University homepage | Suomeksi | På svenska | In English

Book Review – Lisa Zunshine’s Why We Read Fiction: Theory of Mind and the Novel – Reviewed by Merja Polvinen

The Electronic Journal of the Department of English
at the University of Helsinki

ISSN 1457-9960

Volume 5, 2009

Emotions Issue

© 2009 Merja Polvinen



Why We Read Fiction: Theory of Mind and the Novel 

by Lisa Zunshine

Columbus: Ohio State University Press, 2006. 198 pages


Merja Polvinen

Department of English, University of Helsinki



Lisa Zunshine’s volume begins with a question: Why did Peter Walsh, in Virginia Woolf’s novel Mrs. Dalloway, tremble when meeting Clarissa? Or more precisely: “[W]hat allows Woolf to assume that we will automatically read a character’s body language as indicative of his thoughts and feelings?” (3). The answer Zunshine offers in Why We Read Fiction: Theory of Mind and the Novel is that making inferences regarding other people’s mental states is so natural to human beings that we do it automatically, whether those people are fictional or real, actually present or just being spoken about.

The title of this book, however, offers an even more extensive claim. Not only is the ubiquitous “mind reading” a way of explaining how readers postulate the emotions and motivations of fictional characters, but Zunshine also suggests that such an activity is the basis for the very existence of the novel as we know it. The reason why we read fiction, she contends, is because it exercises our mind-reading ability:

As a sustained representation of numerous interacting minds, the novel feeds the powerful, representation-hungry complex of cognitive adaptations whose very condition of being is a constant stimulation delivered either by direct interactions with other people or by imaginary approximation of such interactions (10).

Zunshine thus approaches fiction not through speech-act theory or questions of pretended beliefs, nor through mimesis and represented worlds, but through the stimulation and exercise of human sociability.

In order to make her argument, Zunshine concentrates on two theoretical concepts: the Theory of Mind (or ToM) and metarepresentation. The first of these, an idea much discussed in cognitive psychology, refers to the ability to explain behaviour by assuming that it is motivated by a mind resembling our own, a mind which experiences emotion, holds beliefs, and reasons in roughly similar ways as we ourselves would. This capacity, dependent both on a predisposition we are born with and on our learned interaction with other human beings, is for most people so automatic that it feels absurd to try and explain how I know that the man-shaped thing turning its head left and right on the other side of the street is a person planning to cross without being hit by cars. By relying on their Theory of Mind, Zunshine argues, readers construct similar inferences of fictional characters. Thus from the brief description given by Woolf in her novel, readers make the assumption that Peter Walsh is a fully conscious human being whose trembling is caused by an emotional reaction to meeting a woman he has loved.

Metarepresentation, which Zunshine derives from the work of evolutionary psychologists Leda Cosmides and John Tooby (4-5), refers to readers’ ability to keep track of the layers of information in a novel. Perhaps the major pleasure in reading Jane Austen is in exercising that ability and being aware of the changes in trustworthiness of the representations provided in the novels. Think of the difference in Pride and Prejudice between the representations given through the prism of an Elizabeth Bennet who trusts Mr. Wickham, and the one who trusts Mr. Darcy. Her first account of events assumes a proud, unfair and selfish Darcy, a view which is reinforced by Wickham’s story of how cruelly Darcy treated him. But as soon as Wickham’s true nature is revealed and his story receives the stamp of untrustworthiness, it also helps overturn Elizabeth’s view of Darcy. What views readers have of both men is obviously influenced by the text’s layers of irony as well as prior knowledge of the plot, and their experience of Austen’s narrative depends on which opinions and which perspectives (including those of Elizabeth and the narrator) they trust and which they take “under advisement” (61). Thus metarepresentation, as Zunshine explains it, consists of judgements made on the basis of the Theory of Mind:

Broadly speaking, whereas our Theory of Mind makes it possible for us to invest literary characters with a potential for a broad array of thoughts, desires, intentions, and feelings and then to look for textual cues that allow us to figure out their states of mind and thus predict their behaviour, our metarepresentational ability allows us to discriminate among the streams of information coming at us via all this mind-reading. It allows us to assign differently weighed truth-values to representations originating from different sources (that is, characters, including the narrator) under specific circumstances. (60)

Zunshine adapts the research into metarepresentation to make brief forays into the implications this narrative “source-monitoring” (5) might have for unreliable narration (77-79), the implied author (79-82) and the difference between fiction and history (65-73). But in terms of the general focus of this issue of Helsinki English Studies, perhaps the most intriguing aspect of Zunshine’s work is that among the different aspects of cognition engaged by fiction, emotion comes across as a central element. Many of her examples of the use of ToM come from moments in fiction when readers engage specifically with a character’s emotions, and metarepresentational judgements are shown to be crucial in readers’ own emotional reactions towards various fictional situations.

Zunshine declares her book to be primarily aimed at students and the general audience, and as a result a scholar may find herself wishing for more in-depth discussion of the pivotal theoretical questions, but the meat and potatoes of Why We Read Fiction can be found in the three main sections of literary analysis (II:10 and 11, III: 3) of Richardson’s Clarissa, Nabokov’s Lolita and the conventions of detective stories, respectively. I particularly appreciated Zunshine’s ability to shed light on the conflicting readings of Nabokov’s narrator, which see him either as a manipulative paedophile or a modern and slightly twisted version of a tragic lover.1 By pointing out how Nabokov overloads his readers’ source-monitoring capability, Zunshine is able to explain why the novel draws readers into adopting Humbert’s twisted perspective on his relationship to Lolita. That perspectival alignment could account for the sympathy apparently felt by so many critics for a narrator whose actions, if seen from any other point of view, would be abhorred. Rather than presenting the entire story explicitly from Humbert’s perspective (which, Zunshine notes, would have made readers easily wary of the information he provides), Nabokov’s narrative structure seems to give several external view-points to the events, view-points that support Humbert’s presentation of Lolita and thus justify both his desire for her and his actions in fulfilling that desire. Through various sleights-of-hand, Humbert erases himself as the source of these representations, and as a result readers lose track of the fact that the seemingly universal view of Lolita as a lustful nymphet is only created by Humbert’s descriptions of what he imagines the other men around Lolita to see and feel.

A similar universalisation occurs, Zunshine suggests, in Clarissa, where Lovelace presents the thoughts and behaviours of the people around him through the distorting prism of his own plans to rape Clarissa. Lovelace wants Clarissa to see him as a rake reformed by the love of a good woman, but that fantasy sometimes fools even Lovelace himself into forgetting that it is his fantasy, and that Clarissa’s behaviour may not be motivated by such a vision at all. Lovelace, like Humbert, is an unreliable narrator, particularly because he “lose[s] track of himself as the source of his representations of the other person’s mental world” (89), making it very difficult for readers to tell the difference between his real insights into Clarissa’s behaviour and his wishful thinking. As a consequence Richardson’s novel pushes the source-monitoring capabilities of its readers to their limits.

While these first two examples deal with stories that play with readers’ inclination to trust a narrator’s point of view, detective stories, according to Zunshine, “want us to disbelieve, from the very beginning and for as long as possible, the words of pretty much every personage we encounter” (124). These stories give readers an opportunity to enter states of maximum metarepresentational doubt and still, at the end of the narrative, provide a solution where truths and lies are neatly separated. Detective fictions thus exercise readers’ ability to deal with multiple levels of metarepresentation – a skill that, for Zunshine, forms the cornerstone of the success of not only detective fiction but also the novel as a literary form. Where Theory of Mind is a set of cognitive skills for understanding our social world, the novel, by engaging those skills in sorting out metarepresentations, is a focused and distilled version of that same cognitive activity. This emphasis on narrative fiction as an offshoot of human social skills also leads Zunshine, first implicitly and towards the very end of the book also explicitly, to regard cognition and emotion as inextricable, rather than forming some sort of opposition between rationality and irrationality (163-164). In this, again, she is representative of larger trends in cognitive studies, where embodiment and emotion are drawing more and more attention at the expense of views of cognition that are based mainly on formal logic.2

Objections to Zunshine’s arguments could be raised, firstly, because of the very wide scope of her general claims, and, secondly, because of the emphasis she puts on fictional characters as the part of fiction that justifies those claims. Certainly the question of fictionality in general, not just narrative fiction, is one that cannot properly be delineated by focusing on fictional characters alone. But Zunshine is careful to hedge her suggestion that ToM explains the existence of all fiction with statements which remind her readers that such an argument is only the starting point for the development of further analytical tools (e.g. 36, 163).  Furthermore, her use of narratology keeps the focus on characters as textual constructs (rather than real people), and she is fairly clear about when she is extrapolating on her own reading experiences instead of presenting them as universal reactions.

If experientiality is the central feature of narrative, as Monika Fludernik has suggested,3 human cognition, emotion and narrative forms must be seen as an inextricable whole. However, they do not form an unchanging monolith, and thus it is vital to pay attention to the development of narrative forms, arguably the quickest of the three to show change. Thus, while many cognitive theories postulate an unchanging universal human reaction, Zunshine should also be applauded for emphasising a diachronic perspective and calling for further research into the different expectations and reactions readers have had to the developing forms of narrative fiction. Even during the relatively short history of the novel form, readers’ ability to enjoy the complex interplay of represented minds has always depended on how used they are to various narrative conventions, as well as on the kinds of emotions they are willing to attribute to fictional characters. This volume presents a clear and readable introduction to the ways in which the combination of cognitive-affective studies and narratology can approach such developments.



1 Zunshine (101-102) refers to an overview of critical reactions in Brian Boyd. Vladimir Nabokov: The American Years. Princeton: Princeton University Press, 1991, and to Peter Rabinowitz. “Lolita: Solipsized or Sodomized?; or, Against Abstraction – in General.” A Companion to Rhetoric and Rhetorical Criticism. Eds. Walter Jost and Wendy Olmsted. Molden, MA: Blackwell, 2004. 352-339.

2 See e.g. Evan Thompson, Mind in Life: Biology, Phenomenology and the Sciences of Mind, Cambridge, MA: Harvard University Press, 2007.

3 Toward a ‘Natural’ Narratology. London and New York: Routledge, 1996.