As the ripples in space-time blow through dust…

The last time a big announcement in science was followed by an at least partly successful furor to invalidate it was when physicists at the Gran Sasso National Laboratory, Italy, claimed to have recorded a few neutrinos travelling at faster than the speed of light. In this case, most if not all scientists know something had to be wrong. That nothing can clock such speeds except electromagnetic radiation is set in stone for all practical purposes.


Although astronomers from Harvard University’s Center for Astrophysics (CfA) made a more plausible claim on March 17 on having found evidence of primordial gravitational waves, they do have something in common with the superluminal-neutrinos announcement: prematurity. Since the announcement, it has emerged that the CfA team didn’t account for some observations that would’ve seriously disputed their claims even though, presumably, they were aware that such observations existed. Something like willful negligence…

Imagine receiving a tight slap to the right side of face. If there was good enough contact, the slapper’s fingers should be visible for some time on your right cheek before fading away. Your left cheek should bear mostly no signs of you having just been slapped. The CfA astronomers were trying to look for a similar fingerprint in a sea of energy found throughout the universe. If they found the fingerprint, they’d know the energy was polarized, or ‘slapped’, by primordial gravitational waves more than 13 billion years ago. To be specific, the gravitational waves – which are ripples in space-time – would only have polarized one of two components the energy contains: the B-mode (‘right cheek’), the other being the E-mode (‘left cheek’).

The Dark Sector Lab (DSL), located 3/4 of a mile from the Geographic South Pole, houses the BICEP2 telescope (left) and the South Pole Telescope (right).
The Dark Sector Lab (DSL), located 3/4 of a mile from the Geographic South Pole, houses the BICEP2 telescope (left) and the South Pole Telescope (right). Image:

On March 17, CfA astronomers made the announcement that they’d found evidence of B-mode polarization using a telescope situated at the South Pole called BICEP2, hallelujah! Everyone was excited. Many journalists wrote articles without exercising sufficient caution, including me. Then, just the next day I found an astronomy blog that basically said, “Hold on right there…” The author’s contention was that CfA had looked only at certain parts of the sea of energy to come to their conclusion. The rest of the ‘cheek’ was still unexplored, and the blogger believed that if they checked out those areas, the fingerprints actually might not be there (for the life of me I can’t find the blog right now).

“Right from the time of BICEP2 announcement, some important lacunae have been nagging the serious-minded,” N.D. Hari Dass, an adjunct professor at Chennai Mathematical Institute told me. From the instrumental side, he said, there was the possibility of cross-talk between measurements of polarization and of temperature, and between measurements on the E-mode and on the B-mode. On the observational front, CfA simply hadn’t studied all parts of the sky – just one patch above the South Pole where B-mode polarization seemed significant. And they had studied that one patch by filtering for one specific temperature, not a range of temperatures.

“The effect should be frequency-independent if it were truly galactic,” Prof. Dass said.

The Milky Way galaxy’s magnetic fingerprint according to observations by the Planck space telescope. Image: ESA

But the biggest challenge came from quarters that questioned how CfA could confirm the ‘slappers’ were indeed primordial gravitational waves and not something else. Subir Sarkar, a physicist at Oxford University, and his colleagues were able to show that what BICEP2 saw to be B-mode polarization could actually have been from diffuse radio emissions from the Milky Way and magnetized dust. The pot was stirred further when the Planck space telescope team released a newly composed map of magnetic fields across the sky but blanked out the region where BICEP2 had made its observations.

There was reasonable, and it persists… More Planck data is expected by the end of the year and that might lay some contentions to rest.

On June 3, physicist Paul Steinhardt made a provocative claim in Nature: “The inflationary paradigm” – which accounts for B-mode polarization due to gravitational waves – “is fundamentally untestable, and hence scientifically meaningless”. Steinhardt was saying that the theory supposed to back the CfA quest was more like a tautology and that it would be true no matter the outcome. I asked Prof. Dass about this and he agreed.

A tautology at work.
A tautology at work.

“Inflation is a very big saga with various dimensions and consequences. One of Steinhardt’s
points is that the multiverse aspect” – which it allows for – “can never be falsified as every conceivable value predict will manifest,” he explained. “In other words, there are no predictions.” Turns out the Nature claim wasn’t provocative at all, implying CfA did not set itself well-defined goals to overcome these ‘axiomatic’ pitfalls or that it did but fell prey to sampling bias. At this point, Prof. Dass said, “current debates have reached some startling professional lows with each side blowing their own trumpets.

It wasn’t as if BICEP2 was the only telescope making these observations. Even in the week leading up to March 17, in fact, another South Pole telescope named Polarbear announced that it had found some evidence for B-mode polarization in the sky (see tweet below). The right thing to do now, then, would be to do what we’re starting to find very hard: be patient and be critical.

About Me

I’m a science editor and writer in India, interested in high-energy and condensed-matter physics, research misconduct, pseudoscience, science’s relationship with society, epic fantasy, open source/access/knowledge systems, H.R. Giger’s art, Goundamani’s comedy, Factorio, and most things that require a lot of time to get the hang of.