Text comprehension in multilingual children : mental representation and narrative text structure.

Abstract

Investigations of children‘s narrative text structure are necessary because narrative abilities are linked to literacy development and academic achievement (Dickinson & Tabors, 2001). Information contained in a narrative may reflect a child‘s use of decontextualised, literate language features (Curenton & Justice, 2004). Moreover, storytelling may lead to higher-order thinking including causal reasoning because children not only recall events from text or film but also generate inferences therefrom. Investigations that consider children‘s narrative text structure are vital in multicultural and multi-linguistic societies ―in order to not only preserve cultural and linguistic diversity, but also to support students who find themselves in educational environments in which their cultural and linguistic practices are misaligned with the language(s) of teaching and learning‖ (Tappe & Hara, 2013, p. 299). This study set out to examine the effects of language and medium of presentation on the narrative text structure in the [re]tellings of multilingual children through the use of the Narrative Scoring Scheme (NSS) (Heilmann et al., 2010a, 2010b). This study also aimed to investigate whether the [re]tellings by children with Chichewa as their primary language (L1) and English as their language of teaching and learning conform to the NSS, in other words, whether their [re]tellings conform to canonical scoring schemas that are widely used as diagnostic tools to measure the narrative competence of children – irrespective of the children‘s primary language(s). The participants in this study numbered 127 children (64 female, 63 male) whose age range was 10 to 12 years (44 10-year-olds, 40 11-year-olds and 43 12-year-olds). The results have revealed that the two languages themselves (Chichewa and English) that were used in story production and the medium of the stimulus presentation do not have a significant influence on the frequency of realisation of the canonical narrative text structure elements (Heilmann et al., 2010a, 2010b) in the children‘s [re]tellings. The children obtained low mean scores for certain elements (i.e. elements 1 and 3) and high mean scores for other elements (i.e. elements 2 and 7) irrespective of the language or medium of presentation or the school type. Importantly, the results have demonstrated that the [re]tellings by children with a Southern African primary language (Chichewa) do not conform to the Narrative Scoring Scheme (that is, the canonical scoring schemas). The results reveal that the children seem to possess a story

grammar (i.e. a Southern African story grammar) that has strong leanings towards elements that are associated with Southern African folktales. The Southern African story grammar appears to be different from Stein and Glenn‘s (1979) story grammar and other versions of story grammar that researchers developed from Stein and Glenn‘s (1979) story grammar (see Anderson & Evans, 1996). Two main arguments have been made in this study. The first argument is that the ‗universality‘ of the canonical narrative text structure may not be valid because the children‘s [re]tellings seem to have been significantly influenced by elements from Southern African folktales. In the consequence the children demonstrated limited performance in some of the elements associated with the canonical scoring schemas. The second argument is that the canonical scoring schemas for narrative text structure available in the literature may not be appropriate when analysing stories narrated by children with a Southern African language as their primary language. This study recommends that further research be done to investigate narrative skills of Southern African children in order to explore the Southern African story grammar proposal in greater depth. Additionally, it recommends that further research be conducted in languages which have been under-represented in or absent from text comprehension research. Existing research has not concentrated enough on macrostructural differences between texts produced in different languages; more research is therefore required to assess language- and culture-specific narrative text structure elements.