This is a therapsid, the ancient animal of which a hot, new paleontology thesis is making the rounds this week regarding the emergence of the much-vaunted brain module, the neocortex. The neocortex is unique to mammals among vertebrates, but it is difficult to imagine what a brain would be like without one. It comprises about 75% of the human brain, and gives rise to many of the functions that we think about when we think about brains: primary sensory regions for vision, touch and audition, language, planning, emotional processing, semantic memory, etc. It plays less of a primary role in the brain functions that we tend to forget about and ignore: controlling the heart rate, breathing, sleeping, swallowing. Since, the neocortex does so much of the “cool” stuff, there has been a lot of interest in the evolution of that part of the brain in particular.
What could have been…
As a big geek, I definitely see the intrinsic appeal in trying to learn about the history of animals on our planet in general, and particularly about our ancestors. I think that the news articles are also trying to appeal to that geekdom (even though mammals didn’t evolve from the line of therapsids that this paper is about). They were still close enough to our ancestors to be interesting, though, especially if they almost evolved a brain that kind of resembled early mammalian brains. However, before we get too excited about how much we can apparently learn from studying small pieces of bone from hundreds of millions of years ago, we should first reflect on how research like this could overstep its data.
The dangers of backwards inference and the fun guessing game of evolutionary neuroscience
The paper that is reviewed here actually stays away from too much speculation, and even includes nice, cautionary statements where it tries to make any inferences. For example:
“Because soft-tissue is not preserved in extinct vertebrates, inferring the presence of neocortex or a brain structure analogous to the neocortex is always hypothetical and can only be deduced from the shape and the volume of the brain endocast. Accordingly, there are three lines of evidence, which are commonly used as indication for the presence of neocortex: (1) the presence of two distinctive, clearly inflated cerebral hemispheres, (2) the presence of a rhinal fissure on the brain endocast, and (3) an increased EQ (e.g., Kielan-Jaworowska et al., 2004; Quiroga, 1980, 1984; Rodrigues et al., 2013; Rowe et al., 2011). The same criteria will be used here to discuss the question whether Kawingasaurus possessed a structure analogous to the mammalian neocortex or not.”
As is too often the case, this is a far cry from the popular press report, (here from Neuroscience News):
“Obviously, a neocortex-like structure at the forebrain similar to the mammalian neocortex was present in this animal.”
In my mind, the bigger problem in morphology research is speculation about function, rather than speculation about the morphology. First, there is the reverse inference problem, which is a problem both for making inferences about brains from 255 million years ago and a problem for making inferences about brains today. “Reverse inference” is when we take what we think we know about something, and then project that on to some other result. It’s easy to do this with morphology.
For example, I read that one of my heroes, the rock-climber Alex Honnold, who recently climbed El Capitan without any rope, had been assessed in an MRI.
The researchers found that Honnold had no amygdala activation to “high-arousal” photographs. The amygdala is known to be involved in threat response and fear. Here comes the reverse inference: because Honnold has no amygdala activation, he has no threat response and probably doesn’t experience fear when he is climbing. This contrasts with Honnold’s own description of his feelings. The reverse inference can lead to a decent hypothesis, and in some cases that hypothesis is testable. But the reverse inference itself is not a test of that hypothesis and can easily lead someone astray.
Unfortunately, in the case of long extinct animals, a hypothesis regarding function is much more difficult to test. We would need indirect evidence of the brain function to correlate with the morphology. Therefore, to make any claims about the function of the brain areas of a long extinct animal stands on thin ice.
This leads to the second pitfall of evolutionary neuroscience. I often read fun and imaginative speculation about what sorts of things could be responsible for what sorts of phenotypes in both ancient and extant animals. Actually, I find reading a lot of these ideas of possible pasts quite enjoyable, in the same way that I love reading about possible futures in science fiction. However, we should be careful to separate this imaginative speculation from the science itself, and we should always consider whether any of the speculation is testable.
In this case, we should take a moment to separate what is the science (that one type of therapsid had a bigger brain than many others) from our runaway imagination (that due to the reptilian giants patrolling the lands of Earth like ruthless soldiers, one of our closest cousins at the time needed to adapt quickly, gaining superb sensory abilities to detect the ravenous beasts by upgrading their cognitive hardware, giving them the ability to outsmart the brutes…at least for a while).
Image 1: By Dmitry Bogdanov - firstname.lastname@example.org, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=2872440
Image 2: By Niccolò Caranti - Own work, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=32502296