If one is committed to computationalism, one must ask: why? Is it because of a fondness for the particular formalisms with which we currently understand computation, and the belief that these will provide an account of cognition? Or is it because of an intuition that there is something special about computers, when compared to other human artefacts, and that whatever explains that specialness will also explain the specialness of human cognition? Those who opt for the latter have a means of responding to objections to computationalism, a means that those who opt for the former lack. But these transparent computationalists thereby incur the responsibility of identifying phenomena that call for new theories of computation, and of producing such theories.