Preprints in debate… six years later

By Adrian Gurza Lavalle

Image: Riho Kroll.

Preprints began to be discussed by social science journals editors just over six years ago, with suspicion and astonishment at a practice that seemed foreign to their disciplinary tradition.1 Borrowing Umberto Eco’s well-known formula, it could be said that editors in the field were divided between a majority of apocalyptic ones, fearful of what seemed to them to be an imposition with undesirable consequences, and a minority of integrators, who identified preprints as an innovation with promising benefits. A few years on, and with the SciELO Preprints server now in place, it is possible to look back at the “threats” and “promises” of those times, benefit from a more sober and fact-informed appraisal and answer to the question: what has actually been achieved in this short but intense period of time?

Reminiscence is not a fictitious resource at the service of argument. In 2017, the traditional ANPOCS journals forum, at its 41^st Meeting, organized the Colóquio Novas Mudanças, Experiências e Estratégias no Campo da Editoração Científica nas Ciências Sociais (Colloquium New Changes, Experiences, and Strategies in the Field of Scientific Publishing in the Social Sciences). Abel Packer was invited to talk about preprints and their progressive implementation as part of the SciELO collection’s inclusion and permanence policies. He accepted and, thanks to a round trip on the same day to Caxambu, MG, Brazil, took part in what I believe was the first public debate with scientific editors in the field about the subject. As expected, the critical tone was predominant and Abel’s final summary was premonitory: there will be resistance, but it will pass, and they will end up adopting preprints.²

In general terms, mistrust and criticism revolved around preprints themselves and some of their consequences, which were considered predictable. Firstly, preprints were seen as a practice alien to the social sciences, driven by the celerity requirements of other areas. The practice was also exotic to the disciplines in the area because, in them, the importance of writing in the preparation of results and the notion of the strongly authorial nature of the text inhibited its “early” circulation. On the other hand, as far as the consequences are concerned, preprints would jeopardize anonymity, compromising the process of impartial assessment of manuscripts and, ultimately, contributing to eroding the role of journals, which were already undermined by trends in scientific publishing. Thus assessed, preprints seemed deleterious, and in most critical evaluations, their adoption was seen as an imposition.

In retrospect, it is surprising how much the reading of the “threats” depended on a mistaken assumption, namely the exotic nature of preprints. A different movement would have been possible by asking about the specificity of preprints in the social sciences. Working papers (WPs) or discussion texts (DTs), as well as the events proceedings, are forms of preprints with a long presence in scholarly communication practices in the field. It is no coincidence that at this year’s meeting ANPOCS adopted the SciELO preprint server for submitting Working Group papers. The initiative is fortunate and allowed by the homology between preprints and proceedings.

It should be noted that these types of preprints are not only uncontroversial but are also considered good practices for the circulation and production of scientific knowledge. WPs, for example, have several functions: to widely circulate research results and ideas in order to gather criticism and improve them with a view to publication – a function analogous to that of preprints – to offer detailed methodological information or detailed empirical results in order to be able to refer to them in scientific articles where, due to space limitations, such detailing is not possible. Certainly, the reasons why preprints are important in areas such as medicine or experimental sciences, notably the relevance of the precedence of discovery in the definition of funding, differs from the functions they perform in the social sciences, but this does not make them a practice alien to their tradition.

On the other hand, the promises of preprints made at the time have only been partially fulfilled so far. Three advantages have been exploited. Firstly, preprints would contribute to improving the knowledge produced by the scientific community thanks to public criticism, not restricted to referees or close peers, but open to the community, in a freer logic similar to the exchange of ideas on social networks. A timely analogy with WPs/DTs would have advised caution regarding these expectations. We know that the benefits of enrichment through debate and open criticism depend on additional, very demanding work. Publishing in itself does not bring about debate, it is merely a necessary condition. The most successful WP series are linked to well-funded dissemination machines run by the research institutions to which they belong. Without the work of dissemination, WPs, like preprints, give rise to little or no debate. Obviously, it is unrealistic and unreasonable to expect SciELO Preprints to invest heavily in dissemination. In this respect, however, the PREreview request feature, a recent innovation in SciELO Preprints to allow evaluation, is a step forward.

Secondly, journals in the field would benefit, as they could proactively select better manuscripts on the preprint server, already enriched by the debate, with clear indications of relevance to the respective subfield, and could even incorporate the previous citations of the manuscript in its various versions as a preprint. I do not know of any publishers in the field who proactively select preprints, although the SciELO Server offers alternative metrics for the circulation and use of manuscripts. In any case, an important part of these benefits still depends on the previous point. In turn, the incorporation of previous citations is a point that, I venture to conjecture, is still dark for social science publishers. It is not uncommon for WPs and papers in event proceedings, when they have the rare fortune of gaining traction in the debate in their subfields, to circulate with a “double life” after being published in their final versions in scientific journals: the journal version and the version that first gained circulation, the latter usually continuing to be cited. It is unclear whether and how preprints can circumvent the risk of “double life”, which discourages their publication in journals, especially considering that the DOI is assigned to unique digital objects and that, in theory, the preprint published in a journal should be given a new DOI.

Finally, the advantage that has been fully realized is the speediness of publication. In fact, in a six year period the preprint server has been implemented, is operating regularly, and making it possible to publish scientific manuscripts almost immediately, as well as identifying them persistently by assigning them a DOI. This is undoubtedly a major step forward. Although the precedence of knowledge is a less relevant factor in scientific output in the field, manuscripts that are sensitive to changes in the situation clearly benefit from fast publication.

I identify two unexpected benefits, at least in the light of the most recurrent ideas in that first debate. The acceptance of preprints in the journals of the SciELO collection and the movement to reconsider the role of anonymity in the manuscripts assessment – as part of the open science movement – have led to experimentation with different possibilities and forms of disclosing review reports. Instead of a single model, we have seen editors-in-chief innovating by combining different openness possibilities and measures. The diversity in this experimentation is valuable and will allow us to make better decisions.

A second unexpected benefit concerns the preservation of the memory of the science production in the country for research purposes. Fundamental debates that took place between 1977 and 1980 at the first ANPOCS meetings were only partially accessible through the publication of the “Social Sciences Today” collection, but some of it has been lost. Changes and updates in the technology for storing and making available the proceedings of scientific catalog meetings, usually under the custody of scientific and professional associations, often produce similar effects of memory loss. Initiatives such as that of ANPOCS and SciELO, which has organized the submission of papers from the Meeting as preprints, can not only broaden access to these papers, but also make this access permanent for those who will be researching the evolution of our disciplines in the future.

Notes

1. This text summarizes the presentation made at the ANPOCS Periodicals Forum on 25/10/2023: “Preprint in debate”, organized by Rúrion Melo (USP) | ANPOCS. Also speaking at the forum were Alex Mendonça (SciELO), Adrian Gurza Lavalle (USP|Br. P. Sc. Review) and Isabel Rocha de Siquieira (PUC-Rio|RBCS)

2. I quote this phrase by heart and, given the vagaries of memory, I do not use quotation marks…

External links

47º Encontro Anual da ANPOCS | SciELO Preprints: https://preprints.scielo.org/index.php/scielo/preprints/section/47-anpocs

PREreview: https://prereview.org/

SciELO Preprints: https://preprints.scielo.org/

About Adrian Gurza Lavalle

Professor of the Department of Political Science at the University of São Paulo (USP), President of the Centro Brasileiro de Análise e Planejamento (Cebrap), Vice-Director of the Centro de Pesquisa, Inovação e Difusão de Estudos da Metrópole (CEM) and researcher at Cebrap and CEM. He is editor-in-chief of the journal Brazilian Political Science Review.