NOTE: This is the third and final installment in a three-part series on earning and maintaining a reader’s trust. The first part focuses on earning initial trust just at the start of a story, while the second part focuses on how world-building, consequential plotting, and story structure/pacing affect the reader’s trust. This part deals with character consistency/reliability, and I know it’s long. I do apologize for that, but there’s really a lot to talk about here.
Consistent Characterization and Reader Trust
When we write, we create a wide variety of characters, each of whom has different degrees of complexity. Like real people, our characters’ choices, attitudes, personalities, and decisions are shaped by their experiences. The most memorable characters are those that are shown to be complex, to have foibles and flaws as everyone else. Readers appreciate flawed characters, but what matters is that their flaws and behavior are consistent with the events of the story. Some people claim that they like to be surprised by characters, but there is a big difference between letting the plot surprise us, and letting a character do so. Character actions should be an inevitable consequence of their natures, and their experiences before and during the story.
Every decision a character makes must logically follow from the experiences our reader has observed through the story. In Les Misérables, Hugo shows us the moral quandary that Marius Pontmercy finds himself in on the eve of revolution: should he join his friends on the barricades or escape with the love of his life, Cosette? Hugo establishes this as a real choice for Marius, one that forces him to choose between two equally “right” values (according to his own value system). As the reader, we understand that he can believably go either way on the choice. Which makes his final decision and the reveal both a surprise (either of the two options would have been) and satisfying.
The seeds of every major (and most minor) choice should be planted well in advance. The hard part, is to plant seeds that allow for branches of equal probability. If the character only has one plausible recourse, then where does tension derive from? This is one of my most frequent complaints about portal/quest fantasies, in particular those of the “prophesied monarch” sub-type. As I’ve grown older (and more curmudgeonly) I have found it very difficult to get any emotional tension out of this kind of story. They become predictable and dull, because I know a priori that every complication the hero runs into will at some point be resolved, and that every mistake he makes will be fixed by the end. The prophecy (which is all too rarely actually ambiguous) will take care of matters in the end.
By setting up characters who have real choices to make internally, and who have real conflicts amongst themselves, we maintain the reader’s interest in the underlying story – which is a prerequisite for maintaining their trust. If the character does something that was not adequately prepared for, something so surprising that it comes out of left field, then the reader may be shocked into losing all trust in the story.
Consider for a moment Star Wars. Would the story have worked if in Return of the Jedi Princess Leia betrayed the rebels to the Empire? No. It might have worked (though yielded a very different story) if Han Solo had done so (for money), or if Luke turned to the Dark Side. But Leia? There was nothing in her character to make such a choice remotely plausible. It would have been a bridge too far, a leap of faith that the audience would have been unwilling to follow.
However, this does not mean that characters need to always be reliable. In fact, one of my favorite methods of playing with reader trust is the use of an unreliable narrator/character.
Structures that Enable Trusting in Unreliable Storytelling
Whether it’s in film (Rashomon, The Usual Suspects, or Citizen Kane) or in prose (Akutagawa’s In a Grove, Nabokov’s Lolita, or Larbalestier’s Liar), I love unreliable narration/characterization. It’s a lever on which my entire understanding of a story can hinge. Executed skillfully, it offers an exponentially broader story experience. But how does the reader trust in a story when the storyteller is shown to be a liar? That’s a question that has been on my mind quite a lot recently, as one of my current WIPs deals extensively with the concept of deception.
In thinking it through, I think I’ve managed to identify five different modes of unreliable storytelling, each of which plays with reader trust in different ways. This list is by no means exhaustive, and I’d love to know other examples:
|The narrative voice is strongly imbued with the characteristics of the story’s narrator, thus adopting the inherent biases, idiosyncrasies, or fallacies held by the narrator. These may include a childlike naivete, a desire to rationalize/justify morally reprehensible actions, or strongly held opinions that color the narrator’s perspective. What matters in such stories is that the narrative voice must go beyond the merely functional, and present a particularly close “lived-in” feel.
The book’s character/narrator will likely be the most memorable aspect of the book, and this stems entirely from a reliance on the narrator’s voice. The source of the narrator’s underlying unreliability affects our emotional position vis á vis the character: we love Huck Finn or Evie Spooner for their childlike innocence, even as that innocence is shattered by their experiences. We love to recoil from Humbert Humbert, and his beguiling rationalization of his monstrous deeds. We judge Chaucer’s Merchant, and the Wife of Bath for the positions and opinions they hold. At no point in the story itself do we as the author insert ourselves and tell the reader about the veracity of our narrator’s statements. Our job is to present the story as if it were the narrator’s, with whatever inaccuracies or ugliness that entails.
The success or failure of these stories rests on their ability to draw us into the character/narrator’s viewpoint. To aid in this process, such stories are often told in first-person to accelerate us into the reader’s head, though that is by no means a requirement. Typically, the reader’s enjoyment derives from a multi-layered interpretation of the text. On the one hand, we can enjoy the events unfold as we share in the narrator’s experiences. On the other hand, we have an intellectual and emotional response to our own interpretation of those experiences based upon our own value systems. We take the facts of the story as given, and generally we do not dispute them. However the moral and emotional implications will be drawn from the reader’s own values and opinions.
So long as the narrator is consistent in terms of voice and characterization, the reader will trust that the narrator is supplying the basic facts of the events accurately. However, if the voice is distinct enough, the reader gains that degree of separation that enables that multi-layered interpretation. This makes apparent the fact that the reader is expected to have value judgments that are independent of the narrators’. As a result, the reader will supply their own emotional/moral “truth” , based upon the facts filtered by the narrator. In essence, the reader is expected to trust the facts of the story, but to question the narrator’s interpretation of those facts. The reader’s “trust” in this type of unreliable narrator rests entirely on that narrator’s voice, and its distinctiveness and attractiveness (even if the character is reprehensible).
|When we combine multiple implicitly unreliable POVs, the result is often an interesting structure which throws into doubt the facts of the story. Here, it is not only the moral/emotional implications which need to be supplied by the user, but the facts of the story as well. Most frequently this story relies on presenting a series of narrators, each of whom recounts the same or closely-related events from their own highly subjective perspective. This structure creates a more complex intellectual puzzle than most implicitly unreliable stories, as it requires the reader to parse and analyze what characters want, what they say, and what they do not say. By analyzing the gaps between what different narrators tell us, the reader can infer the “true” facts of the story.
The facts of the story are themselves in flux in such stories, and will never be explicitly stated by any of the narrators. And because the underlying facts are ambiguous, so too are the emotional and moral implications of the story as well. This mode relies more heavily on a careful analysis of the details included in particular narrators’ versions of the story. Who includes what details, who mentions what, who justifies what actions, who lies, and when they do so are all vital factors that we need to have carefully mapped out as we write the story. The reader’s trust relies on the non-obvious nature of the “truth”. Because this type of story is a puzzle-box, readers who figure out or intuit the puzzle within the first couple of chapters may lose interest: their intellectual investment will have been wasted. To maintain the reader’s trust, balance must be maintained between all of the perspective characters, in terms of level of detail offered and the reader’s expected emotional investment.
|In many stories, we are explicitly told that we should not trust the narrator: that their words cannot necessarily be taken at face value. This may be because the character is a self-avowed liar, or is insane, or because a framing device tells us a priori that the story is untrue. In each case, this model puts the reader on alert that they are dealing with an unreliable narrator and forces the reader into a “problem solving” mode of story consumption.
This model seems to be especially popular in speculative fiction, where a number of authors (Gene Wolfe in particular) execute admirably. When we read stories like There Are Doors or Soldier of the Mist, we question every statement (however banal) that the narrator makes. In essence, we’re playing a perennial game of gotcha with a chimerical narrator: it is only through the narrator that we can glean insight into the author’s intent, and by catching the narrator out we hope to deepen our understanding of the story’s thematic implications.
This is my personal favorite type of unreliable narrator, as when done well, it leaves almost infinite room for conjecture. Books like Soldier of the Mist or Justine Larbalestier’s Liar leave us room for hours and hours of discussion and examination. These stories rely on a combination of factors to maintain reader trust:
First, we must balance the narrator’s stated unreliability against the need to ground our reader in the story. This balancing act rests on the idea of uncertainty. In Soldier of the Mist, Latro’s interactions with gods and monsters may be a result of the head wound that gave him his anterograde amnesia. The character acts as though they are real, because he is unaware of this uncertainty. We as the reader need to choose which interpretation we believe: the explicitly stated, skeptical viewpoint? Or the character’s credulous one? We face the same choice, though further complicated, in Justine Larbalestier’s Liar where on the first page our narrator tells us that she is a pathological liar:
We are at first told that the narrator is a liar, explicitly letting us know that she is not to be trusted. But a moment later we are told that from here on in, everything she says is the truth (“No lies, no omissions.”). At first blush, we might be tempted to believe this. But then two paragraphs further we find two short words which again make us doubt: “This time I truly mean it” (emphasis mine). This push-me/pull-you dynamic is characteristic of these kinds of stories, even when the narrator’s unreliability is more of a background condition (as in Kesey’s One Flew Over the Cuckoo’s Nest).
In the stories where this explicit unreliability is effective and maintains the reader’s trust, there exists a thematic consistency that encourages dueling interpretations of the text. We can almost look at it has having two separate ways of reading the stories: one where we accept the narrator’s word, and one where we disbelieve most of it. If each interpretation wrestles with similar themes, and if each remains plausible based on the text, the reader’s trust will be maintained. By making the narrator explicitly unreliable, we are entering into a contract with the reader, promising that we will consciously play with “truth” in the story.
This is one reason why I was disappointed by James Clemens’ The Banned and the Banished series: in the first novel, Wit’ch Fire, Clemens uses an interesting framing device to establish that the story we are about to read is (ostensibly) false, with hints that this stems from political revisionism. However, as we get into the meat of the story this frame becomes practically forgotten: the story devolves into a fairly standard portal/quest fantasy, with marginal attempts at exploring the ambiguity introduced by the book’s forward. For those of us who like unreliable narrators, ambiguity is like an awesome toy. If the author puts it on the table, we want to play with it. If it’s there, but we aren’t allowed to play, then we feel cheated.
However, it is possible to have too much of a good thing. Much as I admire Larbalestier’s Liar, I found the constant retractions and restatements tiring. When we craft an explicitly unreliable narrator, we’re asking our reader to pay constant attention to the various plausible interpretations we offer them. The more variants the reader must store in their head, the more tiring the experience will become. In most successful cases, the author introduces the narrator’s unreliability and then leaves us with just two ways of interpreting the story: either based on the narrator’s prima facie interpretation, or taking the narrator’s statements with a grain of salt.
In Larbalestier’s case, her story rested upon a narrator who explicitly contradicted her story some twelve times (by my count). This created a swirling cloud of possible interpretations, with many fractal branches to consider. Of course, this was Larbalestier’s thematic goal. However, neither the voice nor the story’s underlying conflict were strong enough to justify the significant investment of effort demanded of me. This may simply be a question of taste, and my own ability to identify with Larbalestier’s character. Regardless of how much I might admire the book’s structural ambitions (which – unquestionably – Larbalestier delivers on excellently), the narrator’s voice was not quite strong enough to maintain my trust.
Bottom line: for explicitly unreliable narrators, make sure that their unreliability relates directly to the story’s thematic concerns, be careful of asking the reader to keep too many plausible interpretations in their heads, and try to offset the inherent complexity through an engaging voice and conflicts.
|Revealed unreliability relies on a moment of anagnorisis or discovery regarding the narrator. While more commonly an element in film than in novels, this probably owes its origins to Agatha Christie’s classic mystery The Murder of Roger Ackroyd. Stories with revealed unreliability rely on a twist ending where some fact is revealed about the narrator (e.g. their identity, mental state, etc.) which forces the audience to re-evaluate the entirety of the preceding story.
Twist endings of this kind are very controversial, and difficult to pull off. Badly rendered twists (text: “It was all a dream!” author: Bahahaha!) are considered a trite cliché. Debate still abounds in the mystery community around whether Christie’s classic is good or bad. That book relies on a narrator who purposefully leaves out vital clues and inserts many red herrings to obscure the killer’s identify – up until the very end of the book, when the killer is revealed. As an early form of this kind of mystery, The Murder of Roger Ackroyd is ostentatious by most measures. Even astute and experienced readers of the genre have difficulty figuring out the mystery before the killer is revealed, which is no doubt one source of readers’ frustration.
Revealed unreliability has become much more accepted – particularly in the literary community. Nevertheless, it is a difficult feat to pull off effectively. To function properly, the author must take great care to lay the seeds of inevitability such that the “answer” becomes apparent on subsequent readings of the same story. Palahniuk’s Fight Club does this particularly well by establishing the tonal uncertainty of the narrator’s own mind: at no point before the reveal is the narrator explicitly shown to be unreliable, but the narrator’s own doubts as to his reliability create the possibility of the reveal near the end.
From what I can see, a revealed unreliability is easier to pull off in film, where the use of visual images can rapidly communicate the revelation to to the audience. Because the train moves very fast, the audience doesn’t have time to feel cheated. Prime examples of this include The Usual Suspects, A Beautiful Mind, and Christopher Nolan’s The Prestige.
In revealed unreliability, it is very easy for the audience to lose trust at the moment of revelation. The reader’s trust relies on a sense of not being cheated. This relies on the author salting the preceding events with enough hints that before the revelation are innocuous enough, but after the revelation make it seem inevitable. Furthermore, the revelation must fit smoothly and plausibly into the preceding events. If it does not (“It was all a dream!”) then the reader will feel like the author pulled a fast one, and cheated them of a satisfying experience.
Everything Relies on Everything Else
In conclusion, everything about reader trust relies on consistently and smoothly introducing the story’s building blocks such that the reader does not notice. I think the train metaphor is a good one for that trust: If the reader can count the rivets on the train, then the train isn’t moving fast enough. The speed at which it moves is only partially a question of pacing. The train’s engine is stoked by the cultural touchstones it relies upon, the narrative voice it is told in, and the author’s precise use of language. It runs on rails of world-building, and story structure, and consistent plotting. And it’s driven by characters who are internally consistent, whether they are reliable or not. If they’re not reliable, then at the very least they need to be functionally unreliable, to have that reliability carefully mapped out by the author so that the reader’s trust is maintained.
Many good stories fall short on one or more of these components. And that’s okay. Honestly, I can’t think of any “perfect” stories that nail every aspect of this. It might be impossible (at least by us mere mortals). But even if they’re not all equally solid, the components do need to balance and work together to earn the reader’s initial investment, to earn their trust, and to keep them turning pages. Which is ultimately the goal: readers show us their trust by turning to the next page.