Physicists Can't Agree What 95% of the Universe Is — And That's the Point
June 21, 2026 · Stephen Hawking, A Brief History of Time~6 min read
Ask the world's physicists what most of the universe is made of, and you will not get one answer. You won't even get a majority. The largest survey of its kind just made that official — and the honest reading of it is not "scientists are lost," but "this is what the frontier looks like when nobody is pretending."
A vote with no winners
In May 2026, Niayesh Afshordi of the Perimeter Institute and the University of Waterloo, with his co-author Phil Halper, published the largest survey ever taken of physicists on the foundational questions of their field — done in collaboration with the American Physical Society's Physics magazine. The striking thing is not which idea won. It's that, on most of the "standard answers" textbooks treat as settled, nothing won. Take dark matter, the invisible stuff that outweighs every star: 17% of physicists bet on undiscovered low-mass particles, 12% think we should modify gravity instead, and 21% suspect it's some combination — three camps, no majority. Quantum gravity, the dream of marrying Einstein to the quantum, splits the same way: string theory 19%, loop quantum gravity 12%, and a stubborn 18% who suspect gravity simply cannot be quantized at all. Even ΛCDM, the standard model of cosmology, failed to win majority support — partly because results from the DESI survey hint that dark energy may be changing over time rather than holding constant, as the model assumes.
The two things they do agree on
Only two statements crossed the halfway line, and both are quieter and deeper than the headlines about disagreement. First, a clear 68% majority accepts that the Big Bang does not necessarily mark the beginning of time — it describes the universe expanding from a hot, dense early state, which is a very different claim from "time started here." Second, a bare 51% agree that the early universe underwent cosmic inflation, a fleeting burst of staggering expansion. Notice what these two survivors have in common: they are descriptions of how the universe behaved, anchored in evidence, not bets on what it ultimately is. Consensus clusters around the charted coastline. The blank interior is where the arguing happens.
A map with a coast and a blank middle
That coastline-and-interior image is the most useful way to hold all this. Picture the field as an old map: the shore is drawn in firm ink — the universe is expanding, it was once hot and dense, galaxies recede, the light from the early cosmos still hums around us. None of that is in doubt; the survey doesn't touch it. But sail inland and the map goes blank. What is 95% of the universe — the dark matter and dark energy that together dwarf everything we can see — actually made of? Here the cartographers scrawl competing guesses in the margins and cross each other out. The disagreement isn't smeared evenly across physics. It is concentrated, precisely, on the parts we have not yet charted.
The largest-ever survey of physicists (Afshordi & Halper, with APS Physics magazine, ~2026) found only two statements with majority support: the Big Bang need not mark the beginning of time (68%) and early-universe inflation (51%). On dark matter (17/12/21%) and quantum gravity (19/12/18%), no answer wins a majority. Read: at the frontier, disagreement is the normal state — not a crisis. Bars to scale; the dashed gold line marks 50%.
Disagreement is not ignorance
It would be easy, and wrong, to read all this as "the experts know nothing." They know an enormous amount — that is precisely why they can disagree so productively. You cannot argue about whether dark matter is a particle or a flaw in our gravity unless you have already pinned down, to many decimal places, how galaxies rotate and how the cosmos clusters. The split is not the noise of confusion; it is the sound of a real question being attacked from several sides at once, where the evidence is genuinely strong enough to support more than one serious answer but not yet sharp enough to kill all but one. The settled core — expansion, a hot dense beginning — is rock. The contested frontier is where the rock is still being quarried.
Hawking's own book was a record of changing his mind
This is exactly the spirit of the book that taught a generation to love cosmology. A Brief History of Time is often remembered as Stephen Hawking handing the public "the answers." Read it again and it's the opposite: a story of the picture being revised. Hawking made his name proving the universe must begin in a singularity, a point of infinite density where the equations break — and then spent later years doubting that very conclusion, proposing a cosmos with "no boundary," no first moment at all. The man who supposedly explained the Big Bang spent his career unsure whether it was the beginning of anything. That isn't a flaw in the work. That is the work. Revising the map is not what happens when science fails; it is what science is.
The settled core is solid — the universe is expanding, it began hot and dense. The 95% we can't name is not a failure of physics; it is its open frontier, and disagreement there is the engine, not the breakdown.
A good map shows its blank spaces honestly.
How to read a "physicists discover" headline
So the next time a headline announces that physicists have "finally explained" dark matter or "solved" the cosmos, hold it up against this survey and run three quick checks. Ask whether the claim lives on the charted coast or the blank interior — expansion is coast, the identity of dark energy is interior, and the second deserves far more caution than the first. Ask whether you're being shown a consensus or one camp's flag planted loudly; "a study suggests" is not "physicists agree." And when you meet honest uncertainty — a survey where no answer wins, a famous physicist doubting his own theorem — read it as strength, not weakness. The universe is mostly made of stuff we cannot yet name. The people who study it for a living openly disagree about what it is. Both of those sentences are true, and neither is bad news. That two-handed grip — holding the solid coast and the open sea at once — is, more than any single result, what it means to think like a scientist.
Sources: Niayesh Afshordi (Perimeter Institute / University of Waterloo) and Phil Halper conducted the largest-ever survey of physicists on foundational questions, in collaboration with the American Physical Society's Physics magazine (reported ~May 2026): no majority on dark matter (low-mass particles 17%, modify gravity 12%, a combination 21%) or quantum gravity (string theory 19%, loop quantum gravity 12%, "cannot be quantized" 18%); ΛCDM won no majority, with DESI results hinting dark energy may evolve over time; only two statements gained majorities — the Big Bang need not mark the beginning of time (68%) and cosmic inflation (51%). Framework paraphrased from Stephen Hawking, A Brief History of Time. Popular-science interpretation, not a research claim.