Zombie misinformation

In which some extremely old bad advice goes viral in the days leading up to the No Kings protests

this doesn’t actually need to go viral, because it’s a bad idea

In the weeks leading up to the June 14th, 2025 No Kings protests, memes advising protesters to spontaneously sit down if violence breaks out in their vicinity circulated widely on multiple social media platforms. As many people pointed out, this is an extremely unwise idea unless one actively desires to be harmed, as it makes a person unable to flee the area, more vulnerable to being injured by police horses or vehicles, and more likely to be struck in the head with less-lethal projectiles, among other problems. (Comparisons with sit-ins are inaccurate, as sit-ins are coordinated sit-down protests, not spontaneous responses to chaos.)

As it turns out, this particular bit of bad advice is not new; it has been circulating online here and there for roughly eight years, although it has failed to get traction during prior large protests.

regardless of the year, Facebook is full of bad ideas

The first occurrence of the “sit down wave” meme that I ran across via various text and image searches was posted to Facebook on December 19th, 2017, followed by a handful of additional posts toward the end of December 2017 and in the early months of 2018. (It is possible that the post pictured above is not the first and that the meme therefore emerged earlier, but if earlier instances exist, I was unable to find them via the search methods used.) Although some of these old posts individually received decent engagement, they do not seem to have been copied widely or triggered any sort of trend at the time.

various “sit down wave” image posts from May and June 2025

As Donald Trump’s second term in the White House got underway in 2025 and protests began, the “sit down wave” meme reappeared here and there on various platforms, including but not limited to Facebook, Instagram, Threads, Bluesky, X, and TikTok. The “sit down” posts are highly repetitive, with the original text from the 2017 posts resurfaced as both plain text and variously colored images. A new “sit down wave” image containing a five-point list also popped up, and was shared on multiple social apps and websites.

unfortunately, not everything posted by popular accounts is reliable

On May 30th, the “sit down wave” meme was posted by the Occupy Democrats organization to multiple social media platforms, garnering massive engagement and causing the meme to go viral among liberal users discussing the upcoming No Kings protests. Many of the posts from that point onward used screenshots of one of the Occupy Democrats posts, usually the Instagram version. The Occupy Democrats Facebook page proceeded to post additional variants of the meme over the ensuing days.

note that posts with no text content are not included

One of the platforms the “sit down wave” meme was spread on, Bluesky, is built on an open protocol, allowing data about public posts and their spread to be freely downloaded and studied. For the sake of this analysis, posts from May 1st 2025 onward containing both “sit down” and “protest” were harvested, along with their associated reposts. This data shows that the first significant pre-protest Bluesky activity containing the relevant terms took place on May 31st, the day after the Occupy Democrats Facebook and Instagram posts. Activity peaked on June 12th, and then dropped off somewhat on the 13th and 14th. (It would be nice to be able to perform the same analysis on all platforms, but we can’t presently have nice things.)

accounts that originally posted the “sit down” advice but issued corrections prior to the protest are in the “debunking” category (green)

The repost data can be used to construct a network graph of the spread and debunking of the “sit down wave” meme, at least for accounts that included the search terms “sit down” and “protest” in their posts. The accounts that promoted the meme are mostly clustered in the same area of the graph because they were often reposted by the same accounts, and in several cases by each other. Note that this graph is likely incomplete, due to the absence of image-only posts and deleted posts from the dataset. (For those of you who may be interested in producing Bluesky network graphs of this sort, I plan to describe the process I used and include example code in an upcoming article).