Raises some key issues needing to be addressed by indexing practitioners and scholars in order to deal more effectively with the problems of unacademic indexing. Terminological confusion over the notion of subject and the continuing debate over its permanent or indeterminate nature provide no clear nor consistent explanation of what indexers might actually look for when indexing a document. These problems are highlighted by empirical evidence on measures of inter-indexer consistency suggesting that substantial inconsistency is the rule rather than the exception. Little attention has been paid to how indexers actually determine the subjects of documents or what guides them in establishing the aboutness. While the literature identifies some broad approaches to subject analysis there is little supporting empirical evidence and few attempts to explicate any specifiable procedures. A productive step forward with indexing research would be to begin by examining how indexers actually undertake the process of subject analysis and to explore systematically factors that guide and influence this process. This would shed some light on a theory of subject analysis, clarify some of the central concepts of indexing, and provide an intelligent knowledge-base for effective, academic indexing practice.