This research posits that socially constructed spatial relations address concepts of interactions instead of intersections, human/tool agents instead of physical processes, and broader ranges of geographical outcomes. The hypothesis is that social space can be represented by using patterns of logic relations between sets of entities. The data corpus of spatial relations was extracted from geographic term definitions. The relations were further analyzed as primitives using Case Grammar Matrix models. These findings are being related to Web Ontology Language (OWL) properties. This approach allows an extensive range of natural language terms to instantiate ontology sub-types, while supporting inferences to study their logical implications.