Heterogeneous Inference with Maps

MARIELA AGUILERA; FEDERICO CASTELLANO

Ann Arbor, Michigan

Congreso; 44th Annual Meeting of the Society for Philosophy and Psychology; 2018

The Society for Philosophy and Psychology

It is widely argued that maps have the resources for inferences (see AUTHOR, XXXX; Camp, 2007; Casatti and Varzi, 1999, Rescorla, 2009). These proposals have been mainly focused on the capacities of maps for making inferences within cartographic systems. Here, instead, we are more interested in cases of heterogeneous inferences where cartographic representations interact with sentences. Heterogeneous inferences are inferences that combine representations that belong to different representational systems. There are remarkable studies on heterogeneous inferences in the literature (see Barwise & Ethemendy?s Hyperproof in 1996). However, most of these studies have been focused on the interaction between linguistic representations and diagrams (see Shin, 1994 and Allwein & Barwise, 1996) and images (see Barceló, 2012). Based on these studies, we want to make a contribution to the case of heterogeneous inferences involving cartographic representations. More specifically, we want to argue that maps can participate in heterogeneous inferences, but only as singular premises. To do that, firstly we argue that maps have a logical structure. In particular, maps decompose into referential and predicative elements. So, for example, the property of being a station can be represented using circles and particular stations are named with labels, such as ?Victoria?. So, a circle labeled Victoria represents Victoria Station. By virtue of having this logical form, maps can interact with sentences in rational ways. Secondly we argue that maps do not have the resources for quantification. But, as long as maps lack a quantificational apparatus, they cannot express general thoughts such as ?all subways stations are open? and ?some stations are open? and, hence, cannot play the role of general premises. However, we will show that maps can play a more modest role as singular premises, that is, premises that contain neither variables nor quantifiers, and they are composed only of predicates and constants that functions as names for particular entities.Despite being limited in this ways, we argue that heterogeneous inferences are an extraordinary cognitive tool. One of the main virtues of this kind of inferences is that when maps are combined with a sentence, it can help us obtain information that it is neither explicitly nor implicitly represented in maps. In addition, unlike the case of purely sentential inferences, it is possible to draw many conclusions in an efficient way from the combinations of one map and one sentence.