Lexicalized Aspectual Usage in Oral Proficiency Interviews RICHARD M. ROBIN The George Washington University Department of Romance, German, and Slavic Languages and Literatures Phillips 509 Washington, DC 20052 Email:
[email protected] This study suggests that Intermediate High and Advanced speakers produce aspectually valid constructions in Oral Proficiency Interviews (OPIs) in large part because they are doing more than assigning aspect to lexical categories (Lexical aspect hypothesis), but because they are assigning lexicalized meaning to discrete verbs, for example govorit’ (imperfective)—‘to talk,’ but skazat’ (perfective)—‘to say.’ Analysis of the data from 55 OPIs conducted with second language (L2) Russian speakers ranging from Intermediate High to Superior confirms three hypotheses: (a) the proportion of aspectual utterances that owe their correctness to lexicalization decreases as proficiency increases; (b) aspectual competence as demonstrated by the greater number of aspectually non-lexicalized verbs in the speech of L2 Russian speakers rises with proficiency level; (c) a decrease in the proportion of lexicalized verbs in an examinee’s speech correlates positively with an increase in the variety of verbs used. Other findings include a relative paucity of unprefixed verbs of motion in the corpus and a high ratio of lexicalized forms of skazat’ ‘to say,’ and their distribution did not correlate with proficiency levels in either direction. The high ratio of aspectually lexicalized items in the corpus and the aspectual inaccuracies to which they lead have implications for teaching students who are approaching Intermediate High and Advanced levels of proficiency.
SECOND LANGUAGE (L2) SPEAKERS WHO have reached the American Council on the Teaching of Foreign Languages (ACTFL) Advanced level “demonstrate the ability to narrate and describe in all major time frames (past, present, and future).” Advanced Low speakers demonstrate “some control of aspect,” but Advanced Mid speakers demonstrate “good control of aspect” (ACTFL, 2012). But what constitutes good control of aspect? Does a ratable sample at Advanced Mid confirm that an L2 speaker has essentially mastered a language’s aspectual system? Specifically do American speakers of Russian with Advanced ratings make correct aspectual choices?
The Modern Language Journal, 96, 1, (2012) DOI: 10.1111/j.1540-4781.2012.01292.x 0026-7902/12/34–50 $1.50/0 C 2012 The Modern Language Journal
The question of aspectual choices addresses competence and is not the same as “Do speakers use the right aspect?” As we shall see shortly, students with Advanced ratings in Russian display overwhelmingly accurate control of aspect. Even Intermediate High speakers in our sample produced aspectually adequate verb forms 86.91% of the time. (See Appendix for samples of rated speech from this corpus.) That trend in turn suggests that learners at that level have mastered a universal grammar of aspect,1 whether one presented explicitly in pedagogical form that is conceptually based (e.g., Janda, 2003) or through rules of thumb, either taught or deduced through non-pedagogical acquisition. At the same time, Rassudova (1975) contended that even after “several years of study the student’s understanding [my emphasis] of verbal aspect and how it functions in speech communication falls short of our present scientific understanding of it” (p. 139) This article tests the notion that students narrating events
35
Richard M. Robin in Russian at the Advanced level are actually making consistent aspectual choices. It will be proposed that in many situations, they are assigning fixed lexical meanings to verbs of one aspect or another. Furthermore, it is hypothesized that (1) the proportion of aspectual utterances that owe their correctness to lexicalization decreases as proficiency increases and (2) the number of instances in which true aspectual choices are to be made rises with proficiency level, and (3) a decrease in the proportion of lexicalized verbs in a an examinee’s speech correlates positively with an increase in the variety of verbs used. RUSSIAN ASPECT Like all Slavic languages, Russian expresses aspect morphologically in surface verb forms except the present tense. Nearly all verbs have perfective and imperfective paradigms. The copular/existential byt ‘to be’ has no perfective and no surface present tense. About 650 other verbs are listed by dictionaries as bi-aspectual (Gladney, 1982), for example, ispolzovat ‘to make use of,’ uqastvovat ‘to participate,’ or bi-aspectual when not iterative, like organizovat ‘to organize’ for example. A traditional approach views aspect as a privative opposition in which the imperfective verb is the unmarked member, while the perfective is marked for completion (Forsyth, 1970, pp. 82–102). The finite forms of an imperfective verb are present tense. Conjugated perfective verbs give perfective future. The imperfective future is formed by combining the conjugated future tense of ‘to be’ with an imperfective infinitive. However, the issue of aspectual pairedness, both as a general principle, and for individual classes of verbs, is one that runs deep in Slavic aspectology. See Isaˇcenko (1962, pp. 360–362), Forsyth (1970, pp. 32–58), and Dickey (2000, pp. 44–48) for summaries. PERCEPTIONS OF ASPECTUAL ERRORS Rifkin’s (1995) polling for perception of errors among speakers of Russian found learners’ mistakes in aspect to be in the middle of the spectrum of error gravity. In the subpopulation of Russians living in Russia who were not teachers of Russian (the likely target audience for learners at the Advanced level), errors in aspect ranked 3 on an eight-point scale (1 to 9) of acceptability. Of verb categories—aspect, tense, conjugation, and voice (realized as the presence or absence of a reflexive particle)—aspectual inaccuracy was found to be the least “offensive” error in that population.
HOW ASPECT IS TAUGHT Much of Russian grammar is a matter of memorization: case morphology and environment. Russian aspect, on the other hand, is probably the most abstract grammatical construction for nonnatives and therefore could become a prime target for concept-based instruction (CBI) as has been proposed for issues of aspect in other languages (Negueruela, 2003; Negueruela & Lantolf, 2006). However, the basic privative opposition in which the perfective is marked for wholeness and completion, while the imperfective is unmarked results in corollaries that most students at first find illogical and arbitrary. As a result, most presentations of aspect begin with a general statement about the delimiting nature of the perfective, which, both immediately and over time, is modified with a series of rule-of-thumb amendments, often targeted at various forms of the verb. For example, “Imperfective infinitives are typically used to advise against an action in general . . . . In some contexts it is also possible to use negated perfective verbs, but they typically indicate the possibility of an unintended negative result rather than a negation of the action itself” (Rosengrant, 2007, p. 117). To date only Janda (2003) has proposed a unified treatment of aspect (an analogy between solids and fluids) that approaches something resembling CBI. REVIEW OF THE LITERATURE The 1990s and early 2000s witnessed an expansion of studies of the acquisition of L2 aspect, in both natural and classroom settings. By the late 1980s, evidence began to emerge that supported a model of L2 aspect acquisition based on the lexical aspect hypothesis, in which learners generalized L2 surface morphology to clear-cut aspectual categories of inherent aspect, such as punctual versus durative. For example, Kaplan (1987) showed that college-age learners of French used pass´e compos´e consistently as a perfective but used the present tense to mark non-events as imperfective. Robison (1990) provided further evidence of the “defective tense hypothesis” by showing that ESL learners used target verbal morphology (past tense markers) for punctual verbs and –ing for durative verbs. Anderson (1986, 1991) and Bardovi-Harlig (1992) were early proponents of a similar “aspect hypothesis of acquisition,” which asserts that the semantic categories of verbs based on Vendler (1967) dictate the degrees of aspectual control by L2 learners. Learners for the languages initially
36 studied (English, Spanish) first learn to assign verbs perceived as punctual and telic (e.g., broke) to perfective/preterit categories, while verbs perceived a stative (e.g., had) atelic, and durative (talk, sleep) and dynamic (played) are more quickly assigned to aspects with imperfective meanings. The Vendler classification has undergone further modification. For Russian, Chvany (1990) based adaptations of Vendler to Russian by Kuˇcera (1983) and Brecht (1985) and arrived at a discourse saliency hierarchy ranging from accomplishments/culminations (4) through states (0). Bardovi-Harlig and Reynolds (1995) used the Vendler (1967) framework to show that tutored adult ESL learners acquired English past tense/aspect control in stages according to a verb’s lexical aspect, with correct responses emerging first for accomplishments and achievements. Evidence for the aspect hypothesis has been confirmed across a number of languages to varying degrees (Ayoun & Salaberry, 2008; Comajoan, 2002; Salaberry, 2003; Shibata, 2000). The lexical aspect hypothesis stands in opposition to the discourse hypothesis, in which a verb’s narrative context, not its lexical category, is seen as key to an L2 speaker’s choice of tense and aspect. According to the discourse hypothesis, speakers indicate foregrounding (the main timeline) and backgrounding (the supporting elaboration or evaluation) through verbal/aspect morphology. For example, previous research suggests that L2 speakers use interlanguage tensing as attempts to delineate backgrounding and foregrounding. Kumpf (1984) found significant tensing in the background narration but not in the foreground of a Japanese speaker of English. Flashner (1989) found that Russian speakers of English used past tense for foregrounded information and present tense forms for the background. Bardovi-Harlig (1998) found that foregrounded accomplishments were more likely to be inflected in the past than accomplishments in the background. Finally Wulff et al. (2009) demonstrated support for a constructivist account, one in which the acquisition of an abstract concept like aspect represents a confluence of factors involving input and the cognitive perception of the learner. Specifically it was found that previous studies of learner acquisition of English tense-aspect showed a correlation between speed of acquisition on the one hand and a token’s place in input on the other: its frequency, salience (the form’s morphological visibility), and prototypicality as represented by the degree of telicity. This research echoes earlier
The Modern Language Journal 96 (2012) studies by Ellis (2006, 2008) showing that lexically encoded tense/aspect contexts for input cues in ESL (e.g., temporal adverbs like “yesterday” or “tomorrow”) were more easily learned than contexts in which the tense/aspect encoding was entirely morphological. The research that maps aspect to proficiency is not entirely conclusive, largely due to the small number of proficiency interviews for study. The work cited above (Kumpf, 1984 and Flashner, 1989) suggest that discourse-sensitive tense marking in English appears early on in L2 English speakers, while the earliest use of tense morphology marks lexical aspect as defined by BardoviHarlig (1994). Liskin-Gasparro’s (2000) study of the oral narratives of eight learners of Spanish, ranging from ACTFL Intermediate High (n = 3) through Advanced2 (n = 3), Advanced High (n = 1) to Su´ perior (n = 1), as well as Lopez Ortega’s (2000) study of four Moroccan immigrants to Spain, confirmed elements of both the discourse and lexical aspect hypotheses. In fact, Liskin-Gasparro found four factors that dictated the L2 speakers’ aspectual morphology: the nature of the narrative task, lexical aspect, the role of the narrator in constructing discourse, and the impact of L2 instruction. However, what emerged in the post narrative interviews that Liskin-Gasparro conducted with the participants was their explicit assignment of definitive aspect to certain “safe” lexical items: certain verbs (e.g., erai ‘was’) are always to be encoded in the imperfect and others (e.g., dije, dijop ‘said’) in the preterite (p. 837).3 Such evidence suggests that for Russian lexicalized aspect may play as much a role as the wider lexical aspect hypothesis. Lexical aspect assumes broad categories affecting entire classes of verbs—crucially for English learners of Romance languages in past tense. Russian aspect morphology is not embedded in tense markers but treated in dual verb paradigms. Evidence for lexicalization comes from a number of seemingly punctual verbs such as verba percipiendi videt i ‘to see’ and slyxat ‘to hear,’ which novice and intermediate learners of Russian lexicalize quickly in the imperfective, while the perfective forms uvidet and uslyxat usually appear much later. All in all, learners of Russian quickly amass a small number of verbs corresponding closely to English lexical items. These verbs are classifiable into Vendlerlike categories such as achievements, duratives, statives, and punctuals, discussed earlier, but students are usually given explicit lexical definitions: Achievements (attempt vs. success): sdavat i /sdat p –to take (a test)/to pass;
37
Richard M. Robin i
p
postupat /postupit –‘to apply’/‘to enroll’; rexat i /rexit p –‘to work (on a problem)’/‘to solve.’ Durative vs. punctual: govorit i /skazat p –‘to speak,’ ‘to talk’/‘to tell,’ ‘to say.’ Stative vs. punctual (inceptive): znat i /uznat p –‘to know’/‘to find out;’ ponimat i /pont p –‘to understand’/‘to realize.’
Initially learners acquire additional verb forms as lexical units only, for example, da(te)! p – ‘give,’ dava(te) i !–‘come on!’; idi(te) i !–‘go!’ These are accompanied by past-tense punctuals that form the backbone of memorized narrations that appear at the Novice level: ‘was born’ (rodils, rodilas p ), ‘grew up’ (vyros, vyrosla p ), ‘graduated’ (zakonqil/a p xkolu, universitet), ‘found a job’ (ustroils, ustroilas p na rabotu), and so on. The ubiquity of those and similar verbs suggests that the reason for such high rates of aspect accuracy at the ACTFL Advanced levels owes itself to the number of verbs with lexicalized aspect in a proficiency interview. METHOD Fifty-five interviews with rating ranging from Intermediate High through Superior were analyzed. Thirty-two interviews came from the database of the American Council of Teachers of Russian. Another twenty-two were conducted for this project and were double-rated by ACTFL certified testers. One speech sample came not from an ACTFL interview but from an interview of an American official conducted on a Russian radio station. The interviewee had been rated on the Interagency Language Roundtable (ILR) at S-4+ (beyond ACTFL Distinguished), and the structure of the narration was quite close to that of an Oral Proficiency Interview (OPI). In addition, interviews with two native speakers, conducted by an ACTFL tester according to the OPI protocol, were used as controls for what constitutes normal verb selection and aspect use. The interviews were transcribed. A native speaker coded all verbs for acceptability in aspect. The verbs were then coded for aspectual accuracy, regardless of evidence of lexicalization into correct imperfective choice and correct perfective choice. Correct imperfective choice. The speaker used an imperfective verb in the correct environment. The context might have required imperfective (Prepodavanie men vsegda interesovalo i . ‘Teaching always interestedi me’)or the
speaker might have chosen imperfective although perfective was possible, for example, Mo uqitelnica prixla i smotrela i na mo rabotu. ‘My teacher came and lookedi at my work,’ when perfective posmotrela ‘took a look at’ would have been possible. Correct perfective choice. The speaker correctly used perfective: Oni rexili p v vozraste vosemnadcati let, qto oni hott byt urnalistami. ‘They had decided at eighteen that they wanted to be journalists.’ (The punctual ‘decided’ must be perfective.) As with the category of correct imperfective choice, occasionally both aspects would have been appropriate in a given context: . . . ,Ubistva ne videla, i nikto men ubit p [ubivat i permissible] ne hotel. ‘I never saw any murders, and no one wanted to murderp me.’ (Negation here makes imperfective possible.) Incorrect aspect choice (perfective or imperfective). [Moego druga] vy moete srazu uznavat i → p . ‘You can recognizei → p [my friend] instantly.’ (Modals of possibility generally require perfective nondurative infinitives, especially when modified by adverbs emphasizing instantaneousness.) The texts were then examined for verbs that were likely to have been aspectually lexicalized. Verbs were counted as lexicalized if, overall, speakers were deemed unlikely to have used the other aspect under any circumstances. The reasons for considering an aspectual form to have been lexicalized vary. All lexicalized verbs are of high frequency. Some forms are clearly frozen in early stages of classroom acquisition (e.g., rodils, rodilas ‘was born’ or the morphologically salient vyros/la ‘grew up’).4 Others are assigned explicit English lexical meaning (e.g., postupat i /postupit p –‘to apply/to enroll’). We can classify a number of verb forms as lexicalized because the nonlexicalized member of the pair either does not occur in its assigned lexical meaning in Russian (e.g., stat p ‘to begin’ or ‘to become’ but, stanovits i can only mean ‘to become’) or because the corpus does not reflect its use by Advanced speakers of Russian (e.g., stanovilos i ‘to become,’ perestavali ‘to stopi doing something’). Lexically motivated inaccuracies strengthen the evidence for lexicalization (e.g., Takie ldi vsegda perestali kurit i → p – ‘People like that always had quitp → i smoking’). Moreover, the paucity of inherent semantic content in the less used member of a suspected lexicalization serves as strong evidence. For example, American students quickly associate the past tense
38 perfective zabyl to the colloquial English presenttense form ‘forget’ (“I forget when the test is”), but the imperfective and less frequent zabyval is virtually limited to iterativity. (The progressive meaning ‘was forgetting’ would be hard to come by.) Nevertheless, a researcher’s ultimate decision to declare a verb form or an entire paradigm an aspectual lexicalization is somewhat subjective. For example, for the purposes of this study, the perfective verbs naqat ‘to begin’ and zakonqit ‘to end’ were not included in the list of lexicalized verbs despite the fact that there are no uses of the imperfective members of these pairs. The fact remains that native speakers often use both in apparently punctual meanings (e.g., Kako universitet ona zakanqivala i ? ‘What college did she graduate from?’). It was decided that a conservative approach would serve best. A more forceful approach would have dictated the inclusion of more verbs and would have resulted in even higher correlations in the data. LEXICALIZED IMPERFECTIVE VERBS Following the principles laid out above, verbs were considered to have been lexicalized if they belonged to any of the following categories: (a) truly biaspectual verbs; (b) common activities, which would have been assigned lexical meanings early on in acquisition; and (c) commonly occurring states of mind, many of which introduce subordinate clauses. The following list provides an additional foundation for considering such forms to have been lexicalized. All Truly Biaspectual Verbs This corpus includes, among others, suwestvovat ‘to exist,’ ispolzovat ‘to make use of,’ issledovat ‘to investigate,’ prepodavat ‘to teach,’ zavedovat ‘to be in charge of,’ imet ‘to have,’ reklamirovat ‘to advertise,’ borots ‘to struggle,’ razgovarivat ‘to converse,’ oidat ‘to expect,’ kontrolirovat ‘to check,’ ‘to control,’ immigrirovat ‘to immigrate,’ and organizovat ‘to organize’ (although ogranizovyvat exists as an iterative). Verbs With Common English Meanings Assigned to Imperfective The cases where lexicalization is limited to a single tense are noted. Lexicalized past tenses are given in the masculine.
The Modern Language Journal 96 (2012) Activities. ovorit ‘to talk,’ ‘to speak.’ The perfective pogovorit is a durative for ‘to talk’ and appears only occasionally. The perfective skazat can mean ‘to talk about something’ (e.g., Ob tom skau p potom–‘That’s something I’ll talkp about later’), but its use in that context is infrequent and does not appear in the corpus. (See the discussion of lexicalzied perfective verbs.) Rabotat ‘to work.’ Both possible perfectives, porabotat ‘to work a bit’ and prorabotat + time period in accusative ‘to work a quantity of time’ are delimited duratives. One instance of porabotat occurred in the interviews (Advanced Mid). Chloe,5 an ESL teacher, tells how few people showed up to her first class: Qerez, moet byt, 15 minut ewe odin molodo qelovek prixel. My qut-qut porabotali p , a potom ih ot–otpuskala i → p I to . . . nu, koneqno, interesny pervy den . . . na rabote. ‘And maybe about fifteen minutes later, one more young man arrived. We worked a bit, and then I le–was letting themi → p them go.’ Zanimats ‘to study,’ ‘to do homework,’ a ‘to be involved in.’ The less frequent perfective zants means ‘to take up’ (an interest, hobby, activity, etc.) and did not appear in the interviews. Uqits ‘to be in school.’ Outside of its English lexical assignment, uqits i /nauqits p ‘studyi /learnp ’ did not occur in the interviews. Izuqat ‘to take a course in. . . . ’ The assigned lexicalized imperfective meaning is used for academic disciplines. Non-lexicalized aspectual choice is represented by the activity/achievement pair izuqat i /izuqit p ‘to make a study of . . . i /to complete a study of. . . . p ’ Uqit ‘to learn’ is a synonym to izuqat. Delat ‘to do’ in most qto ‘what’ questions, especially in the infinitive and future tense, e.g., Qto delat? ‘What is to be done?’ Qto X budet delat? ‘What is X going to do?’ Qtointerrogatives with the perfective sdelat are low-frequency with specialized semantic weight (e.g., What will X manage to get done?) and did not occur in the interviews. States of Mind, Verbs of Perception (Verba Percipiendi), Quotes. Dumal ‘thought,’ especially introducing dependent clauses. Only Superior speakers used the rarer punctual or semelfactive podumal, which usually signals a subsequent reevaluation: Nu, snaqala podumala p , qto podam na politologi. ‘At first I thought that I would apply for political science.’6 An Advanced Mid speaker resorted lexically determined dumala i when podumala p
39
Richard M. Robin was required: dumala i (should be podumala p ), qto oni umerli p , potomu qto oni oqen medlenno ili praktiqeski sovsem ne dvigalis i v vode. ‘I thought they had died since they were movingi very slowly in the water or practically not at all.’ Kazalos ‘it seemed’: The perfective past pokazalos is semelfactive ‘(at first) it seemed.’ Sqital(, qto) ‘was of the belief (that . . . )’ The perfective sql is a low-frequency form and almost always used in combination with an adjective (e.g., sql nunym ‘considered it necessary’). This form was absent from the corpus. Znal ‘knew.’ The lexicographically assigned perfective uznat means ‘to find out.’ Hotel ‘wanted.’ The lexicographical past tense perfective zahotel is limited to an inceptive or individuated meanings in the past tense, except when negated, when it means ‘did not seem to want.’ Zahotet as an individuated past tense occurred only in one Advanced Mid interview: Mne bylo oqen trudno, kogda pereehala v Tehas, kogda prosto zahotela p kofe zakazat i stola v oqeredi dest minut poka ldi peredo mno boltali o qem-to. . . . ‘It was hard for me when I moved to Texas, when I just would wantp to order some coffee and I’d be standing in line for ten minutes while people in front of me were gabbing about something. . . . ’ Lbit ‘to love.’ The perfective inceptive polbit ‘to come to love’ was found in one Advanced Mid interview: Mono . . . mono ustraivat takie festivali, takie sobyti, kotorye by privlekli takih lde, moet byt na nedel ili na dve nedeli. I postepenno mono nadets, qto oni polbt p gorod i zahott tam it. ‘One could organize festivals, events that would attract people, maybe for a week or two weeks. And you could gradually hope that they would come to lovep the city and want to live there.’ Cobirats ‘to plan’ (to do something). The perfective sobrats means only ‘to congregate.’ Slyxal ‘heard (tell that).’ Uslyxalp is analogous to uvidelp ‘caught sight of.’ Verbs of sound/sight perception have their own set of aspectual peculiarities (Lubensky, 1984), which appear to be emerging at the Advanced level, at least for videt/uvidet ‘to see.’ On the one hand, the perfective of that verb appears in four interviews, three times correctly in the infinitive and once (questionably) in the past tense—enough to suggest that videt, if lexicalized at earlier levels of proficiency, was now undergoing grammaticization. On the other hand, beyond the meaning of ‘heard tell,’ imperfective slyxal ‘was able to
hear’ occurs twice, both times, in Superior level interviews. That factor suggests that for Advanced speakers, slyxali , in its punctual meaning, is lexicalized. Bols ‘was afraid,’ especially introducing dependent clauses, e.g., Rhoda (Intermediate High): bolas i , qto . . . ne zna, ne dumala, qto oni hoteli problemy ili . . . no prosto oni oqen zlye i to ponima i ponla togda . . . ‘I was afraidi that . . . I don’t know; I didn’t think they wanted problems or . . . It’s just that they were really mean, and I now realize, and I realized then . . . ’ Volnovats ‘to worry.’ The perfective zavolnovats is inceptive only and absent from the corpus. It should be noted that the imperfective verbs of perception that readily introduce dependent clauses (e.g., Oni sqitali, qto–‘They believed that . . . ’; my dumali, qto. . –‘we thought that . . . ’) contrast with the lexicalized perfective skazat ‘to say’ (discussed in depth below). A student who had not lexicalized these forms into the imperfective could be expected to analogize perfective skazat ‘to say’ to verbs of perception and produce utterances such as Oni podumali p , qto . . . ‘they believedp that. . . . ’ The rare occurrence of such perfective forms suggests that students are acquiring each such common verb on a case by case basis, rather than constructing an entire aspect system based on analogy apart from lexicalized assignment of meaning.
Lexicalized Perfective Verbs Skazat in all forms is deeply encoded as ‘to say’ or ‘to tell’ in English speakers of Russian, most likely because the pair govorit/skazat is suppletive. The lexicalization of skazat becomes clear through numerous errors from Advanced Mid speakers. Chad: Oni budut skazat. ‘They will say.’ (Imperfective formant budut must take an imperfective infinitive govorit.) Margaret: Mo uqitelnica prixla i smotrela na mo rabotu i skazala Margaret, ty qto delaex? Dime pomoga, vot i vse. I ona skazala, Margaret, ty ponimaex, qto to ego domaxn rabota, a skau p → i(present) , o izvinite, ne znala, ne znala! ‘ My teacher came and looked at my work and said, Margaret, what are you doing? I’m just helping out Dima. And she said, Margaret, you do realize that that’s his homework? And I will sayp → i(present) , Oh, I’m sorry. I didn’t know, I didn’t know!’
40 Reese, in a roleplay, explaining that her rental car was defective: Vse ravno! Ne rabotaet. Vam ne nado skazat p → i , kak dolen rabotat. ‘It doesn’t matter! It doesn’t work. You shouldn’t tellp → i me how it’s supposed to work.’ (Negated infinitives in modal constructions that are equivalent to a negative imperative are with a few exceptions imperfective.) Colleen, describing sporadic television news reports on an ongoing terrorist attack: vklqila televizor, tam bylo niqego. Inogda bylo pt minut, oni qto-to skazali p → i . Tolko potom stalo izvestno qto sluqilos. ‘I turned on the TV. There was nothing there. Sometimes it was five minutes, they saidp → i (would say) something.’ For ‘said’ we see a grammaticized aspectual choice made in Erich’s Superior-level speech: My govorili i postonno, qto tot konflikt putem voenno sily ne rexit. ‘We have saidi consistently that this conflict cannot be solved by military force.’ Erich goes on to talk about previous statements on U.S.-Russian relations: Oba kandidata ue dovolno sno govorili i , qto oni ponimat vanost horoxih konstruktivnyh otnoxeni s Rossie. ‘Both candidates have saidi clearly that they understand the importance of constructive relations with Russia.’ In a semantically easier situation, Advanced Mid Brett opts for ‘said’ as imperfective (perfective would have been permitted as a punctual) complaining about a malodorous car interior. Kak govorili , ne hoqu alovats, prosto s syrom nesereznoe delo, no est bolxe sereznyh del s to maxino. ‘As I was sayingi I don’t want to complain, it’s just that this cheese [smell] is not a big deal, but there is more serious business with this car.’ The presence of the lexicalized perfective skazat (past, future, infinitive, imperative) is pervasive. It accounts for 59.72% of all lexicalized perfectives in the corpus. However, no correlation was found between the incidence of skazat as a lexical item and changing levels of proficiency. Lexicalized skazat occurred evenly across proficiency levels. Biographical Lexical Items Rodils ‘was born’ is usually learned early as part of a memorized spiel. The imperfective rodats can be used only in durative and iterative meanings and is not part of this corpus. Vyros ‘grew up’ is also learned lexically as part of a life story spiel. The imperfective ros means ‘was in a state of growth’ or is iterative and was not part of this corpus. Consider Penelope’s (Interme-
The Modern Language Journal 96 (2012) diate High) improper use of lexicalized vyrosli as ‘they matured,’ most likely in an iterative sense. Penelope, older than her peers, attended a graduate program long enough to see several sets of undergrads: Potom, poka seqas v aspiranture studenty tam vyrosli p i oni dobrye, dobrye so mno. ‘Then, while I’m now in graduate school, the undergrads have grown up there, and they’re kind, kind with me.’
The interviewee’s meaning is further obscured by her use of the temporal while for the intended logical while it may be: “So while it may be I’m now in grad school, the undergrads were [always] getting more mature.” Evaluatives/Quotes Rexil (, qto . . . )—literally, ‘decided (that . . . ).’ In fact, the sense of this lexicalized perfective is ‘concluded that . . . ,’ ‘thought that . . . ’ and is often synonymous with the imperfective dumal, qto ‘thought that,’ except that rexil can be argued to have a summative, Akzionart. Beyond this special past tense meaning rexit in the sense of ‘to arrive at a decision’ and ‘to solve a problem’ were also counted as a separate lexicalized verb. In each of these cases, the imperfective rexat (when not iterative) represents an attempt. That meaning did not occur in the corpus. Pont, especially in the past tense, is lexicalized as ‘to realize.’ In those contexts the speakers use it with consistent accuracy. This is not true of the meaning ‘to understand,’ commonly assigned to both aspects. Consider Margaret’s (Advanced Mid) correct and incorrect use of the verb in the same paragraph: Amerikanskoe detstvo ne prigotovilo men k tomu, qtoby vybrat professi v tom vozraste, i ponla p , qto togo ne hotela. [After a few seconds talking about cultural differences] Cnaqala, oqen malo ponla p → i . mogla govorit po-russki posle neskolkih mescev, no kulturu ne sovsem ponla p → i . My American childhood had not prepared me to choose a profession by that age. And I realizedp that that was not what I wanted. [After a few seconds talking about cultural differences] At first, I understoodp → i very little. I could speak Russian after several months, but I didn’t understandp → i the culture at all.
Sumet is lexicalized to ‘manage to.’ The imperfective umet ‘to know how’ was absent from the corpus.
41
Richard M. Robin Beginning and Ending Perestat + infinitive ‘to stop doing something’ is usually acquired relatively early. The imperfective perestavat is a low-frequency item, absent from the corpus, and practically limited to an iterative meaning or, more rarely, nullified action or action disconnected with the present. Stal + infinitive as a synonym for naqat p ‘to begin’ is lexicalized. The imperfective stanovils can only mean ‘was becoming’ or ‘used to become.’ Stat + instrumental for ‘become + profession’ (e.g., stal vraqom ‘became a doctor’) was also counted as a lexicalized perfective because contexts for stanovits i + profession is rare; rarer even than relatively low-frequency expressions such as stanovits qelovekom ‘to become a person.’
Resultative State Resultative state verbs, as categorized by Forsyth (1970) and Comrie (1976), are often rendered in English by alternative tenses and/or parts of speech. Privyk ‘got used to’ is often lexicalized to ‘is used to.’ Umer ‘died’ is often lexicalized as ‘dead.’ The imperfective must be progressive (was dying) or iterative and usually plural, e.g., umirali ‘people died/were dying.’ Ustal ‘got tired’ is usually lexicalized as ‘tired,’ as evidenced by the frequent L2 error ∗ byl ustal (an attempt to translate was tired literally, which, in fact, is ‘was got tired’—not encountered in this corpus.) Zabyl ‘forgot’; the imperfective zabyval must be iterative.
Punctuals Vernuts is lexicalized as ‘to return.’ Vozvrawats i not recorded in the corpus. The attempted use of vernuts in the present tense (required imperfective) is a common mistake in beginners and occurs in this sample of Advanced Mid speech: Mo sosedka po komnate obyqno vernulsi → i,present domo v dva–v dvuh ili treh. ‘My roommate usually will get backi → i,present home at two– in two or three.’ Poloit is lexicalized as ‘to put,’ ‘to set.’ The imperfective suppletive form klact (usually acquired later) is not present in the corpus. A common error in the Russian of beginners is the substitution of perfective future polou ‘I will
put’ for present tense kladu, but that error was not evident in the corpus. Poznakomits, especially in the past ‘to get acquainted.’ The imperfective znakomits does not occur in the interviews. Sluqilos ‘happened,’ especially in the phrase qto sluqilos ‘what happened.’ The imperfective sluqalos is rare. Accomplishments. The most common accomplishments are lexicalized to specific English verbs, which distinguish them from the imperfective attempts. Nati ‘to find.’ The imperfective nahodit is not frequent outside the present tense (Shteinfel’dt, 1963, as cited by Forsyth, 1970, p. 48). Nati can also be considered the achievement (successful outcome) of iskat i ‘to seek.’ Evidence of ossified nati is found in Casey’s Advanced Mid interview when she describes a computer center in Russia: Esli my govorim o politike, esli to bylo oqen legko ispolzovat internet i nati p → i svobodnye komptery. ‘If we’re talking about politics, it was very easy to use the Internet and findp → i available computers.’ Postupit ‘to enroll,’ ‘to get into (an educational institution).’ The imperfective is usually lexicalized to ‘to apply.’ Vyleqit ‘to cure.’ The imperfective leqit is usually lexicalized as ‘to treat.’ Uznat is lexically assigned to ‘to find out.’ The secondary imperfective uznavat does not occur in the corpus. Happenstance. Happenstance or “accident” verbs are naturally lexicalized in the perfective. Potert ‘to lose (an object)’ in the classroom is almost always acquired first in the past tense. The imperfective tert comes much later (and was not present in the corpus). Poluqits (future tense) can be assumed to be lexicalized to ‘it will work out.’ The imperfective budet poluqats is a low-frequency, mostly iterative form, which does not appear in the interviews. Popast ‘to end up at,’ ‘to arrive at.’ The imperfective popadat is fairly rare, almost always likely to denote repeated action in the past and future, and absent from the interviews. Imperatives Most neutral affirmative imperatives occur in the perfective. Izvinite and prostite ‘forgive me/pardon me’ follow that rule.7 However, both occurred often enough and are of such high
42 frequency that they were counted as lexicalized items. (Their imperfective counterparts in affirmative commands are virtually nonexistent.) Excluded Verbs All other verbs outside the present tense were viewed as subject to non-lexical aspectual choice with a few exceptions: Forms repeated from the interviewer (Uptake). Interviewer: zna, qto dl togo, qtoby popast v takie agentstva nado proti dovolno strogi process dl dostupa. Vy prohodili i tot process? ‘I know that to get into agencies like that you have to go through a fairly rigorous security clearance. Did you go throughi such a process?’
Interviewee: Prohodila i , prohodila i , i oqevidno. ‘I went through it . Obviously I went through iti .’ Badly misformed verbs or significantly lexically or semantically inappropriate verbs. When improper morphology clouded the question of the speakers’ aspectual choice, the verb was excluded, e.g., mogu ∗ kipt ? [acceptable: prokiptit p ] kiptit i /vskiptit p , vodu.‘I can boil water’; mne povezlo p [required: udalos p ] kupit maxinu. ‘I was luckyp [required: managedp ] to buy a car.’ Unprefixed verbs of motion. Unprefixed verbs of motion (mostly “go”) have an additional aspectual layer of multi/unidirectionality. For this reason, they were excluded from the data. In fact, they accounted for 161 of 6,178 verbs in the corpus (2.6%). The overall use of unprefixed verbs of motion produced 28 mistakes (17.4% of verb of motion attempts, and 5.6% of all aspectual errors). The proportion of verbs of motion in the corpus is significantly different from zero (t = 7.22, df = 54, p = 0001), but qualitatively small, especially when compared to other verbs. For example, skazat p ‘to say’ represents 57.83% of lexicalized perfective verbs throughout the corpus. Such figures suggest that the quicksand represented by unprefixed verbs of motion, at least at the Advanced level, may be overplayed in pedagogy. DATA AND ANALYSIS A preliminary glance at the data in Table 1 suggests a number of trends. 1. As the proficiency of speakers increases, the proportion of true aspectual choices (non-
The Modern Language Journal 96 (2012)
Total
IH
6
84.47
Percent
correct imperfectives and perfectives (all types)
AL AM AH S All
5 27 9 8 55
89.36 90.62 91.97 91.98 89.93
of aspectually valid utterances.
Lexicalized verbs
IH AL AM AH S All
6 5 27 9 8 55
46.82 41.73 41.01 37.50 26.68 39.40
Percent of all verbs used by all students are lexicalized.
lexicalized forms) increases. In imperfective verbs this figure goes from just over 28% in Intermediate High speakers to more than 57% in Superior speakers. For perfective verbs, non-lexicalized forms make up a greater proportion of the verb forms used by all speakers. Nevertheless, a steady increase throughout the proficiency scale is observable, from just over half the verbs in Intermediate High speakers to almost three-fourths of the verbs in Superior speakers. 2. On the one hand, for both aspects, the reliance on lexicalized verbs drops steadily from just under half in Intermediate High speakers to 37.5% in Advanced High speakers. On the other hand, the greatest single decrease in the reliance in lexicalized verbs—10 points—is seen at the Advanced High to Superior borderline. 3. Aspectual accuracy, whether for the corpus taken as a whole or specifically for non-lexicalized verbs, increases steadily from Intermediate High to Advanced High but plateaus at the Advanced High—Superior borderline—except in the case of true choices involving the imperfective. In the tables that follow, we examine the significance of the correlations between proficiency value (the ACTFL proficiency rating) and decrease of use of lexicalized verbs. In lower levels of proficiency (Proficiency Value), more lexicalized verbs (imperfectives and perfectives) are used than at higher levels of proficiency. Correlations were computed between Proficiency value (rating) and the proportion that each of the above quantities represents of the number of total verbs used in the corpus (6,178). Table 2 documents that correlations between the number of lexicalized imperfective and perfective verbs and proficiency rating are not
43
Richard M. Robin TABLE 1 Use Verbs in Corpus (All Levels Together) Totals
Prof.
N
%
Correct choices made up n% of all uses of imperfectives (right and wrong)
IH AL AM AH S All IH AL AM AH S All IH AL AM AH S All IH AL AM AH S All IH AL AM AH S All IH AL AM AH S All
6 5 27 9 8 55 6 5 27 9 8 55 6 5 27 9 8 55 6 5 27 9 8 55 6 5 27 9 8 55 6 5 27 9 8 55
28.23 37.28 39.13 44.35 57.63 40.54 50.25 58.97 63.79 70.86 73.52 63.34 71.19 74.77 83.25 87.26 90.83 81.47 66.34 85.87 88.65 94.08 93.62 87.66 91.15 89.62 92.99 94.01 94.83 91.88 89.82 90.56 92.35 92.78 93.02 91.78
Proficiency Value
Frequency
Percent
Valid Percent
Cumulative Percent
6 7 8 9 10
6 5 27 8 9 55
10.9 9.1 49.1 14.5 16.4 100.0
10.9 9.1 49.1 14.5 16.4 100.0
10.9 20.0 69.1 83.6 100.0
Correct choices made up n% of all uses of perfectives (right and wrong)
Correct choices made up n% of all uses of imperfectives (right and wrong)
Correct choices made up n% of all uses of perfectives (right and wrong)
Total correct imperfectives including lexicalized imperfectives
Total correct perfectives including lexicalized perfectives
Comment When students opted for a correct imperfective verb, they were making a real choice this percentage of the time.
When students opted for a correct perfective verb, they were making a real choice this percentage of the time.
Students making a real choice in choosing imperfective produced aspectually valid utterances this percent of the time. Students making a real choice in choosing perfective produced aspectually valid utterances this percent of the time. Students choosing imperfective (whether a choice or lexical) aspectually valid utterances this percentage of the time. Students choosing imperfective (whether a choice or lexical) aspectually valid utterances this percentage of the time.
TABLE 2 Proficiency Levels of Students in the Analysis ACTFL Proficiency Rating IH AL AM AH S Total
significant. However, the correlation of the proportion of each subject’s verbs that are lexicalized imperfectives or lexicalized perfectives is large (r = –.450) and significant (p = .001). This confirms the hypothesis that as proficiency ratings in-
crease, the proportion of lexicalized verbs (both imperfective and perfectives) decreases. Importantly, the proportion of lexicalized imperfectives decreases at more than twice the rate of lexicalized perfectives (–.396 vs. –.183).
44
The Modern Language Journal 96 (2012)
TABLE 3 Correlations
Proficiency rating
∗∗
Lexicalized Imperfective
Lexicalized Perfective
Lexicalized total (LEXTOT)
APROPLE XIM
APROPLE XPF
APROPLE XTOT
.006
.107
.062
−.396∗∗
−.183
−.450∗∗
.967
.438
.653
.003
.181
.001
55
55
55
55
55
55
Pearson Correlation Sig. (2tailed) N
Correlation is significant at the 0.01 level (2-tailed).
TABLE 4 Correlations Between Correct IM, PF, Total of Both IM and PF, and the Proportions That Each are of Total Verbs Correct Proportion Proportion Proportion aspectual of Correct of Correct of Correct choices Imperfective Perfective Aspectual Choices Correct Correct CORRC APROP APROP APROP imperfective perfective TIMPF CORRECT IM CORRECT PF CORRECT IMPF Proficiency Pearson rating Correlation Sig. (2tailed) N ∗∗
.534∗∗
.466∗∗
.507∗∗
.487∗∗
.404∗∗
.516∗∗
.000
.000
.000
.000
.002
.000
55
55
55
55
55
55
Correlation is significant at the 0.01 level (2-tailed).
TABLE 5 Correlation Between Non-lexicalized Imperfective/Perfective and Proportion of Total Verbs That Are Imperfective and Perfective With Proficiency Rating
Proficiency rating
Pearson Correlation Sig. (2tailed) N
Imperfective
Perfective
Imperfective as a total portion
Perfective as a total portion
.530∗∗
.424∗∗
.522∗∗
.272∗
.000
.001
.000
.045
55
55
55
55
∗∗ ∗
Correlation is significant at the 0.01 level (2-tailed). Correlation is significant at the 0.05 level (2-tailed).
It was hypothesized that as proficiency level would correlate positively with the number of verbs requiring true aspectual choice. Tables 4 and 5 document a strong relationships between both the absolute number and the proportion of non-lexicalized verbs against proficiency level.
DISCUSSION The method of solicitation of narratives has been demonstrated to influence the use of discourse type, verbal aspect and surface morphology (Bardovi-Harlig, 1992; Kumpf, 1984; Selinker & Douglas, 1985, 1994; V´eronique,
Richard M. Robin 1987). The OPI protocol almost always includes a ritualized autobiography that requires lexicalized perfectives rodils ‘was born,’ vyros ‘grew up,’ stat ‘to become,’ as well as pereehal ‘moved,’ and zakanqival/okonqil universitet ‘graduated,’ both of which are likely to have been lexicalized, usually correctly. In short, speakers’ aspectual control during an OPI is likely in part to stem from at least five minutes of floor time in each interview with high-density past tense and infinitive verb content that they have sufficiently vetted for aspect so as to be near perfect. On the other hand, a common initial probe to Advanced is a description of daily routine, often in past time. (“So what was a typical day like when you were living in Omsk?”) Not surprisingly, the answer to this question normally requires an imperfective sequence of events8 and therefore leads L2 speakers into aspectual badlands, replete with improperly lexicalized perfectives: Rhoda (Intermediate High): Obyqno vstala p → i rano, okolo vosmi qasov, i my s sosedkami inogda my eli i , no obyqno my pili i tolko qa, koneqno. ‘I usually had gotten upp → i early, around 8:00, and my roommates and I, sometimes we atei , but usually we dranki only tea of course.’ Evan (Advanced Mid): Vesno poqti kady den vstalp → i oqen rano. ‘In the spring, nearly every day, I had gotten upp → i very early.’ Lois (Advanced Mid): , navernoe, vstala p → i oqen rano . . . obyqno priehala p → i v universitet okolo devti, byla tam ves den, zanti, vot ti soqineni studenqeskie, delala i svo rabotu, navernoe poxla p → i v sportzal ili na urok po baletu i obyqno priehala p → i domo okolo 9 ili 10 veqera i potom nado bylo uin gotovit i . . . ‘I probably had gotten upp → i very early and usually arrivedp → i at the university around 9:00, was there all day. My classes . . . I did my work, these student compositions. Then I probably had gonep → i to the gym around 9:00 or 10:00 in the evening. And after that, I would have to preparei dinner.’ Verbs such as vstat p ‘to get up’ are simplex verbs of action. Durst-Anderson (1992) called that “surprising” (p. 59). Perhaps perfective simplex verbs are more subject to lexicalization. Such verbs (others are dat ‘to give,’ vernuts ‘to come back,’ and kupit ‘to buy’) are almost always learned first in perfective contexts. The assumptions made here about which verbs L2 Russian speakers lexicalize and the supporting
45 data analysis strongly suggests that aspectual lexicalization goes a long way in explaining high accuracy rates among Advanced and even Intermediate High speakers of Russian. But the data also show that speakers at this level monitor the use of aspect, both for lexicalized and non-lexicalized verbs, as the occasional case of aspectual repair shows. Brett (Advanced Mid): v buduwem budu opt v Rossii, vot, navernoe, budu postupat i . . . postupit p → i v aspiranturu. ‘In the future I’m going to probably applyi . . . have gotten intop → i graduate school again.’ (Brett repairs the correct aspect to the wrong aspect.) Louise (Advanced Mid): Vy by ?dumali i → p preferrred, to Kep Kod, no my ue delali i , sdelali p nekotorye izmeneni v tomu domu. ‘You might thinki → p preferrred that it was Cape Cod, but we had already been makingi , had already madep several changes this year.’(It is unclear whether Louise’s repair was to correct the time frame or whether the speaker decided that the adverb ue ‘already’ made the perfective a safer bet.) Indeed, in a number of cases involving nonlexicalized aspect (true choice), we cannot determine with certainty whether or not the speaker “hit” the target; in some situations we can assume that speakers may have gotten the aspect right without realizing what the attending meaning was. Narrative often provides a wide choice of free variation. In some cases, the choice of aspect is almost inconsequential. (Consider the English example “I saw you run down the street” vs. “I saw you running down the street.”) In other situations, aspectual choice is more semantically laden, but we cannot, without further analysis beyond the scope of this article, demonstrate that learners actually were saying what they intended. In a few cases, knowing what we know generally about a speaker’s level, we can guess that an utterance came out correctly accidentally. Consider Penelope (Intermediate High), who tells about friends who left the old country and now send money back home: zna celu sem, kotory lde uezali i iz stranu. ‘I know an entire family whose people had lefti the country.’ The speaker mostly likely meant uehali p ‘left (for good)’ rather than the imperfective of nullified action ‘had left temporarily.’ The linguistic competence indicated by the use of the imperfective of nullification in a telic verb is unexpected in Intermediate High speakers. But such instances were few and unlikely to have skewed the data.
46 Implications for Pedagogy The confirmation of the hypothesis that lexicalization diminishes steadily with progress through the Advanced level suggests that the acquisition and increased use of a variety of verbs lies at the heart of actual demonstrated control over aspect. At the same time, aspectual problems never disappear entirely as borne out by clear errors, even at the highest level. Consider the utterance of a speaker with an ILR rating of 4+ who said On zapretilp? nam, otec vsegda nam ∗ zapretilp → i skazat p to slovo. ‘He forbadep? us, always forbadep → i us to say that word.’ At lower levels of proficiency, however, we should consider whether our results call for a major overhaul in the teaching of aspect or some minor adjustments. For Romance languages, Blythe (1997) called for a revision of the teaching of aspect based on constructivist principles; similarly Negueruela (2003) and Negueruela and Lantolf (2006) have recommended replacing aspectual rules of thumb with a comprehensive CBI approach. However, two considerations serve as a counterargument: (a) surface use of aspect, even at Intermediate High, is largely in place; and (b) target Russian audiences perceive aspectual errors as the least grave of all errors involving verbs. In short, we seem to be doing something right. However, because the most common aspect errors are likely to involve improper lexicalization of specific verbs—mostly simplex perfective verbs in past tense narration about repeated events—we would do best to concentrate our efforts on them, starting in the Intermediate range, when memorized chunks give way to free-standing sentencelength utterances or to longer memorized paragraphed chunks in which verbs denoting repetition are “frozen” in place. I am aware that chunking that smacks of memorization as students move from Intermediate to Advanced is suspect. However, I agree with Rifkin (2003) and Shekhtman and Sibrina (2005) that the preparation of rehearsed texts can serve as one of a number of tools to raise proficiency, even at higher levels, in this case, with correct, generalizable aspect. Furthermore, one could make the argument that because Advanced speakers of Russian prefer durative verbs in the imperfective even where a narrative perfective (aorist—see Timberlake, 19859 ) is appropriate (e.g., proil tam pt let ‘lived there five years’), that such uses be taught explicitly. However, an equally valid argu-
The Modern Language Journal 96 (2012) ment can be made that such explicit instruction is not the most effective use of classroom time given the fact that the learner preference for the lexicalized imperfective rarely produces aspectually invalid utterances. This last issue goes to the crux of the question surrounding CBI versus rule of thumb approaches to presenting aspect, touched upon at the beginning of this article. A successful concept-based approach to aspect presentation has the potential advantage of linguistic truth: loyalty to it should produce no aspectually invalid utterances. However, failure to grasp the concept, which in Slavic languages covers far more formal territory than in Romance languages, is likely to result in a great number of aspect errors. Rules of thumb may atomize Russian aspect for the learner resulting in something less than linguistic truth. However, in terms of proficiency, it is ultimately performance, not competence, that counts. A view of aspect instruction must ask how many contextually valid utterances are produced, not how many valid utterances could be produced—a tantalizing subject for future research. CONCLUSION The evidence presented suggests that, as students begin to produce paragraphed speech in Russian, the aspectual accuracy of what they say, while quite high, is less a matter of aspectual choice than of common lexicalized vocabulary— nearly half of all aspectual choices were lexicalized in Intermediate High speech. The use of lexicalized vocabulary trends downward as proficiency level (and the number of available verbs) increases. More “decisions” are apparent in perfective verbs because durative and activity verbs, at least in oral proficiency, are so prevalent. Yet even at Advanced Mid, 41% of aspectual choices were found to be a matter of vocabulary, not aspect grammar. The widespread lexicalization of verb forms goes a long way in explaining the surprising rate of accuracy in a grammatical category that is perceived as the bane of the English-speaking learner of Russian. At the same time, it gives us reasons to keep on searching for evidence pertaining to the actual aspectual competence of learners of Russian.
ACKNOWLEDGMENTS This research was made possible by International Research and Studies Grant to the National Capital Language Resource Center, Award number P017A050036–06.
Richard M. Robin NOTES 1 Long and Gor (2009) consider a correct response rate of 80% on perception production tests or higher an indication of mastery. 2 Advanced Low was not added until 1999. 3 Superscripts “p” and “i” are used for perfective (preterite) and imperfect(ive), respectively. In quoted passages from OPIs, the superscript notation “i → p” means “imperfective used where perfective is required.” The verbs English translations are similarly marked and where possible are rendered to reflect the inappropriate aspect. 4 The past tense is normally marked by -l [l] plus a gender marker (zero for masculine). Vyros [‘vÈr´s] ‘grew up’ (masculine) is one of a number of verbs, where in the masculine form, the [l] is absent. 5 This research plan was approved by the George Washington University’s Institutional Review Board for Title 45 of the Code of Federal Regulations, Part 46. In accordance all names and identifying details have been changed. Pseudonyms were assigned after they were randomized. However, the pseudonyms accurately reflect the speaker’s gender. 6 In fact, an argument is to be made that podumat in this past tense context is a semelfactive, a single rapid occurrence of an action (cf. prygnut ‘to jump once’). Moreover, nullified result of podumat here resembles the result state of imperfective “two-way” (Forsythe, 1970) verbs. 7 It can be argued that other ritualized imperatives such as vstava(te) ‘get up,’ sadis, sadites ‘have a seat,’ lois, loites ‘lie down,’ zahodi(te) ‘come on in,’ and razdevas , razdevates ‘take off your coat,’ all of which are exhortatives, are lexicalized to the imperfective in L2 speech, but their absence from the corpus renders the point moot in this study. 8 Under certain circumstances a series of perfective future verbs is possible. The use of the perfective future as an equivalent of the English “would” imperfective is rare and characteristic of near-native control (e.g., Prekrasno vs bylo: utrom vstanu, podu na pl kupats . . . ‘It was so wonderful: I’d get up in the morning and go down to the beach for a swim . . . ’). No such occurrences are part of this corpus. 9 Here “aorist” refers to use of the perfective in the meaning of the simple past tense, in which the aspectual contexts that promulgate the use of imperfective (duration, repetition) are not relevant, while the basic perfective meaning of completed action is deemphasized. Russian inherited a true aorist tense from Common Slavic (and from earlier Indo-European) but gradually lost it during the twelfth century.
REFERENCES American Council on the Teaching of Foreign Languages. (2012). ACTFL Proficiency Guidelines— Speaking (2012). Retrieved January 2, 2009,
47 from http://www.actfl.org/files/public/ACTFL ProficiencyGuidelines2012_FINAL.pdf Anderson, R. (1986). El desarrollo de la morfolog´ıa en el espa˜ nol como segundo idioma [The development of morphology in Spanish as a second language]. In J. M. Meisel, (Ed.), Adquisici´on del lenguaje/Aquisic¸a˜ o da linguagem[Language acquisition] (pp. 115–138). Frankfort: Vervuert Verlag. Anderson R. (1991). Developmental sequences: The emergence of aspect marking in second language acquisition. In T. Huebner & C. A. Ferguson (Eds.), Crosscurrents in second language acquisition and linguistic theories. Amsterdam: Benjamins. Ayoun, D. & Salaberry, M. R. (2008). Acquisition of English tense-aspect morphology by advanced French instructed learners. Language Learning , 58, 555–595. Bardovi-Harlig, K. (1992). The relationship of form and meaning: A cross-sectional study of tense and aspect in the interlanguage of learners of English as a second language. Applied Linguistics, 16 , 253– 278. Bardovi-Harlig, K. (1994). Anecdote or evidence? Evaluating support for hypotheses concerning the development of tense and aspect. In E. Tarone, S. Gass, & A. Cohen (Eds.), Research methodology in second language acquisition (pp. 41–60). Hillsdale, NJ: Erlbaum. Bardovi-Harlig, K. (1998). Narrative structure and lexical aspect: Conspiring factors in second language acquisition of tense–aspect morphology. Studies in Second Language Acquisition, 26 , 471–508. Bardovi-Harlig, K. & Reynolds, D. (1995). The role of lexical aspect in the acquisition of tense and aspect. TESOL Quarterly, 29 , 107–131. Blyth, C. (1997). A constructivist approach to grammar: Teaching teachers to teach aspect. Modern Language Journal , 81, 50–66. Brecht, R. D. (1985). The form and function of aspect in Russian. In M. S. Flier & R. D. Brecht (Eds.), Issues in Russian morphosyntax (pp. 9–34). Bloomington, IN: Slavica Publishers. Chvany, C. V. (1990). Verbal aspect, discourse saliency, and the so-called “perfect of result” in modern Russian. In N. B. Thelin (Ed.), Verbal aspect in discourse (pp. 213–235). Amsterdam: Benjamins. Comajoan, L. C. (2002). The acquisition of catalan L2 past morphology: Evidence for the aspect and discourse hypotheses affiliation (Doctoral dissertation). Retrieved from Dissertation abstracts international, A: The humanities and social sciences. (2740-A ISSN 0419–4209.) Dickey, S. M. (2000). Parameters of Slavic aspect: A cognitive approach. Stanford, CA: CSLI Publications. Durst-Anderson, P. (1992). Mental grammar: Russian aspect and related issues. Bloomington, IN: Slavica Publishers. Ellis, N. C. (2006). Selective attention and transfer phenomena in SLA: Contingency, cue competition, salience, interference, overshadowing, blocking, and perceptual learning. Applied Linguistics, 27 , 164–194.
48 Ellis, N. C. (2008). The dynamics of second language emergence: Cycles of language use, language change, and language acquisition. Modern Language Journal , 92, 232–249. Flashner, V. E. (1989). Transfer of aspect in English oral narratives of native Russian speakers. In H. Dechert & M. Raupach (Eds.), Transfer in Language Production (pp. 71–97). Norwood, NJ: Ablex. Forsyth, J. (1970). A Grammar of aspect: Usage and meaning in the Russian verb. Cambridge: Cambridge University Press. Gladney, F. Y. (1982). Biaspectual verbs and the syntax of aspect in Russian. The Slavic and East European Journal , 26 , 202–215. Isaˇcenko, A.V. (1962). Die russische sprache der gegenwart [Contemporary Russian]. Formenlehre. Halle (Saale): Niemeyer. Janda, L. (2003). Russian aspect at your fingertips: Why what you know about matter matters, or a userfriendly conceptualization of verbal aspect in Russian. Slavic and East European Journal , 47 , 251–281. Kaplan, M. A. (1987). Developmental patterns of past tense acquisition among foreign language learners of French. In B. VanPatten, T. R. Dvorak, & J. F. Leeds (Eds.), Foreign language learning: A research perspective (pp. 52–60). Cambridge, MA: Newbury House. Kuˇcera, H. (1983). A semantic model of verbal aspect. In M. S. Flier (Ed.), American contributions to the IXth congress of slavists, Kiev, USSR (pp. 171–184). Bloomington, IN: Slavica Publishers. Kumpf, L. (1984). Temporal systems and universality in interlanguage. A case study. In F. Eckman, L. Bell, & D. Nelson (Eds.), Universals of second language acquisition (pp. 132–143). Rowley, MA: Newbury House. Liskin-Gasparro, J. (2000). The use of tense-aspect morphology in Spanish oral narratives: Exploring the perceptions of advanced learners. Hispania, 83, 830–844. Long, M. & Gor, K. (2009, February). Linguistic correlates of proficiency: Towards a research-informed Russian curriculum. Plenary Session of the Interagency Language Roundtable, Arlington, VA. ´ Lopez-Ortega, N. R. (2000). Tense, aspect, and narrative structure in Spanish as a second language. Hispania, 83, 488–502. Lubensky, S. (1984). The aspectual properties of verba percipiendi. In Flier M. & A. Timberlake (Eds.), The scope of Slavic aspect (pp. 76–93). Bloomington, IN: Slavica Publishers. Negueruela, E. (2003). A sociocultural approach to teaching and researching second languages: Systemic-theoretical instruction and second language development. Unpublished doctoral dissertation. State College: Pennsylvania State University. Negueruela, E., & Lantolf, J. P. (2006). Concept-based instruction and the acquisition of L2 Spanish. In R. Salaberry & R. Lafford (Eds.), Spanish second language acquisition: From research to praxis (pp. 79–102). Washington, DC: Georgetown University Press.
The Modern Language Journal 96 (2012) Rassudova, O. P. (1975). Aspectual meaning and aspectual context in the teaching of Russian verbal aspect. Slavic and East European Journal , 19 , 139– 144. Rifkin, B. (1995). Error gravity in learners’ spoken Russian: A preliminary study. Modern Language Journal , 79 , 477–490. Rifkin, B. (2003). Oral proficiency learning outcomes and curricular design. Foreign Language Annals, 36 , 582–588. Robison, R. E. (1990). The primacy of aspect: Aspectual marking in English interlanguage. Studies in Second Language Acquisition, 12, 315–330. Rosengrant, S. F. (2007). Russian in use: An interactive approach to advanced communicative competence. New Haven, CT: Yale University Press. Salaberry, R. (2003). Tense aspect in verbal morphology. Hispania, 86 , 559–573. Selinker L., & Douglas, D. (1985). Wrestling with context in interlanguage theory. Applied Linguistics, 6 , 190–204. Selinker L., & Douglas, D. (1994). Research methodology in context-based second-language research. In E. Tarone, S. Gass, & A. Cohen (Eds.), Research methodology in second language acquisition (pp. 119– 132). Hillsdale, NJ: Erlbaum. Shekhtman, B., & Sibrina, S. Imitaci reqi nositel zyka—osnovno instrument v obuqenii prodvinutyh studentov [Imitation of native speech—advanced students’ main tool]. In I. Dubinsky & R. Robin (Eds.), Teaching and learning to near-native levels of proficiency II: Proceedings of the fall 2004 annual conference of the Coalition of Distinguished Language Centers (pp. 55–58). Salinas, CA: MSI Press. Shibata, M. (2000). Comparing lexical aspect and narrative discourse in second language learners tense-aspect morphology: A cross sectional study of Japanese as a second language Unpublished doctoral dissertation. Tempe: University of Arizona. Shteinfel’dt, E. A. (1963). Qastotny slovar sovremennogo literaturnogo russkogo zyka [Frequency dictionary of contemporary literary Russian]. Tallinn: Tallinn. Timberlake, A. (1985). The temporal schemata of Russian predicates. In M. S. Flier & R. D. Brecht (Eds.), Issues in Russian morphosyntax (pp. 35–57). Bloomington, IN: Slavica Publishers. Vendler, Z. (1967). Verbs and times. In Z. Vendler (Ed.), Linguistics and philosophy (pp. 97–121). Ithaca, NY: Cornell University Press. V´eronique, D. (1987). Reference to past events and actions in narratives in L2: Insights from North African learners of French. In C. W. Plaff (Ed.), First and second language acquisition processes (pp. 252–272). Cambridge, MA: Newbury House. Wulff, S, Ellis N., R¨omer, U., Bardovi-Harling, K., & Leblanc, C. J. (2009). The acquisition of tense– aspect: Converging evidence from corpora and telicity ratings. Modern Language Journal , 93, 354– 369.
49
Richard M. Robin APPENDIX Samples of ACTFL Speech Levels From the Corpus These abbreviated descriptions provided here are condensed from ACTFL, 2012. Superior speakers participate fully and effectively in conversations on a variety of topics in formal and informal settings from both concrete and abstract perspectives. They explain complex matters in detail and provide lengthy and coherent narrations. They support their opinions on social and political topics of importance to them with structured argument to support their opinions. Such discourse, while coherent, may still be influenced by the Superior speakers’ own language patterns, rather than those of the target language. On the topic of Russian corruption:
Oqen trudno skazat legkimi slovami to, qto nado delat, potomu qto korrupci kak my govorili zaloena v sistemu seqas i nastolko stala qast gosudarstvennyh struktur i pressa nastolko zadavlena, qto ue oqen trudno s ne borots, no tem ne menee sqita, qto nado vse-taki skazat nunye slova, qto nuno govorit kak mono qawe o tom, qto pressa dolna byt svobodno, qto processy dolny byt otkrytye, qto milici dolna byt nepodkupna, qto dolno byt doverie naroda v gradanskom obwestve. to naivnye slova, no mne kaets, qto seqas, 20 let spust, naivnye slova Regana imeli kakoe-to vlinie. I hot ne podderivala ego politiku SXA, no duma, on v to vrem govoril to, qto nado bylo govorit gensekam Sovetskogo
‘It’s very hard to say in a few simple words what is to be done because corruption, as we were saying, is embedded in the system, and it has become such a part of government entities, and the press is under such constraints, that it has become really hard to fight. That notwithstanding, I believe that we should nevertheless say what we have to say and say it as often as possible, that the press should be free, that processes should be open, that the police should be unbribable, that the people should be able to trust in civil society. Those are na¨ıve words, but it seems to me that now, 20 years on, Reagan’s na¨ıve words exerted an influence. And although I didn’t support his policy in the United States, think that at the time he said what had to be said to the General Secretaries of the Soviet Union. And it seems to me that no matter what, it’s most important that we not stop talking about freedom of the press and about transparency in the judiciary system.
Soza. I mne kaets, qto nado, ne ostanavlivas, vse ravno govorit vanee vsego pro svobodu pressy i pro otkrytost sudebno sistemy. Nu Bog znaet, qto mono sdelat s korrupcie na gosudarstvennom urovne.
But God only knows what can be done about corruption at the level of the State.’
Advanced High speakers perform like Superior speakers, but inconsistently. Advanced Mid speakers demonstrate the ability to narrate and describe in all major time frames (past, present, and future) by providing a full account, with good control of aspect; . . . Narration and description tend to be combined and interwoven to relate relevant and supporting facts in connected, paragraph-length discourse . . . . They lack the ability consistently to provide a structured argument in extended discourse. On why the speaker likes silent films;
Vy znaete, to oqen trudno skazat, to oqen trudno ob snit svoe mnenie, kogda to bolxe vsego svzano s quvstvom i ne svzano kak-to s umom. duma, qto to iz-za togo, qto mnogomnogo let zanimalas baletom, 17 let zanimalas baletom i govort, qto kino blie k baletu qem naprimer k teatru i duma qto iz-za togo, qto mogu sidet smotret na nemoe kino i prosto poluqit p → i udovolstvie ot dvieni i kak oni svzany drug s drugom i ne nado imet kako-to rasskaz dl men, to kak idei iskusstva konstruktivizma, prosto vse ti ivopisi, vse vot ti dvieni, qto mono vot to qitat i ne nado imet kak logiqny rasskaz.
‘You know it’s very hard to say. It’s very hard to explain your opinion when it’s mostly connected with a feeling and not connected somehow with your mind. I think that it might be the fault of my taking ballet for so many years. I did ballet for 17 years, and they say that film is closer to ballet than, say, theater, and I think that that’s why I can sit and watch silent film and just getp → i pleasure from the motion and how they are connected. And there doesn’t have to be a story for me. It’s like the ideas of the art of constructivism, just all of those paintings, all those movements, that you can read it and not have it as a logical story.’
50
The Modern Language Journal 96 (2012)
Advanced Low speakers perform in a way similar to Advanced Mid speakers, although somewhat haltingly. Their participation in work-related activities is less developed. When pressed for fuller narrations and descriptions, they tend to grope and rely on minimal discourse. Their utterances are typically not longer than a single paragraph. Control of aspect may be lacking at times. On getting out of jail for illegal proselytizing:
My sideli v trme neskolko dne. Vzli naxi dokumenty, pasporty, i skazali, qto budet sud, kak, qto oni rexat, qto delat s nami. My pozvonili direktor i on skazal nam, horoxo, u nas est risty, kotorye ivut v Γermanii i nuno sprosit ih, qto nam nuno sdelat p → i i togda oni poslali kak . . . ne zna, qto oni sdelali, no qerez neskolko dne my vygnali.
‘We spent several days in jail. They took our documents, our passport, and told us that there would be a trial, that they were deciding what to do with us. We called to our director and he told us, okay, we have lawyers who live in Germany and they would have to ask them what must get done, and then they sent how. . . I don’t know what thing they did, but several days later we drove them out.’
The performance of Intermediate High speakers is similar to that of Advanced Low speakers, but it is inconsistent.
