Any person who has taught Spanish as a second language or who has interacted with a non-native speaker of Spanish can easily tell that mastering the correct use of the copulas ser and estar is very difficult in both spoken and written production. But L2 learners are not alone. The Spanish copulas also present difficulty and frustration for L2 instructors of Spanish, since most pedagogical explanations of the uses of ser and estar provided in textbooks are incomplete and inaccurate. However, the acquisition of copular constructions has not received the attention it deserves in the acquisition literature, making a special issue of Bilingualism: Language and Cognition dedicated to this topic particularly welcome. A reason for the scarcity of research in this area may be related to both the linguistic complexity of Spanish ser and estar and the inadequacy of many available theoretical treatments to explain their complementary distribution. Although constructions with ser and estar are highly frequent in the input, they are grammatically quite intricate, straddling between the levels of morphology, syntax, semantics and pragmatics. Accordingly, Bruhn de Garavito and Valenzuela correctly frame their study within the current debate on interface vulnerability in language development in general, and in adult L2 acquisition in particular (Sorace, Reference Sorace2004; White, in press). In a nutshell, grammatical areas which require the integration of different levels of linguistic knowledge (e.g., syntax–discourse, syntax–morphology, morphology–semantics, etc.) for processing, production, or interpretation, show developmental delays and instability in monolingual and bilingual acquisition. In the case of monolingual acquisition, instability or non-target-like behavior is temporary, but in L2 grammars, instability can persist up to very advanced levels of proficiency, eventually leading to fossilization.Footnote ¹

Bruhn de Garavito and Valenzuela's study is a very important contribution to the acquisition of Spanish ser and estar, both at the methodological and the empirical levels. Existing studies of the L2 acquisition of ser and estar have largely focused on correct written or oral production of these copulas, as Bruhn de Garavito and Valenzuela review in their article. Approaching the L2 acquisition of ser and estar from linguistic theory, Bruhn de Garavito and Valenzuela investigated instead the syntactic distribution and semantic interpretations of these copulas in different pragmatic and syntactic contexts through the use of a written grammaticality judgment task and a written sentence selection task. At the empirical level, Bruhn de Garavito and Valenzuela capitalized on the overt semantic and syntactic distinction between verbal and adjectival passive structures, which in Spanish are overtly and differentially realized through the two copulas. To my knowledge, verbal and adjectival passives have not been investigated in the acquisition of Spanish before, thereby making this study doubly valuable and relevant to the field. In their study, Bruhn de Garavito and Valenzuela assume an aspectual analysis of ser and estar (Luján, Reference Luján1981; Schmitt, Reference Schmitt1992) which, in their view, is clearly manifested in the use of ser with verbal passives and estar with adjectival passives. The research question they pursue is whether English-speaking L2 learners of Spanish acquire the aspectual difference between ser and estar and are able to extend this knowledge to the syntactic and semantic constraints of passives in Spanish. Although Bruhn de Garavito and Valenzuela conceive the problem of ser and estar as an interface phenomenon, and interface phenomena show delay acquisition and/or fossilization in general, they hypothesize that the L2 learners will converge on the target grammar, without much elaboration or justification for their particular position.

The study compared a group of 21 advanced learners of Spanish with a group of 21 Spanish native speakers. The grammaticality judgment task tested ser and estar with adjectives, in passives with by phrases, with preterit and imperfect past tenses, and with adjectives taking the prefix in-. The results of this task showed that, in general, the L2 learners were quite accurate on the use of the copulas with specific adjectives, but contrary to Bruhn de Garavito and Valenzuela's hypothesis predicting successful acquisition, the L2 learners were not very accurate on the other three structures. Bruhn de Garavito and Valenzuela explain away the L2 learners' treatment of passives with the “by”-phrase due to their rarity in the input. Yet, given the aspectual analysis of the copulas assumed in this study, Bruhn de Garavito and Valenzuela do not say much on what I see as the most revealing result of this test: that the L2 learners were not very accurate on the preterit/imperfect distinction with the copulas. This finding suggests that either these participants may not have acquired the basic aspectual distinction of the copulas or they have not yet figured out how different meanings are associated with different overt forms in Spanish. The sentence selection task tested the semantic interpretation of the subject in copular constructions with ser and estar. Do the L2 learners know that the subject of sentences with ser + adjective/participle gets a generic interpretation but if the copula is estar the subject receives a specific interpretation? Bruhn de Garavito and Valenzuela found that the L2 learners had no problem interpreting the subject of estar as specific but were only 46.1% accurate at interpreting the subject of ser as generic. Clearly, these advanced L2 learners have not acquired the subtle interaction between the forms of the copulas (morphology), their syntactic environment (syntax), and their interpretation (semantics). If they know the basic morphological distinction between the two copulas, as the sentences with adjectives in the grammaticality judgment task appear to indicate, they have not yet integrated their meaning implications. Bruhn de Garavito and Valenzuela suggest that the problem is not with the grammatical representations for aspect, copulas and passives, but with processing, although it is hard to see how this explanation works given the methodology employed in this study, as the researchers correctly acknowledge.

Can the problems with the copulas for these learners be due to L1 transfer instead of processing? In what follows, I suggest that an alternative L1-based interpretation is likely, especially if one considers how English and Spanish map meanings to forms in the aspectual domain. In fact, I suggest that the patterns of results found by Bruhn de Garavito and Valenzuela are very similar to the patterns found by Montrul and Slabakova (Reference Montrul, Slabakova, Pérez-Léroux and Liceras2002, Reference Slabakova and Montrul2003) in their studies of the preterit/imperfect contrast in Spanish.

According to the aspectual analysis adopted by Bruhn de Garavito and Valenzuela, ser and estar are allomorphs: the overt morphological expression of the aspectual distinction [±perfective]. Therefore, ser represents the [–perfective] value and estar the [+perfective] value (see also Roby, Reference Roby2007). The problem for English speakers is that copula be is ambiguous or neutral as to the perfective/imperfective distinction, as are stative verbs in general (Giorgi and Pianesi, Reference Giorgi and Pianesi1997). Perfective and imperfective values are obtained through aspectual morphology or through compositionality with other elements in the sentence. Compare the following examples:

1. (1)

Sentence (1a) describes the state of being drunk, which typically occurs within one delimited time period, whose beginning or end (or both) is implied. The English copula is interpreted as [+perfective], and the Spanish equivalent is expressed with estar. Sentence (1b) describes the event of being drunk as a quality of Robert, and is thus interpreted as [–perfective]. Note that in English the NP takes an indefinite article, which contributes to the [–perfective] interpretation of the sentence in English. In Spanish, the presence of the indefinite article is optional, while the natural copula to use is ser. The adjective silly in sentence (1c) is sufficient to denote a [–perfective] interpretation in English, and the Spanish equivalent also takes ser. But in (1d), progressive -ing is needed to turn a [–perfective] state in English into [+perfective] with the same adjective. Once again, Spanish expresses this [+perfective] interpretation with estar. These sentences illustrate how the potential aspectual ambiguity of the English copula is resolved in English via morphology or compositionality (i.e. through other elements in the sentence). By contrast, to interpret aspect in Spanish copular constructions, English-speaking L2 learners must pay attention to the copulas, and match [–perfective] to ser and [+perfective] to estar.

In my view, ser and estar present the same learning problem as the acquisition of the preterit and imperfect in Spanish. While English has one simple past tense form, Spanish has two. The two languages express the same semantic notions: one-time event, ongoing, habitual and generic, but these are encoded differently. In English, the simple past with eventive predicates expresses a perfective (Peter robbed a lady in the bus) or a habitual event (Peter robbed banks). The progressive is expressed with the -ing form (Peter was robbing a lady when a man shouted HELP!). In Spanish, a perfective event is expressed with the preterit (Pedro le robó a una señora en el autobús); progressive and habitual events take the imperfect (Pedro robaba bancos, Pedro le robaba a una señora en el autobús cuando un hombre gritó AYUDA!). In restructuring their grammars, English-speaking learners of Spanish have to first learn the two new morphological forms, realize that they are in complementary distribution, and then reassemble the way they map meaning onto form. Such a three-stage task involves the coordination of grammatical knowledge at the interface between morphology, semantics and syntax, and causes obvious delays in acquisition. Indeed, the patterns of errors found by Bruhn de Garavito and Valenzuela are very similar to those reported by Slabakova and Montrul (Reference Slabakova, Montrul, Salaberry and Shirai2002, Reference Slabakova and Montrul2003) and Montrul and Slabakova (Reference Montrul, Slabakova, Pérez-Léroux and Liceras2002, Reference Slabakova and Montrul2003) in their studies on the interpretive properties of Asp(ect)P(hrase) and its morphological realization through preterit and imperfect tense morphology.

Montrul and Slabakova (Reference Montrul, Slabakova, Pérez-Léroux and Liceras2002) tested 71 adult English-speaking L2 learners of Spanish, from low to advanced proficiency. They were first asked to complete a task to see whether they recognized the morphological distinction between preterit and imperfect. If they know the morphology, will they know the semantics associated with it or does this knowledge come later? To answer this question, Montrul and Slabakova administered a sentence conjunction judgment testing the semantic implications of preterit and imperfect with accomplishment, achievement and stative predicates. We found that those advanced and intermediate learners who scored above 80% accuracy on the morphology recognition task had also acquired the semantic implications of preterit and imperfect. Those learners who still had difficulty discriminating between the two morphological forms did not have intuitions about their interpretation. Bruhn de Garavito and Valenzuela's results are very similar: their advanced subjects know the morphological difference between the two copulas, but they are not necessarily sure about their implications as perfective, imperfective and in generic contexts.

Are aspectual interpretations subject to permanent indeterminacy and possibly fossilization? Montrul and Slabakova's (Reference Montrul and Slabakova2003) study is relevant to this question. Montrul and Slabakova (Reference Montrul and Slabakova2003) included different levels of advanced learners – advanced, superior and near-native, classified on the basis of an oral interview and a standardized written proficiency task. This study is also comparable to Bruhn de Garavito and Valenzuela's study because it tested subject interpretation as a function of aspect. In Spanish, impersonal pronouns and 2nd person pronouns can take a generic or specific interpretation depending on the aspectual form of the verb: when the verb is in the imperfect, the interpretation is generic and specific (since specific is part of generic), but when the verb is in the preterit, the interpretation can only be specific.

1. (2)

2. (3)

Montrul and Slabakova (Reference Montrul and Slabakova2003) tested the interpretation of sentences like (2a, b) and (3a, b) in a truth-value judgment task. They found that the advanced and superior groups did not see that the imperfect could have both interpretations (marked by # in (2b) and (3b)). However, the near-native speakers performed at the level of the native speakers, correctly interpreting the subject of imperfect sentences as having both generic and specific interpretations and the subject of preterit sentences as being only specific. Therefore, these results suggest that the acquisition of subtle semantic properties associated with aspect in Spanish are eventually acquirable by learners of English, but it takes a while. L2 learners must reorganize the way they encode morphologically and syntactically the same meanings they have available from their L1.

In conclusion, Bruhn de Garavito and Valenzuela's study adds another interesting dimension to the complex issue of aspectual composition in a second language. Although the results are limited because they only focused on one group of advanced speakers, they still provide us with a new direction for further research in this exciting area.