The custom officer's eyes opened wide. She was viewing an x-ray image of my two suitcases. Both were packed with plastic containers small and large, all double-wrapped individually and each carrying a soft mass in clear liquid. “Are you bringing fresh cheese?” she asked me. It was June 2012, and I was returning from South Africa to Brazil through the international airport in São Paulo. A Portuguese couple ahead of me had just been caught sneaking in prohibited fresh cheese, which could contain live pathogens harmful to local cattle.

“No, it's just brains,” I replied. I had no fear of having those precious organs confiscated. I knew I had done everything correctly, and so I waited patiently, sitting by my still unopened suitcases, while she went to fetch the appropriate officer. Somebody would have quite a story to tell that evening at dinner.

As the confused officials tried to remember what to do when someone brought brains into the country, I presented the customs agent with a thick stack of permits in several languages—including documentation that declared my specimens posed no biological threat and had no commercial value. I was bringing in brains of giraffes, various antelopes, lions, hyenas, one minke whale and dozens of smaller African rodent species that my collaborator had collected in South Africa, the Democratic Republic of the Congo, Saudi Arabia, Denmark and Iceland. I handed the agent my paperwork, and she let me go without ever opening my suitcases. Part of me was sorry: I wished she had seen my cool collection.

I have been in the business of importing animal brains for about 10 years now, carrying them from the laboratories of collaborators in many countries. My research interest is finding out how many neurons each contains, how that relates to brain size and how it compares with the human brain in particular.

Fourteen years ago, while at the Federal University of Rio de Janeiro, I developed a method that has allowed me to count neurons from creatures large and small, human and otherwise, something that could not be done before in vertebrate brains. My procedure? Turning brains into soup. The numbers we have obtained have overturned some of the old myths of human exceptionality and revealed that our brains are both uniquely powerful and surprisingly predictable in the context of other primates. In fact, comparing us with our evolutionary cousins suggests that it was technology rather than anatomy that allowed us to fully realize our neuronal capabilities.

How to Make Brain Soup

For more than 50 years scientists have been trying to number the brain's cells. Pioneered by anatomist Hans Elias in the early 1960s, the classic and most widely used approach to unbiased counting was stereology: one would “fix” brain tissues, turning them hard in formaldehyde, and then slice them finely. Chemical stains would then make the cells visible under a microscope, and a careful sampling scheme allowed scientists to extrapolate the total number of cells in a brain structure in just a few counts.

The problem with counting cells in this way was that it could only be done properly in well-defined, homogeneous brain regions. The procedure was painstaking and time-consuming. Although it was very accurate if used properly, it was also prone to user errors. Applying that procedure to whole brains, and particularly to large brains, would take forever.

In the 1970s some researchers observed that because there is a set amount of DNA content associated with each brain cell's nucleus—and there is just one nucleus per cell—it should be possible to extract all the DNA in a brain and use it to calculate the total number of cells. This idea inspired me: What if instead of extracting DNA from the nuclei, I were to extract the nuclei themselves?

Herculano-Houzel (left) with her packed suitcase full of specimens in 2012. She has studied hundreds of brains, including those of African elephants (right), to examine how the number of neurons varies across animal species. Credit: Suzana Herculano-Houzel

I figured that nuclei could be freed from cells, like pits out of peaches, and once liberated, one could mix them in a known volume of liquid until they were distributed evenly, then count them under the microscope without an elaborate sampling scheme. As I learned later, I was not the first person to count free nuclei. In 1963 comparative anatomist John Zachary Young counted and stained brain tissue in liquid to estimate the number of neurons in the brain and arms of an octopus. His tally: 500 million.

We now know that figure is likely too low. The reason my kind of brain soup worked accurately, whereas others missed cells, is that I started from fixed, not fresh, tissue, which hardened the nuclei and made them resistant to the process of liquefaction. Of course, it did not work at first. I initially borrowed from biochemists a common method to free nuclei: flash-freezing brain tissue in liquid nitrogen, then cracking it in a blender. Predictably I had frozen pieces of brain hurled around the lab. My mother advised: “You have to keep the lid on, silly.” But it was still no good; there were too many bits and pieces stuck to the walls of the blender. I had to make sure I could collect every last nucleus.

In 2003 I struck gold using detergent to dissolve brains that had been well fixed. Sloshing the tissue around in detergent inside a glass tube, I turned heterogeneously distributed neurons into a soup of evenly distributed free cell nuclei.

I could then easily and quickly count free nuclei under a microscope and—because of the one-nucleus-per-cell rule—that was as good as counting cells. For a rat brain, I needed only about half a cup of liquid to suspend all the nuclei and then tally them up. I could do that in a morning. Even an elephant brain can be enumerated in about six months through a few gallons of elephant brain soup.

Since 2003 research teams at both Vanderbilt University and the University of Nevada, Reno, have shown that brain soup gives comparable results to traditional stereology where both methods can be easily applied. Researchers in Canada, Australia, Germany, Hong Kong, the Czech Republic, Brazil and the U.S. have studied brain soups from birds, fish, mammals and invertebrates.

Are Humans Really Special?

Very quickly my method began yielding insights. I discovered that the numbers of neurons in the mammalian cerebral cortex, the brain's outer layer of tissue, is enormously variable—from just a few million to several billion neurons—in different species. These cells are responsible for sensory integration, movement generation, personality, temperament, pattern finding, logic reasoning and planning for the future, making behavior more than simple reactions to stimuli.

Related animals, such as ungulates, primates or rodents, have similar brain architecture but can have different numbers of neurons. Credit: Ignacio Palacios Getty Images (ungulate); Anup Shah Getty Images (orangutan); Getty Images (chipmunk)

Further, in the dozens of studies my colleagues and I have conducted to date, we have not found a single, universal relation between the size of a cerebral cortex and the number of neurons therein. In a 2014 review of our findings thus far, we concluded that different rules apply to primate and nonprimate animals, with much larger numbers of smaller neurons fitting inconspicuously in the cortices of primates than in, say, rodents or ungulates of similar size.

A baboon cortex, for instance, has 10 times more neurons than the similarly sized cortex of an antelope. The number of neurons, therefore, could not be surmised simply from the size of a cortex. The human cortex, meanwhile, has a whopping 16 billion neurons, which may seem out of the ordinary for our brain size—but only when we are compared with nonprimates.

The human brain has long been seen as an evolutionary outlier: too big for its body by the largest amount for any species. But according to my numbers, the human brain is actually just a scaled-up primate brain.

In 2014 we took a look at the elephant cortex, which is twice as large as ours, and found it has only about a third as many neurons: 5.6 billion. Even the largest whales, by our accounts, do not have much more than three billion to five billion cortical neurons. Most mammals have less than one billion.

For many years scientists also suspected that humans have a disproportionately large prefrontal region, the part of the cerebral cortex that deals with complex, associative functions beyond simply integrating sensory information and generating movements. Yet in 2016 we found evidence to the contrary. My colleagues and I looked at the distribution of neurons along the cortex of eight primate species and discovered that the human prefrontal cortex has only about 8 percent of all cortical neurons—the same proportion as in other primates. But because our cortex has the most neurons overall, that 8 percent translates into the largest number of such neurons in any primate.

The jury is still out on the number of prefrontal neurons in other brains, particularly in elephants and whales. These species have a lot of gray matter but a comparatively small number of neurons in the cortex when compared with us. Furthermore, their prefrontal cortex appears to amount to just a sliver of the brain, whereas it takes up proportionally more space in humans, making it likely that the human brain has the most higher-order, prefrontal neurons.

What does that matter? If neurons are the basic information-processing units of the brain, then the more neurons in a cerebral cortex, the more capable it should be, regardless of the overall size of the structure. We have, by far, the most cortical neurons of any single brain on earth. That, I believe, is the simplest explanation for our remarkable cognitive abilities, given that our brain's overall connectivity and distribution of functions are actually pretty typical for a mammalian brain.

Food for Thought

Brain cells are costly, so it is impressive that we can afford so many cortical neurons. Although the brain represents only 2 percent of our body mass, it chugs about 25 percent of all the energy required to operate the body each day.

Yet here again our brain is just a scaled-up primate's. By dividing estimates of how much energy different brains cost by our calculations of the number of neurons therein, we have found that rodent and primate brains alike cost about six kilocalories per billion neurons a day, regardless of their size. With 86 billion neurons on average, the human brain has an expected cost of 516 kilocalories a day, very close to its actual measured daily cost of about 500 kilocalories.

Bigger is not always better. Whales and elephants are large-brained but appear to have relatively few of the cortical neurons that assist humans in higher-order thinking. Parrots and songbirds, meanwhile, pack neurons more densely than mammals. Getty Images (whale); Annick Vanderschelden Getty Images (elephant); Frans Lanting Getty Images (parrot)

Humans have just the number of neurons and precisely the brain mass one would expect, given our bodies. In many Old and New World monkeys, as well as smaller primates, the brain represents 2 percent of body mass—just as in humans. It is great apes that stand out, with brains that amount to less than 0.5 percent of body mass. Gorillas, in particular, are animals that can weigh up to three times as much as humans. Because larger animals tend to have bigger brains, we would expect the biggest primates to have larger brains than us. Yet, to the contrary, the human brain weighs about three times more than the gorilla or orangutan brain.

Gorillas, with their large, expensive bodies, may have reached a point where they cannot afford the energy to support as many neurons as we do. With these findings, I could turn a key question of comparative neuroscience on its head: What if, because of energetic constraints, it was great apes that had brains too small for their bodies rather than humans having brains too large for their bodies?

Whatever energy is available to support brain and body, it has to come from what an animal eats. In 2012 Karina Fonseca-Azevedo and I published a paper—based on work done when she was my undergraduate student—with a few calculations, including how much energy different primate brains and bodies require, how much they receive from their natural diets, and how long it takes them to find and ingest that energy.

We found that at their current body mass, gorillas and orangutans could not afford any more neurons in their brains than they already have. These are animals that forage and eat for about eight hours a day, and they lose weight when the food they pick and ingest for eight hours is not enough, for example, during the dry season.

Eating longer hours to afford more neurons is not an option for a primate that also has to sleep for seven to eight hours daily and take care of other business, such as defending territory or enforcing social status. There is not much time left for anything else. A college education is an impossible dream for someone who needs to forage and eat for that long. Because energy intake is limited, there is a trade-off between body mass and number of neurons. In the case of great apes, their brain is only as large as the body still allows.

A similar limitation applies to us: On a comparable diet as other primates, our ancestors should have spent almost 9.5 hours a day looking for food and eating—something that would be prohibitive for a larger primate. If they, and we humans, still fed like other primates do, we could not have survived.

And yet here we are. If our ancestors did not skimp on neurons, did not have inordinately cheap brains and could not spend most of their waking lives eating, the only way out of the energetic limitation to brain cells was a radical change in diet. My colleagues and I argue that our ancestors found a way around that limitation about three million years ago. They improved on the lucky evolutionary innovation of bipedality—which extends the range that one can roam looking for food—by creating tools to cut, slash, dice, mince, crush and pound.

In 2016 research by paleoanthropologist Daniel E. Lieberman and his group at Harvard University showed that modifying food prior to eating—by my definition, cooking—increases its energetic yield. Lieberman's team conducted a series of experiments, including a setup in which participants had to gnaw on raw goat meat, that demonstrated how Paleolithic technologies, such as stone tools for slicing and pounding, altered food enough to make energy-rich, chewy meats easy to swallow.

Simply put: our ancestors, and ours alone, cooked. Homo culinarius, I like to call them—and maybe that would be a better name for our modern selves rather than the presumptuous and improbable sapiens, which implies that no other species thinks or knows. To the list of technological implements that could modify food before it was eaten, our ancestors later added fire, about one million years ago. Primatologist Richard Wrangham, also at Harvard, has previously proposed cooking with fire as a watershed in human evolution.

Great apes such as gorillas and orangutans need to spend hours foraging to have enough energy to sustain their large frames. Under similar constraints, we would need 9.5 hours of seeking and consuming calories from raw plants. Cooking gives us a higher caloric yield from the same foods and also makes it easier to consume meat, which allows our body to afford a larger number of neurons. Anup Shah Getty Images (gorilla); Christian Jegou Science Source (early humans)

Our research shows that had there not been a radical change in diet that tremendously increased the caloric intake of our ancestors, we could not feed our brain and therefore would not be here. Cooking, first without fire and later with it, was most likely that change.

The Legacy of Homo culinarius

Like any technology—by definition, objects, systems or procedures that help to solve problems—cooking freed time for our ancestors. They could put those now affordable extra neurons to other uses.

Once free time was no longer a rare commodity, our ancestors could develop technological innovations and share these discoveries with others. New tools begot further progress. Our species grew in culture and complexity. In the process, we probably pushed our brain ever further toward more neurons, and with them, we expanded our mind's capabilities.

Yet capabilities are not abilities. Judging from cranial size, our modern 16 billion neurons or so have been with us forat least 200,000 years. Our amazing cognitive feats—building, writing, investigating ourselves and the universe—are much more recent.

It takes a lifetime to sculpt a newborn human brain into a learned, mature brain of impressive abilities. In modern times, our collective wisdom and achievements are no longer within the grasp of a sole individual. Without enough people to hold that knowledge collectively and without cultural transmission, all our hard-earned gains could vanish in a single generation, despite the fact that we would remain capable of these deeds. We therefore must cultivate, document, and pass on knowledge and crafts through culture and formal education to ensure that our capabilities will give rise to the abilities of future generations.

Our species' achievements are many, and the potential of our collective thinking is tremendous. We have certainly distinguished ourselves from all other animals. But we have never stopped being primates.