Researchers at Université de Bordeaux and UCLouvain found that pupil dilation indicates people begin solving simple arithmetic problems before all numbers are presented [1, 2].
This discovery challenges previous assumptions about how the brain processes mathematical information. It suggests that the human mind does not wait for a complete set of data before initiating a solution, a finding that could alter the understanding of cognitive load and mental processing speed.
The study, released in April 2026, involved collaborative efforts between the Université de Bordeaux in France and UCLouvain in Belgium [1, 2]. The teams said they aimed to better understand the mental processes that underpin simple arithmetic [1, 2]. By monitoring pupil dilation, the researchers were able to track the precise moment the brain engages with a mathematical task.
Physiological changes in the eye often correlate with mental effort. In this instance, the dilation occurred as soon as the problem began, rather than after the final digit was revealed [1, 3]. This indicates a "shortcut" in mental math where the brain anticipates the operation based on the initial numbers provided.
The research team said that this preemptive solving occurs during basic arithmetic [1, 2]. The data suggests that the brain begins calculating in real-time as information arrives, rather than storing all digits in short-term memory before starting the computation [1, 2].
This method of observation allows scientists to see the "invisible" start of a mental process. Because pupil dilation happens involuntarily, it provides a more accurate timestamp of cognitive activity than a participant's verbal response [1, 3].
“People start solving simple arithmetic problems before all numbers are presented.”
This study suggests that human cognition is more proactive and less linear than previously thought. By demonstrating that the brain initiates problem-solving before a data set is complete, the research highlights a level of predictive processing in basic mathematics that may have implications for how educators teach math and how AI models simulate human thought processes.
