clock menu more-arrow no yes mobile
Javier Zarracina/Vox

Filed under:

What humpback whales can teach us about alien languages

How to recognize a potential alien language, in two steps.

Brian Resnick is Vox’s science and health editor, and is the co-creator of Unexplainable, Vox's podcast about unanswered questions in science. Previously, Brian was a reporter at Vox and at National Journal.

For decades, a group of researchers on a quest to find extraterrestrial life have been listening patiently to the cosmos. From their hub at the SETI Institute (a.k.a. the “Search for Extraterrestrial Intelligence”) in Mountain View, California, the researchers are searching for a signal, a radio transmission, from an extraterrestrial species.

Though there is a plausible scientific case that other life could exist out there somewhere, it could be a very long time before we detect anything. And in the meantime, there’s an important question they need to figure out: How can we recognize an alien language when we have no idea what it may sound like? How can we be sure the signal we’re receiving is a language, and not just some random noise?

These are the big questions Laurance Doyle, a research scientist at SETI, is dedicated to answering. He believes it will be possible to recognize an alien language as a language. And he bases that belief on his studies on how animals communicate with one another on Earth. Really.

According to Doyle, there are two main steps to figure out if a communication from beyond is a language. All it takes is some know-how of a branch of math called “information theory.” Recently, Doyle walked me through it.

Step 1: Does the communication follow a special pattern called Zipf’s law?

Have you ever heard a humpback whale sing? It sounds like this: squeaky, resonant, and otherworldly. If there’s an analog to extraterrestrial communication on planet Earth, it might as well be whale songs.

Humpback whales sing to communicate with one another. They sing to facilitate hunts, to socialize, and so on. Bottlenose dolphins, on the other hand, make chirpy calls to one another in a pod, and it’s believed these chirps help them work together as a group.

Scientists don’t know what exactly these whales and dolphins are saying to one another. But they suspect what they are vocalizing is, perhaps, a language, because it mimics a mathematical pattern shared by all languages on Earth.

The pattern is called Zipf’s law, named after George Kingsley Zipf, an American linguist who lived in the first half of the 20th century.

Zipf was interested in figuring out which letters and words are the most common in the English language, and in all languages. He started off by analyzing all the letters used in James Joyce’s novel Ulysses — the number of E’s, the number of A’s, even all the Q’s — as Doyle explained to me.

If you plot the frequency of the most common letters to the least common letters on a logarithmic scale (one that increases or decreases exponentially), you get a simple negative sloping line, going down at a 45-degree angle. It means our most common letters (E, A) are used exponentially more often than our least common letters.

The same pattern holds true when you look at words: The most commonly used words in English — like “the,” “a,” and “I” — are used way more often than less common words like “appropriate.” This is true for all other human languages on Earth.


Zipf’s law is a reflection of the careful balancing act a language needs between variety and simplicity. It shows language has syntax: a consistent way to arrange the order of words. We use easy, common words — articles, pronouns, prepositions — to lay the scaffolding of our language. Then we adorn it with more complicated, specific words.

In his work, Doyle finds that the squeaks of bottlenose dolphins follow Zipf’s law, meaning when you record the individual sounds they make — their signals — some are used exponentially more often than others. Humpback whale songs don’t quite hit the same ratio at Zipf’s law, but they come close. This doesn’t mean these creatures are the most intelligent mammals after us (though they very well may be). But it’s a clue there’s a complex system, a complex brain, animating their communications.

Overall, Doyle has found that Zipf’s law “is a necessary, but not sufficient, system for proving complexity” in a communication system, he says. “It’s very quick intelligence filter.”

Which means Zipf’s law would be a good place to start in listening to ET. Analysts at SETI could analyze the patterns of transmissions from the alien source and see if some of the signals occur exponentially more often than others. That would be a big clue that we’re listening to a language.

Step 2: Does the communication contain “conditional probabilities”?

Zipf’s law wouldn’t be enough to determine if an alien transmission is, indeed, a language. A signal following the pattern of Zipf’s law could just be a coincidence.

For a language to be a language, it also needs something called conditional probabilities.

If I give you the letter Q and ask you what letter probably comes after it, you’re likely to say U. That’s a conditional probability, meaning there is a correlational structure to our language. U’s usually follow Q’s. But it also works for whole words. Conditional probabilities mean that if I write you a note with a word smudged out, “How are *** today?” you might guess I meant to write “you.”

Assessing if an alien language has conditional probability isn’t an easy task. But to figure out how, Doyle returned again to humpback whale songs.

He believes the humpback whales have conditional probabilities in their language. When there’s noise in the water, like a boat sound, some of the whale songs get garbled, but they can still understand the song, he says. And they slow down their communication in the presence of noise — but not as much as you would think if they cared about getting every note clearly broadcast.

It’s possible that extraterrestrial life could exhibit a similar pattern to deal with interference in the cosmos, as the whales do. And that’s something you could quantify and measure. With the math of “information theory,” you can potentially figure out how many orders of entropy a language has by listening for how the language deals with interference or noise. A higher-order entropy in a signal would signify a more complex language.

But we first need more data from animals to see if there’s a universal pattern to look out for, one that signals the language contains these conditional probabilities.

The goal: Can we tell how smart a creature is just by recording the patterns of their communication?

Here’s the goal of Doyle’s work on animal communication: It would be nice to build a scale that equates the complexity of an animal’s communication with the size of its brain. If he could do that for animals, he reasons, he could do it for ET. It would be a good way to figure out if an alien transmission is from a species that is much smarter than us or more on our level.

Doyle says this may work similarly to what’s called the “encephalization quotient.” That’s a scale that explains that the intelligence of a given species generally increases as the relative size of their brain increases. Human anatomy devotes more space in the skull to the brain than a chimp’s does. And humans are smarter.

“We think we can come up with a communication encephalization quotient ... a measure of the communication intelligence of a species,” Doyle says. This quotient, which doesn’t yet exist, could help us understand that the bee communication traces itself back to a tiny bee brain, and that whale communication traces itself back to a bigger whale brain.

“I’d like to see a full distribution of species of life on Earth and their communication systems,” Doyle says.

Again, even if we had such a quotient at the ready, we wouldn’t know what exactly ET is saying. But we’d know, from analyzing their conversation patterns, how complex their language is, and, therefore, how complex they are.

We could potentially learn that their communications are vastly more advanced than ours — operating with layers of complex rules we couldn’t hope to understand. Wouldn’t that be fun, and a bit terrifying?

“What if an alien says, ‘By this time, we will have to be had done it,’” Doyle says. “We can’t handle it.”

Basically: We could find out that their language is to us, as our language is to a dog. Dogs can understand a few words and follow basic commands. But their brains are not equipped to follow our syntax. (Note: If your dog can understand past participles in English, or in another language, please email me.)

Why this thought experiment isn’t completely pointless

If all of this sounds a bit too hypothetical, well, to a degree it is. We haven’t received any transmissions from the great beyond. And it may seem silly that people would spend so much time thinking about this what-if.

But a lot of people, including serious scientists, believe it would be ridiculous if human beings are the only life, or intelligent life, in the galaxy, or the universe. The late physicist Stephen Hawking said he preferred to believe (though without perfect certainty) that “there are other forms of intelligent life out there,” and they just haven’t noticed us yet.

If something’s out there, trying to reach us, we might as well be listening. And if we’re listening, we might as well try to figure out what to listen for.

In another real way, in recent years, this question of “what’s out there” has become less abstract. In the past 20 years, thanks to exoplanet-hunting missions like Kepler, we’ve learned that there are more planets than stars in our galaxy.

Some are gas giants, like Jupiter or Saturn. Others are small and rocky, like our own Earth or our neighbor Mars. Some, like many of the planets in the Trappist-1 system, are in their solar system’s habitable zone, meaning it’s conceivable for liquid water to flow on their surfaces. We don’t know how many could contain life, but these discoveries inflame our imaginations about what these worlds might be like, and what could live there.

This work is also meaningful because it directs us to look at the world around us and better understand how animals communicate with one another. It’s possible that in deeper study of communication systems on Earth, scientists will find new, interesting patterns separate from Zipf’s law.

Take the slime mold, for example. It’s a superorganism — a single body composed of thousands of individual amoeba-like creatures — that seems to communicate and have intelligence despite the fact that it doesn’t have a single neuron. How does a slime mold communicate? Scientists are still figuring it out. And in figuring it out, they could discover something new and wonderful that could help them better decipher alien communication.

Overall, this work is a reminder that there are many wonderful natural things to study on the Earth, and potentially communicate with. “If humpback has as complex a rule structure as English, we should be to translate English things into humpback someday,” Doyle says. And, heck, if we can learn to talk to whales, we’re one step more prepared to talk to aliens if they were ever motivated to pick up the phone and call.

“We’re looking up in the sky going, ‘Are we alone?’ and there are humpback whales going, ‘Woo-hoo’,” he says. In preparation for a conversation with ET, he says, “we have a million communication systems on Earth to look at.”


The “new abnormal”: The rise of extreme flooding, briefly explained


9 ocean mysteries scientists haven’t solved yet

Down to Earth

Scientists will unleash an army of crabs to help save Florida’s dying reef

View all stories in Science

Sign up for the newsletter Sign up for Vox Recommends

Get curated picks of the best Vox journalism to read, watch, and listen to every week, from our editors.