sciencebriefs
13:00in productionCh. 1 · Three puzzles about the early universe/ 13:00 · ceiling 15 min
Physics · Astronomy & space

Cosmic inflation

1981

Alan Guth's 1980 fix for three specific puzzles about the early universe made a precise prediction that satellites later confirmed, and it also implies an eternally inflating multiverse that one of the theory's own inventors now argues against.

By the late 1970s, cosmologists faced three unexplained puzzles: why distant regions of the universe look so alike despite having no time to interact, why the universe's density sits so close to a precise critical value, and why predicted magnetic monopoles have never been observed. Alan Guth, developing the idea at Cornell and submitting his paper in August 1980, proposed that the early universe underwent an extraordinarily rapid exponential expansion from a supercooled state, addressing all three problems at once. His original mechanism needed fixing, refined by Andrei Linde, Andreas Albrecht and Paul Steinhardt into slow-roll inflation around 1982. Later observations, including the COBE satellite in 1992 and the Planck spacecraft, matched the theory's specific predictions about the universe's flatness and structure closely. But the particle responsible for inflation has never been identified, the theory requires its own fine-tuning, and many versions imply an eternally inflating multiverse, an implication Paul Steinhardt, one of the theory's original architects, now argues against.

Chapters & takeaways6
  1. 0:08
    Three puzzles about the early universe

    Standard Big Bang cosmology could not explain three specific observations.

  2. 2:10
    A supercooled false vacuum

    Guth's 1980 mechanism proposed an extraordinarily rapid early expansion.

  3. 4:20
    Refined into something workable

    Later physicists fixed a fine-tuning flaw in Guth's original version.

  4. 6:30
    What the microwave background shows

    Satellite measurements matched the theory's specific numerical predictions.

  5. 8:40
    What still isn't nailed down

    The driving particle is unidentified, and eternal inflation divides even its founders.

  6. 10:50
    A Kavli Prize, an open debate

    Recognition for the theory coexists with genuine, unresolved scientific dissent.

Worth your time?

Yes. Study the whole thing.

4/ 5
What works
  • the theory's predictions about flatness and structure have matched satellite observations closely
  • the fix from old inflation to slow-roll inflation addressed a real, acknowledged flaw
  • the theory turned cosmology into a field making testable, quantitative predictions
What does not
  • the specific particle mechanism believed to drive inflation has never been identified
  • eternal inflation and its multiverse implication remain genuinely disputed, even among the theory's founders
Study it if
  • anyone who wants a clear split between confirmed prediction and open theoretical debate
  • readers curious how satellite data can test an idea about the universe's first fraction of a second
  • people interested in a theory whose own inventor later became one of its critics
Skip it if
  • readers wanting a settled, uncontested account of the early universe
  • anyone looking for the detailed particle physics of the hypothetical inflaton field
The written brief3 min read

Three puzzles about the early universe

By the late 1970s, cosmologists faced three specific puzzles the standard Big Bang picture could not explain on its own. The horizon problem asked why regions of the universe on opposite sides of the observable sky look so nearly identical in temperature and structure, despite having had, under the standard expansion rate, no time to exchange light or heat and reach that similarity naturally. The flatness problem asked why the universe’s overall density sits so close to a specific critical value, a balance that would have needed tuning in the very early universe to within about one part in ten to the sixty-second power to still look nearly flat today. The magnetic monopole problem noted that grand unified theories of particle physics predicted an abundance of stable magnetic monopoles that had simply never been observed.

A supercooled false vacuum

Alan Guth, then working at Cornell, began developing a solution after hearing Robert Dicke describe the flatness problem in 1978 and Steven Weinberg discuss grand unified theories and early-universe phase transitions in early 1979. He gave a first seminar on the idea at SLAC in January 1980 and submitted his paper, “Inflationary universe: A possible solution to the horizon and flatness problems,” to Physical Review in August 1980, formally proposing cosmic inflation the following year. His mechanism involved the early universe becoming trapped in a supercooled false vacuum state that then decayed, driving an extraordinarily rapid exponential expansion in which distances between points doubled roughly every ten to the minus thirty-seventh power of a second, over a period lasting at least ten to the minus thirty-fifth power of a second.

Refined into something workable

Guth’s original version, since called old inflation, turned out to have a serious internal problem, one Guth himself acknowledged by 1983 as requiring extreme fine-tuning of its parameters, because the process by which the false vacuum decayed did not naturally bring inflation to a clean, uniform end. Andrei Linde, working independently in December 1981, and separately Andreas Albrecht and Paul Steinhardt around the same period, developed a refined version called slow-roll inflation that resolved this specific flaw, and it is this later, corrected version that underlies most modern treatments of the theory rather than Guth’s original mechanism.

What the microwave background shows

Later observation gave the theory real, checkable support. The COBE satellite’s 1992 measurements of the cosmic microwave background found temperature variations following a nearly scale-invariant pattern, matching what inflation predicted for the seeds of later galaxy formation. Decades later, the Planck spacecraft found the universe flat to within about half a percent, homogeneous and isotropic to about one part in a hundred thousand, and measured a specific statistical property of that structure, the spectral index, at 0.968 plus or minus 0.006, squarely inside the range inflationary models had predicted well before the measurement was made.

What still isn’t nailed down

Real open problems remain. The specific particle physics mechanism thought to drive inflation, generally called the inflaton field, has never been identified, though researchers have at least ruled out the Higgs field as a candidate. The theory’s own required starting conditions raise a fine-tuning question of their own, and the energy scale at which inflation is believed to have occurred, roughly ten to the sixteenth power of a giga-electronvolt, sits at a small fraction of the Planck energy, a gap some physicists consider itself unexplained. Many versions of the theory also predict that inflation never fully stops everywhere, continuing eternally in some regions and generating an effectively infinite multiverse, an implication that has drawn real dissent: Paul Steinhardt, one of the theory’s own original architects, has become one of its most outspoken critics specifically over what eternal inflation implies.

A Kavli Prize, an open debate

Despite these unresolved questions, inflation transformed cosmology from a field relying on unexplained starting assumptions into one making specific, quantitative predictions later confirmed by satellite observation, and it remains the dominant framework for explaining why the universe looks as uniform and as structured as it does. Guth, Linde and Starobinsky shared the 2014 Kavli Prize in Astrophysics for originating the theory, recognition that sits alongside, rather than resolving, the continuing debate among physicists over whether some of the theory’s deepest assumptions are on solid ground or still fundamentally open. This is worth an hour for how clearly it separates confirmed prediction from genuinely unresolved theory within one still-live scientific idea.

Same field · Physics4 of 183
Up next in Science

CRISPR gene editing

· 13:00

CRISPR-Cas9, built from a bacterial virus-defence system decades in the discovering, turned gene editing into a routine laboratory tool now behind approved sickle-cell therapy and an unresolved fight over editing embryos.

13:00