Written a paper on how we can compare and even integrate space syntax with spatial interaction. You can get it here by clicking on this link or on the image above. This is based on the notion that we need to develop a way of disentangling the underlying planar graph of the street network into components that when put together either lead to operations on the planar graph itself as in spatial interaction or on the dual of this graph between the streets which is the graph used in space syntax. The link is obvious when developed in this way but the ways of integration are somewhat convoluted. We develop these ideas for some simple hypothetical graphs and make comparisons of the various accessibilities associated with these graphs which appear both in space syntax and spatial interaction. We then develop a semi-real application using data on the nearest neighbour generic graph for Greater London and this reveals the problems of specifying this graph in the first place. Our comparisons with real data are disappointing for many obvious reasons but what I think this paper does is throw light on space syntax and on ideas about accessibility, suggesting that we need a sustained effort to develop the right sorts of underlying graphs from which space syntax can be developed. We are only at the beginning of the process and in space syntax, we have not really explored the properties of the underlying graphs in any depth hitherto. This establishes directions for further research. Click here for the paper.