sciencebriefs
13:00in productionCh. 1 · Two forty-minute talks/ 13:00 · ceiling 15 min
Astronomy & space

Great Debate (astronomy)

1920

In 1920 two astronomers each gave a forty-minute talk on whether the Milky Way was the whole universe or one galaxy among many, and neither had quite enough evidence to win outright, until a third astronomer settled it a few years later.

At a 1920 meeting of the National Academy of Sciences in Washington, Harlow Shapley and Heber Curtis each argued a different picture of the cosmos: Shapley for a single enormous Milky Way containing everything, including the spiral nebulae, and Curtis for a smaller Milky Way sitting alongside other, separate galaxies. Both drew on real but incomplete evidence, and both got important pieces wrong, Shapley's galaxy was too large in his estimate, Curtis's too small, and one of Shapley's key data points, a claimed observation of a spiral nebula rotating, later turned out to be mistaken. Edwin Hubble resolved the central question a few years later by finding Cepheid variable stars in Andromeda far too faint, and therefore far too distant, to belong to the Milky Way, confirming Curtis's basic claim that the universe holds many separate galaxies.

Chapters & takeaways6
  1. 0:08
    Two forty-minute talks

    A formal 1920 debate set two rival pictures of the universe side by side.

  2. 2:10
    One galaxy or many

    Shapley argued for a single vast Milky Way; Curtis for many separate galaxies.

  3. 4:20
    The evidence each side brought

    Globular clusters and a rotating-nebula claim faced off against novae counts and recession speeds.

  4. 6:30
    Who was right about what

    Each debater got a central point right and a key figure wrong.

  5. 8:40
    Hubble settles it

    Cepheid variables in Andromeda placed it far outside the Milky Way.

  6. 10:50
    Is it worth your time

    A rare case of a live debate mapping onto a real, later-resolved question.

Worth your time?

Yes. Study the whole thing.

4/ 5
What works
  • Shapley's conclusion that the Sun sits away from the galaxy's centre has held up
  • Curtis's claim that spiral nebulae are separate galaxies was later confirmed
  • the debate is well documented, with both men's actual arguments preserved
What does not
  • Shapley's estimate of the Milky Way's size was too large, and Curtis's too small
  • a key piece of Shapley's evidence, a claimed observed rotation in a spiral nebula, did not hold up
Study it if
  • anyone curious how astronomers argued before the size of the universe was settled
  • readers who like seeing two flawed positions each contain a correct piece
  • people interested in how Hubble's later work closed an open scientific argument
Skip it if
  • readers wanting a full technical account of Cepheid distance measurement
  • anyone looking for a single clean winner rather than a mixed verdict
The written brief3 min read

Two forty-minute talks

On 26 April 1920, at a meeting of the National Academy of Sciences held at the US National Museum in Washington, two astronomers, Harlow Shapley and Heber Curtis, each gave a forty-minute presentation on the scale of the universe. The question at stake was whether the Milky Way was the entirety of the cosmos or one galaxy among many. Shapley argued for a single, enormous Milky Way, roughly 300,000 light years across, with spiral nebulae such as Andromeda no more than features on its outer edge. Curtis argued the opposite: a considerably smaller Milky Way, around 30,000 light years across, coexisting with spiral nebulae that were themselves separate, distant galaxies comparable in scale to our own.

One galaxy or many

Shapley built his case on globular clusters and Cepheid variable stars, using the period-luminosity relation established by Henrietta Swan Leavitt to gauge distances and map how the clusters were distributed around the galaxy, from which he placed the Sun well away from the centre. He also cited a bright nova seen in Andromeda in 1885, which briefly outshone the entire nebula, and a claimed observation by Adriaan van Maanen that a spiral nebula was visibly rotating, a result that would be essentially impossible if the nebula were as large and distant as Curtis proposed. Curtis, in turn, pointed to the far greater number of novae recorded in Andromeda than in the Milky Way, to the unexpectedly large recession speeds measured for spiral nebulae, and to dust obscuration as a shared explanation for why few nebulae appeared along the Milky Way’s own plane.

The evidence each side brought

Parts of each man’s case have held up well. Shapley’s placement of the Sun away from the galactic centre, derived from the spread of globular clusters, remains accepted and reshaped where humanity sits within its own galaxy. Curtis’s central claim, that the spiral nebulae are independent galaxies rather than features of the Milky Way, was likewise vindicated. The debate itself is remembered less for either man being simply right than for how directly it set two testable pictures of the universe against each other in public, using the observational evidence each side actually had rather than appeals to authority or intuition, which is unusual enough in scientific history to be worth noting on its own.

Who was right about what

Neither man’s specific numbers survived. Shapley’s estimate that the Milky Way spanned some 300,000 light years proved too large, and Curtis’s estimate of roughly 30,000 light years proved too small; the true figure fell between the two. Shapley’s broader claim that the Milky Way constituted the entire universe turned out to be wrong, and one of his supporting data points, van Maanen’s claimed detection of a spiral nebula rotating, did not hold up either, since a real rotation of that kind was incompatible with the nebula being as distant as it in fact is. The 1885 nova that Shapley leaned on also had a different explanation than either debater assumed: it was later recognised as a supernova, extraordinarily bright for reasons unrelated to the nebula’s true size.

Hubble settles it

The central question was resolved within a few years, not by further debate but by better measurement. Edwin Hubble found Cepheid variable stars in the Andromeda nebula and used their known period-luminosity relation to calculate a distance far too great for Andromeda to be part of the Milky Way, confirming Curtis’s basic claim that spiral nebulae are separate galaxies. That single result reframed the Milky Way from the whole of creation into one galaxy among an enormous population, now estimated at somewhere between roughly 200 billion and 2 trillion galaxies in the observable universe. It marks the practical beginning of extragalactic astronomy as a field, and of cosmology as a science built on measured distances rather than assumed scale.

Is it worth your time

This is a good subject for anyone who likes watching an argument actually get settled rather than simply fade away. Both debaters were reasoning carefully from real evidence, and both were still substantially wrong about the numbers that mattered, which is a more honest picture of how science often proceeds than a story with a single clear winner from the outset. Shapley’s own graceful acceptance of Hubble’s later evidence against his position adds a human note to what could otherwise read as a purely technical dispute. It rewards an hour from anyone curious about the moment the universe stopped being one galaxy and became billions.

Same field · Astronomy & space4 of 78
Up next in Science

Great Pacific Garbage Patch

Charles Moore · 1980 · 9:30

Moore didn’t discover the garbage patch — he named it, after NOAA had already measured it.

9:30