Generic Tasks in KEW

In this paper we describe an experiment in which we cast Generic Tasks into the mold provided by KEW. The result was advantageous for both GT theory and KEW-GT benefitted by being formalized, and by gaining a computer implementation. KEW benefitted by having its strategy vindicated on a new target theory, and having its software more thoroughly tested. The experiment also exposed a weakness in the KEW meta-methodology, which might have implications beyond its use in KEW.