Bangs big and small in cosmic origins debate

There's an idea in the area of physics known as quantum mechanics that suggests one can't measure a phenomenon without influencing the result.

It turns out that sometimes this same "observer effect" crops up in science journalism.

The story goes that, right now, there is a quiet debate happening that could have implications for how the Universe as we know it came to be - and what came before.

And the debate is being driven in part by the fact that news outlets including BBC News took a small peek into the machine of modern-day astrophysics.

It started in a fairly pedestrian way: I spotted a paper authored by someone with a familiar name, outlining analyses of what is known as the cosmic microwave background, or CMB.

Professor Sir Roger Penrose, along with his colleague Vahe Gurzadyan, had crunched through the publicly-available data on this ever-so-slightly jumbled glow of light that permeates the whole of the cosmos.

They found neat, circular rings of order in the CMB, a feature which would support a theory of Professor Penrose's: that the Big Bang is just the latest in an endless cycle, rather than a beginning per se.

The initial paper suggested these rings were echoes from before the Big Bang

As is common among cosmologists, the researchers published the idea on Arxiv.org, a repository for scientific papers before they go through the publishing process - where I found the manuscript and wrote a story on it, among other news outlets.

Three weeks later, two papers were posted to the Arxiv site refuting Professor Penrose's hypothesis: one by Hans Kristian Eriksen and Ingunn Wehus at the University of Oslo and another by Douglas Scott, Jim Zibin, and Adam Moss from the University of British Columbia.

Just two days later, a third refutation by Amir Hajian of the University of Toronto appeared.

It is rare that science works in this way; rarer still that it works at this speed. And there is some indication it's going this way in part because the BBC and others shed light on the initial paper.

In any case, the process is a remarkable, almost-but-not-quite public airing - kind of like a high-speed film of how scientific discourse can go on - with footnotes.

Anomalous

There are a number of factors driving this unusual debate. Perhaps the most obvious is the "celebrity science" factor.

"What's going on is that out of many outrageous claims made in cosmology each year, one was singled out recently for a huge amount of attention," Douglas Scott told BBC News.

"That would not have been the case had it come from someone without the reputation of Sir Roger."

It should also be said that Professor Penrose is championing a theory that, if correct, would completely upend the "inflationary cosmology" that is the current, widely-accepted best guess as to the Universe's origins.

Or, as Hans Eriksen put it: "The story contained a rather explosive mix of pre-Big Bang physics, black holes and Roger Penrose - any science journalist has got to love that."

And so we did. But Dr Eriksen hinted also at another factor surrounding the story - that the public nature of the data made the analysis easy to repeat - and refute.

That is, the current best-available data on the CMB are free to download; for now, that comes from the Wilkinson Microwave Anisotropy Probe, or Wmap.

Jo Dunkley, a University of Oxford astrophysicist and part of the Wmap team, explained that the hunt for "anomalies" like that posited by Professor Penrose is a common pursuit.

"That's a whole industry in our field, to look for unusual things in the maps," Dr Dunkley tells me.

"In the map of the sky, you can see an 's' and an 'h' if you look carefully. People have joked that it's Stephen Hawking's signature on the sky, but obviously we don't think that's significant.

"The nice thing about Penrose's theory is that it's a theory that you then go and test; that's exactly what we should be doing."

She's right. While all of this is going on, another paper is posted on the Arxiv that purports to see "bruises" of other universes in the CMB. That, too, should be tested for significance - perhaps before it gets media attention.

Error bars

The standard first test is to plug the new idea into simulations of the sky; you create a model on the basis of everything that is known and the result should replicate what observations bear out in the real CMB data.

It is here that the authors of the three rebuttal manuscripts suggest that Gurzadyan and Penrose have got it wrong. They say the original paper shows nothing more than what your common-or-garden inflationary cosmologist would have predicted on the basis of existing theory.

To prove the point, Douglas Scott and colleagues showed that a hunt for triangles in the CMB was statistically as successful as that for the circles that support Professor Penrose's theory.

After that, but before the third paper was published, Penrose and Gurzadyan had posted their riposte; it is clear that the subtleties of the argument are still being discussed, principally through the medium of the Arxiv site.

For some corners of science, this "preprint" culture is faster than the traditional refereed journal approach, in which results are sent to journals, who choose appropriate but anonymous experts to sign off the veracity or merit of the work.

The Planck telescope will soon be providing significantly more detailed data on the CMB

But preprint servers are to that process as YouTube is to Cannes: you can find more, faster, but you can't know much about the quality of what you'll see.

"The new technology raises new problems but also provides the means to resolve them quickly," Jim Zibin told BBC News. "In a sense... the old refereeing system is not fast enough to respond, and a new, informal form of refereeing appears to fill the void.

Preprints and these debates can be seen as a workshop, then - but Amir Hajian says it is important to have a home for the finished product.

"The debates finally end - that is the nature of science," he told BBC News.

"Unknowns will be knowns sooner or later; what is important is to have trusted refereed journals to publish the final polished version of each work."

In the interim, Dr Zibin hinted that media interest kicks off the "observer effect" in the delicate experimental setup.

"As far as the experts are concerned, incorrect work will either be corrected via Arxiv comments, the standard review process, or more slowly over many years and many papers, or will simply go unnoticed and forgotten.

"But as far as the public is concerned, in a situation like [this one] the publicity may happen too fast to get out the news that there were big errors with the work. Once the errors are found, the story has died down and the media have moved on."