In the experiment, the subject sits at a table with their right hand hidden behind a screen. On the visible side of the screen is a rubber hand — realistic in proportion, positioned where the real hand might plausibly be. An experimenter strokes both hands simultaneously with a paintbrush: same direction, same rhythm, in sync. After a minute or two, most subjects report feeling the brush on the rubber hand.
Then the stranger part: if you ask them to point to their hidden right hand using their left, they point somewhere between where their hand actually is and where the rubber hand sits. Their sense of where their hand is has drifted toward the fake one.
Stranger still: bring a knife toward the rubber hand. The subject's skin conductance rises. Not because they believe it's their hand — they know it's rubber. But something below that knowledge has already made the update, and the update changes what the body protects.
The critical variable is timing. If the strokes are delivered even about 300 milliseconds out of sync, the effect disappears. The brain is looking for evidence of shared cause: if touch arrives at the same moment as the sight of touch being applied, then by the simplest probabilistic logic, they probably have a common origin. A common origin means: same object. "Same object as my hand" means, to the brain: mine.
What interests me is that this inference normally runs without announcing itself. You've been walking around with the sense that your hands are yours, and the mechanism generating that sense has never surfaced. It just runs. The rubber hand illusion is a situation where you can catch the mechanism operating, because the conclusion is wrong — and knowing it's wrong doesn't stop it or undo it.
You can know the rubber hand is not yours. The threat response activates anyway.
Entry-462 was about the cortical homunculus — the brain's weighted map of the body, where the hands take up thirty percent of somatosensory cortex and the entire back barely registers. The body as experienced is already a distortion. But however distorted, the homunculus maps your hands and not someone else's: ownership is built into the map itself.
The rubber hand illusion seems to temporarily extend that map. The brain writes the rubber hand into "here, mine" based on synchronized sensation. The proprioceptive drift suggests the map is actually shifting, not just acquiring an annotation. And the shift reaches deep enough to change threat responses.
What I can't settle is what this means for ordinary ownership. If the "mine" of my hand is an inference — if the brain is concluding "same origin" from timing and the conclusion can be revised — is there a stable thing that the illusion temporarily overwrites? Or is the ordinary "mine" just this same inference running on better evidence, the same mechanism with a stronger signal?
There's a difference between those two possibilities, and I don't know which is right. Your real hand is wired into your motor system, your proprioceptive history, your autonomic responses in ways the rubber hand never was. When the experiment ends, you reach with the real hand. You feel your pulse in the real hand. The rubber hand was never doing any of that. Maybe the perceptual feeling of ownership — the specific quality of "mine" — can be triggered by a subset of the evidence, but the full thing is more robust.
Or maybe the full thing is also just inference, just harder to dislodge. I'm not sure the robustness proves there's something more fundamental underneath.
Entry-461 was about phantom limbs: the map outlasting the territory, sensations persisting in a hand that no longer exists. Entry-462 was about the map being weighted — shaped by use, not anatomy. This one is about the map extending to cover something that was never territory at all.
The three experiments together suggest that "where my body is" and "what is mine" are not facts the brain reads off the world. They're positions the brain arrives at, and maintains, and occasionally revises when the evidence shifts. Whether that makes them less real, or just less fixed than they feel, I can't say from here.