Bilingual first language acquisition and the mechanisms of substrate influence

1
Bilingual first language acquisition and the
mechanismsof substrate influence

• Stephen Matthews,
• University of Hong Kong
• Virginia Yip,
• Chinese University of Hong Kong

2
1. Theoretical background

• Linguists who study language contact often
  seek to describe changes at the level of
  linguistics systems in isolation and abstraction
  from speakers. Sometimes they tend to treat the
  outcome of bilingual interaction in static rather
  than dynamic terms, and lose sight of the fact
  that the bilingual individual is the ultimate
  locus of language contact (Romaine 1996 573,
  our emphasis)

3
1.1 Mechanisms of contact-induced language change
(Thomason 2001)

• 1. Code-switching
• 2. Code alternation
• 3. Passive familiarity
• 4. Negotiation (approximation)
• 5. Second language acquisition strategies
  (interference/transfer)
• 6. Bilingual first language acquisition
• 7. Deliberate decision (language
  planning/engineering)
• Synergy in situations of widespread
• bi-/multilingualism, any combination of these
  factors
• could be operating.

4
Thomason's example of bilingual first language
acquisition

• In bilingual acquisition of French and German,
  frequencies of some French word orders are
  affected, relative to monolingual children
• Also in French/Dutch (Hulk van der Linden 1996)
  
• (1) live lire (Anouk, 20520)
• book read
• (2) Anouk riz manger (Anouk, 20520)
• Anouk rice eat-INFIN

5

The Hong Kong Bilingual Child Language
Corpus ?????????

• subjects exposed to Cantonese and English from
  birth in one parent one language families
• longitudinal data for Timmy (15-36), Sophie
  (106-30), Kathryn (36-46) and Llywelyn
  (20-304)
• Total of 191 tagged files in 2 languages
• special features digital audio and video demo
  files

6
Corpus information

• project homepage
• http//www.cuhk.edu.hk/ils/home/
• bilingual.htm
• corpus available at CHILDES (Child Language Data
  Exchange System ) archive http//childes.psy.cmu.
  edu

7
Subject Information
8
1.2 Mechanisms of creole formation

• 1.2.1 Theories invoking child first language
  acquisition
• The Language Bioprogram Hypothesis (Bickerton
  1981, 1984)
• Little discussion of the role of bilingual
  children

9
1.2.2 Theories invoking adultsecond language
acquisition

• Relexification substratal features are
  transferred into the creole by means of
  relexification... this mental process applies in
  a situation which involves second language
  acquisition (Lefebvre 1998 34-5)
• the creators of a creole, adult native speakers
  of substratum languages, use the properties of
  their native lexicons, the parametric values and
  the semantic interpretation rules of their native
  grammars in creating the creole. (Lefebvre 2001
  186-7, emphasis added)

10
Theories invoking adult second language
acquisition

• Contact ecology attestations of transfers from
  substrate languages in several creoles are among
  convincing evidence against the central role of
  children in their development. (Mufwene 2001
  131)
• Assumption of these approaches adults, not
  children are agents of substrate influence

11
A role for bilingual children?

• Additional possibilities transfer in bilingual
  first language acquisition (BFLA) and/or child
  second language acquisition (child SLA)
• Transfer is relatively well understood in adult
  SLA, still poorly understood in BFLA (e.g.
  directionality of transfer, determining factors)
• The epistemological relationship and demarcation
  between BFLA and child SLA remain unclear (Yip,
  to appear)

12
2.1 Factors determining transfer

• Language dominance development of Language A is
  ahead of that of Language B. Can be measured by
  Mean Length of Utterance (MLU).
• e.g. wh-in situ interrogatives develop early in
  Cantonese and are transferred to English in
  Cantonese-dominant children (Yip Matthews
  2000)
• (3) You go to the what? (Timmy, 205)
• (4) Daddy, you writing what? (Sophie, 35)

13
Dominance of Cantonese reflected in Timmys
MLUw between age 201 208.
14
(No Transcript)
15
2.1.2 Input ambiguity

• sentences in the adult input to children may
  allow more than one analysis, one based on
  language A and one based on language B
• (5) Input sentence I want to eat
• English (target) analysis I want to eat
  (something)
• Chinese-based analysis TOPICi I want to eat
  xi
• ( I
  want to eat this/that)
• -gt This ambiguity allows the child to hypothesize
  that English allows null objects, like Cantonese

16
2.Bilingual first language acquisition

• both languages acquired simultaneously from
  birth (De Houwer 1990, 1995)
• Possible courses of development (both attested)
• (i) Separate development (De Houwer 1990)
• (ii) Interactive development (Döpke 2000,
• Yip Matthews 2000, among others)

17
Null objects in bilingual data

• (6) You get, I eat (Timmy, 20203)
• father takes chocolates off shelf
• (7) Daddy Timmy, do you want the rest of this?
• Timmy I dont want. (20707)
• (8) Dont break! cautioning the adult not to
• break a toy cup (Sophie, 30606)

18
(No Transcript)
19
2.2 Resolution of transfer-based structures

• It is not the case that errors or innovations
  in a childs grammar survive into adulthood.
  Instead, childrens errors which presumably
  manifest a grammar (or lexicon) different from
  that of their parents tend to disappear in a
  later phase of language acquisition (Croft
  2000 47)
• Do developmental errors/innovations go away? How?
  
• Is this the result of normal acquisition
  processes,
• or of schooling and literacy?
• What happens when a community of such
  bi/multi-lingual children develops?

20
Possible outcomes in the individual

• 1. Preemption e.g. wh-movement largely replaces
  wh-in situ between age 3-4
• 2. Persistence e.g. null objects persist to age
  7 and beyond (Yip Matthews 2000)
• (9) Alicia I want to put. bringing jar of face
  cream
• Sophie You want to put on your face? (61110)
  
• Alicia Yah. (20810)

21
Differential outcomes in the individual

• These differential outcomes (preemption vs.
  persistence) are determined in part by
  considerations of learnability,
• e.g. input ambiguity favours null objects (Yip
  Matthews 2003)

22
Possible outcomes in a speech community

• 1. Replacement transfer-based structure(s)
  ironed out by adult community
• 2. Persistence transfer-based structure persists
  as innovation in adult usage
• These differential outcomes must be determined in
  part by language-internal considerations such as
  learnability, and in part by external factors
  such as ecology (Mufwene 2001)

23
Transfer in bilingual development and substrate
influence

• 
  
• Interaction of English with southern Chinese
  dialects leads to similar results in Singapore
  Colloquial English (SCE) and in Hong Kong
  bilingual children

24
(No Transcript)
25
already marking perfective aspect

• (10) She wake already. (Sophie 206)
• (11) He go already he go already the monster
  (Sophie 21021)
• (12) Keoi seng-zo laa (adult Cantonese)
• S/he wake-PFV PRT
• Shes woken up

26
(No Transcript)
27
already marking perfective aspect in SCE (Bao
1995, 2002)

• (13) I wash my hand already
• I have washed/washed my hand
• (14) The tongue red already
• The tongue has turned/turned red/
• The tongue was red
• Despite the syntactic difference, V-le and S-le
  in Chinese and S already in Singapore English,
  the substrate source of already is unmistakable.
  (Bao 2002 9)

28
(No Transcript)
29
give as passive marker

• (15) Here is give Timmy scratch.
• (points to scratched leg) (Sophie, 306)
• (16) Daddy, I already give the mosquito to bite
  (shows bite on tummy) (Sophie, 309)

30
one as a nominalizer in relative clauses

• (20) Sophie I also want.
• Father What do you want?
• Sophie Timmy said that one. (30821)
• the child has been asking for a piggy-bank
• (21) Ngo jiu Timmy gong go go (adult Cantonese)
• I want Timmy talk that CL
• I want what Timmy was talking about

31
Relative clauses with one as head in adult SCE
(Alsagoff and Ho 1998)

• (22) They grow one very sweet.
• The fruit that they grow is very sweet.
• (23) Don't have car one, I don't want.
• I don't want a man who does not own a car.

32
give-passives in SCE (Bao Wee 1999)

• (17) John give his boss scold
• (18) John bei loubaan laau (adult Cantonese)
• John give boss scold
• John is scolded by his boss

33
4. Discussion

• SCE (almost) a creole? (Ho Platt 1993, Gupta
  1994, Bao 2002)
• - no longer an issue if creole is not seen as a
  unique structural type (cf. e.g. Corne 1999,
  DeGraff 2001, Mufwene 2001)

34
Ecology in which SCE developed (Gupta 1994 33)

• (i) English-medium schools
• (ii) Racially mixed districts (in which most of
  these schools were located) including Eurasians,
  Jews, Armenians and Straits Chinese
• Both these ecologies involve child bilingualism,
  hence possibility of transfer in BFLA/child SLA
  (ii) may resemble circumstances of creole
  formation, e.g. Baba Malay in Malacca (Ansaldo
  Matthews 1999).

35
5. Implications

• Bilingual first language acquisition is a
  possible route for substrate influence, both in
  general and specifically in creole formation.
• Parallel phenomena in bilingual development (HK)
  and in Singapore Colloquial English (SCE)
  illustrate this possibility.

36
Implications BFLA and/or SLA?

• These effects occur alongside classical second
  language acquisition, e.g. Chinese substrate
  features are incorporated into SCE in the course
  of both bilingual first language and child/adult
  second language acquisition.
• In principle such effects are only possible in
  cases where interactive development occurs the
  conditions for this to occur still need further
  investigation. Language dominance and input
  ambiguity both appear to be factors favoring
  interaction (Yip Matthews 2000).

37
Implications creole formation

• Balanced development with little or no
  interaction is also possible, at least in ideal
  circumstances where the input from both languages
  is both balanced (resulting in no clear pattern
  of language dominance) and separate (e.g. by
  domains of use). But such ideal situations are
  most unlikely to prevail in complex contact
  ecologies such as those of creole formation.

38
Acknowledgements

• Hong Kong Research Grants Council (RGC),
• ref. nos. HKU336/94H,CUHK 4002/97H, CUHK4014/02H
  and CUHK Direct Grant 01/02
• Special thanks to members of our research team,
  especially Huang Yue-Yuan, Peng Ling-Ling, Bella
  Leung, Simon Huang Pai-Yuan, Gene Chu, Chen Ee
  San, Michelle Li and Uta Lam for their dedication
  and commitment at various stages of the project.