The chapter examines Fred Dretske’s teleosemantic theory of mental representation, focusing on his innovative account of misrepresentation as dysfunction. Through analysis of his conception of representational function as an indicating function, it demonstrates how Dretske resolves the hard problem of content by naturalizing satisfaction conditions. The discussion reveals how his approach distinguishes itself from causal-informational theories by allowing for genuine representational error while maintaining naturalistic credentials. Critical evaluation shows both strengths and limitations of Dretske’s framework. While successfully addressing the disjunction problem and providing a sophisticated account of representation at personal and subpersonal levels, his theory faces challenges regarding the distinction between mental and perceptual representations based on individual learning. Using the example of frog predation, the analysis demonstrates how his account handles cases of misrepresentation while revealing tensions in his views about cognitive flexibility and learning. The discussion establishes that although Dretske’s theory offers valuable insights into the nature of representational error and cognitive architecture, it inadequately addresses how systems detect their own errors. His framework provides foundations for understanding mental representation while indicating areas requiring further theoretical development, particularly regarding productivity and error detection capabilities of cognitive systems.