Summary
A bioRxiv preprint reports that 19 human patients used the same cortical regions and oscillatory architecture during easy and difficult visual-search tasks. The researchers suggest that alpha rhythms adjust the speed of this shared circuit rather than switching between dedicated systems.
A bioRxiv preprint reports that fast and slow visual search can be produced by the same cortical architecture, with the brain changing how quickly the circuit operates. The result came from intracranial EEG recordings collected while 19 human patients performed two visual-search tasks: an easy search for a uniquely coloured item and a more difficult search based on an item's orientation.
The study, posted on September 15, 2026, challenges the idea that different search speeds require separate, dedicated neural systems. The authors report that both tasks engaged the same cortical regions, oscillatory frequencies and processing hierarchy. The manuscript is a preprint, so its findings are presented as research evidence before peer review.
What the researchers compared
Visual search requires selecting a target from other items in a scene. In the easier task, the target was a colour singleton—an item that stood out by colour. In the harder task, participants searched for a target defined by orientation, such as a differing angle among similar distractors.
These tasks produced a clear difference in performance speed. The researchers used intracranial EEG, which records electrical activity through electrodes positioned within the skull, to examine how neural activity unfolded during each search. Rather than finding separate sets of regions or frequencies for the two tasks, they observed a shared pattern across the measured cortical network.
A shared neural trajectory at different speeds
The researchers compared the timing of neural activity using dynamic time warping, a method that aligns two patterns when similar events occur at different rates. This analysis indicated that the activity sequence for one task could be treated as a time-stretched version of the other.
The same timing relationship also appeared within each task: faster and slower individual trials followed similar neural trajectories, but at different speeds. In the authors' interpretation, response speed therefore tracked the rate at which a common processing sequence unfolded, rather than identifying which of two separate circuits had been selected.
This distinction matters because it offers a different account of flexible behaviour. A fixed anatomical network could support multiple operating modes by changing its temporal dynamics. For the two search tasks studied here, the reported evidence favours that shared-speed-control account.
How alpha rhythms may set the pace
The proposed timing mechanism involved alpha-band oscillations at approximately 8–10 hertz. The researchers identified two coordinated roles for these rhythms.
First, alpha activity synchronised processing across a posterior cortical network involved in the search tasks. Second, it controlled the timing—or phase—at which local high-frequency activity in the 70–150 hertz range was most likely to occur. This phase relationship was reported at the same preferred point in the alpha cycle for both the easy and difficult searches.
Together, these observations connect the overall speed of the neural trajectory with coordination between brain regions and the timing of local activity. Alpha oscillations are therefore presented not simply as a signal that differs between conditions, but as part of a mechanism for regulating when processing progresses through a shared network.
What the result could change
The findings suggest that some apparent dual-process systems may be better understood as one circuit operating at different speeds. That possibility could influence models of visual attention and other forms of flexible behaviour, where distinct behavioural modes have often been linked to distinct neural circuits.
The evidence in this report is limited to 19 patients and the two visual-search tasks tested. The authors present applying the same principle to other search tasks—and to other domains in which separate processing systems have been proposed—as a testable prediction for future studies.