'Tone' and assessing writing: applying Bernstein and Halliday to the problem of implicit value assumptions in test constructs

The value assumptions of language test constructs are inherently difficult to identify. While language test scoring descriptors refer to concepts such as tone and appropriateness, the role of socially determined value assumptions in written language assessment has not been adequately modelled or discussed. This paper presents a framework to link the results of analysis of written test scripts with the value assumptions of the pedagogic discourse of the test.