The lack of both specificity and consensus about definitions for homosexual, homosexuality, gay, and lesbian are first shown to confound comparative research and cumulative understanding because criteria for inclusion within the subject populations are often not consistent. The Description section examines sociolinguistic variables which determine patterns of preferred choice of terminology, and considers how these might impact gay and lesbian studies. Attitudes and style are found to influence word choice. These results are used in the second section to devise recommended definitional limits which would satisfy both communication needs and methodological purposes, especially those of sampling.