In discussing the processes of development so far we have concentrated upon theories which are set at the regional level of resolution. This tends to be problematic for at least three reasons. First, it is unclear what is being referred to in the designation ‘regional’ in these theories. We do not know whether they are intended to refer to, for example, areas defined in the way that planning regions are, or administrative units such as states as in the USA, or whether they are perhaps defined in the process of analysing development patterns in space. Second, it seems difficult to abstract regional processes from urban, in the sense that since all regional growth is theorized as occurring in urban concentrations of some kind, every regional develop­ ment theory offers some, albeit general, rationale for urban development too. And third, and most importantly, designations such as ‘regional’ and ‘urban’ clearly reflect a particular, spatially-dominated, way of thinking about processes which are not themselves primarily spatial, but social (Anderson, 1975; Massey, 1978). They tend to divert attention away from the mechanisms which lead to space being consumed in variable ways, and, as we have seen in the case of disequilibrium theory, can result in a mis­ leading causal analysis which endows space with capacities which are, in reality, social. Some of these problems are clearly recognized in attempts to reproduce the variety of spatial scales of development in terms such as city-region, sub-region, conurbation, urban realm, megalopolis and so on, but these verbal containers obfuscate rather than clarify distinctions, a feature which stems from a reluctance to relinquish the initial spatial categorization.