From Nothing to Everything: How Our Universe Emerges from the Dance of Vacuum & Energy

Ask a physicist how the universe began and you get a careful answer about the first fraction of a second. Ask what came before, or why there is anything at all, and the careful answer runs out. That is not a smaller question. It is the largest one we have, and physics has not answered it.

So we measured how hard we are trying. Searching arXiv year by year for papers whose abstracts mention quantum cosmology returns 86 in 2010 and 67 in 2025. In between: 75, 91, 65, 67, 74, 79, 53, 91, 94, 70, 82, 64, 66 and 69. Sixteen years, an average near 74 a year, no trend either way.

That flat line is the most interesting number here, because over the same years the fields around it exploded. The question did not stop mattering. It stopped receiving new data, and a field without new data does not grow — it circulates.

What Physicists Mean When They Say “Nothing”

The word does most of the damage. In ordinary speech, nothing means the absence of everything. In physics it means something narrower and stranger: the quietest state the fields can settle into. That state has structure and measurable effects. It is not an absence.

This is not pedantry. Getting it wrong is how popular accounts go astray. A universe emerging from the quantum vacuum has not emerged from nothing in the everyday sense. It emerged from an arrangement of physical law, fields and space — and whether that arrangement needs explaining is a separate question physics cannot settle.

Fields, not particles, are what actually exist

Physics describes the world as fields rather than objects. An electron is not a tiny ball but a ripple in a field that exists everywhere, and the same goes for every particle we know. Empty space is simply the state where all those fields are as quiet as they can be.

Quantum mechanics will not let them be perfectly still, so some residual jitter always remains — usually called zero-point energy. It is surrounded by confusion, some of it commercial, which we separated from the marketing in our piece on zero-point energy.

Virtual particles are bookkeeping, not weather

Popular accounts describe the vacuum as boiling with particles popping in and out of existence. Vivid, and misleading. Those particles are bookkeeping entries in a calculation, not objects that could in principle be photographed.

What is real is the effect. Work out how an electron behaves and the answer only matches experiment when those entries are included. The vacuum contributes to things we can measure, even where the picture we paint of it does not.

The Evidence That the Vacuum Is Real

Three measurements carry most of the weight, and they are worth stating precisely, because “the vacuum has energy” sounds speculative and is not.

The first is the Lamb shift. In hydrogen, two states that theory says should have identical energy do not. Lamb and Retherford found the splitting in 1947, measurable by microwave spectroscopy. The gap exists because the electron interacts with vacuum fluctuations of the electromagnetic field, and quantum electrodynamics predicts its size correctly.

The second is the Casimir effect. Place two uncharged metal plates very close together and the space between them can hold less field activity than the space outside, so they are pushed together. Steven Lamoreaux measured it directly in 1997 between roughly 0.6 and 6 micrometres, matching theory to within a few per cent.

The third is how strongly an electron responds to a magnetic field. Theory predicts one value; the measurement differs in the twelfth decimal place, and the vacuum accounts for the difference exactly. It is among the most precise agreements in science, and the strongest evidence that the vacuum is a physical thing.

The Ledger: Why a Universe Might Sum to Zero

If the universe began with no stock of energy, the books have to balance. Everything you can point at carries positive energy, so something must carry an equal negative amount.

Gravity is the candidate. Pulling two objects apart takes energy, which means a bound system holds less than a scattered one — gravitational energy counts as negative. Extend that to every mass in the observable universe and the negative total is enormous.

The zero-energy idea says the two sides cancel exactly — matter and radiation on one, gravity on the other, summing to nothing. If true, creating a universe would break no conservation law, because there would be nothing to conserve. Elegant, and it has a serious problem.

Pair production makes the principle concrete

Turning energy into matter is not theoretical. An energetic photon passing near an atomic nucleus can become an electron and a positron, routinely. Paul Dirac predicted antimatter from the shape of his equation before anyone saw it, a story we tell in our piece on Dirac and the prediction of antimatter.

Pair production shows matter is a form energy can take, not a separate substance needing separate creation. That premise the zero-energy argument needs is solidly established. It is also the easy part.

Why “The Total Energy” Is Harder to Define Than It Sounds

The argument rests on adding two quantities and getting zero. In general relativity that addition is not straightforward, and the answer may not even be well defined. This is the weakest joint in the chain, and popular accounts almost never mention it.

Energy conservation is not a free-standing law. It follows from the rules of physics being the same today as yesterday — and an expanding universe breaks that sameness, because space itself is changing. We can add up the energy of an isolated star, because there is empty space around it to measure against. The universe has no outside, so there is nowhere to stand and add everything up.

So “the total energy of the universe is zero” is not straightforwardly true or false. Under some reasonable ways of setting up the sum, the total is not zero so much as undefined. Several physicists read the claim as a suggestive idea rather than a proven one.

The Measurement That Makes the Idea Testable

The cancellation only works for a particular geometry. In general relativity the total energy density fixes spatial curvature, so a universe summing to zero must be spatially flat: Euclidean at large scales, parallel lines staying parallel, triangles closing at 180 degrees.

That turns a philosophical proposal into a measurement. Curvature can be read off the sky, because a curved universe would change how large the features in the microwave background appear.

The Planck 2018 cosmological results, combined with baryon acoustic oscillation data, give Ω_K = 0.001 ± 0.002. The universe is flat to within about a fifth of one per cent. That single number is the strongest observational support the zero-energy hypothesis has, and it is a prediction confirmed rather than a fit after the fact.

Two cautions belong beside it. Consistent with zero is not the same as zero; the data still allow a very slight curve too large to detect. And inflation predicts flatness for its own reasons, so a flat universe supports both ideas without separating them.

Inflation: The Mechanism That Would Have Done It

Two things about our universe resist explanation by ordinary expansion. It is remarkably even: the microwave background is almost the same temperature everywhere, including regions that could never have exchanged anything. And it is flat, which in ordinary cosmology is unstable — any early deviation grows, so flat today means impossibly flat at the start.

Alan Guth’s proposal, set out in his 1981 paper, solves both. If the early universe briefly expanded at an enormous rate, one tiny connected patch was stretched to cover everything we see. Evenness follows because it was all in contact beforehand; flatness follows because stretching any surface far enough makes it look flat.

Inflation earns its keep on a third count, and this is what made it dominant. The quantum jitter would have been stretched too, leaving faint denser and thinner patches that gravity later pulled into galaxies. The predicted pattern matches what Planck measured. Everything we see, including the stars that formed, traces back to that magnified noise.

What the Microwave Background Actually Shows

For the first few hundred thousand years the universe was too hot for atoms to hold together. Loose electrons scatter light relentlessly, so the universe glowed like fog rather than being clear. As it cooled, electrons and protons joined into hydrogen, the fog lifted, and light travelled freely for the first time.

That release is what we detect as the cosmic microwave background. It is not a picture of the Big Bang. It is a picture of the moment the universe turned clear, about 380,000 years in, and it is the earliest thing we can actually see. Everything before it has to be worked out indirectly.

Why the temperature map carries so much information

The background is astonishingly even — the same temperature in every direction to about one part in a hundred thousand. The interest is in that last digit.

Before the fog lifted, gravity was pulling matter into clumps while the pressure of trapped light pushed back, setting off sound waves through the early universe. They all began at roughly the same moment, so they moved in step, and when the fog lifted they were frozen mid-swing.

The spacing of those frozen waves across the sky tells you what the early universe was made of and what shape it has. A curved universe would magnify or shrink them, so measuring their apparent size measures curvature. This is where the flatness figure comes from — an angle, measured.

The Planck 2018 release fits six numbers to this pattern and derives the rest: the age of 13.8 billion years, how much matter and dark energy there is, and the slight tilt inflation predicted. Impressive physics, with an important limit. It tells us what the universe held at 380,000 years, and reaches earlier only through models.

The Low-Entropy Problem Nobody Has Solved

There is a deeper puzzle than what caused the beginning, and it gets a fraction of the attention. The early universe was not merely hot and dense. It was in an extraordinarily special state, and no origin account explains why.

Entropy increases. Run that backwards and the past must have had lower entropy than the present, lower the further back you go. The beginning must therefore have been extraordinarily low-entropy — which is to say extraordinarily improbable, selected from an overwhelmingly larger set of possibilities.

The counter-intuitive part is that a hot, even plasma sounds disordered, because a gas spread evenly through a room is as disordered as it gets. Gravity turns this around. When gravity dominates, smooth is the orderly state and clumpy the disordered one. The smooth early universe was therefore extraordinarily orderly.

Roger Penrose has pressed this hardest, arguing in The Emperor’s New Mind that the odds against such a starting state are so long the number cannot be written out inside a universe this size. It is a rough estimate and the figure is debated, but nobody disputes that the fine-tuning is severe.

Inflation is often credited with fixing this, and does not. It explains why our region is smooth given a patch that was already inflating — but that patch had to start orderly for inflation to begin. The puzzle moves rather than dissolving.

We go deeper in our piece on the arrow of time. Here the point is narrow: any complete story of cosmic origins owes an account of why the initial state was so special, and none of the proposals in this article provides one.

The Problem With Inflation: 118 Models, One Universe

Inflation is not a theory the way general relativity is a theory. It is a framework. General relativity has no dials: tell it what the matter is doing and the equations answer. Inflation needs a field with a particular shape to its energy, and nothing fixes that shape.

Cosmologists have taken this seriously enough to count. The Encyclopædia Inflationaris tests the simplest family of inflation models against Planck data, and its software supports 118 of them. One hundred and eighteen ways for the universe to have begun, all within the simplest family. The trouble is what would tell them apart.

The number that would settle it has never been measured

Inflation makes one prediction that could separate the models sharply: gravitational waves from the first instant. That violent expansion should have shaken space itself, leaving a faint twist in the microwave background. Its strength is written r, and different models predict values differing enormously.

No B-mode signal of primordial origin has been detected. BICEP/Keck’s 2021 analysis, combining its own data with Planck, places r below 0.036 at 95% confidence. That is a bound, not a measurement. It has ruled out a substantial fraction of the 118, including textbook favourites, and leaves the remainder standing.

The situation is unusual. Inflation’s one testable prediction has been squeezed downward for twenty years without ever being confirmed, and each squeeze kills models rather than the framework. A sceptic calls that unfalsifiable; a defender calls it normal science narrowing the field. Both readings come from the same data.

Did Time Itself Have a Beginning?

A common move in popular cosmology is to say inflation removes the need for a beginning: if inflation is eternal, universes bubble off forever and the question dissolves. The mathematics does not support this as cleanly as the retelling suggests.

Borde, Guth and Vilenkin showed in 2003 that any spacetime expanding on average throughout its history cannot be extended infinitely into the past. The theorem is remarkably general: it assumes neither Einstein’s equations nor any particular matter content, and it covers inflationary models including eternal ones. Given only average expansion, past-directed paths terminate.

What it does not say matters just as much. It does not prove the universe began to exist, and it does not prove a creator. It shows that one kind of description cannot run backwards forever, so it fails at some finite point in the past. What replaces it there is exactly what nobody knows — the theorem marks the edge of a map, not the edge of reality.

The Quantum Proposals, and Their Standing

Two serious attempts exist to describe the origin quantum-mechanically. Both date from the early 1980s, both are mathematically substantial, and neither has been tested. It is not obvious how either could be.

The no-boundary proposal

James Hartle and Stephen Hawking’s Wave Function of the Universe (1983) treats the whole universe as a quantum system and asks what holds at its edge. Their answer: there is no edge. Near the origin, time behaves like another direction of space, so there is no first moment, in the way there is no point south of the South Pole.

It is elegant, and it dissolves the question of what came before rather than answering it. No boundary, no before. Hawking spent much of his later career on the consequences, covered in our piece on A Brief History of Time, and the same instinct runs through his work on black hole radiation.

Tunnelling from nothing

Alexander Vilenkin’s Creation of Universes from Nothing (1982) takes another route: a universe appearing the way a particle sometimes slips through a barrier it has no business crossing. The starting state contains no space or time at all.

The two predict slightly different universes, and in principle the difference might show in the pattern of early ripples. In practice it sits far below anything a planned experiment could measure, and both depend on a quantum theory of gravity nobody has finished writing.

It is worth being blunt about status. These are not fringe ideas — cited thousands of times, by physicists of the first rank. They are also, forty years on, unsupported by a single measurement. The flat arXiv line at the top of this article is what that looks like from outside.

The Number That Refuses to Behave

If the vacuum has energy, that energy should gravitate. General relativity says every form of energy curves spacetime, and vacuum energy is no exception. Calculating how much the quantum vacuum contributes is therefore a direct test of whether we understand it at all.

We fail it comprehensively. Add up what the known fields should contribute and the answer is vastly larger than what is measured — the gap is roughly a one followed by 120 zeros. Weinberg’s 1989 review laid out the problem, and nothing fundamental has changed since.

This is the largest quantitative disagreement between theory and observation in physics, and it sits directly underneath every claim in this article. The vacuum whose fluctuations we invoke to explain the origin of the universe is the same vacuum whose energy we cannot calculate to within a hundred and twenty orders of magnitude.

Nor is the measured value zero, which would at least hint at a hidden rule. It is small and positive, and it drives the accelerating expansion we call dark energy. Any full account owes us both why it is so small and why it is not quite zero.

In March 2014 the BICEP2 collaboration announced a detection of those primordial gravitational waves, at a press conference, before peer review. It was reported worldwide as direct evidence of inflation and a probable Nobel Prize.

It was dust. Polarised emission from aligned grains in our own galaxy produces a signal in the same band with a similar signature, and BICEP2 had underestimated it. A joint analysis with Planck in 2015 showed the observed signal was consistent with galactic foreground alone. The claimed detection did not survive.

The episode is worth studying, and not because anyone was foolish. The correction came from the field itself, fast, published in the same journal as the claim. The failure was in judging the dust rather than in the instrument, which is still the hard part of this measurement today.

The third lesson belongs here. The result everyone wanted was announced, celebrated and withdrawn inside eleven months, and the field’s publication rate barely noticed. A real detection would have transformed the subject; its absence changed nothing, because there was nothing new to work on either way.

How We Would Know If We Were Wrong

A claim that cannot fail any test is not doing scientific work. So it is worth setting out, for each proposition here, what observation would count against it — and where nothing would, saying so.

Claims with live tests

• The universe is flat. A measurement of curvature clearly different from zero would disprove it. Current limits are tight, and future surveys will tighten them.
• Inflation happened. Detecting those primordial gravitational waves would strongly support it; failing to find them down to r of about 0.001 would kill the simplest models and leave the framework in trouble.
• The vacuum has real structure. Tested repeatedly and passed. A failure of quantum theory at higher precision would count against it; none has appeared.
• Galaxies grew from quantum ripples. Testable through how galaxies cluster and how the microwave background varies. The wrong statistical signature would rule out the simplest models.

Claims with no current test

• The universe’s total energy is exactly zero. No experiment is proposed, and part of the difficulty is that the sum may not be definable.
• The no-boundary proposal, or tunnelling from nothing. Both predict in principle; neither predicts anything we could tell apart with foreseeable instruments.
• A multiverse of other regions. Some inflation models imply endless other universes. No proposed observation has found one.
• Why there is something rather than nothing. No observation bears on it. This is the one the older traditions identified correctly as being of a different kind.

Sorting claims this way is more useful than ranking them by how exciting they sound. The first list is science in progress. The second contains ideas that may well be true, held by serious people for serious reasons, but currently doing philosophical rather than scientific work. Presenting the second as the first is the central failure of popular cosmology.

What “Nothing” Meant Before Physics

The question of why anything exists is older than the equipment. Every literate civilisation produced an answer, and the answers are more varied, and considerably more careful, than the habit of treating them as naive first drafts allows.

This is not an argument that ancient texts anticipated modern cosmology. They did not, and claims that some scripture encoded the Big Bang work backwards from the answer. The reason to look is different: these are the most sustained attempts people made before measurement was possible.

The hymn that ends in doubt

The most striking is the Nāsadīya Sūkta, the creation hymn at Rigveda 10.129, composed somewhere in the second millennium BCE. It opens by refusing both alternatives: neither existence nor non-existence, neither death nor immortality, no distinction between day and night. What existed, it says, was breathing without breath.

What makes it extraordinary is not its cosmology but its epistemology. It does not end in assertion. It ends by asking who could possibly know, notes that the gods came after the creation and so cannot have witnessed it, and closes on the possibility that even the highest overseer does not know. In Ralph Griffith’s 1896 translation:

“Who verily knows and who can here declare it, whence it was born and whence comes this creation? The Gods are later than this world’s production. Who knows then whence it first came into being?”— Rigveda 10.129, translated by Ralph T. H. Griffith, 1896. For a modern scholarly rendering see Jamison and Brereton’s Oxford translation.

Three thousand years later, that is close to where the physics sits. The hymn identifies the structural problem exactly: any witness to the origin is part of what originated, so there is no outside vantage point. Modern proposals try to work around that constraint with mathematics rather than resignation.

Creation from nothing was a late and contested idea

Creation from nothing at all is often assumed to be the default religious position. Historically it is neither default nor early. Genesis opens on a formless earth with waters already there, closer to ordering than to creating from absence. It was worked out as formal doctrine in the first centuries CE, partly against Greek philosophy, which broadly held that nothing comes from nothing.

Greek thought split. Parmenides argued you cannot coherently speak of non-being, so an absolute beginning makes no sense. Aristotle held the cosmos had always existed. The atomists proposed everlasting particles in everlasting void. Each makes a move still made today: something has to be taken as given, and the argument is about what.

Buddhist and Jain traditions largely declined the question. Several Buddhist texts class the origin of the world among matters not conducive to liberation and set it aside deliberately — not from incuriosity, but on the view that the question is malformed. That is closer to the no-boundary proposal than the distance suggests.

Four more answers, and the shape they share

Chinese cosmology took another route. The Daodejing describes something undifferentiated and complete existing before heaven and earth, from which the one arises, then the two, then the ten thousand things. The emphasis falls on differentiation out of an undivided state rather than creation out of absence — closer in structure to a phase transition than to a manufacture.

Egyptian cosmogony began with Nun, the primordial waters: not nothing, but an inert chaos from which a mound emerges. Norse accounts start with Ginnungagap, a yawning void bounded by fire and ice. In both, the starting point is a formless something rather than an absence — the same move Genesis makes, and the same move the quantum vacuum makes.

Medieval Islamic philosophy took the opposite line: an endless chain of past events is impossible, so the universe must have had a beginning and therefore a cause. Al-Ghazālī developed it in the eleventh century, and it has been revived in modern philosophy of religion, often with the Borde-Guth-Vilenkin theorem attached. The theorem does not carry that weight, but whether an infinite past makes sense is still open.

What the comparison actually shows

Read together, the older traditions cluster around a few answers. Something has always existed. Something outside time holds up what is inside it. The question is badly posed. Or nobody can know. Modern cosmology, with the mathematics stripped away, offers the same short list.

That convergence is not evidence the ancients were doing physics. It is evidence the constraint is logical rather than technological. You cannot get an explanation without something to explain from, and whatever you start with is unexplained. Better instruments move where the line falls. They have not moved the fact that there is one.

Where Science Stops and Something Else Begins

The most useful criticism of the “universe from nothing” genre came from a philosopher of physics rather than a theologian. David Albert, reviewing the book in The New York Times in 2012, made a point that is difficult to answer.

A vacuum in physics, he wrote, is no less an arrangement of physical stuff than a table is. The true equivalent of there being nothing at all is not one arrangement rather than another — it is the fields not being there at all. A theory that starts with fields and derives a universe has explained how this universe arose from an earlier physical state. It has not explained why there is any.

This is not an anti-science argument, and Albert is no creationist. It is a claim about the scope of a kind of explanation. Physics explains transitions between states. Why there is any state at all is not a transition, and no amount of better physics obviously converts it into one.

Two responses are respectable. The first is that the question is malformed — “why is there something rather than nothing” may ask nothing answerable, the way “what is north of the North Pole” does. The second is that it is real but outside the reach of experiment. What is not respectable is presenting the first as a finding, or the second as physics.

The fine-tuning question, and why it resists both sides

Next to the origin question sits another that generates more heat than light. Several constants of nature sit in narrow ranges that allow complex structure. Change the force holding nuclei together by a few per cent and stars make a different periodic table.

Three explanations compete, each with a real weakness. Design puts the tuning outside physics, explaining it at the cost of raising the same question one level up. The multiverse makes it a selection effect, coherent but unprovable. Necessity holds that a deeper theory will show the constants could not have been otherwise — possible, with no candidate theory behind it.

The honest observation is that this is an argument about probability, and we have one universe. Putting odds on a single case with nothing to compare it against is not a calculation; it is an intuition wearing the clothes of one. That cuts against confidence in every direction, which is why the argument has run four decades without moving anybody.

The Honest Position

Here is what can be said with confidence. The vacuum has real, measurable structure, confirmed three separate ways. The universe is flat to within a fifth of a per cent. Matter and energy convert into each other, so matter needs no separate creation. The universe was hot and dense 13.8 billion years ago and has expanded since.

Here is what cannot. Whether the total energy is exactly zero, which depends on defining gravitational energy in general relativity. Which of the 118 inflation models is right, or whether any is. What preceded the earliest moment we can model. Why the vacuum energy has the value it does. Whether ultimate origin has a physical answer at all.

The gap between those two lists has barely narrowed in forty years, and the flat publication record is the honest measure of it. That is not a failure of the people doing the work. It is what a question looks like when the observations that would settle it are out of reach. The Hubble tension shows what progress looks like when data do arrive. Nothing like it has happened here.

The universe may well have emerged from a quantum vacuum, through a process that breaks no law of physics and needed no cause outside itself. It is coherent, and supported by real measurements along the chain. It is also not established, its central mechanism has 118 versions, and “nothing” in the popular telling does not mean what readers assume. Holding all of that at once is the accurate position.

Frequently Asked Questions

Sources

• Planck Collaboration — Planck 2018 results VI: Cosmological parameters, Astronomy & Astrophysics (2020)
• Martin, Ringeval & Vennin — Encyclopædia Inflationaris, Physics of the Dark Universe (2014)
• BICEP/Keck Collaboration — Improved Constraints on Primordial Gravitational Waves, Physical Review Letters 127 (2021)
• BICEP2/Keck and Planck Collaborations — Joint Analysis of BICEP2/Keck Array and Planck Data, Physical Review Letters 114 (2015)
• Alan H. Guth — Inflationary universe: A possible solution to the horizon and flatness problems, Physical Review D 23 (1981)
• Borde, Guth & Vilenkin — Inflationary Spacetimes Are Incomplete in Past Directions, Physical Review Letters 90 (2003)
• Hartle & Hawking — Wave function of the Universe, Physical Review D 28 (1983)
• Alexander Vilenkin — Creation of universes from nothing, Physics Letters B 117 (1982)
• Steven Weinberg — The cosmological constant problem, Reviews of Modern Physics 61 (1989)
• Gibbons & Hawking — Cosmological event horizons, thermodynamics, and particle creation, Physical Review D 15 (1977)
• S. K. Lamoreaux — Demonstration of the Casimir Force in the 0.6 to 6 µm Range, Physical Review Letters 78 (1997)
• Lamb & Retherford — Fine Structure of the Hydrogen Atom by a Microwave Method, Physical Review 72 (1947)
• David Albert — On the Origin of Everything, The New York Times (25 March 2012)
• Roger Penrose — The Emperor’s New Mind, Oxford University Press (1989)
• Simons Observatory Collaboration — The Simons Observatory: science goals and forecasts, JCAP 02 (2019) 056
• Jamison & Brereton — The Rigveda: The Earliest Religious Poetry of India, Oxford University Press (2014)
• CERN — The early universe
• CERN — Antimatter
• European Space Agency — Planck mission