A Multidisciplinary Approach to Applied Linguistics and Education

2.1 CLIL research on agreement morphology errors and null subjects in Spain

CLIL instruction emerged in the 1990s to cater for a need in Europe to promote foreign language teaching through innovative methods (see Marsh 2002; Eurydice 2006: 8). The term ‘CLIL’ was coined in 1994 (Marsh 1994) as an overarching term to cover those dual-focused educational approaches where the foreign language is integrated into the content classes, with the content subject being taught not ‘in a foreign language but with and through a foreign language’¹ (Eurydice 2006: 7; Coyle et al. 2010) (see Cenoz et al. 2014 for a discussion on the term). CLIL instruction thus presents itself as a very convenient pedagogical approach to promote the learners’ proficiency in the foreign language through increased exposure and, importantly, without detriment to the school curriculum, which undoubtedly explains its increasing popularity and rapid upsurge in the last decades.

CLIL instruction in Spanish contexts has proved to have a highly beneficial effect for the learners’ proficiency in English. Learners enrolled in CLIL programmes have thus been shown to outperform age-matched (Jiménez Catalán et al. 2006; Lasagabaster 2008) and even older (Ruiz de Zarobe 2008; Navés and Victori 2010) learners receiving only EFL instruction. With a few exceptions (see Martínez Adrián and Gutiérrez Mangado 2009; Villarreal Olaizola and García Mayo 2009), there is less evidence of a significantly positive impact of a CLIL approach on the acquisition of morphosyntax. In fact, most of the research – mainly carried out in Secondary Schools – has observed a very limited impact of CLIL compared to exclusively EFL instruction on the omission and commission error rate of verbal inflection and subject contexts in English (Martínez Adrián and Gutiérrez Mangado 2009, 2015a; García Mayo and Villarreal Olaizola 2011; Villarreal Olaizola 2011). This has also been confirmed for Primary School (Grade 6) learners, as it was found out that a very early (five/six years old) and long-term (five/six years) additional CLIL exposure resulted in some non-significant but, nonetheless, incipient positive trends (e.g. lower rate of placeholder is and null subjects) compared to the oral production of the EFL learners (Fernández-Pena and Gallardo-del-Puerto 2021). The (pseudo-)longitudinal studies by Vraciu (2020) and Martínez-Adrián and Nieva-Marroquín (2022) attest a limited longitudinal impact of CLIL instruction on the learners’ production accuracy of verb morphology, with the former finding a positive effect on the range of verb inflections used and the longitudinal progress of affixal omission.

Prior research has also confirmed that both CLIL (Villarreal Olaizola and García Mayo 2009; García Mayo and Villarreal Olaizola 2011; Villarreal Olaizola 2011; Vraciu 2020) and EFL (Ionin and Wexler 2002) groups show two well-attested asymmetries in the acquisition of L2 English morphosyntax. The first one concerns the earlier mastery of suppletive inflection – particularly copula be over the auxiliary – compared to affixal morphology. Most of the previous research has thus observed a higher omission of affixal inflection over suppletive inflection and a negligible rate of commission errors (i.e., wrongly inflected forms, e.g., *he are eating). Non-CLIL learners, typically at a less advanced stage of acquisition, have also been found to make considerable use of placeholder is, that is, of the suppletive form is to hold the inflection of the main verb (e.g., *he is sleep). The second asymmetry pertains to the earlier acquisition of referential subjects in contrast to expletive subjects (see Fernández-Pena and Gallardo-del-Puerto 2021 for CLIL contexts and Ruiz de Zarobe 1997, 1998, 2000 for EFL contexts). Both CLIL and non-CLIL Spanish learners of English, having a pro-drop L1, have proved to have great problems with the provision of obligatory subjects in general (Martínez Adrián and Gutiérrez Mangado 2009; Martínez-Adrián and Nieva-Marroquín 2022), but particularly of the purely grammatical and semantically empty expletive subjects that Spanish lacks.

2.2 CLIL intensity

Prior research on the impact of CLIL instruction on L2 English morphosyntax has controlled factors such as the learners’ English and CLIL onset age as well as their age at testing (Villarreal Olaizola 2011; Martínez Adrián and Gutiérrez Mangado 2015a), the long-term effects of the CLIL programme (Villarreal Olaizola 2011; Lázaro 2012) and the amount of exposure (Basterrechea and García Mayo 2013), among others. To this date, however, very few studies have considered an important variable that has proved to be crucial for a more effective implementation of the CLIL programme and a subsequently greater benefit for the L2 English learners’ proficiency, namely the degree of intensity of the CLIL programme. Precursors to those few studies are the investigations carried out in Canada around the 1990s on the effects of time distribution on language learning in both EFL settings and in French immersion contexts. Regarding the former, Spada and Lightbown (1989), and Ligthbown and Spada (1994) confirm the greater benefit of intensive EFL programmes that are communicatively oriented over regular and extensive EFL instruction focused on rote learning and the accuracy of learners’ production (see also Curtain 2000 for L2 Spanish programmes in the US). As for the latter, Wesche (2002) claims that intensive early immersion models are generally more effective for learning French than those models providing the same amount of language instruction over a longer span of time. More recently and in Spanish settings, Serrano and Muñoz (2007), and Serrano (2010) show that concentrating the hours of instruction in more intensive programmes (i.e., more hours per week in a more reduced time period) grants a greater benefit than increasing the duration of the programme at the expense of reducing the number of hours per week.

A limited number of studies gauge the effects of CLIL instruction intensity on foreign language learning in Spain. Unlike Serrano and Muñoz (2007), and Serrano (2010), here ‘intensity’ involves more CLIL hours per week in the same time period as less intensive programmes. Their results reveal a positive correlation between the greater amount of exposure granted in more intensive CLIL programmes in secondary education and (i) higher levels of proficiency (Ruiz de Zarobe 2008, 2010; Merino and Lasagabaster 2018) and (ii) motivation (Somers and Llinares 2021), as compared to either less intensive CLIL programmes or traditional EFL instruction. Ruiz de Zarobe (2008) finds not only that CLIL learners’ oral production and lexical richness are significantly higher than those of their non-CLIL homologous but also that students enrolled in an intensive CLIL programme outperform those in the less intensive one in pronunciation, vocabulary, grammar, fluency and content. Ruiz de Zarobe (2010) reveals that the advantage of the increased exposure in the intensive programmes is less significant in written production, with less marked differences between CLIL and non-CLIL students, as well as between intensive CLIL and less intensive CLIL groups. Nonetheless, Ruiz de Zarobe (2010) contends that the benefit of CLIL instruction is undeniable and is translated into a faster rate of acquisition, with CLIL groups outperforming older EFL learners, and also into better results in the long term, as greater differences between the groups are attested with increased content instruction. Merino and Lasagabaster (2018) confirm the longitudinal effects of intensive CLIL programmes over less intensive programmes and traditional EFL instruction, and the subsequent higher proficiency of the former in reading, writing, listening and speaking skills. These investigations, however, explore adult instruction (eighteen to twenty-four-year-olds). To the best of our knowledge, there are not previous studies on the impact of the degree of intensity presented by the CLIL programme on primary education L2 learners’ grammar, particularly on specific linguistic features, a gap that our investigation aims to fill.

4.1 Participants

The sample investigated consists of forty-three primary education (Grade 6) students of L2 English from three schools in Cantabria (Spain), from a middle-high socioeconomic status area. The participants were divided into two age-matched (eleven-/twelve-year-old) groups: two schools belong to the CLIL+ group (n=28), which comprises the students enrolled in the more intensive CLIL programme, while the other school conforms to the group with a lower exposure to the target language, CLIL– (n=15). As shown in Table 1, both groups started learning English when they were five/six years old and received – until the time the data were collected – a mean of 617.5 hours of English as a Foreign Language at a rate of around 2.5–3.5 hours per week. As regards their CLIL instruction, their onset age was also around five/six years old, but their additional exposure to the target language differs substantially. The CLIL+ group was taught some content classes – i.e., arts and crafts, music, natural sciences and physical education – exclusively in English at a rate of two/three hours per week, thus receiving a mean of 508 CLIL hours by the time the data were collected. The CLIL– students’ additional exposure, however, was more limited, amounting to only 376 CLIL hours. More specifically, they were exposed to the L2 in some content classes (i.e. physical education, science) they received at a rate of four hours a week, where the L2 and L1 were used roughly in the same proportion, that is, two hours each approximately, but where the L2 was used for teaching content in the process of carrying out specific interdisciplinary projects focused on a particular event or topic. It must be noted that the teachers of both CLIL and EFL lessons are not native speakers of English.