Is the Human Brain Special? A Conversation with Suzana Herculano-Houzel
The human brain is sometimes called the “most complex thing in the universe”. But what, if anything, makes it unique? And if it is so useful, why didn't other animals evolved one, too?
The human brain is sometimes called the “most complex thing in the universe”. It is what allows us to study ourselves, other animals, and the cosmos itself. Indeed, we often think of it as a pinnacle of evolution.
But vague notions of superiority aside, what do we actually know about the human brain? Is it somehow special? How different is it from the brain of an elephant? A chimpanzee? A raccoon? What, if anything, makes it worth the hype?
To discuss this topic, I was joined by the Brazilian neuroscientist Suzana Herculano-Houzel. Author of The Human Advantage, Herculano-Houzel has done a lot — perhaps more than anyone else — to help us understand these questions.
You are reading the On Humans newsletter. This is a written breakdown of my conversation with Suzana Herculano Houzel. You can listen to the episode on Spotify or wherever you get your shows. Just find On Humans and episode 27. Or you can continue reading
I find Herculano-Houzel’s work both clarifying and thought-provoking. It explains why the human brain is remarkable, yes, but not a biological oddity. Whatever makes our brains remarkable is firmly rooted in wider patterns in the natural world. Indeed, her calculations put our brain to be exactly where we would expect from biological models. (That’s unlike the elephant, the raccoon, or the chimpanzee. Their brains are truly special in various ways.)
Herculano-Houzel’s work also suggests an answer to what might be the biggest question in human evolution: If the human brain is such a successful creation, why didn’t any other animals develop one, too?
The Classic View
It has long been known that the human brain is very big compared to other great apes. The natural conclusion? Our big brain is linked to our (relatively) high intelligence. But what exactly is the link? Our brain is not the biggest. Elephants and whales have brains much bigger than us. Herculano-Houzel put the problem to me like this: “Why do we study elephants and not the other way around?”
The standard answer is that what matters is not brain size but brain-to-body ratio, an idea developed by Harry Jerrison in his 1973 book The Evolution of Brain and Intelligence. But the classic answer is problematic. As Herculano-Houzel explained:
“I've had my share of this fun asking engineers, especiallu biomedical engineers ‘Do you need more controlling units to operate a bigger body?’”1
And the answers are unanimous:
“Of course you don't! If you're not changing the shape of the body, if you're not changing the number of joints, then you can run a tiny kid-sized [robot] and a T-Rex sized robot with exactly the same number of control units”.
To see the problem in real data, consider primates. The brain-to-body ratio of humans is quite close to that of small monkeys (around 20%). So monkeys should be almost as clever as humans. On the other hand, the great apes and elephants would be far below. Yet these are both incredibly smart species on many metrics. (She even told me that, if we took Jerrison’s formula literally, we would need to conclude that great apes and elephants lack the brain mass required even just to run their bodies. This makes no sense. So we should forget about Jerrison’s metrics.) 2
The New Findings
The argument so far says the following: Intelligence is not about brain size, nor about brain-to-body ratio. What is it about?
This is where Herculano-Houzel’s own work comes in. She is famous for developing a method to reliably count the amount of neurons in a given brain. So far, her method has given us a reliable neuron count for the brains of 267 species.
And there is a pattern.
What really matters for intelligence is not brain size but the amount of neurons. This makes sense: we don’t hype computers that are big in size. We hype computers that fit a lot of transistors into a small space. Same with brains, she argues.
This is not a novel idea. We always acepted that neurons do the hard work of intelligene. But before having a means ot count them, we couldn’t do much with this knowledge. So we just weighed brains as a proxy. But it turned out to be a bad proxy. And now we know it.
But where do humans land on this metric?
Not at the top. The human brain was long assumed to have 100 billion neurons. That was not a bad guess. The right number, her research showed, was roughly 86 billion. This is a lot, sure. For example, baboons have “only” 15 billion neurons in their brain. But 86 billion pales in comparison with the elephant, whose brain hosts over 200 billion neurons. And remember, Herculano-Houzel argues that the huge body size of elephants makes this no less impressive.
Again, humans are not at the top. So should we simply accept that our brain is not special? Yes. Partially.
But there is still something unique about the human brain.
The Human Advantage
Remember the point about robots, controlling units, and body size? The engineer’s response was that size does not matter. But the shape does matter! And we all know that elephants have a funky shape. And this is important, for what mattered most in body shape is the amount of joints. Well, the elephant trunk has no joints. Unlike our limbs, the trunk can quite literally move however it wants. In robotics jargon, it has infinite degrees of freedom.
Could this explain why elephants have evolved an abnormally neuron-rich brain without also becoming the masters of mathematics? Perhaps. And some data hints at this idea. For the elephant brain is oddly shaped - a true biological anomaly, with 98% of its neurons located in the cerebellum, compared to 80% in most mammals. We don’t know exactly what this means. Indeed, the cerebellum is perhaps the most mysterious part of the brain. But Herculano-Houzel links this to the wonderful design of the elephant trunk. Her idea, as I read it, is that a lot of the elephant neurons are serving the movements of the trunk, not mathematical reasoning (or zoological studies on humans).
So what about humans? Unlike the elephant brain, our brain is not an anomaly. It is not oddly shaped. It does not have an abnormal amount of neurons in any one part of the brain. Contrary to my earlier beliefs, we don’t have an abnormally large cortex. Our brain is just a “generic primate brain”. But this statement is more significant than it sounds.
First, primate brains are special. They are the supercomputers of the animal world, able to pack a lot of neurons in a small space. And more importantly, they are very good at scaling. Herculano-Houzel explained:
“A primate with 10 times more neurons only has a 10 times bigger brain. A non-primate with 10 times the neurons has a 40 times bigger brain.”
Furthermore, humans are not just a generic primate. We are a big generic primate. Due to the scaling ability of the primate brain, a big primate “should” have a lot of neurons in their brain. And we do: 86 billion neurons is a lot. Granted, it is not as much as the elephant. But unlike the elephant, these neurons are distributed in a “normal” way. According to this generic mammalian plan, a bit less than 20% of neurons should be in the cortex. For a brain with 86 billion neurons, that number is around 16 billion. And that’s us. And here is the kick: that’s more than in any other species studied.
So here we are. The human brain is not oddly shaped. It is not oddly big. It is not even the most neuron-rich. It is big compared to our body size but this is probably a red herring.
What really matters is that our brain has more neurons in the cortex than anyone else. And this is no small thing: The cortex plays a vital part in intelligence. As Herculano-Houzel explained to me:
“The cerebral cortex literally saddles the rest of the brain, exactly like a saddle on a horse. I like the horse analogy because a horse can function without a saddle, but once you have a saddle on top of the horse you have the means to direct [the horse], you have this added layer of complexity to the horse's behaviour. And that's the exact same thing in a brain that has a cortex. You don't need the cortex, but once you have it saddling the brain, you have the possibility for much more complex and flexible behaviour. And that is because the cortex is this structure that receives copies of everything else that's going on in the other parts of the brain.” **3
She continued:
“The cerebral cortex has these ‘horizontal’ connectivities. It talks to itself. A lot of the neurons in the cortex simply talk to each other. They exchange signals. They build information together as they exchange signals with one another. And the result of this crosstalk is what we call ‘associative processing’. It means that the cortex is capable of putting things together that the rest of the brain can't do.”
I got excited at this point, suggesting that this explains possible links between the cortex and complex cognition, even consciousness. She nodded but suggested that we keep it basic – probably a wise suggestion.
“So with this associative connectivity the cerebral cortex has the capability for building patterns, for building associations, for finding associations between your actions and something else that happened out there. It puts those signals together. It creates new associations. It finds patterns in the past. It can use those patterns to simulate what’s next. So it can start representing a future. It can represent things that are not available to your senses anymore. You can start acting now towards futures that haven't happened yet.”
This was the clearest explanation I’d heard on the role of the cortex. Things clicked.
I continued to comment that this is important because many would say that these cortex-dependant skills, like thinking about future, are uniquely human. And they are not. But there is a kernel of truth in our hubristic intuition, as we we do have more neurons in the cortex than any other species. And so, we are probably more equipped to do these things than any other species on earth.
She agreed and continued positioning humans vis-a-vis other animals. The most neuron-rich cortices were, in descending order:
Humans (16 billion)
Great apes (6-8 billion)
Elephants (5-6 billion)
Dolphins & whales (1-4 billion, based on estimations)
Baboons (2- 3 billion),
Smaller monkeys (1-3 billion)
Raccoons (over 1 billion)
Crows (a notch less than 1 billion)
Raccoons are one of the more interesting animals on this list. They – not humans – are a true outlier.
“The raccoon is an outlier in a really cool way because every single part of the raccoon brain has twice the number of neurons that you would predict in a carnivore and brain of its size. It's like the entire raccoon brain underwent a doubling of neurons. And the consequence is that you have this creature that has a fairly small brain but it has twice the number of neurons that you would expect. ”If you gave me the numbers for a raccoon, I would tell you, “oh, what a cool primate brain!” But it's not a primate, it's a raccoon. And that makes perfect sense, of course, with all the anecdotes of how smart these creatures are.”
But let’s return to the list. You might be puzzled by one thing: the great apes are below us. But their bodies are huge, often bigger than humans. So aren’t we special after all? Why does Herculano-Houzel then call us a “generic primate”?
The answer is simple. Humans fit the pattern. Great apes don’t. Their brains are much smaller than they “should” be.
This might sound like a play of words. But the point is crucial. It gives us a hint of what happened in a critical chapter in human evolution.
Brains take energy. Energy means eating. And eating takes time. Herculano-Houzel suggests that as apes grew in size, their energy demands hit a ceiling. To keep their brains growing, they would have had to eat for over 10 hours per day — a bad idea. And so, the growth of their bodies was detached from the growth of their brains. The outcome: their brains are “too small” for a primate of their size.
Our ancestors were on this path too. But they discovered a cheat:
“And the cheat is that we have means to absorb more energy from the foods that we put in our mouths. And that's called pre-processing or pre-digesting your food. Or let me put it in the blunt way: Our food is pre-chewed. That is what cooking is.“
She was careful to note that she does not just mean cooking with heat. Fire helps. But fire is not necessary. Stone tools slicing meat is an early example. Smoothies are a modern-day one.
So this is our story: We are a primate. The primate brain is neuron-dense and good at scaling up in size. In theory, any big primate should have a brain like ours. But the others don’t: they cannot eat enough to fuel it. But we can. And that is due to cooking, whether with knives, stones, or fire.
Lessons for Darwinism
There was a lot in our conversation – more than is captured here. For example, I asked her to defend her basic premise that there is such a thing as “intelligene” that we can meaningfully study across species. Her answer was great. (To hear it, check the last 10 minutes of the episode.) I won’t go through it here. But before we end, I want to address one general lesson from Herculano-Houzel’s work. This is a lesson about evolution. The lesson is:
Mysterious are the ways of natural history.
Indeed, Herculano-Houzel repeatedly reminded me not to assume that everything has a clear adaptive story. No amount of armchair logic would have led people to know why humans, and not other mammals, developed a brain like ours. To know about it, one had to be a humble student of brain architecture and the body's energy demands. And to drive the point home, she told me that there is a deeply mysterious link between her studies and another key topic in human evolution: longevity.
I’ve done a whole episode on human longevity. That was episode number 6. My guest was Kristen Hawkes, famous for her theory on why humans evolved to outlive menopause. Her theory has to do with the benefits of having grandparents around. The theory is nuanced and plausible. It is backed by some exciting data. It might be right. But as Herculano-Houzel pointed out, we only need a theory if humans are a biological anomaly. Yet she has found a striking correlation — 72% — between the longevity of a warm-blooded animal and the number of neurons in their cortex. And again, humans fit right on the graph.
“If you had a warm-blooded vertebrate with 16 billion neurons in the cortex, [approximately the amount we have], you would predict that species to live a life of 90-something years. That’s us! You would also expect that species to only reach sexual maturity about 12 years after birth. That’s us!”
As far as I can tell, this link came completely out of the blue. And we have no idea about the mechanism. On that front, Herculano-Houzel was optimistic about finding an answer soon. But this is secondary.
“The point is: whatever gives an animal more neurons in the cortex also slows down their pace of life, slows down their growth, and allows them to live longer lives.”
What should we do with this information? I don’t know. Should we throw the grandmother hypothesis out of the window? I don’t know. But at least, we should add one more data point to the long list of the curious ways in which natural history has, once again, defied our expectations. I appreciate any work that reminds us of this all-important lesson. I also appreciate any work that links human capacities to the rest of the animal kingdom. Herculano-Houzel’s work does both. It was a wonderful opportunity to converse with her.
The full conversation is a bit over 1 hour. It includes all of the above, and much more: things from the brain of the T-Rex to Herculano-Houzel’s reflections on humanity. You can listen to it wherever you get your podcasts. There are some links below. Just find the show and episode 27. I hope you enjoy it!
Thank you for reading!
Take care,
Ilari
MORE LINKS
You can listen to On Humans on Spotify, Apple Podcasts, or other major platforms.
To you get your To get more pieces like this, you can subscribe to On Humans on Substack.
All quotations have been simplified by deleting filler words, filler phrases, and parts not critical for the argument. I have decided to hide all these edits to ensure legibility.
Does anyone know of existing defenders of Jerrison’s theory? Let me know!
To confuse the picture a little bit, she does admit that the cerebellum does the same thing. This adds to the mystery of the cerebellum. I hope to return to the topic one day. But let’s leave this complication aside.