Presentation on theme: "Introduction to second language acquisition and Specific Language Impairments in Children 37-975-01 Challenges to Language Acquisition: Bilingualism and."— Presentation transcript:
Introduction to second language acquisition and Specific Language Impairments in Children Challenges to Language Acquisition: Bilingualism and Language Impairment Dr. Sharon Armon-Lotem Bar Ilan University
1. Bilingualism and Second Language Acquisition Who is bilingual? How does one become bilingual? Is it a homogeneous group? Are these facts related? Why? How? Related issues How are bilinguals different from monolinguals?
Who is bilingual? A bilingual knows two languages A bilingual speaks two languages A bilingual is native or near-native in two languages Functional Bilingualism (Kohnert 2008)
How does one become bilingual? Immigrants Indigenous minorities Bidialectal populations Privileged populations (e.g. Anglophones in Canadian French immersion programs, Israelis who have returned from extended stays in North America) By parental choice
Is it a homogeneous group? Age of acquisition Birth order and family size Order of L1/L2 acquisition (simultaneous/ sequential) Acquisition context (e.g. one parent for each language/L1 at home and L2 at school).
Are these facts related? Why? How? Age – most L2 learners are older than L1 learners Degree of success attained by the learner Fossilization – L2 learners often get stuck at a point short of native-like grammar
Related issues Simultaneous vs. sequential bilingualism Critical period Transfer vs. access
How are bilinguals different from monolinguals?
Simultaneous bilingual (age 3;7( 1. *EFR:Do you want to read the Jungle Book? 2. *YAR:I can see Mowgli going. 3. *EFR:what can you see here? 4. *YAR:Bagheera take him to the animals. 5. *EFR:really? who are these? 6. *YAR:the wolfim. 7. *EFR:and here? 8. *YAR:I can see Baloo and the Mowgli. 9. *EFR:what are they doing? 10. *YAR:they throwing nuts. 11. *EFR:they throwing nuts. 12. *EFR:and now? 13. *YAR:Mowgli going quickly and Bagheera’s sleeping. 14. *EFR:oh. 15. *YAR:now Baloo want to eat the monkeys. 16. *EFR:and now? 17. *YAR:here Mowgli with Shere Khan. 18. *EFR:what happened to Mowgli? 19. *YAR:and he is doing fire to Shere Khan. 20. *EFR:Shere Khan is scared. 21. *YAR:why? 22. *EFR:he is afraid of Mowgli. 23. *YAR:yeah, from the esh.
L1 Hebrew, L2 English (age 5;1, LoE 6 mos) 1. INV: do you want to tell me what happened yesterday at school? 2. GAL: yeh. 3. INV: what happened? 4. GAL: we are do a project. 5. INV: you did a project, yes. 6. GAL: and we we play with play-doe in Miss Pam. 7. INV: oh, you played with play-doe? 8. GAL: yes. 9. GAL: and Miss Lilach come back. 10. INV: ah, Miss Lilach came back? 11. GAL: yeh. 12. GAL: and INV: did she play with you? 14. GAL: no. 15. INV: did she tell you a story? 16. GAL: no. 17. INV: what did she do? 18. GAL: they, they show a picture of his baby and give us stickers. 19. com: she her
L1 Russian, L2 Hebrew (Age 4;7) 1. LitalAni roca she taasi li meshulash. 2. Int 1Bevakasha. 3. LitalAval gam et ha-ceva shelo taasi. 4. Int 1Kaxa? 5. Lital Ve gam po ve gam po. Ribua. Kaze davar. ….. Kmo xalon ze. 6. Int 1Naxon. 7. LitalAz tavi li ani ecayer lax mashu. 8. Int 1Ma at taasi? 9. LitalAni ose im ze ceva kaxol i ceva yarok. 10. Int 1Boi ani aazor lax. 11. LitalOd lo asitnu et ha-ribua. 12. Int 1Az ma ze? ze ribua? 13. LitalAval ze ktana ribua. 14. Int 1Az eze at roca? 15. LitalGdola gdola
2. Specific Language Impairment Language impairment with no hearing loss (no history of otitis media), no emotional and behavioral problems, no below average non- verbal IQ(>=85), no neurological problems, and no oral or facial defects (Tallal & Stark 1981). A developmental language disorder characterized by Gleason (2001, p. 504) as involving ‘delayed or deviant language development in a child who exhibits no cognitive, neurological or social impairment’ (Radford 2006).
Sample narrative (MoSLI) אמא הכינה לילדים שלה אוכל ואכלו ואכלו אח"כ בא לו זבוב. אח"כ הוא כעס אח"כ שמו לה בייגלה בזנב אח"כ שמו לה בשערות משהו חם אח"כ ניקו אותה וזהו. Mom prepared food for her children and pro ate.pl and pro ate.pl Then, came a fly. Then, he was angry Then, pro put.pl a pretzel on her tail. Then, pro put.pl something hot in her hair Then, pro cleaned.pl her and that’s it
Major Issues Frequency of SLI Genetic basis of SLI Neurological basis of SLI Overall characteristics of SLI
Sentences produced by children with SLI (Radford 2006) Sentences produced by the SLI children in the Leonard files on the childes data-base.
Expressive vs. receptive deficit SLI children typically show some (or all) of the following types of impairment: Phonological (e.g. problems with consonant clusters and syllable-final consonants) Lexical (delayed acquisition of words – e.g. first word appears around 23 months in SLI children, but around 11 months in TD children; SLI children also have word-finding problems) Semantic (problems in determining the linguistic meaning of words, phrases and sentences, and understanding the meaning of metaphors) Grammatical (e.g. problems with affixes/inflections and articles/particles, complex syntax) Pragmatic (e.g. problems in the use of language in appropriate contexts) Reading problems
Delay versus Deviance Delay: Protracted acquisition of language, following typical developmental pattern. Deviance: Different developmental sequences and processes. Delay Plateau Profile differences Abnormal frequency of errors Qualitative difference
“Two of a Kind” ? Some parallels are found between the language of sequential bilingual children and the language of children with SLI – e.g., both use bare verbs (*He go). Paradis & Crago while children with SLI tend to omit the auxiliary in past or future periphrastic verb constructions, L2 children substitute the auxiliary with the base or present tense form. Paradis (2008) - only L2 children generalize the use of BE, in order to fill a gap between their communicative demands and their knowledge of the L2 with a morphosyntactic expression. Both the high proportions of substitution errors and the overgeneralization of BE single out L2 children with TLD from children with SLI.
3. Bilingual SLI Bedore, L.M. & Pena, E.D. (2008). Assessment of Bilingual Children for Identification of Language Impairment: Current Findings and Implications for Practice. International Journal of Bilingual Education & Bilingualism, 11,1, 1-29.
L1 Russian, L2 Hebrew (4;7) 1. Int 1Ma ata mecayer 2. Child Lo yodeya. 3. Int 1Ata roce be-acmexa o she ani azor? 4. Child Izo. 5. Int 1Izor? Keilu ani aazor? 6. Child Ken. 7. Int 1Aval tesaper li ma ata roe ba sefer tov? 8. Child Ze. 9. Int 1Mi ze? 10. Child Yosenet. 11. Com %In the pictures he sees a boy. 12. Int 1Ma? Yoshenet. 13. Child Osenet po. Oshenet po. 14. Int 1Ma maxzik ha-yeled? 15. ChildYeled mazik bi ze…… shinaim. 16. Int 1Ma? 17. Child Et ze. 18. Int 1Ve ma ze? 19. ChildXatul. 20. Int 1Ma hu ose sham? 21. ChildEe…. Pes kelev. 22. Int 1Ma hu ose? Ani roca ladaat ani lo yodaat,… Hu mecayer. 23. Child Ken.
24. Int 1Ve ma hu roe ba xalon. 25. Child Ba-xalon i bait. 26. Int 1Ba-xalon i ibait. 27. Child Ze kelev. 28. Int 1Ve ma ze? Bait shel kelev? 29. Child Ze.. Ine ze kelev bait. 30. Int 1Ve ze? 31. Child Ze…bait kelev. Olex kaxa. 32. Int 1Ze bait kelev olex kaxa. Ma od ata roe ma ze? 33. Child Be.. Ec. 34. Int 1Naxon ec, kol ha-kavod,….. Ata o(h)ev ecim yesh lexa ba-bait ec? 35. Child Lo. 36. Int 1Lama? 37. Child Li yesh ba-bait peax. 38. Int 1Ve mi metapel ba-perax? 39. Child Eee…ima. 40. Int 1Ma hi osa? 41. ChildHi osa.. Main mm peax. 42. Int 1Hi osa maim praex …tov.… Ve mi zot? 43. Child Lo yodeya. 44. Int 1Ulai zot maxshefa?
45. Child Aval ze uga yom -uledet. 46. Int 1Naxon. 47. Child Li yesh yom –uledet. 48. Int 1Matai yesh lexa yom –uledet. 49. Child Ein li …aya. 50. Int 1Aaa… aya lexa, evanti, ve ex aya? 51. Child Aya kef. 52. Int 1Ma ata omer?, ivi(h)u lexa uga? 53. Child Ken, ima. 54. Int 1Ima? 55. Child Ken. 56. Int 1Kan o ba-bait? 57. Child Ba-bait. 58. Int 1 Ma ata omer?….ve (h)ayu lexa orxim? 59. Child Ken orxim ve ba-bait. 60. Int 1Ve mi ba? 61. Child Saba vee lo yodeya…. Ima veeee.. Lo yodeya. 62. comChild sees a car in the picture 63. Child Oto!!! 64. Int 1Oto! Ata o(h)ev oto? 65. Child Ma ze ? Ma ose? 66. Int 1Ma hu ose? Hu xoshev, hu xoshev ma lecayer. Naxon she ata ciyarta dvarim? 67. Child Ken. 68. Child Ken… hu cayer.
Paradis (1999) and Crago & Paradis (2000) L1 and L2 French-speaking SLI children A range of measures related to the ‘optional infinitive’ phenomenon “Significant similarities” between SLI and L2 learners,: Tense marking Avoidance of object clitics Verb diversity Use of general purpose verbs (e.g. do, make). tense-marking may not be an effective clinical indicator of SLI for second language learners.
Naming (Simonsen, 2002) 6 years old Swedish- Finish BL Naming task (Renfrew Word Finding Vocabulary test) Scores Total naming time ControlsSLI BL ML ControlsSLI 06:0206:26BL 04:5304:11ML
Findings MSLI- phonological naming problems more often than the other groups (can explain their fast naming speed) Substitution of phonemes: MSLI > BSLI The bilingual children have difficulty in finding words BLC is slow in naming, does not find the target word as accurately as the MLC, but uses strategies that are pragmatically efficient: describes, chooses a Finnish word, or uses gestures.
The use of definite article (Scheaffer at al., 2003) In children with SLI (14 Subjects: 3;11-4;10), pragmatic principles develop normally as a function of age, rather than as a function of grammar developmental stage. Grammatically, 4-year old children with SLI make errors comparable to younger normally developing children.
Predictions Omission of the indefinite article in English will be found both for TD and for at-risk children, and cannot be considered indicative of SLI. Omission of definite articles in both languages will be found for at-risk children, and can be indicative of SLI.
Armon-Lotem, Danon & Walters (2006) 12 sequential bilinguals, ages 4-7 (mean age 5;9), from English- speaking homes exposed to L2 Hebrew in Hebrew-speaking pre-school programs for over two years, who attend “language preschools”. 6 age matched TD bilinguals, from English-speaking homes, who attend regular preschools. 10 are younger TD bilinguals (mean age 4;1) from bilingual or monolingual non-Hebrew-speaking homes, who attend regular preschools. Most children come from the same neighbourhood and all from the same (middle-high) SES Children are screened for both languages and are classified as children with typical development (TD), Hebrew typical development (H-TD), English typical development (E-TD), atypical development (A-TD). H-TD, E-TD, and A-TD are considered to be At-Risk for SLI. This division reflects their linguistic abilities as diagnosed by standardized tests (e.g., CELF Preschool for English, Goralnik for Hebrew). TD is less than 1.5 SD below the norm.
Method Spontaneous data were collected and supplemented by an elicited production task (Thornton 1996, Scheaffer 1997 for definiteness). All data were collected in separate sessions for each language by native speakers. Findings were analyzed for each child separately, yielding individual language profiles, which, when combined, yield group profiles
Naturalistic Samples – Young TD Frequency of correct use, omission and commission (wrong article) of articles in Hebrew and English
Naturalistic sample – At Risk 12% 8% 20% Use, omission and commission (wrong article) of articles in Hebrew and English (raw numbers)
Spontaneous sample - Summary Both TD and At-Risk bilingual children omit the indefinite article in English. This error reflects the typological difference between Hebrew and English, and the process of feature reassembly. Only the At-Risk children drop the definite article in obligatory contexts in both languages: 8% in English naturalistic 12% in Hebrew naturalistic data. This error resembles Scheaffer at al’s (2003) for English monolingual SLI children
The projects Challenges to Language Acquisition: Bilingualism and Language Impairment Dr. Sharon Armon-Lotem Bar Ilan University
Syntax and morphology Syntax and morphology are sources of linguistic indicators of SLI and a central focus of ongoing research on bilingual SLI (Armon-Lotem et al. 2008; Chilla & Barbur 2008; Jacobson & Schwartz 2002; de Jong et al. 2007; Marinis 2007; Papadopoulou 2009; Roeper 2004; Rothweiler et al. 2007). This project will target morphosyntactic and syntactic phenomena in both languages that have been shown to be vulnerable in monolingual children with SLI, e.g. verbal inflections (third person s He walks), auxiliaries (such as in He is walking), plural marking on nouns and adjectives (such as the suffix im in yeladim ktanim ‘little children’ in Hebrew or ‘the little children’ in English, determiners (such as the in The boy walked), prepositions (such as at in He laughed at the girl, or on in He turned on the light), and case marking (such as Russsian On dal mashinu mal'chiku ‘He gave car to boy’). Omission and/or substitution of such morphemes is often taken to be an indicator of SLI, but in bilingual contexts such errors could reflect L2 characteristics and/or crosslinguistic influence. For example, Russian does not have definite articles, and Russian-Hebrew bilingual children often omit the definite article in Hebrew. In terms of syntax, we include sentences with non-canonical word- order, e.g. passives (The elephant was pushed by the giraffe), wh-questions (Who did the elephant push), and relative clauses (The elephant who the giraffe pushed ran away), but leave room for other phenomena, based on the typology of the language pairs. In this context, a distinction between code interference errors and errors which cannot be traced to code interference has been shown to be an effective marker for distinguishing TDL and SLI bilingualism (Armon-Lotem & Walters 2008, 2009).
Sentence Repetition (SR) Sentence Repetition (SR), effective in distinguishing typically-developing children from children with language impairment (Conti- Ramsden et al. 2001) will be used to study syntax and its interface with morphology by identifying structures which are less sensitive to crosslinguistic differences, and are vulnerable for monolingual and bilingual children with SLI, but not for typically developing bilingual children.
Level 1 – 20 sentences 1. SVO with one auxiliary X2, SVO with one modal x2 1. The kitten is chasing the rat up and down. noun-noun, 9 words, 3 lexical, 6 functional, 11 syllables 2. They are eating the bananas in the park. pronoun – noun, 8 words, 3 lexical, 5 functional, 11 syllables 3. The boy must sweep the floor in the kitchen. noun – noun, 9 words, 4 lexical, 5 functional, 10 syllables 4. She can bring the glass to the table. pronoun - noun, 8 words, 3 lexical, 5 functional, 9 syllables
Level 2 – 20 sentences 3. Long actional and non-actional passives 4X 9. The sandwich was eaten by the postman. (actional, noun, 7 words, 3 lexical, 4 functional, 10 syllables) 10. He was kicked in the leg by the donkey. (actional, pronoun, 9 words, 3 lexical, 6 functional, 10 syllables) 11. The bear was feared by the boy in the park. (non-actional, 10 words, 4 lexical, 6 functional, 10 syllables) 12.She was seen by the doctor in the morning. (non-actional, pronoun, 9 words, 3 lexical, 6 functional, 11 syllables) 4. wh-object which question 2X and indirect object wh-questions 2x 13. Which drink did the milkman spill in the house? (noun, 9 words, 4 lexical, 5 functional, 10 syllables) 14. Which picture did he paint at home yesterday? (pronoun, 8 words, 3 lexical, 5 functional, 11 syllables) 15. Who did the father cook the meal for today? (noun, 9 words, 3 lexical, 6 functional, 11 syllables) 16. Who did she give the beautiful rose to? (pronoun, 8 words, 3 lexical, 5 functional, 10 syllables)
Level 3 – 20 sentences 2. SO relative clause – centre embedded 4x 5. The swan that the dear chased knocked over the plant. (noun – noun – noun, 10 words, 5 lexical, 5 functional, 11 syllables) 6. The horse that the farmer pushed kicked him in the back. (noun – noun - pronoun, 11 words, 5 lexical, 6 functional, 12 syllables) 7. The boy that the milkman helped has lost his way. (noun – noun - noun, 10 words, 5 lexical, 5 functional, 11 syllables) 8. The bee that the man swallowed had hurt him. (noun – noun – pronoun, 9 words, 4 lexical, 5 functional, 10 syllables) 3. Sentence Complex sentences with conditionals (2 simple, 2 complex) 9. The people will get a present if they clean the house. (11 words, 5 lexical, 6 functional, 13 syllables) 10. If the kids behave we will go in the garden. (10 words, 4 lexical, 6 functional, 12 syllables) 11. He wouldn’t have brought his friend if she was nasty. (10 words, 4 lexical, 6 functional, 12 syllables) 12. If she was poorly she would go to the nurse. (10 words, 4 lexical, 6 functional, 11 syllables)
NWR Phonological processing and auditory memory, claimed to be impaired in children with SLI (Graf-Estes et al. 2007), should be intact in bilingual children with TLD, offering a promising direction for disentangling the two. Previous research has shown that monolingual and bilingual children with SLI perform poorly on non- word repetition (NWR) tasks (Gathercole & Pickering 2000; Girbau & Schwartz 2007). This task requires children to repeat nonce words and primarily taps phonological memory, but can also address lexical processing when the words are designed to reflect syllable structure, stress patterns and phonotactic rules similar to words in the target language. The task has been claimed to relate to vocabulary development (Gathercole 2006, and possibly to the development of syntax too (Stokes et al. 2006). The project will use a non-word repetition (NWR) task, developed in collaboration with Chiat (Chiat 2006; Roy & Chiat 2004) which include both non-words and pseudo words to make it possible to tap both on auditory memory and linguistic knowledge in both languages. Words will vary in syllable length (from one to four syllables) to tap on short term memory and in terms of clusterhood (no cluster, initial cluster and medial cluster), since clusters often pose greater difficulty for children with SLI (Marshall et al. 2009).
The Russian-Hebrew task Version 1: 48 words were used in each language. Variables: Length (1-4 syllables) Clusterhood (with or without a cluster, initial cluster or medial cluster) Degree of similarity to the target language (target like/pseudo words vs. non-target like/non-words). 14 categories. Each category (with two exceptions) was represented by three items. Version 2: A subset of 16 items – 8 pseudo words, 8 non-words, 1-4 syllables, with and without a cluster.
Narrative and discourse abilities Narrative and discourse abilities pose difficulties for children with SLI, since the ability to construct a narrative relies on a range of linguistic skills, including lexical, grammatical and discourse abilities (e.g. temporality, cohesion, etc.). SLI children generally use fewer connectives, more lexical ties and more unclear reference, and find it difficult to gain entry to an existing dyadic interproject (Thompson, Craig & Washington 2004). Research on bilingual narrative skills is still limited (e.g. Pearson 2001, 2002; Fiestas & Peña 2004), and even more limited among BISLI children (Guttierez-Clellan et al. 2008; Uccelli & Paez 2007). Potentially diagnostic features are informed by Ravid and Berman (2006) and will be targeted from: 1. lexicon (lexical diversity, general purpose verbs; 2. morphosyntax, i.e. tense/aspect markers found in narrative discourse; 3. syntax, e.g. subordination and other means to distinguish main ideas from details; 4. narrative structure (e.g. story grammar categories, connectives, clause sequencing; 5. discourse features, e.g. information density, elaborations, topic maintenance, explicitness; 6. fluency features including repetitions, false starts, pauses, discourse markers; and 7. frequency, locus and directionality of codeswitching.
Story Retelling The project will make use of a bilingual retelling task in which a child is asked to retell stories in L1-L2, L2-L1 and codeswitched conditions to listeners with different language preferences (Raichlin & Walters 2007; Iluz- Cohen 2008). This task has been found effective in distinguishing between bilinguals with TLD and those with SLI (Iluz-Cohen 2008; Iluz-Cohen et al. 2009). Retelling permits examination of the range of features mentioned above, assessment of language dominance and codeswitching patterns based on: frequency, locus and directionality (L1->L2 vs. L2->L1) and pragmatic differences in codeswitching related to story content and listener's preferred language.
The Story retelling Task (Stein & Glen, 1979) Three conditions (Walters & Reichlin 2005): 1. Story told in Hebrew (L2) was retold in English (L1) to an English monolingual puppet. 2. Story told in English (L1) was retold in Hebrew (L2) to a Hebrew monolingual puppet. 3. Story told in a switched manner was retold to a bilingual puppet. Language of retelling depended on the child's choice. Child is recorded and transcribed.
Home settingSchool settingNeutral setting
Executive Functions Executive functions, among other cognitive skills, seem to offer a promising direction for disentangling bilingualism and SLI. Monolingual children with SLI perform worse than typically developing children on tasks tapping executive functions (e.g., Montgomery 2002; Baddeley & Hitch 1974; Baddeley 2007), and this suggests that they have a deficit in memory-related executive functions. On the other hand, recent research on adult bilinguals has demonstrated enhanced abilities in executive functions tapping inhibition and shifting (Bialystok & Martin 2004), which relate to monitoring two languages at the same time and being able to switch between the two languages. The question is whether bilingual children with SLI will be able to take advantage of the better performance found with adult bilinguals or whether their language impairment will have a weakening effect on their performance in executive function tasks.
This project will target executive functions in bilingual children with SLI in language and non-language oriented tasks. Cognitive (non-linguistic) tasks include the Embedded Figures Task (Piaget & Inhelder, 1971; Pascual-Leone, 1989) which tests inhibition, and classification tasks adapted to test shifting in bilingual preschool children (Smidts et al. 2004). Tasks of this type are also found in standardized tests such as the Cambridge Neuropsychological Test Automated Battery (CANTAB) and the Wisconsin Card Sort test. Impairment in executive function could influence language abilities which have direct manifestations in bilinguals. A bilingual verbal fluency task (Luk & Bialystok, 2008) taps language control abilities, and will also serve as a measure of proficiency in both languages. A bilingual picture naming task (Hernandez et al., 2001; Festman et al., in press; Biran & Friedmann, 2005) will be used to test language control, that is, interference of the non- target language.
Cognitive executive control Inhibition The Embedded Figures Task (based on Piaget & Inhelder, 1966; Pascual-Leone, 1989; De-Avila & Ducan, 1980). Ten pictures were presented, each includes an embedded mouse, which the child was asked to spot as fast as possible. Pictures were presented in a gradually increasing level of difficulty. Time was measured, and errors, failures and successful turns were noted. Degree of inhibition ability: the number of correct answers, ranging from 0 to 10.
Where is the mouse?
Shifting The Classification Task (based on Ben-Zeev, 1977) 18 cards were presented. 3 different shapes (circle, triangle, square) 3 different patterns (no color, partial-color, full-color). 1 vs. 3 items of each shape.
Procedure & Analysis Child is requires to classify them, and reclassify them in a different way, and in a third way. Scoring for each classification: Immediate success: 3 First clue: 2 Second clue: 1 No success: 0 Time is measured, and successful, unsuccessful, and aided trials were noted.
Language Control The Verbal Fluency Task (Luk & Bialystok, 2007; Festman et al., in press). Subjects produce as many words as they could of one semantic category within one minute, twice in either of the languages. English – food and clothing. Hebrew – animals and body parts. Instructions are given in alternating languages Degree of language control: the ratio between the number of CSs and the number of words.
2. The Bilingual Picture Naming Task (Hernandez et al., 2001, Festman et al., in press; Biran & Friedmann, 2005). Name pictures of objects in English (L1; 20), and Hebrew (L2; 20). High frequency words were chosen excluding cognates and compounds. Order of presentation: two Hebrew words (blue background), and two English words (yellow background).
Errors are retested in the end to check familiarity with the items. Child is recorded and transcribed.