The Top 10 Science Papers of 2018 Are an Apocalyptic Nightmare

This used to be more fun. What happened?

Drunk person with spilled liquor bottle, destruction after hurricane, bleached coral, injured sea turtle
Photo illustration by Slate. Photos by Matt Cardy/Getty Images, Ricardo Arduengo/AFP/Getty Images, Prisma Bildagentur/UIG via Getty Images, Equinac/Barcroft Images/Getty Images.

It’s Top 10 season at Slate, with all the usual enumerations. We’ve got lists of the year’s 10 best books, 10 best jazz albums, 10 best TV shows, 10 best movies, 10 best albums that might or might not be jazz, and so forth. Today, I’d like to call attention to a different sort of list, a profoundly disturbing and depressing one. I mean, of course, it’s time to look at the top 10 scientific papers of 2018.

The leaderboard in question comes from the data-crunching firm Altmetric, which last week posted its annual tally of the year’s 100 most attention-grabbing studies—the ones that caught the notice of policymakers, the public, and the press. If one chooses to view the top of the list (as I have done) as a monumental portrait of the year in science, if not the grand résumé of its most important findings, then I’m afraid that one may end up (as I have done) in deep despair. The research output of the past 12 months, as represented here, has been nothing less than apocalyptic.

Start with No. 10 on the list, an epic study from the Proceedings of the National Academy of Sciences that came out in May, entitled “The Biomass Distribution on Earth.” While other studies have addressed the role of humans in decreasing global biodiversity, this one was the first to estimate the sheer, living tonnage that we’ve wiped out. Here’s the headline finding, per the Guardian: “Humans Just 0.01% of All Life but Have Destroyed 83% of Wild Mammals.” Oof.

Next up the list, at No. 9, is a Nature paper that came out in April, on the “unprecedented,” climate-driven devastation of the Great Barrier Reef: One-third of its corals perished in the heat over just a nine-month stretch. At No. 8 we have a paper on the dangers posed by complementary medicine: Among cancer patients, the use of herbs and acupuncture is associated with a doubled risk of death. Next up at No. 7, a study of the “rapidly accumulating plastic” in the Great Pacific Garbage Patch, which now appears to be many times bigger than we thought.

Papers No. 6 and No. 5 provide a brief respite, with less alarming research into healthy diets and the benefits of exercise. But then we’re back into what looks like the table of contents for a special issue of the Journal of Armageddon Studies. The fourth-buzziest science paper of 2018 was a warning that we may soon find ourselves in an irreversible state of climate breakdown—the so-called Hothouse Earth scenario (headline: “Runaway Global Warming Could Be Just Decades Way, Say Scientists”). The third-buzziest: an epidemiological report on alcohol consumption, which makes the controversial claim that the only safe amount of drinking is not to drink at all. At No. 2, a Science paper on our post-factual predicament, featuring data-driven proof that lies spread much faster than the truth on Twitter. And finally, at No. 1, the Top Science Paper of the Year: the summer’s somber follow-up report on Puerto Rico’s devastation at the hands of Hurricane Maria, which says, despite the official double-digit death count, that the storm likely claimed closer to 5,000 lives.

This grim top-10 tableau of death and destruction is all the more striking in the context of its predecessor lists. Just four years ago, the Altmetric rankings told a very different story, about a very different sort of world. Yes, there was a whiff of worry in the findings listed then. One entry, for example, showed that artificial sweeteners, like those in diet soda, mess with microbiomes and exacerbate obesity. (N.B. that study didn’t really make much sense.) Another from that year used creepy methods to conclude that good and bad emotions spread on Facebook.

But the other specimens on the 2014 list seem embedded in the amber of a prior, placid age of human history, when the end of the world didn’t feel so close at hand. There were feats of scientific progress—a breakthrough in producing stem cells (later shown to be a fraud, but whatever), and a step toward machine-assisted telepathy. The rest were simply weird, wonderful, or absurd: There was a study showing that dogs can sense the Earth’s magnetic field (and poop accordingly); one that measured how long it took for chocolates to be eaten, when left out in hospital waiting rooms; one on the drinking habits of James Bond.

This rank silliness and optimism has since faded from the list. To look back now at the top-ranked science paper from 2015, on a novel antibiotic that can even kill off drug-resistant superbugs like MRSA, yields the sense of watching scratchy newsreel clips in black and white. (It’s hard to get your 2018 head around that degree of positivity.) Or else that year’s No. 2: a major study that debunked the link between MMR vaccines and autism. (Update from the future: The bunk abides.) The 2016 list includes a pair of major astrophysical discoveries, awesome in a way I only half-remember now, just two years later.

Yet that 2016 list also showed a dawning sense of dread. There were papers on the spread of light pollution, a growing inequality of life expectancy, the effects of Zika virus, and the shocking, deadly prevalence of medical error.

Last year, in the 2017 list, we once again had some hopeful breakthroughs mixed amid a stew of signs that things are getting worse. There were papers on the progress toward an artificial womb, a vaccine for Ebola, and the use of CRISPR in addressing hereditary illness. Another analyzed a feathered dinosaur tail discovered in Myanmar. But these sat side by side with studies of the tenfold rise in global rates of obesity, the 75 percent decline in total flying insect biomass in protected areas since 1989, and the fact that even little girls act according to their learned belief that little boys are smarter.

This year, the glooming is complete. The top 10 science papers of 2018 describe a world in flames.

To try to illustrate this trend, I’ve categorized each of the top 10 papers from the past five years as being “hopeful,” “worrisome,” or neither. Some entries were a little tricky to categorize. Should a paper on the ineffectiveness of wearable technologies for weight loss count as hopeful news, since it means we’ve figured out a bogus treatment doesn’t work? Or is it worrisome that these technologies were so widespread? (I chose the latter.) For others, the classifications seemed self-evident. A paper called “Entering the Sixth Extinction”? Worrisome. One called “Correction of a Pathogenic Gene Mutation in Human Embryos”? Hopeful. Then I counted up the number of papers in each category and plotted out the difference between the “hopeful” and “worrisome” totals for each year:

Graph of worrisome index of scientific papers of the year
Daniel Engber

To be clear about the methods: The Altmetric “Attention Scores” on which these lists are based don’t aim to show the sum total of news coverage around any given paper or provide a simple tally of its reach on social media. They aren’t Q Scores for science, nor are the papers cited likely to match up, one-to-one, with those mentioned in 2018’s “most read” science stories from, say, the New York Times or USA Today. Rather they reflect Altmetric’s own analysis, derived from a proprietary potpourri of sources including policy documents, media outlets, Facebook, Twitter, and various “scholarly and non-scholarly forums.”

Altmetric’s scoring product is part of an important, broader movement within scientific scholarship (which itself is called “altmetrics”) to improve upon the classic methods of assessing any given research paper’s influence. Traditionally, this has been done by counting up all the other research papers that end up citing it, or by keeping track of the number of times its PDF gets downloaded, or perhaps by checking how often it’s been recommended by an expert panel in the field. These measures may be useful, but they tend to lag behind the pace of modern scholarship, and they don’t reflect a fuller range of online interaction, from data sharing to post-publication peer review. The rankings from Altmetric represent one of many efforts to produce a better way to filter through the research literature, a worthwhile endeavor.

Given the above, it should come as no surprise that Altmetric’s yearly science-paper rankings are tilted toward a hardcore research audience. You’ll find some papers on the list that seem of interest only to a narrow set of scientists in academia—and maybe no one else in the world. Thus, the eighth-ranked paper from 2015: “An Efficiency Comparison of Document Preparation Systems Used in Academic Research and Development.” Or the No. 2 from 2017: “Work Organization and Mental Health Problems in PhD Students.”

If these rankings weren’t quite so specialized—if they were based on Facebook shares across the board, for instance—then the change in focus that we’ve seen might be just another index of a broader market shift in viral headlines. Five years ago, we were riding out a boom in wonder-sucker clickbait (e.g., “You won’t believe what they discovered next”). Now we’ve landed in a different stage of social headlines, where feel-good teasers have been swapped for ones that make you scared or angry.

But that’s not all that’s going on in top 10 science, I suspect. This list doesn’t measure social reach; it tries to get at something closer to the core concerns of researchers. And its rapid evolution, from the OMG-LOL-WTF experiments of 2014 to today’s crib sheet for the Book of Revelation, suggests that this spotlight may have shifted, darkening the awe in favor of anxiety. Worries can be purposeful, of course, and it may well be more important that we talk about the latest mass extinction than the fact that dogs prefer to poop with their bodies lined up north to south. But that doesn’t mean it’s not a bummer. The larger takeaway, this year, might be that soft regret: Wonder has a value, too; I really hope it makes a comeback.