Abstract
Although cataloging in libraries has been done for decades with the oft-stated intention of serving user needs, there is little evidence that those needs have been empirically assessed. This paper proposes a domain analytic model for systematically making such assessments, and for altering existing cataloging practice to meet the assessed needs. The author's dissertation research into the catalog needs of a single, narrowly-defined domain is used as an example of the type of study that may be repeated with an infinite number of domains, thereby providing a type of data regarding catalog needs not previously available.