How Smart Could Dinosaurs Get?

Abstract

This commentary discusses the main points made in the Reiner (2023) paper on the prospect that some theropod dinosaur could have given rise to a lineage that achieved a human level of intelligence, and those made in the Herculano-Houzel (2023) paper on the potentially monkey-like numbers of neurons in the pallium of large theropods, and the implications of this for their intelligence.

Like many scientists working in the area of biology, my interest began with my childhood fascination with dinosaurs. They were big, they were unlike any living animals, and they must thus have been awesome to see. Certainly, they were awesome to imagine. Moreover, they were around for a really, really long time, but then suddenly they were gone for good many millions of years ago. All of them. Why? Were they so dull-witted that those then-tiny but crafty mammals were able to evict them from dominance? Well, we now know this explanation for their extinction was wrong on two counts. First, we know that the time of dinosaurs, and that of many other large vertebrates, came to an end due to a catastrophic meteor strike. Secondly, dinosaurs were not doomed by brain power that was too insufficient (compared to mammals) to be able continue their reign. Rather, as emphasized in my own paper (Reiner, 2023), numerous published studies have presented various lines of evidence that many dinosaur species functioned at least at an avian level of behavioral and cognitive complexity. In her own paper, Dr. Herculano-Houzel takes this a step further and from what we know about the relationship of brain size and neuron numbers suggests that larger theropod dinosaurs had an abundance of pallial (“cortical”) neurons comparable to that in some non-ape primates such as baboons (Herculano-Houzel, 2023a). If total pallial neuron number matters for cognitive power, and it certainly should, this at the very least further emphasizes that many dinosaur types were not dull-witted.

With the extinction of the dinosaurs, the small, mobile and/or burrowing vertebrates who were able to survive evolved to occupy the now-vacant niches and repopulate the earth. Important among these were mammals, leading ultimately to the evolution of humans in the primate lineage. What would have happened, however, without the meteor strike? In my own paper (Reiner, 2023), I imply there was nothing inevitable about the decline of dinosaurs and the concomitant rise of mammals from the viewpoint of brain capacity. Although their pallial regions were extremely likely to have different designs – nuclear in dinosaurs like in extant birds versus cortical in mammals – the nuclear design in and of itself works as well as the cortical. For example, Alex the African gray parrot was capable of cognitive feats well beyond those of, say, a rat or opossum. So, the dinosaur reign could have continued indefinitely without the mass extinction event caused by the cataclysmic meteor strike at what is now the Yucatán Peninsula in Mexico. Given this additional time on earth, was it inevitable then that some dinosaur lineage could have given rise to a species with a human level of intelligence? Certainly, innumerable works of science fiction, as noted in my paper, presume so. But science fiction is called that for a reason. Could dinosaur humanoids have evolved? In my paper, I lay out lines of thought that lead me to conclude that this would have been unlikely (Reiner, 2023). The reasons deal with a potential wiring inefficiency inherent to the nuclear pallial plan that would have been exacerbated with the pallial expansion needed to reach a pallial complexity and neuron number equivalent to that in humans. Dr. Finger in his commentary on my paper and that of Dr. Herculano-Houzel has nicely put the inefficiency in terms of computer design (Finger, 2023). In her own commentary, Dr. Herculano-Houzel has also raised the possibility of bioenergetic limitations in expanding the avian nuclear design to a human number of pallial neurons and cortical areas (Herculano-Houzel, 2023b).

My proposal, like that of Dr. Herculano-Houzel (Herculano-Houzel, 2023a), is based on assumptions, and can be evaluated based on the merits of those assumptions. As noted, chief among these is that the avian nuclear pallial design has a latent wiring inefficiency that is exposed when pallial size grows beyond a certain point. The presumed inefficiency, however, is based on the assumption that the neuron constituents of a cortical column function best when close together, as seen in the mammalian cortical design, and that this proximity is critical to the operation of the thinking part of the brain – neocortex in mammals and what is called pallium in birds and reptiles. Pull them increasingly far apart, as would happen with a nuclear pallial design expanded to human size, and the inefficiency should grow to self-defeating size. Is this, however, the case? Is there something special about the neuronal proximities in a cortical column as seen in the mammalian design? Perhaps there isn’t. Or perhaps there are some unique solutions that birds, for example, have achieved to overcome the supposed inefficiency - solutions a dinosaur humanoid might have been able to achieve as well. As part of addressing this inefficiency, it would be useful to determine the scaling rules for the distances between the neuronal equivalents of a cortical column in a nuclear design, by collecting the needed data in living birds, to determine the magnitude of the supposed inefficiency.

Even if I am right and a dinosaur humanoid was unlikely to have evolved, we are still left with the suggestion of Dr. Herculano-Houzel in her paper that large theropod dinosaurs could have possessed a monkey-like number of pallial neurons, suggesting an impressive intelligence level for these large theropods (Herculano-Houzel, 2023a). But how intelligent? To address this, a few additional consideration need to be taken into account. First, does pallial neuron density decline with increasing body size? There is evidence that it does (Tower, 1954; Roth and Dicke, 2005). If this is the case, one needs to take this into account in extrapolating pallial neuron abundance in large species from that in small species. Secondly, the late Harry Jerison noted that large species have larger brains because of the need for enough neurons to control the sensory and motor functions of the large body. For this reason, he developed the concept of encephalization quotient to extract an intelligence index from brain size – how much bigger a brain is than needed for a body that size, on average (Jerison, 1973). A similar consideration needs to be applied in the case of pallial neuron abundance in large theropods. It thus may be that cortical/pallial neuron abundance, per se, is not the determinant of intelligence. Consistent with this, several recent studies have suggested that some large whale species have more cortical neurons than humans (Mortenson et al., 2014; Ridgway et al., 2019), and certainly none of them are contributing papers to the Journal of Comparative Neurology.

Note that I hope that I am not being perceived as dogmatic in my proposal that a dinosaur humanoid could not have evolved. I cannot know for sure, but have tried to build a case in my article for why this is a reasonable conclusion. Moreover, I certainly find it intriguing to think that there is only one way to build a creature with a human level of intelligence, and in our case, as evolution would have it, it is all due to a little accident in pallial design that appeared hundreds of millions of years ago. On a final note though, I must emphasize that I am perfectly happy if experimental scrutiny of my premises proves them faulty and my idea thus without proper basis. Under these circumstances, a better understanding of the functioning of the cortical versus nuclear pallial designs will have emerged, and we will have a better understanding of how the brain works in different species. Plus, I and many other science fiction fans will have the contentment of knowing that the Gorn commander in the original Star Trek series was possible after all.

Data Availability Statement:

This is a commentary and any data mentioned has already been published by other authors, and the cited papers are publicly available.