Science › Creation theories
Slow roll: energy rolling down a hill
1982· Theory
Well supportedStrong evidence or math, but not fully tested.
The mechanism that lets inflation last and then end cleanly: the inflaton creeps down a very gentle energy slope, its energy nearly constant.
See it on the Science map →How it works
Picture the inflaton as a ball on a long, nearly flat plateau that ends at a drop. The rapid expansion acts like thick syrup, slowing the ball, so it creeps while its energy, and therefore the inflation, barely changes. At the steep edge it speeds up, rolls to the bottom and oscillates there, passing its energy to particles (reheating). Within one rolling region, inflation ends smoothly and almost everywhere at once, so no bubbles need to merge.
Why it matters
It rescued inflation from Guth's failed first version, and the ball's slight speed-up as it rolls predicts a measurable tilt in the cosmic ripples.
How we know
Its key forecast, ripples slightly weaker on small scales, was measured: Planck (2018) gives a tilt of 0.9649 ± 0.0042, well below the perfectly even value of 1.
Still a mystery
Why is the slope so gentle, when quantum effects tend to steepen such hills? In 2025 the Atacama Cosmology Telescope, with DESI data, nudged the tilt to about 0.974, squeezing some favorite models.
Connections (4)
▸The inflaton field rolls down its energy hill in Slow roll: energy rolling down a hillWell supported
The relationship
Expansion itself acts as friction on the inflaton, called Hubble friction: the faster space grows, the more it slows the field. On a gentle slope the field settles at a crawling terminal speed, as a falling raindrop does in air, and its energy stays nearly constant. Slow roll needs two things: a shallow slope, and a slope that changes only gradually.
How the link was found
Linde and, separately, Albrecht and Steinhardt introduced it in 1982. The general slow-roll approximation was formalized through the 1980s and early 1990s by Steinhardt, Michael Turner, Andrew Liddle, David Lyth and others.
Is it proven?
The equations of a field in expanding space are standard physics. They show exactly how the slope and its curvature set the ripples' strength and tilt, which Planck has measured.
Open questions and mysteries
Quantum corrections tend to steepen such hills, in supergravity theories usually enough to spoil slow roll (the 'eta problem'). What keeps the hill flat, perhaps a hidden symmetry, is unknown.
▸Slow roll: energy rolling down a hill sustains and smoothly ends Cosmic inflationWell supported
The relationship
To flatten space and even out the sky, inflation must last at least about 60 'e-folds', each multiplying size by 2.7, for 10²⁶-fold growth. Slow roll provides that and ends smoothly. Because the field speeds up slightly, later ripples, now on smaller scales, are a bit weaker. The tilt depends on how long inflation ran after today's visible scales were stretched; in many models, 60 e-folds gives about 0.965.
How the link was found
Linde and Albrecht–Steinhardt showed in 1982 that slow roll lasts long enough. Viatcheslav Mukhanov and Gennady Chibisov predicted the slight tilt in 1981.
Is it proven?
Planck (2013) measured the tilt below 1 at over five standard deviations, as predicted three decades earlier. It is one of inflation's strongest successes.
Open questions and mysteries
Inflation could have lasted far longer than 60 e-folds; if so, every trace of what came before is stretched beyond view. Why it lasted at least that long is unexplained.
▸Slow roll: energy rolling down a hill ends in Reheating: inflation's energy becomes hot matterWell supported
The relationship
Slow roll lasts while the field's energy hill is gentle, so its stored energy barely changes as it moves. Near the bottom the slope steepens, the field speeds up, and inflation stops once its energy of motion rivals its stored energy. The field then overshoots and swings around the low point, and those swings power reheating. The shape of the hill's bottom controls how reheating goes.
How the link was found
Built into the 1982 slow-roll models of Andrei Linde and of Andreas Albrecht and Paul Steinhardt, whose fields ended in oscillations; the decay was worked out that year with Michael Turner and Frank Wilczek.
Is it proven?
The slow-roll phase fits microwave-background ripple data well. The exit and oscillations follow from the same math, but no observation confirms them yet.
Open questions and mysteries
Does the hill end in a smooth bowl, a sharp drop, or a handoff to another field? Ripple data weakly limit how long reheating took, but not how it worked.
▸Andrei Linde fixed old inflation with Slow roll: energy rolling down a hillWell supported
The relationship
Linde saw that if the energy hill near the false vacuum were very flat, the field inside a newly formed bubble would keep inflating while rolling slowly down. One bubble alone would grow far larger than our observable universe, so bubbles never needed to merge. Inflation would end smoothly, releasing its energy evenly.
How the link was found
Andrei Linde in Moscow, submitted October 1981, published in Physics Letters B in February 1982. Andreas Albrecht and Paul Steinhardt at the University of Pennsylvania reached the same answer independently (April 1982). It became 'new inflation'.
Is it proven?
The fix works mathematically. But the specific 1982 models, tied to grand unified theories, predicted ripples far too strong and were dropped; the slow-roll principle survived in later models that fit Planck.
Open questions and mysteries
Why the hill should be so flat was unexplained then and still is. Linde himself soon moved on to chaotic inflation, which needs no false vacuum at all.