Figure 1 presents a simple model of communication for GIS that aims to illuminate some key locations in the GIS production process (from software development to interpretation and use of GIS output) at which bias may be introduced (Bunch 2001a). This model is informed by the Critical GIS literature and earlier work by Robinson and Petchenik (1976) and Chrisman (1987). Its purpose is to embody some of the understanding in the Critical GIS literature, acting as a heuristic device to help bridge the gap between conceptualization and formalization in Critical GIS identified by Schuurman (2006), so that GIS practice avoids marginalization, and so that its use may instead lead to empowerment of stakeholder groups and communities. We use this model to inform discussion of the case studies presented later in this paper.