Despite the uniqueness of each individual's experience, approximate communication is possible in the margin of overlap between the speaker/writer's experience and that of the hearer/reader. This vague phrase, "margin of overlap," can be clarified and given an empirical base through application of the measure of vagueness developed in Labov's study, "The Boundaries of Words and Their Meanings." Labov shows that for words that refer to simple objects, the non-simple range of ascription can be described with great accuracy through study of actual speakers' uses of a word. A modal description of an object can be constructed that is linked to a list of key elements affecting ascriptions of the word. For technical discussions, this strict definition (modal with variables) gives an empirical base on which to ground discussion. Such strict definitions could be immediately useful in discussing terms like "art" and "novel," where Labov's model for strict definitions could be used to construct empirically valid, verifiable bases for such contested concepts. From these strict definitions it should be possible to construct loose, general definitions that meet speaker intuitions and incorporate all important variables, as is the case with words referring to simple objects. Such intuitively valid, empirically verifiable definitions could ground studies in margin of overlap, thereby grounding empirical rhetoric. (Author/DS)