Passive' unaccusative errors in L2 English revisited

• Published in: INQUIRIES IN LINGUISTIC DEVELOPMENT: IN HONOR OF LYDIA WHITE, Slabakova, Roumyana, Montrul, Silvina A., & Prevost, Philippe [Eds], Amsterdam: John Benjamins, 2006, pp 17-39
• Main Author: Hirakawa, Makiko
• Format: Book Chapter
• Language: English
• Published: The Netherlands John Benjamins Publishing Company 2006
• Subjects: LANGUAGE & LINGUISTICS
• ISBN: 9789027232328; 9027232326
• DOI: 10.1075/z.133

Previous studies have reported that L2 learners of English with various L1 backgrounds produce/accept non-target forms in which unaccusative verbs appear in passives such as John was arrived late (Hirakawa 2000; Oshita 1997; Zobl 1989). It has been argued that the errors indicate learners observe unaccusativity, correctly generating the single argument of an unaccusative verb in object position but overgeneralizing passive morphology when promoting the argument to subject position. Yusa (2003) however challenges this NP-movement analysis & claims that the use of auxiliary 'be' by L2 learners is a reflection of the auxiliary 'be' selection with unaccusatives in perfective aspect & past tense observed in languages like Italian (e.g., 'Maria e arrivata' vs. 'Maria ha telefonato'). Including different types of unaccusatives & unergatives along the universal Auxiliary Selection Hierarchy (Sorace 2000), Yusa found that Japanese learners of English incorrectly accepted 'be' for core, telic unaccusatives (e.g., arrive) more than for peripheral, atelic unaccusatives (e.g., stay). He argued that learners selected auxiliary 'be' in accordance with telicity. if 'passive' unaccusative errors reflect perfective auxiliary selection, we should not expect 'be' errors in present tense. This issue is investigated in the present study, involving 25 Japanese learners of English who took a grammaticality judgment task. Overall results show that (1) there were generally no significant differences in learners' acceptance of 'passive' forms in the present vs. perfective/past tense, (2) no significant differences were found between telic vs. atelic unaccusatives, & that (3) learners were more accurate in rejecting 'passive' core unergatives than 'passive' atelic unaccusatives. We argue that the results are consistent with the NP-movement analysis. Tables, Figures, Appendixes, References. Adapted from the source document