Metalinguistic interaction in multilingual learning settings as a predictor for language awareness and its significance for language formation in German, foreign and heritage languages (individual project)


Metalinguistic awareness can be defined as the ability to reflect on language and its utilisation in order to use it purposefully and consciously for accomplishing intended linguistic behaviour. Its development and enhancement is therefore a critical task in language lessons. To date, little is known about the development of metalinguistic awareness or the factors that promote it. In the literature, it is postulated that metalinguistic awareness first correlates with primary language development and second with general cognitive development. What is more, multilingualism is assumed to also play a role in the development of metalinguistic awareness. Yet, so far, there have been no empirical studies that broadly investigate these hypotheses.

This project investigated the metalinguistic awareness of mono- and multilingual primary school children on the basis of their linguistic and cognitive development. The study aimed, firstly, to capture how children processed linguistic structures and, secondly, to investigate correlations between the language skills of mono- and multilingual children and their metalinguistic awareness.

Because metalinguistic awareness is a mental construct which cannot be directly observed, verbal data were generated in order to obtain language-related reflections. A procedure was developed as part of this project to guide children in making linguistic reflections, without influencing or limiting the content of their reflections.

Against this backdrop, the first research question of the investigation was as follows:  How can metalinguistic awareness be assessed and described on the basis of metalinguistic expressions?

Because the relevant literature has not yet conclusively reported on the role of linguistic and cognitive development, the second research question was thus: Is there any relation between the language skills in the first and second languages and metalinguistic awareness?

Furthermore, the role of multilingualism in the development of metalinguistic awareness has also not been widely investigated, although it is often assumed to be a resource for solving metalinguistic problems. The third research question was thus: How do multilingual primary school children use their first language abilities to deal with metalinguistic tasks?

It was hypothesised that multilingual primary school children, drawing on their linguistic resources in their heritage language, may show qualitative as well as quantitative variability in their metalinguistic expressions.

What was investigated and how?

Data measurement I

Firstly, demographic data concerning gender, age, birthplace (child and parents) as well as language usage in the family and among peers were collected in interviews with the participating children (N=400).

Then, their general cognitive abilities and language skills in German as well as in Turkish and Russian (for the children who speak these heritage languages) were assessed. For this, an adapted version of the diagnostic instrument Tulpenbeet was used (Reich, Roth & Gantefort, 2008).

Data measurement II

Metalinguistic expressions were then assessed as indicators of metalinguistic awareness. For this purpose, a new procedure called M-SPRA was developed which uses six reflection prompts. The participants had to solve metalinguistic tasks, which prompted them to make language-related hypotheses and observations and to verbalise their language knowledge.

A typical sequence in this procedure was as follows: Two children (forming an interaction team) operated the multilingual software My First Stories (2013). Together, they listened to and read the story Maddox The Magician, which is available in five languages (German, English, Spanish, Russian, Turkish). They had the option of switching from one language to another at any time. While the children listened to/read the story, the test administrator (i.e. member of the research team) asked questions and provided prompts for the children to express their language-related thoughts. They did this by interacting with their tandem partner and the test administrator. Interactive settings of this kind were recorded on video and then analysed to identify the metalinguistic expressions – i.e. the linguistic levels to which the participants refer – and examine the complexity of their expressions, that is, the depth of their reflection on language and the degree of analysis.

Example: Count the words!

One example of such a prompted interaction sequence involved children comparing the number of words in sentences presented to them in different languages. Children were shown the same sentence from Maddox The Magician in the five available languages (Figure 2). The test administrator then asked them why the number of words differed for each language. The children expressed and justified their assumptions as to why this might be the case. Their expressions were recorded and analysed.


A wide spectrum of metalinguistic expressions was elicited and qualitatively analysed to identify indicators of metalinguistic awareness. Using the four-field model (Bredel, 2007), which distinguishes situation-related from non-situation-related linguistic reflections, the expressions were divided into two groups.

Situation-related metalinguistic expressions include the expressions that children spontaneously voiced while working with the multilingual software. This category comprises: self- and external corrections (e.g. “No, this should be pronounced ”); language-related evaluations that concern language skills and/or attitudes or emotional judgements regarding other languages (e.g. “I can say almost nothing in Turkish because I am not Turkish”), and language-related descriptions that simply reproduce aspects discerned in the materials (e.g. “No capital letters are used here”).

Non-situation-related metalinguistic expressions were found using the M-SPRA assessment tool. These expressions can also be divided into three hierarchical categories, depending on the amount and degree of metalinguistic reflection: The lowest category contains language-related statements, followed by explanations and, finally, by analyses.

Metalinguistic expressions (Wildemann et al. 2016)

Taken altogether, the statements, explanations and analyses made by each child are considered to constitute a global value indicating his or her level of metalinguistic awareness. These levels were also controlled for with respect to general cognitive and language abilities. German-language ability, general cognitive abilities and metalinguistic awareness showed significant positive correlations.

At the same time, these correlations were too weak to explain metalinguistic awareness among the primary school children. It can thus be assumed that the development of metalinguistic awareness may be influenced by the school context and language classes. No significant correlation was found between Russian and Turkish heritage language skills and metalinguistic awareness. This indicates that German, as the language of schooling, plays a more important role in the formation of metalinguistic awareness among primary school children than languages spoken only in the family or with peers. In this regard, the role of written language acquisition and school-related linguistic reflection ought to be investigated with respect to the development of linguistic knowledge.

To see whether multilingualism impacts metalinguistic awareness, the metalingual awareness values of multilingual children were compared with those of monolingual German speakers. Our analyses show that multilingual children produced a larger amount of metalinguistic expressions than their monolingual classmates, after German-language skills, general cognitive ability and age had been controlled for. With regard to non-situation-related expressions, the analyses show that multilingual children reflected on language on a higher metalinguistic level than their monolingual German classmates. It can therefore be assumed that children raised with more than one language are more capable of linguistic analyses than those who cannot draw on a second, comparative language. We can further assume that the children who spoke Russian or Turkish referred more often to their heritage languages than the multilingual children whose heritage languages were not available in the software program. Thus, the availability of a particular language somehow guided access to a student’s own linguistic resources. More studies with larger sample sizes are, however, necessary to test these assumptions.

What does this mean for educational practice?

The results of the study show that primary school children display a wide spectrum of metalinguistic abilities that may be useful both in and outside of school. Multilingualism can be – and is indeed – used as a resource for language reflection and language comparisons. What is more, German, as majority language, has a greater effect on metalinguistic awareness than the children’s first languages (where applicable). Other differences observed in this study lead us to believe that the pedagogical approach has a large influence on the development of metalinguistic awareness. In this regard, language lessons in primary school may have two functions: first, the development of metalinguistic awareness as an ability to make language and linguistic behaviour a subject of discussion and, second, the inclusion of existing multilingual resources in language reflection and language comparisons for the benefit of all children.