About the project
“Decoding Antisemitism” is an interdisciplinary and transnational research project that examines online antisemitism, focusing on the political mainstream of selected European societies — specifically the UK, France, and Germany — as well as the US. Antisemitic discourse on social media offers valuable insights into the present and future of this ideology of hate, which, due to its adaptability, infiltrates all social strata and is currently reaching new heights. This surge is driven in part by the unique nature and communication conditions of digital communication.
To effectively analyze antisemitism online, the research design must account for both its nuanced manifestations and its broader dissemination. Consequently, the project employs a multi-stage analytical approach.
In the first step – qualitative content analyses – data sets are examined and categorized by experts in linguistics, discourse analysis, and antisemitism research. The data comprises user comments responding to reports by leading media outlets that have the potential to trigger antisemitic reactions. These comments are categorized based on their antisemitic content and the linguistic or visual patterns used to convey them. In mainstream political milieus, antisemitic statements are often implicit or disguised through puns, allusions, irony, or rhetorical questions. Comparing the content and structural aspects of a comment is therefore crucial to uncover its antisemitic meaning and to identify broader trends in the social acceptability of certain antisemitic concepts.
To facilitate the analysis of antisemitic discourse at the level of both content and structure, a code system and a guidebook has been in development since the project’s inception. These tools enable coders to document, organize, and define the wide variety of antisemitic concepts and their linguistic and visual expressions. The guidebook ensures antisemitic content is identified consistently and allows for cross-national comparability.
In the second step, the analyzed data is used to train models that imitate the decisions of our experts using a supervised machine learning approach. These models learn to recognize how antisemitic concepts are linguistically articulated in specific online contexts. The automated classifications are continuously reviewed and refined by our teams to improve the accuracy of the classifiers.
Step 3 involves complementary quantitative analyses, capturing antisemitism by measuring the frequencies and combinations of cataloged words and phrases across various datasets.
Integrating the results from these three steps allows us to describe and understand the manifestations of antisemitism and their communicative dynamics in online debates with far greater precision than would be possible using any single method alone.
The project’s findings will be shared not only with the academic community but also with stakeholders in politics, media, education, law, and security. Results will be disseminated through workshops, lectures, regular discourse reports, and other open-access publications available on our website. The insights gained into how antisemitism manifests online today—and how it may evolve in the future—will play a key role in shaping effective counter-strategies.
The project is funded by the Alfred Landecker Foundation.
