A RARE example

The most difficult question to answer in science, as in life, is why? Here, I outline the key requirements for tool use, as I see them. I leave the thorny issue of actually defining ‘tool use’ for another post, and instead concentrate on what needs to happen in order to make tool use even possible.

At around the same time that the scientific world was waking up to the idea that wild animals regularly use tools, Dutch ornithologist Niko Tinbergen was formulating his own response to the why problem. In 1973 he co-won the Nobel Prize in Physiology or Medicine, in part for his thinking on this topic.

In a 1963 paper dedicated to fellow ethologist Konrad Lorenz (co-winner of that same Nobel Prize) Tinbergen outlined what have come to be termed his ‘four questions’. They looked at different scales of explanation for why an animal does something—whether eating, playing, building, fighting, mating, migrating and more—anything at all that is observable to the ethologist.

His questions were:

1. How does the animal actually physically carry out an action? Calling this ‘causation’, Tinbergen referred to the complicated interaction of muscles, nerves and biochemical processes required to sense and act. These days this explanatory level is more usually termed ‘mechanism’.

2. How did the animal develop this behaviour? Under the label ‘ontogeny’, it covers the way that individual animals change and learn about their environment over time, for example by avoiding toxic plants after experiencing their ill-effects.

3. How did the species evolve the behaviour? Tinbergen naturally called this ‘evolution’, and it takes a wider, long-term perspective on why animal behaviour differs between related species. A given behaviour may not have evolved in the same environment that exists now.

4. What is the fitness benefit? Tinbergen termed this ‘survival value’, and it refers to the way that specific outcomes, such as the sudden appearance of a predator causing an animal to flee or freeze, increase or decrease its chances of survival.

The first two questions are proximate: they refer to things happening in real time that can be tracked by a patient biologist. The second two are ultimate: they are best examined across groups of individuals or species over longer time-frames.

Tinbergen technology

We can turn Tinbergen’s questions to tool use, and ask the following:

• What internal abilities prompt and allow an animal to use a tool? As an example, let’s take a long-tailed macaque using a hand-held stone to break open oysters in coastal Thailand. The macaque is able to perceive the presence of closed oysters at low tide, move to find a suitable stone, then balance to strike repeatedly at the shell until it breaks, and use fine motor control to prise out the meat inside.

• How does tool use develop in an individual? We know that long-tailed macaques take time and practice over several years to become proficient at breaking oysters, and even at holding and precisely striking objects with a stone tool. Each macaque has its own developmental path, but building strength is a key factor.

• How did tool use evolve in a species? Not all long-tailed macaques use stone tools, just those in some coastal environments. And their close relatives don’t use tools in the wild at all. Evolution in this case is therefore linked to specific local clues, and likely involved the co-evolution of social cues that prompt (or at least don’t actively interfere with) tool use.

• What is the benefit of tool use? This one is fairly straightforward: the consumption of oysters brings in far more energy and nutrients than are expended in the use of the tool, a net gain for the animal. Plus, the noise created by macaque tool use doesn’t usually attract predators.

Under this scheme, the answer to why an animal uses tools is therefore fairly comprehensive—because their body allows them to, they grew up in conditions that permitted it, they’re a member of a species that evolved the capacity or tendency to do so, and they gain a benefit that aids their survival.

Introducing the RARE framework

This catalogue of explanatory levels has been extremely useful to ethologists, behavioural ecologists, neurobiologists and more over the past decades. However, as I’ve worked with wild capuchins, crows, macaques, sea otters and chimpanzees across continents and years, I’ve come to see the question of why animals use tools in a slightly different way.

The primary driver of my framework—which has the acronym RARE to make it easier for me to remember—has been whether or not it helps me understand why only certain species or sub-species use tools proficiently. It also needed to offer some insight into why a particular member of a tool-using community may learn faster than others, or not adopt tool-use at all.

The four-part framework is:

• Recognition. An animal must be able to differentiate a tool from the background noise of the environment, even if the tool is part of a larger structure (a stick still attached to a tree, or a stone flake embedded in a larger stone block). If a chimpanzee can’t identify a leaf as a potential sponge for collecting water, rather than litter or food, tool-use doesn’t even begin.

• Agility. An animal must be able to pick up, carry, manipulate and orientate the tool. The ability to smash open nuts, eggs and hollow branches would give many animals direct access to otherwise inaccessible food, but wielding a large wooden club or stone hammer is always going to be tricky for a snake.

• Reliability. The right tool must reliably co-occur with the conditions in which it is beneficial, both so an animal can learn its affordances, and so that the tool doesn’t need to be permanently carried around. Long sticks, for example, would certainly help with reaching burrowing prey in a desert, but that’s an environment often short on useful plant life. Carrying tools is one of the ways that humans distinguish themselves from other animals, although even for us there are weight and bulk limits.

• Effectiveness. The tool needs to make previously impossible tasks possible, or happen more quickly, or give greater reward, otherwise it’s a burden. The reward could be in mating success, or improved shelter, or social status, rather than just getting higher quality food, although to be fair a lot of animal tool use is all about a better lunch.

The RARE framework doesn’t replace Tinbergen’s questions, but it does focus them on what I see as the most productive channels. It also leans more openly on social influences, for example in identifying what is or isn’t (or or can or can’t be) a tool. Groups of animals can also increase the reliability of tool/target co-occurrence by concentrating tools in the landscape, an example of niche construction that doesn’t neatly fit into Tinbergen’s list. Of course, the need for effectiveness or efficiency, a benefit that sustains and perhaps originates tool use, is directly linked to Tinbergen’s notion of survival value.

Niche construction aside, the reliability component drifts further from the four questions, in that it’s usually out of an animal’s control. It encourages us to think about how landscapes themselves develop and change over time, as rivers, coastlines, forests, volcanic surfaces and more shift and evolve. It also ensures that tool-use is not simply a story of can-do attitude and internal drive, which is sometimes how human technology is portrayed. The fact that you don’t have Mozart’s concertos without his ready access to musical instruments, or Zuckerberg’s facebook without his ready access to electricity and silicon chips, should not be forgotten.

A final guiding principle for this framework was that it needed to help us consider the pathways that an animal takes in any given tool-use task. From finding and selecting a suitable tool, to wielding it correctly, to gaining from its use, an animal follows a kind of algorithm, a sequence of repeatable steps that lead to an expected outcome.

I will be discussing algorithms more in future posts, along with the fact that humans and other animals often follow quite similar simple routines when using tools. As a hint, an important difference between our own species and others may lie in the history and complexity stored within our tools themselves. But that’s a tale for another time.

Sources: Tinbergen, N. (1963) On aims and methods of Ethology. Zeitschrift für Tierpsychologie 20:410-433. || Bateson, P. & K. Laland (2013) Tinbergen’s four questions: an appreciation and an update. Trends in Ecology & Evolution 28: 712-718.