Pretty interesting, but it feels like the vague category 'literary fiction' is being used to make the results seem more surprising than they really are. You cou

Pretty interesting, but it feels like the vague category 'literary fiction' is being used to make the results seem more surprising than they really are. You could also say, "fiction which requires readers to model the internal mental states of characters improves empathy"—and the results seem pretty straightforward: modeling mental states of others improves ability to model mental states of others. I mean, it is pretty interesting that it works with fictional characters—but still: nothing I read in the article lead me to believe this is due to some elusive quality of the intentionally undefined category 'literary.' If anything, these results give a little more info. on how we actually do define 'literary': that a typical requirement for a work to be called 'literary' is that it pushes readers to model mental states of characters.