Abstract
This article examines the dynamics of anomalies as part of scientific work. Several types of anomaly are identified: mistakes, artifacts, fraud, and discovery; and typical trajectories for artifacts are described: the establishment of suspected artifacts, changes from unacceptable to acceptable, changes in significance, visibility, and means of control. The conditions under which an anomaly changes status are examined. A detailed example of an anomaly trajectory is presented, analyzing an anomaly in neuroscience research and tracing its career from 1870 to the present.