Content uploaded by Ethan Kutlu
Author content
All content in this area was uploaded by Ethan Kutlu on Jun 05, 2017
Content may be subject to copyright.
The Processing of Ditransitive Verbs in Turkish-English
Bilinguals
!"#$%&'(&)*"+*,-&.$*/$&0$12*/3%4-&)*5362&7*/$8*93:
,;%3<=/83">&2?&@+2/3A$B&4;%3<=/83">&2?&C""$D$B&:E$/+="2%&;%3<=/83">
A=%6*"+*F*?+(=A*
!"#$%&$
'()G2H6B&I(&)(&J,KLMN(&0>%"$H"3H&O=/838"=%H=&3%&+$%9*$9=&O/2A*H"32%(&Cognitive)PsychologyB&,LB&:PPQ:LR(*()E#/38"2??=+8B&S(B&T&'=&U/22"B&V(&J4WWXN(&E25O2%=%"8&2?&835*+"$%=2*8&3%"=/O/="3%9Y&E25O$/3%9&3%"=/O/="3%9&D3"#&
8#$A2D3%9&$%A&O$/$O#/$83%9( Bilingualism:)Language)and)CognitionB&R&J:NB&44RZ4XW(&+()A=&G2"B&)(&J,KK4N&V&G3+3%9*$+&[/2H=883%9&72A=+Y&.=<=+"\8 ]0O=$63%9\&72A=+&VA$O"=A(&VOO+3=A&.3%9*38"3H8B&,:Y&,Z4X(&,()G/$%39$%B&^(B&
[3H6=/3%9B&7(B&T&E+=+$%AB&V(&J4WWWN(&0>%"$H"3H&H2Z2/A3%$"32%&3%&A3$+29*=(&CognitionB&RPB&G,:QG4P(&-()E+=+$%AB&V(&V(B&T&[3H6=/3%9B&7(&I(&J4WW:N(&_#=&*8=&2?&+=`3H$+&$%A&8>%"$H"3H&3%?2/5$"32%&3%&+$%9*$9=&O/2A*H"32%Y&!<3A=%H=&
?/25&"#=&O/353%9&2?&%2*%ZO#/$8=&8"/*H"*/=(&Journal)of)Memory)and)LanguageB&XKB&4,XQ4:W(&.()^$/"8*36=/B&a(&I(B&)2+6B&^(&^(&I(B&T&^*386$5OB&[(&J,KKKN(&[/353%9&D2/A&2/A=/&3%&8=%"=%H=&O/2A*H"32%(&I2*/%$+&2?&!`O=/35=%"$+&
[8>H#2+29>B&P4JVNB&,4KQ,XR(&/()^$/"8*36=/B&a(&I(B&T&b=8"=%1=/9B&E(&J4WWWN(&b2/A&2/A=/&O/353%9&3%&D/3""=%&$%A&8O26=%&8=%"=%H=&O/2A*H"32%(&E29%3"32%B&RPB&G4RQG:K(&0()[3H6=/3%9B&7(B&T&@=//=3/$B&c(&J4WWLN(&0"/*H"*/$+&
O/353%9Y&V&H/3"3H$+&/=<3=D(&Psychological)BulletinB&134J:NB&X4RQXPK(&
123242#&25
Research Question
Do bilinguals of typologically different languages
show evidence of a shared syntax for ditransitive
verbs?
uL2 sentence processing focuses on two different views:
shared syntax (Bock, 1986; Branigan et al., 2000) or
separate syntax (Hartsuiker et al., 1999)
• Typologically similar languages may have a
higher degree of shared syntax (Bock, 1986;
Pickering et al., 2008)
• e.g., languages such as Dutch, German, and
English share the same word order for many
structures (i.e., ditransitive verbs; complex noun
phrases; active / passive sentences)
• Advanced speakers of these languages have
been tested by using a production based syntactic
priming methodology
• When there is a word order dissimilarity, priming
effects are not observed, indicating that the syntax
may not be shared in typologically different
languages
uThis study tests the shared syntax account by
extending previous studies in three ways:
1. Typologically different languages: Turkish and
English
• English has SVO word order, Turkish has SOV
word order, however, it is mainly free in word
order
2. A different participant group: Students of
simultaneous interpretation programs in Turkey
• Empirically, simultaneous interpreters (SIs)
have been shown to use syntactic priming as a
strategy during interpretation (de Bot, 1992)
• Simultaneous interpreting is a cognitively
demanding job because it requires
simultaneous comprehension and
production àit will be ideal if SIs have a
shared syntax
• It is worth noting that many simultaneous
interpreters are late bilinguals
3. A comprehension task instead of a production
one
uIn this study, a shared syntax view would be
supported if evidence is found that the syntax
of one language affects that of the other:
• e.g., higher knowledge in English may interact
with L1 Turkish knowledge causing a change in
word order preference cross-linguistically
uEnglish ditransitive verbs have two structures:
they mean the same, but they do not have the
same syntax:
1. I sent the flowers to Mary.PO (Prepositional Obj)
2. I sent Mary the flowers. DO (Double Obj)
uTurkish ditransitive verbs have one structure
with two word orders (they differ in case
marking) that means the same:
1. Ali kitab-ı Ayşe-ye ver-di.
Ali book-ACC Ayse-DAT give-PAST3SG
‘Ali gave the book to Ayse.’ PO word order
2. Ali Ayşe-ye kitab-ı ver-di.
Ali Ayse-DAT book-ACC give-PAST3SG
‘Ali gave Ayse the book.’ DO word order
6#$4"78&$9"#
Method
uAuditory syntactic priming paradigm
uConditions: match [PO-PO/DO-DO], mismatch [PO-DO-DO-PO)
u4 comprehension based tasks were conducted using PO and DO sentences. Participants judged each condition by pressing YES/NO:
uTurkish-English speakers heard two sentences consecutively, and they were asked “Are these two sentences the same?”
1. Within-English Sameness Judgment Task [Match: Steve gave the box to Mary. / Steve gave the box to Mary. Mismatch: Steve
gave Mary the box. / Steve gave the box to Mary.]
2. Within-Turkish Sameness Judgment Task [Match: Steve box-ACC Mary-DAT gave. / Steve box-ACC Mary-DAT gave.
Mismatch: Steve box-ACC Mary-DAT gave. / Steve Mary-DAT box-ACC gave.]
uTurkish-English speakers heard two sentences consecutively, and they were asked “Was the second sentence a correct
translation of the first one?”
1. Translation Judgment Task from English to Turkish
2. Translation Judgment Task from Turkish to English
Hypotheses: (1) High Proficiency SI students will judge mismatch conditions (PO-DO or DO-PO) as being less similar than
match conditions (DO-DO or PO-PO), indicating that their L1 and L2 interact and thus that their syntax is shared. (2) Syntactic
effects will decrease with low proficiency unless it was purposefully manipulated (i.e., native speakers who were asked about
the sameness of the two sentences).
:$9;8<9
uHyptohesis 1 ✓: High Proficiency SI students judged mismatch conditions (PO-DO or DO-PO) as being
less similar than match conditions (DO-DO or PO-PO), indicating that their L1 and L2 interact and thus that
their syntax is shared.
uHypothesis 2 ✓: Native speakers of English (asked about the meaning of the two sentences), MP
bilinguals, and LP bilinguals judged mismatch conditions (PO-DO/DO-PO) to be as similar as match
conditions (PO-PO/DO-DO). Syntactic effects decreased with low proficiency. Proficiency may play a role
in shared-syntax.
uTurkish-English bilinguals chose DO sentences in Turkish over PO sentences àAnimacy can play a role in
PO and DO sentences
uIt has been shown that animate objects are placed higher at the sentence level in typologically different languages (Verhoeven,
2014) àSharedness of animacy effect
!"#&<859"#
uAre simultaneous interpretation students the same as regular bilinguals? They may have better Working
Memory due to their training (Christoffels et al., 2004)
uAdd a control group of monolingual native speakers of Turkish
uAddress methodological issues (i.e., interpretation of the question (semantics vs. syntax), working memory
interference)
=8$842)>942&$9"#5
?@A249;2#$%<)>259B#
1258<$5
Participants
53 Turkish-English simultaneous interpretation students
oTested in Turkey as the experimental group
oPerformed all four tasks
oEnglish Cloze test (Brown, 1980) was used to assess
proficiency
• High Proficiency (HP) (n = 13)
• Mid Proficiency (MP) (n = 21)
• Low Proficiency (LP) (n = 11)
15 native speakers of English
oTested as the control group in Ottawa (Canada)
oPerformed only the Within-English Sameness Judgment
Task
oTwo different questions were asked:
• Are they the same (n = 8)
• Do these two sentences mean the same (n = 7)
W
4W
XW
MW
LW
,WW
d$"3<=&
J$/=&"#=>&
"#=&8$5=N
^[ 7[ .[ d$"3<=&JA2&
"#=>&5=$%&
"#=&8$5=N
C2%#):%;2#255)D87B;2#$5)EFG
H%4$9&9A%#$)I4"8A5
![CZ![C
!'CZ!'C
![CZ!'C
!'CZ![C
Oe&(WW,
Within English Sameness Judgments F (2, 23) = 24.1, p < .001
EPO-EDO pairs less similar than EDO-EPO pairs (MD = -1.8, p = .002)
W
4WW
XWW
MWW
LWW
,WWW
,4WW
,XWW
^[ 7[ .[
C2%#)12%&$9"#)J9;2)E;5G
H4"39&92#&K
![CZ![C
!'CZ!'C
![CZ!'C
!'CZ![C
Within English Sameness Judgment RTs F (2, 23) = 7.9, p < .001
Match conditions (EPO-EPO and EDO-EDO) quickly > mismatch (EPO-EDO
and EDO-EPO) pairs (MD = 385.9, p = .003).
W
4W
XW
MW
LW
,WW
^[ 7[ .[
C2%#):%;2#255)D87B;2#$5)EFG
H4"39&92#&K
_[CZ_[C
_'CZ_'C
_[CZ_'C
_'CZ_[C
Within Turkish Sameness Judgment F (3, 54) = 36.1, p < .001
TPO-TDO more similar >TDO-TPO (MD = 14.2, SE = 3.8, p = .009)
W
4W
XW
MW
LW
,WW
^[ 7[ .[
C2%#)!"442&$)J4%#5<%$9"#)
D87B;2#$5)EFG
H4"39&92#&K
![CZ_[C
!'CZ_'C
![CZ_'C
!'CZ_[C
Translation Judgment English-Turkish F (3, 108) = 8.1, p < .001
EDO-TDO correct translation >EPO-TPO (MD = -9.9, p = .013).
L&M#"N<27B2;2#$5
S&D2*+A&+36=&"2&$H6%2D+=A9=&"#=&5=51=/8&2?&!U!E!cB&S0E!cB&$%A&_;E!c&3%&_*/6=>&?2/&"#=3/&O$/"3H3O$"32%&$%A&9/=$"&8*OO2/"f&S&$+82&D2*+A&+36=&"2&"#$%6&$++&5=51=/8&2?&!a[.3%9
.$1&?2/&"#=3/&$5$g3%9&H2++$12/$"32%8(&