Title VI Anniversary conferenceTitle VI Anniversary conference Washington, DC, March 20, 2009

What heritage speakersWhat heritage speakers know…

Maria PolinskyHarvard University

big picturebig picture

• What does it mean to know a language?What does it mean to know a language?

Wh t th i l t i l• What are the crucial stages in language development and attrition?

• Is it possible to map out language p p g gsubsystems that are stable and that are most susceptible to change?p g

defining ‘heritage’defining heritage• Broad conception: those who have been raisedBroad conception: those who have been raised

with a strong cultural connection to a particular language, usu. through family interaction (Fishman 2001; Van Deusen-Scholl 2003)

• Narrow conception: those who have been exposed to a particular language in childhood but did not learn it to full capacity

b th l b d i t– because another language became dominant

who are heritage speakers?who are heritage speakers?

Tamarine Tanasugarn defeats Jelena Jankovic 6-3, 6-2Tamarine Tanasugarn defeats Jelena Jankovic 6 3, 6 2

Tamarine Tanasugarn defeats Jelena J k i 6 3 6 2Jankovic 6-3, 6-2

Q: Were you aware that Jelena was t li ith i j t th ?struggling with an injury out there?

TAMARINE TANASUGARN Y h lTAMARINE TANASUGARN: Yeah, also, well, the way of her movement a little bit, but I guess probably, I probably noticebut I guess probably, I probably notice when she have to do her serve. She maked a lot of, quite some double-fault at the end of the second set so I g essthe end of the second set, so I guess probably that's, uhm, bother her. So was it kind of maybe good for me somehow.kind of maybe good for me somehow.

typical perception of heritage kspeakers

• It is embarrassing/frustrating how manyIt is embarrassing/frustrating how many things they don’t know (get wrong)

the glass is ¾ fullthe glass is ¾ full

• It is rewarding/encouraging how muchIt is rewarding/encouraging how much heritage speakers already know

• The goal is to release this implicit• The goal is to release this implicit knowledge and build on it

heritage speakers as a distinct population

that deserves special attention

heritage speakers are not the same as…

• Heritage speakers have a different profile than g p puninterrupted L1 speakers

• Heritage speakers have a different profile than L2 speakersL2 speakers

• Heritage speakers are different from balanced (stable) bilinguals

• Heritage speakers are different from “forgetters” (stopped using L1 in adulthood, show aging effects)e ects)

• Heritage speakers are different from “fossilized” child language speakers

HL is like L1…HL is like L1…

• Early exposure to languageEarly exposure to language• Naturalistic setting (auditory input)

G d t l f f t i d l i• Good control of features acquired early in life (phonology, everyday lexicon, some t t )structures)

• Developmental errors

HL is different from L1HL is different from L1HL L1

Abundant continuous input

Successful and complete outcome of acquisitionComplex grammatical & pragmatic structures (associated with schooling)Fossilization ?

HL is like L2…HL is like L2…

• Varying amount and scope of inputVarying amount and scope of input• Resulting grammar is incomplete

D l t l d t f ff t• Developmental errors and transfer effects• Variable proficiency• Fossilized errors

HL is different from L2HL is different from L2HL L2HL L2

Late exposure to languagep g g

Problems with phonology (“accent”)

Instructed setting

Experience with literacy, formal registersregisters

HL and bilingualismHL and bilingualism

• HL: sequential not simultaneousHL: sequential, not simultaneous, bilingualism

• HL: unbalanced and unstable bilingualism• HL: unbalanced and unstable bilingualism, no diglossia

so who are heritage speakers?so who are heritage speakers?

• they simply forgot their languagethey simply forgot their language…

th f ili d t t f• they are fossilized at some stage of child language acquisition…

heritage speaker vs. forgetterheritage speaker vs. forgetterN=12

90%100% N=5

60%70%80%90%

fl i b

30%40%50%60% reflexive verb

subject gerundPr complement

0%10%20%

HS FGHS FG

Heritage speakers and forgetters, performance on grammaticality judgment tasks (GJT): Russian

HL and language forgettingHL and language forgetting

• Heritage speakers are different fromHeritage speakers are different from forgetters; main indicator of differences: performance on grammaticality judgmentperformance on grammaticality judgment task:

HSs are at chance FGs are close to controls– HSs are at chance, FGs are close to controls• Heritage speakers: competence problems.

Forgetters: performance deficiencyForgetters: performance deficiency

HL and fossilization

N ll h pothesis Ad ltNull hypothesis: Adult incomplete grammar = fossilized childfossilized child language, with the level of fossilization roughly g ycorresponding to the age of interruption

adult heritage speaker vs.child

Correct use of classifiers, Mandarin Chinese

80

100

40

60adjacent CLFstranded CLF

0

20

HS adult HS childN = 22 N = 16

heritage speaker is not a i hildcryogenic child

Evidence :Evidence :• Narrative structure (children outperform HS)• Classifiers(children outperform HS)• Classifiers(children outperform HS)• Semantics of number/plurality (HS outperform

children)children)• Relative clause comprehension• Aspect (children and HS show different errors)• Aspect (children and HS show different errors)

heritage language ≠ fossilized g g gchild language

• Hypothesis: Adult ypheritage grammar = fossilized child language, with the level of fossilization roughly corresponding to thecorresponding to the age of interruption

what exactly do adult HSs know?what exactly do adult HSs know?

• HSs differ from forgettersHSs differ from forgetters• HSs differ from children acquiring their first

languagelanguage• HSs make systematic errors which are

i il t f d i th i lsimilar to errors found in other special populations– Relative clause interpretation matches that of

Broca’s aphasics

heritage languageheritage language

• is a coherent structured systemis a coherent structured system• is different from the baseline language

what exactly do they know?what exactly do they know?

• Null Hypothesis:• Null Hypothesis:heritage speakers do not control ado not control a comprehensive language system; g g y ;instead, they retain a random collection of “language chunks”

what exactly do they know?y y• An incomplete

grammar differs fromgrammar differs from the grammar of the respective full planguage in a systematic, rather than random waythan random way

• The differences have twofold motivation:– Reanalysis in face of

insufficient dataLack of automatic– Lack of automatic access

addressing the deficitsaddressing the deficits• Reanalysis in face of insufficient data:Reanalysis in face of insufficient data:

expanding the range of data in order to facilitate the implicit revision of patternsp p

• Lack of automatic access: re-learning

re-learningre learning

• Heritage speakers are successful at re-Heritage speakers are successful at relearning because they already have a significant body of language data to add tosignificant body of language data to add to

• This is the strength that needs to be utilized more preferably at a younger ageutilized more, preferably at a younger age

• The social matrix of the heritage group is i th l d t h tunique: these people do not have to use

their L1 but it is still around them

back to big pictureback to big picture

• What does it mean to know a language?What does it mean to know a language?• What are the crucial stages in language

development?development?– In addition to the upward trajectory (which we

are used to in language acquisition field) weare used to in language acquisition field) we need to consider the downward trajectory of attrition and reanalysisattrition and reanalysis

acknowledgmentsacknowledgments• Ivano Caponigro, Bernard Comrie, Hana Filip, Michael p g p

Flier, Vlad Gitt, Gaby Hermon, Tanya Ionin, Olga Kagan, Tracy King, Robert Kluender, Soon-Hee Lee, Beth Levin, Roger Levy, Silvina Montrul, Andrew Nevins, Cathy g y yO’Connor, Yanny Siu, Antonella Sorace, Amy Weinberg, Alex Yanovsky

• Center for Research in Language UCSD• Center for Research in Language, UCSD• Center for World Languages, UCLA• Davis Center Harvard UniversityDavis Center, Harvard University• National Science Foundation• U.S. Department of EducationU.S. Department of Education

thank you!thank you!