For the past few years, I've wandered around narrative conferences muttering. "Please, not another talk about the implied author," "Can't we banish arguments about the implied author from the conference? .... How about a moratorium on the term 'implied author' for next year?" Those cranky comments were calling not for the erasure of a term but for the admission of a stalemate: I believed that the longstanding debates about the implied author had reached a point of diminishing returns. I am hoping that the current issue of Style will prove me wrong. Meanwhile, hoisted with my own petard to participate in a conversation I'd proclaimed we should banish, I offer a set of propositions about the non-entity that continues to haunt narratological discourse. That I adopt a quasi-theological language is a conscious choice that attempts to render the spirit of this prolonged inquiry.
Let me clarify that I don't consider the issue of implied authorship unimportant, just stuck. The manifesto I offer here is an effort to articulate a set of principles that might lead if not to consensus, then to a shift in our approach, a move neither away from nor toward the implied author but toward a different critical practice that might tell us whether--and, if so, how--divergent understandings of implied authorship affect the poetic-hermeneutic enterprise. My propositions are agnostic not only because I myself remain so with respect to the concept of the implied author, but because I hope they will speak to theorists on both sides of the 1A divide. As steps toward praxis, these premises build on one another and are thus best read in sequence.
1. The implied author is not an empirical entity. It is neither an identifiable textual voice nor a demonstrable material being. It therefore has no place in a model of communication.
Nearly all literary scholars agree that the implied author is neither the historical ("flesh-and-blood") author of a text. nor a narrator, nor any other textually identifiable persona. But we do not always act in keeping with this axiom. Even though Seymour Chatman, for instance, argues that the implied author"has no voice" (41), he famously includes the implied author as a figure in his chain of communication. Shlomith Rimmon-Kenan criticized this inclusion almost immediately (89), and there has been no successful refutation of her position, yet Chatman's model continues to be evoked even by narrative theorists who disagree with it, which perpetuates its visual power. As Ansgar Nunning reminds us, no one has ever been able to tell us where the implied author resides, except in ways that conflate the implied author with the "totality of meanings that can be inferred from a text" (Bal 18).
2. Since the implied author has no material being or textual identity, it is necessarily a reading effect, it is something that happens rather than something that is, and it happens in the wake of reading rather than prior to it.
This axiom holds true no matter how one defines the implied author and no matter what one believes about the relationship between the implied author and the historical author. As Bal argues, the implied author is necessarily "the result of the investigation of the meaning of a text." In other words, "only after interpreting the text on the basis of a text description can the implied author be inferred and discussed" (18). This necessarily sequential--and thus dependent--position of the implied author distinguishes it dramatically from the historical author. Without a reader, there is no implied authorship. Even if the implied author is "contained" in the work, as Wolf Schmid recognizes, it is "not represented"; it has "only a virtual existence" (26) and therefore must be created from the work rather than the other way around.
3. The belief that a particular text is an intentional human discourse provides the only feasible rationale for deploying the concept of implied author. …