Things that happen on a weekly basis:
- Trash pickup.
- Church (if you’re into that).
- Saturday morning brunch with Susan. (I’m sorry I missed last week, Susan. I’ll do better.)
- Somebody discovers a new dinosaur.
Yes, dinosaurs. Michael Benton, professor of vertebrate paleontology at the University of Bristol in England, says a new dinosaur species is named about once a week, on average. Another paleontologist, Thomas R. Holtz Jr. of the University of Maryland, keeps a running tally of new dinos each year for an encyclopedia he is in the process of updating. He’s up to 14 this year. In 2015, he hit 45.
REASON | TOTAL INVALIDATED SPECIES THROUGH 2004 |
---|---|
Scant fossil evidence | 340 |
Already had a name | 230 |
Not actually a dinosaur | 58 |
No evidence at all | 47 |
But while that rate of discovery might delight 8-year-olds, it’s not necessarily an accurate reflection of the ancient world. Eight years ago, Benton published two papers on the error rate in dinosaur species identification and found that 48.2 percent of “new dinosaurs” are eventually cast aside, deemed invalid for a variety of reasons (see the chart below). That’s far above the rate for living species, which is only 20 percent.
Dinosaurs put the mega in charismatic megafauna. But they aren’t fairy-tale creatures that we can relish with all skepticism suspended. They’re real animals studied by real people. And those people make real mistakes.
Benton told me the 48 percent error rate applies to dinosaurs named between 1850 and 1980. That’s because there’s been enough time since then for other researchers to cock an eyebrow, poke at the data and declare it wrong. Everything that’s been found in the last 36 years, in contrast, is more of a mystery. We think it has a lower error rate than the older research, but we don’t know for certain because there hasn’t been enough time to thoroughly vet it. Meanwhile, in that same period, the number of dinosaurs being named has exploded like an asteroid in the Yucatan.
You can see the impact in this chart, which shows the number of named genuses1 compared with the number of genuses still considered valid.

If a species is like your first name, genus is like your last name — there are fewer of them in the world, and they describe a bigger, extended clan. (In an organism’s name, though, genus comes first and species second.) Fewer dinosaur genuses are rejected than species, but the errors are important, because many dinosaurs, unlike a lot of other animals, are known only by their genus. Think Prince. Or Rihanna. Or Tyrannosaurus2. The differences among genuses are typically bigger than those among species and they’re debated less often, which is why you see a higher error rate at the species level. But this all gets messy with dinosaurs.
Consider our friend Tyrannosaurus. It’s a genus. It’s also used as shorthand for the genus’s only species, Tyrannosaurus rex. And there are a bunch of related genuses, including the horse-sized Timurlengia euotica, whose discovery was just announced on March 14. There’s a lot of debate among scientists about whether some of these genuses should be merged, discarded, or downgraded to the status of species, under the umbrella of a different genus.
In particular, Benton said, scientists have hotly contested the existence of Nanotyrannus lancensis. This tiny relative of the mighty T. rex was discovered in 1942 in Montana and has been designated a distinct species since 1988. But many scientists disagree with that designation. Increasingly, Benton told me, researchers think N. lancensis was just a baby T. rex. At some point in the future, he said, you’ll likely see the name Nanotyrannus lancensis thrown out, its museum labels replaced with the more familiar Tyrannosaurus rex.
The other reason Tyrannosaurs matter? They are representative of a massive imbalance in the dinosaur world. Tyrannosaurs are theropods, a suborder that encompasses all the chicken-esque dinosaurs that stand on two legs and have stubby, comically ineffectual little forearms. (If genus is an organism’s last name, think of theropoda, the T. rex’s suborder, as its nationality.) Theropods are meat eaters. In living, healthy ecosystems, there are way fewer meat eaters than plant eaters. But that’s not the case in the dinosaur fossil ecosystem. “There are huge numbers of them,” Benton said. “There should be one-tenth the number of carnivores to herbivores, but half the named dinosaurs are theropods.”

Facts like this make paleontology seem hopelessly flawed. But there are good reasons to think that we’re getting better at naming dinosaurs, not worse, Benton said. Compared with 50 years ago, dinosaur names are now based on larger quantities of fossil evidence, and that evidence is evaluated in far more detailed, scientific ways. The theropod-herbivore imbalance suggests there is still something deeply wrong, but it’s not unfixable.
The trouble is, problems take a long time to correct. Scientists can go around looking as if they’re doing a great job of naming dinosaurs only to have their reputations tarnished years later, even after their deaths. Take a look at this chart showing the most successful dinosaur namers in history, with each paleontologist’s Paleoscore, a new dinostat my colleague Neil Paine created that adjusts success rate for the era in which fossils were discovered.3 The higher the better.

You’ll notice that the people with the highest rates of success tend to have done their work most recently. The further you move into the past, the more likely a prolific paleontologist is to have had many of his species invalidated. That’s not a coincidence, Benton told me. Xu Xing, for instance, has a perfect track record … for fossils discovered between 1999 and 2006. Looking at the historical precedent, it’s extremely unlikely that Xu can maintain both his fierce pace of naming and his blemish-free success rate. “The people who name most are least successful. Boom-boom! What more do you need to know?” Benton said.
What’s that mean for amateur dino fans? It’s crucial context. New dinosaurs aren’t a rarity, and when they happen they may not last. It’s easy to hype a new dinosaur. It’s harder to prove that dinosaur actually existed.
Neil Paine contributed statistical analysis to this article.