The process of operationalizing computer-constructed "hot spots" for a 1988-1989 crime experiment in Minneapolis illustrated mutual effects between experimental design requirements and practical concerns of both field research and operational policy. Experimental results identified at least three different decision points where abstract concepts of space had to be negotiated in operational terms: (1) in the nature of human techniques and practices that assign activities to particular addresses in official records; (2) in the attribution of public space; and (3) in conflict over the nature of boundaries, distinct and discrete in computerized representations but invisible and fluid under field conditions. 11 references