Position Paper for Dagstuhl Seminar 10131: Spatial Representation and Reasoning in Language; Theme 5: Integrating temporal and spatial ontologies and logics for reasoning about motion and change Image‐schematic ontology integration Werner Kuhn Muenster Semantic Interoperability Lab (MUSIL) Institute for Geoinformatics, University of Muenster Abstract Image schemas have been identified as possibly universal abstractions from linguistic representations of space and time, including abstract spaces. This suggests that they can serve for ontology mapping. I will show an image‐ schematic formalization of basic spatio‐temporal relations and processes like containment, support, and motion, explaining how it supports ontology integration. Problem Geospatial information is typically separated into spatial and temporal components. The resulting map‐ and snapshot‐based views of the world hinder information integration and make reasoning about motion and change difficult. Since maps show what is, rather than what is happening, they have to be linked to process models built into specialized application software in order to monitor processes like land slides or oil spills1. The static view of the world inherent in spatio‐temporal data has also contaminated ontologies of space and time, pushing the integration problem a level up, from data to ontologies. For example, the standards and ontologies in the sensor web area are biased toward static geographic features (Stasch et al. 2009), despite the fact that all observations capture properties and effects of processes. They encourage data models where, for example, weather stations are observed phenomena rather than observers of weather processes, since they are handy static features with fixed positions. Similar map‐centered views of the world, rooted more strongly in geometry than physics, can be found in some evolving ontologies of mapping agencies. Approach and Hypothesis In order to integrate static spatial (and temporal) ontologies with those supporting dynamic reasoning, it is necessary to tie them to some kind of overarching structure. Cognitive science suggests that image schemas may provide such an integrative approach. These schemas have been identified as (possibly universal) patterns of space‐time conceptualization and reasoning in natural language (Johnson 1987). Rather than being about images, as their name might suggest, they combine spatial and temporal perspectives to process‐ 1 See, for example, the ENVISION project pilots, http://www.envision‐project.eu/