According to the traditional conception of the mind, semantical content is perhaps the most important feature distinguishing mental from non-mental systems. And this traditional conception has been incorporated into the foundations of contemporary scientific approaches to the mind, insofar as the notion of ‘mental representation’ is adopted as a primary theoretical device. Symbolic representations are posited as the internal structures that carry the information utilized by intelligent systems, and they also comprise the formal elements over which cognitive computations are performed. But a fundamental tension is built into the picture - to the extent that symbolic ‘representations’ are formal elements of computation, their alleged content is completely gratuitous. I argue that the computational paradigm is thematically inconsistent with the search for content or its supposed ‘vehicles’. Instead, the concern of computational models of cognition should be with the processing structures that yield the right kinds of input/output profiles, and with how these structures can be implemented in the brain.