In his paper “Fanciful Arguments For Realism,” Alan H. Goldman offers what he terms a “local case” argument as part of a more general defence of “semantic realism” against the anti-realist manoeuvres of Michael Dummett and Hilary Putnam. Semantic realism, here, is the thesis that sentences in a natural language L may have content that transcends any verification or assertability conditions associated with these sentences by competent speakers of L: an adequate semantics, the realist maintains, must equate the content of an assertoric sentence with its realist truth-conditions. Goldman contends that his local case argument demonstrates both the need to accord a central role to a realist conception of truth in an account of linguistic meaning, and the manner in which we acquire such a conception. He further maintains that the argument cannot be countered by any of the strategies that anti-realists are wont to deploy against “global case” arguments for realism (“evil demon” and “brain in a vat” scenarios, for example), and that, given the local case argument, we can recognise the fallacy in such anti-realist strategies and thereby rehabilitate the global case arguments that a full-blown semantic realism requires. I argue that Goldman's local case cannot fulfil its intended function in the overall economy of his argument, and that his strategy for defending semantic realism is fundamentally flawed. In so doing, I attempt to curb the realist's flight of fancy before it leaves the ground.