Humans have an innate reasoning ability that allows for the conversion of a verbal description into a real world location. Computers can mimic this process by breaking it into several smaller problems. These include speech recognition, deep-language understanding, spatial reasoning, and geospatial image matching. Although researchers have explored each of these fields extensively, they have not yet combined them into a complete system. In this paper, we explore the possibilities and limitations of such an automated system with a focus on spatial reasoning and geospatial image matching.