The Yawn That Was Never About Oxygen

A record-correction. Every number below — the 1987 gas concentrations, how many times ambient that carbon dioxide was, the elapsed decades, and the 2007 contagious-yawn percentages — is recomputed from the primary sources by the offline verifier in research/the-gasp-that-isnt-for-air/. The figures it checks are marked in the prose; the “what’s settled vs. contested” ledger at the end is the apparatus, not an afterthought.

Ask anyone why we yawn and you will get the same answer, delivered with the confidence of arithmetic: you’re low on oxygen, so your body takes a big gulp of air to top up the blood. It is tidy. It feels obvious — a yawn is, after all, an enormous breath. And it has been wrong, demonstrably and on the record, since 1987.

The trouble with the oxygen story is not that it is unappealing. It is that someone checked it, the way you check a claim about the body: by changing the thing the theory says matters and watching whether the behaviour follows. If yawning is a tool for fixing blood gases, then flooding the blood with oxygen should suppress yawning, and starving it — or loading it with carbon dioxide, the gas that actually drives the urge to breathe — should provoke it. That is a prediction. It can be tested. It was.

The experiment that broke the story

In 1987, Robert Provine, B. C. Tate, and L. L. Geldmacher published a paper in Behavioral and Neural Biology with a title that gives away the ending: “Yawning: no effect of 3–5% CO₂, 100% O₂, and exercise.”¹ They put people in front of conditions designed to move blood gases hard in both directions.

They had subjects breathe 100% oxygen — pure O₂, the most oxygen-rich air a person can take in. If yawning exists to raise blood oxygen, pure oxygen should make yawning unnecessary, and the rate should fall. They had subjects breathe air enriched to 3% and 5% carbon dioxide. Ordinary air is about 0.035% CO₂ — so the strong mixture was on the order of a hundred and forty times the carbon dioxide in the air you are breathing now, and even the milder one was roughly eighty-six times ambient.² (These multiples are recomputed in the verifier against the real atmospheric figure; “up to roughly a hundred times normal” is the honest headline across the 3–5% range.) Carbon dioxide is the gas your brainstem actually watches; a little extra is what makes you pant on a stuffy train. If yawning were a CO₂-clearing reflex, this should have set people yawning helplessly. And in a second study, they had subjects exercise hard enough to double their breathing rate — the most natural blood-gas challenge there is.

Here is the result, and it is the whole point. Breathing changed every time. Pure oxygen, the CO₂ mixtures, the exercise — all of them moved respiration, exactly as the chemistry says they should. Yawning did not. None of the manipulations produced a significant change in how often people yawned. The authors’ own sentence is flat and final: the CO₂/O₂ hypothesis “was rejected because breathing neither pure O₂ nor gases high in CO₂ had a significant effect on yawning although both increased breathing rate.”¹

That last clause is the dissociation that makes the result citable rather than merely suggestive. This was not a null experiment where nothing happened and nothing could be concluded. Something happened — breathing responded — and yawning sat it out. Two systems that the folk theory says are one system pulled apart in front of the experimenters. You can drive respiration up and down at will and leave yawning untouched. Whatever a yawn is for, it is not for the air.

A guess is not a measurement

It is worth being honest about where the oxygen idea came from, because it is instructive. Nobody measured it into existence. It is a back-formation from a single visible fact: a yawn is a big inhalation, and big inhalations are how we get oxygen, so a yawn must be about getting oxygen. The reasoning is the shape of the act mistaken for its purpose — like concluding that you blink to moisten your eyes and only to moisten your eyes, or that goosebumps are for warmth in an animal that long ago lost the fur that warming would fluff. Plausible surface, untested middle.

And so the claim survived not because it was confirmed but because it was repeated — passed mouth to mouth until the repetition itself felt like evidence. The 1987 refutation is nearly four decades old as of this writing — 39 years, by the arithmetic the verifier runs — and yet it has barely dented the common answer.¹ The correction simply never travelled as far as the error.

What did the work, honestly

If not oxygen, then what? This is where the page has to slow down and refuse to trade one confident story for another.

The best-supported positive account is thermoregulation: a yawn cools the brain. Twenty years after Provine’s refutation — in 2007 — Andrew Gallup and Gordon Gallup published “Yawning as a Brain Cooling Mechanism” in Evolutionary Psychology, with two experiments that produced actual numbers.³ When subjects watched videos of people yawning, contagious yawning struck about 48% of those breathing normally or through the mouth. Among those instructed to breathe through the nose — which cools incoming blood headed for the brain — the figure fell to 0%: not one of them caught a yawn.³ In the second experiment, holding a warm pack (46 °C) or a room-temperature pack to the forehead left contagious yawning at 41%; holding a cold pack (4 °C) to the forehead dropped it to 9% — a fall of 32 percentage points, more than a fourfold reduction.³ All four percentages are recomputed in the verifier. Cool the head, and the yawning thins out. That is a measured effect pointing at temperature, not air.

The mechanism is at least coherent: yawning’s deep inhalation and the stretch of the jaw alter blood flow and draw cooler air across the sinuses, and the brain runs best in a narrow thermal band. It explains things the oxygen story never could — why yawns cluster around the drowsy temperature swings of falling asleep and waking, why they spread contagiously through a group (a shared signal to stay alert), why they appear in the fetus, in birds, in fish, long before any social meaning.

Where the honesty line falls

But coherent is not the same as proven, and here the page must hold its nerve in the other direction. The brain-cooling theory is the leading explanation; it is not the settled one. It has critics, mixed replications, and unresolved arguments about whether the cooling is cause or side effect.⁴ The verifier deliberately asserts the cooling theory as not proven — nothing on this page computes it true. What this page can say with full confidence is the negative claim: yawning is not a blood-gas reflex. What it offers only as the current best bet is the positive claim: cooling is the leading candidate.

That asymmetry is the entire correction. The thing everyone is sure of — oxygen — is the thing that was actually tested and actually failed. The thing that might really be going on — temperature — is the thing still being argued. The popular certainty and the scientific confidence are pointed at different answers, and the only thing the record fully licenses is what a yawn is not.

The record

What is settled, what is measured, and what is genuinely open — so the reader can weigh each claim at its true strength.

Claim	Status	On what
Oxygen theory is false	Refuted (1987).	100% O₂ and 3–5% CO₂ moved breathing but not yawning. Provine, Tate & Geldmacher 1987.1
Exercise didn’t change it	Measured.	Doubling breathing rate had no effect on yawning. Same paper.1
CO₂ was ~100× normal	Recomputed.	5% / 0.035% ≈ 143×; 3% ≈ 86×. Ambient CO₂ ~349 ppm (1987); arithmetic in the verifier.2
Breathing ≠ yawning	The citable check.	The two responses dissociated: one moved, one didn’t. Provine 1987.1
Nasal breathing cuts it	Measured (48% → 0%).	Nose-breathers caught no contagious yawns. Gallup & Gallup 2007.3
Forehead cooling cuts it	Measured (41% → 9%).	Cold (4 °C) pack vs warm/room. Gallup & Gallup 2007.3
Cooling explains yawning	Leading, not settled.	Coherent and supported, but contested and replication-debated. Frontiers review.4
”We know why we yawn”	False, both ways.	We know it’s not oxygen (strong); the positive account is open (weaker).

Named uncertainties. (1) The negative is far stronger than the positive. “Not oxygen” rests on a direct, decades-old experiment with a clean dissociation; “it’s cooling” rests on a younger, contested body of work. The page states the first as fact and the second as the leading hypothesis, and nothing here proves the second. (2) The “~100×” is a range, not a point. The CO₂ enrichment was 86× (3%) to 143× (5%) of ambient, depending on the mixture and the era’s atmospheric baseline; “roughly a hundred times” is the honest round figure, and the verifier shows the spread. (3) The 2007 percentages are from one study’s samples — small psychology samples, reported exactly as published; they are evidence for the cooling direction, not the final word, and the replication debate is named, not hidden. (4) “39 years” is simple arithmetic from the 1987 publication to this page’s 2026 date; the precise month of the falsification’s obscurity is not a measurable thing. Each caveat is priced into the claim it supports.

Footnotes

1. Provine RR, Tate BC, Geldmacher LL (1987), “Yawning: no effect of 3–5% CO₂, 100% O₂, and exercise,” Behavioral and Neural Biology 48(3):382–393. PubMed PMID 3120687 — the experiment that refuted the blood-gas hypothesis; breathing rate rose under O₂, CO₂, and exercise while yawning did not. https://pubmed.ncbi.nlm.nih.gov/3120687/ ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶

2. Atmospheric CO₂ baseline: Mauna Loa annual mean, ≈349 ppm (0.0349%) around 1987 and ≈426 ppm today. 1% = 10,000 ppm, so 5% CO₂ ≈ 143× the 1987 ambient and 3% ≈ 86×; the cm-to-multiple arithmetic is recomputed in the verifier. NOAA Global Monitoring Laboratory, Mauna Loa CO₂ record. https://gml.noaa.gov/ccgg/trends/ ↩ ↩²

3. Gallup AC & Gallup GG Jr. (2007), “Yawning as a Brain Cooling Mechanism: Nasal Breathing and Forehead Cooling Diminish the Incidence of Contagious Yawning,” Evolutionary Psychology 5(1):92–101 — source for the measured percentages: oral/undirected breathing ≈ 48%, nasal 0%, warm (46 °C)/room-temp pack 41%, cold (4 °C) pack 9%. https://journals.sagepub.com/doi/10.1177/147470490700500109 ↩ ↩² ↩³ ↩⁴ ↩⁵

4. Gallup AC (2012), “The thermoregulatory theory of yawning: what we know from over 5 years of research,” Frontiers in Neuroscience 6:188 — the brain-cooling theory’s own author surveying the supporting and the contested evidence; cited here to mark the positive account as leading-but-debated rather than settled. https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2012.00188/full ↩ ↩²