Saturday, December 17, 2016

Severe inaccuracies in media stories about Verlinde's speculations

Last month, Erik Verlinde released another speculative paper that he intended to sell as a breakthrough to the media. It's his version of what I coined as the holographic MOND paradigm. Dark matter doesn't really exist.

Instead, the effects attributed to it result from some modification of the usual laws of gravity that is guaranteed by "fundamental physics" in combination with dark energy (or cosmological constant). One consequence is that the parameters controlling the observations displaying "dark matter" and those displaying "dark energy" aren't independent. Verlinde uses different details in the justification but they're as speculative as holographic MOND and have similar observational consequences as other MOND papers.

The Dutch media have persuaded themselves that it's the greatest event in science since the Big Bang. The journalistic class of the Netherlands – and other countries – is an echo chamber where some amazing group think is nurtured. I've been asked for interviews by 4 Dutch science journalists and rejected those offers for various reasons. The hype in the English-speaking media is much weaker than the hype in the Netherlands but it's still excessive.

To show some serious problems with the current state of the science journalism, let me pick two articles that quote a sentence of mine:
Nude Socialist, Mark Anderson: First test of rival to Einstein’s gravity kills off dark matter, 2 days ago

Science Alert, Fiona Macdonald: A controversial new gravity hypothesis has passed its first test, today
One thing you can notice is that these two articles are extremely far from being independent of each other.

They copy not only the plausible "reporting" about the "news". They share most of the mistakes, bizarrely picked details, and misinformation. To see that, start with the sentences that contain my name. New Socialist wrote:
String theorist Lubos Motl savaged Verlinde’s ideas in a recent blog post: “I wouldn’t okay this wrong piece of work as an undergraduate term paper.”
Science Alert told us:
String theorist Lubos Motl recently took down Verlinde's ideas a blog post, saying: "I wouldn’t okay this wrong piece of work as an undergraduate term paper."
The difference is that one journalist wrote "savaged" and the other one wrote "took down" while the word "recent" was changed to "recently" and moved to a different place of the sentence. ;-) That's the kind of "qualified editing work" for which these journalists are being paid hundreds of dollars per day.

Try to quantify the probability that these two sentences arise from two independent journalists' evaluation of the actual stories that they observe in the real world. Needless to say, this degree of similarity is implausible. There are over 7,000 blog posts on this website, a dozen is about Verlinde's speculative ideas, and each of them contains a large number of sentences.

And I am obviously not the only person who has criticized Verlinde's ideas.

What is your estimate of the probability that two journalists pick exactly the same sentence from the same critic of Verlinde's ideas and introduce it in almost the same way? The probability is tiny. In other words, the evidence that what is going on is actually a form of plagiarism is extremely strong, much stronger than the evidence for any speculative ideas of Verlinde's.

I don't want to single out Fiona Macdonald. Journalists – and perhaps especially science journalists – are copying stuff from each other all the time. Most of the stuff they copy from each other is garbage.

OK, I was critical and they could have informed the readers about the criticism. Another problem is that the content of the criticism is reported completely incorrectly by both of them. If you read my actual text about the term paper, you will see:
I wouldn't okay this wrong piece of work as an undergraduate term paper but he got 6.5 millions of euros for this absolutely worthless pile of feces so many people who are impressed by the money but don't have an idea about science – which includes virtually all journalists – started to think that Verlinde is a top physicist.
The point is that the comment about the "term paper" wasn't a comment about the recent Verlinde's paper at all. It was a comment about an old idea – namely the "entropic gravity" that Verlinde has gotten lots of money for years ago. You don't need to read the whole context of my blog post. The sentence they "quoted" is absolutely self-sufficient and makes it clear that I am talking about some older work by Verlinde. They just ignore this self-evident fact. Independently of my sentences, both of these journalists have been completely unable to notice that it's two totally inequivalent ideas by Erik Verlinde that we're looking at. They have completely conflated them.

In 2010, Erik Verlinde incorrectly wrote that gravity arises from the physical objects' desire to increase the entropy i.e. gravity is an entropic force. A month ago, he wrote about emergent gravity that unifies the dark matter and dark energy phenomena. The words "entropic" and "emergent" may sound the same to a very sloppy journalist but they're not the same at all. These two papers are in no way equivalent.

And the newer paper is in no way building on the assumptions made by the earlier paper. The 2010 paper is the reference [14] in the new paper and the only place where the paper [14] is referred to in the new paper are the sentences:
In this way it was proven [10, 11] that the entanglement entropy indeed obeys (1.1), when the vacuum state is divided into two parts separated by a Killing horizon. This fact was afterwards used to extend earlier work on the emergence of gravity [12, 13, 14] by deriving the (linearized) Einstein equations from general quantum information theoretic principles [15, 16, 17].
So a month ago, Verlinde was just reviewing some vaguely related speculations about the possible non-fundamental character of gravity and his 2010 paper was one of them. But this 2010 paper isn't really an "important pillar" underlying the new 2016 paper. In fact, they're probably incompatible with each other. You may conclude that the journalists are writing about two totally different speculations and they don't see the difference. It's like a journalist who writes about Einstein's special relativity and identifies it with Einstein's theory of the photoelectric effect.

This sloppy culture of journalism may be seen not only in science journalism. These days, a really crappy journalist in a no-longer-credible daily, e.g. the Washington Post, writes something and tons of similar inkspillers copy it into all of their own media. Many stupid readers think that when the media and journalists have many names, the information was confirmed by independent sources and must be true. But that's a totally wrong assumption. The writing of these cliques of journalists totally fail to be independent. In fact, you can often be almost certain that a piece of illogical junk will be uncritically copied from AB to CD and EF and GH if AB happens to write it.

Now, the Nude Socialist and the derived Science Alert stories celebrate a fresh paper
First test of Verlinde's theory of Emergent Gravity using Weak Gravitational Lensing measurements
by Margot Brouwer, Konrad Kuijken, and 20 other astronomers. They say that a graph one may extract from some lensing is consistent with the predictions of Verlinde's 2016 paper – which has no parameters. Note that this Brouwer et al. 2016 paper only builds on Verlinde 2016. The paper Verlinde 2010 or 2011 (on entropic gravity) is just vaguely mentioned along with Ted Jacobson's papers and others, just like it's just vaguely mentioned in Verlinde 2016.

What is the character of the agreement? Look e.g. at Figure 3 on Page 11 of the 22 astronomers' papers. There are four graphs of ESD, the excess surface density, as a function of the radius. The empirical data points (crosses) agree with the predicted curves – basically straight decreasing lines with the same slope – rather well. But the agreement in all the quantities amounts to a "correctly predicted ESD up to a factor of two in one direction or another".

I think that this test is basically equivalent to the usual "numerological" tests of MOND theories and doesn't strengthen the case for MOND or any version of it – like Verlinde's version – at all. I do think that a MOND-like paradigm is plausible and may have a justification. But the whole MOND paradigm also has somewhat serious problems that could very well be proofs that the paradigm is wrong. Papers like this one don't really change the situation if you look at it carefully. They're looking at the same class of observations where MOND has "sort of worked" and, not too surprisingly, still "sort of works" – and they are satisfied with the usual unimpressive accuracy – while they ignore the places where "MOND doesn't seem to be too healthy".

The positivity of the message in the Nude Socialist and derived Science Alert articles is excessive. But the journalists aren't experts and maybe they should be forgiven, you might suggest. Maybe they're not deliberately deceptive, they just don't quite understand what's going on. However, some titles are just outrageous and you may be sure that they were used to deliberately deceive the readers. The title of the Nude Socialist article reads:
First test of rival to Einstein’s gravity kills off dark matter
Oh, really? So Verlinde's ideas are so great that you just pick the first test and it not only validates Verlinde's theory but also "kills off dark matter" as a side effect – one of the most famous concepts of the 20th century cosmology is just ruled out easily. The test surely cures cancer, too. Is the Nude Socialist's summary of the article by 22 astronomers accurate? Well, let's ask one of the 22 astronomers who actually wrote the paper:

The headline is totally misleading, Kuijken says. A short discussion under his tweet agrees that the quality of Nude Socialist wasn't good for many years and the headline is a typical example of "click bait". The paper by 22 astronomers clearly doesn't make any statement that would in any way imply that they have "killed off dark matter".

The dark matter paradigm – or a model with a somewhat larger number of parameters than Verlinde's MOND – is obviously consistent with the lensing data studied by the 22 astronomers. Nude Socialist and Science Alert know that. Brouwer has told Nude Socialist that Verlinde's MOND has fewer parameters but dark matter fits the data better than Verlinde's formula. If it fits better, it couldn't have been possibly "killed off", could it?

"Click bait" has become omnipresent in the media that pretend to be science media – such as Nude Socialist. They are trying to attract as many readers and clickers as possible. Titles such as "quantum mechanics or string theory or supersymmetry or dark matter is killed off" sound like "intriguing stories" for them play an analogous role as the headlines "look at the celebrities who have aged badly" – except that the texts about the celebrities, while annoying, are much more true than the claims about physics. The "science journalists" just don't give a damn that their titles are complete lies.

So you can easily prove that both pop-science authors knew that the observations are compatible with dark matter. But that didn't prevent Nude Socialist from picking the title saying that the test "killed off dark matter". They are just lying straight into your face. And this is extremely far from being the only major lie in these articles. They also tell you, among other things, that Verlinde basically discovered MOND. The quality of the science journalism in Nude Socialist, Science Alert, and lots of similar "mainstream" media has collapsed beneath all tolerable threshold. These inkspillers have turned into a stinky worthless cesspool.

No comments:

Post a Comment