1. Introduction
1.1 L2 Spanish in the world
As a major international language with approximately 400 million speakers (Lewis Reference Lewis2009), Spanish is nowadays studied all over the world. The number of students of L2 Spanish is estimated at 14 million (Instituto Cervantes Reference Instituto Cervantes2006). In the last two decades there has been a global expansion of educational contexts for the teaching of Spanish, and, consequently, an increasing interest in the study of learning and teaching Spanish. There are several reasons for this. One important factor has been the institutional recognition of Spanish as a major international language and the promotion of teaching Spanish globally from institutions such as Instituto Cervantes and the Spanish government (Del Valle Reference Del Valle, Mar Molinero and Stewart2006; Mar Molinero Reference Mar Molinero, Mar Molinero and Stewart2006). Spanish and Latin American universities have seen the rapid growth of their Spanish for foreigners programs as well as their teacher preparation and applied linguistics programs in L2 Spanish. At the same time, Spain has a large new immigrant population, which has created a demand for teaching Spanish to immigrant adults and children. In North America, enrollments in Spanish classes have also surged dramatically which, together with the increasing size of the Latino population, have created opportunities for experimenting with language programs (immersion, content-based, classes for heritage learners, etc.). In Latin America there has been a renewed interest in bilingual education for indigenous populations who learn Spanish as their second language, as well as a new emphasis on teaching L2 Spanish to foreigners.
1.2 Criteria for inclusion of studies
The diversity of educational contexts mentioned above has prompted a proliferation of studies shared through conferences and publications. In light of the increase in the number of publications on L2 Spanish since the 1990s, it seems timely to conduct a broad review of the work that has been done in recent years. This review presents a panoramic picture of a wide variety of studies in relation to L2 Spanish, representing a variety of learners and learning contexts around the world. Along with studies based in North America, where research on L2 Spanish has a well-established tradition, the review highlights research conducted in Spanish-speaking countries (Spain and Latin America) for its novelty and vigor. Studies from other countries, being less numerous and less accessible to this reviewer, are represented to a lesser degree. In preparation for this review, electronic and manual searches yielded over 260 works for consideration, including books, articles in major international and national journals, doctoral dissertations, and conference proceedings. The works selected for review were published between 2000 and 2008 in English or Spanish, and focus on learning L2 Spanish through formal instruction, the context most relevant for language teaching professionals. Empirical, evidence-based, quantitative and qualitative research is emphasized over descriptions or reflections on pedagogical practice. However, acknowledging the different academic traditions of Spanish-speaking countries, some non-empirical works have been included. Research on bilingual education is included only if there is explicit focus on L2 Spanish. Given the large geographical and thematic scope of this review and space restrictions, some areas of research have been entirely excluded, such as advances in service learning and language programs for specific purposes. I present a subjective selection of studies that, in my view, paint a good picture of advances in the study of learning and teaching L2 Spanish.
The text is divided into four main sections according to learner type and context of learning (traditional L2 learners, heritage learners, immigrants, and indigenous populations in Latin America). The bulk of the research has been with traditional learners; for that reason, the next section is further divided into subsections (the language system, learning processes and contexts).
2. L2 Spanish for traditional classroom learners
Teaching and learning Spanish has a long tradition in North America. The American Association of Teachers of Spanish and Portuguese has had an important role in the promotion of L2 Spanish studies for nearly 100 years. Lafford (Reference Lafford2000) reviewed publications in Spanish applied linguistics in the twentieth century. She traced the developments in the field from descriptive studies of problem areas in the first half of the century, to contrastive analysis and error analysis from the 1950s to the 1970s, evolving towards a focus on products, processes and contexts of acquisition from the 1980s onwards, a trend which has continued into the first years of the twenty-first century. Studies have continued to focus on traditional problem areas for learners (for example, phonetics and phonology, the distinction between ser/estar, preterite/imperfect and lexical acquisition), along with other less traditional areas such as the development of language skills and the acquisition of pragmatic competence. The study of the effect of the context of learning has prompted a number of publications on such areas as study abroad and immersion contexts, use of technology, and classroom discourse. The recognition of the learner as an individual with variable needs and aptitudes has prompted researchers to look into the effect of motivation, anxiety and students’ perspectives on their learning experience. As far as theory is concerned, studies of L2 Spanish in North America are grounded on theories of learning and language acquisition, particularly universal grammar, cognitive approaches, and sociocultural theory. Apart from Lafford (Reference Lafford2000), there have been some important volumes (Lafford & Salaberry Reference Lafford and Salaberry2003, Salaberry & Lafford Reference Salaberry and Lafford2006, and Lacorte Reference Lacorte2007) providing state-of-the-art reviews on topics related to L2 Spanish.
Growing interest worldwide in learning Spanish and the promotion of Spanish at the institutional level, mainly through the work of Instituto Cervantes, have given an extraordinary impetus to L2 Spanish in Spain. There has been a proliferation of new didactic materials, as well as teacher training manuals, surveys on areas of applied linguistics and language acquisition, and original studies on aspects of learning L2 Spanish. In Spain, as in the rest of Europe, the teaching of Spanish is framed by the guidelines of the Common European Framework of Reference for Languages (CEFR) (Council of Europe 2001). Much of the published work is geared towards pedagogical issues. Some general volumes have been useful in bringing Spanish L2 professionals up to speed with advances in second language acquisition (SLA) and teaching. Such is the case with Sánchez Lobato & Santos Gargallo's (Reference Sánchez Lobato and Santos Gargallo2004) comprehensive survey of a vast array of content and methodology areas relevant for teachers of Spanish.
As in Spain, teaching Spanish to foreign speakers in Latin America is a recent phenomenon. The development of transnational ties in Latin America in the form of economic agreements such as NAFTA and Mercosur has increased the demand for teaching Spanish since the mid 1990s, which, in turn, has resulted in professional conferences and publications dealing with the challenges of the new situation.
In what follows, I review studies that have continued to explore the acquisition of the language system, processes in learning, and individual and social variables that affect learning.
2.1 The language system
2.1.1 Phonology
Continuing a historical trend in Spanish language acquisition, publications in the area of phonology have been scant in recent years. The main areas of study have been the description of pronunciation problems experienced by L2 learners, processes in the acquisition of L2 pronunciation, and pronunciation instruction.
Analogy emerges as an important process in learning pronunciation. Bullock & Lord (Reference Bullock, Lord, Pérez-Leroux and Roberge2003) explored the role of analogical processes in the acquisition of L2 stress through experimental research with English L1 learners at beginner, intermediate and advanced level, and native speakers. A timed reading test with sentences containing real words and nonsense words led to the conclusion that learners resort to analogy when pronouncing unfamiliar words. Analogy is more successful as learners improve proficiency because their growing lexicon increases the probability of better matches. Intralingual analogy and L1 transfer accounted for errors in stress assignment among L2 Spanish learners at a Taiwanese university (Cortés Moreno Reference Cortés Moreno2003a, Reference Cortés Moreno2003b). Cortés Moreno (Reference Cortés Moreno2003a) summarized the results of the research undertaken for his doctoral dissertation on the perception of stress assignment by Taiwanese learners of Spanish. Learners were asked to identify the stressed syllable in each word read to them. Descriptive quantitative results pointed at good levels of stress perception from 70–80% across proficiency levels. Most difficulties were experienced with last-syllable stress. This study presents a pedagogically useful picture of the learners. However, as far as the study of acquisition of pronunciation is concerned, adding a discussion of the results in the light of SLA theories would help to provide a more comprehensive picture of phonological acquisition.
The role of instruction in the acquisition of pronunciation motivated Lord's (Reference Lord2005) experimental study, which concluded that explicit instruction has a positive effect on the acquisition of L2 Spanish sounds. The effect of explicit phonetic instruction for problematic sounds for L1 English speakers was investigated with learners in a Spanish phonetics class. Computerized voice analysis of /p/ /t/ and /k/, and comparative spectrographic analysis of native speech samples and participants, revealed that students made some gains after instruction, which consisted of the study of contrasts between Spanish/English sounds, oral practice, phonetic transcription, and self-analysis of recorded speech. Students’ perceptions of the effect of the course on their acquisition of native-like pronunciation patterns were also positive. However, because there was no control group in this study, there may have been reasons other than instruction for students’ improvement.
Researchers generally highlight the lack of attention paid to the teaching of pronunciation, at least at lower levels of proficiency. Arteaga (Reference Arteaga2000) advocated the inclusion of pronunciation instruction in current L2 teaching and made practical suggestions for first-year classes. Stokes (Reference Stokes2004) gave some pedagogical ideas, including communicative and cultural activities for advanced Spanish phonetics and phonology courses. For example, a recorded interview with a native speaker is a source of cultural, communicative and phonological practice. One reason why teaching pronunciation is downplayed in language classes is that teachers often lack appropriate training. González Bueno (Reference González-Bueno2001) stressed the importance of teacher training and described a model course for teaching pronunciation within a teacher preparation program. She suggested that, in teaching pronunciation, exposure to new sounds and discriminatory listening must precede imitative speaking. Other activities recommended are sound discrimination and production using visuals (diagrams, spectrograms, etc.), practice in communicative drills, and free production in communicative activities.
2.1.2 Morphology, syntax and semantics
2.1.2.1 Gender
The acquisition of Spanish gender morphology by adult learners has been a focus of recent studies. Alarcón (Reference Alarcón2005) measured gender agreement accuracy in an experimental computerized completion task with sentences that included a complex subject and a predicate adjective. Statistical analysis of accuracy results from 139 learners at three proficiency levels correlated with seven linguistic variables indicated that there were high rates of accuracy at all levels, but also differences in development. Overall, learners were more accurate in gender agreement when the head noun was masculine and overtly marked, and also when the two subject nouns had the same gender. Vocabulary knowledge was the third variable that positively affected accuracy at all levels. McCowen & Alvord (Reference McCowen, Alvord, Klee and Face2006) studied the accuracy of gender markings for human referents used by beginners in a variety of structured and unstructured tasks (oral interview, written description of video clip, dictogloss, written translation and questionnaire on gender assignment). Overall, learners were confused about gender markings and overused both forms, although task type was important. Interestingly, responses in the questionnaire indicated that when in doubt there was tendency to use masculine, suggesting that learners were aware that masculine is the unmarked form. The most common simplification strategies were overgeneralization and avoidance.
2.1.2.2 Clitics
The acquisition of clitics has received much attention since the 1980s, particularly from an input processing perspective. VanPatten (Reference VanPatten2004) discussed the results of input processing studies of clitic acquisition and the prevalence of the first noun strategy in interpreting clitics. Processing instruction, which takes into account the strategies adopted by learners in processing input, proved effective in helping learners interpret and produce clitics. Torres (Reference Torres, Montrul and Ordóñez2003) took a discourse functional approach to the use of clitics in learners’ conversational discourse. Speech data from English-speaking study-abroad students at the intermediate proficiency level provided evidence for strategies other than SVO order, specifically overuse of nouns (which is an avoidance strategy). In general, learners used few clitics, and there were few errors with clitics. There was also a high percentage of nouns to compensate for lack of use of clitics.
2.1.2.3 Tense and aspect
The teaching and acquisition of Spanish preterite and imperfect has been a major topic in applied linguistics for decades (Lafford Reference Lafford2000). Montrul & Salaberry (Reference Montrul, Salaberry, Lafford and Salaberry2003) reviewed studies on this topic from different theoretical perspectives: lexical semantics, generative, discourse, communicative and cognitive-perceptual perspectives. Recent research has mostly focused on providing empirical evidence for these. Liskin-Gasparro (Reference Liskin-Gasparro2000) analyzed the verbal morphology used in oral narratives by advanced learners of Spanish. Two types of narratives were included: film narration and personal experience narration. Narrative tasks were followed by retrospective protocols for the use of aspect. The results indicated that aspect choice was guided by four elements: the nature of the narrative task, lexical aspect, the role of the narrator in constructing the discourse, and the impact of instruction. However, studies conducted with lower-level learners in formal settings yielded different results. In a series of studies with beginners and intermediate learners, Salaberry (Reference Salaberry2000, Reference Salaberry2003, Reference Salaberry, Lubbers Quesada and Maldonado2005) found that beginners use a default past marker (preterite or imperfect) and that task type (fictional or personal narrative) may have an effect on the choice of aspect. For example, Salaberry (Reference Salaberry, Lubbers Quesada and Maldonado2005) analyzed the development of tense/aspect in four intermediate-level groups of learners in relation to the Lexical Aspect Hypothesis, which predicts that the preterite will be acquired first with achievement verbs while the imperfect will appear first with verbs of state. Data from four pedagogical activities (fill-in-the-blank text, multiple choice, guided oral narrative based on a movie and improvised oral production) did not categorically support the hypothesis, at least in production tasks. Learners in this particular context seemed to use the preterite as a default marker for the past, particularly under conditions that limited planning and monitoring. In a fill-in-the-blank task, results were in accordance with the Lexical Aspect Hypothesis. However, in oral production activities there was a tendency to use one past marker, regardless of the lexical meaning of the verb. As to why preterite would be the preferred past default marker, the author hypothesized that it has wider uses and it is the narrative verb form par excellence. However, as stated by the author, the results of this study are limited by the small number of participants and their low levels of acquisition, as well as by the lack of longitudinal data. A longitudinal study with L2 Spanish learners at four different levels of proficiency and native speaker control groups showed similar results (Salaberry Reference Salaberry2000). While beginners and intermediate learners did not seem to be guided by aspect in oral or written tasks, advanced learners showed an effect of lexical aspect, which was stronger as proficiency levels increased. This study confirms that beginners use preterite as a default marker of past tense. Similarly, in comparison to a native speaker group, Granda (Reference Granda2005) found that English-speaking learners of Spanish in Mexico used preterite forms successfully with telic verbs, but also used preterite in contexts where the native speaker group chose imperfect. Thus L2 Spanish data challenge the explanatory power of the Lexical Aspect Hypothesis, at least at beginner levels and in formal instruction contexts.
Learners’ metalinguistic knowledge of verbal aspect was discussed by Cadierno (Reference Cadierno2000–2001), who highlighted the cues Danish learners of Spanish followed when they explained their choice of aspect in oral, written and fill-in-the-blank tasks. Learners tended to identify one aspect with specific actions. In addition, they showed awareness of semantic and discursive factors that affect aspect choice, but generally followed cues at the sentence level (presence of adverbs, for instance) over discursive cues. Finally, focusing on self-correction during the production of oral narratives, Camps (Reference Camps2003) argued that limited attentional resources come into play when learners try to focus on person, number, tense and aspect at the same time, accounting for higher success in self-correction of person–number than tense–aspect.
Other studies have analyzed the effect of instruction. The effectiveness of concept-based grammatical instruction of verbal aspect was the object of Negueruela & Lantolf's (Reference Negueruela, Lantolf, Salaberry and Lafford2006) sociocultural study. Verbal aspect was presented through diagrams which guided learners on verbal aspect selection, followed by reflective activities. In a sense, verbal reflection forces learners to self-explain concepts and their motives for linguistic choices. Qualitative pre- and post-instruction comparative data from 12 university-level intermediate learners of Spanish demonstrated understanding of aspect and improvement of verbal aspect accuracy. By highlighting the role of verbal reflection as a learning experience mediated by language, this study pioneers a sociocultural approach to the study of verbal aspect acquisition.
The conclusions of studies on the acquisition of Spanish tense and aspect are not convergent, due in part to the use of different variables in the studies (levels of proficiency, context of learning, variety in tasks). Most studies have been short-term and conducted with small numbers of learners. Future studies should pay careful attention to variables in the learning environment as well as data collection and analysis so that their results may be comparable to existing studies.
2.1.2.4 Subjunctive
The acquisition of the subjunctive has continued to be a focus of research, particularly with regard to its dependence on the acquisition of complex syntactical structures and the effectiveness of processing instruction techniques compared with other instructional methods. Comparisons of the effectiveness of processing instruction and output instruction on the acquisition of the subjunctive have so far yielded contradictory results. While Farley (Reference Farley2001) found that processing instruction had an overall greater effect in comprehension and production tasks, Collentine (Reference Collentine1998) did not find differences between these two types of instruction. The divergence in results is attributed to methodological differences in the implementation of the curriculum.
Regarding the relation between the acquisition of complex syntax and the subjunctive, Collentine (Reference Collentine, Leow and Sanz2000) convincingly argued that, in light of the limited effectiveness of instruction, most beginner and intermediate programs in L2 Spanish in the US devote too much time to this grammatical point. He studied the relationship between the development of verbal morphology and syntactic abilities of L2 Spanish learners compared to a control group of native speakers. Quantitative data indicated that the learners’ ability to understand verbal morphology was similar to their ability to understand syntactic structure, and similar to that of native speakers. An interesting difference with the native speaker control group was that L2 learners experienced more difficulty in holding complex sentences in short-term memory, which affected their ability to interpret verbal morphology. Further exploring the hypothesis that complex syntactic development is necessary for the acquisition of subjunctive mood, Collentine et al. (Reference Collentine, Collentine, Clark, Friginal, Lee, Geeslin and Clements2002) reported on an experimental study to measure the effect of complex syntax instruction on the effectiveness of subjunctive instruction among three groups of learners. Post-tests showed that instruction on complex syntax increased the benefit from subjunctive instruction. Although both the subjunctive treatment group and the syntactic–subjunctive treatment group could predict equally well when the subjunctive was needed, the group that also received syntactic instruction could identify better the syntactic structures where subjunctive was not possible.
Experimental studies like those reported above have been important in identifying conditions and instructional methods that facilitate language acquisition. Some shortcomings of their results are that they are based on a small number of participants, that treatments and post-tests are often given over a very short time, and that language is tested at the sentence level in non-authentic contexts. Future studies should overcome these limitations by collecting longitudinal data of naturally occurring learners’ language and analyzing it through a different methodological lens that may enlighten external factors affecting internal cognitive development of the L2 linguistic system.
2.1.2.5 Copula
Studies during the 1980s and 90s discovered fairly consistent stages of acquisition of copula choice in Spanish (VanPatten Reference VanPatten, Van Patten, Dvorak and Lee1987, Guntermann Reference Guntermann1992). In a series of recent studies, Geeslin (Reference Geeslin, Leow and Sanz2000, Reference Geeslin, Bonch-Bruevich, Crawford, Hellermann, Higgins and Nguyen2001, Reference Geeslin2005) has moved away from error analysis in the study of copula acquisition and has focused instead on learners’ use of copula attending to adjective type. Geeslin (Reference Geeslin, Leow and Sanz2000) tried to identify the adjectival semantic features that would predict the use of estar, regardless of whether the use is prescriptively correct or not. Data from three different tasks (semi-structured interview, picture description and grammaticality judgment) performed by 77 adolescents at four levels of proficiency showed consistent results with previous studies with regard to the omission of copula at lower levels and an increasing use of estar as proficiency level increases. Adjectival semantic features such as dynamicity and perfectivity predicted use of estar at all levels, while others differed. For instance, susceptibility to change only predicted use of estar at lower levels while modality and frame of reference were only relevant at higher levels of proficiency. Reviewing research from a variety of linguistic disciplines, Geeslin's book (Reference Geeslin2005) presents a comprehensive overview of Spanish copula acquisition. Refining her multivariate analysis approach to copula choice, Geeslin compared her previous research on copula use by high school adolescents (Geeslin Reference Geeslin, Leow and Sanz2000) with that of adult advanced learners and native speakers (Guijarro-Fuentes & Geeslin Reference Guijarro-Fuentes and Geeslin2003). The inclusion of a native speaker baseline allows for an assessment of degrees of approximation to native-speaker usage. Advanced learners, for the most part, responded to contextual cues in similar ways to native speakers. However, differences in some variables suggested that native speakers adhered more strictly to semantic constraints, while advanced learners were more sensitive to the pragmatic situation. The reduction of data collection to only one task (contextualized preference task) resulted in an improvement of the variables as significant predictors of copula choice. Geeslin's multivariate analysis of learners’ copula choice represents a refreshing and promising way to look at old issues in the acquisition of Spanish. Her refinement of data collection procedures and the addition of a native speaker baseline sharpened the methodology.
Regarding pedagogical instruction, Chung Cheng (Reference Chung Cheng2002) reported on a comparative study of processing instruction of ser/estar vs. traditional instruction. Similar results in the use of ser/estar were found for both the traditional and the processing instruction group when compared to the control group. However, when only estar was considered, the traditional group improved only on a sentence production task and a guided composition test, while the processing instruction group improved on interpretation and production tasks.
Most studies of the acquisition of grammatical features have been quantitative. Experimental designs are tightly controlled with background information and pre-tests to ensure the homogeneity of the participant population, treatments and tests are carefully planned, and statistical analysis is used, which shows how the discipline has become increasingly aware of variables that can affect the interpretation of results. Nevertheless, the number of participants tends to be small, and the treatment period short. It is recommended that a mixed methods approach to data collection and data analysis (quantitative–qualitative) be encouraged in more studies and that longitudinal studies be conducted in order to corroborate current findings. Because learning and teaching grammar has been approached from a variety of theoretical frameworks (mostly generative and cognitive), dialogue among researchers working within different frameworks may result in greater advancements.
2.1.3 Vocabulary
Given the importance of lexical knowledge for successful communication, it is not surprising that much attention has been given to pedagogical and theoretical aspects of the acquisition of vocabulary. Several monographs published in Spain review theoretical perspectives on the lexicon, its acquisition and implications for teaching (Alvar Ezquerra Reference Alvar Ezquerra2003, Morante Vallejo Reference Morante Vallejo2005, Santamaría Pérez Reference Santamaría Pérez2006). Others offer practical suggestions for teaching vocabulary. Martín Mohedano (Reference Martín Mohedano2004) and Cruz Piñol (Reference Cruz Piñol2002) discussed the use of the Internet as a resource for teaching vocabulary. Álvarez Cavanillas & Chacón Beltrán (Reference Alvarez Cavanillas and Chacón Beltran2003) described linguistic aspects of lexical collocations and called teachers’ attention to its pedagogy. Barcroft (Reference Barcroft2005) illustrated the processing instruction model for teaching vocabulary.
Empirical studies of the acquisition of vocabulary have been mostly concerned with the effectiveness of learning strategies. Within a cognitive input-processing perspective, Barcroft (Reference Barcroft2007) examined how copying words and word fragments affects vocabulary learning. First semester students (N = 21) were asked to learn words under three conditions: word writing, fragment writing and no writing. Participants were given a pre-test, computer presentation of word–picture pairs and word–translation pairs, and productive and receptive post-tests. Scores were higher in the no-writing condition and in the receptive task (Spanish to English). These results corroborate previous studies and are consistent with the resource depletion for output hypothesis, that is, copying words may decrease vocabulary learning by using processing resources that could otherwise be used to make form-meaning connections. From a pedagogical standpoint, the results underscore the importance of processing words as input. In a large-scale and well designed study of lexical processing, Sagarra & Alba (Reference Sagarra and Alba2006) compared the effect of three strategies for vocabulary learning by the same L2 Spanish beginners (N = 778): memorization, semantic mapping and the keyword method. In this experiment, the keyword method, which involves form and meaning connections at a deeper level, proved significantly superior to the other two strategies for vocabulary retention. Quantitative results were consistent with learners’ perceptions. Subjects ranked the keyword method as most effective (92%), followed by memorization (8%) and semantic mapping (0%).
Text glossing is a common technique used to facilitate text comprehension and foster incidental vocabulary learning. Within an attentional framework, Bowles (Reference Bowles2004) conducted a comparative experiment on the effect of computerized vs. text glosses for text comprehension and vocabulary development. Think aloud protocols, comprehension tasks, word recognition tasks and written tasks performed by first-year learners of Spanish showed that glossing has significant benefits for comprehension and vocabulary acquisition when compared to a control group, but there were no significant differences regarding the medium of delivery. Also on incidental vocabulary learning, Frantzen (Reference Frantzen2003) investigated factors that influence the deriving of word meaning from context. Pre-tests and post-tests performed by third-year learners of Spanish in association with a reading text revealed that, although context may have a positive effect in guessing word meaning, it did not help for 70% of the words in this text. Several factors came into play in the lack of success in guessing word meaning from context, including context ambiguity, difficulty of the text, and learners’ behavior, particularly lack of attention to context and overconfidence that they understood the word. It was also noticed that glossing and context may lead students to deduce incorrect meanings.
The use of derivational morphology as a strategy to increase breadth and depth of vocabulary was studied by Morin (Reference Morin2006). A treatment group of Spanish learners received explicit instruction in derivational morphology over the course of the semester; a control group did not. Productive and receptive knowledge test results indicated that explicit strategy instruction may yield benefits in receptive and productive knowledge of Spanish derivational morphology, but not in vocabulary size.
The role of interaction and negotiation of meaning in vocabulary acquisition is relevant for communicative-oriented instruction. De La Fuente (Reference De la Fuente, Montrul and Ordóñez2003) compared online versus face-to-face negotiated interaction in the acquisition of vocabulary. Two groups of learners were subject to experimental conditions involving pairs of students participating in either oral interaction activities or a virtual chat. Quantitative results revealed gains for both groups, but the group that participated in oral interaction activities outperformed the rest in listening comprehension and oral production tests as well as in written recognition and written production tests. Based on examination of the content and quality of interaction during the task, the author hypothesized that oral interaction seems to promote more noticing of lexical forms and thus a deeper level of processing.
Overall, these studies are consistent with those based on other languages. Results suggest the importance of processing vocabulary as input, and support the use of vocabulary learning strategies, particularly those that involve deeper processing. Case studies, ethnographies and other qualitative methods could add a socio-psychological dimension to experimental studies by helping elucidate individual and socio-cultural factors affecting vocabulary learning.
2.1.4 Pragmatics
The field of pragmatics has gained increased attention among language teaching researchers and educators in recent years because of its central role in the development of L2 communicative competence. Pons Bordería (Reference Pons Bordería2005) suggested areas of pragmatics that could be incorporated into teaching materials beyond the speech acts and discursive markers commonly found in textbooks, such as implicature, politeness and intercultural pragmatics. Also with a pedagogical aim, Mir (Reference Mir2001) proposed a three-stage model for teaching speech acts in Spanish using film, from diagnostic interpretation of the speech act to presentation of its authentic use through film clips and analysis of cultural interpretation of speech acts. García Castro (Reference García Castro2004) dealt specifically with pedagogical issues of teaching turn-taking patterns.
Empirical studies have advanced our understanding of the acquisition of pragmatics in a classroom environment. Koike & Pearson (Reference Koike and Pearson2005) investigated the effectiveness of teaching suggestions using explicit/implicit instruction and feedback to three groups of third-semester English-speaking learners of Spanish. The authors were able to show that the group that received explicit instruction and explicit feedback performed better in a multiple-choice task, while the group that received implicit instruction was superior in the open-ended task. Gains, however, were not clearly maintained in the long term. Examination of the type of suggestions used by learners in the open-ended task revealed that only a few of the suggestion strategies taught were actually used. While the control group tended to use more statements and commands, the groups that received explicit or implicit pragmatic instruction generally used fewer direct strategies. The results of this study were limited by the short period of time of the treatment and the small number of students in each group (N = 4). In addition, it seems difficult to separate the effect of treatment from that of regular instruction and exposure to dialogues in textbooks. The fact that the learners had a very low level of proficiency calls for comparative or replication studies with more advanced learners. In a subsequent study, Pearson (Reference Pearson2006) investigated further the acquisition of Spanish directives, particularly commands and polite requests, by 131 learners with low-level proficiency (second semester) in Spanish. Two groups received instruction in Spanish speech acts. The first group engaged in metapragmatic discussions of the speech acts in video scenes. The second group received additional exposure to the video scenes but no discussion. A third group acted as control and did not receive any explicit instruction. The quantitative results of the tests showed no significant differences. However, qualitative analysis of developmental patterns in directive strategies in relation to factors such as instruction, L1 transfer, and L2 grammatical competence suggests that – as a result of increasing L2 grammatical competence – learners replace low-level direct strategies (want, need statements) with more indirect strategies and morphologically complex verb forms. Finally, L1 pragmatic transfer seems to occur in processing and interpreting L2 data. Despite the low impact of pragmatic instruction at this level, the author recommends introducing pragmatics in instruction, as studies show that pragmatic competence does not naturally develop with higher proficiency.
González Lloret's (Reference González-Lloret2008) study of the Spanish address system through synchronous computer-mediated communication provides evidence for the acquisition of pragmatic and intercultural competence through this medium. The author collected longitudinal data of chat interactions between Spanish university students in Spain and L2 Spanish learners in Hawaii over a period of eight weeks. One group engaged in free conversation while the other completed a project-based task. Interaction was then analyzed using a mixed qualitative and quantitative methodology. Whereas other similar studies have used a sociocultural or interactionist approach to data analysis, this one is framed within a Conversation Analysis (CA) emic approach to longitudinal data, which then leads to a quantitative analysis of found patterns. Microanalysis of developmental patterns in the use of the Spanish address system revealed that for those students who displayed variation in address forms at the onset of the study, it was possible to acquire native-speaker usage of informal pronouns and the corresponding verb morphology if they had multiple interactions with the native speakers. Student predisposition to address form usage and explicit feedback also had an effect on the acquisition of native norms. Students with a good command of the L2 address system at the beginning of the study were able to imitate the native speakers using a range of interactive mechanisms to express closeness or distance in the interaction (laughter, sharing personal problems, use of slang, etc.). This study takes a novel methodological approach to the areas of pragmatics and computer-mediated interaction by applying CA as a first approximation to data and by skillfully integrating quantitative and qualitative results in order to explore the topic in greater depth.
2.2 Learning processes and contexts
2.2.1 Error analysis
A good overview of advances in the study and classification of student errors is provided by Torijano (Reference Torijano2004), who takes the position that errors evidence progress in interlanguage. In Latin America, much of the work has been conducted with Brazilian L1 learners. For example, Bertolotti & Masello (Reference Bertolotti, Masello, Masello and Bertolotti2002) focused on pronouns of address from a contrastive analysis perspective. Bocca & Compagnoni (Reference Bocca, Compagnoni and Pozzo2006) analyzed written texts produced by L1 German and English learners in Argentina, illustrating psycholinguistic processes that cause errors.
Contrastive and error analysis studies in Spain have analyzed the L2 production of learners from different L1 backgrounds as well. An interesting study for its careful description of data is Lin's doctoral dissertation (Reference Lin2003) analyzing the use of articles in written compositions by Chinese L1 learners. The frequency of errors ranged from 32% to 42% across proficiency levels, which indicates an overall high percentage of errors, article omission being the main source of errors, particularly at low levels of proficiency. Interestingly, while errors of article omission notably decreased in upper levels of proficiency, errors of article insertion and number slightly increased. Lin provided a thorough description and exemplification of errors and descriptive statistics that illustrate the development of this feature of learners’ interlanguage. The interpretation of the sources of errors pointed at L1 transfer, but also intralingual causes, although it is always difficult to identify unequivocally a single source for any given error. Following a similar error analysis methodology, Pastor Cesteros (Reference Pastor Cesteros2001) conducted a frequency count of agreement errors in a grammar test administered to four groups of learners of different L1 backgrounds in Spain. Errors with noun–determiner and noun–adjective agreement disappeared or considerably decreased in the upper levels. However, gender and number agreement persisted at all levels, as well as complex subject–verb agreement.
Error analysis studies with learners of different L1 backgrounds have obvious practical applications in helping teachers design a curriculum that will take into account the difficulties learners are likely to run into. In addition, the results of these studies could be further used to gain greater insight into interlanguage stages and learning processes irrespective of L1 background.
2.2.2 Speaking, reading, listening and writing skills
Understanding how learners develop their language skills and helping them improve those skills have been the subjects of a good number of publications. In Spain, several monographs survey research findings in skills development and guide teachers with strategies for classroom use (on writing, Cassany Reference Cassany2005, on speaking, Cestero Mancera Reference Cestero Mancera2005, on listening, López García Reference López García2002).
Empirical studies have aimed at understanding the processes involved in the development of language skills. On conversational competence, García García (Reference García García2005) analyzed recorded conversations from 11 learners of Spanish in Germany at the intermediate–high or above proficiency level. Participants held free conversations with each other outside the academic setting. Qualitative results showed that participants overemphasized accuracy. There was a general absence of discourse markers as well as a lack of natural overlapping and use of connectors in turn-taking and topic shifts. She concluded that, overall, participants lack native-like interactional strategies and called for conversation to be taught explicitly in the classroom. Roebuck & Wagner (Reference Roebuck and Wagner2004) also focused on teaching conversational ability. Taking a sociocultural stance, they investigated whether learners can be taught to use repetition as a cognitive and communicative tool to create cohesion and involvement in the discourse. Analysis of students’ productions in different tasks suggests that they were able to use repetition for cognitive and communicative functions as a result of instruction and practice.
The bulk of research on reading has concentrated on the investigation of variable effects on reading comprehension and the development of reading skills. Within the input processing research tradition, Lee (Reference Lee, Leow and Sanz2000) investigated whether beginners in Spanish who have never been exposed to direct object pronouns were able to comprehend and process input from a reading passage. Results showed that text comprehension was superior to the input processing of pronoun forms, which indicates that when reading, learners focus more on meaning than on form–meaning connections. Brantmeier (Reference Brantmeier2005) studied the effect of readers’ subject knowledge, use of analogy and test type on reading comprehension in the first and second language. A group of L2 English learners in Costa Rica and two groups of intermediate and advanced learners of L2 Spanish in the US read two versions of scientific texts. The results of three comprehension tests (multiple choice, written recall and sentence completion) revealed that knowledge of subject matter does affect reading comprehension, but analogy seems to have a negative effect on reading comprehension. Asención's (Reference Asención2006) quantitative study of the effect of the rhetorical organization of text indicated that the ability to summarize was not affected by text type (compare/contrast and descriptive) or perceived L1 skills, but there was a weak relation to L2 achievement.
In the area of writing, Ramírez (Reference Ramírez, Leow and Sanz2000) illustrated how narrative competence in L2 Spanish develops with language proficiency. He adopted a multidimensional approach to the analysis of narratives based on a picture-story sequence by L2 Spanish learners at four levels of proficiency. Focusing specifically on the construction of the narrative text, articulation of narrative elements and development of narrative discourse, the narrative texts exhibited not only an increase in length as the level of proficiency increased, but also an increase in time indicators, use of past verbs, increase in morphological and semantic verb use, and lexical cohesive devices. In terms of story construction, all narrative texts displayed the basic components of a story (orientation, complication, evaluation and resolution). The effect of coded feedback on L2 Spanish writing was investigated by Greenslade & Félix-Brasdefer (Reference Greenslade, Félix-Brasdefer, Klee and Face2006). Nineteen participants at the intermediate/advanced level of proficiency completed two in-class narrative tasks. In the first composition, errors in syntax, lexis and mechanics were underlined, while in the second, errors were underlined and also coded. Then students were asked to revise their first draft in class and to complete a questionnaire about their perception of the feedback and improvement in their writing. Quantitative analysis of error correction yielded significant improvement in the second draft when feedback was coded, although both types had positive effects. Consistent with these results, students’ perceptions were more favorable towards coded feedback.
2.2.3 Interaction and feedback
Current attention to the process of learning has prompted research on classroom interaction between teachers and learners and among learners. It is important that teachers be aware of how classroom discourse may affect learning. Blanco (Reference Blanco2005) argued for reflective practice by teachers on their use of language in the classroom as a way of understanding how their personality, beliefs and previous experience affect their teaching. Blanco analyzed transcriptions from two class sessions by two teachers in Brazil. One class focused on a writing activity while the other dealt with grammar practice. While the first class was intended as interactive and participatory, with students and teachers composing a text collaboratively, discourse analysis revealed that interaction was dominated by the teacher. Discourse analysis from the second class showed how the teacher took complete control of the activities and interaction in the classroom. The discursive structure of class discussion in an advanced L2 Spanish literature course was analyzed by Donato & Brooks (Reference Donato and Brooks2004). They focused on teachers’ questions, verb tense distribution and student uptake in order to investigate whether discussion of literary readings resulted in advanced-level discourse by students, as evidenced by description and narration in different time frames, expression of opinion and hypothesis, and the ability to sustain extended discourse. The authors concluded that the predominance of the Initiation–Response–Feedback (IRF) pattern inhibited free expression and expanded discourse. Additionally, the prevalence of teachers’ rhetorical and evaluative questions – instead of open-ended questions – diminished opportunities for divergent opinions and complex critical thinking.
Contextualized grammar instruction within communicative activities poses a challenge for language teachers because the double objective of grammatical accuracy and meaningful interaction often leaves learners confused about the instructor's intentions. Framing his study in relevance theory, Toth (Reference Toth2004) compared the degree of classroom discourse cohesion in two lessons by the same teacher. In one, common conversational topics were discussed, while the other dealt with L2 grammatical forms. Poor cohesion was signaled by extended latency gaps and repair sequences during learners’ turns. Transcripts of videotaped lessons showed that grammatically-oriented discourse produced longer latencies and a higher number of remedy sequences. The grammar lesson included more discourse shifts, which accounted for the greater frequency of extended latencies. Five of the participants observed the videotaped lessons and explained their latencies. Participants indicated that they had greater difficulty understanding the instructor in the grammatically-oriented lesson than in the conversational one. Based on the results, the author argued that focus on form should be transparent to students as an instructional objective, at least at low levels of proficiency, because learners are more dependent on broad discourse to interpret meaning and respond appropriately.
Group work in the L2 classroom is widely considered a viable way to provide students with increased input and output opportunities in the classroom. However, teachers often remain skeptical about its use for fear that it will increase the use of L1 and off-task talk in the classroom. Alley (Reference Alley2005) addressed these concerns in a study that analyzed talk among five groups of high school learners of Spanish in the US. Results showed that the majority of discourse was indeed conducted in L1 (71%) and there was off-task behavior (21%), but most of the talk was task-related. Survey data collected from the same students showed positive attitudes towards group work. One caveat in the interpretation of these results is that the activities required learners to prepare a role-play situation for enactment later, which is different from asking them to converse in Spanish. It is therefore possible that the high percentage of L1 use may have been caused by task type.
The social and cognitive aspects of error repair during classroom interaction among learners of Spanish was the focus of Buckwalter (Reference Buckwalter2001), who, looking at interaction from a sociocultural framework, tried to unveil the processes at play when learners face language difficulties and struggle to overcome them. The study of video and audio recordings of first- and second-year learners showed a marked preference for self-repair, much as with L1 speakers. Collaborative repairs tended to be lexical, mostly to improve comprehension. During interaction, learners showed an ability to recognize errors and fix them, for which they used private speech, repetition and rhetorical questions as self-regulatory mechanisms. The study underscores the benefits of group interaction in the classroom for the opportunities for self-repair and output refinement it provides.
Cognitive-oriented input processing research has paid particular attention to the kinds of feedback and manipulation of input that aid acquisition. In that tradition, Rosa & Leow's (Reference Rosa and Leow2004) experiment with 100 advanced learners of L2 Spanish investigated the effect of computerized problem-solving task conditions on short and long-term recognition and production of a target structure. The design included exposition of Spanish past conditional sentences under six conditions with different degrees of ‘explicitness’ through a combination of pre-task with grammatical explanation, feedback concurrent with input processing, and variation of implicit/explicit feedback. Processing input under more explicit conditions had a positive effect on production of the target structure. The results suggest that explicit feedback could be beneficial not only in comprehension but also in production. Building on previous research on feedback and attentional processes, Leeman's (Reference Leeman2003) well-informed experimental study of recasts during an oral interaction task with first-year learners of Spanish suggested that recasts are most effective when they involve enhanced salience of positive evidence. Tight control of experimental conditions allowed Leeman to go further than previous studies in identifying those features of recasts that are quantitatively most significant in aiding acquisition.
In general, interaction and feedback studies based on L2 Spanish data have made great strides in helping the field of SLA understand learning affordances during teacher–learner and peer interaction. Cognitive theory and sociocultural theory have proved to be productive frameworks for the study of interaction. The innovative use of relevance theory (Toth Reference Toth2004) opens a new door for the understanding of this topic.
2.2.4 Technology
The Internet is a source of information as well as a means of communicative exchanges in the L2. The Internet as a powerful resource for language teaching is the theme of Arrarte & Sánchez's book (Reference Arrarte and Sánchez de Villapadierna2001), which lists useful Internet resources for teachers and students of Spanish together with guidelines for designing Internet-based activities.
Studies on the effects of technology use have been abundant. Students’ attitudes towards the use of e-mail in a distance-learning Spanish language course in New Zealand were reported in González Tuñón (Reference González Tuñón2004). Survey responses from the 16 students on the course indicated that most students valued the use of e-mail for its convenience and ease of communication, rapid feedback, increased L2 input and opportunities for authentic communication with the teacher.
The degree of effectiveness of online language courses has been a cause of concern among language teaching professionals. Despain (Reference Despain2003) compared differences in achievement and attrition rate between a beginner-level Spanish course and an online version of the same course delivered through streaming video of the regular course. Several tasks were designed so that students in the online group maintained their engagement with the course through daily e-mail submissions, phone conversations and submission of voice samples. Similar achievement test scores for all students who completed the course indicated that Internet technology can provide an almost identical experience for motivated students. However, the attrition rate was significantly higher on the Internet course. Because the study did not control for differences among the groups in initial language achievement and motivation, results must be interpreted with caution. Blake & Delforge (Reference Blake, Delforge, Salaberry and Lafford2007) conducted another comparative study on the effectiveness of an online university Spanish course versus a traditional one. Based on results of a multiple choice test on grammatical accuracy, short compositions, attitude surveys and sound/text chat interactions from 21 online students and 46 traditional students, the researchers reported that online students had higher scores on grammar, but there were no significant differences in the writing samples. The responses to surveys were very few in number (eight students); overall, students valued being able to work at their own pace and the low level of anxiety in online courses. The surveys also indicated that there were important differences in motivation among groups. The traditional students were mostly taking the course to satisfy a language requirement, while the online students declared personal interest as their primary motivation. The results indicate that online courses may provide a positive experience for language learners, a finding consistent with previous studies in this area. However, the small number of participants in this study and the differences in student motivation and class size also call for caution in interpreting the results.
Darhower (Reference Darhower2006) analyzed chat-room communication between university L1 Spanish learners of English in Puerto Rico and L1 English learners of Spanish in the US. Student responses to a questionnaire and transcripts of chat sessions showed how learners became integrated into a bilingual speech community and how technology provided opportunities for intercultural communication and integration of other foreign language standards (ACTFL 1996). In a subsequent article, Darhower (Reference Darhower2007) analyzed group dynamics in the same bilingual chat. While one group was able to establish a cooperative learning environment, another did not. Looking at their interaction through the metaphor of communities of practice cast light on the linguistic and social behavioral patterns that were at play in the creation of a community. Full participation in virtual communities and the achievement of a degree of camaraderie were dependent upon the degree of cohesion reached by its members. Similar conclusions were reached by Vinagre Laranjeira's (Reference Vinagre Laranjeira2005) study of weekly e-mail exchanges between L2 English learners in Spain and L2 Spanish learners in the US. In her study of learners’ perceptions of chat exchanges with native speakers, Lee (Reference Lee2004) also concluded that differences in language proficiency as well as in age and computer skills condition the quality of the interaction.
In the area of technology, there is a pressing need for solid, carefully controlled, large-scale comparative studies of online versus classroom learning. The results would be important for pedagogical practice and program administration. Computer-mediated interaction studies show great promise for the potential effect of technology on language and cultural learning. As there is considerable overlap with studies of classroom interaction, pragmatics and intercultural competence, future studies should make explicit connections with these areas and explore further the theoretical contribution of interaction through technology to SLA.
2.2.5 Motivation and individual factors
Motivation is considered a major factor in language learning. Hernández (Reference Hernández2006) investigated the effects of three variables – integrative motivation, instrumental motivation and the need to fulfill a university language requirement – to predict Simulated Oral Proficiency Interview (SOPI) scores and intention to continue language study. A questionnaire was administered to 130 university-level students. Results revealed that integrative motivation was a significant predictor of scores and of a desire to continue language study. However, there was a negative relation between completing the language requirement and a desire to continue language study. Based on these results, the author recommended classroom activities that promote integrative motivation by interacting with the community through interviews, service learning, immersion experiences and authentic materials from the L2 culture.
Anxiety in the L2 classroom is believed to inhibit learning. Starting with the belief that students’ emotional concerns must be addressed, Ewald (Reference Ewald2007) conducted a qualitative study of classroom anxiety among advanced learners of Spanish. Questionnaire responses revealed that most students experience anxiety in upper-level courses and that the teacher plays an important role in both producing and relieving anxiety. Sources of anxiety were a lack of confidence in their ability in Spanish, perceived high language ability in peers, and teachers’ attitudes towards error correction. An original aspect of this study is that students from an upper-level methods class were engaged as co-researchers in the analysis of their peers' responses to questionnaires. Practical suggestions for the creation of a low-anxiety student-centered learning environment were listed. Among them were demonstration of the teacher's support and encouragement, and open discussion of anxiety.
Schulz & Elliot (Reference Schulz and Elliott2000) focused on the acquisition of Spanish by one of the authors, an older adult, during a five-month stay in Colombia. Data from diary entries and language proficiency pre-tests and post-tests showed that learning took place despite the author experiencing some cognitive-aging related difficulties such as slow word retrieval, interference from other languages and oral comprehension problems. Post-tests showed progress in proficiency level (from low intermediate to high intermediate) and overall improvement in lexical and verb tense usage. Diary entries also recorded the learner's reactions to classroom activities, in particular, negative emotions and frustration with role-play activities with non-native peers, and insufficient feedback and error correction. The authors suggest that older adults may be more critical of teaching approaches that do not meet their needs and expectations.
Studies in this area draw attention to the importance of the socio-affective aspects of language learning. Methodologically, many rely on descriptively rich qualitative interpretations, which may not be generalizable to other contexts and may not provide solid evidence of causal relations. More research studies should be encouraged to adopt mixed quantitative and qualitative methods to explore both the breadth and depth of these issues.
2.2.6 Study abroad and immersion contexts
An increased emphasis on internationalization and large numbers of students attending language programs abroad has alerted attention to the effect of study-abroad experiences on language learning. There is a general belief that study abroad is beneficial for language learning because of the opportunities it provides for immersion and increased exposure to the language. In the studies included here, special attention has been placed on measuring gains and understanding the processes at play during study-abroad learning.
A combination of linguistic, psychological and social variables is responsible for the effect of study abroad on language learning. This point is illustrated by Bacon (Reference Bacon2002), who set out to study the interplay between experiences in and out of the classroom through an ethnographic study of one student during one semester of a year-long stay in Mexico. Qualitative analysis of three oral interviews, three written samples and participant observation in classes and field trips revealed language development and cultural and academic adjustment that resulted from a dynamic relation between formal and informal contexts of learning. Although non-academic contexts were the source of much of the learning experienced by the student, that learning was transferred back into the classroom. Based on the positive changes experienced by this student, the author recommends encouraging students to talk and write about their experiences to facilitate cultural adjustment and language growth. In an interesting and comprehensive study of the impact of one semester abroad in Argentina on five American students, Isabelli-García (Reference Isabelli-Garcia2004) found that two factors stand out as leading to acquisition in a study abroad context: motivation and significant interaction with native speakers. Quantitative data from SOPIs and informal interviews, and qualitative data from weekly diary entries and network contact logs, showed that four learners improved their ratings. The student who stayed at the same level improved linguistically, but her responses were short and not elaborate. Diaries showed that she was sensitive to rejection from the host culture and that episodes of rejection had decreased participation with native speakers. Variation in linguistic improvement was therefore related to interconnectedness of motivation, attitudes and social networks.
A study of comparative attitudes towards short-term study abroad programs held by learners of Spanish from the US in Mexico and European students in Spain was reported by García Laborda & Bejarano (Reference García Laborda and Bejarano2008). Questionnaire responses showed positive attitudes towards the study-abroad experience. A concern, particularly among the European learners, was the lack of opportunities to interact with local residents. Due to uneven group size and time that elapsed between the experience and data collection, the results can only be suggestive, but the study of similarities and differences between study-abroad experiences in America and Europe is a good step towards increased communication and the creation of joint research programs on both sides of the Atlantic.
The studies discussed above provide evidence of changes in students who participate in study abroad, but say nothing about how their language development differs from those who study Spanish in regular programs in their home country. A battery of comparative studies of at-home versus study-abroad learning were published in a special issue of Studies in Second Language Acquisition (Collentine & Freed Reference Collentine and Freed2004). Data from the same group of participants in a semester-long study-abroad program were analyzed for linguistic gains in oral proficiency, oral fluency, grammar, vocabulary, pronunciation and communicative strategies. Segalowitz & Freed (Reference Segalowitz and Freed2004) compared gains in oral proficiency and oral fluency in the two groups of students. Oral proficiency interview (OPI) scores were significantly higher for the study-abroad group. Oral fluency, as measured by comparing rate of speech, mean length of utterance without fillers, and length of longest utterance, improved only in the study abroad group. Looking at recorded speech samples, Collentine (Reference Collentine2004) conducted a corpus-based multivariate comparative analysis of grammatical and lexical gains in the two groups. A lexical item count and analysis of selected grammatical features revealed advantages for the at-home group. However, when looking at narrative ability (use of subordinate clauses) and lexical density, the study-abroad group outperformed the at-home students. In the area of pronunciation, Díaz Campos (Reference Díaz Campos2004) found no difference in a reading test between the two groups. Lafford (Reference Lafford2004) compared the use of communicative strategies during two-minute excerpts from OPIs by the same at-home and study-abroad learners. Quantitative results showed that both groups reduced their use of communicative strategies over time, but the study-abroad group showed a significantly decreased use, which seems to have been caused by a more developed pragmatic ability following more interactional practice with native speakers. Overall, the study-abroad group showed increased narrative ability, discursive abilities, increased fluency and less focus on form, whereas the at-home group showed greater grammatical ability. This project provides an excellent model of collaborative research that yields sound comparative results from the same set of learners.
Framed within the larger issue of acquisition of intercultural competence – currently a topic of great interest in Europe – Rollin (Reference Rollin2006) discussed the training that British learners received in ethnography and other tools for the study of culture in preparation for a year of study in Spain, which was then followed by an ethnographic case study that the students developed while abroad. Since the preparation of students for study abroad is not a widespread practice, this article is useful in drawing attention to its importance and illustrating effective ways to equip students for intercultural experiences. Future studies of the impact of training could provide insights on how to maximize the value of cultural immersion experiences.
2.2.7 Assessment
In the area of assessment I wish to highlight three recent developments: the creation of international standardized proficiency tests, language program assessment as a catalyst for change in students’ assessment, and the application of dynamic assessment techniques to L2 Spanish contexts.
Widely accepted national and supranational professional standards and curricular plans (ACTFL 1996, Council of Europe 2001) have continued to drive the development of assessment techniques. A good overview of historical trends in assessment is provided by Bordón (Reference Bordón2004), who noted, in particular, alternative assessment methods such as oral interviews and the use of portfolios. The European Language Portfolio is novel in the way it balances external and internal evaluation (through reflection on the learning experience and self-evaluation of language proficiency). Parrondo (Reference Parrondo2004) reviewed the issue of ethics in testing, which has played an important role in assessment since the 1990s, resulting in higher standards for tester certification and a code of ethical practice that has been adopted by testing institutions like ALTE (Association of Language Testers of Europe), Instituto Cervantes and the Educational Testing Service.
The past decade has witnessed the development of new international Spanish proficiency tests, including the Certificado de Español Lengua y Uso (CELU) in Argentina, and the Examen de Posesión de la Lengua Española (EPLE) and Certificado de Español como Lengua Adicional (CELA) in Mexico, in addition to the existing Oral Proficiency Interview (OPI) developed by the American Council on the Teaching of Foreign Languages in the US in the 1980s and the Diploma de Español como Lengua Extranjera (DELE) in Spain in the 1990s. The 1991 DELE exams have undergone important revisions since 2002 as a result of the requirement for compatibility with the Common European Framework of Reference (Prieto Hernández, Díez Santos, Domínguez López & Martín Maestre Reference Prieto Hernández, Díez Santos, Domínguez López and Martín Maestre2004). An exciting initiative, described by Parrondo (Reference Parrondo2008), is the current effort to create an international system of certification of L2 Spanish (Sistema Internacional de Certificación de Español como Lengua Extranjera, SICELE), which will unify criteria and standards for Spanish proficiency tests from different countries along European and North/Latin America guidelines.
Alongside the interest in large-scale standardized tests, language program assessment has attracted recent attention in the US. Prompted by accountability models as well as calls from professional organizations (Glisan & Phillips Reference Glisan and Phillips1996, MLA Reference MLA2007)Footnote 2, university language programs have started to redesign their curricula and to measure learning outcomes in their graduates. Oxford (Reference Oxford2008) described revisions to a traditional Spanish language major that involved collaboration among faculty in revising the curriculum so that it would align better with the National Standards for Foreign Language Learning (ACTFL 1996). The creation of a departmental curriculum committee that oversees the program's coherence in course structure was instrumental in the improvement of the program. Moreover, the introduction of written and oral proficiency exams, attention to language development in all courses, and a research requirement before graduation were deployed as instruments of program and student assessment. The use of students’ electronic portfolios in program assessment was the subject of Morris & Cooke-Plagwitz (Reference Morris and Cooke-Plagwitz2008). The creation of individual electronic portfolios, including one artifact from each course and reflective pieces on each artifact, was a requirement for graduation. Portfolios were evaluated for cultural competence, reading, speaking and writing ability. Despite some technical problems and initial resentment of the requirement by students, the program assessed that over 90% of their graduates were able to meet their standards. Another innovative method of student assessment is through an experiential learning requirement (study abroad or service learning). Long & Macián (Reference Long and Macián2008) described how students who do not study abroad are required to participate in 100 hours of voluntary service and cultural activities using Spanish. All of these are exciting developments in that the effectiveness of the program is now being assessed by students’ performance, which in turn brings about curricular changes.
Favored by recent attention to the role of interaction in L2 assessment, dynamic assessment (DA) is beginning to attract interest and gain acceptance in L2 education circles. DA, which has its roots in sociocultural theory, relies on the agentive role of the examiner during a testing situation and places great value on the quality of the interaction between examiner and examinee in order to assess not only actual competence, but also learning potential. Spanish has been among the first L2 contexts to which DA has been applied. For example, Oskoz (Reference Oskoz2005) analyzed peer-scaffolded interactions among learners of Spanish while completing collaborative tasks in a chat environment. She suggested that the process of task completion was more revealing of the learners’ capabilities than the product. The use of DA in L2 contexts is still scarce, but it is an area worthy of attention in the coming years for its potential to provide a more complete assessment of learners’ abilities.
Overall, assessment of L2 Spanish tends to be local, and the results of large-scale standardized testing have not attracted much attention from researchers. The combination of using international standardized testing with other qualitative assessments has great potential for comprehensive and comparable assessments of L2 Spanish learners.
3. Heritage learners and bilingual programs
The large number of Hispanics in the US has created the need for instruction geared towards speakers who have had contact with the language outside the classroom. Heritage learners have a set of linguistic needs that differ from those of other learners; thus, existing programs are inadequate for them. Most heritage learners have not developed proficiency in registers of Spanish that will allow them to function in academic and professional settings. Professional organizations and institutions such as the American Association of Teachers of Spanish and Portuguese, the National Heritage Language Resource Center and the Center for Applied Linguistics have placed particular emphasis on these learners through conferences and the publication of research and teaching materials. The celebration of a National Conference on Heritage Languages in 1999 and 2001 and the creation in 2003 of an online journal, The Heritage Language Journal, have promoted academic study of this area of inquiry. The Heritage Language Research Priorities Report (UCLA Steering Committee Reference UCLA Steering Committee2000) identified a series of issues around the heritage speaker: family, community, language focus, policies, programs and assessment. Many of the studies conducted in recent years are connected with these issues.
Because of its important effect on motivation and learning, the sociolinguistic status of heritage learners’ language and attitudes towards it have been investigated in several studies. Bernal-Enríquez & Hernández-Chávez (Reference Bernal-Enríquez, Hernández-Chávez, Roca and Colombi2003) pointed out the detrimental effects of negative attitudes towards Chicano language by teachers. To increase students’ motivation they proposed including Chicano language variety and culture in classrooms as a point of departure for the acquisition of the standard variety. Beckstead & Toribio (Reference Beckstead, Toribio, Roca and Colombi2003) incorporated students’ perspectives by studying language attitudes of Latino high school students in a bilingual program in California. Results of a survey indicated that the students place high value on Spanish as a symbol of identity and on English for its practical value.
The description of heritage learners’ grammatical abilities and the acquisition of academic discourse have been at the center of research on heritage learners for the past decade. Fairclough & Mrak (Reference Fairclough, Mrak, Roca and Colombi2003) interviewed bilingual Mexican–Americans who were born in the US or who came from Mexico before the age of six. They found a low percentage of morphosyntactic errors and use of code-switching in their speech, and no difference in the number of errors between those who had some formal instruction (less than four years) and those who did not. In a particularly well-developed study, Fairclough (Reference Fairclough2005) investigated the effects of formal instruction on the acquisition of conditionality by Hispanic university students learning the prestige variety of Spanish (a second dialect). The cross-sectional study included a wide variety of written and oral data sources (short paragraphs, a cloze test, an acceptability judgment test and a semi-structured interview) from 141 heritage learners (HL) and 142 foreign language learners (FL). Differences, both quantitative and qualitative, emerged between the groups: the HL group produced a higher percentage of target language hypotheticals, but the FL group displayed marked improvement between levels. In errors, there was more variety in the FL group, particularly uses of the indicative, while the HL group used either target language forms or non-standard forms. Forced output tasks produced a lower percentage of target language forms, but there was avoidance of target forms in open tasks. What is interesting is that non-standard forms do not disappear; thus, acquisition of a second dialect seems to be additive, creating a more complex linguistic system, not a separate system. Although this study is limited to the acquisition of only one grammatical feature, it represents an important step towards understanding the process of acquisition by heritage learners and it points to some significant differences between L2 and heritage learners. The variety of data sources and the thoroughness of analysis provide a good model for research of other aspects of learners’ language. Further, longitudinal, studies would be a good complement to this kind of work.
The acquisition of academic discourse is the main objective of most programs for heritage learners. Achugar (Reference Achugar, Roca and Colombi2003) justified this objective explaining that, because language is a basis for the assessment of a person's identity, bilinguals who have not been exposed to academic registers will be at a disadvantage in trying to establish themselves as experts. Achugar's recordings of oral presentations in two graduate seminars showed differences in linguistic and discursive features that marked differences in knowledge of the academic register by Spanish monolinguals and bilinguals. Colombi (Reference Colombi2001) followed longitudinally the development of academic discourse in 30 heritage Spanish speakers. Grounded in functional systemic linguistics, the study revealed that the use of grammatical metaphor, mainly nominalization, was the most effective resource for creating academic register.
The search for appropriate methods and effective programs for heritage learners has also been an important area of research. Looking at programs for heritage learners from a language maintenance perspective, Valdés, Fishman, Chávez & Pérez (Reference Valdés, Fishman, Chávez and Pérez2008) reported on survey data from 104 high school and college programs in California. Based on the institutional profile that emerged from the surveys, the authors warned that the current focus on the traditional academic goals of linguistic correctness and acquisition of the standard dialect might accelerate the abandonment of the home language. A departure from traditional methods was proposed by Nichols & Colón (Reference Nichols and Colón2000), who discussed the use of code-switching in the classroom by bilingual students and teachers as a linguistic resource, instead of a liability. The authors examined instructional uses of teacher talk in the students’ L1 to make directions clear, to give explanations, to summarize, and to establish and maintain rapport. They concluded that the teacher's acceptance and use of code-switching created an academic environment that validated students’ language patterns and contributed to their academic success. The AATSP Handbook Spanish for Native Speakers (AATSP 2000) and Potowski, Berné, Clark & Hammerand (Reference Potowski, Berné, Clark and Hammerand2008) include model activities for heritage learners that will be useful for teachers who wish to include heritage learners’ language and culture in the classroom.
Among bilingual programs, content-based immersion is considered the best way of developing language proficiency. In a longitudinal study of a dual immersion school, Potowski (Reference Potowski2007) showed positive results in terms of understanding of culture, attitudes towards Spanish and Hispanics, standardized tests, respectable levels of proficiency in Spanish and investment in the language. However, her analysis of language use and accuracy showed some negative results, the most outstanding of which was the failure to produce balanced bilingualism. The study combined quantitative analysis of language production patterns and language proficiency with qualitative analysis of student identity investment, which provided a deeper picture of the social dynamics at play in language learning and language use. Participant observation and classroom recordings showed that in fifth grade Spanish was used only 56% of the time, mostly when the students were talking to the teacher about academic topics. This percentage decreased dramatically by the time these same students reached eighth grade. Participant observation, journals, questionnaires and interviews with four students, parents, teachers and administrators revealed that investment in Spanish identity, not language contact with native speakers, leads to more use of Spanish. In terms of the robustness and novelty of the methodology and the richness of the results, this research provides a good model for future studies of bilingual programs and heritage language learners.
Despite the recent attention given to heritage learners, researchers are alarmed by the scarcity of theoretical and empirical research, together with a lack of validated teaching practices and professional training for teachers (Valdés Reference Valdés, Salaberry and Lafford2006, Valdés et al. Reference Valdés, Fishman, Chávez and Pérez2008). The studies reviewed in this section represent an advance towards remedying these deficiencies. Future research should keep in mind the need for carefully designed studies that contribute to the development of a sound theory of heritage language acquisition and, hence, improved ways of teaching. To that end, the research agenda delineated by Valdés (Reference Valdés, Salaberry and Lafford2006), emphasizing in particular the study of the linguistic systems of heritage learners, effect of instruction, and proficiency assessment, seems appropriate. Research should also be expanded to other countries with a growing presence of heritage learners, such as Canada, where different social conditions may raise other issues. Advancement in this area will require a concerted effort among researchers to conduct collaborative projects or, at least, to build on each other's results and research methodology.
4. Immigrant language education in Spain
Since Spain joined the European Union in 1986 and entered an era of economic growth, the country has experienced a rapid increase in immigrants from various ethnic, socioeconomic and educational backgrounds. The presence of immigrant children in the public education system brought L2 Spanish into the curriculum. Adult immigrants also created a need for L2 Spanish classes. Existing materials for teaching Spanish to educated foreigners have proven inadequate for teaching immigrants from low socioeconomic backgrounds. Teaching Spanish to immigrant adults and children presents many challenges, from the choice of appropriate methods and materials, through teacher education, to developing intercultural understanding. Several books on the topic have appeared, together with special journal issues, and a web forum, active since 2001 and hosted by the Instituto Cervantes. An electronic journal devoted to language learning by the immigrant population (Segundas lenguas e inmigración) was started in 2008. The studies selected for this section reflect the challenges faced by language teaching professionals.
The sociolinguistic background of immigrants emerges as an important topic when establishing curricular needs and designing effective methods and materials. In a long-term study of Sub-Saharan African immigrants, Ambadiang (Reference Ambadiang2003) underscored the importance of understanding their sociocultural background when creating relevant and motivating language programs. The prevalence of communicative goals in their use of other languages causes imperfect learning of the grammatical system and heavy reliance on negotiation of meaning with interlocutors. Soto Aranda & El-Madkouri (Reference Soto Aranda and El-Madkouri2002) also called attention to the contextual factors that affect L2 Spanish learning by Moroccan immigrants. They argued that age of arrival, gender, the degree of integration in the host culture and their employment situation play a role in the degree of proficiency that immigrants attain, since they determine the opportunities to interact in the L2.
The relation between linguistic proficiency in L2 Spanish and sociolinguistic factors was explored by Navarro Sierra & Huguet Canalís (Reference Navarro Sierra and Huguet Canalís2003) in an extensive study of 49 immigrant schoolchildren and a control group of 44 local schoolchildren in Huesca (Spain). Questionnaires, language proficiency tests and teacher surveys provided data on language gains and their relation to relevant variables. Significant differences were found in language proficiency levels among L1 Spanish immigrants from Latin America, L2 Spanish immigrants and local children. Language gains were higher among children who immigrated before the age of ten and those who had been in the host country for six or more years. Results with regard to age of arrival and number of years in the country are particularly important because they show that it takes years to develop the cognitive academic language proficiency necessary to succeed in school. This is an important study for its systematic approach to data collection, analysis and reporting of results. Some of the same points relating to the effect of age and the acquisition of cognitive academic linguistic proficiency were emphasized by Soto Aranda (Reference Soto Aranda2008) with respect to Moroccan immigrants. For children who immigrated before age six, the main issues have to do with L1 loss or maintenance, and the interplay of L2 Spanish and other L2 co-official languages in bilingual communities.
The variables that affected language acquisition by recent immigrants in secondary schools in León (Spain) was the object of a study by Cabañas Martínez (Reference Cabañas Martínez2008). The number of immigrant children in this area was small (around 1% of students), which is probably the cause of the unpreparedness of schools and lack of adequate teacher training and materials revealed by questionnaires and interviews with teachers and school administrators. Despite the small number of participants in the study (n = 36), the study shows the value of this approach to gathering information that can be used by educational institutions to better serve these learners.
Several publications have attempted to fill the gap by providing methodological and curricular guidelines. Villalba Martínez & Hernández García (Reference Villalba Martínez and Hernández García2001) published two function-based curricular plans (one for primary and one for secondary schools). García Mateos (Reference García Mateos2004) presented a curriculum designed with adult literate and illiterate immigrants in mind. The curriculum is grounded in the sociolinguistic context of immigrant learners, including descriptions of the learning preferences and challenges of each community based on teachers’ observations. Significantly, the book emphasizes the need to individualize instruction for groups of learners sharing common characteristics (L1 similarities, illiteracy) and to prioritize oral skills and functional language to enable immigrant learners to function in their new society.
Also important for curriculum design is a description of the speech immigrant children encounter in mainstream classrooms. In a data-driven study, Villalba Martínez & Hernández García (Reference Villalba Martínez and Hernández García2004) analyzed the expository speech of six high school teachers. The authors noted that transmission of information is often interwoven with language that serves other purposes, from answering questions to disciplining students. A key factor is the students’ ability to recognize supporting language that facilitates information processing, language that anticipates main points, comprehension requests (¿de acuerdo?), and language that helps organize the text (lo primero. . .). This description is useful for L2 teachers and materials designers because it calls attention to discourse features that should be included in the L2 Spanish curriculum if learners are to be successful in immersion school experiences. At the same time, the description of teacher talk invites future studies on learners’ comprehension of language and content as well as investigation of the discursive mechanisms that facilitate language learning in mainstream classrooms.
In conclusion, almost two decades after the wave of immigration began in Spain, educational institutions are still struggling to find ways of helping immigrant learners acquire the L2. Recent publications have familiarized practitioners with principles of language acquisition, the sociolinguistic characteristics of immigrant groups that may affect language learning, and methods and materials that may help in achieving that learning. In future, this area of study would benefit from extensive empirical studies that test the theories, curricular proposals and pedagogical techniques that have been suggested. The results should be compared with the experiences of others who are teaching language to immigrants in Europe and in English-speaking countries.
5. Bilingual programs for indigenous populations
Bilingual education programs for indigenous populations exist in several Latin American countries. López's (Reference López2006) historical overview reminds us that indigenous societies in Latin America are still underprivileged, even in countries where the indigenous population is in the majority, such as Guatemala and Bolivia. There are over 40 million indigenous people throughout the continent: 10% of the population. With over 400 different languages used in the region, bilingual education is now enjoying the best period in its 70-year history, with programs in 17 countries. A good overview of the history and current state of bilingual education in Latin America (1994–2004) is the book edited by López & Rojas (Reference López and Rojas2006), which includes chapters on bilingual education in Bolivia, Ecuador, Guatemala and Peru. Particular attention is given to the need for greater emphasis on teaching L2 Spanish. At the same time, indigenous organizations and families are increasingly asking governments to refocus objectives, from the transition to L2 Spanish to L1 maintenance. Educators, researchers and policy makers will benefit from this comprehensive account of the state of intercultural bilingual programs in four countries with a large indigenous population. At the same time, researchers in the various countries should be encouraged to continue building on each other's research and to refine research methodologies, so that sound comparisons of bilingual programs may guide their future development.
The case of bilingual education in Paraguay stands out because of its traditional situation of widespread stable bilingualism and positive attitudes towards the indigenous language, Guaraní. A comparative study of large-scale attitudinal surveys conducted among educators in 1995 – a year after the implementation of bilingual programs in the country – and again in 2001, shows an increase in the number of bilingual speakers and a shift towards more positive attitudes about bilingualism in general and about literacy in Guaraní (Gynan Reference Gynan Shaw, Sayahi and Westmoreland2005). However, Choi (Reference Choi2003) warned that attitudes and use do not always go hand in hand. Her survey of parents and high school students in Asunción concluded that despite positive attitudes towards learning Guaraní, most students (85–97%) felt more comfortable using Spanish and use Guaraní mostly in Guaraní classes and when visiting rural areas. Thus, with the caveat that the study is based on a survey of participants’ perceptions and not actual behavior, it can be concluded from these studies that bilingual education in Paraguay has been most effective in increasing positive attitudes, but that actual language use seems to be shifting towards Spanish.
In assessing the effectiveness of bilingual programs, great emphasis has been placed on the comparison of academic and language achievement by monolingual and bilingual students.
Whereas socio-political factors that are outside the scope of this review affect the state of bilingual education in Latin America, the studies reviewed below have examined levels of achievement in L2 Spanish and offer assessments of methods, classroom instruction and materials. Based on observations of bilingual programs on the Nicaraguan Caribbean coast in 1994, repeated more recently, McClean (Reference McClean and Muñoz2006) concluded that no significant changes were observed in that period of time. Although progress had been made in L2 Spanish textbook design with changes from traditional to more communicative materials, bilingual programs suffer from a lack of proficiency in Spanish on the part of teachers, a lack of opportunity to use Spanish in the community, and a lack of teacher training in methods of teaching L2 Spanish. López (Reference López2005) used information from 153 Bolivian rural communities collected through informal interviews and focus groups with teachers, school administrators, children and the community to report on fifteen years of intercultural bilingual programs. There was general agreement that the educational reform brought about by intercultural bilingual programs since 1994 had a very positive effect on children's self-esteem and greater participation in the classroom. Family concerns remained regarding the development of writing abilities and of proficiency in L2 Spanish. This is an important expectation that, López admits, intercultural bilingual programs have yet to meet satisfactorily.
King (Reference King2001) reported on a year-long ethnographic study conducted in two Quechua communities in Ecuador undergoing language shift towards Spanish. Her ethnography provides us with an inside look at language instruction in the bilingual schools of these communities. Regarding language learning, she found that most instruction was conducted in Spanish. However, an examination of classroom activities revealed that copying, handwriting and other activities that do not promote language development took up most of the time. King's description gives concrete examples of how the effectiveness of the programs is limited by, for example, a lack of materials, unsound pedagogical practices and a lack of trained teachers.
Given the methodological shortcomings mentioned above, Arévalo, Pardo & Vigil (Reference Arévalo, Pardo and Vigil2004) and Walqui & Galdamés (Reference Walqui and Galdamés2005) provide teachers with a much-needed introduction to language teaching tailored to their specific context. Arévalo, Pardo & Vigil's (Reference Arévalo, Pardo and Vigil2004) book highlights the importance of teachers’ understanding of the social context of learning in their particular community and provides tools (questionnaires, observation guidelines, etc.) to enable teachers to research the psycholinguistic and sociolinguistic factors that may affect learning, guidelines for the use of didactic materials created for bilingual intercultural programs, and sample assessment instruments.
Two original field studies uncover other important factors in bilingual program implementation. Limachi Pérez (Reference Limachi Pérez2000) conducted a qualitative study in the Cutusuma aymara-speaking community of Bolivia on the effect of administrative procedures and classroom management on the acquisition of L2 Spanish. Limachi argued that, while efforts to improve the situation have been directed towards materials development and methods workshops for limited numbers of teachers, other aspects of program implementation affect students’ low level of competency in Spanish. In particular, the lack of a coherent administrative organization and communication between parents, teachers and administrators, in addition to lack of access to L2 Spanish materials and teacher training are to blame for inappropriate implementation of national guidelines. On the other hand, Marzana Téllez's classroom ethnography (Reference Marzana Téllez2005) is illuminating in its understanding of the effect of attitudes. Her detailed study of three bilingual teachers in a Quechua community in Bolivia showed how the teachers’ positive attitudes towards the principles behind the teaching program resulted in a favorable implementation, despite lack of materials, training and other resources.
In conclusion, the teaching of Spanish in Latin American bilingual programs has moved from the rather unsuccessful submersion methods of the past to more up-to-date theoretical and methodological approaches to L2 teaching. Since assessment studies show that implementation of government-proposed approaches is still far from complete, a great deal of effort has been put into familiarizing teachers with theoretical principles, appropriate techniques and materials for the teaching of L2 Spanish. The improvement and assessment of L2 Spanish language proficiency stands out as an important issue for bilingual programs at this time. It is hoped that in the coming years more empirical studies will be conducted assessing the effect of specific interventions on the development of Spanish language skills among students in bilingual programs. Detailed and rigorous quantitative and qualitative studies are also needed of students’ proficiency in Spanish so that instruction and materials may better address their needs. A promising future line of research, for theoretical and practical purposes, would be the interplay of children's L1 and L2 in the acquisition process.
6. Conclusion and future directions
The richness and variety of contexts for teaching and learning L2 Spanish provide fertile ground for important contributions to our understanding of the processes, products and variable nature of SLA in the coming years.
This review has shown that the considerable amount of work on L2 Spanish represented by the publications reviewed has contributed to and expanded the wider field of L2 learning and teaching. Overall, research on L2 Spanish in North America, due to its longer history, follows research agendas that are both theoretically grounded and informed by research on other languages. In Spain and Latin America, where the field of L2 Spanish is relatively new, the diversity of learner backgrounds and learning situations presents an opportunity for advancing our knowledge of learning processes through empirical (experimental-psychometric and socio-scientific) research that builds on current SLA studies.
Some new lines of research hold promise for the future, such as the effect of technology (in the form of global communication) on the development of communicative and intercultural competence, and the effect of motivation, contextual and individual variables on the acquisition process. Other areas with great research potential that remain comparatively unexplored include the acquisition of L2 Spanish in bilingual programs in Latin America (e.g. the levels of proficiency attained through different programs, the interplay between motivation/attitudes/acquisition, the effect of discursive practices in and out of the classroom, and the comparison of these programs with bilingual programs in other countries) and the acquisition of L2 Spanish by immigrants in Spain, particularly children in Spanish schools. This line of research is closely related to current research on ‘generation 1.5’ in the English-speaking world (Harklau, Losey & Siegal Reference Harklau, Losey and Siegal1999). Similarly, research into learning Spanish as a foreign language in both Spain and Latin America may add insights into the effect of long-term study abroad programs, cultural and linguistic immersion on the acquisition of linguistic and intercultural competence.
Regarding research methodology, more studies would benefit from widening perspectives by combining quantitative and qualitative approaches. For instance, the study of acquisition processes should not be completely detached from the social context of learning and other individual variables that are difficult to capture quantitatively. A mixed-method approach is increasingly being used and the richness of its analysis is illustrated by several studies included in this review. Overall, more evidence-based quantitative and qualitative studies are to be encouraged. In addition, methodological refinement of study design by taking baseline measures, tightly controlling for variables, and reporting results systematically is still needed in a larger number of studies, particularly in those countries where empirical research on language acquisition and teaching is relatively new. Current emphasis on localized practical issues in some areas of study, such as error analysis, should now look towards constructing theory upon which sound methodology can be based.
Overall, the field would benefit from more exchanges between researchers in different geographical areas, and collaborative research projects that examine the same topic in different learning contexts.
Acknowledgements
I owe special thanks to Lluís Agustí (Instituto Cervantes New York), Oscar Alatriste (CEPE-UNAM, Chicago) and Carol Klee for directing me to bibliographical sources. I also thank Fred DiCamilla for his constructive comments, and my students, Adam Ballart and Wendy Garringer, for their bibliographical and editorial assistance. I am also grateful to the anonymous reviewers for their careful reading and helpful suggestions. All remaining errors are mine.
Marta Antón is Associate Professor of Spanish at Indiana University–Purdue University Indianapolis, USA, and Research Fellow at the Indiana Center for Intercultural Communication. She has taught L2 Spanish in Spain and the United States, and is an examiner for the Diploma de Español como Lengua Extranjera. She has published articles and book chapters on classroom interaction, sociocultural theory, dynamic assessment and Spanish sociolinguistics.
