In this paper we argue that cartography is profitably conceived as a processual, rather than representational, science. Building on recent analysis concerning the philosophical underpinnings of cartography we question the ontological security of maps, contending that it is productive to rethink cartography as ontogenetic in nature; that is maps emerge through practices and have no secure ontological status. Drawing on the concepts of transduction and technicity we contend that maps are of-the-moment, brought into being through practices (embodied, social, technical); that mapping is a process of constant reterritorialization. Maps are never fully formed and their work is never complete. Maps are transitory and fleeting, being contingent, relational and context-dependent; they are always mappings; spatial practices enacted to solve relational problems (eg, how best to create a spatial representation, how to understand a spatial distribution, how to get between A and B, and so on). Such a rethinking, we contend, provides a fresh perspective on cartographic epistemology, and could work to provide a common framework for those who undertake mapping as applied knowledge (asking technical questions) and those that seek to critique such mapping as a form of power/knowledge (asking ideological questions). We illustrate our argument through an analysis of mapping practices.