Elevers spr\ue5k i naturvetenskapliga f\uf6rklaringar. Analys av skriftliga svar fr\ue5n \ue5rskurs 8 i TIMSS 2007

AbstractAnalyses of the language used by Swedish grade eight students in their explanations about matter and changes in matter are reported in this licentiate thesis. The linguistic perspective taken is motivated by an interest in the role of language in science teaching and learning. In order to understand the use of language in the specific case of tests in science education, answers to items from the international test Trends in Matematics and Science Studies (TIMSS) 2007 were analyzed. With the assumption made that the relevant discourse students normally participate in will leave traces that can be observed, the following main questions were put to the material:To what degree do students use linguistic features typical for scientific language in their written explanations and do this use differ depending on gender or achievement groups? What lexical and grammatical resources are used to distinguish between and relate different representational levels in explanations that take the submicro level of representation into account?Theoretical framework drawn upon were the sociosemiotic perspective of Systemic Functional Linguistics (SFL), where language is seen as constitutive of as well as being constituted by the social context. Language is seen as stratified in context, meaning and lexicogrammatics, with meaning in a specific context realised by lexical and grammatical choices. With analytical tools from the same perspective, students written explanations were analysed regarding features typical for scientific language and other lexicogrammatical resources used.The results show a limited use of the analysed typical linguistic features technical terms, nominalisations and passive verbs. Students with high total scores on the test use significantly more technical terms than the students with low total scores, in some case the use is twice as high for the students with high scores. A wide range of different verbs is used, especially on the submicro level of representation. This could indicate a confusion about what processes are possible on this level, and points to a possible mechanism of adding new, scientific language use to previous everyday language use. At the same time it demonstrates creativity in the construction of meaning about an unknown, abstract world.The report is written in Swedish and the two papers in English