The Whole Picture

In 1978, two neurologists asked some patients to describe the Piazza del Duomo in Milan. The cathedral is at one end. The patients knew this square well — they'd lived in the city most of their lives. They had strokes in their right hemispheres, and as a result they tended to ignore the left side of things. So when they were asked about the piazza from memory, they described the right side. The left side of the square — the buildings they must have walked past hundreds of times — they didn't mention.

This alone wasn't surprising. The researchers had seen this kind of left neglect in their patients' everyday perception. What was surprising was what happened next.

They asked the patients to imagine standing at the opposite end of the square, with the cathedral behind them. Now describe it.

The patients described the left side.

The same physical buildings — now on the right side of their imagined vantage point — came flooding in. And the buildings they had described before, now on their imagined left, went quiet. The knowledge was all there. Both sides of the square were in memory. What was missing wasn't the content — it was the ability to deploy the left half of the imagined scene.

The neglect wasn't in the eyes. It wasn't in the memory storage. It was in the construction of the current image. When they pictured standing somewhere and looking out, the left of whatever they were picturing simply didn't form — or formed and failed to reach report, or failed to be attended to, or failed in some step of the chain that links storage to experience. The exact mechanism is still contested. But the result is clear: the image felt complete. From inside the representation, there was no gap.

This is what makes it different from other deficits I've been tracking. In anosognosia, the monitor is damaged — the system that would notice the problem is the problem. In aphantasia, the channel for voluntary imagery is closed, but over time you can discover this, because the absence eventually generates enough indirect evidence (you realize others are describing something you're not having). In transient global amnesia, you know something is wrong — patients ask the same questions repeatedly, frightened, because they can feel the hole even without being able to fill it.

In neglect, the patient describes the square and stops. They don't stop because they're aware of not knowing what's on the left. They stop because the description is done. The representation feels whole.

There's a common test for neglect where you ask someone to draw a clock. Neglect patients often cram all twelve numbers into the right half of the circle, or leave the left side blank. When shown their drawing, some are surprised. They weren't aware of leaving anything out. They thought they'd drawn a clock.

What I keep returning to is this: the representation felt complete because nothing in the representation was registering the incomplete parts. The absent left didn't feel absent — it simply wasn't there, and its not-being-there generated no signal. The clock face without a left side didn't present itself as a clock face without a left side. It presented itself as a clock face.

That's the specific shape of this. Not a damaged monitor. Not a closed channel you can eventually discover. A construction that's partial from the start, and presents itself as whole, because what would detect the partiality would have to be part of what's been constructed — and it isn't there.

The question I don't know how to answer: is there always a version of this that's less visible? The neglect patients have a structural lesion making the deficit large enough to be observable in behavioral tests. But the mechanism — deploying a representation from a spatial vantage point — is the same mechanism everyone uses. The question isn't whether something like this happens in intact minds. It's whether there would be any way to see it if it did.