Spreading Galaxies Gospel on Facebook

Galaxy rotation curves: NOT flat (Noordermeer et al, 2007)


Paolo Salucci has a bone to pick with the community. The Trieste-based astronomer is fed up with his colleagues’ misconceptions about galaxy rotation curves and has decided to Do Something About It. In his short paper posted to astro-ph last Friday, he describes the experiment he’s set up to convince the world that galaxy rotation curves are not flat (oh sorry, that’s: NOT flat) – he has set up a Facebook group with plots, background, links and discussion, to orchestrate a change in the hearts and minds of astronomers around the world, to stamp out once and for all the damaging “hoax” of flat galaxy rotation curves.

Rotation curves describe how the rotation velocity in galaxies changes with increasing distance from the galactic centre. For spiral galaxies these curves are useful for learning about the galaxies’ matter distribution and, importantly, the presence of dark matter. But I’m not going to talk about rotation curves – it’s not my field and I’m happy to accept that they’re not flat.

I wanted to write about Salucci’s chosen approach of starting a group on Facebook. His reasoning is entirely justified: social networking, he argues, has become a true force in society, as a platform for rapidly spreading news and ideas through a large community. Despite the fact that the web itself originated in the physical sciences,  social networking hasn’t been widely adopted here. Indeed, it’s my impression that the biological and biomedical sciences have embraced the web 2.0 far more enthusiastically (although it may just seem that way as those communities are much bigger). Studying the potential of social networking for spreading scientific ideas through a global community is certainly an interesting proposition. So I completely agree with the core of his idea, and it’s a great experiment.

Other aspects of his argument, however, I find completely bizarre. First, he gives numbers to show the magnitude of the problem he wants to address:

Our thesis is proved if on a timescale of months there will be a substantial reduction in the number of submitted papers to astrophysical journals with a FRC [flat rotation curves] statement, (currently, these amount to ~15% of those published in the DMAG [dark matter, its alternatives, and galaxy formation and present-day properties] subject) or with ambiguous or no statement (~20%), the extinction of papers based on such paradigm (~5%). We also want the corresponding increase of works framed within the correct phenomenology (~60%). Furthermore, we expect a radical shift of thinking also in those many people that, although do not work in the DMAG field, still express FRC convictions in their seminars, lectures, public outreach events and articles, books, reviews and web pages sometime just loosely related to the subject.

If I understand these numbers correctly, he argues that 5% of papers in the category of “DMAG” are based on the entirely false notion that galaxy rotation curves are flat, 15% of papers contain a specifially erroneous statement, and a further 20% are ambiguous. Where do these numbers come from? He doesn’t describe how this count was carried out, over what timescale, or how the subject was defined beyond the “DMAG” description of all papers related to dark matter, its alternatives, galaxy formation and present day properties. In its broadest sense, that could encompass a substantial fraction of all astrophysics papers. Did he really count them all up?

Second, he expects an outcome “on a timescale of months”, which seems a little naive. Even if he convinced all these foolhardy flat rotation curvers on the very day the Facebook group launched, given the time to publication of scientific papers, the shortest timescales on which this kind of shift can be detected must be at least a year. Third and in keeping with the rest of his methodology, Salucci doesn’t discuss at all how he intends to measure the outcome of this experiment, or over what timescales. What does a “substantial reduction” mean? Isn’t a quantifiable outcome somewhat essential to running a successful experiment? Even something as simple as a user survey might tell you something useful. Science is science, even if it’s over Facebook, and this is no way to approach it.

Yet, despite the lack of clear methodology or quantifiable measures of success, he expects a dramatic and conclusive outcome that will have far reaching implications for social networking in science:

The test case is organized in a way that the failure of the mission leaves no hope to SNs to be of any help for astrophysics at the present time[…] The success of the experiment may mark the opening of a new phase in Astrophysics: a truly global community in which results and ideas are really shared and discussed and in which there is less space for prejudices and dogmas.

Even with a perfectly rigorous methodology, isn’t this a little overly optimistic/pessimistic? While both success and failure of this “experiment” would give some interesting food for thought, I doubt that it would prove anything conclusively. Like science itself, progress in research methods and adoption of new technology advances in baby steps. Also, the idea that using social networking would lead to less “prejudice and dogma” in the community leads me to think that Salucci has not spent much time on the internet! Social networking has huge potential for engaging with a large global audience – but sadly that applies to both “good” and “bad” science. With the best methods in the world, this experiment will not be the final word on anything.

The tone of the paper actually suggests to me that Salucci doesn’t think much of social networks to begin with. In his final discussion section, he poses the following question:

Are SNs undermining the peer review system?

Wait, what? Isn’t the fact that the notion of flat rotation curves has persisted in the literature despite being disproved over and over again observationally a surefire proof that peer review has failed? Why would SNs undermine that? The numbers quoted above, although it’s impossible to interpret them accurately without knowing how they were produced, would suggest that at least some referees who know that galaxy rotation curves are not flat did not make corrections to papers that stated the opposite. And:

Why would we want to establish new arenas for scientific debate and who will control them?

Why wouldn’t we want to establish new arenas for scientific debate? Particularly spaces without barriers in hierarchy, geography, financial status or even education level? That’s the fantastic thing about social networking: anyone can connect and participate, everyone has a voice, and everyone’s voice is equal. As for control, yes, some moderation is required. But from what I’ve seen on, say, the Galaxy Zoo forum (surely a great attractor for crackpottery!) this is perfectly manageable for a small number of people.

Finally, as a small aside, I’d like to comment on Salucci’s statement that “[…] in astrophysics, SNs and blogs do not play an important role in forming the prevalent scientific views, in influencing the scientific policy and in supporting specific ideas”. I’ve noticed on twitter that the discussion of the funding cuts in the British Science and Technology Facilities Council (STFC) has continued since last December between several prominent senior physicists and astronomers, PhD students, postdocs, and, importantly, the British science minister Lord Drayson and several more politicians. These funding cuts, together with other issues (e.g. the controversial sacking of a respected scientist from a government advisory committee on drugs policy and the use of libel laws to stifle academic debate), have led scientists to rally to make science a big voting issue in the upcoming election. Social networking platforms such as blogs and twitter are instrumental in this. So yes, I do cautiously believe social networking can have an effect on science policy – maybe not for astrophysics on its own, but on broader issues we do have a voice.

I think Salucci’s idea is really good. But the way it’s currently outlined, it won’t allow us to conclude anything. If Salucci would like this to be considered a real experiment, I’d like to see it described as one.

An update [13/04]

I actually had a nice chat with Paolo Salucci yesterday, he called me up to clarify some points I’d brought up here, which I appreciated. He does actually have recorded data on paper counts and intends to be more explicit about the methods of the experiment in a future publication, which I think is essential to be able to draw any conclusions from it. Salucci is very passionate, not only about rotation curves, but also about openness in science and better communication – and certainly I have no arguments with that. His paper was actually submitted to the open-access Journal of Science Communication so will be peer-reviewed and updated in due course – I’ve also modified the citation below to reflect that.  I look forward to seeing how the paper and the experiment pan out.


Paolo Salucci (2010). Can Social Networks help the progress of Astrophysics and Cosmology? An experiment in the field of Galaxy Kinematics Submitted to JCOM arXiv: 1004.1190v1


  1. Paolo was my advisor when I was at SISSA in Trieste. I remember well his arguments and admire his tenacity. I worked with him and Massimo Persic for almost 2 years on a sample (I still have) or over 1000 rotation curves…

  2. Kurtis W. says:

    This preprint raised two big flags in my mind. The first you mentioned, that the experiment is ill-defined. The second is that the topic, as far as I understand the issue, is more one of semantics than actual science. I have not heard anyone make a serious claim that rotation curves are purely flat (maybe because it is not my specialization), and I take it for granted that all rotation curves have some sort of structure. To me, the term “flat rotation curve” means “non-Keplarian” or more generally “not following the luminous matter”. A truly flat rotation curve is just a spherical cow, a starting point for teaching DM halos and singular isothermal spheres. Perhaps we are guilty of being sloppy with language, but I don’t know that discussions over vocabulary choices are really a good test of the scientific utility of social networking.

    The combination of the lack of discussion of the experimental design and the topic led the majority of astronomers I know who saw this paper to wonder if it was a joke, perhaps a late April Fool’s prank on the arXiv.

    We are all scientists, and we are trained and knowledgeable in the craft of designing rigorous experiments on well-defined topics. As-is, this paper falls short on both counts. The publishing of a non-rigorous experiment only hurts efforts to have web 2.0 efforts taken seriously by the larger astronomical community.

  3. Hi Kurtis,
    Yes I think you’re right – to the vast majority of us, the issue is one of semantics rather than bad science. But I can appreciate that we’re all a little more sensitive to semantics in our own field of work, I admit I’m pretty pedantic when it comes to language myself. Salucci also pointed out to me that scientists in other fields like particle phyics have adopted this “flat rotation curve” expression, presumable from astro papers, and may not have the background to understand the semantics. So you could maybe argue that our sloppy word choices can result in misconceptions in other fields.

    Re. web 2.0 “research” – again, agreed. Too many papers or preprints are published about the role of social networking in science that use tiny sample sizes, anecdotal evidence or poor surveys to make grand statements. It would be really good to see some proper rigorous work done on the subject.

  4. Though it’s true that not much science happens on social network sites like Facebook (though I did get an invitation to the Moriond conference on FB) – doesn’t CosmoCoffee count as a “a platform for rapidly spreading news and ideas through a large community”? In my view CosmoCoffee has proven to be highly successful with over 1300 members (which is a high proportion of the cosmological community).

  5. I was originally going to use CosmoCoffee as an example of a successful SN platform for discussion in astronomy. I don’t know it quite well enough to comment on its success, but it’s always been my impression that it’s a very useful and well-run resource for cosmology (that doesn’t seem to take up lots of people’s time to moderate…).

  6. oh holy crap! I’d never heard of or seen CosmoCoffee before! That discussion board looks awesome! Who’s the admin? I’d love a post on AstroBetter about this site…and their thoughts on implementing in for other sub-disciplines.

  7. John Gizis says:

    Uh, I don’t get it. Those rotation curves are flat.


  1. […] it accounts for the cosmic microwave background and its minute fluctuations, the non-Keplerian (but not flat) rotation curves of spiral galaxies. In 2006, X-ray observations of the colliding galaxy clusters […]