Mining the Author-Narrator Gap

Tuesday, May 28, 2013

If creative nonfictionists build a persona, can persona-building also become a source of conflict and dynamism in writing? Can building a less-than-reliable persona be a deliberate strategy, much like the use of unreliable narrators in fiction, such as Nabokov's Humbert Humbert? Or does any kind of unreliability in the narrator undermine the entire premise of creative nonfiction? In this five part TriQuarterly series, five writers of nonfiction and one writer of fiction brainstorm creative ways for writers to make themselves unreliable narrators—with playful, conflicted, and imaginative results.


“My thinking seems something separate from me.” So wrote Italo Svevo in his 1923 novel Confessions of Zeno.

“I can see it”—Svevo wrote of his “thinking”—“it rises and falls . . . but that is its only activity. To remind it that it is my thinking and that its duty is to make itself evident, I grasp the pencil. Now my brow does wrinkle, because each word is made up of so many letters and the imperious present looks up and blots out the past.”

Svevo’s novel, a fake autobiography by a fictional character named Zeno, is as fine a spoof as any of some of memoir’s worst foibles—its whimsy as to the truth or falsity of its content, its notion that one’s own thinking is somehow outside one’s accountability, its fetishistic caressing of the notion of writing one’s thoughts for no reason other than simply to write them down. In the words of Claire Messud, the book is, though fictional, a “liar’s memoir.”

When I first read this passage by Svevo, I was struck by its similarity to a real memoir, Confessions of an English Opium Eater, first published by Thomas de Quincey in London Magazine in 1821. This passage came to mind:

I trust that [this] will prove not merely an interesting record, but in a considerable degree useful and instructive. . . . [T]hat [is] my apology for breaking through that delicate and honourable reserve which, for the most part, restrains us from the public exposure of our own errors and infirmities. . . . All this I feel so forcibly, and so nervously am I alive to reproach of this tendency, that I have for many months hesitated about the propriety of allowing this or any part of my narrative to come before the public eye until after my death.

Both are emblematic examples of unreliable narration, one fictional, one nonfictional. Each, in its own way, reads as a kind of a joke: in the first case, a spoof on bad memoir; in the second, a document of a writer perhaps too close to his “confession” to be aware of its own transparency. The liar in the fictional version seems precisely a parody of the supposedly reliable—and yet deluded— narrator of the autobiography: “How I first came to be a regular opium-eater”—De Quincey also writes—“was not for the purpose of creating pleasure, but of mitigating pain in the severest degree. [This is how] I first began to use opium as an article of daily diet.”

Reading these passages over one beside the other, I began to wonder about not only their fundamental, generic, differences but the similarities between their narration styles. I began to think about what this suggested about the possibilities for the nonfiction writer, how the fun and play in Confessions of Zeno might be available for the rest of us as well.

DeQuincey’s confessions, of course, work better for the reader—or this reader, anyway—as “found object,” art by accident. As in traditional, fictional, unreliable narration, the de Quincy can be read ironically. It’s as if reader and publisher are having a joke on author/narrator. This mirrors the traditional Author-Narrator-Reader triangle of fiction. In unreliable narration, author and reader are in cahoots against narrator. Everyone—author included—knows the narrator is deluded, except for the narrator himself. As James Wood puts it, speaking of Svevo’s Zeno, its comedy “resid[es] in the incongruity between our concepts and objective reality.”

But perhaps accident is not the only possibility for the nonfiction writer meaning to exploit narrative instability in the ironic style of novelist Svevo. Perhaps nonfiction can also intentionally manipulate that incongruity? Should it, could it? By writing from a persona, the nonfiction writer always plays with the author–narrator divide. How far can that gap be expanded while one is still building a successful text? Can it be taken to the extreme of humor?

In e-mails back and forth with nonfiction writers Mimi Schwartz. Mike Steinberg, Clay Benjamin, and Tom Larson, I discussed whether we as nonfiction writers might attempt a narration style as outrageously unreliable as Nabokov’s Humbert Humbert, writing a narrator situated at so great a distance from author that it would be if author and reader were having a joke on narrator, one deeply and obviously deluded. While we disagreed on the extent to which the author–narrator divide might be manipulated, we all agreed there is a gap between author and narrator. Whether as large and humorous as in fiction—that, more or less, was the topic of our conversation.

Our project came about after I first read Joseph Epstein’s thoughts on unreliability in nonfiction in his introduction to the Norton Book of Personal Essays. He writes, summarily: “There are no unreliable narrators in personal essays; in a personal essay an unreliable narrator is just another name for a bad writer. We believe—we have to believe—what the writer tells us.”

His statement had always struck me as too narrow, too limiting. And while Epstein is prescriptive, the idea of memoir in particular as the province of unmediated “truth” is often presented descriptively, for instance when Charles Baxter writes: “In a culture afflicted with data-nausea, I thought, everyone begins to hoard and value the autobiographical, a refuge safe from irony.”

Is autobiography really a “refuge safe from irony”? Unpacking this statement, I think what Baxter is getting at is that truth is opposed to irony, which it is, since irony is by definition “the expression of one’s meaning to signify the opposite."

But the lie at the heart of the statement—and Epstein’s—is the assumption that autobiography resides categorically in the province of truth.

If the liar in fiction is a parody of the supposedly reliable narrator of the autobiography, that parody has such power because the autobiographer is in many senses inherently unreliable. I’m interested in the ways that autobiographers might make use of, rather than deny, that inherent unreliability.

I certainly do so in my own current project. This became especially clear to me recently, when my agent asked me to change the prologue to my recently completed memoir from its original, set-in-the-present opening. In the original, this depicted an unraveling of things for my family during the summer of 2011, which I’d set as the “present moment” of the narration for the rest of the book. This 2011 material set up an opening into an exploration of childhood and the past. My agent said this opening was a downer. I decided she was right, and changed the present moment opening to a point two years in the past of “the present” (i.e., 2009). Once I did that, I had to go through the memoir rewriting the present-moment reflections so that they accurately reflected this different starting/vantage point. Then I decided that the present moment needn’t be static but could move: from 2009 to 2011. So now, in my memoir, the unraveling of 2011 is revealed at the end of the book. The present-moment voice develops from beginning to end, from 2009 to 2011.

This reminded me that the present-moment narrator was an unreliable one, one who was ignorant of what would come though I, the author, already knew. It became easier to present the material with this more distant “I”—easier to see her as a character, another “me” of the past just as the child “me” in the sections set in the “deep past” was a character.

Like many memoirists, I write nonfiction from the point of view of knowing that the narrator “I” is different from the author “I.” In the gap between the two resides unreliability—or, in any case, fallibility. As in the Heisenberg principle, what I see shifts, reliably so, when my vantage shifts.

Perhaps these words illuminate the issue of unreliability somewhat, reliably, or unreliably.