Critiquing the tendency to look at how physically accurate a map was at the expense of examining the politics in the map, Soja argues that"In all these approaches, spatiality is reduced to physical objects and forms, and naturalized back to a first nature so as to become susceptible to prevailing scientific explanation in the form of orderly, reproduceable description and the discovery of empirical regularities (largely in the spatial co-variation of phenomenonal appearances). Such a short-sighted approach to space has proved productive in the accumulation of accurate geographical information and seductive as a legitimization for a presumed science of geography. It becomes illuseive, however, when geographical description is substituted for explanation of the social production of space and the spatial organization of society, in other words, when geographical appearances are asserted as the source of an epistemology of spatiality" (123).
Such a position does not argue that maps should be in-accurate, or that "orderly, reproducible description" is a bad thing. But we cannot allow such concerns to replace critical, social thought.