General
Home
News
IDL Speed Tips
Web Statistics

Science
Publications
Variable Stars In Globular Clusters

DanIDL
Features
Installation Instructions
Release History
v3.0

DanIDL-Lite
Features
Installation Instructions
Release History
v3.0

DanDIA
v1.0 - Coming Sometime

Blog
I dislike ArXiv...
I dislike Mobile Phone Companies...

Contact
Daniel Bramich
dan.bramich "AT" hotmail.co.uk
    

I dislike ArXiv...

What is "arXiv"?

A website where scientists (for the most part) upload papers and documents in the subjects of Physics, Mathematics, Computer Science, Quantitative Biology, Quantitative Finance and Statistics for wider distribution usually before actual publication. There is no proper screening or peer review of uploaded articles, and so one may find instances of plagiarism and poor quality articles (or crack-pot ideas) that are never officially published (because they fail peer review). Furthermore, most articles differ from the final published versions which means that one can never depend on references to arXiv articles.

Why do I dislike arXiv?

Apart from the problems with arXiv that I mention above, which are acceptable to some extent so long as one is aware of and accounts for them, the system is run in an opaque and discriminatory manner. This only becomes clear after various experiences of using the system.

As a scientist, it is important to disseminate my work as quickly as possible, especially in a competitive environment. From the moment a publication is accepted at the end of the peer review process, it can take 1-2 months to actually appear as published in the corresponding journal. Hence posting peer-reviewed work on arXiv brings it into the public domain with a wide audience as soon as possible (and sadly many scientists actually abuse the system by posting their articles before the peer review process begins). So I have been using arXiv since 2005.

My problems with arXiv started when I found an article by Safonova & Stalin that plagiarises text from my paper on NGC6981 (Bramich et al. 2011, MNRAS, 413, 1275). I immediately contacted arXiv to request that they look into the issue and disable the posting until it is resolved:

On 24/05/2011 04:42 AM, Daniel Bramich wrote:

Dear arXiv-moderation,

I was browsing astro-ph and came across the following posted paper:

http://de.arxiv.org/abs/1105.4363

Sections 2.2 and 2.4 contain phrases and even paragraphs directly lifted from my paper on NGC6981. Section 2.4 is entitled "Caveats of DIA" and Section 2.3 of my NGC6981 paper is also headed "Caveats of difference image analysis" (DIA stands for Difference Image Analysis).

My NGC6981 paper is already published with MNRAS as:

Bramich et al. 2011, MNRAS, 413, 1275

I request that this posting is disabled while you investigate this issue, and if you confirm the plagiarism I have detected, I would hope that at least it is requested that the relevant sections are rewritten in the author's own words.

Please let me know how this will be dealt with,

Yours Sincerely
Dr Daniel Bramich

Example:

NGC6981 paper Bramich et al:

Section 2.2, 3rd para:

"In each reference image, we measured the fluxes (referred to as reference fluxes) and positions of all PSF-like objects (stars) by extracting a spatially variable (with polynomial degree 3) empirical PSF from the image and fitting this PSF to each detected object."

NGC5024 paper by Safanova:

Section 2.2, 2nd para:

"In the RF, the fluxes (referred to as reference fluxes) and positions of all PSF-like objects were measured by extracting a spatially variable (with polynomial degree 3) empirical PSF from the image and fitting this PSF to each detected object."

The response from arXiv was surprising:

On 24/05/2011 05:05 PM, [www-admin] wrote:

Dear Daniel Bramich,

While we agree that their reuse of your text arXiv:1012.5652 is inappropriate in roughly 15 places corresponding to about 2.5% of their article, we note that they did reference your article. The reused a roughly equal amount of text from arXiv:0709.2728, also referenced. It is likely they are unaware that this reusage is inappropriate.

Unfortunately, there are many examples of much greater text copying in our database, including many without citation, and we are not currently set up to moderate these issues. We encourage you to contact the authors directly.

-- arXiv admin

So, arXiv confirmed that the plagiarism was not only from my article but also from another article by different authors. Furthermore, it is clear from their response that they have the automatic tools to detect and quantify plagiarism when it occurs. However, the arXiv policy is to wash their hands of the issue and leave it to me, the prejudiced party, to try and resolve it directly with the authors concerned! At least if an author like myself takes the time to report the plagiarism and it is confirmed by arXiv, then arXiv should be responsible enough to take action. My simple request to disable the posting until the issue is resolved requires very little effort on the part of arXiv.

What is most worrying is that it was only by good fortune that I read the offending article and recognised my own writing. This begs the question as to how many other articles exist on arXiv that contain plagiarised material of which the original authors are not aware? The answer is most likely many!

I replied to the above response from arXiv saying that their stance was unacceptable, but I did not get a satisfactory answer (i.e. still no action). In the mean time, I also contacted the authors Safonova & Stalin. Their response to the issue was to simply drop the relevant sections from their paper and upload a new version to arXiv (at least they took some remedying action!).

However, the situation at this point was still not acceptable to me because the original version of the Safonova & Stalin paper still existed on arXiv. So again I wrote to arXiv:

Date: Mon, 6 Jun 2011 11:40:10 +0200

Dear arXiv admin,

The authors have replaced a new version of this paper on astro-ph without the text from my paper which satisfies my original request to some extent.

However, there is still a version available (v1) with my copyrighted material. Can this version be disabled?

Thanks
Daniel Bramich

Sadly I never got a response to my email and the original version of the Safonova & Stalin paper continues to be available in the public domain via the arXiv website. I am not surprised by arXiv's lack of response because it has no accountability for its actions (which became really clear after my next problem).

I'll try to keep the description of the next issue quite brief even though it is a complicated one. In October 2011 I was contacted by a university friend claiming to have worked out a possible explanation for the apparent superluminal velocity of neutrinos in the OPERA experiment within the framework of special relativity. At that point in time there was no independent confirmation of the neutrino result and such a verification would be quite some time forthcoming, and so there was a rush of non-peer-reviewed papers appearing on arXiv trying to explain the result with exotic theories. The beauty of my friend's idea was that the experimental result could supposedly be explained without inventing any new physics.

I took about a week of my time to study my friend's arguments, perform some (pre-)peer review, and help him write the result into a format that could potentially be published. And given the explosion in the number of article's on arXiv dealing with this topic, I advised him that in this specific case it might be worth putting the idea out to the scientific community for wider discussion by uploading it to arXiv.

Immediately my friend had problems with the uploading process since he is not a physicist working at a recognised institution and consequently he was unable to register as a user on arXiv. This is fair enough since this sort of filtering helps prevent against crack-pots uploading anything to the site. In fact arXiv have an "Endorsement system" whereby a scientist with a proven scientific record (like myself) can endorse a fellow colleague to enable them to upload to arXiv. Having done some pre-refereeing on the article we wanted to upload, I carried out the endorsement process for my friend.

Unfortunately my endorsement was not good enough for arXiv and they put my friend's article on hold, and then it was rejected (by my good friend arXiv-admin):

From: [www-admin]
Date: 31 October 2011 13:40
Subject: 'arXiv submit/346060'

Your submission has been removed upon a notice from our moderators, who determined it inappropriate for arXiv. Please send to a conventional journal instead for the requisite feedback.

If you disagree with this determination, please do not resubmit the submission to any archive until you first explain the reason to moderation@arxiv.org and receive a positive response.

Please direct all questions and concerns regarding moderation to the moderation@arxiv.org address. More information about our moderation policies can be found at:

http://arxiv.org/help/moderation

-- arXiv admin

Indignified at this decision after having spent time carefully checking my friend's paper and taking responsibility over the fact that I was making an endorsement to allow someone else to post to arXiv, I wrote to ask for an explanation of the decision from arXiv. Independently, and rightly so, my friend also wrote to arXiv requesting an explanation for the rejection. Neither of us received a satisfactory response. The response to my request was totally uninformative:

From: moderation@arxiv.org
Date: 01 November 2011 15:27:14

Dear Dan Bramich,

Our moderators maintain that submit/0346060 is not appropriate for arXiv and should be submitted to a conventional journal. Our moderators are not referees and do not provide any feedback for submissions.

-- arXiv moderation

As you can see, no reason has been given as to why the paper is not appropriate. And since I already followed the endorsement procedure, the initial refereeing had been done. In fact, why did I even bother to waste my time when on the first time I endorse someone it is immediately overturned without any reason?

Further follow-up emails from myself about the fact that no clear reason had been given for the rejection were just ignored by arXiv (do you see a pattern here?).

Since the whole endorsement process is opaque to the outside user, I can only conjecture that my friend was treated with prejudice because his email is not associated with a university or another scientific institution. Furthermore, arXiv gets away with all of the problems I have described because it has no accountability. It holds the monopoly on the publication of non-peer-reviewed scientific publications and it can do what it wants without fear of repercussions.

Interestingly, arXiv is moving to a new funding model that asks for contributions from the institutes with the highest site usage. This will (hopefully) be the game changer. At some point arXiv will learn that it cannot treat its customers with disrespect since they are members of the institutes that are in control of the purse strings. Personally, I wouldn't send them a single euro!

By the way, it will be interesting to see if I am banned from arXiv for writing this webpage...

Conclusion

Unfortunately I still have to continue using arXiv as it has the monopoly on pre-prints in my scientific area. However I have adopted the following policies towards them:

  • I have never uploaded a paper to arXiv that has not previously been accepted for publication and hopefully I will never need to.
  • I refuse to include an arXiv reference to a paper in my publications. Until the arXiv article appears as an article in a peer-reviewed journal, there is no guarantee it will be accepted for publication, and even if it is, substantial changes may occur. In this way I never generate any references to arXiv material.
  • I have taken plenty of opportunities to discuss the above problems in discussion forums and also at science meetings. This way scientists may become more aware of the issues with arXiv policies.
I urge other scientists to also treat arXiv with caution and skepticism.


This site is © Copyright Daniel Bramich 2021, All Rights Reserved.
Free web templates