The Empathy Question: What Reading Fiction Does and Does Not Do

The claim is appealing in its tidiness: reading literary fiction makes you more empathetic. It shows up in op-eds, in arguments for humanities education, and in the self-justifications of devoted readers everywhere. The research behind it is real but far more qualified than its popular presentations usually suggest — and understanding where the evidence actually points is more interesting than the simplified version. The study that launched a thousand headlines came from a team at the New School for Social Research, published in Science in 2013. David Comer Kidd and Emanuele Castano had participants read short literary fiction, popular genre fiction, or nonfiction, then tested them on the Reading the Mind in the Eyes Test — a measure of the ability to infer emotional states from photographs of people's eyes. Literary fiction readers scored higher than the other groups. The effect was modest but statistically significant, and it was interpreted as evidence that literary fiction, with its complex and ambiguous characters, trains the cognitive skills associated with understanding other minds.

The Replication Problem

What happened next is a case study in how science actually works. Multiple subsequent attempts to replicate the original finding produced inconsistent results. Some studies found the effect, others did not, and some found it under specific conditions but not others. A meta-analysis examining the accumulated evidence found that the relationship between fiction reading and empathy was real but smaller than the original study suggested, and was complicated by the fact that people who are already higher in empathy and theory-of-mind ability tend to be more drawn to literary fiction in the first place. Separating cause from selection effect is genuinely difficult. A research group at the University of Toronto, including Raymond Mar, has conducted some of the most careful longitudinal work on this question. Their studies suggest that habitual fiction reading is associated with better social understanding, but that the direction of causality is hard to establish from correlational data. People may read more fiction because they are socially curious, and social curiosity may independently predict both reading habits and empathy scores.

What Might Be Happening Anyway

Even with the replication complications, there are good theoretical reasons to think that sustained engagement with literary fiction does something useful for social cognition, even if the effect size is more modest than advocates claim. Literary fiction, unlike most genre fiction, tends to create characters whose motivations are opaque, ambiguous, and not fully resolved. Reading such characters requires active inferential work — the reader has to construct an interior life for the character from partial evidence. This is structurally similar to what we do in real social situations, where other people's inner states are never fully transparent. The practice-makes-better logic would suggest that doing this kind of inferential work repeatedly, across many characters and situations, should strengthen the relevant cognitive muscles. Whether that strengthening transfers to real-world social situations in a measurable way is the open question. The evidence for transfer is weaker than the evidence for the activity itself. There is also an affective dimension that the cognitive tests may not fully capture. Stories allow readers to experience emotional situations from the inside of a perspective not their own — to know something about grief from the position of a character experiencing it, without having suffered that particular loss. Whether this experiential simulation counts as genuine empathy development or something more limited is partly a definitional question. What the research does not support is the strongest version of the claim: that picking up literary novels is a reliable empathy intervention with consistent, durable effects. What it suggests instead is something more modest and more honest: that reading widely and with genuine attention is probably good for your understanding of other people, in ways that are real but harder to measure cleanly than a TED Talk would have you believe.