This study is a discussion of, rather than a contribution to, generative phonology. The central question posed, is: Does linguistic theory provide a basis for choosing between competing grammars — that is, an evaluation procedure for grammars? If so, then what is its form? If not, then how are we to interpret controversies between linguists as to the relative merits of competing grammars? These issues will be discussed in relation to a particular problem of evaluation in the treatment of the morphonology of final segments in Modern French.