Towards an Ontology of Landforms

The paper begins with the question “Do mountains exist?” It shows that providing an answer to this question is surprisingly difficult, and that the answer which one gives depends on the context in which the question is posed. Mountains clearly exist as real correlates of everyday human thought and action, and they form the archetype for geographic objects. Yet individual mountains lack many of the properties that characterize bona fide objects, and mountains as a category also lack many of the properties that characterize natural kinds. In the context of scientific modeling of the environment, especially of such phenomena as surface hydrology and fluvial erosion and deposition, mountains are not picked out as constituents of reality in their own right at all; rather they are just parts of the field of elevations whose gradients shape the direction of runoff and influence the intensity of erosion. Thus while an object-based ontology of mountains and other landforms is required to do justice to our everyday conceptions of the environment, and to support spatial reasoning and natural language processing, topographic databases designed to support environmental modeling can be field-based at geographic scales.