Cosmic inflation

• 1. Introduction

  ›

  Cosmic inflation is the idea that the universe, in its very first fractions of a second, expanded at a rate so extreme that distances between points doubled every ten to the negative thirty-seven seconds. Not slowly, not gradually. In an interval shorter than any clock could ever measure, a region smaller than an atom ballooned to roughly the size of a grapefruit. The theory was first proposed by particle physicist Alan Guth in 1979, while he was investigating a seemingly unrelated puzzle: why has no one ever found a magnetic monopole? The answer he found pointed to something far larger than magnets. It pointed to the shape, temperature, and structure of everything that exists. The questions this documentary will explore are: why did physicists think the Big Bang alone was not enough to explain the universe we see? Who built this theory, and what price did it pay to work? And what happens to a model of the cosmos when it predicts something that might inflate forever?

• 2. Problems Without Inflation

  ›

  In the early 1970s, Yakov Zeldovich noticed that the Big Bang model carried within it two deep embarrassments. The first is called the flatness problem. Observations show the universe is flat to within a few percent, meaning its total energy density sits almost exactly at a critical value. In the equations governing cosmic expansion, any deviation from that critical density would grow larger over time, not smaller. Working backwards, the early universe must have matched the critical density to within one part in ten to the sixty-second power. No known physical process demanded that precision. It simply had to be assumed. The second embarrassment is the horizon problem. When scientists map the faint afterglow of the Big Bang, the cosmic microwave background, they find it almost perfectly uniform in every direction. Yet in a standard Big Bang without inflation, regions on opposite sides of the observable universe could never have exchanged heat or information. They were too far apart, moving away from each other faster than light could travel between them. They should look different. They do not. A third problem motivated Guth directly. Grand Unified Theories of particle physics predict that the early universe should have produced vast numbers of magnetic monopoles, a kind of stable, heavy magnetic charge. Not a single one has ever been found. Martin Rees captured the awkwardness of relying on this last problem when he wrote that skeptics might not be impressed by a theoretical argument to explain the absence of particles that are themselves only hypothetical.

• 3. Building the Theory

  ›

  Alexei Starobinsky at the Landau Institute for Theoretical Physics was thinking about quantum corrections to general relativity when he found that those corrections, applied to the early universe, would generate an effective cosmological constant. That constant would drive an exponential expansion, an inflationary de Sitter phase, and it made specific, calculable predictions about corrections to the microwave background. His work preceded Guth's proposal but came from a different direction: modified gravity rather than particle physics. Guth's 1979 insight was that a positive-energy false vacuum would, according to general relativity, produce exponential expansion of space. He recognized this could explain why magnetic monopoles are absent: they would be separated from one another as the universe expanded around them, lowering their density by many orders of magnitude. In January 1981, Guth formally proposed the model and coined the word "inflation". Around the same time, Demosthenes Kazanas suggested that exponential expansion could eliminate the particle horizon, and Katsuhiko Sato suggested it could eliminate domain walls. Guth's original version had a serious flaw. The model required the early universe to tunnel out of a false vacuum through the formation of bubbles of true vacuum. Those bubbles expanded at the speed of light, but if inflation lasted long enough to solve the initial conditions problems, bubble collisions became exceedingly rare. In any single causal patch, likely only one bubble would form. The energy stored in inflation would never efficiently convert back into matter and radiation. The model did not reheat properly. This problem was solved by Andrei Linde, and independently by Andreas Albrecht and Paul Steinhardt, in a model called new inflation or slow-roll inflation. Instead of tunneling, the inflaton field rolled slowly down a potential energy hill. When the field rolled slowly compared to the expansion of the universe, inflation occurred. When the hill steepened, inflation ended and reheating could begin. Linde later proposed a further refinement called chaotic inflation, arguing that inflation would occur generically in any universe beginning in a chaotic, high-energy state with a scalar field of unbounded potential energy.

• 4. Quantum Seeds of Structure

  ›

  Once slow-roll inflation was established, physicists found that it did not produce a perfectly uniform universe. Tiny quantum fluctuations in the inflaton field were stretched to cosmic scales during the inflationary epoch, and those microscopic variations became the seeds of every galaxy, every cluster of galaxies, every large-scale filament visible in the universe today. These fluctuations were first calculated by Viatcheslav Mukhanov and G. V. Chibisov in the context of Starobinsky's model. A parallel calculation occurred at the three-week 1982 Nuffield Workshop on the Very Early Universe at Cambridge University. Four independent groups worked out the fluctuations over the course of that workshop: Stephen Hawking; Starobinsky; Alan Guth and So-Young Pi; and James Bardeen, Paul Steinhardt and Michael Turner. The specific form of the spectrum inflation predicts is called a nearly-scale-invariant Gaussian random field. It has only two free parameters: the amplitude of the perturbations and the spectral index, which measures the slight deviation from perfect scale invariance. Inflation predicts the spectral index, called n, lies between 0.92 and 0.98. Data from the Planck spacecraft place it at 0.968 plus or minus 0.006. That match is not a coincidence. The Planck results also confirmed that the perturbations are adiabatic, as inflation requires, and that the universe is flat to within a percent, homogeneous and isotropic to one part in 100,000. Temperature anisotropies observed by the COBE satellite in 1992 had already shown the nearly scale-invariant spectrum predicted by the inflationary paradigm.

• 5. The Eternal Inflation Problem

  ›

  Paul Steinhardt introduced the first example of eternal inflation in 1983. He showed that inflation could proceed forever in at least some regions of the universe by producing bubbles of non-inflating space, each filled with hot matter and radiation, surrounded by empty space that continues to inflate. The bubbles could not grow fast enough to keep up with the surrounding inflation. Later that same year, Alexander Vilenkin showed that eternal inflation is generic to many inflationary models. The mechanism is straightforward in principle. Quantum fluctuations can push the inflaton field upward in potential energy in some regions. Those regions expand faster than regions where the field is rolling downward. The volume of inflating space grows without bound. This leads to a picture in which the universe, at the largest scales, is an infinite, never-ending patchwork of inflating and non-inflating regions. Different regions may settle into different vacuum states with different laws of physics, producing what is often called a multiverse. Steinhardt, one of the architects of slow-roll inflation, later became one of the model's most vocal critics for exactly this reason. Together with Anna Ijjas and Abraham Loeb, he argued that the inflationary paradigm is in trouble in view of data from the Planck satellite. He pointed out that bad inflation, meaning a period of accelerated expansion whose outcome conflicts with observations, is more likely than good inflation, and that no inflation at all is more likely than either. Guth, David Kaiser, Yasunori Nomura, and Linde offered counter-arguments, asserting that cosmic inflation is on a stronger footing than ever before.

• 6. Challenges and Criticisms

  ›

  Roger Penrose has argued from 1986 onward that inflation does not actually solve the initial conditions problem. It displaces it. Inflation requires extremely specific initial conditions of its own. Penrose calculated that among all possible configurations of the inflaton and gravitational fields, obtaining a flat universe without inflation is more likely than obtaining one with inflation by a factor of ten to the googol power. At a conference in 2015, Penrose said that inflation is not falsifiable, and that it is falsified. In 1999, John Earman and Jesús Mosterín published a critical review of inflationary cosmology concluding that there are not yet good grounds for admitting any of the models of inflation into the standard core of cosmology. A recurring concern is that the inflaton field does not correspond to any known physical field, and that its potential energy curve can be shaped to accommodate almost any observational data. The fine-tuning problems within inflation itself are substantial. In slow-roll models, the inflaton potential must be exceptionally flat compared to the vacuum energy, and the inflaton particles must have very small mass. These conditions require explanation, not just assumption. A separate difficulty called the trans-Planckian problem arises because some of the quantum fluctuations that seeded cosmic structure were smaller than the Planck length before inflation stretched them. At that scale, the known equations of quantum field theory break down, and corrections from an unknown quantum theory of gravity should in principle apply. Disagreement remains about whether those corrections would be detectable or negligible. Don Page argued that the thermodynamic arrow of time requires low-entropy initial conditions, and that the reheating at the end of inflation increases entropy, making the initial state of the universe necessarily more ordered than in Big Bang models without inflation. The controversy over the BICEP2 announcement in March 2014 illustrated the difficulty of observational tests. The BICEP2 team announced detection of B-mode polarization in the cosmic microwave background, which would constitute evidence for gravitational waves produced during inflation. By the 30th of January 2015, confidence in the finding had been significantly reduced. By 2018, additional data suggested with 95% confidence that the tensor-to-scalar ratio is 0.06 or lower, consistent with both the null hypothesis and many remaining inflation models.

• 7. Recognition and What Remains Open

  ›

  Starobinsky, Guth, and Linde won the 2014 Kavli Prize for pioneering the theory of cosmic inflation. In 2002, Guth of M.I.T., Linde of Stanford, and Steinhardt of Princeton shared the Dirac Prize for development of the concept of inflation in cosmology. In 2012, Guth and Linde received the Breakthrough Prize in Fundamental Physics for their invention and development of inflationary cosmology. The particle physics identity of the inflaton remains unknown. Guth's early proposal suggested it might be the Higgs field, the field responsible for the mass of elementary particles, but current thinking holds that the Higgs field faces tension with electroweak-scale data that makes this identification problematic. Several alternative frameworks exist. The ekpyrotic and cyclic models propose that the horizon problem was solved by an expanding epoch well before the Big Bang, with the observed density perturbations generated during a contracting phase ending in a Big Crunch. String gas cosmology, proposed by Robert Brandenberger and Cumrun Vafa, models the early universe as a hot gas of strings and uses a mechanism called the Brandenberger-Vafa mechanism to explain why three spatial dimensions expanded while others remained compactified. Loop quantum cosmology offers another route, proposing that if energy density exceeds what a quantized spacetime can hold, it bounces rather than collapses. The history of what happened between the end of inflation and roughly one second after the Big Bang, a period dominated by the reheating process, remains largely unknown. Upcoming measurements of 21-centimeter radiation, emitted and absorbed by neutral hydrogen before the first stars formed, may eventually map the primordial power spectrum with greater precision than current microwave background surveys, though whether interference from terrestrial and galactic radio sources will permit those measurements is still an open question.