japanese, marathi and other asian languages...1. introduction •the aim of this presentation: to...

Synchronic exploration in search of diachronic paths: A contrastive study of

the grammaticalization of “PUT/KEEP” in Japanese, Marathi and other Asian

languages

Prashant PARDESHI

NINJAL INTERNATIONAL SYMPOSIUM 2013

MYSTERIES OF VERB-VERB COMPLEXES IN ASIAN LANGUAGES

15th Dec. 2013

Beginning remarks

• Empirical work on the vector verb PUT/KEEP

• [V1+PUT]Genuine compound verb: clears all the standard diagnostic tests (I am not confused!!)

• Type 4 of the position paper

• Popular: at least in this symposium

3 poster presentations on the vector PUT

2

Beginning remarks

• Hiroko Koto and Nerida Jarkey (University of Sydney): "The nature of V1 in the -te oku construction in Japanese“

• 一色舞子（北海道大学）:日本語の「-ておく」と韓国語の「-e twuta/-e nohta」における史的対照研究

• 日髙晋介（東京外国語大学大学院／国立国語研究所）:日本語「ておく」とウズベク語-(i)b qo'y-の対照

3

1. Introduction • The aim of this presentation: to explore the

process of grammaticalization of the “vector verb” PUT/KEEP in Asian languages with a special focus on Japanese (ok-u) and Marathi (Thev-Ne).

• Methodology: synchronic contrastive study of [V1+ PUT/KEEP] concatenation

Macro-level study: contrastive study of Japanese with selected N-E, Central and South Asian lgs.

Micro-level study: contrastive study of Japanese with Marathi

4

1. Introduction • Assumption: the direction of semantic change is

the same but that the speed of the change may vary from one language to other.

• If so, a comparative study of vector verbs across the languages of Northeast, Central, and South Asia provides us an opportunity to examine the process of grammaticalization at different points in apparent time and to follow it through successive phases (see Hook 1988).

5

1. Introduction

• Such an exploration in “space/geography” may provide a glimpse into the “time/history” of the trajectory of change.

• I will focus on the co-occurrence restrictions of a given v2 with the V1

• The co-occurrence restrictions may be taken as a barometer of the degree of semantic bleaching or grammaticalization of the v2 in question.

6

2. Background: What is an ECV? • Masica (1976: 141-158)

Explicator compound verb (ECV)

Main verb (V1)+ vector verb (v2)

tabe-te ok-u (eat-CP keep)

tabe-te shima-u (eat-CP finish)

tabe-te age-ru (eat-CP give) ….

• one of the hallmarks of the languages spoken in Northeast, Central, and South Asia (Type 4 as per the position paper).

7

2. Background: Distribution of ECV

• Masica (1976: 151)：Explicator compound verb

8

2. Background: Form

• ECV: V1(non-finite)＋v2(finite)

• V1＝Primary/Polar/Main verb

• v2＝Secondary/Explicator/Vector verb

• The set of v2s is a closed class：GO, COME, GIVE, TAKE, THROW, STRIKE, FALL, SIT, GET UP, PUT/KEEP, etc.

• ECV [V1+v2] alternates with its non-ECV counterpart [V1]:

mi-ru (see) : mi-te ok-u (see + put/keep)

9

2. Background: Semantics & Function • Semantics： V1 is the semantic center of gravity and

v2 is semantically bleached.

• Function of a v2: the function of V2 is varied and manifold: it connotes completion, suddenness, directionality, benefaction, intensity, violence, stubbornness, reluctance, regret, forethought, thoroughness etc. depending on the item involved and on the circumstances (Masica 1976: 143)

(Cf. Mastumoto’s handout: subjective color perspectiveal, interactional, attitudinal etc.)

• PUT/KEEP: forethought, future use in view, determinedness

10

2. Background: Distribution of PUT/KEEP

• Masica (1976: 146)：Chief explicator auxiliaries

11

2. Background: Distribution of PUT/KEEP

• Masica (1976: 147)：Chief explicator auxiliaries

12

2. Background: PUT/KEEP

• Among vector verbs, PUT/KEEP is special in the sense that it often conveys almost the same meaning across a large number of languages: perform V1 with the expression of forethought or future use in view as one of its functions.

(1) J kangae-te ok-e asoko-ni ok-e

think-CVB put-IMP there-LOC put-IMP

(2) M vichar.kar-un Thev tithe Thev

thought.do-CVB put-IMP there-LOC put-IMP

Think it over (beforehand)! Put it there!

13

See Zeisler (handout): exx. 18, 23

2. Background: Decategorialization

• Hopper & Traugott (1993: 103): consider vector verb as a case of decategorialization.

• Decategorialization: spread of grammaticalization along a path or cline of structural properties, from a morphologically “heavier” unit to one that is lighter, that is, from one that tends to be phonologically longer and more distinct (e.g. with stressed vowels) to one that tends to be less distinct and shorter.”

14

2. Background: Decategorialization • A verb-to-affix cline [Hopper & Traugott 1993: 108-

109]

• full verb > (vector verb >) auxiliary > clitic > affix

• She keeps indoors on cold days. (full lexical verb)

• yoku kangae-te ok-e (vector verb ) [Japanese]

well think-CP put/keep-IMP

Think over (it) well beforehand.

• She keeps watering the tomatoes. (auxiliary)

• We’ve built a new garage. (clitic)

• ils passer-ont (affix<cliticized aux ‘have’) [French]

‘They will pass’

15

2. Background: Decategorialization

• The size and diversity of the set, that is the low degree of specialization as auxiliaries displayed by its members, is one of the factors that points the need to think of them as intermediate between full verbs and auxiliaries. [Hopper & Traugott 1993: 109]

• The compound verb as a whole can be followed by an auxiliary verb. The aux verb shows grammatical distinctions of tense, aspect, person, and number.

tyaa.naa kaLav-un Thev-le aahe-s kaa?

them-to inform-CVB put-PST.NSg be.PRS-2PSg Q

Have you informed them? 16

2. Background: co-occurrence restrictions

• The concatenation of [V1+v2] is lexically selective: A given v2 combines only with such V1 as are compatible with it, or to put it another way, as have semantic potential for it (Masica 1976: 143).

• Little enough is known about intra-linguistic co-occurrence restrictions of a given v2 with the V1.

17

3. Explorations in co-occurrence restrictions

• Two studies on the co-occurrence restrictions of the vector verb (v2) PUT/KEEP with the main verb (V1)

• (1) Macro-level study: Languages of Northeast, Central and South Asia

Data source: questionnaire-based elicitation

• (2) Micro-level study: Japanese and Marathi

Data source: Large-scale corpus (BCCWJ, app.104 million words) for Japanese and on-line data for Marathi (supplemented with my intuition)

18

• Macro-level study: Languages under consideration

• Northeast Asia: Japanese, Korean, Mongolian (of Inner Mangolia)

• Central Asia: Kazakh, Uzbek, Kyrgyz, Tajik, Uyghur

• South Asia: Hindi, Marathi, Telugu, Tamil, Malayalam

19


• Macro-level study: Questionnaire based elicitation from native speakers

• Consultants：Native speakers of the concerned languages (and also fluent in Japanese)

• No. of consultants： 1 per language（5 for Korean）

• The data represents the dialect/idiolect of the consultant.

20


• Macro-level study: Consultants

• Korean: Dr. J-M, Kim, Dr. H-J Chung, Y-N Li, Y-J Kwon and J-H Kwon

• Monogolian: Badema (Wūlánchábù City, Inner Monogolia, China)

• Uzbek: Mukhamediev Nozim Mufazzalovich (Tashkent, Uzbekistan）

• Tajik: Fariza Abidova Mirmuhammadovna(Samarqand, Uzbekistan)

• Kyrgyz: Djamilia Soltobaeva (Ak Tuz, Kyrgyzstan)

• Kazakh: Saltanat Abdildina (Astana, Kazakhstan)

• Uyghur: Gulizar Keyim (Ghulja City, Uyghur Autonomous Regions, China)

• Hindi：Sunil Lakhera (New Delhi, India)

• Telugu: Dr. SAR Reddy (Kakinada, India)

• Tamil: Verrappan Veerasureshkumar (Madurai, India)

• Malayalam: Dr. P. A. George (Thiruvananthapuram, India)

Thank you very much !!!

21


22

V1

1. transitive, exx. 2-12

2. ingestive-type semi-transitive,

exx. 13-17

3. reflexive-type semi-transitive,

exx. 18-21

4. unergative, exx. 22-31

5. unaccusative, ex. 32 (to die)

See appendices for a legible version and the relevant data.


• V1 classification: 5 groups in the order of decreasing transitivity:

• Transitive (to clean, to put on light, to cook, to write, to wash clothes, to hide something, to reserve, to convey, to think about s.t., to put/keep) exx. 2-12 [exx 11, 12: put+put]

• Ingestive-type Semi-transitive (to listen/hear, to see, to drink, to eat) exx. 13-17

• Reflexive-type semi-transitive (to wash oneself, to wash one’s hands/face, to brush one’s teeth) exx. 18-21

• Intransitive: unergative (to swim, to sleep, to hide oneself, to run away, to jog, to run, to stand, to sit) exx. 22-31

• Intransitive: unaccusative (e.g. die) ex. 32 Check if the above can co-occur with the vector PUT/KEEP.

23


3.1: Findings

24

3.1: Findings

25

3.1: Findings: Summary

26

• The range of V1 in Japanese is broader than that in all other languages under consideration.

• Transitive verbs: all the languages under discussion allow V1+PUT combination.

• Combination of PUT+PUT (o-ite-oku>oitoku): Northeast and Central Asian languages allow such combinations while South Asian languages do not (cf. Hindi: le le [take + take], de de [give + give])

3.1: Findings: Summary

27

• Ingestive-type semi-transitive, reflexive-type semi-transitive, unergative verbs: very few languages barring Japanese and Korean permit concatenation with PUT/KEEP.

• Unaccusative verbs: only Japanese allow concatenation of unaccusative verb and PUT/KEEP (only combination found so far: shin-de oku [die + put]).

• Japanese data: V1-te/de-oku

BCCWJ corpus (Approx. 104 million words)

Search tool used: NINJAL-LWP for BCCWJ (NLB)

28

4. Micro-level study: Japanese vs. Marathi

http://nlb.ninjal.ac.jp/

http://nlb.ninjal.ac.jp/

• Marathi data: V1-un Thev-Ne

Online search, my native intuition

29

4. Micro-level study: Japanese vs. Marathi

dar-veLes-pramaaNe kaahi uttara benc-var lih-un Thev-to

every-time-like some answers bench-on write-CVB put-PRS.M.Sg

Like I do every time (I) will write down some answers (of the questions that are

most likely to appear in the examination paper) on the bench.

(www.tarunbharat.net)

4. Micro-level study: Japanese & Marathi

30

Two environments investigated:

1. Range of V1 co-occurring with oku classified on the basis of their transitivity

2. Negation of V1 (in contrast to shimau for the Japanese data and TaakNe for the Marathi data)

4. Micro-level study: Japanese

64%

26%

2% 8%

Transitivity of V1

Tr

Tr/Intr

Intr

Unclear

31

Japanese: V1 [V1-te/de-(oku)] (affirmative V1)

Transitivity V1 Freq (tokens)

Percentage

Tr 18,282 63.7%

Tr/Intr 7,361 25.7%

Intr 750 2.6%

Unclear 2,302 8.0%

Total 28,695 100%

Data source: BCCWJ (104 million words, search tool: NLB

1. Range of V1 co-occurring with oku classified on the basis of their transitivity

Ssemi-transitive verbs: EAT & DRINK (ingestive)

（3）今のうちにおいしいものを食べておきたいし。。。

ima-no-uchini oishii-mono tabe-te ok.i-tai-shi..

now-GEN-within delicious-things eat-CVB put-optative-and

(I) would like to eat delicious things before it is too late.

（4）牛乳をもっと飲んでおけばよかった。

gyuunyuu motto nonde-de ok-eba yokatta

milk more drink-CVB put-COND good.PST

It would have been better to drink more milk.

32


Intransitive (unergative verbs)

（5）たっぷり寝ておけ

tappuri ne-te ok-e

amply sleep-CVB put-IMP

Get plenty of sleep!

（6）ショー開始１０分前にはぜひとも座っておこう

shouu-kaishi-10-pun-mae-ni-wa zehitomo

show-start-10-minutes-before-at-TOP by all means

suwa-tte ok-ou

sit-CVB put-volitional

By all means (I should) take a seat at least 10 minutes before the show starts.

33


Intransitive (unaccusative)

（7）私はもっと前に死んでおけばよかったのです。

watashi-wa motto maeni shin-de ok-eba yo.katta-no desu

I-TOP much before die-CVB put-COND good.PST.NOML COP

It would have been good had I died way before.

(村上春樹著『ねじまき鳥クロニクル』, 1994, 913)

（8）もう少し早くやっておくべき、思いついておくべき実験だったかもしれない。

mou-sukoshi-hayaku ya-tte-oku-beki, omoitui-te-oku-beki

little-more-early do-CVB-keep-ought strike-CVB-put-ought

jikken-datta-kamoshirenai

experiment-be.PST-perhaps

Probably it was the experiment which (I) should have conducted, conceived/thought of a bit early.

(http://blog.livedoor.jp/cdc2kinase/archives/2012-02.html)

34


35

Main verb Japanese (NLB) tokens

Marathi (Google) tokens

Marathi (Google) tokens

PUT (oku) PUT (ThevNe) TAKE (gheNe)

DO (TR) 3486 Few hundreds Few hundreds

WRITE (TR) 369 Few hundreds Few hundreds

EAT (SEMI-TR) 31 1 30

DRINK (SEMI-TR) 19 1 20

SLEEP (INTR-UNERG) 12 0 7

SIT (INTR: UNERG) 3 0 17

DIE (INTR: UNACC) 2 0

2 (both in poems)

4. Micro-level study: Japanese & Marathi

1. Range of V1 co-occurring with ThevNe classified on the basis of their transitivity

I guess that the amount of data available on Google for Marathi would be more

than the sample I used for Japanese taken from the BCCWJ (104 million words).


36

Japanese: negation of V1 [V1-nai-de-oku]

Transitivity V1 Freq (tokens)

Percentage

Tr 59 43.4%

Tr/Intr 56 41.2%

Intr 16 11.8%

Unclear 5 3.7%

Total 136 100%

2. Negation of V1

43.40%

41.20%

11.80%

3.70%

Transitivity of V1

Tr

Tr/Intr

Intr

Unclear


37

(9) 来年は入札しないでおこう。

rainen-wa nyuusatsu shi-nai-de ok-ou

next.year-TOP bid do-NEG-CVB put-VOL

The next year (I) should not tender a bid.

(10)今は何も言わないでおこう。

ima-wa nanimo i-wanai-de ok-ou

now-TOP nothing tell-NEG-CVB put-VOL

At the moment let us not tell (it) to anyone.

(11)君の仲間は殺さないでおいたよ。

kimi-no-nakama-wa koros-anai-de o-ita-yo

you-GEN-buddies-TOP kill-NEG-CVB put-PST-SFP

(I) did not kill your buddies.

2. Negation of V1 with oku (KEEP) vs. shimau (FINISH)

38


Transitivity V1-nai-de-oku V1-nai-de-shimau

Tr 59 43.4% 1 25%

Tr/Intr 56 41.2% 0 0%

Intr 16 11.8% 3

75%

Unclear 5 3.7% 0 0%

Total 136 100% 4 100%


39

The top ten most frequent V1s in Japanese:

V1 with Positive polarity V1 with Negative polarity

N = 28695 N = 136

する suru （3357） ‘to do’ する suru (22) ‘to do’

入れる ireru（790）‘to put in’ 言う iu (22) ‘to tell/say’

言う iu （729）‘to tell/say’ 出す dasu (5) ‘to take out’

放つ hanatsu（667）‘to release’ 忘れる(4) wasureru ‘to forget’

知る shiru（588）‘to know’ 書く kaku (4) ‘to write’

置く oku （548）‘to put/keep’ ふれる fureru（2）‘to touch’

覚える oboeru（507）‘to remember’ 会う au（2）‘to meet’

つける tsukeru（412）‘to attach’ 入れる ireru（2）‘to put in’

見る miru（377）‘to see/look’ 切る kiru（2）‘to cut’

書く kaku（361）‘to write’ 吸う suu（2）‘to smoke’


• Passive in V1 (cf. Kageyama’s handout Table 2)

• Type 4: Passive on V1: OK

40

Passive on V1

Form V1 token

frequency V1 type

frequency Type to token

frequency ratio

V1+(ra)rete-oku 1 1 1.0000

V1+(ra)rete-shimau 4,813 850 0.1766

V1+te-oka-reru 49 12 0.2449

V1+te-shimawa-reru 38 36 0.9474

4. Micro-level study: Marathi 2. Negation of V1: ThevNe (KEEP) vs. TaakNe (THROW)

41

Main verb Marathi (Google) tokens

Marathi (Google)

tokens

Marathi (Google)

tokens

Marathi (Google)

tokens

V1-ThevNe (PUT)

V1-NEG-ThevNe (PUT)

V1-TaakNe (THROW)

V1-NEG-TaakNe (THROW)

DO (TR) Few hundreds 2 Few hundreds 1

WRITE (TR) Few hundreds 20 to 30 30 to 40 0

EAT (SEMI-TR) 1 0 70 to 80 1

DRINK (SEMI-TR) 1 0 10 to 15 0

SLEEP (INTR) 0 0 0 0

SIT (INTR) 0 0 0 0

DIE (INTR) 0 0 0 0

4. Micro-level study: Marathi

42

• The negation of V1 (V1-nai-de oku) is more wide spread [extended to all degrees of transitivity of V1] and relatively more frequent than it is in Marathi (V1-un Thev-la/li/la/Naar naahi, Thev-u nakaa/nako/naye etc.).

(12) tyaa-ne aatmahatye-puurvi ciTThi-hi

he-ERG suicide-before note-EMPH

lih-un Thev-li naahi

write-CVB put-PST.F not

He did not even write a note before committing suicide.

(maharashtratimes.indiatimes.com/.../5916518.cms)

• Speaker is reporting to a normal expectation which has been broken.

5. Summary & Conclusions

• The results of the macro cross-linguistic study:

• Transitive verbs: almost all the languages under discussion behave alike.

• Combination of PUT+PUT (o-ite-oku>oitoku): Northeast and Central Asian languages allow such combinations while South Asian languages do not (cf. le lenaa [TAKE+TAKE], de denaa [GIVE+GIVE], *rakh rakhnaa [PUT+PUT] in Hindi).

43


• Semi-transitive (ingestive, reflexive) and unergative verbs: very few languages barring Japanese and Korean permit concatenation with PUT/KEEP.

• Unaccusative verbs: Although not so widespread, only Japanese allows unaccusative+PUT (DIE+PUT) combination.

44


• The results of the micro study (Jpn vs. Mar):

V1+PUT combinations

• The range of V1s in Japanese is more broad than in Marathi (transitive, semit-transitive [ingestive, reflexive] and intransitive verbs).

• The range of V1s with negation in Japanese is slightly broader than in Marathi.

45

• Further, the vector oku (PUT/KEEP) in Japanese, unlike its counterparts in other Asian languages including Korean, undergoes phonetic erosion/attrition: o-ite-oku>oitoku (Bybee and Pagliuca 1985: 59-60).

46



• From these facts, I argue that on the continuum of grammaticalization the vector PUT/KEEP in Japanese has advanced further as compared to its other North-East Asian, Central Asian and South Asian counterparts.

• SALs > Central Asian> Other North-East Asian > Japanese

47

5. Summary & Conclusions • Type 4 V-V complexes are lexically selective.

• More work needs to be done on each vector in terms of (i) syntactic valency of the V1s, (ii) semantic class of V1s, (iii) Type frequency of V1s, (iv) token frequency of V1+V2 combination etc., both within a single language and across languages.

• More attention needs to be given to the context (specific constructions) in which CVs are favoured and dis-favoured (see Hook 1991).

48

• A more fine-grained analysis of synchronic data from Asian languages (involving elicitation from a range of speakers, attested examples from large scale corpora etc.) is required to confirm or to falsify the conclusions drawn in this study.

• Comparison with other vector verbs within the same language is also necessary.

• Wherever possible diachronic study should be carried out.

49


감사합니다 Thank you very much धन्यवाद !

50

[SEE + PUT] Mt. Fuji today!!!!

• Appendix 1: Questionnaire

• 14. When I had been to Japan I should have seen Mt. Fuji.

• 日本へいった時に富士山を見ておくべきだった。

51

Mt. Fuji view from NIJAL (4F) 15th Dec 2013, 9:10 AM

52

Mt. Fuji view from NIJAL (4F) 15th Dec 2013, 9:10 AM

53

Thank you very much !!!

• Comments, criticisms, suggestions welcome !!

• Acknowledgement

The research reported here is supported in part by grants from (i) the Japan Society for the Promotion of Science (#18520314, 18520290) and (ii) NINJAL.

• Contact e-mail address:

[email protected]

54

Selected references • Bybee, Joan and William Pagliuca. 1985. Cross-linguistic comparison and the

development of grammatical meaning. In Jacek Fisiak (ed.) Historical semantics, historical word formation. Berlin: de Gruyter. Pp. 59-83.

• Hook, Peter. 1988. Paradigmaticization: A Case Study from South Asia. In Proceedings of the Fourteenth Annual Meeting of the Berkeley Linguistics Society. Pp 293-303.

• Hook, Peter. 1991. The Emergence of Perfective Aspect in Indo-Aryan. In Approaches to Grammaticalization. B. Heine and E. Traugott (eds). Amsterdam: Benjamins. Pp. 59-89.

• Hook, Peter. 2001. Where do Compound Verbs Come from? And where are they Going? In The yearbook of South Asian Languages and Linguistics 2001. Tokyo Symposium on South Asian Languages, Contact, Convergence and Typology. Delhi: SAGE. Pp. 101-130.

• Hook, P., and P. Pardeshi. MS. Inflation in the Marathi Compound Verb: 1300-2000.

• Hook Peter and Prashant Pardeshi. 2005. “Are Vector Verbs Eternal? ” Presented at the South Asian Linguistic Analysis (SALA)-25 Conference, University of Illinois, Urbana-Champaign, USA, 16-18 September, 2005. (Revised version presented at the 10th International Conference on South Asian Languages (ICOSAL-10), Moscow State University, Moscow, Russia, 5-7 July, 2012.

• Hopper, Paul and Elizabeth Traugott. 1993. Grammaticalization. Cambridge: Cambridge University Press.

• Taro Kageyama, Peter Hook and Prashant Pardeshi. 2013. Position paper: NINJAL International Symposium 2013: Mysteries of verb-verb complexes in Asian languages. (Available on: http://www.ninjal.ac.jp/vvsympo/files/NINJAL_VVsymposium2013.pdf)

• Masica, Colin. 1976. Defining a linguistic area: South Asia. Univ. of Chicago Press. 55

Appendix-1: English translation of the questionnaire in Japanese (1/3)

• 1. He/She put the book on the table.

• 2. Since the guests are going to come tomorrow I cleaned the room beforehand.

• 3. Since the guests were to come soon I switched on the light beforehand.

• 4. Since I was going return home late toady I prepared dinner beforehand.

• 5. I wrote the telephone number on the paper so as not to forget it.

• 6. Since I was to go on a trip from tomorrow I washed the cloths beforehand.

• 7. Mother hid the chocolates beforehand so as not be found by the kids.

• 8. I booked the ticket beforehand.

• 9. I will think about it beforehand.

• 10. Please convey him that I will go tomorrow.

• 11. Put that book on the table.

• 12. I put that file on the desktop.

56


• 13. Since the examination is tomorrow I listened the casette tape2/3 times beforehand.

• 14. When I had been to Japan I should have seen Mt. Fuji.

• 15. It is better to drink (take) medicine before getting sick.

• 16. I drank a lot of tea beforehand so as not to fall asleep later.

• 17. Today I may not be able to take lunch hence I ate/had a heavy breakfast beforehand.

• 18. Since I wanted to go out I took shower beforehand.

• 19. As I was going to eat food I washed my hands.

• 20. In order to put on makeup I washed my face.

• 21. Let me brush my teeth before going to bed.

• 22. I cannot swim tomorrow so I will swim today.

• 23. If you swim in the swimming pool in the morning you can stay cool even in the hot summer.

57


• 24. After reaching there I would have to start the work immediately so I got enough sleep in the plane beforehand.

• 25. Hide yourself so he cannot find you.

• 26. I felt danger so I decided to run away.

• 27. It looks like it is going to rain later so I better jog in the morning.

• 28. While training it is better to run the distance more than is required to be run in the race.

• 29. Those who want to ask questions should line up in front of the microphone.

• 30 It is a long journey so I should better sit down.

• 31. It is important to sit when one can.

• 32. I do not want to live long. It is best to die when there are still people around to regret your death.

58

Appendix-2 (1/8)

59

Uzbek (Central Asia) 6 Ertaga sayil-ga chiq-ib ketish-im bois kiyim-lar-im-ni

tomorrow journey-to leave-CP go-1S so cloth-PL-1S-ACC yuv-ib qo’y-d-im wash-CP keep-PAST-1S

Malayalam (South Asia) 6 Nale muthal yatrakkyu pokunnathu-kon-du vasuthram

tomorrow from journey.to go-take-GER cloths.ACC

alaki vecchu wash.CP keep.PAST Since I was to go on a trip from tomorrow I washed the cloths beforehand.

Exx. 2-9: In all the language at hand the combination of

[WASH+ KEEP/PUT] is permitted.

1. Canonical transitive, exx. 2-12 (see appendix 1)

Appendix-2 (2/8)

60

Mongolian (N-E Asia)

9 tere učir-yin toqai bodulqila- ya/ *[bodulqila-ĵu talbi-ya]

that thing-GEN about think-FUT think-CP keep-FUT

Telugu (South Asia)

9 Aa vishayam gurunchi aalochi-staanu/ aalochi-nchi-uncutaanu

that topic about think-1S.FUTURE think-CP-keep.FUTURE

I will think about it beforehand. あのことについて考えておきます。

[THINK＋KEEP/PUT ] combination variation is observed

across lgs.


Appendix-2 (3/8)

61


Tajik 12 in faila desktop-ba mon-da mon-d-am

that file.ACC desktop-on keep-CP keep-PST-1S

Hindi 12 vo fail desktop-pe (*rakh) rakhi

that file.ACC.F desktop-on keep.CP keep.PST.F あのファイルをデスクトップに置いておいた。 I kept that file on the desktop.

Exx. 11, 12：[KEEP/PUT+ KEEP/PUT ] is permitted in East

and Central Asian lgs. But not in South Asian lgs.

Appendix-2 (4/8)

62

Exx. 13, 14：[LISTEN+ KEEP/PUT ], [LOOK/SEE+

KEEP/PUT ]: variation across languages

2. Semi-transitive/ingestive, exx. 13-17 (see appendix 1) Uyghur

13 Ete imtixan bolgha-chqa plastinka-ni 2-3-qétim angla-p qoy-d-um

tomorrow exam be-so tape-ACC 2-3-times listen-CP keep-PAST-1SG

Kazakh

13 erteng ekzamen bol-a-dy sol ushyn kasseta-ny 2-3 ret

tomorrow examination be-FUT-3S that for cassette-ACC 2-3 times

tynda-p {*khoi-d-um / al-d-ym}

listen-CP keep-PAST-1SG take-PAST-1SG

Since the examination is tomorrow I heard the 2 to 3 times tape beforehand.

Appendix-2 (5/8)

63

2. Semi-transitive/ingestive, exx. 13-17 (see appendix 1)

Exx. 15-17：[EAT+ KEEP/PUT ], [DRINK+KEEP/PUT]: only

permitted in Japanese.

Kyrgyz 16 anan uiku-m kel-bes uchun chai-dan kop

later sleep-1S become-NEG so as to tea-ACC lot ichi-p {*koy-d-um/ al-d-ym} drink-CP keep-PAST-1SG take-PAST-1SG

Marathi 16 nantar dzhop yeu-naye mhaNun bharapur cahaa

later sleep come-NEG so as to lot tea.ACC pi-un {*Thev-laa/ ghet-laa} drink-CP keep-PF.M.Sg. take-PF.M.Sg. I drank a lot of tea so as not to fall asleep later.

Appendix-2 (6/8)

64

3. Reflexive, exx. 18-20 and 23 (see appendix 1)

Exx. 18-20, 23：[{TAKE BATH / WASH HANDS or FACE /

BRUSH TEETH}+ KEEP/PUT ]: not permitted in any lg.

except Japanese

Uyghur

19 tamaq–ni yéyésh üchün qolum-ni yu-du-m/*yuyu-p qoy-d-um

food-ACC eat.INF for hand-ACC wash-PAST-1S/wash-CP keep-PAST-1S

Tamil 19 saappaaDu saappiDum mun kai.yai kazhuvinEn/*kazhuvi-vaittEn

food eat.PTCPL before hand.ACC wash.PAST.1S/wash.CP-keep.PAST.1S I washed my hands for eating food.

Appendix-2 (7/8)

65

4. Unergative, exx. 21, 22, 24-31 (see appendix 1)

Mongolian 30 qula-ĵam ayala-qu tola saGu-ĵu {bayi-ya/ *talbi-ya/?ab-ya}

long-road walk-adj so sit-CP be-VOL keep-VOL/take-VOL It’s a long journey hence let us sit.

Marathi 30 laamb-tsaa pravaas asalyaane bas-un {*thevaa-va/ ghyaa-va}

long-GEN.M journey.M Being-INS sit-CP keep-VOL take-VOL 長旅なので座っておこう。

Exx. 21, 22, 24-31: [ {SWIM / SIT / STAND / HIDE / RUN } +

KEEP/PUT ]: only permitted in Japanese.

Appendix-2 (8/8)

66

5. Unaccusative, ex. 32 (see appendix 1)

Telugu 32 naa-ku ekkuwa kaalam bratakaalani leedu.

I-DAT long time wish to live not janam tiTTukoka munde chan-i {powaDam/*uncaTam} manchidi people commenting before die-CP going/keeping good

Kazakh

32 men-in uzakh omyr sur-g-ym kel-mei-dy on-da barlykh

1SG.GEN long life live-NOML come-NEG-3SG this-in all

adam-dar sen-y aiai-dy ol-yp Khal/*khoi-ghan jakhsy

person-PL 2SG-GEN regret-FUT.3P die-CP stay/put-NOML good

I do not want to live long. It is best to die when people around regret your death.

Ex. 32：[DIE+KEEP/PUT ]: Only possible in Japanese

japanese, marathi and other asian languages...1. introduction •the aim of this presentation: to...

Documents