This thesis argues that John Searle's theory of institutional facts, put forth in The Construction of Social Reality (1995) and related articles, is defective in several respects. After outlining the essentials of the theory, we criticize a central tenet: that collective intentionality is an irreducible and biologically primitive phenomenon straddling the gap between animal and human societies. We argue that a theoretical model developed by the cognitive anthropologist Dan Sperber provides better concepts of the social and the cultural, as well as an interesting alternative analysis of institutions. Next, a real-life historical example is used to test Bruno Celano's claim that Searle's theory fails to handle false beliefs. We then suggest that the theory relies too heavily on deontic commitment, and that the theory of coordination games as pioneered by David Lewis can explain more with less. Some implications for the social sciences are briefly noted. The final section argues that Searle ignores the need for an epistemological point of view in defending his naturalistic premise, and that the concept of 'fact' upon which he builds his theory is problematic.