Can anecdotes be evidence? In general, “anecdotal evidence” is deemed potentially misleading (if not an oxymoron). But there are certain situations, and certain fields, in which anecdotes register phenomena that can be hard to capture, control, or quantify. Anecdotes are often the only way to record chance occurrences or account for inner experience. Because they can carry moral values as well, anecdotes also trouble the is/ought divide in scientific discourse. A history of efforts to collect, analyze, and use anecdotes—as well as the opposition to such efforts—might thus provide a kind of counter-history to the coalescence of evidentiary standards in modern science. This paper addresses the authority of anecdotes in the context of animal psychology, specifically as the field developed in the latter half of the nineteenth century. With the rise of experimentalism and the decline of introspection, animal anecdotes occupied a vexed middle ground in the work of such figures as George John Romanes, Conwy Lloyd Morgan, and Edward Thorndike. On the one hand, anecdotes about “other minds” were seen as superior to introspective data; on the other hand, such evidence resisted experimental control, thwarting attempts to standardize research subjects. Through a case study of the American beaver, this paper recovers the role of anecdotes both in scientific studies of animal reason and in the unmaking of situated knowledge, showing how “anecdotal” became a bad word that was used to reject some scientific claims in favor of others.
