Representations of Cognitive Geography for Navigating Sustainable Neighbourhoods

forthcoming

At the heart of quantitative geographic analysis exists a critical disconnect. Digital representations of space, upon which all spatial data are based, do not reflect how geographic spaces are remembered by people. This fundamental issue of spatial representation has profound implications for how we analyse, model, and ultimately, design our cities and societies. If we fail to accurately represent cities as people see them, how can we expect our computational methods to accurately understand them? The ReCoGNiSe project will create a step-change in the digital representation of geographic space to better reflect the facets of spatial cognition that everyday shape human behaviour. In doing so, the project will build methodologies, evidence, and tools that integrate spatial cognition and GIScience. It will build new representations of spatial features that capture the multidimensionality of spatial learning. It will build representations of spatial relations that replicate how they are encoded in the brain. It will produce spatial representations and code that contribute towards the rapid advances in open science in geography. But going further, ReCoGNiSe will make manifest contributions to how we understand spatial cognition in the context of sustainable active travel. Following an extensive mixed methods analysis of the spatial cognition of walking, the project will use a range of experimental methods to test the validity of these new representations of space in predicting walking behaviour. This phase will grow the evidence base on walking behaviour; but also provide a new way forward for its effective design and promotion. ReCoGNiSe sets an ambitious target, building representations that can unite geographers and neuroscientists on common ground – geographic space. Achieving this objective could create a new paradigm in cognitive geography study, with significant implications for how we shape and design spaces – for people.