Home > News > What Twitter can tell us about the Jerusalem protests
271 views 8 min 0 Comment

What Twitter can tell us about the Jerusalem protests

- August 28, 2017
Palestinian Muslims wave a national flag and flash the victory gesture in front of the Dome of the Rock in the Haram al-Sharif compound, known to Jews as the Temple Mount, in the Old City of Jerusalem on July 27. (Ahmad Gharabli/Agence France-Presse/Getty Images)

The recent tensions in Jerusalem triggered very significant online engagement among the Muslim public sphere. Global digital political engagement trends through two weeks of the Temple Mount clashes give us three main insights on the future of foreign policy and social media.

First, Arabic hashtags, rather than English ones, were shared most widely, driving the digital expression of the Al-Aqsa crisis. This gives us some insight on the role of digital language, as well as local versus external agenda-setting diffusion in digital crises. Second, countries with already high levels of digital political activity (instead of the origin point of incidents) tend to dominate and even lead online agenda during global crises. Third — and perhaps most important — internationalization of crises on digital platforms significantly increases the role of digital spoilers: trolls, bots and fake news.

Driving hashtags: English, Arabic, Turkish

As the Al-Aqsa crisis erupted, I mapped out the spatial distribution of geotagged Twitter hashtags related to the Al-Aqsa clashes in English, Arabic and Hebrew through a two-week period in July. While studying #AlAqsa alone may not be statistically meaningful, comparing it to five other hashtags provides us with a politically meaningful comparison of global digital political engagement.

Only half of the six widest-shared hashtags on Temple Mount clashes were in English: #AlAqsa, #TempleMount, #FreeQuds. Two were in Arabic: القدس# (Al Quds) and فلسطين (Palestinian territories). One was in Turkish: #MescidiAksa (Al-Aqsa Mosque). Hebrew hashtags were virtually nonexistent during the crisis.

The most frequently recurring hashtag during the early hours of the crisis were #AlAqsa and its combinations with #Jerusalem and #Palestinian territories. On the day of July 14, as three gunmen attacked an Israeli police checkpoint in the Lions’ Gate, the first few hashtags began appearing in and around Jerusalem, first receiving foreign attention in Turkey, Qatar and the United Arab Emirates. In the next day, #AlAqsa paired with #IsraeliAtrocities expanded into South Africa. It is mostly with the July 21 clashes at around 1 p.m. — when the Al-Aqsa clashes, Halamish stabbing attack and Mahmoud Abbas’s order to suspend contact with Israel occurred — that global engagement with the crisis through the hashtag #AlAqsa became truly international.

Trolls, bots and fake news

The Arabic hashtag of Al-Aqsa (#المسجد_الاقصي) was more popular in terms of quantity, but not as widely distributed compared to its English version. Based on existing methodological literature on bot measurement, this hashtag contains significant evidence on bot-driven engagement, with regard to its rapid peak on July 14 and its subsequent disappearance within a matter of 30 minutes.

Who was controlling these bots? It’s hard to say. Even though bot accounts or their individual tweets may contain geolocation information, this information doesn’t give out the location of their programmer.

These messages are usually misinformation — either by containing misleading information or through redistributing old information as if it was happening anew. Users either engage with emotionally charged propaganda — or opponents take no action, which confuses the general public. Usually in a two-sided bot conflict, both sides attempt to generate engagement on their side and force the opposing side into inaction.

Foreign countries dominate online engagement

A uniquely Western hashtag was #TempleMount with clusters in the United States, U.K., France and Israel. It was responsive to the Lions’ Gate attack, Al-Aqsa clashes and Umm al-Fahm funeral — peaks during the crisis — but it was virtually nonexistent in the Middle East.

Perhaps the most interesting hashtag was #FreeQuds. The Turkish pro-government Twittersphere used this hashtag frequently, and also contained disproportionately higher percentage of geolocated tweets. The #FreeQuds hashtag also contains evidence on bot-driven engagement, with regard to its diffusion patterns, as well as its rapid disappearance from digital agenda. #FreeQuds was sensitive only to one single peak — that of the July 18 afternoon protests. Another major Turkish hashtag, #MescidiAksa, was also mainly bot-driven and isolated in Turkey and parts of Europe with high presence of Turkish Diaspora.

Two of the most popular hashtags — القدس# (Al Quds) and فلسطين# (Palestinian territories) — emerged gradually from the Middle East (mainly the Gulf area) and they spread to North Africa by July 20, peaking with the July 21 incidents. The two Arabic hashtags were well-distributed along the Middle East, as well as Europe and parts of the U.S. Clustered in Egypt, Israel, Qatar and Kuwait, these hashtags had two peaks: on the Friday protests of July 21 and July 28, with the removal of metal detectors around Al-Aqsa and the Grand Mufti’s declaration that Muslims could return for Friday prayers.

Digital engagement and information-seeking trends

“Al-Aqsa” and its Arabic ‘الاقصى’ became significantly popular searches on Google and Wikipedia during the crisis, both peaking during the July 21 protests. “Temple Mount” was mostly prevalent in the United States. With added search terms, the crisis generated more interest on Benjamin Netanyahu than Mahmoud Abbas.

Bot-driven fake news during international crises tends to contribute to their escalation and prolongation. These factors may have real-life policy outcomes in a hypothetical future platform, which establishes real-time connections between legislators, officials and their constituencies. Imagine a congressional debate on foreign arms transfers or military intervention where legislators are live-connected to their voters as the two-way communication system gets flooded by extremely emotional fake news, sent by a legislator’s diaspora-heavy constituency.

Although digital technologies are improving political information and knowledge diffusion, they are also enabling digital spoilers and distractors to infiltrate global information flow systems. Although information spoilers and organized automated distraction have so far been studies in relation to domestic politics and elections, another effect lies in foreign policy and international crises.

Akin Unver is an assistant professor of international relations at Kadir Has University and a dual-visiting fellow at the Oxford Internet Institute and Alan Turing Institute in London.