Cross-Linguistic Influence On L2 Grammar

jhh19970724

TheFacilitativeRoleofL1InfluenceinTenseAs.pdf

Home >English homework help >Cross-Linguistic Influence On L2 Grammar

The Facilitative Role of L1 Influence in Tense–Aspect Marking: A Comparison of Hispanophone and Anglophone Learners of French JESÚS IZQUIERDO1

Universidad Juárez Autónoma de Tabasco Licenciatura en Idiomas División Académica de Educación y Artes Francisco Sarabia # 3 Jalpa de Méndez, Tabasco C.P. 86200 México Email: [email protected]

LAURA COLLINS Concordia University TESL Centre, Department of Education 1455 de Maisonneuve Blvd. West Montreal, Quebec, H3G 1M8 Canada Email: [email protected]

English learners of French whose first language (L1) does not mark the perfective/imperfective distinction have shown verb semantic influence and an overall preference for perfective over imperfective in their use of second language (L2) tense–aspect markers. This study investi- gated whether learners whose L1 marks the perfective/imperfective distinction would exhibit similar acquisition profiles. Hispanophones (n = 17) and Anglophones (n = 15) at simi- lar levels of French L2 proficiency completed a 68-item cloze task with equal numbers of perfective and imperfective contexts distributed across 4 semantic categories: stative, activity, accomplishment, and achievements. In a 20-minute retrospective interview, a subsample of the participants (8 Hispanophones, 11 Anglophones) commented on factors influencing their tense–aspect choices. An ANOVA of 1,012 predicates revealed that unlike the Anglophones, the Hispanophones did not prefer perfective over imperfective, and they were also less influenced by verb semantics. The learners’ comments suggest that the Hispanophones made effective use of L1–L2 similarities, whereas the Anglophones appealed to verb semantics and partially understood pedagogical rules, which were frequently associated with inappropriate uses of the forms.

English speaker: We don’t make such a distinc- tion in English, so I didn’t know whether, here, in French one would say il y avait [imperfec- tive] or il y a eu [perfective].

Spanish speaker: Here, I used il y a eu [perfective], because, here, in Spanish we say

1Correspondence to the first author should be mailed to: Jesús Izquierdo, Francisco Sarabia # 3, Jalpa de Méndez, Tabasco, C.P. 86200, México.

. . . hubo [perfective]; we would not say . . . habı́a [imperfective].

These comments were made by two adult sec- ond language (L2) learners of French of similar overall proficiency who had been asked to com- ment on the tense–aspect forms they had pro- duced during a controlled production task. The passé composé (perfective) and imparfait (imper- fective) distinction addressed in the comments is a well-documented challenge for learners of French, but the Spanish speaker (i.e., His- panophone) reports that recourse to his first lan- guage (L1), in which the perfective/imperfective distinction also exists, influenced his choice of what was ultimately the correct tense–aspect form for the context. The English speaker (i.e.,

浦原喜助�

Jesús Izquierdo and Laura Collins 351

Anglophone), however, reports that the absence of such a distinction in his L1 made it difficult to choose between the two forms.

Studies of tense–aspect acquisition have pro- vided cross-linguistic evidence in support of com- mon patterns of development among L2 learn- ers, with respect to the influence of verb se- mantics and narrative grounding on the use of temporal morphology. There is also evi- dence that these influences can be mediated by knowledge of previously learned languages (Salaberry, 2005). To date, most research has focussed on the negative effects of L1 influ- ence (Collins, 2002, 2004; Housen, 2000; Kihl- stedt, 2002). The degree to which L1–L2 simi- larities may also facilitate acquisition—potentially providing a rate or route advantage, or both— is not well documented, however. This study was designed to address this issue, through an investigation of the acquisition of the perfec- tive/imperfective in French by Hispanophones and Anglophones.

BACKGROUND

The Acquisition of L2 Temporal Morphology in French

Learners of French need to acquire different sets of morphological markers to contrast not only tense but also grammatical aspect. The for- mer points to the time when an action occurs (e.g., present, past, future), whereas the latter renders the manner in which this action takes place (e.g., habitual, progressive, completed; see Comrie, 1976, 1985; Klein, 1994). Both sentences in Example 1 refer to the past action of plant- ing strawberries, but they differ in the perspective taken on the past event.

EXAMPLE 1 French Perfective and Imperfective Marking

1(a) Pendant l’été, j’ai planté des fraises ‘During the summer, I planted [perfective] strawberries.’

1(b) Pendant l’été, je plantais des fraises ‘During the summer, I planted [imperfec- tive] strawberries.’

In French, the perfective is marked with the passé composé 1 and renders the completeness of a past event (Howard, 2002; Kihlstedt, 2002). In Example 1(a), the action of planting strawberries is described as an event that was fully achieved in the past. The imperfective is marked with the imparfait and may express habit, duration, itera- tion, or progression in the past tense2 and thus,

in Example 1(b), the use of the imperfective with the predicate presents the planting of strawber- ries as taking place either habitually or repeatedly during a past period of time.

Among L2-instructed learners of French, two consistent findings with respect to the acquisition of these past tense forms are (a) that they are acquired at different rates (with the perfective emerging before the imperfective) and (b) that they are distributed differently across verb types. The perfective emerges with dynamic predicates (e.g., tomber ‘to fall,’ nager un kilomètre ‘to swim a kilometre,’ manger ‘to eat’), and the imperfective with a limited set of statives, namely être ‘to be’ and avoir ‘to have’ in particular, along with aimer ‘to like,’ vouloir ‘to want,’ and pouvoir ‘to be able to’ (Ayoun, 2001, 2004; Bardovi-Harlig & Bergström, 1996; Harley, 1992; Harley & Swain, 1978; Kaplan, 1987; Salaberry, 1998).

Studies conducted in the content-based Cana- dian French immersion programmes have shown that young learners who achieved impressive lev- els of communicative competence in their L2 may still operate with a simplified past tense mark- ing system (Harley, 1992; Harley & Swain, 1978), which is characterised by use of the perfective with dynamic verbs, even in obligatory contexts for imperfective, and a more restricted use of im- perfective with statives. Adult learners in different types of French L2 programmes have also shown a preference for using the perfective with dynamic predicates that have an inherent end (e.g., telics, such as nager un kilomètre ‘to swim a kilometre’ and tomber ‘to fall’) and the imperfective with states and dynamic situations (activities) that do not have an inherent end (e.g., atelics, such as manger à la maison ‘to eat at home,’ regarder la télé ‘to watch TV’) (Ayoun, 2005; Bardovi-Harlig & Bergström, 1996; Clemance, 2004; Howard, 2001, 2002; Salaberry, 1998).

These patterns of past tense marking are con- sistent with the predictions of the aspect hypoth- esis (Andersen, 2002; Andersen & Shirai, 1994), according to which the acquisition of perfective and imperfective marking is influenced by se- mantic properties inherent in verbal predicates (i.e., lexical aspect). These semantic properties are dynamicity, telicity, and punctuality. Dynamic situations require energy; telic situations have an inherent endpoint; and punctual situations are perceived to occur instantaneously3 (Ayoun & Sal- aberry, 2005; Bardovi-Harlig, 2000; Salaberry & Shirai, 2002).

According to the aspect hypothesis, similarities between the semantic properties of verbs and the functions of tense–aspect markers lead learners to

浦原喜助�

352 The Modern Language Journal 92 (2008)

associate some verbs with specific temporal mor- phology. A verbal predicate such as fermer la porte ‘to close the door,’ which is telic and punctual, is a strong candidate for perfective marking but a very weak candidate for imperfective morphol- ogy because the action of closing the door needs to reach completion in order to occur. Perfective marking is used to indicate the completeness of an event and results in the generation of sentences such as j’ai fermé la porte ‘I closed [perfective] the door.’ However, learners also need to learn that a different form is required when telic events are ha- bitual, repeated, or incomplete: je fermais la porte ‘I closed [imperfective] the door.’

The aspect hypothesis makes three predictions about learner acquisition of the perfective and the imperfective: (a) Learners will acquire the perfec- tive before the imperfective; (b) they will prefer to use the perfective with telic verbs (i.e., verbs of achievement, e.g., tomber ‘to fall’; fermer ‘to close’; and verbs of accomplishment, e.g., nager un kilomètre ‘to swim a kilometre,’ aller à l’école ‘to go to school’), a preference that then spreads to atelic dynamic verbs (i.e., verbs of activity; marcher dans la forêt ‘to walk in the woods,’ manger à la maison ‘to eat at home’) and, finally, to atelic non- dynamic verbs (i.e., verbs of states: être ‘to be,’ aimer ‘to like’); and (c) then they will acquire the imperfective first with states and then activi- ties, accomplishments, and achievements. In this study, we examined the degree to which these pre- dictions held when the learners’ L1 marked the imperfective/perfective distinction.

L1 Influence and the Acquisition of Temporal Morphology

In all of the research previously reviewed, the participants were Anglophone learners of French. Unlike French, English does not grammaticalise the distinction between perfective and imper- fective aspects; most of the functional values of the French imperfective, along with that of the French perfective, can be rendered in English by the simple past, as in the sentence “During the summer, I planted strawberries.” In order to differ- entiate between a single instance or multiple oc- currences of the action in this sentence, one needs knowledge of the context, of adverbials, or of the use of the optional habitual aspect markers used to or would . In French, this difference is clearly established through temporal morphology (i.e., grammatical aspect), as illustrated in Example 1. The progressive value of the French imperfective is rendered in English by the past progressive, but there are times when even this function can be

expressed through the simple past as well. For ex- ample, the sentences “I hit my head as I fell down the stairs” and “I hit my head as I was falling down the stairs” can both express a progressive mean- ing. (For further discussion, see Ayoun, 2005; Cle- mance, 2004.) This similarity between the English past progressive and the progressive value of the French imperfective appears to mislead learners, causing them to restrict the use of the French im- perfective to its progressive use at the expense of its other values (Ayoun, 2004, 2005; Harley, 1992).

Research has shown that L1 influence can affect the acquisition of temporal morphology. Collins (2002) pointed out that among French learners of English, use of the present perfect (a com- pound construction similar in form but not in function to the French passé composé ) competed with use of the simple past when the learners worked with telic verbs, a tendency that was not ob- served among learners of mixed L1 backgrounds, virtually all of whom spoke languages in which the compound form does not exist (Bardovi-Harlig & Reynolds, 1995). This finding was corroborated by a cross-sectional comparison of Japanese- and French-speaking learners’ use of the English sim- ple past. In this comparison, the Francophones’ inappropriate use of the present perfect in con- texts that were obligatory for the simple past re- sulted in a significant difference between the two L1 groups in the category of telic achievement verbs (Collins, 2004).

In a longitudinal study of the development of tense–aspect marking of a Dutch learner of En- glish, Housen (2000) concluded that L1–L2 dif- ferences may increase a learner’s reliance on the temporality inherent in the meaning of verbs. Dutch learners need to acquire the distinction between the simple past and the past progressive in English because this distinction is not marked in their L1. Housen found that his participant’s use of the past progressive and the simple past for regular verbs was in line with the inherent tem- porality of the verbs, but her use of the simple past for irregular verbs was not. Housen argued that this finding was the result of the learner’s re- liance on the temporality in the meaning of the verbs in cases where she needed to function on a rule-based system for tense–aspect marking, but not when she needed to work with irregular verbs that she had already acquired.

There is also evidence that L1 influence may continue to affect the use of the functional val- ues of temporal morphology among more pro- ficient L2 learners. Kihlstedt (2002) found that advanced Swedish learners of French, whose L1 does not mark grammatical aspect, differed from

浦原喜助�

Jesús Izquierdo and Laura Collins 353

French native speakers in the functional use of tense–aspect marking in nonprototypical contexts for tense–aspect forms, such as using the imparfait with telics and the passé composé with states.

In light of the impact that L1–L2 differences can have on the acquisition of L2 features, one question that arises is whether the learning of the perfective/imperfective distinction in an L2 is facilitated when the learner’s L1 already marks the forms. Spanish, like French, marks the perfec- tive and imperfective distinction, as illustrated in Example 2.

EXAMPLE 2 Spanish Perfective and Imperfective Marking

2(a) Durante el verano, sembré fresas. ‘During the summer, I planted [perfective] strawberries.’

2(b) Durante el verano, sembraba fresas. ‘During the summer, I planted [imperfec- tive] strawberries.’

The expression of termination or complete- ness of an event (the passé composé in French) is rendered in Spanish through the pretérito.4 The imperfective verbal morphology of French and Spanish also is similar in that it may express ha- bituality, duration, and frequency. In Spanish, the progressive value of the French imperfective can be rendered through the pasado progresivo (see De Lorenzo Roselló, 2001; Salaberry, 2000, 2005).

The facilitative role of L1 influence on the acquisition of L2 temporal morphology has not received much research attention, but some re- searchers have hypothesised a potential benefit to learners when there are L1–L2 or L2–L3 similar- ities in tense–aspect forms (De Lorenzo Roselló, 2001; Kihlstedt, 2002; Salaberry, 2005). This bene- fit would be consistent with evidence from studies on the acquisition of other features that indi- cate that the L1 can have a facilitative effect on the acquisition of an L2 (Kleinmann, 1977; Laufer & Eliasson, 1993; Sabourin, 2001; Zobl, 1980a, 1980b, 1982). Sabourin, for instance, in- vestigated the ability of learners from various L1 backgrounds to deal with gender agreement be- tween determiners and nouns in Dutch. The find- ings showed greater accuracy in gender marking among learners whose L1 was similar to Dutch in terms of noun gender and determiner–noun phrase gender agreement than among learners whose L1 was not similar to Dutch.

RESEARCH QUESTION AND HYPOTHESES

The research question addressed in this study was how the existence of the perfective/

imperfective aspectual distinction in the L1 would benefit the acquisition of these forms in an L2. Two hypotheses were entertained, one related to the distribution of perfective and imperfec- tive markers overall and one related to the use of tense–aspect markers in the more challenging nonprototypical contexts. Both hypotheses com- pared English-speaking learners of French with Spanish-speaking learners of French at the same level of L2 proficiency as follows.

Hypothesis 1: The Spanish-speaking learners of French will not show a preference for perfective over imperfective marking.

Hypothesis 2: The Spanish-speaking learners will be less influenced by lexical aspect in the use of perfective and imper- fective markers than the English- speaking learners.

METHODOLOGY

Participants

A total of 77 young adult low-intermediate learners of French volunteered for the present study. The Anglophones (n = 46) were enrolled in French classes at a university in Montreal and the Hispanophones (n = 31) were attending French classes at the university level in southeast Mex- ico. To control for potential differences in access to the target language (i.e., L2 context in French Canada, French foreign language context in Mex- ico) that could influence the performance of the learners in the testing tasks (Howard, 2001; Light- bown, 1985), 32 English L1 and 28 Spanish L1 participants were selected by matching linguistic and biographical profile data. This information was elicited through a questionnaire and informal interviews about access to the target language out- side the classroom. The participants were similar in age (ranging from 19 to 21 years), in their use of French outside the classroom,5 in the number of French study hours outside the classroom (be- tween 0 and 2 hours), and in the amount of time learning French prior to their current course (an average of 18 months). None of the participants reported having previously learned another Ro- mance language or having studied French in a French-dominant context.

Instrument and Procedures

In some studies of temporal morphology, the participants have been grouped according to overall knowledge of the tense–aspect marker

浦原喜助�

354 The Modern Language Journal 92 (2008)

under investigation, and the subsequent analyses focussed on how this knowledge was distributed across the variables of interest (i.e., verb types or narrative grounding; e.g., Bardovi-Harlig, 1998; Bardovi-Harlig & Bergström, 1996; Collins, 2002, 2004). In this study, because the research question focussed on the use of two tense–aspect markers and the hypotheses predicted greater appropriate use for one of two L1 groups with the forms, it was important to equate the groups on overall level of L2 proficiency.

Proficiency Test (Grouping Instrument). The pro- ficiency test was developed to place learners into one of six levels at one of the participating universities.6 For each level, there is a 10- minute interview followed by a 15-minute, 35- item multiple-choice test covering a range of forms (tenses, articles, gender agreement, etc.) in French. The participants in the present study completed the Level-III test, which is described by the institution as low intermediate. Only those participants whose scores fell within the institu- tional range for the low-intermediate level (60– 90%) were retained for the study. There were 17 Spanish L1 learners and 15 English L1 learners.7 A t -test found no significant differences between the Spanish L1 group (M = 76.6%, SD = 10.1) and the English L1 group (M = 76.6%, SD = 6.6), t (30) = .022, p = .98. Results from the oral interviews showed that all the participants were able to communicate using the past, present, and future tenses.

Tense–Aspect Test (Data Elicitation Instrument 1). Following the design of previous tense–aspect studies, a cloze passage test was developed (e.g., Bardovi-Harlig & Reynolds, 1995; Collins, 2002). Two sample passages appear in the Appendix. This type of elicitation instrument allows for con- trol over the numbers of verbal categories and obligatory contexts for the perfective and imper- fective aspects. In order to elicit spontaneous use of tense–aspect marking from the learners, re- sponse choices were not provided.

The test included an equal number of obliga- tory perfective and imperfective contexts (28 for each) and 18 distractors (contexts for other tense–aspect forms) distributed across 20 pas- sages. Each of the activity, accomplishment, and achievement verb types included five target verbs, which were required once in the perfective con- text and once in the imperfective context. There were six stative verbs (five statives were required twice and one stative was required three times in the perfective and imperfective contexts in the

passage). Because some scholars have hypothe- sised that the frequency of statives in the input may play a role in the use of temporal morphol- ogy (Ayoun, 2004; Harley, 1989; Harley & Swain, 1978), the selection of verbs other than être and avoir was based on their frequency of use in an online corpus, with a total of 1,110,392 words, taken from the French newspaper Le Monde (The Compleat Lexical Tutor, French section: http:// 132.208.224.131/concordancers/). Out of 15 sta- tives used in the corpus, the following statives with the highest frequency profiles were selected: penser ‘to think’ (269 imperfective cases in the corpus), sembler ‘to seem’ (43 imperfective cases), and appartenir ‘to belong’ and sentir ‘to feel’ (8 imperfective cases each). Our earliest test was first piloted with nine university-educated Franco- phones (aged 28–50 years) who also spoke English and Spanish. Items yielding less than 90% agree- ment on the required grammatical aspect (perfec- tive or imperfective) were excluded from the test. The new version was piloted a second time with a different group of five Francophones who were in 100% agreement for all target items.8 Following Collins’s (2002) cross-sectional study design, two counterbalanced versions of the revised test were created (the same items were presented in reverse order).

Cloze Passage Data Coding and Analyses. Fol- lowing the methods used in Bardovi-Harlig and Bergström’s (1996) study, each test was scored for the appropriate use of the perfective and the im- perfective. Instances where the learners used an inappropriate auxiliary to build the compound form of the perfective were considered appropri- ate attempts at the perfective (e.g., ∗il a retourné à la maison ‘he + incorrect auxiliary choice + returned to the house’ instead of the accurate form il est retourné à la maison ‘he + correct aux- iliary choice + returned to the house’). Similarly, errors in agreement between the imperfective in- flection and the head noun phrase also were ig- nored (e.g., ∗les édifices tombait ‘the buildings + fall with wrong imperfective inflection’ instead of les édifices tombaient ‘the buildings + fall with correct imperfective inflection’).

To ensure coding reliability, appropriate at- tempts at perfective and imperfective markers were coded twice by the same rater. In cases where differences between the first and the second cod- ing were found (10 items out of the 1,012 target items that were coded), a native speaker of French recoded the tests, and the differences were re- solved through discussion.

Jesús Izquierdo and Laura Collins 355

Reorganisation of the Target Items. The verbs ap- partenir ‘to belong’ and être ‘to be’ elicited a high number of nonresponses from both L1 groups. During the interviews, some Anglophones indi- cated that they did not understand the verb ap- partenir or its context; other Anglophones did not know which auxiliary to use, avoir or être . Two His- panophones did not use the verb appartenir in any of the contexts where it was required. Our Franco- phone reviewers indicated, during the interrating process, that in one perfective context for être , the sentence also would have been correct in the imperfective but would have rendered a different meaning. There was a second perfective context with être in which the Anglophones did not sup- ply temporal morphology. In this case, it was not possible to determine the reason. We therefore discarded appartenir from the statistical analyses and retained only two instances of être (once in an imperfective context and once in a perfec- tive context), reducing the target items in the test from 56 to 50. As a consequence, five stative verbs were considered for the statistical analyses, each of which had the same number of obligatory contexts for the perfective as for the imperfec- tive: être (one item each tense), avoir (two items each tense), penser (two items each tense), sembler (three items each tense), and sentir (two items each tense).

Data Collection Instrument 2 (Retrospective In- terview). Retrospective interviews are an impor- tant elicitation instrument in L2 research because they can contribute to our understanding of the processes that underpin L2 acquisition (Gass & Mackey, 2000). For example, interviews follow- ing an oral narrative task allowed Liskin-Gasparro (2000) to identify factors associated with tense– aspect choices among advanced Spanish L2 learn- ers. Collins (2005) showed that stimulated recall (where the participants were shown a selection of their responses on a timed computerized task) provided insights into the mental representations learners hold with respect to L2 tense–aspect marking and the input variables that potentially influence its development. In order to minimize validity problems inherent in self-report method- ology (notably the fading from memory of the thoughts that guided the original performance), it is important to conduct the interviews as soon as the task is completed (Ericsson & Simon, 1993) and to provide cues from the participants’ perfor- mance on the task (Egi, 2004; Taylor & Dionne, 2000). Accordingly, the subset of learners who agreed to be interviewed (8 Hispanophones and 11 Anglophones) for the present study met with

the first author for 20 minutes immediately after completing the cloze test. The interviews focussed on the learners’ choice of temporal morphology for the same 13 tokens, all of which required non- prototypical marking (see Appendix). To stimu- late their recall, the learners were shown the target passages they had responded to and were asked to comment on their thinking during the task. The interviews were conducted in the L1 of the learners.

RESULTS: CLOZE TEST

This section presents the results of the two-way mixed between–within subjects analyses of vari- ance (L1 group between; tense–aspect within) on both the overall and the appropriate use of tense–aspect markers in the cloze test. Overall use represents the percentage of the perfective or im- perfective across all tokens in the test, whereas appropriate use represents the percentage of the perfective and imperfective in obligatory con- texts.

Main and interaction effects are reported in line with the hypotheses of the study; when a significant interaction was found, simple main effects were tested using Bonferroni probabil- ity adjustments for multiple comparisons. Effect size is reported as Hedges’s d (Hedges & Olkin, 1985), calculated as the difference between the two group means divided by the pooled within- group variance of the two groups, whose result is then multiplied by an adjustment factor for small samples: 1 − (3/(4N − 9)).

Hypothesis 1: Rate of Perfective and Imperfective Marking

The statistical analyses of the overall use of tense–aspect markers in the cloze test yielded a sig- nificant interaction effect between L1 and gram- matical aspect, F (1, 30) = 4696.7, p = .01. Simple main effect contrasts within groups revealed a sig- nificant mean difference between the use of the perfective (M = 57.4%, SD = 6.49) and the imper- fective (M = 40.3%, SD = 6.9, p = .01, d = 2.53) within the Anglophone group, but not for the mean difference between the use of the perfec- tive (M = 47.7%, SD = 8.41) and the imperfective (M = 47.7%, SD = 6.9) within the Hispanophone group. Simple main effect contrasts for groups within each level of grammatical aspect revealed significant differences, in the use of both perfec- tive marking (Anglophone group: M = 57.4%, SD = 6.49; Hispanophone group: M = 47.7%, SD = 8.41, p = .01, d = −1.25) and imperfective

356 The Modern Language Journal 92 (2008)

marking (Anglophone group: M = 40.3%, SD = 6.9; Hispanophone group: M = 47.7%, SD = 10.1, p = .01, d = 0.82).

Analyses of the appropriate use of tense–aspect markers also showed a significant interaction ef- fect between L1 and grammatical aspect, F (1, 30) = 5013, p = .01. Simple main effects con- trasts for grammatical aspect within groups re- vealed a significant mean difference between the appropriate use of the perfective (M = 73.5%, SD = 12.52) and the imperfective (M = 56.5%, SD = 16.11, p = .01, d = 1.15) within the Anglo- phone group, but not for the difference between appropriate perfective marking (M = 72.8%, SD = 15.8) and imperfective marking (M = 73.5%, SD = 15.7) within the Hispanophone group. Contrasts for groups within each level of grammatical aspect showed no significant differ- ence in appropriate use of the perfective between groups (Anglophone group: M = 73.5%, SD = 12.52; Hispanophone group: M = 72.8%; SD = 15.8), but significant group differences in ap- propriate use of imperfective morphology (An- glophone group: M = 56.5%, SD = 16.11; His- panophone group: M = 73.5%; SD = 15.73, p = .01, d = 1.04).

TABLE 1 Percentage Distribution of Grammatical Aspect Markers With Verb Types

Grammatical Stative Activity Accomplishment Achievement L1 Aspect (SD) (SD) (SD) (SD)

Spanish Perfective 40.59% 44.71% 48.82% 56.47% n = 17 (14.98) (10.68) (10.54) (18.69)

Imperfective 54.12% 51.18% 47.65% 37.65 (18.40) (12.18) (11.47) (20.17)

English Perfective 44.33% 46.67% 58.67% 80.00 n = 15 (13.47) (12.91) (11.87) (14.64)

Imperfective 51.00% 50.00% 40.00% 20.00% (14.17) (13.09) (10.70) (14.64)

TABLE 2 Appropriate Use Percentage of Grammatical Aspect Markers With Verb Types

Grammatical Stative Activity Accomplishment Achievement L1 Aspect (SD) (SD) (SD) (SD)

Spanish Perfective 60.59% 75.29% 82.35% 72.94% n = 17 (23.04) (13.38) (17.14) (26.40)

Imperfective 71.76% 83.53% 83.53% 55.29% (21.86) (24.72) (21.58) (26.00)

English Perfective 54.00% 62.67% 85.33% 92.00% n = 15 (18.04) (23.74) (17.68) (16.56)

Imperfective 56.67% 69.33% 68.00% 32.00% (18.40) (21.20) (21.10) (22.42)

Hypothesis 2: Effects of Lexical Aspect on Tense–Aspect Marking Among L1 Groups

The three-way mixed between–within subjects (L1 group between; grammatical aspect and lex- ical aspect within) ANOVA revealed a significant interaction effect among the overall use of gram- matical aspect, lexical aspect, and L1, F (3, 30) = 3504.5, p = .03. The analyses performed on the ap- propriate use of tense–aspect markers also yielded a significant interaction among grammatical as- pect, lexical aspect, and L1 F (3, 30) = 3797.5, p = .03. The effects of lexical aspect on the over- all and appropriate use of tense–aspect marking are reported first within groups and then between groups. The overall and appropriate use of tense– aspect markers is displayed in Tables 1 and 2, respectively.

Grammatical Aspect Marking With Verb Types by Anglophones

Overall Use of Tense–Aspect Markers. Figure 1 shows that the Anglophones exhibited the highest rate of perfective marking on verbs of achieve- ment, followed by verbs of accomplishment,

Jesús Izquierdo and Laura Collins 357

FIGURE 1 Overall Distribution of Tense–Aspect Markers Across and Within Verb Types by Anglophones

44.33 46.67 58.67

80.00

51.00 50.00

40.00

20.00

100

Stative Activity Accomplishment Achievement

P er

ce n

ta ge

T es

t It

em s

Imperfective Perfective

Notes. Arrows highlight significant difference.

verbs of activity, and statives. Simple main effect contrasts on mean differences indicated that the use of the perfective with verbs of achievement differed significantly from its use with verbs of accomplishment, p = .01, d = 1.56; with verbs of activity, p = .01, d = 2.28; and with statives, p = .01, d = 2.47. The overall use of the perfec- tive with accomplishment verbs proved to be sig- nificantly different from its use with activity verbs, p = .03, d = .94; and with statives, p = .03, d = 1.10.

With respect to the overall use of imperfective marking, the Anglophones showed the highest rate of use for this grammatical aspect with sta- tives, closely followed by verbs of activity, and to a lesser degree by verbs of accomplishment. Verbs of achievement exhibited the lowest rate of im- perfective marking. Mean differences in imper- fective marking within durative verbs (i.e., states, activities, and accomplishments) were not signif- icant. Simple main effect contrasts revealed that the low rate of imperfective marking with verbs of achievement was significantly different from its rate of use with verbs of accomplishment, p = .01, d = –1.52; and of activity, p = .01, d = –2.10; and with statives, p = .01, d = –2.09.

Appropriate Use of Tense–Aspect Markers. Sim- ple main effect comparisons of the appropriate use of tense–aspect marking within the Anglo- phone group revealed that use of the perfective with achievement verbs differed significantly from its use with activity verbs, p = .01, d = 1.39 and

with statives, p = .01, d = 2.14. The appropri- ate use of the perfective with accomplishment verbs also differed from its use with activity verbs, p = .01, d = 1.05 and with statives, p = .01, d = 1.71. Mean comparisons of the appropriate use of the imperfective within the Anglophone group revealed significant differences between achieve- ment verbs and all durative verbs (i.e., statives, p = .03, d = 1.15; activity verbs, p = .01, d = 1.65; and accomplishment verbs, p = .01, d = 1.59).

Grammatical Aspect Marking With Verb Types by Hispanophones

Overall Use of Tense–Aspect Markers. Figure 2 shows that verbs of achievement elicited the high- est rate of perfective marking from the His- panophones. Within the durative verbs, verbs of accomplishment attracted the highest rate of use for the perfective, followed by activity verbs and statives, but mean differences in the use of the perfective were significant only between statives and achievement verbs, p = .04, d = 0.92. The Hispanophones exhibited the highest rate of im- perfective use with statives, followed by verbs of ac- tivity, accomplishment, and achievement. Pairwise comparisons yielded significant mean differences only between statives and verbs of achievement, p = .04, d = 0.83.

Appropriate Use of Tense–Aspect Markers. Simple main effect contrasts for the use of the perfective

358 The Modern Language Journal 92 (2008)

FIGURE 2 Overall Distribution of Tense–Aspect Markers Across and Within Verb Types by Hispanophones

40.59 44.71 48.82

56.47

54.12 51.18

47.65 37.65

100

Stative Activity Accomplishment Achievement

P er

ce n

ta ge

T es

t It

em s

Imperfective

Perfective

Notes. Arrows highlight significant differences.

with verb types among the Hispanophones re- vealed significant mean differences in the appro- priate use of the perfective only between statives and verbs of accomplishment, p = .01, d = 0.44. In appropriate imperfective marking, significant differences showed between verbs of achievement and accomplishment, p = .01, d = 1.24 and be- tween verbs of achievement and activity, p = .01, d = 1.09.

Between-Group Differences in the Use of Grammatical Aspects

Figures 3 and 4 illustrate differences between the two groups in overall marking and appropri- ate use of grammatical aspects within verb types.

Stative Verbs. Figure 3 illustrates the tendency for the Hispanophones to supply the imperfective with statives more often than the Anglophones and the tendency for the Anglophones to sup- ply the perfective with statives. However, mean contrasts revealed no significant between-group differences. Figure 4 shows that, overall, the His- panophones tended to mark both grammatical aspects more appropriately with stative verbs than did the Anglophones, but the only statistically sig- nificant between-group mean difference was for imperfective marking with statives, p = .04, d = 0.72.

Verbs of Activity. Patterns of overall tense– aspect marking with activity verbs were similar to those with stative verbs. The Hispanophones showed more frequent use of the imperfective with activity verbs than did the Anglophones. The Anglophones showed more use of the perfective with activity verbs than the Hispanophones. Yet, between-group mean differences were not signifi- cant. In comparison to the Anglophones, the His- panophones showed larger means in appropriate use of the perfective and imperfective, but mean contrasts revealed no significant differences be- tween groups.

Verbs of Accomplishment . The Anglophones supplied the perfective with accomplishment verbs significantly more often than did the His- panophones, p = .02, d = –0.86. Contrasts of the appropriate use of the perfective for accomplish- ment verbs showed no significant differences be- tween groups, however. The mean difference in the overall use of the imperfective for accom- plishment verbs between groups was not signif- icant. Yet, mean comparison of the appropriate use of the imperfective for accomplishment verbs showed that the Hispanophones supplied the im- perfective more appropriately than the Anglo- phones did, p = .04, d = 0.71.

Verbs of Achievement . The Anglophones mar- ked achievement verbs with the perfective more

Jesús Izquierdo and Laura Collins 359

FIGURE 3 Between-Group Differences in the Overall Use of Tense–Aspect Markers With Verb Types

100

Stative Activity Accomplishment Achievement

P er

ce n

ta ge

T es

t It

em s

Perfective by Hispanophones

Imperfective by Hispanophones

Perfective by Anglophones

Imperfective by Anglophones

Notes. Arrows highlight significant differences.

FIGURE 4 Between-Group Differences in the Appropriate Use of Tense–Aspect Marking With Verb Types

100

Stative Activity Accomplishment Achievement

P er

ce n

ta ge

O b

li ga

to ry

C o

n te

xt s

Perfective by Hispanophones

Imperfective by Hispanophones

Perfective by Anglophones

Imperfective by Anglophones

Notes. Arrows highlight significant differences.

often than the Hispanophones, and the result- ing mean difference was significant, p = .01, d = –1.36. Similarly, mean contrasts on the ap- propriate use of the perfective with achieve- ment verbs revealed that the Anglophones supplied the perfective for achievement verbs in obligatory contexts significantly more often than the Hispanophones, p = .01, d = –0.83. The Hispanophones supplied the imperfective for achievement verbs more often than the Anglophones, and the resulting mean difference

proved to be significant, p = .01, d = 0.91. In addi- tion, the Hispanophones showed a greater appro- priate use of the imperfective with this verb type than the Anglophones, and the resulting mean difference was significant, p = .01, d = 0.93.

To summarise this section with respect to Hy- pothesis 1, the Hispanophones did not show an overall preference for the use of the perfective over the imperfective, and they exhibited a similarly appropriate use of both tense–aspect markers.

360 The Modern Language Journal 92 (2008)

With respect to Hypothesis 2, simple main effect contrasts within groups revealed a large number of significant interactions between lexical aspect and grammatical aspect within the Anglophone group. Their use of the perfective differed signifi- cantly within dynamic verbs (i.e., verbs of activity, accomplishment, and achievement) and between statives and telics (i.e., verbs of accomplishment and achievement). Use of imperfective marking with achievement verbs differed significantly from its use with all durative verbs (i.e., statives, activity verbs, and accomplishment verbs).

Within the Hispanophone group, significant differences in the use of the perfective and the imperfective were found only between sta- tive and achievement verbs. In addition, the Hispanophone participants showed more appro- priate use of the imperfective with statives and telics and correspondingly less reliance on the perfective to mark aspect with telics than the Anglophones. The Anglophones, by contrast, marked almost all achievement verbs with the perfective. This tendency, in turn, led the An- glophones to use this tense–aspect marker more appropriately in the obligatory contexts than the Hispanophones.

RESULTS: INTERVIEWS

The interviews in English were transcribed by an Anglophone research assistant and the inter- views in Spanish by a Hispanophone research assistant. After preliminary analyses of the inter- views, eight categories were elaborated for the tense–aspect comments made by the learners: (a) reference to the L1, (b) lexical aspect, (c) form–function mapping, (d) adverbials, (e) over- reliance on a pedagogical rule, (f) learner’s rule,

TABLE 3 Distribution of Explanations by L1 Groups

Anglophones Hispanophones

Raw % Instance Participants Raw % Instance Participants Explanation Instances Distribution (n = 11) Instances Distribution (n = 8) Reference to L1 20 23.81 9 27 30.34 6 Form–Function Mapping 15 17.86 6 21 23.60 6 Adverbial 13 15.48 6 12 13.48 6 Pedagogical Rule 15 17.86 7 10 11.24 4 Lexical Aspect 15 17.86 9 6 6.74 5 Discourse 2 2.38 2 5 5.62 2 Learner’s Rule 3 3.57 3 4 4.49 3 L2 Frequency 1 1.19 1 4 4.49 2 Total 84 100 89 100

Note . L1 = first language; L2 = second language

(g) discourse, and (h) L2 frequency. These cat- egories are explained in more detail in the next subsections.

In all, 173 explanations were coded (89 by the 8 Hispanophones and 84 by the 11 Anglophones) and assigned to a category. All interviews were coded twice by the first author in reverse order. Coding differences were found for 14 comments, yielding a 92.18% coding reliability. These differ- ences were resolved through discussion with the second author. Table 3 summarises the results for each category in raw and percentage numbers, in- dicating the number of participants contributing to each category. The following subsections de- scribe the established categories and the results.

Reference to the L1

This category included explanations where the learners reported that they used direct transla- tion from their L1 or referred to the tense–aspect marking of their L1. Table 3 shows that the His- panophone participants tended to refer to their L1 more often than the Anglophones. In this cat- egory, the Hispanophones included explanations such as: That’s the way we say it in Spanish; It doesn’t sound good in Spanish; I thought it in Spanish, and they made explicit L1–L2 contrasts as illustrated at the beginning of this article. The Hispanophones’ L1-related explanations spanned the various val- ues of the imperfective. In contrast, the Anglo- phones referred to their L1 to acknowledge the hindering effects of L1–L2 differences on their learning of past tense marking in French (see the quotation at the beginning of this article). Only two Anglophones referred to their L1 to map the progressive value of the imperfective with the past progressive.

Jesús Izquierdo and Laura Collins 361

Form–Function Mapping

Through explanations in this category, the learners showed their knowledge of the functions of the two grammatical aspects, as illustrated in Example 3 by Krista’s9 (L1 English) appropriate use of the perfective for the verb trembler ’to trem- ble/quake/shake’ and by her explanation.

EXAMPLE 3 Appropriate Perfective Use (Trembler ), Krista’s Cloze Test

A: J’ai entendu à la radio ce matin qu’en 1986, la terre (trembler) a tremblé très fort à Mexico. ‘I heard on the radio this morning that in 1986, there was a big earthquake [lit. the earth shake + perfective very hard] in Mexico.’

Krista’s Explanation During the Interview Interviewer: Okay. Why did you go for passé com-

posé here? Krista: Because, eh, in this situation, it was

one shaking of the earth, not mul- tiple and continued shaking. In this case, passé composé is needed.

The Hispanophones provided more form– function mapping explanations than their En- glish counterparts. Through their explanations, the Hispanophones retrieved the various values of imperfective and perfective morphology and were able to identify them in the obligatory contexts of the test. The Anglophone explanations were re- stricted to the use of the perfective to mark com- pletion and the use of the imperfective to mark progression and habituality.

Lexical Aspect

This category included explanations in which the learners referred to the meanings of the verbs to supply L2 temporal morphology, as illustrated in Example 4. In this example, David’s (L1 En- glish) explanation for his inappropriate use of the perfective centred on the meaning of the verb tomber ‘to fall.’ He pointed out that falling is an action that needs to reach completion, or that this action occurs in such a short period of time that there is no time for anything else to happen.

EXAMPLE 4 Inappropriate Perfective Use (Tomber ), David’s Cloze Test

Les gens dans les gratte-ciels du centre ville essayaient de s’échapper pendant que les édifices (tomber) sont tombé par terre.

‘People in the downtown skyscrapers tried to es- cape while the buildings fall + perfective down.’ David’s Explanation During the Interview Interviewer: Why did you use passé composé here? David: I thought that this is saying that

the buildings fell, right? So, buildings fall and that’s it [punctual aspect of achievement verbs].

The percentage of explanations coded for lexical aspect among the Anglophone group (17.86%) contrasts with that exhibited by the His- panophones (6.74%). Lexical aspect-type expla- nations were coded at least once in the transcripts of 9 of the 11 Anglophones interviewed, or 81% of the participant group sample. Despite percentage differences between the L1 groups, lexical aspect explanations consistently revealed that learners from both L1 groups associated the punctuality inherent in the meaning of achievement verbs with the value of completion rendered by the perfective.

Overreliance on a Pedagogical Rule

This category included explanations based on a rule that the learners explicitly acknowledged to have learned in class and that they appeared to use without reflecting on its applicability, as illustrated in Example 5, Kerri’s (English L1) inappropriate use of the imperfective with penser ‘to think.’

EXAMPLE 5 Inappropriate Imperfective Use (Penser ), Kerri’s Cloze Task

Puis, il a vu du monde courir partout et il (penser) pensait à se cacher sous la table. ‘Then he saw people running everywhere and think + imperfective about hiding under the table.’

Kerri’s Explanation During the Interview Interviewer: Then, here, why did you use impar-

fait? Kerri: That, just because . . . pensait because

I learned in class that verbs like penser ‘to think’ or croire ‘to believe’; like de- scriptive words, that they’re usually in the imparfait .

Within the Anglophone group, 17.86% of the explanations fell into this category; however, this explanation was less frequent among the His- panophones. Some Anglophones showed aware- ness of the limitations of the pedagogical rules on which they were relying; yet, they used them as “rules of thumb.” For example, one Anglophone

362 The Modern Language Journal 92 (2008)

participant said, after a research assistant pointed out that stative verbs can be used in the perfec- tive: “I know there are cases where these verbs [statives] are needed in passé composé , but I can’t figure them out, so at this point, I feel safe using imparfait with penser .”

Adverbials

This category included explanations in which the learners indicated that they supplied tense– aspect marking based on the presence of an ad- verb in the sentence, without further analysing the context of the target item. In Example 6, Eva (Spanish L1) provides this type of explanation for her inappropriate use of the perfective with the verb se lever ‘to get up.’

EXAMPLE 6 Inappropriate Perfective Use (Se Lever ), Eva’s Cloze Task

Il était 8 heures et ma grand-mère (se lever) s’est levé quand elle s’est rendu compte qu’il y avait un très fort tremblement de terre. ‘It was 8 a.m. and my grandmother get + perfec- tive up when she realized that they were having [lit. that there was] a very strong earthquake.’

Eva’s Explanation During the Interview Interviewer: Now here, why did you choose passé

composé ? Eva: Because a specific time is given. Here

it says: It was 8 a.m.

Both L1 groups reported the impact of adver- bials on the use of perfective and imperfective markers with similar frequency. In obligatory con- texts for the use of the imperfective, the learners justified their use of perfective markers by point- ing to a specific date, to a time, or to sequence adverbs appearing in the cloze passage.

Less Frequent Explanations

Other explanations included some self- generated rules that participants from both L1 groups seemed to operate on and for which it was, in some cases, possible to establish a source, such as discourse marker analyses and L2 frequency. Discourse explanations included in- stances in which the learners referred to elements from other sentences in the paragraph to supply tense–aspect marking. These elements were, for instance, tense–aspect markers in the surround- ing sentences and sequence markers (e.g., d’abord ‘first,’ après ‘then,’ finalement ‘finally’).

The L2 influence category included explana- tions in which the learners’ use of tense and aspect was based on patterns of L2 use that they said they had noticed in the L2 input. For example, Perla (L1 Spanish) made the same mistake on the cloze test with the verb se lever as Eva did in Example 6. When asked why she used the perfective instead of the imperfective, Perla said, “Se lever is a pronom- inal verb. See, I’ve seen pronominal verbs in passé composé only.”

Some explanations suggested that the partici- pants’ use of tense–aspect markers was based on a rule for which the origin could not be identified. These explanations were coded as Learner’s Rule, as is the case in Example 7, provided by Luis (L1 Spanish). Explanations of this type were more fre- quent among the Anglophones than among the Hispanophones.

EXAMPLE 7 Inappropriate Imperfective Use (Avoir ), Luis’s Cloze Task

B: Oui, ça a été une grande catastrophe. Les en- fants qui se dirigeaient vers l’école (avoir) avait la peur de leur vie. ‘Yes, it was a huge catastrophe. Children who were on their way to school fear + imperfective [lit. have + imperfective fear] for their lives.’

Luis’s Explanation During the Interview Interviewer: Then you used avoir in imparfait. Luis: Here I wasn’t sure, so I used imparfait.

I didn’t know whether, here, I should say il y avait [‘to have’ in French im- perfective] or il y a eu [‘to have’ in French perfective]. But it was avoir , so I used imparfait.

Findings in the retrospective interview with re- spect to Hypothesis 1 are summarised as follows: The Hispanophones made more frequent refer- ence to their L1 to explain their appropriate choices of L2 tense–aspect markers and, over- all, showed a better understanding of the form– function mapping of the two aspects than their Anglophone counterparts. The Anglophones’ ref- erence to the L1 emphasised the difficulties they had when they tried to compare tense–aspect in French and English.

Explanations from both groups were also con- gruent with Hypothesis 2 because they pointed out that the participants in the Anglophone group were more likely to use tense–aspect mark- ers as they related to the semantics of the verbs than were the Hispanophones. In addition, the Anglophones reported that they relied on

浦原喜助�

Jesús Izquierdo and Laura Collins 363

partially understood pedagogical rules and adver- bials more often than the Hispanophones.

SUMMARY OF OVERALL FINDINGS

Hypothesis 1 stated that Hispanophone learn- ers of French would not exhibit a preference for perfective over imperfective marking. The ANOVA analyses revealed no difference in the use of the perfective and imperfective marking among Hispanophones, but it revealed a sig- nificant difference between the use of the two tense–aspect markers among the Anglophones. The Hispanophones also made more appropriate use of the imperfective than the Anglophones. The qualitative interview data showed that many of the Hispanophones were aware of similari- ties between the grammatical aspects of Span- ish and French and, perhaps because the cloze task allowed them time to access this knowledge, they also reported benefiting from this aware- ness. In light of this evidence, Hypothesis 1 was retained.

Hypothesis 2 stated that the use of tense–aspect markers with verb types would be less influenced by lexical aspect among the Hispanophone learn- ers than among the Anglophones. The ANOVA analyses of the interaction of lexical aspect and grammatical aspect revealed fewer interactions within the Hispanophone group than within the Anglophone group. Between-group analyses on the overall and appropriate use of tense–aspect markers suggested that the Hispanophones were better able to identify nonprototypical contexts for the use of the imperfective and relied less on the perfective to mark aspect with telics than the Anglophones. Congruent with these quantitative findings, the interview data pointed out that the Anglophones needed to appeal to lexical aspect and rules of thumb in order to decide between the two forms. In light of this evidence, Hypothesis 2 was retained.

DISCUSSION

A quick glance at Figures 1 and 2 and at the differences between the lexical aspect and L1 reference comments in Table 3 clearly reveals the differences found between the two groups of learners. Lexical aspect played a very impor- tant role for the Anglophones, whereas the His- panophones appeared to benefit from the facili- tative effect of L1 influence. The contrast in the number of significant differences in Figures 1 and 2 and in the analyses points to a large number of complex interactions between grammatical as-

pect marking and verb semantics within the An- glophone group.

The influence of semantic categories on the An- glophones’ performance is congruent with pre- dictions of the aspect hypothesis and previous research among other Anglophone learners of Romance languages (Andersen, 1991; Bardovi- Harlig & Bergström, 1996; Comajoan, 2005; Co- majoan & Reid, 2005; Harley, 1989, 1992; Harley & Swain, 1978; Salaberry, 1998), including dif- ferences in the distribution and appropriate use of the perfective that related not only to telicity but also to punctuality; that is, among the Anglophones, the use of perfective mark- ing was influenced by telic–atelic differences, and within telic verbs, it was influenced by the dura- tive/nondurative contrast distinguishing achieve- ment verbs from accomplishment verbs. The Anglophone participants’ use of the imperfec- tive adhered strongly to the durative features of verb types, which led them to a less appropriate use of this grammatical aspect with achievement verbs than was found among their Hispanophone counterparts. Among the Hispanophones, how- ever, the lexical aspect influence led to differential effects only between the prototypical extremes of the aspectual categories: stative and achievement verbs.

One striking similarity between the two groups was the effect of telicity and punctuality on the use of the perfective. The impact of punctual- ity on the preference for the perfective was ob- served both in the participants’ production (the cloze test) and in their comments on their pro- duction. When the learners needed to explain their preference for the perfective with achieve- ment verbs (e.g., arriver ‘to arrive’ and tomber ‘to fall’) that were given in contexts obligatory for the imperfective, their responses to our state- ments that the imperfective was needed with these verbs included: “But you don’t arrive once and again, you arrive only once”; or “Buildings fall and that’s it.” The latter reaction often was accom- panied with gestures indicating the immediate completion of the event. These reactions indicate that the learners held representations in which the nondurative and nonrepetitive properties of achievement verbs strongly influenced their pref- erence for perfective marking with this verb type.

Similarities and differences in the impact of tense–aspect marking choice with verb types among L1 groups suggest that although the influ- ence of some lexical properties on tense–aspect marking choice can be mitigated by L1–L2 simi- larities, the lexical properties of punctuality and

浦原喜助�

364 The Modern Language Journal 92 (2008)

nondynamic duration are a powerful influence on the acquisition of tense–aspect markers irrespec- tive of L1–L2 similarities/differences.

The findings of a facilitative effect for the His- panophones are in line with post hoc explana- tions for results obtained in previous research (De Lorenzo Roselló, 2001; Kihlstedt, 2002; Sal- aberry, 2005), which indicated a benefit for the development of the perfective and imperfective distinction. This L1 influence on L2 tense–aspect marking was apparent at the outset of the re- search in the initial organisation of the partici- pants for the analyses. As explained previously, we grouped the participants by their overall pro- ficiency in French, as measured by the in-house placement test of one of the participating uni- versities. Initially, however, we had attempted to group the Hispanophone and Anglophone par- ticipants by similar overall knowledge of tense– aspect markers (as has been done in some previ- ous research; e.g., Bardovi-Harlig, 1998; Collins, 2002, 2004). When we did so, there was a signif- icant difference on the proficiency test scores, in favour of the Anglophones. This finding sug- gests that Hispanophones have a rate advantage in acquiring the imperfective. In our study, it was not possible to determine, however, whether L1 influence had a positive impact on the emer- gence of imperfective morphology or at what level of L2 tense–aspect marking proficiency the beneficial effects of L1 influence become oper- ative. We are currently investigating this issue in a cross-sectional study in which the overall use of the perfective and imperfective and their dis- tribution across verb types are being examined among Hispanophone learners at different levels of French L2 tense–aspect proficiency (Izquierdo, 2008).

The distribution of temporal morphology in the cloze test and the learners’ explanations for their choices clearly indicate that both the His- panophones and the Anglophones held men- tal representations in which the temporality in- herent in the meaning of verbs was associated with functions of grammatical aspect markers, but the Hispanophones made this association to a lesser degree. This finding lends support to Collins’s (2004) argument that L1 influence operates within the attested order of acquisi- tion of tense–aspect markers, affecting the rate but not the distribution of grammatical aspect markers. The findings for the role of lexical as- pect for the Anglophones are consistent with Housen’s (2000) suggestion that the absence of a grammatical aspect may increase L2 learners’ reliance on the inherent temporality in the mean-

ing of verbs during the acquisition of L2 temporal morphology.

It is also important to note that not all of the Hispanophone learners appeared to be aware of L1–L2 similarities. (Contrast the comment at the beginning of the article with Example 7.) Andersen’s (1983) principle, transfer to somewhere , points out that, for transfer to occur, learners need some knowledge that allows them to estab- lish differences or similarities between features in their L1 and those in their L2 or in their interlan- guage system. Although the relationship between metalinguistic knowledge and noticing/acquiring language is not straightforward (e.g., Lightbown, 2000), there is nevertheless some research on the effective use of L1–L2 contrasts to help learn- ers overcome incorrect hypotheses about the L2 based on perceived differences between the L1 and the L2 (Kupferberg & Olshtain, 1996; Spada, Lightbown, & White, 2005; White, Muñoz, & Collins, 2007). One issue raised by our findings is whether these contrasts may also help learn- ers make better use of key similarities between the L1 and the target language. If L1 knowledge facilitates L2 tense–aspect mapping among His- panophone learners of French, language aware- ness tasks, such as those discussed by Alegrı́a de la Colina and Garcı́a-Mayo (2006) and Simard (2007), may assist them in their construction of the L2 tense–aspect system.

Previous research shows that the use of the perfective with être and avoir (i.e., in nonproto- typical contexts) constitutes a challenge for L2 learners of French (Ayoun, 2004; Harley, 1989, 1992; Harley & Swain, 1978). Comments made during the retrospective interviews showed that participants in both L2 groups in the present study also found this use of the perfective challeng- ing. Some participants first supplied the perfec- tive for être , as required by the obligatory context, but then crossed it out and replaced it with the imperfective. In explaining this change, one His- panophone reported that he first thought that the perfective felt appropriate because the context de- scribed a single completed event. However, he was unsure of ever having seen être in the perfective and he also reported learning that imperfective should be used with states. He therefore opted, in- correctly, for the imperfective. These findings sug- gest that if Hispanophones are benefiting from L1 influence to supply nonprototypical markers with stative verbs and if Anglophones have reached a point for the potential development of this type of marking, the rules they hold and the type of input available to them may hinder interlanguage change.

浦原喜助�

Jesús Izquierdo and Laura Collins 365

CONCLUSION

This study demonstrated that L1–L2 similarities in tense–aspect marking in Spanish and French facilitated the distribution and appropriate use of grammatical aspect markers with some verb types (dynamic-durative verbs) for Hispanophone learners of French. The use of tense–aspect mark- ing in the least prototypical contexts (stative and accomplishment verbs), however, posed a prob- lem. Challenges arose in these contexts for both the Anglophones and the Hispanophones be- cause nonprototypical contexts require learners to tease apart the temporality inherent in the meaning of the verbs and to recognize the range of semantic values that tense–aspect markers of- fer to a speaker. These findings point to the need for pedagogical intervention in which learners are provided with opportunities to process nonproto- typical marking (Blyth, 2005; Housen, 2002) and to receive corrective feedback that draws their at- tention to the restricted semantic scope of tense– aspect forms in their interlanguage. Previous re- search suggests that prompts may be a particu- larly effective feedback type for pushing students to differentiate past tense marking values from verb semantics in their L2 production (Lyster, 2004a, 2004b, 2007; Lyster & Mori, 2006; Lyster & Ranta, 1997; Ranta & Lyster, 2007). Learners also may benefit from externalising their under- standing of the L2 temporal system at differ- ent points in the developmental process (Collins, 2005, 2007; Trévise, 1996). There is also some ev- idence that computer-assisted language learning modules that simultaneously provide visual, tex- tual, and aural exposure to the perfective and imperfective markers may provide the type of meaningful learning opportunities that can pro- mote noticing of the forms (Izquierdo, 2007; Izquierdo & Lyster, 2007). A computerised mul- timedia learning environment also may allow for the provision of repeated and sustained interven- tion that may facilitate the acquisition of this chal- lenging aspect of L2 grammar.

ACKNOWLEDGMENTS

We would like to thank our research assistants Véronique Fortier and Susan Ballinger in Canada and Jaime Córdova and Nohemi Lezama in Mexico for assistance with the data collection and analyses and our research associate Randall Halter for patiently advising us on the statistical analyses. We are also grateful for the helpful comments of Roy Lyster, Daphnée Simard, Dalila Ayoun, and three anonymous reviewers of The Modern Language Journal on previous versions of this manuscript.

Preliminary results from this study were presented at the annual meeting of the Canadian Association of Applied Linguistics (2005) at the University of Western Ontario, the Second Language Research Forum (2005) at Columbia University, and the Segundo Encuentro de Profesionales en Lenguas Extranjeras EPLE (2005) of the Universidad Juárez Autónoma de Tabasco. Funding for the research was provided by the Social Sciences and Humanities Research Council of Canada (a standard research grant 410-2002-0988 awarded to Roy Lyster, and a Doctoral Fellowship awarded to Jesús Izquierdo) and from the Fonds québecois de la recherche sur la société et la culture (a Quebec Ministry of Education grant awarded to Laura Collins).

NOTES

1 The perfective aspect in French also can be marked through the passé simple. This form is increasingly less common in daily discourse.

2 The number and description of the functional val- ues of the imperfective vary across authors (cf. Abrate, 1983; Ayoun, 2004; Comrie, 1976; Howard, 2002; Kihl- stedt, 2002). In our view, the habit, duration, iteration, and progressive values of the imperfective capture the essential functions of the form.

3 Because these categories and the resulting verb types have been described in a number of recent publi- cations, we have provided a brief description only. More comprehensive accounts can be found in studies by Bardovi-Harlig (2000), Collins (2002), and Ayoun and Salaberry (2005).

4 One difference between these languages lies in the morphosyntactic construction of the French perfective. French requires a compound form, whereas Spanish operates on a series of inflections in the main verb. Both languages mark the imperfective with a series of inflections in the main verb.

5 English L1 learners functioned primarily in English in Montreal. All learners were enrolled at an English- speaking university and lived in an English-speaking area of the city, close to campus. During telephone calls to set interview appointments, the participants or their room- mates always answered the phone in English and asked the researcher to switch from French to English. Back- ground television and radio noise was always in English as well.

6 We are grateful to the French Department and the Coordinators of the Language Centre of the participat- ing university in Mexico for granting us access to the pro- ficiency test, as well as for their support for this project. Due to our research regulations, however, it is not pos- sible to acknowledge the name of this university openly.

7 In the English L1 group, there were 3 male and 12 female participants. In the Spanish L1 group, there were 4 male and 13 female participants.

8 In previous studies (e.g., Ayoun, 2005; Salaberry, 1998), the use of tense–aspect by native speakers was used as baseline data to determine deviant patterns of L2 tense–aspect marking among L2 users. Our study, how- ever, addressed differences in L2 tense–aspect marking

浦原喜助�

366 The Modern Language Journal 92 (2008)

between two distinct L1 groups. Therefore, comparisons with native speakers were not made.

9 All participant names are pseudonyms.

REFERENCES

Abrate, J. (1983). An approach to teaching the past tenses in French. French Review, 56, 546–553.

Alegrı́a de la Colina, A., & Garcı́a-Mayo, M. d. P. (2006). Attention to form across collaborative tasks by low-proficiency learners in an EFL setting. In M. d. P. Garcı́a-Mayo (Ed.), Investigating tasks in formal language learning (second language acquisi- tion) (pp. 91–116). Clevedon: UK: Multilingual Matters.

Andersen, R. (1983). Transfer to somewhere. In S. Gass & L. Selinker (Eds.), Language transfer in lan- guage learning (pp. 195–215). Rowley, MA: New- bury House.

Andersen, R. (1991). Developmental sequences: The emergence of aspect markers in second language acquisition. In T. Huebner & C. Ferguson (Eds.), Crosscurrents in second language acquisition and linguistic theory (pp. 305–324). Amsterdam: Ben- jamins.

Andersen, R. (2002). The dimensions of pastness. In R. Salaberry & Y. Shirai (Eds.), The L2 acquisition of tense –aspect morphology (pp. 79–105). Amsterdam: Benjamins.

Andersen, R., & Shirai, Y. (1994). Discourse motivations for some cognitive acquisition principles. Studies in Second Language Acquisition, 16, 133–156.

Ayoun, D. (2001). The role of negative and positive feed- back in the second language acquisition of the passé composé and imparfait . Modern Language Jour- nal, 85, 226–243.

Ayoun, D. (2004). The effectiveness of written recasts in the second language acquisition of aspectual distinctions in French: A follow-up study. Modern Language Journal, 88, 31–51.

Ayoun, D. (2005). The acquisition of tense and aspect from a universal grammar perspective. In D. Ay- oun & R. Salaberry (Eds.), Tense and aspect in Ro- mance languages (pp. 79–128). Amsterdam: Ben- jamins.

Ayoun, D., & Salaberry, R. (Eds.). (2005). Tense and as- pect in Romance languages. Amsterdam: Benjamins.

Bardovi-Harlig, K. (1998). Narrative structure and lexi- cal aspect: Conspiring factors in second language acquisition of tense–aspect morphology. Studies in Second Language Acquisition, 20, 341–382.

Bardovi-Harlig, K. (2000). Tense and aspect in second lan- guage acquisition: Form, meaning and use . Malden, MA: Blackwell.

Bardovi-Harlig, K., & Bergström, A. (1996). Acquisition of tense and aspect in second language and for- eign language learning: Learner narratives in ESL and FFL. Canadian Modern Language Review, 52, 308–330.

Bardovi-Harlig, K., & Reynolds, D. (1995). The role of

lexical aspect in the acquisition of tense and as- pect. TESOL Quarterly, 29, 107–131.

Blyth, C. (2005). From empirical findings to the teach- ing of aspectual distinctions. In D. Ayoun & R. Salaberry (Eds.), Tense and aspect in Romance lan- guages (pp. 211–252). Amsterdam: Benjamins.

Clemance, C. (2004). L’acquisition de l’aspect dans les temps du passé par des apprenants anglophones [The acquisition of aspect in past tenses by Anglophone learners]. Unpublished master’s thesis, The Uni- versity of Western Ontario, London, Ontario, Canada.

Collins, L. (2002). The role of L1 influence and lexical aspect in the acquisition of temporal morphology. Language Learning, 52, 43–94.

Collins, L. (2004). The particulars on universals: A com- parison of the acquisition of tense–aspect mor- phology among Japanese and French-speaking learners of English. Canadian Modern Language Review, 61, 251–274.

Collins, L. (2005). Accessing second language learners’ understanding of temporal morphology. Language Awareness, 14, 207–220.

Collins, L. (2007). L1 differences and L2 similarities: Teaching verb tenses in English. English Language Teaching Journal, 61, 295–303.

Comajoan, L. (2005). The acquisition of perfective and imperfective morphology and the marking of dis- course grounding in Catalan. In D. Ayoun & R. Salaberry (Eds.), Tense and aspect in Romance lan- guages (pp. 35–78). Amsterdam: Benjamins.

Comajoan, L., & Reid, L. (2005, October). The aspect hypothesis and the acquisition of L2 Spanish imper- fect . Paper presented at the Second Language Re- search Forum, Columbia University, NY.

Comrie, B. (1976). Aspect . Cambridge: Cambridge Uni- versity Press.

Comrie, B. (1985). Tense . Cambridge: Cambridge Uni- versity Press.

De Lorenzo Roselló, C. (2001). Relations temporo- aspectuelles entre les procès dans le récit oral en français et castillan [Aspectual/temporal relations between French and Castilian in oral discourse]. In Y. Macchi (Ed.), Panorama de la linguistique his- panique (pp. 251–268). Lille, France: Publication de l’université Lille III.

Egi, T. (2004). Verbal reports, noticing, and SLA re- search. Language Awareness, 13, 243–264.

Ericsson, K. A., & Simon, H. A. (1993). Protocol analysis: Verbal reports as data. Boston: MIT Press.

Gass, S., & Mackey, A. (2000). Stimulated recall method- ology in second language research. Hillsdale, NJ: Lawrence Erlbaum.

Harley, B. (1989). Functional grammar in French im- mersion: A classroom experiment. Applied Linguis- tics, 10, 331–359.

Harley, B. (1992). Patterns of second language in French immersion. French Language Studies, 2, 35– 79.

Harley, B., & Swain, M. (1978). An analysis of the verb system used by young learners of French. Interlan- guage Studies Bulletin, 3, 35–79.

Jesús Izquierdo and Laura Collins 367

Hedges, L. V., & Olkin, I. (1985). Statistical methods for meta-analysis. Orlando, FL: Academic Press.

Housen, A. (2000). Verb semantics and the acquisition of tense–aspect morphology in English. Studia Lin- guistica, 54, 249–259.

Housen, A. (2002). The development of tense–aspect in English as a second language and the variable influence of inherent aspect. In R. Salaberry & Y. Shirai (Eds.), The L2 acquisition of tense –aspect morphology (pp. 155–197). Amsterdam: Benjamins.

Howard, M. (2001). The effects of study abroad on the L2 learner’s structural skills. EUROSLA Yearbook, 1, 123–141.

Howard, M. (2002). Prototypical and non-prototypical marking in the advanced learner’s aspectuo- temporal system. EUROSLA Yearbook, 2, 87–113.

Izquierdo, J. (2008). Tense –aspect development in French L2: Benefits and limitations of L1–L2 similarities. Universidad Juárez Autónoma de Tabasco, Villa- hermosa, Mexico. Manuscript in preparation.

Izquierdo, J. (2007). Multimedia environments in the foreign language classroom: Effects on the acquisition of the French perfective and imperfective distinction. Unpublished doctoral dissertation, McGill Univer- sity, Montreal, Québec, Canada.

Izquierdo, J., & Lyster, R. (2007). Apprenons le français: Histoires et contes [Learning French: Tales and stories. Computer software]. Montreal, Canada: McGill University.

Kaplan, M. (1987). Developmental patterns of past tense acquisition among foreign language learners of French. In B. VanPatten, T. Dvorak, & J. Lee (Eds.), Foreign language learning: A research perspec- tive (pp. 41–60). New York: Newbury House.

Kihlstedt, M. (2002). Reference to past events in dia- logue: The acquisition of tense and aspect by ad- vanced learners of French. In R. Salaberry & Y. Shirai (Eds.), The L2 acquisition of tense –aspect mor- phology (pp. 323–363). Amsterdam: Benjamins.

Klein, W. (1994). Time in language . London: Routledge. Kleinmann, H. H. (1977). Avoidance behaviour in adult

second language acquisition. Language Learning, 27, 98–107.

Kupferberg, I., & Olshtain, E. (1996). Explicit con- trastive instruction facilitates the acquisition of dif- ficult L2 forms. Language Awareness, 5, 149–165.

Laufer, B., & Eliasson, S. (1993). What causes avoidance in L2 learning: L1–L2 difference, L1–L2 similar- ity, or L2 complexity? Studies in Second Language Acquisition, 15, 35–48.

Lightbown, P. M. (1985). Input and acquisition for sec- ond language learners in and out of classrooms. Applied Linguistics, 6, 263–273.

Lightbown, P. M. (2000). Anniversary article: Classroom SLA research and second language teaching. Ap- plied Linguistics, 21, 431–462.

Liskin-Gasparro, J. (2000). The use of tense–aspect mor- phology in Spanish oral narratives: Exploring the perceptions of advanced learners. Hispania, 83, 830–844.

Lyster, R. (2004a). Differential effects of prompts and

recasts in form-focused instruction. Studies in Sec- ond Language Acquisition, 26, 399–432.

Lyster, R. (2004b). Research on form-focused instruc- tion in immersion classrooms: Implications for theory and practice. Journal of French Language Studies, 14, 321–341.

Lyster, R. (2007). Learning and teaching languages through content: A counterbalance approach. Amster- dam: John Benjamins.

Lyster, R., & Mori, H. (2006). Interactional feedback and instructional counterbalance. Studies in Sec- ond Language Acquisition, 28, 269–300.

Lyster, R., & Ranta, L. (1997). Corrective feedback and learner uptake. Negotiation of form in commu- nicative classrooms. Studies in Second Language Ac- quisition, 20, 37–66.

Ranta, L., & Lyster, R. (2007). A cognitive approach to improving immersion students’ oral language abil- ities: The awareness-practice-feedback sequence. In R. DeKeyser (Ed.), Practice in a second language (pp. 141–160). New York: Cambridge University Press.

Sabourin, L. (2001). L1 effects on the processing of grammatical gender in L2. EUROSLA Yearbook, 1, 159–169.

Salaberry, M. R. (1998). The development of aspec- tual distinctions in L2 French classroom learning. Canadian Modern Language Review, 54, 508–541.

Salaberry, M. R. (2000). The development of past tense morphology in L2 Spanish. Amsterdam: Benjamins.

Salaberry, M. R. (2005). Evidence for transfer of knowl- edge of aspect from L2 Spanish to L3 portuguese. In D. Ayoun & R. Salaberry (Eds.), Tense and aspect in Romance languages (pp. 179–210). Amsterdam: Benjamins.

Salaberry, M. R., & Shirai, Y. (Eds.). (2002). The L2 acqui- sition of tense –aspect morphology. Amsterdam: Ben- jamins.

Simard, D. (2007). L’éveil au langue en classe de langue seconde: De la sensibilisation langagière à l’enseignment de la forme [Language aware- ness in the second language classroom: From metalinguistic reflection to form-focused instruc- tion]. Journal of French Language Studies, 17, 187– 206.

Spada, N., Lightbown, P., & White, J. (2005). The im- portance of form/meaning mappings in explicit form-focused instruction. In A. Housen & M. Pier- rard (Eds.), Investigations in instructed second lan- guage acquisition (pp. 199–234). Berlin: Mouton de Gruyter.

Taylor, K. L., & Dionne, J.P. (2000). Accessing problem- solving strategy knowledge: The complemen- tary use of concurrent verbal protocols and retrospective debriefing. Journal of Educational Psychology, 92, 413–425.

Trévise, A. (1996). Contrastive metalinguistic represen- tations: The case of “very French” learners of En- glish. Language Awareness, 5, 188–195.

White, J., Muñoz, C., & Collins, L. (2007). The his/her challenge: Making progress in a ‘‘regular’’ second

368 The Modern Language Journal 92 (2008)

language program. Language Awareness, 16, 278– 299.

Zobl, H. (1980a). Developmental and transfer errors: Their common bases and (possibly) differential effects on subsequent learning. TESOL Quarterly, 14, 469–479.

Zobl, H. (1980b). The formal and developmental selec- tivity of L1 influence on L2 acquisition. Language Learning, 30, 43–57.

Zobl, H. (1982). A direction for contrastive analysis: The comparative study of developmental sequences. TESOL Quarterly, 16, 169–183.

APPENDIX Excerpts from Cloze Task

Note. Verbs in italics are target items, and the remaining verbs are distractors. This formatting is for illustration purposes; in the learners’ version of the task no such distinction was made.

Situation 12 A: Hein! J’ai entendu dire que ton père a gagné le loto. B: Oui. Durant le week-end, je (visionner ) mon film favori quand ma soeur m’a téléphoné pour

m’annoncer la nouvelle. A: Que va-t-il faire avec son argent? B: Je ne sais pas, mais ça fait des années que mes parents (rêver) de s’acheter une maison en Floride.

Situation 13 Mes amis et moi avons suivi un entraı̂nement intensif pour nous préparer à escalader le mont Everest. Malgré cet entraı̂nement, l’année passée, pendant que je (monter ) le mont Everest, soudainement mon corps et ma tête (sembler ) ne plus répondre. J’ai demandé à mes amis d’arrêter pour me reposer un peu. Je (dormir ) , puis mes amis m’ont dit qu’ils voulaient continuer la marche jusqu’au coucher du soleil. Mais moi, je (se sentir ) très fatigué. Deux jours plus tard, nous avons décidé d’abandonner et de revenir chez nous, décision que nous (regretter) maintenant! Pendant que nous (quitter ) le mont Everest, je me suis promis d’y retourner un jour, et de finir la montée. Si d’autres personnes (déjà/ monter )

l’Everest, pourquoi pas nous?

Translation of Situation 12 A: Hey, I heard your dad won the lottery. B: Yes. Last weekend, I (watch) my favourite film when my sister rang me up to share the news. A: What will he do with the money? B: I don’t know, but my parents (dream) of a house in Florida for many years.

Translation of Situation 13 Last year, my friends and I trained hard to climb up Mount Everest. Despite all the training, while I (climb)

up the mountain, my body (seem) not to respond anymore. I asked my friends to stop for a while to catch my breath. I (sleep) , then my friends wanted to go on climbing till sunset. But, I (feel) very exhausted. Two days later, we decided not to continue climbing and went back home. Now, we (regret) our decision. I remember as we (leave) Mount Everest, I promised to myself I’d come back one day and finish climbing. If some other people (climb) up it, we can also do it.