Communication scientist Kasper Welbers has been awarded a Vidi grant by the Dutch Research Council (NWO). In his project, 'Reading Different Realities: How Do You Write News for a Polarised Society?', he will investigate how different groups interpret the same news and what role journalists’ language choices play in that process.
One news story, different realities
News shapes how we see the world, but the reverse is also true: our existing beliefs influence how we interpret the news. Especially on polarised issues, people often look, consciously or unconsciously, for an interpretation that fits what they already believe. A varied media diet is therefore not automatically enough to reduce polarisation.
Welbers will investigate how journalists can communicate the core message of a news story as effectively as possible to audiences with different beliefs. Through large-scale research into Dutch news, he aims to map how different groups read the same stories and how language choices influence those interpretations. “I want to understand which formulations can help people with different worldviews still take away the same core message from a news story.”
Polysemic content analysis
To do this, Welbers is developing a new method: polysemic content analysis. Rather than assuming that a text has one fixed meaning, the method maps how different groups can understand the same text differently. “In traditional research, you try to measure texts as unambiguously as possible. Differences in interpretation can then quickly be seen as a problem for the measurement. In my research, those differences are the starting point: I want to understand which groups read the same text differently, and why.” This will allow him to identify which formulations lead to different interpretations and which language choices help preserve the core message of a news story across diverse audiences.
A tool for journalists
Together with journalists, Welbers wants to translate the findings into practical guidelines for what he calls ‘polysemic competence’: the ability to use language in such a way that an intended message comes across as clearly as possible to a large and diverse audience. He will also investigate which strategies are both practical and desirable in journalistic practice.
In addition, he is developing a practical tool that uses advanced language models to predict how different audience groups are likely to interpret a text. The models will be trained and tested using data from the project. The aim is to help journalists gauge how susceptible a text is to unintended interpretations that may obscure or distort its core message.