RE: https://scicomm.xyz/@JohnBarentine/116296832763875772

Ah. A Nature article that I can actually read. I still don't understand why scientists continue to submit their papers to companies like Nature or Elsevier which hoard vast amounts of research behind expensive pay walls. Can anyone explain?

@galaxy_map Honestly? Because in a lot of cases, they own the journals that are the highest-impact and which most people trust (rightly or wrongly). So publishing in Journal XYZ which has no name recognition _also_ means that people tend to trust the work less, cite it less, etc.
@galaxy_map Also, in a lot of cases, the large publishers have largely cornered entire fields (again, in terms of journals people actually _trust_). It shouldn't be this way, but it is right now (and has been for a long time).

@chiraag But their impact ratings would fall through the floor as soon as scientists stop publishing in those journals. Perhaps the solution is that governments should simply ban publishing publicly funded research in paywalled journals.

Normally I would have expected that the smart people who publish in these journals would have figured out a workable solution themselves. But since they have failed to do so decades after paywalls stopped making any sense, maybe governments need to step in?

@galaxy_map @chiraag

To a lot of the people publishing papers in these journals the paywalls are invisible. They work at universities and the like that pay extortionate institutional subscription fees to the publishers to obtain unlimited access for computers connected to the university network. They know that the paywalls exist, of course, and should care more about them, but they're not confronted with them on a regular basis.

Second, people did figure out a workaround: "pre-print" servers such as https://arxiv.org have been around a long time. Usage varies by field, but in astronomy at least most papers get uploaded to the arXiv prior to official publication where they can be accessed for free. A secondary benefit of widespread use of services like arXiv is that it matters less which journal you publish a paper in; if most readers are discovering papers through arXiv it doesn't make much difference how much traffic the journal's website gets.

More recently a number of free, open access journals have been created that use the arXiv infrastructure as their back end, e.g. The Open Journal of Astrophysics is an arXiv overlay journal that is free not just to read, but to publish in too.

That last point is important. One of the things that can hold authors back from publishing in open access journals is that they normally charge higher publishing fees than journals that charge readers for access. The administrators that hold the purse strings may not approve payment of these higher fees.

arXiv.org e-Print archive

@spacelizard @chiraag Yes, I get essentially all the research papers I use for astronomy from arXiv. Fortunately once those papers are published in most astronomy journals like Astronomy and Astrophysics they stay accessible. Astronomy as a discipline has *almost* slain the paywall monster. Unfortunately occasionally researchers will still choose to publish in Nature Astronomy which is often paywalled and therefore invisible to me.
@galaxy_map @chiraag Yeah, Nature Group journals are... special in various annoying ways, not letting authors post the papers on arXiv being one of them. Unfortunately many universities, funding bodies, etc. still consider Nature papers as extra good, so there is pressure on researchers to put up with Nature Group's bullshit and publish in their journals if they can.
@galaxy_map @chiraag Funnily enough if you actually want accurate information then you're probably better off ignoring the "elite" journals anyway. In their attempts to publish only the highest impact, most groundbreaking research they end up publishing a higher proportion of stuff that ultimately turns out to be wrong.
@spacelizard @galaxy_map @chiraag That is true, but that is true for high-risk high-gain results in general and on its own not an argument against a journal in general. The problem is more the hyping of results that are borderline

@spacelizard @galaxy_map @chiraag I thought you can in principle publish on ArXiv, but I am not sure. Most people choose not to.

This paper of us https://ui.adsabs.harvard.edu/abs/2016Natur.531...70B/abstract 10 years ago appeared on ArXiv

A large light-mass component of cosmic rays at 10<SUP>17</SUP>-10<SUP>17.5</SUP> electronvolts from radio observations

Cosmic rays are the highest-energy particles found in nature. Measurements of the mass composition of cosmic rays with energies of 10<SUP>17</SUP>-10<SUP>18</SUP> electronvolts are essential to understanding whether they have galactic or extragalactic sources. It has also been proposed that the astrophysical neutrino signal comes from accelerators capable of producing cosmic rays of these energies. Cosmic rays initiate air showers—cascades of secondary particles in the atmosphere—and their masses can be inferred from measurements of the atmospheric depth of the shower maximum (X<SUB>max</SUB>; the depth of the air shower when it contains the most particles) or of the composition of shower particles reaching the ground. Current measurements have either high uncertainty, or a low duty cycle and a high energy threshold. Radio detection of cosmic rays is a rapidly developing technique for determining X<SUB>max</SUB> (refs 10, 11) with a duty cycle of, in principle, nearly 100 per cent. The radiation is generated by the separation of relativistic electrons and positrons in the geomagnetic field and a negative charge excess in the shower front. Here we report radio measurements of X<SUB>max</SUB> with a mean uncertainty of 16 grams per square centimetre for air showers initiated by cosmic rays with energies of 10<SUP>17</SUP>-10<SUP>17.5</SUP> electronvolts. This high resolution in X<SUB>max</SUB> enables us to determine the mass spectrum of the cosmic rays: we find a mixed composition, with a light-mass fraction (protons and helium nuclei) of about 80 per cent. Unless, contrary to current expectations, the extragalactic component of cosmic rays contributes substantially to the total flux below 10<SUP>17.5</SUP> electronvolts, our measurements indicate the existence of an additional galactic component, to account for the light composition that we measured in the 10<SUP>17</SUP>-10<SUP>17.5</SUP> electronvolt range.

ADS
@hfalcke @spacelizard @chiraag I think the key is to keep the peer review process. Currently most researchers I follow wait until their paper is peer reviewed and accepted by a journal before they upload to arXiv.

@galaxy_map
Because publishing a paper in Nature is seen as more prestigious than in other journals. People believe that these papers will get more citations (maybe just a bias because people publish their most interesting papers there). And finally, selection committees tend to prefer applicants who have papers in Nature.

I don't like publishing there. It's much more work, and the style required for Nature often makes papers worse. But for early career scientists, it's different..

@brunthal @galaxy_map But this was not even Nature (the main journal) but Nature Astronomy (the rubbish bin for the main journal). I never got why colleagues allow their papers to be downgraded like that.🤷‍♂️
In this case it's the editorial, so I guess for the issue at hand it's good to have that published, and get some extra visibility of the problems mentioned.