Swedish preschools and schools are welfare institutions required to achieve and demonstrate high quality. The teachers’ extensive responsibility for quality development and evaluation is stated in the national curriculums. Many studies have contributed with quality definitions and suggestions on how quality can be achieved and measured. This paper discusses how discourse analysis can contribute to the studies of quality and quality work in school. Short extracts of data from the on-going project Quality work in Swedish pre-school are used in order to give an insight in the method’s potential and limitations. Our methodological framework is inspired by Foucault’s theories of discourse and knowledge relations, in which discourses are regarded as more or less systematic statements about how we should think about the world. According to Foucault discourses shape and created meaning systems which may gain the status of 'truth', and hereby dominate how we define and organize ourselves and our social world. Some discourses can constrain and dissent the production of knowledge, while others may enable 'new' knowledge. Discourses lead to mechanisms that control us, by construing some manners of speaking [or writing] and behaving as appropriate, true and good, while others are construed as inappropriate, untrue and undesired. Consequently the study of discourses is closely related to the study of power relations, focusing on how these relations are operating in the spoken [or written] words. The questions that arise within this framework are e.g. How do some discourses maintain their authority? Why do some 'voices' get heard whilst others are silenced? Who benefits and who does not? Critical discourse analysis is a method used to study discourses and their relations to sociocultural practices, which allows for a critical perspective on the society. Critical discourse analysis is especially suitable for studies of texts such as steering documents, interview transcripts etc. The analyses focus on how discourses express themselves, e.g. in teachers’ statements about quality. The method provides a number of analytical tools such as: Normalization [What is expressed as a ’truth’? What does this ‘truth’ encompass?]; Pronouns used [Who are made responsible?]; and Modality [How forceful are the statements?]

The short extracts from our analyses demonstrate how language contributes to create, consolidate and sometimes change the perceptions of quality in socio-political institutions such as the Swedish municipal preschool. We exemplify how power relations coexist or concur within this discourse, and illustrate the discursive shift which seems to have occurred.

Conclusively, discourse analysis neither describe quality nor prescribe how it can be achieved. It contributes to our knowledge about the ‘truths’ and power relations that are embedded in the spoken and written ‘talk’ about quality. This knowledge is not merely of academic interest. It can also prove useful for teachers trying to understand and relate to the circumstances under which they are expected to create and evaluate quality in a school setting.