BILINGUAL SENTENCE PROCESSING EVA M. FERN

1
BILINGUALSENTENCE PROCESSINGEVA M. FERNÁNDEZ

• QUEENS COLLEGE GRADUATE CENTERCITY UNIVERSITY
  OF NEW YORK

2
MY INTEREST IN THE MUNDANE

• Or Im bilingualam I normal?
• Humans are genetically predisposed to use
  Language we do so by designBILINGUALISM THE
  INDIVIDUAL
• But bilingualism is a product of the
  environmentBILINGUALISM ENVIRONMENT

3
BILINGUALISM ENVIRONMENT

• Q Most of the worlds children today
• grow up exposed to
• one language
• more than one language

?
The building of the tower of Babel, Pieter
Bruegel, 1563 Oil on oak panel,
Kunsthistorisches Museum Wien http//www.khm.at/ho
meE3.html
4
BILINGUALISM ENVIRONMENT

• Bilingualism is widespread, perhaps not so much
  in so-called BILINGUAL NATIONS Canada, Belgium
• Its pervasive even in places where there is an
  OFFICIAL LANGUAGE, e.g., Spain
• 1 official language 6 (recognized) regional
  languages
• 13 local languages 9 immigrant languages
• bilingualism and bidialectalism are common

(Source http//www.ethnologue.com)
5
2000, UNITED STATES
AND QUEENS COUNTY, NY
18 speak a LOTE at home, and speak English
54 speak a LOTE at home, and speak English
LOTE language other than English
(Source http//factfinder.census.gov)
6
BILINGUALISM THE INDIVIDUAL

• Bilingual communities are stable if there is a
  mutually exclusive need for Lx and
  Lyfunctional compartmentalization
• Functional compartmentalization doesnt apply
• But a great deal of research focuses on precisely
  the opposite problem
• How does a bilingual compare to the ideal speaker
  of two languages?

7
BILINGUALISM THE INDIVIDUAL

• Monolingual processing
• The range of bilingual linguistic behavior
• Bilingual competence
• Bilingual performance

8
monolingual processing
music

SIGNAL
LANGUAGE COMPETENCE PERFORMANCE
MEANING
math
knowledge about the real world
logic

9
monolingual processing

PERCEPTION
grammar lexicon
PRODUCTION
10
production

• Ill give you my undevoted attention!
• Youll earn her eternal grapefruit.
• Put the oven on at a very low speed.

• taddle tennis
• a glear plue sky

• They roasted a cook.
• If you give the nipple an infant
• You ordered up ending.

lexical retrieval
structural assignment
phonological encoding
11
perception

structure building
lexical access
phonological decoding
lexical retrieval
structural assignment
phonological encoding
12
and for the bilingual
?2 ??
13
BILINGUAL BEHAVIOR

• ABILITIES TO BE ACCOUNTED FOR
• Separate use of Lx and Ly (unilingual mode)
• Use of both Lx and Ly in the same discourse or
  same sentence (code-switching, bilingual mode)
• Translation Lx ? Ly (a specialized bilingual
  mode)
• Accuracy in grammaticality judgments in Lx and Ly

14
BILINGUAL SPEECH

• Accent in Lx or Ly, or both!(phonological
  transfer)? or ? or ?
• Structures taken from the other
  language(syntactic transfer)
• Hans fängt um drei Uhr an zum Arbeiten jeden
  Tag.(Hans fängt um drei Uhr zum Arbeiten jeden
  Tag an.)
• Yo me gusta esa casa, mami.(A mí me gusta esa
  casa, mami.)

Please call Stella.  Ask her to bring these
things with her from the store  Six spoons of
fresh snow peas, five thick slabs of blue
cheese, and maybe a snack for her brother
Bob. The Speech Accent Archive,
http//accent.gmu.edu/
Hans goes to work at three oclock every day.
(Myers-Scotton, 2006)
I like that house, mom. (Myers-Scotton, 2006)
15
BILINGUAL SPEECH

• UNINTENTIONAL? (incomplete acquisition,
  attrition)
• Transfer (phonological, syntactic)
• INTENTIONAL! (part of communicative repertoire)
• Lexical borrowing and calquing
• Mi building es el más grande del bloque.
• Code-switching
• Jétais certain que cétait pas la real thing.
• No porque quiera dispressare a mi language Italian

My building is the biggest on the block. (EMF)
I was certain that it wasnt the real thing (M-S,
2006)
Not that I want to undervalue my Italian language
(M-S, 2006)
16
BILINGUAL SPEECH
Language in Laredo A South Texas border
community http//www.tamiu.edu/kdegarcia/texmex/

• A  Ya no están, están en special
• B  You know what I saw by Burger King? Ahí no
  dice free cellulars, free pagers and everything?
• A  Pos ya habíamos ido ahí no?
• B  No, we didn't go there
• A  Which one do you say?
• B  Nex--- it's in between Taco Palenque and
  Burger King.
• A  Cuánto valen? no sabes?
• B  No sé --- it says there free cellulars, free
  pagers, pero I don't know
• A  When are you gonna go?
• B  Pues, ouch, pues what is, what is... it
  tomorrow is Wednesday, tomorrow es cuando tienes
  más tiempo tú no?
• A  Yes, but I have to do some homework I dont
  know if we have time pos we can go
• B  Ahí para comprarlo para este weekend pa
  este weekend
• A Pero antes de la semana que entra

A  Theyre no longer on special B  You know
what I saw by Burger King? Doesnt it say there
free cellulars, free pagers and everything? A 
Wed already gone there, right? B  No, we
didn't go there A  Which one do you say? B 
Nex--- it's in between Taco Palenque and Burger
King. A  How much do they cost, do you
know? B  I dont know --- it says there free
cellulars, free pagers, but I don't know A  When
are you gonna go? B  Well, ouch, well what is,
what is it tomorrow is Wednesday, tomorrow
iswhen you have more time, right? A  Yes, but
I have to do some homework otherwise, if we have
time so we can go B  There to buy it this
weekend for this weekend A But before the
coming week
17
bilingual processing

• Size (proficiency?) based on
• age of acquisition
• frequency of use
• etc.

PERCEPTION
PRODUCTION

• The two competence repositories are connected
• Lx may have Ly-like rules
• Lx lexical items will be linked to Ly lexical
  items

18
Exploring the model empirically

• Separate but interconnected competence
  repositories
• Is Lx active during unilingual Ly processing?
• Do Lx lexical representations affect structural
  decisions during Ly processing?
• Unitary and language-independent performance
  mechanisms
• Exploit cross-linguistic differences in
  monolingual processing to examine processing in
  Lx and Ly, in bilinguals

19
DATA COLLECTION

• Off-line, UNTIMED techniques (questionnaires)
• Grammaticality judgment tasks
• Sentence comprehension tasks
• On-line, TIMED techniques
• Behavioral methods
• Cross-modal priming, probe recognition,
  rapid-serial visual presentation, self-paced
  reading, self-paced listening, etc.
• Eyetracking techniques (eyes are the windows to
  the world)
• Reading Visual world paradigm
• Neurophysiological techniques
• ERP fMRI

20
DATA COLLECTION

• Variables to consider
• Cost availability
• Degree of invasiveness
• Ecological validity (?!)
• Contrasts to manipulate
• Participants
• Language history, working memory, reading skill,
  sex, handedness
• Materials
• Phonological, morphological, syntactic,
  semantic, pragmatic contrasts
• Visual world contrasts

21
FRENCK-MESTRE PYNTE, 1997

• One example (among others) of how sentence
  processing routines can be compared in
  monolinguals and bilinguals
• Indeterminate terminology
• Native speakers, L1 speakers, monolingualsv.
  non-native speakers, L2 speakers, bilinguals ?
  whats the difference?
• Two experiments
• Experiment 1 syntactic ambiguity similar in L1
  L2
• Experiment 2 ambiguity only in one language

22
E1 Similar construction, L1 and L2

• They accused the ambassador
• of espionage but nothing came of it.
• of Indonesia but nothing came of it.
• He rejected the manuscript
• on purpose because he hated its author.
• on horses because he hated its author.
• N 16 x 2 participants N 24 x 2 items
• L1 v. L2 French speakers attachment
  transitivity

23
Empirical questions

• Does the structural processor (parser) operate
  the same way in L1 and L2?
• Will L2-ers produce similar patterns of eye
  movements as L1-ers?
• (Qualitative differences)
• Is there a cost in processing L2?
• Will L2-ers take more time than L1ers?
• (Quantitative differences)

24
Apparatus
25
Eye movements

• Visual span
• 9 letters, periphery
• Focus (fixations) move (saccades
• eyetracker records focus

• Measurements
• First pass gazes
• First pass fixations
• Regressions
• Second pass gazes

http//gandalf.psych.umn.edu/gellab/mrchips/chips
2d.html animation by Steve Mansfield Tim Klitz
26
Data to discuss

• First pass gazes left-to-right fixations within
  a region that had not been previously read, plus
  all within-zone regressions (p. 126)
• Second pass gazes all fixations not included in
  the first pass analysis and often comprising
  more than one re-reading of the sentence (p.
  130)

They accused the ambassador of / espionage / but
/ nothing / came of it.
/ Noun of PP / N 1 / N 2 /
27
E1 First pass gaze durations
Monolinguals Bilinguals Mean


28
E1 Second pass reading times
Monolinguals Bilinguals Mean

29
E1 Main findings

• Processing is qualitatively similar in L1 and L2,
  particularly in first-pass measures
• Processing time is quantitatively different
  increased reading times in L2, particularly in
  second-pass measure

30
E2 Ambiguity in English only

• Every time the dog obeyed the pretty little girl
  showed her approval.
• Chaque fois que le chien obéissait la jolie
  petite fille montrait sa joie.
• Every time the dog barked the pretty little girl
  showed her approval.
• Chaque fois que le chien aboyait la jolie petite
  fille montrait sa joie.
• N 16 x 2 participants N 16 x 2 items
• Edom v. Fdom bilinguals Transitive (EN)
  v. intransitive verb

31
E2 Ambiguity in English only

• Every time the dog obeyed the pretty little girl
  showed her approval.
• Chaque fois que le chien obéissait la jolie
  petite fille montrait sa joie.
• Every time the dog barked the pretty little girl
  showed her approval.
• Chaque fois que le chien aboyait la jolie petite
  fille montrait sa joie.
• N 16 x 2 participants N 16 x 2 items
• Edom v. Fdom bilinguals Transitive (EN)
  v. intransitive verb

32
First pass, at subordinate verb
-ish

33
First pass, at beginning of matrix NP
?


34
Second pass, means




35
E2 Main findings

• Influence of language not being processed, if
  non-dominant, during lexical access (first-pass
  gazes at subordinate verb)
• Otherwise, very similar patterns of eye movement
  behavior reflecting operations at work during
  syntactic ambiguity resolution (first-pass gazes
  at beginning of NP, second pass times)
• How should findings like this inform a model for
  the bilingual linguistic architecture?

36
bilingual processing

PERCEPTION
PRODUCTION