If I Could Talk With The Animals...
Most animals on Earth engage in some form of communication. Baboons rub feet in each other’s faces to signify “I am in the mood for sex,” herring gulls tap their beaks on the ground to let the young know “I have food,” and cats sharpen their claws on your ankles to make sure you know “you ain’t all that.”
My favourite mode of animal communication however is easily that employed by honey bees. When scouts want to describe their nectar finds to the rest of the hive, they perform what are genuinely called in the literature a “waggle-dance”. They shake their rears around in a figure eight with the length of dance indicating distance from the hive, and the angle they make to the vertical axis of the hive translating to the angle between the sun’s position in the sky and the food source. They also secrete pheromones to indicate how good the source is, meaning rival bees have to argue for whose find is superior. That's right. Bees communicate using bum-dance trigonometry battles.
Sadly, humans have not mastered this subtle art, but we have invented something truly remarkable for sharing ideas and information: languages. Six thousand five hundred of them are known to our species alone, so is it possible that other species could develop something similar?
First off, I think we can argue that many other species have “words” – unique noises which convey a meaning. Chickens for instance have distinct clucking sounds for “predator approaching from above” and “predator approaching on the ground”, indicating that the noise is not just a panic - it is telling other chickens vital information.
Squirrels take this even further with their barks; they are more likely to make a warning call if members of their family are close and less likely to do so if there is a rival squirrel in the area i.e. some animals can not only use sound to convey information, they can change their noises depending on who is listening. There is even a fascinating project being carried out at the University of Washington called Deep-Squeak which aims to build a computer capable of translating mouse-squeaks into English.
You might consider these noises to be nothing like words because they are just simple sounds, but I would immediately dispute that. Consider Silbo, a language which is entirely whistled, allowing shepherds to communicate across the valleys of La Gomera island. Or take the Taa language of West Africa which contains 164 letters, 111 of which are clicking sounds. Or how about the Wakashan language of British Columbia which features throat-grunts as well as vowels. If we consider clicks, whistles and grunts to be legitimate word sounds, why not the noises animals make too?
But Is It Language?
This all sounds pretty optimistic but there is something really important we need to consider. As the linguistic philosopher Noam Chomsky pointed out when addressing this issue, language is more than just a collection of words – it is also the rules for how those words can be combined.
A vocabulary is not the same as a language, in the same way a dictionary is not the same as a play by Shakespeare. In fact, the English language contains over 171,000 words and the average English-speaking adult speaks 60,000 of them, meaning the average English-speaking adult only knows 35% of their own language. Clearly there is more to a language than just "knowing the words".
For example, here is a sentence I doubt anyone has ever written: In Antarctica there are a species of pink pandas who eat wood shavings. That sentence is not one you have seen before, so you cannot simply be recognising the combination. Yet you still know exactly what the sentence means. Language is not just memorising and regurgitating words. It allows us to generate new combinations that are still meaningful.
Another key feature of language is that as we increase the length of word combinations (the sentence) we increase the information contained within them. For instance:
1) I like hats.
2) Janet said “I like hats”.
3) According to Frank the fishmonger, Janet said “I like hats.”
4) According to Frank the fishmonger, who really should not be trusted given the fact he is a Twilight fan, Janet said “I like hats”.
The more words we put in, the more information we convey and we can do this infinitely. Then of course we have to consider the order the words come in. The sentence “Margaret likes Jeff and hates Richard,” means something different to “Margaret hates Richard and likes Jeff”.
These are the kinds of features we do not find comparisons for in other species. Animal noises are mostly isolated and combinations of them contain no new information. A chicken can utter the squawk “predator approaching” over and over, but this longer sentence does not increase the meaning, it is just repetition. Animals also do not seem to invent new sentences or have a grammar to their limited sounds, but this makes sense from a neurological perspective because, as it turns out, human brains are genuinely different to those of most animals.
Dawn of the Language of the Apes
I’m about to horrendously simplify half a century’s careful study into the field of linguistic neurology, but hopefully you should pick up the gist even if my words are not precise. There’s another feature of human languages - a distinction between literal and implied meaning.
The human brain has two main centres for processing language (found on the left side of the brain for 90% of the population) called Broca’s area and Wernicke’s area. Put crudely, Wernicke’s area is the part which deals with comprehension while Broca’s area deals with remembering words and generating combinations.
People who suffer damage to Broca’s area are still able to follow complex instructions and listen attentively to what people are saying, but speak in a halting, staggered fashion. “Me…want…food” etc.
By contrast, people who suffer damage to Wernicke’s area are able to speak fluently and elaborately, but their sentences are meaningless word-salads: “There wasn’t four parsons undulating to birefringent celery opacity for plums with your mesmerisation.”
Most other animals do not even have a Broca's and Wernicke's area, so language is physically beyond their capability, with the exception of the ape species of course. Chimpanzees, orangutans, gorillas and bonobos have very small Broca’s and Wernicke’s areas. Nowhere near as develeoped as ours, but these emerging structures might imply that apes can learn something akin to language.
The first attempt to teach an ape to speak was made by Catherine and Keith Hayes in 1951 with their chimpanzee Viki. By rewarding the chimpanzee and moving its mouth to encourage certain sounds, the Hayes were able to get Viki to “say” four words: mama, papa, up and cup. And yes, listening to recordings of Viki is every bit as disturbing as you might imagine.
A lot of debate raged over why Viki could only master those four words for several years, until someone pointed out the freaking obvious: chimpanzees don’t have the vocal chords needed. In fact, no other animals do.
While many animals have a larynx and tongue, the arrangement of them inside the human throat allows us to make a wider variety of sounds than any other creature. There are obviously animals which can make noises we can’t - pistol shrimps produce screaming sounds which reach 200 dB - but those are the only sounds these animals can make. It isn’t just our brains which are unique, but also our mouths and throats.
However, just because apes cannot make sophisticated vocalisations does not mean they cannot learn language. After all, there is a whole category of human languages that involve no sound whatsoever: sign languages.
Sign of The Times
Sign languages have all the same features that “verbal” languages have. American Sign Language for instance has over 50,000 words as well as grammatical rules and syntax. Word-order matters in ASL, longer sentences contain more information, new sentences can be invented, different people sign with different "hand-accents" and they can be used to express metaphor, write poetry and tell jokes.
In fact, deaf and mute babies start “babbling” with their hands at the same age hearing and speaking children start babbling with their mouths. They begin to wave their hands in an incoherent fashion as they grasp the language they are exposed to…which means language is not really about moving your throat and engaging your ears, it is about understanding the meaning behind the signifiers.
And, perhaps most importantly, Broca’s and Wernicke’s areas light up in brain-scans just as much for deaf and mute people as for speaking and hearing people. These brain regions are not really about the ears or the throat, they are about the far more abstract notion of processing and creating meaning for symbols. So whichever muscles your language uses are irrelevant. Your brain still allows you to speak. So obviously we tried it with apes.
The first chimpanzee to learn sign language was named Washoe. In 1967 she was adopted by Beatrix and Allen Gardner who taught her 350 words in American Sign Language, and there were many remarkable findings. Washoe could hold basic conversations with the Gardners and was observed “talking to herself” i.e. signing when nobody else was around.
More remarkable was that Washoe, on a few occasions, was apparently able to join words together to form new combinations. When presented with a picture of a duck for instance, Washoe signed “water-bird” which at the time seemed astonishing.
However, the primatologist Herb Terrace was skeptical of these claims and by studying both Washoe and his own chimp fluent in ASL (which he named Nim Chimpsky), he discovered that what the chimps were doing was not really language.
Apes can only make two-three word combinations at most and are unable to increase the amount of information by extending a sentence. For instance, one of Washoe’s favourite phrases was “tickle me” (Chimpanzees love to be tickled) but if the Gardners refused to tickle her she would simply repeat the phrase over and over: “tickle me, tickle me, tickle me, tickle me”. She could not handle a longer sentence like “tickle me now” or “tickle me or I will be sad.”
Washoe was also not able to get word-order correct. Just as often as “tickle me” she was liable to sign “me tickle”. So her signing ability had no syntax and no extension of meaning. In fact, as Terrace pointed out, even the water-bird phenomenon was nothing special. Washoe could just as easily have been signing “water” because there was water in the picture and then “bird” because there was a bird. The sign combination “water-bird” did not mean “bird that floats on water” but simply “there is some water…there is a bird.”
Whereas children start inventing new sentences (and sometimes words) around 15 months old, the chimps never did. They were just repeating the physical signs they had been taught and were not understanding them the same way we do. Don’t get me wrong, it’s still impressive that a chimpanzee was able to look at a picture of a bird on water, process the information and sign the correct symbols…but it’s not what we would call a language. Sign language does not allow apes to speak, unless we somehow strap them into some kind of robotic talking device...
The Koko Kontroversy
Perhaps even more famous than the Washoe experiments was the work of Penny Patterson, a former Stanford psychology doctoral student who, in 1979, decided to recreate the Washoe trials with a female gorilla called Koko, borrowed from San Francisco Zoo (“borrowed” is a generous term because Patterson actually refused to return Koko after the agreed lease was over, claiming Koko did not want to live among gorillas anymore and had fully acclimated to humans…which many people considered a form of animal cruelty, isolating Koko from her kin).
Patterson taught Koko a lot of signs (she claims over 1000) and it appeared for a long time that Koko’s verbal skills were even greater than Washoe’s. Koko could identify colours, answer simple questions and even expressed sadness at the death of Robin Williams (a celebrity she had met many years prior). Koko would also issue Christmas cards online through Patterson’s website, wishing the world peace and love. This is a gorilla supposedly making abstract comments about an entire species. It seemed too good to be true. And of course, as Herb Terrace began to demonstrate, it was.
The first suspicious thing was that studies published by Patterson were non-existent. She did not publish data or describe any controlled experiments, preferring to communicate everything through press conferences and edited online appearances. It’s easy to get “oohs” and “aahs” from an audience, but this does not prove Koko was doing the things Patterson claimed. In fact, when Terrace got hold of the original videos of Patterson communicating with Koko, the story which emerged was quite absurd and a little worrying. Here’s the kind of thing that would happen:
Patterson might hold up an object like a banana and ask Koko to sign it. Koko would then sign something like the word for “building,” to which Patterson might respond “come on Koko, stop being silly, what is it?” Koko would then make the sign for “trousers” and Patterson would laugh and say “she’s being funny, come on Koko what is it?” And then, after a bit of cue-ing from Patterson herself, Koko would finally symbol something like plant and Patterson would go “well done Koko it is a plant! What kind of plant?” Koko would then symbol “pain” and Patterson would respond with “yes that’s right, if you eat too many plants your stomach can be in pain! Well done!”
Patterson would sometimes claim Koko was being ironic when signalling the wrong words (I’m not kidding) or that the end of October was tough on her because it was the anniversary of another gorilla’s death. I don’t think anyone would dispute that Koko could be sad when remembering the death of another animal, but Patterson was claiming Koko knew how to use the Gregorian calendar and acknowledged anniversaries the same way humans do. Gradually the scientific community began to distance themselves from Patterson and she was accused of delusion, misrepresenting data and, by her harshest critics, mistreating Koko as some sort of party-trick animal.
Also, bizarre true story: Patterson claimed Koko had an obsession with nipples and there were several charges of sexual harassment against her, claiming she would instruct her students to expose their nipples to Koko (as she would regularly do herself).
From a scientific point of view the Washoe and Koko experiments are super-cool but they don’t prove apes have the capacity for language. In fact they seem to prove the opposite. We could maybe go so far as to say apes have proto-language ability and there is one bonobo currently being studied (Kanzi) who seems to show word-combination skill. But, I’m afraid if we are honest, we cannot justify saying that apes have language. However there is one other place we should consider looking.
Under The Sea
Humans do not have the biggest brains in the animal kingdom by a long shot, but this is not necessarily the most important thing to look at. Elephants have huge brains, but considering the size of their bodies they need them just to move around. Instead, it makes more sense to consider what is called the encephalisation quotient, which measures how big the animal is in relation to its body volume. On this scale humans have the highest score with apes coming in third place and then, sitting in between them, are the cetaceans: whales, dolphins and porpoises. This is where the research gets really interesting.
In 2018, Stephanie King working in Shark Bay, made the discovery that dolphins appear to have what we might consider “names”. A specific sound can be uttered by a member of the pod and only one dolphin repeats it back. When King recorded the same sounds and played them herself, the same individual dolphin echoed it and none of the others paid attention. It’s almost like the Dolphins are shouting “You there Harry?” and the other one shouts back “This is Harry.”
Then there was an intriguing study carried out in 2016 by Vyacheslav Ryabov who analysed the clicks and whistles exchanged between two Black Sea bottle nosed dolphins named Yasha and Yana. What he found is that the noises were broken down into as many as five distinct sound-chunks which he likens to five-word sentences. Even more crucially, he discovered that dolphins do not interrupt each other when doing this.
If you bring two chimpanzees together who have both been taught sign language (as has happened) they do not exchange information. They “talk” over each other constantly, and repeat the same symbols back and forth. Dolphins however, pause when the other dolphin is making their noises and they do not repeat the same sounds. This could genuinely be a form of language.
Then there is whale song which is a total mystery. A pod of whales will sing patterns of notes which can carry several kilometers across the water to a different group who can modify and send it back, or pass it on to another pod. Some researchers have suggested that whale songs are transferred across great distances like a whale internet and everything from mating calls to story telling has been touted as a possible explanation. Although obviously I’ve seen Star Trek IV, so I know exactly what's going on.
Why Cetaceans Are Tricky
Unfortunately we know hardly anything about cetacean neurology because of the obvious problem…all the methods we use to analyse the brain cannot be used under water. We tend to figure out how brains work by performing brain scans on live animals, observing the behaviour of brain-damaged individuals, observing the effects of medication, or by carrying out various tests in a controlled environment.
Brain-scanners are out of the question because (funnily enough) sensitive electrical equipment doesn’t work under water. We also cannot tell if a whale has been brain-damaged or suffered a stroke because all we can observe is how they swim about. We cannot ethically give them psychiatric medication either and because they move in vast arenas (it’s the sea) it is not easy to perform any kind of controlled experiment.
The only things we can sensibly do are make deductions about their behaviour in captivity or analyse the brain once the animal has died. But even that is difficult because you either have to wait for a carcass to wash up on shore and hope the brain is in tact by the time you extract it (a rare occurence) or you hunt and kill a whale yourself. Curiously, people who feel passionate about cetaceans and want to study them are not the same people who want to hunt them.
On the rare instances we do get hold of a cetacean brain in good condition, there is still a limit to what we can find out about it. We cannot watch it in action, so we can only make statements about things like size, mass and chemical composition, which is like trying to figure out what software a computer was running by looking at the hard drive after it has broken down. And even when we do this we run into a huge problem: cetacean brains are not put together the same way ours are. It isn’t known if they even have Broca’s and Wernicke’s areas. Their brains look different so, simply put, nobody has a clue what’s going on inside them.
Personally, I feel there is just enough evidence to answer the overall question of animal language with a hard "maybe". Whales and dolphins are our best bet and, given that the field of cetacean linguistics is new, we could be in for some exciting surprises over the next few years.
Maybe if we can learn to speak dolphin we will get an insight into how another intelligent creature views reality. Maybe we can learn how our own minds work from studying those which are drastically different. And maybe, just maybe, if we can prove these wonderful creatures are capable of language it will persuade those who hunt them to reconsider what they are doing. Maybe Science won't just save our species, but others too!
I love science, let me tell you why.