How We Know the Universe Is Expanding

In 1929, Edwin Hubble plotted the distances of galaxies against the speeds at which they appeared to be moving and saw something strange: the farther a galaxy was, the faster it was receding. The relationship was roughly linear. A galaxy twice as far away was moving away twice as fast. This was not what a static universe should look like, and it was not what a universe with random galactic motions should look like either. It looked like the entire fabric of space was stretching, carrying galaxies along with it like raisins in rising dough.

The core measurement behind Hubble's plot was redshift — the lengthening of light's wavelength as its source moves away from the observer, or, more precisely in cosmology, as the space the light travels through expands beneath it. Astronomers identify redshift by looking at spectral lines, the dark absorption features produced by specific elements in a star's atmosphere. Hydrogen, calcium, and sodium each leave a fingerprint at known wavelengths in a laboratory. When that fingerprint appears in a distant galaxy's spectrum shifted toward the red end, the shift is unambiguous and quantitative. You are not guessing how far the lines moved; you are measuring it.

But redshift alone cannot tell you the universe is expanding. A galaxy could be redshifted simply because it is flying away through static space. To get expansion, you need to pair redshift with distance, and distance is the harder problem. Astronomers use standard candles — objects whose intrinsic brightness is known, so that their apparent brightness reveals how far away they are. Cepheid variable stars pulse at a rate tied to their luminosity; Type Ia supernovae detonate with a characteristic peak brightness. By measuring how dim such an object appears, you can infer its distance with reasonable precision. Hubble's original sample relied on Cepheids. Modern measurements at cosmological distances rely heavily on Type Ia supernovae, which are bright enough to be seen across billions of light-years.

The pairing matters. Redshift gives velocity. Standard candles give distance. Plot them together across thousands of galaxies and the linear relationship Hubble first sketched holds, and holds across scales he could not have reached. This is what convinced astronomers that the recession was not a local accident but a property of space itself.

A third, independent line of evidence arrived in 1965, when Arno Penzias and Robert Wilson detected a faint microwave glow coming from every direction in the sky. This is the cosmic microwave background, the relic radiation released when the early universe cooled enough for atoms to form and light to travel freely. Its temperature today is about 2.7 kelvin. If the universe had always been the size it is now, this radiation should not exist; if the universe were expanding from a hotter, denser past, it should look almost exactly as it does. The cosmic microwave background does not measure expansion directly, but it confirms the hot dense past that expansion implies.

The convergence is the point. Redshift surveys, standard-candle distances, and the microwave background are three different kinds of measurement, sensitive to different physics, vulnerable to different errors. They could have disagreed. They do not. Each, on its own, would be suggestive but contestable. A redshift could be reinterpreted; a standard candle could be miscalibrated; the microwave background could in principle be explained by something else. What is hard to explain away is three independent measurements pointing to the same picture: a universe that was smaller and hotter, and is now larger and cooler, and is still stretching.

This is also why the recent disagreement over the precise rate of expansion — the so-called Hubble tension, where measurements from the early universe and the late universe yield slightly different numbers — is taken so seriously. The fact of expansion is not in question. The exact value of the expansion rate is, and the discrepancy is small enough that it might be a measurement artifact and large enough that it might be telling us something is missing from the standard cosmological model. Knowing the universe is expanding is settled. Knowing exactly how it expands is still live work.