The chimpanzee's clever use of sticks to fish for termites is fairly well known. In 1964, Jane Goodall announced her groundbreaking discovery to the world, writing in the journal Nature, "During three years in the Gombe Stream Chimpanzee Reserve in Tanganyika, East Africa, I saw chimpanzees use natural objects as tools on many occasions. These objects consisted of sticks, stalks, stems and twigs, which were used mainly in connexion with eating insects, and leaves which were used as ‘drinking tools’ and for wiping various parts of the body." While the collection of non-human animals known to use tools has expanded significantly in the nearly fifty years since Goodall watched her first Gombe chimp, the act of shoving sticks into termite mounds and snacking on the termites clinging to it has become the iconic example of non-human tool use, the model by which we judge all other possible forms. Perhaps lesser known is that another ape, the Sumatran orangutan (Pongo abelii), also fishes with sticks. Instead of fishing for termites, they fish for honey.

Like humans and chimpanzees, communities of orangutans on the Indonesian island of Sumatra have different traditions. In particular, only those living west of the Alas River have ever been observed fishing for honey with sticks. The behavior has never been observed in the wild among the orangutans living east of the river.

There is general consensus among cognitive psychologists that genetic differences between communities of the same species aren't sufficient to explain the variation in tool use that's been seen. Nor are the environmental differences from community to community typically large enough to account for the emergence of particular forms of tool use in some groups but not others. The explanation that remains is uniquely cognitive. That is, cultural differences must be explained by a social learning process. After an individual spontaneously uses a certain tool for a certain purpose, the behavior would then spread throughout the group. Naive or inexperienced individuals would learn about tools from watching older or more experienced individuals successfully use them. Over time, the behavior would become stable within that particular social grouping.

The implication of this tool use hypothesis is that individuals from communities in which tool use has not been observed are indeed cognitively capable of tool use; it isn't that they're somehow less mentally apt. Instead, the spontaneous use of a tool simply hasn't occurred in the first place, or if it has, circumstances have made it impossible to spread throughout the group.

While this account makes logical sense, it hasn't been easy to prove. Social learning among apes such as chimpanzees or orangutans has been heavily documented in captivity, and among apes born in the wild that were later rescued and raised in sanctuaries. However researchers have been unable to design an experiment that could prove that behaviors "deemed cultural in the wild" were spread through social learning. Now, a group of researchers led by Thibaud Gruber of the University of Zurich may have done just that.

Life is tough for the orangutans of Sumatra, primarily thanks to habitat loss due to deforestation caused by the logging and palm oil industries. As a result, lots of orangs wind up at the Batu Mbelin Care Center, in a town called Sibolangit in Northern Sumatra. There, wild-born orangutans that have been rescued or confiscated are rehabilitated (first in quarantine and then in social groupings) before being released into parts of the Sumatran forests that are safer for them.

Since orangutans from the honey-fishing communities west of the Alas River as well as from communities east of the river that don't use sticks to collect honey both wind up at Batu Mbelin, it was a perfect place for Gruber's team to assess the tool-using capabilities of the red apes. It was also much safer for the researchers, due to political unrest in other parts of Sumatra.

They gave the orangutans two different types of stick-related tasks.

In the honey task, they were given a vertical wooden log with a hole in the center, along with a wooden branch that had been stripped of most of its leaves. The orangutans watched while a keeper poured honey into the hole. The researchers then watched for ten minutes to see if the apes would dip their sticks into the hole to retrieve the honey. Of thirteen individuals originally from west of the river, nine were successful, compared with only two out of ten individuals from the east. This was consistent with the field observations that honey-dipping was prevalent to the west and all but absent in the east.

Next came the raking task.

The orangutans were presented with two sticks: a straight one and a curved one. The test was whether they attempted to use either of the tools to retrieve a piece of food that was out of reach. Ten of the thirteen western orangutans were successful, along with four of the ten from the east. While the margin between 77% success for the western orangutans and 40% for the eastern ones seems large, it is not statistically significant. Therefore, both groups can be thought of as similarly adept when it comes to using a stick as a rake.

Based on these results alone, it might seem reasonable to conclude that honey-dipping could be explained by cognitive rather than genetic factors, since both groups were able to do the raking task. The orangutans from the east weren't less skilled at both stick-related tasks, only the honey task. However, the results aren't so clear-cut. The researchers hadn't completely ruled out genetics as a possible factor leading to efficient tool use, only hinted at it.

They reasoned that if genetics was indeed behind the ability to use sticks as tools in general, then success in the raking task should predict success in the honey task. Using a statistical model, Gruber determined that neither age, sex, age at arrival at the center, or success in the raking test were predictive of success in the honey task. In fact, the only variable that predicted success in the honey task was geographic origin. This makes the case that social learning, rather than genetics or environment, was the key to tool use.

"These results," Gruber writes, "support the hypothesis that the Alas River constitutes a geographical barrier to the spread of the honey-dipping cultural variant." Interestingly, the forests west of the river are also capable of supporting more orangutans per square kilometer than those east of the river, lending even further support to the social learning hypothesis. New behaviors can only propogate efficiently through a group via social learning when the population density is sufficiently high.

There's more to this story. According to the researchers, most of the orangutans who participated in their study and were successful at the honey task arrived at Batu Mbelin at an age when they were too young to have experimented with honey-dipping in the wild. Field observations have indicated that juvenile orangutans do not use tools themselves until age 4, and do not become proficient until age 6 or 7. Of the honey-dipping orangutans at the care center, only two may have reached this age before being separated from their groups and brought to Batu Mbelin. Since a genetic explanation had already been ruled out because of the raking task, the researchers speculate that the two groups of orangutans arrived at the center with "different acquired knowledge," but that their knowledge was not the result of behaviors that had already completely developed or been extensively practiced.

Instead they argue that cultural knowledge, such as the use of sticks to fish for honey, might be thought of more abstractly as ideas, rather than more concretely as specific behaviors. Perhaps the orangutans from the west had inferred the idea that sticks could be used to retrieve honey by watching older individuals before they were physically capable of performing those behaviors. The idea, while highly speculative, is not without a reasonable theoretical basis. By altering the demands of different cognitive tasks, developmental psychologists have found that human infants are capable of high-level reasoning before their relatively immature anatomy provides them with the ability to act upon that reasoning. For example, a young infant might succeed in a task that measures eye gaze but not in a version of the same task that requires overt reaching behavior. (This is sometimes referred to as a distinction between performance and competence, and while supported by many developmental psychologists, it is not universally accepted.)

By analogy, it is possible that despite not having sufficient practice with honey-dipping in the wild, the orangutans from the west did have enough exposure to it before being relocated to Batu Mbelin in order to develop an understanding of the tool at the level of ideas, thus preparing them sufficiently for the task. Is culture acquired "at the representational level," as Gruber writes, "rather than the behavioral level"? The argument rests on the notion that orangutans aren't proficient tool users until age 6 or 7, but that is derived from a single observational study. More field observations would be necessary to strengthen the argument.

The ability of researchers to collect those observations, however, like the survival of the Sumatran orangutans themselves, is becoming increasingly threatened due to habitat loss. Over the last 75 years, the Sumatran orangutans have lost almost 80 percent of their population. Just 7000-7500 individuals remain. Unless more can be done to protect those that have survived, orangutan culture will simply fade away.

Gruber T., Singleton I. & van Schaik C. (2012). Sumatran Orangutans Differ in Their Cultural Knowledge but Not in Their Cognitive Abilities, Current Biology, 22 (23) 2231-2235. DOI:

Header photo via Flickr/Restless Mind.