There has been much talk in scientific circles recently about a 2013 paper by Anna Ijjas, Paul J. Steinhardt and Abraham Loeb, titled, Inflationary paradigm in trouble after Planck2013. The authors of the paper question the cosmological theory of inflation, which postulates that the universe underwent a period of extremely rapid expansion shortly after the big bang, and that it has been expanding at a slower rate ever since. What I think their paper does instead is lend powerful support to the fine-tuning argument, which claims that the physical constants, initial conditions and laws of the universe were designed by God. This conclusion follows naturally if we assume that the Intelligent Designer of the cosmos wanted to not only make a universe that is hospitable to intelligent beings like ourselves, but also to send a clear signal of His existence to these intelligent beings.
What is the theory of inflation?
In this post, I’m going to quote some non-technical excerpts from the new paper by Ijjas, Steinhardt and Loeb, in order to convey the gist of the authors’ argument. But first of all, I’d like to say a little bit about the theory of inflation, which Wikipedia, summarizes as follows:
In physical cosmology, cosmic inflation, cosmological inflation, or just inflation is the extremely rapid exponential expansion of the early universe by a factor of at least 1078 in volume, driven by a negative-pressure vacuum energy density…
The inflationary hypothesis was originally proposed in 1980 by American physicist Alan Guth, who named it “inflation”. It was also proposed by Katsuhiko Sato in 1981…
Inflation answers the classic conundrum of the Big Bang cosmology: why does the universe appear flat, homogeneous, and isotropic in accordance with the cosmological principle when one would expect, on the basis of the physics of the Big Bang, a highly curved, heterogeneous universe? Inflation also explains the origin of the large-scale structure of the cosmos. Quantum fluctuations in the microscopic inflationary region, magnified to cosmic size, become the seeds for the growth of structure in the universe…
The basic process of inflation consists of three steps:
- Prior to the expansion period, the inflaton field was at a higher-energy state.
- Random quantum fluctuations triggered a phase transition whereby the inflaton field released its potential energy as matter and radiation as it settled to its lowest-energy state.
- This action generated a repulsive force that drove the portion of the universe that is observable to us today to expand from approximately 10−50 metres in radius at 10−35 seconds to almost 1 metre in radius at 10−34 seconds.
Following the inflationary period, the universe continued to expand, but at a slower rate…
The mission of the Planck satellite
Max Planck (1858-1947), the founder of quantum theory who won the Nobel Prize for physics in 1918. This picture was taken by A.B. Lagrelius & Westphal at the Nobel ceremony, and was published in Sweden in 1919. Image courtesy of Wikipedia.
The Planck satellite (pictured at top, courtesy of NASA and Wikipedia) is a space observatory operated by the European Space Agency (ESA). Named after the Nobel Prize-winning German physicist Max Planck (1858-1947) who founded quantum theory, the satellite was designed to observe and measure the tiny non-uniformities in the cosmic microwave background (CMB) at microwave and infra-red frequencies. The satellite was launched in May 2009, and after completing a successful survey, it commenced a second all-sky survey in February 2010. On 21 March 2013, the mission’s all-sky map of the cosmic microwave background was released. In the paper cited below, this new data is simply referred to as the Planck2013 data.
What did the Planck satellite find out?
And now, without further ado, let’s have a look at the new paper by Ijjas, Steinhardt and Loeb. The authors get to the nitty-gritty of the matter in their two opening paragraphs:
The Planck satellite data reported in 2013  shows with high precision that we live in a remarkably simple universe. The measured spatial curvature is small; the spectrum of fluctuations is nearly scale-invariant; there is a small spectral tilt, consistent with there having been a simple dynamical mechanism that caused the smoothing and flattening; and the fluctuations are nearly Gaussian, eliminating exotic and complicated dynamical possibilities, such as inflationary models with noncanonical kinetic energy and multiple fields. (In this Letter, we will not discuss the marginal deviations from isotropy on large scales reported by the Planck Collaboration .) The results not only impose tight quantitative constraints on all cosmological parameters , but, qualitatively, they call for a cosmological paradigm whose simplicity and parsimony matches the nature of the observed universe.
The Planck Collaboration attempted to make this point by describing the data as supporting the simplest inflationary models [4, 5, 6]. However, the models most favored by their data (combined with earlier results from WMAP, ACT, SPT and other observations ) are simple by only one criterion: an inflaton potential with a single scalar field suffices to fit the data. By several other important criteria described in this Letter, the favored models are anything but simple: Namely, they suffer from exacerbated forms of initial conditions and multiverse problems, and they create a new difficulty that we call the inflationary “unlikeliness problem.” That is, the favored inflaton potentials are exponentially unlikely according to the logic of the inflationary paradigm itself. The unlikeliness problem arises even if we assume ideal initial conditions for beginning inflation, ignore the lack of predictive power stemming from eternal inflation and the multiverse, and make no comparison with alternatives. Thus, the three problems are all independent, all emerge as a result of the data, and all point to the inflationary paradigm encountering troubles that it did not have before.
This sounds like very bad news for the theory of cosmological inflation. But as we’ll see, it’s nothing of the sort: it’s actually good news for the theory that the cosmos was fine-tuned for intelligent beings like ourselves.
The authors preface their discussion of the problems confronting inflation theory with a summary of the new Planck satellite data:
Planck2013 has added impressively to previous results in three ways. First, it has shown that the non-Gaussianity is small. This eliminates a wide spectrum of more complex inflationary models and favors models with a single scalar field.
…[A] second contribution of Planck2013  has been to independently confirm the results obtained previously by combining WMAP with other observations. The data disfavors by 1.5-sigma or more all the simplest inflation models: power-law potential and chaotic inflation , exponential potential and power-law inflation , inverse power-law potential [10, 11]. Third, the r-ns plot favors instead a special subclass of inflationary models with plateau-like inflaton potentials…
An obvious difference between plateau-like models like this and the simplest inflationary models, like V(φ)=λ.φ4, is that the simplest models require only one parameter and absolutely no tuning of parameters to obtain 60 or more e-folds [or a factor of about 1026 – VJT] of inflation while the plateau-like models require three or more parameters and must be fine-tuned to obtain even a minimal amount of inflation.
At this point, I imagine readers may be feeling puzzled. The new satellite data rule out the simplest models of inflation and instead support a model that requires a lot of tweaking to make it work. To many people, that sounds inelegant: surely, a Deity Who was a Master Mathematician wouldn’t make a cosmos like that. Or would He?
In their paper, Ijjas, Steinhardt and Loeb describe the three main problems that the new satellite data poses for the theory of inflation.
Problem #1: inflation can only smooth out the universe if it’s already very smooth to begin with
Back in the 1980s, the theory of inflation was supposed to be able to turn any old lumpy universe into a smooth one. That’s what the theory was designed for in the first place. Now, in the light of the new satellite data, it turns out that inflation can only smooth out the universe if it starts off very smooth in the first place! In other words, the initial conditions of the universe are not random, after all: they have to be tweaked for inflation to work.
As originally imagined, inflation was supposed to smooth and flatten the universe beginning from arbitrary initial conditions after the big bang . However, this view had to be abandoned as it was realized that large inflaton kinetic energy and gradients within a Hubble-sized patch prevent inflation from starting. While some used statistical mechanical reasoning to argue that the initial conditions required for inflation are exponentially rare [23, 24], the almost “universally accepted”  assumption for decades, originally due to Linde [8, 26, 27, 28, 29, 30, 31, 32, 33, 34], has been that the natural initial condition when the universe first emerged from the big bang and reached the Planck density is having all different energy forms of the same order.
After Planck2013, the very same argument used to defend inflation now becomes a strong argument against it. [B]eginning from roughly equal kinetic and gradient energy, gradients and inhomogeneities quickly dominate and the combination blocks inflation from occurring…
In sum, by favoring only plateau-like models, the Planck2013 data creates a serious new challenge for the inflationary paradigm: the universally accepted assumption about initial conditions no longer leads to inflation; instead, inflation can only begin to smooth the universe if the universe is unexpectedly smooth to begin with!
Problem #2: the inflationary models favored by the satellite data are models which are highly unlikely if inflation theory is true
The second oddity uncovered by the Planck satellite data is that the version of inflation they support is one that involves so-called plateau-like models, which themselves require a lot of fine-tuning in order to make them work. This is odd, because other things being equal, another version of inflation, called power-law inflation, is exponentially more likely then plateau-like inflation. What’s more, power-law inflation doesn’t require any fine-tuning to make it work, either. Ijjas, Steinhardt and Loeb find this all very perplexing:
All inflationary potentials are not created equal. The odd situation after Planck2013 is that inflation is only favored for a special class of models that is exponentially unlikely according to the inner logic of the inflationary paradigm itself. The situation is independent of the initial conditions problem described above; even assuming ideal conditions for initiating inflation, the fact that only plateau-like models are favored is paradoxical because inflation requires more tuning, occurs for a narrower range of parameters, and produces exponentially less plateaulike inflation than the now-disfavored models with power-law potentials…
…[G]iven the much larger field-range for φ, and larger amount of expansion, inflation from the power-law side is exponentially more likely according to the inflationary paradigm; yet Planck2013 forbids the power-law inflation and only allows the unlikely plateau-like inflation. This is what we call the inflationary unlikeliness problem.
The authors derive an important scientific lesson from the new satellite findings. It isn’t enough, they say, for a theory to agree with the observations predicted by a mathematical model of the theory. In addition, they say, that model itself has to be one which we’d expect to be true, on the basis of the scientific paradigm (or basic assumptions) contained within the theory itself:
Therefore, post-Planck2013 inflationary cosmology faces an odd dilemma. The usual test for a theory is whether experiment agrees with model predictions. Obviously, inflationary plateaulike models pass this test. However, this cannot be described as a success for the inflationary paradigm, since, according to inflationary reasoning, this particular class of models is highly unlikely to describe reality. The unlikeliness problem is an alarm warning us that a paradigm can fail even though observations favor a class of models if the paradigm predicts the class of models is unlikely.
I hope by now that some readers will have spotted the flaw in the authors’ scientific reasoning. What they are illicitly invoking here is the question-begging principle of methodological naturalism: the idea that we should do science as if the universe were a self-contained entity. According to this principle, scientific theories are meant to explain everything about our observations, without any need to invoke the supernatural. On this logic, any scientific theory which needs to be supplemented by extraneous ad hoc assumptions in order to make it work is a bad theory, which should be jettisoned. Ijjas, Steinhardt and Loeb are arguing that inflation is just such a theory.
Problem #3: inflation, once it starts, gives rise to a multiverse where anything can happen, but the new data agrees perfectly with the most naive predictions, which ignore the multiverse
Grace Kelly in High Society. Studio publicity still, 1956. Image courtesy of MGM and Wikipedia.
But there’s more. The whole logic of inflation theory is that it implies the existence of a multiverse, a vast (and perhaps infinite) ensemble of universes, including our own universe. But in a multiverse, where the parameters can vary in any possible way, you would never expect to find a universe whose parameters all had typical values.
In order to see why, let’s think about faces. Faces vary in many ways that we can measure – height, width, ratio of nose length to chin length, and so on – and each of these quantities has an average value. But hardly any face is perfectly average: nearly every face has some irregularity or oddity about it, which makes it different from the norm. Interestingly, faces that are perfectly average turn out to be surprisingly beautiful, like the face of Grace Kelly (pictured above).
The odd thing about the new satellite data is that all of the parameters measured for our universe agree with the values that one would naively expect them to have. None of them are odd, or unusual. And that in itself is an oddity. It’s not what one would expect, given a multiverse: instead, we’d expect a universe that was off-kilter in some way. Instead, it seems we live in a “Grace Kelly” universe:
A well-known property of almost all inflationary models is that, once inflation begins, it continues eternally producing a multiverse [36, 37] in which “anything that can happen will happen, and it will happen an infinite number of times” . A result is that all cosmological possibilities (flat or curved, scale-invariant or not, Gaussian or not, etc.) and any combination thereof are equally possible, potentially rendering inflationary theory totally unpredictive. Attempts to introduce a measure principle [39, 40, 41, 42, 43, 44] or anthropic principle [45, 46, 47] to restore predictive power have met with difficulty. For example, the most natural kind of measure, weighting by volume, does not predict our universe to be likely. Younger patches [48, 49] and Boltzmann brains/babies [50, 51] are exponentially favored.
Planck2013 results lead to a new twist on the multiverse problem that is independent of the initial conditions and unlikeliness problems described above. The plateau-like potentials selected by Planck2013 are in the class of eternally inflating models, so the multiverse and its effects on predictions must be considered. In a multiverse, each measured cosmological parameter represents an independent test of the multiverse in the sense one could expect large deviations from any one of the naive predictions. The more observables one tests, the greater the chance of many-sigma deviations from the naive predictions. Hence, it is surprising that the Planck2013 data agrees so precisely with the naive predictions derived by totally ignoring the multiverse and assuming purely uniform slow-roll down the potential.
But if our universe was intelligently designed, instead of being just one of countless universes in an infinite multiverse, this agreement with “naive predictions” is precisely what we might expect to find. What it tells us is that the universe was designed to be beautiful.
Is there any escape from these new problems?
The authors of the paper consider ways in which one might try to retain the theory of inflation, and at the same time avoid the three problems they have raised. One thing they are quite adamant about, however: invoking the anthropic principle won’t help. According to the anthropic principle, the universe has to have the properties it has, because if it didn’t, we wouldn’t be living it. On a very strong version of this principle, the universe was intentionally designed to be hospitable to life. Ijjas, Steinhardt and Loeb won’t have a bar of it. Even if that were so, they argue, we’d still expect the simplest (power-law) models of inflation to hold true instead of the plateau-like model, because power-law models are (as far as we know) perfectly compatible with the existence of intelligent life-forms, like ourselves:
In the previous sections we introduced three independent problems stemming from the Planck2013 observations: a new initial conditions problem, a worsening multiverse unpredictability problem, and a novel kind of discrepancy between data and paradigm that we termed the unlikeliness problem. It is reasonable to ask: is there any easy way to escape these problems?
One approach that cannot work is the anthropic principle since the new problems discussed in this Letter all derive from the fact that Planck2013 disfavors the simplest inflationary potentials while there is nothing anthropically disadvantageous about those models or their predictions.
The authors go on to dismiss other proposals for avoiding the problems generated by the new satellite data, and conclude that it is the theory of inflation itself which needs to be junked:
The multiverse-unpredictability problem has been known for three decades before Planck2013 and, thus far, lacks a solution. For example, weighting by volume and bubble counting, the most natural measures by the inner logic of the inflationary paradigm, fail.
By contrast, one might imagine the unlikeliness problem first brought on by Planck2013 could be evaded by a different choice of potential… [N]one of these … cases evades the unlikeliness problem. At the same time, it is clear that none does anything to evade the new initial conditions problem caused by Planck2013. In each case, the plateau-like inflation begins well after the big bang, enabling kinetic and gradient energy to dominate right after the big bang.
…If the only way the inflationary paradigm will work is by delicately designing all the test criteria and data into the potential, this is trouble for the paradigm.
Is the Designer trying to send us a message? If so, what kind?
Back in 1993, Walter Remine published a book called The Biotic Message. Evolution versus Message Theory (see here and here for reviews), in which he argued that life was designed in order to convey a simple message: that life was created by a single Designer.
What I’m suggesting here is that the universe was designed not only in order to be hospitable to intelligent life, but also to convey a message: that the Designer exists. And that’s it. Nothing more. It s reasonable to suppose that if an Intelligent Designer created a universe fit for intelligent life-forms like us, that Designer would want these life-forms to be aware of the Designer’s existence. Notice that I am not assuming here that the Designer wants our thanks, prayers, love or adoration. All I am assuming is that the Designer wants to be known.
Assuming that, what would be an ideal way for the Designer to make himself known? A very good way to do that would be to design a fine-tuned universe, which was balanced on a knife-edge, and whose physical parameters and initial conditions needed to be very carefully tweaked, in order for it to have developed the way it did, and in order for it to support life.
In my post, Night Vision: A new version of the fine-tuning argument (February 22, 2013), I discussed a new version of the fine-tuning argument, proposed by philosophy professor John T. Roberts, of the University of North Carolina, Chapel Hill. In jargon-free language, the argument goes like this:
Premise 1+: we already know that life exists in our universe.
Premise 2+: if we already know that life exists, then the probability that life would require fine-tuning is higher if there’s a Designer than it would be if everything is ultimately the product of blind chance.
Step 3: since we know that life exists in our universe, it follows that the probability that life would require fine-tuning is higher if there’s a Designer than it would be if everything is ultimately the product of blind chance.
Conclusion: therefore the discovery of fine-tuning favors the hypothesis that there’s a Designer over the hypothesis that everything is ultimately the product of blind chance.
The crucial step in this argument was premise 2+, which said that the probability that life would require fine-tuning is higher if there’s a Designer than it would be if everything is ultimately the product of blind chance. Roberts’ justification for this premise was that even if it were unlikely that a Designer would make a finely-tuned cosmos that was balanced on a knife-edge, rather than another kind of cosmos, the occurrence of fine-tuning is still more likely if the parameters describing our universe are set by a Designer than if the parameters are set by chance. Which is fair enough, so far as it goes, but it still leaves us with a nagging question: why would a Designer make a fine-tuned universe, in the first place?
If the proposal I am making is correct, then the answer is very simple: the Designer wants His existence to be known, and to that end, He has created a universe whose fundamental parameters are set up in such a way as to point to the existence of a Designer.
(Some atheists might want to ask at this point: “If that were the case, wouldn’t there be an even better way for the Designer to make His existence known – such as writing messages in the sky?” Perhaps. But someone might wonder if the messages were being written by mischievous aliens. If, however, the very warp and woof of the universe itself appears to be fine-tuned, then we have to posit a more Fundamental Cause: the Cause of the universe itself. In that respect, fine-tuning is a better signal. And as I have pointed out previously, even if our universe turns out to be one of many inside a larger multiverse, that multiverse would itself need to be fine-tuned.)
If my proposal is correct, then we shouldn’t expect the Designer to make the simplest possible universe. Instead, we should expect him to make the simplest universe which at the same time is capable of clearly conveying the message that it was intelligently designed – i.e. fine-tuned.
More trouble for inflation from the Large Hadron Collider?
Compact Muon Solenoid (CMS) detector for the Large Hadron Collider. Image courtesy of Wikipedia.
In their paper, Ijjas, Steinhardt and Loeb hint that experimental results from the Large Hadron Collider might spell more trouble for the theory of cosmological inflation:
Thus far, we have only focused on recent results from Planck2013, but recent measurements of the top quark and Higgs mass at the LHC and the absence of evidence for physics beyond the standard model could be a new source of trouble for the inflationary paradigm and big bang cosmology generally [57, 58]. Namely, the current data suggests that the current symmetry-breaking vacuum is metastable with a modest-sized energy barrier ((10^12 GeV)4) protecting us from decay to a true vacuum with large negative vacuum density …
The predicted lifetime of the metastable vacuum is large compared to the time since the big bang, so there is no sharp conflict with observations. The new problem is explaining how the universe managed to become trapped in this false vacuum whose barriers are tiny (by a factor of 1028!) compared to the Planck density when it is obviously much more probable for the field to lie outside the barriers than within them… Even in the unlikely case that the Higgs started off trapped in its false vacuum and inflation began, the inflaton would induce de Sitter-like fluctuations in all degrees of freedom that are light compared to the Hubble scale during inflation. These tend to kick the Higgs field out of the false vacuum, unless the Hubble constant during inflation is smaller than the barrier height . Curiously, a way to evade the kick-out is if all inflation (not just the last 60 e-folds) occurs at low energies where the de Sitter fluctuations are smaller than the barrier height. This would be possible if the only possible inflaton potentials are plateau-like with sufficiently low plateaus: the very same potentials that have the initial conditions and multiverse problems.
In the light of the suggestion I made above, we can now see that these discoveries, far from disconfirming inflation, provide support for inflation, set up by a fine-tuning cosmic Designer.
Alternatives to inflation?
According to a recent article in New Scientist magazine, titled, No need for inflation if cosmos was a bouncing baby by Lisa Grossman (New Scientist, 23 October 2013), Paul Steinhardt of Princeton University and his colleagues have developed a new alternative to inflation: the cyclic universe:
In their theory, a previous universe went through a phase of slow
contraction, crunching space-time. Then something reversed the process and it expanded again to make a new universe.
Compression explains cosmic smoothness without the need for inflation, because high pressures during the crunch would iron out most wrinkles. However, tiny quantum fluctuations could be carried over from the previous universe to provide the seeds of large-scale structure.
Previous studies had shown that these “before” and “after” pictures work mathematically. But no one was sure what was happening during the bounce. Some models require the universe to shrink to a singularity before rebounding, and we would need something we don’t have – a complete theory of quantum gravity – to describe it.
Now Steinhardt and his colleagues have built a model where, before the universe can collapse to a point, an unknown field with negative pressure shoves everything back outwards. The team used a “ghost” field, a physically unrealistic but mathematically simple field with negative kinetic energy.
“Programming a more realistic energy source would not change the outcome, but would require more complicated equations that would slow down the simulation,” says Steinhardt.
The ghost field is weak enough that it can be ignored, except during the bounce, when the universe is very small and dense. Using the ghostly equations, computational tools that describe space-time under general relativity show that tiny fluctuations from the dying universe can indeed be carried over into the reborn cosmos (Physical Review D, doi.org/pcd).
“Paul’s paper shows that things go through beautifully,” says Burt Ovrut at the University of Pennsylvania in Philadelphia. “It means that these alternatives to inflation are alive and well.”
Steinhardt admits that this bounce model has its demons. For one, the ghost field is just a placeholder, and the true nature of the one that gave us a push is as mysterious as whatever would have driven inflation.
Well, it seems to me that Steinhardt and his colleagues have given up God for … a ghost. They’ve produced some very nice mathematics, but it would be folly to call it science.
What do readers think?