Multiliteracy: the interrelation between language abilities in the first- and second languages and extra-linguistic factors


This project investigates the interrelation of writing abilities in the first and second languages of bilingual 9th and 10th graders with the heritage languages Turkish, Italian or Greek. The study aims to examine the impacts of extra-linguistic factors (such as language attitude, literacy practices, language use) and metalinguistic awareness on written discourse competences in both languages. The following hypotheses were tested:

  • Writing abilities in L1 and L2 influence one another. A high level of competence in the L1 is imperative to high levels in the L2.
  • Extra-linguistic factors and metalinguistic awareness influence writing abilities in the L1 and L2.

With regard to metalinguistic awareness, we focus on linguistic aspects and include cognitive as well as performance levels (see James/Garrett 1991): i) Cognitive level: Knowledge of grammar, rules und functions of language, ii) Performance level: Language use, strategies of communication and the practice of talking about language while using formal metalanguage (Fehling 2006: 86).

What was investigated and how?

Different survey instruments were developed in order to examine writing abilities, extra-linguistic factors and metalinguistic awareness.

Writing assignments

Assignments were developed to elicit narrative and argumentative texts. For the narrative texts, pictures were selected and presented to participants who were tasked with composing fictional stories in both the L1 and L2. For the argumentative texts, two different letters were composed: one on the subject of a ban on foreign languages in the schoolyard (L1), another on a mobile phone ban at school (L2). The assignments were piloted and then repeated at 4-week intervals.

Language Awareness Test (LAT)

Based on the speech awareness test by Fehling (2005), a test on metalinguistic awareness (LAT) was developed to capture pragmatic, semantic, and textual knowledge in both the L1 and L2. The aim was to examine the level of linguistic variation and addressee orientation on the pragmatic level, as well as to examine the use of synonyms and universal terms on the semantic level. On the textual level, knowledge of textual organisation coherence and cohesion could be captured.

Language biographical interviews with students in L1 and L2

Interviews were conducted in that participants’ L1 and L2 in which questions language attitudes, language use (oral and written) were addressed. The data elicited from the interviews could help to explain the influence of extra-linguistic factors on writing abilities. The aim of conducting the interviews in both languages was to interpret students’ competency levels in those languages and to determine whether language attitudes are communicated differently depending on the respective language in use.

Interviews with parents

Further sociolinguistic data - such as literacy practices at home, language use, language input - was gathered through parental interviews.


Developing an analytical framework: In order to assess writing abilities in the respective languages, a model was developed to capture overall writing patterns. This model factors in macro- and microstructures, discourse modes (oral conceptual vs. written structural), and communicative modes (distant vs. involved). Based on these criteria, a detailed analytical framework including five writing levels was constructed for each type of text. All texts were analysed by three independent evaluators.

Analysis of the Language Awareness Tests: Answers to the LAT were ranked on a scale of 1 to 4 with regard to adequacy. The semantic, pragmatic and textual levels, as well as the overall test outcome, were thereby evaluated for both L1 and L2.

Analysis of the language biographical data: The language biographical interviews were transcribed orthographically. Based on the speakers’ statements, profiles were created to show the differences in language use between the students. The data were evaluated statistically using the free programming language for statistical computing and graphics, R (cf. R Core Team 2013) - in accordance with the generalised linear mixed model.


The results show that participants representing all three heritage language groups reach a higher level in the argumentative texts in the L2 than in their L1. The argumentative texts composed in the L1 differ significantly, however, with regard to text structure (macrostructure), which can be explained by culture-specific formalities not acquired in the respective language. Compared with the argumentative texts, participants attain high levels in writing competence in the narrative texts in both languages. Generally speaking, those who show good writing abilities in their L1 also show good abilities in the L2.

With regard to metalinguistic awareness, the results indicate a correlation with writing abilities. Participants with poor writing abilities in the L1 demonstrate awareness of the essential aspects of a text (for instance, cohesion and coherence) and that they are generally able to estimate register-specific norms correctly (such as how to appropriately address the listener). From a didactical point-of-view, assignments to promote metalinguistic awareness should be created to support these young multilinguals in their writing development. In addition, a concept in terms of contrastive language didactics should be developed to compare different text, argumentation and stylistic patterns in various languages.

The statistical analysis of the sociolinguistic data indicates that heritage language education may have a positive impact on writing abilities in the L1 only after seven years. However, this may also relate to other factors, such as lesson design (especially when taken as extra-curricular). This was confirmed by teachers and pupils in interviews. Nevertheless, it can also be shown that L1 lessons, even though they do not have a significantly positive effect, do not have a negative one either. Rather, literacy practices at home, language attitudes and use among the study participants appear to have a highly positive effect on their writing abilities.

What does this mean for educational practice?
  • Linguistic support for multilingual children must be carried out in both languages.
  • The L2 does not suffer when the L1 is supported; rather, it can be strengthened.
  • Support for multiple languages includes the acquisition of literacy as part of explicit knowledge.
  • The promotion of metalinguistic awareness, in other words a differentiated knowledge of linguistic structures and language usage, should be taken more into account in formal education.