#ElectionWatch: Catalans Tweet Tweety For Freedom

How a cartoon character became a separatist symbol

Meme from pro-independence account @CriaDemocracia. Archived on September 25, 2017. (Source: Twitter)

On September 24, the political showdown in Spain between independence-minded Catalans and the central government in Madrid brought in an unlikely hero: Warner Brothers’ cartoon character Tweety.

Known as “Piolín” in Spanish, from the noun “un pío” (a tweet, as in the noise, rather than the social media post, which is “un tuit”), Tweety’s name quickly became a rallying cry for separatists, under the hashtag #FreePiolin.

In less than twelve hours, the hashtag racked up almost 170,00 tweets — nearly twice the volume of traffic on the word “Catalunya” (Spanish for “Catalonia”) over the same period.

Tweets on “FreePiolin” and “Catalunya”, September 25, 2017, 14:00–23:00, from machine scan.

This was more than a moment of light relief in an increasingly tense standoff ahead of Catalonia’s planned referendum on independence, scheduled for October 1. The hashtag and accompanying memes were used to accuse the Spanish government of muzzling democracy and were partially driven by automated “bot” accounts.

A ship with a view

The events, which led to the hashtag, began when the Spanish government deployed gendarmes — the Guardia Civil (“Civil Guard”) — to the port of Barcelona, Catalonia’s capital.

To accommodate them, officials chartered passenger ships. One of them was decorated with images of Warner Brothers characters, including Tweety, as social media users pointed out on September 21.

“Consider that while a good number of Catalans are sleeping at the head of the Lluis Companys Passeig [in Barcelona], the Guardia Civil has come on board two luxury cruisers.” (Source: Twitter / @jacintfelip)

The ship’s name is the Moby Dada (IMO number 7911533), an Italian-registered roll-on / roll-off passenger vessel, as a comparison with the below image from shipspotting.com shows.

Image of the Moby Dada from July 16, 2017. (Source: shipspotting.com / Eric Houri)

As of September 25, the Moby Dada was berthed in Barcelona harbor, according to vesselfinder.com.

(Source: vesselfinder.com)

The decoration met with mockery, and a wide variety of memes and images spread online. Some of the posts used the hashtag #1O, referring to the October 1 referendum date.

“BREAKING: our reinforcements to control the #1O arrive by train.” (Source: @cescsaladrigs / Twitter)

While some appeared light-hearted, others had an edge, referring to the Guardia Civil deployment as repression.

“First images of the cabins on the ship chartered by the police to repress the Catalan people.” (Source: Twitter / @PereForcada)

At this stage, while traffic on some of the posts was significant, reaching into the hundreds or low thousands, no specific hashtag stuck out.

Masking the meme

On September 24, however, the cartoon images on the ships were covered over. The change was reported by Catalan media, most significantly this tweet from daily DiariARA, posted at 14:34 UTC.

“Important news at the Port: Piolí / Piuet / Tweeti ‘censored’?” (Source: Twitter / @diariARA)

Some of the replies were humorous and referred to other Warner Brothers catch-phrases, such as “That’s all, folks” and the “ACME” brand.

“Is the canvas the ACME brand?” (Source: Twitter / @CATexit61)

However, within three minutes, Twitter user @Chyavanaprasha had replied with the hashtag #FreePiolin.

(Source: Twitter / @Chyavanaprasha)

The same minute, another user, @Piulador, retweeted the post, using three hashtags: #FreePiolin, #FreeTweeti, and #FreePiuet (“Piuet” is Catalan for Tweety).

(Source: @Piulador / Twitter)

Another user, @jan_xc, replied to the @DiariARA post, copying Warner Brothers, and adding the hashtags #FreePiuet, #FreePiolin, and #FreeTweety.

(Source: Twitter / @jan_xc)

Traffic on these other hashtags did not perform well. In 24 hours, #FreeTweety was posted 7400 times. #FreePiuet was posted 593 times, while #FreeTweeti was posted just 55 times.

Piolín goes viral

However, #FreePiolin went viral. In the space of an hour and a half, traffic accelerated from zero to a peak of over 750 tweets per minute by 16:00 UTC.

Twitter traffic on #FreePiolín from 14:30 UTC on September 25, from machine scan.

This is a high level of traffic and a rapid acceleration; however, it should be noted that the takeoff does not show the near-vertical spike characteristic of purely bot-driven traffic. Compare with this timeline of traffic on a post driven by a botnet.

Traffic on #StopNGOSoros, July 22.

@DFRLab conducted a machine scan of traffic on the first 100,000 tweets to use the hashtag. This showed that the posts had come from 31,629 users, for an average of 3.1 posts per user. This is higher than the typical range of a purely organic series of posts, but well below the range of 5–10 tweets per user that @DFRLab has identified in exclusively bot-driven hashtag drives.

Results of machine scan, showing total number of posts and users, and the top users.

The number of original posts was also high. Of the almost 100,000 tweets posted, a little over 15,000 were original posts, as opposed to retweets, indicating that relatively large numbers of users were engaged.

Some of the traffic was driven by accounts with very high levels of activity and low levels of creativity, posting hundreds of retweets but few or no original tweets, suggesting that they are either automated “bots” or human-operated amplifiers used to push a political message.

The ten most active users on #FreePiolin, from machine scan from 14:30 UTC to 18:30 UTC on September 24. Note the high number of tweets in a four-hour period.

For example, the most active of all, @321BCN, posted the hashtag 337 times in the space of two and a half hours, at an average of one post every 26 seconds; all but one of its posts was a retweet. This is botlike behavior.

The account was created in January 2016, but only began large-scale tweeting on September 11, 2017. It posted 652 tweets on September 24. Its avatar and background picture were also clearly altered to refer to the Tweety story.

Profile page for @321BCN. (Source: Twitter)

Similarly, @freed4catalonia and @jaumesol each posted 111 times, all of them retweets, in an hour and a half. @LLUISAGC and @IvanKirchen each posted over 200 times in two and a half hours, averaging almost two posts per minute, the majority being retweets. This behavior is characteristic of bots or of dedicated human operators trying to make a hashtag trend.

Original posts, humor and anger

However, the rise of #FreePiolin should not be ascribed solely to automated bots, or to obsessive human users. According to a machine scan, the top fifty accounts posted 5,475 tweets, while the top 10 posted 2,020 — a high number, but a small proportion of the overall traffic of almost 100,000 posts.

Moreover, the tone of the posts which were amplified varied markedly, suggesting that their spread was at least partly organic, rather than the result of a single campaign. Many appeared mocking, or called for a humorous response.

“Now seriously: the situation is ugly… but humor is the main weapon against fear. So let’s carry on!” (Source: Twitter / @AlbanoDante76)
“Just when you think that they can’t make it any more ridiculous, there’s a plot twist that shows they’ll always find the way.” (Source: Twitter / @sergiolesc)
“One. Free. Warner.” (Source: Twitter / @Hqtorzinc)

One sharp-eyed user noted that the images had only been covered on one side of the ship.

“As the colleague who took this photo said, ‘There’s no budget to cover it on the other side’.” (Source: Twitter/ @Bartserk)

One user even altered a genuine post from the Guardia Civil. The original read,

Destiny whispered to the warrior: “You cannot endure the storm.” And the warrior turned and whispered back, “I am the storm.”

The edited version ran:

Destiny whispered to the warrior: “You cannot endure the storm.” And the warrior turned and whispered back, “Has anyone seen a cute little pussy-cat?”
Left, original post from Guardia Civil. Right, altered version from @marcbeldata. (Source: Twitter)

However, others were more directly political. One major pro-independence account, for example, turned the story into a collection of different-colored memes.

“In defense of the rights of everyone.” (Source: Twitter / @CridaDemocracia)

The account is linked to a pro-independence web-page which offers a variety of memes for supporters, all featuring characters with red slashes over their mouths and the tag, “democracy,” in reference to separatists’ view that the Spanish government’s refusal to allow a referendum is undemocratic.

(Source: Crida Democràcia)

The most popular tweet of all proclaimed, “Tweety lives, the fight goes on.”

(Source: Twitter / @gabrielrufian)

Another popular tweet attacked the Partido Popular (PP), Spain’s governing conservative party.

“The PP is a mess, covering things up, whether it’s Tweety or corruption.” (Source: Twitter / @Sr_Dios)

Others invoked Spain’s former dictator, General Francisco Franco, such as the post below.

“Not even Franco banned Tweety. They’re losing their grip.” (Source: Twitter / @pagocre)

This combination of responses matches the combination of human and apparently automated users. The traffic does not show a uniform tone or a uniform message; it does not resemble a coordinated campaign launched by a small group, but a larger and more heterogeneous movement with a mixture of user engagement and bot amplification.

The traffic also underlines — if any underlining were needed — the intensity of online activity in the build-up to October 1. The hashtag #FreePiolin amassed 100,000 posts, and over 15,000 original tweets, in a few hours; within 24 hours, it had passed 200,000 posts. This is a high volume of traffic, and of engagement, especially given the relatively trivial nature of the incident which sparked it.

Online traffic is only likely to increase as October 1 approaches. @DFRLab will continue to monitor the situation and report on significant developments.


Ben Nimmo is Senior Fellow for Information Defense at the Atlantic Council’s Digital Forensic Research Lab (@DFRLab).

Follow along for more in-depth analysis from our #ElectionWatch.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.