María J. Arche , Laura Domínguez & Florence Myles

1. The L2 Acquisition of the semantics and morphology of Aspect: a study of the acquisition of the Spanish imperfect-perfective contrast by native speakers of English María J. Arche, Laura Domínguez & Florence Myles

2. This talk • To investigate the L2 acquisition of Grammatical Aspect by looking into the acquisition of the semantics of the Spanish imperfect by native speakers of English. • To evaluate whether the predictions of the Aspect Hypothesis (Andersen 1991, Andersen & Shirai 1996) can be extended to the acquisition of meaning. • To examine the path of emergence of the semantics of the imperfect. • To present new Spanish L2 cross-sectional comprehension data.

3. Grammatical Aspect • Semantic category represented in the Syntax. • Conveys information about • whether eventualities are in progress, finished or about to start. • the number of occasions an eventuality takes place.

4. Grammatical Aspect • Comrie 1976, Smith 1991, Verkuyl 1993, Demirdache & Uribe-Etxebarria 2000, Arche 2006.

5. Syntax of Aspect (Arche 2006) Progressive Habitual AspP AspP TT Asp’ Asp’ TT Ordering Predicate Asp Q<occ> Q<occ> Asp within within >1 |1| VP e Continuous AspP Perfective AspP TT TT Asp’ Asp’ Asp Asp Q<occ> Q<occ> after within Quantifier Occasions ∃ |1|

6. Grammatical Aspect and Inner Aspect • Inner Aspect/Situation Aspect: internal temporal structure of eventualities (e.g. duration, culmination, or delimitation). • Vendler 1967, Verkuyl 1993, Smith 1991, a.o.

7. Aspect Hypothesis(Andersen 1991; Andersen &Shirai, 1996; Bardovi-Harlig, 2002; Sugaya&Shirai, 2007): 1. Learners first use (perfective) past marking with achievement and accomplishment verbs, eventually extending use to activity and state verbs. 2. In languages that encode the perfective-imperfective distinction morphologically, imperfective past appears later than perfective past, and imperfective past marking begins with states and activities (i.e., atelic verbs), then extends to accomplishments and achievements (i.e., telic verbs. 3. In languages that have progressive aspect, use of progressive marking begins with activity verbs and then extends to accomplishment and achievement verbs. 4. Learners do not incorrectly attach progressive marking to stative verbs.

8. Research Questions • Which of the three meanings associated with the Spanish imperfect is acquired first? • Can the acquisition of the meanings of the Spanish imperfect be explained by the LAH? • What is the acquisition task regarding Spanish aspect for native speakers of English?

9. Grammatical Aspect • Understood as an ordering predicate (Demirdache & Uribe-Etxebarría 1999 - 2007). • Tense orders the Topic Time with respect to the Speech Time (before, after, within). • Aspect orders Topic Time with respect to the Event time. • Aspect is syntactically located at a lower position than Tense.

10. Aspect in English and Spanish (Summary) • Both languages have the same syntax and semantics for Aspect. • Spanish & English temporal inflection includes tense and aspect information.

11. The learning task • We assume all the relevant semantic features can be transferred from L1. • The learning task is defined as the mapping of the semantic features onto L2 morphology. • In learners’ L1 there is a one to one form to meaning correspondence for imperfect meanings, while in the L2 one single form can correspond to three different meanings. • L1 has ambiguous forms (past forms with states). • John was ill when I visited him (imperfect) • John was ill the whole winter of 2001 (perfective)

12. Predictions • If semantics to morphology mapping is the issue, states will be the most difficult case where the correct meaning to form mapping can be established. • If emergence and development of imperfect is guided by lexical aspect properties of the verb (at least in the path suggested by AH), imperfect should be acquired with states early on.

13. The experiment

14. Participants

15. Comprehension task • Sentence-context matching task • 32 sentences

16. Comprehension Task Learners were given the prompt in English Five-point Likert scale Test measures both acceptance of the correct form and rejection of the incorrect one

17. Overall Results

18. Results • Beginners do not show acquisition of imperfect. • Rates of correct acceptance of the imperfect and correct rejection of the preterit were significantly lower in continuous contexts for the intermediate and advanced groups. 2. No statistical differences in acceptance of the imperfect and rejection of the preterit were found according to type of predicates (stative or eventive).

19. Continuous • Intermediate learners had lower scores than advanced learners but not significantly lower. • There is no significant difference between the mean in the continuous and the mean in the progressive (eventive) for any of the learner groups.

20. Habitual •  According to the paired t-test, there is no significant difference between the mean in the habitual eventive and the mean in the habitual stative tasks for any of the learner groups. • This means that none of the groups’ results were influenced by the aspectual properties of the verbs. (Being eventive or stative did not affect the participants’ choices)

21. Discussion and conclusions • Acquisition looks gradual and attainable (advanced group behaved mostly native-like in some scenarios). • Beginners don't seem to distinguish between the meanings. • Intermediate and advanced learners do distinguish between the meanings of the imperfect and are better with some of them than with others. Not all of the meanings of the imperfect are equally problematic. • The habitual meaning seems to be the earliest and best acquired.

22. Continuous meaning (available only with states) is the one where learners perform the worst. • Difficulty cannot be explained by lexical aspect properties of the predicate, but by the need of establishing a new semantic-morphology mapping with no morphological equivalent in L1. 6. Event type does not have an impact on the correct acceptance of imperfect and rejection of the preterit. • The semantics of the imperfect is not first acquired with states. Lexical properties do not seem to be at the root of learner’s choices.

23. Acknowledgments The SPLLOC 2 project is supported by ESRC research grant (RES-062-23-1075) We would like to thank all the participants in the project, including subjects, transcribers and fieldworkers www.splloc.soton.ac.uk

24. Spare slides

25. Semantics of Tense & Aspect • Interval-based semantics • Temporal categories ORDER one interval with respect to another (Reichenbach 1947, Lemmon 1967, Zagona 1990, Stowell 1993). • Relevant intervals: • Speech time: the reference time with respect to which every event is ultimately understood. • Event time: the interval in which the event takes place • Topic Time: the interval that is referred to in the sentence (Partee 1973, Stowell 1993, Klein 1994). • Maria was cuddling the baby {when John entered the room} TT.

26. Syntax of Tense & Aspect • Before: future • After: past • Within: present Before/after/within • Before: prospective • After: perfective • Within: imperfect (progressive/habitual/continuous) Before/after/within Stowell 1993 - 2007; Demirdache&Uribe-Etxebarría 1998 - 2007.

27. The quantifier over occasions (Verkuyl 1993) enables us to account for the differences found within the imperfect. Progressive, habitual and continuous share the ordering property (within) but differ in the amount of instances of the event they refer to (one, more than one or, possibly none, ).

28. Illustrative examples (2) In his twenties, John used to go by bus to uni. ------[x------x--------x-------x---------x------UTT (3) When I called him, John was going to uni ---------[-----x------------------UTT (6) In his twenties, John was grey-haired. ------------[---------------------UTT (7) John was sick the whole of 2002 ------------[--------]---------UTT Red font text: Topic Time