Into the wild
In the water with dolphins, you are out of your element. And they know it.
“Dolphins are clearly intelligent,” Kershenbaum says, recalling his time swimming with them. “They’re very interested in you. They’re aware that in this underwater environment, you’re pathetic. You can barely swim, compared to their abilities.”
Five dolphin whistles
Dolphins are the only animals we know of that give each other names. They play games with one another, like ‘catch the fish’ and ‘tag, you’re it’. But communicating underwater is a steep challenge. In such a noisy environment, transmitting information over longer distances is fraught with difficulty. With a series of clicks, dolphins can ‘see’ in sonar - an ability known as echolocation. At close range, they can chat comfortably, but have to adapt the sounds they make when sending messages over long distances.
A surprising parallel can be found in another species Kershenbaum studies: wolves. Wolves share the dolphins’ problem of reliably transferring information in a confusing environment.
“Dolphins and wolves have a lot in common. Long-range communication has some physical constraints: a lot of nuanced information is lost when you send an acoustic signal over long distances. That’s the same if you’re trying to communicate from one mountaintop to another, or trying to call your dolphin friend several leagues away.
“Both dolphins and wolves have evolved a much simpler, robust use of acoustics to overcome these physical restraints on long-range communication.”
This is why wolves howl. Howls are adapted to travel. They remain detectable and distinct over long distances. As far as Kershenbaum can tell, howls fulfil at least three different roles: marking out territory, keeping in touch with other pack members, and, adorably, satisfying the animal's love of howling.
Eastern Timber Grey Wolf howl
Whether this kind of long-range communication lacks the subtlety and complexity required for what we call language is contentious. Ian Roberts, Professor of Theoretical Linguistics at the Faculty of Modern and Medieval Languages and Linguistics, thinks that the systems Kershenbaum describes are limited.
“Human languages all rely on two separate systems,” Roberts says. “The first combines meaningless sounds into meaningful sentences. The second combines meaningful words into potentially unlimited sentences. The communication systems Arik describes don’t seem to work in this way.”
Kershenbaum is less convinced that all potential languages would need words and sentences. He remains open-minded on animals’ ability to communicate information, even if long-range methods demand more basic messages. There does seem to be more information transmitted in the short-range communication of animals, even if we can’t call it ‘language’.
Dolphins and wolves have this kind of short-range communication (whistles, barks and growls), but studying this poses a greater challenge. Getting close enough to eavesdrop on a wolf pack isn’t usually possible. One animal that Kershenbaum has got closer to is the hyrax.
Hyraxes are small, rotund, furry mammals, who live in large groups. Their songs are long and loud. As Kershenbaum found out, their calls have specific meanings.
“Hyraxes keep guard. One of them keeps a lookout for predators, and calls out when one is sighted. Depending on the type of call, the group can come together to mob this potential predator.
“In a particular case, I was doing some research close to a hyrax colony. I played a recording of a hyrax distress call, and they all came running at me. There were about 20, and they were very threatening.
Call of the hyrax
“This just illustrates that communication in the social context plays an adaptive role. It's really important to be able to say ‘here's a predator, come and help me’, and that needs to be a different call from ‘I see a predator, run and hide’. It becomes very clear that hyrax messages have different meanings.”
More widely, the problem of understanding meaning in animal communication is daunting and controversial.
“Animals convey a lot of information in their communication, but it’s probably encoded very differently to how we do things. They almost certainly do not use an equivalent of words, for example. Before you can hope to understand what they’re saying, you have to understand where the information is. These are not trivial problems.”
To help with this thorny information task, groups like Earth Species Project and Project CETI (the Cetacean Translation Initiative) are developing new algorithms and machine learning techniques. Kershenbaum acknowledges the benefits of these efforts, but sees problems with relying too heavily on data.
“If you hope to solve wider problems merely by looking at the data, and not the animals in the wild, you’re wrong. Some things can be done by data alone, but to understand meanings, you have to be able to see the behavioural context. It’s that context that drives all communication.”
“Pitching such efforts as ‘translation’ is misleading. There isn’t a word-to-word correlation between different animals’ communication. But we should use new technologies to figure out what information is in there - and that’s very valuable.”
