Sizing up the soul’s seat

Abstract

As we are getting to understand the brain better and better on a microscopic scale, its macroscopic evolution — especially brain size — still poses some riddles. Florian Maderspacher inquires. As we are getting to understand the brain better and better on a microscopic scale, its macroscopic evolution — especially brain size — still poses some riddles. Florian Maderspacher inquires. “It is for certain the principal organ of our soul and the instrument with which it executes formidable things: the soul believes that it has penetrated everything outside of itself, such that there is no limit to its knowledge; however, when the soul enters its own house, it would not know how to describe it, and it does not know itself anymore. One only has to dissect the big mass of matter that makes up the brain to have reason to lament this ignorance. On the surface, you see an admirable diversity, but when have entered it, you cannot see a thing.” Thus wrote the Danish polymath and later cleric Nicolaus Steno in his Discours sur l’Anatomie du Cerveau in 1669. Steno inhabited an age that had only just acknowledged that the brain is really where it’s at — the seat of our perceptions, thoughts and feelings. This was in part because the heart, prime contender for that space since antiquity, had just recently been shown to be but a prosaic pumping engine. Now, there was a flurry of interest in the brain. Anatomists went on to describe the exterior shape and structure of the brain, most prominently Thomas Willis in England, who with Christopher Wren produced some of the most staggering early brain images. But the brain proved a tough nut to crack. It wasn’t like the heart, whose function could be inferred from looking at it carefully enough and with some clever experiments. Beyond the brain’s macroscopic anatomy lay an impenetrable gooey mass that did not in any way offer an glimpse into how it worked. No doubt Steno would be in awe were he to witness the astounding level of detail we can see in the brain today. We understand the matter that to him seemed impenetrable down to its subcellular structures. Just last year, in an epic effort that had spanned the better part of a decade, a tiny bit of mouse cortex (around 1,500 cubic micrometers) was reconstructed in three dimensions. This provided new biological insights, for instance into whether the sheer proximity of axons and dendrites might predict where synapses form. (It doesn’t.) But it also installed a sense of awe into just how complex brain tissue is. Here, a nutshell really contains an infinite space. Around 1,400 axons are crammed into the tiny space. Yet, the speck that was reconstructed corresponds to the entire mouse brain in the same way ten paving tiles correspond to the area of London. On that scale, the human brain, around 2,800 times larger, would be roughly the size of the EU. And a single pyramidal neuron may receive over of 30,000 inputs, a cerebellar Purkinje cell in the range of 200,000. The brain remains bewilderingly complex. But our understanding of how this most complex form of matter is organized will keep growing fast. Surprisingly, it is perhaps our understanding of some of the brain’s macroscopic aspects that has been lagging behind. Especially when the human brain is viewed in the context of other mammalian brains. Steno already realized that “the brain is different in different animal species, which is all the more reason to examine them all; the brain of birds and fishes is very different from that of a human; and in the animals that have brains closest to ours, I have not seen a single one where I could not find some significant difference.” This is an astute observation right at the foundation of comparative neuroanatomy, but just how significant is the difference really? The conundrum is this: if the brain is the site of cognition — which it is — how come the brain of humans — which are the most sapient species known to, well, humans — isn’t fully reflecting that special status? If a pickled human brain is placed among the brains of other mammals, it doesn’t jump out immediately. Not in a way that would adequately reflect our cognitive accomplishments. If an alien were to stroll through a museum of brains in vats and asked to pick the brain of the most intelligent species — the one that was capable of pickling all the other brains — which one would she pick? An intuitive response would probably be to go for the biggest brain. In our collection that would be that of the sperm whale, weighing in at roughly 8 kg, and for land animals that of the African elephant, at up to 6 kg still four times larger than the human brain. Elephants certainly do show some evidence of a rich cognitive life — they live in complex societies, they are able to remember other elephants and may even have some notion of death — but why so much brain hardware would be needed is not immediately evident at all. Naturally, brain size is a crude measure. Not all brain cells are concerned with cognitive tasks; half of all human brain cells are glia. And brain size is itself linked to body size. A large animal, like the elephant or the sperm whale, will on average also have a large brain. So if an animal lineage experiences selection for a large body, a large brain may just follow suit. Traditionally, scientists have thus resorted to relative brain size as a measure; specifically, the so-called encephalization quotient, which is the actual brain mass divided by a hypothetical brain mass that would be expected based on the brain–body scaling relationship of animals in the respective taxonomic group. Here, humans score a record high, with a value around seven when compared with other mammals, and around three when compared with other primates. This may in no small part have contributed to the popularity of this measure. Seemingly, the encephalization quotient does reflect the special status of the human brain and has increased continuously over the course of human evolution. But there are also some abnormalities where a high encephalization quotient does not seem to be reflecting high intelligence — great apes, for instance, score lower on the encephalization quotient than some smaller brained monkeys, such as the capuchin. Other animals, most notably dolphins, with an encephalization quotient of four to five, come precariously close to humans in terms of their relative brain size. To avid readers of Douglas Adams, the high degree of encephalization in dolphins may come as no surprise. And there is no doubt that dolphins are capable of some remarkable cognitive feats: they can form abstract numerical concepts or use learned ‘signature whistles’ to address each other. But again, none of these feats stands out as something other animals cannot do and as unique as, say, human language. Thus, whether the cognitive skills of dolphins — impressive as they may be — reflect in fact the extraordinarily high degree of encephalization is hard to tell. This has even led to the suggestion that the brains of dolphins and other toothed whales evolved relatively large brains for reasons unrelated to cognition. The so-called thermogenic hypothesis, for instance, claims that the brains of these animals enlarged as oceans were getting cooler and contain a high proportion of glia to effectively generate heat. Of course, heat was mainly generated within the dolphin cognition community by the implicit suggestion that dolphins might perhaps not be quite as clever as they had been made out to be. However, the possibility that large brains and high encephalization indices evolve because of selective pressures that have nothing to do with cognition is worth considering nonetheless. In the evolutionary history of toothed whales, the high encephalization quotient is in part owing to a reduction in body size, much as the very low encephalization quotient of baleen whales, such as the blue whale, is due to their attaining gigantic dimensions. This is one of the key difficulties with using relative measures of brain size. The encephalization quotient depends just as much on the size of the body as on that of the brain. But why should the size of the body say anything about the cognitive power of a brain? And, even more fundamentally, how do you assess brain power? Ultimately, the computational power of a brain must in some way depend on the number of its processing units, the neurons. But how do you count neurons? Usually, neuron numbers for whole brains had been inferred by extrapolations from small slices of brain. This approach comes with sizable uncertainty, as some brain parts consist mainly of fibres, while others have densely packed cell bodies. About a decade ago, scientists devised a new and ingeniously simple way to count brain cells. In essence, the brain is mashed up, so that only the nuclei remain intact. In the solution, the concentration of nuclei will be the same for each sample volume, they can then be counted in small volume and the number in the whole brain calculated. This technique has put cell numbers on a large number of brains — though largely confined to mammals and some other vertebrates — and has led to some fascinating insights. Take, for instance, the elephant brain. It contains a staggering 257 billion neurons, three times as many as the human brain with its 86 billion. 250 billion of these neurons (98%), however, are found in the cerebellum. Likewise the human cerebellum takes up a vast number of neurons, nearly 70 billion. Why so much computing hardware should be allocated to the cerebellum seems unclear. The cerebellum — the ‘small brain’ — is consigned to relatively mundane tasks to do with coordination of movement sequences and routines. So, perhaps it is the elephant’s highly moveable trunk that requires extraordinary cerebellar brain power. But whether it is in any way involved in functions related to higher cognitive skills is not clear. Fair to say, the cerebellum has been living its life as a subject in neuroscience largely in the shadow of the all important cortex. Interestingly, in humans and great apes, the cerebellum does appear to have undergone an evolutionary increase in size, both in absolute terms and in relation to the cortex. Perhaps an adaptation for dexterity in tool use. Bizarrely, humans (we don’t know about elephants) can live without a cerebellum, and the movement and speech defects are not entirely unmanageable — a testament to the great plasticity of the brain. So what does neuron number mean for brain computing power, when life is possible without 80% of a brain’s neurons. As with other measures of brain complexity, neuron number needs to be put into context. Just like brain size, neuron number scales with body mass. However, the scaling relationships depend crucially on the evolutionary context. The brain of a rook, for instance, a member of the corvid family famed for their intelligence, weighs a little less than the brain of a marmoset, yet it contains nearly three times as many neurons. This is most likely due to constraints placed on body and brain size evolution in the bird due to their need to stay aloft. When the human brain is considered in the context of its evolutionary relatives, the primates, it turns out that its number of cells — 86 billion neurons and roughly the same number of glial cells — is pretty much what one would expect of a typical primate of our size. In terms of scaling relationships, it looks just like an ordinary primate brain. This non-uniqueness remains when the brain is broken down into regions. The number of neurons in the cortex, whose relative expansion in size is often cited as one of the organic foundations of human cognitive superiority, is typical for a primate of our size. Also specific cortical areas, such as the prefrontal cortex, which is highly elaborated in primates and presumed to orchestrate higher cognitive functions, our internal goals and dispositions, seems to show no signs of a specific increase in neuron number. This leaves but one quantitative measure of human uniqueness, and that is simply the sheer number of neurons in the cortex, around 16 billion. It is this number alone that sets us apart from other animals. (Although dolphins might have even more, according to one estimate, which based on a different counting procedure comes up with 37 billion neurons for the pilot whale cortex.) This is hard to fathom, because neuron number is a continuous parameter to which seemingly discontinuous features, such as some of our unique cognitive skills, are hard to assign. It is like the old paradox: can a guy with one, three, fifty, 300 or 3,000 hairs on his head still be called bald? Is there a threshold number of neurons that, once crossed, all of a sudden leads to the emergence of uniquely human cognitive features. Or is our mind really just a quantitatively up-scaled version of the ape mind? If already the quantitative patterns and parameters that describe brain evolution are so hard to come to terms with, how about the causes? Logically, explanations for the forces driving brain evolution have been sought in cognition — after all, it is what brains do. Because brain size is so seductively easy to measure, a veritable scientific cottage industry has been at work to establish countless correlations between brain size and cognitive or ecological measures in a large number of animal groups. Sure enough, some patterns have emerged, such as an association of larger brains with more predation, or in the case of birds an unpredictable environment. But equally the uncertainty over the value of such associations has risen, especially over which measures best reflect cognitive performance. While such measures may be applied to individual species, comparing between species, especially over large evolutionary distances, is notoriously difficult, if not impossible given the incommensurability of some behaviours. The one association that has received the most attention is that between relative brain size and the social environment, especially the size of the groups an animal typically lives in. Known as the ‘social brain hypothesis’, the idea seems to make intuitive sense. Living in groups, while advantageous for many reasons, such as better resource use or shelter, comes with its own set of cognitive challenges. Every teenager is living testament. Members of a society need to recognize and remember each other and past encounters, they need to make decisions with whom to ally or to compete. There is no doubt that humans, as a particularly social species, excel in all of these domains. Humans are even able to put themselves into another’s mind, to infer their predispositions and intentions. In fact, the very ability to share and work towards common goals is perhaps one of the hallmarks of human cognition and evolutionary success. Likewise, there is ample evidence for dedicated hardware in the brain that handles social tasks, such as recognizing faces and the emotions they express or inferring others’ mental states. The correlation between group size and relative brain size is, however, mainly found in primates. In other mammalian taxa, such as carnivores, which also can have sophisticated social lives, it doesn’t seem to hold. For birds, rather than with flock size, larger brains seem to be correlated with pair-bonding — a cognitive challenge of its own. And for the insects, which have produced the largest animal societies on the planet, there is equally no clear correlation, though insect social life may be considerably less colourful than that of primates and involves behavioural routines that may require little additional brain power. Another, perhaps more fruitful way is, instead of looking at the cognitive benefits of larger brains, to look at their costs. Brains are energetically very costly. In resting humans, the brain consumes about one fifth to one quarter of all energy, even though it makes up only 2% of the body mass; in most other primates, relative energy consumption is probably less than half that. In order for the dramatic increase in brain size to be evolutionarily feasible, either energy needs to be funnelled away from other parts of the body towards the brain, or more energy needs be taken in overall. Just as, if you want to buy an expensive car you need to either cut down on fancy holidays or get a raise. It is becoming clear now that in the course of evolution the human lineage took the latter approach. Both basal metabolic rate and total energy expenditure are much higher in humans than in their closest relatives, the apes. Such a high basal turnover comes at a risk of course; if demands cannot be met because food is scarce, you wind up in dire straits quickly, just as if your monthly car payment is too high. Humans have met this challenge in two ways: by evolving more fat depots and by becoming defter at acquiring high-energy food, for instance through cooking and collaborating with others. This need to be able to maintain a high energy-throughput — which also allowed us to reproduce at an inordinately high rate — may thus have created precisely the selective environment in which the cognitive advantage of a larger brain may have played out. It is perhaps sobering that features so seemingly low-key as energy and metabolism seem to provide more explanatory power to understand the evolution of brains than fancy cognitive scenarios. No doubt this is due to our still lamentable ignorance of how brain size and structure correspond to brain function, and how these functions might provide selective advantages. As we are beginning to understand the brain better and better on its smallest scale, we still don’t know much about how microscopic architecture translates onto the larger scales. And we probably have a natural tendency, inherent in our presumption as the smartest guys on the planet, to ascribe any evolutionary change in brain size to cognitive adaptations. As long as we don’t understand these relationships better, we are still — as far as brain evolution is concerned — in the same boat as Steno, whose motto was “beautiful is what we see, more beautiful is what we understand, but the most beautiful is what we don’t understand” — yet!