African scientists launch their own preprint server
The free, online outlet is one of a growing number where academics on the continent can share their work
25 June 2018
A group of open science advocates have launched the first preprint repository aimed exclusively at African scientists. AfricArxiv seeks to improve the visibility of African science by helping academics share their work quickly, say co-founders Justin Ahinon, a web developer and student in applied statistics at the National School of Statistics, Planning and Demography in Parakou, Benin, West Africa, and Jo Havemann, a trainer at the science communication consultancy, Access 2 Perspectives, based in Berlin, Germany.
They hope the preprint server will increase collaboration among researchers, and make knowledge more accessible to policymakers, entrepreneurs, medical staff, farmers, journalists, among other stakeholders.
The platform will be hosted on the Open Science Framework (OSF), a free, open-source software that allows researchers to connect and share their work. It will support preprints, postprints, code and data, and welcomes submissions from all African languages, including Akan, Twi, Swahili and Xhosa.
AfricArxiv is the latest of a number of open publishing platforms launched since the beginning of 2018. In March, the African Institute for Mathematical Sciences, and Elsevier announced that they would create the open-access journal Scientific African, and in April, the African Academy of Sciences (AAS) and F1000 launched AAS Open Research, which publishes manuscripts that go through an open peer review process. AAS Open Research has published 17 articles since going live, with eight more being revised.
The idea for AfricArxiv was generated by tweets by attendants at the open science summit in Kumasi, Ghana in April 2018. Only three months later, AfricArxiv has a platform, a Facebook page, Twitter account, and a team of 12 scientists, who have volunteered their time to promote the service, and check whether the content submitted is appropriate.
The speed of delivery is largely due to the ease with which the Center for Open Science can customize its OSF platform. AfricArxiv is one of 21 community preprint services built on the OSF, including Arabixiv, which disseminates knowledge in Arabic, and INA-Rxiv for Indonesian scientists.
The latter is among their best-performing archives, says Rusty Speidel, marketing director of COS. Since launching in August 2017, it has accrued 2,920 preprints.
For comparison, the well-established physics archive arXiv receives 10,000 submissions a month, and bioRxiv reached 1,000 monthly submissions, some four years after its creation in 2013.
The success of AfricArxiv will depend on the size and willingness of the community to share their work with each other and the world, says Speidel.
Some communities expressed initial concern about their work being scooped, but have been persuaded of the benefits of getting early feedback from their peers, he says, especially since many journals have begun to encourage researchers to share their manuscripts in repositories. Preprints are also assigned a formal DOI, which he says mitigates the risk of someone else taking credit for one’s work.
“The easy part is getting the platform up. The hard part is growing it, supporting it and moving it forward,” says Speidel.
Access and engagement
Tolu Odumosu, a science technology and society researcher at the University of Virginia, says access is a sore point among researchers in the region. While preprints can partly address the problem, other barriers also need attention, such as the prohibitive subscription rates of established journals and fees for publication. Institutions also need to invest more time, materials and space to research, says Odumosu. He would be interested in contributing his work to AfricArxiv if he knew that it would reach an active community of scientists in Africa, he says.
Another concern is one of recognition. “People are still stuck in the world of impact,” says Nelson Torto, analytical chemist and executive director of the AAS. “It will take a while for people to shift from so-called high-impact journals to a platform where they are able to share with their colleagues as quickly as possible.” In the case of preprints, however, the two need not be mutually exclusive.
For many researchers in the region, the more outlets the better. But some worry that scientists might confuse preprints with peer-reviewed journals. “There are always merits of sharing information,” says Torto. “But people need to know in what form that information is in, and what they can use it for.”
Ahinon and Havemann hope that clear definitions and guidelines will pre-empt any misconceptions.
Correction 26 June 2018: Nelson Torto is not currently affiliated with the University of Botswana.