The paper explores the concept of "fisheye views" as a method for displaying large information structures, such as computer programs, databases, and online text. Fisheye views balance local detail with global context by showing more detailed information about nearby elements while reducing detail for more distant elements. The author, George W. Furnas, draws on naturalistic studies to support the effectiveness of fisheye views and proposes a formal framework for creating them. This framework uses a "Degree of Interest" (DOI) function that combines a priori importance and distance to determine which elements to display at different levels of detail. The paper includes examples of fisheye views in various contexts, such as tree structures, and evaluates their effectiveness through experiments. The results show that fisheye views significantly improve navigation and understanding of complex structures. The paper concludes by discussing the potential applications and limitations of fisheye views in different domains.The paper explores the concept of "fisheye views" as a method for displaying large information structures, such as computer programs, databases, and online text. Fisheye views balance local detail with global context by showing more detailed information about nearby elements while reducing detail for more distant elements. The author, George W. Furnas, draws on naturalistic studies to support the effectiveness of fisheye views and proposes a formal framework for creating them. This framework uses a "Degree of Interest" (DOI) function that combines a priori importance and distance to determine which elements to display at different levels of detail. The paper includes examples of fisheye views in various contexts, such as tree structures, and evaluates their effectiveness through experiments. The results show that fisheye views significantly improve navigation and understanding of complex structures. The paper concludes by discussing the potential applications and limitations of fisheye views in different domains.