The potentiality exists for a multitude of different geographic objects to be established from the same census dataset. Faced with a variety of aggregation procedures it is critical that we develop research into geographic objects based on a standard methodology which satisfies a set of specific aggregation criteria. If not anarchy will prevail in the form of copious new output geographies. This would be a disaster in that geographic objects are of fundamental importance to the study of both physical and human geographic structures. In census geography the study of geographic objects primarily aims at acquiring vital information on differences between cities, the evolution of those cities and the study of urban management. After initially discussing the issue of which aggregation procedure to use this paper adopts a standard procedure, that of the fragmentation and objectification analysis, which is applied to a race/ethnicity dataset for Los Angeles. As displayed in the Los Angeles application the number of types of major fragments and geographic objects that exist in this city are very limited, but the number ofgeographic objects are many. Los Angeles beyond the two citadels of the Whites and the barrio of the Hispanics is a very fragmented city.