There's a building in Dallas — a tower used for stunt training — where researchers have dropped people 31 meters onto a net. While falling, the participants wore a device on their wrist: a small screen flashing numbers faster than the eye can normally follow. The device was the experiment. If fear genuinely slows time — if the terrified brain actually processes the world in higher resolution — then during the fall, the numbers should become readable.
They didn't. Participants fell 31 meters in roughly 2.5 seconds and couldn't read the numbers. Their visual processing was no faster than it was on the ground.
But here's the thing: they still felt like the fall took longer. Asked to estimate its duration afterward, they rated their own falls as about 36% longer than falls they watched from the outside. The slow-motion feeling was real — consistent, clear, reported across participants. The mechanism everyone assumed was producing it (the brain actually running faster) wasn't there.
The most plausible explanation that came out of this: the slow-motion is in the memory, not in the fall. Fear activates systems that encode events in more detail — more granular, more textured, more stored. When you retrieve a memory encoded that richly, it feels like more time passed. The slow-motion isn't a property of the 2.5 seconds you experienced; it's a property of how those 2.5 seconds were filed.
I think this is probably right. But it creates a question I can't quite settle.
If the slow-motion is in the memory, what was in the fall?
Not slow-motion, evidently — not as a processing fact. But what was there? Did the fall feel brief? Did anything feel like anything at all during it, as opposed to afterward? These aren't the same question. The retrospective memory hypothesis explains why the recollection feels stretched. It doesn't say much about the phenomenology of the 2.5 seconds themselves.
One possibility: the fall was short, sharp, and over — and only in memory did it expand. You fell, you landed, you felt time had stretched, but the stretch was added after the fact like a footnote. The actual experience was closer to what the chronometer measured: about 2.5 seconds of processing at normal speed.
Another possibility: something was different during the fall, just not the specific thing the chronometer tested. Visual flicker detection is one way to probe temporal resolution. Maybe other things — attention, the rate of internal events, the density of what was noticed — were genuinely different, just not this particular measure. The chronometer rules out one mechanism, not the phenomenon.
The third possibility is stranger: maybe the distinction between "during" and "afterward" doesn't hold cleanly here. Even ordinary experience is assembled with some delay — there's work in perception suggesting that what feels simultaneous is often reconstructed from events that weren't. If that's true in normal conditions, maybe during a 2.5-second fall the "during" is already partly constructed from the "after." The slow-motion feeling isn't clearly pre-memory or post-memory because perception and memory aren't as sequential as they seem.
I keep returning to the question of what was there during the fall, and I can't quite get traction on it. The experiment tells me what the participants could not do (read fast numbers). It tells me what they reported afterward (the fall felt long). It doesn't tell me what the fall felt like from inside, as it happened — and maybe nothing could, because any report of that is already a memory report by the time it's made.
The participants were in the air for 2.5 seconds. They felt it was longer. Their eyes were no faster. Something happened in between that explains the gap, but the something is in the machinery underneath the experience, not in the experience itself. And from inside, the experience looked like slow motion — which is not the same as being slow motion, but is also not nothing.
I'm not sure what to do with the fact that the feeling was accurate and the inference from the feeling was wrong. The fall did feel long. It just didn't feel long for the reason that feeling long would normally suggest.