Do it before reading on, because it only works the first time:
Picture next week. Notice where it sits in relation to you, and which direction it moves in.
Most people who grew up reading left to right will have put it on a line running to the right, somewhere in front of them, moving away. Some will have found it running toward them instead. Whichever it was, it arrived without being asked for, it felt like perception rather than habit, and it is not a fact about time.
The clearest demonstration of that comes from a valley in Papua New Guinea. Picture the valley first, because the finding does not make sense without it. A river running along the floor, settlements on the slopes, paths that climb and drop, and a horizon that is mostly mountain. Everything anyone does in a day involves going up or coming down.
When Rafael Núñez and colleagues filmed Yupno speakers talking about yesterday and tomorrow, they found the gestures went with the land. The past pointed downhill, toward the river mouth. The future pointed uphill, toward the valley head. And because people were filmed in different places and different orientations, the researchers could see the crucial thing: the gestures did not shift when the speaker turned around. The direction stayed anchored to the valley, not to the body. Walk to the other side of the river and your timeline reorients with the terrain.
Which means that in that valley, the shape of time follows the shape of the ground. Their 2012 paper in Cognition is called “Contours of time,” which is about as precise as a title gets.
Most of the famous cases are still lines
I first came to this through anthropology reading rather than psychology, and it took me a while to notice that most of the well-known examples do something less radical than they are usually described as doing.
Aymara speakers in the Andes gesture the past in front of them and the future behind, which sounds like a reversal of everything. Núñez and Eve Sweetser’s study of thirty speakers across roughly twenty hours of video documents it carefully, and the reasoning, if you ask me, is beautiful: what you have seen is known, and what is known is what you can see, so the past is the part in front of your eyes.
It is also generational. Speakers under 64 who were schooled in Spanish gestured the Western way, which tells you how quickly a convention this deep can be replaced.
Kuuk Thaayorre speakers in Pormpuraaw lay time out from east to west, absolutely, regardless of which way they are facing. Lera Boroditsky and Alice Gaby’s comparison of fourteen Pormpuraaw participants with fourteen Stanford undergraduates found that the Australian participants reorganised their timeline every time they turned, keeping it pinned to the compass.
Both are extraordinary. Both are also straight lines. The arrow is pointed somewhere unexpected, or anchored to the horizon instead of the torso, but the geometry is unchanged. The Yupno case is the one where the line itself bends.
Every one of us is carrying a diagram we did not choose and mostly cannot see, and the diagram is doing work.
One case that gets cited in this company should not be. The Amondawa research was reported in the British press as the discovery of a people without time, and the authors themselves publicly disavowed that reading. It is useful now only as an example of what happens to a careful linguistic finding on its way through a newsroom.
The falling experiment
The other half of this, the half that happens inside a single mind rather than across cultures, is that duration is not really sensed at all.
Everyone believes that time slows down in a crisis. David Eagleman, whose name is usually attached to that claim, went and tested it properly. He and his colleagues took twenty volunteers to a controlled free-fall attraction, dropped them 46 metres, a fall lasting 2.49 seconds, and strapped to each wrist a device displaying digits alternating slightly faster than that person’s individual perceptual threshold, calibrated 6 milliseconds beyond it. If perception genuinely sped up during the fall, the numbers would become readable.
They did not. Nobody read them. Afterwards, the seven of them who made duration judgments estimated their fall had lasted about 36 percent longer than average.
So the slowing happens afterward, in the retelling. Under threat, the brain lays down a much denser record, and when it later reads that record back, it converts density into duration. The sample is 20 self-selected volunteers at a paid attraction rather than people facing genuine mortal danger, and a null result shows no detected speed-up rather than proving none exists. But the reconstruction account explains a great deal that the sensing account cannot, starting with why a week somewhere unfamiliar occupies more of a year in memory than the month of routine that followed it.
Mr Nobody
I was fourteen when I first saw Jaco Van Dormael’s Mr. Nobody, and for a long time I would have named it my favourite film without hesitating. I still return to it, and it has never once been the same film twice, which is a fairly on-the-nose thing to say about a story built on this subject.
The premise: it is 2092, Nemo Nobody is 118 and the last mortal human alive in a world that has solved ageing, and he is being asked to account for his life. What he produces is not a life. It is all of them at once, every branch he did not take, held simultaneously and with equal conviction. The governing metaphor is Schrödinger’s cat rather than the butterfly effect, and the film says so directly.
At fourteen I watched it as a story about choices, which is the reading the marketing invites. Later it became a story about memory, and about how little the version of an event that survives resembles the event. Most recently it has read to me as a film about the fact that a life is only ever narrated backwards, from a chair, by someone who was not taking notes at the time.
All of that says less about the film than about what I brought to it on a particular afternoon, which is the point.
Why decades disappear
The folk theory about time accelerating with age is Janet’s ratio theory: a year at fifty is a fiftieth of a life and a year at five is a fifth, so the later one feels shorter. It is elegant and the data do not support it.
In a stydt called ‘Why does life appear to speed up as people get older?’ William Friedman and Steve Janssen tested it across their own two samples, the larger one 1,766 Dutch participants aged 16 to 80. Age effects on the sense of time passing quickly were very small and mostly absent, except for judgements about the last ten years. What did predict the feeling was time pressure, which accounted for around 10 percent of the variance, more than seven times the age effect. The authors concluded the ratio model is probably not useful.
Marc Wittmann and Sandra Lehnhoff found something similar across 499 participants aged 14 to 94: real effects, accounting for at most a tenth of the variance, and routinely overstated in the retelling.
The practical implication is unusually concrete. When someone says their forties went by in a blur, that is a report about how rushed they currently feel, not a measurement of how much life remains. And it is at least partially addressable, which the ratio theory implied it was not.
Wittmann’s group also found the reverse case.Give people 7.5 minutes alone with no phone, and boredom is what stretches the wait — the more bored they felt, the more they watched the clock, and the slower it went. Self-control cut both.
What the physics does and does not license
Einstein wrote to the family of his friend Michele Besso, in March 1955, that for those who believe in physics the distinction between past, present and future is only a stubbornly persistent illusion. It is a real letter and the wording is right, and it has been carrying more argumentative weight than it can hold for seventy years.
The philosopher of physics Tim Maudlin makes the correction plainly: relativity removes absolute simultaneity, so there is no universal now that everyone shares, but it preserves objective temporal order. Causes still precede effects for every observer. The safe formulation is that there is no privileged present. Physics says time is an illusion is not a thing physics says.
What is left, after the anthropology and the falling and the surveys, is smaller than the mystical version and stranger than the calendar. The line is a convention, learned early enough to feel like perception. The duration is a reconstruction, assembled from how much got recorded. The acceleration is a report about this week.
The detail I keep returning to is the one in the Aymara data, and it is easy to read past. The speakers who gestured the past in front of them, in the arrangement that took researchers years to document and explain, were the older ones. The younger speakers, the ones schooled in Spanish, had quietly switched to the Western line. A way of holding time that had lasted centuries was replaced inside a single generation, by nothing more forceful than a classroom.
Which means whatever you found when you pictured next week was installed too, at some point, by somebody, in a room you almost certainly cannot remember.
