Six practical questions organise the theory this chapter uses, and each strand below opens with the question it answers before naming its source.
Whose head am I in, and is that the same as who is talking?
Craft advice about “point of view” often fails because it runs two questions together. Gérard Genette separated them in Narrative Discourse (1972; English translation by Jane E. Lewin, 1980): who speaks? is the question of voice, the narrating agent, and who perceives? is the question of focalisation, the position from which the narrative’s information is regulated. Genette distinguished zero focalisation, in which the narration knows more than any character; internal focalisation, in which it is restricted to what one character knows; and external focalisation, in which it reports behaviour without interiority, like a camera. He then divided internal focalisation by how it behaves across a whole text. Fixed focalisation keeps one focaliser throughout, as The Remains of the Day does. Variable focalisation changes focaliser one at a time, as most multi-viewpoint novels do, and buys coverage: each vantage sees different material and the reader assembles a wider world. Multiple focalisation tells the same events through different characters in turn, as The Sound and the Fury does, and buys doubt: the reader assembles a contradiction. The difference between those two last kinds is what separates the ordinary multi-viewpoint novel from the proposed sixth domain in The Taxonomy.
Mieke Bal (Narratology, 1985) added that focalisation is a relation with two ends, a focaliser and what is focalised. To say a chapter is Nadia’s is half a description; the other half is what Nadia is allowed to look at, and a viewpoint character pointed only at plot information is a fault that Bal’s formulation makes visible.
Why do some viewpoint characters feel inhabited and others only reported?
Dorrit Cohn’s Transparent Minds (1978) names three ways third-person narration renders a mind. Psycho-narration reports a consciousness in the narrator’s language: She had never trusted her brother’s silences. Narrated monologue, the mode usually called free indirect style, gives the character’s own idiom in the third person and the narrative tense: Her brother’s silences. What was she supposed to do with those? Quoted monologue gives the thought in the first person: I have no idea what he is not telling me. The modes cost different amounts of the reader’s attention. Psycho-narration reports a mind without asking the reader to inhabit it, and a book can afford it for a dozen characters; narrated monologue asks the reader to adopt a character’s judgements, and is the mode in which a viewpoint is truly held. A draft that “has five viewpoint characters” may have one inhabited mind and four reported ones, and the diagnosis is invisible until the modes are named.
What happens to the people who get no viewpoint?
Alex Woloch’s The One vs. the Many (2003) supplies the rigour behind the metaphor this chapter calls the empathy budget: reader identification treated as a finite pool divided among the viewpoint characters. A novel implies many people and has room to attend to few. Woloch calls the delimited position each character occupies in the narrative its character-space, and the arrangement of all those spaces, in which the attention given to one is a function of the attention withheld from others, the character-system. His most useful claim for a writer is that minor characters are not small but squeezed, and that the squeeze shows: the eccentric with one repeated gesture, the type, the walk-on defined by a job. This edition calls that effect character compression, to keep it apart from the compression of page-space against story-weight that Scene and Sequel (Chapter 2) measures. A repeated tag is what a person becomes when a book needs her to be recognisable and has no room to make her knowable. Woloch’s further point is that the distribution argues something: deciding whose interior the reader will inhabit is deciding whose experience the book treats as fully human. His evidence is the nineteenth-century realist novel, with its large cast and single protagonist, and the chapter uses the model as an instrument for noticing distribution, not as a theory of value.
Why does each new viewpoint cost more than the last?
Rolf Zwaan and Gabriel Radvansky’s review of research on situation models (1998), the reader’s running model of the state of affairs a text describes, holds that readers track protagonist, space, time, causation and intention, and pay a measurable processing cost when any of them changes. A change of viewpoint usually changes the protagonist and drags space, time and intention with it, which is why it is among the more expensive moves a book can make; that it is the most expensive of the five is this edition’s inference, not their finding. Catherine Emmott’s work on contextual frames (1997) supplies the remedy: once a book has declared a pattern, the reader recalls a frame rather than building one, so the twelfth alternation costs less than the first. Lisa Zunshine (Why We Read Fiction, 2006) adds the idea of levels of intentionality, I believe that she wants him to think, and reports from earlier studies that readers manage about four comfortably. A grant turns a character from someone the reader guesses at into someone whose view of every other character must be held as well, so the load plausibly grows faster than the cast; that last step is conjecture, offered because it matches what dilution feels like.
Will going inside a character make readers sympathise?
Wayne Booth (The Rhetoric of Fiction, 1961) observed that sustained inside views tend to align a reader with a mind more or less regardless of what it contains, so that a writer who grants a viewpoint has already argued for the character. Suzanne Keen (Empathy and the Novel, 2007) supplies the limit: narrative empathy, feeling with characters while reading, is real and measurable, but the evidence that it changes behaviour afterwards is weak. The best-publicised experiment, Kidd and Castano’s 2013 report that reading literary fiction improved a theory-of-mind task, failed in a large replication by Panero and colleagues (2016). What survives is the narrower claim this chapter needs: inside views produce alignment within the book.
Is head-hopping wrong?
For most of the novel’s history in English, French and Russian, the default was a mobile narrator who moved between minds when the material called for it. Henry James’s prefaces (collected as The Art of the Novel, 1934) argued for the reflector, a single registering intelligence, as a gain in intensity bought by a loss of scope; Percy Lubbock’s The Craft of Fiction (1921) turned that preference into doctrine; and E. M. Forster’s Aspects of the Novel (1927) dissented at once, holding that a novelist may shift vantage if he can make the reader accept it. The prohibition then hardened in mid-century Anglo-American teaching and editing. The convention is real, and a reader trained by recent fiction does feel an unsignalled change as a wobble, so it has a cost to be paid. The cautions in this chapter describe that convention, not narrative as such.
See also: the Reference theory, with Stanzel’s typological circle, Uspensky’s four planes, Palmer’s fictional minds and the historical table.