Creating a Narrator

Point-of-view taxonomies are unhelpful. By no means do they encapsulate all possibilities for narrating a story.

Dale Stromberg

Writers often create taxonomies of point-of-view: close this, distant that, limited this, omniscient that. Such categories are unhelpful—by no means do they encapsulate all possibilities for narrating a story.

It might be more helpful if we think of creating a narrator. We often see narrative viewpoint presented in prepackaged, plug-and-play form; why not approach it as a series of choices instead?

Eight Questions for Creating a Narrator

1. Is the narrator part of the story? Or outside it?

If she exists within the story, it is fairly certain we will be using the grammatical first person, though the second person might also work. If she is outside it, the convention is to use the third person, though authors don’t always respect this: for example, Vonnegut’s narrator in Slaughterhouse-Five stands outside the action, but still comments on it in the first person on occasion.

Where the narrator must stand in relation to the story can dictate grammatical person. Rather than beginning arbitrarily with “I” or “he/she”, we can allow our narrative needs to tell us the person to use.

2. Can the narrator read thoughts? If so, whose?

The latter question is especially important: If we decide the narrator can read every character’s thoughts, it becomes very hard to keep secrets from the reader. This could work wonderfully for us; or, depending on the story, it could hobble us.

Allowing the narrator to read thoughts introduces a kind of obligation: If a character, whose thoughts can be read, knows something relevant to the tale, and your narrator neglects to mention this at an appropriate time, and furthermore does this for the sake of a ‘surprise, surprise’ plot twist later, you’re cheating. Readers tend to feel they’ve been played a lousy trick.

If we decide the narrator can read the thoughts of only some characters, we must define this limitation consciously and carefully—both so we can be sure we ourselves know the reason for it, and so we can make certain we’re following it.

3. Does the narrator know the future? Or things the characters don’t know?

Just as Gloria is setting out in the car, is the narrator permitted to tell the reader, “Little did Gloria know this drive would change her life forever”? Or, “Because the stereo blasted the instant Gloria started the engine, she didn’t hear the warning ding that the door was ajar”?

Although the practice of switching between characters’ consciousnesses is sometimes given as the criterion for determining whether or not third-person narration is ‘omni’, having the narrator point out what a character has failed to notice, or forgotten to do, or can only learn later, is equally omni—it goes beyond what the point-of-view character herself can know.

There is no logical inconsistency in a first-person narrator having such knowledge at a given point in the story, if it’s feasible they could have found out later, and if the story is being told retrospectively—say, by a narrator who, in old age, is recounting a youthful experience.

4. Does the narrator exist in space, or outside space?

Stories in film or television cannot be told without a camera, and the camera must exist at some point in space—a tremendous narrative limitation. A fact often forgotten by contemporary writers of fiction is that we do not require a camera.

Considering the countless childhood hours we spend as viewers, even before we learn to read, it is no wonder many of us presume that fiction must be written as though being viewed from some point in space. But the author always has the option of writing narration as if from nowhere in particular, and indeed one finds plenty of such writing from antiquity up to the end of the nineteenth century, if not later.

The more we conscientiously insist on locating every physical observation from an identifiable point—often conceived of as, yes, a camera mounted over a character’s shoulder—the more we will be writing in a ‘limited’ point of view. An important decision to make when creating a narrative voice is whether it must behave like a camera, located somewhere in every scene, or whether it may exist outside of space, free to describe things from an ‘anywhere’ that only nonvisual storytelling can make possible.

5. Will the narrator intrude?

It goes without saying that a first-person narrator can impose her own ideas on the story. But what about other forms of narrator?

If a third-person narrator tells us that Kevin was wearing a marvelously silly hat, he’s butting in to share his opinion with us. I don’t refer to a ‘limited’ narrator who follows a single character exclusively, in which case readers will logically conclude that the character finds Kevin’s hat silly, with the narration simply reporting this. I mean a narrator who stands farther away from things, but still makes free to, say, criticise fashion choices.

Dickens did this chronically—in Little Dorrit, for example, the narrator puts the plot on hold for quite a spell to gripe about ineffectual government. Not all narrators need be as chatty as Dickens’s; we must simply determine beforehand whether value judgements of any kind will be allowed into the narration. This can go beautifully if done advisedly, as in much humor writing, but it seriously weakens the story if it slips in without our knowing about it.

And there is another way that a narrator can slip in ideas that do not originate from any character: We can grant our narrator poetic license. To allow the description of a mud-caked horse, or a convenience store aisle, or what have you, to include wordplay, or irony, or madcap grandiloquence—these are enjoyable ways the narrator might intrude. (This overlaps quite a bit with the concept of ‘voice’.)

A notable example that combines these two forms of intrusion is the opening line of Pride and Prejudice: “It is a truth universally acknowledged, that a single man in possession of a good fortune, must be in want of a wife.” This is, of course, a disembodied opinion, posed not by a character in the story but by the story itself. Yet we also somehow know, though yet ignorant of the rest of the tale, that the narrator doesn’t really mean this – she is being ironic, and thus invites the reader to eye what follows with similar ironic distance.

6. Will the whole story be told by the same narrator?

It need not be. Brontë’s Wuthering Heights switches narrators periodically to allow the characters to tell Heathcliff’s tale piecemeal; and in Faulkner’s The Sound and the Fury, four narrators tell parts of one story, each in a radically different style. Something about your story may likewise demand more than one teller.

7. Is the narrator reliable?

The archetype of the unreliable narrator is one so self-absorbed as to fail to understand the import of his own story; Nabokov is notable for using this. The narrator tells the story one way, but the perceptive reader understands it another way.

Other forms of unreliability are possible. The narrator could lack knowledge or understanding that the reader nevertheless supplies: Imagine a story narrated by (or from the point of view of) a young girl whose parents are on the brink of divorce but conceal this from her. Double-meanings, subtle cues in gesture, and things unsaid will conspire to create dramatic irony.

8. If the teller is in the story, is she a central character, or some other observer?

We may decide to have a protagonist tell his own story, like Cather’s Jim Burden or Ellison’s unnamed invisible man. But telling the story through the eyes of a minor character can shed light on the protagonist in ways that asking her to tell her own story might not. If Sherlock Holmes were to narrate his own tales, his process of ratiocination would have to be revealed to the reader step by step; telling the story through Watson allows Doyle to keep secrets and then wallop the reader at the end.

The Only ‘Ought’ of This Post

Perhaps you will create your narrator before you begin; or perhaps, as I often do, you’ll discover your narrator as you write and then revise to make it consistent. Either way, consistency is key.

Imagine a man who has a poor grasp of the concept of musical key sitting down to improvise at a piano. As he tickles the ivories, he’s liable to play an off-key note from time to time. A listener, even if she has no musical training, will say to herself, “Hmm? That doesn’t sound right…”

No single note is, in and of itself, ‘out of key’. But enough notes played within one key will create expectations in the listener’s ear—and subsequent deviations can sound wrong.

Likewise, once we’ve written enough paragraphs in one narrative mode, the reader’s mind makes unconscious assumptions about the voice telling the story: perhaps without realizing, she begins to expect that this voice won’t change.

If an author, while mostly writing in one point-of-view, occasionally slips out of it, even readers who cannot pinpoint why they feel this way may find that something doesn’t seem right. (Sort of like when one musician in the band plays a bum note. Perhaps you can’t say who played it, but it sticks out.) This can reduce their trust in the story—and implicitly trusting the author is one key to enjoying fiction.

This is the only ‘ought’ of this article: point-of-view ought to be consistent.

Now, just as a composer may shift keys on purpose, an author can certainly shift points of view within a story—but it must be done advisedly; and even then, after the shift, consistency is crucial.

So, if the reader finds, on page 181 of a novel, that, for the first time in the story...
– the narration’s ‘camera’ is showing us the main character from a distance, rather than exclusively looking over her shoulder, or
– the narration is suddenly cracking wise rather than maintaining a detached and sober tone, or
– the narrator is capable of telling us what the villain is thinking of doing next, despite having kept mum on this till now, or
– the narrator, taking pity on any readers not up to speed, is giving a potted history of an industry or nation 
...then the reader has caught the author being sloppy with point-of-view. (Please again note that this is happening for the first time a fair way through the book.) The effect is not unlike taking your eighth or ninth sip of a given pint of beer and tasting maple syrup.

If we want our narrator to have the authority to provide potted histories, we must establish this early in the book, and then make it a not infrequent practice thereafter. Likewise the use of shifting ‘camera angles’, or an edgy and irreverent tone, or the liberal exercise of mind-reading, or what have you. Which means that, if we decide midway through a novel that we need to, say, allow access to characters’ thoughts to make the story work, we had better go back to the start and revise to have the narrator doing so throughout.

Even if we will mix and match the narrative approaches above, consistent (i.e. careful and intentional) mixing will give the best, or least jarring, results. Failing this, to slip willy-nilly from one narrative mode to another simply makes the writer look bad.

Member submitted content is © individual members.
Other material ©2003-2022