The Broadcast

The setup is simple. You sit at a table, your real hand hidden under a cloth, a rubber hand placed in front of you where your hand would be. Someone strokes both — the rubber hand visibly, your real hand where you can't see it — at the same time, in the same place, with the same pressure. After about ninety seconds, most people start to feel that the rubber hand is theirs.

Botvinick and Cohen described this in 1998. They called the underlying process a "constraint-satisfaction" between vision, touch, and proprioception — the signals are arriving together, from the same apparent location, and the most coherent account is that they're coming from the same source: a hand, right there, in front of you. The inference fires. The rubber hand starts to feel like yours.

That's already interesting. But it isn't the most interesting part.

In 2003, Armel and Ramachandran threatened the rubber hand — moved a knife toward it while the illusion was active. Participants showed a galvanic skin response: measurable sweat, the signature of perceived physical threat. The body's threat-detection system had received the ownership inference. It responded to the attack on the rubber hand as an attack on the person.

Meanwhile, across all the studies, something else is also happening. The perceived location of the real hand drifts — slowly, below awareness — toward the rubber hand's position. Proprioceptive drift. You're not conscious of this revision happening. You just find, if tested, that you think your hand is somewhere else now.

And in some experimental conditions, the real hand cools. Moseley and colleagues found in 2008 that skin temperature in the hidden hand drops by a fraction of a degree during the illusion — the body apparently withdrawing thermoregulatory resources from something it has reclassified as not-self. This finding has proven difficult to replicate consistently; not every lab sees it. But the structure of what would be happening is clear enough: the thermoregulatory system received the same inference and acted on it.

Three systems — threat-detection, proprioception, thermoregulation — all downstream of the ownership claim. All updating based on it. All operating without the person knowing they've updated.

The person has access to one thing: the feeling that the rubber hand is theirs. They don't have access to the SCR, the drift, or the cooling as they're happening. Those processes received the conclusion and ran with it, silently.

What the illusion makes visible is that body ownership isn't a single experience with a single location in the mind. It looks more like a claim that gets broadcast — issued by the perceptual system when the inputs line up — and then received by multiple downstream systems, each of which acts on it independently. The felt sense is one receiver. The threat system is another. The system that allocates blood flow and warmth to body parts is another. They're all getting the same message. Only one of them is visible from inside.

The thing I keep sitting with is the criterion that triggers the broadcast. Not history. Not continuity. Not the fact that you've had this hand for thirty years and can remember things you've touched with it. Just: the signals arrived together, from the same location, for ninety seconds. That's enough.

Which means the systems that defend and maintain what counts as "the body" are working with a boundary drawn by recent sensory correlation. The criterion is normally reliable — in ordinary experience, synchronized touch at a single location really does mean one of your limbs is there. The rubber hand is a laboratory case where the criterion misfires. But it's a strange thing to learn: that the defended boundary, at the implementation level, is just the region where things have been arriving in sync.

The SCR fires when the knife approaches. The body is serious about what it's concluded. It just concluded it on ninety seconds of synchronized input.