Wrong On The Internet

Just when I thought the Weinstein/Einstein kerfuffle had wound down, it seems to have decided to go another round.  Jennifer Ouellette updated her commentary with more good sense (including links to my earlier post – thanks!)  Peter Woit also weighed in, linking the Weinstein circus to some of his previous commentary. Woit’s question was why extensive news coverage of another Oxford event — a talk by Laura Mersini-Houghton on apparent observational evidence for a multiverse — did not provoke a similar outcry.  

Here it is in The Australian (possibly paywalled — but google for it, as it was widely reproduced) and I was asked about it by people who had seen it online and wanted to know if it was “really true”.  Mersini-Houghton’s paper has been around for a while; it claims that anomalously cold spots in the microwave background and large flows in the distribution of galaxies are evidence that our “universe” interacted with other parts of a multiverse immediately after the Big Bang. The abstract contains the memorable phrase

Placed in this context, the discovery of the bulk flow by Kashlinsky et al. becomes even more interesting as it may provide a probe of the preinflationary physics and a window onto the landscape multiverse.

— http://arxiv.org/pdf/0810.5388.pdf [Note the published version has a more circumspect abstract, but retains strong claims in the text.]

​Testable signatures of the “multiverse” are clearly newsworthy, so it is hard to fault the media for running the story.  Beyond the technical arguments, however, an immediate problem is that “dark flow” — large scale motions of galaxies coherent on scales of a billion light years or more — as claimed by Kashlinsky and others has always been regarded with skepticism, so it has never made sense to speak of its “discovery”.  

Moreover, other analyses failed to detect any anomalous large scale flow and the recent Planck results appear to rule it out with even greater confidence.  Kashlinsky is holding on, but it looks to be an increasingly quixotic mission for him and his collaborators.  And if dark flow really constitutes evidence for a multiverse, the absence of anomalous dark flow in our universe is, at best, a null result.

So Woit is right that the media coverage of the multiverse story left a lot to be desired. Woit may also be objecting to the idea multiverse itself, and there we would part company — many otherwise testable cosmological theories seem to imply the possible existence of a “multiverse” and cosmologists have a duty to follow their ideas wherever they may lead. As for the du Sautoy story, though, the objection was not that it was bad journalism, since du Sautoy is not a journalist — he gets to blog at the Guardian because he is a scientist, and publicly hyping work his colleagues have not seen is bad science

Weinstein himself will be talking again at Oxford today.  He is  steamed about the contretemps, if his twitter feed is anything to go by. What’s more, it looks like we might be waiting a while for the paper, since revolutionizing science is “just a […] hobby”.  Hopefully that will help everyone keep things in perspective. 

View fullsize


Trainwrecks I Have Seen…

[EDITED: The original description of Weinstein’s talk at Oxford said the physicists were not informed it was taking place — turns out they were sent an email, but no-one paid it a lot of attention.  The text below is edited according.]

​Yesterday a remarkable blog post by Marcus du Sautoy at the Guardian appeared multiple times in my Twitter feed.  Du Sautoy claimed that Eric Weinstein — a PhD mathematical physicist who now works largely in theoretical finance — had resolved a raft of fundamental problems in physics and cosmology.  On the face of it, this falls into the “interesting if true” category — a bold claim in an area where bold claims are common, and commonly wrong.  Shortly afterward, angry physics/astro friends lit up my Facebook feed, while Andrew Pontzen and Jennifer Ouellette took to their blogs to complain in a more measured fashion.  So why aren’t we happy? 

My own suspicions ​were raised by the inevitable picture of Einstein that accompanied the Guardian story, whose caption read, “Eric Weinstein’s theory is the first major challenge to the validity of Albert Einstein’s Field Equations”. This is flat wrong: hardly a day goes by without serious people posing serious challenges to Einstein. Problem is, none of them stand up. But the caption might have been written by a sub-editor, so I hit the reset button on my bullshit detector, and ploughed into the article.  

Turns out this is a theory proposed by someone well-versed in physics, but who now largely operates outside of formal academia. No real problem with that, but the signal-to-noise from this sector of the community is on the low side, Ramanujan notwithstanding.  But du Sautoy offered no details, and the theory has not been described in seminars or preprints, much less subject to peer-review.  Since we can’t see the details, discussion turned to the meta-story.  

​As Jennifer Ouellette pointed out, we can compare this to the approach of Tom Zhang, the mathematician who recently made huge progress on the twin prime conjecture, a long-standing problem in number theory.  He wrote a detailed paper, sent it to a top journal, got it accepted, gave a seminar at Harvard and then talked to journalists.  (Googling for coverage of this story, I love this blog at the AARP website — a lobby group for the over 50s in the US — which spins it as counterexample to the cliché that mathematicians all do their best work in their 20s.)  My own favorite example of this sort of rectitude is the discovery of the microwave background, which was announced in a paper entitled A Measurement of Excess Antenna Temperature at 4080 Megacycles per Second ​as opposed to We Have Discovered the Birth of the Universe, Now Can we Please Have a Nobel Prize.  Really good work usually sells itself. Conversely, over-hyped proposals typically under-deliver.

So if I had to guess on the basis of what we currently know, Weinstein is going to be filed alongside “surfer-dude” Garrett Lisi (see this New Yorker puff-piece for more) or Stephen Wolfram and his New Kind of Science — well-educated people now outside the profession, who make overblown claims that gain attention independently of their intrinsic merit.  Like Lisi Weinstein has an “interesting” story: he was as an economist who rang alarm bells before the Global Financial Crisis.  Like Lisi he has also found a few champions inside academia (du Sautoy for Weinstein, Smolin and Baez for Lisi).  Like Wolfram, Weinstein is super-smart, and has a significant reputation outside physics. But Wolfram still managed to produce a book one critic described as “a rare blend of monster raving egomania and utter batshit insanity” which failed to have any discernible impact on the old kind of science.  

The one source on the actual science is this Guardian article.  As a cosmologist, my eye is caught by the claim that Weinstein’s theory correlates the smallness of the cosmological constant with the overall curvature of spacetime, a quantity that changes as the universe evolves. A cosmological “constant” that depends on a property of the universe that can change with time sounds like what is known in the trade as dynamical dark energy. All astrophysical data suggests that our universe has a constant cosmological constant, rather than some more baroque form of dynamical dark energy — so it really would be good to see the details.  ​

Interestingly, looking at Weinstein’s blurb at the “Institute for New Economic Thinking” one finds this:

Dr. Weinstein has frequently championed scientific risk taking, the importance of self-taught scientists in their role of connecting fields, and the need to limit academic credentialism and rent-seeking behavior from hindering foundational breakthroughs in basic research.

— http://ineteconomics.org/people/eric-weinstein

So perhaps my concerns are just academic credentialism and an effort to preserve my own standing, hindering Weinstein’s foundational breakthrough. Or perhaps Weinstein has been breathing his own fumes for too long, and has managed to persuade du Sautoy that the usual standards of reproducibility, peer-review and the scrutiny of one’s colleagues are just for the little people. And here Weinstein really is challenging Einstein: the Swiss-German patent clerk played by the rules.

​Another thing we can be sure of is that du Sautoy has failed in his role as a public advocate for science.   If Weinstein turns out to be right, du Sautoy gets a pass. But if Weinstein is wrong, we will have yet another reminder of why ​science works best when we back our claims with argument, analysis and data, and when we let our colleagues test and reproduce our conclusions. 

I am writing this blog in Invercargill, a small city at the bottom of New Zealand’s South Island. I am here for the New Zealand Royal Astronomical Society’s annual conference. It’s a “pro-am” event: in this morning’s session I was struck by the talent, passion and commitment of amateur astronomers who are doing professional-level work with equipment they build themselves. These people genuinely get science, and could give a few tips to the Simonyi Professor for the Public Understanding of Science at the University of Oxford.

​Personally, I would be overjoyed if Weinstein is right — fundamental physics needs good ideas, and new ones are few and far between.  But in the meantime, my feelings are best summarized by The Wire’s Senator Clay Davies: