WASHINGTON (WASHINGTON POST) - Twitter, used by 126 million people daily and now ubiquitous in some industries, has vowed to reform itself after being enlisted as a tool of misinformation and hate.
But new evidence shows that the platform may be inflicting harm at an even more basic level. It could be making its users, well, a bit witless.
The finding by a team of Italian researchers is not necessarily that the crush of hashtags, likes and retweets destroys brain cells; that's a question for neuroscientists, they said.
Rather, Twitter not only fails to enhance intellectual attainment but substantially undermines it, the economists said in a working paper published this month by the economics and finance department at the Catholic University of the Sacred Heart in Milan.
"It's quite detrimental," Gian Paolo Barbetta, a professor of economic policy at the private research university and the paper's lead author, said in an interview with The Washington Post. "I can't say whether something is changing in the mind, but I can say that something is definitely changing in the behavior and the performance."
To the best of his knowledge, Barbetta's study is the largest and most rigorous examination of Twitter's effect on student achievement, with applications to learning and information retention in other areas of life.
The investigation drew on a sample of roughly 1,500 students attending 70 Italian high schools during the 2016-17 academic year.
Half of the students used Twitter to analyse The Late Mattia Pascal, the 1904 novel by Italian Nobel laureate Luigi Pirandello, which satirises issues of self-knowledge and self-destruction. They posted quotes and their own reflections, commenting on tweets written by their classmates. Teachers weighed in to stimulate the online discussion.
The other half relied on traditional classroom teaching methods. Performance was assessed based on a test measuring understanding, comprehension and memorisation of the book.
Using Twitter reduced performance on the test by about 25 to 40 per cent of a standard deviation from the average result, as the paper explains.
Jeff Hancock, the founding director of the Stanford Social Media Lab, described these as "pretty big effects".
Notably, the decline was sharpest among higher-achieving students, including women, those born in Italy and those who had scored higher on a baseline test. This finding, the paper notes, bolsters the conclusion that blogs and social networking sites actively impair performance, rather than simply failing to augment learning.
A spokesman for Twitter declined to comment on the study.
The company does not purport to make its users smarter. But its mission statement sets forth goals not so different from those of a literature course - "to give everyone the power to create and share ideas and information instantly without barriers".
And in describing the platform as a "digital public square", Twitter's chief executive Jack Dorsey appeared to embrace civic and social aspirations, saying last year that the standard to which the company should be held is "building a systemic framework to help encourage more healthy debate, conversations, and critical thinking".
Barbetta said more results were necessary to draw definitive conclusions about the "possible negative effects" of Twitter on learning. "As results accumulate, we definitely should be more wary about how we use social platforms," he said.
The study focused narrowly on high school literature students. But that approach gave the researchers access to a large sample, as several hundred Italian schools had already adopted a framework for Twitter-based conversations about literary masterpieces, called "TwLetteratura".
The method also allowed them to avoid problems plaguing past studies, some of which allowed participants to opt in to social networking, skewing the data towards those with an aptitude or particular interest in online engagement.
The relevance of literature and reading comprehension to evaluations of digital communication was underscored on Wednesday, when Special Counsel Robert Mueller seemed to enjoin the nation to heed the warnings in his 448-page report - in other words, to do the reading instead of consuming sound bites on social media.
Karen North, a professor at the University of Southern California's Annenberg School for Communication and Journalism, said the study had stark implications for politics, adding that its findings were hardly surprising.
"It's the same problem that we have with the level of political discussion," said North, who previously worked on science and technology policy in Washington, both in the White House and on Capitol Hill. "People get 280 characters, and it's not enough. Without the full background, you're more likely to be led astray."
Recent analysis has suggested that broad swathes of the electorate are not as engaged online as is an especially vocal cohort of digital aficionados.
Still, Twitter, which is President Donald Trump's preferred medium of communication, has cemented its place in the political ecosystem, and its role is only like to expand in advance of the 2020 election.
Twitter is where candidates go to issue announcements and respond instantly to news developments. It's also where pundits react in real time.
But the platform doesn't lend itself to explanation or in-depth analysis, North said. "Remember when we were debating whether people have the attention span to consume 280 characters, instead of just 140?" she recalled.
While social media shouldn't be dismissed as a learning tool, more thought is required to determine the strengths of different technologies and their proper audience, she said.
Above all, the communications professor emphasised, platforms like Twitter should not replace more traditional methods of engagement, especially in grappling with complex topics - whether that means a presidential election or the plot of The Late Mattia Pascal. The problem, said Barbetta, is that people will take a shortcut if it's given to them.
"But a shortcut won't take you to the destination in this case," he added. "It will take you somewhere different."
As the study indicates, Twitter is the ultimate shortcut. Barbetta suggested that declining performance among students who had used the social networking site to study the novel was a result of two factors.
The first was a mistaken belief on the part of students that they had absorbed the book by circulating tweets about its contents. The second was that time spent on social media simply replaced time spent actually poring over the book.
The study contributes to growing skepticism that human activities - and learning, specifically - can be transferred to cyberspace without a cost. For instance, analysis has found that screen-based reading lends itself to skimming.
In a 2016 study, it was discovered that test scores were lower among American undergraduates assigned to classrooms where computers were allowed than among those required to resort to pen and paper.
In the case of Twitter and Italian literature, the initial assumption of the study turned out to be faulty. "We thought we were testing a positive intervention," Barbetta said.
Among some researchers, the urgent question is now whether social media - once embraced uncritically - is a net positive; indeed, whether it is capable of accurately reflecting reality.
It's a problem once captured by Luigi Pirandello, the author of the literary text used in the Italian study.
"There is someone who is living my life," the Nobel laureate wrote in a diary entry in 1934. "And I know nothing about him."