Twitter-Induced Degradation

In which I discuss the manner in which and the degree to which Twitter is ruining the media.

————

Yesterday, Kevin Drum posted saying that Twitter is ruining political journalism, calling out its role in solidifying media groupthink before events are even completed. That seemed like a pretty good criticism to me, but like a true squishy liberal, Kevin later retracted his comments in the face of criticism. I saw the second link come across Kevin’s Twitter feed, and responded there, saying:

You were right the first time. But Twitter’s only the latest step in the continuing degradation of political journalism.

Kevin replied:

But has Twitter degraded physics? Inquiring minds need to know.

Which is a fair question, and a decent subject for a blog post to take my mind off other things for a few minutes.

The short answer, which I gave on Twitter last night, is basically to link back to this old post and op-ed about the OPERA fast neutrino story, which I think is a good illustration of the way Twitter, and social media more generally, has caused some degradation in physics media.

If you don’t want to follow the link, my claim is basically that the OPERA people behaved as they should have when confronted with a weird result: they did the obvious internal consistency checks, couldn’t find the error, and then they showed their preliminary results to other physicists, to see if anybody else could find the error. This sort of thing goes on all the time, and is part of the tutelary myths we teach our students– in my own field of laser cooling, my old group at NIST famously found some “impossible” results, which they quietly shared with other groups, by distributing pre-prints and giving talks at institutions where there were other people doing laser cooling research.

So, what happened to OPERA? Well, they did basically the same thing: they posted a preprint on the arxiv (the modern method of distributing such things), and arranged to give a presentation at CERN, where the people most likely to have relevant expertise are concentrated. That’s exactly what they should do. The problem is, in the current media climate, that’s as good as holding a press conference: science journalists look at the arxiv, and they follow scientists on Twitter, so word got out early, and the whole thing turned into a media circus.

You saw something similar with a dark-matter rumor a year or two earlier, where frantic Twitter activity helped blow an inconclusive result from CDMS up into a huge Thing, and with a couple of rounds of Higgs announcements. Stories that would’ve quietly unfolded at their natural pace ten or twenty years ago wound up confronting exaggerated expectations and the scientists involved had to scramble to do a bit of damage control.

So, yes, Twitter has done some damage to scientific media, in more or less the same way that it’s done damage to political media. The degradation isn’t quite as bad, though, because buzz-worthy scientific stories aren’t quite as common as political news, and the community of people who follow science news via social media isn’t quite as big and frantic as the community of people who follow political news via social media.

I also said in my response to Kevin that I think Twitter’s just the latest step in the process, which began back when I was a high school and college student, with the launch of 24-hour cable news, which greatly accelerated the pace and hype of news stories, compared to the prior world of evening newscasts and daily papers. The next step was political blogs, which are even more frantic than cable news networks, and then Twitter, which is to political blogging what political blogging was to cable news.

Along the way, lots of people have sort of decoupled from the system at different stages of the process. There were people who found cable news to be a bit too much back in the day, and others who could never take blogs. I went along until well into the blogging phase, before the pace of outrage-of-the-moment blogging got too much for me; now, I don’t even watch cable news.

About the only political blogs I read with any regularity these days are Kevin’s, Fred Clark’s Slacktivist, and Nate Silver’s Five Thirty Eight. In each case, I like them because they follow a pretty reasonable schedule: Fred and Kevin post a small handful of pieces daily, and Nate posts roughly one article per day. That’s about right for real political news– the natural time scale for political stories is days and weeks, not hours or minutes. Even at one post per day, Nate Silver seems to spend about 70% of his time reminding people that nothing has actually changed since the previous day.

Now, this doesn’t mean that I think Twitter needs to be razed to the ground and the earth beneath it sown with salt. I’ve had a good deal of fun following things like the debates via Twitter, as anybody who follows me would guess from the string of retweeted jokes and so forth. But it’s important to recognize it as a quick diversion, and not mistake it for actual information.

The problem with the media use that Kevin originally criticized is that the rapid-fire Twitter traffic and absorption in social media helps a set frame for an evolving story coalesce more rapidly than it should. Back in 2000, it was hours after the debates before the “Al Gore sighs” narrative took hold. In 2012, they were maybe a third of the way through the first debate before the “Obama is asleep” story was completely locked in, and nothing changed after that. That’s kind of a problem, and produces a sort of rigid lockstep among the moderately sane media (Fox News and its nuttier brethren are largely decoupled from this, off in their own bubble universe) that’s bad for political discourse.

Of course, I don’t think there’s any way to change that, and I suspect it will get worse before it gets better. I hold out some hope that as things get faster and crazier, more people will start to back away, and some degree of sanity will be restored through a recognition that not every story evolves in ten-second 140-character bursts, and Twitter is better understood as entertainment than information. But we’re probably a long way from that.