In this talk I will consider what is known about how the human mind and brain represents individuals non-linguistically, and what the implications are for our theories of linguistic interpretation. In standard model-theoretic semantics, individuals are simply taken for granted as part of the world model; and I too will assume that the world itself contains individual entities. But if we take seriously the goal of developing a theory of the natural language semantics that is instantiated in the human mind, we need to know something about how the mind manages to *approximate* the world in mental representation, because it is this mental representation of the world to which natural language has to be mapped. As Bach (1989) noted, this is as challenging for individuals as for anything else. Although cognitive neuroscience is still in its infancy, today some rough outlines of how the brain represents individuals are becoming clearer. There appear to be multiple, redundant, non-linguistic representations of individuals in circuits that evolved for different purposes and which differ in format and temporal continuity. Some of our representations of individuals are fleeting, serving the purpose of helping us to accurately represent the structure of a current situation or interaction while it is happening. Others, supported by different brain circuits, are designed to last a longer period of time, allowing us to re-identify a recently encountered individual and successfully integrate new knowledge about them with prior knowledge. Yet a different form of representation for known individuals is included within the semi-permanent long-term knowledge base that also supports generalized knowledge. Evidence from neuropsychological patients implies that we cannot rely on introspection to tell us which of these circuits we are drawing on in any given instance of thought or reasoning. The diversity of individual representation formats suggested by cognitive neuroscience raises new and interesting questions for semantic theory about what we should take to be the internal world-model in which sentences are interpreted, and can inspire new ideas about how human language was shaped by its interface with these pre-existing representational formats.