In this paper, I explore the relationship between Relevance Theory and Jaszczolt's Default Semantics, framing this debate within the picture of massive modularity tempered by the idea of brain plasticity ( Perkins, 2007). While Relevance Theory focuses on processing (see cognitive efforts and contextual effects interplay), Default Semantics focuses on types of sources from which addressees draw information and types of processes that interact in providing it. In particular, I argue that Relevance Theory interacts with default semantics by standardizing inferences which are ultimately compressed (to use a term by Bach, 1998) into a default semantics. I briefly discuss potential obstacles to the idea of default semantics coming from the experimental pragmatics literature (e.g. Noveck and Sperber, 2007; Breheny et al., 2005) and I support further the idea of the division of labor default inferences and the inferences derivable through the Principle of Relevance. In the end, I compare Relevance Theory and Default Semantics, in an attempt to come to a more unified picture.