The same trick, taken to the limit
In the cosmology rung just behind you, you ran the film backward once and reached a powerful conclusion: because space itself is expanding, the universe was denser and hotter long ago, and that hot dense beginning — the [[hot-big-bang|hot Big Bang]] — left two relics we can still measure, the microwave glow and the primordial helium. This rung does nothing new in principle. It simply keeps winding the same handle, frame by frame, all the way back toward time zero, and asks at each step: how hot, how dense, and do we still trust the physics here?
Two warnings carry over, and they are worth restating because the whole rung depends on them. First, none of this is an explosion from a point. The hot early state filled all of space at once; there is no center, no edge, no surrounding void to expand into. Second, the cooling is a direct consequence of the stretching: lengthen every wavelength of light along with the expanding space, and the radiation that fills the universe drops in temperature. So 'winding back' literally means undoing that stretch — squeezing the light back to shorter wavelengths, and the universe back to a furnace.
The timeline, read forward
It is easiest to lay the eras out forward, from the earliest instant to today, and then remember we discovered them by going the other way. The very first sliver — before about 10⁻⁴³ of a second, the [[planck-era|Planck epoch]] — is the frontier, where our theories of gravity and of the quantum world both apply at once and neither alone suffices; we will be honest below about how little we can say there. Then, within the first tiny fraction of a second, comes a proposed burst of staggeringly fast stretching called [[cosmic-inflation|cosmic inflation]] — well-motivated and supported by clues, but still a model under active test, not settled fact.
From there the universe is a seething soup of fundamental particles, hotter than any star's core. As it expands and cools through the first second, it crosses a series of thresholds: a slight excess of matter over antimatter survives (baryogenesis, itself only partly understood), and the soup eventually settles into the protons and neutrons that the next chapters care about. Between roughly one second and a few minutes, those particles fuse — [[big-bang-nucleosynthesis|Big Bang nucleosynthesis]], which you met as a pillar of evidence — leaving the universe about three-quarters hydrogen and one-quarter helium, and then the furnace cools too far to fuse and shuts off.
For the next several hundred thousand years the universe is a hot, opaque plasma — bright but blind, because light keeps scattering off loose electrons and cannot travel far. Then, around 380,000 years in, cooling to roughly 3,000 kelvin lets electrons and protons combine into neutral atoms: [[recombination|recombination]] clears the fog, and the light set free at that instant becomes, after billions of years of stretching, the cosmic microwave background we now measure at about 2.7 kelvin. After that the universe is transparent but starless — the long, dark stretch we will cover next as the [[cosmic-dark-ages|cosmic dark ages]] — until gravity gathers the first gas into the [[first-stars|first stars]], and the cosmic dawn breaks.
A timeline you can hold in your head
The numbers span an almost comic range — from a tiny fraction of a second to billions of years — so it helps to see them stacked up. Notice how lopsided the scale is: nearly all the named eras happen in the first second, and yet that first second sets the stage for everything after. The temperatures are approximate and the boundaries are soft transitions, not sharp walls, but the ordering and the rough scales are solid and well-measured for everything from nucleosynthesis onward.
COSMIC TIMELINE (winding FORWARD from t = 0)
time since start event / era rough temperature confidence
------------------ --------------------- ----------------- ----------
< 10^-43 s Planck epoch ~10^32 K frontier
~10^-36 to 10^-32 s inflation (proposed) extreme model
first ~1 second particle soup, > 10^10 K well-tested
baryogenesis
~1 s to ~3 min nucleosynthesis ~10^9 K well-tested
(forges H + He)
~380,000 years recombination ~3,000 K measured (CMB)
-> universe clears
~0.4 to ~1 billion yr cosmic dark ages falling inferred
-> first stars, dawn
~13.8 billion years today ~2.7 K (the CMB) measured
(temperatures approximate; boundaries are gradual transitions, not sharp walls)Read that table again with one question in mind: how far back can we actually check our story? The cleanest answer is that two independent relics pin it down. The light-element recipe from nucleosynthesis tells us the universe behaved exactly as known physics predicts when it was about one second old and billions of kelvin. The microwave background gives us a detailed snapshot at 380,000 years. Everything between and after those two anchors is on firm experimental ground. It is what lies before them that we must treat with more care.
Where the map runs out
Now we reach the part this guide most wants you to take seriously: the boundary between tested physics and honest speculation. Wind the clock back past one second and we lose the steadying handrails of direct relics, but our well-tested particle physics still applies for a while longer, into energies probed in laboratory accelerators. Keep winding, though, and you approach a temperature so extreme — around 10³² kelvin, at the [[planck-era|Planck epoch]] — that gravity itself becomes a quantum affair, and we simply do not have a theory that works there. General relativity, our theory of gravity, and quantum mechanics, our theory of the very small, give contradictory answers. The map runs out.
This is why you should be wary of any confident sentence about 'the moment of the Big Bang' or 'what happened before time'. The honest scientific position is humbler and more interesting: the hot Big Bang model describes a universe expanding and cooling from a hot dense state, and it works beautifully back to a tiny fraction of a second. It does not describe the very first instant, and it does not say 'everything came from nothing'. The phrase 'the singularity at t = 0' is best read not as a real physical event we understand, but as a flashing sign that says: known physics fails here, bring a better theory.
Why the earliest instants are worth the trouble
It would be reasonable to ask why anyone bothers with an era we cannot directly observe. The answer is that the earliest moments wrote the initial conditions for everything since. The exact balance of matter over antimatter, the seeds of density that later grew into galaxies, the overall smoothness and flatness of the cosmos — all of these were set in that first sliver of a second and then merely played out. To explain the universe we do see, we are forced to ask what the universe we cannot see was doing. That is the real reason inflation was proposed: not for spectacle, but to solve concrete puzzles the later, tested universe hands us.
There is also a beautiful loop hidden here, and it is the deepest reason cosmology works at all. The early universe was so hot and dense that it was governed by the physics of the very small — the same particle physics we study in accelerators on Earth. So the largest thing there is, the whole cosmos, becomes a laboratory for the smallest things there are. To read the first second, we lean on what we learned smashing particles together; and the early universe, in turn, reaches energies no earthly machine can match. The big and the small meet in the early universe, which is exactly what makes it the edge of known science — and the most exciting place a curious mind can stand.
That is the map for the rest of this rung. Having wound the universe all the way back and marked where confidence ends, we can now walk forward through the eras one at a time: the proposed burst of inflation and the puzzles it was built to solve; the first minutes that forged the elements in your body; the relic glow and the rich information hidden in its tiny temperature ripples; and finally the long dark ages giving way to the first stars and the cosmic dawn. Keep the confidence column from the table in mind throughout — it is the difference between knowing and guessing, and this field is honest about which is which.