We still don't know how the brain turns a few jumpy snapshots per second into one smooth visual movie. But this paper gets us closer. Every time your eyes flick from a coffee mug to a friend's face, the image on your retina lurches around like a badly managed stock chart. And yet your experience stays stable. The brain is paying for that.
The new study by Kämmer and colleagues asks a simple question: before your eyes land on something, does your brain already pre-load some of that object's visual features into the foveal cortex - the bit of early visual cortex that usually handles what you are looking at dead center? In economic terms, this is hedging. Why wait for the market open if you can price the asset a few milliseconds early?
They tested this with a clever setup in an fMRI scanner. Participants prepared eye movements toward natural images shown in peripheral vision. But right before the eyes could land on the target, the image disappeared. So the fovea never actually got direct visual input from that object. If the foveal cortex still carried information about the missing target, the signal had to get there by feedback - top-down communication, not plain old incoming light.
That is exactly what they found. The researchers could decode which target people were about to look at from foveal regions of early visual cortex, even though nothing ever showed up there directly. The feedback cared about shape, not semantic category. So this was less "ah yes, a bird" and more "I am expecting this bundle of edges and contours." Your visual system, ever the thrifty accountant, seems to preload the receipt before the item arrives. Kämmer et al., 2026
Not Mind Reading - More Like Inventory Forecasting
This is where the paper gets fun. The result does not say your foveal cortex is secretly doing object recognition in advance. It says the system seems to push low-to-mid-level visual features back into the foveal map during saccade preparation. Think less psychic powers, more warehouse management.
That lines up with earlier behavioral work showing that the fovea seems to anticipate defining features of an upcoming saccade target before the eye movement finishes. In other words, the visual system does not just wait passively for the high-resolution input to arrive. It places a small bet first. Kroell and Rolfs, 2022
It also fits a larger literature on presaccadic attention: right before you move your eyes, perception at the future landing spot gets a temporary upgrade. That process overlaps with attention, but it is not identical to ordinary covert attention, which is a useful reminder that the brain enjoys running several overlapping systems with suspiciously similar names. Recent reviews and experiments have sharpened that distinction and linked it to different computations and circuits. Hanning et al., 2021, Li et al., 2021, Hanning et al., 2023
Why This Matters Outside the Scanner
If this result holds up, it gives us a cleaner mechanism for visual stability. Your brain may keep perception stable not by rebuilding the world from scratch after every eye movement, but by running a rolling prediction market. Peripheral vision supplies the teaser trailer. Higher areas place a quick forecast. Early foveal cortex gets primed with likely shape information. Then when the eye lands, the incoming signal meets a market that is already open.
That matters for everyday life because saccades are constant. Reading, driving, scanning a crowded sidewalk, and judging whether the thing on the stove is a pancake or a small disaster all depend on moving your eyes fast while keeping perception coherent. A system that can pre-position likely visual features could shave time, reduce confusion, and help the brain decide what changed versus what merely moved on the retina. It also gives vision scientists a more specific circuit to test when that sense of stability breaks down.
The Catch, Because There Is Always a Catch
This study used fMRI decoding, which is powerful but indirect. It tells us that information about the target was present in foveal cortex, not exactly how individual neurons carried it moment to moment. The feedback was also shape-sensitive rather than category-sensitive, so we are probably looking at an early-stage prediction, not a full object file with a name tag.
Still, the paper tightens the story nicely. The intraparietal sulcus tracked fluctuations in how decodable that foveal feedback was, making it a plausible upstream player in sending the forecast.
The big takeaway is simple: when your eyes prepare to jump, your brain may not be waiting for the next frame. It may already be penciling it in.
References
Kämmer L, Kroell LM, Knapen T, Rolfs M, Hebart MN. Feedback of peripheral saccade targets to early foveal cortex. eLife. 2026;14:RP107053. DOI: https://doi.org/10.7554/eLife.107053
Kroell LM, Rolfs M. Foveal vision anticipates defining features of eye movement targets. eLife. 2022;11:e78106. DOI: https://doi.org/10.7554/eLife.78106
Hanning NM, Deubel H, Li HH, Carrasco M. To look or not to look: dissociating presaccadic and covert spatial attention. Trends Cogn Sci. 2021;44(8):669-686. DOI: https://doi.org/10.1016/j.tics.2021.05.002. PMCID: https://pmc.ncbi.nlm.nih.gov/articles/PMC8552810/
Li HH, Pan J, Carrasco M. Different computations underlie overt presaccadic and covert spatial attention. Nat Hum Behav. 2021;5:1418-1431. DOI: https://doi.org/10.1038/s41562-021-01099-4
Hanning NM, Deubel H, Szinte M. Dissociable roles of human frontal eye fields and early visual cortex in presaccadic attention. Nat Commun. 2023;14:5381. DOI: https://doi.org/10.1038/s41467-023-40678-z
Disclaimer: The image accompanying this article is for illustrative purposes only and does not depict actual experimental results, data, or biological mechanisms.