MEG-SKoRe – Multilingualism as a linguistic and cognitive resource in English language acquisition in primary school

Introduction

In educational contexts, multilingualism is frequently seen as an obstacle to, rather than a resource, for successful learning. Against this backdrop, the research project MEG-SKoRe investigates how multilingualism can constitute a resource for the early acquisition of English in primary school. In the first project cycle (2014-2017), we conducted a longitudinal study in which we identified critical success factors of multilingualism in the context of English language acquisition. In the second cycle, (2017-2020) we apply those factors to classroom settings. The project focuses on two central research questions:

  • Can multilingualism become a resource in early academic foreign language learning?
  • How can teachers make use of this resource didactically in order to support foreign language learning?

Previous research shows mixed results concerning the role of multilingualism in early foreign language learning. Studies that compare the English skills of monolingual Germans with those of multilingual primary school students have either found no group differences in those skills or weaker skills for multilingual students in reading and listening comprehension (see Keßler & Paulick, 2010, for overview). Possible reasons for these heterogeneous findings may be individual differences in the social backgrounds of students as well as varying skills in the languages previously acquired by multilingual students. Indeed, differences in linguistic (e.g. language combinations, language awareness), cognitive (e.g. working memory) and social (e.g. socioeconomic or familial) factors can impact the acquisition of English (e.g. Maluch et al., 2015; Wilden & Porsch, 2015). MEG-SKoRe considers these factors systematically in order to identify critical success factors in early foreign language learning.

What was investigated and how?

In MEG-SKoRe I, we examined whether primary school students with German as a second language (L2) had different learning outcomes in the early acquisition of English compared to their monolingual peers. To this end, we investigated (a) which individual linguistic, cognitive and social factors influence the acquisition of English, and (b) in which respect multilingualism acts as a resource in early foreign language learning. The project consisted of two parts. Part 1 focused on linguistic transfer and to which extent the L1 and/or L2 affect the acquisition of English. Part 2 explored whether metalinguistic awareness has a positive impact on English language skills.

Part 1: Vocabulary and Grammar

For the first part, we assessed general vocabulary and grammar skills as well as specific grammatical phenomena, i.e. article realisation, subject realisation and word order. For general skills, we collected data in German, English and the respective non-German L1 of the multilingual students. To measure receptive vocabulary size, we used the British Picture Vocabulary Scale (BPVS3; Dunn et al., 2009), in which students heard an English word which they then had to match to one of four pictures. Furthermore, we determined productive vocabulary size with a category fluency task (following Delis, Kaplan & Kramer, 2001). Students had to name, within one minute, as many items as possible relating to a single semantic category, such as ‘food’ or ‘clothes’. To assess receptive grammatical knowledge, we administered the Test for Reception of Grammar (TROG-2; Bishop, 2003). The set-up was parallel to the BPVS except, in this case, students had to match an English sentence to one of four picture choices.

Besides these general language skills, we analysed the production of specific grammatical structures, i.e. article realisation, subject realisation, and word order, as part of a computer-based sentence repetition task. Here, students listened to grammatically correct and incorrect sentences in English. They then had to repeat the sentences exactly as they heard them. Depending on whether students corrected the grammatically incorrect sentences, we could assess acquisition of the respective structure. This way, we could determine, for instance, whether the use of articles in English is different for those learners whose L1 uses articles (e.g. Italian) or does not use them (e.g. Turkish, Russian).

Part 2: Language Awareness

In the second part of the project, we examined different aspects of language awareness. First, we assessed students’ levels of phonological awareness by asking them to segment English words into phonemes and manipulate them (e.g. “What remains when you delete the last sound in ‘green’?“). Second, students explicitly reflected on and talked about language(s) in a structured metalinguistic interview. Students answered questions about their language learning experiences and discussed linguistic contrasts between English, German and their respective heritage languages.

Participants

Overall, 200 students (88 monolingual; 112 multilingual) from six public primary schools in south-west Germany took part in the study. They were assessed at two intervals: initially at the end of 3rd grade and then at the end of 4th grade, at which point 184 students remained in the sample (81 monolingual; 103 multilingual). The following heritage languages of multilingual students were represented in the sample: Afghan languages, Albanian, Arabic, Bosnian, Bulgarian, Chinese, French, Greek, Italian, Croatian, Kurdish, Persian, Polish, Romani, Romanian, Russian, Serbian, Spanish, Tamil, Turkish, Hungarian and Vietnamese. The largest subgroups were speakers of Turkish (40), Kurdish (11), Albanian (10) und Italian (8).

Results

Group comparisons

The results from 3rd grade show that, compared with their monolingual peers, multilingual students had statistically significantly lower skills in English vocabulary, phonological awareness and working memory. In sentence repetition, multilingual students also showed lower performance than the monolingual German students. At first glance, these results seem to confirm findings from previous studies that did not find a multilingual advantage in foreign language learning.

Individual factors

In further analyses, we factored in cognitive, social and educational variables. Figure 1 illustrates which of these factors significantly affect English vocabulary size. A multilevel regression analysis shows that both social variation at the school level, as well as individual differences between students, impact English skills. Importantly, multilingualism also has a positive impact on English vocabulary knowledge (Hopp et al., forthcoming).

Figure 1: Contributing factors for English receptive vocabulary (BPVS) at institutional and individual level (mixed linear regression). Non-significant factors in faint font.

Similarly, multilingualism contributes positively to English grammar skills in 4th grade for a subgroup of multilingual students (Figure 3), although many other factors show larger contributions (Hopp et al., forthcoming).

Figure 2: Contributing factors for receptive English grammar (TROG-2) at institutional and individual level (mixed linear regression). Non-significant factors in faint font.

Development from 3rd to 4th grade

When looking at the development of English vocabulary skills from 3rd grade to 4th grade, multilingualism loses its positive impact; in fact, its impact becomes negative as skills in German become increasingly important. These results suggest that multilingual resources do not remain stable over time, possibly due to the lack of support in the foreign language classroom. 

The role of language awareness

As Figure 2 shows, (phonological) awareness positively affects English grammar. Moreover, a higher degree of phonological awareness implicates higher English vocabulary skills for both monolingual and multilingual students (Hopp et al., 2017). Finally, the degree to which students can reflect on language(s), as measured in their responses to questions in the metalinguistic interview, also positively affects their English skills.

What does this mean for educational practice?

MEG-SKoRe I shows that multilingualism as such constitutes neither a general resource nor a global disadvantage in the early foreign language classroom. Rather, multilingualism may become a resource in early foreign language learning when additional individual factors are considered - such as a high degree of language awareness, a large vocabulary in the L1 as well as superior cognitive skills. The follow-up project MEG-SKoRe II therefore focuses on how those resources may be addressed and promoted during English lessons in the multilingual classroom.