Researchers Sara Carta, Emina Aličković, Johannes Zaar, Alejandro López Valdés, and Giovanni M. Di Liberto published findings in PLOS Biology on July 16, 2026 revealing how the human cortex manages attention switching in multi-talker environments. Using EEG recordings from normal-hearing adults cued to switch attention between speech streams every 15–30 s, they measured neural encoding via Temporal Response Functions (TRF).

The study uncovered asymmetric disengagement and engagement processes: neural tracking of the new target stream emerges before disengaging from the previous target, demonstrating transient simultaneous encoding of two competing speech streams. This transition correlated with a reduction in EEG alpha power, indexing cognitive effort during attention switches.

The team further isolated cortical activity reflecting lexical prediction mechanisms, comparing four context-accumulation strategies constructed using Large Language Models. Results suggest listeners perform a reset in lexical context after switching attention, elucidating both temporal and contextual mechanisms underlying auditory attention shifts in complex listening environments.