However, Physics World hired Neil Turok, Science Friday interviewed Roger Penrose, and Nature now asked Paul Steinhardt to inform us about the status and the future of cosmology.
This is really amusing, shocking, or hysterical, depending on your temperament. It's like the following situation: It's April 1945. The Red Army arrives to Berlin and CNN, MSNBC, and the New York Times interview Adolf Hitler, Joseph Goebbels, and the Japanese emperor (the order isn't necessarily the same as the order of the three physicists at the top!) about their plans for the future of Europe, Asia, and the world. ;-)
This would be silly, wouldn't it? The real-world leaders of the Axis responded to the Soviet visit to Berlin more realistically than Steinhardt, of course. For example, Adolf Hitler committed suicide.
Readers could perhaps be interested in the opinions of the winners – and perhaps their arguments why the discovery of these waves could imply more far-reaching things such as the multiverse – but instead, we are told what the losers think.
At any rate, Paul Steinhardt's diatribe is written under the title
In the first part of the text, Steinhardt clearly takes it for granted that two speculative preprints criticizing the BICEP2's discovery must be right and nothing is left out of the BICEP2's discovery. As I discussed previously, much of the criticism is vague and lacks beef, and that's true for the two particular preprints. See Matt Strassler's characteristically "balanced" (alibistic) account with the links.
The things that BICEP2 is criticized for are things that every experiment like BICEP2 has to work with, anyway, and BICEP2 was well aware of all these things. Still, BICEP2 followed a strategy that maximally minimized the influence of things as the galactic dust in the Milky Way and they have obviously focused a lot of energy on efforts to decide whether their whole signal could be due to the dust. Their answer is "No" and they have numerous reasons to think so. The critics seem to ignore pretty much everything that BICEP2 has done to decide whether the dust is to be blamed for (nearly) the whole signal – with the answer that it almost certainly not – and they just return to the general talking points that the dust could be behind everything. I find the work by BICEP2 to be much more concrete, clever, and convincing. Of course that I am not quite 100.0000% sure that they are right but my belief is safely above 50%.
Steinhardt complains that the experimental work by BICEP2 has already ignited talks about the Nobel prize and grant decisions. Well, that's surely true and for a very good reason. If BICEP2 gets confirmed, and chances are very high that it will, it is surely a Nobel-prize-scale discovery – and I think it's fair to say that it would be not just an average one but one of the most spectacular Nobel prizes in recent decades. It's equally obvious that such a discovery – and even the significant chance that it is right, if I use a very modest language – is a sufficient reason for sponsors to adjust their decisions what should be funded. How it could not matter? Cosmology and physics are empirical sciences so of course that major experimental results such as this one do matter a great deal.
Also, Steinhardt mindlessly worships the peer review process and argues that no one should have been excited and no one should have celebrated the BICEP2 result before it was peer-reviewed and published in a prestigious journal. You know, I do think that in most disciplines, peer review significantly improves the average quality of the published results. But it's just a method to achieve this goal, not a holy ritual. Steinhardt's idea that a result like that is perfectly valid once a reviewer says "Yes" but the work is completely worthless before that is utterly silly. The BICEP2 people are arguably the world's super elite when it comes to the measurement of the B-modes and by their having chosen and performed the strategy so cleverly, they at least partially justify the claim that they don't have any true peers in this business – I mean only in these very special measurements.
But even if they had peers, a reviewer's "Yes" doesn't guarantee that a paper is right. Reviewers may share the misconceptions and errors with the authors and they often tend to be less careful, anyway (because they're less responsible for the paper than the authors – and they can benefit from the paper less than the authors) so they often do endorse a wrong paper. And on the contrary, good and important papers sometimes get rejected by referees who are misguided. Assuming that the referee's "Yes" is the most important argument that decides about the value and validity of some scientific work is largely a belief in religious rituals. It's the evidence, and not the partly random verdict of a randomly chosen person, that matters in science.
Now, other cosmologists like Seljak and Spergel are very good, and so on, but it's still true that they haven't made this experimental discovery – although they have made other discoveries (concerning the measurements of the cosmological parameters and, in Seljak's case, the theoretical calculations of B-modes) so it's fair to assume that a part of their criticism is due to jealousy. But one simply cannot eliminate a potentially very important experimental result as soon as he sees first two papers that are skeptical. Science doesn't work like that, sorry.
The BICEP2 folks have suggested that the peer review process is/was proceeding in a standard way. Many – sometimes sufficiently major – corrections have been forced upon the BICEP2 authors. But the claim about the discovery has survived. People like Steinhardt who seem to be "sure" that the discovery is wrong just because some critical papers have appeared are obviously prejudiced.
Falsifiability of inflation
But the second part of Steinhardt's rant is about the "falsifiability" of inflation. He has heard that the cosmic inflation would be fine even if the primordial gravitational waves would be measured to be undetectably weak. And he finds it stunning. He concludes that "the cosmic inflation is clearly unfalsifiable" because of that.
Needless to say, this conclusion doesn't follow from the facts at all. It has only followed in Steinhardt's sloppy brain because the wheels and gears inside the device are not doing what they should be doing, at least not correctly. What actually follows from the facts is that the cosmic inflation as a paradigm doesn't make unambiguous predictions about (at least) one physical quantity, namely the strength of the primordial gravitational waves.
However, the inflationary cosmology is making tons of other predictions! The whole remarkable measured "WMAP curve" which agrees with the theory so wonderfully does require inflation or "something closely related" to be true. The true alternatives predict entirely different curves – not too different from the curves on the picture above (although they are actually curves for different values of the cosmological constant and a different proton/neutron ratio).
Steinhardt, like many other amateur warriors against science, seems to be convinced that a theory must predict "absolutely everything" and "absolutely unambiguously" for it to have any scientific value. But that's not true at all. Theories generally don't predict everything. And only a theory of everything, string/M-theory, is in principle capable of achieving such a thing. In reality, because we don't understand the rules that govern the vacuum selection – and whether there are any rules at all – such a prediction of everything is only possible once some additional data about the compactification are inserted into the theory.
These individuals seem to be obsessed by the idea that "the more easily a theory gets falsified, the better theory it is". But that's completely false, too. It's great if a theory is highly predictive. But science is ultimately looking for the most correct theories, not the most predictive theories! The only real condition is that theories predict certain experimental results or their patterns. But if the predictions are wrong, the hypothesis is dead and comments that it was highly predictive about "almost everything" just can't help. Other, possibly less predictive theories supersede the falsified predictive competitor!
Steinhardt's and other men's idea about the "ideal theory" are analogous to his opinions about the best soldier. The best soldier, according to Steinhardt et al., is one who jumps in front of an enemy tank and screams:
All of you are motherf*ckers except for your leader who is a grandmotherf*cker!That's bold and predictive – and the response is largely predictable – but I think that the army may still prefer a different kind of soldiers. Wars and science just don't work and can't work in the way that Steinhardt and his soulmates are suggesting. The soldiers and theories that ultimately win are more subtle, more careful, and they're not risking their life in "every possible dangerous situation". Audacity just isn't enough. Competent armies fight battles they have a reasonable chance to win; and viable theories manage to focus on sharp claims about quantities and patterns that have a reasonable chance to be predicted and nailed down.
The knowledge in science is increasing gradually. It's just not true that "we must learn everything or nothing" once a new theory is proposed. If I describe this point from the opposite side, if a theory gives you some freedom (like the precise potential for the inflaton field[s]), it just doesn't mean that it gives you a complete freedom about everything and that it is a completely vacuous theory! In particular, cosmological theories in which scalar fields and their potential energy curves play a decisive role for the character of the later universe form a very special subclass of hypotheses about the universe, and even if they were not forming a very special subclass, Guth and Linde (and others) would have to be celebrated for their discovery of the importance and unusual, dominant abilities of the scalar fields!
With this being said, I must emphasize that the inflationary cosmology is remarkably predictive. Whole books may be written about this point.
This idiotic "falsifiability" talking point could be equally well used as a talking point against Darwin's theory of evolution (and probably many other, very important theories). The Bible says exactly what humans look like and how quickly they were created. Darwin's theory doesn't predict what the humans look like, how many legs they should have, and so on. It is completely ambiguous. It doesn't really predict anything, Steinhardt would say. The organisms may take pretty much any form. Evolution isn't science according to his philosophy.
However, the fact that the organisms may take pretty much "any" form that turns out to be viable – and the viability is redefined with the hindsight, anyway, so that it always looks circular to a certain extent – is really the main point of Darwin's evolution. It's why we celebrate it and it's ultimately a reason why it's so powerful. Evolution isn't a particular technical theory explaining one parameter such as the size of a bone in a particular animal. Evolution is a powerful framework, one that makes all such minor, technical explanations possible. Evolution is giving us tools to produce a huge, exponential diversity of life whose precise composition inevitably depends on historical accidents. In fact, because we have this evolutionary evidence that the size of the bone resulted from some historical accidents, we may self-confidently claim that all other theories calculating the size of the bones "more unambiguously" (and therefore more impressively from Steinhardt's viewpoint) are just wrong! The quantity cannot be calculated from scratch. Evolution tells us why.
The cosmic inflation is analogous. It is giving us a tool to produce an exponentially large, nearly flat, homogeneous universe – these are amazing, verified, universal predictions – and allows many expected things to exist in the cosmos that emerges from the inflationary expansion. That's what we need. This basic setup doesn't answer all questions, e.g. about the magnitude of gravitational and other waves, but it explains and answers tons of fundamental questions.
Because of the agreement between the inflationary predictions and observations when it comes to these general enough questions, we believe that the cosmic inflation is basically right and a more specific theory that also predicts the right magnitude of various waves etc. must be a special case or a minor modification of inflation. In other words, more specific theories are being looked for pretty much as subsets of theories that may be described as inflation. This gradual progress – increasingly more accurate focusing on the promising theories – is what science is all about. Steinhardt doesn't seem to understand it. Steinhardt doesn't seem to understand the scientific method.
I have said that because of the successes of inflation, more accurate theories are being looked for in the class of "special types of inflation". Guth and Linde (and others) have therefore started a huge industry, an extensive dynasty with many followers. They have seeded a large tree with many branches and subbranches. Steinhardt who has made early contributions to the cosmic inflation but who is not a genuine "forefather" apparently doesn't like the idea that folks like Guth and Linde are "forefathers", founders of a big industry, while he is not one. And perhaps for this simple psychological reason, he refuses to become "another follower".
The truth is that whenever he did something that made at least some sense, he has been one, anyway. Even his and Turok's ideas about ekpyrotic and cyclic universes are just small special cases of inflation. Even if the ekpyrotic and cyclic universes of this kind turned out to be right at the end – which can't be the case if BICEP2 is right – it would still be true that it's effectively a Guth-Linde idea with a small appendix that was added by Steinhardt and Turok that would be established. The distance between the ekpyrotic branes is pretty much the inflaton field. It's just being treated in a less field-theoretical way – the higher-dimensional dynamics of this field is assumed to be more important. But the reason why the ekpyrotic universe agrees with inflation concerning some correct predictions is that it is pretty much a special type of inflation, or at least a modest generalization of inflation. Attempts to distort the terminology can't hide this fact.