Since then Elon Musk has ended his Twitter deal, and the network’s diehard users have taken steps to praise him. People have downloaded their own archives from Twitter. Others have started the discussion with screenshots of their all-time favorite tweets. And there’s an ongoing Google Doc cataloging Twitter trends and memes, a guide that could one day be used to decode the app’s hieroglyphics.
Whether Twitter goes bankrupt (as Musk himself puts it) or becomes an unpassable stream of hate speech and deceptive parody accounts, the future of the network is uncertain. But there are concerns that Twitter’s treasure trove of content, important to history and political influence (and good laughs), could be lost. Twitter’s founding premise — a 140-character (now 280) one-liner — was not a good fit for the archive. This is partly because capturing content streams that grow by thousands per minute is a technical nightmare, but also because of ethical concerns that not all tweets are created equally. Some are fired by world leaders who incite violence, while others are fired by unknown ordinary citizens, if not for their affinity for bird apps. Both types of tweets can go viral and have a lasting impact.
“I think it’s really important to think carefully about the data you collect,” said Miles McCain of PolitiTweet, a service that archives tweets from public figures and influential organizations. “When you try to archive anything, you end up with a whole bunch of irrelevant information.”
An attempt by the Library of Congress to record every public tweet in 2010 failed. Tweets evolved from short texts to regularly include photos, videos and live links. Seven years later, the library ended the Sisyphean project, saying it would only archive selected accounts. In 2012, the library said it archived 500 million tweets a day. A library spokesman had no comment to WIRED prior to publication of this story.
Elisabeth Fondren, a journalism professor at St. John’s University in New York City, said the failure of the archiving project demonstrated a huge missed opportunity to preserve a rich dataset of political discourse and communication trends. Currently, the need for social media archiving is gaining attention and exposing the instability of hosting a public square on a private company’s servers.
“If it had been successful, we would have it by now,” Fondren said. “It really undermines attempts by researchers to assess the social impact of media on society.”
For years, smaller third-party services have sought to archive more specific content. ProPublica maintains a list of deleted tweets by politicians in its Politwoops database. PolitiTweet has a database tracking 1,500 accounts. These document statements and news stories by key figures in government and politics, but the projects are not intended to capture the mass discourse of online communication.
Twitter is designed to capture moments, and finding or viewing old tweets in the early days wasn’t easy and didn’t seem to matter. But by 2014, Twitter improved its public tweet search tool. The move helped researchers, but it also breathed new life into long-forgotten tweets that had been moved around the timeline without much afterthought. The change proved problematic for some Twitter users, such as those who started pronouncing 140 characters aloud as teenagers but later became college students or young professionals. Their tweets don’t always go stale, especially as the era of cancel culture begins.
