Robert Inkpen (Principal Lecturer in Geography at Portmouth University) interient sur le thème "Finding causes in physical geography".

Abstract : This presentation uses case studies from environmental reconstruction and geomorphology to illustrate the problems associated with trying to find causes in physical geography. Finding causes in physical geography involves a complex marshalling of the interplay between theory, evidence and belief. Theory underpins and directs the search for evidence but the evidence collected is itself judged in the light of the belief associated with the theory and with other evidence related to the theory. Physical geography is both an historical and a field science focused on finding evidence for causation in the field. This field becomes the arena in which evidence of causation is in competition with information destroying processes that vary in both time and space. Field evidence is unlikely to provide a sufficient basis for deciding between competing theories and causes; field evidence will always have issues of underdetermination. Although experimental techniques can be employed to supplement field evidence, such field experiments are themselves likely to be open to the similar issues of ambiguity in interpretation. In finding causes in physical geography the degree of belief in evidence and the network of arguments used to support this belief is a key aspect in resolving this underdetermination, although such resolution is always temporary and conditional.

Suggested Reading

Inkpen, R. and Wilson, G. 2009. Explaining the past: abductive and Bayesian reasoning. The Holocene, 19, 329-334. 
Turner, D. 2005. "Local underdetermination in the historical science". Philosophy of Science, 72, 209-230