Stop Hate Speech explores the spread of hate speech online and possible courses of action for dealing with hostilities and insults that target a person or group based on their identity.
Many studies point to an increasingly serious problem in online debates: hate speech, a term that refers to derogatory or discriminatory language about people or groups based on their identity, for example their gender, ethnicity, or religion. A disproportionately high number of online hate speech incidents target women or members of ethnic minorities. Online hate speech often has severe offline consequences for those affected. And, although hate speech generally concentrates on individuals, it also has a far-reaching impact on the communities targeted and on society as a whole.
Research and opinions differ on the best strategies to curb hate speech. Moderation of posts and comments by social media platforms and traditional media or legal provisions stipulated by governments may be effective, yet at the same time infringe on freedom of speech. ‘Counterspeech’ is an alternative tactic that involves politically active people presenting a counter narrative to put a stop to hate speech without constraining freedom of speech.
In the Stop Hate Speech research project, randomized field experiments investigate the effectiveness of various counterspeech strategies in combating hate speech on the internet. The research team first developed methods based on artificial intelligence to detect hate speech automatically. In the second project phase, users engaging in hate speech encountered one of the counterspeech strategies. One of those tactics attempts to spark empathy with those affected. Another strategy highlights the offline consequences that such online behavior can have for users who post hate speech.
The experimental set-up, designed to ensure the counterspeech strategies were allocated at random, allowed the researchers to calculate precisely how much the various interventions influenced perpetrators’ future behavior. This included appraising any changes in the frequency or toxicity of hate speech tweets, as well as noting deletion of such tweets and the extent to which other users also adopted counterspeech methods. The results of an initial experiment revealed that empathy-based strategies in particular can provide an effective way to combat hate speech.
To create digital posters linked to Stop Hate Speech, students from the Interaction Design BA at Zurich University of the Arts (ZHdK) used word play with anagrams. The aim was for viewers to interact with the posters in public spaces.
In small teams, ideas about "Stop Hate Speech" were researched, discussed, and provoked through playful design and creation techniques. The students picked examples of hate speech from social media and transformed the original intention by shifting the letters around to form an anagram. In addition, they also selected terms related to these issues and juxtaposed them with anagrams of those words. HATE becomes HEAT, GARCE (bitch) turns into GRÂCE, HURE (whore) is transformed into RUHE (calm).
The research work of UHZ and ETH served as a basis for the exploration and implementation of the interactive posters. Anagrams of hate speech were formulated to recontextualise them and subvert their original meaning, before being further investigated through a form-finding process. Students explored issues of reduction, expression, conciseness and boldness to realize posters that interact with the viewer. After the analogue design process, the interactive posters were programmed in Javascript.
Team Science:
Project leaders: Prof. Dr. Karsten Donnay, Prof. Dr. Fabrizio Gilardi, Department of Political Science, UZH, Prof. Dr. Dominik Hangartner, Public Policy Group, ETH Zurich
Project managers: Dr. Philip Grech, Selina Kurer, Public Policy Group, ETH Zurich
Postdocs: Dr. Laura Bronner, Dr. Gloria Gennaro, Public Policy Group, ETH Zurich, Dr. Ana Kotarcic, Department of Political Science, UZH
PhD: Maël Kubli, Department of Political Science, UZH
Predoc: Laurenz Derksen, Public Policy Group, ETH Zurich
Data scientist: Dr. Bruno Wüest, Forschungsstelle Sotomo
Auxiliary assistants: Osama Abdullah, Department of Political Science, UZH, Nick Bachmann, Sotomo Research Center, Maxime Bataillard, Florian Curvaia, Public Policy Group, ETH Zurich, Florian Eblenkamp, Marc Eggenberger, Stephanie Fierz, Department of Political Science, UZH, Rachel Kunstmann, Public Policy Group, ETH Zurich,
Felicia Mändli, Fabio Melliger, Mattia Mettler, Paula Moser, Alexandra Nagel, Jonathan Progin, Jennifer Roberts, Department of Political Science, UZH
Intern: Cyril Bosshard, Sotomo Research Center
Team ZHdK:
Module leaders Reactive Signs: Luke Franzke, Rebecca Morganti-Pfaffhauser, Interaction Design, ZHdK
Design Interactive Posters: Lukman Aščić & Audrey-Meret Lohmann, Mo Bünzli & Carina Good, Lars Ziegler & Matthias Naegeli, Sonja Cowley & Giovanna Yanireth León Briceno, Loïc Hommel & Nanthatchaporn Janthasom, Luis Praxmarer & Tanja Landolt, Elena Walther & Lea Bischoff, Lyvia Muniz Gomes Wägli & Benjamin Eggstein, Students Interaction Design, ZHdK
Team Civil Society:
Project leaders: Sophie Achermann, Kathrin Bertschy, alliance F
Project staff: Morgane Bonvallat, Sasha Rosenstein, alliance F