The history of American fluvial geomorphology over the past century is viewed as one of conflict and crises. From 1888 to 1938, a controversy arose between (1) a rational approach to understanding landscape genesis and history, with its roots in geology, and (2) a spatial-analytical approach to landscape classification and description, with its roots in geography. By the 1960s, geomorphology, led by fluvial studies, had changed its emphasis from historical studies to process studies, and the geology/geography dispute became irrelevant. Since the 1960s, a new conflict has arisen between (1) problem-oriented studies of landform genesis and (2) method-oriented studies. The latter emphasize useful predictions and a methodology that generates respect from other scientific and engineering disciplines. In extreme cases, approach 2 may bypass the understanding of phenomena in order to generate useful predictions of systems assumed to embody the behavior of those phenomena. In order to achieve its goal of intellectually satisfying understanding of phenomena, approach 1 may require the stimulus of the occasional outrageous hypothesis, thereby posing a seeming anathema to an existing scientific program. The identification and explanation of anomalies is critical to approach 1. Because of the inherent conflict in these approaches to fluvial geomorphology, there is a need to balance opposing tendencies.