Gutt's relevance theoretic approach to translation offers a cognition-based theoretical framework for the study of translation. However, as Tirkkonen-Condit notes, it does not go far enough, since it fails to show what is unique to translation and what role interlinguality plays in translation. The present paper represents an effort to make one step forward by elaborating the concept of secondary communication situation and by following up some of the problems raised by or naturally arising from Gutt's approach. The paper accepts the view that translation is a prototype concept ranging from simple repetition in the same language at the same time through intralingual translation to written interlingual translation. Ironically, the role of language change in translation, neglected in some theories of translation, has to be emphasised again. Similarly, the author finds that the role of constraints on achieving relevance in various types of communication must receive more attention.