After breaking through 2 million papers, arXiv gradually lost popularity?

Written by | Zhang Qingdan

At the beginning of 2022, the preprint server arXiv received a New Year's gift, and scientists around the world broke through 2 million papers published by it.

It was supposed to be a time of celebration, but Scientific American published a negative report analyzing the many problems that arXiv is currently exposing, which has aroused heated discussion in the scientific community.

As The country with the largest number of papers in the world, the importance of arXiv is self-evident. On arXiv, the number of papers from China has also ranked second in the world.

Nowadays, many Chinese scientists watch arXiv every day at work, which has become a habit. Therefore, they also complained a lot about arXiv's pain points, especially the papers sent to arXiv were reviewed for a long time, delaying the release time.

The academic controversy brought by arXiv is fun and exciting

After more than 30 years of rapid development, arXiv has become an important academic communication platform for mathematics, physics, astronomy, computer and other disciplines.

Because at arXiv, researchers can independently post papers without peer review, publish research results as soon as possible, and seize ownership of scientific discoveries; those published papers can also be posted on arXiv to let more research peers understand and communicate.

Usually, there are three ways to post preprints in arXiv: first, post to arXiv at the same time as the paper is submitted, expecting more peers to see and carry out academic discussion and exchange; second, post to arXiv after the paper is about to be received or received, considering the limited number of audiences on the publishing platform or the need to purchase a database to read, posting the paper that has been finalized to a free public platform will undoubtedly increase the amount of reading; third, purely to express their academic views, Such manuscripts are generally unlikely to be accepted by formal journals, but arXiv provides a platform where you can speak freely.

Either way, it makes up for the lack of scientific and technological literature publication. This is also the reason why arXiv is on fire.

In traditional journals, the review cycle and publication cycle of papers are particularly long, averaging more than 30 days, and some even take several years. If, during this period, the author chooses not to publish the results of the paper, the research progress in this field will appear "outdated". Obviously, the way arXiv preprints make academic exchanges smoother, and some hot fields can develop rapidly.

Luo Huiqian, an associate researcher at the Institute of Physics of the Chinese Academy of Sciences, said in an interview with China Science Daily, according to his research experience: "In the research of iron-based superconducting materials in 2008, chinese scientists announced new superconducting materials and refreshed the critical temperature record in days, and the fast review and publication speed of traditional journals could not keep up with the speed of preprints on arXiv. Because of this, in the torrent of iron-based superconductivity research, the Chinese community of scientists quickly led the world's forefront. ”

Luo Huiqian pointed out that the rapid publication of arXiv preprints is an important way to seize credit, and many important discoveries nowadays are almost based on the time of preprints, which has become the "unspoken rule" of scientific research circles.

But posting a preprint in advance at arXiv can also bring "annoyance" to the author.

For example, a large number of questions and discussions may lead to the article being retracted before it reaches the official publication; the results of the paper are published prematurely, and in the highly competitive frontier field of science and technology, some people may take advantage of the time difference between the publications of different journals to seize the opportunity, resulting in the author losing ownership of the results; there are often many scientific errors in the preprints, and scientists who often publish immature preprints have a decline in trust in the eyes of their peers and are not good at standing.

Luo Huiqian gave a more specific example. At the end of 2014, the German scientist A. P. Drozdov and M. I. Eremets et al. discovered a paper on 203 K superconductivity under high-pressure hydrogen sulfide, which was first posted in arXiv, but the scientific community has been questioning this, and even at academic conferences, almost everyone present has tacitly accepted that this result may be wrong. Later, the results of this article were repeatedly reviewed and supplemented with experimental evidence before they were officially published in Nature, and it took half a year for the study to receive great attention.

The field of superconductivity research of high-pressure metal hydrides has developed rapidly in recent years, constantly refreshing the world record for superconducting critical temperature. Many papers were first published on arXiv, but some of them have not been officially published because of poor data quality or exaggerated conclusions.

Interestingly, in 2020, the research group of American scientist Ranga P. Dias found near-room temperature superconductivity (Tc=287.7 K) in a similar carbon-hydrogen-sulfur system, and they chose to submit the paper directly to Nature without pre-posting on arXiv.

But the data in the Dias et al. paper was widely criticized for being "too pretty." One of the most strongly opposed was J. E. Hirsch (who himself invented the H index, a citation evaluation method), has posted more than a dozen papers at arXiv discussing the experimental process and data analysis problems of hydride superconductivity in Dias and all scientists (including Eremets, etc.), some of which were officially published in Nature and Physiological Review Letters.

Dias and Eremets, among others, have also responded at arXiv, and the scientific community has been hotly debated at several academic conferences. Undoubtedly, the existence of arXiv provides a very suitable venue for this hydride superconducting "series", and the academic controversy continues to be widespread, and it also maintains the high attention of this field.

"There are many such phenomena in the scientific research circle, which is really interesting and exciting." Ji Yang, a researcher at the Institute of Semiconductors of the Chinese Academy of Sciences, said in an interview with China Science News, "I think the significance of arXiv's existence is to promote scientific exchanges and let scientific research continue to move forward in the controversy." ”

However, Ji Yang believes that arXiv's current approach has forgotten this original intention, which has caused many problems.

"ArXiv shouldn't be an academic referee"

The most reflected issue at the moment is auditing. One researcher complained that he pitched arXiv a fairly normal paper in 2017, with no problems with either the title or the language. Before that, he had posted more than 100 papers on arXiv, but this time he was strangely vetted for two weeks and finally approved without any changes.

"Fast speed and high efficiency" was arXiv's original chip for sucking powder. But in recent years, with the frequent intervention of auditors, papers have been delayed by days or weeks, and some have even been rejected. According to statistics, about 6% of submissions on arXiv will be shelved, and about 2% of submissions will be rejected directly.

The number of papers on arXiv by Chinese researchers has ranked second in the world, but there is no Chinese in the team of auditors. As the world's top scientific research paper library, most of arXiv's auditors are from research institutions in the United States, and the imbalance of auditor nationality will bring great challenges to academic fairness.

The audit rules given by arXiv are also biased in the eyes of many researchers. "Auditors are now in power, and they can directly decide the life and death of papers, but because of the limitations of their field, papers can be mistakenly killed in the review because of personal preferences." Ji Yang pointed out.

In August 2021, pan jianwei and Lu Chaoyang, two scientists who have made outstanding contributions to the field of quantum research in the mainland, were also rejected by arXiv. This shocked the scientific community, and Roberto Casadio, a theoretical physicist at the University of Bologna in Italy, evaluated the paper and wondered why the rejection was rejected.

"ArXiv shouldn't have a vetting system, it's a preprint platform in itself, it shouldn't be an academic referee, they don't have that ability." Ji Yang stressed.

After all, he adds, preprint servers are not journals, and reviews are not peer-reviewed. "If you want to go through the same review cycle as the journal, instead of tossing and turning on the preprint, it is better to go directly to the core journal to submit, even if it takes half a year, it is more 'fragrant' than in arXiv."

However, in Luo Huiqian's view, arXiv rejection and academic journal rejection are two natures, and the two cannot be equated. Just like posting articles on forums such as Zhihu, the platform has the full right to delete posts, and the author will not know it, not because the article itself has scientific problems. Over the years, in the massive number of submissions, arXiv rejected a large number of very unprofessional manuscripts, which have no value for academic exchange, and rejected manuscripts to save everyone's time and energy.

"Of course, it is undeniable that in today's sharp increase in manuscripts, this review mechanism will also cause manslaughter, but after all, it is an isolated incident." Luo Huiqian said.

Personally, he believes that arXiv's audit mechanism works well and does not need to be corrected. However, it is important for readers to be clear that the preprint paper on arXiv is not the same as official publication, and although it represents the author's point of view, it does not represent the recognition of the academic community.

Will you still contribute to arXiv?

Following the trend is a phenomenon that exists in every field, and as arXiv's influence grows, it has also spawned other preprint servers, such as viXra, bioRxiv and medRxiv, which accept submissions at a low threshold in order to compete with arXiv.

An elephant was already standing in a room, and there was more room left for others. "Those who are latecomers, in order to develop themselves, so there are no restrictions, but there is not much room for development." And if you can make scale in the future, you may forget the original intention like arXiv. Ji Yang said.

When asked whether he would still submit articles to arXiv in the future, Luo Huiqian said that "definitely", in his view, posting a preprint on arXiv is not equivalent to submitting to an academic journal, but only equivalent to a "free academic advertisement".

He also reminded everyone that since the arXiv platform does not assume the responsibility of peer review or formal publication, there is no obligation to correct scientific errors. This means that when reading arXiv papers, authors need to have sufficient screening ability to identify some papers with obvious scientific errors.

For the reasons put forward by Steinn Sigurdsson, director of arXiv Science, such as "no recruitment", "lack of funds", "huge audit workload", etc., in Ji Yang's view, these are pure excuses. "Now there are more than 2 million papers in the library, as long as the database is done well and the referees are removed, it is not a matter at all."

Ji Yang hopes that arXiv will not forget its original intention, it should belong to the majority of scholars, belong to the entire scientific community, and contribute to academic resonance and scientific progress. "If arXiv insists on going his own way, it will eventually lose popularity and will definitely be replaced in the future."

This article is reprinted with permission from the WeChat public account "Science Network", edited by Fang Yuan.

Special mention

1. Enter the "Boutique Column" at the bottom menu of the "Return to Simplicity" WeChat public account to view the series of popular science articles on different topics.

2. "Return to Park" provides the function of retrieving articles on a monthly basis. Follow the official account, reply to the four-digit year + month, such as "1903", you can get the index of articles in March 2019, and so on.