© Adam Batchelor
The human brain is special. Just not that special. To understand animal minds, and our own place in the living world, we should remove ourselves from centre stage, argues Peter Aldhous.
Tommy the chimpanzee got his day in court on 8 October 2014. He was unable to attend the hearing in ‘person’ – spending the day, like any other, in a cage at a used trailer sales lot in Gloversville, New York. But an hour’s drive away, in a courtroom in the state capital of Albany, Steven Wise of the Nonhuman Rights Project argued that Tommy should indeed be considered a person under New York state law. If so, Patrick and Diane Lavery of Circle L Trailer Sales could be summoned to determine whether they are imprisoning him illegally.
Central to Wise’s arguments in Tommy’s case, and similar suits his organisation has filed on behalf of other captive chimpanzees, is the assertion that apes are highly intelligent and self-aware beings with complex emotional lives. “The uncontroverted facts demonstrate that chimpanzees possess the autonomy and self-determination that are supreme common law values,” Wise told the five judges hearing the case.
It is a bold legal move – and so far unsuccessful. The court in Albany, like a lower court before it, rejected the idea that Tommy has legal rights of personhood. But Wise intends to fight on, taking Tommy’s case to the state’s ultimate arbiter, the New York Court of Appeals.
Events elsewhere in New York State stand in stark contrast to its courts’ willingness to consider the legal implications of the science of animal cognition. In March 2014, the Rip Van Winkle Rod and Gun Club in Palenville, a hamlet of some 1,000 people on the Hudson River, held the fourth instalment of an annual festival that makes a competitive sport out of shooting down creatures that – judged by objective measures of their mental abilities – are arguably just as deserving of personhood as Tommy.
Those creatures are crows, targeted with abandon at the Palenville ‘Crow Down’. In recent years, members of the corvid family – including crows, ravens, jays and magpies – have been found to possess cognitive skills once thought to be the exclusive domain of people and the great apes. They make and use tools. They remember details about the past and plan for the future. They even seem to respond to one another’s knowledge and desires. “For all the studies that have been compared directly so far, the corvids seem to perform as well as the chimpanzees,” says Nicky Clayton of the University of Cambridge, in whose lab some of the most exciting discoveries have been made.
© Adam Batchelor
We gaze into the eyes of a chimp and see a reflection of ourselves. We glance at a crow and see an alien being that under some jurisdictions can be exterminated with impunity – bringing a sinister second meaning to the phrase “a murder of crows”. Such biases affect ordinary people and academic experts alike, skewing our understanding of what non-human intelligence looks like. Apes are undoubtedly very smart, but their cognitive abilities have also been scrutinised more closely than those of other species – largely using experiments designed to explore the mental prowess of human children. Creatures that share our basic body plan and sensory world have big advantages when taking these tests. If we want to understand the diversity of animal minds – and by doing so perhaps understand ourselves better, too – we have to judge them on their own terms.
That’s hard to do. Even knowing how clever they are, when I met Clayton’s corvids this past summer, I felt little connection. It was a very different story a few weeks later when I had the chance to interact with captive wolves. I was fully aware that we can trick ourselves into thinking that dogs and wolves are smarter than they are by projecting our own thoughts and feelings onto them – but I did it, just the same.
My views on animals are as conflicted as the next person’s. I live with a dog – until a recent bereavement, two. I eat meat. I believe that some animal experiments are justified to advance human and veterinary medicine. But my experiences as a PhD student studying mouse behaviour, back in the late 1980s, left marks on my psyche. Before launching into these studies, I didn’t properly consider that I’d have to dispose of dozens of mice once they were no longer needed for an experiment. The first few times I killed, my heart raced, my mouth went dry, and I felt sick and depraved. I probably should have told my supervisor that I needed to change project. But I continued with my research, and the killing became easier – mundane, even. In retrospect, that worries me more than my initial visceral reaction.
Mice feature on nobody’s list of unexpectedly smart creatures. Still, my experiences trying to understand the social lives of animals living in a world dominated by odours led me to question the relevance of experiments designed to probe the abilities of more sophisticated animal minds. The best known is the ‘mirror test’ of self-awareness, devised in 1970 by Gordon Gallup (now at the University at Albany, State University of New York). After giving chimps a while to interact with a mirror, Gallup marked them with a dye while anaesthetised, and then studied their reaction to their reflections after they came round. The chimps looked into the mirror while touching the marked area, from which Gallup concluded that they knew they were looking at themselves. A concept of self is considered one of the hallmarks of an advanced mind, and human infants adopt similar behaviour from about 18 months of age.
My issue with the mirror test is not the meaning of success, which has since been claimed for other ‘intelligent’ species such as elephants, dolphins and magpies, but rather how we should interpret failure. Do most animals fail because they lack a concept of self, or is the test largely irrelevant to species that don’t use visual appearance for social communication? How would you design a similar test using odours for a bloodhound or ultrasound for an echolocating bat? A bat hearing an altered version of its own echoes might well be thinking: “Yes, that’s me…although there’s something weird going on…” – but how would we know?
I don’t blame researchers for concentrating their energies on what separates people from our closest living relatives. The human mind is clearly special: chimpanzees and other great apes don’t compose symphonies, design aircraft or devise experiments to probe one another’s mental abilities. The difficulty comes when we try to break down our mental prowess into its component parts, with the aim of working out exactly why we’re different.
Once there was a long list of cognitive attributes thought to be uniquely human: they included language, morality, solving problems by inventing tools, projecting our thoughts into the past or the future, and a ‘theory of mind’ – the ability to attribute desires, knowledge and other mental states onto others, realising that these may be different to our own. Over the past few decades, the edifice of human cognitive uniqueness has slowly been chipped away as first great apes and then other species, including elephants and dolphins, have passed landmark tests of cognitive skill. At the same time, the gap between these ‘clever’ species and those that perch far away from us on the tree of life has arguably already been closed.
Folklore in several cultures associates crows with wisdom and cunning, but it’s only relatively recently that science has caught up with this idea. The first recognition that members of the crow family may rival the great apes for smarts came in the mid-1990s, when Gavin Hunt of Massey University in Palmerston North, New Zealand, reported his field observations of crows native to the forests of New Caledonia – an archipelago in the South Pacific.
Hunt watched the crows fashion twigs into hooks to capture insect prey from holes in dead wood, or snag insects using barbed tools cut from the leaves of pandanus trees. Subsequent work with captive New Caledonian crows has shown that Hunt’s observations were no fluke. In one famous experiment at the University of Oxford, a crow named Betty quickly figured out how to bend wire into a hook to lift a small bucket containing food out of a tube, after her companion took off with a wire that had already been shaped this way.
New Caledonian crows aren’t just adept toolmakers. They seem to understand cause and effect, and are capable of astounding feats of meta-tool use: using one tool to manipulate another in order to achieve an ultimate goal. And, if anything, New Caledonian crows seem better than chimps at translating their skill in solving one problem to another that is conceptually similar but looks a little different. One standard test is the trap-tube test, in which an animal must work out how to use a stick to obtain food from a tube, realising that pushing it in one direction gets the reward but pushing it in the other causes the morsel to fall into a trap. New Caledonian crows that have already solved this problem do better than naïve birds when presented with the trap-table test, where they are given a choice between two food items that can be raked towards them across a flat surface – one of which is doomed to fall into a similar trap. For chimps and other great apes, prior experience with the trap tube gives no advantage on the trap table. Apes don’t seem to get that both games have the same rule: Don’t let the food drop down the hole!
These sorts of problem-solving tasks are staples of comparative psychology, frequently used to compare the cognitive abilities of young children and chimpanzees. They are fairly easy to adapt to crows, which have good vision and a beak that is almost as dexterous as an infant’s or ape’s hands. But understanding the wider scope of corvid cognition required a leap of imagination, entering the birds’ world to devise experiments to probe the mental processes behind their everyday behaviour.
Nicky Clayton made that leap while at the University of California, Davis, in the late 1990s. A psychologist colleague, Tony Dickinson of the University of Cambridge, had told her confidently that animals lack ‘episodic’ memory – an autobiographical ability to remember the ‘what, where and when’ of past events. That was the orthodox view, but Clayton had big doubts: “To my knowledge, nobody’s tested that,” she recalls saying. Clayton was studying foraging behaviour in western scrub jays, and she realised that the birds’ habit of burying food in caches provided a prime opportunity to see how their memories shaped up.
She and Dickinson first let the jays cache perishable waxworms and non-perishable peanuts in sand-filled trays. The birds quickly learned that the insect larvae became unpalatable after a few days, while the nuts remained good to eat. In subsequent experiments, the jays sought out the locations where they had hidden their favourite food – the waxworms – if just four hours had elapsed from the last time they had seen the caching trays, but went for the hidden peanuts if several days had passed. And it wasn’t just that they could smell rotting waxworms: even if the food items were removed and the trays filled with fresh sand, the jays sought out the locations where they had cached waxworms only when the insect larvae were likely to be fresh.
There is no way of knowing whether scrub jays’ memories of when and where they hid specific food items involve consciously projecting their thoughts into the past, which is how the human mind would approach the problem. Still, it was as impressive a demonstration as science could provide of what Clayton calls “episodic-like” memory in an animal.
Clayton’s scrub jays today reside in an aviary in Madingley, a sleepy village outside Cambridge that hosts the university’s Sub-Department of Animal Behaviour. When I visited, evidence of their caching behaviour was clear to see. They don’t just hide food: stones, corks and a nail had all been jammed between their enclosure’s wire mesh and its wooden frame. It was a while before I saw one of the birds bury an item of food – apparently my presence had interrupted the usual order of business. “Because you’re new, and they don’t know you, they are checking you out,” Clayton told me. I stared back, struck by the gulf between us. I’d read the scientific papers; I knew how cognitively sophisticated scrub jays are. Yet I experienced none of the emotional connection I have felt each time I have come face to face with a captive chimp.
Clayton, it seems, has no difficulties relating to her experimental animals. In part, she attributes this to a long-standing wish to fly. That motivated a lifelong devotion to ballet training, which she believes brings her as close as a human being can get to that experience. And she is convinced that spending much of her leisure time thinking like an artist made it easier to think about animal minds on their own terms. “We are limited by the fact that, as humans, we do see things in a particular way,” she says. “But that doesn’t mean we can’t try and minimise those constraints by taking ourselves out of the picture – which is what artists do all the time.”
Whatever the explanation, the discoveries from Clayton’s lab have come thick and fast. Recent experiments have exploited the fact that jays, like people, can get sated with one food item but still desire a different food type. (This is why we can still tuck into dessert even when we would turn down a second main course.) Working with a PhD student, Lucy Cheke, Clayton has found that even if Eurasian jays are sated on a particular food item, they will ignore their current desires and selectively cache that food if they have learned it is likely to be scarce in future. So now we apparently have jays planning ahead, in addition to acting on specific memories of the past.
It seems that jays also take account of the knowledge and desires of their companions. After male Eurasian jays watched their partners eat either waxworms or mealworms to satiety, Clayton’s team found that the birds selectively fetched the other food item for their partners. The males weren’t responding to some subtle “I want waxworms!” behavioural cue, because they only provided that preferred offering if they had actually seen their mates gorging on mealworms.
Earlier experiments with scrub jays had revealed that the birds change their behaviour when they realise that their caches are likely to be pilfered. If watched by another scrub jay while caching, they later move their stashes around when given some privacy. Crucially, the birds only engaged in this crime-prevention behaviour if they had themselves previously raided other scrub jays’ caches. Once they know from personal experience that stealing is something that jays can do, it seems that they then react accordingly when another bird’s knowledge about their caches poses a threat. Or, as Clayton puts it: “It takes a thief to know a thief.”
Together, these results suggest that jays possess something similar to what, in a human child, would be called a theory of mind – thinking about the mental states of others, and recognising that they are autonomous beings with their own knowledge and motivations.
© Adam Batchelor
So, if creative scientists think themselves into the mental worlds of other species to design experiments to probe their impressive mental abilities, we will get a more complete picture of animal intelligence, right? Perhaps, but humanity’s self-absorption seems also to trick us into seeing some animals as more special than they really are. I’m convinced that’s happening with man’s best friend, where some experiments arguably reveal more about the scientists running them than about the mental abilities of the animals involved. Dogs have even been put into MRI machines to see if they love us, prompting the scientist behind the research, Gregory Berns of Emory University in Atlanta, to claim in The New York Times: “Dogs Are People, Too.”
What’s special about dogs, the theory goes, is that they have been bred over tens of thousands of years to be unusually receptive to interacting with people. For instance, dogs seem exceptionally good at reading human gaze and pointing gestures to locate hidden food. In 2002, researchers led by Brian Hare, then at Harvard University, found that domestic dogs consistently beat both human-raised wolves and chimpanzees on these tasks, strengthening the idea that these skills were the result of many generations of selective breeding.
At Wolf Park in Battle Ground, Indiana, Hare’s findings caused some raised eyebrows. Pat Goodmann, senior wolf handler at the park, was deeply sceptical. “I could remember quite a few instances where I’d been pointing, and I was signalling to another human, but a wolf picked up on it,” she says. Those incidents had stuck in her mind, because often they involved an item – such as a beer can in a lake in the park’s main enclosure – that she really didn’t want the animals to get hold of.
Some years later, word of the doubts at Wolf Park about Hare’s work reached Clive Wynne, then at the University of Florida in Gainesville, who had recently started investigating dog behaviour with his PhD student, Monique Udell. Intrigued, Wynne and Udell flew to Indiana and ran similar pointing experiments, with one key difference: the people doing the pointing were inside the enclosures with the wolves, rather than outside, separated by a fence. Under these circumstances, the wolves outperformed pet dogs, which did well indoors but struggled when tested outside. Shelter dogs, which had less prior experience of interacting with people, were pretty useless. “All these animals have the capacity to notice relationships between things that people do and consequences that matter to them,” says Wynne. “What differs is the kinds of experiences they have in their lives.”
Wynne’s conclusion: Dogs’ skills in social cognition weren’t shaped by humans through selective breeding; they were already present in the wolf pack. He is also unconvinced that these abilities involve anything more sophisticated than simple learning.
When Wynne and Udell ran their initial experiments, they stayed outside the enclosures and let Goodmann and her colleagues do the pointing. But last August, I took the opportunity to get up close and personal with one of the stars of the show, a female wolf named Marion. Already nine years old when put through her point-reading paces, Marion has become the elder stateswoman of Wolf Park. She’s 16 now, almost snow white, and although the companions she once lived with have passed away, she can’t be united with any of the other mini-packs at Wolf Park: her alpha status means she’d tear any rival apart – or die trying.
Unfamiliar people are welcome, though, once they have been instructed in wolf etiquette: let Marion come to you, give her a scratch, but ease up from time to time. It turns out that hand-raised wolves feel compelled to stay put if someone is touching them, but if that’s not what they really want to do, they can get annoyed. I’m delighted to say that my encounters with Marion and the other residents of Wolf Park were entirely amicable. We even tried an impromptu recreation of the pointing experiment – although by that time Marion had realised that the bag attached to Goodmann’s belt contained some tasty treats, so she wasn’t very interested in having her attention diverted to anything else.
My day at Wolf Park reinforced how hard it is to do what Clayton urges and take ourselves out of the picture when considering animal minds. As Marion leant forward to lick my face, I was mentally back home with our dogs. Later, wandering through the main enclosure with the park’s youngest residents – Bicho, Kanti and Fiona – I saw a ferocious display of aggression from the three siblings towards their parents, housed in the neighbouring enclosure. So when Kanti, a powerful male, tested me out by leaning into my legs, I couldn’t help but feel a frisson of fear.
“It’s not about you,” my ex-scientist’s inner voice was saying, reminding me to observe what the animals were doing, rather than being led by my emotional reactions to them. But in my mind, at that moment, it was all about my responses to the wolves, and there was little I could do about that.
Maybe we would find it easier to assess evidence for animal intelligence on its own terms if it emerged from minds so alien that we wouldn’t even try to project ourselves onto them. That may be coming, from studies of cephalopods – octopuses, squid and their relatives. Here we are talking about invertebrates with brains organised radically differently from our own: they have a distributed nervous system, with mini-brains in their tentacles, as well as the main organ. But Clayton has found that cuttlefish seem to remember past events, while octopuses have been observed carrying around coconut shells to use as shelters if attacked, which suggests forward planning in tool use.
Some researchers working on vertebrate cognition, meanwhile, are starting to reject the field’s anthropocentric biases. In Thailand’s Golden Triangle, Josh Plotnik of the University of Cambridge works at a luxury resort that is home to a group of elephants which, when not giving rides to tourists, take part in his research. Plotnik started with the usual roster of experiments tried on young children and chimps, including the mirror test. But he now realises that he needs to better understand the elephants’ sensory world – dominated by odours and low-frequency sounds – before he can work out how to explore the full scope of their cognitive abilities.
“It would be very unethical of me to take all of the chimp experiments and just run them on the elephants,” Plotnik says. “I’d be publishing all these negative results, saying: ‘Elephants can’t do this. Elephants can’t do that.’ When in fact, they probably could, if we asked the questions the right way.”
This will probably mean moving beyond simple ‘pass-or-fail’ tests of human-inspired cognitive skills and instead looking in detail at what Alex Taylor of the University of Auckland in New Zealand, a colleague of Clayton’s, calls cognitive ‘signatures’. He means not just recording whether or not a mental task can be performed, but also noting error rates and studying the circumstances under which the ability breaks down – ideally also using brain imaging to look at the underlying neural activity involved.
It’s hard to predict where this approach to studying animal minds might lead, but wouldn’t it be interesting if we found that some animals are thinking in ways that avoid stereotypically human mistakes? We are bad at judging when it makes sense to take economic risks, for example, typically placing too much value on assets we already own – even when it would make sense to risk them for a bigger payoff. I can also imagine that probing specific animal mental skills could have applications in artificial intelligence, though we’re still a long way from being able to replicate the cognitive abilities of non-human animals. And as we search the skies for signs of extraterrestrial intelligence, we surely don’t want to be limited to thinking about what ‘intelligence’ means in purely human terms.
For now, the only minds available for us to study are those on our own planet. Let’s not squander that opportunity by getting caught up in arguments over whether Tommy the chimpanzee, our pet dogs, or any other species should be seen as people, and instead adopt the mantra: “People are animals, too.”
Damned interesting animals, for sure. But we don’t have a monopoly on that.