My dissertation questions the interstices between oral and written modes of communication through a theoretical study on the interactions of structure, context and subject in literary texts that attempt to convey a spoken discourse. It focuses on the function of language in three twentieth-century Latin American Novels: João Guimarães Rosa's Grande Sertão:Veredas (1956), Julio Cortázar's Rayuela (1963), and Guillermo Cabrera Infante's Tres tristes tigres (1967). The study illustrates how these self-reflexive novels use language to develop narrative patterns that inscribe in the content the configuration of the form. The primary aim of the dissertation is to understand the written in terms of the oral, to account for how linguistic codes produce orality in written expression. The first chapter of the dissertation, "On Narrative," provides the critical framework that sustains the thematic analysis of the following chapters. The chapter discusses the make-up of narrative expression by identifying parallel theoretical concepts in oral and written narrative styles. The overlap between the basic formal features of these discourses (events, text, telling and time) is highlighted to show how and why it is possible for a written narrative to emulate spoken discourse. Additionally, the chapter includes a reflection on how reading can be transformed into a listening experience. The concepts of literary dialect and eye dialect are explained to demonstrate that the audible component of a text is found in the visible structure of the writing. The second chapter, "Narrative Consciousness," examines the use of literary dialogue as a structuring device in Brazilian novelist João Guimarães Rosa's Grande Sertão: Veredas. Written entirely as an exchange between a first person narrator and a silent interlocutor, Rosa's novel develops under a conversational framework, instead of following the traditional format of the genre. The chapter demonstrates how the portrayal of spoken discourse as the telling of a story by a mindful narrator reveals a commentary on the function and form of narration. The analysis centers on the interplay between orality and memory. The narrator's discourse is analyzed to show how the protagonist theorizes about narrative by way of the narrating act, and how his theory develops, within a fictional context, a particular conception on the uses of language and the limits of representation. The third chapter, "Articulating Authorship," addresses the post-structuralist debate on the question of authorship in a study of the main works of Argentinean novelist Julio Cortázar and Cuban author Guillermo Cabrera Infante. This section explores, within the experimental contexts of Rayuela and Tres tristes tigres, how the portrayal of the image of the author corresponds to a perspective on the production of literature. The meta-discursive commentary present in both novels is viewed as constituting paradigmatic proposals for the development of the novel and the role of the author. The chapter centers on the idea of dialogue as action, and on the interactions between representation and criticism. It establishes a comparison between the declaratory remarks made by the fictional authors and the evaluation of these claims by non authorial figures. The chapter also analyzes how the authoritative role of the fictional authors in these novels is not contingent on their continuous presence in the narratives. The final chapter of the dissertation, "Voices that Echo," conceptualizes narration as articulation. The chapter is about technique and the representation of style in Cabrera Infante's Tres tristes tigres. Specifically, the chapter studies the simulation of José Lezama Lima's literary aesthetics within Cabrera's text through a close reading of the vignette entitled "Nuncupatoria de un cruzado," to show that the underlying principles of a work are defined within discourse. Cabrera Infante reinvents recognizable narrative strategies and challenges the idea of authentic individual expression. The discussion illustrates how multiple discursive voices coexist within the narration and the ways in which these can be identified. The study exemplifies the emulation of another's voice and reflects on the changes that narrative undergoes when imitation and innovation are brought together. My dissertation shows that meaning is inscribed in the structure of discourses and that literary texts that challenge the limits between the oral and the written correlate form and content. Furthermore, it promotes a change in the way literature is read critically by not imposing theoretical frameworks on the text, and instead, questioning literature from within. The overall goal of the dissertation is to develop a working theory on how to read the oral in the written, and to value the text as the primary source of critique.