Highly social lemurs are better thieves than their less-social cousins. On its surface, this isn't particularly surprising. The social intelligence hypothesis claims that the evolution of primate intelligence was driven by the need to predict and manipulate the behavior of others. So the more a primate species has to keep track of social relationships, the smarter those primates ought to be, the better they should be at stealing from you.

One problem with the social intelligence hypothesis is that most comparisons between species have been made just between two closely related species that live in groups of different sizes, such as among different types of corvids or monkeys. A more ideal test would be a comparison among several different closely related species.

Another problem is that intelligence is often - not always - correlated with relative brain size. The most "intelligent" species, such as humans, dolphins, and elephants, have brains that are significantly larger than would be expected given the size of their bodies. Human brains, for example, are seven times larger than their body mass predicts, while chimpanzee brains are just twice as large as would be expected. It is therefore unclear whether sociality is causally related to brain size, leaving intelligence as a byproduct of this relationship, or whether sociality is causally related to intelligence, making brain size the byproduct.

Also, it is not entirely clear whether sociality is related to intelligence in it's most general form, or whether sociality is only related to social intelligence and not to other forms of animal smarts.

In an attempt to address each of these problems, Duke University evolutionary anthropologist Evan L. MacLean turned to six lemur species. MacLean and his and colleagues published a paper today in PLoS ONE describing their findings.

While the different lemur species live in groups of different sizes, the average sizes of their brains controlled for body size are not tremendously different. This is in stark contrast to the the anthropoid primates (monkeys and apes), who do have differently-sized brains.

MacLean looked at six different kinds of lemurs, all housed in groups of typical size for their species at the Duke Lemur Center: ring-tailed lemurs, who live in groups of approximately 15 individuals; black lemurs, who live with nine or ten others; brown lemurs, who live in groups of eight to nine; Coquerel’s sifakas, which come in groups of six; ruffed lemurs, which come in groups of five; and mongoose lemurs who typically reside in groups of just two or three individuals.

MacLean gave ten different lemurs of each species one of two tests.

The test of social cognition was based on a common task used for human infants and non-human animals. Pieces of food are placed in front of two human researchers, but only one of the researchers makes eye-contact with the lemur. One researcher might face the lemur, while the other researcher might face away, or one researcher might have his eyes covered with a headband, while the second researcher's mouth is covered with the headband, leaving his eyes open. A lemur with more social-cognitive intelligence would be more likely to steal the food from the researcher who isn't paying attention.

The non-social test measured inhibitory control by using what is called a "detour task." The lemurs were trained to remove a piece of food from an opaque cylinder that was only open on the two ends. In order to get the food, the lemurs had to approach the cylinder from the side instead of from the front. Then, during the test phase, the opaque cylinder was replaced with a transparent one. The smarter lemurs would remember how to get the food out of the container, despite the fact that they had to first increase their distance from the food reward - that is, they had to take a detour - before they could access it.

The researchers discovered that a species' group size was predictive of performance in the social task, but not in the non-social detour task. Lemurs from species who maintain larger groups, like ring-tailed lemurs, were better thieves in the social task than those lemurs from species with smaller groups, like mongoose lemurs. Group size didn't matter, however, when it came to non-social reasoning. In fact, it was the ruffed lemurs who performed best at this task. There were no statistical relationships between brain size and performance on either task.

These findings shed light on the social intelligence hypothesis in two important ways. They show that it is social reasoning, specifically, that is related to sociality, not intelligence more generally. This makes good sense. The mind more likely evolved as a mosaic, with different skills changing independently of each other, in response to different sorts of selection pressures. There is no reason that skills such as those involved in spatial navigation would be subject to the same pressures as social reasoning.

These findings also show that brain size is not always a good proxy for cognitive sophistication. In fact, this study adds to a growing set of findings in which smaller-brained species outperform larger-brained species: for example, dogs routinely outperform wolves, especially at social-cognitive tasks, and bonobos sometimes outperform chimpanzees.

What does this mean for the evolution of human intelligence?

Possibly nothing. While group sizes and relative brain sizes are not correlated for lemurs, they are correlated for monkeys and apes, suggesting that brain evolution may simply have taken different paths for the different groups of primates. It could be that the largest of the lemur social groups, 15 ring-tailed lemurs, is still too small to pose the same demands on social reasoning as the larger social networks typical for monkeys and apes, including humans. Or, it could be that lemur social interactions, typically occurring between pairs of individuals, require less mental sophistication than the more complex alliance-based social interactions that tend to occur among monkeys and apes.

On the other hand, at least one thing is certain: these results should give researchers pause if they intend to use brain size as a metric for intelligence, even for monkeys and apes. As the researchers write, "the accumulation of findings violating a strict 1-to-1 relationship between these traits warrants great caution when making inferences about cognition based solely on brain size."

MacLean E.L., Sandel A.A., Bray J., Oldenkamp R.E., Reddy R.B. & Hare B.A. (2013). Group Size Predicts Social but Not Nonsocial Cognition

in Lemurs, PLoS ONE, 8 (6) DOI:

For more on lemurs:

Ringtailed Lemurs Look Where You’re Looking

Numerical Cognition and Hidden Grapes

Book Review: Mireya Mayor’s “Pink Boots and a Machete”

Header photo: Brown Lemur (Eulemur fulvus) via Wikimedia Commons/David Dennis. Experiment photo via Evan MacLean/Duke University.