In today’s digital world, news reaches us through websites, social media feeds, and algorithmic recommendations. While news is still expected to present objective facts rather than subjective opinions, it is increasingly common to encounter messages containing subjective evaluations, especially on social media. This dissertation investigates how subjectivity is expressed in British online news and how readers perceive it. Computational methods are used to analyze the language of news articles and Facebook posts from major British outlets and compare levels of subjectivity across platforms and source types. The results show that increased use of subjective language on social media is mainly characteristic of popular news outlets, while quality outlets tend to maintain a more objective reporting style. Experimental studies demonstrate that habitual readers are better at distinguishing between subjective and objective news content. While subjective framing does not necessarily increase readers’ intention to engage with the content further, it comes at a cost: news outlets that use subjective language in reporting are perceived as less trustworthy. Moreover, when subjective evaluations appear in quotations, some readers interpret them as reflecting the journalist’s own stance. Together, these findings show how subjectivity operates in today’s platform-driven news environment and how it shapes readers’ judgements.
Elena Savinova (1997) obtained her Bachelor’s degree in Fundamental and Computational Linguistics from HSE University Moscow, followed by a Research Master’s degree in Linguistics from Utrecht University, where she began working on subjectivity in language. In 2021, she started her PhD at Radboud University, focusing on subjectivity in British online news. Her doctoral research combines computational linguistics, corpus analysis, and experimental methods to examine how subjective language is expressed across news platforms and how it influences reader perceptions. Elena has presented her work at leading international conferences in computational linguistics and has published in established journals in linguistics and media studies. Her research interests include discourse processing, stance attribution, and the role of language technologies in shaping how people interpret and respond to information in digital communication.