Schumm, S.A., and Lichty, R.A. 1965. Time, Space,and Causality in Geomorphology. American Jour-nal of Science 263:110–19.Schweber, S.S. 1993. Physics, Community, and theCrisis in Physical Theory. Physics Today 46:34–39.von Bertalanffy, L. 1951. An Outline of General Sys-tems Theory. Journal of the British PhilosophicalSociety 1:134–65.———. 1968. General Systems Theory: Foundations,Development, and Applications. New York: GeorgeBraziller.Methodology in Geomorphology 687Correspondence: Dept. of Geography, University of Southern California, Los Angeles, CA 90089, [email protected] (Bauer); Dept. of Geography, University of Colorado, Boulder, CO 80309, email [email protected] (Veblen); Dept. of Geography, Michigan State University, East Lansing, MI 48824-1115, [email protected] (Winkler).Iwish that Richard Hartshorne were alive. Hewould know what to write in an essay onmethodology—even one concerning geo-morphological methodology! He had a strong,unshakeable and clear vision of the nature ofgeography, and what the proper methodologicalapproaches ought to be. His disciplinary visionincluded the place of physical geography, andgeomorphology in particular, within the schol-arly framework of geography. Few would havedared to assail his pronouncements, and thosethat did would have been excoriated in reply.With one hand clutching a tattered volume ofHettner as a shield, Hartshorne would onceagain be driving geography’s methodological bustoward intellectual oblivion, pushing geomor-phology as a vanguard. Perhaps some would dareget off. A very few jumped from a similar busabout half a century ago, keeping a remnant ofgeomorphology viable in American geography.Geomorphology is the science concernedwith landforms. Geomorphologists are con-cerned with the evolution of landforms, andtheir geometry, origins, and locations. Geomor-phologists, like all scientists, really, are reduc-tionist, at least in the sense that we do not treatthe form of the entire earth as the entity for ourresearch. Indeed, the titles of many of our sub-disciplines indicate a first-order reduction of theunit of study: fluvial, tropical, glacial, arid,coastal, soils, and so forth. The unification ofthe subdisciplines is through the generic of land-form, and through commonly accepted norms ofmethodology. With few exceptions, geomor-phologists are trained either as geologists or asphysical geographers, lending great methodolog-ical diversity to the discipline. My concern hereis mainly with physical geographers, but thearguments are relevant for both schools.Hartshorne (1939) viewed physical geogra-phy, with special attention to geomorphology, asa subdiscipline of geography (in literal and figu-rative senses), necessary only as its productsrelate directly to human geography. He did notthink much of geomorphology as a part of geog-raphy, at least to the extent that the former isconcerned with landforms as entities or with theorigins (e.g., processes) of landforms. Indeed, hewent so far as to denounce geographers who aregeomorphologists for “the confusion that manyof them have introduced into methodologicalthought in geography” (Hartshorne 1939:423–24). Hartshorne was methodologically pre-scriptive in the sense that, following Longino(1990), he prescribed the proper methodologyto be employed by geographers in meeting theirdisciplinary goals. His prescription for geograph-ical geomorphologists required that landformsbe studied only as a set of interrelated objects.Specifically, Hartshorne (1939:425–26) arguesthat “it is immediately clear that the interest ofthe geographer is not in the phenomena them-selves, their origins and processes, but in therelations which they have to other geographicfeatures.” Reduced to a logical minimum, thismeans that geographically oriented geomorphol-ogists make maps. Many scholars have sincesubscribed to and defended versions of thisMethodology in Geomorphology: Traditions and HypocrisyDouglas J. ShermanDepartment of Geography, University of Southern Californiaperspective, especially during geomorphology’sdisenfranchisement from geography through themiddle of this century. This prescription resultsin a sterile discipline, however, one that I wouldrather not practice.Sauer (1941) was one of the first to note someof the holes in Hartshorne’s arguments concern-ing physical geography and relevant processes,especially their supposed derivation from Het-tner. In reaction to Hartshorne’s prescription,Sauer (1941:45) wrote that it is a fallacy tobelieve “that descriptive studies, done withoutregard to process, i.e., genesis and function, canadd up to a science, either physical or social.”Geomorphology without process was certainlynot a position advocated by Hettner (e.g.,Butzer 1989:37), and Hartshorne’s representa-tion of that view appears misleading. The per-spectives described by Hartshorne and Sauer arethe extremes of a methodological spectrum formodern geomorphology. At one end is geomor-phological enquiry as the description of formand location, relative and absolute. At the otherend is geomorphological enquiry as the under-standing of space-independent processes andthe resulting responses of the earth’s surface.Between these extremes, the spectrum opens aplace for each of us. Closing doors, á laHartshorne’s manifesto, just takes us backwards.By undertaking this essay, I have accepted anopportunity offered by the associate editors ofthe Annals to discuss modern methodologies ingeomorphology. In particular, I was charged toaddress three main questions. First, what are thekey methodologies in geomorphology? Second,what are the practical implications of makingcertain methodological choices for the produc-tion of knowledge in geomorphology (how havemethodologies influenced where we have beenand where we are going)? And third, to whatextent can methodological discussions enable usto achieve both an integrative and interdiscipli-nary perspective on physical geography andgeography at large? The key methodologicalstances taken by past and present geomorpholo-gists (e.g., positivism, realism, functionalism,idealism, and so on) have been reviewed or dis-cussed at several levels of detail (e.g., Marston1989; Bishop 1980; Schumm 1991; Tinkler1985; Rhoads and Thorn 1996; Gregory 1985;Phillips and Renwick 1992; a thorough review ispresented in Beckensale and Chorley 1991).Approaches to the second question must beinextricably bound to what will always be