The evolutionary mystery of gigantic human brains

By Cathleen O'Grady

Modern art, eh?
Enlarge / Modern art, eh?

In the Scottish National Gallery of Modern Art is a work that’s simply a list of names: every person the artist, Douglas Gordon, can remember meeting in his life. The list pours on and on in interminable columns, down a corridor and across walls that are multiple storeys high. It's dizzying, yet far from complete, since Gordon is only 52 years old.

The list is an abstract idea about humans made concrete in black paint: we are an intensely, bewilderingly social species. Our brains somehow have a vast vault for storing details about other humans, even when those details entail little more than face or name recognition.

It’s possible that this vast vault, along with a host of our brains' other cognitive abilities, are there precisely because of the intense sociality of our species. One of the most prominent explanations for the evolution of big brains is that large social groups lead to problem-solving challenges, which in turn create an evolutionary pressure for smart, big-brained individuals capable of navigating social situations.

Proponents of the social brain hypothesis point to reams of evidence to back it up, but a string of recent publications raises questions about whether it should dominate thinking in its field. It’s true that in primates, big brains generally go along with big social groups, but that correlation has started to look shakier when researchers have poked at it.

Definitive answers are hard to come by, as, when it comes to primate brains and behavior, what we don’t know dwarfs what we do. The precise combination of factors that drove evolutionary changes are lost to history. That has left the field struggling to test this hypothesis with data that's far less complete than it would like—a challenge shared with many other fields of science.

Yet even in this gray zone, progress can be made—a gradual thinning of the murk, as researchers focus on answering one small question at a time. Even when the most sorely needed information is out of reach, science as a whole can edge forward.

How do you solve a problem like cognition?

Big brains might not seem like such a big mystery. You have a bigger brain, you’re smarter, you take care of your survival needs more easily, and you can start to spend time painting pretty antelopes on cave walls. Boom, solved, it's only a matter of millennia before we’re on the Moon.

The mystery lies in the fact that big brains are hungry. They’re the organ equivalent of a muscle car’s engine, guzzling obscene amounts of fuel, when, for many animals, a Prius seems to do just fine. To justify its existence, the intellectual horsepower produced by the bigger brain had better be worth its weight in calories. And it’s not just humans who have big brains: all our primate relatives seem to be on the brainy side, including our closer great ape cousins like chimps and gorillas. More distant relatives like monkeys and lemurs are quite brainy, too.

Ring-tailed lemurs eat frozen pastry because they have big brains and know frozen pastry is delicious.
Enlarge / Ring-tailed lemurs eat frozen pastry because they have big brains and know frozen pastry is delicious.

Primates in general have brains that are weirdly big enough to demand an explanation—what in a lemur’s world makes it worthwhile to cart around a calorie black hole?

Larger social groups are what makes it worthwhile, according to the social brain hypothesis. Big groups are thought to have evolutionary advantages like protection against predators. But living in close social relationships with other individuals brings with it a host of cognitive challenges. Those animals that can handle the cognitive challenges survive and thrive; over generations, that means a species with bigger brains.

And that’s what evidence has pointed toward: primates with bigger social groups also have bigger brains. As evidence has mounted, the idea has become well-established among scientists in the field. Similar results have been found in other families of mammals, like ungulates (deer, camels, and the like) and cetaceans (whales and dolphins). Researchers have reported that individual magpies that are more social are also better at solving puzzles. “It has definitely been the consensus,” said Rob Barton, a researcher who studies brain evolution, in a phone call with Ars.

Revisiting the data

Barton had long been an advocate of the social brain, even co-authoring work on the hypothesis with its main proponent, Robin Dunbar. But he was worried about problems with the data and wanted to take another look. “It occurred to me that this hadn’t really been done fully since the old days,” he explained. He started by taking a cursory look, and when the results were unexpected, he passed the project to a doctoral student, Lauren Powell, to delve into the work more thoroughly.

“I honestly had no agenda about this,” Barton told Ars. “I was expecting to find a correlation with social group size.” But that’s not what they found. Instead, they found that brain size was more closely linked to factors in a primate species’ environment, things like how big their territories were and what kind of food they ate. Unsurprisingly, more calorific diets went along with bigger brains.

Meanwhile, a different group of researchers, led by primatologist Alex DeCasien, published a paper with a similar finding: fruit-eating primates had larger brains than leaf-eating primates. This pattern matched the data on brain size better than the primates’ social lives did.

Should we pivot from the social brain hypothesis to the ecological brain? Not so fast. Powell, Barton, and their colleague Karin Isler compared different data sets and found that the results looked different depending on how they sliced the data.

Small, noisy data sets make a clear answer impossible to pin down. But all the data sets currently available to the field are, well, small and noisy—the data the researchers would love to get their hands on is still well out of reach.

At San Diego's <a href="">Digital Brain Library</a>, each lacy slice of brain is about 70 micrometers thick, or roughly as wide as a human hair.
Enlarge / At San Diego's Digital Brain Library, each lacy slice of brain is about 70 micrometers thick, or roughly as wide as a human hair.

Dream the impossible data dream

Researcher Susanne Shultz, a proponent of the social brain hypothesis, was impressed by Powell’s analysis. “It’s very good. They've tested things rigorously, and their conclusions are very justified,” she said in a phone call with Ars. But, she points out, there should be two different correlations at play in the conversation: there’s the correlation between group size and brain size and a correlation between group size and a region of the brain called the neocortex.

The neocortex receives and integrates information from the external world, organizing it and cooperating with other brain regions to transform that information into behavior. In large-brained animals like primates, the neocortex makes up a disproportionately large part of the overall brain. Scientists are constantly improving their understanding of the neocortex, including figuring out more about how it communicates with other areas, but "the view still clings on that the neocortex is the 'intelligent' bit of the brain," says Barton.

Powell and her colleagues looked at the whole brain and didn’t find support for the social brain hypothesis. The jury’s still out on the neocortex correlation, but Barton points out that the neocortex makes up such a large proportion of the total brain that if the neocortex correlates with another factor (like social group size), then we should expect overall brain size to correlate, too. So, if we don't find the correlation for total brain size, it seems less likely that the correlation would be there for the neocortex.

Getting the verdict requires autopsies—lots of them. And opportunities for primate autopsies are, unsurprisingly, not an everyday occurrence. “Not all of us can go and dissect the brains of every animal that we're doing work with,” said Shultz.

It’s difficult to pin down exactly how many primate species there are, but it’s safe to say it’s currently more than 200; the data set with decent information on brain regions, published by Heinz Stephan and his colleagues in 1981, covers only around 40 species. That’s the data providing support for the neocortex findings.

Barton is actually a cheerleader for the idea that looking at specific brain regions is a better idea than looking at the brain as if it were a big lump—he points to nocturnal species that have enlarged olfactory regions that deal with smell and diurnal species that have bigger visual cortices. If you looked just at the whole brain of those groups, you wouldn’t get the full picture of how it's specialized. But because of the limited data on primates, Powell, Isler, and Barton were able to get whole-brain data on 114 species, so that’s what they used.

When they narrowed their analysis down to just the species used in the Stephan data set, the social brain seemed to be back in the running. The problem is that the data set is skewed in some very important ways—it focuses more on “Old World” African and Asian monkeys, says Shultz, and less on the “New World” monkeys of the Americas. Among those Old World species, there’s a heavy emphasis on the fruit-eaters. Powell also points to a lack of good data on great apes, like gorillas and orangutans, in the data set.

It’s possible that the neocortex correlation shows up because the skew in the data coughs up a false positive. It’s also possible that it’s real. If a primatologist ever won the lottery and funded the dream data set, we’d find out.

Listing image by Kate Shaw

Page 2

Having friends, a social group, may be good for your mind.
Enlarge / Having friends, a social group, may be good for your mind.

Our data woes extend beyond autopsies: even if we knew everything we wanted to know about the brains of different primate species, we'd need to match it up with information about their group behavior. Group size, though, is a nebulous concept. If you were asked how large your social group is, you’d likely be left spluttering about the definition of “group.” You would run into the same problems trying to pin down how many friends an orangutan has.

It’s pretty clear that a mother and her young form a social unit. It also seems fair to say that a giant herd of wildebeest or a thousands-strong shoal of fish is a loose group that’s not really a social network. But in between, it gets hard to say exactly how we should think about social group size. Ideally, “it’s about the social richness of an animal’s environment,” Shultz says. “Counting heads doesn’t tell us anything about that.”

To get good data on social group size, you ideally need to agree on what “social group” means in the first place.

On top of that, there’s the issue of time. “Things like social network size vary massively,” Powell explained, “even across windows of 40 days in the same individual. You’re looking at this monkey, and you’re taking a static snapshot of this monkey’s social circumstances, and it lives for like 20 years, with its ecological circumstances changing all the time. We need a lot more depth.”

The social brain hypothesis also depends on there being cognitive challenges of social living—but what are they, precisely? Different researchers talk about these challenges in different ways: learning to out-compete and out-maneuver competitors, learning cultural behaviors, or cooperating to solve problems. It’s clear that not everyone is talking about the same thing when they talk about the social brain.

“We don't really know what is cognitively challenging about being in a social group,” Shultz said. She favors the idea that the tricky thing is “solving ecological problems in a social context”—that is, how do you take care of your own needs while also managing your social relationships?

There might be ways forward to try to answer that question based on data, Shultz says: “What can a group of five do that they find really hard to do in a group of 10? I would really like to give animals foraging experiments: how much harder did it get if you have four animals with you or if you do it by yourself? That won’t necessarily tell us all that much about the social brain hypothesis, but it will tell us what’s hard about being in these kinds of groups.”

Experimental data could be useful, but that comes with its own problems, she adds, because the question is about multiple species at the same time. If you want to give a researcher nightmares about experiment design, ask them to come up with a task that everything from Capuchins to gorillas can do.

Big groups, big brains, big appetites

At this point, the conversation revolves around correlations—between brains, social groups, and environmental factors—which leaves certain questions frustratingly unanswerable. Short of being able to reconstruct the common ancestor of all primates and let some evolutionary experiments run for a few million years, it’s impossible to know for sure what the evolutionary chain of events was.

There are a few different ways that the possible scenarios could have played out. Here’s the story favored by proponents of the social brain hypothesis like Shultz and Dunbar: some factor, like a high risk of being eaten by predators, caused big groups of primates to form. This created an evolutionary pressure for the species to get social and, therefore, get smart. But in order for their brains to grow, they needed extra calories, forcing them to cover a wider territory looking for food and switch to high-calorie foods like fruit.

In this story, the social challenges are what cause the brain growth, and the calories are just a constraint holding back brain growth until more food could be acquired.

But we don’t know for sure that social groups were actually the cause of brain growth, rather than just something that happened alongside it. Here's an alternative story: something forced the primates to cover a wider territory to find food, and the new challenges of finding food created an evolutionary pressure for problem-solving abilities. Then, with their new-found smarts, the social challenges of big groups were no longer an issue, allowing big groups to emerge, along with benefits like safety in numbers.

In this story—the ecological brain hypothesis—the ecological challenge is the cause, and brain growth (or intelligence) is a constraint on group size. Without the magical primate evolution experiment, how do you know which is the real causal story?

In the ecological brain hypothesis, something forced primates to cover a wider territory to find food, and the new challenges of finding food created an evolutionary pressure for problem-solving abilities. So... digging for food led to better brains?
Enlarge / In the ecological brain hypothesis, something forced primates to cover a wider territory to find food, and the new challenges of finding food created an evolutionary pressure for problem-solving abilities. So... digging for food led to better brains?

Thinking causally

For Shultz and Dunbar, the essential point in favor of the first scenario is that living in groups is difficult, something that most humans who’ve been on a vacation with a large group of relatives can probably agree on. It “increases competition,” they write, and “induces costs in terms of time required for foraging, travel… and social bonding.” Even something as simple as moving a group to a new foraging spot is tough, Shultz pointed out to Ars. “If you look at group dynamics, one of the hardest things is to get a group of people to agree on something. It’s like a pub crawl problem, and animals are doing it on a day-to-day basis with no language.”

Because living in large groups is a pretty tough gig, Dunbar and Shultz write, it’s not something that’s likely to just pop up once animals are smart enough to cope with it. There has to actually be some reason it’s adaptive in its own right, like safety from predators. That’s why they think the first story is the one that makes sense—the group size expands first, social thinking has to improve, and calorie intake increases to keep up with the demands. And it's also why finding a correlation between high-calorie diets and big brains isn’t a convincing takedown of the social brain hypothesis, Shultz said—it just confirms that “poor-quality diets are associated with small brains.” That doesn’t tell us anything about the tangled chain of causal events that led to primate brain growth.

With robust data on primate brains and behavior, it might be possible to string together enough information on the primate family tree to get a sense of which developments came first—group size, brain size, and ecological changes. But short of that kind of data, experimental research might be the best bet. Even then, experimental work with primates is incredibly challenging.

“We need better brain data for sure,” said Shultz. “But I think a lot of things will stay correlational just because of the kinds of information that we can get. That doesn't make it useless. It just needs to be qualified.”

The socio-ecological brain

Consensus may be shifting, but the social brain is far from dead. The disagreement, at this point, revolves more around matters of emphasis than anything else. “It’s perfectly plausible that sociality was an important influence on cognitive evolution,” said Barton. “It would be bonkers to take our results and say that we’ve disproved that. I’d be very surprised if sociality had nothing to do with cognitive evolution, but it has been massively over-emphasized at the expense of other things, like tool use.”

“I think it’s misleading to look at it in a dichotomous way, as social or ecological,” said Powell. “Biology is inherently messy. An animal’s niche is part of a continuum, with lots and lots of dimensions. Although that isn’t very sexy… I think it’s more true.”

As for Shultz, she readily agrees that “it’s partly ecological, partly social; why can't we just call it the socio-ecological brain and we would all be done?”

That makes it sound like a closed case, but the precise details of that causal story will continue to dog the argument.

More data is on its way—not in a flood that would answer these myriad questions in the next few years but in a trickle that allows one small step at a time. A team of researchers in the Netherlands recently published a new dataset, measuring 16 brain areas from 39 primate species, including 20 that didn’t appear in the literature before. But in most cases, they managed to get data from just one individual member of each species, meaning that we don’t know how much variation there is in a species.

The impasse might eventually dissipate, especially if a trickle of brain data coincides with at least a trickle of behavioral data. For both, research funding is a crucial problem. “People are rarely sent out into the forest for a long time looking at a specific species, but that would pull the field up so much and go a long way towards allowing us to test much more nuanced hypotheses,” Powell said to Ars. “You really do need people out in the forest, looking up at the trees, counting monkeys.”