A research paper on drying coffee stains had unusual and unexpected applications. Credit: Getty

What makes good science? And how do Nature’seditors select papers to publish? The answers to both questions are many and varied. But they have one thing in common: neither is necessarily reflected in citations.

Nature publishes about 800 papers each year. Over, say, two years following publication, the pattern of citations typically ranges from a few papers with citations in the hundreds, to a large number with tens of citations, and a tail with single figures.

We are pleased when our papers make an impact. But there is much more to scientific impact than citations. For example, last week, in an ‘In Retrospect’ article in News & Views, Ronald Larson described a remarkable story (R. Larson Nature 550, 466–467; 2017). In 1997, Nature published a paper by Robert Deegan and his colleaguesthat provided an explanation of the ‘coffee-ring effect’ in spilt liquids, based on considerations of evaporation and surface interactions (R. D. Deegan et al. Nature 389, 827–829; 1997). For several years, the paper sat proudly in our pages, typically gathering about 20 citations per year. In 2006, as new implications and applications became clear, the rate picked up to well over 100 a year. So far, it has attracted about 4,000 citations. The paper is worth highlighting as an example of the varied types of judgement that Nature editors use to select papers. 

Our most highly cited papers are indeed often key landmarks in their fields. But there are papers that turn out to have low citations that we are equally happy to have published. The work by Deegan et al. was selected not because of any editorial ability to anticipate advances years in the future, but because, at the time, we considered it to be a noteworthy and pleasing piece of insight. Nothing more, nothing less. The developments celebrated by Larson are an editor’s unexpected bonus.

Most papers that we publish, with the invaluable help of our reviewers, are selected with a view to their scientific significance, whether as a powerful insight or an unusually empowering resource. And often that will correlate closely with citations (although citation patterns differ across disciplines). But it’s important also, for editors in all the disciplines from which we publish, sometimes to appreciate the interest in a paper using quite different criteria. It may be compelling for its sheer creativity or logical elegance, for making the reader stop and think very differently about a question, or for a stimulating and even mysterious observation. Many of these may be slow burners citation-wise — or simply be textbook examples that never get taken up in abundance. Here are other examples, drawn from the physical sciences, that, despite low citations, we like to celebrate. 

One such paper illustrated how images could be taken using X-rays radiated when sticky tape was peeled (C. G. Camara et al. Nature 455, 1089–1092; 2008). The citations are not huge by physics standards (165 since 2008) but we still love it, and we did not fully anticipate how it would go viral on social media. Another (11 citations) reported an actual sample of Cretaceous seawater from 145 million years ago. (W. E. Sanford et al. Nature 503, 252–256; 2013). And finally, a theoretical paper providing an exact textbook solution for the capacity of noisy quantum communication channels has been cited just six times since 2013 (G. Smith and J. A. Smolin Nature 504, 263–267; 2013), but we value it for what it is and anticipate that its take-up could increase as research moves from idealized ‘noise-free’ systems to more realistic noisy ones.

There are examples in other disciplines too. Why highlight such papers? Because we are glad to have published them. And because it’s perhaps salutary to appreciate just how unrelated scientific interest (at least, as we at Nature see it) and citation numbers can be.