Professor Stake feels that Professor Glaser's paper does not give adequate attention to how descriptions and judgements of the instructional process can be used to facilitate educational decisions. He also asserts that behavioral specification, as it is usually defined, is often impractical and unnecessary, and he challenges the primacy of goal specification as a first logical step in instruction or evaluation. Objectives must be reported, but attention must be given to how they change over time and the priority given by different groups to different objectives. (Author)