Tense Aspect in Early Stages of Child Language Acquisition

2. Tense - aspect in early stages of child L2 acquisition Suzanne Schlyter Lund University Sweden Eurosla 18 Aix-en-Provence sept 2008

3. Child second language acquisition - chL2 Child second language acquisition, chL2 : start 3 8 years Is chL2 more like L1 or more like adL2 acquisition? Role of the Age of Onset of the Acquisition? Unsworth (2005), Meisel (2006, 2008), Bonnesen (2008), Thoma (2008) etc

4. Structure of the speech Study 1) chL2 - L1 - 2L1 of same age: various phenomena Study 2) chL2 - 2L1 - adL2 in initial stages: Tense-Aspect

5. Study 1, GSK07 Granfeldt, Schlyter & Kihlstedt 2007: French as cL2, 2L1 and L1 in pre-school children PERLES 24, SOL, Lund university

6. Background Open question whether chL2 proceeds like 2L1 (same age, same input, but AOA birth) ( Unsworth 2005; Kroffke & Rothweiler 2004; Meisel 2007, 2008; Bonnesen 2008; Tracy, Gawlizek & Thoma) Meisel (2006-7): adL2 fundamentally different from L1 - Where is the cut off point, age 3-4-5-6?

7. Study 1 (GSK07) Research questions, chL2 acquisition vs L1 and 2L1 , in children of same age, same input? If so, in what phenomena? Studied here: in French L1, 2L1, chL2 6 ys Finite and non-finite forms Marking of past tense reference Object clitics Gender

8. Study 1: General summary of Granfeldt, Schlyter & Kihlstedt 2007 French in child L2 patterns with adult L2 , but differs from 2L1 of the same age. These phenomena are: Non-finite verb forms Past tense marking Object clitics Gender

9. STUDY 2 Tense and Aspect development in adL2, 2L1 and cL2 Use of French Tense-Aspect marking in initial stages of acquisition, Swedish-French adL2 (19-50 ys) 2L1 children (1;10 4 ys) chL2 children (6 ys)

10. Differences TA in L1 adL2 Past reference (2)L1 : start with morphology; very correct development: all past contexts are marked; adverbs appear later; discourse very late adL2 : start with discourse (PNO etc.); temporal adverbs; past contexts often not marked by verb morphology (Meisel 1985, 1987; Rieckborn 2007) Weist (2002): For untutored L2 learners, the inflectional morphology is the last thing to be acquired in stark contrast to first language learning children. Hence, from this global perspective, L1 and L2 acquisition are as different as they can get.

11. AH the Aspect-before-Tense Hypothesis Ayoun & Salaberry 2008:556 The AH states that, in the early stages of acquisition, verbal morphology encodes inherent semantic aspectual distinctions, i.e. does not encode tense or grammatical aspect (). () the initial stages of development of tense and aspect marking are constrained by lexical aspectual classes: states, activities, accomplishments and achievements () AH is considered similar in L1 and adL2 acquisition (e.g. Ayoun 2008)

12. Study 2: Comparisons 2L1, adL2 and chL2 Matching point: the first moment from which the learners clearly refer to past using at least some past tense forms (PC or Impf). (cf. Rieckborn 2007) Children chL2 Viola, Patrick, Hannes, Valentina (6 ys) are compared to: 2L1 Jean, Dany, Anne, Mimi (2 ys) adL2 Henry, Bjrn, Sara, Martin (>20 ys)

13. adL2 - Adult second language learners

14. adL2 forms for past reference, from first occ of marked past ref Adult L2 learners: HENRY, BJRN, MARTIN, SARA Studied: Mean Length of Utterance (MLU) Rate of past tense marking Different verb forms used Lexical aspect (Telic, Activity, State) Temporal reference (PF = Perfect/ Imm past, vs Pa = Remote past)

15. Learner HENRY Exposure months MLU %past marking Henry1 2-3 3,3 43% Henry2 3-4 3,9 58% Learner BJRN Exposure months MLU %past marking Bjrn 1 3m 5,3 25% Bjrn 2 7m (7,0) 30%

16. Exemple adL2 production (5) Bjrn 1, MLU 5,3, exposure to French 3 months: *INT: tu peux raconter de de ce voyage? *BJ: voyage # eh nous eh fait le auto auto-stop. *BJ: et c'est - he he - plus difficile. *BJ: eh # eh nous commence en Porte de la Chapelle. *BJ: et prenE auto. c'est un petit eh auto eh avec un homme et une fille et ils eh # hm il fait (). *BJ: et eh eh airport. () *BJ: donc eh nous eh # allons , oui? () *BJ: nous allons aller # une le un xx. *BJ: c'est eh peut-tre eh # psh trente-cinq cinq kilomtres. () *BJ: un bon eh # place. *BJ: c'tait eh pas le bon place. *BJ: pour # c'tait # eh pas votre dir direction. *BJ: c'tait un autre direction.

17. adL2 learners verb forms for past reference Learner HENRY Exp mo mlu % past mk PC IMPF Long Form PRS sho rt default Henry1 2-3 3,3 43% 1T ai oubli PF 2T jai a bu Pa a comprendre Pa 1A ai contrl Pa 4S tait avais 1T ? compren- dE 2T entrE tlphon 2A tudiE fonctionnE all Pa 4S est jai Pa connat Pa 3A marche Pa parle Pa cherche Pa 1T? traduis Pa Henry2 3-4 3,9 58% 5 T jai oubli jai appris PF a vu Pa 2 A a travaill Pa 4S avais tait 1T oubliE Pa 4A - travaillE Pa mangE Pa dormi Pa 3S est

18. adL2 learners verb forms for past reference Learner BJRN Expo- sure mlu %pa mk PC Impf LongForms Present Bjrn 1 3m 5,3 25% (cf *allons allE T) 3S tait 1T prenE 2A tudiE jouE (+many LF for prs) 3S cest 2A fait 4T prend allons V allons Bjrn 2 7m (7,0) 30% 4T jai crit PF ai compos x PF chang trouv Pa 7A tude PF dorm dormE Pa e jou Pa essay couch Pa (aux: nous avons) 2S tait 1A jouait 2T allE demandE 6A jouE couchE continuE faire 1S avoir 14S 12cest avons 4A tude attende 5T prend allons

19. Result adL2 learners adL2 learners behave very similarly to what is shown in many studies of (French) L2 acquisition. The AH is not evident: Telic verbs in PRS and IMPF, Activity verbs in PC PC and IMPF is used from start to refer to remote past (i.e. Tense is marked early)

20. 2L1 children with two first languages, age 2 3 years (French stronger language)

21. 2L1 forms for past reference, from first occ of marked past ref 2L1 - two first languages = simultaneous bilingual children : ANNE, MIMI, JEAN, DANY Studied: Mean Length of Utterance (MLU) Rate of past marking Different verb forms used Lexical aspect (Telic, Activity, State) Temporal reference (PF = Perfect/Imm past, vs Pa = Remote past)

22. Exemples 2L1 production Dany 2;6 (MLU 1,8) Dany: (a) pati (kako) Pre: tu es parti lcole Watching ski competition on tv, Child sees someone falling: Dany: (a) tomb! Father: il est tomb Dany: est tomb Father: il est tomb

23. Child ANNE age MLU %past marking Anne3 2;8 2,7 83 Anne4 2;10 2,4 100 Anne5 2;11 3;2 91 Child MIMI Mimi2 age 2;2 MLU 3,2 % past marking 88 Mimi3 2;6 3,5 94 Mimi4 2;10 3,5 100

24. Child ANNE age mlu %past mark PC IMPF LongF PRS Anne3 2;8 2,7 83 7T+ cass PF tomb PF 3T fait PF - - - Anne4 2;10 2,4 100 1T+ fini PF 2T donn PF coll PF 2A mang PF - - - Anne5 2;11 3;2 91 7T+ parti PF pass PF trouv PF 1S avait 1A il jouE - -

25. 2L1 forms for past reference, from first occ of PC with avoir Child MIMI Mimi2 age 2;2 mlu 3,2 % past marking 88 Pass Compos 3 T+ vu PF tomb PF trouv PF 4 T crit PF fait x PF Impf - LongForm 1T crit PF Present - Mimi3 2;6 3,5 94 3 T+ cass tomb PF 4 T mang x PF fait x PF renvers PF 5 AT fait x PF ? mang PF 1 A jou Pa ai fait Pa - - 1T +TombePF Mimi4 2;10 3,5 100 2T+ tomb PF 8T fait x PF effac mis PF enlv PF - - -

26. Results 2L1 children Behave very similarly to what is shown in many studies of (French) L1 or 2L1 acquisition (Antinucci & Miller, Meisel, Rieckborn, Weist ...) No overextension of default forms for past ref Evidence for AH : 1-1 relation Telic verbs and PC ( est cass, est tomb, a trouv, a fini etc.) PC is used initially only for Resulting state (=Perfect) or Immediate past

27. chL2 - Child L2 learners, AOA ca 6 years like (2)L1 or adL2 ?

28. Child L2 forms for past reference, from first occ of PC with avoir Children: VIOLA, PATRICK, VALENTINA, HANNES Studied: Mean Length of Utterance (MLU) Rate of past marking Different verb forms used Lexical aspect (Telic, Activity, State) Temporal reference (PF = Perfect/Imm past, vs Pa = Remote past)

29. Child VIO Age / expo MLU %past marking Viola1 6;11/ 7 mo 2,3 (66%) Viola2 7;7/ 12mo 3,6 50% Child PAT Age /expo MLU %past marking Patrik1 6;2/ 7 mo 3,5 44% Patrik2 6;11/ 12 mo 2,6 44%

30. Child L2 : forms for reference to past, from first occ of PC with avoir Child VIO age mlu %past mark PC IMPF LF PRS Viola1 6;11 2,3 (66%) 2 T e venu Pa - - 1A mange Viola2 7;7 3,6 50% 4 T e venu a all * a achet x Pa 3 A a regard a mang a dans Pa 2 S sappellaient (imit) Pa voulait 1A coutE Pa 6 S veut V sont 1A regarde 1T va vont V

31. Child L2 : forms for reference to past from first occ of PC with avoir Child PAT age Mlu % past mk PC IMPF LF PRS Patrik1 6;2 3,5 44% 4 T a vu a crit PF a revenu Pa - 2T venE Pa posE PF 3T coupe PF ouvre PF ferme PF Patrik2 6;11 2,6 44% 8T e all, a vu Pa e (de)venu Pa e arveill Pa 2A a rigol Pa a regard Pa 1S y a t Pa 1T trouvait Pa 2S trouv Adj voulu 2A dansE chantE 5T all venu araveill se couchE 2S il y a Pa sont Adj Ps 2T va Pa

32. Exemples cL2 Patrick1, exp 7 months: *INV: et qu+est+ce+qu ' il a fait avec la porte ? *CHI: il ouvre . *INV: et l qu+est+ce+qu ' il a fait avec la porte ? *CHI: euh il & freme [= ferme] . Viola 1, exp 7 months: *INV: qu+est+ce+que tu as fait au caf ? tu es alle boire des cafs ? *CHI: non . *INV: tu as mang du caf ? *CHI: non . *INV: qu+est+ce+que tu as fait ? *CHI: mange . *INV: tu as mang ? mhm d ' accord . Viola2, exp 13 mois: *INV: elles taient alles la plage aprs ? *CHI: oui . *INV: et qu+ est+ce+qu ' elles y avaient fait ? *CHI: elle [alt=elles] a regard de soleil .

33. Results chL2 children AH is not evident: Telic verbs in PRS and IMPF, Activity verbs in PC PC and IMPF used from start to refer to remote past (i.e Tense marked early) >> similar to adL2 learners

34. Discussion Why cL2 like adL2 rather than 2L1 ? Hypotheses: - Conceptual development (Weist 2002) - Syntactic development (White 2003, Giorgi & Pianesi 1997) GSK07: When the child has developed the entire syntax (DP > VP > IP > CP) then cL2 is like adL2

35. MERCI! THANK YOU!

36. Acknowledgements Jean-Luc Montois, Sylvie Renard The children and their parents Elisabeth Rausing Memorial Foundation for Research (grant nov 2007)

37. Some references Ayoun, D. & Salaberry, R. (2008). Acquisition of English Tense-Aspect Morphology by Advanced French Instructed Learners. Language learning 58:3, sept 2008 Meisel, J.M. (2008) Child second language acquisition or successive first language acquisition? In B. Haznedar & E. Gavruseva (eds.) Current Trends in Child Second Language Acquisition: A Generative Perspective . Amsterdam: John Benjamins. Rieckborn, S. (2008) Erst- und Zweitsprachenerwerb im Vergleich. Eine Studie zum Erwerb von Tempus und Aspekt im Deutschen und Franzsischen. PHILOLOGIA, Bd 99, Hamburg. Unsworth, S. (2005) Child L2, Adult L2, Child L1: Differences and Similarities. A Study on the Acquisition of Direct Object S crambling in Dutch . Utrecht: LOT. Doctoral dissertation. Weist, R. (2002) The first language acquisition of tense and aspect: A review. Salaberry, R.& Shirai,Y. (eds) The L2 acquisition of Tense-Aspect Morphology . Benjamins