Relevance Theory (RT: Sperber&Wilson, 1986) argues that human interpretative processes maximize relevance and postulates that there is a relevance-based procedure that a listener follows when trying to understand utterances. However, Mazzone (2013) points out that RT fails to explain how speaker-related information, such as the speaker’s abilities or preferences, is incorporated into pragmatic processes. He proposes that situational or goal schemata, with the speaker represented as a component, are sufficient to activate the hearer’s speaker-related knowledge and further asserts that human communication is driven by goal management and action rather than relevance maximization. Yet Mazzone cannot fully explain how linguistic meaning and speaker-related knowledge are integrated within a modular system. Based on RT’s cognitive requirements and contemporary cognitive theory, we argue that this integration is realized within working memory via production-like conversational rules with which the constructed utterance interpretation should be consistent, and present a simple model of this process.