control structures in korean: syntax and processingaccent.ucsd.edu/pdf/korean_polinsky_zas_05.pdf1...
TRANSCRIPT
![Page 1: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/1.jpg)
1
Control structures in Korean: Syntax and processing
Maria PolinskyNa-Young KwonRobert Kluender
(UCSD)
![Page 2: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/2.jpg)
2
Introductory remarks
![Page 3: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/3.jpg)
3
ControlA dependency between two argument positions in
which the referential properties of the overt controller determine the referential properties of the silent controllee:
Craig Venteri tried [___i to capture the code of life]controller controllee
![Page 4: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/4.jpg)
4
Subject ControlCraig Venteri tried [___i to escape] controller controlleesubject subject
![Page 5: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/5.jpg)
5
Object ControlVenture capitalists persuaded Craig Venteri [___i to capture the code]
controller controlleeobject subject
![Page 6: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/6.jpg)
6
Prevalent theoretical assumptions� (Overt) controller is structurally higher than
(silent) controllee
� Base-generated analysis of control, with an invisible subject or no subject at all in the complement clause (depending on the theory)
![Page 7: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/7.jpg)
7
Traditional analyses predict that… (at least) the following structures are impossible:
(1) ___i tried [Craig Venteri to capture the code…](Backward control)
(2) Craig Venteri tried that hei captured the code…(Copy control)
![Page 8: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/8.jpg)
8
Traditional analyses too restrictiveEmpirically attested:(1) ___i tried [Craig Venteri to capture …]controllee controller
(Tsez, Malagasy, Jakaltec, Zapotec)
(2) Venture capitalists persuaded ___i [Ventericontrollee controller
to work on the code of life](Brazilian Portuguese, Malagasy, Korean)Backward control is empirically possible
![Page 9: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/9.jpg)
9
What’s the source of the problem?�These new data are misanalyzed:
Apparent cases of backward control are amenable to an account that maintains the base-generated analysis of control
�The theory needs to be changed
![Page 10: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/10.jpg)
10
Roadmap of the talk� Two (2.5) object control constructions in Korean
and their properties� Two possible analyses of Korean control
� Syntactic control� Semantic control
� Which analysis is superior?� Structural evidence � Processing evidence
� Conclusions and outstanding questions
![Page 11: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/11.jpg)
11
Object control in Korean
![Page 12: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/12.jpg)
12
Korean object control� Complement clause headed by the
complementizer -tolok (Kim 1978, 1984), embedded under such verbs as ‘persuade’, ‘order’
� Apparent controller is in the accusative (or dative) case (ACC)� Controller precedes the complement clause (ACC1)� Controller follows the complement clause (ACC2)
� Apparent controller is in the nominative case (NOM)
![Page 13: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/13.jpg)
13
Korean object controlController in the accusative case:
(1) Chelswu-ka Yenghi-lul [Yenghi-ka ACC1Chelswu-NOM Yenghi-ACC Y-NOMhakkyo-lul ttena-tolok] seltukhaysstaschool-ACC quit-COMPL persuaded
(2) Chelswu-ka [Yenghi-ka hakkyo-lul ACC2Chelswu-NOM Yenghi-NOM school-ACCttena-tolok] Yenghi-lul seltukhaysstaquit-COMPL Yenghi-ACC persuaded
‘Chelswu persuaded Yenghi to quit school.’
![Page 14: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/14.jpg)
14
Korean object controlController in the nominative case(3) Chelswu-ka Yenghi-lul [Yenghi-ka NOM
Chelswu-NOM Y-ACC Yenghi-NOMhakkyo-lul ttena-tolok] seltukhaysstaschool-ACC quit-COMPL persuaded ‘Chelswu persuaded Yenghi to quit school.’
(3’) Chelswu-ka [Yenghi-ka hakkyo-lulChelswu-NOM Yenghi-NOM school-ACCttena-tolok] Yenghi-lul seltukhaysstaquit-COMPL Y-ACC persuaded ‘Chelswu persuaded Yenghi to quit school.’
The difference between the base and scrambled positions is unclear
![Page 15: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/15.jpg)
15
The scope of alternation�A number of predicates participate in the
alternation between ACC and NOM
�Corpus data (Seejong corpus 2002)
![Page 16: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/16.jpg)
16
Representative predicateskangyohata ‘force’kwunyuhata ‘recommend’kwuenhata ‘recommend’myenglyenghata ‘order’pwuthakhata ‘ask (as a favor)’selthukhata ‘persuade’yokwuhata ‘ask, request’congyonghata ‘recommend/encourage’ cisihata ‘order’ thailuta ‘implore’pwuchwukita ‘encourage’
![Page 17: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/17.jpg)
17
Properties of the constructionsProperties relevant for both ACC and NOM:�Evidence of the control relation�Evidence that the structure is biclausal,
with a matrix control verb �Evidence of obligatory control
![Page 18: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/18.jpg)
18
Properties of the constructionsProperties relevant for both ACC and NOM:�Evidence of the control relation�Evidence that the structure is biclausal,
with a matrix control verb �Obligatory control
![Page 19: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/19.jpg)
19
Evidence of control� selectional restrictions#Chelswu-nun tol-i/ul tteleci-tolok seltukha-ess-taChelswu-TOP rock-NOM/ACC fall-COMP persuade-PAST-DECL(‘Chelswu persuaded the rocks to fall.’)� idiom chunks impossible#sin-un pal ep-nun mal-i/mal-ul chenliGod-TOP feet not.exist-REL horse-NOM/ACC 10000kmka-tolok myenglyenghaesstago-COMP ordered
(‘God ordered the news to travel fast (lit.: … the footless horse to go 10,000 km).’)
![Page 20: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/20.jpg)
20
Properties of the constructionsProperties relevant for both ACC and NOM:�Evidence of the control relation ��Evidence that the structure is biclausal,
with a matrix control verb �Obligatory control
![Page 21: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/21.jpg)
21
Biclausal structure� event quantification� scrambling patterns�NPI licensing (will be discussed later)� (ellipsis: control complement is treated as
a constituent)
![Page 22: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/22.jpg)
22
Biclausal structure: Event quantification� event quantificationACC1/ACC2: Yesterday John persuaded Mary-ACC
[to leave tomorrow] NOM:Yesterday John persuaded
[Mary-NOM to leave tomorrow]
![Page 23: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/23.jpg)
23
Biclausal structure: Scrambling� scrambling patterns: ACCChelswu-ka Mary-lul [nayil hakkyoey ka-tolok] seltukhaesstaChelswu-NOM Mary-ACC tomorrow to.school go-COMP persuaded‘Chelswu persuaded Mary to go to school tomorrow.’
Chelswu-ka Mary-lul [hakkyoey nayil ka-tolok] seltukhaessta*Chelswu-ka Mary-lul [hakkyoey ka-tolok nayil] seltukhaessta*Chelswu-ka nayil Mary-lul [hakkyoey ka-tolok] seltukhaessta
![Page 24: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/24.jpg)
24
Biclausal structure: Scrambling� scrambling patterns: NOMChelswu-ka [Mary-ka nayil hakkyoey ka-tolok] seltukhaesstaChelswu-NOM Mary-NOM tomorrow to.school go-COMP persuaded‘Chelswu persuaded Mary to go to school tomorrow.’
Chelswu-ka [Mary-ka hakkyoey nayil ka-tolok] seltukhaesstaChelswu-ka [hakkyoey Mary-ka nayil ka-tolok] seltukhaessta*Chelswu-ka [Mary-ka hakkyoey ka-tolok] nayil seltukhaessta*Chelswu-ka [nayil hakkyoey ka-tolok] Mary-ka seltukhaessta
![Page 25: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/25.jpg)
25
Properties of the constructionsProperties relevant for both ACC and NOM:�Evidence of the control relation ��Evidence that the structure is biclausal,
with a matrix control verb ��Obligatory control
![Page 26: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/26.jpg)
26
Obligatory controlDoes the silent element obligatorily take a unique
antecedent?� Obligatory control: yes� Non-obligatory control: no(Williams 1980, Koster 1984, Hornstein 2003, Jackendoff and Culicover 2003, and many others)
![Page 27: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/27.jpg)
27
Obligatory control� these constructions instantiate obligatory control
��non-local antecedent��de se reading
��non-c-commanding antecedent��strict reading under ellipsis��arbitrary interpretation of null controllerNOMACC
![Page 28: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/28.jpg)
28
Properties of the constructionsProperties relevant for both ACC and NOM:�Evidence of the control relation ��Evidence that the structure is biclausal,
with a matrix control verb ��Obligatory control �
![Page 29: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/29.jpg)
29
Interim summary�selthuhata ‘persuade’V [__ DP CP/IP [tolok]]
� ACC1/ACC2 and NOM instantiate obligatory object control
![Page 30: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/30.jpg)
30
Properties of the NOM construction� Evidence of the control relation �� Evidence that the structure is biclausal, with the
control verb as matrix �� Evidence that the overt DP is in the embedded
clause� Evidence that there is a silent element in the
matrix clause
![Page 31: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/31.jpg)
31
Overt controller downstairs� case-marking� scrambling�NPI licensing� subject honorific agreement on the
embedded predicate
![Page 32: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/32.jpg)
32
Overt controller downstairs: Case marking�case-marking determined by the lower verb
Chelswu-TOP [Yenghi-NOM leave-COMP] persuaded
‘Chelswu persuaded Yenghi to leave.’
![Page 33: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/33.jpg)
33
Overt controller downstairs:Scrambling� the entire complement clause scrambles as a constituent[Yenghi-NOM tomorrow leave-Comp] Chelswu-NOM __ persuaded
'Chelswu persuaded Yenghi to leave tomorrow.'
![Page 34: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/34.jpg)
34
Overt controller downstairs: Scrambling� overt NP scrambles with constituents of the complement clause
[tomorrow Yenghi-NOM leave-Comp] yesterday Chelswu-NOM persuaded
![Page 35: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/35.jpg)
35
Overt controller downstairs: Scrambling� overt NP scrambles with constituents of the complement clause
[tomorrow Yenghi-NOM leave-Comp] yesterday Chelswu-NOM persuaded�…. but not with constituents of the matrix clause
* Chelswu-NOM [tomorrow leave-Comp] yesterday Yenghi-NOM persuaded
![Page 36: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/36.jpg)
36
Overt controller downstairs: NPI licensing�Negative polarity items (NPIs) are licensed
by clause-mate negation (Sohn 1996, Shi 1997)�NPI in NOM is licensed by the embedded
negation:Chelswu-ka [amwuto ka-ci anh-tolok] seltukhaesstaChelswu-NOM NPI go-INF NEG-COMP persuaded‘Chelswu persuaded nobody to go.’(lit.: Chelswu persuaded nobody not to go)
![Page 37: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/37.jpg)
37
Overt controller downstairs: Honorific agreement� Honorific agreement is local, triggered by subject:sensayng-nim-i ka-si-ess-tateacher-RESP-NOM go-HON-PAST-DEC‘The teacher went.’� Embedded verb shows subject honorification in NOM:Chelswu-nun [sensayng-nim-i ka-si-tolok] seltukhaesstaChelswu-TOP teacher-RESP-NOM go-HON-COMP persuaded� … matrix verb does not:*Chelswu-nun [sensayng-nim-i ka-si-tolok] seltukha-si-esstaChelswu-TOP teacher-RESP-NOM go-HON-COMP persuaded-HON
![Page 38: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/38.jpg)
38
Honorific agreement consistent across all three constructions � ACC1:Chelswu-nun sensayng-nim-ul [___ ka-si-tolok] seltukhaesstaChelswu-TOP teacher-RESP-ACC go-HON-COMP persuaded� ACC 2:Chelswu-nun [___ ka-si-tolok] sensayng-nim-ul seltukhaesstaChelswu-TOP go-HON-COMP teacher-RESP-ACC persuaded� NOM: Chelswu-nun [sensayng-nim-i ka-si-tolok] seltukhaesstaChelswu-TOP teacher-RESP-NOM go-HON-COMP persuaded
![Page 39: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/39.jpg)
39
Overt controller downstairs (summary)� case-marking� scrambling�NPI licensing� subject honorific agreement on the
embedded predicate
![Page 40: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/40.jpg)
40
Properties of the NOM construction� Evidence of the control relation �� Evidence that the structure is biclausal, with the
control verb as matrix �� Evidence that the overt DP is in the embedded
clause �� Evidence that there is a silent element in the
matrix clause
![Page 41: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/41.jpg)
41
The sound of silenceProposed structure: null upstairs controllee Chelswu-NOM __i [Yenghii-NOM leave-COMP] persuadedChelswu-NOM [Yenghii-NOM leave-COMP] __i persuaded
Evidence:�Binding�Quantifier float
![Page 42: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/42.jpg)
42
The sound of silence: Binding� Reflexive binding is local (Yoon 1989)*Chelswu-ka [Yenghii-ka hakkyo-ey kaessta-ko]Chelswu-NOM Yenghi-NOM school-DAT went-COMPkunye casini-uy chinkwu-eykey malhaesstaherself-GEN friend-DAT said‘Chelswu said to herselfi’s friend(s) hat Yenghiiwent to school.’
The embedded DP cannot bind a reflexive in the matrix clause
![Page 43: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/43.jpg)
43
The sound of silence: Binding�Reflexive binding is local �The silent controllee binds a local reflexiveChelswu-ka ___i [Yenghi-ka ka-tolok] Chelswu-NOM Yenghi-NOM go-COMPkunye casini -uy cipeyse seltukhaesstaherself-GEN at home persuaded’Chelswu persuaded Yenghi, at her house, to go.’
![Page 44: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/44.jpg)
44
The sound of silenceEvidence:�Binding ��Quantifier float
![Page 45: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/45.jpg)
45
The sound of silence: Quantifier float� If a quantifier follows the DP it modifies, the two
must agree in case (Gerdts 1987, Choi 1988, Cho 2000)
haksayng-tul-i twul-i/*ul/*Ø kaesstastudent-PL-NOM two-NOM/*ACC/*no case went‘Two students went.’
� Postnominal quantifier can be separated from the host DP (quantifier float)
![Page 46: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/46.jpg)
46
Quantifier float restrictions� Quantifier float is strictly local
(Kang 2002, Miyagawa 2005)
*Chelswu-ka [haksayng-i hakkyo-ey kaessta-ko] sey-myeung-iChelswu-NOM student-NOM school-DAT went-COMP three-CL-NOM
malhaesstasaid(‘Chelswu said that three students went to school.’)
� Case-matching quantifier must follow its host DP:*twul-i haksayng-tul-i kaesstatwo-NOM student-PL-NOM went
(‘Two students went.’)
![Page 47: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/47.jpg)
47
The sound of silence: Postnominal quantifier� The silent element licenses a case-marked quantifier (floated quantifier)
� The case of the quantifier is determined by the matrix verb (not the embedded verb)
kunye-ka [ai-tul-i ka-tolok] motwu-lul/*motwu-kashe-NOM child-PL-NOM go-COMP all-ACC/*all-NOM
seltukhaesstapersuaded
‘She persuaded all the children to go.’
__
![Page 48: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/48.jpg)
48
The sound of silence: Quantifier float� Restriction: the floated quantifier must follow
the control complement*kunye-ka motwu-lul [ai-tul-i ka-tolok] she-NOM all-ACC child-PL-NOM go-COMP seltukhaesstapersuaded
‘She persuaded all the children to go.’� Why? Seems unexpected on the analysis where
the gap precedes the control complement:Chelswu-NOM __i all [childreni-NOM leave-COMP] persuaded
![Page 49: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/49.jpg)
49
The sound of silence: Quantifier float� Quantifiers float only to the right in Korean
Cf. in ACC:kunye-ka ai-tul-ul [__ ka-tolok] motwu-lul seltukhaesstashe-NOM child-PL-NOM go-COMP all-ACC persuaded
‘She persuaded all the children to go.’
kunye-ka motwu-lul [__ ka-tolok] ai-tul-ul seltukhaesstashe-NOM all-ACC go-COMP child-PL-NOM persuaded� A floated quantifier must follow an overt DP?� The position of the floated quantifier in NOM is
not determined by the placement of the gap
??*
![Page 50: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/50.jpg)
50
Properties of the NOM construction� Evidence of the control relation �� Evidence that the structure is biclausal, with the
control verb as matrix �� Evidence that the overt DP is in the embedded
clause �� Evidence that there is a silent element in the
matrix clause �
![Page 51: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/51.jpg)
51
Interim summaryTwo patterns in Korean object control: � Matrix controller, silent embedded controllee
(ACC1, ACC2)� Embedded overt controller, silent matrix
controllee (NOM)ACC1: John Mary-ACC [___ leave] persuadedACC2: John [___ leave] Mary-ACC persuadedNOM: John __ [Mary-NOM leave] persuadedNOM: John [Mary-NOM leave]__ persuaded
![Page 52: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/52.jpg)
52
Question�What is the appropriate analysis of these constructions?
![Page 53: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/53.jpg)
53
Roadmap of the talk� Two (2.5) object control constructions in Korean
and their properties �� Two possible analyses of Korean control
� Syntactic control� Semantic control
� Which analysis is superior?� Structural evidence � Processing evidence
� Conclusions and outstanding questions
![Page 54: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/54.jpg)
54
Analytical possibilities
![Page 55: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/55.jpg)
55
An impossible analysisBase-generated control structures(1) John Mary-ACC [PRO leave] persuaded(2) *John PRO [Mary-NOM leave] persuaded
� Problems with (2):� PRO is ungoverned but does not receive arbitrary interpretation
� Condition C violation� Base-generated analysis of the backward pattern (NOM) is untenable
![Page 56: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/56.jpg)
56
Two possible analyses�Desideratum: analysis must be able to handle both forward and backward patterns�Syntactic control (Polinsky and Potsdam 2002, Monahan 2004)
�Semantic control (Cormack and Smith 2002, 2004)
![Page 57: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/57.jpg)
57
Syntactic control� Matrix and embedded DP form an A-chain;
Control is raising into a theta-position� ACC: the tail of the chain is deleted �
Forward ControlJohn Mary-ACC [Mary-NOM leave] persuaded� NOM: the head of the chain is deleted�
Backward ControlJohn Mary-ACC [Mary-NOM leave] persuaded
![Page 58: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/58.jpg)
58
Syntactic control�ACC1John [VP Maryk-ACC [CP [IP __k [VP leave]]-COMP] persuaded]
�ACC2 (possibly scrambled?)John [XP [CP [IP _k [VP leave]]-COMP]j [VP Maryk-ACC tj persuaded]
�NOM John [VP __k [CP [IP Maryk-NOM [VP leave]]-COMP] persuaded]
A-chain
A-chain
A-chain
![Page 59: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/59.jpg)
59
Syntactic control
NOM:Backward control
ACC 1/ACC 2:Forward control
Delete head (higher element) of movement chain
Delete tail (lower element) of movement chain
� The difference between the two forward patterns is due to scrambling; it is unclear which pattern is basic
� Main question: What motivates the deletion of the higher element in the movement chain?
![Page 60: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/60.jpg)
60
Semantic control�Korean has subject and object pro-drop;
the silent element in all three constructions is a null pronominal
�Overt DP is co-indexed with a null pronominal, via a meaning postulate
![Page 61: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/61.jpg)
61
Semantic control� Unmarked structure:
control complement is in the specifier of VP, DP (including null pronominal) adjoined to V’
John [VP [CP Mary1-NOM leave-COMP] [V’[pro2] persuaded]John [VP [CP pro1 leave-COMP] [V’ [Mary2-ACC] persuaded]� Shifted structure:
accusative DP is in the specifier of VP, control complement adjoined to V’
John [VP [Mary1-ACC] [V’ [CP DP2 leave-COMP] persuaded]
![Page 62: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/62.jpg)
62
Semantic control� Unmarked structure:
control complement is in the specifier of VP, DP (including null pronominal) adjoined to V’
John [VP [CP [IP Mary1-NOM leave-COMP]] [V’pro2 ] persuaded]
John [VP [CP [IP pro1 leave-COMP]] [V’ Mary2-ACC] persuaded]co-indexation
co-indexation
![Page 63: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/63.jpg)
63
Semantic control with pro� Shifted structure:
accusative DP is in [spec, VP], control complement adjoined to V’
John [VP [DPMary1-ACC] [V’ [CP [IP pro2 leave-COMP]]] persuaded]co-indexation
![Page 64: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/64.jpg)
64
Semantic control
ACC 1(DP before CP)
ACC 2 (CP before DP)
pro in the embedded clause
impossible because of Condition C violation
NOM Controlpro in the matrix clause
DP in [spec, VP]CP adjoined to V’
CP in [spec,VP]DP adjoined to V’
![Page 65: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/65.jpg)
65
The two analyses� Convergence: The syntactic and semantic analyses yield the same
interpretation:‘John persuaded Mary to go.’
� Divergence:The two analyses make different structural
predictions
![Page 66: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/66.jpg)
66
Roadmap of the talk� Two (2.5) object control constructions in Korean
and their properties �� Two possible analyses of Korean control �
� Syntactic control� Semantic control
� Which analysis is superior?� Structural evidence � Processing evidence
� Conclusions and outstanding questions
![Page 67: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/67.jpg)
67
Syntactic vs. semantic analysis
Structural differences
![Page 68: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/68.jpg)
68
Relevant structural properties� c-command effects—discussed here
� (representation of verb frames)
![Page 69: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/69.jpg)
69
Relevant structural properties: c-command
��c-command between the matrix argument and the embedded subject
Semantic analysis
Syntactic analysis
![Page 70: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/70.jpg)
70
c-command effects� embedded subject restriction:
� Only the embedded subject, overt or silent, can be co-indexed with the matrix element
� intervening material: � An intervening clause disrupting the c-command chain should be impossible
� distributive quantifiers: � Distributive quantifiers that c-command pronouns construed as bound variables should be possible in control structures, including the NOM construction
![Page 71: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/71.jpg)
71
c-command effects in the two analyses
impossiblepossibledistributive quantifiers
possibleimpossibledisruption of c-command
does not applyappliesembedded subject restriction
Semantic analysis of NOM (backward pattern)
Syntactic analysis of NOM (backward pattern)
![Page 72: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/72.jpg)
72
Embedded subject restriction� Syntactic analysis:
If a matrix empty category c-commands a constituent of the embedded CP, only the embedded subject could be co-indexed with it
� Semantic analysis: Since no c-command holds, the meaning postulate should allow for the embedded agent, regardless of grammatical function, to be co-indexed with the matrix null pronominal
![Page 73: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/73.jpg)
73
Embedded subject restrictionChelswu-nun [Swuyeng-ij Yenghi-eykeykChelswu-Top Swuyeng-Nom Yenghi-Datintephyu pat-tolok] __j/*k seltukhaysstainterview pass-Comp persuaded‘Chelwsu persuaded Sueng to be interviewed by
Yenghi.’*‘Chelswu persuaded Yenghi that she interview
Swueng.’� Embedded subject restriction supports
the syntactic analysis
![Page 74: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/74.jpg)
74
c-command effects� embedded subject restriction �� intervening material� distributive quantifiers
![Page 75: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/75.jpg)
75
Intervening material� Syntactic analysis:
If there is a matrix empty category c-commanding the embedded subject of CP, the command chain cannot skip intervening clauses
cf. in English:Johnj decided [that there was a plan [__*j to evacuate]]John convinced Maryj [that there was a plan [__*j to evacuate]]
![Page 76: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/76.jpg)
76
Intervening material� Semantic analysis:
Since no c-command holds, the null pronominal and its identifying expression can be separated by another clause
cf. in English:Johni was shocked [that Mary said [that hei was a liar]]
![Page 77: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/77.jpg)
77
Intervening materialBoth analyses can handle:Chelswu-ka [[cipey Yenghii–ka o-tolok ] Chelswu-NOM home Yenghi-NOM come-COMP
__i/proi kyelsimha-tolok] __i/proi seltukhaesstadecide-COMP persuaded
‘Chelswu persuaded Yenghi [to decide [to come home]].’
![Page 78: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/78.jpg)
78
Intervening materialBoth analyses can handle:*Chelswu-ka [[cipey __i/proi o-tolok ] Chelswu-NOM home come-COMP
Yenghii–ka kyelsimha-tolok] __i/proi seltukhaysstaYenghi-NOM decide-COMP persuaded
(‘Chelswu persuaded Yenghi to decide to come home.’)
![Page 79: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/79.jpg)
79
Intervening materialOnly the syntactic analysis can handle:*Chelswu-ka [Yenghii–ka onul [__i/proi cipey
Chelswu-NOM Yenghi-NOM today home
nayil ka-tolok ] kyelsimha-tolok] __i/proi seltukhaysstatomorrow go-COMP decide-COMP persuaded(‘Chelswu persuaded Yenghi [to decide today [to go home tomorrow]].’)
Semantic analysis: co-indexation should be possible with scramblingSyntactic analysis: predicts ungrammaticality
co-indexation
![Page 80: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/80.jpg)
80
Intervening material� The restriction against intervening
material supports the syntactic analysis
![Page 81: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/81.jpg)
81
c-command effects� embedded subject restriction �� intervening material �� distributive quantifiers
![Page 82: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/82.jpg)
82
Distributive quantifiers�Syntactic analysis: Distributive quantifiers should be possible�Semantic analysis: True distributive quantifiers should be impossible because they would bind a pronominal
![Page 83: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/83.jpg)
83
Distributive quantifiers…… are possible in NOM (backward pattern):Chelswu-nun [ai-ka may-kaChelswu-Top child-Nom every-Nomswukcay-lul ha-tolok] seltukhaesstahomework-Acc do-Comp persuaded‘Chelswu persuaded every child to do the homework.’
� Distributive quantifier evidence supports the syntactic analysis
![Page 84: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/84.jpg)
84
c-command effects� embedded subject restriction �� intervening material �� distributive quantifiers �
� Primary linguistic evidence based on c-command relations supports the syntactic analysis of Korean object control
![Page 85: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/85.jpg)
85
Conclusions� The semantic analysis of Korean object control
makes a number of incorrect predictions� Primary linguistic data support the syntactic
analysis of Korean object control� Korean control patterns are accounted for within
current theoretical assumptions:� Control as movement into a thematic position� Copy and delete theory of movement
![Page 86: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/86.jpg)
86
Syntactic vs. semantic analysis
Processing differences
![Page 87: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/87.jpg)
87
The three control constructions� ACC1: Forward pattern, DP before CP� ACC2: Forward pattern, CP before DP� NOM: Backward pattern
![Page 88: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/88.jpg)
88
Korean object controlController is in the accusative case(1) Chelswu-ka Yenghi-lul [Yenghi-ka ACC1
Chelswu-NOM Yenghi-ACC Yenghi-NOMhakkyo-lul ttena-tolok] seltukhaysstaschool-ACC quit-COMPL persuaded
(2) Chelswu-ka [Yenghi-ka hakkyo-lul ACC2Chelswu-NOM Yenghi-NOM school-ACCttena-tolok] Yenghi-lul seltukhaysstaquit-COMPL Yenghi-ACC persuaded
‘Chelswu persuaded Yenghi to quit school.’
![Page 89: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/89.jpg)
89
Korean object controlController is in the nominative case
(3) Chelswu-ka Yenghi-lul [Yenghi-ka NOMChelswu-NOM Yenghi-ACC Yenghi-NOMhakkyo-lul ttena-tolok] seltukhaysstaschool-ACC quit-COMP persuaded
(3’) Chelswu-ka [Yenghi-ka hakkyo-lulChelswu-NOM Yenghi-NOM school-ACCttena-tolok] Yenghi-lul seltukhaysstaquit-COMPL Yenghi-ACC persuaded ‘Chelswu persuaded Yenghi to quit school.’
The difference between the base and scrambled positions is unclear
![Page 90: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/90.jpg)
90
Reading time study� Self-paced reading time study
� 40 sentences per condition (70 filler sentences)� 23 native Korean participants
Example target sentence:The marketing department persuaded the leading actress to appear on a popular talk show to advertise the movie.
![Page 91: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/91.jpg)
91
Opening frame…
“The marketing department of the production, to advertise the movie, …”
W6W5W4W3W2W1
foradvertising-ACC
movie marketing-dept-NOM
production-GEN
that
wuyhay hongpo-lul yenghwa hongpothim-i yenghwasa-uyku
![Page 92: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/92.jpg)
92
… target sentences“…persuaded the leading actress to appear on a popular talk show”
W11W10W9W8W7
persuadedheroine-ACC
go-comp talk-show-to
popularACC2
persuadedgo-comp talk-show-to
popularheroine-NOM
NOM
persuadedgo-comp talk-show-to
popularheroine-ACC
ACC1
![Page 93: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/93.jpg)
93
Where’s the gap?� ACC1:John-NOM Mary-ACC [GAP leave] persuaded
� ACC2:John-NOM [GAP leave] Mary-ACC persuaded
� NOM:John-NOM GAP [Mary-NOM leave] persuaded
orJohn-NOM [Mary-NOM leave] GAP persuaded
![Page 94: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/94.jpg)
94
Direct comparison of ACC1 and NOM� Because of word order differences between ACC2 and the other two constructions (NOM/ACC1), word-by-word comparisons were possible only between ACC1 and NOM
� Nonetheless, ACC1 and ACC2 patterned alike in that they were read faster than NOM in terms of � total reading time across the sentence� total reading time across the 2nd half of the sentence� reading time at final matrix predicate (W11)
![Page 95: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/95.jpg)
95
Direct comparison of ACC1 and NOM“…persuaded the leading actress to appear on a popular talk show”
W11W10W9W8W7
persuadedgo-comp talk-show-to
popularheroine-NOM
NOM
persuadedgo-comp talk-show-to
popularheroine-ACC
ACC1
![Page 96: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/96.jpg)
96
Predictions� The constructions are initially analyzed as
mono-clausal� But at some point, the structure has to be reanalyzed as bi-clausal, which entails a processing cost
� The constructions are initially analyzed as not containing a gap� But at some point, the structure has to be reanalyzed as containing a gap, which entails a processing cost
![Page 97: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/97.jpg)
97
Predictions (ACC1)John-NOM Mary-ACC [ GAP leave-COMP ] persuaded � initially processed as mono-clausal� NP-ACC (W7) interpreted as matrix object� when the parser reaches leave-COMP (W10),
the sentence� has to be reanalyzed as bi-clausal, and � a gap is posited in the embedded clause
� slowdown in reading time should occur at leave-COMP position (W10)
![Page 98: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/98.jpg)
98
Predictions (ACC1)“…persuaded the leading actress to appear on a popular talk show”
W11W10SLOW
W9W8W7
persuadedgo-comp talk-show-to
popularheroine-ACC
ACC1
![Page 99: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/99.jpg)
99
Predictions (NOM)John-NOM (GAP) [Mary-NOM leave-COMP] (GAP) persuaded � initially processed as mono-clausal� when the parser reaches the 2nd NP-NOM (W7),
� the sentence has to be reanalyzed as bi-clausal� a gap could logically be posited in the main clause (but native speakers find this highly implausible)
� slowdown (mon0- to bi-clausal reanalysis) should occur prior to W10
![Page 100: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/100.jpg)
100
Predictions: Gap positing in NOM“first resort” gap positing:� if a gap is posited at W7 (2nd NP-NOM),
then all of the hard processing work should be over by W10 (leave-COMP)
“last resort” gap positing:� if a gap is not posited until W10 (leave-COMP),
there should be an additional slowdown at W10
![Page 101: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/101.jpg)
101
NOM: “first resort” gap positing“…persuaded the leading actress to appear on a popular talk show”
W11W10W9W8W7SLOW
persuadedgo-comp talk-show-to
popularheroine-NOM
NOM
![Page 102: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/102.jpg)
102
NOM: “last resort” gap positing“…persuaded the leading actress to appear on a popular talk show”
W11W10SLOW
W9W8W7SLOW
persuadedgo-comp talk-show-to
popularheroine-NOM
NOM
![Page 103: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/103.jpg)
103
Predictions: Summary
W10W7W10Gap positing
W7W7W10Mono- to bi-clausal reanalysis
NOM: last resort gap
NOM: first resort gap
ACC1
![Page 104: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/104.jpg)
104
Reading times: ACC1 and NOM
300
400
500
600
700
800
900
W1 W2 W3 W4 W5 W6 W7 W8 W9 W10 W11
ACC1
NOM
![Page 105: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/105.jpg)
105
Direct comparison of ACC1 and NOM“…persuaded the leading actress to appear on a popular talk show”
W11W10W9W8W7
persuadedgo-comp talk-show-to
popularheroine-NOM
NOM
persuadedgo-comp talk-show-to
popularheroine-ACC
ACC1
![Page 106: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/106.jpg)
106
Direct comparison of ACC1 and NOM“…persuaded the leading actress to appear on a popular talk show”
W11W10W9W8W7NOM SLOW
persuadedgo-comp talk-show-to
popularheroine-NOM
NOM
persuadedgo-comp talk-show-to
popularheroine-ACC
ACC1
![Page 107: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/107.jpg)
107
Reading time at W7
ACC1 < NOM (p <0.002)
638.705
757.23
500
550
600
650
700
750
800
ACC1 NOM
![Page 108: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/108.jpg)
108
What takes NOM longer at W7?What happens when the parser reaches the 2nd
nominative (NP-NOM)?Processing effects:
� clause-boundary effect (Miyamoto 2002, 2003)� second NP-NOM marks the beginning of a new clause,
which increases processing load
� similarity effect at second nominative (Uehara 1997)� difficulty in discriminating between two NP-NOMs awaiting
structural assignment also delays processing
![Page 109: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/109.jpg)
109
Reading times: ACC1 and NOM
300
400
500
600
700
800
900
W1 W2 W3 W4 W5 W6 W7 W8 W9 W10 W11
ACC1
NOM
![Page 110: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/110.jpg)
110
Direct comparison of ACC1 and NOM“…persuaded the leading actress to appear on a popular talk show”
W11W10W9W8W7NOM SLOW
persuadedgo-comp talk-show-to
popularheroine-NOM
NOM
persuadedgo-comp talk-show-to
popularheroine-ACC
ACC1
![Page 111: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/111.jpg)
111
Direct comparison of ACC1 and NOM“…persuaded the leading actress to appear on a popular talk show”
W11W10NOM SLOW
W9W8W7NOM SLOW
persuadedgo-comp talk-show-to
popularheroine-NOM
NOM
persuadedgo-comp talk-show-to
popularheroine-ACC
ACC1
![Page 112: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/112.jpg)
112
Predictions: Summary
W10W7W10Gap positing
W7W7W10Mono- to bi-clausal reanalysis
NOM: last resort gap
NOM: first resort gap
ACC1
![Page 113: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/113.jpg)
113
Results: Summary
W10W7W10Gap positing
W7W7W10Mono- to bi-clausal reanalysis
NOM: last resort gap
NOM: first resort gap
ACC1
![Page 114: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/114.jpg)
114
Reading time at W10
ACC1 < NOM (p < 0.003)
492.788
572.537
440
460
480
500
520
540
560
580
ACC1 NOM
![Page 115: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/115.jpg)
115
W10: Predictions for NOM� when the parser reaches the 2nd NP-NOM (W7),
the sentence� has to be reanalyzed as bi-clausal, and� a gap could logically be posited in the main clause
� if a gap is posited at W7 (2nd NP-NOM), processing work should be over at W10
� if a gap is not posited until W10, there should be an additional slowdown at W10
![Page 116: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/116.jpg)
116
What takes NOM longer at W10?Clear reading time evidence for:� bi-clausal reanalysis at W7� no difference from ACC1 at W8 and W9� some additional processing cost at W10
� should not be for bi-clausal reanalysis at this point� so must be for gap positing and filler-gap association
![Page 117: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/117.jpg)
117
W10: Predictions for ACC1� when the parser reaches W10, the sentence
� has to be reanalyzed as bi-clausal, and � a gap is posited in the embedded clause
� slowdown in reading time should occur at W10
![Page 118: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/118.jpg)
118
What’s going on with ACC1 at W10?� predictions were for both
� bi-clausal reanalysis� gap positing and filler-gap associationat this point
� yet ACC1 was read faster than NOM at W10, which� does not require bi-clausal reanalysis� only requires gap positing and filler-gap association
![Page 119: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/119.jpg)
119
Reading time at W10
ACC1 < NOM (p < 0.003)
492.788
572.537
440
460
480
500
520
540
560
580
ACC1 NOM
![Page 120: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/120.jpg)
120
What’s going on at W10?Clearly, something about
� gap positing and filler-gap association is more difficult in NOM at W10 than
� bi-clausal reanalysis and� gap positing and filler-gap association
in ACC1 at W10
![Page 121: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/121.jpg)
121
What’s going on at W10?� In other words, one might expect a greater
processing cost for ACC1 than for NOM at W10� But the results are the opposite: NOM > ACC1� Why? What extra factor makes NOM slower?
![Page 122: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/122.jpg)
122
What kind of dependency is it?� Syntactic analysis of ACC1:
� deletion of tail of A-chainHans-NOM Peteri-ACC [Peteri-NOM gehen-COMP] überzeugte
� Syntactic analysis of NOM: � deletion of head of A-chain
Hans-NOM Peteri-ACC [ Peteri-NOM gehen-COMP ] überzeugteOR
Hans-NOM [ Peteri-NOM gehen-COMP ] Peteri-ACC überzeugte
![Page 123: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/123.jpg)
123
What kind of dependency is it?� Semantic analysis of ACC1:
� forward co-indexation � marked “lightest first” ordering of arguments
Hans-NOM [VP [CP Peteri-ACC] [V’ [CP proi gehen-COMP]] überzeugte]
� Semantic analysis of NOM:� forward co-indexation� unmarked “heaviest first” ordering of arguments
Hans-NOM [VP [CP Peteri-NOM gehen-COMP] [V’ [proi-ACC]überzeugte]
![Page 124: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/124.jpg)
124
What kind of dependency is it?All analyses of ACC1 and NOM posit the same
filler-gap dependency EXCEPT the syntactic analysis of NOM (backward
control), which posits a gap-filler dependency in one variant
![Page 125: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/125.jpg)
125
What kind of dependency is it?� Syntactic analysis of NOM:
� deletion of head of A-chainHans-NOM Peteri-ACC [ Peteri-NOM gehen-COMP ] überzeugte
ORHans-NOM [ Peteri-NOM gehen-COMP ] Peteri-ACC überzeugte
� Semantic analysis of NOM:� forward co-indexation� unmarked “heaviest first” ordering of arguments
Hans-NOM [VP [CP Peteri-NOM gehen-COMP] [V’ [proi-ACC] überzeugte]
![Page 126: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/126.jpg)
126
Sorting out the analyses of NOM� One syntactic analysis
Hans-NOM GAPi [ Peteri-NOM gehen-COMP ] überzeugte[gap-filler dependency]
� Other syntactic analysis and semantic analysisHans-NOM [ Peteri-NOM gehen-COMP] GAPiüberzeugte[filler-gap dependency]
![Page 127: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/127.jpg)
127
Syntactic vs. semantic analysis� The semantic analysis predicts ACC1 to be slower
than NOM because of the “marked” pattern� The syntactic analysis correctly predicts that
NOM should be slower because of the gap-filler dependency
� The reading time results are consistent with the structure proposed by the syntactic analysis in which the gap precedes the complement clause
![Page 128: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/128.jpg)
128
Could this be a frequency effect?Perhaps NOM control is simply less frequent
than ACC1 0r ACC2 controlThis might account for the slowdown in
reading time
![Page 129: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/129.jpg)
129
Frequency data analysis�Two sets of corpus statistics:
� Total number of tokens for each construction� Total number of obligatory control tokens for each construction
�Data from the Seejong corpus (2002)
![Page 130: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/130.jpg)
130
Corpus distribution: All instances
NOM ACC1
ACC2
![Page 131: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/131.jpg)
131
Corpus distribution: OC only
NOMACC1
ACC2
![Page 132: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/132.jpg)
132
Could this be a frequency effect?� The NP-NOM1 NP-NOM2 configuration would
seem more likely to cause a slowdown in reading time at NP-NOM2 (W7) than at the embedded verb (W10)
� Recall the additional, separate effect at W10� The effect at W10 was unlikely due to bi-clausal
reanalysis� Therefore, the W10 effect had to be related in
some way to gap positing and gap-filler association
![Page 133: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/133.jpg)
133
NOM causes processing difficulty
300
400
500
600
700
800
900
W1 W2 W3 W4 W5 W6 W7 W8 W9 W10 W11
ACC1
NOM
![Page 134: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/134.jpg)
134
NOM causes processing difficulty
W10W7W10Gap positing
W7W7W10Mono- to bi-clausal reanalysis
NOM: last resort gap
NOM: first resort gap
ACC1
![Page 135: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/135.jpg)
135
Processing conclusions� The syntactic analysis makes correct processing predictions, while the semantic analysis does not
� The parser thus seems to adopt a “last resort” strategy for positing gaps in Korean NOM control structures
� The same strategy applies in Korean pre-nominal relative clauses (ambiguous with pro-drop clauses), which also contain gap-filler dependencies
![Page 136: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/136.jpg)
136
“Last resort” gap positing in RCs
300
500
700
900
1100
1300
W1 W2 W3 W4 W5 W6 W7 W8 W9 W10 W11
ST
OT
SP
OP
W7: embedded clause verbW8: head noun of main clause
![Page 137: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/137.jpg)
137
Processing conclusions� The parser thus seems to adopt a “last resort” strategy for positing gaps in Korean NOM control structures
� The same strategy applies in Korean pre-nominal relative clauses (ambiguous with pro-drop clauses), which also contain gap-filler dependencies
� Head-final languages do have filler-gap dependencies (e.g. leftward scrambling in Japanese, which invokes a “first resort” strategy for positing gaps )
� The fact that Korean seems to adopt a “last resort” strategy for object control with a NOM controller suggests that this is a gap-filler dependency
![Page 138: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/138.jpg)
138
Roadmap of the talk� Two (2.5) object control constructions in Korean
and their properties� Two possible analyses of Korean control
� Syntactic control� Semantic control
� Which analysis is superior?� Structural evidence � Processing evidence
� Conclusions and outstanding questions
![Page 139: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/139.jpg)
139
Conclusions� The alternation in Korean complement-taking
predicates can be accounted for as an alternation between forward and backward object control
� Korean object control alternations support the growing body of empirical evidence for backward control
![Page 140: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/140.jpg)
140
Conclusions�Backward control is possible within
current theoretical assumptions:� Control is movement into a thematic positionJohn Mary [Mary to leave] persuadedJohn Mary [Mary to leave] persuaded� Control and raising are instances of a single phenomenon: a referential dependency between two elements, one of which can be deleted
� That is, one can serve as filler, and one as gap
![Page 141: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/141.jpg)
141
Conclusions� The fact that Korean seems to adopt a “last
resort” gap-positing strategy for object control with a NOM controller suggests that this is a gap-filler dependency, thus:John Mary [Mary to leave] persuaded
![Page 142: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/142.jpg)
142
Outstanding questions: Korean� What accounts for the restriction that floated
quantifiers must follow the complement clause in NOM?
� What motivates the choice between the constructions examined here? � Preliminary evidence that the NOM and ACC constructions have differences in interpretation
� Why are most of the verbs allowing the object alternation ambiguous between control and non-control predicates?
![Page 143: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/143.jpg)
143
Outstanding questions� Theory-internal: On the copy and delete analysis
of backward control, what forces the deletion of the higher copy?
� Processing: Can processing data shed more light on the choice between the semantic and syntactic analyses?
� Cross-linguistic: Now that we know where to look, can more “backward” predicates be found?
![Page 144: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/144.jpg)
144
__ hören jetzt auf, [wir zu reden]
Und wir danken für Ihre Aufmerksamkeit!
![Page 145: Control structures in Korean: Syntax and processingaccent.ucsd.edu/pdf/Korean_Polinsky_ZAS_05.pdf1 Control structures in Korean: Syntax and processing Maria Polinsky Na-Young Kwon](https://reader033.vdocument.in/reader033/viewer/2022041516/5e2b25e7c0c24c3f0b056813/html5/thumbnails/145.jpg)
145
Acknowledgments� Annabel Cormack� Shin Fukuda� Norbert Hornstein� Laura Kertz� Ron Langacker� Beth Levin� Phil Monahan� Colin Phillips� Eric Potsdam� Peter Sells� Barbara Stiebels
� Funding:� National Science Foundation
� Max-Planck Institute