Captive orangutan (Pongo pygmaeus) stone flaking

What’s the difference between doing something, and being made to do it? And how does that distinction translate into animal behaviour, when a captive member of a species can be trained or coerced into actions that have never been seen in the wild? Do we learn something meaningful about the ‘capacities’ of a species by prompting unusual behaviour, or are we just playing at being circus ringleaders, essentially pushing bears to ride bicycles?

Today we examine a 50-year-old experiment, from March and April of 1971. It involved a juvenile orangutan named Abang, who was wild when captured in Sarawak as an infant, then in 1967 flown to Bristol Zoo in the United Kingdom. Using the published report of that experiment, which was performed by archaeologist (and Bristol-native) Richard Wright, we’ll take a journey into the potentially ethically fuzzy area of captive animal research.

A simple question

Wright began his research with a hypothetical: what if Jane Goodall had not only found plant tool use among the Gombe chimpanzees, but also stone tool flaking, and apes using those flakes as cutting tools? Many types of stone leave a sharp edge when violently struck, a consequence of fine-grained raw material and the way that force projected into a stone block travels through and then exits that block. In theory, therefore, any animal with sufficient strength and coordination can make such tools.

However, cutting tools are considered a human-only invention, one of the ways that our ancestors were able to extract meat to feed our rapidly growing brains. As Wright spelled out clearly:

The sharpest reaction to my imaginary discovery would be from those who need a single and simple qualitative trait to be stamped, as a hallmark, onto the complex notion ‘human’. The apes are unwelcome applicants for admission to the human state.

The question was whether non-human animals were cognitively or physically incapable of making and using sharp-edged stone tools, or whether they were insufficiently motivated or provisioned. Initially Wright (then based at the University of Sydney) planned to find captive chimpanzees or gorillas with which to conduct an experiment into this question. After all, the African great apes are the closest living relatives to humans. However, ‘matters of housing and pregnancy’ forced him to turn instead to Bristol Zoo’s orangutans.

Starting with two juveniles, male Abang and female Dayang, Wright found that they distracted each other too much for him to carry out his research. So he isolated Abang, at the time about five and half years old, on ‘the hunch that he had more potential than Dayang’. That’s Abang in the image at the top of this post, taken from a video made of Wright’s experiment by the BBC.

Cutting the cord

How did the experiment work? Abang was initially given small flakes of English flint, a siliceous rock from chalk deposits that can break in an almost glass-like fashion. Wright made these flakes away from the orangutan’s sight, and demonstrated their ability to cut through a short nylon cord holding closed an aluminium food box. The idea was to see whether Abang caught on to the notion that the flakes were a step to getting food.

Here’s Abang with a flake that he’s just manipulated with his mouth, and the portal through which he needs to reach to cut the cord. It’s inside an enclosure so he can’t just bite through the string, something he did still try:

After nine human demonstrations over two days, and 12 failures to open the box, Abang finally suceeded. In the final failure before that success, the zoo keeper actually held and guided Abang’s hand directly to the string. In subsequent trials the orangutan was much more successful, although he also had to be restrained by the keeper at one point when the food box was being filled, and another time he sliced open his own finger with the blade.

In the second part of the experiment, Wright tested whether Abang could be made not just into a tool-user, but a tool-maker. The orangutan was given a block of flint, and a river cobble with which to strike it, in the hope that this would result in Abang making his own small flake tools. The wooden enclosure floor was too hard to stop the flint from sliding around, so Wright used a succession of plaster and other adhesives, and finally a leather strap, to attach it to a board. Here’s another image to help you visualise the new set-up:

Again, success was only achieved after a series of failures where Abang either failed to make a suitable flake, or failed to use a flake to open the box. In the meantime, Wright repeatedly demonstrated how to break the flint core with the hammer stone. Eventually Abang succeeded, on the sixth day of this second phase, at a time when his entire daily meal was locked in the problem box.

A week and a half later the BBC filmed Abang and Wright, in the final session of the experiment before Wright flew back to Australia. Here’s that video—the raw footage was apparently destroyed accidentally by the BBC:

Learning from captivity

What should we make of this? To start, we can say that it is definitely within the capacity of a juvenile orangutan to break off a piece of flint, and then use that flint to cut a string and get to its food. In that broad sense, orangutans have the potential to be stone tool users. However, the circumstances are highly unusual.

Abang was taken from his wild family at around the age of one, and flown to a place that must be close to the exact social, visual, climatic, acoustic, olfactory and tactile opposite of the warm forest of northwest Borneo: a cage in Bristol. Four years later a British archaeologist from Sydney dropped by and showed Abang the cutting potential of flint flakes, while also inducing hunger in the ape. The unfamiliar aluminium box was accompanied by a succession of different presentations of a flint core.

Should Abang’s eventual success at this highly artificial task be considered relevant to what wild orangutans, or other apes, actually do out in the natural world? To his credit, Wright took a pragmatic approach, noting that:

I think that the orangutan’s achievements would not have been widely predicted. Nevertheless the work does not throw new light on the capabilities of apes, whose exploits under training have for long transcended Abang’s simple actions.

Elsewhere he makes explicit the circus comparison, recognising that ‘The imitative capabilities of chimpanzees are exploited by circuses and space scientists’. Instead, Wright saw his work more as a sign that stone tool making was simpler than foreseen, and so perhaps within the range of our small-brained and un-opposable-thumbed Australopithecine ancestors.

Some of the behaviours that Wright saw—especially Abang’s manipulation of the stone flakes in is mouth, and even attempts to cut the cord while holding the flake in his mouth—have been since found to be part of the tool-use of wild orangutans. However, all known wild tool use by orangutans involves plant parts such as leaves, branches and sticks, rather than stones. For an arboreal genus that spends much of its time in the trees, orangutan adaptation to stones (which are famously not found in trees) would make little practical sense.

In some work I did a few years ago for my co-edited issue of the Philosophical Transactions of the Royal Society B, I looked at the topic of primate tool use in captivity versus the wild. What I found was a ‘captivity bias’, especially for well-studied species like capuchins, macaques, and the apes. This bias was seen in the much greater diversity of tool use forms seen in captive animals than witnessed in the wild. In that work I attributed the difference to things like the safety and free time afforded to captive animals, their regular interaction with humans, and their direct contact with other tool users in their group. Labs and zoos are inextricably artificial, and without accounting for that we cannot reliably understand any behaviour we see there.

The tussle over the greater ecological and social validity of wild reports, and the greater control gained in captive contexts, is longstanding and not going to be resolved any time soon. This is a topic that we will return to on occasion as I post more on this site, and so I will not go further into it here. For now, it is worth perhaps taking a minute or two to consider the arguments you would make for either side, including the ethical question of keeping caged animals in the first place, and the value of the precise comparative data we get from access to controlled animal behaviour. It’s a difficult puzzle box, and we are still struggling to find a way to cut its cord.

Sources: Wright, R. (1972) Imitative Learning of a Flaked Stone Technology—The Case of an Orangutan. Mankind 8: 296-306. || Haslam, M. (2013) ‘Captivity bias’ in animal tool use and its implications for the evolution of hominin technology. Philosophical Transactions of the Royal Society B 368: 20120421

All image credits: BBC & UNE Archaeology, https://www.youtube.com/watch?v=3exAOxSKYCE