What Does It Mean to Explain the Whole Universe?

Here’s a strange thing about cosmology (the science of the universe as a whole): it’s the only science that studies something that happens only once. There’s only one universe. You can’t run an experiment on it, make a second copy, or see what happens if you change the initial settings. You can’t even see most of it—the universe is so huge that light from most places hasn’t had time to reach us yet.

And yet cosmologists claim to know quite a lot: that the universe started with a Big Bang about 13.7 billion years ago, that it’s mostly made of stuff we can’t see (dark matter and dark energy), that it’s expanding faster and faster, and that it will probably keep expanding forever.

How can they possibly know all that? And what kind of “explanation” could ever satisfy us when we’re talking about the origin of everything?

The Standard Model of the Universe

Let’s start with what cosmologists actually think they know. The standard picture—sometimes called the “concordance model” or ΛCDM model—goes roughly like this.

The universe began in an extremely hot, dense state. It has been expanding and cooling ever since. Gravity is the main force that shapes its large-scale structure, and general relativity (Einstein’s theory of gravity) describes how that works. The universe looks basically the same in every direction we look—it’s remarkably uniform. And it’s filled with a faint glow of radiation, the cosmic microwave background, which is the leftover heat from the early stages.

But here’s where it gets weird. To make this picture work with the observations, cosmologists have had to add three ingredients that nobody has ever directly detected:

Dark matter – some kind of invisible stuff that doesn’t emit or absorb light but has gravity. We can see its effects on how galaxies rotate and how light bends around clusters of galaxies. There’s about five times as much dark matter as ordinary matter.

Dark energy – something that’s pushing the universe apart faster and faster. We don’t know what it is. It acts like Einstein’s “cosmological constant” (a term he added to his equations and later called his biggest mistake). Dark energy makes up about 70% of everything in the universe.

The inflaton field – a hypothetical field that supposedly caused a brief period of ridiculously fast expansion in the first fraction of a second after the Big Bang. This period, called “inflation,” would explain why the universe is so uniform and flat.

So here’s the first puzzle: the best model we have for the universe depends on three things nobody has ever observed directly. Is that a problem? Or is it just how science works when you’re studying something unique?

Why It’s So Hard to Know

Cosmology faces a special kind of difficulty that other sciences don’t. Philosophers call it “underdetermination”—the idea that evidence alone isn’t enough to force you to choose one theory over another. In most sciences, you can do experiments to break the tie. In cosmology, you can’t.

There are at least three different ways this shows up.

We Can’t See the Whole Thing

The speed of light is finite, and the universe has only existed for 13.7 billion years. That means we can only see objects whose light has had time to reach us. There’s an “observable universe” beyond which we just can’t see. And there might be parts of the universe so far away that they aren’t even causally connected to us—meaning nothing that happens there could ever affect us, and nothing we do could ever affect them.

What’s out there beyond the horizon? We have no direct evidence. The universe could be infinite, or finite but much bigger than what we see, or it could have some strange shape that loops back on itself. Different possibilities lead to different global properties of the universe, but we might never be able to tell which is right.

The Physics Horizon

Most of the interesting things that happened in the early universe—like inflation, or the moments right after the Big Bang—involved energies far higher than any particle accelerator on Earth can produce. We can’t reproduce those conditions in a lab. The only way to test our ideas about them is through their effects on what we see in the sky today.

This means that a proposal like inflation is simultaneously offering an explanation and asking us to accept new physics we can’t independently test. It’s a package deal. As the Soviet physicist Yakov Zel’dovich memorably put it, the early universe is the “poor man’s accelerator”—but it’s also the only accelerator we’ve got.

Only One Universe

Here’s a problem that’s unique to cosmology. In particle physics, if you want to test a prediction, you run the experiment many times. You get a distribution of outcomes. You can calculate probabilities. In cosmology, you can’t. There’s exactly one universe, and you happen to be in it.

How do you know whether some weird feature you observe – say, a cold spot in the cosmic microwave background – is a genuine anomaly that requires explanation, or just a random statistical fluctuation? In any other science, you’d check other instances. In cosmology, you can’t. This is sometimes called the “cosmic variance” problem.

The Problem of Origins

Perhaps the deepest philosophical puzzle in cosmology is about how to explain the beginning of the universe.

The standard model says that as you go backward in time, the universe gets smaller, hotter, and denser. Eventually you hit a point – the Big Bang singularity – where the equations break down. The density goes to infinity. The temperature goes to infinity. Time itself seems to stop making sense. The singularity theorems (proved in the 1960s by Stephen Hawking and Roger Penrose) show that this isn’t just an artifact of the simple models. Under very general assumptions, any realistic universe that’s expanding now must have had a past singularity.

So what does that mean? It means that general relativity, our best theory of gravity, predicts its own failure. It can describe the universe up to the singularity, but not the singularity itself. To go further, we’d need a theory of quantum gravity, which we don’t yet have.

But there’s an even deeper question: What would count as an explanation of the origin of the universe? In other sciences, explanations work by placing an event in a larger context. But what larger context could there be for the universe itself? The universe is, by definition, everything that exists. So any explanation of its origin seems to require appealing to something outside it – which is a strange thing for science to do.

Some cosmologists have proposed that the universe “bounced” from a previous contracting phase, avoiding the singularity altogether. Others have suggested that our universe is just one of many in a “multiverse.” Others still think the initial state was very special and that we need a new kind of law – not a law that governs how things evolve, but a law that constrains what the initial state could be.

Notice that none of these proposals is testable in the usual sense. They’re more like philosophical commitments about what kind of explanation would be satisfying.

The Multiverse and Anthropic Reasoning

One of the most controversial ideas in cosmology is that our universe is just one among an enormous (possibly infinite) number of universes, each with different physical laws and properties. This is called the multiverse.

The idea arises naturally from some versions of inflation. In “eternal inflation,” different parts of space stop inflating at different times, creating “pocket universes” that are causally disconnected from each other. If inflation is eternal in this way, then there are infinitely many pocket universes, and anything that can happen does happen somewhere.

Why would anyone believe this? The main argument goes like this: The fundamental constants of physics seem “fine-tuned” for life. If you changed the strength of gravity by just a tiny amount, stars wouldn’t form. If you changed the mass of the neutron, there would be no carbon. And so on. If there’s only one universe, the fact that these constants are just right for us seems like an incredible coincidence. But if there are infinitely many universes with different constants, then it’s no surprise that at least one of them supports life – and of course we find ourselves in that one.

This is called “anthropic reasoning.” The idea, originally due to the physicist Brandon Carter, is that we have to account for the fact that our observations are filtered by the requirement that observers exist to make them. The astronomer Fred Hoyle once said the universe looks like “a put-up job” – as if it were designed for us. Anthropic reasoning is an attempt to explain why it looks that way without appealing to a designer.

But anthropic reasoning is philosophically messy. How do you define the “typical” observer? If there are infinitely many observers spread across an infinite multiverse, which one is typical? What counts as an observer anyway? And how do you assign probabilities when everything happens somewhere? These are known as the “measure problem” and the “reference class problem,” and nobody has solved them to general satisfaction.

Where This Leaves Us

Cosmology occupies a strange place among the sciences. It relies on enormous extrapolations of physics we’ve tested only in our little corner of the universe. It makes claims about things we can never observe directly. It struggles with the fact that its subject matter is unique. And its most interesting questions – about origins, fine-tuning, and the multiverse – push right up against the boundaries of what it means to give a scientific explanation.

None of this means cosmology is not a real science. The standard model has been remarkably successful, passing a huge number of observational tests. But it does mean that cosmology forces us to think about what we’re doing when we try to explain things. What counts as a good explanation? How do we know when we’ve really understood something? What should we do when the evidence underdetermines the theory?

These are not just scientific questions. They’re philosophical questions that the sciences force on us when they push to their limits. And cosmology, which tries to explain literally everything that exists, pushes harder than any other science.

Appendix

Key Terms

Term	What it does in the debate
Underdetermination	The idea that evidence alone may not be enough to choose between competing theories
Singularity	A point where the equations of physics break down, like the beginning of the Big Bang
Multiverse	The hypothesis that our universe is one among many, each possibly with different laws
Anthropic reasoning	Taking into account that our observations are filtered by the fact that we exist
Inflation	A proposed period of extremely rapid expansion in the first fraction of a second after the Big Bang
Dark matter	Invisible stuff that has gravity but doesn’t interact with light
Dark energy	A mysterious force causing the universe’s expansion to accelerate
Fine-tuning	The observation that fundamental constants seem precisely set to allow life to exist
Copernican principle	The assumption that our location in the universe is not special

Key People

Stephen Hawking – a British physicist who proved theorems showing that the universe must have had a beginning. He later changed his mind and proposed ways the universe could be eternal.
Roger Penrose – a British physicist who argued that the initial state of the universe must have been extremely special (low entropy) to explain the arrow of time.
Brandon Carter – a French physicist who introduced the “anthropic principle” in the 1970s.
Steven Weinberg – an American physicist who used anthropic reasoning to predict the value of the cosmological constant before it was measured.
Yakov Zel’dovich – a Soviet physicist who called the early universe the “poor man’s accelerator” because it lets us study physics at energies we can’t reach on Earth.

Things to Think About

If there were infinitely many universes with different laws, would it make sense to say that ours is “fine-tuned”? Or does the multiverse just push the question back one step?
Suppose you could prove that the universe is infinite. What would that mean for how we should think about the probability of any event happening somewhere?
How do we decide whether a new scientific idea (like dark matter or inflation) is a legitimate discovery or just a “saving the phenomena” – a way to make the data fit our assumptions?
If science can’t explain why the universe exists at all, or why it has the particular laws it does, is that a failure of science? Or is it just recognizing the limits of what science can do?

Where This Shows Up

Particle physics – the search for dark matter particles in underground detectors is closely tied to cosmology.
Climate science – similar questions about underdetermination arise when you can only run one experiment (the Earth).
Evolutionary biology – anthropic reasoning is similar to the idea that we shouldn’t be surprised to find ourselves in a world that supports life, because if it didn’t, we wouldn’t be here to notice.
Everyday reasoning – a version of the anthropic principle comes up whenever someone says “isn’t it amazing that the universe is so perfectly suited for us?” The question is whether that amazement is justified or naive.