Herculano-Houzel is no passive dinosaur observer. The Vanderbilt University neuroscientist is an expert in comparative neuroanatomy, as well as editor-in-chief of The Journal of Comparative Neurology. In 2023 that journal published a paper she authored about the intelligence of theropods, a clade of dinosaurs which includes Allosauruses, Spinosauruses, Giganotosauruses and of course the iconic Tyrannosaurus.
By analyzing data from T. rex remains as well as data from the theropod's closest living bird relatives, including emus and ostriches, Herculano-Houzel concluded that theropod brains had in excess of three billion neurons. That would mean they had more neurons than baboons, with perhaps even enough intelligence to use tools — the T. rex's notoriously tiny arms notwithstanding.
Yet other scientists quickly pounced on Herculano-Houzel's paper, with a recent study in the journal The Anatomical Record criticizing the argument's supposed "several crucial shortcomings regarding analysis and interpretation." Perhaps most importantly, the group of paleontologists, biologists, geologists and other scientists disagree with Herculano-Houzel comparing theropods to birds in her analysis, saying that she instead should have used lizards as her basis. The introduction refers to "the consensus of crocodile-like cognition in these animals, a position informed by comparative anatomical data." From there, the paper made other assumptions that the scientists felt needed to be publicly challenged.
"The methods used in the original paper are characterized by several shortcomings such as assuming the brains of many dinosaurs were very densely packed with neurons because they were warm-blooded," says one of the new paper's co-authors Hady George, a PhD student of Earth Sciences at the University of Bristol. "This led to over-exaggerated neuron number estimates, and this was then argued to be evidence for many dinosaurs, including T. rex, to be capable of complex behaviors, such as tool use."
The authors also contest Herculano-Houzel's assumption that one can predict different dinosaurs' metabolisms, aging rates and other life history traits based on the existing data.
"We are not inherently against their conclusions, as it is fascinating to think about dinosaurs in this new light, but the shortcomings of the methods used and the lack of consideration for other lines of evidence that more accurately predict metabolism and life history traits in fossil animals render their conclusions highly questionable," George says.
Herculano-Houzel is defiant against such criticism, taking particular umbrage by a statement by paper co-author and University of Southampton paleozoologist Dr. Darren Naish that T. rex were "more like smart, giant crocodiles, and that’s just as fascinating.” From Herculano-Houzel's perspective, it is both inaccurate and unfair to lump in the T. rex with its distant crocodilian relatives, regardless of the qualification that they would have been "smart" crocodiles.
"I absolutely stand by my original findings, and I want to do right by dinosaurs," says Herculano-Houzel. "They do not deserve to go back to being considered as dumb as crocodiles just because a group of paleontologists used their credentials (which I don't have; I am a neuroscientist) to back up an erroneous apples-and-oranges comparison due to a beginner's mistake in their analysis."
Herculano-Houzel also said that "this is a smear job on T. rex," one she claims the authors committed because "they had an opinion from the get-go: that T. rex had crocodile-like cognition."
Simply put, Herculano-Houzel says previous paleontologists take it for granted that "an animal that large could not have been smart, even with a respectably monkey-sized brain."
"I have already shown that body size is irrelevant, but never mind that," said Herculano-Houzel, observing that her original study "followed the data" and was correct in comparing T. rex to close bird relatives like ostriches, chickens and ducks. Although her critics say this is like comparing apples to oranges, Herculano-Houzel argues that her results prove "they were all applies."
"Think of T. rex as a scaled-up ostrich in body and brain size as well as number of neurons," Herculano-Houzel said. She argues her critics inaccurately divide all birds into two groups, thereby mixing theropods' closest cousins with more distant relatives like pelicans, egrets, albatrosses and penguins. Using a twist on the "apples and oranges" expression, Herculano-Houzel explained why she believes their reasoning is flawed.
"It's like mixing apples and oranges in the juicer then complaining that the drink doesn't taste like an apple, so it couldn't have been an apple going in," says Herculano-Houzel. "Given their decision to mix theropod dinosaurs with pelicans, albatrosses and penguins, the result is that theropod dinosaurs then appear to not have had bird-like brains already. But go to their graph knowing what the species are, and you will see T. rex exactly along the line together with ostriches and chickens and ducks — exactly like I showed. Apples."
George did not characterize the scientists as entirely opposed to the hypothesis that T. rex could have been intelligent. He instead argued they provided evidence "that dinosaurs did not have neuron estimates as high as those estimated in the original paper."
In addition to what George described as their "convincing argument" that the neurons in dinosaur brains were not as densely packed as those in modern bird brains, the scientists also argued "various other lines of evidence like gross anatomy and trace fossils can collectively better inform on dinosaur metabolism and life history traits than neurological variables such as relative brain size alone." The authors did not definitively conclude one way or the other on the question of dinosaur intelligence, aside from arguing that it is "a very complicated topic that cannot be understood using only neuron number estimates."
While that last observation may seem counterintuitive, George said that there are some bird species with large neuron counts that do not have the capacity to use complex tools, even though other bird species with comparable neuron counts do have that ability. Similarly, although some dolphin species have billions more neurons than humans, George said there is no evidence that they are more intelligent than humans.
"Overall, neuron counts do inform on intelligence, but are very limited in telling us what exactly animals are capable of," said George. He instead pointed to other kinds of evidence, such as the fact that many dinosaur species had elaborate head crests and feather fans. "Both of these adaptations have been argued to be socio-sexual display structures," said George. "Additionally, many dinosaurs have been found together, suggesting they might have once lived together. Many dinosaurs probably engaged in sophisticated communication with each other to live in groups and find mates, and this indicates at least some level of impressive intelligence."
As far as Herculano-Houzel is concerned, though, "the absolute number of neurons in the cortex (or telencephalon) certainly is by far the best known correlate of cognitive capacity." She is unmoved by a paper that Smithsonian Magazine declared "official refutes" her own.
"It's really a pity," says Herculano-Houzel. "Like I said, T. rex deserves better. My hope is that open-minded paleontologists who read my original paper have already restarted revisiting the data on T. rex-associated fossil record and reconsidering it with new eyes that are accepting of the possibility that maybe they did, yes, make and use tools, and maybe even had a culture."
