Among the jargon that obfuscated what “functional” or “behavioral” optometrists were trying to achieve in the 1950s and 1960s was the mantra that “vision is motor”. While no doubt cerebellar research has accentuated the importance of motor control to the visual process and vice-versa, to equate vision with motor is self-limiting (and some might say unnecessary if not counter-productive).
Nevertheless, it’s always good to read research about the motor components of vision that continue to inform our understanding of the process in its entirety. In that regard, I thank Dr. Gary Williams for sharing this press release from the University of Rochester titled Small, involuntary eye movements help us see a stable world.
Our eyes are always moving, even when we are fixated on an image. “The human brain has a very precise knowledge of how the eyes move, even if humans are not aware of moving them, and they use this knowledge to infer spatial relations and perceive the world not as blurry but as stable,” says Rochester researcher Michele Rucci. (Getty Images photo / Carmelo Geraci)
Although the press release contains the hyperlink to the group’s published research in Nature Communcations, I’ll reproduce the hyperlink here as well to the article, Inferring visual space from ultra-fine extra-retinal knowledge of gaze position.
The VisionHelp Blog 
Very interesting study, and while I don’t refute that motor information is probably important for visual processing and function, a part of the experimental design I didn’t see questioned was whether or not the subjects received spatial information based on the moment the second vernier entered the aperture, as it would enter the aperture from the right or left and for (however infinitesimally) a small time be only partially visible on the right or left side of the aperture, unless I misunderstood the design, and the second vernier only became visible once it filled the entire aperture. I think this would qualify as spatial information from the retina and not purely extra-retinal information, again unless I misunderstood and they did control for this in the design.
That question appears to be addressed in the Discussion section, which states in part: “the stimulus duration was too short—more than one order of magnitude—to provide useful motion signals. Our custom display was designed to switch on/off within tens of microseconds (Fig. 2A), and the median displacement of the stimulus on the retina during the resulting brief exposures was only ~14 arcseconds, well below the thresholds reported in the literature for similar tasks. Results did not change when selecting only the trials with minimal displacement during exposure, and the instantaneous velocities measured around the times of bar exposures were only weakly correlated with perceptual reports (Fig. 4B). All these observations indicate that retinal image motion played no role in our experiments.” It would be interesting however for you to pose that question to the corresponding author, Michele Rucci, whose email address is: mrucci@ur.rochester.edu