Building on Relevance Theory (Sperber & Wilson 1986/1995) and connectionist principles (Elman et al. 1996), this article presents the rationale and assesses the plausibility of a dynamic cognitive model of translator’s competence proposed by Gonçalves (2003). It differentiates between a specific and a general translator’s competence, supports the amalgamation of procedural and declarative knowledge as constitutive instances of translation competence, and argues for an increment of awareness-raising, meta-cognitive processes to develop this kind of competence. Translator’s competence is, therefore, considered to be a particular cognitive configuration that allows translators to establish a balance between the periphery and the central layers of cognitive systems and, from a situated perspective, enables them to attain an inferentially driven interpretive resemblance between source and target texts.