basque word orders, psycholinguistic and neurolinguistic research
DESCRIPTION
Basque Word Orders, Psycholinguistic and Neurolinguistic Research. Author: Kepa Erdozia Advisor: Itziar Laka. Quote (Chomsky 1986: 3-4). - PowerPoint PPT PresentationTRANSCRIPT
Basque Word Orders, Psycholinguistic and
Neurolinguistic Research
Author: Kepa Erdozia
Advisor: Itziar Laka
Quote (Chomsky 1986: 3-4)
The study of generative grammar represented a significant shift of focus in the approach to problems of language. Put in the simplest terms, to be elaborated below, the shift of focus was from behavior or the products of behavior to states of the mind/brain that enter into behavior. If one chooses to focus attention on this latter topic, the central concern becomes knowledge of language: its nature, origins, and use.
The answer to the first question is given by a particular generative grammar, a theory concerned with the state of the mind/brain of the person who knows a particular language. The answer to the second is given by a specification of UG along with an account of the ways in which its principles interact with experience to yield a particular language; UG is a theory of the “initial state” of the language faculty, prior to any linguistic experience. The answer to the third question would be a theory of how the knowledge of language attained enters into the expressions of thought and the understanding of presented specimens of language, and derivatively, into communication and other special uses of language.
The three basic questions that arise, then, are these:(1) (i) What constitutes knowledge of language?
(ii) How is knowledge of language acquired?(iii) How is knowledge of language put to use?
TALK PLANNING
What constitutes knowledge of language?
How is knowledge of language acquired?
How is knowledge of language put to use?
Linguistic research about word order in Basque
How the children acquire word order in Basque
How humans use their knowledge of language to generate and process word orders in Basque
Free Word Ordera. PP S IO O V[Afaldu ondoren] [Mikelek] [Elenari] [gerriko berria] [oparitu dio][After dinner-PP] [Mikel-S] [Elena-IO] [the new belt-O] [given has-V]‘After dinner, Mikel has given the new belt to Elena’
5 constituent sentence; P5 = 120 sentences.
Nearly, all constituent permutation are grammatical in Basque
d. IO V O PP S[Elenari] [oparitu dio] [gerriko berria] [afaldu ondoren] [Mikelek]
e. ...
c. S O PP V IO[Mikelek] [gerriko berria] [afaldu ondoren] [oparitu dio] [Elenari]
b. O PP IO S V[Gerriko berria] [afaldu ondoren] [Elenari] [Mikelek] [oparitu dio]
a. PP S IO O V[Afaldu ondoren] [Mikelek] [Elenari] [gerriko berria] [oparitu dio][After dinner-PP] [Mikel-S] [Elena-IO] [the new belt-O] [given has-V]‘After dinner, Mikel has given the new belt to Elena’
Free Word Order
Previous Research on Word Order in Basque
Generative GrammarSOV (De Rijk 1969, Eguzkitza 1986, Ortiz de Urbina 1989, Laka 1990, Artiagoitia 1995, Fernandez 1998, A. Elordieta 2001, Arregi 2001 among others …)
Previous Research on Word Order in Basque
Generative GrammarSOV (De Rijk 1969, Eguzkitza 1986, Ortiz de Urbina 1989, Laka 1990, Artiagoitia 1995, Fernandez 1998, A. Elordieta 2001, Arregi 2001 among others …)
SVO (Ormazabal et al 1994, G. Elordieta 1997, Haddican 2004)
Informational and Statistical approaches
Experimental Psycholinguistics: language acquistion
Osa 1990, Hidalgo 1994, Aldezabal et al 2003
Bronckart & Idiazabal 1982
De Rijk 1969
Statistical analysis: SOV Tales Play Narrative Total
SOV 66% 44% 61% 57%
SVO 23% 37% 31% 30%
OVS 5% 9% 5% 6%
OSV 2.5% 7% 1.5% 4%
VSO 3% 2% 1.5% 2.5%
VOS 0.5% 1% 0% 0.5%
Analyzed sentences
209 183 67 459
Addapted from De Rijk 1969: 16
De Rijk 1969
Following Greenberg: SOV
Statistical analysis: SOV
Relative clauses: SOV
Tales Play Narrative Total
SOV 66% 44% 61% 57%
SVO 23% 37% 31% 30%
OVS 5% 9% 5% 6%
OSV 2.5% 7% 1.5% 4%
VSO 3% 2% 1.5% 2.5%
VOS 0.5% 1% 0% 0.5%
Analyzed sentences
209 183 67 459
Addapted from De Rijk 1969: 16
Postpositions V>Aux
NP
NP NP
Y Y Verb
Ortiz de Urbina 1989
Subject/Object asymetries
In the hierarchical configuration of Basque subjects are hierarchically higher than objects
INFL
INFL’’
INFL’Otsoak
VP’’
ardia V’
jan
duOtsoak ardia jan du
Ortiz de Urbina 1989Otsoak ardia jan du Ardia otsoak jan du
INFL
INFL’’
INFL’Otsoak
VP’’
ardia V’
jan
du
Ortiz de Urbina 1989Otsoak ardia jan du Ardia otsoak jan du
INFL’
VP’’ INFL
V’
tk
ti
tj
INFL’’
CP
otsoaki
ardiaj
C’
C
jan duk
CP
INFL
INFL’’
INFL’Otsoak
VP’’
ardia V’
jan
du
A. Elordieta 2001Otsoak ardia jan du
CP
T
C TP
AuxP
Aux
DPsub
v
vP
vP
AspP
VP Asp
DPobj V
OSV: Ardia otsoak jan duSOV
Diplaced the subject to focus position and the verb to CP position; and
displaced the object to the topic position
[TopObjj [FocSubji [CPjan du]V-aux [TP ti tj tV-aux]]]
Antisymmetry (Kayne 1994, 2004)
Kayne: syntactic structure is universally and without exceptions of the form S[pecifier]-H[ead]-C[omplement]. The complement of a head invariably follows that head. The associated specifier invariably precedes both head and complement (2004: 3)
XP
HeadSpecifier Complement
Kayne: The question is whether Japanese [Basque] objects ever surface within VP, in complement position of V. Antisymmetry says no, given OV order (2004: 5)
All languages are based generated as SVO
Ormazabal, Uriagereka and Uribe-Etxebarria 1994
Mary-ga sono hon-o yonda
Mary-S book that-O read-V
Japanese
Mirenek liburu hori irakurri du
Mary-S book that-O read-V
BasqueDeclaratives
Mary-wa nani-o yonda ka?
Mary-S what-O read-V Q-marker
Japanese
Zer irakurri du Mirenek?
What read-V Mary-S
BasqueInterrogatives
Ormazabal, Uriagereka and Uribe-Etxebarria 1994
In Basque interrogative sentences, the WH-word raises the CP leaving behind the IP
CP
C
IP
tIP
C’In declarative sentences, IP moves to the specifier position of CP in the both languages
Neuter SOV order two possibilities
CP
C
WH
IP
C’
a) Before the movement of IP to CP, move the verb to C
b) Extract the arguments from the IP which is in CP
G. Elordieta 1997
CP
VP
T
DP V’
C
ModP
NegPMod
TP
Neg
DPV
a) In functional projections above the VP
How derived SOV order:
b) Agreement features are present in the verb from the start of the numeration and languages choose whether to spell-out or not morphologically
Otsoak jan du ardiaSVO
Haddican 2004
CaseP
VP
Case’Zorraki
ordaindu ti
Declarative sentences: V>Aux
Negative sentences: Neg>Aux>V
Polarity Phrase (PolP)
From VO to OV
PolP
TP
Aux T’
Mod(evid)P
Mod(evid)’
Pol’
omen
!VP
Zorrak ordaindu
T tm
But, his system allowed focus construction which are ungrammatical
*JONEK Miren ikusi du
Word Order and Comunicative Abilities (Osa 1990)
The canonical word order in Basque is Subjet-Object-Verb
a) Less presupositions are elicited by SOV
b) Prosodically flat
c) It replies to a What happened? question
d) All the sentence could be new information
Functionalist point of view
Informational structures of Focus provide word order variability
Word Order and Statistics (Hidalgo 1994)
Source: Hidalgo 1994
XVIIth Century
Statistically analyzed following De Rijk’s sentence selection criterion
Word Order and Statistics (Hidalgo 1994)
Source: Hidalgo 1994
Statistically analyzed following De Rijk’s sentence selection criterion
XIXth Century
Word Order and Statistics (Hidalgo 1994)
Source: Hidalgo 1994
Popular oral tales collections
Statistically analyzed following De Rijk’s sentence selection criterion
Word Order and Statistics (Hidalgo 1994)
Source: Hidalgo 1994
Statistically analyzed following De Rijk’s sentence selection criterion
Oral testimony from 1994
The Hidalgo’s statistical research continues but he changed the sentence selection criterion and then the sentences are not the same sentences that we followed in our research.
Word Order and Statistics (Aldezabal et al. 2003)
The corpus of the Euskaldunon Egunkaria (from January 1999 to May 2000)14,557 declarative sentences where 512 sentences had spelled out the subject, the object and the verb
Source: Aldezabal et al. 2003
Psycholinguistics: Bronckart & Idiazabal 1982
Participants: 7 Groups of different age people
Aim: to analyze the Acquisition of different structures in Basque, and the processing strategies of these structures
Group 2: 4-5 years
Group 3: 5-6 years
Group 4: 6-7 years
Group 5: 7-8 years
Group 6: 10-11 years
Group 7: adults
Group 1: 3-4 years
Task: To represent the listened sentence with some toys
Psycholinguistics: Bronckart & Idiazabal 1982
Zakurrak neska bota duNeska zakurrak bota du
Zakurrak bota du neskaNeska bota du zakurrak
4-5 y.o. 5-6 y.o. 6-7 y.o. 7-8 y.o. 10-11 y.o. adults3-4 y.o.
15 1
14 3
19 113 7
18 2 20 011 9 11 9
19 1 20 0
11 9 18 2 20 0
20 0
15 1
11 7
19 115 5
17 3 20 015 5 16 4
20 0 20 0
14 6 20 0 19 1
20 0
Subject first sentences were comprehended well
Object first sentences were comprehended worse untill the age of 8
It seems that younger children understood the first constituent as subject and the second as object
INTERNAL SUMMARY 1
GENERATIVE GRAMMAR
Most linguists:
Basic word order Canonical word order
SOV SOV
Antisymmetrists: SVO SOV
FUNCTIONALISTS: SOV
STATISTIC RESEARCH:
De RijkHidalgoAldezabal et al
PSYCHOLINGUISTICS (ACQUISTION):
SOV
SOV
SOV/SVO
SVO
Previous Research on Word Order in Basque
Most frequent word order
Earliest acquired word order
Quote:
“Just as the theory of grammar has as its goals an account of Universal Grammar and parameters of language variation, the theory of sentence processing has as its goal the characterization of the universal parser,
the human sentence processing mechanism” Sekerina 2003: 302
Psycolinguistics Tecnique in Syntax
Reaction times
Participants perform the experiment at their own pace. To move from one element to the next element, participants had to press the space bar of the computer keyboard, one press for each element. Thus, participants decided the time they needed in order to process each element of the sentence, and therefore they decide the time they needed to comprehend the whole sentence.
SELF PACED READING MOVING WINDOW
Psycholinguistic Experiments in Basque: Method
The comprehension task allowed us to be sure that participants had understood the sentences they read. The task consisted in a yes-or-no question after each sentence. The answer of half of questions of each word order was “yes” and the other half was “no”.
COMPREHENSION TASK
Experiment 1: SOV-OSV
Goal: to determine whether OSV sentences have a higher processing cost than SOV sentences:
(a) longer reading times
(b) comprehension problems
23 participants (13 w and 10 m) Age-range was 18 to 36 (mean 25; SD ± 5).
Participants
Materials32 sentences in SOV and 32 sentences in OSV. 2 lists: 16 SOV and 16 OSV sentences per condition. 32 fillers (the same for two lists). Experimental conditions and fillers contained 4 words
emakume-ak gizon-a ikus-i du
woman-the/Subj. man-the/Obj. seen has
‘the woman has seen the man’
gizon-a emakume-ak ikus-i du
man-the/Obj. woman-the/Subj. seen has
‘the woman has seen the man’
Subject
Object
Verb Aux
Subject
Object
Verb Aux
Experiment 1: Material
Filler sentences consisted in one argument sentences
Manu futbolari bikaina da.
‘Manu is an excellent soccer player’.
Experiment 1: Material
Experiment 1: Recording
Recording
The EXPE6 (Pallier et al. 1997) recorded the reaction times and the answers of the participants:
(i) time to read each word of the sentence(ii) the time to perform the comprehension task (read and answer)(iii) whether the answer to the question is correct or not.
ExpectationsThe derived OSV word order sentences
(i) would require longer reading time(ii) would require longer reading time in the comprehension task(iii) would induce more errors in the comprehension task.
Experiment 1: Results
*
Reading time of two word orders
3500
3600
3700
3800
3900
4000
4100
4200
SOV OSV
Sentence Type
Tim
e (m
s)
p<0.005
Mean Reading Times of Sentences: Global Score
•SOV order is processed faster than OSV order
Reaction Times of Comprehension Task
2300235024002450250025502600265027002750
SOV OSV
Senteces Type
Tim
es (m
s)
p<0.002
Reaction Times in the Comprehension Task
• Questions about OSV word order elicited longer reading time
Experiment 1: Results
Experiment 1: Results
•OSV order elicited more errors than SOV order
Errors in the Comprehension Task
Comprehension Task: Errors
0%2%4%6%
8%10%12%14%16%
SOV OSV
Sentence Type
Per
cen
tag
e
p<0.001
Experiment 1: ResultsMean reading times Word by Word
Unmarked form processed faster than marked OSV requires a reanalysis of syntactic structure at subject position
Reading Times Word by Word
700
800
900
1000
1100
1200
1300
DP DP Vb Aux
Sentence Element
Tim
e (
ms
)
sov
osv
p<0.05
p<0.01
p<0.05p<0.05
interaction between first two DPs of the sentences F = 12.9; p < 0.002
Experiment 2: Ambiguous Chains
Goal: to determine how the ambiguous chains were processing and how the syntactic disambiguation happened.
(a) whether ambiguous chains were processed as canonical SOV sentences
(b) syntactic disambiguation elicited a syntactic reanalysis of the sentences
MORPHOLOGICAL AMBIGUITY
OBJECT PLURAL
SUBJECT SINGULAR
EMAKUME-AK‘WOMAN-X’
Emakume-ak gizon-ak ikusi ditu woman-X man-X see has
‘The woman has seen the men’ o ‘The man has seen the women’
Experiment 2: MethodMETHOD
Participants23 subjects (3 man and 20 woman; mean age 20.4, SD = 2.5).
MaterialsThree conditions (48 sentences per condition):
Three lists: one version of each item was assigned to one of the two listsList 1: 16 SOV / 16 OSV / 16 AMB + 48 fillers ( = 96 sentences)List 2: 16 SOV / 16 OSV / 16 AMB + 48 fillers ( = 96 sentences)List 3: 16 SOV / 16 OSV / 16 AMB + 48 fillers ( = 96 sentences)
48 filler sentences, the same for every list.
SOV condition
OSV condition
AMB condition
As in the previous experiment
The new condition
Emakume-ek gizon-ak ikus-i dituzte
women the Subj. men the Obj. seen have
‘the women have seen the men’
Emakume-ak gizon-ak ikus-i ditu woman-X man-X seen has‘the woman has seen the men’ or ‘the man has seen the women’
Gizon-ak emakume-ek ikus-i dituzte
man-X women-the Subj. seen have
‘the women have seen the men’
Subject
Object
Verb+aux
Subject
Object
Verb+aux
Ambiguous Chain
Experiment 2: Material
Total Reading Time
4000
4500
5000
5500
SOV AMB OSV
Sentence Type
Tim
e (
ms)
Experiment 2: Results
n.s.
p<0.001
Ambiguous chain is processed as SOV sentence
Mean Reading Times of Sentences: Global Score
Experiment 2: Results
The objects are processed faster than subjects
At subject second position, the reanalysis of the structure
SOV vs OSV, word by word
900
1100
1300
1500
1700
DP1 DP2 Verb Aux
Sentence Element
Tim
e (
ms
)
SOV
OSV
p<0.05 p<0.05 p<0.001p<0.001
interaction between first two DPs of the sentences F=17,137; p<0.001)
Experiment 2: Results
No evidences of syntactic reanalysis. No interaction
Ambiguous chains are processed as SOV order sentencesSOV is the simplest processing solution.
SOV vs AMB, word by word
900
1100
1300
1500
DP1 DP2 Verb Aux
Sentences Element
Tim
e (
ms
)
SOV
AMB
p<0.01p<0.05n.s. n.s.
Experiment 3: Verb-Medial Sentences
Goal: To determine how the verb-medial sentences were processing. These sentences are considered derived by linguists
Participants24 participants (2 men, 22 women; mean age 20, DS ± 3.21)
MaterialsFive conditions (100 sentences per condition):
SUBJECTsg-VERB-OBJECTsg
OBJECTsg-VERB-SUBJECTsg
SUBJECTpl-VERB-OBJECTpl
OBJECTpl-VERB-SUBJECTpl
AMBIGUOUS-VERB-AMBIGUOUS
100 filler sentences, the same for every list.
a. Gizonak ikusi du emakumea [Sak-V-Oa] Man-S see has woman-O
‘The man has seen the woman’
b. Emakumea ikusi du gizonak [Oa-V-Sak] Woman-O see has man-S
‘The man has seen the woman’
c. Gizonak ikusi ditu emakumeak [Amb-V-Amb] Man-X see has woman-X
‘The man has seen the women’ or ‘The woman has seen the men’
d. Gizonek ikusi dituzte emakumeak [Sek-V-Oak] Men-S see have women-O ‘The men have seen the women’
e. Emakumeak ikusi dituzte gizonek [Oak-V-Sek] women-O see have men-S
‘The men have seen the women’
Experiment 3: Material
total reading times
2900
3100
3300
3500
3700
Sak-V-Oa Oa-V-Sak AMBak-V-AMBak Sek-V-Oak Oak-V-Sek
sentences type
tim
e (m
s)
Experiment 3: Results
The sentences in singular were processed faster than the sentences in plural and the ambiguous chains (p<0.001).
There were no differences between SVO and OVS sentences; thus we considered the two structures derived.
Mean Reading Times of Sentences: Global Score
n.s.
n.s. n.s.
p<0.001
Experiment 3: Results
COMPREHENSION TASK: TIME
2000
2200
2400
2600
2800
3000
3200
Sak-V-Oa Oa-V-Sak AMBak-V-AMBak Sek-V-Oak Oak-V-Sek
sentence type
tim
e (m
s)Reaction Times in the Comprehension Task
Differences between the singular sentences and the remainning conditions
p<0.001
p<0.007
p<0.01
No differences between the plural conditions
Experiment 3: Results
comprehension task: errors
0123456789
Sak-V-Oa Oa-V-Sak AMBak-V-AMBak Sek-V-Oak Oak-V-Sek
sentence type
erro
rs (
/20)
Errors in the Comprehension Task
Ambiguous condition elicited most errors in the comprehension task (p<0.001 comparing to any condition)
Sak-V-Oa condition elicited fewest errors in the comprehension task
Experiment 3: ResultsMean reading times Word by Word
Reading Times Word by Word
500
600
700
800
900
1000
1100
1200
DP V Aux DP
sentence element
tim
e (
ms)
Sak-V-Oa
Oa-V-Sak
n.s. n.s. n.s.p<.025
The object was read faster than the subject
But, there was not any other difference, suggesting that verb-medial sentences are derived
Experiment 3: ResultsMean reading times Word by Word
Reading Times Word by Word
600
800
1000
1200
1400
DP V Aux DP
sentence element
tim
e (
ms)
Sek-V-Oak
Oak-V-Sek
There was no difference, suggesting that verb-medial sentences are derived
Experiment 3: ResultsMean reading times Word by Word
Reading Times Word by Word
500
700
900
1100
1300
1500
1700
DP V Aux DP
sentence element
tim
e (
ms)
Sek-V-Oak
AMBak-V-AMBak
P<0.008
Interaction between verb-auxiliary and sentence type. (F=5,924 p<.02)
The fact that the sentences’ first argument is considered the subject of the sentences could explain the differences between the SVO and the OVS without postulating that one of them (SVO/OVS) is more basic than the other
BEHAVIORAL RESULTS
a) SOV word order is processed faster and easier than OSV.
b) Object is read faster than the subject
c) OSV sentences require a syntactic reanalysis at subject position
d) Ambiguous chains were processed as canonical SOV.
e) No difference was found in verb-medial SVO and OVS word orders.
BEHAVIORAL EXPERIMENTS SUGGEST THAT THE
CANONICAL WORD ORDER IS SOV IN BASQUE.
NEUROLINGUISTICS
Language is processed by means of biological mechanisms
Knowledge of Language:
Nature
Acquisition
Use
Neurolinguistics
“Linguists seek a characterization of the nature of linguistic knowledge;
psycholinguists are after modeling the algorithms that implement this knowledge
when language users speak and understand; and neurolinguists are
interested in neural mechanisms that realize these algorithms, and their
cerebral localization. ”(Grodzinsky 2003)
ERP Components Related to Language
Early Left Anterior Negativity
Left Anterior Negativity
N400
P600
ERP Components Related to Language
Early Left Anterior Negativity
Das Baby wurde gefüttertThe baby was fed
*Das Baby wurde im gefüttertThe baby was in the fed
Phrase Structure Violations
Left Anterior Negativity
N400
P600
Hahne & Friederici 1999
ERP Components Related to Language
Early Left Anterior Negativity
Left Anterior Negativity
N400
P600
S-O
O-S
Matzke et al 2002
Agreement and Case Violations
Object relatives vs Subject relatives
Object questions vs Subject questions
Word Order variation in German relatives: SOV vs OSV
ERP Components Related to Language
Early Left Anterior Negativity
Left Anterior Negativity
N400
P600
Kutas & Van Petten 1988
Semantic Component
ERP Components Related to Language
Early Left Anterior Negativity
Left Anterior Negativity
N400
P600
Matzke et al 2002
Syntactic Violations
Syntactic Reanalysis
To seek the different brain responses to different word orders
Experiment 4: ERP evidences
Goal:
4 conditions
Materials
Canonical SOV conditionNon Canonical OSV condition
Unambiguous
Temporally Ambiguous
Canonical SOV conditionNon Canonical OSV condition
Subject
Verb auxObject
Subject Verb aux
Object
‘the wolf has eaten the sheep’
otso-ak
wolf-the/Subj
Ardi-a
Sheep-the/Obj
jan
eaten
du
has
‘the wolves have eaten the sheep(pl)’
Otso-ek
Wolves/the/Subj
ardi-ak
sheep-the/Obj
jan
eaten
dituzte
have
Experiment 4: Non Ambiguous
SOV temporally ambiguous
OSV temporally ambiguous
‘the wolf has eaten the sheep(pl)’
Otso-ak
Wolf-X
ardi-ak
sheep-X
‘the wolf has eaten the sheep(pl)’
otso-ak
wolf-X
Ardi-ak
Sheep-X
jan
eaten
ditu
has
jan
eaten
ditu
has
Experiment 4: Ambiguous
1 question after each block of 8 sentences was presented.
Design
Experiment 4: Method
240 sentences per condition were created (Total = 960 sentences) 4 lists: material were divided in 4 lists in order to avoid sentence repetition across conditions (like in behavioral expes)
30 blocks: Each lists contained 30 blocks of 8 sentences (2 per condition). Sentences into blocks and blocks were mixed randomly every experimental session
Experimental sentences were automatically presented word by word in the middle of the screen (words 300 ms; intervals 200 ms)
At the end of each sentence participants were asked to blink, and a green dash informed that a new sentence was going to start
Experiment 4: ParticipantsMETHOD
Participants
24 neurologically healthy and right handed native speakers of Basque, mean age 26 (SD ± 4.7) years; 8 males and 18 females
Experiment 4: Analysis
Parasagital
Temporal
Midline
Anovas were established in three regions:
Factors were:
Sentence type (four conditions)
Hemisphere of electrodes
Anterior/Posterior positions
Experiment 4: Results
Behavioral Results
The behavioral data showed that participants performed well in the experiment. In the comprehension task they performed correctly in 91% of trials (SD = ± 7.8)
LAN 375-425 ms
Experiment 4: Results
OTSOEK ardiak jan dituzteARDIA otsoak jan du
SOVOSV
COMPARING UNAMBIGUOUS SENTENCES
Left Anterior Negativity
Experiment 4: ResultsCOMPARING UNAMBIGUOUS SENTENCES
Otsoek ARDIAK jan dituzteArdia OTSOAK jan du
SOVOSV
Left Anterior Negativity
Experiment 4: ResultsCOMPARING UNAMBIGUOUS SENTENCES
P600
Otsoek ardiak JAN dituzteArdia otsoak JAN du
SOVOSV
Experiment 4: ResultsCOMPARING AMBIGUOUS CHAINS
ARDIAK otsoak jan dituOTSOAK ardiak jan ditu
NothingSENTENCE FIRST POSITION
AMB-SOVAMB-OSV
SENTENCE SECOND POSITION
AMB-SOVAMB-OSV
Ardiak OTSOAK jan dituOtsoak ARDIAK jan ditu
Experiment 4: ResultsCOMPARING AMBIGUOUS CHAINS
Frontal Negativity
Otsoak ardiak JAN DITUArdiak otsoak JAN DITU
AMB-SOVAMB-OSV
Experiment 4: ResultsCOMPARING UNAMBIGUOUS SENTENCES AND AMBIGUOUS CHAINS
OTSOAK ardiak jan ditu [AMB]OTSOEK ardiak jan dituzte [SOV]
600-800 Temporal: ST x H, F(1,23) = 4.51 p(HF) < 0.05 ; ST x H x AP, F(2,46) =
8.88 p(HF) < 0.001
Total Reading Time
4000
4500
5000
5500
SOV AMB OSV
Sentence Type
Tim
e (
ms)
Experiment 2
900
1100
1300
1500
DP1 DP2 Verb Aux
Sentences Element
Tim
e (
ms
)
SOV
AMB
p<0.01
p<0.05n.s. n.s.
SOV vs AMB
Experiment 4: ResultsCOMPARING UNAMBIGUOUS SENTENCES AND AMBIGUOUS CHAINS
ARDIAK otsoak jan ditu [AMB]ARDIA otsoak jan du [OSV]
Left Anterior Negativity
-a vs –ek and –ak = LAN
Experiment 4: ResultsCOMPARING UNAMBIGUOUS SENTENCES AND AMBIGUOUS CHAINS
SOV vs AMB-SOV
SOV vs AMB-OSV
Posterior positivity (P600)
Frontal Negativity
The differences observed in the comparison of the canonical non-ambiguous condition and both ambiguous conditions leads us to conclude that ambiguous chains are processed as canonical SOV sentences except when there is a disambiguating element which generates a revision of the syntactic structure
Verb Position
ERP RESULTS SUMMARY
a) LAN component in object first position.b) LAN-like component in subject second position.c) P600 component in verb position of OSV condition.
d) Nothing in ambiguously marked constituent positions.e) Frontal Negativity in verb and auxiliary position due to world knowledge disambiguated and syntactic reanalysis required condition
ERP EXPERIMENT SUGGESTS THAT THE CANONICAL
WORD ORDER IS SOV IN BASQUE.
UNAMBIGUOUS CONDITIONS
TEMPORALY AMBIGUOUS CONDITIONS
CONCLUSIONSPREVIOUS RESEARCH ON WORD ORDER IN BASQUE
BEHAVIORAL EXPERIMENTS
ERP EXPERIMENT
Most linguists: SOV SOV
Antisymmetrists: SVO SOV
Basic word order Canonical word order
AcquisitionSOV/SVO
Earliest acquired word order
Experiment 1: In derived OSV a syntactic reanalysis processExperiment 2: Ambiguous chains are processed like SOV sentences
Experiment 3: Verb-medial sentences didn’t show differences indicating that they could be derived in Basque
Ergatives and absolutives are processed differently, LAN
Verb position of derived OSV required a syntactic reanalysis, P600Temporally ambiguous chains disambiguated by means of world knowledge and posterior syntactic reanalysis elicited a Frontal Negativity