Skip to content
The numbersVisual atlas

The six numbers that build a cosmos

Martin Rees's dials β€” gravity, the nuclear force, dark energy, and three more β€” and what each one does.

Lukasz Szramuk Β· Β· updated Β· 7 min read

In Just Six Numbers, the astronomer Martin Rees argued that our whole universe hangs on roughly six quantities. Get any badly wrong and you get no stars, no atoms, or no time to make either. These are the exact dials in our Universe Tuner. Rees wrote the book in 1999 as a tour of 'the deep forces that shape the universe'; a quarter-century of precision cosmology has only made his list look better chosen.

NgravityΞ΅fusionΞ©matterΞ›dark energyQlumpinessDdimensions
Rees's six dials β€” gravity, fusion, matter, dark energy, lumpiness, and the number of dimensions. Each sits at a particular setting; move one and the cosmos forgets how to make you.

N β€” gravity's grip

N measures gravity against the forces inside atoms: about 10⁻³⁢. Gravity is so weak that a small magnet lifts a paperclip against the pull of an entire planet β€” and that weakness is exactly what stars need. A star is a slow-motion negotiation between gravity squeezing in and fusion pushing out; weak gravity keeps the negotiation running for billions of years.

Make N a few powers of ten stronger and the same negotiation ends in a rout: stars become smaller, hotter, and done in millions of years β€” no eons for planets to cool or chemistry to experiment. Make it weaker and gas never compresses enough to ignite at all. Either way the sky goes dark.

Ξ΅ β€” the nuclear glue

How efficiently hydrogen fuses into helium β€” about 0.007 of its mass becomes energy. A hair lower and nothing fuses; a hair higher and hydrogen burns away at birth, leaving no water and no long-lived suns.

The same number governs the famous triple-alpha bottleneck: carbon forms only because 0.007 lands close enough to a resonance of carbon-12 for stars to build it three helium nuclei at a time. Fred Hoyle predicted that resonance from our existence before anyone measured it β€” the full story is in our essay on the anthropic principle.

Ξ© β€” the amount of stuff

Ξ© compares the universe's actual density to the critical density where expansion and gravity exactly balance. Ours is suspiciously close to 1 β€” and that is itself a tuning puzzle, because Ξ©=1 is an unstable equilibrium: any early deviation grows. To sit near 1 today, the infant universe had to be flat to better than one part in 10⁡⁹. Inflation was invented to explain precisely this.

Too much matter and the expansion reverses into a Big Crunch before structure matures. Too little and gravity never wins anywhere β€” no galaxies, just thinning gas. The Tuner plays both deaths for you.

Ξ› β€” the number nobody ordered

Ξ› is the energy of empty space, and the observed value is absurdly small: about 10⁻¹²² in the natural units where quantum theory predicts something enormous. Physicists have called the mismatch the worst prediction in the history of science.

Yet small is not zero β€” we measured Ξ› in 1998 because distant supernovae showed the expansion speeding up. Steven Weinberg had argued a decade earlier that observers can only exist where Ξ› stays below a narrow ceiling; we live within a factor of a few of that ceiling. Coincidence, selection, or an undiscovered law β€” the three answers compete over this number most of all.

Q β€” the texture of the early sky

Q is the size of the seed ripples in the young universe: about one part in 100,000, now photographed as the mottling of the cosmic microwave background. Everything large β€” galaxies, clusters, the cosmic web β€” grew from those wrinkles by gravitational amplification.

Shrink Q tenfold and the ripples never gather: smooth gas forever. Grow it tenfold and matter overshoots, collapsing into black holes instead of galaxies. The window is wider than Ξ΅'s, but the direction of both failures is physics we can watch β€” turn the dial in the Tuner and the fog and the collapse are both rendered.

D β€” three dimensions

D is the quietest number: the count of space's large dimensions. In 1917 Paul Ehrenfest pointed out that stable orbits β€” planets around suns, electrons around nuclei β€” exist only in three. In four or more, inverse-square laws become inverse-cube or worse, and orbits spiral in or fly off.

Two dimensions starve complexity differently: waves travel cleanly but networks tangle β€” a digestive tract would cut a flat creature in half. Three is not merely permitted; it is privileged, and we still don't know why.

Beyond the six

The Tuner adds three dials Rees didn't need for his list but we couldn't leave out: Ξ·, the tiny excess of matter over antimatter (without it, no stuff at all); Ο‡, the dark matter that scaffolds galaxies; and Ξ±, the fine-structure constant behind all chemistry. Their failure modes are too instructive to skip.

What precision cosmology added

When Rees published in 1999, several of his six were known to a digit or two. Then came WMAP and Planck: Ξ©, Ξ›, and Q are now measured to percent-level precision from the microwave background, and the cosmic expansion rate is argued over in the third significant figure. The tuning question didn't go away β€” it grew error bars. We no longer ask merely whether the numbers are special; we ask it about values written down in data tables.

Could fewer numbers do?

Six dials look independent, but a deeper theory might set several at once. The fusion efficiency Ξ΅ traces back to the strong force and quark masses; the cosmological constant may someday fall out of quantum gravity rather than stand alone. Every time physics has unified numbers, the roster of free ones shrank β€” the periodic table was once dozens of unexplained elements. 'Six' is a floor for today, not a ceiling for tomorrow, and shrinking it is exactly what a theory of everything would be for.

Rees himself has admitted the list is partly a storyteller's choice β€” six is a memorable count, and other physicists would draft a slightly different roster: quark masses, neutrino masses, the strength of the weak force. The exact membership matters less than the shape of the claim: a short list of numbers you cannot derive from first principles (yet), on which everything you have ever seen depends.

Rees returned to the theme in Our Cosmic Habitat (2001), sharpening the point that matters most here: whether or not the numbers are tuned, they are the right syllabus. Learn what each of the six does and you own a working map of why there is something rather than nothing interesting. That is the syllabus the Tuner teaches with your hands instead of a chalkboard.

A measured number is not automatically a free parameter. Ξ© and Q describe the universe's state as well as its laws, while N and Ξ΅ package deeper couplings into memorable ratios. The list mixes categories on purpose: it is a map of cosmic consequences, not a claim that a final equation must contain exactly six independent knobs.

Where the toy simplifies

Six dials look independent on a control panel; in real physics they correlate β€” the same underlying theory would set many of them at once. The windows in the toy are honest about scale and direction, dishonest only in their sharpness. The full confession is on the methods page.

Six numbers. Nudge any of them and the universe forgets how to make you.

Next experiment

Compare low, ours and high.

A constant matters in both directions. Probe both edges instead of treating our value as the only interesting point.

  1. Step 1Choose the constant discussed here and reset its companions.
  2. Step 2Record one low-side and one high-side failure.
  3. Step 3Compare which cosmic epoch disappears first in each run.
Open the Universe Tuner β†’