A series of computer programs to handle natural language retrieval and analysis in a manner analogous to that of a statistical package is discussed. The document presents an overview of several language analysis projects currently underway, and of several research approaches to problems in language analysis. The manner in which the Language Analysis Package (LAP) could be used with each approach is considered, and design specifications for a state-of-the-art system are presented. A preliminary implementation of such a system using programs currently available at the Southwest Regional Laboratory is suggested. An overview of the design considerations and algorithms for LAP is also included. (Author/DGC)