Back in January, in a baptismal post, BuzzFeed's Silicon Valley Bureau Chief Mat Honan christened the next trend to take over the social web: the "screenshort." He explained the phenomenon — "a chunk of text, screen-shotted, and embedded in a tweet" — and prescribed, firmly: "The bottom line is, if you want someone to read something on Twitter, don't just link to it. Post a screenshort as well."
The premise behind the screenshort is old, but the hype is new, and Honan wasn't the only one trumpeting it. By late last year, bloggers were calling it a trend, and by the beginning of this one, major outlets such as the New York Times, BuzzFeed, Politico — even The Onion — were adopting the 'short as a convenient work-around to social media (mostly Twitter)'s character restrictions. In a post from March, BuzzFeed's Charlie Warzel wrote of the screenshort: "It's a work-around to Twitter's 140-character limit that doesn't abuse the constraint as much as enhance it." In April, The Next Web noted that celebrities such as Justin Bieber, Pink, and Lorde had adopted the screenshort as a bare-bones blogging tool, ultimately declaring that the traditional blog had been "killed by screenshorts." At this point, a cursory scroll through Twitter — and, to a lesser extent, Instagram and Facebook — yields a consistent stream of these pictures of text.
There's one problem, though: The screenshort is a bad idea.
The screenshort is what a programmer or an architect — or, incidentally, a blind person — might call a universal design failure. For the blind, the issue is obvious: Unlike normal text, pictures can't be interpreted by the screen readers many — if not most — visually impaired people (including me) use to navigate the web. If screenshorts continue to gain popularity the way so many breathless blog posts predict they will, the internet will become significantly less accessible to hundreds of millions of people all over the world.
But not only are screenshorts near-useless for the blind and visually impaired, these static, un-highlightable, un-copyable, un-searchable blocks of text threaten to "blind" the internet writ large. Put simply: When you take a screenshot of text, the information therein is lost. This makes screenshorts effectively invisible to blind people's screen readers — but it also prevents the text from being found easily by anyone who wants to locate or index it later. The real important data — the text data — is stripped, and if the text isn't published somewhere else, it can never be found in a search engine, either.
The screenshort has gone mainstream. https://t.co/TCIWrIH5L2 https://t.co/8F2CNPIEhg https://t.co/FzS6Rwlaz4 https://t.co/j6l0LVKf4Q
Last summer was when screenshorts — or at least a version of them — first gained some legitimacy on Twitter, as the platform itself began experimenting with an official feature to enhance tweets with images of other tweets (this has now become "retweet with comment"). At the time, Mashable attributed the move to Twitter's push toward becoming "a more visual service," much in the way of Facebook and other social networks.
Though that feature wasn't exactly a screenshort, its rollout was enough to worry many members of the blind community, including assistive technology consultant Jonathan Mosen, who wrote a blog post suggesting that the new trend toward a more visual Twitter could signify the end of a "golden era" for Twitter's accessibility. At the time, Twitter was an ideal social network for the blind: linear, text-heavy, streamlined, and digestible, easily consumed via screen reader. For that reason, Mosen wrote, "Twitter ... represents one of the most empowering aspects of the Net for disabled people." Mosen said he hoped Twitter would keep this in mind as it developed the feature. "I think it's better to raise these issues now, and seek clarification, than to play ostrich and then complain about the degraded accessibility when it's too late."
Twitter was listening. Two days after Mosen's post, Twitter addressed the issue head-on, breaking its policy of not commenting on experimental features to reach out to Mosen directly. A Twitter spokesperson told him that if any official updates were made to the service, they would be entirely accessible to the blind. "It's reassuring that Twitter is so cognisant of accessibility issues," Mosen wrote.
But if the blind community dodged a bullet in Twitter's official rollout of graphic retweets, something much bigger was coming in the form of the widespread adoption of screenshorts. In December, M.G. Siegler published a post on Medium called "Hacking the Tweet Stream," advocating for the method and pointing out that it was spreading to not just techies, but celebrities and pro athletes.
Adrian Roselli, a veteran web developer and interface designer, responded to Siegler with with a post titled "Don't Tweet Pictures of Text," pointing out that there are plenty of reasons not to turn textual data into images — not just for the sake of blind people. Converting text to image, Roselli points out, not only creates problems when it comes to resizing and reading out loud, but sets up new roadblocks against translation, small devices, foreign devices, and limited data plans.
"This isn't an accessibility issue," Roselli wrote on his blog in December, "it's a usability issue and an engagement risk." And as it turns out, it's a much more complicated problem than it appears to be.
OneShot is a new app that actually specializes in screenshorts. Interestingly enough, the company says it is keenly aware of the accessibility problem, but says it is not necessarily on its end. Ian Ownbey, the app's one-man tech team, told BuzzFeed News that OneShot already captures data using OCR (optical character recognition, the same technology used by the blind). But it can't use the data it has.
"We have the text," he said, and suggested that it'd be easy to implant it into the screenshort's metadata, making it machine-readable. So why doesn't OneShot do that?
The problem, Ownbey said, is that almost all social networks strip metadata from photos as soon as they're uploaded. There are different ways to attach text information to a photo using tags, but without any agreed-upon standard for accessible photo metadata, it's hard for Ownbey to make a coherent case to any social network about which metadata is sacred.
After puzzling it over and finding no simple solution, Ownbey has come to think that perhaps the companies creating screen reading software need to get with the times. "I was actually kind of surprised that screen readers don't have OCR software," he told BuzzFeed News. "I think as we move towards mobile more and more, images for better or for worse are just going to become more pervasive, so it seems like something that will continue cropping up."
But as mainstream media continues to embrace the work-around presented by screenshorts, it's worth remembering it's just that — a work-around, not a solution. And it's working not with Twitter's user interface, but in spite of it. This is why screenshorts will be a bane not just to the blind, but to everyone else as well.
Imagine an internet strewn with the data equivalent of dark matter. Convert enough text to images, and the algorithms used by Google, Twitter, Facebook, and any other service will stop working as well. Old screenshorts will soon enough be artifacts of the ancient web— a massive vault of photographs, unlabeled, unsorted, and sourced from every corner of the world.
It's entirely not hard to imagine a future in which, like looking through a telescope into space, we won't even know what we're not seeing. Google, Facebook, Twitter, Yahoo, Reddit — none of their algorithms would be worth a penny if those indexes didn't contain textual data. Precious, precious SEO — the art form that so many marketers, programmers, and high-level officers have spread as gospel — feeds on plain text, not images.
When it comes to the internet, we're all, in a sense, blind. You can't just go on the internet and "see" what you're looking for on the horizon. You have to search around until you're actually close enough to "touch" it, so to speak. This is why we rely so heavily on the web's equivalent of "guide dogs": the search engines, forums, apps, and social networks that help us get where we mean to go — and, not insignificantly, use text to get us there.
It's easy to assume that people with disabilities live a simpler, less technology-saturated life — but actually, that couldn't be further from the truth. The blind, in particular, are often technological obsessives, early adopters, and in most parts of life must be methodical, thoughtful, and purposeful. We can — we must — cut through the GIFs and photos and bells and whistles to use the web in its simplest and most purely functional form. For that reason, blind internet users are the canary in information technology's coal mine. And now we're emerging, covered in soot, to report that if screenshorts proliferate as they have, they could collapse the web.