What do the arts mean to our lives? To at least some researchers, they're a way that we learn how the people around us think. Previous studies have concluded that reading fiction is correlated with various measures of empathy — as you learn how characters interact, you can transfer that to the real world. But for David Kidd and Emanuele Castano of the New School for Social Research, some types of fiction may be better at this than others. The results of their experiments, published today in Science, suggest that reading "literary" fiction, as opposed to its more mainstream or pulpy counterparts, could especially prime people to understand others' thoughts and emotions.

The authors' theory isn't that literary fiction is more artistically important or of better quality. It's that the writing we commonly describe as literary is designed to make readers actively interpret how characters think and feel, mirroring a psychological concept called theory of mind. If reading any fiction flexes our empathetic muscles, Kidd and Castano believe reading more ambiguous or difficult literary work could be the equivalent of lifting heavier weights.

"What's the value of having students read about a 19th-century Russian axe murderer?"

Testing social concepts with empirical research is hard, but this notion is particularly fraught. How do you even define literary fiction, which can be considered everything from a genre to a way of writing to a marker of prestige? And why draw the distinction at all? At least in part, it's to provide reference material in the ongoing war over what schools teach. Existing tests measure fiction reading in comparison to non-fiction or no reading at all, but Kidd and Castano wanted to specifically test the sort of books that wouldn't necessarily be picked up for fun. "There have been doubts about the value of fiction. What's the value of having high school students read about a 19th-century Russian axe murderer?" asks Kidd, referencing Crime and Punishment. "Why have them read that rather than, say, Malcolm Gladwell?"

With this paper, Kidd and Castano are sparring with "practical education" proponents on their own turf. "We shouldn't dismiss the arts without having a good sense of what they actually contribute to our lives," says Kidd. "We believe that we should have an evidence-based debate." To define literariness, they turned to theories about fiction that describe relatively straightforward delivery of plot, character, and situation as "readerly" and work that actively requires filling in gaps and inferring meaning as "writerly." Literary fiction, by that token, is writerly — to them, it's fiction that forces readers to grapple with the text and characters. More practically, literary fiction is what people say is literary fiction: if it wins a literary award or is described by a large sample of participants as literary, it probably counts.

A story by Alice Munro could go up against one by Robert Heinlein

From there, Kidd and Castano needed to come up with a way to pick examples of both kinds of fiction and to measure how readers responded. Testing how well people can interpret others' feelings and thoughts isn't a clear-cut business, but there's a framework for it, says Kidd. On one side of the equation, there are tests meant to measure something like emotional empathy: one gives subjects a face with only eyes showing, then asks them to pick the face's emotion from a list. Measuring how well people can interpret thought processes, however, proved harder. One method they tried was meant for young children — "all of our adults basically aced the test," says Kidd. They later used a more sophisticated system called the Yoni task, in which participants must guess what a cartoon face is thinking of or referencing.

Over five tests, groups of between 70 and 350 participants were asked to read samples of either literary fiction, genre or popular fiction, or nonfiction. The "literary" category was drawn from the PEN / O. Henry fiction awards, while the genre fiction was drawn from bestseller lists and anthologies. In some experiments, that could pit something like an award-winning Alice Munro story against one by Robert Heinlein or successful romance writer Rosamunde Pilcher. Since Kidd and Castano found that one of the easiest ways to find "literary" fiction was to look at examples picked by experts like the awards panel, they wanted correspondingly lauded genre examples. "We didn't want to just compare something that was well-written to fiction that was not well-written," says Kidd.

"There are so many good reasons to read, and this is just one of them."

After reading the sample, participants took the aforementioned theory of mind tests. Some of the results were inconclusive, especially those that involved interpreting thought processes. But in other tests, particularly those where people interpreted emotions, participants with "literary" samples did significantly better than those with genre samples. To Kidd and Castano, this suggests that we could see a difference over time as well. If previous work has shown a correlation between reading fiction and interpreting behavior, and their priming experiments show that there could be both a causal relationship and a difference between styles, there's reason to believe that specifically reading literary fiction can have long-term effects.

This, though, is just a preliminary study, working with a concept that will always remain amorphous. Kidd and Castano admit that they've just scratched the surface of figuring out what could actually be causing the difference in results — "literariness" covers a broad range of techniques, and the results were mixed enough that it's still hard to pinpoint exactly what reading makes you better at. Like other sociological studies, it's also difficult to say how much this would affect day-to-day life. Even if literary readers are, on average, slightly better at reading others, the variation between individuals would mean it might be hard to see the difference. And no study is going to satisfactorily settle the debate over what counts as literary, or what schools should be teaching.

No matter what the results, Kidd doesn't want to turn reading into another kind of exercise or professional training. The study's implications "are not in any way that people should not read popular or genre fiction, nor do we mean to suggest that the only value of literary fiction is to improve theory of mind," he says — though his study may end up exacerbating the existing tension between some genre and literary writers and readers. "There are so many good reasons to read, and this is just one of them."