Illustration by Claire Scully


In the beginning

Cosmology has been on a long, hot streak, racking up one imaginative and scientific triumph after another. Is it over?

by Ross Andersen + BIO

Illustration by Claire Scully

One crisp day last March, Harvard professor John Kovac walked out of his office and into a taxicab that whisked him across town, to a building on the edge of the MIT campus. People were paying attention to Kovac’s comings and goings that week. He was the subject of a fast-spreading rumour. Kovac is an experimental cosmologist midway through the prime of a charmed career. He did his doctoral work at the University of Chicago and a postdoc at Caltech before landing a professorship at Harvard. He is a blue chip. And since 2009, he has been principal investigator of BICEP2, an ingenious scientific experiment at the South Pole.

Kovac had come to MIT to visit Alan Guth, a world-renowned theoretical cosmologist, who made his name more than 30 years ago when he devised the theory of inflation. Guth told Kovac to take the back steps up to his office, to avoid being seen. If Guth’s colleagues caught a glimpse of the two men talking, the whispers swirling around Kovac would have swelled to a roar.

The science of cosmology has achieved wonders in recent centuries. It has enlarged the world we can see and think about by ontological orders of magnitude. Cosmology wrenched the Earth from the centre of the Universe, and heaved it, like a discus, into its whirling orbit around one unremarkable star among the billions that speed around the black-hole centre of our galaxy, a galaxy that floats in deep space with billions of others, all of them colliding and combining, before they fly apart from each other for all eternity. Art, literature, religion and philosophy ignore cosmology at their peril.

But cosmology’s hot streak has stalled. Cosmologists have looked deep into time, almost all the way back to the Big Bang itself, but they don’t know what came before it. They don’t know whether the Big Bang was the beginning, or merely one of many beginnings. Something entirely unimaginable might have preceded it. Cosmologists don’t know if the world we see around us is spatially infinite, or if there are other kinds of worlds beyond our horizon, or in other dimensions. And then the big mystery, the one that keeps the priests and the physicists up at night: no cosmologist has a clue why there is something rather than nothing.

To solve these mysteries, cosmologists must make guesses about events that are absurdly remote from us. Guth’s theory of inflation is one such guess. It tells us that our Universe expanded, exponentially, a trillionth of a trillionth of a trillionth of a second after the Big Bang. In most models of this process, inflation’s expansive kick is eternal. It might cease in particular parts of the cosmos, as it did in our region, after only a fraction of a second, when inflation’s energy transformed into ordinary matter and radiation, which time would sculpt into galaxies. But somewhere outside our region, inflation continued, generating an infinite number of new regions, including those that are roaring into existence at this very moment.

Not all these regions are alike. Quantum mechanics puts a slot-machine spin on the cosmic conditions of every region, so that each has its own physical peculiarities. Some contain galaxies, stars, planets, and maybe even people. Others are entirely devoid of complex structures. Many are too alien to imagine. The slice of space and time we can see from Earth is 90 billion light years across. Today’s inflationary models tell us that this enormous expanse is only one small section of one tiny bubble that floats along in a frothy sea whose proportions defy comprehension. This vision of the world is wondrous, in its vastness and variety, in the sheer range of possibilities it suggests to the mind. But could it ever be proved?

John Kovac had come to MIT to deliver good news. In 2009, Kovac and colleagues installed a telescope at the bottom of the Earth, and with it caught some of the oldest light in the Universe. He’d come to tell Guth that this light bore scars from time’s violent beginning, scars that strongly suggested the theory of inflation is true.

If the BICEP2 discovery held up, it would mint Nobel Prizes, and his would be the first

That same week, Chao-Lin Kuo, one of Kovac’s collaborators, paid a similar visit to Andrei Linde, another pioneer of inflationary theory. Kuo surprised Linde at his home, not far from Stanford’s sunny Silicon Valley campus. He brought a cameraman to record the moment for posterity, and a bottle of Champagne. When he knocked on Linde’s door, Linde and his wife answered. ‘I have a surprise for you,’ Kuo said. Linde’s wife, Renata Kallosh, who is also a physicist, was the first to react. She closed her eyes and hugged Kuo. Linde was stunned. ‘What?!’ he said, before asking Kuo to repeat the data. Soon, they were drinking Champagne, and Linde was effusive. ‘If this is true,’ he said, ‘this is a moment of understanding of nature of such magnitude, it just overwhelms.’

Back at MIT, Guth grilled Kovac with question after question, feeling around for weaknesses in the data. Guth would want to be sure. If the BICEP2 discovery held up, it would mint Nobel Prizes, and his would be the first. It would mean that an extraordinary idea entered human culture by way of his imagination. After more than an hour of interrogation, Guth relented. He could find no fault with the data.

A week later, the BICEP2 team went public, sparking a rare media event for the cerebral science of cosmology. In a front-page story for The New York Times Magazine, Kovac was quoted saying there was a one-in-10-million chance that the result was a fluke. The MIT physicist Max Tegmark told the Times that Kovac’s work would be one of the greatest discoveries in the history of science, ‘if [it] stays true’. For a time, it seemed as though cosmology had once again delivered a new cosmos.

When you read that word cosmos, you might begin to imagine the most expansive physical world your mind can build. Deep fields of glittering, star-filled galaxies stretching out in every direction, and maybe into forever. But even that image represents only the barest sliver of what is meant by ‘cosmos’. To build a cosmos, you have to extend your imagination to all of space and all of time. Only one of Earth’s creatures can pull off that cognitive trick. All living things are attuned to their environment: bacteria can sense chemical shifts in their immediate surroundings; migrating birds know our planet well enough to wing annually across its whole face; dung beetles navigate by the light of the Milky Way. But only the human being lives inside a cosmos, and only recently.

By the end of the last Ice Age, humans had travelled to every continent on Earth except Antarctica. At some point during these prehistoric wanderings, we began to pay close attention to the celestial realm. There are hints of this in Paleolithic cave art, where we find the first etchings of the Moon and its phases, its cycling from silver sliver to illuminated whole, and back again. We see it in the stone pillars that humans hauled across landscapes, to form rings that tracked the Sun’s seasonal arcs through the sky.

But these clues are few and far between. They aren’t enough for us to sketch the wider cosmos that the prehistoric mind inhabited. The first cosmos we can confidently describe comes down to us from the Bronze Age, whose belief systems were caught and preserved, in a newly invented cultural amber called writing. Even these, we know only crudely. Just well enough to identify a few of their common elements.

The ancient cosmos was not a complex mathematical structure. It was a sensory world, stitched together from people’s everyday experiences, people who had never seen Earth’s curves from orbit, or the night sky as magnified by a telescope. The ancient cosmos had a beginning, a birth out of a formless state, usually an infinite liquid realm, or a chaotic void that would suddenly separate into opposites, like light and darkness, or fire and ice, or earth and sky. This separation concept is still with us today in scientific creation stories, which often invoke a primordial splitting of symmetries. But the ancient versions were much more vivid. In the sacred Sanskrit text the Rig Veda, the universe begins as a symmetrical orb of pure potential, an egg surrounded by an infinite amniotic sea, which splits into two bowls of earth and sky, with the yolk-like sun hovering somewhere in the middle.

The earth that emerged from this primordial separation was usually a flat, round disc, wrinkled by mountains, cut through with rivers, and surrounded by ocean on every side. Above this disc was the closed dome of the sky, and below it was an underground realm of equivalent size. Together, they formed a sphere. Every night, the sun would travel through the invisible underworld after teetering over the horizon’s edge. The ancients knew this because the sun reappeared at dawn on the earth’s opposite side.

The ancients imbued the cosmos with consciousness and agency. The sun became a person, and so did the ocean

Of course, the idea that there is a singular ‘ancient cosmos’ is a gross simplification. Not all were spheres. Each had its cultural quirks and oddities. But they did have one thing in common: they were all teeming with gods.

When ancient peoples sought to explain the more mystifying aspects of their environment, they projected their own nature onto the cosmos. They imbued the cosmos with consciousness and agency. The sun became a person, and so did the ocean. The ancient gods had human frailties, the sort you’d expect from a highly intelligent, highly social animal. They acted impulsively. They were jealous. Their mood swings determined events in the world. Earthquakes, droughts, storms, floods, and rainbows. Their conflicts would eventually bring about the world’s end, in a fiery battle between the twin human concepts of good and evil.

In some cases, a phoenix universe would arise from the ashes, taking its place in a series that had no beginning or end, an infinite multiverse in time. This cosmic cycle of destruction and rebirth would sometimes occur on human time scales, but in more imaginative traditions, it could span billions or even trillions of years. Among today’s physicists, there are some who still believe the cosmos cycles in and out of being in this way.

Last November, I spent a week with Paul Steinhardt, the director of the Princeton Center for Theoretical Sciences. Steinhardt was one of the first high-profile physicists to question BICEP2’s findings in public. In a column for Nature last June, he said that the team’s analysis was seriously flawed. Few cosmologists were surprised. Steinhardt is inflationary theory’s most vocal critic, and has been for years. But perhaps critic is the wrong word. Apostate might be better, for Steinhardt was present at inflation’s birth. You might even say he midwifed it.

Steinhardt learned the art of theory at Caltech, at the feet of Richard Feynman, the charismatic Nobel Prize-winner and Manhattan Project veteran. As an evangelist for science, Feynman was second only to Carl Sagan and, even then, it’s a matter of taste. Feynman took Steinhardt under his wing, serving as his student thesis advisor, and his personal mentor. The two of them created a course together, a weekly meeting called Physics X where students would propose a question or some unexplained phenomenon, and then watch as Feynman riffed, hopping back and forth between disciplines with ease. The meetings were held in an old lecture hall with creaky wood benches, which is now named for Feynman. Steinhardt makes it sound like a grove outside Athens.

Steinhardt didn’t give much thought to cosmology until his postdoc days at Harvard, when he attended a talk by Guth. This was back in 1982, when Guth was a postdoc at Stanford, a ‘struggling postdoc’, according to Steinhardt.

‘Guth gave the most wonderful talk,’ Steinhardt told me. ‘He detailed his new theory of inflation from the ground up, including the basics of Big Bang cosmology, which I had never been exposed to.’

Guth explained that there were problems with Big Bang cosmology. For one, the Universe is mysteriously uniform in all directions. If you position telescopes at the North and South poles, and point each of them at a dark patch of sky, you can catch light from opposite ends of the Universe. If you measure the temperature of light from these regions, all the way out to eight digits, you’ll see the same number. This is mysterious because the two regions are separated by more than 20 billion light years, too far to have ever interacted in a way that would lead to such extraordinary equilibrium. It’s possible to generate a uniform universe such as ours within the standard Big Bang framework, but you have to carefully calibrate its initial conditions. You have to ‘fine-tune’ it.

Guth said the Big Bang’s problems could be avoided if the early universe had expanded, exponentially, so that its structure stretched and smoothed. He also said particle physics provided a mechanism for this expansion. But there was a catch: Guth couldn’t figure out how the expansion would end.

Within a few years, ‘eternal inflation’ was ascendant. A decade later, it was the sexiest idea in cosmology. Today, it is virtually a paradigm

‘It was the most exciting and most depressing talk I had ever been to,’ Steinhardt told me. ‘I couldn’t believe that such a sweet idea would have such a sour ending.’

Steinhardt decided to take a few weeks off, to brush up on astrophysics and cosmology, and to see if he could come up with a workaround for Guth’s problem. Weeks turned into months. In the meantime, Steinhardt landed a professorship at the University of Pennsylvania, where he picked up a talented grad student named Andreas Albrecht. Together, the two men developed an inflationary model that allowed inflation to continue forever, generating an infinite, bubbly multiverse as it went along. Linde hit upon a similar idea a few months before, but he was in Moscow at the time. His work was still hidden behind the Iron Curtain. When Guth debuted the theory of inflation, it was widely seen as stillborn, but Linde, Albrecht and Steinhardt breathed new life into it. Within a few years, ‘eternal inflation’ was ascendant. A decade later, it was the sexiest idea in cosmology. Today, it is virtually a paradigm.

Steinhardt had everything to gain by continuing to champion inflation, but before long, the theory’s flaws began to nag at him. ‘I made the first eternal inflation model, but I left the problems for someone else to solve,’ he told me. When decades passed with no solutions, Steinhardt’s doubts grew. He began to wonder whether there might be another way to fix Big Bang cosmology. He wondered if there were previous Big Bangs before our Big Bang. He began working on a new theory. So far, the work has been lonely.

‘The last 30 years is a very unusual period in the history of fundamental physics and cosmology,’ Steinhardt told me. ‘There’s confusion, and maybe even a certain amount of fear. People are wedded to these ideas, because they grew up with them. Scientists don’t like to change ideas unless they’re forced to. They get involved with a theory. They get emotionally attached to it. When an idea is looking shaky, they go into defensive mode. If you’re working on something besides inflation, you find yourself outside the social network, and you don’t get many citations. Only a few brave souls are willing to risk that.’

I teased Steinhardt, pointing out that he hadn’t exactly been hauled before the Inquisition. Steinhardt is fully tenured, a lion of Princeton’s storied physics department. He walks the same leafy streets that Albert Einstein walked. Indeed, his official title at Princeton is Albert Einstein Professor in Science. Still, he feels overmatched. He told me he has asked for help from outside the field.

‘The outside community isn’t recognising the problem,’ he said. ‘This whole BICEP2 thing has made some people more aware of it. It’s been nice to have that aired out. But most people give us too much respect. They think we know what we’re doing. They take too seriously these voices that say inflation is established theory.’

I asked him who might help. What cavalry was he calling for?

‘I wish the philosophers would get involved,’ he said.

In his magisterial history Conceptions of Cosmos (2006), the Danish historian Helge Kragh anoints cosmology ‘the most philosophical of sciences’. To create a cosmos, a story that encompasses the origins and ultimate fate of all that is, you have to leave established science behind. You have to face down the cold void of the unknown. Philosophers are always in a dogfight to prove their utility to society, but this is something they do well.

It was the philosophers of ancient Greece who first began to drive gods out of the cosmos. By explaining phenomena with natural laws instead of human-like personalities, Greek philosophers kicked off a slow rolling process that would, over millennia, reduce the world’s gods from many to a few. A trinity, perhaps. Or a lone deity. Or something more impoverished still: a first cause sitting outside space and time, its nature forever unknowable. For atheists, even this chilly, alien god is too much.

Thales of Miletus is generally thought to be the first natural philosopher. According to Plato, Thales was so dazzled by the stars that he once fell into a well while walking. According to Herodotus, the Greek father of history, Thales predicted a solar eclipse two years in advance. No one knows if these legends about Thales are true, but we do know that his habits of mind inspired a group of Greek philosophers who would, over the course of several centuries, develop a number of radical ideas about the cosmos. These philosophers were the first upright primates to understand that they stood on the surface of a sphere. Some of them suggested it might be spinning. Some knew the Moon wasn’t luminous, but merely a mirror for sunlight. A few famously argued that all of these things – the Earth, Moon, Sun, stars, and every other material body in existence – were composed of atoms that were too small to see, all moving around in a void.

Unlike Thales and Aristotle, Plato had no affection for the stars. He regarded them as mere ephemera compared with his pristine realm of ideas

When it came time to craft his cosmos, Aristotle took on a number of these ideas, but not all. He preferred the five elements of earth, water, air, fire and ether to a void filled with atoms. Aristotle’s cosmos is best imagined as a series of concentric spheres. The Earth was fixed at the centre. Whirling around it were spheres containing the Moon, Sun and stars. Aristotle’s Earth was made of degraded, decaying materials, but these outer spheres belonged to a separate, exalted realm. The outermost sphere of stars was most perfect of all, because nothing lay beyond it, and according to Aristotle, ‘that which contains is greatest’. Sealed in by the eternal stars, Aristotle’s cosmos was singular and complete. It was the only thing that ever was, and the only thing that ever would be.

Unlike Thales and Aristotle, Plato had no affection for the stars. He regarded them as mere ephemera compared with his pristine realm of ideas. He was a theorist’s theorist. ‘We shall dispense with the starry heavens if we propose to obtain a real knowledge of astronomy,’ he wrote in The Republic. And yet, according to Simplicius, it was Plato who saw the anomaly at the heart of the Earth-centred cosmos.

To see what Plato saw, imagine you were standing on the surface of Aristotle’s fixed Earth, gazing through the clear, rotating spheres that contained the Moon and the Sun, all the way out to the final sphere of stars. In the dark skies of antiquity, those stars would have been uncountable, but some of them would have stood out from the others. The most conspicuous was Venus, the star that shines brightest against dusk’s orange stripe, or when held aloft by dawn’s rosy fingers. The Greeks referred to Venus as a wanderer, a πλανήτης (planitis). Unlike the rest of the stars, which moved around in perfect, orderly circles, Venus would sometimes zig-zag as it made its way through the sky.

According to Simplicius, Plato challenged the stargazing philosophers of antiquity to reduce the ugly wanderings of Venus and the other planites to circles. Aristotle and the natural philosophers who followed him solved Plato’s problem by fine-tuning their cosmos in various ways. But none of their solutions satisfied. For nearly 20 centuries, the wandering of the planites would gnaw on astronomer’s minds. It would hang like a loose thread from the clean logical stitching of Greek cosmology, until Copernicus gave it a tug, and set the whole thing unravelling.

Last October, John Kovac boarded an 18-hour flight from the United States to New Zealand, as he has every year since 1990. After touching down in Christchurch, he drove to a large warehouse, to be fitted with a fluffy red parka, military-grade snow boots, mittens, goggles and other extreme-cold weather gear. The next morning, he crossed the Southern Ocean in a military transport plane, before landing at McMurdo research station on the coast of Antarctica. The Southern Ocean is the violent, iceberg-strewn moat that encircles Antarctica. It is the only latitudinal band on Earth where there is no land to stop ocean winds from whipping furiously around the planet, stirring up storms as they go. These storms had waylaid Kovac at McMurdo before, but this year the weather was mild. He was cleared to leave the next morning, to complete the last leg of his trip.

No matter how jetlagged, Kovac always makes it a point to stay awake for this final flight. During the first two hours, the plane passes over hundreds of miles of blocky, blue-veined glaciers, before cresting over the Transantarctic Mountains, whose barren peaks once played host to thick forests and, during the early Jurassic, some of the first large dinosaurs.

The final hour is more monotonous, because Central Antarctica is an enormous plateau, blanketed by ice so thick, it conceals whole mountain ranges. Few microorganisms can survive there. Birds fly over only if they are blown off-course by a Southern Ocean storm. Human beings are the only resident land animals. Kovac told me the plateau’s featureless terrain makes him feel like he’s flying over a frozen white sea that seems to stretch forever. At the end of the flight, the plane dips and a long building on stilts appears in the porthole window, a building whose slate grey exterior and sleek Scandinavian angles make it look like a villain’s lair in a spy thriller. That’s when Kovac knows he’s reached South Pole Station.

The politics that govern land use on Antarctica are radical, relative to Earth’s other six continents. The first human visitors to Antarctica came to claim territory for crown or country but, in 1959, 12 nations signed a treaty that demilitarised Antarctica indefinitely, in order to preserve it for peaceful, scientific purposes. One of the finest achievements of Cold War diplomacy, the treaty transformed an object of imperial conquest into a commons for the collective human mind.

Kovac can still remember his first few trips to Antarctica, back in the early 1990s, when he slept in insulated tents left over from the Korean War. Now he bunks down at the new South Pole Station, a 65,000-square-foot facility, whose galley serves flank steaks, and vegetables grown locally in a hydroponic greenhouse on the station’s first floor, down the hall from the reading room and the sauna. Life at the South Pole still has its hardships. The scientists sleep on narrow beds in tiny, cubicle-like cabins, and are allowed only two short showers per week. Kovac told me his team puts in long hours, but he said he doesn’t mind the gruelling schedule, because his tasks are specific and discrete. He feels like he’s on a special mission.

If you’re at the bottom of the Earth, the view into the Universe doesn’t change much as the planet moves around in its orbit

One Friday last November, Kovac called me from South Pole Station, where it was 3am. I asked him how the night sky looked. I’d heard that the Milky Way was something to behold from Antarctica, especially when the auroras were dancing. Kovac gently reminded me that there is only one ‘night’ at the South Pole, and that it begins in February and ends in September. He stayed over at the Pole for one of these winter stretches, back in the 1990s, before he had kids. Winter temperatures on the plateau are too frigid for planes to fly, meaning there is no traffic in or out during the dark months. ‘It was like being on a submarine,’ Kovac told me. ‘Once was enough for me.’

Kovac did his winter stint back when he was a graduate student, when he was beginning to distinguish himself as a skilled designer of observatories. Instrument design is in Kovac’s blood. His late father Michael was dean of the engineering school at the University of Southern Florida. When John was 12, a teacher at his school gave him his first telescope. He set it up in his backyard, but quickly grew tired of using it. Gazing at celestial objects that others had already seen bored him, especially when he could see spectacular photos of those same objects in books. ‘But I was fascinated by the technology,’ he said. ‘I wanted to know everything about the construction of the telescope itself, and how it focused light.’

Kovac called me from a conference room that overlooks the main runway at South Pole Station. There wasn’t much action outside at that hour. He could see clear across the landing strip to the Dark Sector, a special zone where electromagnetic radiation is forbidden. Some of the world’s most sophisticated telescopes have been hauled down to the Dark Sector at great expense, because it’s one of the best places on this planet to do astronomy. If you’re at the bottom of the Earth, the view into the Universe doesn’t change much as the planet moves around in its orbit. You can train your telescope on the same celestial objects for months at a time, and you have a clear view, because Antarctica is a desert. Any mist that manages to levitate from its surface quickly freezes into ice crystals that free-fall back to Earth. If you think of our planet as an eyeball, the Antarctic plateau is its iris, and the Dark Sector its pupil. At its centre, Kovac had installed an exquisite telescope, and with it, he hoped to peer deeper into time than any human being in history.

The cosmos that Kovac was peering into has changed radically since Classical antiquity. Aristotle’s elements of earth, air, water, fire and ether are gone, replaced by complex chemistry. The still Earth now spins. The spheres are in ruins. But we needn’t weep for the Greeks. They had a good run. Greek cosmology waned in influence during Rome’s slow fall, and was nearly lost during the Early Middle Ages, but it remained unsurpassed in sophistication until well into the 15th century.

The Christian church that came to power during late antiquity scoffed at Greek learning, especially natural philosophy. ‘What has Athens to do with Jerusalem?’ asked Tertullian, one of the Early Church fathers, a group whose most learned members knew ‘pitifully little’ about astronomy, according to historian Helge Kragh. When it came to cosmic matters, the Bible became the final authority. A few even regressed to belief in a flat Earth.

Only in the 12th century did learned Christendom read Aristotle and Ptolemy, and only because Islamic scholars rescued their works from oblivion. (Many of the stars in our star maps have Arabic names.) A century later, Christian thinkers were calling Aristotle ‘The Philosopher’. Greek natural philosophy was like a smooth, black monolith that suddenly appeared in the midst of the Christian West, a gift from a futuristic people who happened to have lived 15 centuries prior.

Historians have fixed a permanent marriage between the name Copernicus and the word ‘revolution’, and rightly so, for Copernicus asked his readers to entertain radical notions. He knocked the Greek cosmos, saying it was needlessly complex, and therefore ‘insufficiently pleasing to the mind’. He also rebelled against an even more powerful and entrenched intellectual foe: human sense perception. Ever since the dawn of consciousness, humans had stood on firm ground, watching, as the Sun moved in its daily path from east to west, its arc across the sky rising in summer, and falling in winter. Copernicus said this was an illusion. It was the Earth that moved, at blazing speeds, around its own axis and around the Sun. These motions explained the zig-zagging of the planites simply, and besides, Copernicus argued: what better place for the Sun, ‘the lamp of this most beautiful world’, than at the centre, where it could illuminate the entire cosmos at once?

Copernicus was not the first to propose a heliocentric cosmos. The Greek astronomer Aristarchus devised a similar system in the 3rd century BC. But Copernicus had an advantage over Aristarchus. Copernicus had the luxury of living in the century that preceded Galileo’s first glance through the telescope. His writings persuaded Galileo, the first major figure of technological cosmology, to test his strange ideas.

Newton showed that Heaven and Earth belonged to the same jurisdiction: the cosmos was one

Copernicus didn’t dismantle the Greek cosmos entirely: he enlarged it, and dislocated the Earth from its centre, but he left the spheres intact. The real demolition work began when Galileo fixed his telescope on the ghostly stripe of light that rose into the sky above Florence on clear nights. The ancients compared this smear of light to spilled milk, but through his telescope Galileo saw that it was dense with stars that belonged to a larger structure, one that would come to be understood as an enormous disc, a galaxy in which the Sun enjoyed no privileged position.

Around this same time, Galileo’s friend Johannes Kepler began studying the planets, until he knew their movements well enough to establish his three laws of planetary motion. The third of these laws lodged itself in the extraordinary mind of Isaac Newton, and played muse to his universal law of gravitation, a theory whose philosophical import cannot be overstated. With a single, elegant equation, Newton explained the orbits of the planets and the falling of apples, and in doing so, dissolved Aristotle’s distinction between Earth and Heaven. Newton showed that both realms belonged to the same jurisdiction. The cosmos was one.

As telescopes swelled in size, from small tubes you could lift to your eye with one arm, to large wooden observatories that had to be navigated by ladder, they revealed new objects in the sky, including blurry blobs of light called ‘nebulae’. Many astronomers thought these mysterious nebulae were single stars surrounded by luminous fluid. They had forgotten Galileo’s lesson about milky white light: it often magnifies into stars.

In 1755, the German philosopher Immanuel Kant offered an alternative explanation for the nebulae, in an extraordinary new theory of the cosmos, which anticipated so much 20th-century science that it could rightly be called a vision. Kant’s cosmos began with a sea of particles, all at rest in an infinite void. As time passed, the denser particles in this sea attracted others, forming tiny clusters that scaled into orderly structures such as our planet, and its Sun, and the stars, and our galaxy – and the other galaxies, which were so distant that they appeared to us as faint and formless nebulae. Kant said that all these structures would decay over time, until the cosmos returned to its primordial state, before reconstituting itself again. He said the cosmos would do this an infinite number of times, with each cycle taking ‘whole myriads of millions of centuries’.

During the early decades of the 20th century, astronomers built enormous mountaintop observatories, to peer at nebulae through thin alpine air, to confirm that they were indeed galaxies, separated by voids so vast that photons take millions of years to cross them. Like the ancients, cosmologists wondered whether their cosmos had always existed, or if it evolved as Kant suggested. Aristotle thought the cosmos was eternal. The Stoics disagreed, arguing that erosion would have already planed down Earth’s mountains if the world had always existed in its present state. When Medieval Christians took up the Greek cosmos, they sided with the Stoics. They needed the world to have a starting point in time, to stay faithful to the Genesis creation myth. Perhaps, then, we should not be surprised that it was a priest, a Belgian named Georges Lemaître, who gave our cosmos its beginning, which eventually came to be called the Big Bang.

By the early 20th century, telescopes were seeing the deep sky in high definition. Galaxies were sharpening into gorgeous ellipses and spirals, like M51, whose whirlpool swirls inspired Van Gogh’s Starry Night (1889). Galaxies were also giving off a special kind of light, a downshifted hue that suggested they were speeding away from Earth, in the same way that a waning ambulance siren tells you it’s moving away. Using Einstein’s equations, Lemaître deduced the relationship between a galaxy’s speed of retreat and its distance. The farther they were, he discovered, the faster they were flying away, and that meant the Universe was expanding. Remarkably, Lemaître also claimed this expansion was accelerating, a prediction that would not be confirmed until the close of the 20th century, when astronomers lofted a telescope above the mountaintops, and into outer space, in order to clock the speed of receding galaxies, by catching light from their exploding stars.

Today’s cosmologists use computer simulations to fast-forward this accelerating expansion, to show that it will one day rip every galaxy from view, stranding us in the same lonely cosmos we lived in before the telescope was invented. Lemaître was more interested in what happened when you rewound the tape. If galaxies were flying apart, he reasoned, they must have once been closer together. Push further back in time and the Universe would have been hotter and denser still. Rewind all the way to the beginning, and every planet, star and galaxy would be compressed into a ‘cosmic egg’ that would detonate at the moment of creation, with a flash so bright that its light would fill all of space, where it would linger, cooling and thinning, before vanishing altogether, one trillion years from now.

When he is not at the South Pole, Kovac teaches a class at Harvard called Applied Astrophysics 191. Students in this class are required to build a radio antenna with basic, inexpensive materials, mostly stuff you can buy at RadioShack. For the penultimate class meeting, Kovac invites the emeritus physicist Robert Wilson to guest-lecture. The students listen carefully to Wilson. With their shoestring detectors, they are trying to replicate an experiment that Wilson and his colleague Arno Penzias conducted in 1964, when they set out to measure the invisible radio waves that stream from the Milky Way’s centre.

To make this measurement, Wilson and Penzias had to account for interference from other sources of radiation. There was one electromagnetic source they couldn’t identify, or get rid of. They tried pointing their antenna at different parts of the sky, but this radiation was stubborn. It showed up everywhere. It seemed to permeate the entire Universe. When they measured its temperature, Wilson and Penzias realised it matched a signal that theorists first predicted back in the mid-1940s. By sheer accident, they’d found the thinning, cooling flash from Lemaître’s Big Bang.

The Big Bang is a story about entropy. Rewind time, Lemaître said, and the cosmos got hotter, denser, more energetic. During its first few hundred thousand years, it was a frenzied plasma, a stew of elementary particles that was too hot and chaotic to consolidate into atoms. Only after the cosmos cooled and expanded could electrons join protons to form hydrogen atoms, creating space for light to roam free. The photons that poured into the void at that moment have been travelling across the Universe ever since. Catch them with a detector and you can see, with your own eyes, the afterglow of the Big Bang. The physicist Paul Davies once quipped that this light has taught humanity ‘more about the creation and organisation of the Universe than 1,000 years of religion and philosophy’.

Kovac made his name as a scientist by helping to make extremely precise measurements of this afterglow, amid the rigours of the South Pole environment. That’s why he was tapped to lead BICEP2, an experiment that would train a telescope on a patch of sky for three years, in order to collect an exposure, a slowly compiled sheet of photons from the Big Bang’s afterglow.

According to most inflationary models, those photons would have encountered a universe that was still reeling from inflation, when they beamed off the primordial plasma. The Universe would have been awash in gravity waves, distortions in space-time that would spin the photons into a polarisation pattern, whose swirls would show up on Kovac’s sheet of light. You’d have to squint to see them. The swirling effect would be visible only in one out of every 30 million photons. To detect something so faint and ancient in the sky was ludicrous, a wild-goose chase, according to the late Andrew Lange, the beloved Caltech cosmologist who helped conceive the experiment. But it was a goose worth chasing. To see swirls in the afterglow would be to see through it, back to the Big Bang itself. That’s why so many experiments are hunting for them. And that’s why the BICEP2 team worried about being scooped when swirling patterns started to show up in their exposure in 2010, almost immediately after the telescope saw first light.

Kovac thinks of himself as a cautious scientist. He told me he was initially skeptical of the swirling signal. But then a year went by, and another, with each trip to the Pole bringing better news. The signal was strengthening. In a series of meetings, the BICEP2 team brainstormed ways to test and retest the signal. They took turns playing devil’s advocate, trying to come up with alternative explanations for the swirls. In December 2013, Kovac convened a group call to discuss the possibility of publishing. He was at the South Pole at the time. ‘I wanted to make sure everything was done properly,’ he told me. ‘I even wrote up notes about what I wanted to say.’ By the end of the call, the team felt they’d exhausted every possible challenge to the signal. ‘We decided it was real and it was on the sky,’ Kovac said.

A few months later, Kovac showed the finished paper to Guth, and a few days after that, Harvard’s astrophysics department released a statement saying it would hold a press conference to ‘announce a major discovery’. The statement set off a frenzy of speculation on Twitter and Facebook, and in physics departments worldwide. When Kovac and his team announced, the media was waiting, and the following day, Kovac’s name was on front pages across the globe.

Kovac told me his inbox was useless for weeks afterward. Of course it was. BICEP2 was a feel-good story. In an age of accelerators and telescopes that cost billions, BICEP2 made one of the most precise measurements in the history of astrophysics, for less than $10 million. And they did it with a tiny team. It took hundreds of scientists to hunt down the Higgs Boson, but BICEP2 needed only a few dozen to spot swirls in the Big Bang’s afterglow, swirls that Linde had described as ‘a smoking gun for inflation’ in a Stanford press release. People were calling it the discovery of the century. Time magazine reshuffled its annual list of the world’s 100 most influential people to include Kovac. He attended the gala in black tie.

BICEP2’s signal could have been contaminated by the dust that hangs between stars

But that was in April 2014, during the honeymoon period. In May, a handful of prominent cosmologists began to question the BICEP2 team’s interpretation of its results, and their decision to go public before their paper was peer-reviewed. In June, Kovac was invited to the World Science Festival, to sit on a panel with Guth, Linde and everyone’s favourite inflation skeptic, Steinhardt. Most of the discussion was a victory lap for BICEP2 and inflation. But toward the end, Steinhardt and Kovac had a tense exchange, when Steinhardt asked Kovac if he was still confident that his signal was caused by gravity waves from inflation, ‘now that [he’d had] some feedback’.

The previous month, Steinhardt’s colleague David Spergel co-authored a paper suggesting that BICEP2’s signal could have been contaminated by the dust that hangs between stars in our galaxy. Spergel is one of cosmology’s alpha dogs. When he barks, people listen. His paper explained that the Milky Way’s magnetic fields spin galactic dust into swirling patterns. Starlight ricochets off this dust as a dim, swirling glow. Some of the photons from this glow might have slipped into BICEP2’s telescope undetected, and mixed with the Big Bang’s afterglow.

Most of this came as no surprise to Kovac. Like many cosmologists, he would love to put a telescope outside the Milky Way entirely. But we are stuck in our galaxy, so Kovac’s team picked a patch of sky that was thought to be relatively clear, and they spent months making models to account for what little dust was there. None of the models made a dent in the signal, but the models were speculative. They were based on dust estimates from past studies, whose uncertainties couldn’t be quantified. They were too weak to discount alternative interpretations of the data. Kovac was repeatedly asked about the dust in the weeks following the BICEP2 announcement. He always gave the same answer. He said it was disfavoured as an explanation ‘through multiple lines of reasoning’. He was wrong.

The BICEP2 team’s interpretation was already looking shaky in late June, when Steinhardt asked Kovac if his confidence had slipped. The month prior, a team from a competing experiment released new data on the dust. In September, they released another data set, which persuaded most cosmologists that BICEP2 could no longer distinguish its signal from the dust glow. This February, Kovac’s team published a paper admitting as much.

When Kovac went public last March, a few news reports noted that it was almost 50 years after Wilson and Penzias announced their discovery of the Big Bang’s afterglow. Now, a darker irony links the two announcements. Wilson and Penzias were trying to detect a signal from the centre of our galaxy, but they picked up a signal from the deep cosmos. The BICEP2 team made the opposite mistake.

Kovac took some heat in the media after his signal lost its revolutionary luster. But not all of it was deserved. You could blame the BICEP2 team for putting too much faith in their data analysis. You could blame them for throwing a bit of a party for themselves. But you had to admire the precision and economy of their measurement, and you couldn’t accuse them of scientific malpractice. These were first-class scientists, and they acted like it when the chips were down. They welcomed peer review. They revised their work quickly. The whole BICEP2 saga was, in many ways, a triumph for science. It showed precisely why science has become our supreme means of obtaining knowledge about the natural world.

When I visited Steinhardt in November, I expected him to come down hard on Kovac and his team. The original BICEP2 announcement was unwelcome news for inflation skeptics such as Steinhardt. He’d been suspicious from the outset. Even in June, when the narrative had barely begun to turn, he hadn’t hesitated to go after Kovac onstage. What would Steinhardt say now? Not much, it turned out: he went relatively easy on BICEP2. He saved his real ire for the theorists.

The BICEP2 signal caught inflationary theorists by surprise. That’s why Linde squawked ‘What?’ when Kuo came knocking on his door. That’s why he asked Kuo to repeat the data. Remember, there were several experiments looking for this result. So far, they hadn’t found much. In fact, they had established a low upper limit on swirls in the Big Bang’s afterglow – lower than predicted by most models of inflation, a theory that doesn’t make many testable predictions. When data from those previous experiments was released in 2013, Steinhardt pounced. He said it showed that inflation was in trouble. ‘They said I should stop being so negative,’ Steinhardt told me. ‘They said it was no problem. They said they could make models without gravitational waves.’

That’s why Steinhardt was surprised to see inflationary theorists clinking glasses when BICEP2 announced a high swirls figure. ‘They declared victory,’ he told me. ‘They said it was smoking-gun proof! Just what they expected!’

But then a few months passed and BICEP2’s interpretation started to look wobbly. In June, Linde told New Scientist that he didn’t like the way BICEP2’s swirls were being treated as a smoking gun for inflation. In July, Guth made similar statements to the Washington Post. Steinhardt was furious. He thought it was flip-flopping. He began to wonder if any data would disturb the serene certainty of inflationary theorists. ‘It was Andre Linde who used the “smoking gun” language in the first place,’ he told me. ‘Now he says it doesn’t make a difference what BICEP2 says. How can it be that not seeing gravitational waves is fine, and then seeing them is a smoking gun, and then not seeing them is fine again?’

Steinhardt told me that this flip-flopping on gravity waves is emblematic of inflation’s deeper flaws. Remember, inflation was originally designed to patch another theory’s fine-tuning issues. To produce the strange universe we see around us, you had to fine-tune the Big Bang. Inflation fixed that problem with a theoretical mechanism that briefly blew up the Universe like a balloon. But to produce a universe like ours, inflation’s initial conditions must also be precisely calibrated.

Eternal inflation is often invoked as a solution to inflation’s fine-tuning problem, because it spits out a multiverse, an infinite sea of cosmic regions, each with its own physical peculiarities. One with our peculiarities, our tuned initial conditions, is bound to show up somewhere. And even if such regions are rare, we are bound to inhabit one of them, for the simple reason that observers will only arise in regions with ‘Goldilocks’ conditions, just right to give rise to observers. In the lifeless regions, nature is not called ‘tuned’, or ‘designed’, or ‘beautiful’. She is not called Mother, because there is no one there to call her anything. But we observers should expect our region to look tuned. All observed regions of the cosmos look tuned.

The real trouble with the multiverse is that it can’t be tested, not yet. We can’t put a telescope in the regions outside ours

There are other theories of nature that treat fine-tuning as evidence in this way. Proponents of these theories will often trot out aspects of the natural world that seem too good to be true, and use them as evidence for an entity that can’t be sensed directly. Something as marvellous as the human eye could not have simply emerged from nature, they will say. It must have been crafted and honed by a mind like my own. Except it wasn’t. Eyes evolved, independently, on more than 40 branches of life’s tree. The eye looks designed to you because you do not understand the deeper properties of the world you inhabit. This is what usually happens to evidential fine-tuning. Science dissolves it into the clean, purring operations of nature’s fundamental laws. Fine-tuning usually signals weakness in a theory, not strength. When fine-tuning is used as evidence for a grand metaphysical apparatus capable of making anything and everything, it usually means that something has gone amiss.

There are other reasons one might be suspicious of the multiverse. This idea that the very existence of observers tells us something deep about the cosmos bears a disturbing resemblance to ancient anthropomorphic thinking. Once again, we find ourselves making grand, cosmic extrapolations from our own existence. Once again, the world is made in our image. The British philosopher Bertrand Russell had a great line on this sort of thinking: ‘All such philosophies spring from self-importance, and are best corrected by a little astronomy.’

But these are not knock-down arguments. They rely on innuendo and guilt by association. The real trouble with eternal inflation’s multiverse is that it can’t be tested, at least not yet. We can’t put a telescope in the regions outside ours. We have to look for evidence of the multiverse in our region. What should we look for? It’s hard to say, because the multiverse explores every combination of cosmic conditions an endless number of times. It’s not clear that any combination is likelier than any other. Theorists are trying to determine whether some conditions are more probable than others, but they haven’t succeeded yet, and there’s no guarantee they will. In the meantime, it’s hard to know whether inflation’s fine-tuning problems are genuine explanatory gaps that need exploring, or quirky outcomes of the quantum slot machine. The theory’s weaknesses can be explained away with the same glib shrug that accompanies the retort: ‘God just made it that way.’

A dominant, infinitely flexible multiverse theory could make it easy not to strain for the next leap forward. It could lead to a chilling effect on new ideas in cosmology, or worse, a creative crisis. Steinhardt thinks we’re already there. ‘Andre Linde has become associated with eternal inflation because he thinks the multiverse is a good idea,’ he told me. ‘But I invented it, too, and I think it’s a horrible idea. It’s an emperor’s new clothes story. Except in that story, it’s a child who points out that the Emperor has no clothes. In this case, it’s the tailors themselves telling us that the theory is not testable. It’s Guth and Linde.’

Steinhardt worries that science itself could be compromised. Science freed the imagination from cave shadows and shibboleths. Science let the mind run wild with radical ideas, ecstatic visions and new worlds, so long as those ideas explained what we actually see when we gaze out into nature. The Earth moves around the Sun: look how Venus wanders and you’ll see. The nebulae are distant galaxies, brimming with stars: magnify them and you’ll see. The Universe was once hot and dense, and has been expanding ever since: catch photons from its primordial flash and you’ll see. Science owes its epistemological gravitas to its stern insistence that every idea faces the firing squad of experiment. That is its philosophical backbone. That’s the methodology that gifted us the shimmering, intricate, expansive cosmos we live in today.

cosmologists should be searching for an alternative theory. They should not be waving away problems with the multiverse

That doesn’t mean that theorists should shackle their imaginations to the limits of today’s instruments. Atoms and black holes were both theoretical entities before they were observed. Reality is always grander than the world we can see. The Caltech cosmologist Sean Carroll has argued, persuasively, that we shouldn’t refuse to contemplate the existence of what we cannot sense directly ‘on the grounds of some a priori principle’ such as testability. Especially not in the theoretical realm, which is speculative by nature. But nor should we be blind to where a field’s leading theory is leading us.

Inflation could turn out to be right in the end. Some of its predictions have come true. At the moment, there is no alternative theory of the early Universe that explains more. But cosmologists should be searching for one. They should not be waving away inflation’s fine-tuning problems with the multiverse. Until eternal inflation is testable and tested, successfully, again and again, cosmologists should not allow it to monopolise the collective theoretical imagination. Inflation is a speculative theory, and it should be treated as such.

Steinhardt looks out on his field, and sees a generation of theorists tinkering with models, wasting whole careers fiddling at the edges of a 30-year-old idea. ‘I know why they’re doing it,’ he says. ‘It’s easy to do. You can make hundreds of these models, and you can tweak them so they fit the data. But usually, those fixes aren’t the answer. Usually, you have to do something new.’

Steinhardt is trying to do something new. He spends most of his research time working on an alternative cosmological theory. He thinks the Big Bang might have been a reaction to a contraction, a bounce, perhaps one in a sequence of bounces that extends deep into the past and maybe into eternity. He is trying to figure out whether a bounce could have yielded a smoothed, stretched, uniform cosmos such as ours. Sometimes he feels isolated, but he knows how to chip away. His sense of possibility powers him through. He told me he thinks we might be edging up to a transformative idea. Something that could rearrange reality as we know it. Something of Copernican magnitude.

‘We should be excited that inflation is in trouble,’ he said. ‘That usually means we’re on the brink of discovery. It means we’re missing some idea, a really important idea. Something that’s going to take over when it hits. Don’t people want to be there for that?’

As I walked out of Steinhardt’s office for the last time, it occurred to me that our cosmos is once again a sphere. Our Earth has been demoted in recent centuries. It no longer enjoys its former status as the still centre of all that is. But it does sit in the middle of our observable cosmos, the sphere of light that we can detect with our telescopes. Gaze into this sphere’s reaches from any point on Earth’s surface, and you can see light coming toward you in layers, from stars and the planets that circle them, from the billions of galaxies beyond, and the final layer of light, the afterglow of the Big Bang.

We might be trapped in this snow globe of photons forever. The expansion of the Universe is pulling light away from us at a furious pace. And even if it weren’t, not everything that exists can be observed. There are more things in Heaven and Earth than are dreamt of in our philosophies. There always will be. Science has limits. One day, we might feel ourselves pressing up against those limits, and at that point, it might be necessary to retreat into the realm of ideas. It might be necessary to ‘dispense with the starry heavens’, as Plato suggested. It might be necessary to settle for untestable theories. But not yet. Not when we have just begun to build telescopes. Not when we have just awakened into this cosmos, as from a dream.