Conflict between visual motion and vestibular input is the main cause
Aliases: sensory conflict · neural mismatch · cybersickness · vection conflict
What it is
The person is sitting still; the virtual carriage accelerates: the eyes see optic flow receding, the canals and otoliths report zero acceleration. Two bodies of evidence for self-motion disagree. The most common sickness in a headset — stomach lift, sweat, yawning, wanting it off immediately — has visual–vestibular conflict as its main cause, the sensory conflict / neural mismatch of Reason and Oman. It is not too few polygons, and it is not “psychological.” While the conflict is on, the symptom path has fuel; take the conflict away (eyes closed, teleport, the world pinned again) and the path loses fuel.
Main cause is not sole cause. It explains why experiences of “looks like I am moving, body is not” are especially nauseogenic, and why a headset does this more readily than a small-screen film.
Why it happens
Self-motion is estimated jointly from visual flow, vestibular signals (rotation and linear acceleration), and proprioception. Evolution kept the three highly correlated: walking should show the ground streaming back, with matching forces at the feet and otoliths. A headset can drive vision alone: the whole retina is occupied by a virtual world whose flow writes “you are going forward,” while the chair and the inner ear write “you are sitting.” The mismatch is treated as a cousin of a toxin signal — historically, visual–vestibular havoc often followed ingestion — and the protective set of nausea, cold sweat, and saliva follows.
The conflict can also run the other way: a real head turn, the picture late or under-rotated, vestibule present, vision not caught up. That is delay and gain error, still the same mismatch. What matters is that the two channels are telling different motion stories. A cinema screen is a small field; vision is never taken as complete evidence of whole-body motion, so conflict is weaker. A headset makes vision into “I am in it,” and the vestibular veto becomes loud.
Studying it
The classic move is to change visual motion only, body still: an optokinetic drum, virtual forward acceleration, a scene yawing, collecting SSQ or VRSQ and time to abort. Controls are vision and vestibule aligned (a real rotating chair, real walking) or visual motion removed (eyes closed, a stable reference frame).
Independent variables: presence and direction of visual self-motion, whether the body is really moving, duration of conflict. Dependent variables: nausea and related symptom scores, physiology (skin conductance, electrogastrography), abort time.
Split “the picture is moving” (object motion) from “I am moving” (self-motion flow). Objects flying past need not fire this conflict; the whole environment moving is self-motion evidence. Playing a third-person cutscene is not a test of the same mechanism.
Where it stops holding
Postural instability theory holds that sickness comes from failing to keep the body stable, and that conflict is a correlate. The two theories predict the same thing on many stimuli, and the lab rarely keeps only one. “Main cause” means the path most operable in HCI and most often replicated, not closed physiology. Discomfort with no self-motion, only eye strain and pressure, should not be labelled cybersickness. People with vestibular damage show a different conflict pattern: some resist certain visual motions, some are more sensitive to residual conflict, and they are not a cap for general users. Short exposures, a small field, and an explicit “I am looking at a screen” frame downgrade vision as self-motion evidence and weaken conflict — a boundary of the mechanism, not “the content is well made.”
Applying it
- First map the moments when vision says “you are moving” and the body is not: virtual acceleration, smooth turns, being flown. Treat those moments as conflict sources, not as story peaks.
- Prefer methods that do not generate self-motion flow (teleport, a cut, letting the person walk) over stacking effects on top of conflict. Keeping continuous visual motion is an admission that conflict budget is being spent.
- Do not explain with “it looks so real, of course they get sick.” Realness is not the cause. Whether vision and vestibule are telling the same motion story is.
- How to check: replace a stretch of continuous forward motion with a teleport while the body stays still, leaving other image quality untouched. If abort rate and nausea fall as the conflict is removed, the main cause is being explained by this path. If they do not fall at all, look at delay and at fatigue outside conflict; do not add “more real” flow.