#ElectionWatch: Fake Photos in Catalonia?

Assessing the Spanish Foreign Minister’s claim of “fake news”

Left, the original image (Source: Twitter / @AminaHekmet). Right, the doctored image (Source: Twitter / @MainatJM).

On October 22, Spanish Foreign Minister Alfonso Dastis told the BBC’s Andrew Marr that “many” of the pictures of police violence shared online during Catalonia’s contested independence referendum on October 1 “have been proven to be fake pictures.”

Open sources confirm that a number of images were indeed taken from earlier confrontations, and a number of claims made about participants were false. However, many dramatic images, both still and video, are authentic. The focus on fakes thus undermines the genuine evidence of police violence, as well as the reporters and researchers who documented it.

The claim

Dastis’ initially argued:

I think, by now, many of those pictures have been proven to be fake pictures, and if there was any use of force, it was a limited one, and prompted by the fact that the law and order agencies were prevented from discharging the orders of the courts.

http://www.bbc.com/news/av/world-europe-41712725/spain-fm-many-police-violence-pictures-fake

Challenged by Marr on whether “those pictures that people saw of Spanish police intervening aggressively in polling stations are all fake pictures”, Dastis clarified:

I’m not saying that all are fake pictures, but some of them are, and you know, there have been a lot of fake news and alternative facts here. (…) If there was at all — according to the pictures, there was — some use of force, it was not a deliberate use of force, it was a provoked use of force.

Dastis’ precise claim was therefore that some of the images circulating of violence and its victims were fake, but that there was some use of force, and that use of force was “provoked”.

Fake images

The first claim is true as far as it goes: a number of images which allegedly showed the victims of police violence had, indeed, been taken from earlier incidents and misattributed. This is one of the most common methods of generating fake content and is widely known from countless crises, elections, and other political moments.

Spanish fact-checking group Maldito Bulo (literally, “damned hoax”) exposed a number of fakes during and after the referendum.

For example, on October 1, Twitter account @PersianRose1 shared a video described as, “Spanish police attacking Catalan voters.”

Tweet from @PersianRose1, archived on October 2, 2017. (Source: Twitter)

The Maldito Bulo team pointed out that this footage came from an earlier confrontation, identifying the police force as Catalonia’s “Mossos” and dating it to a Spanish general strike of 2014.

Translated from Spanish: “NO. This video of a charge against a child which is doing the rounds is not from today. It’s from 2014, and shows Mossos charging in the general strike.” (Source: Twitter)

In fact, both attributions appear to be incorrect. The video was uploaded to YouTube by Spanish outlet Ceres TV on November 15, 2012, in the context of a general strike the day before. A comparison of stills from the @PersianRose1 tweet and the Ceres TV video confirms the match.

Left, screenshot from the Ceres TV video noting the date at bottom left (Source: YouTube). Right, screenshot from the @PersianRose1 tweet (Source: Twitter).

The error by Maldito Bulo is a minor one, and understandable for an open-source researcher on a day of intense drama and increased volume of content. The misidentification by @PersianRose1, or its source, can only be viewed as fake.

The same day, @Twitter user @MainatJM shared a photo of police pushing back against demonstrators, under an enormous Catalan flag.

Translated from Spanish: “Pulitzer photo. Sorry, I don’t know who the author is.” Archived on October 2, 2017. (Source: Twitter)

Within an hour and a half, Maldito Bulo tweeted that the image was a fake, with the flag included using photoshop.

Translated from Spanish: “NO. The photo is false. The star-spangled banner is a montage.” (Source: Twitter)

Reverse searching the image confirms that the flag was indeed superimposed on another image from the same day, widely shared on Twitter.

Left, the original image (source: Twitter). Right, the doctored image (source: Twitter).

However, the Maldito Bulo tweet was retweeted over 3,700 times; the fake was retweeted over 12,600 times, showing the difficulty of fake-busting in such a heated and viral information environment.

Fakes were spread to discredit both sides. One Facebook post from a user called Fernando Garcia accused demonstrators of attacking a policeman.

Facebook post by Fernando Garcia. The text includes the phrase translated from Spanish: “Let the whole world know what these undesirables are doing with one of our NATIONAL POLICE. SAVAGES, COWARDS AND TRAITORS!!” (Source: Facebook)

In fact, the photo was published by leading Spanish daily El Pais in 2008.

(Source: El Pais)

An article published by website okdiario.com on October 4 claimed an old woman carried away by the riot police, whose photo had gone viral, was a friend of Basque separatist Arnaldo Otegi.

Translated from Spanish: “The peaceful granny’s ‘show’ is uncovered: she’s a friend of the terrorist Otegi.” (Source: OKdiario.com)

The Maldito Bulo team said that they had identified and contacted the woman in the referendum image, and proven that she was not the woman photographed with Otegi.

Translated from Spanish: “NO. It’s not the same lady. We spoke with the one from the Guardia Civil photo. She’s called Fina, lives in Sant Julia de Ramis. She doesn’t have a twin sister. The other one is a lady from Bilbao.” (Source: Twitter)

Okdiario.com also wrote that various images shared online by left-wing Spanish leader Pablo Iglesias in a tweet to Spanish Prime Minister Mariano Rajoy were “manipulating reality and showing scenes which do not correspond with the facts”. The outlet quoted this tweet.

Translated from Spanish: “Is this your victory, @MarianoRajoy? #NotInMyName.” (Source: Twitter)

The tweet shared four images, showing various aspects of the police action, of which three showed varying degrees of the use of force.

The images shared in Iglesias’ tweet. (Source: Twitter)

These images appear to be genuine; none appears to have been posted before October 1, according to reverse image searches, and each is cross-referenced by mainstream media or other posts. The woman with the bleeding forehead was featured in a separate @Instagram post on October 1 from a user who called her his mother and said she was hit while trying to vote; the photo of the man in the bloodstained shirt was sent to a BBC journalist by an eyewitness and showed a man “caught up in scuffles at a polling station”; the old woman being carried was also photographed by Getty Images near a polling station.

In this case, the accusation of “fake” appears to be fake.

Taken together, these cases confirm Dastis’ contention that some of the images shared of violence and police action were fake, and that some “fake news” was circulating — on at least one occasion — by claiming that genuine images were faked.

The examples also confirm his admission that other images were genuine. Spanish police did use force against demonstrators; some of the demonstrators were injured as a result; the victims included both women and men, old and young. Critically, these incidents largely occurred as police moved in against polling stations, and the victims most portrayed on social media appear to have been attempting to vote.

In the eyes of the Spanish government, the holding of the referendum itself was illegal (it has also been termed a “coup d’état” and “rebellion and disobedience”); this may be the logic behind Dastis’ final claim that the violence was provoked.

Thus, Dastis’ first two claims are correct; the third largely depends on how “provocation” is defined. In a narrow view, his comments were unexceptionable. More broadly, however, they are troubling.

The reference to “many of those pictures” being fake, even with the subsequent caveat that not all were, undermines the genuine evidence of the use of force by Spanish police on October 1. Dastis’ claim was not false, and it should not, therefore, be termed “fake news” itself; but by casting doubt on the evidence, it undermines the work of the reporters and activists who documented genuine incidents of the use of force, and their results.


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 #DigitalSherlocks.