2 Cross-linguistic evidence for UH Grammaticality and - PowerPoint PPT Presentation

Cristóbal Lozano: SLADG presentation, University of Essex, 15 th March 2001 2 Cross-linguistic evidence for UH Grammaticality and [-interpretable] features in Spanish SLA: ! Unaccusatives manifest in different ways in different languages (but out the unergative/unaccusative distinction of scope of this study): ! SV/VS orders (Spanish, Italian, Romanian, partially in English) 1 Assumptions ! Unaccusatives select auxiliary ‘be’ (rather than ‘have’) in German, Italian, French, or even Middle English : Intransitives Unergative Unaccusative I have swam I have arrived * I am swam I am arrived (a) Unergatives (b) Unaccusatives 3 Word order: [top]/ [foc] and verb type TP TP ! Neutral focus : DP T’ (1) Unaccusatives . Context: Your friend Manuel and you are at a party at pro i T’ his place. While Manuel is in the kitchen getting a beer, a neighbour you T VP T VP don’t t know comes in to complain about the loud music. When Manuel Spec V’ comes back from the kitchen, he asks you: ‘What happened?’ You answer: Spec VP a. Spanish: Vino un vecino (VS) V DP i b. English: A neighbour came (SV) V (2) Unergatives . Context: You are at a restaurant with your friend Maria. External argument Maria goes to the toilet and in the meanwhile you can see a woman shouting No external argument [AGENT] in the street. When Sole comes back, she asks you: ‘What happened in the street?’ You answer: a. Spanish: Una mujer gritó (SV) b. English: A woman shouted (SV) ! Focused subject : Internal argument [THEME] (3) Unergatives. Context: Last night you went to the disco with your friends. It was boring because only a girl danced. Today, your mum phones ! Unaccusative Hypothesis (UH) is universal, Burzio (1986) ! two different you and asks: Who danced last night? You answer: types of intransitive verbs: unergatives and unaccusatives : a. Spanish: Bailó una chica (VS) ! Unergatives : subject base-generates in subject position (SV) b. English: A girl danced (SV) Laura estornudó (4) Unaccusatives . Context: You are at a party with your friend Maria. ‘Laura sneezed’ While Maria is in the toilet, a man you don’t know arrives. When Maria comes ! Unaccusatives : subject base-generates in object position (VS) . back from the toilet, she realises there’s somebody else but doesn’t know Unaccusatives normally express movement ( arrive, come, go, etc). who, so she asks you: ‘Who arrived?’ You answer: Llegó Laura a. Spanish: Llegó un hombre (VS) ‘(There) arrived Laura’ b. English: A man arrived (SV) Doc: Handouts for SLADG presentation unaccusatives Page 1 of 5

! Unaccusatives: ! [± interpretable] features and word order: TP ! Neutral contexts ! [-interpretable] due to universal lexical features of unaccs (VS) vs unergs (SV) Spec T’ ! Focused Subject contexts ! [+interpretable] due to T VP movement to satisfy principle [TOP] [FOC] in Spanish (Zubizarreta, 1998) ! therefore VS order with both verbs. Spec V’ ! Unergatives : pro i XP V DP llegó j TP X TP DP T’ DP T’ t j Laura i T VP T VP estornudó j Spec V’ Spec V Laura Laura (a) What happened? VS Or (b) Who arrived? (VS) t j estornudó j 4 Aims of our study t j t j ! Original aim: to test whether learners of Spanish behave differently with [+interpretable] and [-interpretable] verbal (and also pronominal) features: (a) What happened? (SV) (b) Who sneezed? (VS) ! [+interpretable] more prone to L1 transfer: learners would prefer SV order with both unacc/unerg (instead of expected VS order). ! [-interpretable] less prone to L1 transfer as they are universal: learners would prefer in neutral contexts: SV with unergatives and VS with unaccusatives. ! These expectations originated from project on attrition (Tsimpli, Sorace & Heycock) ! Provisional hypothesis (for today) : ! (H1) if learners show knowledge of [-interpretable] features, then abstract verbal features located in lexicon are accessible after puberty (Full Access). ! (H0) if learners do not show knowledge of [-interpretable] features, then abstract verbal features located in lexicon are not accessible after puberty. ! I get a somewhat mixed result (not very good result) Doc: Handouts for SLADG presentation unaccusatives Page 2 of 5

! Design : 4 conditions: 5 Previous SLA studies ! No studies on the nature of [±interpretable] and verb type. However: [-interpretable]: [+interpretable]: ! Hertel (2000) English learners of Spanish prefer/produce VS with neutral “what happened?” Foc S “who arrived?” unaccusatives more than with other types of verbs (a) Susana estornudó (a) ! Susana estornudó Unerg ! De Miguel (1993) English native learners of Spanish produce and accept (b) ! Estornudó Susana (b) Estornudó Susana VS order more with unaccusatives than with any other verb type. ! Hertel & Pérez-Leroux (1999) English native learners of Spanish accept (a) ! Susana llegó (a) ! Susana llegó VS order more with unaccusatives than with unergatives. Unacc ! Hirakawa (1999) English native learners of Japanese are sensitive to the (b) Llegó Susana (b) Llegó Susana different interpretation of grammatical subjects with unaccusatives vs. unergatives. ! Balcom (1997) Chinese native learners of English preferred passive ! ! ! ! 6 items for each condition (after ‘Darwin’ test): morphology more with unaccusatives than with any other verb type. Unergatives Unaccusatives ! Zobl (1989) learners of English inverted SV order to VS order with bailar ‘to dance’ entrar ‘to come in’ unaccusatives (independent of whether this was possible in their L1). gritar ‘to shout ’ llegar ‘to arrive’ dormir ‘to sleep’ salir ‘to leave’ 6 Subjects in our study reir ‘to laugh’ venir ‘to come’ llorar ‘to cry’ volver ‘to return’ ! Experimental: L1 English learners of L2 Spanish at the Uni. of Essex estornudar ‘sneeze’ escapar ‘to escape’ ! Advanced level (University of Wisconsin PT) ! 6 [-interpretable] and 6 [+interpretable] unacussative conditions ! n=16 ! 6 [-interpretable] and 6 [-interpretable] unergative conditions ! Control group : Spanish natives (n=15) ! All items are randomised (‘ blocking’ procedure, Cowart, 1997) ! There are 2 versions for each task, so as to minimise order-of-presentation Method 7 effects. ! Paired GJT : Example: You are at a party with your friend Laura. Laura leaves the room and at that 8 Data analysis moment the police arrive because the party is too noisy. When Laura comes ! Data were coded in MS Excel (v. 97) and analysed with SPSS (v.9.0). back, she asks you: ¿Quién llegó? You answer: ! Tests I haven’t performed yet: non-parametric tests to check sample’s (a) La policía llegó. –2 –1 0 + 1 + 2 normality of distribution . (b) Llegó la policía. –2 –1 0 + 1 + 2 ! Within-group analysis (for both word orders in a group): paired samples t- test ! Between-group analysis (for each word order in a condition): GLM univariate with posthocs (in my date I’ve got 3 groups: Spanish, English and Greek, though Greeks won’t be analysed here) Doc: H Handouts f s for S SLADG p pres esen entation un on unac accu cusatives es Page 3 Pa ge 3 of 5 of 5

9 Results ! Unergs [-interpretable] ! ! ! ! ! ! ! Unaccs [-interpretable] 2.0 2.0 1.5 1.5 1.0 1.0 .5 .5 0.0 0.0 -.5 -.5 -1.0 -1.0 -1.5 Unerg Neutral: !VS -1.5 Unac Neutral: VS Mean Mean -2.0 Unerg Neutral: SV -2.0 Unac Neutral: !SV Spanish English Spanish English Group Group ! Spanish: !VS different from SV (p<0.01) ! Spanish: VS different from !SV (p=0.01) ! English: VS not different from !SV (p=0.15) ! English: VS not different from !SV (p=0.34) ! HOWEVER , let’s focus on grammatical/ungrammatical ! HOWEVER , let’s focus on grammatical/ungrammatical ! !VS different between Spanish-English (p=0.01) ! VS not different between Spanish-English (p=0.84) ! SV different between Spanish-English (p<0.03) but not too significant ! !SV different between Spanish-English (p=0.03) [not excessively] ! CONCLUSION : ! CONCLUSION : ! Spanish natives show knowledge of [-interp] features since they ! Spanish natives show knowledge of [-interp] features since they distinguish gramm./ungramm. word order distinguish gramm./ungramm. word order ! English natives don’t show complete knowledge of [-interp], although ! English natives don’t show complete knowledge of [-interp], although they show preference like Spanish natives. they show preference like Spanish natives. ! Grammatical: both groups show just about same knowledge. ! Grammatical: both groups show same knowledge ! Ungrammatical: Spanish group according to hypothesis, English group ! Ungrammatical: Spanish group according to hypothesis, English group indeterminate intuition indeterminate intuition ! Feature transfer? Difficult to decide in this case since SV is grammatical. ! Feature transfer? If so, we’d expect English group to prefer !SV order to ! Interesting: English natives do prefer !VS with unergatives VS order, which is not exactly the case. Doc: Handouts for SLADG presentation unaccusatives Page 4 of 5