This book examines the dialectical role of semantic analysis within metaphysical inquiry. It argues that semantic analysis ought to be modest in its metaphysical pretensions in the sense that linguistic and conceptual analysis should not be expected to yield deep insight into either what exists or the nature of what exists. The argument turns on distinctions among narrowly linguistic semantics in the generative tradition and two varieties of broadly philosophical semantics which correspond to broad approaches to semantically infused metaphysical inquiry. In particular it distinguishes ideational semantics and metaphysical inquiry via the way of ideas, on the one hand, from referential semantics and metaphysical inquiry via the way of reference, on the other. It is argued that foundational assumptions of the generative framework are insufficient on their own to support the drawing of metaphysically immodest conclusions from the narrowly semantic premises. But it is shown that if we are determined to bridge the gap between narrowly semantic premise and metaphysical conclusion, we must augment our semantics with additional metasemantic premises. Such additional premises may come either from ideationalist or referentialist metasemantics. A number of arguments for preferring referential metasemantics over ideational metasemantics are offered. It is argued pursuing referentialist metasemantics as opposed to ideationalist metasemantics yields a semantics that is metaphysically modest. Finally it is argued that metaphysically modest should regarded as a feature rather than a bug of a semantic theory, one that serves to bring semantics into closer alignment with the special sciences generally.