Is arXiv a monopoly bully in scientific publication?

After decades of ever-increasing dominance, has become the largest and most popular storage space or eprint service for scientific publications in physics and other related fields. It would have been the most beneficial to the physics community had arXiv sticked to its original principles for sharing new ideas and works quickly. Sadly, is becoming more and more arrogant these days instead. Veiled censorship by arXiv is making it just like another giant refereed journal system but without any transparency.

This is especially bad because of two reasons. First, arXiv is the largest preprint archive in physics and the widest user community established since 1990s makes it indispensable. It does possess a monopoly status in physics publication, as there is no other similar service that can compete. Second, arXiv’s moderation is much worse than a normal journal’s peer review system as it is anonymous, no valuable response, and operated in a black-box way. There is no real mechanism to regulate arXiv’s bully behaviors as a monopoly. As such, arXiv has made new (especially non-orthodox or disruptive) ideas hard to spread.

There have already been quite some reports that exposed arXiv’s censorship issues. One of the most high-profile ones may be the experience by Brian Josephson (Nobel Prize laureate in 1973), as described in “Covert censorship by the physics preprint archive“.

I have endured similar issues with arXiv. I am blacklisted as well. I am no longer allowed to post any article to the HEP categories. The several articles I posted there before were moved to the general category: gen-ph, including a couple of them that were published in HEP journals (Phys. Lett. B and Phys. Rev. D). Even when I recently tried to post articles to the “allowed” gen-ph category, they’ve been on hold for nearly two months so far. I would not be surprised if they eventually deny both as they did for my other submissions earlier. Ironically, I wrote one of them for the Gravity Research Foundation 2021 Essay Competition and it has just earned Honorable Mention. What’s funnier is that at least 13 other papers submitted at the same time for this renowned competition got accepted by arXiv but have earned nothing in the competition. I am not saying that those 13 articles do not deserve to be in arXiv, they do; but why did arXiv block mine which the GRF referees regard scientific and better?

The best discussion on arXiv’s censorship could be the anonymous comments I found on the web. Indeed, it is not arXiv’s business to do the peer review on submissions. Instead, they should provide the best platform for helping “scientific startups” to grow into the next “scientific giants” (like in the business world). After all, it is the researchers who post their ideas on arXiv that bear all the risks.

What would be better ways of running arXiv? Most importantly, it should stick to its genuine original commitment – quick sharing and storage of new ideas. It should accept all honest scientific publications meeting the typical social (no abuse, prejudice, violence, plagiarism, etc.) and minimal scientific standards (excluding only clearly non-scientific works). Such minimal scientific standards are like the basic requirements for a business license. The business world has the most successful experience of fostering ample startup companies which occasionally grow into the next leaders of the market. The academic world, especially arXiv and funding agencies, need to learn from the business counterparts.

Many other issues on arXiv including the endorsement one originate from its over-regulation or arrogance. I guess that they consider themselves the defenders of scientific orthodoxy, just like the medieval Church to a lesser extent. Maybe we should be looking forward to the next true scientific Renaissance that will hopefully stop such pro-orthodox practices all together.

Just to show another example for arXiv’s arrogance, the forceful latex-source-submitting policy is getting more and more annoying. They might have a point in forcing people to use their latex system if they have the best and most-up-to-date latex compiling system. Instead they have an old system, missing features and packages. Even if one gets a manuscript well compiled under one’s own latex system, it may not work the same way in arXiv’s system. In such cases, one has to live with the ill-formatted pdf file generated by arXiv and sometimes it may not even compile under the arXiv’s system at all. But arXiv is still bullying you to use their system no matter what. They even adopt very aggressive steps to make sure that you can’t get around it. This is much worse than what Microsoft or any other business monopoly has done. You could still get around the restrictions (e.g., browsers) set by business giants. But you can not when it is arXiv.

In the end, it is the authors of the intellectual work who should have the right on how the work is presented. Unlike the regular journals, arXiv is not a real publication system, but a storage one. They could ask authors to upload their source files and MEANWHILE give them the option to use their own pdf file for the final presentation. By no means, arXiv should force one to use their compiling system, especially a poor one.

They list five reasons of forcing the latex-source upload which do nothing but expose their arrogance and absurdity:

  1. “latex is plain text and easy for future document migrations.” This is probably the most credible excuse. But the most important thing for a scientific work is to keep it authentic. Most likely, authors  have checked the pdf file they generated themselves much more carefully than what arXiv created. The original author-generated pdf file should be automatically in the upload. If there is any doubt, that should be the reference.
  2. “Using emerging new technology -> hyperlinks” How many times have they reprocessed the old manuscripts? Do people really care? Are the old stuff/links still working? What if the reprocessing messed up things – you’d need the original pdf file for comparison.
  3. This reason just states latex can generate pdf. There is no point if author-generated pdf file is provided.
  4. “There is no single Postscript standard” is for ps files but not for pdf files
  5. “Cross-referencing within arXiv is added automatically with hyperlinked Postscript” again for ps files not pdf files.

All these questionable / bullying practices by arXiv have caused some physicists’ counteractions. In particular, a more lenient archive system viXra was established for unorthodox articles. A website of was created against arXiv’s blocking activities and censorship. There are also other preprint services emerging, e.g., OSF Preprints,, etc. However, none of them has comparable influence as arXiv does yet. For healthy advancement of science, we need a better service for preprints. Reform on this and other aspects (e.g., funding) by advocating the principles of open science will be critical for avoiding the stifled progress in science.

Author: Wanpeng Tan

I share my ideas and thoughts mainly about mirror matter theory and open science on this blog. Under the new theory, we live in the universe with a mirror (hidden) sector of particles. A perfectly imperfect (minimally broken) mirror symmetry is the key to unlock the beauty and elegance of our universe. Click on the menu links for a popular introduction, a technical summary, and list of my papers on the new mirror matter theory.