In June, Twitter was criticised by disabled advocates for releasing a new voice tweets feature without captions. They apologised, said “accessibility should not be an afterthought”, then rolled out the inaccessible tool to more users.
Despite their best intentions, Twitter still has the controversy of voice tweets lingering in the background of any future conversations around accessibility.
As a handful of employees at the social media platform chaired a panel on the subject in the latest in their #BehindThePlatform series on Tuesday, the aftermath of the backlash still lives on in the creation of two new accessibility teams in response to the criticism. The Accessibility Center for Excellence “will set goals, drive progress, consult and partner with groups across our core business […] functions to help make aspects of Twitter more accessible”, while the Experience Accessibility Team will work “to ensure [the platform is] held accountable in identifying and filling accessibility gaps”.
Away from managerial changes, the only direct commitment to improving the situation with voice tweets came with a plan to add automated captions to audio and video on the platform “by early 2021”. A few weeks after the blog post about the teams was published in September, Twitter Support tweeted that they have “taken your feedback seriously and are working to have transcription available to make voice Tweets more accessible”.
It was, however, in the same 271 characters that they revealed that it would be expanded to more users on iOS. The “feedback” clearly wasn’t taken on board.
Twitter cannot simply spout the disability mantra of “accessibility should not be an afterthought”, when their continued rollout of an inaccessible feature screams ‘put it out now, worry about discrimination later’. Don’t appropriate our language only to disregard it weeks or months later.
“To clarify, the reason is so the team can continue to learn about how people use audio,” Twitter developer Amro Mousa replies, to a tweet of mine on Twitter. “Simultaneously, our teams are working to bring automatic captions to Voice Tweets (and more) — they’re doing their best to do that quickly, but it’s tricky at scale and w/ multiple languages.”
My follow-up tweet did not receive a response. If the further rollout was designed to see how people use audio, then what’s stopping Twitter from analysing the usage of those who had voice tweets from the beginning? Boosting the sample size is no doubt helpful for monitoring, but not when allowing more accounts to use the tool also increases inaccessibility.
A spokesperson for the social media platform declined to comment further beyond what was said by Amro and Twitter Support previously. A detailed rationale for extending an – at present – discriminatory feature to more users is still yet to be revealed.
To think Twitter had learned its lesson in the summer…
Liam O'Dell 