For years, there has been a debate about the most effective strategy to achieve universal open access to scholarly publications. This has unnecessarily pitted two approaches against each other: the ¡°gold¡± approach, based on open-access journals, and the ¡°green¡± approach, based on open-access repositories. While there are strengths and drawbacks to each, both are critical ¨C for the moment, at least.

The Plan S requirement for immediate open access has further fuelled the debate. In a recent article published in ³ÉÈËVRÊÓƵ, Springer Nature's chief publishing officer, Stephen Inchcoombe, argues that transformative agreements (TAs) are the fastest route towards full open access. Yet these agreements ¨C which allow researchers to both access journals¡¯ subscription content and to publish in them open access ¨C are only available to institutions or countries with the substantial funds required to pay for them. In addition, they sometimes take years to negotiate and, because most institutions will not be able to afford TAs with all publishers, they lock researchers into publishing in specific venues.

Moreover, TAs do not transform journals to open access, but make individual articles available for a fee. While such ¡°transformative journals¡± are supposedly on a path to becoming fully open access, the real direction of travel is questionable. As such, transformative agreements alone will only result in a slow and partial transition to open access, with content remaining siloed in various publisher platforms.

That underlines the case for a parallel green route. Inchcoombe claims that repository content is of lower quality and has less visibility than the publishers¡¯ version. But the accepted manuscript (the most common version found in repositories) contains the same content as the published version. In addition, articles in repositories such as arXiv, Pubmed Central and Zenodo, as well as many institutional repositories, are both .

³ÉÈËVRÊÓƵ

Moreover, the repository route is much more than just a parallel system. It represents an investment in public research infrastructure that will expand over time and be responsive to the evolving needs of the research community. To optimise research communication, we need to have a diversity of content widely accessible for text and data mining. And not just articles: all valuable research outputs, such as data and code, should be made available through an interoperable network of trusted repositories.

Collectively, repositories represent a highly collaborative community. They are mostly hosted by long-lived research institutions and libraries whose missions are aligned with the aims of research and scholarship. And they are well positioned to support the innovation that is long overdue in article publishing itself. By linking articles to external peer review services and , repositories offer a low-cost and to traditional academic publishing.

³ÉÈËVRÊÓƵ

Moving towards a new ethos in scholarly publishing is important. The commodification of research outputs has not only led to unnecessarily high prices (first to access articles, and now to publish) and increasing market consolidation, it has also contributed to a significant decline in diversity and multilingualism in the academic publishing sphere, something critical for a healthy ecosystem. As the commercial publishers sought to increase their market share, they have taken over the smaller, more niche publications (often non-English journals, based at universities or academic societies) and transformed them into the kinds of ¡°international¡±, English journals that are more likely to be recognised in the databases heavily used by universities and funders for research assessment and rankings.

The existence of a free version of a paper puts pressure on the publisher to reduce prices and improve services. This, of course, is why publishers resist the green route. They have a huge financial stake in transitioning the system to a pay-to-publish model to maintain their significant profits.

Of course, we could quickly get to universal open access if we paid publishers the vast sums in open access fees they demand. But, in a system where the actual costs of publishing represent a , wouldn¡¯t this money be better spent if it were invested back into research itself?

At the end of the day, we should be less concerned about the publishing industry¡¯s desire to maintain its huge profits and focus on optimising publishing infrastructure and scientific funding to facilitate new discoveries and solve global problems.

³ÉÈËVRÊÓƵ

As with Plan S, funders, which have an interest in maximising the impact of science, should set the terms and conditions for the availability of their funded research outputs. And universities and funders should redraw their assessment measures to recognise the quality of the research contained in the article, rather than the venue in which it is published.

The service providers will just have to adapt or perish.

Kathleen Shearer is executive director of the Confederation of Open Access Repositories.