“Study the Past if You Would Define the Future.”—Confucius

In a previous Function Editorial,¹ Editor-In-Chief Ole Petersen wrote an erudite piece lamenting the loss of intellectual rigor, gauged through a growing tendency to reclaim old discoveries made by others. He attributed this to a lack of appreciation of the history of a field and a penchant for citing earlier reviews in lieu of any historical perspective or understanding, perhaps in the hope that a perfunctory read of a cited review might have missed a chronological dissection of the field. Petersen gives the example of the discovery of calcium-activated Cl^- channels, first reported by the late Sir Michael Berridge in 1975, but widely attributed in more recent reviews to publications from others almost a decade later.¹ The Berridge example is not an isolated one but could perhaps be rationalized by the inconvenience of accessing old papers not stored on digital archives, a postmodern scepticism of discoveries made in the past using less sophisticated techniques and without molecular biological interventions, limitations to the numbers of references that are imposed by many journals, and the difficulty in tracing original discoveries when these are not alluded to in high-profile reviews.

Whilst one can try to provide some sort of justification for not citing key discoveries from several years ago, more disconcerting is the growing habit of airbrushing recent discoveries perhaps to control the narrative over a field as the key to scientific legitimacy and leadership. A case close to home relates to our discovery that Ca²⁺ nanodomains near open store-operated Orai1 CRAC channels activate the transcription factor NFAT1, whereas Orai2 and Orai3 are considerably less effective. Orai1 has privileged communication with NFAT because it forms a reversible signalosome with the scaffolding protein AKAP79, which in turn binds calcineurin and NFAT. Calcineurin is stimulated with high fidelity by the local Ca²⁺ signal, resulting in NFAT dephosphorylation and activation. Despite this work being published between 2011 and 2015,^2–4 subsequent published work by some other groups from 2019 onward not only fail to cite our original discovery of an Orai1-AKAP79 interaction but presents this as their discovery, and then builds on this in subsequent papers, citing their own work. Recent conversations with colleagues in both Europe and the USA reveal that ours is not an isolated or rare occurrence, and there is the sense that rediscovering a discovery is on the up.

Why is this? Pressure to publish in high-profile journals, with the attendant increase in grant success, could certainly contribute. Journals all too readily reject papers on the basis of perceived novelty, and therefore giving credit to others on whose discoveries the paper is based could be a form of self-flagellation. With fields becoming more and more focused, editors, even in more specialized journals, understandably lack the expertise to identify striking omissions and instead rely largely on the expert reviewers. With the proliferation in journals, finding suitable reviewers is often an arduous, and in many cases an unsuccessful process. The vanishingly small time a reviewer is given to assess a manuscript, typically less than two weeks, means careful assessment of the references is simply not possible. And a further conflating factor is the expectation that heads of research groups provide mentoring to their trainees, which often includes co-reviewing manuscripts. Trainees generally produce rigorous and detailed reviews but tend to focus on the data and interpretation rather than on the prior art in the field, particularly as they often are somewhat unfamiliar with the history of a field. And of course, as in all human endeavors, a discovery might be overlooked because of the need to present the work as entirely novel for a promotion, to strengthen the chances of securing research funding, or simply because of a personal vendetta.

Recent studies further suggest gender and racial/ethnic bias also contribute to the likelihood of a researcher being cited. Research driven by men (which was measured by the gender of the first or last author) is generally cited more often by other male colleagues as well as by researchers who are new to, or less familiar with, the field⁵. Men also cite themselves more than women do.⁵ Furthermore, it is thought that an additional factor for citation bias comes from the so-called first mover advantage.⁶ In this view, for similar data/conclusions, men are more willing to press ahead and publish their work before women, giving them an advantage in receiving more citations, and being regarded as the first to make the discovery. However, some fields have a higher proportion of men than women, and this will inevitably lead to the difference in relative citations.

One recent study looked at the reference lists of papers published in five leading Neuroscience journals (Nature Neuroscience, Neuron, Journal of Neuroscience, Brain, and NeuroImage) over a 25-yr period (1995–2020).⁷ This amounted to 63,677 articles. The authors reported that there were more papers with a white first or last author than would be expected if race and ethnicity (Asian, Black, Hispanic) were unrelated to referencing. It would be interesting to analyze this further and see whether nonwhite scientists are less likely to be cited even when they are reporting research from European and US-based laboratories.

Although important, gender and race cannot wholly account for the rediscovery of old discoveries or the omission of key references. Many white male scientists find their contributions overlooked, unreferenced, and often ignored. This is therefore not an issue related to one group; it impacts all of us.

It would of course be unwise for journals or research fields to introduce some sort of quota-based system or affirmative action to ensure references cited reflect a diverse community. Papers differ in the extent of rigor. For two studies that reach the same conclusion, one could be detailed, address alternative hypotheses, and use direct and elegant experimental approaches, whereas the other happens to draw one particular conclusion when others are viable, is less well controlled, and uses indirect experimental methods. Clearly, only the former paper should be cited. And who will police the level of diversity within a field? By its very nature, a field is dynamic with new researchers coming in and others moving out as interests change. The extent of diversity will therefore fluctuate, and monitoring this will be a Sisyphean task. Journals often impose a limit on the number of references cited; it would be damaging to substitute high-quality papers for weaker ones simply to reflect diversity within the field. Finally, science flourishes when independent, free to pursue an intellectually interesting problem unencumbered by socio-political constraints. Mandating who should be cited would certainly undermine the independence and autonomy of science.

Can we learn from other disciplines? It is instructive to consider the process of reviewing in the Humanities, as opposed to the Social Sciences, which are closer to the Natural Sciences and often suffer from the same issues raised here. First, the Humanities have a keen sense of history and this is reflected in published papers. Several articles published in the highly regarded journal History of Political Thought, for example, include citations to original papers published in the 1960–1980s. The opening paragraphs place the paper in a historical context, and credit those whose works are being developed. Manuscripts can be rejected outright if key references are omitted or credit is falsely assumed by the authors. Second, there is generally no immediate social relevance in the Humanities as there is in the Sciences. Therefore, there is no pressure for immediate publication. Reviewers are given appropriate time to read the manuscript carefully-typically 4–6 weeks rather than the 10–14  days imposed by many scientific journals. Whilst scientific discoveries can be of immediate societal benefit, as seen during the recent pandemic, this is more the exception than the rule. Whether a paper is published immediately or a few weeks later is generally of little consequence. A faster review process should not compromise rigor. We are currently in a period of human history where experts are now often disregarded and many sections of society absorb information, often of dubious veracity, from social media. It is vital that Science maintains high-quality peer review. Taking a few weeks longer to achieve this is a small price to pay for the potential disaster of high-profile flawed research. Third, research in the Humanities is less factual and based more on a line of argument. Marx’s theory of revolution and Kant’s formulation of the categorical imperatives are not derived from facts or empirical data but are abstractions reasoned from a central belief. This has an important consequence; reviewers are perforce required to provide a substantial critique of the manuscript, including an assessment of the author’s founding premises. Reviews are therefore detailed and comprehensive; and it would be unheard for a reviewer to have the temerity to submit just a few sentences with general comments on the paper. Finally, the Humanities encourage long footnotes where many issues can be mentioned and discussed, something not common in the natural sciences.

In all human endeavors, there is an element of subjectivity. One might argue that a rediscovery is in fact THE discovery because the original discovery was not, in the eyes of the rediscoverers, convincing. But if the rediscovery confirms what was found earlier, then the original discovery should stand. No scientific paper is complete of its own; there are always additional experiments that could be done to confirm further the findings, but this can become an endless process. One of the innovative features of Function, introduced by Ole Petersen, is that Editors make a decision on the paper without asking for numerous additional experiments. Are the data sufficiently strong to support the main conclusions? If yes, then the paper is acceptable pending minor changes. Substantial and time-consuming revisions, which often add little to the overall conclusions, are not necessary for Function.

Journals often impose limitations on the number of citations an original research article can include. But this should not compromise scientific accuracy nor provide a means for not citing key papers on which the study is based. Reviewers need to police this rigorously. It is rare these days to see reviews pointing out omissions in references. It has almost become a secondary issue. With limitations on the number of references permitted in a research article, authors are understandably citing review articles in lieu of original papers. Therefore, the accuracy of the review, in attributing key discoveries to the appropriate scientists and in offering the historic narrative of the field, becomes of critical importance. The onus is on those who write review articles, on those who review the article for a journal, on the Editor, and on the research community as a whole to ensure the field upholds the highest academic standards and ensures discoveries are indeed discoveries. As Mahatma Gandhi said: “The future depends on what you do today.”

Conflict of Interest statement

A.B.P. holds the position of Executive Editor for Function and is blinded from reviewing or making decisions for the manuscript. The authors declare they have no conflict of interest.