Why Do Literary People Hate Science?

Henry Oliver disparages excellent science without understanding it.

Long experience has taught me that literary people are, in the aggregate, hostile to scientific approaches to the study of literature. By literary people, I mean not only professors and university literature students, but also non-academic literature fans like Henry Oliver, author of the popular Substack, “The Common Reader.” As someone who completed a PhD in comparative literature in 2002, I have had much occasion to brood on this issue. I focused on the potential of evolutionary psychology to illuminate the nature and genesis of literary texts, and I found only a handful of scholars in my field who supported my work. I spent most of my time talking with psychologists.

One blatant expression of this hostility towards science is a 2013 rant called “Crimes Against the Humanities” by Leon Wieseltier, then the literary editor of The New Republic. Wieseltier inveighs against efforts to interpret literature from a scientific perspective. As the subhead reads, “Now science wants to invade the liberal arts. Don't let it happen.” However, usually the bias is more subtle. In a previous article on literature and critical thinking, I gave prominent literary scholar John Guillory as an example. Guillory argues that the study of literature inculcates critical thinking without referencing any of the readily available scientific evidence for this effect. I could give many other examples of the same curious omission.

As time passes, this enmity seems ever more perverse. A wide and rich array of psychological research has now established that reading literature makes us better social reasoners and critical thinkers. In my previous article, I surveyed evidence that literature readers are, among other things, more sensitive than non-literature readers to the complexities of reality and less prone to stereotyping and leaping to conclusions.

One would have thought that literary people would eagerly embrace these scientific findings and call loudly for more such research to be done, especially since reading literature is on the decline, and literature departments at universities are seeing a drop in enrollment and funding. And yet they do not.

My fascination with this perverse and self-sabotaging neglect led me to raise the issue in comments on Henry Oliver’s posts. I asked him his opinion about the matter multiple times and linked to my article on the effect of exposure to literature on critical thinking. Oliver’s response to me was always disparaging. He told me that the studies that I discussed were weak and unconvincing. When I asked why he thought that, he did not reply directly to me. (You can find one of our exchanges here.)

However, Oliver replied to me indirectly in his article “Can Fiction Improve You?” Here he discusses psychological studies on how exposure to literature affects theory of mind and critical thinking. In his eagerness to diminish these excellent studies, Oliver thoroughly misrepresents their import and findings and along the way exposes his ignorance of statistics and scientific research methods. It is obvious Oliver read these studies with a closed mind and tried to find excuses to dismiss them rather than to understand and appreciate them.

I should make clear that I am not accusing Oliver and other literary people of hating science as a whole. They are usually at least willing to pay lip service to the value of science. What Oliver and many other literary people hate specifically is the application of scientific methods to the study of literature.

Literature and Theory of Mind

In my last essay, I discussed the evolution of imagination, or the ability to construct mental models of the external world, of our interior state, and of the minds of other beings. This last type of imagination is called “theory of mind” and it refers to the cognitive ability to assess others’ emotions, knowledge, and intentions.

There is now significant support for the theory that exposure to literature augments theory of mind ability. Since challenging literary texts present us with confusing worlds, ambiguous characters, and dense language, it is natural that such texts would strengthen our interpretative skills. As psychologists Daniel Comer Kidd and Emanuele Castano explain:

We submit that fiction affects [theory of mind] processes because it forces us to engage in mind-reading and character construction. Not any kind of fiction achieves that, though. Our proposal is that it is literary fiction that forces the reader to engage in [theory of mind] processes…. [Through] the systematic use of phonological, grammatical, and semantic stylistic devices, literary fiction defamiliarizes its readers. The capacity of literary fiction to unsettle readers’ expectations and challenge their thinking is also reflected in Roland Barthes’s distinction between writerly and readerly texts. Although readerly texts—such as most popular genre fiction—are intended to entertain their mostly passive readers, writerly—or literary—texts engage their readers creatively as writers. Similarly, Mikhail Bakhtin defined literary fiction as polyphonic and proposed that readers of literary fiction must contribute their own to a cacophony of voices. The absence of a single authorial perspective prompts readers to enter a vibrant discourse with the author and her characters.

Oliver reviews four studies that tested the theory that reading literature augments theory of mind. In these studies, subjects read literary and non-literary texts and then took tests that measured theory of mind. The tests gauged subjects’ understanding of what other people know and believe, as well as their ability to infer the emotional state of other people from their facial expressions.

Kidd and Castano gave subjects short texts of literary fiction, science fiction, and non-fiction and found that literary texts augmented theory of mind ability significantly more than the other texts. Maria Chiara Pino and Monica Mazza gave subjects whole books of literary fiction, science fiction, and non-fiction to read and again found a significant effect of literary fiction on theory of mind. A meta-analysis by David Dodell-Feder and Diana I. Tamir, which aggregated the findings of 14 literature and theory of mind studies, also confirmed the effect.¹

The natural conclusion is that the theory is stunningly successful and has been replicated many times. What is particularly impressive is that these studies succeeded despite the difficulties of measuring the effect of our reading habits on the way that we think. There is no experimental design that can measure how reading shapes our psychology over the course of a lifetime, which is the question of true interest. However, even the relatively crude technique of measuring the effect of short-term exposure to different types of texts finds that literary texts are significantly more effective at augmenting theory of mind ability than other kinds. These findings suggest that literature has a very potent effect on the way that we think.

Does Oliver acknowledge and celebrate this stunning success? Of course not! Rather, he searches for reasons to diminish the studies and, in doing so, thoroughly misrepresents their findings and import. He notes that the meta-analysis found that exposure to literature produced only a small improvement in theory of mind. From this he concludes that the effect must be insignificant. But a small effect is not necessarily an insignificant one. As the authors say:

The small size of the effect raises the question of how meaningful it is. We argue that this effect has the potential to be very meaningful. The magnitude of an effect does not determine its practical impact (Cooper, 2008). Social cognitive skills have been shown to positively impact social connection across the lifespan (e.g., Goldstein, Vezich, & Shapiro, 2014; Slaughter, Imuta, Peterson, & Henry, 2015), particularly in clinical populations (e.g., Fett et al., 2011). Strong social connections can significantly improve well-being (Reis, Sheldon, Gable, Roscoe, & Ryan, 2000; Ryan & Deci, 2000), stave off physical illness (Yang et al., 2016), and enhance longevity (Holt-Lunstad, Smith, & Layton, 2010).1 Thus, any method that enhances social cognitive skills in the general population or in individuals with social cognitive deficits is worthwhile and deserving of additional research – especially when this method is cost-effective, easily disseminated, and well-tolerated.

The authors also discuss the difficulties of measuring how reading habits affect our psychology and note that the best designed study, that of Pino and Mazza, showed a stronger effect than the others.

It is also important to consider the possibility that fiction may have an even larger impact with more immersive, longitudinal reading experiences. Almost all of the studies in this meta-analysis required participants to read only one short fiction story. Longer periods of reading may yield larger or longer-lasting effects; indeed, the one study included here that required participants to read an entire book yielded some of the largest effect sizes.

Oliver claims that the effect sizes found in the Pino and Mazza study were “very small.” In fact, the study found an eta squared (η2) of 0.16 for one theory of mind test and 0.17 for another (p. 10). These are large effect sizes.

This is not the only place where Oliver reveals his ignorance of scientific method. For example, he says:

It is likely that the benefits that are showing up in these studies are just the result of other factors, like random effects due to the variation between readings, or the people with better Theory of Mind being more drawn to fiction.

The whole purpose of statistical significance tests is to show that experimental findings are not likely to be the result of random variation. Elsewhere, Oliver complains that some of the studies had small sample sizes. Apparently, he is unaware that the statistical significance tests used in these studies take sample size into account, so that the smaller the sample, the stronger the effect must be to reach significance.

Oliver cites another study that found no effect on theory of mind from reading short literary texts. There are sometimes negative results in science, and one such study does not by itself disprove a hypothesis. Also, the authors of this study leave open the possibility that more prolonged exposure to literature might affect theory of mind. Even this study found one positive result: scores on an author recognition test that was designed to measure lifetime exposure to literature predicted theory of mind ability.

Oliver states that the studies do not establish causality. On his view, all that they prove is that there is a correlation between the strength of people’s theory of mind ability and their exposure to literature. It is true that correlational studies are ineffective at establishing causation, but none of the studies that Oliver cites are, in fact, correlational. Rather these are experimental studies, although the authors do sometimes add in correlational information. The entire purpose of the experimental method is to establish a causal relation between one variable and another. For example, Kidd and Castano presented some subjects with non-fiction or popular fiction texts and others with literary fiction texts, and then compared the performance of the two groups on theory of mind tasks. Because the literary fiction group scored higher, the natural conclusion is that reading literature caused the high score.

Oliver complains that the results of the studies might be due to extraneous factors unrelated to the relationship between exposure to literature and theory of mind. He fails to note that, in all the studies, the scientists gathered information about potentially confounding factors, like education level, age, gender, and so forth and tested whether they were responsible for the results. Oliver makes no substantive critique of the methods employed in these studies.

In a footnote, Oliver also addresses the topic of our previous discussion: the relationship between exposure to literature and critical thinking. He does not link to my article on the subject. Perhaps Oliver feels that his readership needs to be protected from skeptical characters like me! He writes:

As for the old canard that literature improves critical thinking, if you ask Consensus about that you get a load of studies with either zero citations, or with one or two. Why this even need [sic] studying is beyond me. If you want to learn critical thinking, study logical subjects like philosophy and economics. Not poetry.

Oliver links to these results from Consensus, an academic search engine that summarizes findings across studies. It’s hard to imagine a lazier way of researching a topic. The Consensus results show quite massive support for the theory that exposure to literature enhances critical thinking: 91% of studies find a positive relationship. It is true that many of the articles referenced appear in low-quality journals with weak review standards. The quality of the journals does not mean that the results reported by these articles are invalid, however. If Oliver had bothered to read the studies that I discussed in my article, he would have seen that a number of high-quality peer-reviewed studies of the topic exist.

Oliver refers to the theory that reading literature increases critical thinking ability as an “old canard,” which implies that there is no current interest in the subject. In fact, the theory is central to many recent defenses of the value of studying of literature, including those of highly respected thinkers like John Guillory and Martha Nussbaum.

What’s so weird about Oliver’s reaction is that these studies would seem to support his intellectual agenda. His pinned article is a passionate defense of the value of reading classic literature. Surely citing high-quality scientific work on the positive psychological effects of reading literature would bolster his case.

Why do literary people hate science?

I am not arguing that all discussion of literature ought to be based in science, merely that literary people should recognize when scientific research validly addresses questions relevant to literary studies. A number of thinkers, including Leon Wieseltier, Roger Scruton, Raymond Tallis, and William Deresciewicz, have been critical of scientific approaches to the study of literature, which, they argue, do not result in meaningful insights into the nature and value of literary works. While such criticisms are sometimes unfair, it is true that applications of neuroscience, evolutionary psychology, and other scientific disciplines to literary and other art works have often proved unilluminating. However, literary people should learn how to separate the wheat from the chaff.

So why are literary people so profoundly hostile to scientific work on literature? Perhaps they have encountered bad research in this genre and have developed a prejudice against it. Another reason for the hostility is simple lack of education. If you don’t understand how statistical significance tests work and can’t tell the difference between a correlational and an experimental research design, it will be hard for you to assess the meaning and validity scientific research.

However, I think the major reason for the hostility is deep and philosophical. I think that many people want to believe that they have souls that are non-physical, non-deterministic, and not amenable to scientific explanation. They feel that literature and other artistic works put them in touch with this soul and are therefore off limits to science. These people, therefore, regard scientific approaches to the study of the arts as a kind of blasphemous denial of the reality of their souls. I hope address this philosophical hostility to the scientific study of literature soon.

1. Dodell-Feder, D., & Tamir, D. I. (2018). Fiction reading has a small positive impact on social cognition: A meta-analysis. Journal of Experimental Psychology: General, 147(11), 1713–1727. https://doi.org/10.1037/xge0000395