Gold Open Access Publishing Must Not Be Allowed to Retard the Progress of Green Open Access Self-Archiving
Total Page:16
File Type:pdf, Size:1020Kb
LOGOS Gold Open Access Publishing Must Not Be Allowed to Retard the Progress of Green Open Access Self-Archiving Stevan Harnad Introduction Universal Open Access (OA) is fully within the reach of the global research community: research institutions and funders need merely mandate (green) OA self-archiving of the fi nal, refereed drafts of all journal articles immediately upon ac- ceptance for publication. The money to pay for gold OA publishing will only become available if Born in Hungary, Stevan Harnad did his under- universal green OA eventually makes subscriptions graduate work at McGill University and his gradu- unsustainable. Paying for gold OA pre-emptively ate work at Princeton University. Currently Harnad today, without fi rst having mandated green OA is Professor in Electronics and Computer Science at not only squanders scarce money, but it delays the University Southampton, UK, and Canada Research attainment of universal OA. Chair in Cognitive Science at Université du Québec à Montréal. He was founder and editor of CUP’s Be- Open Access (OA) means free online access. What havioral and Brain Sciences journal, past president of made Open Access possible was the advent of the the Society for Philosophy and Psychology, external networked online medium: the Internet, and even- member of the Hungarian Academy of Sciences, and tually the Web, empowered the authors of digital author and contributor to over 300 publications. works to give them away free for all online if they Harnad’s research interests are in Open Access Sci- wished. entometrics, Category Learning, Symbol Grounding The term “Open Access” was fi rst coined by the and Language Origins. One of the most devoted and Budapest Open Access Initiative (BOAI),1 spon- outspoken supporters of Open Access to research, sored by the Open Society Institute (OSI)2 in 2001. Harnad spends a considerable amount of time advo- But the idea of providing free online access – and cating Open Access to scholarship worldwide. the provision of free online access – started much earlier than the BOAI and the adoption of the E-mail: [email protected] name “OA.” The inventors of Unix3 and the In- ternet4 – mostly computer scientists – had already Websites: http://www.ecs.soton.ac.uk/people/harnad been providing OA to their research papers by http://openaccess.eprints.org self-archiving them in “anonymous FTP archives”5 since at least the 1970s. With the invention of the Web 6 in 1990, websites soon became the preferred way of self-archiving papers. High energy physi- cists – who had already been systematically shar- ing their works on paper before the Internet, and DOI: 10.1163/095796511X559972 then via email when it became possible – began 86 LOGOS 21/3-4 © 2010 LOGOS Downloaded from Brill.com10/03/2021 01:50:42AM via free access Gold Open Access Publishing Must Not Be Allowed to Retard the Progress of Green Open Access Self-Archiving self-archiving them in Arxiv,7 a centralized physics science are cited signifi cantly more than non-OA web archive, in 1991. Many individuals from many articles.17 Many subsequent follow-up studies18 con- other disciplines have since followed the lead of fi rmed that this “OA impact advantage” was also the computer scientists and the physicists. present in every other scholarly and scientifi c fi eld tested. But even the OA advantage was not suf- The “Subversive Proposal” to Self-Archive fi cient to induce the 85% of non-self-archiving au- Refereed Journal Articles thors to do so. A “Subversive Proposal”8 to make all refereed journal articles free for all by self-archiving them Mandating Self-Archiving online was posted in 1994. The proposal also iden- It had already been proposed since 1998 in the tifi ed the way to cover the cost of publication if American Scientist Open Access Forum19 that uni- OA self-archiving eventually made subscriptions versities and research funders should mandate OA unsustainable: fees for publishing individual arti- self-archiving (i.e., make it a requirement, as a cles instead of subscription fees for accessing them. natural extension of the requirement to publish-or- The fi rst OA journals began appearing in 1989;9 perish).20 The School of Electronics and Computer most were either the online versions of subscrip- Science at Southampton University (UK) was the tion journals or they were subsidised online-only fi rst in the world to adopt an OA self-archiving journals. mandate, in 2002.21 The fi rst university-wide OA Meanwhile, the 1994 Subversive Proposal to mandate was then adopted by Queensland Univer- self-archive went largely unheeded: For the follow- sity of Technology (Australia)22 and the fi rst Eu- ing decade, the rate of author self-archiving con- ropean university-wide mandate by University of tinued to hover at about 15-20 percent10 of yearly Minho (Portugal),23 both in 2004. refereed research output. The proportion of arti- cles published in OA journals was even lower.11 Providing centralized archives like Arxiv for other In 2001, Steve Lawrence disciplines (e.g., CogPrints12 for the Cognitive Sci- published a paper in ences) likewise failed to increase the rate of OA self-archiving. Nature reporting that In 1999, the Open Archives Initiative (OAI)13 OA articles in computer developed a metadata-tagging protocol in order to make all Open Archives “interoperable,” which science are cited means that depositing locally in any individual signifi cantly more than archive became equivalent to depositing centrally in one global, seamlessly searchable Open Archive. non-OA articles. In 2000, free software (EPrints)14 was designed at the University of Southampton (by adapting the Likewise in 2004, the UK Parliamentary Se- CogPrints software to make it OAI-compliant and lect Committee on Science and Technology rec- generic) to make it possible for all universities to ommended that universities and research funders create their own OAI-compliant Open Archives should mandate OA.24 In the same year, the US (which soon came to be called, instead, “Institu- House Appropriations Committee, too, recom- tional Repositories” [IRs]).15 Many IRs were subse- mended that NIH should mandate OA.25 The UK quently created, worldwide – their growth has been government failed to act on the Committee’s rec- monitored by the University of Southampton’s ommendation, yet within a few years all seven of Registery of Open Access Repositories (ROAR)16 the UK Research Councils nevertheless followed since 2001 – but IRs remained near-empty because it, each adopting a self-archiving mandate of its 85% of researchers still were not self-archiving. own.26 The Wellcome Trust became the fi rst re- In 2001, Steve Lawrence published a paper in search funder to mandate OA in 2005.27 In the Nature reporting that OA articles in computer same year, NIH adopted an OA request instead of a 87 LOGOS 21/3-4 © 2010 LOGOS Downloaded from Brill.com10/03/2021 01:50:42AM via free access Stevan Harnad mandate; that policy failed and was upgraded to a OA (publishing). In fact, the fastest and surest road mandate in 2007.28 to OA is the green road of OA self-archiving,44 A further incentive to mandate and provide OA for two fundamental reasons: (1) providing green was provided by the fact that the outcome of the OA is entirely in the hands (and interests) of the UK Research Assessment Exercise29 – in which providers of the research itself, the global research peers review and rank the research publications community, and (2) green OA can be mandated – of all departments of all UK universities every six whereas gold OA is in the hands of the publishing years – turns out to be highly correlated with the community and cannot be mandated. citation metrics that OA has been shown to in- Hence green OA needs to come fi rst, and it crease.30 The University of Southampton has been needs to be universally mandated, by institutions strongly promoting the development of OA met- as well as funders. It has been a great strategic mis- rics31 to track, evaluate and reward research usage take to wait instead for Gold OA.45 If, despite all and impact, creating Citebase32 as a model for a the benefi ts, most authors are not providing green scientometric engine for research evaluation and OA spontaneously of their own accord, at no cost, navigation and IRStats33 for gathering IR usage and without having to abandon their journal of metrics. choice, then they certainly will not provide gold Two international, cross-disciplinary author sur- OA, for an additional cost, and having to publish veys by Alma Swan in 200534 reported the most in a gold OA journal instead of their journal of fundamental strategic and practical fi nding about choice. Nor will their institutions have the mon- why OA growth had been so slow: although most ey to pay their authors’ gold OA publishing costs authors do not self-archive, over 90% of them in- while those funds are still tied up in paying for dicate that they would self-archive if their funders journal subscriptions. Nor can institutional jour- or institutions mandated it – over 80% of them nal subscriptions be cancelled while the journals’ indicating they would do so willingly. Outcome contents are still not otherwise accessible to the studies35 from Arthur Sale in Australia have since institution’s users. Moreover, the asking price for confi rmed that within two years of mandate adop- gold OA publishing is still much higher than it tion, compliance rates are indeed over 60% and needs to be, while journals are still producing print well on the road toward 100%. ROARMAP36 and online editions.