In the last twenty years geomorphologists have moved away from genetic descriptions, attempting to separate more clearly description from interpretation. However, it remains that descriptions mean little in isolation from the context of explanation of which they are a part, and that description and interpretation are not as sharply distinguishable as convenience might prefer. Just as there are more than two dimensions to landforms, so there is often a third if not a fourth way of dissecting the several dichotomies on which the nature of geomorphology appears to be structured. Geomorphological processes are not easily specified exactly, for several reasons. Even if geomorphological processes are comprehended, this knowledge is not easily integrated into interpretations of landforms. This task is most apparent for ‘homologies’ of landforms. The quantitative element in geomorphology differs depending on whether time, process or form is considered.