Can we know what animals are thinking?
The inner lives of other species may be a lot richer than science once thought
For the first of our new series of Medium stories about the animal kingdom, we’ve chosen a 2015 essay about the minds of animals. Animals think, therefore…?
In 1992, at Tangalooma, off the coast of Queensland, people began to throw fish into the water for the local wild dolphins to eat. In 1998, the dolphins began to feed the humans, throwing fish up onto the jetty for them. The humans thought they were having a bit of fun feeding the animals. What, if anything, did the dolphins think?
Charles Darwin thought the mental capacities of animals and people differed only in degree, not kind — a natural conclusion to reach when armed with the radical new belief that the one evolved from the other. His last great book, “The Expression of Emotions in Man and Animals”, examined joy, love and grief in birds, domestic animals and primates as well as in various human races. But Darwin’s attitude to animals — easily shared by people in everyday contact with dogs, horses, even mice — ran contrary to a long tradition in European thought which held that animals had no minds at all. This way of thinking stemmed from the argument of René Descartes, a great 17th-century philosopher, that people were creatures of reason, linked to the mind of God, while animals were merely machines made of flesh — living robots which, in the words of Nicolas Malebranche, one of his followers, “eat without pleasure, cry without pain, grow without knowing it: they desire nothing, fear nothing, know nothing.”
The organism itself remained a black box: unobservable things such as emotions or thoughts were beyond the scope of objective inquiry
For much of the 20th century biology cleaved closer to Descartes than to Darwin. Students of animal behaviour did not rule out the possibility that animals had minds but thought the question almost irrelevant since it was impossible to answer. One could study an organism’s inputs (such as food or the environment) or outputs (its behaviour). But the organism itself remained a black box: unobservable things such as emotions or thoughts were beyond the scope of objective inquiry. As one such “behaviourist” wrote in 1992, “attributing conscious thought to animals should be strenuously avoided in any serious attempt to understand their behaviour, since it is untestable [and] empty…”.
By then, though, there was ever greater resistance to such strictures. In 1976 a professor at Rockefeller University in New York, Donald Griffen, had taken the bull by the horns (leaving aside what the bull might have felt about this) in a book called “The Question of Animal Awareness”. He argued that animals could indeed think and that their ability to do this could be subjected to proper scientific scrutiny.
In the past 40 years a wide range of work both in the field and the lab has pushed the consensus away from strict behaviourism and towards that Darwin-friendly view. Progress has not been easy or quick; as the behaviourists warned, both sorts of evidence can be misleading. Laboratory tests can be rigorous, but are inevitably based on animals which may not behave as they do in the wild. Field observations can be dismissed as anecdotal. Running them for years or decades and on a large scale goes some way to guarding against that problem, but such studies are rare.
No animals have all the attributes of human minds; but almost all the attributes of human minds are found in some animal or other
Nevertheless, most scientists now feel they can say with confidence that some animals process information and express emotions in ways that are accompanied by conscious mental experience. They agree that animals, from rats and mice to parrots and humpback whales, have complex mental capacities; that a few species have attributes once thought to be unique to people, such as the ability to give objects names and use tools; and that a handful of animals — primates, corvids (the crow family) and cetaceans (whales and dolphins) — have something close to what in humans is seen as culture, in that they develop distinctive ways of doing things which are passed down by imitation and example. No animals have all the attributes of human minds; but almost all the attributes of human minds are found in some animal or other.
Consider Billie, a wild bottlenose dolphin which got injured in a lock at the age of five. She was taken to an aquarium in South Australia for medical treatment, during which she spent three weeks living with captive dolphins which had been taught various tricks. She herself, though, was never trained. After she was returned to the open sea local dolphin-watchers were struck to see her “tailwalking” — a move in which a dolphin stands up above the water by beating its flukes just below the surface, travelling slowly backwards in a vaguely Michael Jackson manner. It was a trick that Billie seemed to have picked up simply by watching her erstwhile pool mates perform. More striking yet, soon afterwards five other dolphins in her pod started to tailwalk, though the behaviour had no practical function and used up a lot of energy.
Such behaviour is hard to understand without imagining a mind that can appreciate what it sees and which intends to mimic the actions of others (see “The imitative dolphin”). That in turn implies things about the brain. If you had to take a bet on things to be found in Billie’s brain, you’d be well advised to put money on “mirror neurons”. Mirror neurons are nerve cells that fire when the sight of someone else’s action triggers a matched response — they seem to be what makes yawning contagious. A lot of learning may require this way of linking perception to action — and it seems that, in people, so may some forms of empathy.
Mirror neurons are important to scientists attempting to find the basis of the way the human mind works, or at least to find correlates of that working, in the anatomy of human brains. The fact that those anatomical correlates keep turning up in non-human brains, too, is one of the current reasons for seeing animals as also being things with minds. There are mirror neurons; there are spindle cells (also called von Economo neurons) which play a role in the expression of empathy and the processing of social information. Chimpanzee brains have parts corresponding to Broca’s area and Wernicke’s area which, in people, are associated with language and communication. Brain mapping reveals that the neurological processes underlying what look like emotions in rats are similar to those behind what clearly are emotions in humans. As a group of neuroscientists seeking to sum the field up put it in 2012, “Humans are not unique in possessing the neurological substrates that generate consciousness. Non-human animals, including all mammals and birds, and many other creatures…also possess these neurological substrates.”
But to say that animals have a biological basis for consciousness is not the same as saying they actually think or feel. Here, ideas from the law may be more helpful than those from neurology. When someone’s state of being is clearly impaired by a calamity of some sort, it can fall to the courts to decide what level of legal protection should apply. In such cases courts apply tests such as: is he or she self-aware? Can he recognise others as individuals? Can he regulate his own behaviour? Does he experience pleasure or suffer pain (that is, show emotion)? Such questions reveal a lot about animals, too.
The most common test of self-awareness is the ability to recognise yourself in a mirror. It implies you are seeing yourself as an individual, separate from other beings. The test was formally developed in 1970 by Gordon Gallup, an American psychologist, though its roots go back further; Darwin wrote about Jenny, an orang-utan, playing with a mirror and being “astonished beyond measure” by her reflection. Dr Gallup daubed an odourless mark on the face of his subjects and waited to see how they would react when they saw their reflection. If they touched the mark, it would seem they realised the image in the mirror was their own, not that of another animal. Most humans show this ability between the ages of one and two. Dr Gallup showed that chimpanzees have it, too. Since then, orang-utans, gorillas, elephants, dolphins and magpies have shown the same ability. Monkeys do not; nor do dogs, perhaps because dogs recognise each other by smell, so the test provides them with no useful information.
Recognising yourself is one thing; what of recognising others — not just as objects, but as things with purposes and desires like one’s own, but aimed at different ends. Some animals clearly pass this test too. Santino is a chimpanzee in Furuvik zoo in Sweden. In the 2000s zookeepers noticed that he was gathering little stockpiles of stones and hiding them around his cage, even constructing covers for them, so that at a later time he would have something to throw at zoo visitors who annoyed him. Mathias Osvath of Lund University argues that this behaviour showed various types of mental sophistication: Santino could remember a specific event in the past (being annoyed by visitors), prepare for an event in the future (throwing stones at them) and mentally construct a new situation (chasing the visitors away).
Chimps also understand that they can manipulate the beliefs of others; they frequently deceive each other in competition for food
Philosophers call the ability to recognise that others have different aims and desires a “theory of mind”. Chimpanzees have this. Santino seemed to have understood that zookeepers would stop him throwing stones if they could. He therefore hid the weapons and inhibited his aggression: he was calm when collecting the stones, though agitated when throwing them. An understanding of the capabilities and interests of others also seems in evidence at the Centre for Great Apes, a sanctuary in Florida, where male chimpanzees living with Knuckles, a 16-year-old with cerebral palsy, do not subject him to their usual dominance displays. Chimps also understand that they can manipulate the beliefs of others; they frequently deceive each other in competition for food.
Another test of legal personhood is the ability to experience pleasure or pain — to feel emotions. This has often been taken as evidence of full sentience, which is why Descartes’s followers thought animals were unable to feel, as well as reason. Peter Singer, an Australian philosopher and doyen of “animal rights”, argues that, of all the emotions, suffering is especially significant because, if animals share this human capacity, people should give consideration to animal suffering as they do to that of their own kind.
Animals obviously show emotions such as fear. But this can be taken to be instinctual, similar to what happens when people cry out in pain. Behaviourists had no trouble with fear, seeing it as a conditioned reflex that they knew full well how to create. The real question is whether animals have feelings which involve some sort of mental experience. This is not easy. No one knows precisely what other people mean when they talk about their emotions; knowing what dumb beasts mean is almost impossible. That said, there are some revealing indications — most notably, evidence for what could be seen as compassion.
Some animals seem to display pity, or at least concern, for diseased and injured members of their group. Stronger chimps help weaker ones to cross roads in the wild. Elephants mourn their dead (see “The grieving elephant”). In a famous experiment, Hal Markowitz, later director of the San Francisco zoo, trained Diana monkeys to get food by putting a token in a slot. When the oldest female could not get the hang of it, a younger unrelated male put her tokens in the slot for her and stood back to let her eat.
There have also been observations of animals going out of their way to help creatures of a different species. In March 2008, Moko, a bottlenose dolphin, guided two pygmy sperm whales out of a maze of sandbars off the coast of New Zealand. The whales had seemed hopelessly disoriented and had stranded themselves four times. There are also well-attested cases of humpback whales rescuing seals from attack by killer whales and dolphins rescuing people from similar attacks. On the face of it, this sort of concern for others looks moral — or at least sentimental.
In a few examples the protecting animals have been seen to pay a price for their compassion. Iain Douglas-Hamilton, who studies elephants, describes a young female which had been so severely injured that she could only walk at a snail’s pace. The rest of her group kept pace with her to protect her from predators for 15 years, though this meant they could not forage so widely. As long ago as 1959, Russell Church of Brown University set up a test which allowed laboratory rats in half of a cage to get food by pressing a lever. The lever also delivered an electric shock to rats in the other half of the cage. When the first group realised that, they stopped pressing the lever, depriving themselves of food. In a similar test on rhesus monkeys reported in the American Journal of Psychiatry in 1964, one monkey stopped giving the signal for food for 12 days after witnessing another receive a shock. There are other examples of animals preferring some sort of feeling over food. In famous studies by an American psychologist, Harry Harlow, rhesus monkeys deprived of their mothers were given a choice between substitutes. One was made of wire and had a feeding bottle, the other was cloth, but without food. The infants spent almost all their time hugging the cloth mother.
If animals are self-aware, aware of others and have some measure of self-control, then they share some of the attributes used to define personhood in law. If they display emotions and feelings in ways that are not purely instinctive, there may also be a case for saying their feelings should be respected in the way that human feelings are. But the attribute most commonly thought of as distinctively human is language. Can animals be said to use language in a meaningful way?
Animals communicate all the time and don’t need big brains to do so. In the 1940s Karl von Frisch, an Austrian ethologist, showed that the “waggle dances” of honeybees pass on information about how far away food is and in what direction. Birds sing long, complex songs either to mark territory or as mating rituals. So do pods of whales (see “The singing whales”). It is hard, though, to say what information, or intention, goes into all this. The bees are more likely to be automatically downloading a report of their recent travels than saying, “There’s pollen thataway, slackers.”
The vocalisations of, say, vervet monkeys have more to them. Vervets make different alarm calls for different predators, demanding different responses. There is one for leopards (skitter up into the highest branches), for eagles (hide in the undergrowth) and for snakes (stand upright and look around). The monkeys need to recognise the different calls and know when to make which one. Animals brought up with humans can do much more. Chaser, a border collie, knows over 1,000 words. She can pull a named toy from a pile of other toys. This shows that she understands that an acoustical pattern stands for a physical object. Noam Chomsky, a linguist, once said only people could do that. Remarkably, if told to fetch a toy with a name she has not heard before placed in a pile of known, named objects, she works out what is being asked for. Betsy, another border collie, will bring back a photograph of something, suggesting she understands that a two-dimensional image can represent a three-dimensional object.
More impressive still are animals such as Washoe, a female chimpanzee which was taught sign language by two researchers at the University of Nevada. Washoe would initiate conversations and ask for things she wanted, like food. But evidence that many animals can, when brought up with humans, tell their thoughts to others using a human language is not quite the same as saying they use language as people do. Few have a smidgen of grammar, for example — that is, the ability to manipulate and combine words to create new meanings. It is true that dolphins in captivity can distinguish between “put the ball in the hoop” and “bring the hoop to the ball”. Alex, an African grey parrot, combined words to make up new ones: he called an apple a “bannery”, for example, a mixture of banana and cherry (see “The chatterbox parrot”). But these are exceptional cases and the result of intense collaboration with humans. The use of grammar — certainly a complex grammar — has not been discerned in the wild. Moreover, animals have no equivalent to the narratives that people tell one another.
If language can still be claimed as uniquely human, can anything else? Until recently, culture would have been held up as a second defining feature of humanity. Complex ways of doing things which are passed down not by genetic inheritance or environmental pressure but by teaching, imitation and conformism have been widely assumed to be unique to people. But it is increasingly clear that other species have their own cultures, too.
In “The Cultural Lives of Whales and Dolphins”, Hal Whitehead of Dalhousie University, Nova Scotia, and Luke Rendell of the University of St Andrews, in Scotland, argue that all cultures have five distinctive features: a characteristic technology; teaching and learning; a moral component, with rules that buttress “the way we do things” and punishments for infraction; an acquired, not innate, distinction between insiders and outsiders; and a cumulative character that builds up over time. These attributes together allow individuals in a group to do things that they would not be able to achieve by themselves.
For the first feature, look no further than the crow. New Caledonian crows are the champion toolmakers of the animal kingdom. They make hooks by snipping off V-shaped twigs and nibbling them into shape. They fashion Pandanus leaves into toothed saws. And in different parts of the island they make their tools in different ways. Studies by Gavin Hunt of the University of Auckland showed that the hooks and saws in two sites on New Caledonia differed systematically in size, in the number of cuts needed to make them and even according to whether they were predominantly left-handed or right-handed. To the extent that culture means “the way we do things around here”, the two groups of crows were culturally distinct.
Chimpanzees are now known to manipulate over two dozen implements: clubs to beat with, pestles to grind with, fly whisks, grass stalks with which to fish for termites, spongy leaves to soak up water, rocks as nutcrackers. Like New Caledonian crows, different groups use them slightly differently. William McGrew of Cambridge University argues that the tool sets of chimpanzees in western Tanzania are just as complex as the simplest human tools, such as early human artefacts found in east Africa or indeed those used in historic times by native peoples in Tasmania.
The skill needed to make and use tools is taught. It is not the only example of teaching that animals have to offer. Meerkats feed on scorpions — an exceptionally dangerous prey which you cannot learn to hunt by trial and error. So older meerkats teach younger ones gradually. First they incapacitate a scorpion and let the young meerkat finish it off. Then they let their students tackle a slightly less damaged specimen, and so on in stages until the young apprentice is ready to hunt a healthy scorpion on its own.
Pretty much all meerkats do this. Elsewhere what is taught can change, with just some animals picking up new tricks. As the story of Billie the tailwalker implies, whales and dolphins can learn fundamentally new behaviours from each other. In 1980, a humpback whale started to catch fish off Cape Cod in a new way. It would slam its flukes down on the surface of the water — lobtailing, as it is known — then dive and swim round emitting a cloud of bubbles. The prey, confused by the noise and scared of the rising circle of bubbles, bunched themselves together for protection. The whale would then surge up through the middle of the bubble cloud with a mouth full of fish.
Bubble feeding is a well known way for whales to freak out their food; so is lobtailing. Making the first a systematic set-up to the second, though, was apparently an innovation — and became very popular. By 1989, just nine years after the first Cape Cod whale started lobtail feeding, almost half the humpbacks in the area were at it. Most were younger whales which, since their mothers did not use the new trick, could not have inherited it. Researchers think young whales copied the first practitioner, spreading the technique through imitation. How the first one got the idea is a mystery — as is the question of whether it is actually a superior way of feeding, or merely an increasingly fashionable one.
Cultures rely not only on technologies, techniques and teaching but on rules of accepted behaviour. That things should be fair seems a widespread requirement among social animals. At a canine research centre at Eotvos Lorand University in Budapest, for example, dogs frequently chosen to take part in tests are shunned by other dogs. It turns out that all the dogs want to take part in these tests because they receive human attention; those which are chosen too often are seen as having got unfair advantage. Capuchin monkeys taking part in experiments keep track of the rewards they are getting. If one is offered a poor reward (such as a slice of cucumber), while another gets a tasty grape, the first will refuse to continue the test. Chimpanzees do this, too.
Most cultures distinguish between outsiders and insiders and animals are no exceptions. Orcas, also known as killer whales, are particularly striking in this regard, having a repertoire of calls which are distinctive to the pod in which they live, a sort of dialect. Dr Whitehead and Dr Rendell compare them to tribal markings. Orcas are unusual in that different pods tend to feed on different prey and rarely interbreed. Most of the time, pods studiously ignore each another. But occasionally one will ferociously attack another. This cannot have anything to do with competition for food or females. Lance Barrett-Lennard of the Vancouver Aquarium attributes it to xenophobia — a particularly extreme and aggressive way of distinguishing between insiders and outsiders.
But if animals display four of the five attributes that go to make up a culture, there is one they do not share. Perhaps the most distinctive thing about human cultures is that they change over time, building upon earlier achievements to produce everything from iPhones and modern medicine to democracy. Nothing like this has been observed in animals. Particular aspects of animal behaviour change in ways that might seem cultural, and disruptive change is certainly possible. In the 1990s, for example, South African culling policies that saw the oldest elephants shot and their children redistributed led to large changes in their normally orderly matriarchal societies. Young elephants became abnormally aggressive, since there were no longer any elders to rein them back. In other cases such disruption can seem, anthropomorphically, not so bad (see “the peaceful baboons”). But whether the shocks are good or bad, animal societies have yet to show steady, adaptive change — any cultural progress. Knowledge accumulates with the oldest individuals — when drought struck Tarangire national park in Tanzania in 1993 the elephant families that survived best were those led by matriarchs which remembered the severe drought of 1958 — but it goes to the graveyard with them.
There is a great deal more to learn about animal minds. Grammatical language can pretty thoroughly be ruled out; learned toolmaking for some species is now indubitable: but many conclusions are in the middle, neither definitively in nor out. Whether you accept them depends partly on the standard of evidence required. If the question of animal empathy were being tested in a criminal court, demanding proof beyond reasonable doubt, you might hesitate to find that it exists. If the trial were a civil one, requiring a preponderance of evidence, you would probably conclude that animals had empathy.
Using that standard, one can hazard three conclusions. First, various animals do have minds, The physiological evidence of brain functions, their communications and the versatility of their responses to their environments all strongly support the idea. Primates, corvids and cetaceans also have attributes of culture, if not language or organised religion (though Jane Goodall, a noted zoologist, sees chimps as expressing a pantheistic pleasure in nature).
Next, animals’ abilities are patchy compared with those of humans. Dogs can learn words but do not recognise their reflections. Clark’s nutcracker, a member of the crow family, buries up to 100,000 seeds in a season and remembers where it put them months later — but does not make tools, as other corvids do. These specific, focused abilities fit with some modern thinking about human minds, which sees them less as engines of pure reason that can be applied in much the same way to all aspects of life as bundles of subroutines for specific tasks. On this analysis a human mind might be a Swiss army knife, an animal mind a corkscrew or pair of tweezers.
This suggests a corollary — that there will be some dimensions in which animal minds exceed humans. Take the example of Ayumu, a young chimpanzee who lives at the Primate Research Institute of the University of Kyoto. Researchers have been teaching Ayumu a memory task in which a random pattern of numbers appears fleetingly on a touchscreen before being covered by electronic squares. Ayumu has to touch the on-screen squares in the same order as the numbers hidden beneath them. Humans get this test right most of the time if there are five numbers and 500 milliseconds or so in which to study them. With nine numbers, or less time, the human success rate declines sharply. Show Ayumu nine numbers flashed up for just 60 milliseconds and he will nonchalantly tap out the numbers in the right order with his knuckles.
There are humans with so called eidetic, or flash, memories who can do something similar — for chimps, though, this seems to be the norm. Is it an attribute that chimps have evolved since their last common ancestor with humans for some reason — or one that humans have lost over the same period of time? More deeply, how might it change what it is for a chimp to have a mind? How different is having minds in a society where everyone remembers such things? Animals might well think in ways that humans cannot yet decipher because they are too different from the ways humans think — adapted to sensory and mental realms utterly unlike that of the human, perhaps realms that have not spurred a need for language. There is, for example, no doubt that octopuses are intelligent; they are ferociously good problem solvers. But can scientists begin to imagine how an octopus might think and feel?
All that said, the third general truth seems to be that there is a link between mind and society which animals display. The wild animals with the highest levels of cognition (primates, cetaceans, elephants, parrots) are, like people, long-lived species that live in complex societies, in which knowledge, social interaction and communication are at a premium. It seems reasonable to speculate that their minds — like human ones — may well have evolved in response to their social environment (see “The lonely orca”). And this may be what allows minds on the two sides of the inter-species gulf to bridge it.
Off Laguna, in southern Brazil, people and bottlenose dolphins have fished together for generations. The dolphins swim towards the beach, driving mullet towards the fishermen. The men wait for a signal from the dolphins — a distinctive dive — before throwing their nets. The dolphins are in charge, initiating the herding and giving the vital signal, though only some do this. The people must learn which dolphins will herd the fish and pay close attention to the signal, or the fishing will fail. Both groups of mammals must learn the necessary skills. Among the humans, these are passed down from father to son; among the dolphins, from mother to calf. In this example, how much do the species differ?