They breathed pure oxygen in 1987 and yawned just the same

In 1987, at the University of Maryland, Robert Provine closed a door that everyone keeps opening. He sat a group of students down breathing through a mask and changed the gas on them without saying which was which: ordinary compressed air, one hundred per cent oxygen, and air loaded with carbon dioxide at three and at five per cent. The yawns came out the same in every round.

Their breathing did change. With the carbon dioxide those students were breathing considerably faster, which is what happens to anyone whose air is enriched with CO2. And there is the result: if yawning served to correct the mixture, pure oxygen should have switched the yawns off and carbon dioxide should have set them going. It did neither. Provine tried again by another route, putting the volunteers through exercise until their breathing rate —the number of breaths a person takes per minute— had doubled, and the yawn counter did not move either. Breathing and yawning share a mouth and not much else.

The five per cent carbon dioxide coming through that mask is more than a hundred times what is in the air of the room where you are reading this, which runs at about 0.04 per cent. That is the dose that failed to produce a single extra yawn.

What started in 1987 is a forty-year-old question nobody has closed. With oxygen ruled out, yawning was left without a known function, and the only solid thing since then is the map of when it happens. It piles up around changes of state: the hour after waking, the hour before going to bed, the stretch where a class starts to drag. It does not show up when air is genuinely short, which is precisely where the old explanation expected it.

The candidate with the most supporters today was put forward by Andrew Gallup in 2007: yawning would cool the brain. The deep intake and the jaw stretch move blood and air through the sinuses of the face, and that would bring the temperature in there down by a few tenths of a degree. It has a couple of odd observations in its favour. In Gallup's own experiment, volunteers given a cold pack to hold against their foreheads yawned contagiously 9 per cent of the time, against 41 per cent among those holding a warm pack or none at all. And people yawn less when it is very hot outside, exactly when ventilating with outside air would stop cooling anything at all. It is an open hypothesis, not a fact.

In the meantime, the one thing yawning does beyond dispute is spread, and it spreads in writing too, with nobody in front of you opening their mouth.

In 2012 a team at Durham University recorded fifteen healthy foetuses on four-dimensional ultrasound, four times each, at 24, 28, 32 and 36 weeks of pregnancy, and counted them yawning. They told yawns apart from simple mouth openings by how the time split between opening the jaw and closing it, and they found the foetuses yawned more the smaller they were. A foetus does not breathe air: its lungs are full of fluid and not a single molecule of oxygen comes in through the mouth. Fifteen foetuses, fifty-eight usable scans, zero breaths and yawns on the count.