Module 43: Semantic Construction Through Twitter in the Colombian Peace Process

Faculty Contact: John Mohr

Research Areas:

In October 2016, more than 6.4 million Colombians voted against peace accords that would end a fifty year war between the government and the FARC. As one of the most popular political parties in Colombia, Centro Democratico achieved this success through a well-organized and articulated campaign, tapping into ideologies and deeply held emotions. The campaign strategy was multimodal, but maintained a significant Twitter component that both reflected traditional propaganda and innovated to reach more tech-savvy voters; the party leader Alvaro Uribe Velez alone tweeted over 17 thousand times in 2016. Novel computational text analysis methods are used to extract semantic associations reinforced through party member Tweets in 2010 through the 2016 peace referendum. The method relies on word embedding algorithms analyzing thousands of Tweets to capture contextual properties that give meaning to words as they change over time. This work shows that constant reiteration of campaign messages emotionally grounded and formed associations between concepts surrounding the nature of the state, legitimacy of leaders, and characterization of the FARC as an inherently anarchic group. 140 character Tweets were well-suited to messages that lacked complex narratives and instead invoked emotions and ideologies that won the hearts and minds of Colombians.

Active Quarters:

  • Fall 2017 and Winter 2018: Devin Cornell