Internationally recognized experts in second language research were asked to provide anecdotes documenting memorable incidents where something had gone awry in their research. The anecdotes proved to emphasize problems of a technical, conceptual, or political nature. Fifteen anecdotes representing these three kinds of problems are presented. In common, the anecdotes express concerns for ... [Show full abstract] scientific realism, that is, second language research was said to go wrong in cases where legitimate claims for knowledge were undermined - by accidents, unforeseen intrusions, or misconceptions. This view of constraints on second language research, based on researchers' experiences (rather than theoretical polemics or methodological prescriptions), may be valuable for learning researchers, users of research results, and critical analyses of second language research processes.