Our research approaches language from a social networks perspective. We show how individuals’ social network structure influences how good individuals are at understanding others and at expressing themselves. We also examine whether one of the reasons that languages differ from each other is because they are spoken by communities with differnet social structures. We examine these questions using a combination of individual differences, experimental, and computational methods, and across different linguistic levels.

LATEST NEWS

Check out our latest papers:

• Lev-Ari, S., Stamp, R., de Vos, C., Yano, U., Nyst, V., & Emmorey, K. (2025) The relationship between community size and iconicity in sign languages. Cognitive Science, 49, 6, e70074.

TDLR Communication is harder in larger communities, so larger languages are often easier to learn and use, to overcome the difficulty. For example, larger spoken languages  are more sound symbolic (see my paper from 2021!). This paper tests whether large sign languages are also more iconic than smaller ones, and finds that they are, at least with regards to non-social signs.

• Lev-Ari, S. (2025) Sorries seem to have the harder words. British Journal of Psychology
  

TDLR Apologies are cheap talk. How can apologizers make them more convincing? By making them more costly. Study 1 shows that people use longer words when apologizing. Study 2 shows that people perceive apologies with longer words as more apologetic. 

• Rühle, M-C. & Lev-Ari, S. (2025) Do native and non-native speakers make different judicial decisions? Bilingualism: Language &and Cognition, 28, 1, 146-153.

TDLR Speakers are less emotional in their foreign than native language. We show that this leads non-native speakers to be less sensitive to emotional mitigating circumstances when making judicial decisions.

• Lev-Ari, S. (2024) The influence of community structure on how communities categorize the world. Journal of Experimental Psychology: Learning, Memory, and Cognition, 50, 8, 1249–1264.

TLDR Languages differ in how they categorize the world. The paper shows that larger communities create more expressive categories that allow them to communicate more successfully. It shows this via agent-based simulations over different types of meaning spaces.

Read and listen to recent media coverage of our research:

• Science Quickly (the podcast of Scientific American), October 2025 – The Science of a Convincing Sorry
• The Guardian, September 2025  – Long-winded apologies seem more sincere, study suggests
• The Times , Septemeber 2025- Repentant, not sorry: why bigger words are better for apologies
• Radio 4 (BBC) PM show with Tom Baker, September 2025 – (starts at 51min:30sec)
• Radio 5 (BBC) Late night show with Tony Livesey, September 2025 –  Interview (starts at 1h:26min)
• Unexpected Elements (BBC radio show), September 2025 – about the sounds of swear words (starting 31:00)
• Wissen aktuell (Austrian science radio show), September 2025 (in German)
• Radio Ulster (daily pm show), October  2025
• Vanity Fair Italy, July 2025 (in Italian) – Come chiedere scusa in modo efficace? Meglio usare parole lunghe e di uso comune
• The Guardian, June 2025 – How sorry are you? Why learning to apologise well could save your relationships
• The Conversation, May 2025 – How to make your apology more effective – new research

Upcoming presentations:

03/12/2025 – Ludwig Maximilian University of Munich (Munich, Germany)