THE%20BILINGUAL%20MIND

1
THE BILINGUAL MIND

• Antonella Sorace
• 7 December 2004

2
Outline

• How successful can you be if you start learning a
  second language as an adult?
• What are the differences between early
  bilingualism in childhood and late bilingualism
  in adulthood?
• What happens to your first language after you
  have been speaking a second language for many
  years?
• Is the bilingual brain different from the
  monolingual brain?
• Do data from second language speakers help to
  understand how language in general works?

3
An interdisciplinary enterprise
RESEARCH ON THE BILINGUAL MIND
EXPERIMENTAL PSYCHOLOGY
LINGUISTICS
COGNITIVE NEUROSCIENCE
4

• How successful can you be if you start learning
  a second language as an adult?

5
A critical period for language?

• In many animal species, failure to learn various
  skills before a certain age makes it difficult or
  even impossible to learn those skills later.
  E.g.
• In ducklings ability to identify and follow the
  mother
• In kittens ability to perceive visual images.
• In sparrows ability to learn the fathers song.

6
Early exposure to language is necessary

• Children raised in conditions of extreme
  isolation and deprivation do not develop normal
  grammatical abilities.
• Deaf children of hearing parents who are
  diagnosed as deaf when they are 2 or 3 are
  impaired in their development of sign language.

7
Why a critical period for language?

• A biological mechanism innately geared to the
  acquisition of language in our species.
• Evolutionary advantages of having the mechanism
  early in life.

8
But what about SECOND language?

• Does this mean that second language learning is
  compromised even if first language development
  was normal?
• Does the fact of already knowing a language help?

9
Near-native speakers

• Speakers who started learning a second language
  as adults and reached an exceptional level of
  ability in it.
• They would be off the scale in the IELTS band of
  English proficiency.

10
Subject pronouns in Italian

• Subject pronouns can be omitted when they refer
  to an entity that is clear in context
• Maria non cè, è andata a casa
• Maria isnt here, she went home
• They cannot be omitted in other cases, for
  example when two entities are contrasted to one
  another
• Maria e Yuri non si capiscono lei parla
  litaliano, lui no.
• Maria and Yuri dont understand each other she
  speaks Italian, he doesnt.

11
Two kinds of knowledge
12
Near-native speakers errors

• Near-native speakers of Italian and Spanish may
  say
• Maria non cè, LEI è andata a casa.
• Maria isnt here, she went home.
• Is this due to interference from English?

13
Cant be (only) interference from English

• English and Spanish non-native speakers of
  Italian make the same mistake.
• They know that in Italian subject pronouns can be
  omitted they know what the contextual conditions
  are.
• In most cases, they use subject pronouns
  correctly.

14
It could be a coordination problem
15
Another interface problem in near-native speakers

• The difference between the sounds /i/ and /I/
• SHEEP - SHIP
• CHEAP - CHIP
• SEEK - SICK
• BEAT - BIT
• DEEP - DIP
• Etc.

16
The near-native speakers dilemma
17
The snickers vs. sneakers problem

• THIS..
• OR
• THIS?

18
More on interfaces auxiliary verbs in Italian

• ESSERE be and AVERE have.
• Maria ha lavorato. Maria has worked
• Maria è partita. Maria has left
• Same distinction as ETRE vs AVOIR in French
• Marie a travaillé.
• Marie est partie.
• In early modern English
• Christ is risen.
• The Lord is come.

19
In Italian as a second language

• Auxiliary ESSERE with verbs such as arrivare
  arrive, venire come, partire leave ------gt
  ACQUIRED EARLY.
• Auxiliary ESSERE with verbs such as rimanere
  stay, bastare suffice, piacere like-------gt
  ACQUIRED MUCH LATER OR NOT ACQUIRED AT ALL, NOT
  EVEN AT THE NEAR-NATIVE LEVEL.

20
Native speakers have gradient intuitions

• Native speakers of Italian, French, German and
  Dutch STRONGLY AGREE on the fact that (the
  equivalents of) verbs such as arrive, leave, come
  select (the equivalents of) BE.
• They DISAGREE, or are UNCERTAIN, on like, stay,
  exist sometimes they like them with BE,
  sometimes with HAVE.

21
The Auxiliary Selection Hierarchy

• The choice of auxiliaries is conditioned not only
  by the grammar, but also by the semantic type of
  verb.
• CHANGE OF LOCATION BE'
• (arrive, come leave, etc.)
• EXISTENCE OF STATE
• (like, stay, be sufficient, etc.)
• HUMAN ACTIVITY 'HAVE'
• (work, talk, play, etc.)

22
Another problematic interface
23
A methodological spin-off how to detect gradience

• If developmental data are gradient, we need a
  method that can detect gradience.
• Magnitude estimation, a method borrowed from
  psychophysics, allows researchers to capture fine
  shades of gray in judgments of linguistic
  acceptability.
• See http//www.webexp.info for a web-based
  application of Magnitude Estimation developed by
  Frank Kellet and Martin Corley.

24
The story so far

• Many properties of grammar can be successfully
  acquired in a second language, but properties
  that involve interfaces between different aspects
  of language may remain non-native even at the
  highest level of attainment.

25

• What happens to your first language after you
  have been speaking a second language for a long
  time?

26
Effects of the second language on the first
language
27
(No Transcript)
28
Ex-native speakers

• Speakers experiencing attrition in their native
  language at Stage 1 have problems with
  constructions that require the integration of
  different types of knowledge, just like
  near-native speakers.
• They also say
• Maria non cè, LEI è andata a casa

29

• Ex-native speakers of Spanish often leave out the
  preposition a with animate direct objects
• Maria vio a mi abuela
• Maria saw my grandmother
• Maria vio la película.
• Maria saw the film.
• This property is also applied inconsistently by
  advanced non-native speakers of Spanish.

30
Interface aspects last in, first out
31

• This research programme that compares acquisition
  and attrition requires the contribution of both
  LINGUISTICS and EXPERIMENTAL PSYCHOLOGY.
• We have both here in Edinburgh.

32
Some of the questions we are working on in PPLS

• Does hearing a language automatically activate
  (prime) the other?
• Can bilinguals be induced into producing
  incorrect word orders in both their languages?
• What are the effects of feedback, correction, and
  explicit knowledge on second language development?

33
The broad view

• Research on bilingual processing helps us to
  understand how human language processing works in
  general.
• Research on bilinguals can inform computational
  models of natural language processing.

34
Bilingual first language acquisition (early
bilingualsim)

• Bilingual children develop two native languages,
  so in general reach higher levels of attainment
  than adult learners.
• They do not normally mix their languages (unless
  they want to!).
• How early do they differentiate the two languages
  they are acquiring?

35
Crossover effects in bilingual children

• The dominant language influences the weaker
  language.
• The language with less complex interface
  conditions influences the language with more
  complex interface conditions.

36
Effects of input

• Bilingual children often hear
• Less input (in both languages) than monolingual
  children.
• Non-native input in the minority language.
• Input resulting from attrition (usually from the
  parent who is a native speaker of the minority
  language).

37
Effects of bilingualism on non-linguistic tasks

• Does the bilinguals experience of constantly
  managing two linguistic systems have an effect on
  coordination in non-linguistic tasks?

38
Cognitive control involves.

• Paying selective attention to the relevant
  aspects of a problem
• Inhibiting attention to irrelevant information
• Switching between competing alternatives.

39

• Bialystok, Craik, Klein Viswanathan (2004)
  bilinguals are better than monolinguals at tasks
  involving cognitive control.
• The advantages are maintained in older age
  bilingualism may help to offset age-related
  cognitive losses.

40
Future research

• Is there a difference between early and late
  bilinguals with respect to cognitive control in
  non-linguistic tasks?
• The answer will bring us closer to understanding
  the relationship between language and other
  cognitive faculties.

41
The bilingual brain

• Structural vs. functional factors what are the
  neural substrates of bilinguals behaviour?
• Does the bilingual brain have a different neural
  organization from the monolingual brain?
• Does the bilingual brain have different neural
  substrates for the native and second language(s)?

42
(No Transcript)
43
Participants

• 6 native speakers of Italian who had been
  screened and categorized as near-native speakers
  of English.

44
Materials used

• NORMAL
• The mother was kissing the child.
• La madre stava baciando il bambino.
• SYNTACTICALLY ANOMALOUS
• The artist was moulding a clays.
• Lartista stava modellando il argilla.
• SEMANTICALLY ANOMALOUS
• The master was teaching the rice.
• Il maestro stava insegnando al riso.

45
Syntactically incorrect sentences, Italian
46
Syntactically incorrect sentences, English
47
Why these differences?

• Explanation A bilinguals need additional
  resources in the second language to compensate
  for their inefficient processing abilities.
• Explanation B bilinguals develop additional
  resources for the second language.

48
Structural scan from Mechelli et al. (2004)
higher density of grey matter in left inferior
parietal lobe.
49

• Our bilinguals
• Mechellis bilinguals

50
What do these results tell us?

• The structure of the brain can be altered by the
  experience of acquiring a second language.
• The neural substrates of second languages in late
  bilinguals are different from those of their
  native language, even if the second language is
  mastered at near-native levels.

51
What next?

• Neural correlates of interface conditions on
  grammatical knowledge.
• Work by Hagoort et al. at the Donders Center for
  Cognitive Neuroscience in Nijmegen is promising.

52
Non-academic implications of research on the
cognition of bilingualism

• Second language classroom learning and teaching
  and computer-assisted language learning.
• Native language maintenance.
• Education of bilingual families.
• (see http//www.lsadc.org for our leaflet on
  Raising bilingual children).

53
To conclude

• The cognitive study of the bilingual mind is an
  exciting interdisciplinary enterprise.

54
Credits

• Ellen Bard
• Adriana Belletti
• Holly Branigan
• Michela Cennamo
• Francesca Filiaci
• Caroline Heycock
• Frank Keller
• Bob Ladd
• Géraldine Legendre
• Pim Levelt
• Martin Meyer
• Mits Ota

• Martin Pickering
• Janet Randall
• Luigi Rizzi
• Herbert Schriefers
• Ludovica Serratrice
• Neil Smith
• Paul Smolensky
• Ianthi Tsimpli
• Nigel Vincent
• Angeliek van Hout
• Lydia White

55
Special thanks

• To my parents
• To my family Bob, Marco, and Carlo.