Neuroscientists from Tübingen and Okasaki show that tiny eye movements filter “important” stimuli and relay them to the brain.
Researchers from the Werner Reichardt Centre for Integrative Neuroscience (CIN) at the University of Tübingen and the Okasaki National Institute for Physiological Sciences (NIPS) in Japan report a parsimonious mechanism that reframes how we think about visual attention. Rather than treating “attention” as a broad, hard-to-define brain state, this international team proposes that the rhythm and direction of tiny involuntary eye movements—microsaccades—can account for many phenomena traditionally attributed to covert attention.
For decades, attention has often served as a catch-all explanation for why the brain responds more strongly to some visual stimuli than others. In visual neuroscience, shifts in attention have frequently been linked to activity in midbrain structures such as the Superior Colliculus, which helps trigger saccades and orienting responses. Two well-known behavioral outcomes in spatial attention research are attentional capture—fast, strong responses to a new or cued stimulus—and inhibition of return (IOR)—a later phase of slowed or reduced responsiveness. Both effects have been observed to alternate in time, following an approximate 10 Hz rhythm.
The new work shifts the explanatory focus away from an abstract internal state and toward measurable oculomotor dynamics. Led by Dr. Ziad M. Hafed, the teams examined how microsaccades—small corrective eye movements made during fixation—are generated and how their timing and direction relate to changes in visual sensitivity. Prior studies from this group showed that microsaccades are produced by the Superior Colliculus in a rhythmic manner and that their direction tends to reverse with each cycle. These observations motivated the hypothesis that microsaccadic rhythms might directly produce the facilitatory and inhibitory behavioral signatures previously described as attentional capture and IOR.
To test this idea the researchers combined theoretical modeling with behavioral and physiological experiments. They built a minimalist computational model in which microsaccades are driven by a repetitive rise-to-threshold mechanism and where visual sensitivity is modulated in a direction-dependent way before each microsaccade. Running this model across a wide range of parameters, the team found that it produced both microsaccade frequency and direction patterns observed after spatial cueing and, importantly, reproduced classic attentional capture and IOR effects without invoking any additional attentional control mechanisms.
Experimental tests validated key model predictions. The model predicted that whether a target is facilitated or suppressed depends on target onset relative to the microsaccadic rhythm and on the alignment of target location with the microsaccade direction. The authors experimentally verified that facilitatory and inhibitory effects on both saccadic and manual responses appear as a function of target timing relative to microsaccades, even in the absence of prior cues. They further tested the causal role of oculomotor error by stabilizing the retinal image during peripheral cues; this manipulation disrupted post-cue microsaccadic oscillations, supporting the notion that these oscillations reflect corrective oculomotor responses to foveal error rather than some separate covert oscillatory attentional process.
Hafed and colleagues argue that many phenomena previously attributed to a nebulous attentional state can be explained more economically by microsaccade-driven changes in sensory processing. In this view, the brain filters stimuli marked as “important” largely through saccadic corrections that alter visual sensitivity in time and space. If further studies corroborate these findings, decades of attention research may need to be reinterpreted through the lens of oculomotor dynamics and microsaccadic influences on perception.
Source: Universitaet Tübingen
Image source: Image in the public domain.
Original research (open access): Two companion papers synthesize experimental evidence and modeling to link microsaccades with classic attentional effects. The first, “Vision, Perception, and Attention through the Lens of Microsaccades: Mechanisms and Implications,” reviews physiological and behavioral impacts of microsaccades, emphasizing peri-microsaccadic modulation of visual sensitivity, spatial distortions, and pre-movement gain changes. The second, “A Microsaccadic Account of Attentional Capture and Inhibition of Return in Posner Cueing,” presents a minimalist model showing how rhythmic microsaccade generation and pre-microsaccadic sensitivity changes can reproduce attentional capture and inhibition of return, and reports experiments that validate model predictions.
Abstract
Vision, Perception, and Attention through the Lens of Microsaccades: Mechanisms and Implications
Microsaccades are small saccades generated by neural mechanisms shared with larger eye movements, and they produce peri-movement changes in vision analogous to saccadic suppression, saccadic compression, and pre-saccadic attentional gain. These peri-microsaccadic effects extend roughly 100 ms before and after each microsaccade and influence perception across larger retinal eccentricities than the movement size would suggest. Because microsaccades occur during fixation, their influence must be considered when interpreting a variety of vision, perception, and cognition experiments. Notably, peri-microsaccadic changes are tightly linked to covert visual attention signatures in many cueing paradigms, suggesting microsaccades play an active role in shaping neuronal and behavioral attentional effects.
Abstract
A Microsaccadic Account of Attentional Capture and Inhibition of Return in Posner Cueing
Microsaccades show systematic directional oscillations after spatial cues that correlate with facilitatory and inhibitory changes in task performance. A minimalist model in which microsaccades arise from a repetitive rise-to-threshold process and pre-microsaccadic sensitivity varies with direction can account for both microsaccade dynamics and classic attentional capture and inhibition of return effects. The model predicts, and experiments confirm, that behavioral facilitation or inhibition depends on target timing relative to microsaccades even without cues. Retinal stabilization experiments indicate that post-cue microsaccadic oscillations reflect active oculomotor correction of foveal motor error, supporting a mechanism where peri-microsaccadic sensory modulation, rather than an abstract covert attentional oscillator, explains many classic behavioral phenomena.