For as long as humans have been around, our innate sense of curiosity compels us to ask questions about the universe. Why are things the way they are? How did they get to be this way? Were the outcomes that we observe inevitable, or could things have turned out differently if we rewound the clock and began things all over again? From subatomic interactions to the grand scale of the entire cosmos, it’s only natural to wonder about all of this and more. For innumerable generations, these were questions that philosophers, theologians, and mythmakers attempted to answer. While their ideas may have been interesting, they were anything but definitive.
The modern alternative, however, provides us with a far more capable way of approaching these puzzles: through the process of science. That’s where this week’s inquiry, courtesy of Jerry Kauffman, comes from, asking:
“It’s always troubling for me to think of the Big Bang as having happened at a single point in [spacetime]… What existed before the Big Bang? And why did the Big Bang happen?”
When it comes to even the biggest questions of all, science provides us with the best answers we can muster, given what we know and what remains unknown, at any point in time. Here and now, these are the best robust conclusions we can reach.
When we look out at the galaxies in the universe today, we find that — on average — the farther away it is, the greater the amount its light is shifted towards redder, longer wavelengths. The longer light spends traveling through the universe before it reaches our eyes, the greater the amount that the expansion of the universe stretches its wavelength; in fact this was how we discovered the expanding universe. Stretched, longer-wavelength light is colder than short-wavelength light, and so, as the universe expands, it also cools. If we extrapolate backwards in time instead of forward, that implies the universe came to be as it is now from an earlier, hotter, denser, more uniform state.
Originally, we took the extrapolation as far back as we could imagine: to infinite temperatures and densities, and an infinitesimally small volume: a singularity. Evolving forward from that initial state, we successfully predicted, and later observed:
- the leftover radiation from the Big Bang, observable as the cosmic microwave background,
- the abundance of the light elements before any stars were formed,
- and the gravitational growth of large-scale structure in the universe.
However, there were also things we observed that we couldn’t explain if the universe began from a singular state, including why there were no leftover relics from the highest-energy epochs, why the universe had the same properties in opposite directions that could never have exchanged information with one another, and why there was absolutely no spatial curvature, leaving the universe indistinguishable from flat.
Whenever we have a scenario such as this — where we observe properties that our leading theories cannot explain or predict — we are left with two options, one of which is scientifically satisfying and one of which is tantamount to saying, “there is no explanation.”
- You can pawn the properties off as “initial conditions.” Why is the universe flat? It was born that way. Why is it the same temperature everywhere? Born that way. Why aren’t there high-energy relics? They must not exist. And so on. There’s no explanation for this; it’s simply the way things are.
- Or you can imagine some sort of dynamics: a mechanism that precedes the state we’ve observed and sets it up, so that it started off with the conditions necessary to create the properties we observe today.
Although it’s a bit controversial to say so, the first option is only acceptable when you are certain that the conditions you could have started off with are sufficiently random. Solar systems form from instabilities in protoplanetary disks around newly-forming stars; that’s random, and so there’s no explanation for why our solar system has the particular set of planets that it possesses. But for the entire universe, choosing that option is tantamount to giving up on dynamics, asserting that there’s no need to even search for a mechanism that could have preceded and set up the hot Big Bang.
Fortunately, however, not everyone fell into that solipsistic logical fallacy. If you want to go beyond your current understanding of how things work, and this is true across all scientific fields, all it takes is a new, superior idea. How do you know if your idea is good enough to make the grade, superseding our old theory and revolutionizing our view of the universe? Believe it or not, there are just three criteria you have to meet.
- It has to reproduce every success that the old theory achieved. Every single one, with no exception.
- It has to succeed where the old theory didn’t, by successfully explaining the phenomena the old theory couldn’t.
- And — this is the big one — it needs to make novel predictions, predictions that differ from the predictions of the old theory, that can then go out and be tested, with the decisive results left to determine the new idea’s failure or success.
That was precisely what, a little more than 40 years ago, the concept of cosmic inflation (sometimes known as cosmological inflation) set out to do. It hypothesized that before the universe was filled with matter and radiation, it was dominated by energy inherent to the fabric of space itself. That energy caused the universe to expand exponentially and relentlessly, which would stretch space so that it was indistinguishable from flat, which would cause all directions to have the same temperature because it was all causally connected in the past, and which would place an upper limit on the maximum temperature achieved in the early universe, preventing the formation of high-energy relics.
The initial model of cosmic inflation succeeded where the Big Bang without inflation failed, but it struggled to meet the first criterion, failing to produce a universe that had uniform properties in all directions. However, with the work of the community, classes models were swiftly discovered that reproduced the Big Bang’s successes, and that led to a rich era of theoretical exploration. We would model cosmic inflation as a field, and then the laws of physics would enable us to extract the properties imprinted on the universe from any particular model we chose. These details were worked out largely during the 1980s and the 1990s, and are found in a variety of textbooks in the field, including:
the last of which became the field-standard text on how cosmic inflation’s imprints are left on the universe, and in particular, in the cosmic microwave background. If you studied cosmology at the graduate level within the past 30 years, these were many of the seminal primary sources that taught you how to extract some key predictions from inflation that would differ from a universe where inflation did not occur.
In particular, there are six major predictions of cosmic inflation that were definitively extracted before they were ever put to the test. Inflation predicts:
- a spectrum of imperfections — density and temperature fluctuations — that are almost, but not perfectly, scale-invariant,
- a universe that’s coarsely indistinguishable from flat, but that has curvature to it at the ~0.001% level or so,
- density imperfections that are 100% adiabatic and 0% isocurvature in nature,
- that there will be fluctuations on super-horizon scales: scales larger than a signal moving at the speed of light in an expanding universe could create,
- there should be a finite maximum temperature to the universe during the hot Big Bang, and it should be significantly smaller than the Planck scale,
- and that a spectrum of gravitational wave fluctuations — tensor fluctuations — should be created as well, with a particular pattern to it.
All six of these predictions were in place long before the first data from the WMAP or Planck satellites came back, allowing us to test cosmic inflation versus a non-inflationary scenario. We’ve since observed strong evidence favoring cosmic inflation for points 1, 3, 4, and 5, and have yet to reach sensitivities that reveal a decisive signal for points 2 and 6. However, going 4-for-4 where we’ve been able to test it has been more than sufficient to validate inflation, rendering it the new consensus explanation for the origin of our universe. Inflation came before and set up the hot Big Bang, with extrapolation back to a singularity having now become an unfounded assumption.
A little deeper
However, as is almost always the case in science, learning something new about the universe only brings about additional questions to explore. What, precisely, is the nature of cosmic inflation? How long was its duration, and what caused the universe to inflate at all? If cosmic inflation is caused by a quantum field — a justifiable assumption to make — then what are the properties of that field? Just as before, if we want to answer these questions, we have to find ways of testing the nature of inflation, and then subject the universe to precisely those tests.
The way we explore this is by building inflationary models — leveraging effective field theories — and extracting the key predictions from various models of inflation. Generically, you have a potential, you get inflation when the ball is “high up on a hill” on the potential, and inflation ends when the ball rolls down from a high point into a “valley” of the potential: a minimum. By calculating various properties of cosmic inflation from these potentials, you can extract predictions for the signals you expect to exist in your universe.
Then, we can go out and measure the universe, such as by measuring some precise and intricate properties of the light that composes the cosmic microwave background, and compare them to the various models we’ve concocted. The ones that remain consistent with the data are still viable; the ones in conflict with the data can be ruled out. This interplay of theory and observation are how all astronomical sciences, including cosmology and the science of the early universe, advance.
In all inflationary models, it’s the final moments of cosmic inflation — the ones that occur just prior to the onset of the hot Big Bang — that leave their imprints on the universe. These final moments always produce two types of fluctuations:
- scalar fluctuations, which show up as density/temperature imperfections and lead to the large-scale structure of the universe,
- and tensor fluctuations, which show up as gravitational waves left over from inflation, and imprint themselves on the polarization of the light from the cosmic microwave background. Specifically, they show up as what we call B-modes: a special type of polarization that happens when light and gravitational waves interact.
How do we determine what the scalar fluctuations and the tensor fluctuations are? As detailed in the aforementioned texts, there are only a few aspects of the inflationary potential that matter. Inflation occurs when you’re high up on the “hill” of a potential; inflation ends when you roll into the “valley” below and stay there. The specific shape of the potential, including its first and second derivatives, determine the values of these fluctuations, while the height of the “high point” versus the “low point” of the potential determines what we call r: the ratios of tensor-to-scalar fluctuations. This measurable quantity, r, can be large, up to ~1, but it can also be very small, down to 10-20 or lower, without any difficulties, or, for that matter, anywhere in between.
It might seem, on the surface, that with such widely disparate possible predictions, that cosmic inflation doesn’t predict anything at all on this front. For the amplitude of the tensor-to-scalar ratio, r, that’s correct, although each model will have its own unique prediction for r. However, there is a very clean and universal prediction that we can extract: what the spectrum of gravitational wave (tensor) fluctuations should look like, and what their magnitude is on any scale we can examine. When we look at the signals that get imprinted on the cosmic microwave background, we can robustly predict what the relative size of these fluctuations are from small angular scales up to large ones. The only thing that’s unconstrained, except by observation, is the absolute “height” of the spectrum, and hence, the magnitude of r.
In the mid-2000s, there was an NASA/NSF/DOE interagency task force that set about planning a new generations of experiments to measure the polarization of the light from the cosmic microwave background on small angular scales, specifically designed to constrain r and either validate or rule out various models of inflation. Numerous observatories and experiments were designed and built to achieve that goal: BICEP, POLARBEAR, SPTpol, ACTPOL and others. The goal, by perhaps the 2030s or so, was to constrain r down to about ~0.001; if the gravitational waves from inflation made a large enough signal, we’d see them; if not, we’d place meaningful constraints and rule out whole classes of inflationary models. Theorists, with new observational data coming, set about making models with large r values: values that would fall in the testing area, and hence would be relevant for these experiments.
In many ways, the best data currently comes from the BICEP collaboration, currently on the third iteration of their experiment. There are only upper-limits on r, now constrainted to be no greater than about 0.03 or so. However, absence of evidence is not evidence of absence; the fact that we haven’t measured this signal doesn’t mean it isn’t there, but rather that if it is there, it’s below our current observational capabilities.
What not finding these tensor fluctuations (yet) definitely, definitely doesn’t mean is that cosmic inflation is wrong. Inflation is well-validated by numerous independent observational tests, and would only be falsified by the data if we did detect these tensor modes, and they didn’t follow the precise spectrum predicted by inflation.
And yet, you’d never know any of this by listening to the scientists associated with BICEP and the public-facing communication they’ve put out into the world. They continue to assert that:
- inflation remains in doubt,
- B-modes (indicating tensor fluctuations) are necessary to validate inflation,
- if there aren’t large magnitude ones, inflation is falsified,
- we are likely on the cusp of a paradigm shift,
- cyclic models are a viable competitor to inflation,
- and that inflation simply moved the “singular Big Bang” to before inflation, rather than immediately preceding the hot Big Bang.
All of these assertions, to be blunt, are both incorrect and irresponsible. Worst of all, every single one of the scientists I’ve spoken to whose made these claims know that they’re incorrect, but they are still advanced — including to the general public through popular treatments — by the very scientists who are running these experiments. There’s no kind way to couch it: if it isn’t self-deception, it’s utter intellectual dishonesty. In fact, when a scientist makes an overblown and premature claim that turns out, on closer inspection, to be completely wrong, some of us in the astronomical community call that, “a BICEP2,” named after the infamous false discovery they announced back in 2014.
Most of all, it’s a pity. These experiments that measure the properties of the cosmic microwave background to such extraordinary precisions are giving us the best information we’ve ever had about the nature of the universe, and of the inflationary epoch that preceded and set up — and caused — the hot Big Bang. Cosmic inflation is thoroughly well-validated as the origin of our universe, and has replaced the non-inflationary, singularity-containing Big Bang as our cosmological standard model for where we all came from. Although there are contrarian alternatives out there, none of them have ever succeeded where cosmic inflation does not, while they all fail to reproduce the full suite of inflation’s successes.
Scientists who value glory and attention over accuracy will no doubt continue to make baseless assertions undercutting what’s actually known about the universe, but you mustn’t allow yourself to be fooled by such claims. At the end of the day, we learn what exists in the universe simply by asking it questions about itself, and listening to what it tells us. As soon as we abandon that approach, we have to admit the uncomfortable truth: we simply aren’t doing science anymore.
Send in your Ask Ethan questions to startswithabang at gmail dot com!