The use of social vs. cognitive approaches to the study of second language acquisition (SLA) has engendered considerable debate in the field. For instance, the recent Modern Language Journal Focus Issue (Lafford, Reference Lafford2007a) reviewed the ongoing debate between scholars espousing socially- and cognitively-grounded approaches to SLA research and explored the extent to which SLA scholars have answered Firth and Wagner's (Reference Firth and Wagner1997, p. 286) call for a reconceptualization of SLA “as a more theoretically and methodologically balanced enterprise that endeavours to attend to, explicate, and explore, in more equal measures and, where possible, in integrated ways, both the social and cognitive dimensions of S/FL [second/foreign language] use and acquisition”. Firth and Wagner (Reference Firth and Wagner1997, p. 286) specifically called for “(a) a significantly enhanced awareness of the contextual and interactional dimensions of language use, (b) an increased emic (i.e., participant-relevant) sensitivity towards fundamental concepts, and (c) the broadening of the traditional SLA data base”.
A more balanced approach to the study of a particular second language (L2) phenomenon would not only carry out classroom cross-sectional cognitive experiments with large groups of subjects, but would also utilize qualitative longitudinal emic approaches that seek to understand how individual learners understand the phenomenon under scrutiny. This new approach would also broaden the SLA database to include conversational interactions gathered in different contexts (e.g., naturalistic [study abroad], immersion settings, and classrooms) in order to demonstrate the inextricable relationship between the acquisition of language and its use in communicative settings. It would also take into consideration issues of learner identity and the complexities of the current multilingual global reality (including regional variation among cultured monolingual and multilingual speakers of the target language) when gauging learner progress against a target “norm”.
Both cognitive and socially-based approaches to the study of SLA may ask different questions and measure different constructs but they all enhance our understanding of the complexity of the processes by which individuals acquire a second language. Central to this debate is the issue of establishing research practices that allow scholars in various theoretical frameworks to gather data that will answer the particular questions posed by their theory of choice.
In this paper on the Spanish copula Woolsey addresses the ongoing challenges of acknowledging the complexity of theoretical conceptualizations and adapting theory to research practices. Although this is essentially a cognitively-based study, the author has adopted a research design that also recognizes and accounts for the influence that social and pragmatic factors may have on copula choice. Due to space limitations I will restrict my remarks in this essay to commentary on some of the methodological choices made by the author and ways in which future research on the acquisition of the Spanish copulas can expand upon Woolsey's work.
The research design of Woolsey's study, based on Woolsey (Reference Woolsey, Klee and Face2006a, Reference Woolsey, Sagarra and Toribiob), was meticulously created and executed. Following Geeslin's (Reference Geeslin2001, Reference Geeslin2003) innovative approach to the study of the Spanish copulas, Woolsey uses a multi-variable approach to investigate the “bundle of features” that combine in various ways to predict the use of estar when followed by adjectives. Due to the complexity of the theoretical conceptualizations that have been proposed for the use of ser and estar with adjectives (e.g., Navas Ruiz, Reference Navas Ruiz1963; Bull, Reference Bull1965; Falk, Reference Falk1979a, Reference Falkb; Lafford, Reference Lafford and de Alba1986), the author wisely restricts his focus to the use of estar with adjectives in only two contexts: (a) to create comparisons within an individual frame of reference (IFOR) and (b) reactions that result from an immediate experience (IE) with the referent.
Typical of a cognitively-based study, the number of participants is large (N = 111) and includes four different proficiency levels. However, important information on these participants is lacking e.g., their native language (L1), their experience speaking Spanish or another language other than English at home, time spent abroad in a Spanish-speaking country, other languages studied. Any of these factors could affect the outcomes of this study and should have been controlled or accounted for in the discussion of the results.
The use of four proficiency levels gives the reader a look at the development of the acquisition of the copulas in “apparent time”: the quantitative, aggregate data show a decline in ser and a gain in the use of estar as proficiency levels rise. However, in order to get an in-depth look at how the use of these copulas changes during the acquisition process, a more qualitative approach to the data is needed to complement the quantitative findings. For example, a longitudinal study could be designed that would investigate the stages of copula acquisition over a period of time in a smaller group of students. Data from pre- and post-tests of copula recognition and use, as well as data from student journals, could be gathered from students studying abroad. Data from their native-speaker (NS) host families would complete the picture by providing information about the copula norms in the speech of people providing input to those students on a daily basis.
One of the limitations of this study mentioned by the author is the lack of a NS control group to compare with the non-native (NNS) L2 learners of Spanish. The inclusion of an NS group is essential to understanding the progress that learners make toward the acquisition of NS norms with regard to copula use. Theoretical claims that have been made about the use of estar with individual frames of reference and with immediate experience have been based on data from native speakers of Spanish. However, as Woolsey notes, recent research (e.g., Geeslin, Reference Geeslin2003; Miller & Schmidt, Reference Miller and Schmitt2005) has also called into question the sustainability of the claim that native speakers use estar in the context of IFOR. As the current study did not gather data from NSs, it is limited in its ability to suggest a change in theoretical conceptualizations of the use of estar + adjectives in IFOR and IE contexts.
Complicating the issue is the fact that scholars have found regional variability among native speakers in their use of the copula verbs. For instance, Silva-Corvalán (Reference Silva-Corvalán1986) and Gutiérrez (Reference Gutiérrez2001) found that Spanish speakers living in bilingual areas (Los Angeles and Houston, respectively) have extended the use of estar + adjectives to contexts in which other dialects would use ser, i.e., to establish a class frame of reference (e.g., Yo nací . . . en un pueblito chiquito, no está muy grande “I was born . . . in a small town, it is not very big”). Other scholars have found the same phenomenon occurring in monolingual dialect areas, e.g., Mexico (Gutiérrez, Reference Gutiérrez1992, Reference Gutiérrez1994) and Venezuela (De Jonge, Reference De Jonge1993). This variability in copula use brings up the following questions when trying to interpret L2 data on the acquisition of ser and estar: Which NS norm do we use to see a progression of “native-like” use of estar in L2 learners of Spanish? What is the regional origin of the instructors of participants at each level? Are they from a dialect area that uses estar in more innovative ways than most other regions? These questions are relevant to the interpretation of the results of this study due to the fact that these instructors probably provide the participants with the majority of the input they receive containing these two verbs.
Another methodological decision that bears discussion is the choice of instruments used to collect the data in Woolsey's study. These data were gathered by a contextualized preference task (CPT; based on Geeslin, Reference Geeslin1999, Reference Geeslin, Leow and Sanz2000) and a picture description (PD) task, which was used to create clear and unambiguous contexts of comparison and reaction. The first task-type, CPT, was based on work by Briscoe (Reference Briscoe1995), Geeslin (Reference Geeslin1999, Reference Geeslin, Leow and Sanz2000), Schütze (Reference Schütze1996), and Thompson (Reference Thompson1991). This use of contextualized prompts constitutes an improvement over typical grammaticality judgment tasks used in prior SLA studies, in that the CPTs provide the participant with information about the context in which the acceptability of the two sentences (one using ser and the other using estar) would be judged. In this way the research design avoids the trap of artificiality often found in cognitively-based studies that use isolated sentences as prompts, and it serves as a model of instrument design for future studies of SLA.
In this study, each item in the CPT contained the following structure: a paragraph in English to provide a pragmatic context for the sentences containing ser and estar, a Spanish prompt, two sentences responding to the prompt: one with ser and the other with estar, and a question (to be answered in English) regarding the participant's preference for one sentence over the other in the given context (see Woolsey's Table 4). Future research studies on copula choice preference could consider using Spanish-language authentic materials to provide the background context for the participant's choice(s). Excerpts from newspapers, essays, and native speaker conversational data could provide a more natural context in which to consider the acceptability of use of one or both verbs.
Although the use of English to provide the context for the use of the copulas may be questionable to some, the author needed to administer the same test to L2 Spanish learners at four different levels of proficiency; therefore, in this case, the use of the informant's L1 was necessary to establish an understandable context for all participants. However, we may call into question the author's choice to avoid the use of a graded scale of acceptability for each of the two sentences, as Schütze (Reference Schütze1996) proposed. The use of such a graded scale would have captured the finer subtleties of preference in the participants' evaluation of the acceptability of the sentences containing the two copulas, and would have provided more insight into the complexity of speaker intent (or point of view) regarding each choice. Finally, the elicitation of the reasons behind the participants' preferences probably provided a good deal of insight into speaker intent regarding copula use; unfortunately, the answers to these questions based on an emic perspective were never discussed in this paper.
The PD task used by Woolsey's provides insight into copula preferences and also raises some important concerns. This PD task differs from those used in previous studies by its selection of referents and the slide design. The referents chosen were four well-known celebrities: Harrison Ford, Britney Spears, Michael Jackson, and Madonna. It is the author's contention that the choice of these famous individuals helped to ensure familiarity with the referent by the researcher and the participants, a factor not controlled for in previous studies. In order to establish a clear IFOR context, pictures of these celebrities taken years ago are put alongside their current pictures; the IE context is created by a photo of the celebrity in an unusual context (e.g., Harrison Ford in a Speedo). Woolsey proposes that the creation of these IFOR and IE pragmatic contexts was possible due to the participants having preconceived ideas about the chosen referents (celebrities). However, it really cannot be assumed that all participants had specific preconceived notions about all four of these celebrities – perhaps some of the students are too young to have established prior conceptions about a “norm” for Madonna, who has changed continuously over the past 25 years. Even if they had certain ideas about these referents before these data were collected, it is possible that there was significant variation in exactly what the preconceived ideas were (e.g., Michael Jackson's appearance has changed so often, it is almost impossible to establish a norm for him so that any given picture of him would seem unusual). In addition, more insight into speaker intent would have been possible with a discussion of the participants' English rendition of their Spanish descriptions; once more, data based on an emic perspective were not included in the discussion.
Future scholars can expand upon the work carried out in Woolsey's solid empirical study in several ways. Although essentially cognitive in design, Woolsey does recognize the need to establish pragmatic contexts for interpreting the copula preferences of the participants. Woolsey's consideration of social factors in his research design is in line with current thinking of finding a “meeting of the minds” between cognitive and social approaches to the study of SLA (Lafford, Reference Lafford2007b). The expanded use of more socially-based theoretical frameworks and qualitative approaches to the data could yield more insights into how copulas are acquired by Spanish L2 learners in NS–NNS and NNS–NNS interactions.
Socially-based frameworks (e.g., socio-cultural and variationist approaches) require gathering dialogic data in natural as well as in classroom contexts from NNSs and NSs alike. To complement classroom research carried out using artificial tasks (e.g., grammaticality judgments and picture descriptions) investigators need to capture the subtlety of copula use by learners in more naturalistic settings (e.g., study abroad contexts) to see how they acquire these copulas in interchanges with more expert speakers. Data gathered from spontaneous conversations between study abroad students and their host families over time would provide insight into the nature of the shifts taking place within the learners' zones of proximal development (Vygotsky, Reference Vygotsky1986) as they acquire the target language. The analysis of these longitudinal data would call for qualitative approaches in which individual paths of acquisition are explored and individual variation is acknowledged. Although this type of case study approach would provide fewer tokens than a larger quantitative study of copula use, it would provide researchers with the potential to gain an in-depth understanding of how learners acquire these verbs in their interactions with NSs in natural contexts.
Finally, future researchers can expand upon Woolsey's study by using a number of innovative techniques to investigate speaker intent. For instance, learners can be video recorded (see Smith & Gorsuch, Reference Smith and Gorsuch2004) while they react to picture descriptions in order to help discern the amount of surprise they actually experience when faced with referents in unusual situations (e.g., the aforementioned Speedo picture). In addition, more elaborate retrospective protocols can be used to gather empirically-based data from learners on what they were trying to convey (their intent) with the use of one copula over the other in a given context.
To conclude, the complexity of the acquisition of the Spanish copulas ser and estar can only be appreciated through the use of several types of theoretical and empirical approaches to the analysis of NS and NNS data gathered in classroom and naturalistic contexts. Lafford (Reference Lafford2007b, p. 751) suggests the adoption of “a dialectical approach that calls for a unified organic synthesis of the insights derived from both cognitively- and socially-oriented SLA research”. In Lantolf's (Reference Lantolf2007, p. 33) words, “a dialectical approach to language, human communication and L2 learning is the best prospect for integrating psychological and social forces into a unified theory of L2 development”. By its use of contextualized grammaticality judgments and its (unanalyzed) elicitation of data that would provide emic insights on speaker intent (e.g., explanations for participants' preferences in the grammaticality judgment task and English renditions of their Spanish picture descriptions), Woolsey's study takes a step towards balancing social and cognitive approaches to the study of second language acquisition. It is only through the triangulation of results of both cognitively- and socially-based studies of the acquisition of the Spanish copula that we can begin to create a more sophisticated understanding of how theoretical conceptions of the copula verbs form, and are informed by research practice.
