Cross-linguistic influence at the level of syntactic ... · • The wh -questions produced by bilingual children are syntactically permissible in the target language. • Bilinguals

Cross-linguistic influence at the level

of syntactic features:

The case of wh-questions*

SYMPOSIUM ON WH-QUESTIONS IN BILINGUAL ACQUISITION

INTERNATIONAL ASSOCIATION FOR THE STUDY OF CHILD LANGUAGE

MONTREAL, CANADA | JULY 20, 2011

The case of wh-questions

Lyn Shan Tieu

<[email protected]>

*Thanks to Diane Lillo-Martin for helpful comments and discussion. This work has been funded in part by SSHRC 752-

2008-2450. This project was also supported in part by Award Number R01DC009263 from the National Institute on

Deafness and Other Communication Disorders (United States). The content is solely the responsibility of the authors

and does not necessarily represent the official views of the National Institute on Deafness and Other Communication

Disorders or the National Institutes of Health (United States).

Wh-questions cross-linguistically

• Languages like English exhibit wh-movement.

– In a single wh-question, the wh-phrase must

undergo movement.

• Languages like Chinese and Japanese are wh-• Languages like Chinese and Japanese are wh-

in-situ languages.

– In a single wh-question, the wh-phrase remains in

situ.

-2-

Wh-questions cross-linguistically

• English-type wh-questions involve movement:

(1) What did you eat __?

• Chinese-type wh-phrases remain in situ:• Chinese-type wh-phrases remain in situ:

(2) lei5 sik6-zo2 mat1je5?

you eat-PERF what

‘What did you eat?’(Yip & Matthews, 2000:195)

-3-

Wh-movement languages

• Traditional Minimalist account (Chomsky, 1995):

– The strong Q feature on the matrix C0 must be

checked.

– The wh-feature on the wh-phrase is attracted to C0 . – The wh-feature on the wh-phrase is attracted to C0 .

– The wh-phrase moves to the Specifier of CP to

check the Q feature.

(3) Whati did John buy __i?

-4-

Wh-in-situCG

• Some wh-movement languages exhibit a restricted kind of wh-in-situ that is pragmatically licensed in certain contexts.

• Example: echo questions

(4) A: Mary ate a skunk.

B: Mary ate WHAT?(Pires & Taylor, 2007)

• Wh-in-situ is possible when the information being requested is expected to be part of the Common Ground (Pires & Taylor, 2007).

-5-

Wh-in-situ languages

• One account:

– Languages like Chinese and Japanese typically have

wh-particles.

– Insertion of such a particle (overt or null Q-

morpheme) checks the Q feature on C0 (Cheng, 1991).morpheme) checks the Q feature on C0 (Cheng, 1991).

– The wh-phrase remains in situ.

(5) Zhangsan mai-le shenme (ne)

Zhangsan buy-PERF what PART

‘What did Zhangsan buy?’

-6-

Wh-in-situ languages

• Other accounts:

– Question complementizers in languages like Chinese and

Japanese have a weak Q feature.

– In these languages, the wh-phrase moves covertly at LF

to check the feature (Huang, 1982).

– The wh-phrase is spelled out “in situ”.– The wh-phrase is spelled out “in situ”.

(6) Zhangsan mai-le shenme (ne)

Zhangsan buy-PERF what PART

‘What did Zhangsan buy?’

-7-

Bilingual production of wh-questions

• What kinds of wh-structures will we find in the

production of bilingual children who are

acquiring two languages that differ in their

wh-options?wh-options?

-8-

Sneak peek

• The wh-questions produced by bilingual children

are syntactically permissible in the target language.

• Bilinguals may produce more wh-in-situ than

monolinguals, but only if wh-in-situ is a syntactic monolinguals, but only if wh-in-situ is a syntactic

possibility in the target language.

• Bilinguals may produce more wh-movement

questions than monolinguals, but only if wh-

movement is a syntactic possibility in the target

language.

-9-

What are the empirical facts?

• Case studies in the literature

– Italian / Indonesian

– Korean / English

– Cantonese / English– Cantonese / English

– ASL / English

– Libras / Brazilian Portuguese

-10-

Case study #1: Indonesian-Italian (Soriente, 2007)

• Jakarta Indonesian– Wh-movement is optional; most wh-forms stay in situ.

– No evidence for obligatory movement to the beginning of the sentence (Cole et al., in press).*

(7) potong rambut di mana?

cut hair LOC wherecut hair LOC where

‘Where did you cut your hair?’

(8) di mana potong rambut?

LOC where cut hair

‘Where did you cut your hair?’(Soriente, 2007:339)

*There is some evidence that wh-movement in Bahasa Indonesian, a wh-in-situ language, is in fact focus movement (Zavitnevich-Beaulac, 2005). The same may hold for Jakarta Indonesian.

-11-


• Italian

– Obligatory wh-movement to sentence-initial position, plus subject-auxiliary inversion, except in echo questions.

(9) cosa fa Gianni?

what do-IND.PRES.3SG Gianni

‘What does Gianni do?’

(Soriente, 2007:337)

-12-


• Monolingual Italian-speaking children do not

produce questions without movement.

– Camilla corpus: atypical wh-in-situ structures are never

produced (Antelmi, 1997).

– Analysis of Italian data in CHILDES: very few in-situ – Analysis of Italian data in CHILDES: very few in-situ

questions found (Soriente, 2007).

• Monolingual Indonesian-speaking children prefer

the in-situ structure, especially in object position.

– Acquisition of wh-questions is error-free (Cole et al., 2000a,b;

Soriente, 2007).

-13-


• Case study of Guglielmo, born in Jakarta,

Indonesia, to Italian mother and Indonesian

father.

• Parents followed one-parent-one-language • Parents followed one-parent-one-language

strategy; more Indonesian input.

• Data: video recording transcriptions (weekly

hour-long recordings from 0;08,00), diary

notes.

-14-


• Guglielmo’s development of wh-questions in

Indonesian (from 1;07) is comparable to that of

monolinguals.

• In his Italian, Guglielmo reproduces structure of • In his Italian, Guglielmo reproduces structure of

Indonesian questions with wh-forms in situ.

• Such sentences are “very frequent” in the

database and in the diary study, especially

between 3;02 and 4;00 (Soriente, 2007).

-15-


(10) l pesci mangiare che cos‘é? [3;02]

DEF-M.PL fish-M.PL eat-INF what thing

‘What are the fish eating?’(Soriente, 2007:346)

(11) Maeldane vuole salutare chi è? [3;07]

Maeldane want-3.SG greet-INF who be-IND.PRES.3SG

‘Who will Maeldane greet?’

(Soriente, 2007:348)

-16-

Case study #2: Korean-English(Park, 2008)

• English– Obligatory wh-movement, except in CG contexts.

• Korean– Wh-in-situ.

(12) What will Betty wear __?

(13) Betty-ka mwu-lul ipulkuya?

Betty-NOM what-ACC wear

‘What will Betty wear?’(Park, 2008:8)

-17-


• Monolinguals (English)

– 23 children (2;03-5;05) (data from CHILDES)

• Bilinguals• Bilinguals

– 7 Korean-English bilingual children residing in the US

(3;04-5;05 at onset of study)

– sequential bilinguals (Korean first, English around

3;00 in preschool)

-18-


• Monolinguals

– produced target wh-movement constructions

– produced some pragmatically licit wh-in-situ (e.g.,

echo questions)echo questions)

• Bilinguals

– produced target-like wh-in-situ in their Korean

– also target-like in their English (no wh-in-situ

constructions observed in English)

-19-


(Park, 2008:72)

(Fig. 3: Percentage

of wh-in-situ to wh-

movement by

group, Park,

2008:72)-20-

Case study #3: Cantonese-English(Tieu, 2010; see also Yip & Matthews, 2000, 2007)

• English is a wh-movement language, except for CG contexts.

• Cantonese Chinese is a wh-in-situ language.

(14) What did you eat __?

(15) lei5 sik6-zo2 mat1je5?

you eat-PERF what

‘What did you eat?’(Yip & Matthews, 2000:195)

-21-


• Monolinguals

– 4 American English-speaking monolingual children

(CHILDES) (range: 1;02 – 5;02).

• Bilinguals• Bilinguals

– 7 Cantonese-English bilingual children residing in

Hong Kong (1;03-3;06 at onset of study).

– MLU differentials (C/E) range from 0.927 to 1.560; by

MLU differential measure, 6 of 7 children were

Cantonese-dominant.

-22-


• Monolingual English-speaking children are target-like in their moved wh-questions; some evidence of pragmatically licit wh-in-situ after about 2;11.

• Monolingual Cantonese-speaking children are target-like in their wh-in-situ (Peng, 1998; Yip & Matthews, 2000).

95.23100.00 100.00 99.19

Figure 1: Monolinguals: Proportion of what-object types

-23-

0

10

20

30

40

50

60

70

80

90

100

Adam Eve Naomi Nina

0.44 0 0 0.274.33

0 00.54

95.23100.00 100.00 99.19

E: Illicit wh-in-situ

E: Licit wh-in-situ

E: Wh-movement


• Bilinguals produced more wh-in-situ in their English compared to monolinguals (both pragmatically licit and illicit wh-in-situ).

• Their Cantonese wh-questions were target-like (in-situ).

98.70

Figure 2: Bilinguals: Proportion of what-object types in

EnglishE: Illicit wh-in-

-24-

3.17 1.30 4.44 6.67 3.33 3.67

51.43

0.00 0.00 4.44 0 0 3.6710.48

96.83 98.7091.11 93.33 96.67

92.66

38.10

0

10

20

30

40

50

60

70

80

90

100

Alicia Charlotte Janet Kathryn Llywelyn Sophie Timmy

E: Illicit wh-in-

situ

E: Licit wh-in-

situ

E: Wh-

movement

Case studies #4-5: ASL-English / Libras-BP(Lillo-Martin, Quadros, and Chen Pichler, ongoing work)

• American Sign Language (ASL) and Brazilian Sign Language (Libras) allow wh-initial, wh-in-situ, wh-final, and doubling constructions (Petronio & Lillo-Martin, 1997).

• Wh-final and doubling are analyzed as emphatic focus constructions (Nunes & Quadros, 2005). constructions (Nunes & Quadros, 2005).

-25-

(Lillo-Martin & Quadros, 2006)

• Monolingual ASL/Libras signing children

– Wh-initial and wh-in-situ structures are prevalent

in spontaneous signing from early on (1;09) (Lillo-

Martin & Quadros, 2006).


Martin & Quadros, 2006).

– In elicited production, high proportion of wh-final

and wh-double structures observed in ASL by 5-6

years (Lillo-Martin, 2000).

-26-

• Spontaneous production of younger ASL-English / Libras-BP bimodal bilingual children (1 ASL, 1 Libras, age range 1;0-3;0):

– ASL/Libras

• preliminary data so far show examples of target-like wh-structures (initial, in situ, doubling) by around 1;11


wh-structures (initial, in situ, doubling) by around 1;11

– English/BP

• wh-in-situ constructions appear earlier compared to monolinguals (around 1;11 compared to 2;11 for English, 3;09 for BP)

• wh-doubling (before 3;0)

• preference for wh-initial questions (after 3;0)

-27-

• Spontaneous production (English):

(16) Where airplane where? (2;00)

‘Where is the airplane?’


‘Where is the airplane?’

(17) Mommy where? (2;00)

‘Where is Mommy?’

(Quadros, Lillo-Martin, & Chen Pichler, in prep.)

-28-

• Elicited production of wh-questions from older ASL-

English / Libras-BP bimodal bilinguals (3 ASL, 13

Libras, age range 5;0-7;0):

– ASL/Libras


– ASL/Libras

• very high proportion (nearly 100%) of wh-initial

structures for all wh-question types

– English/BP

• target-like wh-initial structures

-29-

Summarizing the findings

Table 1. Wh-questions in bilingual development

Language pair 1 Language pair 2 Language pair 3 Language pairs 4-5

Indonesian Italian Korean English Cantonese English ASL/Libras English/BP

Target /

Monolingual

findings

wh-in-situ;

movement

optional (for

focus?)

obligatory

movement;

wh-in-situ

in ‘common

ground’

wh-in-situ obligatory

movement;

wh-in-situ

in ‘common

ground’

wh-in-situ obligatory

movement;

wh-in-situ

in ‘common

ground’

wh-movement;

wh-in-situ; wh-

final and

doubling in

focus

obligatory

movement;

wh-in-situ in

‘common

ground’ ground’

contexts

ground’

contexts

ground’

contexts

focus

constructions

ground’

contexts

Bilingual

findings

target-like increased

wh-in-situ

compared

to

monolings.

target-like target-like target-like increased

wh-in-situ

compared

to

monolings.

younger:

target

structures incl.

movement, in

situ, and

doubling

older:

increased wh-

movement

compared to

monolings.

earlier wh-in-

situ

compared to

monolings.,

some

doubling

-30-

Summarizing the findingsTable 1’. Wh-questions in bilingual development

Language pair 1 Language pair 2 Language pair 3 Language pair 4 Language pair 5

A B A B A B A B A B

Target /

Monoling.

findings

IN SITU MVMT

*IN SITU

IN SITU MVMT

*IN SITU

IN SITU MVMT

*IN SITU

MVMT

IN SITU

FINAL

DOUBLING

MVMT

*IN SITU

MVMT

IN SITU

FINAL

DOUBLING

MVMT

*IN SITU

Bilingual

findings

IN SITU ↑ IN SITU IN SITU MVMT IN SITU ↑ IN SITU younger:

MVMT

earlier IN

SITU

younger:

MVMT

earlier IN

SITUfindings MVMT

INSITU

DOUBLING

older:

↑ MVMT

SITU MVMT

INSITU

DOUBLING

older:

↑ MVMT

SITU

-31-

• If language A is unambiguously in situ, bilinguals are target-like in A (they produce wh-in-situ and not wh-movement).

• Bilinguals may be non-target-like in producing more wh-in-situ OR in producing more wh-movement.

• Having in situ in language A and both in language B does notnecessitate non-target-like rates of in situ in language B.

Generalizing the syntax across languages

• Wh-movement involves movement of the wh-

phrase to Spec,CP to check a strong feature on C0.

– Italian, English, Brazilian Portuguese, ASL, Libras

• Wh-in-situ involves selection of a question • Wh-in-situ involves selection of a question

complementizer bearing a weak Q feature (+LF

movement of the wh-phrase) OR the insertion of

a null question particle.

– Indonesian, Korean, Cantonese, ASL, Libras

– CG: Italian, English, Brazilian Portuguese

-32-

Generalizing wh-in-situ

• Wh-in-situ is syntactically licensed in

languages like English, just as it is in languages

like Chinese.

• The difference is that wh-in-situ in English is • The difference is that wh-in-situ in English is

pragmatically constrained, i.e. subject to

discourse-pragmatic conditions (that don’t really

concern us here).

-33-

What do bilingual children do?

• They produce wh-questions that are

syntactically well-formed – even when they

appear at non-target-like rates.

– Their wh-movement questions are syntactically– Their wh-movement questions are syntactically

well-formed in the target language.

– Their wh-in-situ questions are syntactically well-

formed in the target language.

-34-

What do bilingual children not do?

• They do not produce wh-questions are that syntactically ill-formed.

• What is syntactically ill-formed?

– A wh-movement question in a wh-in-situ language.– A wh-movement question in a wh-in-situ language.

• Syntactic explanations?

– e.g., insertion of null question particle prevents movement of wh-phrase;

– e.g., weak Q feature on C0 cannot trigger overt movement, and movement must be feature-driven.

-35-

What have we defined?

• What is permissible vs. impermissiblei.e. where can and can’t we observe ‘cross-linguistic influence’?

– Bilinguals (observed so far) may produce more wh-movement questions or more wh-in-situ questions in languages that allow both options because these are syntactically permissible in the target language.

– Bilinguals (observed so far) do not produce wh-movement – Bilinguals (observed so far) do not produce wh-movement questions in wh-in-situ languages because these are syntactically impermissible in the target language.

• Some predicted overlaps with accounts that appeal to surface overlap (cf. Hulk & Müller, 2000).– But the present account relies on an overlap at the level of

grammatical features, rather than a surface overlap between the two languages.

-36-

Why?

• We have defined possible ‘transfer/CLI’

contexts (in terms of what is syntactically

permissible in the target language).

• But transfer/CLI does not arise everywhere it • But transfer/CLI does not arise everywhere it

is possible.

• Why does it occur when it does?

-37-

Select

C[weak-Q] C[strong-Q]� ��

-38-

wh-in-situ questions

LANGUAGE A(e.g., Indonesian, Korean, Cantonese)

wh-movement questions

wh-in-situ questions

LANGUAGE B (e.g., Italian, English, BP, ASL, Libras)

Source of the developmental delay

• The option available to both target languages (C[weak-Q]) is arguably Selected more often; the bilingual child with a less discriminating Select might tend to rely more heavily on it.– processing demands– processing demands

– inability to inhibit one of the two options

• If the target language (syntactically) permits both options, the bilingual child might not be as discerning until she masters the discourse-pragmatic constraints on the use of the in-situ option.

-39-

Conclusions• “Cross-linguistic influence” in the case of wh-questions

refers to increased rates of a particular wh-structure compared to monolingual rates – crucially that wh-structure is found in the target language, i.e. it is syntactically permissible.

• Bilingual children produce wh-questions that are syntactically permissible in the target language.

• Bilingual children produce wh-questions that are syntactically permissible in the target language.

• Non-target-like productions by bilinguals still fall within the realm of syntactically permissible structures. – But these structures may not be pragmatically appropriate.

• The source of the developmental delay (e.g., weaker Select operation, discourse-pragmatic delay, etc.) remains to be precisely identified and experimentally tested.

-40-

References• Antelmi. D. 1997. La prima grammatica dell’italiano. Bologna: Il Mulino.

• Cheng, Lisa L.-S. 1991. On the typology of wh-questions. MIT: PhD dissertation. Published in 1997, by Garland Publishers.

• Chomsky, Noam. 1995. The Minimalist Program. Cambridge, Mass.: MIT Press.

• Cole, Peter and Gabriella Hermon. 1994. “Is there LF wh-movement?” Linguistic Inquiry 25(2):239-262.

• Cole, Peter, David Gil, Gabriella Hermon, and Uri Tadmor. 2000a. “The acquisition of WH forms in Jakarta Indonesian.” Handout from the 7th Annual Meeting of the AustronesianFormal Linguistics Association, 13 May 2000.

• Cole, Peter, David Gil, Gabriella Hermon, and Uri Tadmor. 2000b. “The acquisition of in-situ WH-questions and WH-indefinites in Jakarta Indonesian.” Handout from the 25th

Boston University Conference on Language Development, 3 November 2000.

• Huang, James C.T. 1982. “Move WH in a language without WH movement.” Linguistic Review 1:41-80.

• Hulk, Aafke and Natascha Müller. 2000. “Bilingual first language acquisition at the interface between syntax and pragmatics.” Bilingualism: Language and Cognition 3(3):227-244.

• Lillo-Martin, Diane. 2000. “Early and late language acquisition: Aspects of the syntax and acquisition of WH questions in American Sign Language.” In K. Emmorey & H. Lane, eds., The Signs of Language Revisited: An Anthology to Honor Ursula Bellugi and Edward Klima, pp. 401-414. Mahwah, NJ: Lawrence Erlbaum Associates.

-41-

• Lillo-Martin, Diane and Ronice Quadros. 2006. “The position of early WH-elements in American Sign Language and Brazilian Sign Language.” In Proceedings of the Inaugural Conference on Generative Approaches to Language Acquisition in North America, Honolulu, HI, 195-203.

• Nunes, Jairo and Ronice Quadros. 2005. “Duplication of wh-elements in Brazilian Sign Language.” In Proceedings of the 35th Conference of the North Eastern Linguistic Society.

• Park, Sung K. 2008. The acquisition of wh-questions by Korean-English bilingual children: The role of crosslinguistic influence. Masters thesis, Purdue University.

• Petronio, Karen and Diane Lillo-Martin. 1997. “Wh-movement and the position of spec CP: Evidence from American Sign Language.” Language 73:18-57.

• Pires, Acrisio and Heather Taylor. 2007. “The syntax of wh-in-situ and Common Ground.” Romance Languages: Structure, Interfaces, and Microparametric Variation: Proceedings of the 37th Linguistic Symposium on Romance Languages. Amsterdam: John Benjamins.Romance Languages: Structure, Interfaces, and Microparametric Variation: Proceedings of the 37th Linguistic Symposium on Romance Languages. Amsterdam: John Benjamins.

• Soriente, Antonia. 2007. “Cross-linguistic and cognitive aspects in the acquisition of WH-questions in an Italian-Indonesian bilingual child.” In Kecskes, I. and L. Albertazzi, eds., Cognitive Aspects of Bilingualism, 325-362.

• Tieu, Lyn Shan. 2010. “Transfer vs. code-switching in bilingual children’s acquisition of wh-questions.” Handout from the Annual Meeting of the Canadian Linguistic Association.

• Yip, Virginia and Stephen Matthews. 2000. “Syntactic transfer in a Cantonese-English bilingual child.” Bilingualism: Language and Cognition 3(3):193-208.

• Yip, Virginia and Stephen Matthews. 2007. The Bilingual Child: Early development and language contact. Cambridge University Press.

• Zavitnevich-Beaulac, Olga. 2005. “On wh-movement and the nature of wh-phrases – Case re-examined.” SKASE Journal of Theoretical Linguistics 2(3).

-42-

Cross-linguistic influence at the level of syntactic ... · • The wh -questions produced by bilingual children are syntactically permissible in the target language. • Bilinguals

Documents