phonetic features roughly in two main groups: (a) the ‘archaic’ dialects of Leh, Lower Ladakh, Purik and Baltistan, showing initial and final consonant clusters, and (b) the ‘innovative’ dialects of the south-eastern areas of Ladakh, where the clusters have been lost and tonal features might be found (Róna-Tas 1966: 21, Bielmeier 1985). However, on the basis of my recent fieldwork, I would propose to distinguish between
  • the historically younger, but lexico-phonetically conservative Shamskat dialects (wrT Šamskad ‘language of Lower’ La­dakh) in the west and north of Ladakh: Ciktan-Purik, Sham, Nubra, with the Balti dialects as a somewhat separate sub-group,
  • the historically elder, lexico-phonetically partly conservative, partly innovative Kenhat dialects (wrT Gyenskad ‘language of Upper’ Ladakh) in the centre around Leh, in the south and the south-east of Ladakh, obviously linked with the Tibetan dialects of Lahoul-Spiti and Western Tibet,
  • and – as a possibly independent group – the lexico-phonetically innovative tonal Chang­thang dialects for which sufficient data is not yet available.
The exact extension of the Kenhat dialects is unknown to the present date. The main centre apparently lies in the vicinity of Hemis Gompa (ca. 45 km south-east of Leh): Karo, Chemre, Sakti, on the right side of the Indus, Hemis, Upshi, and Martselang on the left side. The dialects extend to the south-east along the right river bank over Igoo and Shara, possibly including the Ronghat area (Roŋskad ‘language of the gorges’) with the villages Hemya (or Himya; located on both sides of the river), Liktse, and Kyunggyam on the right, Tarshit, Tiri, and Kere on the left side of the Indus. To the north-west, the most prominent phonetical feature, fricativisation of initial clusters (cf. Francke 1901: 6), extends only as far as Thiktse and upper She on the right bank (about 25 km an 15 km respectivly south-east of Leh), and perhaps up to Matho on the left bank (about 25 km respectively south of Leh). But the central dialects around Leh show at least regular fricativisation of clusters with voiced velar radical. Although the dialects arround Leh show a pronunciation close to the ¦amskat dialects, all of them share the grammatical feature of genitive agent marking as well as the use of particular evidential auxiliaries with the Kenhat dialects (cf. also table 1 and table 2).

To the south, the Kenhat dialects comprise the dialect of the eastern side valley of Gya-Miru Gyahat (Gyaskad) as well as the central and eastern dialects of the Zanskar valley Zăharhat (wrT Zanskarskad).

Along the Indus, there is a clear geographical boundary between the two main dialect regions. Snyemo (or Nyimo), the first village of the Shamskat area lies in a sharply cut basin, the south-eastern boundary of which is formed by some lines of higher slopes and the very narrow gorge of the Indus river. On the left bank of the Indus river rises a chain of steep mountains, and the Zanskar river coming from the south-west and flowing into the Indus at Snyemo could not serve as an inroad in the past, except in midwinter when it was frozen. The south-eastern slopes and the sand plains behind are and were certainly only a minor obstacle on the route to Leh, but they nevertheless mark a boundary between different climate zones, and thus also between different economic zones: the bassin of Snyemo and Bazgo is an outpost of the apricot horticulture, for which the lower areas are famous (see map).

The Zanskar valley itself is connected by old trade routes over Garzha (Lahaul) in the south to Spiti and Kulu, and by various other passes to the valley of Gya Meru as well as to the Changthang. To the north, the Zanskar valley is connected with the Balti-speaking areas of Purik via the Suru valley, while the lower part of the river is connected by treks via Wanla and Lamayuru to Lower Ladakh. The northern dialects of the Zanskar valley can be expected to be either part of the Shamskat group or to be strongly influenced by that group (cf. LSI p. 52). E.g. the dialect of Yulchung-Nyeraks is clearly Shamskat. The two villages are situated at both sides on the lower part of the river, shortly after the eastward bent near Lingshet. According to the Nyeraks informant (Tashi Angchuk, 1996) they were usually reached by a two to three days’ walk across three passes from Wanla (according to a new map, the road to Hanupatta via Sumdo might shorten the trek by at least half a day). There seem to be suitable geographical boundaries within the area, such as the Panzila on the trek from Padum to the Suru valley and several passes on the trek from Padum to Yulchung-Nyeraks to cut off the populations on each side, but I myself have never been in the Zanskar area.

Data from Kenhat was first obtained in 1996, when I recorded a short version of the Kesar epic in Gya. During my field stay in Ladakh in 2004, I transcribed the narration with the help of a Cemre speaker and discussed with her in some detail the grammar of the Gya dialect as well as of her own dialect. During the field stays in 2005 and 2006, I mainly worked with an informant from Gya-Sasoma, but I also had the possibility to conduct shorter interviews with a speaker from Shara and with a speaker from Hameling in Zanskar (somewhat south of Padum) as well as two speakers of the upper Shayok area. Comparative data is available only for the Manda dialect of Zanskar (Hoshi and Tondup Tshe­ring 1978).

The Kenhat dialects show a strong affinity with the Tibetan varieties of Himachal Pradesh (Spiti, Namkyad, and Tod) and, to a lesser extend, with the varieties of Western Tibet. The paper will discuss the following aspects:
  • phonology:
    • fricativisation of clusters
    • tonogenesis
  • morphophonemics:
    • preservation, generalisation, and eventual loss of the Old Tibetan suffixes -s/-d
    • neutralisation of verb stems
    • the Kenhat substitution of final -s
    • linguistic archaisms: frozen clusters in Gya compounds
  • lexicon:
    • definiteness marker {de}
    • remote deixis
    • first person exclusive plural pronoun /ɦoγo/ ~ /ɦaγo/ ~ /ɦoγa/
  • syntax:
    • bimorphemic case and case neutralisation
    • ablative subject marking
    • tense- and evidentiality-marking

Table 1:   Sound changes in Kenhat

 

(sub-) phone­mic tone

laryngalisation

palata­lisation

fricativisation of cluster

initial radical

me­dial

w > ɦ

y > ɦ vs. ɦ > y

ʃ > ç

voiceless

voiced

Sham

ř

LEH

ř

g

CEM

?

+

ř

+

k,p

g,b

voiceless

SHA

+

+

+

?

+

k

voiceless

GYA

+

+

ɦ-o,u

y-i,e

m-g

voiceless

HML

?

k,t,p

g,d,b

all

MAN

?

?

k,t,p

g,d,b

(all)

Table 2: Kenhat morphophonemics

 

loss of final -s after

genitive

agent

defi­nite­ness marker

evidential mar­ker

consonant

vowel

p,k,m

cluster>ř

um­laut

stem neu­­tralisation

 

‑se gen.

 

-ŋs

-gs

future

past

Sham

{po}

suk, khantsuk

LEH

+

{po}

ok, anok

k(y)ak

CEM

+

+

+

+

+

+

{de}

{kak}

{kanak}

SHA

+

+

+/substitute

+

+

+

{de}

?

?

GYA

+

+

+/substitute

+

+

+

{de}

ak, {kak}

{khanak}

HML

+

+

+

diphtong

+

?

hak

ř

MND

diphtong

?

?

?

?


The Kenhat dialects differ thus in many ways quite substantially from the Shamskat dialects and it does not seem to be suitable to discuss these differences as mere dialectal variance. The differences manifest themselves most obviously on the phonetical level (fricativisation and emerging tone vs. clusters) and on the grammatical level (genitive vs. ergative agent marking, verbal auxiliaries), but also on the semantic level.

From a merely phonetical perspective, the differences between the various Ladakhi dialects appears to be gradual, and it might seem justified to group the Leh dialect with the phonetically conservative Shamskat dialects. However this approach does not account for the essential difference on the grammatical level, due to which the Leh dialect can only be grouped with the phonetically innovative Kenhat dialects. The somewhat unexpected mixed character of the Leh dialect itself can be easily explained by historical facts (Leh as an important point of commercial exchange, repeated settlement of Balti speakers around Leh). Interestingly enough, it is the historically ‘younger’ dialects that have, by and large, influenced the historically ‘older’ one.

The findings do not only show that a classification of dialects cannot be achieved merely according to the phonetical surface, which may suffer more easily from external influences than the grammatical layer. They likewise show that a classification in terms of ‘conservative’ vs. ‘innovative’ might be quite misleading if based solely on phonetics. After all, the Kenhat dialects might have retained more lexical and grammatical archaisms than the Shamskat dialects.

Literature:

  • Bielmeier, Roland. 1985. ‘A survey of the development of western and soutwestern Tibetan dialects’. In: Barbara Nimri Aziz and Matthew Kapstein (eds.), Soundings in Tibetan civilisation. Proceedings of the 1982 seminar of the International Association for Tibetan Studies, Columbia University. Delhi: Manohar: 3–19.
  • Francke, August Hermann. 1901. Sketch of Ladakhi grammar. JASB 70: 1–63. Reprint under the title: Ladakhi and Tibetan grammar. Delhi 1979: Seema Publications.
  • Hoshi, Michiyo and Tondup Tsering. 1978. Zangskar vocabulary. A dialect spoken in Kashmir. (Monumenta Serindica, 5.) Tokyo: Institute for the Study of Languages and Cultures of Asia and Africa.
  • Linguistic Survey of India. Vol. III: Tibeto-Burman family, Part I: General introduction, specimens of the Tibetan dialects, and the North Assam group. Comp. and ed. by George Abraham Grierson. 1909 Calcutta: Superintendent of Government Printing. Reprint 1967, Delhi etc: Motilal Banarsidass.
  • Róna-Tas, András. 1966. Tibeto-Mongolica. The Tibetan loanwords of Monguor and the development of the archaic Tibetan dialects. (Indo-Iranian Monographs, 7.) The Hague etc.: Mouton.
Consultants


top
B11 project page
B11 publications
SFB main page



"Language change and the fossilization of the Old Tibetan b- prefix in Ladakhi and Balti."
A Valency Dictionary of Ladakhi Verbs
Clause types (an overview)

kuʃunaŋ trakuʃu (apples and peaches)on the comparison of comparative expressions in structurally differing languages






 


      

Language archaeology       
stone flowers
‘Steinblumen’, ‘stone flowers’, ‘rdowe mentok’ (found near Khalatse)

"Language change and the fossilization of the Old Tibetan b- prefix in Ladakhi and Balti."
Paper presented at the 12th Colloquium of the International Association for Ladakh, Kargil (Ladakh), 12-15th July 2005.
Consultants



top
B11 project page
B11 publications
SFB main page



A Valency Dictionary of Ladakhi Verbs
Clause types (an overview)
Kenhat

kuʃunaŋ trakuʃu (apples and peaches)on the comparison of comparative expressions in structurally differing languages












kuʃunaŋ trakuʃu (apples and peaches) –

on the comparison of comparative expressions in structurally differing languages
(from the abstract for Linguistic Evidence 2008, poster session, revised)

The German phrase Äpfel und Birnen may signal that one compares things that cannot or should not be compared. Its English counterpart apples and oranges is even more drastic, in so far as the two entities are essentially different, sharing only the shape, whereas the German pair stands for a difference in shape despite essential structural similarities. Similarities in shape often facilitate name derivation for unfamiliar items, cf. German Apfelsinen, similarly trakuʃu ‘peach(es)’ from kuʃu ‘apple(s)’ in Exot-ese (one of the many lesser-known, often structurally quite differing languages of the world).

These examples illustrate a basic dilemma: When comparing non-identical items – which is the main purpose of a comparison – one has to neglect part of the difference and to abstract either from the outer, formal features or from the inner, seemingly more substantial features. Although scientific discourse has often enough looked down upon the naďve mind, capable only to compare the outer shape but not the essence, both approaches may be found in the linguistic literature, e.g. originally functional labels such as the High German Perfekt are transferred on grounds of formal similarity to the southern dialects where the construction in question functions as a preterite, while the perfect function is filled with the so-called Doppelten Perfekt. Conversely, it is claimed by typologists that cross-linguistic comparison can only be based on meaning (Haspelmath 2004), following from which formal differences do not matter (much). Embedded nominalisation, e.g., is thus often equated with relative clauses in English, which is typically the only way to translate it appropriately. Note, however, the asymmetry in both practices: the dialects are described from the perspective of the ‘standard variety’, whereas lesser-known exotic languages are described from the perspective of linguistic ‘standard languages’ such as English (or, in earlier attempts: Latin). Both are not described as entities in their own right, nor are they ever accepted as descriptive models (or challenges).

This is not to say that meaning does not matter. In fact, even the modest tasks of translation or of describing (and thus understanding) an exotic language are based on the fundamental hermeneutic principle that however different the outer appearance (wording or structure), utterances are comparable as long as the intended or the conveyed meaning (the function) is the same, i.e. when referring to the same ‘objective’ situation.

How well this hermeneutic principle actually works may be demonstrated on the basis of a very small segment of linguistic utterances, namely comparative expressions of equality and difference. The situation in the outside world is quite manageable: We have two entities, A (the item to be compared) and S (the standard to which something is compared), to which we ascribe, for the sake of simplicity, a perceptible and measurable, i.e. scalable property X.

In English, scalar properties are typically expressed by adjectives or more precisely: adjectivals with nominal properties, and the relation of equality and similarity are expressed by the relators as … as and like, while the relation of difference is expressed by the relator than and a comparative morpheme -er added to the adjective, hence A is as X as S (equality), A is X like S (similarity), or A is Xer than S (difference).

In Exot-ese, the situation is somewhat more complicated: To start with, this language and its family did not originally possess basic adjectives, but only basic adjective-verbs (verbal adjec­tivals), which imply certain dynamic properties (inchoative, resultative), besides derived adjectives for states. The latter are used in comparative expressions of equality and similarity together with relators that correspond to as and like, but they cannot be used for a relation of difference. In some of the varieries, only certain forms of the verbal adjectival can be used. This holds also for analytic comparative constructions, since the quantitative adjectivals more or less are likewise of verbal character. The speakers of Exot-ese have thus to take refuge to a syntactical solution, namely to add a semantically opaque postposition to the S argument, which functions as a relator. The standard construction is: 

S-Postposition, A Xes. This might be interpreted as ‘In relation to S, A Xes’.

Another, somewhat less felicitous paraphrase, missing out the dynamic character of the property itself as well as the question what the expression really means, could be ‘In relation to S, A differs with respect of a plus in X’.

Exot-ese differs from English in many other respects. E.g. the negation markers are obligatorily bound to a verb or its auxiliary and thus always operate on the whole clause. In the case of constituent negation (nobody, not anybody), an indefinite or limiting quantifier plus an emphatic conjunction must be used, e.g. Anybody / A single person ever does not X in relation to S, but I fear, neither alternative has the same logical entailments as the English sentence Nobody is Xer than S.

In English it is possible, formally at least, to exchange S and the negated item A: A is Xer than nobody. Such sentences are acceptable when Xer than is not meant to express a relation of properties but a direct relation between the items, e.g. Something is better than nothing. But in Exot-ese, we cannot, on formal grounds, exchange the negated item, simply because it does not exist. Like in the case of the comparative construction or the constituent negation, we have to reformulate and reorder the various elements in order to arrive roughly at the intended meaning. Since we need a noun to which we can apply the postposition and since only sentences can be negated, we may take resort to an embedded nominalisation.

However, construction substitutes are often not very perfect matches, they may allow for certain ambiguities that are not there in the model (or vice versa). Even more, structural differences could well be symptoms of functional differences, which might become more evident when viewing the language in its entirety and not only a small segment. Finally, differences that might be still tolerable at an elementary level may accumulate, up to the moment where it is impossible to say, by any interpretative means or formal argument, whether the expressions in question can still be compared in a meaningful way, because it is no longer apparent that they still refer to (roughly) the same situation. In contrast to its English ‘counterpart’, the Exot-ese sentence A Xes in relation to anybody nonexisting is acceptable for some of the speakers even when expressing a relation of properties. This may be mainly due to the fact that it allows alternative interpretations, such as A Xes in relation to anybody else or A is as X as nobody else. But how do we know? If Exot-ese does not have constituent negation and also no comparative, are we not actually comparing here pears and peaches?



Reference

Haspelmath, M. (2004). Does linguistic explanation presuppose linguistic description? Studies in Language 28.3: 554-579.

Consultants




top
B11 project page
B11 publications
SFB main page



"Language change and the fossilization of the Old Tibetan b- prefix in Ladakhi and Balti."
A Valency Dictionary of Ladakhi Verbs
Clause types (an overview)
Kenhat



In a personal matter: A note on the "language" Exotese

This paper describes facts about a language called “Exotese” – unfortunately, I was totally unable to find any information about this language on the Internet as well as in other sources, and I have no other option than to think that this language including all the facts presented is nothing but a fiction of the author(s). [...]  the paper presents absolutely no information about the language resources where the “empirical facts about Exotese” were taken from (corpora, linguistic studies, etc.), and even not a single example of the language is given! [...] Even in the improbable case that Exotese were a truly exiting [! Perhaps rather: 'existing'?] language, this alone would disqualify the paper from presentation as a serious paper on a serious linguistic conference (on the other hand, it definitely could be a nice joke, and I do appreciate the choice of the name “Exotese” and the work the author(s) undertook in fabricating the paper (from an anonymous review to an anonymous abstract).

Admittedly, this reviewer was not the only one who searched in vain for this truely exotic "language" apologies to all that run into this trap!

However, it is one thing, not to recognise a word play (even if the clue to the riddle is given immediately in brackets), but it is another one to get seriously furious at an author if one does not find a quite obviously unrealistic (as admitted in the final remark) cover name in the Internet or elsewhere. Before devising a worst-case scenario of the above kind one could perhaps in dubio pro reo assume that the accused had some more serious reasons to use a pseudonym than just to fabricate a hoax. (And these reasons may perhaps also justify a reply to a rather funny instance of miscommunication, otherwise not being worth to be commented upon.)

In this particular case, the central idea behind the choice of the pseudonym (besides staying anonymous) was that the topic to be discussed was not so much a curious feature of the Ladakhi language, but the more general problem of how we can reasonably compare utterances of languages that differ quite fundamentally in structure from English.

Does it, in such metatheoretical context, really matter if an abstract does not contain any language specific data? How could the reviewer, who by all probability does not know much of Tibetan, not to speak of Ladakhi, how could s/he have been sure that the following example not yet 'officially' discussed in the literature is not fabricated? If it were enough to simply write down that this is real data from a real language, why should it then not suffice to give an abstract representation?

(1)       

den

do_

_rdemo

dug_

_jaŋ,

ʧi-aŋ

met-khan-i

naŋ-ʧig-basaŋ

KHAL

then

that.df

beautiful(adj)

be.exp

and

what-fm

NG2.exist/have/be.n.exp-nom-g

house-lq-rel


      example   meme Tondup Tshering, Khalatse (rec. 2006, discourse on the economic development of the region)

In a way, I would have been all to happy, if this sentence was fabricated, because in that case I could be quite sure of the underlying structure and its intended meaning. But the truth is, that sentences like this one literally threw me into an existential crisis, taking me to the very limits of interpretation. I am still not absolutely sure that I fully understand what this example sentence means for the narrator or any other native Ladakhi speaker.

According to the consultants from whom I elicited similar sentences, there are basically two contradictory interpretations available (a) the modest excess variant: A Xes (only) in relation to a non-X S and (b) the extreme excess variant: A Xes in a way no S can, cf. the following (non-recorded) example:

(2)      SKI su-aŋ riŋmo met-kan-basaŋ Tsheriŋ (riŋmo) duk.
LEHa/b   
su-ʒig-aŋ *(riŋmo) met-kan-esaŋ Tsiriŋ riŋmo duk.
  who-(lq)-fm long(adj) NG2.be.n.exp-nom-rel name (be.long(adj)) be.exp

 

‘In relation to whosoever not being tall (assimilated knowledge), Tshering is (tall) (visual evidence).’
~ modest excess: Tshering is taller than anybody who is not tall. (SKI: %, LEHa: %%, LEHb: *).
~ extreme excess: Tshering is taller than nobody else (i.e. extraordinary tall). (LEHb: o.k., LEHa: %)

The symbol "%" signals that the acceptance or interpretation depends on a supporting context beyond this utterance, "%%" indicates that this context must be made explicit, in oder to rule out the alternative interpretation.

Sentence (1) mainly differs in word order, which is rather free in Ladakhi. It is extremely common to extract constituents or even parts of them and add them after the verb, like an afterthought. So, what is the meaning of (1)? Apparently, the meaning depends on the context, but by stating this I immediately run into a fallacy, since the context also depends on the interpretation of this sentence. The least I can say, is that in this very utterance it is not implied that S does not have any property or is not beautiful at all, and that A is thus not very beautiful itself (modest excess variant). Nor does it imply that A is extremely beautiful (extreme excess variant).

As to the context: The narrator first describes a representative room that although possessing attributes of wealth and modernity is not very beautiful in his eyes, because it is ‘empty’. He then contrasts it with a room of a more traditional house where barley is heaped up in the corners (as if this could make a room more comfortable) and continues with the above sentence (1). In fact, the relational morpheme -(b)asaŋ / -(e)saŋ can be also used to express the contrastive, hence non-comparative relation instead or rather than. A tentative ‘literal’ translation runs as: ‘Then THAT one is/was beautiful, in contrast to / instead of the house that does not have anything.’ While the intended meaning could perhaps be rendered as: ~ Now THIS one is really beautiful, NOT the other house that hasn’t anything.

The main problem, still to be resolved, is: what kind of model sentence should the Ladakhi sentence be related to? Initially, I elicited sentences of this type as more or less formal transformations of the non-sensical English sentence A is Xer than nobody, as indicated in the abstract above. But could it perhaps be the case that the Ladakhi sentence, lacking any kind of overt comparative morphology, simply does not mean what I expected it to mean? Could the disturbing notion of an explicit non-equative comparison, which makes the English sentence uninterpretable, simply be lacking in our Ladakhi sentence, despite the use of a relator? If it is true that any kind of property ascription, such as A is X, automatically implies that a certain threshold level of X must have been reached or surpassed, could it then be the case that the Ladakhi sentence does not contain an expression of non-equative comparison but simply an expression of property ascription: A is X. If this should be the case, than the Ladakhi sentences (1) and (2) could be equivalents only to a model like

A is X – modified by an underspecified negative relation with S.

It seems that I had been, in fact, trying to compare pears and peaches. But how do we, the descriptive linguists, typologists, etc., in the best case: the non-perfect, non-native speakers of a minor language, how do we know what to compare and what not? And what about the most common worst case, namely that we have no further knowledge of the language under discussion: How can we be sure that the analysed sentence does mean what it is said to mean?

The main problem is not the lack of specific data, the main problem is
the data itself...
                                                                                             

... and, of course, (the biased mind of) the researcher.

Abbreviations:

Dialects and interlocutors or language consultants

KHAL       Khalatse (Shamskat), non-elicited data (meme ‘grandfather’ Tondup Tshering)
LEH           Leh (Kenhat), elicited data (a: Thrinlas Wangmo, b: Tsewang Dorjey; siblings)
SKI            Skindiang (Shamskat), elicited data (Choron Angmo)

Descriptive terms

ADJ           derived adjectival of nominal character
DF             definiteness marker
EXP           experiential construction (direct visual evidence, new knowledge)
FM            focus / emphatic marker
G               genitive marker
LQ             limiting quantifier ‘a, some’
N.EXP       non-experiential construction (assimilated knowledge)
NG            negation marker or negated form (NG1 = mi NG2 = ma)
NOM         nominaliser
REL           relational marker (relator)
VA             verbal adjectival
x_ _y          phonetically conditioned features across word boundaries


The general problem outlined here was discussed in a
presentation on the 23rd Scandinavian Conference of Linguistics, Uppsala University, 1-3 October 2008, which is submitted for publication. More of the recalcitrant Ladakhi data will be found in a detailed descriptive study under preparation:

"mindra mindra (incomparably different): Property ascription and expressions for difference (and comparison) in Ladakhi Tibetan", propably appearing in the online journal Himalayan Linguistics Review.

All
consultants


top
B11 project page
B11 publications
SFB main page



"Language change and the fossilization of the Old Tibetan b- prefix in Ladakhi and Balti."
A Valency Dictionary of Ladakhi Verbs
Clause types (an overview)
Kenhat     
kuʃunaŋ trakuʃu (apples and peaches)on the comparison of comparative expressions in structurally differing languages



Technical notes:

"High altitude sickness" of computers and hard drives

Hard drives for electronic notebooks or laptops are commonly not specified for operation in altitudes higher than 3.000 m above sea level. Obviously, the air becomes too thin to support the head at its proper operation heigth. If the air heats up through operation, it becomes even thinner (an effect that in earlier years repeatedly made Indian Airlines, starting from Leh at noon time, unload passengers and luggage, rather than to change its flight schedule - a state enterprise). Although the head might not crash, it might have difficulties to read and write, leading to ongoing fragmentation of all files.

Rather unplanned and with out any gratification for the laboratory rabbit, the field trips to Leh, situated in 3.500 m, turned out to be online experiments, testing the quality of hard drives. Three trips up to now, three hard drives: Toshiba (I), Hitachi, Toshiba (II). Here is the report:

2002, Toshiba (I): two or three days after arriving in Leh, the laptop started producing a very disturbing creaking or squealing noise when accessing the hard drive, as if metal was scratching on metal. Starting up the system and particularly saving the files became slower and slower every day, in the end taking about 5 to 10 minutes for starting up and several minutes for the system's security back up of a small file. Eventually, after only two weeks in the field, the following message appeared on the screen:

"Unknown Hardware or Software Error"

and the laptop stopped working. It was not a total crash, though, the laptop could be started again, after cooling down and would run for about an hour until the ventilator would set in for the first time. I used the laptop thus to backup all the work done in internet cafes. All computers around me were working without any problem.

Back in Delhi (merely 250 m or so above sea level), starting the laptop for a backup, I was immediately startled by the absence of the creaking noise and realised that the starting time had been considerably reduced again. The laptop worked smoothly again, for hours and hours.

2003, Hitachi: Started grumbling in a deep voice after some days, particularly after heating up, obviously also fragmenting the files, but did not become slow in a perceivable manner. Worked three month without problem.

2004, Toshiba (II): Started with the creaking noise after about three weeks of operation. Became considerably slow while saving both the document and the securitiy backups, thus I decided to run the defragmentation in the sixth or seventh week. I am not sure whether that was a wise decision, since the files of the operating system could not be defragmented (and perhaps underwent further fragmentation during the process). Operating the computer thus was going on with increasingly reduced speed (eg opening up the media player sometimes took about three minutes) and much more noise, and the last week I reckoned every hour on having to exchange the hard drive. Document saving, however, worked smoothly again until the end of the field trip.

Recommendations: Although other people's computers and laptops seem to be working without any problem at Leh and similar altitudes, new generations of hard drives might be even more sensitive to thin air, and there is no guaranty that one gets the error message in time before the head crashes. One should thus think of alternatives involving no hard drive at all. If the laptop can boot from the USB drive one could have the operating system on the USB stick, otherwise one could use Linux on CD (Knoppix).

To be continued...

... in fact

like in an ordinary evolutionary setting, a not-so-fit hard drive can also have some advantageous sides. In this case, as it was much more sensitive to heat than the fit hard drive, its unexpected non-functionality within 5 min. of use, turned out to be an important signal that the laptop was constantly overheated.

The reason for this was that the ventilator did not switch on any more, but who would realise, in the middle of one's work, whether or not the ventilator had switched on or of inbetween. Thus in 2006 I had been working happily for about 3 weeks on a laptop with fever. No wonder that the flashcards always felt unusually warm...


Electrone harvesting:

factory
Feeding two mouths with four empty spoons:
2 car batteries, 4 chargers, 1 stabilizator, but
no electricity in Leh.

In theory and as long as the water of the Indus river is not frozen, a hydrel power plant provides Leh with electricity during the day time, while in the evening hours from seven to eleven (when the hydrel power is transmitted to villages in the vicinity of Leh) power is provided by a public diesel generator. (Don't ask me, why this switch is necessary. Its main reason seems to be to make life even more complicated. Before 2005, it regularly caused hundreds of e-mails to be unsent, as the computers in the internet cafés went down.)

Practically, there are many reasons why any of the two plants does not work propperly.

In May and June 2005, for no obvious reason (except perhaps that the snowmelt was late) we did not get any electricity during the day time (except one or two hours on a few days). But it seemed that the people from the power department were trying their best to compensate, as the diesel generator operated for five, sometimes even six hours in the evening.

In July 2006, Ladakh suffered severely from heavy rain falls. As a result of the floods, the hydrel plant went out of work for about two months, because sand and stones had been swept into the canal. There is no kind of lock at the head of the canal, which could prevent such calamities. Sharp tongues assert that this is not because the engeneers did not think of the possibility of floods, but just because they did: who be so 'stupid' and construct something useful that would make his further employment superfluous. The only problem is, that the money for the repair works does not flow as fast as the floods. (Fortunately for all Ladakhis, there was yet the Ladakh festival to come, at the beginning of September, serving the promotion of tourism, and this was reason enough to 'speed up' the process.)

But unlike the year before, the loss at the daytime was by no means compensated in the evening, although everybody would have been happy if the normal quantity (four hours daily) would have been supplied. In the contrary, diesel generated evening power was supplied only every alternate day (in some sections of Leh, as I heard, not even on every alternate day!) between seven and eleven, but hardly more than two hours all together, and with the lowest voltage possible. As a result,
every other evening one could not breathe freely in the bazaar: all shopkeepers were running their own small generators.

The laptop batteries (Panasonic) would each run for three hours (much better batteries are available now for Levono and perhaps other products) and would take the same time to recharge, while working on the computer. Provided, they could be recharged in the night time, I did not need more than about two hours from an external source, i.e. a 12 V car battery (during this time one of the batteries was again partly charged). To recharge a car battery, however, takes about three or four times the working time, as the corresponding chargers are very slow. According to the display, the maximum output is 6 A, but one is told to set the output on maximally 3 A. Even this might be too high, since the voltage provided by either power plant might suddenly rise
and surrogate fuses are not always available in the bazaar. When leaving for dinner, it was thus advisable to turn the ouput down to 1-2 V. Yet, having upgraded my 'power plant' from one to two car batteries with two chargers each, I might have been content with two hours power supply daily...

At this point I would like to acknowledge the solidarity shown by the staff from a neighbouring hotel and by the owner of an internet café, allowing us to work in their rooms while their generators were running.

Something about the prices (2006):
A precharched car battery (Exide) is available for about 3.150 Rs, a good charger for about 1.500 Rs. Generators are available in Leh from 15.000 Rs on (they are hired out for a daily rate of 1.000 Rs!). Small subsidised solar panels (below 20 VA/h) might be available second hand for about 20.000 Rs., but a workable (and unsubsidised) solution would have required an investment of at least 300.000 Rs (exchange rate 1 Euro = 56-57 Rs).

Backpackers' solutions?
A colleague, working in Zanskar in 2008, had spent quite some money on a foldable solar panel equipment, but while the solar cells seem to have been working (at least for some time), the supplied battery had neither any tolerance for
overcharging nor any indication of the charging status (as a matter of course, both features are found with car batteries), nor did the charger have an automatic stop device (as modern chargers for litium batteries have), thus the battery broke down immediately after the first attempt of charging. Find a suitable battery in the bazaar of Fadum! The batteries that were available, were, as it turned out, of no great help. Or perhaps also the solar panel simply did not provide enough energy. Needless to say that there was no technical support from the side of the vendors.

Reasons for optimism?
In 2007 and 2008 we did not suffer any major power cut in Leh.



top
B11 project page
B11 publications
SFB main page



"Language change and the fossilization of the Old Tibetan b- prefix in Ladakhi and Balti."
A Valency Dictionary of Ladakhi Verbs
Clause types (an overview)
Kenhat

kuʃunaŋ trakuʃu (apples and peaches)on the comparison of comparative expressions in structurally differing languages



Layout: Christoph Singer. Responsible for the content: B. Zeisler. Last modified: 24.12.2008