Citation corruption

Published May 4 2016

If I understand things right, journals get scored for citation impact via a metric that considers only two years subsequent to the publication of their papers. Papers that get cited after this two-year window do not enhance the impact factor of the journal in which they were published. For journals, this is big business — they live and die by citation impact.

For scientists, this used to be an afterthought. But boy do Bob Dylan’s famous words ring true in this instance: “The times, they are a changin’.” Whether we know it or not, scientists are judged by the company we keep. If we publish our papers in journals that have high impact factors, the benefit rubs off in many ways. If we publish in journals that have modest citation impact numbers, we suffer.

I see certain flaws in the use of numerical scores to rank the value of a scientist. Some of these flaws have been articulated by other critics of citation impact. Do a simple Google search, and Wikipedia will give you plenty of input regarding the pros and cons of citation impact. Despite obvious flaws, I do not dispute the general thesis that there is a correlative relationship between the value of a paper and the number of times it eventually is cited.

I do choose to critique the two-year time window.

Take the case of a paper published in the Proceedings of the National Academy of Sciences back in 1971. The report was authored by Ronald Konopka and Seymour Benzer and described the results of a forward genetic screen in search of genes that might tell us something about the circadian rhythm of fruit flies. Using ingenious phenotypic assays that allowed timing of both pupal eclosion and locomotor activity, Konopka and Benzer found mutations that lengthened, shortened or eliminated the 24-hour circadian clock of flies. Remarkably, all three categories of mutations mapped to the exact same gene — dubbed by the authors the “Period” gene.

The accompanying figure shows the citation history, according to the Web of Science, of the landmark paper entitled “Clock mutants of Drosophila melanogaster.” It was cited once in 1971, once in 1972 and only 24 times in the decade after its publication. Things sped up in the next decade, during which the paper was cited about seven times per year. Over the past two decades, since Michael Rosbash, Jeffrey Hall and Michael Young cloned the Period gene and it became clear that the Konopka and Benzer discovery was of watershed significance, the paper has been cited roughly 40 to 50 times per year.

The citation history, according to the Web of Science, of the landmark paper “Clock mutants of Drosophila melanogaster”

Few would doubt that the paper by Konopka and Benzer describes one of the most significant discoveries ever achieved in the field of circadian biology. By contrast, citation metrics would have given reward neither to PNAS nor to the authors. Publication of this paper instead would have hurt both the journal and the authors. In the two-year interval subsequent to publication of this paper, it was cited only two times. It is hard to get a worse score than that!

The consequences of the insidious infection of citation impact, at least in its current form, are huge. Both journals and scientists want papers to have immediate impact. After the two-year window, forget it — the stuff is yesterday’s news. The corruptive influences of this flawed system are obvious: We are forced to work on what is faddish or trendy.

If anything, I think a paper should be revered if blessed with the unusual feature of having been ignored in the immediate post-publication window but having gained recognition later. The journal that publishes these sorts of papers should be rewarded, not penalized. Likewise, we all know that Konopka and Benzer were sage scientists. They set the trend decades ahead of others. Is this not the testament of greatness?

In its present form, this measure of citation impact conspires heavily in favor of what is trendy and faddish. Journals will die if they publish Konopka/Benzer-like science. Worse yet, scientists — if scored by these measures — will wither professionally if they fail to follow what is in fashion.

It is hard to think of anything worse for our profession than this insidious form of citation corruption. We should desire and create a system that encourages scientists to take risks and work on new horizons. We should reward and encourage Konopka/Benzer-like creativity. What a pathetic mistake of unintended consequences we have created in allowing science to be led around by the lunacy of citation corruption.

Steven McKnight Steven McKnight is president of the American Society for Biochemistry and Molecular Biology and chairman of the biochemistry department at the University of Texas-Southwestern Medical Center at Dallas.