Evidence in Matters Extraordinary: Numbers, Narratives, and the Census of Hallucinations

This article examines the Society for Psychical Research (SPR) and its efforts to use quantitative methods to establish the evidentiary value of testimonial narratives. It focuses on “The Census of Hallucinations,” a nationwide demographic survey conducted from 1889 to 1892 that sought to determine the frequency with which people were experiencing auditory or visual illusions. The results of this survey were combined with mortality tables to show that people were seeing apparitions, particularly those that coincided with the death of the person thus figured, more often than chance would allow. Drawing on Derrida’s concept of “spectropoetics,” I argue that the SPR’s experiments with the statistical analysis of a large textual corpus reveals the haunted structure of the database as a scholarly form.