BAR ILAN UNIVERSITY RAMAT GAN ISRAEL Verb inflections

BAR ILAN UNIVERSITY RAMAT GAN ISRAEL Verb inflections as indicators of Bilingual SLI Sharon Armon-Lotem, The Bilingual SLI project* Bar Ilan University CLS, July 17 -19, Reading *This project is funded by ISF grant no. 938

Acknowledgement This work has been done in collaboration with: Anat Blass, Jonathan Fine, Efrat Harel, Elinor Saiegh-Haddad, and Joel Walters, Bar-Ilan University Galit Adam, The Open University With the help of: Dori Braude, Michal Giladi, Ruti Litt, Lyle Lustinger, and Efrat Shimon

The bilingual SLI Project - Aim n n n Examining the linguistic production of bilingual children, ages 4 -7, who were diagnosed for SLI in order to assess the relative contribution of each to the child’s linguistic representations and underlying processes. In the overall effort, we look at the interface of SLI and bilingualism, exploring primarily the use of morpho-syntax, pragmatics, and discourse, as well as lexical, phonological, and sub-lexical processing. The present paper focuses on the use of the inflectional verbal system by English -Hebrew bilingual children

Definitions n n Specific Language Impairment (SLI) Children with normal performance IQ, who score 12 months/1 SD below chronological age on standardized language tests, and have no: hearing disabilities, emotional or behavior problems, observed neurological deficit, or severe articulation/phonological deficit. Bilingual children Children with bilingual background who are able to function in two languages (carry a conversation and understand) at a near native level (typical or impaired). This includes both simultaneous bilinguals and sequential bilinguals.

Subject selection n n Preschool children from bilingual or monolingual Englishspeaking homes, who attend regular preschools and special “language preschools”, and have been exposed to Hebrew for at least two years. All children come from the same neighbourhood and same (middle-high) SES Children are screened for both languages and are categorized in accordance with their linguistic abilities as diagnosed by standardized tests (e. g. , CELF Preschool for English, Goralnik for Hebrew), where TD is measured by less than 1. 5 SD below norm. This yields a division into children with typical development in both languages (TD), children with English typical development (E-TD), and children with English atypical development (E-ATD). This later group comprises of children with Hebrew typical development (H-TD), and children with atypical development in both languages (A-TD) – all are considered at-risk for SLI.

TD-children: 6 case studies § 6 case studies, 3 simultaneous, 3 sequential § 3 boys, 3 girls § Age range 5; 5 -6; 5

E-TD children – 5 case studies n n n 5 case studies 3 boys, 2 girls Age range 4; 1 -6; 6

At risk children: 6 case studies n n n 6 case studies, 3 from each sub-group 3 boys, 3 girls Age range 5; 5 -6; 9 H-TD A-TD

Linguistic Measures: Inflections n English: past tense n 3 rd person in the present n n Hebrew: gender and number in present tense n gender, number and person in past and future n

Inflections in Monolingual SLI and TD Bilinguals n n English: Both SLI and bilinguals use root infinitives (RIs), e. g. , David play ball. Hebrew: SLI children find past tense 2 nd person inflection more difficult.

Tasks n n Naturalistic samples (interview, story telling, free play Sentence completion (Based on Dromi et al 1999) Enactment (Based on Dromi et al 1999) Elicited imitation

TD – verb inflections § No errors in Hebrew enactment § In other tasks, error rate is very low, mostly less than 10% and never more then 20% (on the sentence completion task).

At Risk – Verb Inflections

Types of errors - English n n V-ing: The cat hops and the dog hopping Wrong tense: Here the boy jumps and the girl jumped Wrong 3 rd person: The cat hops and the dogs hops Root infinitives: The cat hops and the dog hop

Types of Errors - Hebrew n n n Root infinitives Wrong tense – past for present or vice versa Wrong gender – masculine for feminine Wrong number – singular for plural Wrong person – 1 st person for 2 nd person and or versa

TD - Sentence Completion Task: Frequency of Error Types 5/96 wrong tense, 12/96 person and tense omission (Root Infinitives), (13%) wrong 3 rd person with plural subject 2/252 infinitive, 4/252 wrong tense, 9/256 wrong gender (all present, feminine, plural [N=24])

TD -Imitation Task: Frequency of Errors in Different Linguistic Contexts 5/72 in 3 rd person, 4/72 in past tense. All errors reflected use of root infinitives 1/48 in 1 st person, 6/48 in 2 nd masculine, 5/48 in 2 nd feminine and 12/48 in 2 nd plural. All errors reflected use of 1 st for 2 nd and vice versa

TD - Summary n No errors on Hebrew enactment, up to 20% errors on the sentence completion task, and up to 10% on other tasks § English errors are mostly Root Infinitives (13% of relevant contexts) and wrong 3 rd person with plural subjects (13% of plural subjects) n Hebrew errors are mostly wrong gender in [present, feminine, plural] forms (9 of 24 – 37%) and in [past second person] forms (23 of 144 – 15%).

E-TD and TD - Sentence Completion Task: Frequency of Error Types

TD and E-TD -Imitation Task: Frequency of Errors in Different Linguistic Contexts

E-TD Enactment - Hebrew § E-TD have many errors in person inflection, using 3 rd person

E-TD - Summary n English n n n Sentence completion – All but one child show TD error frequency Imitation – TD error frequency with RIs among the younger children Hebrew n n n Sentence completion – the 3 older children show TD error frequency, the 2 younger ones are a little worse than the TD child with shorter exposure, reflection usage of 3 rd person bare forms Imitation – TD error frequency, with a lot of 3 rd person bare forms rather than 1 st/2 nd person alternations Enactment – A very high rate of errors using 3 rd person bare forms rather than 1 st/2 nd person alternations

At Risk - Sentence completion § H-TD – 60% RIs § A-TD – 50% RIs, 30% wrong 3 rd person § TD error frequency

At Risk - Imitation n § H-TD – 60% RIs § A-TD – 60% RIs n n 2 nd to 1 st person errors A few Sg > Pl and Pl > Sg) H-TD show 40% error rate, A-TD show up to 100% error rate

E-TD and At Risk - Enactment (Hebrew) [N=8] § H-TD and A-TD show near TD profile § E-TD have many errors in person inflection, using 3 rd person

At Risk, E-TD and TD Errors - English n Root infinitives: TD & E-TD: Up to 20% of 3 rd person and past contexts. The younger E-TD have a higher ratio of RIs. H-TD & A-TD: Root Infinitives in 50 -60% of 3 rd person and past contexts n Erroneous tense and erroneous 3 rd person mostly with plural subjects in all groups

At Risk, E-TD and TD errors - Hebrew n n n Sentence completion – At risk, older E-TD and TD have around 10% errors, showing the same variety of errors. The increase ratio of errors among the young E-TD reflects their use of 3 rd person bare forms. Second person triggers substitution E-TD opt for 3 rd person bare forms H-TD and A-TD opt for 1 st person Higher error rate in enactment for E-TD group (60%) Higher error rate in imitation for At-Risk Groups (up to 70%)

Conclusions n n n Studying the inflectional system of 17 English. Hebrew bilinguals, ages 4 -7, we found that: In English, TD and E-TD bilinguals tend to use root infinitive in up to 20% of the relevant contexts. By contrast, At-risk, (like young E-TD) children showed the same kind of errors in 50 -60% of the relevant context. In Hebrew, the TD bilinguals used the wrong person inflection in 15. 5% of the contexts which triggered verbs inflected for 1 st and 2 nd person. By contrast, E-TD children opt for the bare form. At-risk children showed the same kind of error in 50 -60% of the relevant context.

Inflections as indicators for SLI in Bilingual population n The same kind of error was found in both TD and at-risk children, but the quantity was different. n Is the high ratio of root infinitives indicative of SLI in the H-TD and A-TD groups? n Does it mean that the E-TD group is not SLI? n Are difficulties with 2 nd person indicative of SLI in the ETD group?

n n E-TD children are not SLI, but rather slow second language learners, who have not mastered the inflectional system of their L 2 For the At-Risk children, though tense-marking may not be a qualitative clinical indicator of SLI in bilingual populations, the quantity of errors, when manifested in both languages, might be a potential indicator.

Thank you תודה