It is widely understood that to address vexing social problems, learning efforts require collaboration among diverse communities of experts. This has spawned research in the means for data integration and data lineage, or provenance, to better enable sharing of data and workflows. Such research has focused on specific problems in domain representation, such as ontology use, provenance vocabularies and methodologies, and automated workflow management. Management science has also focused on collaboration network success factors. While these contributions are highly relevant to successful collaborative investigation design, none account for all elements of collaborative investigation. Such a model should accommodate both inductive and deductive linking and combining of data and process to take advantage of the diverse array of data and participant networks available to modern research efforts. Information science theory and standards and management science theory regarding collaboration were reviewed to develop a framework for collaborative investigation. The framework suggests basic data patterns and operating principles for social science research with extensive use of geospatial data within large research organizations. The work incorporates concepts from emerging sensor-observation standards and W3C provenance standards. A comparative case study methodology of three real world, socially complex, geospatial investigations provided empirical input to refine the framework. The result is a theoretical framework for Collaborative Geospatial Investigation (CGI), including eight basic content classes and related information design considerations for investigation within this institutional context. A formalization of direction conveyance to enable the articulation of basic rules and guidelines of execution of CGI is defined. Finally, insight into specific leadership and empowerment behaviors which enable collaboration in this context are also put forth. This work can be used to design new Geographic Information Systems (GIS) and system architectures to better enable collaborative, geospatial data intensive, social science research.