Teaching involves assessment. In making decisions about lesson content and sequencing, about materials, learning tasks and so forth, teachers have to determine the strengths and weaknesses of the alternatives available to them. They make selections based on their experience, on their understandings of learning, language development and of language proficiency itself, together with what they consider to be most appropriate and in the best interests of those they teach. Equally, as part of their professional practice, they are always involved in the observation of their learners, which leads to the development of insights about learner progress and judgements about specific learning outcomes and overall performance. In my experience, however, when asked about classroom assessment, teachers will tell you first and foremost about the formal mechanisms that are in place to monitor language achievement, or about the specific assessment procedures that they use. There is a tendency to prioritize the 'formal' and the 'procedural' and to underplay the observation-driven approaches to assessment which is strongly in evidence in their everyday classroom practice, such as language sampling (see Gardner and Rea-Dickins, 2002;Rea-Dickins, 2002). This orientation, I suggest, is mirrored in much of our research in language testing and assessment as evidenced by work on language proficiency testing, the focus À over time À of sustained research. Assessment, with specific reference to teaching and learning in the language classroom, has remained, until recently, relatively unresearched. It is interesting to observe that the recent state-of-theart review (Alderson and Banerjee, 2001) did not include a separate section on either teacher assessment or classroom based assessment. It did, however, contain sub-sections on 'alternative assessment',
