Translation in its typically modern, "Western" format relies on the primacy of a signified, construed as an act of correctly reconstructing the original signified from the (derivative) signifier. This essentially Cartesian conception has been retained even in the critical analysis of translation of literary works, especially by the scholars of the "American school" of comparative literature. Such an approach valorizes communication and fetishizes the source text and the author. A radically different model of a (source) text is presented by linked poetry in medieval Japan, in which a stanza is constantly re-interpreted by a poet who composes a sequential stanza in a dislodging manner so that no part of the text has one, fixed meaning. Hence, Naoki Sakai defines Japanese linked poetry as a "translational" text. In his explication of a "translational text," Sakai refers to a Korean-(Japanese/)American novel as well. While Earl Miner’s rendering of haika and renga successfully captures their spirit of indeterminacy of a meaning by refraining from giving definite signifieds to the translated stanzas, post-colonial texts also typically feature "translatedness" of a text. By examining Japanese linked poetry and postcolonial literature, this article explores the ways to deconstruct the Cartesian model of translation.