The logical status of applied geographical reasoning

Abstract

This commentary is an invitation addressed to policy-oriented human and physical geographers to consider, from a strictly logical point of view, the way that reasoning works in applied geographical thought. The promise hidden in this somewhat unfashionable return to logic is twofold. First, by traversing the distance between our discipline and formal logic we may harvest the reward of a more rigorous understanding of the daily work that we do as policy-oriented geographers than that provided by the earlier scholarship on geography as practice (Dewsbury and Naylor 2002; Lorimer 2003). Second, the arid land of logic offers a novel entry point into answering the formidably difficult question of whether there is any common ground left between applied human geography and applied physical geography (Gatrell and Jensen 2009; Massey 1999; Richards 2003; Bracken and Oughton 2006; Couper 2007; Matthews and Herbert 2004). To be sure, previous scholarship has questioned the operation of logic and reason in geography from a variety of standpoints. These include physical geography (Inkpen 2004; Gregory KJ et al. 2002), positivistic human geography (Bunge 1966; Harvey 1969; Gould 1999), humanistic geography (Buttimer 1993; Adams et al. 2001), non-representational theory (Thrift 2008; Anderson and Harrison 2010; Pile 2010), postmodern and poststructuralist geographies (Soja 1989; Doel 1999; Ley 2003), feminist geographies (Rose 1993; Whatmore 2002; Pratt 2004), as well as a neglected work that stands in a class of its own (Olsson 1975). Part of the value of applied geography is assumed to consist in its unique way of thinking about reality, or, in other words, in the existence of a geographical way of reasoning (Johnston and Sidaway 2004; Gregory et al. 2002; Aitken and Valentine 2006; Gatrell and Jensen 2009). Taking this assumption as given, there remains the further task of specifying and characterising the properties of this type of reasoning. Some may start by invoking the set of concepts that guide applied geographers’ inferring of knowledge about the world (e.g. distance, space, place, territory, landscape, scale, environment, locale, site). According to this view (Holloway et al. 2008) what makes policy-oriented geographical reasoning geographic is the active use of key geographical concepts. The advantage of this position resides in the fact that it brings out in a sufficiently vivid manner what separates applied geography from other related practical disciplines. The disadvantage comes from the fact that, by focusing on substantive semantic issues, this position foregoes the opportunity of characterising applied geographical reasoning in formal terms, qua reasoning (Harman 2008). That is, it sidesteps the epistemically useful exercise of trying to clarify the logical status of policy-oriented geographical reasoning. Logic is the formal study of reasoning. Many scholars equate it with deduction, but deduction is only one of the subfields of logic (Kosko 1993; Shapiro 2007). The study of deduction, however, is useful for understanding what applied geographical reasoning is not (Rips 2008). The shortest path to this goal involves, first, the specification of one of the key properties of deductive logic and, second, the explanation of why applied geographical reasoning fails to have this property. A system of reasoning is said to have the property of monotonicity if the addition of new information to the premises of a valid argument cannot modify the conclusions already drawn in that argument. Monotonicity is the hallmark of deduction and explains both its success and its failure (Rips 2008). Its success and appeal come from the certainty guaranteed by a deductive argument: we live in a changing, uncertain world, and being able to point to a system of reasoning that seems to shelter us from this frightening uncertainty seems reassuring. Beliefs set in stone by The Geographical Journal, Vol. 178, No. 1, March 2012, pp. 9–12, doi: 10.1111/j.1475-4959.2011.00430.x