“Poetry makes nothing happen”. It’s the most often quoted line of W.H. Auden’s famous elegy, “In Memory of W.B. Yeats” – it could even be the most quoted line of his career. People draw on it when they want to denigrate poetry: if one of last century’s great poets thinks poetry is more impotent than important, why should they have to read it?

But these readers tend to forget (or choose to ignore) what comes next: poetry survives, Auden asserts, “in the valley of its making”. It is “a way of happening”, he continues, “a mouth”. Auden was a realist and knew that poetry couldn’t stop the approaching machinery of war – the elegy was written in 1939 – nonetheless he upholds the human need to commune with other humans.

But might literature – novels, plays and, yes, even poetry – be more than a mouthpiece?

Literary aficionados and librarians have long argued the edifying effects of the literary arts, but until now they have been noticeably short on evidence. A recent study at Ohio State University, however, has confirmed that literature does in fact “make things happen”.

In the right situations, the researchers found, reading fiction can lead to measurable changes, if only temporary, in the lives of readers. In jargon that Auden no doubt would have choked on, the researchers coined the term “experience-taking” to describe the phenomenon in which readers feel a character’s emotions, thoughts and beliefs as their own.

“Experience-taking can be a powerful way to change our behaviour and thoughts in meaningful and beneficial ways,” said Lisa Libby, co-author of the study and assistant professor of psychology at Ohio State University.

In one experiment, 70 heterosexual male college students read a story about a day in the life of another student. There were three versions: in one the protagonist was revealed to be gay early in the narrative; in another the protagonist was identified as gay late in the narrative; and in a third the protagonist was identified as heterosexual.

Results showed that students who read the narrative in which the protagonist was identified as gay late in the story reported higher levels of experience-taking than those who read the narrative in which the protagonist’s homosexuality was announced early.

“If participants knew early on that the character was not like them – that he was gay – that prevented them from really experience-taking,” Libby said. “But if they learned late about the character’s homosexuality, they were just as likely to lose themselves in the character as were the people who read about a heterosexual student."

Perhaps more importantly, the version of the story participants read affected how they thought about gays: those who read the gay-late narrative reported significantly more favourable attitudes toward homosexuals after reading the story than did readers of both the gay-early narrative and the heterosexual narrative.

Significantly, those who read the gay-late narrative also relied less on stereotypes of homosexuals – they rated the gay character as less feminine and less emotional than did the readers of the gay-early story.

Similar results were found when white students read about a black student who was identified as black early or late in the narrative.

Experience-taking is different from perspective-taking, where people try to empathise with another person’s experience – but without losing sight of their own identity. “Experience-taking is much more immersive", Libby explains, “you’ve replaced yourself with the other”.

Interestingly, experience-taking only occurs when people are able to “forget” themselves – their self-concept and self-identity – while reading. In a fascinating experiment researchers found that most college students were unable to undergo experience-taking if they were reading in a cubicle that contained a mirror.

When people do undergo experience-taking, however, it can affect their behaviour for days afterwards.

In an experiment which took place several days before the last US presidential election, 82 undergraduates (who were registered and eligible to vote) read one of four versions of a short story about a student who overcomes a series of obstacles (car problems, rain, long lines) on Election Day before arriving at the booth to cast a vote.

After reading the story, the participants completed a questionnaire that measured their level of experience-taking. The results showed that participants who read a first-person narrative about a student at their own university had the highest level of experience-taking. And a full 65 percent of these participants later reported they voted on Election Day. In comparison, only 29 percent of the participants voted if they read the first-person narrative about a student from a different university.

But what are the practical applications of this research?

While the findings would seem to validate the librarian’s clarion call to get reading – for our higher good – other implications are not so heartening.

Might the findings, for example, be used to justify whitewashing, a disturbing practice in which publishers put white models on the covers of books featuring non-white protagonists?

In 2009 Australian author Justine Larbalestier was appalled to find her American publisher, Bloomsbury, had changed the cover model on her novel Liar from black to white in an effort to sell more books. Larbalestier was successful in her campaign to have her publisher to redo the cover, arguing that the perception that covers featuring non-white models do not sell is merely self-fulfilling prophecy. But what if a deeper psychology is at play?

And who is to say that a reader’s experience-taking of less virtuous characters is not an argument for censorship? Might the psychopathy of Patrick Bateman be contagious after all, as censors insist?

Recently fierce arguments have erupted in Germany over whether Hitler’s Mein Kampf has the power to make things happen. Some might argue that the diatribe is more fiction than fact, but this side of history, at least, it is hard to imagine anyone losing themselves in the character of Hitler.

Perhaps to be safe, though, it should be stipulated that the book only be read in cubicles containing a long mirror.