A modal base analysis of Mandarin accomplishment predicates - - PowerPoint PPT Presentation
A modal base analysis of Mandarin accomplishment predicates Xiaoqian ZHANG Yingyi LUO Jianhua HU Institute of Linguistics, Chinese Academy of Social Sciences xiaoqianzhang.diderot@yahoo.com yingyi.luo@gmail.com ctjhu@126.com May 28-30, 2020
Xiaoqian ZHANG Yingyi LUO Jianhua HU
Institute of Linguistics, Chinese Academy of Social Sciences xiaoqianzhang.diderot@yahoo.com yingyi.luo@gmail.com ctjhu@126.com
May 28-30, 2020 The 21st Chinese Lexical Semantics Work (CLSW2020) City University of Hong Kong
1 / 35
1
Non-culminating accomplishments in Mandarin
2
A modal base account
3
The necessity of distinguishing different sources of endpoint
4
Concluding remarks
2 / 35
Accomplishment predicates are known as denoting complex events that are composed of a process and a natural endpoint ensuing from the process. Perfective aspect requires that the event time be included in the topic
perfective aspect, the whole event (process+natural endpoint) should complete within the temporal interval introduced by the topic time. (1) Perfective aspect : λPλtT ∃e[τ(e)⊆tT & P(e)]
3 / 35
Telic reading :
他昨天写了一封信，然后寄出去了。 T¯ a pro3sg zuóti¯ an yesterday xiě-le write-pfv yì
f¯ eng clf xìn, letter ránhòu then jìch¯ uqù-le. send-pfv ‘He wrote a letter yesterday and sent it.’
4 / 35
Telic reading :
他昨天写了一封信，然后寄出去了。 T¯ a pro3sg zuóti¯ an yesterday xiě-le write-pfv yì
f¯ eng clf xìn, letter ránhòu then jìch¯ uqù-le. send-pfv ‘He wrote a letter yesterday and sent it.’
Atelic reading :
他昨天写了一封信，可是没写完。 T¯ a pro3sg zuóti¯ an yesterday xiě-le write-pfv yì
f¯ eng clf xìn, letter kěshì but méi neg xiě-wán. write-finish ‘He wrote a letter yesterday, but he didn’t finish writing it.’
5 / 35
Lexical aspect account : Tai and Chou (1975), Chu (1976), Tai (1984), Sybesma (1997), Soh and Kuo (2005) and Koenig and Chief (2008), etc. Grammatical aspect account : Smith (1991, 1994), Wu (2005), Demirdache et al. (2017) and Martin (2018), etc.
6 / 35
Accomplishments do not exist in Mandarin (Tai, 1984). The Mandarin counterparts of English accomplishments do not imply the culmination of endpoint. Mandarin counterparts of English accomplishment predicates are more like activity predicates. Their lexical representation is a process (2) rather than a complex event comprising a process and a natural endpoint that marks a change of state (3). (2) [ x act ] (3) [[ x act ] cause [ become [ y <state> ]]]
7 / 35
Mandarin accomplishment predicates are like their English counterparts and encode natural endpoint. The atelic reading is induced by the perfective marker verbal -le. Verbal -le can express either termination or completion (Smith, 1991, 1994). When it expresses termination, an atelic reading is triggered ; when it expresses completion, a telic reading is available.
8 / 35
Lexical aspect account :
Although the culmination did not obtain on the atelic reading, it is true that the event initiator had the intention that the natural endpoint culminated. The proposal of lexical aspect fails to capture the intentionality meaning of the atelic reading.
9 / 35
Grammatical aspect account :
The non-culminating reading is also available when verbal -le is
between a terminative reading and a completive reading. (4) 他昨天写过一封信，可是没写完。 T¯ a pro3sg zuóti¯ an yesterday xiě-guò write-exp yì
f¯ eng clf xìn, letter kěshì but méi neg xiě-wán. write-finish ‘He wrote a letter yesterday, but he didn’t finish writing it.’ This solution is rather ad hoc and language-specific, which does not apply to languages that lack overt aspectual markers, such as French, German, Polish, etc.
10 / 35
Our goal : We will propose an alternative account, namely a modal base analysis, for the lexical representation of Mandarin accomplishments. We argue that accomplishments incorporate a modal base containing the set of possible worlds compatible with the initiator’s intentions. The final sub-event and the result state ensuing form the final sub-event are placed in the scope of the modal operator to guarantee that the accomplishment does not necessarily culminate in the actual world.
11 / 35
1
Non-culminating accomplishments in Mandarin
2
A modal base account
3
The necessity of distinguishing different sources of endpoint
4
Concluding remarks
12 / 35
Incremental theme endpoint :
(5) 他吃了一个苹果，没吃光。 T¯ a pro3sg ch¯ ı-le eat-pfv yí
gè clf píngguˇ
apple méi neg ch¯ ı-gu¯ ang. eat-complete ‘He ate an apple. He didn’t eat it up.’
Property endpoint :
(6) 他杀了一只猪，没杀死。 T¯ a pro3sg sh¯ a-le kill-pfv yì
zh¯ ı clf zh¯ u, pig méi neg sh¯ a-sˇ ı. kill-die ‘He killed a pig. The pig was not dead.’
Script endpoint :
(7) 他做了一个实验，没做完。 T¯ a pro3sg zuò-le conduct-pfv yí
gè clf shíyàn, experiment méi neg zuò-wán. conduct-finish ‘He conducted an experiment. He didn’t finish it.’
13 / 35
(8) 他吃了一个苹果，没吃光。 T¯ a pro3sg ch¯ ı-le eat-pfv yí
gè clf píngguˇ
apple méi neg ch¯ ı-gu¯ ang. eat-complete ‘He ate an apple. He didn’t eat it up.’ The verb eat is an incremental verb and the object an apple is thus an incremental theme that provides an endpoint. There is homomorphism between the event described by the incremental verb and the “abstract” scale that measures the change undergone by the incremental theme, cf. Krifka (1989), Krifka (1992), Krifka (1998), Wyngaerd (2001), Rothstein (2004), Wechsler (2005), Beavers (2007), Beavers (2011) and Ramchand (2008).
14 / 35
(9) 他吃了一个苹果，没吃光。 T¯ a pro3sg ch¯ ı-le eat-pfv yí
gè clf píngguˇ
apple méi neg ch¯ ı-gu¯ ang. eat-complete ‘He ate an apple. He didn’t eat it up.’ As the event of eating progressed, more and more parts of the apple were consumed. When the last part of the apple was eaten, the event
The default reading of eat an apple in the context of perfective aspect is a telic reading. The continuation clause appears to cancel the incremental theme endpoint that guang ‘complete’ targets.
15 / 35
(10) 他杀了一只猪，没杀死。 T¯ a pro3sg sh¯ a-le kill-pfv yì
zh¯ ı clf zh¯ u, pig méi neg sh¯ a-sˇ ı. kill-die ‘He killed a pig. The pig was not dead.’ The object a pig is not an incremental theme and does not provide an incremental theme endpoint. It is the pig’s property that supplies a natural endpoint : when the pig entered into the state of being dead, the event of killing a pig reached its culmination point. The negation clause suggests that the secondary predicate si ‘die’ can be used to subtract kill a pig of its property endpoint.
16 / 35
(11) 他做了一个实验，没做完。 T¯ a pro3sg zuò-le conduct-pfv yí
gè clf shíyàn, experiment méi neg zuò-wán. conduct-finish ‘He conducted an experiment. He didn’t finish it.’ The event structure of conduct an experiment is based on an “action script”, which is “prototypical courses of actions performed by the agent” (Kearns, 2007, p. 43). The completion of the last action composing the event of conducting an experiment provides a script endpoint. The event of conducting an experiment unfolded as more and more actions were realised : when the last action was realised, the event in question culminated.
17 / 35
Table – Categorisation of accomplishments on the basis of sources of endpoint
18 / 35
Incremental theme endpoint
(12) 他吃了一个苹果，没吃光。#因为没开始吃。 T¯ a pro3sg ch¯ ı-le eat-pfv yí
gè clf píngguˇ
apple méi neg ch¯ ı-gu¯ ang. eat-complete # Y¯ ınwéi because méi neg ka¯ ıshˇ ı begin ch¯ ı. eat ‘He ate an apple. He didn’t eat it up.#Because he didn’t begin.’
19 / 35
Property endpoint
(13) 他杀了一只猪，没杀死。#因为没开始杀。 T¯ a pro3sg sh¯ a-le kill-pfv yì
zh¯ ı clf zh¯ u, pig méi neg sh¯ a-sˇ ı. kill-die # Y¯ ınwéi because méi neg ka¯ ıshˇ ı begin sh¯ a. kill ‘He tried to kill a pig. The pig was not killed to death. Because he didn’t begin.’
20 / 35
Script endpoint
(14) 他做了一个实验，没做完。#因为没开始做。 T¯ a pro3sg zuò-le conduct-pfv yí
gè clf shíyàn, experiment méi neg zuò-wán. conduct-finish #Y¯ ınwéi because méi neg ka¯ ıshˇ ı begin zuò. conduct ‘He conducted an experiment. He didn’t finish conducting.#Because he didn’t begin.’
21 / 35
The event initiator has the intention that the accomplishment
introduce a modal base in the lexical representation of Mandarin accomplishment predicates. Following Beavers and Lee (2020), we assume that this modal base contains the set of possible worlds compatible with the initiator x’s intentions, i.e. x’s intention set Ix. According to Inman (1993), “an individual x intends P iff P is true in all worlds in x’s intention set Ix”. Therefore, if P is an accomplishment in the intention set of x, then it is true that the event initiator x has an intention that P culminates. A universal modal operator is introduced to quantify over the modal base Ix.
22 / 35
We propose that an accomplishment, as a durative event, is composed of at least three sub-events, i.e. an initial sub-event, an intermediary sub-event and a final sub-event (Beavers, 2007). The different sub-events may occur at different possible worlds. The initial sub-event of an accomplishment must occur in the actual
base Ix. On the contrary, both the final sub-event of the accomplishment and the result state ensuing from it are placed within the scope of the modal base Ix to capture the fact that the event initiator intends that the accomplishment culminates. When the world of evaluation is not contained in the modal base, i.e. the endpoint did not culminate in the actual world, we obtain an atelic reading. By contrast, when the world of evaluation is contained in the modal base, we obtain a telic reading.
23 / 35
(15) [Lisi eat an apple] λe ∃ei [initiator(e)=Lisi ∧ INIE (ei, e) ∧ eat (ei) ∧ theme(ei) = an apple ∧ ILisi ∃ef ∃s [FINE (ef , e) ∧ eat (ef ) ∧ theme(ef ) = an apple ∧ cause(ef , s) ∧ disappear (s) ∧ theme(s) = an apple ]] For any e, e’ ∈ UE, INIE (e’, e) is true iff e’ is initial in e, i.e. there exists no e” < e that precedes e’ in e, and analogously for FINE (e’, e). (Beavers, 2012) (16) [Lisi kill a pig] λe ∃ei [initiator(e)=Lisi ∧ INIE (ei, e) ∧ kill (ei) ∧ theme(ei) = a pig ∧ ILisi ∃ef ∃s [FINE (ef , e) ∧ kill (ef ) ∧ theme(ef ) = a pig ∧ cause(ef , s) ∧ be dead (s) ∧ theme(s) = a pig ]] (17) [Lisi fry a plate of chicken legs] λe ∃ei [initiator(e)=Lisi ∧ INIE (ei, e) ∧ fry (ei) ∧ theme(ei) = a plate of chicken legs ∧ ILisi ∃ef ∃s1 ∃s2 [FINE (ef , e) ∧ fry (ef ) ∧ theme(ef ) = a plate of chicken legs ∧ cause(ef , s1) ∧ be affected (s1) ∧ theme(s1) = a plate of chicken legs ∧ cause(ef , s2) ∧ be cooked (s2) ∧ theme(s2) = a plate of chicken legs]]
24 / 35
1
Non-culminating accomplishments in Mandarin
2
A modal base account
3
The necessity of distinguishing different sources of endpoint
4
Concluding remarks
25 / 35
Previous studies often neglected the distinction among different sources of endpoint. Tai (1984) and Koenig and Chief (2008) did not precise which source
Soh and Kuo (2005) and van Hout et al. (2017) only dealt with one source of endpoint. We will argue that the endpoints cannot be subtracted in the same manner and require being lexically targeted by different words, which justifies the necessity of distinguishing different endpoint sources.
26 / 35
(18)
T¯ a pro3sg tóu-le cast-pfv yì
zh¯ ang clf piào, ballot # kěshì but méi neg tóu-wán. cast-finish ‘He cast a ballot, but didn’t finish voting.’
T¯ a pro3sg tóu-le cast-pfv yì
zh¯ ang clf piào, ballot kěshì but méi neg tóu-shàng. cast-on ‘He cast a ballot, but didn’t succeed.’ The contrast between (18a) and (18b) can only the explained by our proposal according to which different types of endpoint are targeted by different lexical words. More precisely, the property endpoint of cast a ballot cannot be targeted by wan ‘finish’, but only by shang ‘on’.
27 / 35
(19)
Lˇ ısì Lisi xiě-le write-pfv yì
f¯ eng clf xìn. letter Méi neg xiě-wán, write-finish zhˇ ı
xiě-le write-pfv yí
bàn half xìn. letter ‘Lisi wrote a letter. He didn’t finish writing and only half was written.’
Lˇ ısì Lisi kˇ ao-le bake-pfv yì
zh¯ ı clf j¯ ı. chicken # Méi neg kˇ ao-wán, bake-finish zhˇ ı
kˇ ao-le bake-pfv yí
bàn half j¯ ı. chicken ‘Lisi baked a chicken. He didn’t finish baking and only half was baked.’
The continuation clause comprising “only+verb+-le+one half+object” can
endpoint, but not bake a chicken, can co-occur with the continuation clause.
28 / 35
1
Non-culminating accomplishments in Mandarin
2
A modal base account
3
The necessity of distinguishing different sources of endpoint
4
Concluding remarks
29 / 35
Mandarin accomplishments in the perfective context allow both atelic reading and telic reading. There exist three sources of endpoint that contribute to the telic reading of accomplishments. We argue that accomplishments incorporate a modal base containing the set of possible worlds compatible with the initiator’s intentions. The final sub-event and the result state ensuing form the final sub-event are placed in the scope of the modal operator to guarantee that the accomplishment does not necessarily culminate in the actual world.
30 / 35
31 / 35
Beavers, D. (2011). On affectedness. Natural Language and Linguistic Theory, 29(2) :335–370. Beavers, J. (2007). Scalar complexity and the structure of events. In Dölling, J., Heyde-Zybatow, T., and Schäfer, M., editors, Event Structures in Linguistic Form and Interpretation, volume 5, pages 245–265. Berlin : Walter de Gruyter. Beavers, J. (2012). Lexical aspect and multiple incremental themes. In Demonte, V. and McNally, L., editors, Telicity, Change, and State. A Cross-Categorial View of Event Structure, pages 23–59. Oxford University Press. Beavers, J. and Lee, J. (2020). Intentionality, scalar change, and non-culmination in Korean caused change-of-state predicates. Linguistics. Chu, C. C. (1976). Some semantic aspects of action verbs. Lingua, 40 :43–54. Demirdache, H., Liu, J., Martin, F., and Sun, H. (2017). Licensing non-culminating accomplishments in Mandarin. Experimental & theoretical evidence. Paper presented at Workshop on Non-culminating, Irresultative and Atelic Readings of Telic Predicates (TELIC 2017), Universität Stuggart. Inman, M. V. (1993). Semantics and Pragmatics of Colloquial Sinhala Involitive Verbs. PhD thesis, Stanford University.
32 / 35
Kearns, K. (2007). Telic senses of deadjectival verbs. Lingua, 117 :26–66. Koenig, J.-P. and Chief, L.-C. (2008). Scalarity and state-changes in Mandarin (and
Syntax and Semantics, volume 7, pages 241–262. Krifka, M. (1989). Nominal reference, temporal constitution and quantification in event
Semantics and Contextual Expression, volume 11, pages 75–115. Dordrecht : Foris Publications. Krifka, M. (1992). Thematic relations as links between nominal reference and temporal
Chicago : CSLI Publications, Chicago University Press. Krifka, M. (1998). The origins of telicity. In Rothstein, S., editor, Events and Grammar, volume 70, pages 197–235. Dordrecht : Kluwer Academic Publishers. Martin, F. (2018). Non-culminating accomplishments. Manuscript under review. Ramchand, G. C. (2008). Verb Meaning and the Lexicon : A First Phase Syntax. Cambridge : Cambridge University Press.
33 / 35
Rothstein, S. (2004). Structuring Events : A Study in the Semantics of Lexical Aspect. Oxford : Blackwell Publishing. Smith, C. S. (1991). The Parameter of Aspect. Dordrecht : Kluwer Academic Publishers. Smith, C. S. (1994). Aspectual viewpoint and situation type in Mandarin Chinese. Journal of East Asian Linguistics, 3(2) :107–146. Soh, H. L. and Kuo, J. Y.-C. (2005). Perfective aspect and accomplishment situations in Mandarin Chinese. In Verkuyl, H. J., De Swart, H., and Van Hout, A., editors, Perspectives on Aspect, volume 32, pages 199–216. Springer Netherlands. Sybesma, R. (1997). Why Chinese verb-le is a resultative predicate. Journal of East Asian Linguistics, 6(3) :215–261. Tai, J. H.-Y. (1984). Verbs and times in Chinese : Vendler’s four categories. In Testen, D., Mishra, V., and Drogo, J., editors, Papers from the Parasession on Lexical Semantics, Chicago Linguistic Society, volume 20, pages 289–296. Tai, J. H.-Y. and Chou, J. Y. (1975). On the equivalent of ‘kill’ in Mandarin Chinese. Journal of the Chinese Language Teachers Association, 10(2) :48–52.
34 / 35
van Hout, A., Arche, M. J., Demirdache, H., del Real, I. G., Sanz, A. G., Gavarró, A., Marzo, L. G., Hommes, S., Kazanina, N., Liu, Jinhong, L. O., Martin, F., and Strangmann, I. M. (2017). Agent control and the acquisition of event culmination in Basque, Dutch, English, Spanish, and Mandarin. In LaMendola, M. and Scott, J., editors, Proceedings of the 41st annual Boston University Conference on Language Development, pages 323–332. Somerville, MA : Cascadilla Press. Wechsler, S. (2005). Resultatives under the ‘event-argument homomorphism’ model of
Deriving Thematic and Aspectual Interpretation, pages 255–273. Oxford University Press. Wu, J.-S. (2005). The semantics of the perfective le and its context-dependency : An SDRT approach. Journal of East Asian Linguistics, 14(4) :299–336. Wyngaerd, G. V. (2001). Measuring events. Language, 77(1) :61–90.
35 / 35