Syria trends: Mining underused datasets

Ask Google

The Syrian tragedy has produced vast quantities of information, making it arguably the best documented conflict in history. While such information is indispensable for those seeking to document the conflict and untangle its many convoluted threads, its sheer volume can also be overwhelming—seeming, in a sense, to reflect the war’s chaos rather than making sense of it all. Synaps is exploring ways to extract meaningful insight from these mountains of data. Some options are deceptively simple, requiring little more than a web browser and the right mix of empathy and creativity.

A prime example is Google Trends, a tool that enables users to identify patterns in Google searches performed by people around the world. The system tracks keywords as they are “googled” over time, and presents data as a percentile of the strongest show of interest during the period considered. In Google’s own explanation: “A value of 100 is the peak popularity for the term. A value of 50 means that the term is half as popular. Likewise a score of 0 means the term was less than 1% as popular as the peak.”

Applied to the Syrian conflict, this tool can highlight social trends and point toward key junctures in the war’s trajectory. While the insights it offers are naturally imperfect and impressionistic, they can nonetheless prove a useful complement to other forms of research. Here are a few examples of relatively straightforward applications.

Syrian society goes dark

As the uprising kicked off in March 2011, Google searches in Syria reflected unprecedented inquiries regarding “virtual private networks” or VPNs, namely software that anonymizes the internet user, protects him or her from intrusive surveillance, and enables access to banned websites.

Interest in such software proved intense from mid-2011 to mid-2013, a period during which the regime gradually ramped up its suppression of a society that was not yet as geographically and politically fragmented as it is today. June 2011 marked a shift in the regime’s approach from tentative and ambivalent reform and repression to an all-out crackdown; June 2013 constituted another major turning point, with the massive use of chemical weapons in the suburbs of Damascus completing the process whereby President Bashar Assad essentially renounced any prospect of future reconciliation. It may be that, after that date, the segmentation of society was such that covert uses of the Internet declined-with Syrians either submitting to the regime or living in areas with which all bridges had been burned.

Incidentally, the sudden popular recourse to VPNs in the midst of the uprising was seeded by the regime itself in the period that preceded the conflict. In November 2007, the Syrian government officially banned the use of Facebook, citing security issues relating to Israel. In February 2011, it unblocked the social platform, either as a sign of goodwill in the face of mounting popular frustration across the Arab world, or to better equip itself to monitor such discontent, or both. The initial ban did less to deter Syrians than encourage them to explore VPNs. Specifically, they turned in 2009 and, even more vigorously, in 2010 to Hotspot Shield, software that remained popular during the uprisings.

The semi-public space

Regardless of escalating repression and widespread efforts to evade surveillance, Syrian adoption of social media followed patterns similar to—albeit slower than—those in less tightly contained societies. Whereas Western countries generally embraced as early as 2008 social platforms in general, and Facebook with more particular enthusiasm, the same process only kicked-in and accelerated in Syria as a result of the Arab uprisings: a relatively flat curve up to January 2011 became a steep upward slope, gradually eclipsing more traditional online sources of news such as al-Jazeera. The political trigger of this explosion in social media interest suggests the emergence of a “semi-public space” organized, at least initially, around the exchange of politically relevant information.

The elusive peace process

Google Trends offers few insights into rebellious patterns of behavior, due to self-censorship and VPNs. It does, however, present useful indications on how people in Syria respond to the conflict’s more innocuous forms of politics, notably peace talks.

The following chart tracks Syria-based Google searches for cities where key diplomatic events occurred in connection to the Syrian conflict. A striking peak of interest occurred in January 2014, as the United Nations hosted the “Geneva II” peace conference, which was designed to revive, update and operationalize the “Geneva I” agreement brokered in June 2012. Although Geneva II negotiations ended with no deal, they generated far more attention than the previous, more fruitful talks. A likely explanation is that Syrian society combined, at that point, a genuine eagerness for a political solution and some measure of hope that it could still be attained.

By contrast, Geneva III and IV, held respectively in January 2016 and February-March 2017, spurred both marginal and declining interest. The parallel “Vienna Process”—which as of October 2015 brought together key international players such as the United States, Russia, Turkey and Saudi Arabia—generated less curiosity still. The opposition conference held in Riyadh in December 2015 prompted a level of interest slightly superior to the launch of the Vienna Process. Finally, talks organised in the Kazakh capital Astana, which have drawn considerable coverage by mainstream media and become, de facto, the primary venue for Syrian peace talks, have left Syrians cold. Digging deeper into the data and its geographical distribution across the country suggests that what little attention they have garnered, unlike other rounds of negotiations, has been centred in Damascus.

Fear of escalation

A search for Valium or Diazepam—an anti-anxiety medication widely used to manage stress in Syria—brings up a relatively stable graph, with a striking anomaly: on 18 June 2017, interest surged, notably in the Damascus and Aleppo governorates. On that day, two events set eerie precedents: on one side, the US for the first time downed a Syrian warplane; on the other, Iran launched ballistic missiles into Syria, allegedly targeting the Islamic State, but more likely in a show of force aimed at deterring Washington.

Although many Syrians express that a certain numbness has set in surrounding the horrific events in their conflict, this recent spike suggested a still-untapped reservoir of anxiety at the prospect of things getting worse still.

Finding refuge

A search for different keywords suggesting a desire to leave Syria reveals trends that follow, predictably, the arc of the conflict. Syrians initially stayed put, showing, if anything, less of an appetite to leave their country than they did in the preceding period. Interest later grew as violence escalated and hope for the future plummeted. It culminated with the sudden summer dash to Europe in mid-2015, and dropped markedly thereafter—in reaction, no doubt, to radical European countermeasures and a general realization that this exit route was closed. Meanwhile, the spike of interest for the word “visa” in January 2016 is linked to a specific event, namely Turkey’s imposition of a visa regime on Syrians entering by air or sea.

For now, it would seem that Syrians who remain in Syria have largely given up on looking for emigration prospects, at least through Google searches—as opposed to offline networks and various forms of social media, which have developed into increasingly sophisticated tools for orchestrating emigration.

The data provides more unexpected insights also. People googling “ta’shira,” the formal word used for visa, or “Schengen,” happen to be concentrated in Damascus, where the most legitimate forms of travel seem to remain an option in the minds of those enquiring. Elsewhere the colloquial for “visa,” the keyword “emigration” as a generic query, “asylum” and even “smuggling” dominate.

Finally, the data highlights a somewhat counterintuitive seasonality: during the four years where trends were clearest (2012, 2013, 2014 and 2015), peaks occurred in September. This may correspond to the ominous approach of winter, or to a new cohort of young men who, not being registered in university, became susceptible to being drafted.

Again, the insights provided by these data are partial and subject to interpretation—but they are also highly suggestive, raising questions that can be further unpacked through qualitative fieldwork. Synaps will be continuing to explore how this and other open-source data tools can augment our own fieldwork-centric methodology.

31 July 2017

Illustration credits: graphs by Synaps, Syria on the globe by Wikipedia / licensed by CC.