Arrrgh Rumors

Today’s hot issue in my favorite corners of the internet (at least, besides “What’s up with Solange?”) is the possibility that the BICEP2 discovery of the signature of gravitational waves in the CMB might not be right after all. At least, that’s the rumor, spread in this case by Adam Falkowski at Résonaances. The claim is that one of the methods used by the BICEP2 team to estimate its foregrounds (polarization induced by the galaxy and other annoying astrophysical stuff, rather than imprinted on the microwave background from early times) relied on a pdf image of data from the Planck satellite, and that image was misinterpreted.


Is it true? I have no idea. It could be. Or it could be completely inconsequential. (For a very skeptical take, see Sesh Nadathur.) It seems that this was indeed one of the methods used by BICEP2 to estimate foregrounds, but it wasn’t the only one. A big challenge for the collaboration is that BICEP2 only observes in one frequency of microwaves, which makes it very hard to distinguish signals from foregrounds. (Often you can take advantage of the fact that we know the frequency dependence of the CMB, and it’s different from that of the foregrounds — but not if you only measure one frequency.) As excited as we’ve all been about the discovery, it’s important to be cautious, especially when something dramatic has only been found by a single experiment. That’s why most of us have tried hard to include caveats like “if it holds up” every time we wax enthusiastic about what it all means.

However. I have no problem with the blog rumors — it’s great that social media enable scientists to examine and challenge results out in the open, rather than relying on being part of some in-crowd. The problem is when this perfectly normal chit-chat gets elevated to some kind of big news story. To unfairly single someone out, here’s Science NOW, with a headline “Blockbuster Big Bang Result May Fizzle, Rumor Suggests.” The evidence put forward for that fizzling is nothing but the Résonaances blog post, which consists in turn of some anonymous whispers. (Including the idea that “the BICEP team has now admitted to the mistake,” which the team has subsequently strongly denied.)

I would claim that is some bad journalism right there. (Somewhat more nuanced stories appeared at New Scientist and National Geographic.) If a reporter could talk to an actual CMB scientist, who would offer an informed opinion on the record that BICEP2 had made a mistake, that would be well worth reporting (along with the appropriate responses from the BICEP2 team itself). But an unsourced rumor on a blog isn’t news (not even from this blog!). As Peter Coles says, “Rational scepticism is a very good thing. It’s one of the things that makes science what it is. But it all too easily turns into mudslinging.”

We’re having a workshop on the CMB and inflation here at Caltech this weekend, featuring talks from representatives of both BICEP2 and Planck. I was going to wait to talk about this until I actually had some idea of what was going on, which hopefully that workshop will provide. Right now I have no idea what the answer is — I suspect the BICEP2 result is fine, as they did things other than just look at that one pdf file, but I don’t pretend to be an expert, and I’ll quickly change my mind if that’s what the evidence indicates. But other non-experts rely on the media to distinguish between what’s true and what’s merely being gossiped about, and this is an example where they could do a better job.

  1. Sean,

    Assuming this holds up — and, I agree: it needs to be verified, but some very good and reputable scientists have worked very hard and put their reputations on the line, so there’s good reason for cautious optimism — then it would seem to me that this, plus the Higgs and recent observations narrowing down the mass of dark matter WIMPs, would represent the last major missing pieces of hard empirical evidence that theoreticians would need for definitive theories of cosmogenesis and quantum gravity and unification.

    Am I right? Are there more holes you need plugged?

    Or, to put it in a more juvenile manner: Are we there yet? Are we there yet? Are we there yet? Are we there yet?



    Thumb up 7 Thumb down 6

  2. Ya gotta love the thought of basing leading edge Big Bang research on a .pdf you undoubtedly picked up on the internet.
    Very Douglas Adams

    Well-loved. Like or Dislike: Thumb up 9 Thumb down 4

  3. Ya gotta love the thought of basing leading edge Big Bang research on a .pdf you undoubtedly picked up on the internet.

    If I try hard enough, I might be able to construct a not-impossible series of events whereby something like that could have happened. I just can’t bring myself to think that a team like the BICEP2 folks could have made as many overlapping and interlocking mistrakes as would be necessary for that type of scenario.

    I’m almost left wondering if this might not be an instinctive knee-jerk anti-science pro-religion reaction on the part of those spreading the story to further the “narrative” of how scientists never get anything right. Not very charitable of me, I know…but if the foo shits….


    Thumb up 4 Thumb down 6

  4. How very difficult it is to prove anything, science wise or any otherwise. I’ll be happy if Bicep2 is proven correct. Any advancement in science is to be celebrated.

    Like or Dislike: Thumb up 3 Thumb down 2

  5. Why don’t the skeptics do the obvious – ASSUME Planck has discovered large dust polarization in the BICEP2 region and THEN EXPLAIN where the BICEP2 paper went wrong in its analysis? (Since this explanation may be needed in October, can somebody not just do it now?)

    It seems like that is a very non-trivial exercise to me – BICEP2 have presented a rock-solid argument even with the lack of an accurate polarisation dust map – ie THEY ACCOUNTED FOR ALL THE POSSIBILITIES – unless someone can point out a real vulnerability in their analysis (rather than this non-vulnerability of a planck map misinterpretation) it would be best to keep quiet, methinks. (And you can’t just say they only did one frequency blah bla blah – you need to explain why even this one frequency gave such a significant set of measurements – was it just random luck?)

    I have a bad feeling that Planck ain’t gonna be able to convincingly decide this for us – they just never planned for an accurate polarisation map at the regions scanned by the south pole experiments. Could be embarrassing come October.

    Like or Dislike: Thumb up 4 Thumb down 4

  6. The interesting thing about BICEP2 and the blogosphere is that we are “doing journalism” rather than explaining and interpreting, and I suspect a lot of science bloggers have not thought deeply about the threshold for publishing a rumor. In the case of the “admitted mistake” it appears that BICEP2 has admitted no such thing, and I doubt a journalist would have gone to press on this (or been allowed to by their editor) without first soliciting a response from the BICEP people.

    Like or Dislike: Thumb up 1 Thumb down 1

  7. I was also disappointed when the Penrose rings (about 50 of them) turned out to be some sort of obscure instrumentation artifact. That actually happens quite a bit in astronomy.

    Oh, well. As Jack Horkheimer is fond of saying, “Keep looking up!”

    Like or Dislike: Thumb up 4 Thumb down 3

  8. Hidden due to low comment rating. Click here to see.

    Poorly-rated. Like or Dislike: Thumb up 5 Thumb down 13

  9. You reap what you sow. BICEP2 are being treated badly now because of the way they presented their work [complete with that corny youtube video of Linde being given the good news].

    Thumb up 15 Thumb down 14

  10. Well said Elbi: You reap what you sow. Direct-to-media hype and Nobel rumour-mongering will come back and bite you. Especially when it involves “proof of the multiverse”. I think Unzicker hit it on the head with this:

    “This is already a huge problem for the tiny polarization signal seen at the time of the formation of the cosmic microwave background (CMB), assumed to be 380,000 years after the Big Bang, but it renders ridiculous the claims about the first seconds of the universe. How could such subtle information survive in a sizzling hot soup for almost half a million years?”

    Ben: No son, we ain’t there yet. Especially since IMHO there’s going to be some collateral damage coming out of all this. To the thing that is said to explain the uniformity of the CMB, and the non-uniformity. Like: inflation is a dead man walking.

    Thumb up 7 Thumb down 11

  11. I refer you to the latest post on the Resonaances blog, which very well substantiates the earlier post.

    The presentation at Princeton today by Flauger seems like quite a bullet in the chest of the BICEP2 study (see page 39 of the PDF):

    Flauger finds that basically all of the observations of BICEP2 can be explained by ordinary dust and cosmological lensing.

    Thumb up 6 Thumb down 5

  12. James Gallagher – All this came up a few days ago at Peter Woit’s Not Even Wrong blog, and Woit, like Carroll here, pointed to Sesh Nadathur (who also added a post at NEW). One comment pointed out that the two experiments, BICEP2 and Planck, have several researchers in common – which fact would seem to dramatically reduce the possibilities of BICEP2 having a) misinterpreted Planck data that’s out there already, and b) failed to account for other data Planck might gather. On top of that is the essence of Nadathur’s analysis – which is that the effects of misreading or underestimating the effects of foreground dust COMPLETELY still wouldn’t explain BICEP2’s findings (If I’m reading it right, the effect would be to reduce the finding’s reliability by maybe 10% – at most.).

    The radio loops thing has the feel of grasping; it’s not as if we KNOW that radio loops have the potential to skew the BICEP2 findings ‘just so’, or to any material extent; and the odd against the former would seem, well, astronomical.

    There’s also a paper out from Dent & Krauss that’s more in the nature of CAUTIONING we need data before ruling out self-ordering – which fits better the view you’ve expressed here about how others OUGHT to approach the BICEP2 findings.

    I hear you the possibility it may be that expecting confirmation from Planck – which is looking at huge swatches of the sky & not really aiming to address the same issues as BICEP2, except incidentally – may have carry some hazards. But it seems to me any such concerns would have a fairly short romp, given the BICEP2 experiment can be replicated by pointing to some other relatively promising patch of sky, and that could do the trick in a matter of a few years (trivial, really – except in the Internet Age).

    It also should not be forgotten that there are at least two other land-based experiments that have yet to chime in on this, and both would seem promising on their own terms.

    Like or Dislike: Thumb up 3 Thumb down 2

  13. Hidden due to low comment rating. Click here to see.

    Poorly-rated. Like or Dislike: Thumb up 1 Thumb down 6

  14. Avattoir

    my (possibly naive) understanding was that BICEP weren’t hugely concerned about waiting for the Planck dust map since they were convinced their signal couldn’t be due to dust polarisation in any case.

    IF it turns out that the signal is entirely explained by dust polarisation then it’s a pretty sad shocking episode in science – and potentially damaging that it has been played out in public.

    The BICEP team just don’t come across as so cavalier to me, surely they would have waited for the detailed Planck maps if these detailed maps are so crucial to supporting their discovery.

    arrrgh, evidence.

    Like or Dislike: Thumb up 4 Thumb down 4

  15. Pingback: Arrrgh Rumors - The Simple Physicist

  16. Pingback: Allgemeines Live-Blog ab dem 18. Mai 2014 | Skyweek Zwei Punkt Null

  17. Hidden due to low comment rating. Click here to see.

    Poorly-rated. Like or Dislike: Thumb up 0 Thumb down 5

  18. It is amazing how people declare (or insist on) certitude on a result before confirmation or refutation. Whether or not any doubts that have been expressed have any merit, much of the driving force is expressed in the incendiary comments in some of the links supplied by Sean: there appears an ugly pre-existing animosity that has been enraged by the YouTube video of Linde. What’s going on with that? Comments, I strongly suspect, by a particularly persnickety strain of physics students who haven’t yet learned how to maintain themselves without diapers, behaving like brats who think matters of science can be decided by how loudly they shout. Is it at all feasible in this disgusting display to remind people that it is perfectly ok to publish a result that may be mistaken, and that patience is warranted toward confirmation or refutation BEFORE anybody opens their yap declaring foul with the despicable certitude of a god?

    Thumb up 5 Thumb down 5

  19. May 23 Sci-News: “THE UNIVERSE ISN’T EXPANDING AFTER ALL, SAY SCIENTISTS [Eric Lerner & Team at Lawrenceville Plasma Physics]”! …Lay aside a little inadvertent corruption (?) in the B-mode gravitational wave data for a moment — uh, Sean? What in the world is THIS now?

    This must be the week for “Can you top this?” when it comes to science-y party poopers. Not that there would be little upside to this thing NOT being, effectively, an ill-considered or crackpot exercise; but still… Could a few words of expert perspective be addressed to this, too, perhaps?

    EDIT: Oh, wait! A quick Google turns up –> This is old stuff, isn’t it? One minute in on E. Lerner’s home page… and I was already disinclined to read further. Maybe any comment would be better directed toward

    Like or Dislike: Thumb up 0 Thumb down 1

  20. Pingback: Has dust clouded the discovery of gravitational waves? - Technology Org